EVALUATING THE IMPACTS OF CLIMATE...

EVALUATING THE IMPACTS OF CLIMATE CHANGE

AND CLIMATE VARIABILITY ON POTATO

PRODUCTION IN BANISOGOSOGO, FIJI.

by

Moleen Monita Nand

A thesis submitted in fulfilment of the requirements

for the degree of Master of Science in Climate Change

at the University of the South Pacific (USP)

Copyright © 2013 by Moleen Monita Nand

Pacific Centre for Environment and Sustainable Development (PACE-SD)

The University of the South Pacific

September, 2013

xvii

Dedication

Dedicated to the farmers of Fiji

�च�तनीया ह �वपदां आदावेव ��त��या

न कूपखननं यु�त ं�द��ते वि�हना गहेृ

The effect of disasters should be thought of beforehand (that is, before they actually

occur.) It is not appropriate to start digging a well when the house is ablaze with fire

(Sanskrit) (Paranjape, 2013).

xviii

Acknowledgment

Conducting this research had been a journey of intellectual discovery for me and like

most research, I have also benefited from meeting many interesting people. These

people have contributed ideas, co mments as well as considerable professional,

inspirational and motivational support.

First and foremost, I would like to take this opportunity to thank PACE-SD through

the AusAID Climate Leader’s Program for providing me with Masters Scholarship

by thesis which enabled me to conduct this research. Secondly, I am very grateful for

my principal supervisor, Dr. Anjeela Jokhan for her valuable insight in my research

with the thorough revision of my thesis. I would also like to take this opportunity to

thank other members of my team, Dr. Morgan Wairiu who helped me in the initial

stages of proposal write up and Mr. Viliamu Iese, my co-supervisor and DSSAT

adviser, who helped me in my fieldwork, advised me on my thesis write-up and also

for his help in the use of DSSAT v4.5 crop modeling component. Also, I would like

to acknowledge Mrs. Reema Prakash and Dr. Dan Orcherton who helped me in my

write up. I also express my sincere gratitude to Dr. Upendra Singh (my external

supervisor) from International Fertiliser Development Center, Alabama, U.S.A, who

helped me put together my proposal for this research and also helped me calibrate the

DSSAT SUBSTOR Potato model.

Much thanks also goes to Mr. Sai for the logistics of personnel and equipment. I am

very grateful to the Nand Family and the Kumar family for helping me with the acco

mmodation and field-work. I would also like to acknowledge the staff of FSTE, USP;

Mrs. Arieta Banivalu, Mr. Dinesh Kumar, Dr. G. Brodie, Ms. Artila Devi and PACE-

SD staff members: Hilda, Mr. Nasoni Roko, Mr. Sumeet Naidu, Mrs. Vijaya, Mrs.

Nirupa, Dr. Helene and Professor Elisabeth. I would also like to show my

appreciation to the staff of Nacocolevu Research Station and Koronivia Research

Station for providing information on potato experimental plot design and soil data

and staff of Fiji Meteorological Services and Sugar Research Institute of Fiji for

providing the weather data.

I am very grateful to my friends and colleagues: Bipen, Linda, Kilifi, Lucille, Semi,

Sam, Ranjila, Priya, Shirleen, Diana, Sharon, Linda and Ravi. I am also very

xix

appreciative of my research assistant, a wise colleague and a very dear friend, Mr.

Prashneel Kumar, for his valuable help and support throughout this research.

xx

Abstract

This research evaluated how climate change and climate variability affects potato

production in Fiji. It also investigated which crop management strategies optimised

potato yield in both current and future climate scenario. The study was conducted

using the Decision Support System for Agrotechnology Transfer (DSSAT) v4.5

SUBSTOR Potato model. Three experimental replicate plots using the cultivar

Desiree were located in Banisogosogo, Fiji. The values for LAI, fresh and dry weight

of aboveground and belowground plant components (stem, leaves, roots and tubers)

were taken for four progressive harvest (T1, T2, T3 and T4). Prior to the calibration

process, the model’s simulated values did not agree the observed values. This was

because the model had never been tested in the Tropics for Desiree variety. The

DSSAT SUBSTOR Potato model was calibrated using the local experimental field

data, local soil and weather data of the growing season. The calibration steps

involved the recalculation of soil water content and the recalibration of genetic-co-

efficient to suit the temperature and daylength regime similar to the experimental

conditions. The value for R2 for replicate plot 1, replicate plot 2 and replicate plot 3

were R2= 0.88, R2= 0.66 and R2= 0.92 respectively. The calibration showed that the

simulated data was in good agreement with the observed data. The DSSAT

SUBSTOR Potato model was simulated under current climate conditions for

Banisogosogo, Koronivia and Nacocolevu using Desiree variety. The three locations

gave different tuber yield. The potential yield obtained for Banisogosogo, Koronivia

and Nacocolevu were 5561 kg/ha, 9811 kg/ha and 8347 kg/ha respectively while the

non-potential yield for Banisogosogo, Koronivia and Nacocolevu were 4478 kg/ha,

4373 kg/ha and 5405 kg/ha respectively. Other significant findings to emerge from

this simulation was that the optimum row spacing should be 30 cm or 40 cm and the

optimum planting depth of 1.5 cm or 2 cm is required to optimise the yield.

Likewise, under climate variability simulation, the percentage difference for tuber

dry yield shows variation. Neutral year gave the highest yield for Banisogosogo and

Nacocolevu (3527.85 kg/ha and 2880 kg/ha respectively) with El Niño giving a

percentage difference of 61.4% and 29.94% respectively. On the other hand, El Niño

provided the highest yield for Koronivia (4465.2 kg/ha) with La Niña providing a

percentage difference of 31.19%. Furthermore, future climate simulations were also

xxi

conducted with Pacific Climate Change Science Programme for medium (A1B) and

high (A2) emission scenarios for 2030, 2055 and 2090 using Desiree for the three

locations. The three locations showed a similar trend. Under future emission

scenarios, the LAI and the tuber initiation day increased while the tuber production

(tuber dry and fresh weight) decreased. Under Banisogosogo future climate

simulations, no tuber yield was noticed under 2055 potential simulation and 2090

simulations for A1B and A2 emission scenario. Koronivia simulations indicated that

no tuber yield was noticed under 2090 potential simulation for A1B and A2 emission

scenario while under Nacocolevu simulations, the tuber yield was possible under all

emission scenario. The optimisation treatments indicated that optimum planting

depth of 1.5 cm or 4 cm is required to optimise yield. Simulations were also

conducted for Sebago and Russet Burbank variety. This simulation indicated that

Sebago gave higher yields under current and future climate scenario whereas Russet

Burbank gave zero or negligible yield under future climate scenario.

xxii

Abbreviations

A1B Pacific Climate Change Science Program Medium

Emission

A2 Pacific Climate Change Science Program High Emission

AFILE

Average File

B1 Pacific Climate Change Science Program Low Emission

BWAH By-product removed during harvest (kg (dm)/ha)

CEC Cation Exchange Capacity

CIP International Potato Centre

CSIRO

Commonwealth Scientific and the Industrial Research

Organisation

DSSAT Decision Support System for Agrotechnology Transfer

DUL Drained Upper Limit

ENSO El Niño Southern Oscillation

FMS Fiji Meteorological Services

G CMs Global Circulation Models

GHGs

Green House Gases

GDP Gross Domestic Product

xxiii

G2 Leaf Expansion Rate ( cm2/m2/d)

G3 Tuber Growth Rate (g/m2/d)

IBSNAT The International Benchmark Sites Network for

Agrotechnology Transfer

IPCC International Panel on Climate Change

LAI Leaf Area Index

LL Lower Limit

NAO Northern Atlantic Oscillation

NASA National Aeronautics and Space Administration

PACE-SD Pacific Centre for Environment and Sustainable

Development

PAR Photosynthetically Active Radiation

PCCSP Pacific Climate Change Science Program

PD Determinancy

PDO Pacific Decadal Oscillation

pH Percentage Hydrogen

PICs Pacific Island Countries

xxiv

P2 Photoperiod Sensitivity (dimensionless)

R CMs Regional Climate Models

SAM Southern Annular Modes

SDBM Bulk Density

SPCZ SPCZ South Pacific Conversion Zone

SSAT Soil Moisture Upper Limit Saturated

SUBSTOR Simulation for Underground Storage Bulking Organ

SWAD Stem dry weight(kg/ha)

TC Critical Temperature (oC)

TDAT Tuber initiation date (Julian Date Format)

TFILE Time series file

TMAX Maximum Temperature

TMIN Minimum Temperature

UNFCCC United Nations Framework Convention on Climate

Change

USAID United States Agency for International Development

USEPA United States Environmental Protection Agency

xxv

UWAD Tuber dry weight (kg/ha)

UYAD Tuber fresh weight (t/ha)

UWAH Tuber dry weight (kg/ha)

UYAH Tuber fresh weight (t/ha)

xxvi

Table of Contents

Dedication ............................................................................................................... xvii

Acknowledgment ................................................................................................... xviii

Abstract ..................................................................................................................... xx

Abbreviations ......................................................................................................... xxii

Table of Contents .................................................................................................. xxvi

List of Figures ......................................................................................................... xxx

List of Tables ....................................................................................................... xxxiii

Chapter 1 Introduction .............................................................................................. 1

1.0 Background of study .......................................................................................... 1

1.1 Research Objectives and Aims ........................................................................... 2

1.1.1 Aim .............................................................................................................. 2

1.1.2 Objectives .................................................................................................... 2

1.2 Organisation of Thesis ....................................................................................... 3

Chapter 2 Literature review ..................................................................................... 4

2.0 Overview of global climate change and climate variability ............................... 4

2.1 Impacts of climate change and climate variability on food security .................. 5

2.2 Overview of the Fiji Islands ............................................................................... 7

2.2.1 Geographical location .................................................................................. 7

2.2.2 Socio-economic background of Fiji Islands ................................................ 7

2.2.3 Climate of Fiji .............................................................................................. 8

2.2.4 Agriculture in Fiji ...................................................................................... 15

2.2.5 Potato production in Fiji ............................................................................ 16

2.3 Origin of potato ................................................................................................ 17

2.4 Biology of potato .............................................................................................. 17

2.5 Tuber formation and potato nutrient content .................................................... 19

2.6 World potato production .................................................................................. 20

2.7 Potato production in Asia and the Pacific region ............................................. 21

2.8 Cultivation, harvesting and storage of potato. .................................................. 21

2.9 Physiology of potato ......................................................................................... 22

xxvii

2.10 Climate model system in prediction of agriculture and crop growth ............. 24

2.11 Implication of study ....................................................................................... 28

Chapter 3 Calibration of DSSAT SUBSTOR Potato Model ................................ 29

3.0 DSSAT overview ............................................................................................. 29

3.1 Methodology .................................................................................................... 30

3.1.1 Experimental site ....................................................................................... 30

3.1.2 Climate trends of field site ......................................................................... 31

3.1.3 Reason for site selection ............................................................................ 31

3.1.4 Data collection, treatments and importations ............................................ 31

3.1.5 Rainfed and nitrogen response experiment................................................ 37

3.1.6 DSSAT SUBSTOR Potato model calibration ........................................... 38

3.1.7 Genetic co-efficient adjustments ............................................................... 39

3.1.8 Analysis and outputs .................................................................................. 40

3.2 Results .............................................................................................................. 40

3.2.1 Soil information and water content recalculation ...................................... 41

3.2.2 Weather information .................................................................................. 42

3.2.3 Experimental file ....................................................................................... 44

3.2.4 Calibration ................................................................................................. 45

3.3 Discussion ........................................................................................................ 51

3.4 Recommendation .............................................................................................. 57

3.5 Experiment limitation ....................................................................................... 57

Chapter 4 Simulation of Desiree potato variety growth and yield in three

different sites (Banisogosogo, Koronivia and Nacocolevu) in Fiji using the

calibrated DSSAT SUBSTOR Potato Model ......................................................... 59

4.0 Introduction ...................................................................................................... 59

4.1 Methodology .................................................................................................... 61

4.1.1 Simulation sites .......................................................................................... 61

4.1.2 Reason for site selection ............................................................................ 62

4.1.3 Data collection, treatments and importations ............................................ 63

4.1.3 Model simulation under current climatic and potential conditions ........... 63

4.1.4 Optimisation treatment through sensitivity analysis.................................. 65

4.1.5 Climate variability (El Niño Southern Oscillation) ................................... 66

4.1.6 Data analysis .............................................................................................. 67

xxviii

4.2 Results .............................................................................................................. 67

4.2.1 Weather data .............................................................................................. 67

4.2.2 Simulation results for Desiree ................................................................... 69

4.2.3 Optimisation treatments ............................................................................. 77

4.2.4 Climate variability (El Niño Southern Oscillation) simulation ................. 81

4.3 Discussion ........................................................................................................ 85

4.3.1 Desiree potential simulation ...................................................................... 85

4.3.2 Desiree non-potential simulation ............................................................... 85

4.3.3 Optimisation treatments ............................................................................. 90

4.3.4 ENSO effect ............................................................................................... 94

4.4 Recommendations ............................................................................................ 95

4.5 Challenges faced .............................................................................................. 98

Chapter 5 Simulating the impacts of future climatic scenario on potato

production in Fiji using the calibrated DSSAT SUBSTOR Potato model ......... 99

5.0 Introduction ...................................................................................................... 99

5.1 Methodology .................................................................................................. 102

5.1.1 Simulation sites ........................................................................................ 102

5.1.2 Data collection, treatments and importations .......................................... 102

5.1.3 Optimisation treatments ........................................................................... 103

5.1.4 Model output analysis .............................................................................. 103

5.2 Results ............................................................................................................ 103

5.2.1 Future climate simulations ....................................................................... 104


5.3 Discussion ...................................................................................................... 116

5.3.1 Potential simulations ................................................................................ 116

5.3.2 Non-potential simulations ........................................................................ 119


5.4 Recommendations .......................................................................................... 129

5.5 Research limitations ....................................................................................... 130

Chapter 6 Simulating the performance of potential potato varieties under

current and future climates in Banisogosogo, Fiji Islands. ................................ 132

6.0 Introduction .................................................................................................... 132

6.1 Methodology .................................................................................................. 133

xxix

6.1.1 Model simulation under current climate .................................................. 133

6.1.2 Model simulation under future climatic conditions ................................. 134

6.1.3 Model output analysis .............................................................................. 134

6.2 Results ............................................................................................................ 135

6.2.1 Current climate simulations ..................................................................... 135


6.2.3 Climate variability (El Niño Southern Oscillation) simulation ............... 144

6.2.2 Future climate simulations ....................................................................... 147

6.3 Discussion ...................................................................................................... 154

6.3.1 Current climate simulations ..................................................................... 154

6.3.2 Future climate simulation ........................................................................ 159

6.4 Recommendations .......................................................................................... 166

6.4.1 Current climate simulation ...................................................................... 166

6.4.2 Future climate simulation for Sebago variety .......................................... 166

6.4.2 Future climate simulation for Russet Burbank variety ............................ 167

6.5 Research limitations ....................................................................................... 167

Conclusion .............................................................................................................. 168

Bibliography ........................................................................................................... 171

Appendices .............................................................................................................. 198

Appendix 1: Calibration ....................................................................................... 198

Appendix 2: Current climate simulations ............................................................. 201

Appendix 3: Future climate simulations .............................................................. 213

Appendix 4: Simulation of other potato varieties ................................................ 260

xxx

List of Figures

Figure 2.0 shows Fiji’s location in the Pacific…………………………… 8

Figure 2.1 shows the number of cyclones passing within 400km of

Suva…………………………………………………………………………

10

Figure 2.2 shows the southern oscillation graph…………………………… 11

Figure 2.3 shows the South Pacific Convergence Zone and Intertropical

Convergence Zone…………………………………………………………..

11

Figure 2.4 shows the mean annual temperatures and rainfall for Suva and

Nadi………………………………………………………………………….

12

Figure 2.5 shows the atmospheric concentration of atmospheric carbon

dioxide for all three emission scenarios……………………………………..

14

Figure 2.6 shows the transfer and spread of potatoes……………………… 17

Figure 3.0 shows the location of experimental site (Banisogosogo)………. 30

Figure 3.1 showing collection of sample soil and marking

of replicate plot ……………………………………………………………

33

Figure 3.2 shows the design of the experimental replicate plot……………. 34

Figure 3.3 shows tuber initiation during T1……………………………… 35

Figure 3.4 shows measurement of LAI using AccuPAR

LP-80 and tubers during T2……………………………………................

35

Figure 3.5 shows tubers at T2 and pest infections………………………. 35

Figure 3.6 shows nematode infection , beetle infection

(Papuana spp.) and pest Quantula striata………………………………..

36

Figure 3.7 shows tubers during final harvest………………………………. 36

Figure 3.8 shows the daily rainfall for Banisogosogo from

January to September, 2012…………………………………………………

43

Figure 3.9 shows the daily maximum and minimum temperature for

Banisogosogo from January to September, 2012………………………….

44

Figure 3.10 shows the simulated and observed for Desiree

before calibration……………………………………………………………

47

Figure 3.11 shows the potential and observed calibration results for

Desiree ………………………………………………………………

49

Figure 3.12 shows the evaluation of potato yield (dry matter in kg/ha) at

Banisogosogo Fiji for year 2012…………………………………………….

50

Figure 4.0 shows the location of three simulation sites……………………. 62

xxxi

Figure 4.1 shows the monthly average rainfall for Banisogosogo,

Koronivia and Nacocolevu…………………………………………………..

68

Figure 4.2 shows the monthly average maximum and minimum

temperature for Banisogosogo, Koronivia and Nacocolevu……………….

68

Figure 4.3 shows Banisogosogo potential and non-potential simulations

for stem weight, LAI, tuber fresh weight, leaf weight, tops weight and

tuber dry weight……………………………………………………………..

71

Figure 4.4 shows Koronivia potential and non-potential simulations for

LAI, leaf weight, stem weight, tuber fresh weight, tuber dry weight and

tops weight…………………………………………………………………..

72

Figure 4.5 shows Nacocolevu potential and non-potential simulations for


tops weight…………………………………………………………………..

73

Figure 4.6 shows the impact of precipitation, total water and nitrogen

content in soil on non-potential tuber dry yield and LAI for Desiree in

Banisogosogo over the growing season………………………………….

74



Koronivia over the growing season………………………………………

75



Nacocolevu over the growing season……………………………………

76

Figure 5.0 shows the potential future climate simulations for

Banisogosogo for LAI, leaf weight, stem weight, tuber fresh weight, tops

weight, tuber dry weight and tuber fresh weight……………………………

107

Figure 5.1 shows the non-potential future climate simulation for

Banisogosogo under A1B and A2 emission scenario for LAI, leaf weight,

stem weight, tuber fresh weight, tuber dry weight and tops weight………..

108

Figure 5.2 shows the Koronivia potential future climate simulations for


tops weight…………………………………………………………………..

109

Figure 5.3 shows the Koronivia non-potential future climate simulations

for LAI, leaf weight, stem weight, tuber fresh weight, tuber dry weight and

xxxii

tops weight…………………………………………………………………. 110

Figure 5.4 shows the Nacocolevu potential future climate simulations for


tops weight…………………………………………………………………

111

Figure 5.5 shows the Nacocolevu non-potential future climate simulation

for LAI, leaf weight, stem weight, tuber fresh weight, tuber dry weight and

tops weight…………………………………………………………………..

112

Figure 6.0 shows Banisogosogo potential and non-potential simulations

for Sebago for LAI, leaf weight, stem weight, tuber dry and fresh weight

and tops weight……………………………………………………………...

137

Figure 6.1 shows the Banisogosogo potential and non-potential

simulations for Russet Burbank for LAI, leaf weight, stem weight, tuber

dry and fresh weight and tops weight……………………………………….

138


content in soil on non-potential tuber dry yield and LAI for Sebago in

Banisogosogo over the growing season…………………………………….

139


content in soil on non-potential tuber dry yield and LAI for Russet Burbank

in Banisogosogo over the growing season…………………………………..

140

Figure 6.4 shows Banisogosogo future climate potential simulations for

Sebago for LAI, leaf weight, stem weight, tuber dry and fresh weight and

tops weight…………………………………………………………………..

148

Figure 6.5 shows Banisogosogo future climate non-potential simulations

for Sebago for LAI, leaf weight, stem weight, tuber dry and fresh weight

and tops weight……………………………………………………………...

149

Figure 6.6 shows Banisogosogo future climate potential simulations for

Russet Burbank for LAI, leaf weight, stem weight, tops weight, tuber fresh

and tuber dry weight under A1B and A2 emission scenario………………..

152

Figure 6.7 shows Banisogosogo future climate non potential simulations

for Russet Burbank for LAI, leaf weight, stem weight, tops weight, tuber

fresh and tuber dry weight under A1B and A2 emission scenario…………

153

xxxiii

List of Tables

Table 2.0 Shows the climate trends in Fiji from 1961- 2010………………. 13

Table 2.1 shows the projected average annual air temperature change for

Fiji under low emission scenario (B1), medium emission scenario (A1B)

and high emission scenario (B2) ……………………………………………

15

Table 2.2 shows the taxonomy of potato…………………………………… 18

Table 3.0 shows the planting information………………………………….. 39

Table 3.1 shows the DSSAT SUBSTOR Potato model input and data

requirement.………………............................................................................

40

Table 3.2 shows the chemical properties of Banisogosogo soil for each

replicate plot…………………………………………………………………

41

Table 3.3 shows the physical properties of Banisogosogo soil for each

replicate plot…………………………………………………………………

42

Table 3.4 shows the recalculated water content of bulk density, lower limit

of extractable soil water, drained upper limit and saturation for

Banisogosogo soil…………………………………………………………..

42

Table 3.5 shows Banisogosogo monthly average weather data…………… 43

Table 3.6 shows the values for AFile (Average File)……………………… 44

Table 3.7 shows the values for Time Series file (Tfile)…………………… 45

Table 3.8 shows the simulation results for Desiree before calibration…….. 46

Table 3.9 shows the recalibration of genetics co-efficient………………… 48

Table 3.10 shows the simulation results after calibration

for Desiree ………………………………………………………………….

48

Table 4.0 shows the summary of simulations at Banisogosogo, Koronivia

and Nacocolevu……………………………………………………………...

64

Table 4.1 shows the irrigation application during simulations at

Banisogosogo, Koronivia and Nacocolevu………………………………….

65

Table 4.2 shows the fertiliser application during simulations at

Banisogosogo, Koronivia and Nacocolevu…………………………………

65

Table 4.3 shows the average monthly maximum temperature, minimum

temperature and rainfall for Banisogosogo, Koronivia and Nacocolevu….

69

xvii

Table 4.4 shows the potential (non-limiting for water and nitrogen) and non-

potential (limiting for water and nitrogen) simulation results of current climate

for Banisogosogo, Koronivia and Nacocolevu………………………...

70

Table 4.5 shows the yield at different planting time for Banisogosogo,

Koronivia and Nacocolevu……………………………………………………..

77

Table 4.6 shows the yield at different row spacing……………………………. 78

Table 4.7 shows the yield under different irrigation application and irrigation

amount…………………………………………………………………………...

78

Table 4.8 shows the planting depth with corresponding yield………………… 79

Table 4.9 shows the application of fertiliser (banded beneath surface) and

corresponding yield for Banisogosogo using weather data of 2012……………

79

Table 4.10 shows the optimum fertiliser and irrigation management simulation

for Banisogosogo using weather data of 2012………………………………….

80


corresponding yield for Koronivia using weather data of 2010.……………….

80

Table 4.12 shows the corresponding yield of optimum fertiliser and irrigation

management for Koronivia using weather data of 2010………………………..

80


corresponding yield for Nacocolevu using weather data of 2010………………

81


management for Nacocolevu using weather data of 2010………………………

81

Table 4.15 shows the impact of ENSO on potato yield for non-potential

Banisogosogo simulation from 1983-2012……………………………………...

82

Table 4.16 shows the 7 year average for ENSO yield for Banisogosogo……… 82


Koronivia simulation from 1990 to 2010……………………………………….

83

Table 4.18 shows the 5 Year average for ENSO yield for Koronivia………….. 83


Nacocolevu simulation from 1990-2010………………………………………..

84

Table 4.20 shows the 5 Year average for ENSO yield for Nacocolevu………. 84

Table 5.0 shows the environment modification climatic parameters…………... 102

Table 5.1 shows the Banisogosogo simulation results of future climate

simulation………………………………………………………………………..

104

xviii

Table 5.2 shows the Koronivia simulation results of future

climate simulation……………………………………………………….............

105

Table 5.3 shows Nacocolevu simulation results of future climate

simulation……………………………………………………………………….

105

Table 5.4 shows the optimisation treatments at Banisogosogo..……………… 113

Table 5.5 shows the optimisation treatments at Koronivia…………………….. 114

Table 5.6 shows the optimisation treatments at Nacocolevu…………………… 115

Table 6.0 shows the genetic co-efficient of Sebago and

Russet Burbank………………………………………………………………….

133

Table 6.1 Shows Sebago, Russet Burbank and Desiree potential

and non-potential simulation under current climate simulations for

Banisogosogo……………………………………………………………………

135

Table 6.2 shows the yield at different planting time for Banisogosogo……… 141

Table 6.3 shows the yield at different row spacing for Banisogosogo………. 142

Table 6.4 shows the yield under different irrigation application and irrigation

amount for Banisogosogo……………………..................................................

142

Table 6.5 shows the planting depth with corresponding yield for

Banisogosogo……………………………………………………………………

143


corresponding yield for Banisogosogo for Sebago variety………………….....

143

Table 6.7 shows the optimum fertiliser and irrigation management for

Banisogosogo for Sebago variety…………………………………………….....

143


corresponding yield for Banisogosogo for Russet Burbank……………………

144

Table 6.9 shows the optimum fertiliser and irrigation management for

Banisogosogo for Russet Burbank………………………………………………

144

Table 6.10 shows the impact of ENSO on potato yield for Sebago variety at

Banisogosogo from 1983-2012………………………………………………….

145

Table 6.11 shows the 7 year average for ENSO yield for Banisogosogo for

Sebago variety. The highest yield was obtained by the neutral

years……………………………………………………………………………..

146

Table 6.12 shows the impact of ENSO on potato yield for Banisogosogo for

Russet Burbank from 1983-2010……………………………………………….

146

xix

Table 6.13 shows 7 year average for ENSO yield for Banisogosogo for Russet

Burbank. The highest yield was obtained under neutral

year………………………………………………………………………………

147

Table 6.14 shows the Banisogosogo future climate simulation for Sebago

variety……………………………………………………………………………

147

Table 6.15 shows the optimisation treatments for Sebago…………………….. 151

Table 6.16 shows Banisogosogo simulation results of future climate simulation

for Russet Burbank Variety for A1B and A2 emission

scenario…………………………………………………………………………..

151

Table 6.17 shows the optimisation treatments for Russet Burbank…………… 154

1

Chapter 1 Introduction

1.0 Background of study

Potato was first introduced in Fiji by European settlers in 1860’s and since then the

consumption of potatoes has increased. Potato has been cultivated in Fiji for many

years but only on a small scale. Only recently (1960’s onwards), potato was given a

priority in research programmes. Since then, potato trials have been conducted in

Sigatoka Valley, Nadi, Ba and Nadarivatu to investigate suitable cultivar, row

spacing, plant growth and yield, disease tolerance including intercropping. Tests

were conducted to identify suitable varieties for cultivation in Fiji in collaboration

work with the International Potato Centre (CIP) in the 1980’s. Research has also

been conducted on better ways of producing and making accessible seed including

vegetative planting resources for farmers by numerous scientists over the decades in

Fiji (Autar, 2009). Currently, Fiji imports around 16 000 tonnes of potatoes annually

valued at $17 million. Under its Import Substitution programme, the government of

Fiji aims to reduce its potato imports and boost local food production (Mudaliar,

2007; Fiji Times Online, 2010; The Fijian Government, 2010).

This study was set out to evaluate the impacts of climate change and climate

variability on potato production in Fiji and to help identify strategies to optimise

potato yield and identify which varieties of potato can perform better under current

and future climatic conditions of Fiji. Many Pacific Island Countries (PICs) are

among the first to suffer the impacts of climate change. Subsistence and commercial

agriculture are being adversely affected by climate change in PICs (Asian

Development Bank, 2009; Wairiu et al., 2012). Some of the challenges to agriculture

and crop production in PICs include: lack of consideration of climate change in

current development strategies and activities (Quity, 2012). There is also low level of

awareness on climate change, inadequate capacity at community, national and

regional level to develop and implement adaptation strategies. Not only these, but

limited resources and inaccessibility on reliable data on climate change impact on the

region and as well as the lack of resource allocation in agricultural sector are also

some of the challenges faced by the agricultural sector (Wairiu et al., 2012).

2

The motivation for this study arose when I was given an assignment to develop a

research proposal with the theme of climate change and sustainable development.

With the knowledge that food security is an important issue for climate change, I

began to research on the impacts of climate change on staple crops. Coincidently,

potato was being reintroduced to Fiji during that time. However, the fate of potato

industry remained uncertain. Therefore, I decided to look into crop modeling which

not only helps to predict the behavior of the biological system but can also be used

for agronomic efficiency and to promote environmental management. At the same

time, the crop model software Decision Support System for Agrotechnology Transfer

(DSSAT) v4.5 was introduced in Pacific Centre for Environment and Sustainable

Development (PACE-SD) for the students to utilise for research projects. It is hoped

that this research will assist the agricultural sector to improve the yield of potatoes

and to support farmer’s livelihood and to provide food security for Fiji using modern

technology.

1.1 Research Objectives and Aims

1.1.1 Aim

This study investigates how climate change and climate variability affects potato

production in Fiji.

1.1.2 Objectives

The objectives of this study are to:

� calibrate the DSSAT SUBSTOR Potato Model’s performance in

Banisogosogo, Fiji using Desiree variety.

� simulate potato production under current climatic condition and optimise crop

management to maximise potato yield under current conditions.

� simulate potato production under future climatic scenario and to optimise

crop management strategies to maximise potato yield under stressful

conditions.

� identify other potato varieties that may perform better under Fiji’s current and

future conditions using simulation approach.

3

1.2 Organisation of Thesis

This thesis is divided into six chapters. Chapter 1 provides background information

and states the aims and research objectives. Chapter 2 is a literature review and sets

the context of how climate change and climate variability affects food security,

overview of Fiji Islands and potato production in Fiji. It also gives information on

origin, history, distribution and cultivation of potatoes, the use of climate model

systems in prediction of agriculture and crop growth and implication of study.

Chapter 3 calibrates the DSSAT SUBSTOR Potato model in Banisogosogo Fiji using

Desiree variety. Chapter 4 discusses how current climate and climate variability has

had an impact on potato production in Banisogosogo, Koronivia and Nacocolevu and

what strategies are used to maximise potato yield. Chapter 5 describes how future

climate change (Pacific Climate Change Science Program emission scenario for

medium emission (A1B) and high emission (B2) for 2030, 2055 and 2090) will

affect potato production in Banisogosogo, Koronivia and Nacocolevu and what could

be some of the strategies to maximise potato yield. Finally, Chapter 6 states which

cultivar can perform better in Fiji’s current and future climate conditions.

4

Chapter 2 Literature review

2.0 Overview of global climate change and climate variability

According to the United Nations Framework Convention on Climate Change

(UNFCCC), climate change is defined as a “change in the climate which is attributed

directly or indirectly to human activity that alters the composition of global

atmosphere and which is in additional to natural climate variability observed over

comparable time periods” (Intergovernmental Panel on Climate Change, 2001;

United Nations Framework Convention on Climate Change, 2012). Climate

variability is defined as “variation in the mean state and other statistics (such as

standard deviation, the occurrence of extremes) of the climate on all temporal and

spatial scales beyond that of individual weather events over a period of time, for

example, month, season or a year” (Intergovernmental Panel on Climate Change,

2001; World Meteorological Organisation, 2003).

The Intergovernmental Panel on Climate Change (IPCC), in its Climate Change

2007: Synthesis Report, stated that “warming of the climate system is unequivocal,

as is now evident from observations of increases in global average air and ocean,

widespread melting of snow and ice and rising global average sea level”. The IPCC

has acknowledged changes in atmospheric concentration of greenhouse gases

(GHGs), land cover and solar radiation to alter the energy balance of the climate

system and act as the key drivers of climate change (Intergovernmental Panel on

Climate Change, 2007a). The global concentrations of greenhouse gases (GHGs),

such as, carbon dioxide, methane and nitrous dioxide has increased noticeably owing

to anthropogenic causes since 1750’s and now far exceed the pre-industrial values.

The global atmospheric concentration of carbon dioxide has increased from 280 ppm

from pre-industrial times to 379 ppm in 2005 which surpasses by far the natural

range over the last 650, 000 years as determined for ice-cores. The primary reasons

for the global increase in the atmospheric concentration of carbon dioxide are due to

changes in fossil fuel usage and changes in land use patterns whereas the reason for

increase in the concentrations of methane and nitrous oxide is due to agriculture.

Since 1850, eleven of the last twelve years (1995-2006), rank amongst the twelve

warmest years on record. Evidence also exists for the changes in global circulation,

for example, the poleward shift and strengthening of the westerly winds and

noticeable changes in ocean biogeochemistry and salinity. Precipitation events will

5

also be strengthened in a warmer world, with the extensive increase in heavy

precipitation events and increased likelihood of flooding. Warming is also consistent

with sea-level rise. The global average sea level rose at an average rate of 1.8 mm

[1.3 to 2.3 mm] per year over 1961-2003 (Intergovernmental Panel on Climate

Change, 2007b). North Atlantic has noted fluctuations in westerlies and storm track

and these are fluctuations are described by the Northern Atlantic Oscillation (NAO).

In the Southern Hemisphere, due to robust warming over the Antarctic Peninsula

and, to a lesser extent, cooling over the continental parts of Antarctica, there has been

a change in circulation related to an increase in Southern Annular Modes (SAM)

from 1960s to the present (Solomon et al., 2007).

In the Pacific, changes have also been noticed in the ocean-atmosphere interactions.

On interannual time scales, El Niño Southern Oscillation (ENSO) is the dominant

form of global-scale variability. The shift in climate in 1976-1977, is related to the

phase change in Pacific Decadal Oscillation (PDO). This brought about more El

Niño events and variations in the evolution of ENSO, which affected many areas.

Over the 20th century, the Pacific has experienced considerable low frequency

atmospheric variability, with extended periods of weakened circulations (1900-1924,

1947-1976) also periods of strengthened circulations (1925-1946, 1977-2005).

Interannual variability, such as ENSO and North Atlantic Oscillation (NAO), can

also be related to regional average sea level rise (Solomon et al., 2007). Models also

indicate that there is a tendency for more warming in the central and east Pacific than

in the west. Although this east west difference in warming is usually less than a

degree in multi-model ensemble (Boer et al., 2001).

2.1 Impacts of climate change and climate variability on food security

Food security is defined as “a situation when all people, at all times, have physical,

social and economic access to sufficient, safe and nutritious food that meets their

dietary needs and food preferences for an active and healthy life” (Food and

Agriculture Organisation of the United Nations, 2002).

Climate change affects all four dimensions of food security: food

availability/production, food accessibility, food utilisation and food system stability.

Agriculture-based systems is at the immediate risk of crop failure, new pests and

disease and loss of livestock (Food and Agriculture Organisation of the United

6

Nations, 2008a). The rise in global mean average temperature will have an impact on

agronomic productions, such as, changes in growing seasons. Sea level rise will

diminish the quantity of land accessible for agriculture (Darwin, 2001). Due to

climate change agriculture has to compete with other sectors for land, water and also

investment of time and money (Schi mmelpfennig et al., 1996). In developing

countries, almost 11% of arable land be could be affected by climate change (Food

and Agriculture Organisation of the United Nations, 2007). Climate change impacts

can be divided into two impact categories: biophysical impacts and socio-economic

impacts. Biophysical impacts include: physiological effects on crops, pasture, forests

and livestock, changes in the quality and quantity of land, soil and water resources,

increased weed and pest challenges, changes in the spatial and temporal distribution

of impacts, rise in sea level, changes to ocean acidity, increased intensity and

frequency of storms, drought and flooding, altered hydrological cycles and changes

in precipitation have implication for future food availability (Food and Agriculture

Organisation of the United Nations, 2007). On the other hand, socio-economic

impacts consist of: decline in yields and production, reduced marginal GDP from

agriculture, fluctuation in world market prices, increase in the number of people at

risk at geographical distribution of trade regimes, migration and civil unrest (Food

and Agriculture Organisation of the United Nations, 2007). Due to multiple socio-

economic and bio-physical factors affecting food systems and food security, the

capacity for a food system to diminish its vulnerability to climate change is not

uniform as this will require better systems of food production, food distribution and

economic access for food systems to be able to manage with climate change. The key

to addressing these challenges can be conducted through developing a framework

system which deliberates all stresses to the food system such as social stress, political

stress and environmental stress and the communities can learn from past experience

of environmental stress, relevant past adaptation measures and indigenous

knowledge (Gregory et al., 2005).

Climatic variability will bring additional challenges (Gregory et al., 2005). Short

term rainfall variability can pose as a major risk factor. It can lead to soil moisture

deficits, crop damage and crop disease all related to distribution of rainfall and

associated humidity (Ludi, 2009).

7

2.2 Overview of the Fiji Islands

2.2.1 Geographical location

The Fiji Islands comprises of more than 330 islands and has an overall land area of

18 300 km2 and an Economic Exclusive Zone of 1.3 million km2. The geographical

co-ordinates of the Fiji Islands are between longitude 175oEast and 178oWest and

latitude 15o and 22oSouth. It is located about 2100 km north of Auckland, New

Zealand. The biggest island is Viti Levu (10 429 km2), which covers 57% of the total

land area and Vanua Levu which covers 5556 km2. The capital of Fiji is Suva with

the center of politics and economy in Fiji. Other main islands include Taveuni (470

km2), Kadavu (411 km2), Gau (140 km2) and Koro (104 km2) (Pacific Islands

Climate Change Assistance Programme and Fiji Country Team, 2005; Vanualailai

and UNFCCC Consultant, 2008; GEF et al., 2009; Australian Bureau of Meterology

and CSIRO, 2011). The geography of Fiji is dominated by mountainous terrain and

due to earthquakes and volcanic eruptions, the terrain of Fiji is somewhat rugged.

The highest point is Mount Tomanivi which is 1324m. The larger islands of Fiji,

such as Viti Levu and Vanua Levu, are as a result of volcanic activities of the past

whereas the smaller islands are made up of coral reefs. The flora and fauna of the

island is determined by the tropical climate of the islands (Maps of World, 2010,

2011).

2.2.2 Socio-economic background of Fiji Islands

Fiji became independent in 1970 after being a British Colony for nearly a century

(CIA World Factbook, 2012a). It is blessed with natural resources such as timber,

gold, copper, fish, offshore oil and hydro-power (Maps of World, 2010, 2011). The

Fijian population comprises several ethic groups such as Fijian, Indian descendents,

Chinese, European and others (Pacific Islands Climate Change Assistance

Programme and Fiji Country Team, 2005). From July 2011, the estimated population

of Fiji stands at 890, 057 (Index Mundi, 2012). In recent years, there has been an

increasing trend of urbanisation which has led to the development of major squatter

settlements around town areas (Pacific Islands Climate Change Assistance

Programme and Fiji Country Team, 2005). Fiji’s Gross Domestic Product (GDP) at

current basic price for 2010 was estimated to grow at 7.9%. This is an increase from

8.8% from 2009 when the growth rate was considered to be -0.9%. The Transport

8

Storage sector and the Communication sector were the largest contributors to the

growth in GDP which accounted for 15.6% of the GDP followed by Manufacturing

sector and Wholesale and Retail sector which contributed 15.0% and 11% of the

GDP respectively (Bureau of Statistics, 2011).

Figure 2.0 shows Fiji’s location in the Pacific. Source: (Maps of World, 2012)

2.2.3 Climate of Fiji

2.2.3.1 Current climate

The annual average temperature of Fiji is between 20oC-27oC. The variations in the

temperature from one season to another are relatively small and can strongly be

linked to changes in the surrounding ocean temperature. Between the coolest months

(July and August) and the warmest months (January to February), the average

temperature difference is only about 2°C-4°C (Fiji Meterological Services, 2012).

The mean night temperature of Fiji can be as low as 18oC (in the central parts of the

main islands, the temperatures can be as low as 15oC) whereas the maximum mean

day time temperatures can be as high as 32oC. Past records also indicate that extreme

temperatures as low as 8°C and as high as 39.4 °C have been documented in Fiji (Fiji

9

Meterological Services, 2012). The inter-annual variations in temperature are low,

generally ranging from ± 0.5 °C about the long term mean (Pacific Islands Climate

Change Assistance Programme and Fiji Country Team, 2005).

Much of Fiji’s rainfall is closely interconnected to the movement of South Pacific

Convergence Zone (SPCZ) which is closest to Fiji during the wet season. On Viti

Levu and Vanua Levu rainfall is strongly influenced by high mountain peaks of up to

1300m. On the south-eastern slopes of Viti Levu, near Suva, the mean annual rainfall

is around 3000 mm. However, the lowland, on the western side of Viti Levu, are

sheltered by mountain and have an average annual rainfall of 1800 mm with a well-

defined dry season which proves favourable to crops such as sugar cane. Other

smaller islands of the Fiji group receive various amounts of rainfall in accordance

with their location and size, ranging from 1500 mm to 3500 mm. Fiji has a distinct

wet season and dry season from November-April and May-October respectively. The

wettest month of Fiji is usually March and the driest month is almost always July

(Pacific Islands Climate Change Assistance Programme and Fiji Country Team,

2005; Australian Bureau of Meterology and CSIRO, 2011; Pacific Climate Change

Science Program, 2011a; Fiji Meterological Services, 2012). The seasonal cycle is

also strongly affected by the position of SPCZ (Government of the Republic of Fiji,

2012).

Tropical cyclones in Fiji are confined to the months of November to April, with the

highest occurrence around January to February. From 1969 to 2010, 70 tropical

cyclones have passed within 400km of Suva (Pacific Climate Change Science

Program, 2011a), with an average of one to two cyclones per season. On average,

Fiji is affected by ten to fifteen cyclones per decade out of which two to four

cyclones do severe damage. The dominant north-west to south-east tracks gives some

increased risk of damage in the outer lying north-west island groups. Winds over Fiji

are light or moderate. However, from November to April, tropical cyclones and

depression can cause high winds (Fiji Meterological Services, 2012). Tropical

cyclone occurrence risk for Fiji for the 2012/2013 season is high (100%) while the

severe tropical cyclone risk is projected to be low to moderate (38%). For Fiji, 1-2

cyclones projected to occur this season, of which 1 may reach or exceed Category 3

(Fiji Meteorological Services, 2013a). Strong winds are uncommon with the

exception of cyclones which generally occur from November to April (Fiji

10

Meterological Services, 2012). Large scale flooding is mostly associated with the

passage of tropical cyclone or depression which results in extended heavy rainfall

(Pacific Climate Change Science Program, 2011a).

Figure 2.1 shows the number of cyclones passing within 400km of Suva. In the 41 year

period from 1969-2010, 70 tropical cyclones have passed within 400 km of Suva, with an

average of 1-2 cyclones per season. Source: (Pacific Climate Change Science Program,

2011a).

Due to ENSO, Fiji’s climate differs from year to year. There are two extreme phases

of ENSO: El Niño and La Niña. The El Niño phase usually brings dry seasons that

are drier and cooler than normal, while the La Niña phase brings wetter than normal

conditions (Agrawala et al., 2003; Pacific Climate Change Science Program, 2011a).

The El Niño event which causes a shift of SPCZ to northeast is the major cause of

drought in Fiji. During 1997/1998 ENSO event period, the rainfall was strengthened

from April to June of 1997 where the SOI for June reached its lowest value since

1905. In 1997, September recorded 20-50 % below average rainfall in most parts of

the country. The western division recorded less than 10 mm of total rainfall, that is,

below 7 % of the average rainfall. In December, 50-90 % below average rainfall was

recorded (Pacific Islands Climate Change Assistance Programme and Fiji Country

Team, 2005; Australian Bureau of Meterology and CSIRO, 2011). In El Niño years

tropical cyclones are confined to the months of October and May (Pacific Climate

Change Science Program, 2011a). River flooding occurs almost every wet season

and occasionally in the dry season during La Niña. Most meteorological droughts

since 1920 have been closely associated to El Niño events. Recent severe droughts

11

have occurred in 1987, 1992, 1997-98, 2003 and 2010 (Pacific Climate Change

Science Program, 2011a).

Figure 2.2 shows the southern oscillation graph. The SOI is shown in bars while the 5 month

running mean is shown in continuous pink line. The figure shows that SOI has risen from -

3.6 in February to +11.1 in March. Source: (Fiji Meteorological Services, 2013b).

Figure 2.3 shows the South Pacific Convergence Zone and Intertropical Convergence Zone.

The arrow indicates near surface winds while the blue shading represents the band of rainfall

convergence zone, the red dashed oval represents the Western Pacific warm pool while the H

represents high pressure system. Source: (Pacific Climate Change Science Programe,

2011b).

12

Since 1950, both the annual minimum and maximum temperatures have increased in

both Suva and Nadi. The maximum temperature has increased at a rate of 0.15oC per

decade and 0.18 oC per decade in Suva and Nadi Airport respectively. Since 1950,

there is no clear trend in seasonal or annual rainfall in Suva and Nadi (Pacific

Climate Change Science Program, 2011a). There is also increasing variability

between El Niño and La Niña-like conditions in Fiji (Agrawala et al., 2003). Recent

analysis of surface temperature data in Fiji suggests that 1990s were the warmest

years on record (relative to 1961-1990 average) (Lal, 2004).

Figure 2.4 shows the mean annual temperatures and rainfall for Suva and Nadi. Light blue,

dark blue and grey bars indicate El Niño, La Niña and neutral years respectively. Source:

(Pacific Climate Change Science Programe, 2011a).

13

Table 2.0 shows the climate trends in Fiji from 1961- 2010.

Climate Variable Observed Trends (1961-2010)

Rainfall � Very weak positive linear trend in annual rainfall over Fiji.

� Annual increase of 0.65 mm/ year (0.03%) per annum.

� A weak decreasing linear trend in wet season rainfall with a decrease of 1.30

mm/season (approximately 0.08%).

� A weak decreasing linear trend in dry season rainfall with a increase of 0.76

mm/season (approximately 0.11%).

Maximum Air

Temperature

� The average annual air temperature increased by 1.1oC

� The average warm season maximum temperature increased by 1.2oC

� The average cool season maximum temperature increased by 1.0oC

Minimum Air

Temperature

� The annual minimum temperature increased by 0.6oC

� Increasing trend in annual warm season minimum air temperature and increased

by 0.7oC

� The cool season minimum air temperature increased by 0.6oC

Source: (Government of the Republic of Fiji, 2012).

2.2.3.2 Future projected climate change

Projections of future climate change and its impact on the society and the

environment have been crucial for the emergence of climate change as a global

challenge for public policy and decision making. These climate projections are built

on a range of scenarios, models and simulations which contain a number of

embedded assumptions (Dessai et al., 2009; Clarke et al., 2011). The Pacific Climate

Change Science Program (PCCSP) is a vital element of the initiative which is liable

for undertaking fundamental research to support decision making in the 15 partner

countries in partnership with Australia’s Bureau of Meteorology and the

Commonwealth Scientific and the Industrial Research Organisation (CSIRO)

(Pacific Climate Change Science Program, 2011b). Scientists from PCCSP have

assessed twenty four climate models from around the world and eighteen of these

best characterise the climate of western tropical Pacific region. These eighteen

models have been used to develop climate projections for Fiji (Pacific Climate

Change Science Program, 2011a). Under PCCSP Climate Futures, three emission

scenarios are available: B1 (low emission), A1B (medium emission) and A2 (high

emission) for three time periods 2030, 2055 and 2090 with three levels of detail and

these are basic, intermediate and advanced (Pacific Climate Change Science

Program, 2011b). Projection for all emission scenarios indicate that the annual

14

average air temperature and sea surface temperature will increase in the future for

Fiji. It is also projected that increase in mean temperature will also result in an

increase in the number of hot days and warm nights and a decline in cooler weather.

However, there are some uncertainties concerning rainfall and drought projections as

model results are not consistent. Projections from these models indicate that there is

likely to be a decrease in the number of tropical cyclones by the end of 21st century.

However, the intensity of cyclones will increase. There is also likely to be a 2%-11%

increase in the average maximum wind speed of cyclone and an increase of rainfall

intensity of about 20% within the 100km of cyclone centre (Australian Bureau of

Meterology and CSIRO, 2011; Pacific Climate Change Science Program, 2011a).

Recognising the seriousness of climate change concerns, the PICs are implementing

the Pacific Islands Framework on Climate Change 2006-2015. This will aid in

addressing issues of improving the understanding of climate change and providing

training, education and awareness on this matter (Pacific Climate Change Science

Program, 2010). Although the practice of reporting multi-model ensemble climate

projections is well covered, there is still discussion in regards to the most appropriate

levels of evaluating models performance (Irving et al., 2011). Due to the level of

uncertainty attached with climate change projection, it makes it challenging for the

local government to prioritise its commitment to adaptation (SMEC Australia, 2010).

Figure 2.5 shows the atmospheric concentration of atmospheric carbon dioxide for all three

emission scenarios. The projections for amostpheric cardon dioxide concentrations for each

scenario are shown as blue, green and orange. The projections for 2030, 2055 and 2090

(relative to 1990) were calculated using the average value of the 20 year periods 2020-2039,

2046-2065 and 2080-2099 (relative to 1980-1999) to minimise the effect of natural

15

variabilty. The grey bars represent the 20 year periods. Source: (Australian Bureau of

Meterology and CSIRO, 2011; Kumar, 2011).

Table 2.1 shows the projected average annual air temperature change for Fiji under

low emission scenario (B1), medium emission scenario (A1B) and high emission

scenario (B2).

Note: Values represent 90% of the range of models and changes are relative to average

of the period of 1980-1999.

2030 (oC) 2055 (oC) 2090 (oC)

Low Emission Scenario (B1) 0.2-1.0 0.5-1.5 0.7 -2.1

Medium Emission Scenario (A1B) 0.2-1.2 0.9-1.9 1.3-2.9

High Emission Scenario (B2) 0.4-1.0 1.1-1.7 2.0-3.2

Source: (Pacific Climate Change Science Program, 2011a).

2.2.4 Agriculture in Fiji

Agriculture, which was once considered a major stronghold of Fiji’s economy, now

comprises only 16.1 % of the nation’s GDP (CIA World Factbook, 2012b). In 2010

the contribution of the agricultural sector declined by 9.8 % (Bureau of Statistics,

2011). One of the main crops of Fiji is sugar cane which occupies more than 50% of

the arable land and contributes to 9 % of the GDP. The sugar industry engages

almost 13% of the labour force and generates almost 30 % of the export (Rosillo-

Calle and Woods, 2003; Food and Agriculture Organisation, 2009). However, this

industry is facing many challenges such as low investments, uncertain leaseholds and

land ownership rights along with the government’s political instability (CIA World

Factbook, 2012a). Also, the horticultural sector such as ginger, tropical fruits, root

crops and vegetables is now considered the fastest growing agricultural sector.

Traditional tree crops such as cocoa and copra have been deteriorating over the past

decades. Fiji’s import substitution industries consists of rice, dairy, poultry, beef,

pork and tobacco (Food and Agriculture Organisation, 2009).

16

2.2.5 Potato production in Fiji

In recent years, the cultivation of potatoes in lowland tropics, such as Fiji, has gained

popularity. These areas were previously considered unsuitable for potato production

(Iqbal, 1991). More than half of the potatoes in developing countries are cultivated in

warm tropical climates (Midmore and Rhoades, 1988). To maximise potato yield in

the lowland tropics which is characterised by high temperature and high solar

radiation, it is important to maintain high foliage productivity as long as possible

given the fact that growth, development and yield of potatoes is guided by factors

such as high soil and high air temperature, plant density, water stress, availability of

nutrients and the utilisation of solar radiation (Allen, 1978; Ewing, 1981b).

In Fiji, potato was first introduced by European settlers in 1860’s. It was first grown

in Rewa and the crop spread to the Western Division. Between 1937 and 1940, Up

To Date and Early Rose varieties were considered as suitable for local consumption.

Varieties that were suggested in the 1940’s were Brownell and Bismarck (Autar,

2009). Fiji potato acreage in 1981 was around 9 hectares which represented a decline

since 1960 (Iqbal, 1982). Some potential areas for potato production have been

identified and these are Navai, Nadrau Province, Nausori Highlands and Sigatoka

Valley (Autar, 2009). However, one of the major challenges faced by the potato

industry is that of bacterial wilt. To overcome this challenge, the Department of

Agriculture ran a number of unsuccessful experiments with a number of varieties

over the past years. Since the 1970’s, there were introduction of resistant varieties of

potatoes from Mauritius and the International Potato Center (CIP) of Peru. During

1980 and 1991, potatoes worth FJ$2.63 million and FJ$5 million respectively were

imported into Fiji (Autar, 2009). There has always been a potential to grow potatoes

in Fiji, primarily, as a substitute for imports whereby saving foreign exchange.

However, production over the years has varied both in terms of hectare planted and

yield obtained. Variable yield have been obtained in Fiji ranging from 5t/ha to

25t/ha. Production of potatoes is confined to months of May-June to September to

October. In Fiji, Lands Development Authority was a central body that was

responsible for organising large scale planting in mid-1960’s in Nadarivatu, Nausori

Highlands and Sigatoka Valley. Sigatoka Valley has been considered the major

center of production for potatoes with the largest area of production at 134 hectares.

Unfortunately, the Nausori Highlands and Nadarivatu scheme had to come to a stop

17

in 1969 due to high incidence of bacterial wilt. Other challenges to successful growth

of potatoes include: blackleg and root rot, no provision for irrigation during dry

seasons, the attitude of farmers (farmers do not regard potato production as a

business) and potato seeds storage (Iqbal, 1982).

2.3 Origin of potato

Potato, Solanum tuberosum L., is a food crop that is cultivated and consumed

worldwide. It is a basic food source and primary source of income for many of the

communities (Ovchinnikova et al., 2011). Potato made its way around the globe in

the 16th century when the Spanish brought it to Europe from South American Andes.

It later found its way to Asia in the 17th century and to Africa in the 19th century

(Food and Agriculture Organisation of the United Nations, 2008c, d).

Figure 2.6 shows the transfer and spread of potatoes. Source: (The Natural History Museum,

2012).

2.4 Biology of potato

Solanum tuberosum is an herbaceous annual with short vegetative period that grows

up to 100 cm tall and produces a tuber called a potato (Food and Agricultural

Organisation of the United Nations, 2008). It belongs to the family of flowering

plants (Food and Agriculture Organisation of the United Nations, 2008c, d).

18

Table 2.2 shows the taxonomy of potato.

Taxonomic Rank Latin Name

Kingdom Plantae

Phylum Anthophyta

Division Magnoliopsida

Order Solanales

Family Solanaceae

Genus Solanum

Species tuberosa

Common names Potato, tater, spud, tuber

Source: (Bradley, 2009).

In potato, the development of the sprout is dependent on the ageing processes of seed

tuber and is vital for the growth and performance that are taking place in the seed

tuber. This development can be influenced by external factors such as light,

photoperiod, temperature and relative humidity. Many stems can result from a seed

tuber depending on the number of buds. For the duration of the first growing period,

these stems share resources from the same seed tuber and later these stems become

independent units and compete with each other for resources such as light, water and

nutrients. The stems of below-ground part of potatoes are typically round and

massive, whereas in the upper part it is hollow and angular. The potato plant has a

central leaf per node. The early leaves are small but the latter leaves are alternate and

pinnate compound with three or four pairs of large ovate elliptical leaflets with much

smaller leaves found in between. Daylength and temperature also have an influence

on the number of leaves. Long day length and high temperature increase the number

of leaves on secondary stem (Struik, 2007a). At higher temperature, the effect of

photoperiod is stronger. Tuberisation starts before all stolons have been formed. The

tubers begin as stolon swelling and go through different phases of tuber set, tuber

growth and tuber maturation. However, when induction to tuberisation is interrupted,

they can show secondary growth, especially when the plant is exposed to heat or

irregular water supply. The potato has a fibrous root system. The root system is

rather weak and the water and nutrient use efficiency in potato are low. Hence, the

19

crop is very sensitive to drought and poor soil structure. The roots occur not only on

the stems but also on stolons and the tubers (Struik, 2007a).

The potato plant has five growth stages and these are: sprout development, plant

establishment, tuber initiation, tuber bulking and tuber maturation. Depending on the

planting date, physiological age of the seed tubers, cultivar and other environmental

factors, stages I and II last from 30 to 70 days. The period between emergence and

tuber initiation is reduced by short days and temperatures less than 20ºC. The ideal

temperature for the second stage is considered to be between 16ºC-18ºC (van

Heemst, 1986). The third stage, which is tuber initiation, is when tubers start to form

stolon tips. In the fourth stage, the tuber cells start to enlarge due to accumulation of

water, nutrients and carbohydrates. When the conditions are favorable, elongation of

stolons stops and tuber elongation begins (Xu et al., 1998). Depending on the

cultivar, tuber bulking can last up to three months. Tuber maturation is the final stage

where photosynthesis slowly declines, leaves begin to turn yellow, tuber growth

slows and the vines die (Alberta Agriculture Food and Rural Development

Department., 2005).

2.5 Tuber formation and potato nutrient content

Tuber dry matter concentration and tuber size are the two important aspects of

potato. When a tuber is initiated as swelling stolon tip, the dry matter concentration

is about 11%. As the tuber size increases, the dry matter increases along with the

starch. The final dry matter concentration depends on the cultivar, length of growing

seasons, water availability, solar intensity and the average temperature during the

growing season (Haverkort and Verhagen, 2008).

During tuber enlargement, the tuber stores large quantities of carbohydrates (starch

which is 20% of fresh weight) (Fernie and Willmitzer, 2001) and substantial amounts

of proteins and are also low in fat (Food and Agriculture Organisation of the United

Nations, 2008i). It also contains toxins, either natural (example glycoalkaloids) or

food-borne toxins (example acrylamide) (Haase, 2008) and significant amounts of

iron, potassium, zinc, vitamin B and traces of manganese, chromium, selenium and

molybdenum and about half of the daily adult requirement of vitamin C which

enhances iron absorption. An average serving of potatoes with skin has 10 percent of

recommended daily intake of fiber (International Potato Center, 2008a). Potatoes

20

also contribute important amounts of dietary fiber (up to 3.3 %), ascorbic acid (up to

42 mg/100g), potassium (up to 693.8 mg/100g), total caroteniods (up to 2700

mcg/100g) and antioxidant phenols such as chlorogenic acid (up to 1570 mcg/100g)

(Burlingame et al., 2009).

2.6 World potato production

Since the 1960’s, substantial changes have occurred in potato production, types of

potato crops grown and yield in Northern Britain. There is already evidence that

farmers in the northern latitude have already began to adapt to climatic changes such

as warmer temperatures (Plauborg et al., 2010). In the early 1990, majority of the

potatoes were cultivated and consumed in Europe, North America and the former

countries of the Soviet Union. There has been a major increase in potato production

and demand in Asia, Africa and Latin America from 30 million tonnes in 1960 to

more than 165 million tonnes in 2007. One third of the potatoes are now harvested in

China and India (Food and Agricultural Organisation of the United Nations, 2008)

growing 22 % of all potatoes (Jansky et al., 2009). In addition, in harsher climates,

potatoes can produce more nutritious food than any other major crop – up to 85 % of

the crop is edible human food compared to 50 % in cereals. Over the last 10 years,

world potato production has increased at an annual rate of 4.5% and this rate will

increase strongly in the future. This growth has exceeded the production of many

other major food commodities in developing countries (Food and Agricultural

Organisation of the United Nations, 2008).

Potato is regarded as the “frontline in the fight against world hunger and poverty”

(Food and Agricultural Organisation of the United Nations, 2008). The International

Year of Potatoes has contributed to the following United Nations Millennium

Development Goals (MDGs): Goal 1-Eradicate extreme hunger and poverty, Goal 4-

Reduce child mortality, Goal 5-Improve maternal health, Goal 7-Ensure

environmental sustainability and Goal 8-Develop a global partnership for

development (Chiru et al., 2008; Food and Agriculture Organisation of the United

Nations, 2008f, e, b; Thornton et al., 2010). Potato is included in The International

Treaty on Plant Genetic Resources for Food and Agriculture that aims at the

“conservation and sustainable use of crop plant diversity and fair and equitable

21

sharing of benefits derived from their use” (Food and Agriculture Organisation of the

United Nations, 2008a).

2.7 Potato production in Asia and the Pacific region

In many places, such as the highlands of southern China and Vietnam, potato is

emerging as an offseason crop and is being planted as a rotation crop such with rice

and maize. Likewise, in the lowlands of Bangladesh and eastern India, potato is

emerging as a winter cash crop. In Philippines and Indonesia, potato production

satisfies the need for domestic and regional snack food industries. Potato has also

become an important crop in Asia and the Pacific (Changchui, 2008). Three broad

bands of potato production are clearly visible in the Asia-Pacific Region. The first

one runs from west to east across the sup-tropical lowlands of the major river basins

in South Asia and includes Bangladesh, India and Pakistan. The second band runs

from interior of China and ends at Siberian border of Russia. The last band is

characterised by productivity in the Pacific which includes Australia and New

Zealand (Pandey, 2008) and Fiji (Autar, 2009).

In the developing world, potato is crucial to food security. The production of

potatoes has doubled in the last 20 years. Overall the consumption of potatoes has

increased from 9-15 kg/capita and by 2020 the demand for potatoes is anticipated to

double that of 1993. In the Tropics, the farmers can harvest 15-25 tonnes of potatoes

within 90 days of planting. However, there are many challenges that the potato

production faces in the Tropics and Potato Late Blight is one of them affecting 3

million hectares of potato producing land with an estimated damage of $2.75 billion

annually (Food and Agriculture Organisation, 2008).

2.8 Cultivation, harvesting and storage of potato.

Loam or sandy loam soils that are rich in organic matter with good drainage and

aeration and with a pH of 5.2 and 6.4 are appropriate for potato cultivation. In most

cases the soil requires three ploughings with regular hallowing and rolling so that the

soil is well aerated, has good drainage and is soft. Cultivation of potato without

tillage assistances to restore soil fertility, increase potato yields and reduce the need

for fertiliser and fuel (Food and Agriculture Organisation of the United Nations,

2008g). Potato is usually grown from “seed potato”, that are small tubers or pieces of

22

tubers sown to a depth of 5-10 cm. A good seed would usually be disease free, well

sprouted and about 30-40 grams in weight. Small, single eyed seeds are planted 25

cm to 30 cm apart in furrows 60-90 cm apart (Iqbal, 1982). Upon germination, shoots

usually emerge within 21 days. Development of potato canopy takes about 4 weeks

during which weeds must be controlled in order to give the potato crop a

“competitive advantage”. By this time the primary roots have elongated to about 4-6

inches long. During the phase of emergence, nutrients are supplied to the plant

directly from the seed reserves (Lang et al., 1999). Potato cultivation requires very

little labour. Upon maturity, tubers are easily detached from their stolons. The leaves

start to turn yellow indicating maturity. To facilitate harvesting, the potato vines

should be removed two weeks before the potatoes are dug up. Potatoes are usually

harvested using a spading fork. During harvesting it is important to avoid bruising or

other injury which may provide entry points for disease. Proper storage of potatoes is

also vital. If the potatoes are to be stored before consumption, they should be left in

the soil so that the skin of the potato thickens which will prevent storage disease and

shrinkage due to water loss. On the other hand, when left in the ground for too long,

potatoes become susceptible to fungal incrustation called black scurf (Food and

Agricultural Organisation of the United Nations, 2008).

2.9 Physiology of potato

Meteorological factors govern crop productivity by influencing its transpiration,

photosynthesis and respiration in ways to control its growth and development of the

particular crop throughout its physiological cycle at a given site (Pereira et al., 2008).

Among the main environmental factors that govern the growth and development of

plant comprises global solar radiation flux density, air temperature and available soil

water content (Coelho and Dale, 1980), solar radiation, radiation use efficiency,

harvest index (Iqbal, 1991), air temperature, soil temperature, photoperiod, soil

moisture and crop water use (World Meteorological Organisation, 2010).

Potato uses water rather efficiently and has a high harvest index of about 0.75 as

compared to cereals which has a harvest index of 0.5 (Vos and Haverkort, 2007b).

Adequate water is essential for high tuber yield and good tuber quality (World

Meteorological Organisation, 2010). Potatoes usually face water stress due to their

shallow root system (Curwen, 1993). Seventy percent of water uptake occurs at 0.3

23

m and 100 % for upper 0.4 m to 0.6m soil depth (Food and Agriculture Organisation

of the United Nations 2012). For best yields, a 150 day crop requires 500 mm-700

mm of water (Food and Agricultural Organisation of the United Nations, 2008).

Excessive soil moisture can result in seed-piece decay and erratic plant development,

fungal infections such as potato blight, rots and wilts (World Meteorological

Organisation, 2010).

There is a positive relationship between intercepted radiation and the quantity of dry

matter produced (Haverkort, 2007b), the number of tubers (Firman and Daniels,

2011), biomass partitioning (Hoogenboom et al., 1985), leaf area index (LAI) and

the percentage of ground covered by the canopy (Haverkort, 2007a). Tuberisation is

also favourable when plants are exposed to shorter days (Zaag et al., 1986; Garner

and Blake, 1989; Iqbal, 1991; Jackson, 1999; Haverkort, 2007a). LAI is also a

significant property of plant canopy. It is the area of leaves (m2) per area of soil (m2)

(Haverkort, 2007a). Leaf area index of 3-4 is adequate to intercept most of the

incident radiation for maximising the yields of tubers that are grown under temperate

conditions (Harris, 1983).

Temperature also plays a critical role in tuber formation. Potato production is best

suited to temperatures of tropical highlands of 15-18 ºC (Haverkort, 1990). However,

the optimal temperature for potato production is 22 ºC (Burton, 1981). Warm

temperatures favour vegetative growth (Ewing and Struik, 1992) while cool

temperature favour tuber growth (Ewing, 1981c; Khedher and Ewing, 1985). Cool

night temperatures are required by potatoes for tuberisation. Changes in temperature

and rainfall pattern affects root crops (Gawander, 2007). Positive effects of elevated

carbon dioxide have also been observed on the yield (Wheeler et al., 1991;

Schapendonk et al., 1995; Miglietta et al., 1998; Rosenzweig and Hillel, 1998; Te

mmerman et al., 2000; Lawson et al., 2001b; Te mmerman et al., 2002; Persson et

al., 2003; Te mmerman et al., 2007; Bindi, 2008).

Nitrogen uptake efficiency of potatoes under current best practice management is

approximately 65 %, an efficiency rate which is comparable to that of corn and

wheat (Lang et al., 1999). The first in-season nitrogen application should occur after

tuber initiation as during this stage 30-40 % of nitrogen has been taken up. It is also

24

crucial to maintain nitrogen levels during tuber bulking which is considered as an

important stage in potato development (Lang et al., 1999). The requirement of

nitrogen fertiliser ranges from 2.5 to 5.9 kg ha-1 per tonne of tuber yield (Hagman et

al., 2009). To cater for the nitrogen needs of this crop, nitrogen fertilisers can be

supplied in many ways: ammonia, nitrate, urea (Bucher and Kossmann, 2007).

Nitrogen uptake is also cultivar specific (Love et al., 2005; Fontes et al., 2010).

There are also many pests of potatoes that can reduce both tuber yield and quality

(Food and Agriculture Organisation of the United Nations, 2008h), such as Colorado

potato beetle (Leptinotarsa decemlineata), potato tuber moth (Phthorimaea

operculella), leaf miner fly (Liriomyza huidobrensis), serious soil pests of potato in

the temperate regions, such as, Globodera pallid and G. rostochiensis, potato Late

Blight (Phytophthora infestans) (Świeżyński and Zimnoch-Guzowska, 2001;

International Potato Center, 2008a; Cooke et al., 2011), bacterial wilt and Potato

Blackleg (International Potato Center, 2008b) and aphids as virus vectors (Food and

Agriculture Organisation of the United Nations, 2008d). Due to climate change, the

break out period of pest and disease will become earlier than expected (Hannukkala

et al., 2007) with expansion into areas that were previously safe from this disease.

Little effort has been made to document the non-targeted impacts of pesticides on

potatoes (Thornton et al., 2010) .

2.10 Climate model system in prediction of agriculture and crop growth

The most managed ecosystem in the world is agriculture (Stoorvogel et al., 2004).

Agriculture deals with diverse range of issues, such as, sustaining agricultural

livelihoods, protecting water quality and other resources and climate change

mitigation strategies. All these demand the need to create information to make

informed decision on how agricultural systems respond to changes in external

stimuli, such as, changes in technology, policies and climate change (Stoorvogel et

al., 2004). A lot of emphasis is being employed in estimating crop productivity as a

function of climatic elements by different weather models (Pereira et al., 2008). Crop

modeling began since 1960’s (Bouman et al., 1996) and has extensively developed

over the last 30 years (Kumar et al., 2000). Crop simulation models are a

representation of a simplified crop production system made up of non-linear

mathematical equations to provide a systematic analysis of the crop production

25

system (Easterling et al., 1996). Crop growth models are complex as they require a

series of complex interaction between soil, plant, weather and growth reducing

factors such as pest, disease and weeds (Singh, 1999). These need to be accurately

represented in order to produce a reliable picture. The crop models used today are of

a multi-disciplinary approach with environmental impact models and economic

models linked together (Bosello and Zhang, 2005). A total of 104 countries were

reported having utilised modeling done at individual sites within these countries

(Rivington and Koo, 2010). Crop models aim to explain crop development and

behaviour, yield and quality as a function of environmental variables, genetic

variables and disease (MacKerron, 2007). In most crop models, growth and

development are driven by solar radiation. However, in some models, temperature

plays a more dominant role than solar radiation (MacKerron, 2007). Crop modeling

requires input data of weather, soil, crop management data and pest data

(Chakrabarti, 2005). Some of the widely accepted plant physiology models include

CERES-Maize, CERES-Wheat, CERES- Rice (Singh et al., 1993), SOYGRO (Jones

et al., 1988), SIM-POTATO (Bosello and Zhang, 2005), CropSyst (Stockle et al.,

2003), EPIC model (Bulatewicz et al., 2009; Pumijumnong and Arunrat, 2012),

APSIM (Cheeroo-Nayamuth, 1999), Tradeoff Analysis Model (Stoorvogel et al.,

2004) and CANEGRO (Cheeroo-Nayamuth, 1999). Models can also be linked to

other models to gather more information on crop growth and development (Roth et

al., 1991).

Potato crop modeling developed and started in the late 1970’s. The earliest potato

models attempted to represent the typical characteristics of potatoes such as growth

pattern, effect of temperature, day length and cultivar (Ng N and Loomis, 1984).

Later, potato models looked at approximation of potato yield based on solar radiation

and the length of growing season (MacKerron, 2007). Growth rate in many potato

models is calculated on the basis of light interception (Haverkort and MacKerron,

1995). Numerous efforts have been made to simulate potato growth and development

(Te mmerman et al., 2007), leaf area index (Van Delden A et al., 2001; Gonzalez-

Sanpedro et al., 2009), tuber fresh weight (Medany, 2006), comparison of present

yield and predicted yield using the climate change data (Abdrabbo et al., 2010), to

study the interaction of water and nitrogen supply and weather variability (Bowen,

2003), the influence of water supply and potential evaporation on growth and

26

development of potatoes and the effect of length and temperature on potato growth

and development (MacKerron, 2007) and pathogenic fungus infection on potatoes

(Iglesias et al., 2007). A common figure of models is that they are usually location

and cultivar specific. Hence, by limiting the geographical locations, the

environmental inputs required to run the model also becomes limited (Griffin et al.,

1993). Although tuber initiation is sensitive to photoperiod, a number of models

ignore the effect of photoperiod on tuber initiation. Other models, such as POMOD

(Kadaja and Tooming, 2004), SIMPOTATO (Rosenzweig et al., 1996), LINTUL

(Haverkort et al., 2004), AQUACROP (Patel et al., 2008), TUBERPRO and

LINTUL-POTATO (Kooman and Haverkort, 1995) also exist that can be used to

determine potential potato yield.

One of the most widely used crop model is Decision Support System for

Agrotechnology Transfer (DSSAT) which was developed by International

Benchmark Sites Network for Agrotechnology Transfer (IBSNAT). The DSSAT

model incorporates 27 different crop models, such as, cereal grain, legumes and root

crops (Hoogenboom, 2003). DSSAT has been registered in 100 countries (Rivington

and Koo, 2010). The IBSNAT crop models were used to estimate how climate

change will affect crops at 112 sites in 18 countries, representing key production

areas and vulnerable regions at low, mid and high latitudes. DSSAT was successful

in predicting the relative nitrogen rates and potato nitrogen uptake for six different

fertility scenarios centered on different organic and inorganic nitrogen source

combinations (Snapp and Fortuna, 2003). The DSSAT SUBSTOR Potato model has

a great potential to simulate potato growth and evaluating potential changes in

management in many regions (Griffin et al., 1993). It also has the capability to

simulate on a daily basis the growth and development potato crop using data such as

on climate, soil, management and cultivar (Jones et al., 2003; Knox et al., 2010). The

DSSAT SUBSTOR Potato model is branched into four main sub-models which are:

phonological development, biomass formation and portioning, soil water and

nitrogen balance which work simultaneously to provide a realistic description of the

plant-soil-atmosphere system. The genetic-coefficient is used by the model to control

development phases, such as, tuber initiation, leaf area development and tuber

growth rate. The DSSAT SUBSTOR model has been extensively used for global

crop studies and recently for climate change impacts. The reason for its widespread

27

use is due to its ability to simulate canopy response to temperature and radiation

change to incorporate the direct effects of atmospheric carbon dioxide on potato

production (Knox et al., 2010). The DSSAT SUBSTOR Potato model has also been

validated in many areas and used to simulate physiological processes and yield of

potatoes under current climatic and future conditions (Bowen et al., 1998;

International Potato Center, 1999; Fleisher et al., 2003; Štastna and Dufkova, 2008;

Abdrabbo et al., 2010; Knox et al., 2010). To simulate the response of different plant

organs and processes over a wide temperature range, the DSSAT SUBSTOR Potato

model uses zero to one relative temperature functions based on mean daily air or soil

temperature. The timing of tuber initiation in the DSSAT SUBSTOR Potato model is

a function of cultivar response to temperature and photoperiod, plant nitrogen status

and soil water status. The DSSAT SUBSTOR Potato model does not simulate the

development of individual leaves but rather the development of the entire canopy

(Griffin et al., 1993). The model also shows poor performance for LAI (as compared

to tuber initiation) due to lack of disease or insect defoliation subroutines in the

model (Griffin et al., 1993; Parry, 2007). The simulation of tuber yield by

SUBSTOR model is also accurate (Griffin et al., 1993).

Crop simulation models have become accurate enough and have become widely

accepted (Easterling et al., 1996). A number of opportunities are now available

through the use of crop models, such as, determining the potential and attainable

yield for various crops, optimising crop management, evaluating the impact of

climate change, forecasting crop yields, estimation of crop growth and yield under

climate variability and increasing the efficiency of multi-environment testing (Singh,

1999), promotes interdisciplinary collaboration (Chakrabarti, 2005), for decision

support system, education and training, research for crop genetic improvement,

policy development (Rivington and Koo, 2010) and statistical links between weather

and yield can also be useful for economic modeling (Roberts et al., 2013).

While crop models have been successful tools in application, they all have their

weakness and fail under certain circumstances (Graeff et al., 1999). The complexity

of real world cannot fully expressed in any model (Raisanen, 2007). Agricultural

interactions are complex that are not fully understood. Hence, the models are a crude

representation of reality. Model quality is also depended on the quality of scientific

28

data used in model development, calibration and validation. When models are

applied in new situations, model calibration and validation are needed to verify the

model because all processes are not understood and the model contains parameter

adjustments to new situations (Cheeroo-Nayamuth, 1999). Other drawbacks

encountered in the use of crop models include the availability of reliable data as

input (Singh, 1999), lack of skilled manpower, lack of knowledge on computer and

computer language, limited awareness and acceptance towards modeling, existing

complexity of biological systems are hard to replicate in models giving rise to

uncertainties and errors, models developed for one specific region cannot be used in

another region until the model is calibrated (Chakrabarti, 2005) and difficulty in

simulating root mass (MacKerron, 2007).

2.11 Implication of study

This research aims to investigate the impact of climate change and climate variability

on Potato production in Fiji. The DSSAT SUBSTOR Potato model was calibrated for

the first time in Fiji using Desiree variety. Potato growth, development and yield

were assessed under current weather conditions and climate variability. Potato

production was also simulated under future climatic scenario using PCCSP medium

(A1B) and high (A2) emission scenarios. Optimisation treatment for crop

management strategies were also carried out for current and future climatic

conditions to sustain potato yields. Finally, through Sensitivity Analysis the best

cultivar was identified for both current and future climatic conditions. It is hoped that

this research will help us to understand the impacts of climate change and climate

variability on potato production in Fiji. It will advise policy makers and farmers on

suitable adaptation measures to improve potato production and contribute to

improvement in livelihood and food security. Furthermore, since this is the first time

that DSSAT SUBSTOR Potato model was calibrated in the Tropical region, it will

deliver an insight on its performance in simulating potato growth in warmer climates.

This knowledge will also help crop model developers to re-adjust crop models and

their performance to real variables in the Tropics. This research will also contribute

to the awareness of impacts of climate change on root crops in the Pacific Island

Countries which will be important for UNFCCC documents such as IPCC reports.

29

Chapter 3 Calibration of DSSAT SUBSTOR Potato Model

3.0 DSSAT overview DSSAT is a personal computer-based program comprising of crop models and data

base management programs for climate, soil and crop management practices

(Rosenzweig and Parry, 1994). Based on the soil-plant-atmosphere dynamics, the

DSSAT crop models can simulate growth, development and yield (Jones et al.,

2003). For data input, daily weather data, soil surface and soil profile information,

and detail crop management data is required by the model. The weather conditions

that are needed to run the model include daily values for minimum and maximum air

temperature, total precipitation, and total solar radiation. Other variables that are

required include latitude to calculate day length and carbon dioxide concentration of

the atmosphere (Hoogenboom et al., 1985).

Before a model is simulated for a particular region, it should be calibrated and

validated. Calibration and validation are two different processes. Calibrated models

have become a significant tool for economic research and policy analysis (Cooley,

1997). Model calibration is defined as adjusting an already existing model to a

reference system. It is necessary to calibrate the model in order to get reliable results

since the model is based on abstractions, idealisation and many disputable

assumptions (Hofmann, 2005). In calibration, model parameters and coding are

changed in order to obtain the best fit for simulation versus observed data (Boote,

1999).

One important task in working with models is testing their performance in wide

range of circumstances so that one is able to identify their scope of validity and

limitations. This is because simulation models are site and crop specific in nature and

cannot be utilised in other areas unless validated under local conditions (Ahmed and

Fayyaz-ul-Hassan, 2011). However, the process of evaluation can be time consuming

and challenging because it requires the collection of large data sets such as weather

data, soil data, crop and field management information over long time periods

(St’astna et al., 2002). A research conducted in Malawi from 1990-1993 aimed to

validate the CERES-Maize model. The results showed that the model worked

30

reasonably well with grain yield (Singh et al., 2002). Various components of

CERES-Rice model have also been validated (Singh and Ritchie, 1993). To validate

the DSSAT SUBSTOR Potato model, that is, to compare model simulated yield to

observed data, formatted yield data is required (Thorpa et al., 2008). The comparison

of simulated and observed data requires retrieval of observed data and converting it

to DSSAT SUBSTOR Potato model format, and comparing simulated outcomes with

observed values (Abdrabbo et al., 2010).

3.1 Methodology

3.1.1 Experimental site

The potato experimental replicate plot was located in Banisogosogo, Rakiraki

(longitude 17˚ 22' 00" South and latitude 178˚ 9' 00" East) (National Geospatial-

Intelligence Agency, 2012b). Rakiraki, which is on the main island of Viti Levu, is a

district in Ra province which is situated halfway between Suva and Nadi along the

Kings Road on the Northern Coast of Viti Levu, Fiji’s largest Island (Urbita, 2012).

Figure 3.0 shows the location of experimental site (Banisogosogo). The experimental plot is

located in the Western Division of Viti Levu. Source: (GoogleEarth, 2012).

31

3.1.2 Climate trends of field site

Banisogosogo (Rakiraki) is located on the Western side of Viti Levu. The Western

Division is the dry zone and is subjected to large seasonal and inter-annual climatic

variation. The annual average surface air temperature increase for Penang (Rakiraki)

is 0.17 oC, with an increase of 0.05 oC per decade. The average yearly rainfall in

Rakiraki is about 2000 mm. During the winter weather in this area, it is clear or

partly cloudy with plenty of sunshine and a well-developed west to north-westerly

winds by day and south easterlies are prominent wind at night and in all months

except for August to March. From August to March the wind is mostly from north

westerly quarter. The nights are relatively cool in this season. However, during

summer there is thunderstorm activity which is responsible for brief but intense

rainfall (Gawander et al., 2012). According to the hybrid climate data obtained from

Fiji Meteorological Services (FMS) and National Aeronautics and Space

Administration (NASA), the average maximum temperature is 29.09 oC while the

average minimum temperature is 22.4 oC.

3.1.3 Reason for site selection

Banisogosogo was selected as the site for experimental replicate plot to calibrate the

DSSAT SUSTOR Potato model because the Agriculture Department of Fiji has

started to distribute potato seeds to farmers at Rakiraki for potato cultivation. This

site was also selected due to logistic reasons, that is, the land owner of the

experimental replicate plot was a family friend which ensured that the experimental

replicate plot was not disturbed.

3.1.4 Data collection, treatments and importations

3.1.4.1 Weather data

The 30 year climatic data for Penang, Rakiraki (which represented weather data of

the experimental replicate plot), such as, daily maximum temperature (Tmax) (oC)

and minimum temperature (Tmin) (oC), and precipitation ( mm) were collected from

the Sugar Research Institute of Fiji (SRIF) and FMS. Solar radiation of the

experimental site was downloaded from the NASA website. When the weather data

had missing values for rainfall and/or temperature they were replaced with NASA

data creating a “hybrid” weather data. The DSSAT format weather data was created

32

in Excel worksheet. This excel worksheet was then imported into WeatherMan tool

in DSSAT. WeatherMan has the ability to produce complete sets of weather data

historically or synthetically created records (Singh et al., 2002).

3.1.4.2 Soil data

There were three replicate experimental plots. Soil samples were collected from each

replicate plot. Using randomised complete block design, nine random sampling pits

were dug, three in each replicate plot. One sample per layer was taken from each

sampling pit of each replicate plot. The soil samples were taken from 0-40 cm, 40-

100 cm and +100 cm. The soil samples were placed in a bucket and samples were

mixed. Stones, plant materials were removed while mixing. From the bucket, 1 kg of

soil was removed and it was placed in sampling bags and sealed tightly. This was

then given to Koronivia Research Station for chemical analysis (pH, organic carbon

and Cation Exchange Capacity). This was done before the planting began. The soil

data was entered into SBuild tool of DSSAT SUBSTOR Potato model.

The soil characteristics that are needed to run the model include albedo of bare soil,

soil surface runoff (expressed as runoff curve number), first stage soil evaporation,

water permeability and drainage from profile. For each layer the following variables

are required: layer thickness, saturated soil water content, drained upper limit of

extractable plant water (field capacity), permanent wilting point, initial soil water

content at the start of soil water balance simulation and relative root weighing factor

(Hoogenboom et al., 1985).

DSSAT’s SBuild tool estimates the above input based texture and organic matter

when actual measurements are not available. Default Medium Silty Clay from

DSSAT soil database was used to calculate model inputs. A conversion factor based

on measured (Banisogosogo soil) and estimated (default soil) bulk density was used

to recalculate the bulk density, soil lower limit, soil upper drainage limit and the soil

saturation for the Banisogosogo soil (replicate plot 1, replicate plot 2 and replicate

plot 3). The conversion factor was calculated by dividing the Default Medium Silty

Clay soil by the bulk density of replicate plot 1, Banisogosogo soil. This gave us the

conversion factor. The replicate plot 1 bulk density was then multiplied by the

conversion factor to obtain the recalculated bulk density. The same procedure was

33

repeated for Lower Limit, Drained Upper Limit and Saturation. This procedure was

used to ensure the values for soil lower limit, drained upper limit, and the saturated

water content were realistic and not biased by the high clay content of Banisogosogo

soil. The calculated values for the lower limit < than drained upper limit < saturated

water content < pore volume fraction.

3.1.4.3 Experimental replicate plot

The experimental replicate plot was located in Banisogosogo, Rakiraki, Fiji. Potato

seeds were planted in 11 rows and 17 columns with a row spacing of 75 cm and 30

cm within the rows (Agriculture Department of Fiji, 2009). The experimental

replicate plot had a length of 7.5 m and a width of 4.8 m (total area was 36 m2 or

0.0036 hectares). Rows 1 and 11 acted as border plants. Rows 2-4 were replicate plot

1, rows 5-7 were replicate plot 2, rows 8-10 were replicate plot 3. Ten columns

allowed harvesting of 2 plants per row, that is, 6 plants for each replicate plot (T1,

T2, T3) with a total of 18 plants for each sampling time and T4 allowed harvesting of

4 plants per row, that is, 12 plants per replicate plot with a total harvest of 36 plants.

The potatoes were planted on 2nd of July 2012 at a depth of 1.5 cm.

Figure 3.1 shows collection of sample soil (left) and marking of replicate plot (right).

34

Figure 3.2 shows the design of the experimental replicate plot. Potato seeds were planted in

11 rows (R1-R11) and 17 columns (C1-C17). The blue box represents each potato plant

while the red rectangle represents each replicate plot. R1 and R11 and C1, C2, C5, C8, C11,

C16 and C17 acted as border plants. Replicate plot 1 is represented by R2-R4, replicate plot

2 is represented by R5-R7 and replicate plot 3 is represented by R8-R10.

3.1.4.4 Harvesting

The first harvest (T1) of potatoes was in the period of tuber initiation where 50%

plants had at least one tuber greater or equal to 1 cm in diameter. The second harvest

(T2) took place on T1 plus 20 days and the third harvest (T3) took place in T1 plus

40 days. The final harvest (T4) was carried out when the green-leaf canopy cover had

reached 20% of the maximum achieved (Hoogenboom et al., 1999).

T1 was carried out on 7th of August. For T1, columns 3, 4 and rows 2-4 represented

replicate plot 1. Replicate plot 2 was represented by columns 3, 4 and rows 5-7 and

replicate plot 3 was represented by columns 3, 4 and rows 8-10. Columns 1, 2, 5 and

rows 1 and 11 acted as border plants.

35

Figure 3.3 shows tuber initiation during T1.

The second harvest (T2) took place on T1 plus 20 days, that is, 27th of August. For

T2, columns 6, 7 and rows 2-4 represented replicate plot 1. Replicate plot 2 was

represented by columns 6, 7 and rows 5-7 and replicate plot 3 was represented by

columns 6, 7 and rows 8-10. Columns 5 and 8 acted as border plants.

Figure 3.4 shows measurement of LAI using AccuPAR LP-80 (left) and tubers during T2

(right).

Figure 3.5 shows tubers at T2 (left) and pest infections (right).

The third harvest (T3) took place on the 16th of September. For T3, columns 9, 10

and rows 2-4 represented replicate plot 1. Replicate plot 2 was represented by

columns 9, 10 and rows 5-7 and replicate plot 3 was represented by columns 9, 10

and rows 8-10. Columns 8 and 11 acted as border plants.

36

Figure 3.6 shows nematode infection (left), beetle infection (Papuana spp.) (center) and pest

Quantula striata.(right).

T4 was carried out on 20th of September. Columns 12-15 and rows 2-4 represented

replicate plot 1. Replicate plot 2 was represented by columns 12-15 and rows 5-7 and

replicate plot 3 was represented by columns 12-15 and rows 8-10. Columns 11, 16,

17 acted as border plants. For each harvest (T1, T2, T3), 2 plants per row were

harvested from each replicate plot, giving a harvest of 6 plants per replicate plot with

a total of 18 plants harvested for all 3 replicate plots. T4 allowed harvesting of 4

plants per row, that is, 12 plants per replicate plot with a total harvest of 36 plants.

Figure 3.7 shows tubers during final harvest and pest infection (right).

3.1.4.5 Fresh aboveground biomass collection

For the aboveground biomass (stem and leaves), the stems were cut as closely to the

ground as possible with a garden pruner. The above ground mass was cleaned by

washing it in a bucket of water until all the dirt was removed. This was then blotted

dry with tissue paper. The weight of the whole above ground biomass was taken on a

digital scale and recorded. The leaves and stems were then separated and weighed

individually. The number of stems and compound leaves for each plant was also

recorded.

37

3.1.4.6 Fresh belowground biomass collection

To take the below ground mass (roots and tubers), the belowground mass was first

cleaned by washing it in a bucket of water until all the dirt was removed and blotted

dry with tissue paper. The whole belowground mass was recorded. The roots, tubers

and mother tuber were then separated and weighed individually. The number of

tubers per plant was also recorded.

3.1.4.7 Dry biomass

The weight of empty brown paper bag was taken. Individual plant components were

then put in brown paper bag and labeled with a pencil. This was then oven dried in

Ontherm Thermotec 2000 at 70oC regularly until the weight was stabilised.

3.1.4.8 Leaf Area Index (LAI)

A portable leaf area meter (LAW-A) was used to calculate the LAI. The LAI for T1

was calculated by using the formula: LAI = (LAM. N)/A, where, N represents the

number of leaves in the plant and A is the area occupied by one plant (Blanco and

Folegatti, 2003).

From T2, the leaf area index was taken using AccuPAR LP-80. The AccuPAR LP -

80 was first configured by entering information such as longitude, latitude, date/time,

daylight saving time, leaf distribution and the external sensor constant. The external

sensor was then attached to AccuPAR LP-80. When the PAR readings were above

600 µmol/m2s and the probe and the external sensor were leveled and the AccuPAR

was calibrated. Only 3 segments out of 8 segments in the probe were activated and

an external sensor was allowed to make simultaneous readings both above and below

canopy. The probe was extended from one mid row to the next mid row. This

allowed to obtain a good representative sample of the area below and between rows

(Decagon Devices, 2006).

3.1.5 Rainfed and nitrogen response experiment

The Rainfed and Nitrogen Response Experiment were also created to determine the

nitrate and ammonia levels in the soil in order to calibrate the model.

38

3.1.6 DSSAT SUBSTOR Potato model calibration

3.1.6.1. AFile (Average File)

AFile was used to generate the measured data for simulation comparison. The

average of T4 harvest was used to create AFile. Data was entered into Excel

Worksheet and imported into DSSAT SUBSTOR Potato model under Experimental

data. The following are the information that was entered into AFile excel:

� TRNO- treatment number (each replicate plot was considered as a

treatment)

� UWAH- tuber dry weight (kg/ha)

� UYAH- tuber fresh weight (t/ha)

� LAIX- leaf area index, maximum

� BWAH- by product removed during harvest (kg (dm)/ha)

� TDAT- tuber initiation date (Julian Date Format)

� HDAT-harvest date (Julian Date Format)

3.1.6.2. TFile (Time Series File)

TFile, or Time Series Data, was created by taking into consideration the average of

all the T series, that is, T1, T2, T3 and T4. Data was entered into Excel Worksheet.

This data was imported into DSSAT SUBSTOR Potato model under Experimental

data. The following are the information that was entered into TFile excel:

� TRNO- treatment number (for each replicate plot T1, T2, T3 and T4 data was

considered)

� Date- for each Time Series in Julian Date Format

� LAID- leaf area index

� UWAD- tuber dry weight (kg/ha)

� UYAD- tuber fresh weight (t/ha)

� LWAD- leaf dry weight(kg/ha)

� SWAD- stem dry weight(kg/ha)

39

3.1.7 Genetic co-efficient adjustments

Since the DSSAT SUBSTOR Potato model had never been tested in the Tropical

regions before, the genetic co-efficient was adjusted directly in the Genetic File by

comparing the simulation results for phenology (tuber initiation, maturity) and

growth (tuber numbers, yield) with observed results.

A similar procedure for the collection of field-data has been described in previous

studies (Manrique and Bartholomew, 1991; Manrique et al., 1991; Medany, 2006;

Štastna and Dufkova, 2008; Abdrabbo et al., 2010; Knox et al., 2010).

Table 3.0 shows the planting information. This information was required to run

DSSAT SUBSTOR Potato model for calibration.

Variable Information

Cultivar Desiree

Starting Date June 1st 2012

Planting Date July 2nd 2012

Harvest Date September 20th 2012

Planting Method Dry seed

Planting Distribution Rows

Row Spacing 75 cm

Plant Population (m2) 4.4

Planting Depth 1.5 cm

Irrigation 14 mm in 4 application

N-fertiliser (N.P.K fertiliser) 80kg/ha in 2 application (5 grams per plant, banded

beneath the surface at 5 cm depth)

40

Table 3.1 shows the DSSAT SUBSTOR Potato model input and data requirement.

Input

Daily weather Daily Tmax and Tmin (oC), precipitation ( mm) and solar radiation (MJ m-2d-1)

Site Latitude, albedo, surface runoff

Soil (by layer) pH, organic carbon content, depth of each layer ( cm), drained upper limit water

content, saturated water content, bulk density

Cultivar Variety name and genetic co-efficient

Management Planting date, population density, planting depth, row spacing

Development Emergence, floral initiation, maturity date

Water balance Soil water content

Nitrogen Soil nitrogen content

Source: (Singh et al., 1990).

3.1.8 Analysis and outputs

The DSSAT SUBSTOR Potato Model was calibrated using the local experimental

field data, local soil and weather data of the growing season (minimum data set). The

genetic co-efficient was recalibrated directly in the Genetic File (as mentioned in

section 3.1.7) in order to obtain new values for genetic co-efficient parameters. These

inputs were then used in model runs and adjusted in order to obtain a best fit between

simulated and observed data. All summary outputs, graphical outputs and initial

analysis were done in DSSAT 4.5 Output and SBUILD Program. Other statistical

analysis was conducted in 2007 Microsoft Excel Spreadsheet. All these outputs are

provided in the Results section.

3.2 Results

This chapter describes the steps taken to calibrate the DSSAT SUBSTOR Potato

model. The field experiment was carried out at Banisogosogo, Fiji during the potato

growing season (July-September) of 2012 using Desiree variety. Three minimum

data sets were needed to evaluate the model and these included soil data, weather

data and crop management data.

The simulation results before calibration indicated that the simulated values were not

in agreement with the observed values (Table 3.8 and Figure 3.10). The calibration

process included the recalculation of soil water content (Table 3.4) and recalibration

41

of the genetic-co-efficient of Desiree variety (Table 3.9). However, after calibration,

the results showed that there was a good agreement between simulation and observed

values (Table 3.10, Figures 3.11 and 3.12).

3.2.1 Soil information and water content recalculation

The soil samples were collected on 2nd July 2012. For soil data, soil chemical and

physical properties were analysed. The chemical test was conducted for pH,

electrical conductivity (EC), percentage of total carbon and nitrogen, cation

exchange capacity (CEC), exchangeable potassium (K) and moisture levels for each

layer of the three replicate plots. Replicate plot 2 shows the highest percentage of

carbon and nitrogen.

Table 3.2 shows the chemical properties of Banisogosogo soil for each replicate plot.

Replicate Plot pH EC (mS/ cm)

Total C (%)

Total N (%)

CEC ( cmol(+)/kg)

Exchangeable K (me/100g)

Moisture

Replicate plot 1 (0-40 cm)

6.5 0.1 2.32 0.19 40.85 0.75 8.2

Replicate plot 1 (40-100)

6.9 0.07 1.92 0.16 43.43 0.24 7.6


7.0 0.06 1.74 0.14 38.96 0.2 8.3


6.7 0.1 3.51 0.29 41.27 0.84 9.0


7.0 0.06 2.27 0.19 37.43 0.24 6.1


7.1 0.05 1.74 0.14 43.71 0.21 8.1


6.8 0.09 2.14 0.18 36.49 0.56 8.8


7.0 0.06 2.32 0.19 43.49 0.21 8.2

Replicate plot 3 100-110 cm)

7.2 0.06 1.74 0.14 43.94 0.2 8.1

The results for physical properties (Table 3.3) indicated that the overall texture of the

soil was clay. There was very low percentage of silt in the soil.

42

Table 3.3 shows the physical properties of Banisogosogo soil for each replicate plot.

Replicate Plot Clay (%)

( <0.002 mm)

Silt (%)

(0.002-0.06

mm)

Sand (%)

(0.06-2 mm)

Texture

Replicate plot 1 (0-40 cm) 57 23 20 Clay

Replicate plot 1 (40-100) 59 22 19 Clay

Replicate plot 1 (100-110

cm)

44 31 25 Clay


Replicate plot 2 (40-7.1100) 57 22 21 Clay


cm)

56 11 33 Clay


Replicate plot 3 (40-100) 51 26 23 Clay


cm)

63 9 28 Clay

Due to high percentage of clay in the soil, the bulk density, lower limit, drained

upper limit and saturation were recalculated using a conversion factor from Default

Medium Silty Clay to adjust the water level in the soil for model calibration process.

Table 3.4 shows the recalculated water content of bulk density, lower limit of

extractable soil water, drained upper limit and saturation for Banisogosogo soil.

Bulk Density Drained Lower Limit Drained Upper Limit Saturation

Soil

depth

( cm)

Rep 1 Rep 2 Rep 3 Rep 1 Rep 2 Rep 3 Rep 1 Rep 2 Rep 3 Rep 1 Rep 2 Rep 3

0-40 1.24 1.25 1.25 0.26 0.256 0.267 0.40 0.407 0.408 0.51 0.521 0.519

40-

100

1.25 1.23 1.23 0.29 0.294 0.271 0.43 0.435 0.417 0.52 0.516 0.497

100-

110

1.23 1.27 1.22 0.21 0.248 0.273 0.34 0.357 0.357 0.51 0.501 0.49

3.2.2 Weather information

The weather information of the growing season indicated that the highest solar

radiation, Tmax, Tmin and rainfall were received beginning of the year (January,

February and March). On the other hand, during the planting season, which was from

43

July- September, it was seen that the lowest value of solar radiation, Tmax, Tmin and

rainfall were recorded (Table 3.5, Figures 3.2 and 3.3).

Table 3.5 shows Banisogosogo monthly average weather data. Average monthly data

were calculated for solar radiation (MJ/m2/d), Tmax and Tmin (oC) and rainfall ( mm)

from January to September, 2012.

Month SRAD (MJ/m2/d) Tmax (oC) Tmin (oC) Rain ( mm)

January 19.4 30.6 23.6 990.4

February 18.7 30.2 23.0 460.5

March 19.4 31 23.8 761.3

April 14.9 30 23.0 575.6

May 15.0 28.9 22.3 41.3

June 13.1 28.0 21.6 164.9

July 14.2 27.8 20.6 18.6

August 16.2 28.7 21.6 75.3

September 17.7 28.9 22.0 215.1

Rain

Date01-Jan-12 29-Jan-12 28-Feb-12 30-Mar-12 30-Apr-12 01-Jun-12 01-Jul-12 29-Jul-12 28-Aug-12 29-Sep-12

500

450

400

350

300

250

200

150

100

50

0

Figure 3.8 shows the daily rainfall for Banisogosogo from January to September, 2012.

44

TMAXTMIN

Date01-Jan-12 28-Jan-12 28-Feb-12 30-Mar-12 30-Apr-12 31-May-12 01-Jul-12 29-Jul-12 28-Aug-12 29-Sep-12

3432302826242220181614121086420

Figure 3.9 shows the daily maximum and minimum temperature for Banisogosogo from

January to September, 2012.

3.2.3 Experimental file

The crop data, that is, LAI and fresh and dry aboveground and belowground biomass

were entered into the DSSAT SUBSTOR Potato model as an experimental file. The

experimental file included the AFile and the TFile. The AFile was the average final

harvest value (T4) from each treatment (TRNO), for tuber dry weight (UWAH),

tuber fresh weight (UYAH), leaf are index, (LAIX), by product removed during

harvest (BWAH), tuber initiation date (TDAT) and harvest date (HDAT). The date

of TDAT and HDAT was in Julian Date Format. The results showed that the tuber

fresh weight was the highest for treatment 2 and 3 at 19.97 t/ha while the tuber dry

weight is highest for treatment 3 at 2452 kg/ha (Table 3.6).

Table 3.6 shows the values for AFile (Average File).

TRNO UWAH

(kg/ha)

UYAH

(t/ha)

LAIX BWAH

(kg/ha)

TDAT HDAT

1 2196 17.55 0.5 1007 219 263

2 2080 19.97 0.5 989 219 263

3 2452 19.97 0.2 970 219 263

The TFile was the harvest values from each time series (T1, T2, T3 and T4). The

values for tuber fresh weight (UYAD), leaf area index (LAID), tuber dry weight

(UWAD), leaf dry weight (LWAD), stem dry weight (SWAD) and tops dry weight

(CWAD) were recorded in TFile. Table 3.7 indicates that T2 value showed an

45

increase in tuber fresh weight and tuber dry weight. The highest tuber dry weight was

obtained at T3. The leaf, stem and tops weight increased for T2 and then decreased

for T3 and T4.

Table 3.7 shows the values for Time Series file (TFile).

TRNO DATE UYAD

(t/ha)

LAID UWAD

(kg/ha)

LWAD

(kg/ha)

SWAD

(kg/ha)

CWAD

(kg/ha)

1 02219 0.49 0.14 19.4 813 363 1176

1 02239 16.4 2.28 1987 785 518 1303

1 02259 24.9 2.01 2588 354 595 949

1 02263 17.5 0.54 2196 486 521 1007

2 02219 1.09 0.012 95 555 300 855

2 02239 19.2 2.88 2109 1068 737 1805

2 02259 16.9 1.65 4270 340 483 823

2 02263 19.9 0.51 2081 455 534 989

3 02219 1.17 0.22 101 439 350 789

3 02239 16.9 2.41 2200 1132 879 2011

3 02259 11.3 0.64 2967 229 382 611

3 02263 19.4 0.23 2452 405 565 970

3.2.4 Calibration

The DSSAT SUBSTOR Potato Model was calibrated using the local experimental

field data, local soil and weather data of the growing season. When the simulation

was conducted for Desiree before calibration, it was seen that the simulated and the

measured values are not in agreement (Table 3.8 and Figure 3.10).

Figure 3.4 shows the simulated values (represented by line on graph) and observed

values (represented by points on graph) for Desiree variety for LAI (a), leaf weight

(b), stem weight (c), tuber fresh weight (d), tops weight (e) and tuber dry weight (f)

before calibration for the three replicates plot. The simulation indicated that the

variety Desiree has not been calibrated before as the simulated and the observed

values for each variable were not aligned together.

46

For the calibration process, the water level in the soil, that is, bulk density, drained

lower and upper limit and the saturation level were recalculated (Table 3.4) and

genetic co-efficient was also recalibrated (Table 3.9).

Table 3.8 shows the simulation results for Desiree before calibration. Replicate plot 1 Replicate plot 2 Replicate plot 3

Variable Simulated Measured Simulated Measured Simulated Measured

Tuber Initiation Day (dap) 50 35 50 35 50 35

Physiological Maturity Day -99 79 -99 79 -99 79

Tuber Dry Weight (kg\ha)

harvest

1769 2196 1918 2080 1833 2452

Tuber Fresh Weight (t\ha)

harvest

8.55 17.55 9.59 19.97 9.17 19.97

Leaf Area Index (LAI) 4.16 0.5 5.50 0.5 4.52 0.2

The genetic co-efficient of Desiree variety was recalibrated to Desiree-tropics to

achieve a match between the simulated and observed values (Table 3.9). The

recalibration of genetic-co-efficient was done for G2, P2 and TC to suit the

temperature and daylength regime similar to the experimental conditions.

The DSSAT SUBSTOR Potato model was calibrated using data from 2012 season

experimental replicate plot in Banisogosogo, Fiji. The model was calibrated for tuber

dry weight. Since replicate plot 2 had a higher value of soil organic carbon and

nitrogen content (Table 3.2), it simulated the highest tuber dry yield (as shown by

Table 3.10) and the simulation of replicate plot 2 by the model also showed much

more accuracy (that is, the simulated and the observed values are much closer as

indicated by Figure 3.5) as compared to the other replicate plots. However, in the

final harvest, the observed value of T4 (2080 kg/ha), was not in good agreement with

the simulated value (4478 kg/ha). This may be due to the fact that during T3 and T4

harvest, the replicate plot was heavily infected with pest such as nematodes, beetle

(Papuana spp.) and beetle larvae and snail (Quantula striata).

47

Figure 3.10 shows the simulated and observed for Desiree before calibration.

a) LAI b) Leaf weight

c) Stem weight d) Tuber fresh weight

e) Tops weight f) Tuber dry weight

48

Table 3.9 shows the recalibration of genetics co-efficient.

Variety

Name

ECO

Number

G2-Leaf

Expansion

Rate (

cm2/m2/d)

G3-

Tuber

Growth

Rate

(g/m2/d)

G4 PD-

Determinancy

P2- Photoperiod

Sensitivity

(dimensionless)

TC-Critical

Temperature

(oC)

IB0008

Desiree

IB0001 2000 25.0 0.2 0.9 0.6 17.0

FJ0008

Desiree

tropics

IB0001 4000 25.0 0.2 0.9 0.4 18.0

Table 3.10 shows the simulation results after calibration for Desiree. Replicate plot 1 Replicate plot 2 Replicate plot 3

Variable Simulated Observed Simulated Observed Simulated Observed


Physiological Maturity Day -99 79 -99 75 -99 79

Tuber Dry Weight (kg\ha) harvest 2947 2196 4478 2080 3375 2452

Tuber Fresh Weight (t\ha) harvest 14.73 17.55 22.39 19.97 16.88 19.97

Leaf Area Index 0.74 0.5 1.7 0.5 0.89 0.2

Figure 3.11 shows the simulated and observed values for Desiree variety for LAI (a),

leaf weight (b), stem weight (c), tuber fresh weight (d), tops weight (e) and tuber dry

weight (f) after calibration. The simulation indicates the model for Desiree has been

calibrated for tuber dry weight (f). This is because after calibration, the simulated and

the observed values for the variable are aligned together, that is, the simulated and

the observed values appear in good agreement except for T4. During T4 it was

noticed that the replicate plots were heavily infected with pests. The observed values

for this simulation are produced from the TFile.

49

Figure 3.11 shows the potential and observed calibration results for Desiree.


d) Stem weight c) Tuber fresh weight

e) Tops weight f) Tuber dry weight

50

Linear regression function was also calculated for tuber dry weight. The straight line

represents the linear regression function which relates the observed and simulated

tuber yield. The linear regression analysis helps to evaluate the model performance

by providing information on the slope (biasness) and the coefficient of determination

(R2), which provides information on how well the model (simulated values) agrees

with the observed values.

Figure 3.12 shows the evaluation of potato yield (dry matter in kg/ha) at Banisogosogo Fiji

for year 2012. The R2 for replicate plot 1 was 0.88 (y=1.62x-143.02), the R2 for replicate

plot 2 was 0.66 (y=0.99x+674.02) and the R2 for replicate plot 3 was 0.92 (y=1.11x+21.04).

The calibration of the DSSAT SUBSTOR Potato model gave a very close agreement

between the simulated and observed yields. The highest R2 value was for replicate

plot 3 (R2= 0.92) followed by replicate plot 1 (R2= 0.88) and replicate plot 2 (R2=

0.66) indicating a positive relationship between the simulated yield and the observed

yield. Replicate plot 2 had the lowest values for R2. As indicated by Table 3.7, the

value for T4 (2081 kg/ha) was much smaller than T3 value (4270 kg/ha). This may

be due to the impact of pests during T3 and T4. This gives rise to lower value of R2

for replicate plot 2.

51

3.3 Discussion

Trial and error field experiments approach for agrotechnology transfer is not only

time consuming but also costly. Hence, the use of system simulations and decision

support systems for agrotechnology transfer has been adopted and promoted (Singh

et al., 2002). The DSSAT SUBSTOR Potato model is site and cultivar specific.

Hence, before using the model in a new location, it has to be recalibrated (Singh et

al., 1998). The DSSAT SUBSTOR Potato model was calibrated by adapting soil

(water and nitrogen content), weather and management practises into model input

files. The tuber yield and the cultivar were used as evaluation parameters. This was

done to assess the accuracy and sensitivity of the model (Štastna and Dufkova,

2008).

Tables 3.2 and 3.3 show the chemical and physical properties of Banisogosogo soil.

Replicate plot 1 had the lowest pH (6.5-7.0) followed by replicate plot 2 (6.7-7.1)

and replicate plot 3 (6.8-7.2), while the highest organic carbon and nitrogen were

evident in replicate plot 2. The texture of the soil was clay. The optimum pH for

potato cultivation was 5.5-6.0 and the crop preferred high organic matter soils (Singh

et al., 1998).

The minimum weather data input required for the model are solar radiation (MJ

m2/d), maximum and minimum temperature (oC) and precipitation ( mm) (Rezzoug

et al., 2008). From the monthly average temperature of Banisogosogo (Table 3.5)

and the Weatherman Output of rainfall and temperature (Figure 3.8 and 3.9

respectively), the trend suggests that the highest solar radiation (19.4 MJ/m2/d, 18.7

MJ/m2/d and 19.4 MJ/m2/d), maximum temperature (30.6 oC, 30.2 oC and 31 oC),

minimum temperature (23.6 oC, 23.0 oC and 23.8 oC) and rainfall (990.4 mm, 460.5

mm and 761.3 mm) were received in the beginning of the year (January, February

and March respectively) while the lowest temperature (27.8 oC and 20.6 oC for

maximum temperature and minimum temperature respectively) and rainfall (18.6

mm) were recorded in July. Since the lowest average monthly temperature was

recorded in July (27.8 oC day/20.6 oC night), this indicates that July was a good time

for potato cultivation. Cool night temperatures are required by potatoes for

tuberisation. Ideal overnight temperatures are between 15-18 oC (Levy and Veilleux,

52

2007). The optimum amount of rainfall needed for potato cultivation was 50-75 cm

(Singh et al., 1998).

The soil water module used by DSSAT SUBSTOR Potato model is based on

Ritchie’s water balance model. The plant available soil water is determined from

drained upper limit and lower limit of plant extractable soil water for the given soil

(Ritchie, 1981a, b) and the nitrogen balance is simulated using the CERES N model

(Godwin and Singh, 1998) where at each growth stage of the crop, soil water or

nitrogen stress will affect the growth of the crop and hence the final yield (Knox et

al., 2010). The soil water content was recalculated and a nitrogen response

experiment was created to determine the amount of ammonia and nitrate in the soil.

The first calibration step involved adjusting the soil water content, that is, the

Drained Upper Limit (DUL), Lower Limit (LL), Soil Moisture Upper Limit

Saturated (SSAT) and Bulk Density (SDBM). This was manually recalculated using

a conversion factor from Default Medium Silty Clay. The algorithm used in SBuild

was unable to differentiate tropical soils with high clay content and aggregation due

to iron oxides as in Banisogosogo form the soil with similar clay content without

aggregation. Hence, Default Medium Silty Clay with similar water holding capacity

to the Banisogosogo soil was used to calculate a conversion factor for the new values

of bulk density, drained lower limit, drained upper limit and saturation (Table 3.4).

Likewise, AFile and TFile were also created under Experiment Data. AFile

represented the average values of the final harvest (T4) while the TFile represented

the average values of all the harvest, that is, T1, T2, T3 and T4. Table 3.6 of AFile

shows that the highest tuber dry weight (UWAD) was 2452 kg/ha which was

obtained from treatment 3 (replicate plot 3) and tuber fresh weight (UYAH) was

19.97 t/ha which was obtained from treatment 2 (replicate plot 2) and 3 (replicate

plot 3). Table 3.6 also indicates that treatment 3 gave the lowest leaf area index value

(LAIX) at harvest value of 0.2 and also by-product removed from harvest (BWAH)

at 970 kg/ha. On the other hand, TFile (Table 3.7) shows that the tuber fresh weight

starts to increase for each replicate plot from T2. T2 represents biomass partitioning

to competing organs or tubers. The partitioning to tubers is favoured by low

temperature, short photoperiod and low to moderate soil moisture and nitrogen

levels. In the DSSAT SUBSTOR Potato model the tubers are given the first priority

53

to accessible assimilate from photosynthesis and reserved carbohydrate pool. This

eliminates the need to directly estimate partitioning co-efficient to allocate carbon. In

non-limiting conditions, growth demands are met while in limiting conditions the

growth for leaves and vines are reduced and tubers are given the priority (Griffin et

al., 1993). It was also seen that LAI was the highest for T2 for each replicate plot

followed by a decline in T3 and T4. T2 is a stage where 90-100 % of the daily

acquired assimilates are partitioned to tubers. The length of T2 depends on the

relative tuber growth rate, which determines the rate of partitioning of dry matter

between the tubers and the rest of the plant (Haverkort, 2007a). The leaf weight

(except for leaf weight in replicate plot 1) and the stem weight were the highest for

T2 followed by a decline in T3. The model also assumes that the stem and leaves are

main sink for assimilates. However, as plants matures, majority of the assimilates are

transferred to the organ sinks (Singh et al., 1998). The tuber dry weight also

increased for T2 and T3 and declined for T4.

When simulation was conducted for the Desiree variety, it was seen that the

simulated values were not in agreement with the measured values (as shown by

Table 3.8 and Figure 3.10), for example, the simulated tuber initiation day was 50

days after planting while the measured tuber initiation day was 35 days after

planting. The genetic co-efficient of Desiree was recalibrated to Desiree-tropics to

achieve a match between the simulated and observed values of tuber initiation day,

tuber dry yield, tuber fresh yield and leaf area index. Under the genetic co-efficient

parameters, photoperiod sensitivity to tuber initiation is represented by P2 and the

critical temperature above which tuber initiation is inhibited is represented by TC

(oC). G2 (cm 2 m -2 d-1) represents leaf area expansion rate in degree days. G2 is

currently equivalent to 2000 cm 2 m -2 d-1 for all cultivars due to lack of evidence to

contrary. G3 (g m-2 d-1) is potential tuber growth rate. The level of determinacy of the

cultivar is defined by PD (Griffin et al., 1993). There was also notable difference

between the genetic co-efficient of Desiree and Desiree-tropics, that is, between the

original and the calibrated values, for example, changes were made to the leaf

expansion rate (G2) from 2000 cm2 m-2 d-1 to 4000 cm2 m-2 d-1, the value of

photoperiod was changed from 0.6 to 0.4 and the value for critical temperature (TC)

was changed from 17.0-18.0 oC (as shown by Table 3.9). After the calibration of the

genetic co-efficient, the simulated and the measured valued appeared much closer (as

54

shown in Table 3.10). The genetic co-efficient was derived iteratively through the

manipulation of the coefficient to achieve a match between the simulated and

observed values (Fraisse et al., 2001; Rezzoug et al., 2008). The cultivar parameters

were changed individually and the results were compared with the original and

previous run until the cultivar parameter modification gave the best output where the

simulated values matched the observed values. Formal sensitivity analysis was not

conducted in our evaluation but it was done directly in the Genetic Co-efficient file.

The reason for this is that when a formal Sensitivity Analysis was conducted, G2

values were not allowed to go beyond 3000 (maximum value), which shows that the

potato model, particularly the Desiree cultivar, has not been calibrated in the Tropic

regions. Hence, this calibration was done to suit the Tropical region and this brought

about marked improvement in the performance of the model. Most of the species

parameters are developed by modelers from experience of model calibration (Asadi

and Clemente, 2003) while some are developed from optimisation data sets (Gri mm

et al., 1993).

The calibration changes were associated with genetic co-efficient (specific to the

cultivar used) and soil parameters associated with the soil’s upper and lower limits.

The SUBSTOR Potato model subroutines and codes were not changed during the

calibration process. The adjustments to the genetic co-efficient are justified given the

Desiree variety has not been grown under temperature and daylength regime similar

to the experimental conditions. The lowered P2 value indicates greater photoperiod

sensitivity and the increased TC reflects higher temperature tolerance of Desiree

variety. These changes should not affect the performance of Desiree variety under

previously tested conditions. The higher G2 value in Table 3.9 used in the current

study requires verification and multi-location and trials. The simulated results show

that the differences are between the “calibrated” and the “original” model is

associated with changes in genetic co-efficient. The days for the tuber emergence and

harvest are identical; however, the changes are associated with the duration of the

tuber growth- 15 days longer for calibrated model. This combined with the LAI

effect led to higher tuber yield.

It is important for a crop model to accurately predict observed variation in yield,

before modelling climate impact on future yield. The DSSAT SUBSTOR Potato

55

model has been used internationally for crop studies and for climate change impact

studies (Knox et al., 2010). The DSSAT SUBSTOR Potato model has been tested in

various environments and has performed well when simulated data has been

compared with observed data (Bowen, 2003). However, such tests have been

conducted using only a few cultivars under temperate and subtropical/high altitude

environments. The DSSAT SUBSTOR Potato model was calibrated using data from

2012 season experimental replicate plot in Banisogosogo, Fiji. This allowed

comparison of model outputs to observed experimental results and field

measurements of the real system (Huang et al., 2009). Since replicate plot 2 had a

higher value of soil organic carbon content (Table 3.2), it simulated the highest tuber

dry yield (as shown by Table 3.10) and the simulation of replicate plot 2 by the

model also showed much more accuracy (that is, the simulated and the observed

values are much closer as indicated by Figure 3.11) as compared to the other

replicate plots. However, in the final harvest, the measured value of T4 (2080 kg/ha),

was not in good agreement with the simulated value (4478 kg/ha). This may be due

to the fact that during T3 and T4 harvest, the experimental replicate plot was heavily

infected with pest such as nematodes, beetle (Papuana spp.) and beetle larvae and

snail (Quantula striata) (Figures 3.5, 3.6 and 3.7). Due to the lack of disease or lack

of insect defoliation subroutines within the model (Griffin et al., 1993), the simulated

and measured values for T4 were not in agreement. The DSSAT SUBSTOR Potato

model was calibrated for tuber dry yield. The calibration results indicated that the

model is in good agreement with tuber dry weight. The DSSAT model has been

validated in Egypt and was successful in simulating current yield under different

irrigation levels and future yield under A1, A2, B1 and B2 for 2025, 2050, 2075 and

2100 (Medany, 2006). The model has also been validated for tuber initiation which

shows a strong linear relationship was apparent between simulated and observed data

for tuber initiation (Griffin et al., 1993). The model shows great potential for

simulating growth from experimental data (Singh et al., 1998). Since the model has

only been calibrated for tuber dry weight, the leaf area index was overestimated for

T1 and underestimated for T2, T3 and T4 (as shown by Figure 3.11). Leaf area index

is considered a key variable when calibrating models as it accounts for the leaf area

that intercepts the incoming radiation and it also takes part in major physiological

processes such as evapotranspiration, photosynthesis and biogenic emissions. High

LAI is important for potatoes for higher rates of tuber bulking for extended periods

56

which gives rise to higher yields (Geremew et al., 2008). In many crop monitoring

studies, in situ LAI have been used to calibrate crop models but the in situ LAI

measurements seemed inadequate as they are usually available for a limited number

of fields and dates (Gonzalez-Sanpedro et al., 2009). The DSSAT SUBSTOR Potato

model does not take into consideration the simulation of individual leaves but rather

the development of the entire canopy. The simulation of leaf growth is poorer than

the simulation of tuber initiation with a correlation co-efficient of 0.47. This model

tends to over-estimate the leaf area (Griffin et al., 1993; Fleisher et al., 2000). The

leaf weight, stem weight and the tuber fresh weight showed underestimation for T1

and T2 and overestimation for T3, and T4. Replicate plot 2 for top weight showed

different response when compared to all the other replicate plots. Replicate plot 2

was underestimated for T1 and T2 whereas overestimated for T3 and T4. Unlike

CERES-Maize model which has been extensively used in the Tropics (Singh et al.,

1990), the DSSAT SUBSTOR Potato model has been extensively tested in the

temperate regions. It is yet to be tested in the coastal regions, South-East Asia and

the Tropics (Singh et al., 1998).

Figure 3.12 shows the evaluation of potato yield (dry matter in kg/ha) at

Banisogosogo Fiji for the year 2012. The straight line represents the linear regression

function which relates the observed and simulated tuber yield and the coefficient of

determination (R2), which, assesses how well the shape of the simulation matches the

shape of the observed data (Knox et al., 2010). The highest R2 value is for replicate

plot 3 (R2= 0.92) followed by replicate plot 1 (R2= 0.88) and replicate plot 2 (R2=

0.66) indicating a positive relationship between the simulated yield and the observed

yield. It can also be said that each replicate plot is different. Replicate plot 2 has the

lowest values for R2. This might be due to factors not considered in the model, such

as, that of pest and disease (Griffin et al., 1993). As indicated by Table 3.7, the dry

matter (kg/ha) for replicate plot 2, the value for T4 (2081 kg/ha) is much smaller than

T3 value (4270 kg/ha) due to the impact of pests. This gives rise to lower value of R2

for replicate plot 2. Another difference highlighted by (Štastna and Dufkova, 2008) is

due to physiological difference projected into genetic parameters of the cultivar,

change in crop management and nutrient supply. In an experiment by (Štastna and

Dufkova, 2008), the validation for Rosara cultivar showed accuracy between

simulated and observed tuber yield (R2=0.97) while Karin showed lower value (R2=

57

0.43). The authors also stated that the use of model under extreme conditions, such

as, dry years showed that the model tend to underestimate tuber yield (Štastna and

Dufkova, 2008). This accurate simulation of tuber yield is critical as tuber yield are

not only economically important but is also affected by every process simulated in

the model, such as, the rate of development, intercepted radiation and use efficiency,

biomass partitioning, fertility and soil water status.

3.4 Recommendation

Several issues were identified for the usefulness and the improvement of the DSSAT

SUBSTOR Potato model. To begin with, further work is needed to test the DSSAT

SUBSTOR Potato model performance in the Tropics with vigorous multi-location

trials using suitable “tropical” varieties. The simulation of potato growth across the

diverse environment and different cultivars must also take into consideration the

effect of temperature, photoperiod and intercepted radiation. The application of the

model to a specific geographical location reduces the effectiveness of the model by

limiting the environmental inputs (Griffin et al., 1993).

Although the DSSAT SUBSTOR Potato model is now calibrated for tuber dry

weight, it is recommended that studies be conducted to further validate tuber weight,

LAI, leaf weight, stem weight and tuber weight under stress-free/minimal stress

conditions.

3.5 Experiment limitation

The following are the limitations of the experiment. Firstly, the soil water content

had to be manually recalculated using a conversion factor. The SBuild program

should be modified based on the soil data from the tropics. This also highlights the

need for actual measurement of the lower limit of plant extractable soil water and

drained upper limit under the field conditions for tropical soils.

The DSSAT SUBSTOR Potato model does not take into consideration the impact of

pest and disease on the tuber yield (Griffin et al., 1993). As indicated by the results,

T3 andT4 were badly affected by pests. More attention to pest and disease control is

58

needed for experimental trials, particularly those used for model validation since

most models do not simulate pest and disease effect.

Also, the DSSAT SUBSTOR Potato model does not read the year 2012 for the

measured data in PlantGro. The graph had to be exported to Excel and the Date had

to be manually changed from 2009 to 2012. Due to this, the Statics table did not give

any statistical value. It is recommended that the DSSAT SUBSTOR model should be

upgraded so that it is much easier to get the Statics whereby saving time.

The weather data imported into WeatherMan was also a combination of FMS and

NASA data. It was observed that the NASA values were slightly higher than the

FMS values. Investment in automated weather stations that determine daily solar

radiation, temperature, rainfall, humidity and wind would verify the quality of NASA

data.

Furthermore, the LAI was calculated during T1 using leaf area meter. However, from

T2-T4 the LAI was calculated using AccuPAR LP-80. Due to this, the model was not

calibrated for LAI.

59

Chapter 4 Simulation of Desiree potato variety growth and yield in three

different sites (Banisogosogo, Koronivia and Nacocolevu) in Fiji using the

calibrated DSSAT SUBSTOR Potato model

4.0 Introduction

Agricultural decision makers need increasing amount of information to meet

increasing demands for agricultural products and increased pressure on natural

resources for instance land and water (Jones et al., 2003). They also need to better

understand the possible outcomes of their decisions to assist them develop plans and

policies that meet their goals (Jones et al., 1998).

Many crop models have been used to study the growth and development of potatoes.

AquaCrop model has been used to study constraints limiting crop production and

water productivity in India. The model demonstrated good agreement between the

simulated and the observed values of belowground mass. However, the model proved

to be less satisfactory for aboveground mass and when simulating yield under high

water stress (Patel et al., 2008). An agrometeorological model was used to estimate

potential tuber yield in State of São Paulo, Brazil. The simulated and observed tuber

dry matter showed high correlation. However, the model underestimated the yield of

irrigated potato by less than 10 % (Pereira et al., 2008).

DSSAT has been developed to evaluate growth, production, resource use and the

risk linked with crop management practices (Jones et al., 1998). Crop models can

also be used to study variation in cultivar response to the environment (White, 1998).

The CERES crop model in DSSAT has been tested over a wide range of environment

and has also been used to simulate cereal crop growth and development with the

yields being within an acceptable limit of ±5 % to 15 % of measurable yields

(Ritchie et al., 1998).

The DSSAT SUBSTOR Potato model had been developed as a CERES-type model

and uses soil water and nitrogen dynamics that are used in other CERES type

models. The DSSAT SUBSTOR Potato model is successful in predicting potato

production (Abdrabbo et al., 2010) and has been tested in various environments with

good performance through comparison of simulated data with observed data (Bowen,

2003). The DSSAT SUBSTOR Potato model has a great potential to simulate potato

growth and evaluating possible changes in management in many regions (Griffin et

60

al., 1993). This model has also been validated in many areas and used to simulate

physiological processes and yield of potatoes under current climate and future

conditions (Bowen et al., 1998; International Potato Center, 1999; Fleisher et al.,

2003; Štastna and Dufkova, 2008; Abdrabbo et al., 2010; Knox et al., 2010). It has

also been extensively used internationally for crop studies and recently for climate

change impacts (Knox et al., 2010). The DSSAT SUBSTOR Potato model can be

used to simulate on a day-to-day basis the growth and development of the potato

crop using information on climate, soil, management and cultivar (Jones et al., 2003;

Knox et al., 2010). The SUBSTOR Potato model also has the ability to simulate

growth under potential conditions and non-potential (water limiting and water and

nitrogen limiting) conditions (Singh et al., 1998). It assumes that tuber initiation (T1)

is a function of daylength, temperature, water and nitrogen status in soil. Tuber

initiation by early cultivars are less sensitive to high temperatures and long daylength

as compared to late cultivars (Singh et al., 1998). The deficiency or excess of

nitrogen is based on above-ground tissue nitrogen concentrations. There is delay in

development under excess nitrogen conditions (Singh et al., 1998). The impact of

weather variability on a given management option can also be analysed by DSSAT

SUBSTOR Potato model by running the model with many years of weather data

(Bowen, 2003).

Application of crop models for simulating crop growth and yield play a crucial role

in addressing issues of food supply (yield forecasting) and also the risk associated

with weather variability (Thornton and Wilkens, 1998). Simulation models can also

be used as an educational tool for the understanding of biological processes (Ortiz,

1998). The model can also be used to enhance the use of resources for improved crop

production and profitability (Ritchie et al., 1998). This type of agronomic research

increases the productivity of farmers so that they can compete with the global market

and also helps protect the environment through the sustainable use of natural

resources (International Potato Center, 2006). There is a need to simulate potato

production in Fiji using current and historical weather data. This will lead to an

understanding of how current weather data and extreme events affect potato

production around Fiji. This Chapter presents the findings of simulating potato

growth in three sites around Fiji using the calibrated DSSAT SUBSTOR Potato

model. The use of DSSAT on the effect of climate variability on crop production

61

system is now considered part of decision support system in resource management

(United Nations, 2002). The impact of ENSO can also be studied in DSSAT through

multiple runs. These simulations can facilitate the task of optimising crop growth and

yield and suggesting recommendations on these crop management strategies.

Through these, policies and technologies can be identified that can take advantage of

positive opportunities or help in the reduction of adverse impacts of extreme weather

impacts. Also, improvements in forecast skills of ENSO can improve disaster

preparedness and improve production decisions of farmers (Valdivia et al., 2000).

4.1 Methodology

4.1.1 Simulation sites

The simulation sites for current climate included Banisogosogo, Koronivia and

Nacocolevu. Information on geographical location and reason for site selection for

Banisogosogo was provided in Chapter 3. Nacocolevu (also known as Sigatoka

Valley), is located between (18˚05'52.36" S, 177˚31'49.46" E). Sigatoka Valley is

located in Fiji’s western drier zone (Borrower, 2005). The western division is the dry

zone and is subjected to large seasonal and inter-annual climatic variation. Sigatoka

receives annual average to about 2000 mm. During the winter weather in these areas,

it is clear or partly cloudy with abundance of sunshine and a strong west to north-

westerly winds by day and south easterlies are prominent wind at night and in all

months except for August to March, the wind is mostly from north westerly quarter.

The nights are quite cool in this season. However, during summer there is

thunderstorm activity which is responsible for brief but intense rainfall (Gawander et

al., 2012). The average maximum temperature of Nacocolevu is 29.7oC while the

average minimum temperature is 21oC (Fiji Meteorological Services, 2013c)

The third site is Koronivia, in Nausori, which is located between (18˚03'00" S,

178˚32'00" E) (National Geospatial-Intelligence Agency, 2012a). Nausori is located

on the windward side of Viti Levu (Nausori Town Council, 2011). The climate

conditions of Nausori are often warm and humid. Nausori receives annual rainfall of

2000 mm annually while the higher range of Naitasiri receives an average rainfall of

6000 mm annually. During heavy rainfall, low lying areas of Rewa Delta are heavily

affected. The average temperature range of Nausori is from 28-30 oC. The cyclones

are normally confined to the wet season (Nausori Town Council, 2011).

62

4.1.2 Reason for site selection

Banisogosogo and Nacocolevu were selected for simulation purposes because these

are potato cultivating areas. Sigatoka Valley is considered the major center for potato

production at 134 hectares (Iqbal, 1982). The area also has soil which is suitable for

potato production (Autar, 2009; Macfarlane, 2009). Hence, simulation was

conducted to evaluate the impact of climate variability and to investigate how

growth, development and yield are affected by weather and soil of these areas.

Furthermore, Koronivia was selected as a simulation site to see if this site can be

considered as a potato cultivation area. It is also interesting because Koronivia is

located on the southern part of Viti Levu and also where Koronivia Agriculture

Research Station is located.

Figure 4.0 shows the location of three simulation sites. Source: (GoogleEarth, 2012).

63



The 30 year climatic data for Penang (Banisogosogo), Koronivia and Nacocolevu,

such as, daily maximum temperature (oC), minimum temperature (oC) and

precipitation (mm) were collected from the Sugar Research Institute of Fiji (SRIF)

and Fiji Meteorological Services (FMS). Solar radiation of the experimental site was

also downloaded from NASA website. The weather data had missing values which

were replaced with NASA data creating a “hybrid” weather data. The DSSAT format

weather data was created in 2007 Excel worksheet. This Excel worksheet was then

imported into WeatherMan tool in DSSAT.

4.1.3.2 Soil data

As discussed in Chapter 3, soil samples of Banisogosogo were given to Koronivia

Research Station for analysis. The soil data of Koronivia and Nacocolevu were

recorded from secondary sources, that is, written reports which were provided by

Koronivia Research Station. Information on soil physical and chemical properties,

such as, colour, pH, nitrogen and organic carbon content, cation exchange capacity

(CEC) and base saturation for each layer was noted. These soil data were entered into

SBuild tool of DSSAT SUBSTOR Potato model.

4.1.3 Model simulation under current climatic and potential conditions

The calibrated DSSAT SUBSTOR Potato model (from Chapter 3) was used to create

potential and non-potential simulations for Banisogosogo, Koronivia and

Nacocolevu. The simulations were created using site specific current weather

conditions and soil physical and chemical information. Banisogosogo simulation was

conducted with available weather and soil data, that is, weather data of 1960-2012,

with Banisogosogo Replica 2 soil using the calibrated DSSAT SUBSTOR Potato

model. Koronivia potential and non-potential simulations were run using weather

data from 1961-2010 with Koronivia Silt Loam Soil. Nacocolevu potential and non-

potential simulations were conducted using weather data from July 1972-2010 with

Nacocolevu Rolling Phase Soil.

The crop management practices used was the same as in Chapter 3. The date of

irrigation application, the amount of irrigation application and the method of

64

application were similar in all sites. There were four irrigation applications using

Drip or Trickle method. Fertiliser amount of 80 N kg/ha was applied in 2 splits.

Similar fertiliser application techniques were used for all the sites. This information

is summarised in Table 4.0, Table 4.1 and Table 4.2. For potential simulations, the

model assumed that there was no water or nitrogen stress and that the plant was

grown under perfect conditions while non-potential simulation assumed that there

was water and nitrogen stress.

Table 4.0 shows the summary of simulations at Banisogosogo, Koronivia and

Nacocolevu.

Site Banisogosogo Koronivia Nacocolevu

Soil Name Banisogosogo Replica 2

Soil

Koronivia Silt Loam Nacocolevu Soil,

Rolling Phase

Weather Station Penang Nausori Nacocolevu

Cultivar Desiree Desiree Desiree

Planting Date 2nd July 2nd July 2nd July

Planting Method Dry seed Dry seed Dry seed

Planting Distribution Rows Rows Rows

Plant Population

(seedling) (m2)

5 5 5

Row Spacing ( cm) 75 75 75

Planting Depth ( cm) 1.5 1.5 1.5

CO2 Concentration Actual CO2, Mauna

Loa, Hawaii

Actual CO2, Mauna

Loa, Hawaii

Actual CO2, Mauna

Loa, Hawaii

65

Table 4.1 shows the irrigation application during simulations at Banisogosogo,

Koronivia and Nacocolevu.

Site Date (dd/ mm/yyyy) Irrigation ( mm) Operation

Banisogosogo,

Koronivia, Nacocolevu

02/07/2012

13/07/2012

05/08/2012

10/09/2012

1.6

4

4

4

Drip or Trickle

Drip or Trickle

Drip or Trickle

Drip or Trickle

Table 4.2 shows the fertiliser application during simulations at Banisogosogo,

Koronivia and Nacocolevu.

Site Date (dd/

mm/yyyy)

Fetriliser

Material

Fertiliser Application Depth

( cm)

N

Amount

(kg/ha)

Banisogosogo,

Koronivia,

Nacocolevu

02/07/2012

23/07/2012

Diammonium

phosphate

Diammonium

phosphate

Banded beneath

surface

Banded beneath

surface

5

5

40

40

4.1.4 Optimisation treatment through sensitivity analysis

The optimisation treatments were simulated for planting time, row spacing, irrigation

treatment, fertiliser treatment and planting depth to optimise potato yield at each

simulation site using the DOS method in Sensitivity Analysis in DSSAT SUBSTOR

Potato model. The weather data of 2012 was used for all optimisation treatments

(except planting time) for Banisogosogo non-potential simulation with 2nd July 2012

as default run. The optimisation treatments (except planting time) for Koronivia and

Nacocolevu non-potential simulations were run with 2010 weather data with 2nd July

2010 as default run. The planting times for all three sites were simulated with 2009

weather data with 2nd July 2009 as default run. When using 2010 weather data for

optimisation of planting date, it was noted that there were no yield for October,

66

November and December as 2011 weather data was not available. It was assumed

that the crop will be harvested in 90 days. Hence, this optimisation treatment was

conducted with 2009 weather data.

In this simulation, all other crop management details remained the same except for

the optimisation treatment. The optimisation for planting time was simulated on the

first day of the month (except for January which was carried out on the 2nd of

January due to the start of simulation date which needs to be before the planting

date). It was assumed that the harvest day was 90 days after planting. The

optimisations for row spacing were simulated at 30 cm, 40 cm, 50 cm, 80 cm and

100 cm with 75 cm as a default run. For irrigation and fertiliser optimisation, water

and nitrogen stress days were recorded from initial non-potential simulation. The

irrigation optimisations were simulated on a fortnightly basis and three days before

water stress from non-potential simulation. The irrigation treatments were at 1.6 mm,

4.0 mm, 6.4 mm, 8.0 mm, 9.6 mm, 12.0 mm, 14.0 mm, and 16.0 mm with default run

as 1.4 mm and 6.0 mm (refer to Table 4.1). The fertiliser optimisation was carried

out from 60 N kg/ha to 300 N kg/ha at 60 N kg/ha interval (this was applied in 3

splits, one on the day of planting and the other 2 splits were simulated 3 days before

the nitrogen stress from non-potential simulation) with default run as 80 N kg/ha

applied in 2 splits at 5 cm (Table 4.2). The optimisation for planting depth was also

conducted from 2 cm to 10 cm at interval of 2 cm with default run of 1.5 cm. Finally,

the optimum application of fertiliser and irrigation from the fertiliser and irrigation

simulations were noted and these treatments were also simulated together to see if

the yield was optimised.

4.1.5 Climate variability (El Niño Southern Oscillation)

The impact of climate variability, El Niño Southern Oscillation (ENSO) state, on

potato yield in Fiji was also studied. To conduct the simulation for ENSO years, for

the Simulation Options, under General tab, the number of years were changed to

make the model run for that number of years. Banisogosogo simulation was

conducted from 1983-2012 while Koronivia and Nacocolevu simulation was

conducted from 1990-2010. The Fiji Meteorological Services helped to identify the

ENSO state from 1983-2012. The percentage difference was also calculated with

neutral year as a baseline.

67

4.1.6 Data analysis

The above inputs (soil data, weather data and crop management data) were used to

run simulation for the three locations. All summary outputs, graphical and initial

analysis were done in DSSAT v4.5 Output. Other statistical analysis was conducted

in Microsoft 2007 Excel Spreadsheet. These outputs are presented in the Results

section.

4.2 Results

This section describes the simulation of Desiree potato variety growth and yield

under current climate conditions in three different sites, that is, Banisogosogo,

Koronivia and Nacocolevu using the calibrated DSSAT SUBSTOR Potato Model.

The simulations were conducted with the site specific soil data, site specific weather

data and crop management practices. Under potential simulation, Koronivia gives the

highest tuber yield while Nacocolevu gave the highest tuber yield under non-

potential simulation. The impact of climate variability, ENSO state, was also studied

on potato yield for the three locations. Differences in the yield were noticed for

ENSO simulation for the three locations due to differences in weather data and the

water and nitrogen stress faced under El Niño, La Niña and neutral years. In addition

to these, optimisation treatments also investigated which crop management practices,

such as planting date, row spacing, irrigation treatment, fertiliser treatment, optimum

irrigation and optimum fertiliser treatment and planting depth can optimise potato

yield.

4.2.1 Weather data

Site specific “hybrid” weather data for Banisogosogo, Koronivia and Nacocolevu

was imported into WeatherMan tool in DSSAT. The monthly averages for the three

sites indicated that the potato growing season (July-September) received lower

values of average monthly maximum temperature, minimum temperature and rainfall

as compared to months earlier than the potato growing season and later than the

potato growing season (Figures 4.1 and 4.2 and Table 4.3).

68

Figure 4.1 shows the monthly average rainfall for Banisogosogo, Koronivia and Nacocolevu.

Figure 4.2 shows the monthly average maximum and minimum temperature for

Banisogosogo, Koronivia and Nacocolevu.

69

Table 4.3 shows the average monthly maximum temperature, minimum temperature

and rainfall for Banisogosogo, Koronivia and Nacocolevu.

Banisogosogo Koronivia Nacocolevu

Tmax

(oC)

Tmin

(oC)

Rainfall

( mm)

Tmax (oC) Tmin (oC) Rainfall

( mm)

Tmax

(oC)

Tmin

(oC)

Rainfall

( mm)

January 30.4 23.9 409.6 30.3 23.2 354.8 31.3 22.3 174.5

February 30.5 23.8 361.7 30.7 23.4 272.5 31.5 22.7 156.8

March 30.5 23.7 411.7 30.4 23.2 369.8 31.3 22.7 157.4

April 29.7 23.3 302.1 29.4 22.6 343.5 30.5 21.7 160.5

May 28.7 22.3 146.9 28 21.3 228.3 29.3 20 153.9

June 28.1 21.6 99.9 27.3 20.7 163.9 28.5 18.9 152.8

July 27.6 20.7 52.1 26.4 20.0 136.1 27.5 17.9 162.6

August 27.8 20.9 74.3 26.4 19.9 147.7 27.6 17.9 184.5

September 28.4 21.5 101.3 26.8 20.4 174.9 28 18.5 170.7

October 29 22.3 104.4 27.7 21.1 214.5 29.2 19.6 200.1

November 29.7 23 173.6 28.7 21.9 249.8 30.4 20.9 185

December 30.2 23.5 272.9 29.6 22.6 277.2 31.1 21.7 181

Annual Average 29.2 22.54 209.21 28.47 21.69 244.41 29.68 20.4 169.98

Season Average 27.93 21.03 75.9 26.53 20.1 152.9 27.7 18.1 147.48

4.2.2 Simulation results for Desiree

The current climate simulations were conducted for Banisogosogo, Koronivia and

Nacocolevu using Desiree variety. The simulations were run using site-specific soil

data, site-specific weather data and crop management practices. The results indicated

that the potential yield is higher than the non-potential yield. The potential tuber dry

weight for Banisogosogo was 6651 kg/ha while the non-potential tuber dry weight

was 4478 kg/ha. Koronivia simulation indicated that the potential tuber dry weight

was 9811 kg/ha but the non-potential tuber dry was 4373 kg/ha. Under Nacocolevu

70

potential simulation, the tuber dry weight was 8347 kg/ha whereas the non-potential

tuber dry weight was 5405 kg/ha (Table 4.4).

Table 4.4 shows the potential (non-limiting for water and nitrogen) and non-potential

(limiting for water and nitrogen) simulation results of current climate for

Banisogosogo, Koronivia and Nacocolevu.

Banisogosogo Koronivia Nacocolevu Variable Potential

Simulation Non-Potential Simulation

Potential Simulation

Non-Potential Simulation

Potential Simulation

Non-Potential Simulation

Tuber Initiation Day (dap)

38 35 32 30 35 33

Tuber Dry Weight (kg\ha) harvest

6651 4478 9811 4373 8347 5405

Tuber Fresh Weight (t\ha) harvest

33.25 22.39 49.05 21.86 41.74 27.02

Leaf Area Index, maximum

8.29 1.7 9.22 2.72 2.43 1.03

Figures 4.3, 4.4 and 4.5 illustrate the potential and non-potential simulation for

Banisogosogo, Koronivia and Nacocolevu respectively. The figures show that the

under potential simulation, the LAI, aboveground biomass and belowground biomass

increased gradually with time while under non-potential simulations, LAI,

aboveground biomass and belowground biomass showed fluctuation in weight. The

figures also indicate that the potential weight of each variable was higher than the

non-potential weight. This can be due to the fact that under non-potential conditions,

the crop faced water and nitrogen stress.

71

Figure 4.3 shows Banisogosogo potential and non-potential simulations for stem weight,

LAI, tuber fresh weight, leaf weight, tops weight and tuber dry weight.

72

Figure 4.4 shows Koronivia potential and non-potential simulations for LAI, leaf weight,

stem weight, tuber fresh weight, tuber dry weight and tops weight.

73

Figure 4.5 shows Nacocolevu potential and non-potential simulations for LAI, leaf weight,


74

Impact of precipitation, total water content of the soil, nitrate content of the soil and

nitrogen leached from the soil were also studied on non-potential tuber dry weight

and LAI for Banisogosogo (Figure 4.6), Koronivia (Figure 4.7) and Nacocolevu

(Figure 4.8).

Figure 4.6 shows the impact of precipitation, total water and nitrogen content in soil on non-

potential tuber dry yield and LAI for Desiree in Banisogosogo over the growing season.

The Banisogosogo simulation (Figure 4.6) indicated that upon application of nitrogen

fertiliser, the nitrate level in the soil increased. The figure shows that LAI (a) started

to increase a few days after planting (9th of July). The LAI reached its maximum

(1.7) around 20th of August and 26th of September. The LAI decreased around 20th of

a) LAI

b) Tuber dry weight

75

August. Around the same time the tuber dry weight (b) started to increase gradually

until the harvest day.


potential tuber dry yield and LAI for Desiree in Koronivia over the growing season.

For Koronivia simulation (Figure 4.7), it was seen that precipitation levels increased

gradually and around 30th August when the precipitation levels were high and the

nitrate levels decreased in the soil. The LAI (a) showed fluctuation over the growing

season with the maximum LAI (2.5 and 2.72) being achieved on 9th of August and 5th

of September respectively. The tuber dry weight (b) increased when the LAI dropped

the first time (10th of August) and it continued to increase gradually until the 30th of

a) LAI

b) Tuber dry weight

76

August, however, a sharp increase in tuber dry weight occurred after 30th of August.

The figure also shows the trend for total nitrate and water in the soil.


potential tuber dry yield and LAI for Desiree in Nacocolevu over the growing season.

Figure 4.8 illustrates that Nacocolevu water content was consistent throughout the

planting season. The precipitation increased gradually over time and the figure also

indicated that the total nitrate leached from the soil also increased gradually over

time. The LAI (a) increased steadily until the 17th of August where the maximum

a) LAI

b) Tuber dry weight

77

LAI was attained at 1. After the 7th of August the LAI decreased. On the other hand,

the tuber dry weight (b) increased steadily over time. The figure shows that the total

water content in the soil was consistent over time with some nitrogen leaching

experienced as precipitation increased.

4.2.3 Optimisation treatments

The optimisation simulations were conducted to investigate which crop management

option maximised tuber yield. The optimisation simulation was conducted for

planting time, row spacing, irrigation treatment, fertiliser treatment, optimum

fertiliser and optimum irrigation treatment and also planting depth. In general, the

results indicated that different crop management strategies maximised yield in

different locations.

Table 4.5 shows the optimum planting time for Desiree variety for Banisogosogo,

Koronivia and Nacocolevu. The optimum planting time for Banisogosogo was May

while the optimum planting time for Koronivia and Nacocolevu was August.

Table 4.5 shows the yield at different planting time for Banisogosogo, Koronivia and

Nacocolevu.

Yield (kg/ha)

Planting Time Banisogosogo Koronivia Nacocolevu

January 2 18 395 63

February 1 13 607 208

March 1 289 2223 1444

April 1 2109 4117 4006

May 1 5356 1959 1280

June 1 4567 4103 1135

July 2 (default run) 3698 6703 4823

August 1 4329 7956 7612

September 1 967 3824 5899

October 1 2523 7384 3375

November 1 316 2088 3197

December 1 179 1137 234

The table below (Table 4.6) shows the optimum row spacing for Desiree variety for

Banisogosogo, Koronivia and Nacocolevu. The optimisation for row spacing was

78

carried out at 30 cm, 40 cm, 50 cm, 80 cm and 100 cm with 75 cm as a default run.

Banisogosogo had optimum row spacing of 40 cm while Koronivia and Nacocolevu

had optimum row spacing of 30 cm.

Table 4.6 shows the yield at different row spacing.

Yield (kg/ha)

Row Spacing (

cm)

Plant Population

(m2)

Banisogosogo Koronivia Nacocolevu

75 (default run) 5 4478 4373 5405

30 11 4956 4555 7290

40 8 4988 4448 6759

50 6 4837 4496 6170

80 4 4142 4411 4519

100 3 3391 4441 3515

The optimum irrigation management was conducted for 1.6 mm, 4.0 mm, 6.4 mm,

8.0 mm, 9.6 mm and 12.0 mm (Table 4.7). The results indicated that the optimum

irrigation for Banisogosogo was 6.4 mm irrigation per plant while for Koronivia and

Nacocolevu the optimum irrigation amount was 1.6 mm irrigation per plant.

Table 4.7 shows the yield under different irrigation application and irrigation amount.

Yield (kg/ha)

Irrigation ( mm) Banisogosogo Koronivia Nacocolevu

Default Run (1.6 mm

plus 4.0 mm)

4478 4373 5405

1.6 4185 8411 5434

4.0 4601 7932 5434

6.4 6268 8028 5413

8.0 6161 8043 5338

9.6 5062 8052 5338

12.0 3922 8056 5337

14.4 3819 8253 5337

16.0 3850 8379 5335

The optimisation for planting date was simulated from 2 cm to 10 cm with 2 cm

interval with 1.5 cm as the default planting depth (Table 4.8). The results indicated

79

that the optimum planting depth for Banisogosogo and Koronivia was 1.5 cm while

the optimum planting depth for Nacocolevu was 2 cm.

Table 4.8 shows the planting depth with corresponding yield.

Yield (kg/ha)

Planting Depth ( cm) Banisogosogo Koronivia Nacocolevu

1.5 (default run) 4478 4373 5405

2 3992 4044 6119

4 1696 2892 4988

6 1214 3449 3098

8 589 3288 2314

10 1 1111 1130

Table 4.9 shows the optimum fertiliser application for Desiree variety for

Banisogosogo. Application of 300 kg/ha at 6cm gave the highest yield for

Banisogosogo.


corresponding yield for Banisogosogo using weather data of 2012.

Yield (kg/ha) at depth

Default run (80 kg/ha in 2

splits)

4478

Fertiliser (kg/ha) 0 cm 2 cm 4 cm 6 cm 8 cm 10 cm

60 3905 3905 3905 4079 4209 4150

120 4048 4048 4048 4308 4668 4551

180 4068 4068 4068 4914 4944 4862

240 4205 4205 4205 5057 5184 4581

300 4369 4369 4369 5227 4829 4792

80

Table 4.10 shows that the yield is higher than default run and fertiliser optimisation

yield but not higher than irrigation yield.

Table 4.10 shows the optimum fertiliser and irrigation management simulation for

Banisogosogo using weather data of 2012.

Optimum Fertiliser and Irrigation Management

Fertiliser Application (kg/ha at 6 cm) Irrigation Application ( mm) Yield (kg/ha)

300 kg/ha 6.4 mm 5719

The optimum fertiliser application for Desiree variety for Koronivia indicated that

application of 300 kg/ha at 6 cm gave the highest yield for Koronivia (Table 4.11).

Table 4.12 indicated that the yield was higher than default run and fertiliser

optimisation yield but not higher than irrigation yield.


corresponding yield for Koronivia using weather data of 2010.



splits at 5 cm)

4373


60 3987 3987 3987 3972 3989 3987

120 3792 3792 3792 3705 4038 4080

180 4037 4037 4037 3840 3796 3768

240 4406 4406 4406 4144 4344 4368

300 4426 4426 4426 4389 4398 4564


management for Koronivia using weather data of 2010.

Fertiliser Application (kg/ha at

10 cm)

Irrigation Application ( mm) Yield (kg/ha)

300 kg/ha 1.6 4602

81

For Nacocolevu, application of 300 kg/ha at 10 cm gave the highest yield (Table

4.13). The results obtained also indicated that higher yield was obtained under

optimum fertiliser and optimum irrigation treatment (Table 4.14).


corresponding yield for Nacocolevu using weather data of 2010.



splits)

5405


60 2594 2594 2594 2741 2922 3038

120 5036 5036 5036 5295 5551 5659

180 7077 7077 7077 6666 7597 7657

240 7393 7393 7393 7519 7650 7690

300 7649 7649 7649 7801 7895 7925


management for Nacocolevu using weather data of 2010.

Fertiliser Application (kg/ha

at 10 cm)

Irrigation Application ( mm) Yield (kg/ha)

300 kg/ha 1.6 8343

300 kg/ha 4.0 8344

4.2.4 Climate variability (El Niño Southern Oscillation) simulation

The impact of climate variability was also studied on potato yield for the three

locations. Differences in the yield were noticed due to differences in weather data

and the water and nitrogen stress faced under El Niño, La Niña and neutral years for

the three locations. Under Banisogosogo and Nacocolevu simulation, neutral year

received the highest average yield (Tables 4.16, 4.20) while under Koronivia

simulation, El Niño received the highest average yield (Table 4.18).

82

Table 4.15 shows the impact of ENSO on potato yield for non-potential Banisogosogo

simulation from 1983-2012.

El

Niño

Water

Stress

N

Stress

Yield

(kg/ha)

La

Niña

Water

Stress

N

Stress

Yield

(kg/ha)

Neutral Water

Stress

N

Stress

Yield

(kg/ha)

1987 0.62 1.24 2129 1988 0.89 0.40 627 1983 - 0.06 6885

1991 0.15 0.28 779 1996 0.49 0.13 2846 1984 0.45 0.30 3216

1992 0.55 0.23 1934 1998 0.63 0.32 747 1985 0.30 0.14 3816

1993 0.09 0.09 473 1999 0.04 0.07 5500 1986 0.35 0.19 2362

1994 0.08 0.34 957 2000 0.22 0.16 4591 1989 0.33 0.15 5147

1997 0.08 0.21 826 2008 0.61 0.39 1465 1990 0.12 0.16 6179

2002 - 0.07 7744 2012 0.35 0.16 3816 1995 1.05 0.46 642

2004 0.07 0.08 5799 2001 0.58 0.20 3011

2006 0.48 0.10 3692 2003 0.46 0.25 816

2010 0.66 0.28 1407 2005 0.39 0.19 2514

2007 0.8 0.25 766

2009 0.23 0.36 2687

Table 4.16 indicates 7 year average for ENSO yield in Banisogosogo. The results

indicated that highest average yield was obtained under neutral year, with El Niño

giving a percentage difference of 61.4 % while La Niña gave a percentage difference

of 13.88 %.

Table 4.16 shows the 7 year average for ENSO yield for Banisogosogo.

7 Year Average Yield (kg/ha) Percentage Difference (%)

El Niño (1987-2002) 1868.71 61.4

La Niña (1988-2012) 3070 13.88

Neutral Year (1983- 1995) 3527.85

83

Table 4.17 shows the impact of ENSO on potato yield for non-potential Koronivia

simulation from 1990 to 2010.

El

Niño

Water

Stress

N

Stress

Yield La

Niña

Water

Stress

N

Stress

Yield Neutral Water

Stress

N

Stress

Yield

1991 0.41 0.39 6487 1996 0.69 0.18 1651 1990 0.18 0.31 5911

1992 0.58 0.18 4497 1998 0.98 0.12 757 1995 0.45 0.29 3837

1993 0.55 0.21 2056 1999 0.03 0.45 3754 2001 0.45 0.20 5147

1994 0.78 0.34 2709 2000 0.06 0.47 4188 2003 0.75 0.12 3857

1997 0.30 0.42 6577 2008 0.63 0.21 5284 2005 0.50 0.18 2691

2002 0.21 0.38 6107 2007 - - 3964

2004 0.07 0.49 5330 2009 0.46 0.23 5532

2006 0.56 0.23 6257

2010 0.93 0.55 4424

For 5 year average for ENSO yield in Koronivia, the results indicated that highest

average yield was obtained under El Niño at 4465.2 kg/ha with the lowest yield in La

Niña at 3126.8 kg/ha (Table 4.18).

Table 4.18 shows the 5 Year average for ENSO yield for Koronivia.

5 Year Average Yield (kg/ha) Percentage Difference

(%)

El Niño (1991-1997) 4465.2 -4.17

La Niña (1996-2008) 3126.8 31.19

Neutral Year (1990-2005) 4282.6

84

Table 4.19 shows the impact of ENSO on potato yield for non-potential Nacocolevu

simulation from 1990-2010.

El

Niño

Water

Stress

N

Stress

Yield La

Niña

Water

Stress

N

Stress

Yield Neutral Water

Stress

N

Stress

Yield

1991 - 0.05 2686 1996 - 0.09 1954 1990 - 0.16 2434

1992 - 0.06 2292 1998 - 0.08 2456 1995 - 0.05 3223

1993 - 0.09 1833 1999 - 0.09 2890 2001 - 0.05 2546

1994 - 0.09 1998 2000 - 0.05 2587 2003 - 0.06 3084

1997 - 0.08 1839 2008 - 0.32 3000 2005 - 0.05 3114

2002 - 0.04 3283 2007 - 0.09 2292

2004 - 0.06 2445 2009 - 0.07 4022

2006 - 0.09 2689

2010 - 0.07 5352

Table 4.20 indicates 5 year average for ENSO yield in Nacocolevu. The results

indicated that highest average yield was obtained under neutral year, with El Niño

giving a percentage difference of 29.94 % while La Niña gave a percentage

difference of 11.09 %.

Table 4.20 shows the 5 Year average for ENSO yield for Nacocolevu.

5 Year Average Yield (kg/ha) Percentage Difference (%)

El Niño (1991-1997) 2129.6 29.94

La Niña (1996-2008) 2577.4 11.09

Neutral Year (1990-2005) 2880

85

4.3 Discussion

4.3.1 Desiree potential simulation

The potential simulation showed a higher tuber fresh yield, tuber dry yield and LAI

as compared to non-potential simulation for the three areas (Table 4.4). Under

potential simulation, the model assumed that there was no water and nitrogen stress

on the growth and development stages of the plant (refer to Appendix 2A, Table

2.5A, Table 2.6 A, Table 2.9A and 2.10A, Table 2.13A and Table 2.14A). However,

the tuber initiation day (days after planting) for potential simulation occurred after

the non-potential simulation. The model also showed that potato grown in three

location simulated different yield. One reason for the difference in tuber yield for the

three sites can be the soil status and the weather for each site. It was also noticed that

when the aboveground weight (LAI, leaf weight, tops weight and the stem weight)

decreased, the belowground weight (tuber weight) started to increase around the

same time (Figures 4.3, 4.4 and 4.5). This can be due to all the carbohydrates being

partitioned to tuber growth following tuber initiation (Fleisher et al., 2003).

4.3.2 Desiree non-potential simulation

The Banisogosogo non-potential simulation indicated the tuber initiation day was 35

days after planting. The simulated tuber dry weight was 4478 kg/ha and tuber fresh

weight of 22.39 t/ha. The results also indicated that the maximum LAI was 1.7

(Table 4.4). The emergence date was 6th of July which was 7 days after planting

(Appendix 2A, Table 2.7A). The tuber forms around the 9th of August, which was 35

days after planting. The crop was harvested after 80 days, that was, on 20th of

September (Appendix 2A, Table 2.7A). Figure 4.3 and 4.6 indicated that from the

planting date (2nd of July) the LAI increased around the 10th of August. From the

planting day until 10th of August, the LAI, leaf weight, stem weight and tops weight

increased in value. After 10th August to 19th of September, the LAI, leaf weight, stem

weight and tops weight started to decrease while the tuber dry weight started to

gradually increase from 10th of August to 30th of August. From 30th of August to 19th

September there was a sharp increase in the tuber dry weight. The precipitation

increased around 30th of August which also increased the total water content in the

soil. Figure 4.6 also showed an increase in total nitrate rate over time. This can be

due to application of nitrogen fertilisers (on 7th and 23rd of July nitrogen fertiliser was

86

applied which also shows an increase in nitrate rate). The decrease in LAI can be

linked to transfer of photosynthate from the aboveground to the belowground. This

affected the LAI and other aboveground components such as leaf weight and tops

weight. The average leaf area or the ability to maintain complete soil cover by the

canopy during the growing season affects the tuber yield (Boyd et al., 2002). The

stem and leaves are main sink for assimilates. However, as plants mature, these

assimilate are relocated to the organ sinks (Dwelle and Love, 1993; Singh et al.,

1998), in this case the potato tubers. Table 4.3 indicated that the Banisogosogo had

the highest average maximum and minimum temperature for the growing season at

27.93 oC and 21.03 oC respectively with minimum rainfall at 75.9 mm for the

growing season. In tropical highlands, potato production is suitable under

temperature range of 15-18 ºC (Haverkort, 1990) or around 22 ºC (Burton, 1981).

The carbon assimilation and water use efficiency range of potatoes is between 16˚C-

25ºC and the optimal temperature for net photosynthesis at 25 ºC (Ku et al., 1977). In

lowland tropics, the potential yields of potato tubers are highest at daylength less

than 12 hours and average temperature of 20-23˚C (Zaag et al., 1986; Jackson,

1999). For cultivar Desiree, there was an increase in tuber number between the

temperature ranges of 15-23 ºC. Under temperatures of 2 oC and above 30 oC, the

growth and development of potatoes is not possible (Garner and Blake, 1989;

Haverkort, 2007a). The ideal soil temperature for tuber development is around 15oC-

18oC (Levy and Veilleux, 2007). Leaf appearance shows optimal values at an

average daily temperature of about 28 oC (Fleisher et al., 2006a). Regardless of night

temperature, the rate of leaf appearance peaks at 20 oC (Vos, 1995). High

temperatures gave higher specific leaf area as compared to the cool temperatures

(Midmore and Prange, 1991). The optimum temperature is also dependent on

photoperiod (Wheeler, 2006). Kennebec cultivar had a leaf area with a quadratic

relationship with temperature with the largest area at 16.6 oC and 22.1 oC. Main-stem

leaves accounted for > 50 % of the total leaf area at temperatures < 22 oC. However,

the proportion of axillary stem leaf area increased with temperature (Fleisher et al.,

2006b). Higher root temperatures affects the transfer dry matter partitioning to the

haulm (Struik et al., 1989). In cultivar Desiree, cool temperatures gave the largest

number of possible tuber sites whereas high air temperatures decreased the stolon

number and stolon weight (Struik et al., 1989). Results also indicated that there was

presence of nitrogen and water stress (refer to Appendix 2A, Table 2.7A and 2.8A).

87

Nitrogen stress (0.16) began on 9th of August, which was the time that tuber began to

form, followed by 20th of September. Water stress began on the 20th of September,

which was the date of harvest. From the development phase of tuber initiation until

maturity, sufficient water is required for high tuber yield and quality (World

Meteorological Organisation, 2010). Water stress significantly affects stolonisation

and tuberisation as compared to tuber bulking and tuber enlargement stages (Hassan

et al., 2002). Water scarcity can limit yield. Water availability depends on factors

such as the amount and distribution of rainfall, water holding capacity of the soil and

the root-depth (Vos and Haverkort, 2007a). Drought reduces leaf expansion and light

interception affecting dry matter accumulation (Gordon et al., 1997). However, the

initial vegetative stages is not sensitive to early season water stress (Nasseri and

Baramloo, 2009; Ayas and Korukçu, 2010). Application time of nitrogen fertiliser is

also considered important for yield (Hagman et al., 2009). Under current best

management practices, nitrogen uptake efficiency of potatoes is approximately 65 %.

Insufficient level of nitrogen leads to delay in growth whereas excess level of

nitrogen leads to unnecessary increase in the shoot/root ratio that might adversely

impact the gain of other nutrients and water (Bucher and Kossmann, 2007). Altering

nitrogen supply affects tuberisation (Jackson, 1999). Under high nitrogen supply

there is early development and initiation of potatoes crops through seed tuber pre-

sprouting (Möller and Reents, 2007). The protein content of the tubers is also

affected by nitrogen supply (Ozturk et al., 2010). There is rapid nitrogen uptake in

the early season and it progressively slows down or stops (Allison and Allen, 2004).

Under Koronivia non potential simulation, the simulated tuber initiation day was 30

(as shown by Table 4.4). The tuber dry weight was 4373 kg/ha and the tuber fresh

weight was 21.86 t/ha with maximum leaf area index of 2.72. The date of emergence

was 8th of July, which was 6 days after planting (refer to Appendix 2A, Table

2.11A). The tuber started to form around the 1st of August, which was 30 days after

planting. The crop was harvested after 80 days which was 20th of September. Figures

4.4 and 4.7 indicated that from the time of planting (2nd July) the LAI, leaf weight,

stem weight and tops weight increased until the 10th of August. From 10th of August

to 30th August the LAI, leaf weight, stem weight and tops weight decreased and

increased until the 12th of September and decreased until 19th of September. The

tuber dry weight increased when the LAI dropped the first time (10th of August) and

88

it continued to increase gradually until the 30th of August, however, a sharp increase

in tuber dry weight occurred after 30th of August. Precipitation levels increased

gradually and it was also noticed that around 30th August when the precipitation

levels were high and the nitrate levels in the soil decreased. There was presence of

nitrogen and water stress on 1st of August and 20th of September (refer to Appendix

2A, Table 2.12A), during the stages of tuber initiation and harvest respectively.

Depletion of soil water to less than 65% of available water can result in reduction of

yield (Government of Alberta, 2011). The amount of tuber set per stem is decreased

due to increased period of water stress before tuber initiation (MacKerron and

Jefferies, 1986). This can also lead to tuber physiological disorders such as brown

centre, hollow heart, translucent end, secondary growth and growth cracks

(MacKerron and Jefferies, 1985; MacKerron and Waister, 1985; Rex and Mazza,

1989). The first physiological response of the plant, due to water stress, is stomatal

closure to further protect the leaves from more water loss. The canopy temperature

increases as a consequence of reduced evaporative cooling of the leaves and this

leads to reduced carbon dioxide diffusion into the leaves. Due to these,

photosynthesis is reduced whereby reducing the production of starch and sugar and

the transfer of these products from the leaves to the tubers (Davies and J., 1991; King

and Stark, 1997; Bruhn, 2002; Yordanov et al., 2003). The yield and quality of

potato tubers depend on maximising the steady accumulation of photosynthetic

products in the tubers. Lack of water reduces tuber expansion (King and Stark,

1997). Nitrogen is considered the limiting factor in potato growth and development.

Nitrogen deficiency is characterised by yellow leaves, stunted growth and lower

production yields (Bowen et al., 1997).

The non-potential simulation of Nacocolevu indicated that the simulated tuber

initiation day was 33 days after planting. The simulated dry tuber weight was 5405

kg/ha and tuber fresh weight was 27.02 t/ha with maximum leaf area index of 1.03.

The emergence day was 9th of July (refer to Appendix 2A, Table 2.15A). Tuber

initiation started on 4th of August, which was 33 days after planting. The crop was

harvested 80 days after planting, that is, 20th of September. Figures 4.5 and 4.8

showed that the LAI, leaf weight, stem weight and tops weight increased from the

time of planting until 10th August. A drop in LAI, leaf weight, stem weight and tops

weight were noticed after 10th August until the harvest day which was 19th of

89

September. The graph also showed that the tuber dry weight started to increase

sharply after 10th of August. Initially, there was a fast increase in tuber weight

followed by a decrease in growth rate during senescence (Scott and Wilcockson,

1978). The soil water content at Nacocolevu was consistent throughout the planting

season. The precipitation increased gradually over time and the figure also indicated

that the total nitrate leached from the soil also increased gradually over time. Table

4.3 indicated that Nacocolevu had the average maximum and minimum temperature

for the growing season at 27.7 oC and 18.1 oC respectively with rainfall at 147.48

mm for the growing season. Optimum final leaf size is obtained at relatively low

temperatures. Individual leaf size increases with a decrease in temperature well

below the optimum temperature. The leaf appearance rate and leaf expansion rate are

at its optimum below temperatures of 25 oC (Struik, 2007b). The optimum growth

rate of sprout is at 20 ºC or slightly higher (Klemke and Moll, 1990) while a base

temperature of 2 ºC was required for sprout extension (MacKerron and Waister,

1985). A delay and reduction in sprout growth is caused by temperatures lower than

20ºC (Firman et al., 1992). Stolon formation and stolon yield is reduced by high soil

temperatures (Midmore, 1984b). Under 32/27 ºC there was no underground tuber

development or in some cases, the stolons grew out of the hot soil and formed aerial

tubers (Reynolds and Ewing, 1989). A fluctuation of soil temperature can lead to

heat sprouts, chain tubers and secondary growth of tubers (Ewing, 1981a).

Temperature above 25 ºC or more can delay or even impede emergence, reduce plant

survival rate and even reduce the number of main stems per plant (Midmore, 1984a).

The size of root system is reduced at 30 ºC due to a decline in cell division followed

by cessation of root elongation (Sattelmacher et al., 1990). There was nitrogen stress

during the periods of emergence-begin tuber and planting-harvest (refer to Appendix

2A, Table 2.16A). Approximately 85 % of total nitrogen uptake occurs by 45-65

days after emergence (British Potato Council, 2012). Figure 4.8 indicates that there

was consistent water in the soil and the precipitation level also increased over time.

Hence, there was no water stress (refer to Appendix 2A, Table 2.16A). Farmers

should also be mindful of pests also cause reduction in yield (Haverkort, 2007a).

90


The optimisation of planting time was conducted to see which months are suitable

for planting potatoes. Table 4.5 suggested that planting of potato in Banisogosogo

during the months of May and June gave higher yield at 5356 kg/ha and 4567 kg/ha

respectively as compared to other months. These months gave the highest yield

because these months had relatively cool night temperature. The month of May had

average minimum temperature of 21.8 oC during emergence and tuber formation,

21.4 oC between tuber formation and maturity and 21.7oC from planting to harvest.

Under non-potential conditions, the yield at Banisogosogo for the month of July was

4478 kg/ha while under optimisation treatment, the yield of potatoes in year 2009

was 3698 kg/ha. One reason for this can be the difference in high average solar

radiation received during July in 2012. In 2012, the month of July received average

minimum temperature of 20.5 oC during emergence and tuber formation, 21.8 oC

between tuber formation and maturity and 21.3 oC from planting to harvest with

high solar radiation (13.9 MJ/m2, 17.6 MJ/m2, 16.2 MJ/m2 respectively for each

development phase) and rainfall (53.4 mm, 80.5 mm and 134.3 mm respectively for

each development phase) while in 2009, the month of July received average

minimum temperature of 21.6oC during emergence and tuber formation, 21.0 oC

between tuber formation and maturity and 21.3 oC from planting to harvest with

high solar radiation (14.5 MJ/m2, 16.8 MJ/m2, 15.6 MJ/m2 respectively for each

development phase) and rainfall (69.7 mm, 128.7 mm and 203.2 mm respectively for

each development phase). Potatoes require cool night temperatures for tuberisation.

Optimum overnight temperatures are between 15-18 oC, above 22 oC will adversely

affect tuber development (Levy and Veilleux, 2007; Food and Agriculture

Organization of the United Nations, 2008). Cool temperatures also have the highest

rate of relative partitioning of potato tubers (Struik, 2007b). Also, these months

received high solar radiation. Planting of potatoes in months of high solar radiation

will produce more tubers per stem (Firman and Daniels, 2011). On the other hand,

the preferred planting month for Koronivia was August with a tuber yield of 7956

kg/ha. This was because the month of August had low maximum and minimum

temperatures as compared to other months. Under non-potential conditions (weather

data of 2010), the yield at Koronivia for the month of July was 4373 kg/ha while

under optimisation yield for July was 6703 kg/ha. The yield under optimisation

91

treatment for July was higher as compared to non-potential simulation because in

July 2009 the average maximum temperature was lower with high rainfall. The

average maximum temperature in July 2009 was 26.0 oC during emergence and tuber

formation, 26.6oC between tuber formation and maturity and 26.4 oC from planting

to harvest and high rainfall (66.1 mm, 403.4 mm and 506.4 mm respectively for

each development phase) whereas in July 2010 the average maximum temperature

was 26.5 oC during emergence and tuber formation, 27.8oC between tuber formation

and maturity and 27.3 oC from planting to harvest and high rainfall (94.2 mm, 155.2

mm and 253.3 mm respectively for each development phase). Finally, Nacocolevu

optimised planting month was August with a yield of 7612 kg/ha. This may be due to

the fact that August received low average maximum and minimum temperatures and

high average solar radiation as compared to other months. The month of August

received average solar radiation of 15.4 MJ/m2 from emergence to tuber formation,

18.1 MJ/m2 from tuber formation until maturity and from planting to harvest an

average of 17.0 MJ/m2 of solar radiation. The non-potential yield in July (with 2010

weather data) was 5405 kg/ha while under optimisation treatments (weather data of

2009) the yield was 4823 kg/ha. The non-potential simulation received higher

average rainfall as compared to optimisation treatment. The rainfall received in July

2010 was 138.7 mm, 264.4 mm and 475.6 mm for during emergence and tuber

formation, between tuber formation and maturity and from planting to harvest

respectively while in July 2009 the rainfall received was 135.0 mm, 267.6 mm and

424.6 mm for during emergence and tuber formation, between tuber formation and

maturity and from planting to harvest respectively.

Likewise, optimisation was also conducted for row spacing. Table 4.6 showed that

optimised row spacing should be 40 cm in Banisogosogo. Row spacing of 40 cm

gave a yield of 4988 kg/ha whereas a row spacing of 75 cm (as suggested by the

Agricultural Ministry) gave a yield of 4478 kg/ha. Increasing the row spacing beyond

80 cm decreased the yield. Koronivia simulation indicated that a row spacing of 30

cm, produced the highest yield (4555 kg/ha) as compared to the 75 cm of row

spacing which produced a yield of 4373 kg/ha. Nacocolevu simulation also

suggested that the highest yield of potato can be obtained at 30 cm with a yield of

7290 kg/ha. Beyond the optimised plant spacing there was a decrease in yield. It is

uncertain in what way spacing of potato plants will affect growth (Iqbal, 1991).

92

However, studies in Philippines indicated that an increase in plant population and

was proportional to increase in the amount of intercepted solar radiation giving

higher tuber yield (Zaag and Demaganate, 1987; Midmore, 1988).

Potato water use rate begins at 0.4 mm per day while a fully sprouted potato plant

uses 7 mm of water per day (Government of Alberta, 2011). The results indicated

that for the three locations the default irrigation application did not give the highest

yield. Under Banisogosogo irrigation optimisation, irrigation of 6.4 mm per plant

gave the highest yield at 6268 kg/ha. The optimisation treatment for Koronivia and

Nacocolevu indicated that application of 1.6 mm of water per plant gave the highest

yield at 8411 kg/ha and 5434 kg/ha respectively. It was also noticed that under

Nacocolevu irrigation optimisation, there was not much change in potato yield. This

was because no water stress was noted under non-potential run and also in the

optimisation simulations. The total water content of Nacocolevu soil also remained

consistent. Even in normal conditions of watering, water stress occurs at noon due to

high transpiration rates (Posadas et al., 2008). Leaf dry matter is the most responsive

parameter to irrigation treatments (Geremew et al., 2008). It is predicted that world-

wide average crop yield will increase by approximately 50 % if water supply to the

crops is optimised (Posadas et al., 2008).

Furthermore, the optimisation simulation also indicated that planting depth also

affected tuber yield. Table 4.8 suggested that to obtain the highest yield of potato, the

seeds should be planted at 1.5 cm for Banisogosogo and Koronivia. However, at

Nacocolevu, potato seeds should be planted at 2 cm, which gives a yield of 6119

kg/ha. If potato were planted beyond the optimised planting depth, the potato yield

decreased. The planting depth should be adjusted to the most important factor which

is the soil moisture and soil temperature. It should also be noted that deep planting

protects the tuber from disease and pests (Cortbaoui, 1988).

The results of fertiliser treatment indicated that to obtain the highest yield in

Banisogosogo 300 kg/ha of fertiliser should be applied at a depth of 6 cm which

gave a yield of 4972 kg/ha (Table 4.9). Koronivia simulation, Table 4.11, indicated

that application of 300 kg/ha fertiliser at 10 cm depth gave the highest yield of 4564

kg/ha. Under Nacocolevu simulation, Table 4.13, the optimisation treatment

93

indicated that application of 300 kg/ha of fertiliser at 10 cm beneath the surface gave

the highest yield of 7925 kg/ha. It was also noticed that the model gave the same

yield for fertiliser application at 0 cm, 2 cm and 4 cm. Generally, for upland crops,

surface application (0 cm) will be similar to application up to top layer soil. If the

soil layer 1 is 5 cm deep, the model will show very little or no difference between 0-

5 cm applications. However, at 6 cm soil layer, the model to starts to show

difference. The only exception to the above is when soils have high pH (pH >8). The

nitrogen fertiliser requirement ranges from 2.5-5.9 kg ha-1 per tonne of tuber yield

(Hagman et al., 2009). Higher rates of nitrogen increases yield until some maximum

yield is reached but there is delay in the onset of tuber bulking (Vos, 2009).

However, excessive nitrogen application adversely affects the availability of

moisture by elevating salt levels. During the initial stages, excessive nitrogen

application, can delay the transition of shoot accumulation to shoot translocation of

nitrogen to tubers. The first in-season nitrogen application should occur after tuber

initiation as during this stage 30-40% of nitrogen has been taken up. During tuber

bulking, the nitrogen levels should be maintained as this is considered as a crucial

stage in potato development (Lang et al., 1999). Nitrogen treatments should take into

account nutrient reserves and applied organic manures whereby maximising the

economic value of the crop and minimising the environmental impact (Farming and

Wildlife Advisory Group, 2012). Nitrogen application increase tuber quality only to

a certain amount and has an impact on leaf area index, plant height and stem number

(Kuruppuarachi et al., 1989). Nitrogen fertiliser application and soil nitrogen

availability enhances the rate of canopy development and radiation absorption

(Stalham and Allison, 2011). However, the availability of soil nitrogen is determined

by nitrogen mineralisation and nitrogen leaching process (Oliveira, 2000). When the

canopy ceases to take up nitrogen it starts to export nitrogen to the potato tubers

(Allison and Allen, 2004).

The model is also a good measure for studying the interaction of water management

under different management practices and nitrogen supply (Bowen, 2003). When the

optimum nitrogen application and the optimum irrigation application were simulated

together for the three locations, it was noticed that the yield decreased for

Banisogosogo and Koronivia but increased for Nacocolevu. One reason for was that

when the optimum amount of irrigation and fertiliser that has been obtained

94

separately are combined and run as a simulation (both inputs are changed

simultaneously), there is interaction between water and nitrogen and this will not

necessarily give the optimum yield.

4.3.4 ENSO effect

Climate variability measured in inter-annual and inter-season scales is one of the

significant factors that affect agricultural production. Most plants are particularly

sensitive to changes in environmental conditions during ENSO which affect growth

stages such as flowering and grain filling (Baethgen and Magrin, 2000). ENSO has

an impact on weather which in turn affects our crop yield (Legler et al., 1999). The

simulation was conducted for the last 30 years, that is, the simulation was run from

1983. The assumption here was that the fertiliser and the irrigation application

remained same throughout the 30 year period. The results, as shown by Table 4.15,

indicated that there was a high occurrence of neutral years (12 out of 30 years) and a

low occurrence of La Niña years (7 out of 30 years). The neutral year was used as a

baseline to calculate the percentage difference. The average of the first 7 years of the

ENSO state yield was taken and it was found out that neutral year had the highest 7

year average yield (3527 kg/ha) and El Niño had the lowest 7 year average yield

(1868.71 kg/ha with percentage difference at 61.4%) while La Niña simulated an

average yield of 3070 kg/ha with a percentage difference of 13.88% (Table 4.16).

One reason for the low yield in El Niño phase was the high level of water and

nitrogen stress faced by the plant. The warm ENSO periods cause significant

reduction in potato yields (Orlove et al., 2000). In Fiji, El Niño is associated with dry

spells where the rainfall is significantly below average. More than 80% of the El

Niño in Fiji were between 1920 and 2005 and these were associated with drought

(Fiji Meterological Services, 2010). The lowest yields produced were in 1991, 1993,

1994, and 1997 which were El Niño years. One reason for this can be due to decrease

in soil moisture (Poveda et al., 2001). El Niño induced drought can pose a major

agricultural risk for rainfed crops (Yokoyama, 2002). During an El Niño event, the

cyclonic activity increases between Fiji and Tahiti and the drought conditions prevail

from Southwest Pacific to Southeast Asia (Grove and Chappell, 2000). For

Koronivia simulation, the ENSO simulation was run from 1990 to 2010, a total of 21

years. Table 4.18 shows the average over the first 5 years. El Niño year gave the

highest yield at 4465.2 kg/ha, La Niña gave the lowest yield at 3126.8 kg/ha with a

95

percentage difference at 31.19 % and neutral year simulated an average yield of

4282.6 kg/ha. The lowest yield was recorded during 1998 which was also a La Niña

year. This can be due to the fact that heavy rainfall in this year caused the soil

moisture to reach maximum level in response to wetter than normal conditions

(Poveda et al., 2001) whereby causing low yield. For Nacocolevu simulation, it was

found out that neutral year gave the highest average yield at 2880 kg/ha while lowest

average yield was recorded during El Niño period with a yield of 2129.6 kg/ha at

29.94-% percentage difference while La Niña simulated an average yield of 2577.4

kg/ha with a percentage difference of 11.09 %. As shown by Table 4.19 and Table

4.20, El Niño recorded the lowest yield due to high nitrogen stress the crop faced in

this phase. However, one of the limitations of such kind of study is the small number

of years with accessible daily data that is required by the simulation models. The

number of years for each ENSO state is reduced as there is separate consideration of

El Niño and La Niña years. Scientists have improved existing models or developed

new weather generators linked to ENSO phases to overcome such limitations (Singh

et al., 2002). A study conducted in Uruguay highlighted that the yield of maize is

reduced twice as much in La Niña years as compared to neutral years (Baethgen,

1998). During ENSO years, the timing of maize sowing date also had to be adjusted.

El Niño years required early sowing dates while La Niña years required a delay in

sowing dates (Magrin et al., 1999)

4.4 Recommendations

It is recommended that the planting time should be adjusted based on the weather

conditions of a site, such as, maximum and minimum temperature and the amount of

rainfall a site receives which is suitable for potato cultivation. For example,

Banisogososo optimum planting time was April, May and June and the optimum

planting time for Koronivia was September. It can also be said that due to changes in

weather, the optimum planting date for each site can also change. Hence, sound

recommendations should not be provided based on a single year simulation. If the

agricultural officers can simulate the planting date for a particular area they can

advise the farmers on the planting time to obtain the highest yield. Planting too early

or too late in the season can lead to slow emergence, decreased plant vigor, delayed

canopy development and reduced time available in tuber bulking (Dwelle and Love,

1993).

96

From all the simulations conducted, it was found out that to obtain the highest yield,

a row spacing of either 30 cm or 40 cm is recommended. Planting closer than

optimal plant spacing can limit the photosynthetic capacity and bulking rates of

individual tubers would decrease which gives smaller tubers. On the other hand, a

row spacing wider than optimal can lengthen the time for full canopy development

which can reduce carbohydrate supply to tubers (Dwelle and Love, 1993). A

decrease in between-row spacing was effective in increasing tuber yield in hot

tropics (World Meteorological Organisation, 2010).

From the optimisation treatments it is recommended that planting depth should be

1.5 cm at Banisogosogo and Koronivia to obtain the highest yield while the

optimised yield is obtained at 2 cm of planting depth at Nacocolevu.

Nitrogen is considered the limiting factor in potato growth and development. It is

recommended that nitrogen fertiliser should be applied based on the soil nitrogen

content and also the amount of rainfall a site receives. Model simulation can also

predict the time of nitrogen stress. Hence, application of nitrogen fertiliser should

take place a few days before the stress. The simulation shows that to obtain the

highest yield 300 kg/ha of fertiliser should be applied at a depth of 6 cm, 8 cm and 10

cm at Banisogosogo, Koronivia and Nacocolevu respectively. Application of soil

organic matter also improves and stabilises the soil structure so that the soils can

absorb higher amount of water without causing surface run, which can result in soil

erosion and even flooding. Soil organic matter also improves water absorption

capacity of the soil for during extended droughts. A no or low tilled soil is ideal for

preserving the structure of soil for fauna and related macrospores such as

earthworms, termites and root channels. The surface mulch cover protects the soil

from high temperatures and excess evaporation losses and can reduce the water

requirement by 30%. Organic agriculture increases soil organic carbon content,

reduces mineral fertilisation and reduces on farm organic costs (Food and

Agriculture Organization of the United Nations, 2007). Furthermore, management

practices such as planting population density, use of mulch and irrigation can alter

the soil temperature within the root zone that can affect stolonisation, tuber initiation

bulking and tuber enlargement. Application of mulch also reduces soil temperature at

15 cm depth by 1.5-4.5 ºC, which results in faster emergence, earlier canopy

development and higher tuber yields (Mahmood et al., 2002).

97

Furthermore, it is recommended that irrigation should be applied taking into

consideration the weather of the site and the soil structure, for example, for

Banisogosogo the optimised irrigation application is at 6 mm while in Nacocolevu

the optimised irrigation application is at 1.6 mm. When the soil moisture drops below

critical level, it can either reduce or stop canopy or tuber growth and shortens the

tuber bulking period causing a variety of tuber defects. On the other hand, excessive

irrigation can cause a reduction in tuber reduction by restricting plant physiological

activity, nutrient uptake and disease susceptibility (Dwelle and Love, 1993). There

are currently many alternatives available to reduce water loss and enhance water

productivity, such as, that of drip irrigation, improved varieties, growing alternative

crop species and better timing of crop cycle in the season (Vos and Haverkort,

2007b). Factors such as crop appearance, soil water tension, the rate of crop

evapotranspiration, precipitation and the amount of water applied should be kept in

mind. If farmers have knowledge of these factors, irrigation can be well balanced

resulting in high yields of better quality tubers (Pereira and C.C. Shock, 2006). The

optimisation simulations of a particular area through sensitivity analysis can inform

its users of the best irrigation amount and methods that should be applied to get the

best yield. For example, the best irrigation option in Banisogosogo was the

application of 8 mm of irrigation per plant whereas the simulation for Koronivia

indicated that increasing the irrigation amount increased the yield.

It is recommended that El Niño and La Niña years, farmers should be mindful of the

weather conditions, particularly, the amount of rainfall and the average temperature

for a site. Taking these factors into consideration, the farmers should plan their

planting time so that the crop faces minimum stress. Adjustment in cropping calendar

based on crop simulation studies and in response to these climate variability is one

way of managing climate variability in crop production (United Nations, 2002). The

El Niño events have been increasing in frequency over the last 40 years (Yokoyama,

2002). It would be desirable to have experiments conducted over many years (Singh

et al., 2002). These forecast will only be helpful to farmers if they are accurate,

relevant and timely (Stern and Easterling, 1999) and address the current needs of

users (Finan, 1999; Food and Agriculture Organization of the United Nations, 2007).

98

4.5 Challenges faced Given below is a list of challenges that was faced while conducting research on this

section.

� The field site initially selected for this research had to be changed. This is due

to the fact that this site was flood prone and a new site had to be selected.

� The weather data that was received from Fiji Meteorological Services had

missing data. These missing data were replaced with NASA values to allow

successful model simulations.

� The number of years under ENSO simulation was different. Hence, it became

difficult to compare results of different sites.

� Model simulates as high yield for Koronivia as Nacocolevu. More research

needs to be done at both Koronivia and Nacocolevu stations to calibrate and

validate the model outputs.

99

Chapter 5 Simulating the impacts of future climatic scenario on potato

production in Fiji using the calibrated DSSAT SUBSTOR Potato model

5.0 Introduction

Climate change is affecting all areas of development and livelihood, from

subsistence food production in the rural communities for food security to commercial

and industrial development in urban centers. It will also affect economic activities of

PICs and lead to changes in international supply and delivery of food to islands

whereby having implications for food security in the Pacific Island Region

(Secretariat of the Pacific Community and CSIRO, 2011)

However, quantifying the impacts of future climatic projections on Pacific Island

root crops is a major challenge due to limited available tools and high uncertainties

in Global Climate model projections for the PICs. The latter challenges have been

addressed by the results from Pacific Climate Change Science Program (PCCSP).

They managed to deliver better country level estimates of timing and magnitude of

such future potential climate changes (Secretariat of the Pacific Community and

CSIRO, 2011). Scientists from Pacific Climate Change Science Program have

evaluated twenty four global climate models and eighteen of these best characterise

the climate of Western Tropical Pacific region. These eighteen models have used to

develop climate projections for Fiji. Under PCCSP Climate Futures, three emission

scenarios are available and these are B1 (low emission), A1B (medium emission) and

A2 (high emission) for three time period 2030, 2055 and 2090 (Pacific Climate

Change Science Program, 2011b).

It has been observed that the south-west Pacific has become drier while the central

equatorial Pacific is receiving more rainfall. The projections indicate that these

trends will continue over many decades. The region is already experiencing changes

in temperature and rainfall due to annual and decadal variability such as ENSO. In

the next 15-25 years, ENSO is expected to remain the most dominant influence on

the region’s climate and also food production, after which the climate change signal

will become more pronounced. Climate projections also indicate that the intensity of

extreme events will increase. This will increase the risk of crop losses and food

production. However, the degree of loss will also depend on adaptation responses

100

and risk reduction measures put in place across the food productions system.

Postharvest quality of the crop is also going to be affected by climate change. For

example, increases in temperature and carbon dioxide will cause tuber malformation

and occurrence of common scab on potatoes. Harvesting of mature crops can also be

delayed due to unfavorable weather conditions which can promote development of

rot. Storage of food under high temperatures can also promote rot and poor quality

(Secretariat of the Pacific Community and CSIRO, 2011).

CERES-Rice model responds well to changes in temperature, atmospheric

concentration of carbon dioxide, nitrogen management and seasonal and varietal

differences (Singh and Ritchie, 1993). Climate change crop simulation studies

showed that an increase in temperature will shorten the life cycle of maize and rice.

The short growing season will reduce the yield of maize as confirmed by CERES

Maize model while CERES-Rice model predicted an increase in rice. Water stress

and selection of soil will further have an impact on crop’s life cycle. Drought prone

soil with lower water holding capacity will hasten maturity and yield will be reduced

due to water stress. Future precipitation will have an impact on leaching losses of

nitrogen and soil erosion (Singh et al., 1990).

Potatoes will be particularly sensitive to future climate changes, both directly from

changes in temperature and rainfall and indirectly by modifying soil water balance

by affecting the amount of water available to plants and also have an impact on other

soil land management practices such as seedbed preparation, spraying and harvesting

(Knox et al., 2010). The potato cultivation area in the (sub) tropics will suffer a

major reduction in tuber yield with limited adaptation strategies in these regions

(Hijmans, 2003b). The impact of climate change such as carbon dioxide,

precipitation, solar radiation and maximum and minimum temperature levels can be

studied in DSSAT using Environmental Modifications in the WeatherMan (Jones et

al., 2003). The DSSAT SUBSTOR Potato model was employed in Egypt to study

climate change impacts on two cultivars, namely, Dezareah and Valour, under 2050

A1 emission scenario to predict potato yield. The simulation indicated that potato

yield will decrease by 11-13 % under A1 emission scenario compared to 2005/2006

season. The highest tuber yield for both the cultivars were obtained with irrigation

levels of 100 % with CSIRO and Had CM3 models (Abdrabbo et al., 2010). The

101

DSSAT SUBSTOR Potato model and Global Circulation Models (G CM) were used

to evaluate the impact of climate change on crop growth, yield and irrigation

requirements using Maris Piper cultivar in England. The authors stated that if the

current management practices, such as fertiliser application, remain unchanged, the

impact of climate change for 2050 will be relatively minor (+3 % to +6 %).

However, under optimum irrigation and fertiliser management conditions, the

potential yields will increase by an average of 13%-16%. The study also stated that

the seasonal irrigation demand for irrigation will increase by 14-30 % (Knox et al.,

2010). The DSSAT SUBSTOR Potato model and MAGICC/SCENGEN software

were used to study the impact of climate change on tuber fresh yield under different

sowing dates for four climate change scenarios, A1, A2, B1 and B2 for 2025, 2050,

2075 and 2100 in Egypt (using 2005/2006 as a baseline) on Valour cultivar. Under

first cultivation timing (January 1st, January 15th and January 30th) yield reduction

was noted by -1.41 % to -3.98 % while an increase in yield was noted for second

cultivation timing (September 30th, October 15th and October 30th). The loss in yield

was reduced in the second cultivation when the cultivation timing was changed from

January 30th to January 15th. Under all emission scenarios, 100 % water levels gave

the highest yield while 80% water level gave the highest water efficiency (Medany,

2006). The DSSAT SUBSTOR Potato model was also used for adaptation purposes

under controlled environment hydroponic production of potato (cultivar Norland)

under elevated carbon dioxide concentrations in Ohio, New Jersey. Adjustments

were made to cultivar difference, genetic-co-efficient, radiation use efficiency and

light absorption and changes to leaf senescence, carbon mass balancing and the

response of crop growth to elevated carbon dioxide levels (Fleisher et al., 2003).

This chapter will discuss the impact of future climatic scenario, PCCSP A1B and A2

emission scenario, on potato production for 2030, 2055 and 2090. It will also focus

on simulated crop management strategies using the calibrated DSSAT SUBSTOR

Potato model to maximise potato yield under stressful conditions in three sites in Fiji.

102

5.1 Methodology

5.1.1 Simulation sites

The three sites (Banisogosogo, Koronivia and Nacocolevu) presented in Chapter 4

were used for simulations of future climate scenario impacts on potato production.

The outputs of these simulations are presented in this Chapter.



The thirty year weather data used for the simulation of future climate scenario

impacts on potato production for the three studied sites were similar to Chapter 4.

For environment modification of future climate scenario, PCCSP medium (A1B) and

high (A2) emission scenario were used. The temperature, precipitation and the level

of carbon dioxide modification were added under Environment Modification in

DSSAT SUBSTOR Potato model. The table below (Table 5.0) summarises the

environment modification climatic parameters.

Table 5.0 shows the environment modification climatic parameters. These include

minimum temperature, maximum temperature, precipitation and the concentration of

carbon dioxide under each emission scenario that were used for future climate

simulation.

Year Medium Emission (A1B) High Emission (A2)

Minimum

Temperature

(oC)

Maximum

Temperature

(oC)

Precipitation

( mm)

Carbon

Dioxide

(vpm)

Minimum

Temperature

(oC)

Maximum

Temperature

(oC)

Precipitation

( mm)

Carbon

Dioxide

(vpm)

2030 1.2 1.2 1.13 490 1 1 1.15 520

2055 1.9 1.9 1.17 550 1.7 1.7 1.17 600

2090 2.9 2.9 1.18 700 3.2 3.2 1.22 950

5.1.2.2 Soil data

The soil information used for the simulations included Banisogosogo Replica 2 Soil,

Koronivia Silt Loam and Nacocolevu Soil, Rolling Phase (same as Chapter 4).

5.1.2.3 Crop management

The crop management practices for the three sites, such as, planting time and

method, cultivar (Desiree), row spacing, nitrogen application and irrigation

applications were similar to Chapter 4 (Table 4.0).

103


The simulation for optimisation of crop managements (planting date, row spacing,

planting depth and irrigation and fertiliser amount) to maximise Desiree growth and

yield was conducted using Sensitivity Analysis Option in the DSSAT SUBSTOR

Potato model with available weather and soil data. Weather data of 2012 was used to

conduct Banisogosogo optimisation treatments (except for planting time). The

Koronivia and Nacocolevu optimisation was conducted with 2010 weather data. The

planting time for all three sites was conducted with 2009 weather data. The

optimisation treatments and default runs were similar to optimisation treatments in

Chapter 4.

5.1.4 Model output analysis


run the simulations for future climate impacts on potato growth and yield. All

summary outputs, graphical and initial analysis were done in DSSAT 4.5 Output.

Other statistical analysis was conducted in Microsoft 2007 Excel Spreadsheet. These

outputs are presented in the results section.

5.2 Results

The impacts of future climatic scenario were simulated on potato production in Fiji

using calibrated DSSAT SUBSTOR Potato model. In this study, two PSSCP climate

scenarios were used. The emission scenario included medium emission scenario

(A1B) and high emission scenario (A2) for three time periods (2030, 2055 and

2090). The weather data of Banisogosogo, Koronivia and Nacocolevu were modified

under the Environment Modification in DSSAT SUBSTOR Potato model. These

modification included changes in minimum temperature, maximum temperature,

precipitation and the concentration of carbon dioxide under each emission scenario

as stated by PSSCP Country Reports (Fiji). The results are divided into two sections;

(i) the future climate simulations with potential and non-potential production and (ii)

the optimisation treatments.

The future climate simulation, for the three locations, indicated that the number of

days for tuber initiation increased while the tuber fresh and dry weight decreased.

104

The highest LAI were found under 2090 potential medium and high emission

scenario (Tables 5.1, 5.2 and 5.3).

Finally, the optimisation treatment suggested that under each emission scenario

different crop management practices optimised the yield (Table 5.4, Table 5.5 and

Table 5.6).

5.2.1 Future climate simulations

Under Banisogosogo future climate simulation (Table 5.1), the tuber dry and fresh

weight decreased steadily from 2030-2055. For 2055 emission scenario, tuber yield

was only noticed under A1B non-potential simulation. It was also noticed that yield

obtained under A2 was higher than the yield obtained under A1B emission scenario.

The non-potential tuber initial day was earlier than tuber initiation day of potential

simulation. Likewise, the LAI increased under 2030-2090 potential and non-potential

simulation for A1B and A2 emission scenarios.

Table 5.1 shows the Banisogosogo simulation results of future climate simulation. 2030 2055 2090

Potential Non-

Potential

Potential Non-

Potential

Potential Non-

Potential

Variable A1B A2 A1B A2 A1B A2 A1B A2 A1B A2 A1B A2

Tuber Initiation Day

(dap)

57 54 55 51 -99 -99 72 66 -99 -99 -99 -99

Tuber Dry Weight

(kg\ha) harvest

1698 2395 949 1459 0 0 55 301 0 0 0 0

Tuber Fresh Weight

(t\ha) harvest

8.49 11.98 4.75 7.3 0 0 0.27 1.5 0 0 0 0

Leaf Area Index,

Maximum

10.79 9.8 6.83 7.0 20.41 25.93 7.99 8.59 20.67 21.98 8.00 8.20

Under Koronivia potential simulation (Table 5.2), the tuber initiation day increased

from 2030-2090 A1B and A2 emission scenario. However, it was noticed that the

non-potential tuber initiation day was earlier than the tuber initiation day for

potential simulation. The tuber production (tuber dry and fresh weight) decreased

from 2030-2090. No tuber yield was noted for 2090 potential simulation. It was also

noticed that non-potential production was lower than the potential production.

Furthermore, the LAI increased under potential and non-potential simulation from

2030-2090. However, the non-potential LAI was lower than the potential LAI.

105

Table 5.2 shows the Koronivia simulation results of future climate simulation. 2030 2055 2090

Potential Non-Potential Potential Non-

Potential

Potential Non-

Potential



(dap)

42 40 37 36 49 47 45 43 -99 -99 54 62

Tuber Dry Weight

(kg\ha) harvest

5909 6935 2369 2713 2615 4246 1578 2290 0 0 617 233

Tuber Fresh Weight

(t\ha) harvest

29.54 34.68 11.84 13.56 13.08 21.23 7.89 11.45 0 0 3.08 1.16

Leaf Area Index,

Maximum

12.68 11.99 4.03 3.98 16.83 20.14 4.7 4.81 35.41 38.41 6.04 6.92

For Nacocolevu simulations (Table 5.3), tuber yield was possible under 2030, 2055

and 2090 potential and non-potential simulation for A1B and A2 emission scenario.

Eventhough, production was possible under all emission scenario, a steady decrease

in tuber fresh and tuber dry weight was noted for both potential and non-potential

simulation. The results also suggest that non-potential production was lower than

potential production. The tuber initiation day also increased from 2030-2090

emission scenarios. It was also noted that the non-potential tuber initiation day were

earlier than potential simulation. Similarly, the LAI increased from 2030-2090 with

potential simulation showing higher value.

Table 5.3 shows Nacocolevu simulation results of future climate simulation.

2030 2055 2090

Potential Non-Potential Potential Non-

Potential

Potential Non-

Potential



(dap)

41 40 40 39 47 44 45 44 63 69 61 63

Tuber Dry Weight

(kg\ha) harvest

6584 7370 4908 5521 4093 6187 1993 2949 586 93 523 314

Tuber Fresh Weight

(t\ha) harvest

32.92 36.85 24.54 27.61 20.46 30.93 9.96 14.75 2.93 0.46 2.62 1.57

Leaf Area Index,

Maximum

6.87 6.64 3.55 3.19 7.81 8.34 5.00 4.69 12.07 15.64 5.68 5.74

106

Figures 5.0 and 5.1 show the potential and non-potential future climate simulations

for Banisogosogo for LAI, leaf weight, stem weight, tops weight, tuber dry weight

and tuber fresh weight. For potential simulation (Figure 5.0), the highest LAI (a), leaf

weight (b), stem weight (c) and tops weight (d) were obtained under 2055 high

emission scenario while tuber dry (e) and fresh (f) were obtained under 2030 high

emission scenario. Under non-potential simulation (Figure 5.1), the highest LAI (a)

and leaf weight (b) were attained under 2090 high emission scenario, the highest

stem (c) and tops weight (d) were achieved under 2055 high emission scenario while

the highest tuber dry weight (e) and tuber fresh weight (f) were obtained under 2030

high emission scenario

Likewise, Figures 5.2 and 5.3 show the potential and non-potential future climate

simulations for Koronivia for LAI, leaf weight, stem weight, tops weight, tuber dry

weight and tuber fresh weight. For potential and non-potential simulation, the highest

LAI (a), leaf weight (b), stem weight (c) and tops weight (d) were obtained under

2090 high emission scenario while tuber dry (e) and fresh (f) were produced under

2030 high emission scenario. It was also noticed that the tuber dry and tuber fresh

weight decreased in the future while the LAI values increased.

Furthermore, Figures 5.4 and 5.5 illustrate the potential and non-potential simulation

for Nacocolevu. For potential and non-potential simulation, the highest LAI (a), leaf

weight (b), stem weight (c) and tops weight (d) were obtained under 2090 high

emission scenario while tuber dry (e) and fresh (f) were attained under 2030 high

emission scenario.

107

Figure 5.0 shows the potential future climate simulations for Banisogosogo for LAI, leaf

weight, stem weight, tuber fresh weight, tops weight, tuber dry weight and tuber fresh

weight.

The above figure shows that the highest tuber yield (e and f) were obtained under 2030 A1B

and A2 emission scenario. On the other hand, 2030 A1B and A2 emission scenario, gave the

lowest value for LAI, leaf weight and stem weight (a, b, c, and d respectively).


c) Stem weight

e) Tuber dry weight f) Tuber fresh weight

d) Tops weight

108

Figure 5.1 shows the non-potential future climate simulation for Banisogosogo under A1B

and A2 emission scenario for LAI, leaf weight, stem weight, tuber fresh weight, tuber dry

weight and tops weight.

The above figure illustrates that highest tuber weight (e and f) were achieved under 2030 A2,

followed by 2030 A1B, 2055 A2 and 2050 A1B emission scenario.

a) LAI

f) Tuber fresh weight

b) Leaf weight

c) Stem d) Tops weight

e) Tuber dry weight

109

Figure 5.2 shows the Koronivia potential future climate simulations for LAI, leaf weight,


The above figure shows that tuber yield (e and f) is possible under 2030-2055 A1B and A2

emission scenario while the highest LAI was recorded under 2090 A2 emission scenario.


c) Stem weight d) Tops weight


110

Figure 5.3 shows the Koronivia non-potential future climate simulations for LAI, leaf

weight, stem weight, tuber fresh weight, tuber dry weight and tops weight.

The above figure shows that under future climate, tuber yield is possible under 2030-2090

A1B and A2 emission scenario.




111

Figure 5.4 shows the Nacocolevu potential future climate simulations for LAI, leaf weight,


The above figure shows that tuber yield is possible under 2030-2090 A1B and A2 emission

scenario.




112

Figure 5.5 shows the Nacocolevu non-potential future climate simulation for LAI, leaf

weight, stem weight, tuber fresh weight, tuber dry weight and tops weight.

The figure above shows that under future climate, tuber yield is possible under 2030-2090





113


The optimisation simulations were conducted for non-potential simulation for


fertiliser and optimum irrigation treatment and also planting depth (Table 5.4, Table

5.5 and Table 5.6).

The Banisogosogo optimisation treatment (Table 5.4) showed that under 2030 A1B

and A2 emission scenario, the optimum planting time was July and June

respectively. For A1B and A2 2055 emission scenario, the optimum planting time

was July. The optimum row spacing for 2030 A1B and A2 was 75 cm and 40 cm

respectively while the optimum row spacing for 2055 A1B and A2 was 75 cm and

30 cm respectively. Under 2030 A1B and A2 emission scenario and 2055 A1B

emission scenario, default irrigation gave the highest yield while for 2055 A2

emission scenario, 1.6 mm irrigation simulated the highest yield. The optimisation

treatment for fertiliser for 2030 A1B and A2 emission scenario indicated that

application of 300 kg/ha fertiliser at 0-4 cm for both emission scenarios produced the

highest yield but the optimisation treatment for fertiliser for 2055 A1B and A2

emission scenario was the default run. The optimum fertiliser and the optimum

irrigation application indicated that the yield was optimised for 2030 A1B emission

scenario while the yield under 2030 A2, 2055 A1B and A2 emission scenario the

yield was not optimised. Under 2030 A1B and 2055 A2, the optimum planting depth

was 1.5 cm while under 2030 A2 and 2055 A1B the optimum planting depth was 4

cm.

Table 5.4 shows the optimisation treatments at Banisogosogo.

Optimum Treatment, Yield (kg/ha)

Banisogosogo 2030 A1B Emission 2030 A2 Emission 2055 A1B

Emission

2055 A2

Emission

Planting Time July, 2116 June, 2571 July, 757 July, 1423

Row Spacing 75 cm, 1261 40 cm, 2803 75 cm, 148 30 cm, 302

Irrigation Application Default run, 1261 Default run, 2713 Default run,

148

1.6 mm, 301

Fertiliser Application 300 kg/ha fertiliser at 0-

4 cm, 1410

300 kg/ha fertiliser at 0-

4 cm, 2265

Default value,

1166

Default value,

300

Optimum Irrigation and

Fertiliser Application

1475 2221 82 300

Planting Depth 1.5 cm, 1261 4 cm, 1196 4 cm 1196 1.5 cm, 300

114

The Koronivia optimisation treatment (Table 5.5) showed that under 2030 and 2055

AIB and A2 emission scenario, the optimum planting time was August while under

2090 A1B and A2 emission scenario, the optimum planting time changed to July.

The optimum row spacing for 2030 A1B and A2 was 100 cm and 40 cm respectively

while 2055 A1B and A2 emission scenario, the optimum row spacing was 100 cm

and 80 cm respectively. A row spacing of 30 cm was required to produce the highest

yield under 2090 A1B and A2 emission scenario. Under 2030 A1B and A2 emission

scenario, 16.0 mm irrigation per plant gave the highest yield. Under 2055 A1B and

A2 emission scenario, 6.4 mm irrigation and default irrigation produced the highest

yield respectively whereas under 2090 A1B and A2 emission scenario, 1.6 mm per

plant irrigation attained the highest yield. The fertiliser optimisation treatment for

2030 A1B and A2 emission scenario indicated that application of 300 kg/ha fertiliser

at 0-4 cm for both emission scenarios simulated the highest yield. The optimisation

treatment for fertiliser for 2055 A1B and A2 emission scenario was application of 60

kg/ha fertiliser at 0-4 cm and default application of fertiliser respectively while

under 2090 A1B and A2 emission scenario, 60 kg/ha fertiliser produced the highest

yield. The optimum fertiliser and the optimum irrigation application indicated that

the yield was optimised for 2090 emission scenario. Under all emission scenarios,

the optimum planting depth was 1.5 cm.

Table 5.5 shows the optimisation treatments at Koronivia.

Optimum Treatment, Yield (kg/ha) Koronivia 2030 A1B

Emission 2030 A2 Emission

2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

2090 A2 Emission

Planting Time August, 6692

August,7827 August, 3590

August, 5961

July, 1911 July, 1570

Row Spacing 100 cm, 2451

40 cm, 2803

100 cm, 1687

80 cm, 2331

30 cm, 667 30 cm, 291

Irrigation Application

16.0 mm, 3885

16.0 mm, 4911

6.4 mm, 2005

Default run, 2990

1.6 mm, 721

1.6 mm, 251


300 kg/ha fertiliser at 0, 2, 4 cm, 2914



Default run, 2990

60 kg/ha fertiliser at 8 and 10 cm, 773

60 kg/ha fertiliser at 0-6 cm and 8 cm, 351

Optimum Irrigation and Fertiliser Application

3449 4342 1932 2100 779 401

Planting Depth 1.5 cm, 2369

1.5 cm, 2713

1.5 cm, 1578

1.5 cm, 300 1.5 cm, 617 1.5 cm, 233

115

The Nacocolevu optimisation treatment (Table 5.6) shows that under 2030 AIB and

A2 emission scenario, the optimum planting time was August while under 2055-

2090 A1B and A2 emission scenario, June was the optimum planting time. The

optimum row spacing for 2030 A1B and A2 was 75 cm and 100 cm respectively and

optimum row spacing for 2055 A1B and A2 was 100 cm and 40 cm respectively.

The optimum row spacing for 2090 A1B and A2 was 30 cm. Under 2030 A1B and

A2 emission scenario, 6.4 mm irrigation per plant and default irrigation simulated the

highest yield respectively. Under 2055 A1B and A2 emission scenario, 8.0 mm

irrigation and 16.0 mm irrigation gave the highest yield respectively. Under 2090

A1B and A2 emission scenario, 9.6 mm irrigation per plant and 16.0 mm irrigation

per plant produced the highest yield. The optimisation treatment for fertiliser for


at 8 cm and default fertiliser application attained the highest yield respectively. The

optimisation treatment for fertiliser for 2055-2090 A1B and A2 emission scenario

was application of 60 kg/ha fertiliser. The optimum fertiliser and the optimum

irrigation application indicated that the yield was only optimised for 2030 A1B

emission scenario. Under 2030-2090 A1B and A2, the optimum planting depth was

1.5 cm.

Table 5.6 shows the optimisation treatments at Nacocolevu.

Optimum Treatment, Yield (kg/ha) Nacocolevu 2030 A1B


2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

2090 A2 Emission

Planting Time August, 8178

August, 8522

June, 8141 June, 9385 June, 4729 June, 3634

Row Spacing 75 cm, 4908

100 cm, 6417

100 cm, 2182

40 cm, 3145

30 cm, 604 30 cm, 315


6.4 mm, 5125

Default run, 5521

8 mm, 2458 16 mm, 3131

9.6 mm, 680

16 mm, 320


300 kg/ha fertiliser at 8 cm , 4997

Default run, 5611

60 kg/ha fertiliser at 6 cm, 3397


60 kg/ha fertiliser at 0-4 cm, 1144

60 kg/ha fertiliser at 0-4 cm cm, 749


5818 5549 3325 3808 974 612

Planting Depth 1.5 cm, 4908

1.5 cm, 5521

1.5 cm, 1993

1.5 cm, 2949

1.5 cm, 523 1.5 cm, 314

116

5.3 Discussion

5.3.1 Potential simulations

The Banisogosogo Desiree potential tuber yield was only possible under 2030

medium and 2030 high emission scenario. The Banisogosogo potential simulations

showed that under 2030 medium emission scenario the emergence day was 10th of

July (8 days after planting), tuber initiation day was 28th of August (57 days after

planting) with tuber dry yield of 1698 kg/ha, tuber fresh yield of 8.49 t/ha and LAI

was 10.79. Under 2030 high emission scenario, the emergence day was 10th of July,

tuber initiation day was 25th of August (54 days after planting) with tuber dry weight

of 2395 kg/ha, tuber fresh weight of 11.98 t/ha and LAI of 9.8. In other words, the

tuber initiation occurred three days earlier in 2030 high emission scenario compared

to 2030 medium emission scenario. This was because the projected temperature for

high emission scenario (1.2 oC) was less than the projected medium emission

scenario (1 oC). High temperatures caused an increase in abnormal tuber behavior,

chain tuber formation, secondary growth, growth cracks, heat sprouts, heat necrosis

and translucent ends. There is also reduction in starch, sucrose and dry matter content

of the tubers (Struik, 2007b). Above 30 oC, the net assimilation of potatoes fell below

zero and yield reduction occurs. There is reduction in partitioning of assimilation in

the tubers and improved partitioning at the haulm at high temperature (Levy and

Veilleux, 2007). It was also noted that the tuber initiation day and the maximum LAI

decreased from 2030 medium emission scenario to 2030 high emission scenario.

Under 2055 and 2090 medium and high emission scenario there was no tuber yield

but the LAI was higher under high emission scenario (Table 5.1). One reason for no

tuber yield can be due to the high average maximum and minimum temperatures

faced during the growing season which inhibited tuber initiation stage. As shown in

Figure 5.0, under 2030 medium and high emission scenario, the LAI and leaf weight

increased in weight from the planting date (2nd of July) until 25th of August. After

25th of August, the LAI and leaf weight started to decrease until the harvest date

while the tuber weight started to increase from 25th of August to the harvest date.

Under other emission scenarios (2055 and 2090 medium and high emission), there

was steady increase in LAI, leaf weight, stem weight and tops weight while there

was no tuber weight. Positive effects of elevated carbon dioxide has been observed

on potato yield (Schapendonk et al., 1995; Miglietta et al., 1998; Rosenzweig and

117

Hillel, 1998; Temmerman et al., 2000; Craigon et al., 2002; Te mmerman et al.,

2007; Hogy and Fangmeier, 2009a) but these increase in yields are highly species

dependent (Kimball, 1983). Tuber numbers also increased (P< 0.05) under elevated

carbon dioxide, particularly in the smaller size categories, that is, < 35 mm category

(Lawson et al., 2001b; K.Persson et al., 2003). The tuber quality also increased under

elevated carbon dioxide, that is, increase in the dry matter and starch concentration

(Donnelly et al., 2001; Vorne et al., 2002; Bucher and Kossmann, 2007; Te

mmerman et al., 2007), the concentrations of fructose, glucose and total reducing

carbohydrates were also positively related to carbon dioxide levels (Hogy and

Fangmeier, 2009b). One hypothesis that can be drawn from potatoes is that increased

sucrose availability under elevated carbon dioxide feeds forward to the production

and activity of the sink organs whereby resulting in increased yield (Te mmerman et

al., 2002). However, the concentration of nitrogen decreased (Fangmeier et al.,

2002). Doubling of carbon dioxide also caused earlier flowering, increased leaf area,

specific leaf weight, canopy density, plant height and accelerated crop maturation

rate (Krupa and Kickert, 1993). Conversely, elevated carbon dioxide had a negative

impact on tuber quality causing tuber malformation to increase by 62.8 % (Hogy and

Fangmeier, 2009b).

Under Koronivia potential simulations, 2030 medium emission scenario showed that

the emergence day was 9th of July (7 days after planting), tuber initiation day was

13th of August (42 days after planting) with tuber dry weight of 5909 kg/ha, tuber

fresh weight was 29.54 t/ha and LAI of 12.68. Under 2030 high emission scenario,

the emergence day was 9th of July, tuber initiation day was 11th of August (40 days

after planting) with tuber dry weight and fresh weight of 6935 kg/ha and 34.68 t/ha

and LAI of 11.99. The emission scenario of 2030 high emission showed a greater

yield as compared to 2030 medium emission scenario due to difference in

temperature increase and precipitation levels. It can also be said that for 2030 high

emission scenario, the tuber initiation day was earlier than 2030 medium emission

scenario. However, the LAI value was higher under 2030 medium emission scenario.

Under 2055 medium emission scenario it was shown that the emergence day was 10th

of July, tuber initiation day was 20th of August (49 days after planting) with tuber dry

weight and fresh weight of 2615 kg/ha and 13.08 t/ha and LAI of 16.83. The results

of 2055 high emission scenario indicated that the emergence day was 10th of July,

118

tuber initiation day was 18th of August (47 days after planting) with a tuber dry yield

of 4246 kg/ha, tuber fresh weight of 21.33 t/ha and LAI of 20.14. Under 2090

medium and high emission scenario, the emergence date was 12th of July and 13th of

July respectively. However, there were no tuber yield but the value of LAI was very

high (Table 5.2). This can be due to the high average maximum and minimum

temperatures experienced during the growing season which inhibited tuber initiation

development stage. Figure 5.2 indicated that under 2030 medium and high emission

scenario, the LAI, leaf weight and tops weight increased from the day of planting.

However, there was a decline in weight from 10th of August and 12th of August under

2030 medium and 2030 high emission scenario respectively. Under these emission

scenarios, tuber weight started to increase around the same time. For 2055 medium

and high emission scenario, the LAI, leaf weight and tops weight increased from the

day of planting until from 18th of August and 17th of August under 2030 medium and

2030 high emission scenario respectively. Under these emission scenarios, tuber

weight started to increase around the same time. Medium and high emission of 2090

showed gradual increase in LAI, leaf weight, stem weight and tops weight over time

with no tuber yield. Plants that are exposed to higher temperatures have smaller,

shorter, narrower and abundant leaves (Ewing and Struik, 1992). Warm temperatures

promotes vegetative growth while cool temperatures promotes tuber growth (Ewing,

1981c; Khedher and Ewing, 1985).

Nacocolevu produced yields in both medium and high emission scenarios from 2030

to 2090. The highest average temperature was experienced under 2090 medium and

high emission scenario which was 21.8 oC and 22.1 oC respectively. The optimum

temperature for potato production is 22 ºC (Burton, 1981). Table 5.3 indicated that

under 2030 medium emission scenario, the emergence day was 10th of July, tuber

initiation day was 12th of August (41 days after planting), the tuber dry weight was

6584 kg/ha, tuber fresh weight of 32.92 t/ha and LAI was 6.87. The results from

2030 high emission scenario indicated that the emergence day was 10th of July,

initiation of tuber was 11th of August (40 days after planting) with tuber dry weight

of 7370 kg/ha, tuber fresh weight was 36.85 t/ha and LAI of 6.64. The LAI also

decreased from 2030 medium emission to 2030 high emission. Under 2055 medium

emission scenario, the emergence day was 11th of July, tuber initiation day was 18th

of August (47 days after planting) with tuber dry weight and fresh weight was 4093

119

kg/ha and 20.46 t/ha and LAI of 7.81. The high emission of 2055 indicates that the

emergence day was 10th of July, tuber initiation day was 15th of August (44 days

after planting) with tuber dry weight of 6187 kg/ha, tuber fresh weight of 30.93 t/ha

with LAI of 8.34. The tuber dry yield decreased from year 2030 to 2055. The LAI

increased from the emission scenario of 2030 to 2055. Similar results are seen as LAI

increased from 2055 medium emission to 2055 high emission scenario. Finally, the

simulation for 2090 medium scenario indicated that the emergence day was 12th of

July, 3rd of September (63 days after planting) was noted for tuber initiation day.

Under this emission scenario, the tuber dry yield was 586 kg/ha with tuber fresh

weight of 2.93 t/ha and LAI of 12.07. Under 2090 high emission scenario it was seen

that the emergence day was 13th of July, tuber initiation day was 9th of September (63

days after planting) with tuber dry weight of 93 kg/ha, tuber fresh weight of 0.46 t/ha

and LAI of 15.64. Figure 5.4 showed that under 2030 and 2055 medium and high

emission scenario, the LAI, leaf weight, stem weight and tops weight increased from

the day of planting until 31st of August where a decline in weight was noticed. While

the tuber fresh weight started to increase from the 11th of August, 13th of August, 16th

of August, 19th of August and 4th of September for 2030 high, 2030 medium, 2055

high, 2055 medium and 2090 medium emission scenario respectively.

5.3.2 Non-potential simulations

Unlike Banisogosogo Desiree potential simulation which gave tuber yield only under

2030 emission scenario, Banisogosogo Desiree non-potential simulation showed that

tuber yield was sustained under 2030 medium and high emission scenario and 2055

medium and high emission scenario. The growth, development and yield of potatoes

are guided by factors such as high soil and high air temperature, plant density, water

stress, availability of nutrients and the utilisation of solar radiation (Allen, 1978;

Ewing, 1981b). The results from 2030 medium emission indicated that the

emergence day was 11th of July (9 days after planting), tuber initiation day was 26th

of August (55 days after planting) (Table 5.1). The tuber dry weight was 949 kg/ha,

tuber fresh weight of 4.75 t/ha and LAI of 6.83. The non-potential simulation shows

that the crop underwent water stress on 20th September (80 days after planting) and

nitrogen stress on 26th August (55 days after planting) and 20th September (refer to

Appendix 3.A for crop and soil status and environmental and stress factors for

Banisogosogo, Koronivia and Nacocolevu). Under 2030 high emission scenario, the

120

emergence day was 10th of July, tuber initiation day was 22nd of August (51 days

after planting) with tuber dry weight of 1459 kg/ha, tuber fresh weight of 7.0 t/ha and

LAI of 7.0. The crop faced water stress on 20th of September (80 days after planting)

and nitrogen stress on 22nd August (51 days after planting) and 20th of September.

From 2030 medium emission to 2030 high emission scenario, the tuber initiation day

decreased from 55 days after planting to 51 days. It can also be said that the tuber dry

weight and tuber fresh weight increased from 2030 medium emission to 2030 high

emission scenario. Furthermore, under 2055 medium emission scenario it was seen

that the emergence day was 12th of July (10 days after planting), tuber initiation day

was 12th of September (72 days after planting) with tuber dry weight of 55 kg/ha,

tuber fresh weight of 0.27 t/ha and LAI of 7.99. Nitrogen and water stress was

experienced on 12th September (72 days after planting) and 20th September. The LAI

maximum has also increased to 7.99 as compared to 2030 emission scenario. It was

also shown that under 2055 high emission scenario the tuber initiation day was 66

days after planting with a tuber dry weight of 301 kg/ha, tuber fresh weight of 1.5

t/ha and LAI of 8.59. The crop also experienced water and nitrogen stress on 6th

September (66 days after planting) and 20th September. One reason for the higher

yield obtained in 2055 high emission scenario was that under high emission scenario,

the annual increase in surface air temperature was 1.7 oC whereas the annual increase

in surface air temperature of 2055 medium emission scenario was 1.9 oC. Higher

temperatures cause yield reduction as high as 50 % in most vulnerable areas such as

the tropical belt (Bindi, 2008). Higher temperatures reduced dry matter content and

produced tubers with pale skin colour (Haverkort, 1990). There was no tuber

initiation under 2090 medium and high emission scenario. Numerous processes of

tuber formation are affected by high temperatures and these include sprout

development, tuber initiation, stolon formation (Midmore, 1984b; Struik, 2007b),

duration of tuber growth, tuber number, average tuber weight and tuber yield (Struik,

2007b), yield of potatoes (Ewing and Struik, 1992), photosynthesis (Ewing, 1997)

and respiration, partitioning of assimilates (Levy and Veilleux, 2007) and tuber

quality (Ewing and Struik, 1992) and reduction in dry matter content (Thornton et

al., 1996). Reduction in tuber yield also occurs when high soil temperatures are

combined with high ambient air. However, the values of LAI were very high. High

temperature also affects leaf dynamics such as final number of leaves (Firman et al.,

1991) and shape of leaves (Ewing and Struik, 1992; Vos and Haverkort, 2007b).

121

Above 20 oC, every 5 oC increase in leaf temperature, had a reduction of

approximately 25 % in the photosynthesis rate (Levy and Veilleux, 2007). High

temperature also has an effect on the number of stems (Struik, 2007b). Branching of

stems is enhanced at higher temperatures (Struik et al., 1989). High temperatures can

even cause death of stem before production takes place. The optimum temperature

for stem elongation is 25 ºC (Borah and Milthorpe, 1992). Figure 5.1 shows that for

2030 medium and high emission scenario, the LAI, leaf weight and tops weight

increased from the day of planting until 31st of August where a decline in weight was

noticed. The LAI, leaf weight and tops weight again increased from the 4th of

September and decreased around the 18th of September. The tuber fresh weight for

these scenarios started to increase from 2nd of August and 27th of August for 2030

medium and 2030 high respectively. Under 2055 medium and high emission

scenario, the LAI, leaf weight and tops weight increased from the day of planting

until 10th of September and 5th of September respectively. An increase in tuber

weight was noticed around 14th of September and 8th of September for 2055 medium

and high emission scenario. Under 2090 medium and high emission the LAI, leaf

weight, stem weight and tops increased gradually with no tuber yield.

The Koronivia non potential simulations indicated that it was possible to obtain yield

under both medium and high emission scenarios from 2030 to 2090 (Table 5.2).

Under non-potential 2030 medium emission scenario it was shown that the

emergence day was 9th of July, tuber initiation day was 8th of August (37 days after

planting) with dry tuber yield of 2369 kg/ha, fresh weight of 11.84 t/ha and LAI of

4.03. Water and nitrogen stress were evident on 8th August (37 days after planting)

and 20th of September (80 days after planting). Under 2030 high emissions, the

results indicated that the emergence day was 9th of July, tuber initiation day was 7th

of August (36 days after planting) with tuber dry weight of 2713 kg/ha, fresh weight

of 13.56 t/ha and LAI of 3.98. The crop underwent water and nitrogen stress on 7th

August (36 days after planting) and 20th September. Under medium emission

scenario of 2055, the emergence day was 10th of July, tuber initiation day was 16th of

August (45 days after planting). The tuber dry weight was 1578 kg/ha, tuber fresh

weight of 7.89 t/ha and LAI of 4.7. The crop experienced nitrogen and water stress

on 16th August (45 days after planting) and 20th September. The results under 2055

high emission indicated that the emergence day was 10th of July, tuber initiation day

122

was 14th of August (43 days after planting) with tuber dry weight was 2290 kg/ha,

tuber dry weight of 11.45 t/ha and LAI of 4.81. The crop faced nitrogen and water

stress on 14th August (43 days after planting) and 20th September. Furthermore,

under medium emission of 2090 it was noted that the emergence day was 12th of

July, tuber initiation day was 25th of August (54 days after planting) with tuber dry

weight of 617 kg/ha, tuber fresh weight of 3.08 t/ha and LAI of 6.04. The crop

suffered nitrogen and water stress on 25th August (85 days after planting) and 20th

September. Under 2090 high emission scenario, the emergence day was 13th of July,

tuber initiation day was 2nd of September (62 days after planting). The simulated

tuber dry weight was 233 kg/ha with tuber fresh weight of 1.16 t/ha and LAI of 6.92.

Water and nitrogen stress were seen on 2nd September (62 days after planting) and

20th September. Figure 5.3 shows that LAI, leaf weight and stem weight were the

highest under 2090 medium emission scenario and lowest under 2030 high emission

scenario. The final dry matter concentration depends on the cultivar, length of

growing seasons, water availability, solar intensity and the average temperature

during the growing season (Haverkort and Verhagen, 2008). Figure 5.3 showed that

the LAI, stem weight, leaf weight and tops weight increased from the day of planting

until around 9th of August. After 9th of August it experienced a decline in weight 28th

of August, after which it again increased in weight around 5th of September and then

decreased in weight around 5th of September. The figure also showed that the tuber

weight increased from 22nd of August and after 28th of August there was a sharp

increase in tuber weight until harvest. Overall, the tuber fresh and tuber dry weight

was highest under 2030 high emission scenario, followed by 2030 medium emission

scenario, 2055 high emission, 2055 medium emission, 2090 medium emission and

the lowest under 2090 high emission scenario. High temperatures significantly affect

potato growth and development (Food and Agriculture Organization of the United

Nations, 2008). Projections from simulation models indicate that higher temperatures

will reduce agronomic outputs. Farmers need to be mindful of local climate and their

ability to mitigate and adapt to these changing conditions (Mendelsohn and Dinar,

1999). Another significant finding from this study was that under potential

simulations production of potatoes is only possible from 2030-2055 under both

medium and high emission scenario while under non-potential simulation,

production of potato was possible from 2030 to 2090 under both medium and high

emission scenario. It can be said that the tuber initiation day was the earliest under

123

2030 high emission followed by 2030 medium emission, 2055 high emission, 2055

medium emission, 2090 medium emission and finally 2090 high emission. It was

also noticed that LAI increased from 2030-2090. Under elevated carbon dioxide

there is an increase in leaf number due to shift in temporal crop development

(Lawson et al., 2001c).

Nacocolevu non potential simulations show that it was possible to obtain yield under

both medium and high emission scenarios from 2030 to 2090 (Table 5.3). One reason

for this can be that there was no evidence of water stress in Nacocolevu during the

potato growing season. Under non-potential 2030 medium emission scenario it was

shown that emergence day was 10th of July, tuber initiation day was 11th of August

(40 days after planting) with dry tuber yield of 4908 kg/ha, fresh weight of 24.54 t/ha

and LAI of 3.55. Nitrogen stress was evident on 11th August (40 days after planting)

and 20th of September (80 days after planting). Under 2030 high emissions, the

results indicated that the emergence day was 10th of July, tuber initiation day was

10th of August (39 days after planting) with tuber dry weight of 5521 kg/ha, tuber

fresh weight of 27.61 t/ha and LAI of 3.19. The crop underwent nitrogen stress on

10th August (39 days after planting) and 20th September. Under medium emission

scenario of 2055, the emergence day was 11th of July, tuber initiation day was 16th of

August (45 days after planting). The tuber dry weight was 1993 kg/ha, tuber fresh

weight of 9.96 t/ha and LAI of 5.0. The plant underwent nitrogen stress under 16th

August (45 days after planting) and 20th September. For 2055 high emission it was

seen that the emergence day was 11th of July, tuber initiation day was 15th of August

(44 days after planting). The tuber dry weight was 2949 kg/ha, tuber fresh weight of

14.75 t/ha and LAI of 4.69. Nitrogen stress was evident on the plant on 15th of

August (44 days after planting) and 20th of September. Under medium emission of

2090, it was noted that the emergence day was 13th of July, the tuber initiation day

was 1st of September (61 days after planting) with tuber dry weight of 523 kg/ha and

tuber fresh weight of 2.62 t/ha and LAI of 5.68. Nitrogen stress was seen on 1st

September (61 days after planting) and 20th September. Under 2090 high emission

scenario the emergence day was 13th of July, tuber initiation day was 3rd of

September (63 days after planting). The simulated tuber dry weight was 314 kg/ha,

tuber fresh weight of 314 t/ha and LAI of 5.74. Nitrogen stress was seen on 3rd of

September (63 days after planting) and 20th September. Figure 5.5 indicated that the

124

LAI, leaf weight and stem weight increased from the planting date until 19th of

August of 2030 medium and high emission scenario. The tuber weight under these

emission scenarios started to gradually increase from 11th of August until the harvest

day. For 2055 medium and high emission scenario, the LAI, leaf weight and stem

weight increased from the planting date until 2nd September and 19th of August

respectively while the tuber weight under these emission scenarios started to

gradually increase from 17th of August until the harvest day. Under 2090 medium

emission scenario, the LAI, leaf weight and stem weight increased from the planting

date until 2nd September after while there was a decline in weight while the tuber

weight under this emission scenario started to gradually increase from 2nd of

September until the harvest day. The tuber fresh and tuber dry weight was the highest

under 2030 high emission scenario, followed by 2030 medium emission scenario,

2055 high emission scenario, 2055 medium scenario, 2090 high and finally 2090

medium emission scenario. Wild relatives are important gene pool for breeding new

varieties. The increasing temperatures put additional pressure on potatoes wild

relatives. By 2055, due to increase in temperature, 16-22% of all wild potato species

will be threatened with extinction (InfoResources Focus, 2008). Also, an increase in

carbon dioxide levels reduces the growing season leading to earlier senescence and

decline in final biomass accumulation (Makino and Mae, 1999; Lawson et al., 2001a;

Te mmerman et al., 2007). As an isolated factor, an increase in atmospheric

concentration of carbon dioxide, will have a positive effect on crop yields. However,

the combined effects of carbon dioxide and temperature effects are not yet

appropriately examined to make statements regarding their effects on potato tuber

development (Reddy and Hodges, 2000; Food and Agriculture Organization of the

United Nations, 2008).


Farmers in the tropics can harvest potatoes within 50 days of planting, which is

almost one-third of the time it takes in the colder climates (Changchui, 2008). For

Banisogosogo simulation (Table 5.4), the highest yield was obtained under July for

2030 medium emission scenario. This was because the month of July had low

average minimum and maximum temperature with relatively high solar radiation.

Under 2030 high emission scenario, the month of June gave the optimum yield. This

125

was because the month of June had low average maximum and minimum

temperatures. Under 2055 medium and high scenario, the optimum yield was

obtained in July. The month of July had low average minimum and maximum

temperature and high solar radiation. It was also seen that the months of January-

March and November-December did not have tuber yield as the high average

maximum and minimum temperatures in these months inhibited development phase

of tuber initiation (Appendix 3.2A Table 3.56A). Under current climate conditions

(Chapter 4), the optimum month for planting potatoes was May. Under Koronivia

simulations (Table 5.5), for 2030 and 2055 medium and high emission scenario,

August gave the highest yield. The month of August gave the highest yield under

these emission scenarios as it had the lowest average maximum and minimum

temperature with relatively high solar radiation. The results also indicated that under

2030 medium emission scenario January and February gave no tuber and under 2055

medium and high emission scenario, January-March gave no tuber yield. This was

because these months received high average maximum and minimum temperatures

inhibited development phase of tuber initiation. For 2090 medium and high emission

scenario, the optimum planting month switched to July. This was because under

2090 medium and high emission scenario, the month of July received the average

low maximum and minimum temperature and relatively high solar radiation. Under

2090 medium and high emission January-April and November and December gave

no tuber yield (Appendix 3.2A, Table 3.68A). On the other hand, under current

climate conditions August was the optimum month for potato cultivation.

Furthermore, for Nacocolevu simulations, from 2030 medium and high emission

scenario, August gave the highest yield. The month of August received high average

solar radiation and high average rainfall as compared to earlier months and had lower

average maximum and minimum temperatures as compared to months later than

August. For 2055 and 2090 medium and high emission scenario, the optimum

planting month changed to June (Table 5.6) as it had average low minimum

temperatures as compared to other months. It was also noticed that under the planting

time optimisation treatment, December gave no tuber as there was no tuber initiation

(Appendix 3.2A, Table 3.84A). Under current climate condition, August was

considered the optimum month for potato production.

126

Row spacing also had an impact on tuber yield. As indicated in Table 5.4

Banisogosogo 2030 medium emission scenario results of row spacing indicated that

to maximise potato yield, a row spacing of 75 cm was suitable. The results 2030 high

emission scenario of row spacing indicated that a row spacing of 40 cm produced

the highest yield of 2803 kg/ha. Under 2055 medium and high emission scenario, it

was noted that a row spacing of 75 cm and 30 cm gave the highest yield respectively

(Appendix 3A, Table 3.57). Table 5.5 shows that under Koronivia 2030 medium

emission scenario, the simulation of row spacing indicated that maximum yield can

be achieved at a row spacing of 100 cm which simulated a yield of 2451 kg/ha.

Under 2030 high emission scenario, row spacing of 40 cm gave the highest tuber

yield. For 2055 medium emission scenario, row spacing of 100 cm produced the

highest yield. The results of 2055 high emission scenario indicated that a row

spacing of 80 cm produced the highest yield at 2331 kg/ha. For 2090 medium and

high emission scenario, row spacing of 30 cm gave the highest yield (Appendix 3A,

Table 3.69). For Nacocolevu 2030 medium emission scenario, the simulation of row

spacing indicated that maximum yield can be achieved at a row spacing of 75 cm

which gave a yield of 4908 kg/ha. Under 2030 medium emission, the simulation of

row spacing has shown that increasing the row spacing increased the tuber yield.

Hence, at row spacing of 100 cm, the highest yield tuber yield was obtained at 6417

kg/ha. For Nacocolevu 2055 medium emission scenario a row spacing of 100 cm

gave the highest yield. For Nacocolevu 2055 high emission scenario simulation, the

results of row spacing indicated that a row spacing of 40 cm gave the highest yield at

3145 kg/ha. For Nacocolevu 2090 medium emission scenario simulation, the results

of row spacing indicated that row spacing of 30 cm gave the highest yield under

2090 medium emission. Increasing the row spacing decreased the tuber yield. Under

2090 high emission scenario, the row spacing indicated that under 30 cm and 50 cm

of row spacing gave the highest yield (315 kg/ha). The lowest yield was achieved

under 100 cm (Table 5.6 and Appendix 3A, Table 3.85).

Likewise, optimisation simulation was also conducted for fertiliser treatment. The

results for Banisogosogo 2030 medium and high emission scenario fertiliser

treatment indicated that application of 300 kg/ha fertiliser at 0-4 cm simulated the

highest yield (1410 kg/ha and 2265 respectively). Under 2055 medium emission

results for fertiliser treatment indicated that default run gave the highest yield at 148

127

kg/ha while under 2055 high emission scenario, 60 kg/ha at 0-4 cm gave the highest

yield at 304 kg/ha (Appendix 3.2A, Tables 3.60A, 3.62A, 3.64A and 3.66A). For

Koronivia simulations, under 2030 medium and high emission scenario and 2055

medium emission scenario, fertiliser optimisation results indicated that application of

300 kg/ha fertiliser at depth of 0 cm, 2 cm and 4 cm gave maximum yield. For 2055

high emission scenario, default run produced the highest yield at 2990 kg/ha. For

2090 medium emission scenario, application of 60 kg/ha fertiliser at 8 and 10 cm

simulated the highest yield at 773 kg/ha while under 2090 high emission scenario

fertiliser optimisation treatment indicated that maximum yield of tuber can be

obtained under 60 kg/ha of fertiliser treatment at 0-6 cm and 8 cm (Appendix 3.2A,

Tables 3.72A, 3.74A, 3.76A, 3.78A, 3.80A and 3.82A). For Nacocolevu 2030

medium, the fertiliser optimisation results indicated that application of 300 kg/ha

fertiliser at depth of 8 cm gave maximum yield of 4997 kg/ha. Under 2030 high

emission, the default run gave the highest yield. For Nacocolevu 2055 medium

emission scenario simulation, the optimum fertiliser treatment indicated that

application of 60 kg/ha of fertiliser at 6 cm gave maximum yield at 3397 kg/ha.

Under Nacocolevu 2055 high emission scenario simulation, the results of fertiliser

treatment indicated that application of 60 kg/ha fertiliser at 10 cm simulated the

highest yield at 3946 kg/ha. Finally Nacocolevu 2090 and high medium emission

scenario simulation, the fertiliser optimisation treatment indicated that maximum

yield of tuber can be obtained under 60 kg/ha of fertiliser treatment at 0 cm, 2 cm

and 4 cm with tuber yield of 1144 kg/ha and 749 kg/ha respectively (Table 5.6,

Appendix 3.2A, Tables 3.88A, 3.90A, 3.92A, 3.94A, 3.96A and 3.98A).

Furthermore, optimisation was also conducted for irrigation application. Table 5.5

indicates that for Banisogosogo irrigation optimisation, under 2030 medium and high

emission scenario and 2055 medium emission scenario, the default run gave the

highest yield while under 2055 high emission scenario, 1.6 mm of irrigation per plant

gave the highest yield at 301 kg/ha (also refer to Appendix 3.2A, Table 3.58A).

Under 2030 medium and 2030 high emission scenario for Koronivia, the results of

the irrigation optimisation in general suggested that an increase in irrigation levels

increased yield. Hence, the highest yield was obtained at 16 mm and the lowest yield

was obtained at 1.6 mm. Under 2055 medium emission scenario, the results indicated

that application of irrigation of 6.4 mm per plant produced the highest yield at 2005

128

kg/ha. The lowest yield was recorded at 1.6 mm irrigation with tuber yield at 1874

kg/ha. Under 2055 high emission scenario, the results from irrigation simulation

indicated that the highest yield was obtained under default run at 721 kg/ha. For 2090

medium and high emission scenario, the irrigation treatment indicated that increasing

the amount of irrigation decreased the tuber yield hence the highest yield was

obtained under irrigation application of 1.6 mm (Table 5.5 and also refer to

Appendix 3A, Table 3.70A). For all emission scenario at Nacocolevu, under 2030

medium emission scenario, 6.4 mm of irrigation gave the highest yield at 5125 kg/ha

while under 2030 high emission scenario the default run gave the highest yield.

Under 2055 medium emission scenario, application of 8 mm irrigation per plant gave

the highest yield at 2458 kg/ha while under 2055 high emission scenario, application

of 16.0 mm gave the highest yield ay 3131 kg/ha. For 2090 medium and high

emission scenario, application of 9.6 mm and 16.0 mm gave the highest yield

respectively.

In addition to these, optimisation treatment was also conducted for optimum

irrigation and optimum fertiliser simulations. It was noticed that under Banisogosogo

simulations, higher yields were only obtained under 2030 high emission scenario

(Table 5.5). For Koronivia simulations it was noticed that, only 2090 medium and

high emission scenario managed to optimise the tuber yield (Table 5.6). Under

Nacocolevu simulations, only 2030 medium and high emissions optimised the yield.

Finally, optimisation was conducted to investigate how planting depth affected tuber

yield. Banisogosogo 2030 medium emission scenario, planting depth of 1.5 m gave

the highest yield. It was also noted that increasing the planting depth decreased the

yield. Planting depth of 8 m and 10 cm failed to produce any yield. Under 2030 high

emission scenario, a depth of 4 cm simulated the highest yield of 1196 kg/ha.

Planting potatoes at a depth of 1.5 cm and 6 cm gave the same yield at 1166 kg/ha.

Planting of potatoes at a depth of 8 cm gave the lowest yield. Under 2055 medium

emission scenario, planting depth of 4 cm attained the highest yield of 1196 kg/ha.

Planting potatoes at a depth of 8 cm gave the lowest yield at 32 kg/ha. Under 2055

high emission scenario, the best depth to plant potatoes for maximum yield was at

1.5 cm which gave a yield of 300kg/ha. Increasing the planting depth decreased the

yield. The results also indicated that planting potatoes at 4 cm, 6 cm and 8 cm gave

no yield (also refer to Appendix 3A, Table 3.59A). For Koronivia and Nacocolevu

129

simulations it was noticed that under all emission scenarios, potato seeds should be

planted at 1.5 cm to achieve the highest yield. Increasing the depth of planting

decreased the yield of potatoes (also refer to Appendix 3A, Tables 3.71A and Table

3.87A).

5.4 Recommendations

Climate change will lead to changes in planting dates. It is recommended that

cultivation timing for Banisogosogo under 2030 medium emission scenario should

take place in July to obtain the highest yield while under 2030 high emission

scenario, the planting time should be June. Under 2055 medium and high emission

scenario the cultivation timing should be July. For Koronivia, it is recommended that

under 2030-2055 medium and high emission scenario, the planting time should be

August while under 2090 medium and high emission scenario, the planting time

should be July. The cultivation timing for Nacocolevu under 2030 medium and high

emission scenario should be August while under 2055-2090 medium and high

emission scenario, the planting time should be July. Planting time is projected to take

place earlier than anticipated as the thermal environment experienced by the crop

canopies would be more favourable (Singh et al., 2009; Knox et al., 2010). Changes

in the amount photoperiod due to changes in planting time has not been taken into

account (Hijmans, 2003a).

Irrigation levels also have to be adjusted with changes in the climate. In some

climatic conditions, such as, Banisogosogo required minimal irrigation such as that

of default irrigation and irrigation of 1.6 mm per plant. Koronivia 2030 medium and

high emission scenario required irrigation of 16.0 mm per plant while 2055 medium

and high emission scenario required 6.4 mm and default irrigation respectively.

Under 2090 medium and high emission scenario, 1.6 mm of irrigation per plant is

required. Nacocolevu 2030 medium and high emissions, required irrigation of 6.4

mm and default irrigation respectively. The emission scenarios of 2055 medium and

high required 8.0 mm and 16.0 mm of irrigation respectively. Under 2090 medium

and high emission scenario, irrigation application of 9.6 mm and 16.0 mm was

required respectively. If irrigation is not available in the future during the tuber

bulking phase, then the potato industry will not be viable and husbandry issue will

cease to be relevant (Holden and Brereton, 2006). Water management is also seen as

130

a key area that requires adaptation for climate change. A wide range of agricultural

water management practices and technologies are available to buffer the production

risks. Improving water efficiency requires an understanding of how plants balance

their water status in relation to evaporation demand under different soil conditions

(Stalham and Allison, 2011). Improved irrigation technology is one option to

improve water efficiency (Vos and Haverkort, 2007b). Other options include

effective use of rainfall, encouraging deeper rooting of crops, introducing drought

tolerant varieties and soil modifying soil structure to improve water retention (Knox

et al., 2010) and partial root zone drying (Shahnazari et al., 2008; Jovanovica et al.,

2010a). Partial root drying zone treatment resulted in the increase of nitrogen, starch

content and antioxidant activity in the potato tubers (Jovanovica et al., 2010b).

To maximise potato yield in future climatic scenario, fertiliser application levels and

application depth has to be adjusted to suit the change in climatic conditions. In some

climatic conditions, such as, Banisogosogo 2030 medium and high emission

scenario, high fertiliser levels are required while in other future climate scenario,

such as, Koronivia and Nacocolevu 2055-2090 medium and high emission, low

fertiliser levels are required. For potato, halving of nitrogen application rates could

be achieved only if irrigation water was used (Holden and Brereton, 2006).

Finally, it is also important to pay attention to planting depth under future climatic

scenario. In many cases, it was noted that either 1.5 cm or 2 cm were suitable for

planting potatoes. Increasing the planting depth (6 cm and 8 cm) gave no tuber yield.

5.5 Research limitations

One of the major challenges faced in this research was getting hold of the weather

data. Hence, different locations were run with different number of weather data, for

example, Banisogosogo simulation was conducted with weather data of 1960-2012,

Koronivia potential and non-potential simulations were run using weather data from

1961-2010 and Nacocolevu potential and non-potential simulations were conducted

using weather data from July 1972-2010. The different number of weather data used

in these simulations made it difficult to compare one site with another site. The

calibration of the DSSAT SUBSTOR Potato model was conducted only for

Banisogosogo. To obtain more accurate results, the model should have also been

vcalibrated for Koronivia and Nacocolevu.

131

Secondly, under all the simulations, it is assumed that the crop management practices

and the soil physical and chemical properties remained unchanged.

Also, it was noticed that the Plantgro.Output did not provide 2090 high emission

graphics. Hence, Excel was used to merge 2090 high emission scenario graphics into

all the graphs.

Finally, for future research, it is suggested that DSSAT v4.6 be used as it has a better

response to carbon dioxide concentration and better response to radiation use

efficiency.

132

Chapter 6 Simulating the performance of potential potato varieties under

current and future climates in Banisogosogo, Fiji Islands.

6.0 Introduction

There is a need to identify potato varieties in Fiji which may perform better under

current and future climatic conditions. Most of the countries in the Asia and Pacific

region are utilising potato varieties that were bred for growing conditions of Europe

or United States. These varieties are not adapted to the agroclimatic conditions of the

country where they are grown. Since many of these countries do not have suitable

research and development systems for native varieties, the cultivation of exotic

varieties cannot be stopped. However, these exotic varieties should be evaluated first

before they are recommended for large scale production (Singh, 2008).

Climate change poses a great challenge for breeders. To address these challenges,

development of new skills and qualification are required as potato breeding has

become more rooted in basic science and to be more receptive to public concerns

(Bonnel, 2008). Therefore, it is very important to monitor potato growth and forecast

its yield well before the harvest so that planners, decision makers and the

government can focus on its import and export and also redistribute food in times of

disasters. Forecasting the yields is also important in regions with climatic

uncertainties (Bala and Islam, 2008).

Climate change has created the need for adapting to several characteristics at once.

There is an urgent need to breed new potato varieties that are better adapted to

changing climatic conditions. The CIP is focusing on breeding short season potato

varieties for farmers to avoid adverse conditions such as hot, dry periods and heavy

rainfall. The CIP is evaluating the genetic resources of its collection of potatoes and

also of new varieties which are in process of being bred in regards to water and

temperature stress (InfoResources Focus, 2008). The adaptation of potato to a wide

range of environmental conditions is necessary to sustain variation and diversity in

potatoes (Bradshaw, 2007). If such adaptation measures are not taken, the potato

production, under the impact of climate change and global warming may decline by

3.16% in 2020 and 13.72% in 2050 (Singh et al., 2009).

133

This chapter will discuss what other varieties of potato can be cultivated in Fiji under

current climate and future climate conditions.

6.1 Methodology

6.1.1 Model simulation under current climate

6.1.1.1 Simulation site

The site for simulation for other varieties performance in Fiji was Banisogosogo,

Rakiraki (as in Chapters 3 and 4).

6.1.1.2 Data collection, treatments and importations

The weather data and soil data for the studied site used in these simulations was

similar as in Chapter 4. The crop management practices for planting time and

method, row spacing, nitrogen application and irrigation applications presented in

Chapter 4 were also used in these simulations. The baseline variety was Desiree

(Chapters 3, 4 and 5). The other cultivars used for these simulations are Sebago and

Russet Burbank (Genetic co-efficient in Table 6.0)

Table 6.0 shows the genetic co-efficient of Sebago and Russet Burbank.

Variety Name ECO

Number

G2

G3 G4 PD P2 TC

Desiree

tropics

IB0001 4000 25.0 0.2 0.9 0.4 18.0

Sebago IB0001 1500 22.5 0.20 0.90 0.10 19.0

Russet

Burbank

IB0001 2000 26.0 0.20 0.90 0.6 17.0

6.1.1.3 Reasons for cultivar selection

The cultivars selected for the simulation were Sebago and Russet Burbank. Sebago

variety was selected for this simulation because it has been planted in Fiji before and

the Ministry of Agriculture is thinking of reintroducing it in Fiji. Sebago variety is a

90 day crop which can yield up to 9 t/ha (Autar, 2009). Russet Burbank was selected

for this simulation as a possible recommendation for potato production. It is a late

maturity crop and requires 140 to 150 days season to produce maximum yields (The

Potato Association for America, 2009).

134

6.1.1.4 Optimisation treatments

The optimisation of crop managements, such as, planting date, row spacing, planting

depth and irrigation and fertiliser amount to maximise Sebago and Russet Burbank

growth and yield was conducted using Sensitivity Analysis Option in the DSSAT

SUBSTOR Potato model. The Banisogosogo optimisation simulations were

conducted with 2012 weather data except for planting time which was conducted

with 2009 weather data. The baseline variety is Desiree (Chapters 4 and 5).

6.1.2 Model simulation under future climatic conditions

6.1.2.1 Simulation site

The simulation site (Banisogosogo) presented in Chapter 4 was used for simulations

of future climate scenario impacts on potato production. The outputs of these

simulations are presented in the Results section of this Chapter.

6.1.2.2 Data collection, treatments and importations

The weather data for Banisogosogo and the PCCSP climate emission scenarios used

for the studied site in these simulations was the same as presented in Chapter 5. The

soil information for the studied site used in these simulations was the same as in

Replica 2 soil, Banisogosogo in Chapters 4 and 5. The crop management practices,

such as, planting time, nitrogen application and irrigation applications presented in

Chapter 4 were also used in these simulations.


The simulation for optimisation of crop managements (planting date, row spacing,

planting depth, irrigation application and fertiliser amount) to maximise Sebago and

Russet Burbank growth and yield was conducted using Sensitivity Analysis Option

in the DSSAT SUBSTOR Potato model. The Banisogosogo optimisation was

simulated using the 2012 weather data except for planting time which was conducted

with 2009 weather data.

6.1.3 Model output analysis


run the simulations for future climate impacts on potato growth and yield. All

summary outputs, graphical and initial analysis were done in DSSAT 4.5 Output.

135

Other statistical analysis was conducted in Microsoft 2007 Excel Spreadsheet. These

outputs are presented in the Results section.

6.2 Results

This section describes the simulation of Sebago and Russet Burbank potato variety

growth and yield under current climate conditions in Banisogosogo using the DSSAT

SUBSTOR Potato Model. The impact of climate variability, El Niño Southern

Oscillation (ENSO) state, was also studied on potato yield for the two potato

varieties. In addition to these, optimisation treatments were also conducted to

investigate which crop management practices, such as planting date, row spacing,

irrigation treatment, fertiliser treatment, optimum irrigation and optimum fertiliser

treatment and planting depth can optimise potato yield.

6.2.1 Current climate simulations

The current climate simulation was conducted for Banisogosogo for Sebago and

Russet Burbank varieties (Table 6.1). The weight for potential production was higher

as there was no water and nitrogen stress faced by the plant. Under non-potential

conditions, the plant faced water and nitrogen stress which reduced the above-ground

and below-ground biomass. The results show that Sebago variety had earlier tuber

initiation day and higher tuber dry weight and tuber fresh weight under potential

production while Desiree had higher tuber fresh weight and tuber dry weight under

non-potential conditions. Russet Burbank variety had higher LAI.

Table 6.1 Shows Sebago, Russet Burbank and Desiree- tropics potential and non-

potential simulation under current climate simulations for Banisogosogo.

Banisogosogo

Variety Sebago Russet Burbank Desiree

Variable Potential Non-

Potential

Potential Non-

Potential

Potential Non-

Potential


Tuber Dry Weight (kg\ha)

harvest

7777 2686 2885 1922 6651 4478

Tuber Fresh Weight (t\ha)

harvest

38.88 13.43 14.43 9.61 33.25 22.39

Leaf Area Index 6.14 0.45 8.44 5.88 8.29 1.7

136

Figures 6.0 and 6.1 also illustrate the same above information (Table 6.0), that is, the

potential and non-potential simulation for Banisogosogo for Sebago and Russet

Burbank respectively. These figures show that under potential simulation, soon after

the planting date, the LAI, aboveground biomass and belowground biomass

increased gradually with time while under non-potential simulations, LAI,

aboveground biomass and belowground biomass showed fluctuation in weight (due

to water and nitrogen stress). Under the simulation for Russet Burbank (Figure 6.1),

the figure shows that there was not much difference between the values obtained for

potential and non-potential simulation (as compared to Sebago variety).

Impact of precipitation, total water content of the soil, nitrate content of the soil and

nitrogen leached from the soil was also studied on non-potential tuber dry weight and

LAI for Sebago (Figure 6.2) and Russet Burbank (Figure 6.3). Figure 6.2 shows the

impact of precipitation ( mm), total water ( mm) content of the soil, nitrate content of

the soil (kg/ha) and nitrogen leached (kg/ha) from the soil on LAI (a) and tuber dry

weight (b). The LAI (a) achieved its maximum (0.4) on 25th of August and declined

afterwards. The tuber dry weight (b) increased gradually from 2nd of August until the

harvest day. The graph also shows the trend for total nitrogen and water in the soil.

Figure 6.3 indicates the impact of precipitation ( mm), total water ( mm) content of

the soil, nitrate content of the soil (kg/ha) and nitrogen leached (kg/ha) from the soil

on LAI (a) and tuber dry weight (b). The LAI increased gradually with time until the

30th of August after which it experienced a fluctuation in LAI value. This can be due

to water stress (decrease in extractable water) and nitrogen stress (decrease in NO3

level) faced by the plant. The tuber dry weight increased steadily with time after 22nd

of August until the harvest day.

137

Figure 6.0 shows Banisogosogo potential and non-potential simulations for Sebago for LAI,

leaf weight, stem weight, tuber dry and fresh weight and tops weight.

138

Figure 6.1 shows the Banisogosogo potential and non-potential simulations for Russet

Burbank for LAI, leaf weight, stem weight, tuber dry and fresh weight and tops weight.

139


potential tuber dry yield and LAI for Sebago in Banisogosogo over the growing season.

a) LAI

b) Tuber dry weight

140


potential tuber dry yield and LAI for Russet Burbank in Banisogosogo over the growing

season.

a) LAI

b) Tuber dry weight

141


The optimisation simulations were conducted to investigate which crop management

option maximised tuber yield. Optimisation simulations were conducted for planting

time, row spacing, irrigation treatment, fertiliser treatment, optimum fertiliser and

optimum irrigation treatment and also planting depth.

Table 6.2 shows the optimisation for planting time for Sebago and Russet Burbank.

The highest yield for Sebago was obtained in July while the highest yield for Russet

Burbank was obtained in June.

Table 6.2 shows the yield at different planting time for Banisogosogo.

Yield (kg/ha)

Variety Sebago Russet Burbank

January 2 1992 0

February 1 1146 0

March 1 1560 0

April 1 5323 431

May 1 4010 2619

June 1 1994 2728

July 2 (default run) 5975 2644

August 1 4857 1753

September 1 1231 580

October 1 2724 1795

November 1 1877 0

December 1 1359 0

Table 6.3 shows the optimisation for row spacing for Sebago and Russet Burbank. A

row spacing of 30 cm for each variety produced the highest yield.

142

Table 6.3 shows the yield at different row spacing for Banisogosogo.

Yield (kg/ha)


Row Spacing ( cm) Plant Population per m2

75 (default run) 5 2686 1943

30 11 3470 1896

40 8 3190 1902

50 6 2881 1924

80 4 2448 2010

100 3 2178 2035

The optimisation for irrigation treatment for Sebago and Russet Burbank was also

conducted (Table 6.4). For Sebago variety, irrigation application of 16.0 mm per

plant gave the highest yield while for Russet Burbank, the optimum irrigation

application was 1.6 mm per plant.

Table 6.4 shows the yield under different irrigation application and irrigation amount

for Banisogosogo.

Yield (kg/ha)


Default run (1.6 mm and 4.0

mm)

2519 1761

Irrigation ( mm)

1.6 2619 2064

4.0 2884 2064

6.4 3571 1876

8.0 4526 1762

9.6 6871 1713

12.0 7402 1603

14.4 7574 1595

16.0 7598 1535

The table below (Table 6.5) shows the optimisation for planting depth for Sebago

and Russet Burbank. Planting depth of 2 cm and 1.5 cm gave the greatest yield for

Sebago and Russet Burbank respectively.

143

Table 6.5 shows the planting depth with corresponding yield for Banisogosogo.

Yield (kg/ha)


Planting Depth ( cm)

1.5 (default run) 2686 1943

2 3011 1696

4 2739 453

6 1846 325

8 2686 0

10 2686 0

Table 6.6 shows the optimisation for fertiliser treatment for Sebago variety. The

highest yield for Sebago was obtained using 300 kg/ha at 10 cm. The results also

indicated that the yield obtained for optimum fertiliser and optimum irrigation

simulation was higher than default value and optimised irrigation yield but lower

than optimised fertiliser yield (Table 6.7).


corresponding yield for Banisogosogo for Sebago variety.


Default run (80 N kg/ha in

2 splits at 5 cm)

2577


60 2579 2579 2579 2652 2758 2826

120 2622 2622 2622 2787 3004 3136

180 2640 2640 2640 2895 3241 3409

240 2646 2646 2646 2992 3432 3671

300 2648 2648 2648 3109 3607 4359

Table 6.7 shows the optimum fertiliser and irrigation management for Banisogosogo

for Sebago variety.


300 kg/ha 16.0 2645

144

The table below shows the optimisation for fertiliser treatment for Russet Burbank

variety (Table 6.8). The highest yield for Russet Burbank was obtained using 300

kg/ha under 0-4 cm. Under the optimum fertiliser and optimum irrigation treatments,

the yield obtained was higher than the default value but lower than the optimum

fertiliser yield (Table 6.9).


corresponding yield for Banisogosogo for Russet Burbank.


Default run (80 N kg/ha in

2 splits at 5 cm)

1990


60 1915 1915 1915 1903 1888 1877

120 2008 2008 2008 1995 1837 1778

180 2369 2369 2369 2400 2256 2033

240 2402 2402 2402 2305 2132 2126

300 2420 2420 2420 2173 2018 1999

Table 6.9 shows the optimum fertiliser and irrigation management for Banisogosogo

for Russet Burbank.

Fertiliser Application (kg/ha at x cm) Irrigation Application ( mm) Yield (kg/ha)

300 at 0 cm 1.6 1959

300 at 2 cm 1.6 2381

300 at 4 cm 1.6 2381

300 at 0 cm 4.0 1923

300 at 2 cm 4.0 2348

300 at 4 cm 4.0 0

6.2.3 Climate variability (El Niño Southern Oscillation) simulation

The impact of climate variability was also studied on Sebago and Russet Burbank

potato yield for Banisogosogo. Differences in the yield were noticed during El Niño,

La Niña and neutral years. For Sebago cultivar, average El Niño years gave the

lowest yield (7 year average) at 2124 kg/ha (percentage difference 63.43%) while La

Niña produced an average yield of 2883.43 kg/ha at 34.77 % (Table 6.11). On the

145

other hand, La Niña year gave the lowest yield for Russet Burbank (1430.71 kg/ha

with percentage difference of 50.67%) and El Niño had an average yield of 1998.14

kg/ha with a percentage difference of 18.34% (Table 6.13).

Table 6.10 shows the impact of ENSO on potato yield for Sebago variety at

Banisogosogo from 1983-2012.

El

Niño

Water

Stress

N

Stress

Yield

(kg/ha)

La

Niña

Water

Stress

N

Stress

Yield

(kg/ha)

Neutral Water

Stress

N

Stress

Yield

(kg/ha)

1987 0.57 0.17 2044 1988 0.80 0.16 985 1983 - 0.05 7757

1991 0.14 0.27 578 1996 0.32 0.15 1809 1984 0.14 0.18 2365

1992 0.56 0.05 2774 1998 0.58 0.29 721 1985 0.36 0.18 4735

1993 0.09 0.25 438 1999 0.05 0.07 6914 1986 0.36 0.07 2291

1994 0.50 0.31 875 2000 0.28 0.10 4710 1989 0.3 0.09 4433

1997 0.08 0.20 585 2008 0.65 0.21 2545 1990 0.17 0.11 6368

2002 - 0.05 7574 2012 0.21 0.3 3483.4 1995 0.86 0.34 708

2004 0.04 0.09 3788 2001 0.47 0.2 2960

2006 0.47 0.10 3692 2003 0.45 0.26 759

2010 0.72 0.33 1407 2005 0.31 0.16 2085

2007 0.65 0.14 1328

2009 0.36 0.11 4258

Table 6.11 indicates 7 year average for ENSO yield in Banisogosogo for Sebago.

The results indicated that highest average yield was obtained under neutral year, El

Niño had a percentage difference of 63.43% while La Niña gave a percentage

difference of 34.77%.

146

Table 6.11 shows the 7 year average for ENSO yield for Banisogosogo for Sebago

variety. The highest yield was obtained by the neutral years.


(%)

El Niño (1987-2002) 2124 63.43

La Niña (1988-2012) 2883.43 34.77

Neutral Year (1983- 1995) 4096.86

Table 6.12 shows the yield obtained under each ENSO state and water and nitrogen

stress faced during the ENSO state. Russet Burbank variety appeared to be more

stable variety for El Niño years.

Table 6.12 shows the impact of ENSO on potato yield for Banisogosogo for Russet

Burbank from 1983-2010.

El Niño

Water Stress

N Stress

Yield (kg/ha)

La Niña

Water Stress

N Stress

Yield (kg/ha)

Neutral Water Stress

N Stress

Yield (kg/ha)

1987 0.65 0.23 2151 1988 1.00 0.33 292 1983 - 0.02 1786

1991 0.43 0.30 1243 1996 0.32 0.18 2526 1984 0.48 0.28 1658

1992 0.70 0.13 1070 1998 0.73 0.33 698 1985 0.35 0.12 1323

1993 0.10 0.25 543 1999 0.03 0.03 1497 1986 0.35 0.16 2700

1994 0.59 0.31 1055 2000 0.29 0.22 2355 1989 0.31 0.21 4381

1997 0.07 0.17 2152 2008 0.82 0.32 402 1990 0.14 0.14 4434

2002 - 0.06 5773 2012 0.19 0.17 2245 1995 1.00 0.55 530

2004 - 0.07 6746 2001 0.59 0.17 1996

2006 0.53 0.08 3939 2003 0.59 0.28 973

2010 1.00 0.40 1172 2005 0.44 0.18 2727

2007 0.87 0.30 339

2009 0.27 0.34 1689

Table 6.13 indicates 7 year average for ENSO yield in Banisogosogo for Russet

Burbank. The results indicated that highest average yield was obtained under neutral

year, the percentage difference for El Niño was 18.34% while La Niña gave a

percentage difference of 50.67%.

147

Table 6.13 shows 7 year average for ENSO yield for Banisogosogo for Russet Burbank.

The highest yield was obtained under neutral year.


(%)

El Niño (1987-1997) 1998.14 18.34

La Niña (1988-2008) 1430.71 50.67

Neutral Year (1983- 1990) 2401.7

6.2.2 Future climate simulations

The impacts of future climatic scenario on Sebago and Russet Burbank were

simulated in Banisogosogo using DSSAT SUBSTOR Potato model. In this study two

PSSCP climate scenarios were used (similar to Chapter 5). The results are divided

into two sections; (i) the future climate simulations with potential and non-potential

production and (ii) the optimisation treatments.

6.2.2.1 Banisogosogo Sebago simulation

For Banisogosogo potential simulations (Table 6.14), under future emission scenario,

the tuber initiation day and LAI increased while the tuber fresh and dry weight

decreased. However, there was no tuber production noted under 2090 potential

simulation. It was also noticed that the tuber initiation day was earlier for non-

potential simulation as compared to potential simulation. The LAI values for non-

potential simulation were lower when compared to potential simulation.

Table 6.14 shows the Banisogosogo future climate simulation for Sebago variety. 2030 2055 2090 Potential Non-Potential Potential Non-Potential Potential Non-Potential Variable Medium High Medium High Medium High Medium High Medium High Medium High Tuber Initiation Day (dap)

41 39 39 37 53 51 52 47 -99 -99 57 -99


4835 5511 2947 3703 2068 3333 1485 2566 0 0 19 0


24.17 27.56 14.73 18.52 10.34 16.66 7.42 12.83 0 0 0.09 0

Leaf Area Index

6.13 6.12 2.37 2.43 9.33 11.46 4.55 4.71 22.05 23.87 6.93 8.33

148

Figure 6.4 shows Banisogosogo future climate potential simulations for Sebago for LAI, leaf

weight, stem weight, tuber dry and fresh weight and tops weight.

The above simulation indicates that tuber dry yield (e) and tuber fresh weight (f) was

only possible for 2030-2055 A1B and A2 emission scenario. The above figure also

shows that the highest tuber yields were obtained under 2030 A1B and A2 emission

scenario. On the other hand, 2030 A1B and A2 emission scenario, gave the lowest




149

value for LAI, leaf weight and stem weight (a, b, c, and d respectively) while

vegetative growth was highest for 2090 A1B and A2 emission scenario.

Figure 6.5 shows Banisogosogo future climate non-potential simulations for Sebago for LAI,

leaf weight, stem weight, tuber dry and fresh weight and tops weight.




150

The above figure illustrates that tuber dry weight (e) and tuber fresh weight (f) under

non-potential simulation was possible for 2030-2055 A1B and A2 emission scenarios

and 2090 A2 emission scenario. However, the yield for 2090 A2 emission scenario

was very low.


The optimisation treatment was conducted for non-potential simulation for Sebago

and Russet Burbank to investigate which crop management strategies maximised the

tuber yield under future climate scenario. Optimisation simulation was conducted for


fertiliser and optimum irrigation treatment and also planting depth. The results

suggest that under each emission scenario different crop management practices

optimise the yield (Table 6.15 and Table 6.17).

The Banisogosogo optimisation treatment showed that under 2030 AIB and A2

emission scenario, the optimum planting time was June while for 2055 AIB and A2

and 2090 A1B emission scenario, the optimum planting time was July. The optimum

row spacing for 2030 A1B and A2 was 30 cm whereas the optimum row spacing for

2055 A1B and A2 was 40 cm and 100 cm respectively. The optimum row spacing

for 2090 A1B was 30 cm. Under 2030 A1B and A2 emission scenario, 6.4 mm

irrigation produced the highest yield but for 2055 and 2090 A1B emission scenario,

1.6 mm irrigation gave the highest yield. The optimisation treatment for fertiliser for


at 10 cm and application of 300 kg/ha fertiliser at 8 cm simulated the highest yield

respectively. In 2055 A1B and A2 emission scenario, the optimum fertiliser

application was 300 kg/ha fertiliser at 0-4 cm while in 2090 A1B emission scenario,

the optimum fertiliser application was 60 kg/ha fertiliser at 0-4 cm. The optimum

fertiliser and the optimum irrigation application indicated that the yield was not

optimised for 2030 and 2055 A1B and A2 emission scenario. However, the optimum

fertiliser and the optimum irrigation application did optimise the yield for 2090 A1B

emission scenario. Under 2030-2055 A1B and A2 and 2090 A1B emission scenario,

the optimum planting depth was 1.5 cm.

151

Table 6.15 shows the optimisation treatments for Sebago.

Optimum Treatment, Yield (kg/ha) Banisogosogo 2030 A1B


2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

Planting Time June, 5004 June, 6085 July, 2840 July, 4015 July, 597 Row Spacing 30 cm, 2988 30 cm, 4081 40 cm, 1493 100 cm, 2603 30 cm, 18 Irrigation Application

6.4 mm, 5251 6.4 mm, 6465 1.6 mm, 1673 1.6 mm, 2656 1.6 mm, 20



300 kg/ha fertiliser at 8, 5367





4641 5749 2211 3130 911

Planting Depth 1.5 cm, 2947 1.5 cm, 3703 1.5 cm, 1485 1.5 cm, 2566 1.5 cm, 17

6.2.2.3 Russet Burbank variety simulations

For Banisogosogo Russet Burbank potential simulations (Table 6.16) tuber yield was

only possible under 2030 A2 non-potential simulations. Tuber dry weight obtained

under this scenario was 37 kg/ha while the tuber fresh weight obtained was 0.19 t/ha.

However, this tuber yield was negligible. The LAI under all emission scenarios was

very high. Figures 6.6 and 6.7 represent the same tabulated information. These

graphs show that under potential and non-potential simulation there was very high

vegetative or above-ground growth.

Table 6.16 shows Banisogosogo simulation results of future climate simulation for

Russet Burbank Variety for A1B and A2 emission scenario. 2030 2055 2090 Potential Non-

Potential Potential Non-

Potential Potential Non-

Potential Variable A1B A2 A1B A2 A1B A2 A1B A2 A1B A2 A1B A2 Tuber Initiation Day (dap)

-99 -99 -99 74 -99 -99 -99 -99 -99 -99 -99 -99


0 0 0 37 0 0 0 0 0 0 0 0


0 0 0 0.19 0 0 0 0 0 0 0 0

Leaf Area Index

19.14 19.76 9.70 9.26 19.01 24.64 9.01 10.32 19.08 20.65 7.55 7.75

152

Figure 6.6 shows Banisogosogo future climate potential simulations for Russet Burbank for

LAI, leaf weight, stem weight, tops weight, tuber fresh and tuber dry weight under A1B and

A2 emission scenario.




153

The above simulation showed no tuber dry weight (e) or no tuber fresh weight (f) in

any of the emission scenario while LAI (a), leaf weight (b), stem weight (c) and tops

weight (d) showed gradual increase in weight over time with very high values.

Figure 6.7 shows Banisogosogo future climate non potential simulations for Russet Burbank

for LAI, leaf weight, stem weight, tops weight, tuber fresh and tuber dry weight under A1B

and A2 emission scenario.




154

The above simulation shows that tuber dry yield (e) and tuber fresh yield (f) was only

possible under 2030 A2 emission scenario. Even under non-potential conditions,

there was very high vegetative growth (a-d).

6.2.2.4 Optimisation treatments for Russet Burbank variety.

The Banisogosogo optimisation treatment for Russet Burbank showed that under

2030 AIB and A2 emission scenario, the optimum planting time was July while for

2055-2090 A1B and A2 emission scenario, the optimum planting time was

September. The optimum row spacing for A2 was 30 cm and 40 cm respectively.

Under 2030 A1B and A2 emission scenario, 1.6 mm irrigation simulated the highest

yield. The optimisation treatment for fertiliser for 2030 A1B and A2 emission

scenario indicated that application of 60 kg/ha fertiliser at 10 cm for both emission

scenarios gave the highest yield. The optimum fertiliser and the optimum irrigation

application indicated that the yield was not optimised for 2030 A1B emission

scenario while the yield under 2030 A2 emission scenario was optimised. Under

2030 A2, the optimum planting depth was 1.5 cm.

Table 6.17 shows the optimisation treatments for Russet Burbank.

Optimum Treatment, Yield (kg/ha) Banisogosogo 2030 A1B


2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

2090 A2 Emission

Planting Time July, 342 July, 843 September, 68

September, 124

September, 8

September, 2

Row Spacing - 30 cm and 40 cm, 38

- - - -


1.6 mm, 1 1.6 mm, 88 - - - -

Fertiliser

Application

60 kg//ha 0-

10 cm, 7

60 kg/ha 6-

10 cm, 151

- - - -

6.3 Discussion

6.3.1 Current climate simulations

6.3.1.1 Potential simulations

The results of Banisogosogo potential simulation indicated that the emergence day

for Sebago was 7 days after planting which was 9th of July and the tuber initiation

day was 33 days after planting which was 4th of August. The emergence for Russet

Burbank was 7 days after planting and tuber initiation day was 53 days after planting

which was 24th August (for crop stages and development refer to Appendix 4A). The

155

tuber fresh weight for Sebago and Russet Burbank was 38.88 t/ha and 14.43 t/ha with

tuber dry weight as 7777 kg/ha and 2885 kg/ha respectively (refer to Table 6.1). The

leaf area index for Sebago was 6.14 while for Russet Burbank it was 8.44. As shown

in Figure 6.0 for Sebago, LAI, leaf weight, stem weight, tops weight, tuber fresh and

tuber dry weight increased steadily with time. Figure 6.1 indicated that the LAI and

leaf weight for Russet Burbank increased initially and decreased after 23rd of August.

The tuber fresh weight started to increase around 23rd of August. The drop in LAI

and leaf weight can be due to transfer of photosynthate from the leaves to the tubers.

For Russet Burbank, tuber initiation occurred during early flowering (Dwelle and

Love, 1993), the maximum bulking rate has been recorded at 7-8 cwt/A-day and

under optimum growing conditions, the tuber bulking rate is linear from tuber

initiation until maturity or leaf senescence (Kleinkopf et al., 2003).

6.3.1.2 Non-potential simulation

The Banisogosogo Sebago non-potential simulation indicated that the emergence day

was 7 days after planting which was 9th of July. The tuber initiation day was 31 days

after planting which was 2nd of August. Under non-potential conditions the tuber

fresh weight was 13.43 t/ha while the tuber dry weight was 2686 kg/ha with leaf area

index of 0.45. As shown by Figure 6.0, LAI, leaf weight, stem weight and tops

weight all increased from the day of planting until 3rd of August from which there

was a decline in LAI and leaf weight, stem weight and tops weight. On the other

hand, there was an increase in tuber dry and fresh weight after 3rd of August. This

can be due to the transfer of photosynthetic product from the above ground to below

ground. There was also evidence of nitrogen stress on 2nd August and water and

nitrogen stress on 20th September. Figure 6.2 indicated that there was gradual

increase in precipitation and total nitrate rate over time. The increase in total nitrate

can be due to application of nitrogen fertilisers on 7th and 23rd of July. Furthermore,

it also showed that the total nitrate available in the soil increased few days after the

application of fertilisers. This can be due to onset of plant senescence or decay of

aboveground components, such as, leaves and stems, as tubers reached maturity. On

the other hand, Russet Burbank emerged 7 days after planting (9th July). In Russet

Burbank, aged seed accounted for earlier emergence along with an increase of stem

number per plant. However, the seed age did not affect the overall yield (Olsen and

Nolte, 2004). Table 6.1 indicated that the tuber initiation day was 50 days after

156

planting which was 21st of August with tuber fresh weight of 9.61 t/ha, leaf area

index of 5.88 and tuber dry weight of 1922 kg/ha. The optimum soil temperature for

Russet Burbank tuber growth is 16 ºC while the optimum temperature for vine

growth is 25 ºC. At higher temperatures, tuber growth is delayed (Dwelle and Love,

1993). At temperatures of 15 ºC and 24 ºC, Russet Burbank and White Rose, had

higher yield, high starch content and high specific gravity as compared to higher

temperatures (Yamaguchi et al., 1964). Figure 6.1 indicated that LAI, leaf weight

and tops weight increased from the planting day until 30th of August after which it

showed fluctuation in weight and then increased gradually with time. On the other

hand, there was an increase in tuber dry and fresh weight after 23rd of August.

Translocation for Russet Burbank from vine to tubers may account for 10-15 % of

the final yield (Kleinkopf et al., 2003). Figure 6.3 also showed an increase in total

nitrate rate due to application of fertilisers (on 7th and 23rd of July nitrogen fertiliser

was applied which also shows an increase in nitrate rate). However, there was a

decline in nitrate rate and total water content in the soil around 10th of August and the

harvest date. It was seen that the crop underwent water stress on 20th of September

and nitrogen stress on 21st August and 20th September.

It was also noticed from the simulations that there is a huge difference between

potential and non-potential yields for both the varieties. Under potential production,

maximum production was reported with no indication of water and nitrogen stress.

While under non-potential production, the crop was sensitive to water and nitrogen

stress which reduced the crop’s performance and yield.

6.3.1.3 Optimisation treatment

To begin with, optimisation was conducted for planting time to see which month

gave the highest tuber yield. The optimisation for planting time, as shown by Table

6.2, indicated that planting of Sebago variety in Banisogosogo during the month of

July produced the highest yield (5975 kg/ha). This was because the month of July

gave the low average maximum and minimum temperature with adequate rainfall

from planting to harvest. On the other hand, Russet Burbank should be planted in

June which produced tuber dry weight of 2728 kg/ha. The month of June gave the

lowest average maximum and minimum temperature. The results also indicated that

for cultivar Russet Burbank, the months of January, February, March, November and

December produced no tuber. This was because these months had no tuber initiation

157

development phase. As mentioned earlier, low temperature is necessary for tuber

initiation. For cultivar Russet Burbank, which requires more time for tuber initiation,

planting should take place earlier (June) as compared to Sebago (July). Planting too

early in the season can lead to slow emergence, decreased plant vigor, slow tuber

growth rate and susceptibility to seed disease while planting too late in the season

delays canopy development and reduces time available for tuber bulking (Dwelle and

Love, 1993).

Row spacing was also optimised and the yield obtained under each treatment was

noted. The optimisation results for Sebago for Banisogosogo for row spacing

indicated that a row spacing of 30 cm gave the highest yield at 3470 kg/ha. On the

other hand, Russet Burbank simulation in Banisogosogo indicated that a row spacing

of 100 cm produced the highest yield (Table 6.3). Planting closer to optimum reduces

photosynthetic capacity for bulking giving rise to smaller tubers and lower

marketable yields while planting wider than the optimal row spacing can lengthen

the time for full canopy development reducing carbohydrate supply to tubers (Dwelle

and Love, 1993).

Likewise, optimisation was conducted for irrigation application. Banisogosogo

optimisation treatment for irrigation for Sebago indicated that increasing the

irrigation level increased the tuber yield. Hence, the highest tuber yield was obtained

at 16 mm of irrigation while the lowest tuber yield was obtained at 1.6 mm. The

yield of Sebago variety can be increased by maintaining high but safe soil moisture

content throughout the potato growing season (Myhre, 1956). On the other hand, for

Russet Burbank simulation, application of 1.6 mm and 4.0 mm of irrigation per plant

gave the highest yield at 2064 kg/ha (Table 6.4). Water stress can delay the

appearance of particular order of apical branches, the production rate of new leaves,

enhance leaf senescence and there is slowed maturation and compensatory leaf

production upon recovery from transient water stress. As compared to water treated

plants, drought treated plants produced less leaf area 50-83 days after planting (Vos

and Haverkort, 2007b).

The optimisation simulation for Banisogosogo planting depth indicated that planting

of Sebago variety at a depth of 2 cm produced the highest tuber yield (3011 kg/ha).

The lowest tuber yields was obtained at 6 cm (1846 kg/ha) while planting depth of

158

1.5 cm, 6 cm and 8 cm gave the similar tuber yield of 2686 kg/ha. The simulation for

Russet Burbank indicated that increasing the planting depth decreased the tuber yield

(Table 6.5). Planting of Russet Burbank at great depths such as 9 inches (22.86 cm)

greatly reduced tuber yield (Bohl, 2006).

Furthermore, fertiliser treatment was also optimised. As indicated by Table 6.6, the

Banisogosogo fertiliser results for Sebago showed that application of 300 kg/ha

fertiliser at 10 cm gave the highest yield at 4359 kg/ha. Application of fertiliser at 0

cm, 2 cm and 4 cm gave similar yield. On the other hand, Russet Burbank simulation

showed that application of 300 kg/ha fertiliser at 0 cm, 2 cm and 4 cm gave the

highest yield at 2420 kg/ha (Table 6.8). It was also noticed that the yield obtained

under 0-4 cm was similar. This is because surface application (0 cm) will be similar

to application up to top layer soil. The model will show very little or no difference

between first soil layer (0-5 cm) applications. However, at 6 cm soil layer, the model

starts to show difference. For Russet Burbank, LAI of 3.0-3.5 maximises the

interception of sunlight. Under excessive nitrogen fertiliser application for Russet

Burbank, there is increase in vine size above the necessary to maximise the bulking

rate. However, under excessive nitrogen fertiliser application, there is delay in tuber

initiation as much as 10 to 14 days which can reduce yield as much as 90 cwt/A

(Kleinkopf et al., 2003).

Finally optimisation was conducted for optimum fertiliser and optimum irrigation

treatments. When optimum fertiliser application and the optimum irrigation levels for

Sebago and Russet Burbank were combined as a treatment for simulation, the results

suggested that the yield of tuber obtained under this simulation was higher than

default value and optimised irrigation yield but lower than optimised fertiliser yield.

For impacts of ENSO on potato yield simulation neutral year gave the highest tuber

yield for both varieties. For Sebago variety (Table 6.11), average El Niño years gave

the lowest yield (7 year average) at 2124 kg/ha (percentage difference of 63.43%)

while La Niña gave an average yield of 2883.43 kg/ha (percentage difference of

34.77%). El Niño years gave the lowest yield because the crop underwent a lot of

water stress as compared to other ENSO phase (as shown in Table 6.10). On the

other hand, for Russet Burbank variety, La Niña year gave the lowest yield at

1430.71 kg/ha with percentage difference of 50.67% and El Niño gave an average

159

yield of 1998.14 kg/ha with a percentage difference of 18.34% (Table 6.13). As

shown by Table 6.12, La Niña years faced a lot of nitrogen stress which could be one

reason for the low yield obtained. Overall, it can be said that Russet Burbank was a

more stable variety for El Niño events.

6.3.2 Future climate simulation

6.3.2.1 Potential simulations

Banisogosogo Sebago potential simulation (Table 6.14) for 2030 medium emission

scenario indicated that the emergence day was 10th of July (8 days after planting),

tuber initiation day was 12th of August, that is, 41 days after planting. Table 6.14

indicated that the tuber dry weight was 4835 kg/ha with 24.17 t/ha of tuber fresh

weight with LAI of 6.13. For 2030 high emission scenario the results showed that the

emergence day was 10th of July, tuber initiation day was 39 days after planting, that

is, 10th of August, with tuber dry weight of 5511 kg/ha and tuber fresh weight of

27.56 t/ha with LAI of 6.12. The simulation for 2030 high emission scenario showed

that the tuber initiation was 2 days earlier with higher tuber dry and tuber fresh

weight. One reason for this can be the difference in annual temperature increase and

precipitation level in each scenario, that is, 2030 high emission scenario had low

annual temperature increase with higher level of precipitation as compared to 2030

medium emission scenario. Under 2055 medium emission scenario, the emergence

day was 11th of July, tuber initiation day was 53 days after planting, that is 34th of

August, with tuber dry weight of 2068 kg/ha and tuber fresh weight of 10.34 t/ha and

LAI of 9.33. For 2055 high emission scenario, it was seen that the emergence day

was 11th of July, tuber initiation day was 51 days after planting, that is, 22nd of

August. The tuber dry weight was 3333 kg/ha with tuber fresh weight of 16.66 t/ha

with LAI 11.46. There was no tuber yield under 2090 medium and high emission

scenario. As shown by Figure 6.4, under 2030 and 2090 medium and high emission

scenario, the LAI, leaf weight, stem weight and tops weight increased with time

while under 2055 medium and high emission scenario, the LAI increased from the

planting date until the 23rd of August and 21st of August respectively. The tuber

weight increased gradually from 15th August, 11th August, 26th August, 23rd August

for 2030 medium emission, 2030 high emission, 2055 medium emission and 2055

high emission respectively. The figure also showed that 2090 medium emissions had

high vegetative growth while the highest tuber fresh and dry yield was shown by

160

2030 high emission followed by 2030 medium emission, 2055 high emission and

2055 medium emission. On the other hand, it was seen that Russet Burbank under all

emission scenario, the yield was zero with very high vegetative growth and LAI

(Table 6.16 and Figure 6.4).

6.3.2.2 Non potential simulations

Banisogosogo non potential simulations for Sebago for 2030 medium emission

scenario indicated that the emergence day was 11th of July, tuber initiation day was

10th of August (39 days after planting) with tuber dry weight of 2947 kg/ha and tuber

fresh weight of 14.73 t/ha and LAI of 2.37. Water stress was seen on the 20th of

September and nitrogen stress on the 10th of August and 20th of September. For 2030

high emission scenario it was seen that the emergence date was 10th of July, tuber

initiation day was 8th of August (37 days after planting). The harvest for dry weight

was 3703 kg/ha with harvest of fresh weight of 18.52 t/ha and LAI of 2.43. The crop

faced water and nitrogen stress on 8th of August and 20th of September. Availability

of soil nutrients and moisture determines not only the number but also the quality of

tubers that reach maturity (Food and Agricultural Organisation of the United

Nations, 2008). Under elevated carbon dioxide there is limited transpiration which

leads to decreased nitrogen uptake and chlorophyll loss (Bindi et al., 2002). It can be

also seen that the 2030 high emission had an earlier tuber initiation day with higher

dry and fresh tuber yield. Under 2055 medium emission scenario, the emergence day

was 12th of July with tuber initiation at 23rd of August (52 days after planting). The

tuber dry weight was 1485 kg/ha and tuber fresh weight was 7.42 t/ha and LAI of

4.55. It was seen that the crop underwent water and nitrogen stress on 23rd of August

and 20th of September. Furthermore, 2055 high emission scenario indicated that the

emergence day was 11th of July, tuber initiation day was 18th of August (47 days

after planting) with tuber dry weight at 2566 kg/ha, tuber fresh weight at 12.83 t/ha

and LAI of 4.71. The crop underwent water stress on the 20th of September and

nitrogen stress on 18th of August and 20th of September. It can also been seen that

the tuber initiation day increased and the tuber dry weight and fresh weight decreased

under 2055 emission scenario as compared to 2030 emission scenario. For 2090

medium emission scenario, the emergence day was 14th of July, tuber initiation day

was 14th of September (74 days after planting). This emission scenario had very little

tuber yield, with tuber dry weight at 19 kg/ha, tuber fresh weight at 0.09 t/ha and LAI

161

of 6.93. The crop underwent water and nitrogen stress on 14th and 20th of September

respectively. Under 2090 high emission scenario, the emergence day was 14th of

July. However, there was no tuber initiation. Figure 6.5 indicated that LAI, leaf

weight, stem weight and tops weight increased from the day of planting but

experienced a decline in weight on 31st of August, 8th and 13th of September. The

tuber weight increased from 11th of August, 9th of August, 26th of August, 20th of

August for 2030 medium, 2030 high, 2055 medium and 2055 high emission scenario

respectively. The LAI, leaf weight, stem weight and the tops weight were the highest

for 2090 medium emission scenario followed by 2055 high emission scenario, 2055

medium emission scenario, 2030 high emission scenario and 2030 medium emission

scenario. The tuber fresh weight and tuber dry weight indicated that the highest yield

were obtained under 2030 high emission scenario followed by 2030 medium

emission scenario, 2055 high emission scenario, 2055 medium and 2090 medium

emission scenario. Warm temperatures favour vegetative growth while cool

temperature favour tuber growth (Ewing, 1981c; Khedher and Ewing, 1985). On the

other hand, Russet Burbank can only be cultivated under 2030 high emission

scenario. This emission scenario produced negligible yield. The emergence date

under this emission scenario was 10th of July (8 days after planting) with tuber dry

weight of 37 kg/ha, tuber fresh weight of 0.19 t/ha and LAI of 9.26. Under all other

emission scenario, it was observed that there was high LAI but no tuber yield (Table

6.16). Figure 6.7 indicated that under 2030 high emission scenario, the LAI increased

from the day of planting until 13th of September. After 13th of September until the

day of harvest there was a decline in LAI value. For all other scenarios, the LAI

increased gradually with time. The tuber fresh weight for 2030 high emission

scenario increased from 13th of September until 16th of September. From 16th of

September to 17th of September there was no gain in weight. The tuber weight

increased again from 18th of September to the day of harvest. The crop faced water

and nitrogen stress on 20th of September. Finally, it can be said that Sebago was the

best variety to be cultivated in Banisogosogo. The effect of elevated temperature on

potato growth and yield is determined by a combination of factors, such as, between

soil temperature, air temperature, solar radiation and photoperiod (Sarquis et al.,

1996). For Sebago, it was noted that high temperature (32 oC/28 oC, day/night)

promote haulm growth and suppress tuber production (Menzel, 1980, 1983a). At

elevated temperatures dry matter is diverted to shoots rather than tubers (Menzel,

162

1985). Correspondingly, at high shoot and root temperatures tuber formation is

strongly inhibited. Disbudding promoted tuberisation at such cases (Menzel, 1983b).

The yield of potatoes should increase by 10% for every extra 100 ppm CO2 (Bindi,

2008). Under 28 oC and 30 oC, there was a delay in tuber formation for Desiree and

Russet Burbank (Nowak and Colborne, 1989). When Russet Burbank was grown

under 350 of 1000 µmol mol-1, irradiance of 400 or 800µE m-2s-1 photosynthetic

photon flux and photoperiod of 12 or 24 hours, it was seen that tuber yield increased

at elevated carbon dioxide but at lower photosynthetic photon flux (Pereira and C.C.

Shock, 2006). Similarly, another study shows that Russet Burbank tuber yield were

highest at low temperature of 17.5oC, carbon dioxide concentration of 1600µL L-1

and medium light of 455 µmol m-2s-1 (Yandell et al., 1988). The effect of

temperature (25 oC/12 oC and 35 oC/25 oC) on dry matter production of Russet

Burbank, Desiree and DTO-28 was studied for five week, 2 week after tuberisation.

The study showed that tuber growth rate was more affected by high temperature than

the whole plant. The study also showed that all varieties experienced a decline in

tuber dry matter production at high temperature as compared to low temperature with

Russet Burbank experiencing the largest decline (Thornton et al., 1996). When

Russet Burbank is grown under soil temperatures of 15 oC and 24 oC, the yield,

specific gravity and starch content were high while the sugar content was low as

compared to cultivation under higher temperatures (Yamaguchi et al., 1964). An

increase in carbon dioxide concentration leads to 20 % decrease in stomatal

conductance (Olesen and Bindi, 2002; Bradshaw, 2007), enhance water efficiency,

increase C/N ratio and lower dark respiration (Bruhn, 2002; Olesen and Bindi, 2002;

Te mmerman et al., 2007) and increase in photosynthesis (Bindi, 2008). Elevated

concentration of carbon dioxide also causes an imbalance between the source and

sink capacity (Stitt, 1991). Long term exposure to carbon dioxide suppresses

photosynthetic activity and this can be linked to excess carbohydrate accumulation

which leads to repression of photosynthetic genes in the leaf and excess starch which

hinders carbon dioxide diffusion (Makino and Mae, 1999).


To begin with, the planting date was optimised for Sebago and Russet Burbank

variety. The optimisation of planting date for Banisogosogo for Sebago indicates that

under 2030 medium and high emission scenario the month of June simulated the

163

highest yield as this month received low average maximum and minimum

temperatures. For 2055 medium and high emission scenario and 2090 medium

emission scenario, the month of July produced the highest yield. This was because

the month of July received the low maximum and minimum temperature with high

average solar radiation and high average rainfall from planting to harvest. The results

indicated that under 2030 medium emission scenario, January, February and March

had no tuber yield while under 2030 high emission scenario, February gave no tuber

yield. Under 2055 medium emission scenario, January-March and November and

December did not produce any tuber yield while under 2055 high emission scenario,

January-March and December produced no tuber yield. Finally, under 2090 medium

emission scenario, January-May and November and December gave no tuber yield

(Appendix 4.4A, Table 4.26A). These months did not produce any tuber yield

because the high average maximum and minimum temperatures inhibited tuber

initiation stage. Under current climate conditions, the month of July was the

optimum month for potato cultivation. On the other hand, Russet Burbank indicated

that under 2030 medium and high emission scenario, July gave the highest yield.

This was because the month of July had low average maximum and minimum

temperature. While under 2055 and 2090 medium and high emission scenario,

September gave the highest yield. Under 2050 medium emission scenario, only the

month of September had tuber initiation stage while under 2055 high emission

scenario, September received the highest average solar radiation. Under 2090

medium and high emission scenario, the highest yield was produced in September

which was 8 kg/ha and 2 kg/ha respectively. This was because only the month of

September had development phase of tuber initiation. The results also indicated that

under 2055 and 2090 emission scenario the months of January-August (except 2055

high emission scenario in which July gave tuber yield of 16 kg/ha), November and

December gave no tuber yield as there was no tuber formation. (Appendix 4.6A,

Table 4.42A). Under current climate conditions, the month of June was the optimum

month for potato production. When comparing the three varieties, it can be said that

Sebago is the better variety to plant as it gives higher tuber yield.

Secondly, the row spacing was also optimised. For Banisogosogo optimised row

spacing (Table 6.15) under 2030 medium emission scenario, the highest yield for

Sebago was obtained under row spacing of 30 cm while the lowest yield was

164

obtained under row spacing of 75 cm (Appendix 4.0A, Table 4.27A). For 2030 high

emission scenario, the highest yield of potatoes was obtained under row spacing of

30 cm at 4081 kg/ha. The row spacing under 2055 medium emission scenario

indicated that under row spacing of 40 cm gave the highest potato yield at 1493

kg/ha. Increasing the row spacing decreased the yield. Under 2055 high emission

scenario, increasing the row spacing increased the yield. Hence, the highest yield was

obtained under 100 cm of row spacing while the lowest yield was obtained under

row spacing of 30 cm. The optimisation of row spacing under 2090 medium

emission scenario indicated that row spacing of 30 cm and 40 cm gave the highest

yield. On the other hand, as shown by Table 6.17, Russet Burbank indicated that

under 2030 high emission scenario, a row spacing of 30 cm and 40 cm gave a yield

of 38 kg/ha. For Desiree, under 2030 medium emission scenario, a row spacing of 75

cm is suitable to optimise yields. The results 2030 high emission scenario of row

spacing indicated that a row spacing of 40 cm produced the highest yield of 2803

kg/ha. Under 2055 medium and high emission scenario, it was noted that a row

spacing of 30 cm produced the highest yield

Optimisation simulation was also conducted for irrigation application. For

Banisogosogo optimised irrigation, as indicated by Table 6.15, under 2030 medium

and high emission scenario, irrigation application of 6.4 mm per plant gave the

highest yield at 5251 kg/ha and 6465 kg/ha. Under 2055 medium and high emission

scenario and 2090 medium emission scenario, irrigation application of 1.6 mm had

the highest yield. On the other hand, for Russet Burbank simulation under 2030

medium and high emission scenario, application of 1.6 mm of irrigation produced

the highest yield (refer to Table 6.17). Under Desiree, Banisogosogo 2030 medium

and high emission scenario and 2055 medium emission scenario indicated that the

default run gave the highest yield while under 2055 high emission scenario irrigation

application of 1.6 mm gave the highest yield (Appendix 4.0A Table 4.28A and

Table 4.44A).

Furthermore, the optimisation was also conducted for fertiliser application. For

Banisogosogo optimised fertiliser application, under 2030 medium and high

emission scenario, the highest yield was obtained under fertiliser application of 300

kg/ha at depth of 10 cm at 4256 kg/ha. To obtain the highest tuber yield under 2055

medium emission scenario, application of 300 kg/ha fertiliser was required at 8 cm

165

whereas under 2055 high emission scenario, the highest yield was obtained under

fertiliser application of 300 kg/ha at depth of 0-4 cm. The optimisation of fertiliser

application under 2090 medium emission scenario indicated that to obtain the highest

yield, 60 kg/ha of fertiliser should be applied at 0-4 cm. For Russet Burbank, it was

noted that under 2030 medium, application of 60 kg/ha fertiliser at 0-10 cm

simulated the highest yield while under 2030 high emission scenario, application of

60 kg/ha at 10 cm gave the highest tuber yield at 151 kg/ha. The results for Desiree

for 2030 medium and high emission scenario fertiliser treatment indicated that

application of 300 kg/ha fertiliser at 0-4 cm gave the highest yield (1475 kg/ha and

2265 respectively). The optimisation treatment also indicated that under 2050

medium and high emission scenario, the default value gave the highest tuber yield.

For Banisogosogo Sebago simulation, when the optimum fertiliser application and

the optimum irrigation application were combined to run the simulation for 2030

medium and high emission scenario. The simulation indicated that the yield is higher

than default run, optimum fertiliser yield but not higher than the optimum irrigation

yield. Under 2055 medium emission scenario, the results indicated that the yield

obtained under this simulation was higher than the default yield and the optimum

irrigation yield but not higher than the optimum fertiliser yield while for 2055

simulation, the treatments gave higher yield than the default treatment, optimum

fertiliser treatment and optimum irrigation treatment. For 2090 medium emission

scenario, the results indicated that the yield obtained under this simulation was

higher than the default value, optimum irrigation yield and optimum fertiliser yield.

For Russet Burbank 2030 medium emission, the results indicated that the yield for

this simulation was higher than default yield and optimum irrigation while under

2030 high emission simulation, the yield was not optimised. The simulation showed

that the yield obtained was higher than the default yield, optimum irrigation yield but

similar to optimum fertiliser yield. For Desiree, optimum nitrogen fertiliser

application and optimum irrigation application only optimised the yield for 2030

medium emission scenario.

To obtain the highest yield of potatoes in Banisogosogo for all three cultivars under

all emission scenario, potatoes should be planted at a depth of 1.5 cm. Increasing the

depth of planting decreased the yield (Appendix 4.0A, Table 4.29A and Table

4.45A).

166

6.4 Recommendations

It is recommended that the DSSAT model should be used for farm based decision

making (Quity, 2012a). This is because the DSSAT model can be used accurately to

simulate results and comprehensive analysis for different cultivars and different soil

types and other crop management strategies such as irrigation practices, and fertiliser

practices (Jones et al., 2010).

6.4.1 Current climate simulation

Under current climate conditions, the planting time of Sebago variety in

Banisogosogo should be July whereas for the Russet Burbank variety the optimum

planting time should be June. Under current climate conditions, both varieties

produced the highest tuber yield under 30 cm of row spacing. Furthermore, Sebago

variety required 16.0 mm of irrigation per plant to give the highest yield while Russet

Burbank required 1.6 mm of irrigation per plant to produce the highest tuber yield.

Planting depth of 2 cm and 1.5 cm gave the greatest yield for Sebago and Russet

Burbank respectively. The fertiliser application for Sebago stated that application of

300 kg/ha at 10 cm will give the highest yield while for Russet Burbank variety

application of 300 kg/ha fertiliser at 0-4 cm produced the highest yield.

6.4.2 Future climate simulation for Sebago variety

Under future climate scenario for Sebago variety, it is recommended that planting

should take place in cooler months. It was seen that the planting time for each

emission scenario varied. Under 2030 AIB and A2 emission scenario, it is

recommended that planting should take place in June while under 2055 AIB, 2055

A2 and 2090 AIB emission scenario, the recommended planting time changes to

July.

The row spacing under future climate scenarios also needs to be adjusted to

maximise yield. It is recommended that the row spacing for 2030 A1B and A2

should be 30cm while the row spacing for 2055 A1B, 2055 A2 and 2090 A1B should

be 40 cm, 100 cm and 30 cm respectively.

Furthermore, the irrigation levels also require adjustments under future climate

scenario to maximise yield. For example, under 2030 A1B and A2 emission scenario,

it is suggested that that irrigation level should be 6.4 mm irrigation per plant while

167

under 2055 A1B and A2 emission scenario, 1.6 mm irrigation per plant is

recommended.

To achieve high yield with fertiliser application, it is suggested that for 2030-2055

A1B and A2 emission scenario, 300 kg/ha fertiliser should be applied but at different

depth for each emission scenario. For 2090 A1B emission scenario, the

recommended fertiliser application is 60 kg/ha at 0-4 cm.

Finally, under all emission scenarios for Sebago variety, it is recommended that the

planting depth should be 1.5 cm.

6.4.2 Future climate simulation for Russet Burbank variety

The yield obtained for future climate simulations for Russet Burbank was zero or

negligible (non-potential 2030 A2 emission) which indicates that planting of Russet

Burbank in the future is not recommended.

6.5 Research limitations

The DSSAT SUBSTOR model should be evaluated for Sebago and Russet Burbank

potato varieties in Fiji. Based on the results of evaluation, recommendation should be

made to farmers about which variety is suitable for cultivation in which area.

Also, the weather file used for these simulations was also a “hybrid” weather file.

The weather data obtained from Fiji Meteorological Services had missing values

which had to be replaced with NASA values to run the simulation. This was found to

be too time consuming and the NASA values were slightly higher than the data

obtained from FMS.

As mentioned in Chapter 5, the graphics for 2090 A2 emission scenario was not

provided by Plant.Gro. Hence, the data had to be manually merged in Excel for all

the graphs. Also, it is suggested that DSSAT v4.6 be used as it has a better response

to carbon dioxide concentration and better response to radiation use efficiency.

168

Conclusion

This research was conducted to assess the impact of climate change and climate

variability on potato production in Fiji and to investigate which crop management

strategies maximised potato yield. The experiment site was located in Banisogosogo,

which is on the western side of Fiji Islands. Potatoes were grown during the months

of July-September 2012. The DSSAT SUBSTOR Potato model was run with site

specific soil data, weather data and crop management data under present and future

climate scenarios (PSSCP A1B and A2 emission scenario) to predict tuber yield.

The initial step involved the calibration of the DSSAT SUBSTOR Potato model.

This process included recalculation of soil inputs and recalibration of genetic co-

efficient. The soil lower limit, drained upper limit and soil saturation were

recalculated to ensure that these values were not biased by the high clay content of

Banisogosogo soil. The genetic co-efficient was also recalibrated given that Desiree

variety has not been grown under temperature and daylength regime similar to the

Fiji tropical experimental conditions. The DSSAT SUBSTOR Potato model was

calibrated for tuber dry weight using Banisogosogo replica 2 soil and weather data of

2012 (Januray-September). The value for R2 for replicate plot 1, replicate plot 2 and

replicate plot 3 was R2=0.88, R2=0.66 and R2=0.92 respectively. This indicates that

there is a positive relationship between the simulated yield and the observed yield.

Once the model was calibrated, current climate simulations were conducted for

Banisogosogo, Koronivia and Nacocolevu using Desiree variety. Under potential

simulation conditions, Koronivia gave the highest tuber yield (9811 kg/ha) while

under non-potential simulation conditions, Nacocolevu gave the highest yield (5405

kg/ha). Sensitivity analysis was also conducted to optimise crop management

practices for the three locations. The optimised crop management practices were site

specific as these depended on the soil composition and weather of that area. For

example, the optimum planting time for Banisogosogo was May (5356 kg/ha) while

the optimum planting time for Koronivia and Nacocolevu is August (7956 kg/ha and

7612 kg/ha). Also, ENSO years of the three locations were identified from the past

30 year climatic data and the impacts of both El Niño and La Niña events on potato

yield were studied. Under Banisogosogo and Nacocolevu simulation, the neutral

years received the highest average yield while under Koronivia simulation, El Niño

received the highest average yield.

169

Future climate simulations were also conducted for Banisogosogo, Koronivia and

Nacocolevu under PCCSP A1B and A2 emission scenario. For future simulations,

the number of days for tuber initiation increased while the tuber fresh and dry weight

decreased. It was also noticed that the highest LAI were found under 2090 potential

medium and high emission scenario. Under Banisogosogo future climate simulations,

no tuber yield was noticed under 2055 potential simulation and 2090 simulations for

A1B and A2 emission scenario. Koronivia simulations indicated that no tuber yield

was noticed under 2090 potential simulation for A1B and A2 emission scenario

while under Nacocolevu simulations, the tuber yield was possible under all emission

scenario. One major finding from these simulations is that the planting depth should

be 1.5 cm to obtain maximum tuber yield.

Current climate and future climate simulations were also conducted for Sebago and

Russet Burbank variety. The simulations indicated that Sebago gave higher yields

under current and future climate scenario whereas Russet Burbank gave negligible

yield under future climate scenario. However, more research needs to be conducted

with these varieties as the model has not been calibrated for these cultivars and the

genetic coefficient for Sebago and Russet Burbank may need modification to capture

high heat and low photoperiod conditions of the Tropics. In general, the DSSAT

SUBSTOR Potato model indicated that the growth and yield of potato under PCCSP

future climate scenarios had declined due to an increase in temperature, carbon-

dioxide concentration and changes in rainfall pattern.

The findings from this study will primarily assist potato farmers and agricultural

officers to identify the impacts of climate change and climate variability on potato

production in Fiji and to identify crop management strategies that will help farmers

maximise the tuber yield from the adverse impacts of climate change and climate

variability. Agronomists are interested in finding out ways and means of estimating

crop yield. Various modeling tools are used in decision making and planning in

agriculture. Crop models, therefore, can be used to forecast yield (potential and

attainable yield) for various crop with regards to season, year and geographical

location. Crop models also pay a crucial role in evaluating adaptation strategies.

These models can help optimise crop management practices, such as, sowing dates,

row spacing and nutrient management, whereby maximising yield. Another useful

quality of crop models is that it can quantify the impacts of climate change on crop

170

production, that is, the impact of increased carbon-dioxide and temperature on crop

development and yield. Apart from these, it can also manage climate change risk

through application of climate prediction, risk analysis and crop insurance.

It is recommended that the DSSAT SUBSTOR Potato model should be tested with

vigorous multi-location trials in Fiji to calibrate and validate the model outputs. Also,

the same study should be conducted with DSSAT v4.6 to look at possible sources of

error and to test model performance. Traditional environment knowledge should also

be integrated with DSSAT SUBSTOR Potato model to provide the best possible

outputs on crop growth, development and yield. The importance of crop models

cannot be under-valued as they will play an increasingly important role in research

understanding, crop management and policy intervention.

171

Bibliography

Abdrabbo M. A. A., Khalil A. A., Hassanien M. K. K. & Abou-Hadid A. F. 2010.

Sensitivity of Potato Yield to Climate Change. Journal of Applied Science

Research, 6, 751-755.

Agrawala S., Ota T., Risbey J., Hagenstad M., Smith J., Aalst M. v., Koshy K. &

Prasad B. 2003. Development and Climate Change in Fiji: Focus on Coastal

Mangroves. Organisation for Economic Co-operation and Development.

Agriculture Department of Fiji. 2009. Potato Production. The Agronomic Practices

of Producing Quality Potatoes

Ahmed M. & Fayyaz-ul-Hassan. 2011. APSIM and DSSAT Models as Decision

Support Tools. Perth, Australia: 19th International Congress on Modelling

and Simulation.

Alberta Agriculture Food and Rural Development Department. 2005. Botany of the

Potato Plant.

Allen E. J. 1978. Plant Density. In: Harris P. M. (ed.) The Potato Crop. London.

Allison M. F. & Allen E. J. 2004. Evaluation of Soil Nitrogen Supply System-

Opportunities for Further Improvemnets to Nitrogen Economy to GB Potato

Crop. British Potato Council.

Asadi M. E. & Clemente R. S. 2003. Evaluation of CERES-Maize of DSSAT Model

to Simulate Nitrate Leaching, Yield and Soil Moisture Content Under

Tropical Conditions. Food, Agriculture and Environment, 3 and 4, 270-276.

Asian Development Bank. 2009. Building Climate Resilience in the Agriculture

Sector in Asia and the Pacific [Online]. Available:

http://www.adb.org/Documents/Books/Building-Climate-Resilience-

Agriculture-Sector/Building-Climate-Resilience-Agriculture-Sector.pdf

[Accessed 12th March 2013].

Australian Bureau of Meterology & CSIRO. 2011. Climate Change in the Pacific:

Scientific Assessment and New Research.

Autar M. L. 2009. Potato Reserach and Production in Fiji. Ministry of Primary

Industries Fiji.

Ayas S. & Korukçu A. 2010. Water-Yield Relationships in Deficit Irrigated Potato.

Journal of Agricultural Faculty of Uludag University, 24, 23-36.

172

Baethgen W. E. 1998. El Niño and La Niña Impacts in Southeastern South America.

Review on the Causes and Consequences of Cold Events: A La Niña Su

mmit., NCAR.

Baethgen W. E. & Magrin G. O. 2000. Applications of Climate Forecasts in the

Agricultural Sector of South East South America. International Forum on

Climate Prediction, Agriculture and Development. New York: International

Research Institute for Climate Prediction

Bala S. K. & Islam A. K. M. S. 2008. Estimation of Potato Yield In and Around

Munshiganj Using Remopte Sensing NDVI Data. Bangladesh: Institute of

Water and Food Management.

Bindi M. 2008. Gain, Losses as Climate Change [Online]. Available:

http://www.potato2008.org/en/perspectives/bindi.html [Accessed 27th

February 2012].

Bindi M., Hacour A., Vandermeiren K., Craigon J., Ojanpera K., Sellden G., Hogy

P., Finnan J. & Fibbi L. 2002. Chrolophyll Concentration of Potatoes Grown

Under Elevated Carbon Dioxide and/or Ozone Concentrations. European

Journal of Agronomy, 17, 319- 335.

Blanco F. F. & Folegatti M. V. 2003. A New Method for Estimating the Leaf Area

Index of Cu cmber and Tomato Plants. Horticulture Brasileria, 21, 666-669.

Boer G. J., Stouffer R. J., Dix M., Noda A., Senior C. A., Raper S., Yap K. S.,

Cubasch U. & Meehl G. A. 2001. Chapter 9: Projections of Future Climate

Change. In: Kim J. W. & Stone J. (eds.) Climate Change 2001: The Scientific

Basis: Contribution of Working Group I to the Third Assessment Report of

the Intergovernmental Panel .

Bohl W. H. 2006. Tuber Tip: Plant at Correct Depth. Idaho Potato Conference

Idaho.

Bonnel E. 2008. Potato Breeding: A Challenge, As Ever! Potato Research, 51, 327-

332.

Boote K. J. 1999. Concepts for Calibrating Crop Growth Models. In: Hoogenboom

G., Wilkens P. W. & Tsuji G. Y. (eds.) DSSAT Version 3.

Borah M. N. & Milthorpe F. L. 1992. Growth of the Potato as Influenced by

Temperature. Indian Journal of Plant Physiology, 5, 53- 72.

Borrower. 2005. Summary Initial Environmental Examination and Environmental

Imapct Managment and Mitigation Plan

173

Bosello F. & Zhang J. 2005. Assessing Climate Change Impacts: Agriculture. CIP –

Climate Impacts and Policy Division. Working Paper N.02.2007.

Bouman B. A. M., H. Van Keulen, Laar H. H. V. & R. Rabbinge. 1996. The School

of Wit Crop Growth Simulation Models: Pedigree and Historical Overview.

Agricultural Systems, 52, 171-198.

Bowen A., Cabrera H., Barrera V & Baigorri G. 1997. CIP Programme Report

1997-1998.

Bowen W., Cabrera H., V.Barrara & G.Baigorria. 1998. Simulating the Response of

Potato to Applied Nitrogen. CIP.

Bowen W. T. 2003. Water Productivity and Potato Cultivation. International

Fertilizer Development Center, Muscle Shoals, Alabama, USA and

International Potato Center (CIP), Quito, Ecuador.

Boyd N., Gordon R. & Martin R. 2002. Relationship Between Leaf Area Index and

Ground Cover in Potato Under Different Management Conditions. Potato

Research, 45, 117-129.

Bradley A. 2009. Potato (Solanum tuberosum) [Online]. Available:

http://bioweb.uwlax.edu/bio203/s2009/bradley_adam/Classification.htm

[Accessed 23 Mrach 2012].

Bradshaw J. E. 2007. Chapter 8: Potato-Breeding Strategy. In: Vreugdenhil D. (ed.)

Potato Biology and Biotechnology. Advances and Perspectives. Scottish Crop

Research Institute, Invergowrie, Dundee DD2 5DA, United Kingdom:

Elsevier Publication.

British Potato Council. 2012. Soil Management for Potatoes. British Potato Council.

Bruhn D. 2002. Plant Respiration and Climate Change. Botanical Institute,

University of Copenhagen.

Bucher M. & Kossmann J. 2007. Molecular Physiology of the Mineral Nutrition of

the Potato. In: Vreugdenhil D. (ed.) Potato Biology and Biotechnology:

Advances and Perspectives.

Bulatewicz T., Jin W., Staggenborg S., Lauwo S., Miller M., Das S., Andresen D.,

Peterson J., Steward D. R. & Welch S. M. 2009. Calibration of a Crop Model

to Irrigated Water Use Using a Genetic Algorithm. Hydrology and Earth

System Sciences, 13, 1467-1483.

Bureau of Statistics. 2011. Fiji's Gross Domestic Product (GDP) 2010. Suva, Fiji.

174

Burlingame B., Mouille B. & Charrondiere R. 2009. Nutrients, Bioactive Non-

Nutrients and Anti- Nutrients in Potatoes. Journal of Food Composition and

Analysis, 22, 494- 502.

Burton G. W. 1981. Challenge in Stress Physiology in Potato. Americal Journal of

Potato Research, 1, 3-13.

Chakrabarti B. 2005. Crop Simulation Models. New Delhi.

Changchui H. Year. Welcome and Introductory Statement. In: Workshop to Co

mmemorate the International Year of Potato, 2008 BAngkok, Thailand.

Cheeroo-Nayamuth B. 1999. Crop Modelling Simulation/Overview. AMAS, 11-26.

Chiru S. C., Olteanu G. & Struik P. C. 2008. Preface to the Special Issue 51 (3/4):

Potato in a Changing World. Potato Research, 51, 215- 216.

CIA World Factbook. 2012a. Fiji [Online]. Available:

https://www.cia.gov/library/publications/the-world-factbook/geos/fj.html

[Accessed 20th December 2012].

CIA World Factbook. 2012b. GDP Composition by Sector [Online]. Available:

https://www.cia.gov/library/publications/the-world-factbook/fields/2012.html

[Accessed 13 May 2012].

Clarke J. M., Whetton P. H. & Hennessy K. J. 2011. Providing Application-Specific

Climate Projections Datasets: CSIRO’s Climate Futures Framework. Climate

Adaptation Flagship, CSIRO Marine and Atmospheric Research, Aspendale,

Victoria, Australia.

Coelho D. T. & Dale R. F. 1980. An energy-crop growth variable and temperature

function for predicting corn growth and development. Planting to silking.

Agronomy Journal, 72, 503-510.

Cooke L. R., Schepers H. T. A. M., Hermansen A., Bain R. A., Bradshaw N. J.,

Ritchie F., Shaw D. S., Evenhuis A., Kessel G. J. T., Wander J. G. N.,

Andersson B., Hansen J. G., Hannukkala A., Naerstad R. & Nielsen B. J.

2011. Epidemiology and Integrated Control of Potato Late Blight in Europe.

Potato Research, 54, 183- 222.

Cooley T. F. 1997. Calibrated Models. University of Rochester.

Cortbaoui R. 1988. Planting Potatoes. Lima, Peru: International Potato Center.

Craigon J., Fangmeier A., Jones M., Donnelly A., Bindi M., Te mmerman L. D.,

Persson K. & Ojanpera K. 2002. Growth and Marketable-Yield Responses of

175

Potato to Increased CO2 and Ozone. European Journal of Agronomy, 17,

273- 289.

Curwen D. 1993. Water Management. In: Rowe R. C. (ed.) Potato Health

Management.

Darwin P. 2001. Climate Change and Food Security. United States Department of

Agriculture.

Davies W. & J. J. Z. 1991. Root Signals and Regulation of Growth and Developmet

of Plants in Drying Soils. Annual Review of Plant Physiology

Plant Molecular Biology, 42, 55- 76.

Decagon Devices. 2006. AccuPAR PAR/LAI ceptometer Model LP-80 Operator’s

Manual Version 10.

Dessai S., Hulme M., Lempert R. & Pielke-Junior R. 2009. Climate Prediction: A

Limit to Adaptation? In: Adger W. N., Lorenzoni I. & O’Brien. K. L. (eds.)

Adapting to Climate Change: Thresholds, Values, Governance,. Cambridge

University Press.

Donnelly A., Lawsonb T., Craigon J., Black C. R., Colls J. J. & Landon G. 2001.

Effects of Elevated CO2 and O3 on Tuber Quality in Potato (Solanum

tuberosum L.). Agriculture, Ecosystem and Environment, 87, 273- 285.

Dwelle R. B. & Love S. L. 1993. Potato Growth and Development. In: Rowe R. C.

(ed.) Potato Health Management,.

Easterling W. E., X i. C., Haysl C., Brandle J. R. & Zhang H. 1996. Improving the

Validation of Model-Simulated Crop Yield Response to Climate Change: An

Application to the EPIC Model. Climate Research, 6, 263-273.

Ewing E. E. 1981a. Heat Stress and the Tuberization Stimulus. Americal Journal of

Potato Research, 58, 31–49.

Ewing E. E. 1981b. Heat Stress and Tuberisation Stimulus. Americal Potato Journal,

58, 31-49.

Ewing E. E. 1981c. Heat Stress and Tuberisation Stimulus. American Potato

Journal, 58, 31-49.

Ewing E. E. 1997. Potato. The Physiology of Vegetable Crops. CAB International,

Wallingford, UK.

Ewing E. E. & Struik P. C. 1992. Tuber formation in potato: Induction, initiation and

growth. Journal of Horticulture, 14, 89-198.

176

Fangmeier A., Te mmerman L. D., Black C., Persson K. & Vorne V. 2002. Effects of

elevated CO2 and/or ozone on nutrient concentrations and nutrient uptake of

potatoes. European Journal of Agrnonomy, 17, 353-368.

Farming and Wildlife Advisory Group. 2012. Environmental Guidance for Potato

Production.

Fernie A. R. & Willmitzer L. 2001. Molecular and Biochemical Triggers of Potato

Tuber Development. Plant Pathology, 127, 1459- 1465.

Fiji Meteorological Services. 2013a. 2012/13 Tropical Cyclone Season Outlook in

the Regional Specialised Meteorological Centre Nadi – Tropical Cyclone

Centre Area of Responsibility Nadi, Fiji Islands: Fiji Meteorological

Services.

Fiji Meteorological Services. 2013b. Fiji Climate Outlook. April to June and July to

September 2013. Fiji Meterological Services.

Fiji Meteorological Services. 2013c. Fiji Climate Summary May 2013. Nadi Airport,

Fiji: Fiji Meteorological Services.

Fiji Meterological Services. 2010. El Niño Southern Oscillation (ENSO) and Its

Impact on Fiji.

Fiji Meterological Services. 2012. Fiji's Climate [Online]. Nadi, Fiji. Available:

http://www.met.gov.fj/climate_fiji.html [Accessed 13 May 2012].

Fiji Times Online. 2010. Potato Scheme to Reduec Imports. Fiji Times.

Finan T. 1999. Drought and Demagoguery: A Political Ecology of Climate

Variability in Northeast Brazil. Workshop Proceedings from Public

Philosophy, Environment, and Social Justice. October 21-22 1999. New

York, USA.: Carnegie Council on Ethics and International Affairs. .

Firman D. M. & Daniels S. J. 2011. Factors Affecting Tuber Numbers Per Stem

Leading to Improved Seed Rate Recommendations. Cambridge University

Farm.

Firman D. M., O'Brien P. J. & Allen E. J. 1991. Leaf and Flower Initiation in Potato

(Solanum tuberosum) Sprouts and Stems in Relation to Number of Nodes and

Tuber Initiation. The Journal of Agricultural Science, 117, 61-74.

Firman D. M., O'Brien P. J. & Allen E. J. 1992. Predicting the Emergence of Potato

Sprouts. The Journal of Agricultural Science, 118, 55-61.

Fleisher D. H., Cavazzoni J., Giacomelli G. A. & C.Ting K. 2000. Adaptation of

SUBSTOR for Hydroponic, Controlled Environment White Potato

177

Production. ASAE Annual International Meeting Milwaukee, Wisconsin,

USA.

Fleisher D. H., Cavazzoni J., Giacomelli G. A. & Ting K. C. 2003. Adaptation of

SUBSTOR for Controlled Environment Potato Production with Elevated

Carbon Dioxide. American Society of Agricultural Engineers, 46, 531-538.

Fleisher D. H., J.Timlin D. & Reddy V. R. 2006a. Temperature Influence On Potato

Leaf And Branch Distribution And On Canopy Photosynthetic Rate. Agron.

J., 98, 1442-1452.

Fleisher D. H., Timlin D. J. & Reddy V. R. 2006b. Temperature Influence on Potato

Leaf and Branch Distribution and on Canopy Photosynthetic Rate. Agronomy

Journal, 98, 1442-1452.

Fontes P. C. R., Braun H., Busato C. & Cecon P. R. 2010. Economic Optimum

Nitrogen Fertilization Rates and Nitrogen Fertilization Rate Effects on Tuber

Characteristics of Potato Cultivars. Potato Research, 53, 167- 179.

Food and Agricultural Organisation of the United Nations. 2008. International Year

of Potato 2008. A New Light on a Hidden Treasure. Rome: Food and

Agricultural Organisation of the United Nations.

Food and Agriculture Organisation. 2008. Buried Treasure: The Potato. Agriculture,

Biosecurity, Nutrition and Consumer Protection Department Food and

Agriculture Organization of the United Nations.

Food and Agriculture Organisation. 2009. Natures Way Cooperative (Fiji) Ltd : A

Case Study of Agriculture for Growth in the Pacific. Food and Agriculture

Organisation.

Food and Agriculture Organisation of the United Nations. 2002. The State of Food

Insecurity in the World 2001. Rome: FAO.

Food and Agriculture Organisation of the United Nations. 2007. Adaptation to

Climate Change in Agriculture, Forestry and Fisheries:Perspective,

Framework and Priorities. Rome.

Food and Agriculture Organisation of the United Nations. 2008a. Climate Change

and Food Security: A Framework Document. Rome: Food and Agriculture

Organisation of United Nations.

Food and Agriculture Organisation of the United Nations. 2008b. The Global Potato

Economy [Online]. Available: http://www.potato2008.org [Accessed 27th

February 2012].

178

Food and Agriculture Organisation of the United Nations. 2008c. International Year

of Potato 2008. New Light on a Hidden Treasure. Rome: Food and

Agriculture Organisation of the United Nations.

Food and Agriculture Organisation of the United Nations. 2008d. Potato and Climate

Change [Online]. Available: www.potato2008.org [Accessed 27th February

2012].

Food and Agriculture Organisation of the United Nations. 2008e. Potato and Food

Price Inflation [Online]. Available: www.potato2008.org [Accessed 27th

February 2012].

Food and Agriculture Organisation of the United Nations. 2008f. Potato and Gender

[Online]. Available: www.potato2008.org [Accessed 27th February 2012].

Food and Agriculture Organisation of the United Nations. 2008g. Potato and Soil

Conservation [Online]. Available: www.potato2008.org [Accessed Accessed

27th February 2012].

Food and Agriculture Organisation of the United Nations. 2008h. Potato Pest and

Disease Management [Online]. Available: www.potato2008.org [Accessed

27th February 2012].

Food and Agriculture Organisation of the United Nations. 2008i. Potatoes, Nutrition

and Diet

[Online]. Available: www.potato2008.org [Accessed 27th February 2012].

Food and Agriculture Organization of the United Nations. 2007. Adaptation to

Climate Change in Agriculture, Forestry and Fisheries:Perspective,

Framework and Priorities. Rome.

Food and Agriculture Organization of the United Nations. 2008. Potato and Climate

Change [Online]. Available: www.potato2008.org [Accessed 27th February

2012].

Fraisse C. W., Sudduth K. A. & Kitchen N. R. 2001. Calibration of the Cere-Maize

Model for Simulating Site Specific Crop Development and Yield on Claypan

Soils. Applied Engineering in Agriculture, 17, 547-556.

Garner N. & Blake J. 1989. The Induction and Development of Potato Microtubers

in vitro on Media Free of Growth Regulating Substances. Annual of Botany,

63, 663-674.

Gawander J. 2007. Impact of Climate Chnage on Sugar-Cane Production in Fiji.

W.M.O.

179

Gawander J. S., Lal M. & Rounds P. 2012. Baseline Climatology of Four Sugar Mill

Stations in Fiji and Current climatic Trends. Sugar Research Institution of

Fiji.

GEF, GIZ & SPREP. 2009. Pacific Adaptation to Climate Change. Fiji Islands.

Report on In-Country Consultations.

Geremew E. B., Steyn J. M. & Annandale J. G. 2008. Comparison Between

Traditional and Scientific Irrigation Scheduling Practices for Furrow Irrigated

Potatoes (Solanum tuberosum L.) in Ethiopia. South African Journal of Plant

and Soil, 25, 42-48.

Godwin D. C. & Singh U. 1998. Nitrogen Balance and Crop Response to Nitrogen in

Upland and Lowland Cropping Systems. In Understanding Options of

Agricultural Production, The Netherlands, Kluwer Academic Publishers and

International Consortium for Agricultural Systems Applications.

Gonzalez-Sanpedro M. C., Toan T. L., Moreno J., Kergoat L. & Rubio E. 2009.

Seasonal Variations of Leaf Area Index of Agricultural Fields Retrieved from

Landsat Data.: CESBIO.

GoogleEarth. 2012. Location of Experimental Site (Banisogosogo) and Simulation

Sites (Koronivia and Sigatoka). GoogleEarth.

Gordon R., Brown D. & Dixon M. 1997. Estimating Potato Leaf Area Index for

Specific Cultivars. Potato Research, 40, 251-266.

Government of Alberta. 2011. Agri-Facts. Practical Information for Albert's

Agriculture Industry. Irrigation Scheduling for Potato in Southern Alberta. .

Government of the Republic of Fiji. 2012. Republic of Fiji National Climate Change

Policy. Suva, Fiji: Secretariat of the Pacific Community.

Graeff S., Link J., Binder J. & Claupein W. 1999. Crop Models as Decision Support

Systems in Crop Production. University Hohenheim, Crop Science (340a)

Germany.

Gregory P. J., Ingram J. S. I. & Brklacich M. 2005. Climate Change and Food

Security. Philosophical Transaction of The Royal Society, 360, 2139- 2148.

Griffin T. S., Johnson B. S. & Ritchie J. T. 1993. A Simulation Model for Potato

Growth and Development: SUBSTOR-Potato Version 2.0. Honolulu, Hawaii:

Department of Agronomy and Soil Science, College of Tropical Agriculture

and Human Resources, University of Hawaii.

180

Gri mm S. S., J. W. Jones, K. J. Boote & Hesketh J. D. 1993. Parameter Estimation

for Predicting Flowering Date of Soybean Cultivars. Crop Science, 33, 137-

144.

Grove R. H. & Chappell J. 2000. El Nino-History and Crisis. Studies from Asia-

Pacific Region.

Haase N. U. 2008. Healthy Aspects of Potatoes as Part of the Human Diet. Potato

Research, 51, 239- 258.

Hagman J. E., Mårtensson A. & Grandin U. 2009. Cultivation Practices and Potato

Cultivars Suitable for Organic Potato Production. Potato Research, 52, 319-

330.

Hannukkala A. O., Kaukoranta T., Lehtinen A. & Rahkonen A. 2007. Late-Blight

Epidemics on Potato in Finland, 1933–2002; Increased and Earlier

Occurrence of Epidemics Associated

with Climate Change and Lack of Rotation. Plant Pathology, 56, 167- 176.

Harris P. M. 1983. The Potential for Producing Root and Tuber Crops from Seeds.

Lima, Peru.

Hassan A. A., Sarkar A. A., Ali M. H. & Karim N. N. 2002. Effect of Deficit

Irrigation at Different Growth Stages on the Yield of Potato Pakistan Journal

of Biological Sciences, 5, 128-134.

Haverkort A. J. 1990. Ecology of Potato Cropping Systems in Relation to Latitude

and Altitude. Agricultural Systems, 32., 251–272.

Haverkort A. J. 2007a. Potato Crop Response to Radiation and Daylength. In:

Vreugdenhil D., Bradshaw J., Gebhardt C., Makerron D. K. L., Govers F.,

Taylor M. A. & A.Ross H. (eds.) Potato Biology and Biotechnology:

Advances and Perspectives.

Haverkort A. J. 2007b. Potato Crop Response to Radiation and Daylength. In:

Vreugdenhil D. (ed.) Potato Biology and Biotechnology: Advances and

Perspectives.

Haverkort A. J., A V., C G. A. & J U. P. W. 2004. Potato-Zoning: A Decision

Support System on Expanding the Potato Industry Through Agro-Ecological

Zoning Using the LINTUL Simulation Approach. In: L M. D. K. & J H. A.

(eds.) Decision Support Systems in Potato Production: Bringing Models to

Practice. Wageningen: Wageningen Academic.

181

Haverkort A. J. & MacKerron D. K. L. 1995. Potato Ecology and Modelling of

Crops under Conditions Limiting Growth, Kluwer, Dordrecht.

Haverkort A. J. & Verhagen A. 2008. Climate Change and Its Repercussions for the

Potato Supply Chain. Potato Research, 51, 223- 237.

Hijmans R. J. 2003a. The Effect of Climate Change on Global Potato Production.

American Journal of Potato Research, 80, 271- 280.

Hijmans R. J. 2003b. The Effect of Climate Change on Global Potato Production.

American Journal of Potato Research, 80, 271-279.

Hofmann M. 2005. On the Complexity of Parameter Calibration in Simulation

Models. JDMS, 2, 217-226.

Hogy P. & Fangmeier A. 2009a. Atmospheric CO2 Enrichment Affects Potatoes: 1.

Aboveground Biomass Production and Tuber Yield. European Journal of

Agronomy, 30, 78- 84.

Hogy P. & Fangmeier A. 2009b. Atmospheric CO2 Enrichment Affects Potatoes: 2.

Tuber Quality Traits. European Journal of Agronomy, 30, 85- 94.

Holden N. M. & Brereton A. J. 2006. Adaptation of Water and Nitrogen

Management of Spring Barley and Potato as a Response to Possible Climate

Change in Ireland. Agricultural Water Management, 82, 297- 317.

Hoogenboom G. 2003. Crop Growth and Development., New York, The Haworth

Press.

Hoogenboom G., Tsuji G. Y., Pickering N. B., Curry R. B., Jones J. W., Singh U. &

Godwin D. C. 1985. Decision Support Sytem to Study Climate Chnage

Impacts on Crop Production. 59.

Hoogenboom G., Wilkens P. W. & Tsuji G. Y. 1999. Chapter 2. Field Methods.

DSSAT v3 Volume 4. University of Hawaii, Honolulu, Hawaii.

Huang Y., Yu Y., Zhang W., Sun W., Liu S., Jiang J., Wu J., Yu W., Wang. Y. &

Yang Z. 2009. Agro-C: A Biogeophysical Model for Simulating the Carbon

Budget of Agroecosystems. Agricultural and Forest Meteorology, 149, 106-

129.

Iglesias I., Rodrı´guez-Rajo F. J. & Me´ndez J. 2007. Evaluation of the Different

Alternaria Prediction Models on a Potato Crop in A Limia (NW of Spain).

Aerobiologia, 23, 27-34.

182

Index Mundi. 2012. Fiji Demographics Profile 2012 [Online]. Available:

http://www.indexmundi.com/fiji/demographics_profile.html [Accessed 20th

December 2012].

InfoResources Focus. 2008. Potatoes and Climate Change. In: Stäubli B., Wenger R.

& Dach. S. W. v. (eds.).

Intergovernmental Panel on Climate Change. 2001. Climate Change 2001: The

Scientific Basis. Contribution of Working Group I to the Third Assessment

Report of the Intergovernmental Panel on Climate Change. In: Houghton J.

T., Ding Y., Griggs D. J., Noguer M., Linden P. J. v. d., Dai X., K.Maskell &

Johnson C. A. (eds.). Cambridge University Press, Cambridge, United

Kingdom and New York, NY, USA,: Intergovernmental Panel on Climate

Change.

Intergovernmental Panel on Climate Change. 2007a. Climate Change 2007:

Synthesis Report.Contribution of WorkingnGroups I, II and III to the Fourth

Assessment Report of the Intergovernmental Panel on Climate Change. In:

Pachauri R. K. & Reisinger A. (eds.). Geneva, Switzerland:

Intergovernmental Panel on Climate Change.

Intergovernmental Panel on Climate Change. 2007b. Summary for

Policymakers.Climate Change 2007: The Physical Science Basis.

Contribution of Working Group I to the Fourth Assessment Report of the

Intergovernmental Panel on Climate Change In: Solomon S., Qin D.,

Manning M., Chen Z., Marquis M., Averyt K. B., M.Tignor & Miller H. L.

(eds.). Cambridge University Press, Cambridge, United Kingdom New York,

NY, USA.: Intergovernmental Panel on Climate Change,.

International Potato Center. 1999. Scenario Analyses with Process-Based Models.

International Potato Center.

International Potato Center. 2006. LIFE - SIM: Livestock Feeding Strategies

Simulation Models. Peru: Natural Resources Management Division.

International Potato Center. 2008a. Nutrition [Online]. Av La Molina 1895, La

Molina. Available: http://cipotato.org/potato/nutrition [Accessed 5th March

2012].

International Potato Center. 2008b. Pests and Disease [Online]. Av La Molina 1895,

La Molina. . Available: http://cipotato.org/potato/pests-and-disease [Accessed

5th March 2012].

183

Iqbal M. 1982. Potato Production in Fiji- Potentials and Contraints. Research

Division, Department of Agriculture, Sigatoka, Fiji.

Iqbal M. 1991. Optimising Potato Solanum tuberosum Plant Density and

Configuration Asd a Sole Crop or Sugarcane Intercrop. Degree of Doctor in

Philosophy, University of the South Pacific.

Irving D. B., Perkins S. E., Brown J. R., Gupta A. S., Moise A. F., Murphy B. F.,

Muir L. C., Colman R. A., Power S. B., Delage F. P. & Brown J. N. 2011.

Evaluating Global Climate Models for the Pacific Island Region. Climate

Research, 49, 169-187.

Jackson S. D. 1999. Multiple Signaling Pathways Control Tuber Induction in Potato.

Plant Physiology, 119, 1-8.

Jansky S. H., Jin L. P., Xie K. Y., Xie C. H. & Spooner D. M. 2009. Potato

Production and Breeding in China. Potato Research, 52, 57- 65.

Jones J. W., Boote K. J., Jagtap S. S., Hoogenboom G. & Wilkerson G. G. 1988.

SOYGRO v5.41: Soybean Crop Growth Simulation Model User’s Guide.

Florida: Agricultural Experiment Station Journal No.8304, University of

Florida: IFAS.

Jones J. W., Tsuji G. Y., Hoogenboom G., Hunt L. A., Thornton P. K., Wilkens P.

W., Imamura D. T., Bowen W. T. & Singh U. 1998. Decision support system

for agrotechnology transfer: DSSAT v3. In: Tsuji G., Hoogenboom G. &

Thornton P. (eds.) Understanding Options for Agricultural Production.

Springer Netherlands.

Jones J. W., Hoogenboom G., Porter C. H., Boote K. J., Batchelor W. D., Hunt L. A.,

Wilkens P. W., Singh U., Gijsman A. J. & Ritchie R. T. 2003. DSSAT

Cropping System Model. European Journal of Agronomy, 18, 235-265.

Jones J. W., Hoogenboom G., Wilkens P. W., Porter C. H. & Tsuji G. Y. 2010.

Decision Support System for Agrotechnology Transfer Version 4.0. Volume 4.

DSSAT v4.5: Crop Model Documentation, University of Hawaii, Honolulu,

HI.

Jovanovica Z., Stikic R., Vucelic-Radovica B., Milena Paukovica, Brocica Z.,

Matovica G., Rovcanina S. & Mojevicb M. 2010a. Partial Root-Zone Drying

Increases WUE, N and Antioxidant Content in Field Potatoes. European

Journal of Agronomy, 33, 124- 131.

184

Jovanovica Z., Stikic R., Vucelic-Radovica B., Paukovica M., Brocica Z., Matovica

G., Rovcanina S. & Mojevicb M. 2010b. Partial Root-Zone Drying Increases

WUE, N and Antioxidant Content in Field Potatoes. European Journal of

Agronomy, 33, 124- 131.

K.Persson, H.Danielsson, G.Sellde & H.Pleijel. 2003. The Effects of Tropospheric

Zone and Elevated Carbon Dioxide on Potato (Solanum tuberosum L.cv

.Bintje ) Growth and Yield. The Science of Total Environment, 310, 191- 201.

Kadaja J. & Tooming H. 2004. Potato Production Model Based on Principle of

Maximum Plant Productivity. Agricultural and Forest Meteorology, 127, 1-

16.

Khedher M. B. & Ewing E. E. 1985. Growth Analysis of Eleven Potato Cultivars

Grown in Greenhouse under Photoperiods With and Without Heat Stress.

American Potato Journal, 62, 537-554.

Kimball B. A. 1983. Carbon Dioxide and Agricultural Yield: An Assemblage and

Analysis of 430 Prior Observations. Agronomy Journal, 75, 779-788.

King B. A. & Stark J. C. 1997. Potato Irrigation Management, University of Idaho.

Kleinkopf G. E., Brandt T. L. & Olsen N. 2003. Physiology of Tuber Bulking. Idaho

Potato Conference.

Klemke T. & Moll A. 1990. Model for Simulation of Potato Growth from Planting to

Emergence. Agricultural Systems, 32, 295-304.

Knox D. J. W., Daccache D. A., Weatherhead D. E. K. & Stalham D. M. 2010.

Climate Change Impacts on the UK Potato Industry. Agriculture and

Horticulture Development Board.

Kooman P. L. & Haverkort A. J. 1995. Modelling Development and Grown of the

Potato Crop Influenced by Temperature and Daylength: LINTUL-POTATO.

In: Haverkort A. J. (ed.) Potato Ecology and Modeling of Crops Under

Conditions Limiting Grown.

Krupa S. V. & Kickert R. N. 1993. The greenhouse effect: the impacts of Carbon

dioxise (CO2), ultraviolet-B (UV-B) radiation and Ozone (O3) on vegetation

(crops). Vegetation, 104/105, 223-238.

Ku S.-B., Edward G. E. & Tanner C. B. 1977. Effects of Light, Carbon Dioxide, and

Temperature on Photosynthesis, Oxygen Inhibition of Photosynthesis, and

Transpiration in Solanum tuberosum. Plant Physiology, 59, 868-872.

185

Kumar R. 2011. Climate Trends and projections in Fiji. Nadi, Fiji: Fiji

Meteorological Services.

Kumar R. N., B.Sailaja & S.R.Voleti. 2000. Crop Modelling with Special Reference

to Rice Crop. Rajendranagar, Hyderabad: Rice Knowledge Management

Portal.

Kuruppuarachi D. S. P., Harris P. M., Si mmond L. & Gunasena H. P. M. 1989.

Potential Productivity of Intensively Irrigated Potatoes in Low Country

Regosol Belt of Sri Lanka. Tropical Agricultural Research, 1, 1-20.

Lal M. 2004. Climate Change in small island developing countries of South Pacific,

Fijian Studies. Special Issue on Sustainable Development, 2.

Lang N. S., Steven R. G., Thorton R. E., Pan W. L. & Victory S. 1999. Potato

Nutrient Management for Central Washington.

Lawson T., Craigon J., Black C. R., Colls J. J., Tulloch A.-M. & Landon G. 2001a.

Effects of Elevated Carbon Dioxide and Ozone on the Growth and Yield of

Potatoes (Solanum tuberosum) Grown in Open-top Chambers. Environmental

Pollution, 11, 479- 491.

Lawson T., Craigon J., Tulloch A.-M., Black C. R., Colls J. J. & Landon G. 2001b.

Photosynthetic Responses to Elevated CO2 and O3 in Field-Grown Potato

(Solanum tuberosum). Journal of Plant Physiology, 158, 309- 323.

Lawson T., Craigon J., Tulloch A. M., Black C. R., Colls J. J. & Landon G. 2001c.

Photosynthetic response to elevated carbon dioxide and ozone on the growth

and yield of potatoes (Solanum tuberosum L.) grown in open top chambers.

Environmental Pollution, 111, 479-491.

Legler D. M., Bryant K. J. & O’Brien J. J. 1999. Impact of ENSO-related Climate

Anomalies on Crop Yields in the U.S. Climatic Change, 42, 351-375.

Levy D. & Veilleux R. 2007. Adaptation of Potato to High Temperatures and

Salinity-A Review. American Journal of Potato Research, 84, 487-506.

Love S., Stark J. & Salaiz T. 2005. Response of Four Potato Cultivars to Rate and

Timing of Nitrogen Fertilizer. American Journal of Potato Research, 82, 21-

30.

Ludi E. 2009. Climate Change, Water and Food Security. Overseas Development

Institute.

Macfarlane D. C. 2009. Country Pasture/Forage Resource Profiles: Fiji. Food and

Agriculture Organisation.

186

MacKerron D. K. L. 2007. Mathematical Models of Plant Growth and Development.

In: Vreugdenhil D. (ed.) Potato and Biotechnology:Advances and

Perspectives.

MacKerron D. K. L. & Jefferies R. A. 1985. Observations on the Effects of Relief of

Later Water Stress in Potato. Potato Research, 28, 349–359.

MacKerron D. K. L. & Jefferies R. A. 1986. The Influence of Early Soil Moisture

Stress on Tuber Numbers in Potato. Potato Research, 29, 299–312.

MacKerron D. K. L. & Waister P. D. 1985. A Simple Model of Potato Growth and

Yield. 1. Model Development and Sensitivity Analysis. Agricultural and

Forest Meteorology, 34, 241-52.

Magrin G. O., M.O.Grondona, I.Travasso M., Boullón D. R., Rodríguez G. R. &

Messina C. D. 1999. ENSO Impacts on Crop Production in the Argentina’s

Pampas Region. 10th Symposium on Global Change Studies.

Mahmood M. M., Farooq K., Hussain A. & Sher R. 2002. Effect of Mulching on

Growth and Yield of Potato Crop. Asian Journal of Plant Science, 1, 132-

133.

Makino A. & Mae T. 1999. Photosynthesis and Plant Growth at Elevated Levels of

CO2. Plant and Cell Physiology, 40, 999-1006.

Manrique L. A. & Bartholomew D. P. 1991. Growth and Yield Performance of

Potato Grown at Three Elevations in Hawaii: II. Dry Matter Production and

Efficiency of Partitioning. Crop Science, 31, 367-372.

Manrique L. A., Kiniry J. R., Hodges T. & Axness D. S. 1991. Dry Matter

Production and Radiation of Potato. Crop Science, 31, 1044- 1049.

Maps of World. 2010. Fiji Geography [Online]. Available:

http://www.mapsofworld.com/fiji/geography/ [Accessed 20th December

2012].

Maps of World. 2011. Fiji Location [Online]. Available:

http://www.mapsofworld.com/fiji/geography/ [Accessed 20th December

2012].

Maps of World. 2012. Fiji [Online]. Available: http://www.vidiani.com/?p=11392


Medany M. 2006. Assessment of the Impact of Climate Change and Adaptation on

Potato Production. Egypt: Central Laboratory for Agricultural Climate.

187

Mendelsohn R. & Dinar A. 1999. Climate Change, Agriculture, and Developing

Countries: Does Adaptation Matter? The World Bank Research Observer, 14,

277-293.

Menzel C. M. 1980. Tuberisation in Potato at High Temperatures: Responses to

Gibberellin and Growth inhibitors. Annals of Botany, 46, 259-265.

Menzel C. M. 1983a. Tuberisation in Potato at High Temperatures: Gibberellin

Content and Transport from Buds. Annals of Botany, 52, 697-702.

Menzel C. M. 1983b. Tuberisation in Potato at High Temperatures: Interaction

Between Shoot and Root Temperatures. Annals of Botany, 52, 65-69.

Menzel C. M. 1985. Tuberisation in Potato at High Temperatures: Interaction

between Temperature and Irradiance. Annals of Botany, 555, 35-39.

Midmore D. J. 1984a. Potato (Solanum spp.) in the hot tropics I. Soil temperature

effects on emergence, plant development and yield. Field Crops Research, 8,

255-271.

Midmore D. J. 1984b. Potato (Solanum spp.) in the Hot Tropics. I. Soil Temperature

Effects on eEmergence, Plant Development and Yield. Field Crops Research,

8, 255-271.

Midmore D. J. 1988. Potato (Solanum spp) in the Hot Tropics VI. Plant Population

Effectts on Soil Temperature, Plant Development and Tuber Yield. Fields

Crop Research, 19, 183-200.

Midmore D. J. & Prange R. K. 1991. Sources of Heat Tolerance Amongst Potato

Cultivars, Breeding Lines, and Solanum Species. Euphytica, 55, 235-245.

Midmore D. J. & Rhoades R. E. 1988. Applications of Agrometeorology to the

Producytion of Potatoes in Warm Tropics. Acta Horticulture, 214, 103-132.

Miglietta F., Magliulo V., Bindi M., Cerio L., Vaccaxi F. P., Loduca V. & Peressotti

A. 1998. Free Air C02 Enrichmnet of Potato (Solanum tuberosum L.):

Development, Growth and Yield. Global Change Biology, 4, 163- 172.

Möller K. & Reents H.-J. 2007. Impact of Agronomic Strategies (Seed Tuber Pre-

sprouting, Cultivar Choice) to Control Late Blight (<i>Phytophthora

infestans</i>) on Tuber Growth and Yield in Organic Potato (Solanum

tuberosum L.) Crops. Potato Research, 50, 15-29.

Mudaliar L. 2007. Need for Sustainable Agricultural Development in Fiji via

Engineering Technologies. 3rd Session of UNAPCAEM Meeting. Beijing.

188

Myhre D. L. 1956. Irrigation of Sebago Potatoes at Hastings, Florida. Florida

Agricultural Experiment Station Journal Series, 544, 202-204.

Nasseri A. & Baramloo R. 2009. Potato Cultivar Marfuna Yield and Water Use

Efficiency Responses to Early-Season Water Stress. International Journal of

Agriculture and Biology, 11, 201-204.

National Geospatial-Intelligence Agency. 2012a. Koronivia: Fiji [Online]. Available:

http://www.geographic.org/geographic_names/name.php?uni=-

1750779&fid=1806&c=fiji [Accessed 25th December 2012].

National Geospatial-Intelligence Agency. 2012b. Rakiraki: Fiji [Online]. [Accessed

25th December 2012].

Nausori Town Council. 2011. Weather and Climate. Nausori, Fiji: Nausori Town

Council.

Ng N & Loomis R. S. 1984. Simulation of Growth and Yield of the Potato Crop.

Nowak J. & Colborne D. 1989. In vitro Tuberization and Tuber Proteins as

Indicators of Heat Stress Tolerance in Potato. American Potato Journal, 66,

35-45.

Olesen J. E. & Bindi M. 2002. Consequences of Climate Chnage for European

Agricultural Productivity, Land Use and Policy. European Journal of

Agronomy, 16, 239- 262.

Oliveira C. A. D. S. 2000. Potato Crop Growth as Affected by Nitrogen and Plant

Density. 35, 939-950.

Olsen N. & Nolte P. 2004. Maximising Storage Options for Seed Potatoes.

University of Idaho.

Orlove B. S., Chiang J. C. H. & Cane M. A. 2000. Forecasting Andean Rainfall and

Crop Yield from the Infuence of El Nino on Pleiades visibility. Nature, 402.

Ortiz R. A. 1998. Crop Simulation Models as an Educational Tool. In: Tsuji G.,

Hoogenboom G. & Thornton P. (eds.) Understanding Options for

Agricultural Production. Springer Netherlands.

Ovchinnikova A., Krylova E., Gavrilenko T., Smekalova T., Zhuk M., Knapp S. &

Spooner D. 2011. Taxonomy of Cultivated Potatoes (Solanum section Petota:

Solanaceae). Botanical Journal of Linnean Society, 165, 107-155.

Ozturk E., Kavurmacı Z., Kara K. & Polat T. 2010. The Effects of Different Nitrogen

and Phosphorus Rates on Some Quality Traits of Potato. Potato Research, 50,

309- 312.

189

Pacific Climate Change Science Program. 2010. Pacific Climate Change Science

Program Brochure Pacific Climate Change Science Program.

Pacific Climate Change Science Program. 2011a. Current and Future Climate of the

Fiji Islands. Fiji Meteorological Service, Australian Bureau of Meteorology,

Commonwealth Scientific and Industrial Research Organisation (CSIRO).

Pacific Climate Change Science Program. 2011b. PCCSP: Climate Futures. User

Guide- Basic. Pacific Climate Change Science Program.

Pacific Climate Change Science Programe. 2011a. Chapter 5. Fiji islands. Climate

Change in the Pacific: Scientific Assessment and New Research. Pacific

Climate Change Science Programe.

Pacific Climate Change Science Programe. 2011b. PCCSP Climate Futures. Pacific

Climate Change Science Program.

Pacific Islands Climate Change Assistance Programme & Fiji Country Team. 2005.

Climate Change. The Fiji Islands Response. Fijiʼs First National

Communication Under the Framework Convention on Climate Change 2005.

Suva, Fiji islands.

Pandey D. S. K. 2008. Pottao Research and Priorities in Asia ans the Pacific Region.

In: Papademetriou M. K. (ed.). Bangkok, Thailand: FAO.

Paranjape, K. 2013. Sanskrit Quotes. [Online] Avaliable: http://sanskrit-

quote.blogspot.com/ [Accessed 02 September 2013].

Parry M. 2007. The Implications of Climate Change for Crop Yields, Global Food

Supply and Risk of Hunger. 4, 1-16.

Patel N., Kumar P. & Singh N. 2008. Performance Evaluation of AQUACROP in

Simulating Potato Yield under Varying Water Availability Conditions. New

delhi: Indian Agricultural Research Institute.

Pereira A. B. & C.C. Shock. 2006. Development of irrigation best management

practices for potato from a research perspective in the United States.

Pereira A. B., Nova N. A. V., Ramos V. J. & Pereira A. R. 2008. Potato Potential

Yield Based on Climatic Elements and Cultivar Characteristics. Bragantia

67, 327-334.

Persson K., Danielsson H., Sellde G. & Pleijel H. 2003. The Effects of Tropospheric

Zone and Elevated Carbon Dioxide on Potato (Solanum tuberosum L.cv

.Bintje ) Growth and Yield. The Science of Total Environment, 310, 191- 201.

190

Plauborg F., Abrahamsen P., Gjettermann B., Mollerup M., Iversena B. V., Liuc F.,

Andersena M. N. & Hansenb S. 2010. Modelling of Root ABA synthesis,

Stomatal Conductance, Transpiration and Potato Production Under Water

Saving Irrigation Regimes. Agricultural Water Management, 98, 425–439.

Posadas A., Rojas G., Málaga M., Mares V. & Quiroz R. A. 2008. Partial Root-Zone

Drying: An Alternative Irrigation Management to Improve the Water Use

Efficiency of Potato Crops. International Potato Center.

Poveda G., Jaramillo A., Gil M. M., Quiceno N. & Mantilla R. I. 2001. Seasonality

in ENSO-Related Precipitation, River Discharges, Soil Moisture, and

vVegetation Index in Colombia. Water Resources Reseach, 37, 2169-2178.

Pumijumnong N. & Arunrat N. 2012. Reliability and Evaluation of the Potential of

the EPIC Model to Estimate Rice Yields in Thailand. Agricultural Science

Research Journals, 2, 614-622.

Quity G. 2012. Assesing the Ecological Impacts of Climate Change on Roots Crop

Production in High Islands:A Case Study in Santa Isabel, Solomons Islands.

Master of Science in Climate Change, The University of the South Pacific.

Raisanen J. 2007. How Reliable are Climate Models? Tellus, 59A, 2-29.

Reddy V. R. & Hodges. 2000. Climate Change and Global Crop Productivity.

Rex B. L. & Mazza G. 1989. Cause, control and detection of hollow heart in

potatoes: a review. Amercial Journal of Potato Research, 66, 165–183.

Reynolds M. P. & Ewing E. E. 1989. Effects of High Air and Soil Temperature

Stress on Growth and Tuberization in Solanum tuberosum. Annals of Botany,

64, 241–247.

Rezzoug W., Gabrielle B., Suleiman A. & Benabdeli K. 2008. Application and

Evaluation of the DSSAT-Wheat in the Tiaret Region of Algeria. African

Journal of Agricultural Research, 3, 284-296.

Ritchie J. T. 1981a. Soil Water Availability. Plants and Soil, 58, 327-338.

Ritchie J. T. 1981b. Water Dynamics in the Soil-Plant-Atmosphere. Plants and Soil,

58, 81-96.

Ritchie J. T., Singh U., Godwin D. C. & Bowen W. T. 1998. Cereal Growth,

Development and Yield. In: Tsuji G., Hoogenboom G. & Thornton P. (eds.)

Understanding Options for Agricultural Production. Springer Netherlands.

Rivington M. & Koo J. 2010. Climate Change, Agriculture and Food Security

Challenge Program: Report on the Meta-Analysis of Crop Modeling for

191

Climate Change and Food Security Survey. Consultative Group on

International Agricultural Research /Earth Systems Science Partnership

sponsored.

Roberts M. J., Schlenker W. & Eyer J. 2013. Agronomic Weather Measures in

Econometric Models of Crop Yield with Implications for Climate Change.

Americal Journal of Agricultural Economics, 95, 236-243.

Rosenzweig C. & Hillel D. 1998. Climate Change and ,the Global Harvest Potential

Impacts of the Greenhouse Effect on Agriculture., Oxford University Press,

New York.

Rosenzweig C. & Parry M. L. 1994. Potential Impact of Climate Change on World

Food Supply. Nature, 367, 133-138.

Rosenzweig C., Phillips J., Goldberg R. & Hodges’ J. C. T. 1996. Potential Impacts

of Climate Change on Citrus and Potato Production in the US. Agricultural

Systems, 56, 455- 479.

Rosillo-Calle D. F. & Woods J. 2003. Individual Country Biomass Resource

Assessment Profiles for Fiji, Kiribati, Samoa, Tonga, Tuvalu and Vanuatu.

Roth O., Derron J., Fischlin A., Nemecek T. & Ulrich M. 1991. Implementation and

Parameter Adaptation of a Potato Crop Simulation Model Combined with a

Soil Water Subsystem, International Agricultural Centre, Wageningen, The

Netherlands.

Sarquis J. I., H. Gonzalez & Bernal-Lugo I. 1996. Response of Two otato Clones (S.

tuberosum L.) to Contrasting Temperature Regimes in the Field. Amercial

Journal of Potato Research, 73, 285-300.

Sattelmacher B., Marschner H. & R. Kuhne. 1990. Effects of the Temperature of the

Rooting Zone on the Growth and Development of Roots of Potato (Solanum

tuberosum). . Annals of Botany, 65, 27-36.

Schapendonk A. H. C. M., Pot C. S. & Goudriaan J. 1995. Potato Ecology and

Modelling of Crops under Conditions Limiting Growth, Amsterdam., Kluwer

Academic Publisher.

Scott R. K. & Wilcockson S. J. 1978. Application of Physiological and Agronomic

Principles to the Development of Potato Inductry, London.

Secretariat of the Pacific Community & CSIRO. 2011. Food security in the Pacific

and East Timor and its Vulnerability to Climate Change.

192

Shahnazari A., Ahmadia S. H., Laerke P. E., Liu F., Plauborg F., Jacobsen S.-E.,

Jensen C. R. & Andersen M. N. 2008. Nitrogen Dynamics in the Soil-Plant

System Under Deficit and Partial Root-Zone Drying Irrigation Strategies in

Potatoes. European Journal of Agronomy, 28, 65- 73.

Singh A. K. 1999. Crop Growth Simulation Models. New Delhi: Water Technology

Center.

Singh H. P. Year. Policies and Strategies Conducive to Potato Developoment in

Asia and the Pacific Region. In: Workshop to Co mmemorate the

International Year of Potato-2008, 2008 Bangkok, Thailand.

Singh J. P., Lal S. S. & Pandey S. K. 2009. Climate Change and Potato Production

in India. Shimla: Cenral Potato Research Institute.

Singh U., Brink J. E., Thornton P. K. & Christianson C. B. 1993. Linking Crop

Models With a Geographic Information System to Assist Decision Making: A

Prototype for the Indian Semiarid Tropics Muscle Shoals, Alabama: IFDC.

Singh U., Godwin D. C. & Morrison R. J. 1990. Modelling the Impact of Climate

Change on Agricultural Production in the South Pacific. . In: Hughes P. J. &

McGregor J. (eds.) Global Warming-Related Effects of Agriculture and

Human Health and Comfort in the South Pacific. South Pacific Regional

Environmental Program and United Nations Environment Programme,

University of Papua New Guinea, Port Moresby, Papua New Guinea.

Singh U., Matthews R. B., Griffin T. S., Ritchie J. T., Hunt L. A. & Goenaga R.

1998. Modeling Growth and Development of Root and Tuber crops. In: Tsuji

G., Hoogenboom G. & Thornton P. (eds.) Understanding Options for


Singh U. & Ritchie J. T. 1993. Simulating the Impacts of Climate Change on Crop

Growth and Nutrient Dynamics using the Ceres-Rice Model. Journal of

Agricultural Meteorology, 48 819-822.

Singh U., Wilkens P. W., Baethgen W. E. & Bontkes T. S. 2002. Agricultural

System Models in Field Research and Technology Transfer. In: Ahuja L. R.,

Ma L. & Howell T. A. (eds.) Decision Support Tools for Improved Resource

Managemnet and Agricultural Sustainbility. Washington, D.C: Lewis

Publishers

193

SMEC Australia. 2010. Climate Change Adaptation Actions for Local Government.

Australian Government. Department of Climate Change and Energy

Efficiency.

Snapp S. S. & Fortuna A. M. 2003. Predicting Nitrogen Availability in Irrigated

Potato Systems. Hortechnology, 13, 598-604.

Solomon S., Qin D., Manning M., Alley R. B., Berntsen T., Bindoff N. L., Chen Z.,

Chidthaisong A., Gregory J. M., Hegerl G. C., Heimann M., Hewitson B.,

Hoskins B. J., Joos F., Jouzel J., Kattsov V., Lohmann U., Matsuno T.,

Molina M., Nicholls N., Overpeck J., Raga G., Ramaswamy V., Ren J.,

Rusticucci M., Somerville R., Stocker T. F., Whetton P., Wood R. A. &

Wratt D. 2007. Technical Summary. In: Climate Change 2007: The Physical

Science Basis. Contribution of Working Group I to the Fourth Assessment

Report of the Intergovernmental Panel on Climate Change In: [Solomon S.,

D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor and H.L.

Miller (eds.)]. (ed.). Cambridge University Press, Cambridge, United

Kingdom and New York, NY, USA.

St’astna M., Trnka M., Krˇen J., Dubrovsky´ M. & alud Z. Z. 2002. Evaluation of the

CERES Models in Different Production Regions of the Czech Republic.

Rostlinnavyroba, 48, 125-132.

Stalham M. A. & Allison M. F. 2011. Improving Water Use Efficiency Through

Understanding Soil and Plant Water Balance. Agriculture and Horticulture

Development Board.

Štastna M. & Dufkova J. 2008. Potato Simulation Model and its Evaluation in

Selected Central European Country. Agriculturae Conspectus Scientifi cus,

73, 227-234.

SterN P. & Easterling W. 1999. Making Climate Forecasts Matter: Panel on the

Human Dimensions of Seasonal-to-Interannual Climate Variability. Co

mmittee on the Human Dimensions of Global Change. Washington D.C.: Co

mmission of Behavioral and Social Sciences and Education. National

Research Council.

Stitt M. 1991. Rising CO2 levels and their potential significance for carbon flow in

photosynthetic cells. Plant, Cell & Environment, 14, 741-762.

Stockle C. O., Donatelli M. & Nelson R. 2003. CropSyst, A Cropping Systems

Simulation Model. European Journal of Agronomy, 18, 289-307.

194

Stoorvogel J. J., Antle J. M., Crissman C. C. & Bowen W. 2004. The Tradeoff

Analysis Model: Integrated Bio-physical and Economic Modeling of

Agricultural Production Systems. Agricultural Systems, 80, 43-66.

Struik P. C. 2007a. Chapter 11: Above-Ground and Below-Ground Plant

Development. In: Vreugdenhil D. (ed.) Potato Biology and Biotechnology.

Advances and Perspectives. Elsevier.

Struik P. C. 2007b. Responses of the Potato Plant to Temperature. In: Vreugdenhil

D. (ed.) Potato Biology and Biotechnology: Advances and Perspectives.

Struik P. C., Geertsema J. & Custers C. H. M. G. 1989. Effects of shoots, root and

stolon temperarture on the development of Potato (Solanum tuberosum L.)

plant.I. Developmembt of tubers. Potato Research, 32, 133-141.

Świeżyński K. & Zimnoch-Guzowska E. 2001. Breeding Potato Cultivars with

Tubers Resistant to Phytophthora infestans. Potato Research, 44, 97-117.

Te mmerman L. D., Bindi M., Craigon J., Fmgmeier A., Hacow A., H. Pleijel,

Vandermeiren K., Vorne V. & Wolf J. 2000. Changing Climate and Potential

Impacts on Potato Yield and Quality, Veterinary and Agrochemical Research

Centre, Tervuren, Belgium.

Te mmerman L. D., Hacour A. & Guns M. 2002. Changing Climate and Potential

Impacts on Potato Yield and Quality ‘CHIP’: introduction, aims and

methodology. European Journal of Agronomy, 17, 233- 242.

Te mmerman L. D., Vandermeiren K. & Oijen M. v. 2007. Response to the

Environment: Carbon Dioxide. In: Vreugdenhil D. (ed.) Potato Biology and

Biotechnology: Advances and Perspectives.

The Fijian Government. 2010. Potato Farming Aimed to Reduce Imports [Online].

Ministry of Information, National Archives & Library Services of Fiji.

Available: http://www.fiji.gov.fj/Media-Center/Press-Releases/Potato-

farming-aimed-to-reduce-imports.aspx [Accessed 6th of August 2013].

The Natural History Museum. 2012. Potato: Transfer and Spread [Online].

Available: http://www.nhm.ac.uk/nature-online/life/plants-fungi/seeds-of-

trade/page.dsml?section=crops&page=spread&ref=potato [Accessed 27th

January 2013].

The Potato Association for America. 2009. Varieties -Russet Burbank (Solanum

tuberosum) [Online]. Available:

195

http://www.potatoassociation.org/Industry%20Outreach/varieties/Russets/rus

set_burbank.html [Accessed 23rd July 2013].

Thornton M., Miller J., Hutchinson P. & Alvarez J. 2010. Response of Potatoes to

Soil-Applied Insecticides, Fungicides, and Herbicides. Potato Research, 53,

351- 358.

Thornton M. K., Malik N. J. & Dwelle R. B. 1996. Relationship Between Leaf Gas

Exchange Characteristics and Productivity of Potato Clones Grown at

Different Temperatures. Americal Journal of Potato Research, 73, 63–77.

Thornton P. K. & Wilkens P. W. 1998. Risk assessment and food security. In: Tsuji

G., Hoogenboom G. & Thornton P. (eds.) Understanding Options for


Thorpa K. R., DeJongeb K. C., Kaleitac A. L., Batchelord W. D. & Paze J. O. 2008.

Methodology for the Use of DSSAT Models for Precision Agriculture

Decision Support. Computers and Electronics in Agricutlure, 64, 276- 285.

United Nations. 2002. Coping against El Nino for Stabilizing Rainfed Agriculture:

Lessons from Asia and the Pacific. In: Yokoyama S. & Concepcion R. N.

(eds.).

United Nations Framework Convention on Climate Change. 2012. Article 1:

Definitions [Online]. Available:

http://unfccc.int/essential_background/convention/background/items/2536.ph

p [Accessed 25th December 2012].

Urbita. 2012. Rakiraki [Online]. Available: http://urbita.com/destinations/fiji/rakiraki


Valdivia C., Jette C., Quiroz R., Gilles J. & Materer. S. 2000. Peasant Households

Strategies in the Andes and Potential Users of Climate Forecasts: El Niño of

1997-1998. Tampa Florida: Session on Natural Resources and Environmental

Issues in Development. Selected Paper presented at the Meetings of the

American Agricultural Economics Association.

Van Delden A, Kropff M.J & Haverkort J. 2001. Modeling Temperature and

Radiation-Driven Leaf Area Expansion in the Contrasting Crops Potato and

Wheat. Field Crops Research, 72, 112-142.

van Heemst H. D. J. 1986. The Distribution of Dry Matter During Growth of a Potato

Crop. Potato Research, 29, 55–66.

196

Vanualailai D. P. & UNFCCC Consultant. 2008. Climate Change. United Nations

Framework Convention On Climate Change (UNFCCC). In: NCSA Project

Steering Co mmittee, NCSA Project & UNDP Fiji (eds.).

Vorne V., K. Ojanperä, L. De Te mmerman, M. Bindi, P. Högy, M.B. Jones, Lawson

T. & K. Persson. 2002. Effects of Elevated Carbon Dioxide and Ozone on

Potato Tuber Quality in the European Multiple-Site Experiment 'CHIP-

Project'. European Journal of Agrnonomy, 17, 369-381.

Vos J. 1995. Modelling and Parameterization of the Soil-Plant-Atmosphere System.

A Comparison of Potato Growth Models. In: Kabat P., Marshall B., Broek B.

J. v. d., Vos J. & Keulen H. v. (eds.). Wageningen Pers, Wageningen, The

Netherlands.

Vos J. 2009. Nitrogen Responses and Nitrogen Management in Potato. Potato

Research, 52, 305- 317.

Vos J. & Haverkort A. J. 2007a. Water Availability and Potato Crop Performance.

In: Vreugdenhil D. (ed.) Potato Biology and Biotechnology: Advances and

Perspectives.

Vos J. & Haverkort A. J. 2007b. Water Availability and Potato Crop Performance.

In: Vreugdenhil D. (ed.) Potato Biology and Biotechnology: Advances and

Perspectives.

Wairiu M., Lal M. & Iese V. 2012. Climate Change Implications for Crop

Production in Pacific Islands Region. In: Aladjadjiyan A. (ed.) Food

Production – Approaches, Challenges and Tasks. Pacific Centre for

Environment and Sustainable Development, University of the South Pacific,

Fiji.

Wheeler R. M. 2006. Potato and human exploration of space: Some observations

from NASA-sponsored controlled environment studies. Americal Journal of

Potato Research, 49, 67-90.

Wheeler R. M., Tibbitts T. W. & A.H. Fitzpatrick. 1991. Carbon Dioxide Effects on

Potato Growth under Different Photoperiods and Irradiance. . Crop Science

Journal, 31, 1209–1213.

White J. W. 1998. Modeling and crop improvement. In: Tsuji G., Hoogenboom G. &

Thornton P. (eds.) Understanding Options for Agricultural Production.

Springer Netherlands.

197

World Meteorological Organisation. 2003. Frequently Asked Questions (FAQS)

[Online]. Available: http://www.wmo.int/pages/prog/wcp/ccl/faqs.html

[Accessed 3rd March 2012].

World Meteorological Organisation. 2010. Agrometerology of Some Selected Crops.

Guide to Agricultural Meteorological Practices (GAMP) 2010 Edition

(WMO-No.134)

Xu X., Vreugdenhil D. & AAM v. L. 1998. Cell division and Cell Enlargement

during Potato Tuber Formation. Journal of Experimental Botany, 49, 573-

582.

Yamaguchi M., Ti mm H. & Spurr A. R. 1964. Effects of soil temperature on growth

and nutrition of potato plants and tuberization, composition and periderm

structure of tubers. American Society for Horticultural Science, 84, 412–423.

Yandell B. S., Najar A., Wheeler R. & Tibbitts T. W. 1988. Modeling the Effects of

Light, Carbon Dioxide, and Temperature on the Growth of Potato. Crop Sci.,

28, 811-818.

Yokoyama S. 2002. ENSO Impacts on Food Crop Production and the Role of

CGPRT Crops. Un ESCAP.

Yordanov I., Velikova V. & Tsonev T. 2003. Plant Resposes to Drought and Stress

Tolerance. Bulgarian Journal of Plant Physiology, 187- 206.

Zaag P. V. & Demaganate A. L. 1987. Potato (Solanum spp) in an Isohperthermic

Environment I. Agronomic Management. Fields Crop Research, 17, 199-217.

Zaag P. V., Demaganate A. L., Acasio R., Domingo A. & Hagerman H. 1986.

Response of Solanum Potatoes to Mulching During Different Seasons in an

Isohyperthermic Environment in the Philippines. Tropical Agriculture, 3,

229-239.

198

Appendices

Appendix 1: Calibration

Appendix 1 shows the chemical and physical soil properties of Banisogosogo soil.

Replica 2 was selected for calibration process. The growth and development stages

of potato crop and the stress faced by the crop is also shown in the table provided

below.

1.0A Calibration

Table 1.0A shows the chemical properties of Banisogosogo soil.

Replica pH (water)

EC (mS/ cm)

Total C (%)

Total N(%)

CEC ( cmol(+)/kg)

Exchangeable K (me/100g)

Moisture

Replica 1 (0-40 cm)

6.5 0.1 2.32 0.19 40.85 0.75 8.2

Replica 1 (40-100)

6.9 0.07 1.92 0.16 43.43 0.24 7.6

Replica 1 (100-110 cm)

7.0 0.06 1.74 0.14 38.96 0.2 8.3

Replica 2 (0-40 cm)

6.7 0.1 3.51 0.29 41.27 0.84 9.0

Replica 2 (40-7.1100)

7.0 0.06 2.27 0.19 37.43 0.24 6.1


7.1 0.05 1.74 0.14 43.71 0.21 8.1

Replica 3 (0-40 cm)

6.8 0.09 2.14 0.18 36.49 0.56 8.8

Replica 3 (40-100)

7.0 0.06 2.32 0.19 43.49 0.21 8.2


7.2 0.06 1.74 0.14 43.94 0.2 8.1

199

Table 1.1A shows the physical properties of the Banisogosogo soil.

Replica Clay (%) ( <0.002 mm)

Silt (%) (0.002 – 0.06 mm)

Sand (%) (0.06 – 2 mm)

Texture

Replica 1 (0-40 cm)

57 23 20 Clay

Replica 1 (40-100)

59 22 19 Clay


44 31 25 Clay

Replica 2 (0-40 cm)

48 25 27 Clay

Replica 2 (40-7.1100)

57 22 21 Clay


56 11 33 Clay

Replica 3 (0-40 cm)

59 25 16 Clay

Replica 3 (40-100)

51 26 23 Clay


63 9 28 Clay

Table 1.2A shows the simulated crop and soil status at main development stages for

replica 2.

Date Crop

Age Growth Stage

Biomass (kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

9 Jul

7 Emergence 10 0.01 0 4 42.8 0 0 0 0 1

6 Aug

35 Beg Tuber Initiation

307 0.46 0 134 43.7 0 0.17 0 0 2

20 Sep

80 Harvest 3336 0.4 0 645 19.3 0.33 0 0 0 2

200

Table 1.3A shows the main growth and development variables for replica 2.

Variable Simulated Measured Tuber Initiation Day (dap) 35 35 Physiological Maturity Day -99 79 Tuber Dry Weight (kg\ha) harvest

2947 2196


14.73 17.55

Leaf Area Index, Maximum 0.74 0.5 Table 1.4A shows the environmental and stress factors.

Development Stage Environment Stress Tim

e Span Days

Average Cumulative (0=min, 1=max stress)

T max (oC)

T min (oC)

Solar radiation (MJ/m2)

Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth

Emergence-Begin Tuber

28 27.7

20.5

13.9 11.13 392.0

53.4 46.8 0.000 0.004 0.131 0.169

Begin Tuber-Maturity

46 28.9

21.8

17.6 11.62 392.0

80.5 134.5 0.234 0.324 0.000 0.003

Planting-Harvest

81 28.4

21.3

16.2 11.4 392.1

134.3

189.4 0.133 0.185 0.045 0.060

201

Appendix 2: Current climate simulations

Appendix 2 shows the soil chemical and physical properties for Banisogosogo soil,

Koronivia soil and Nacocolevu soil. It also provides the model output for current

climate simulation of these three areas using Desiree variety.

2.0A Current climate simulations

2.1.1A Soil profile information

2.1.1A Soil name: Koronivia Silt Loam

Profile name: KN30

Laboratory number: SB 9604A-E

Elevation (m): 19

Landform: Plateau surface, probably marine-planated

Relief: Flat

Parent material: In-situ rhyolitic outwash and tuffs

Vegetation: Para grass

Drainage: Moderately well drained

Classification: Humoxic Tropohumult, clayey, kaolintic, isohyperthermic

Soil classification:

Ap (0-18 cm): slightly moist, dark brown (10YR 3/3) silt loam, humic staining along

root channels, weakly developed fine and very fine nut with granular structure,

friable to firm, many fine roots, indistinct smooth boundary. Ochricepipedon.

Au (18-41 cm):slightly moist, olive brown (2.5 YR 4/4) and rubbed yellowish brown

(10YR 5/8) silt loam, common medium distinct yellowish red (5YR 4/6) mottles,

weakly developed very fine nut and granular structure, friable, common fine roots,

sharp wavy boundary.

Bt (41- 73): moist, yellowish brown (10YR 5/8) clay loam, few medium prominent

dark red (10Y 3/6) mottles, weakly developed coarse nut breaking to weak very fine

202

blocky structure, firm, sticky, slightly plastic, few faint dark yellowish brown (10YR

4.5/6) clay skins, few very fine roots, indistinct smooth boundary.

Bts (73-100):moist, 75% yellowish brown (10YR 5/8) with 15% very pale brown

(10YR 7/3) clay loam, profuse coarse prominent dark red (10R 3/6) mottles, weakly

developed coarse nut structure, firm, sticky, slightly plastic, common faint strong

brown (7.5 YR 5/6) clay skin, no roots, indistinct smooth boundary.

BC (100-130): moist, very pale brown (10 YR 7/3) silty clay loam, massive, firm,

slightly sticky, few faint yellowish brown (10 YR 5/8) clay skin, in situ moderately

weathered.

Table 2.0A shows the pH and organic matter for each layer of the Koronivia Silt Loam

soil.

SB

Lab No.

Horizon pH Organic Matter

(%)

Depth

( cm)

H2O NAF C N

9604A 0-18 Ap 5.4 2.6

9604B 18-41 Au 5.6 1.8

9604C 41-73 Bt 5.1 1.0

9604D 73-100 Bts 5.0 0.6

9604E 100-130 BC 5.1 0.3

203

Table 2.1A shows the chemical properties of the Koronivia Silt Loam soil.

SB

Lab No.

Horizon CEC (me./100g)

Cation Exchange KCL

Exch. acid

Depth

( cm)

BS (%)

Ca Mg K Na

9604A 0-18 Ap 10.6 6.4 60 5.0 1.00 0.16 0.24

9604B 18-41 Au 8.6 5.2 60 4.4 0.67 0.04 0.08

9604C 41-73 Bt 10.1 4.5 45 4.5 0.86 0.04 0.02

9604D 73-100

Bts 11.4 1.2 11 1.2 0.43 0.05 0.03

9604E 100-130

BC 13.9 1.1 8 1.1 0.34 0.09 0.15 18.2

Table 2.2A shows physical properties of the Koronivia Silt Loam soil.

SB

Lab No.

Horizon Particle size Fine to total clay ratio

15 bar water (%)

Clay (%)

Depth

( cm)

2.0-0.1 mm

2.0-0.2mm

0.2-0.02 mm

0.02-0.002 mm

<0.002 mm

<0.0002 mm

9604A 0-18 Ap 32 13 42 13 32 23 0.72 11.3 0.35

9604B 18-41 Au 29 8 41 10 41 32 0.78 13.4 0.33

9604C 41-73 Bt 13 3 19 9 69 56 0.81 27.4 0.40

9604D 73-100

Bts 6 2 12 10 76 51 0.67 29.1 0.38

9604E 100-130

BC 4 1 18 19 62 40 0.65 23.9 0.39

204

2.1.2A Soil map unit name: Nacocolevu soil, rolling phase

Area: 6 ha

Soil classification: OxicHalplustalf, fine, kaolinitic, isohyperthermic

Parent material: strongly weathered silicified marls and tuff

Geographical distribution: northern part of research station

Physiographical position and slope:rolling and easy rolling crests and ridges of

moderately dissected hill country. Slopes range form 7o-10o.

Vegetation and land use: cut-over bush, small area in Pinuscaribaea, some in

grasses

Elevation: 10-180 m

Annual rainfall: 1025-2973 mm

Soil profile description (Profile Sp 16 on 8o slope):

A (0-21 cm): dark reddish brown (5YR ¾ -3/3) silty clay loam, friable, sticky,

moderately plastic, moderately developed medium and fine nut structure, few fine

pores, many fine and medium rots, few medium (up to 3 cm diameter) strong brown

(7.5 YR 5/8) and yellowish red (5 YR 4/6) weakly weathered gravels coated with

dark reddish brown (5YR 3/3) humus, distinct smooth boundary.

Bt1 (21-40 cm): yellowish red (5YR 4/6) silty clay, friable, sticky, plastic,

moderately developed coarse and fine nut structure, many thin continuous reddish

brown (5 YR 4/4) clay coatings on peds, few dark reddish brown (5YR 3/2) humus

stains decreasing with depth, many fine pores, few fine and medium roots, distinct

smooth boundary.

Bt2 (40-72 cm): red (2.5 YR 4/6) silt loam, friable, slightly sticky, slightly plastic,

moderately developed fine blocky structure, many thin continuous (near red 2.5 YR

5/6) clay skin on peds, many fine distinct yellowish brown to strong brown (10YR-

7.5 YR 5/8) parent material remnants, few fine and medium roots, few fine roots,

distinct wavy boundary.

205

BCt (72-89 cm): red (2.5 YR 4/6) and strong brown (7.5 YR 5/8) silt loam to silt (

weathered parent rock), soft, nonsticky, nonplastic, massive to weakly developed

fine blocky structure, many veins and pockets of white (10 YR 8/2) parent material,

many thin discontinuous clay skins on peds, few fine roots, distinct irregular

boundary.

BC (89-125 cm): reddish brown (7.5 YR 6/6) silt, soft, massive breaking to weakly

developed platy structure, dark red (2.5 YR 3/6) clay coatings on peds, pockets of

white (10 YR 8/1) parent material.

Characteristics soil and site features: the red subsoils are the outstanding fine

characteristics of these soils. Profiles consists of dark reddish brown, friable silty

clay with moderately developed structure overlying yellowish red , friable silty clay

with moderately developed blocky structure and continuous clay skins on peds. This

rest on red, silt loam, friable with moderately developed blocky structure with

distinct clay skin on peds. Pockets of white parent material occur with increasing

depth. The soils occur on broad ridges and spurs mainly in the northern part of the

station.

Inclusion and variants within the map unit: pockets of Kabisi, rolling phase, may

occur (less than 10% of the map unit).

Associated soils: Barara soils on accumulation positions in the hill country. In places

the hill crest consist of Nacocolevu soils, rolling phase, but the associated hilly slope

consists of Kabisi soils, moderately steep to steep phase.

Drainage class: well drained

Erosion and Flooding: moderate soil creep

Chemical features: low base saturation, except directly above the lithic contact (at

146 cm depth in Profile Sp 16) where a medium base saturation was analyzed. Low

exchangeable magnesium and very low exchangeable calcium and potassium.

Medium values of exchangeable sodium. Organic carbon is low.

Soil limitation and potential: low natural nutrient levels. Erosion risk under

intensive uses. Pastoral uses (beef, sheep).

206

Table 2.3A shows the Nacocolevu Rolling Phase soil properties.

Soil and

sample

number

Horiz Organic Cation Exchange

Nacocolevu

soil, rolling

phase

KRS T

1838-43

Depth

( cm)

Silt Clay Carbon

(%)

CEC

(me%)

TEB

(me%)

BS

(%)

Ca

(me%)

Mg

(me%)

K

(me%)

Na(me%)

0-21 A 26 60 3.37 16.0 4.4 28 1.9 1.8 0.2 0.5

21-40 Bt1 25 70 2.18 13.1 2.1 16 0.9 0.7 0.1 0.5

40-72 Bt2 34 64 0.89 13.0 1.9 15 0.8 0.5 0.1 0.5

72-89 BCt 38 53 0.58 13.3 1.9 100 0.7 0.5 0.1 0.5

89-

125

BC 55 41 0.48 12.3 2.2 18 0.9 0.7 0.1 0.6

136-

146

BC N.D 20.2 11.1 55 9.7 0.8 0.1 0.5

Table 2.4A shows the information on weather station and soil for each location.

Location Experiment Number

Weather Station Soil

Koronivia CCRA1250 Non-Potential

FJNU Koronivia Silt Loam

CCRA1255 Potential FJNU Koronivia Silt Loam

Nacocolevu CCRA1260 Non-Potential

FJNC Nacocolevu Soil, Rolling Phase

CCRA1265 Potential FJNC Nacocolevu Soil, Rolling Phase

207

Table 2.5A shows the potential simulated crop and soil status at main development

stages for Banisogosogo.

Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 0

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 0

9 Jul

7 Emergence 0 0 0 0 0 0 0 0 0 0

9 Aug


0 0 0 0 0 0 0 0 0 0

20 Sep

80 Harvest 0 0 0 0 0 0 0 0 0 0

Table 2.6A shows the environmental and stress factors.

Development Stage Environment Stress

Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


31 27.8

20.5

20.7 11.14 392.0

69.4 46.8 0.000 0.000 0.000 0.000


43 28.9

21.8

21.7 11.65 392.0

64.5 134.5 0.000 0.000 0.000 0.000

Planting-Harvest

81 28.4

21.3

16.2 11.4 392.1

134.3

189.4 0.000 0.000 0.000 0.000

208

Table 2.7A shows the non-potential simulated crop and soil status at main development


Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

6 Jul

7 Emergence 0 0.01 0 5 42.8 0 0 0 0 1

9 Aug


0 0.67 0 194 43.9 0 0.16 0 0 2

20 Sep

80 Harvest 0 1.60 0 1162 19.7 0.33 0.03 0 0 2

Table 2.8A shows the environmental and stress factors for Banisogosogo.


Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


28 27.7

20.5

13.9 11.13 392.0

53.4 49.3 0.000 0.004 0.120 0.159


46 28.9

21.8

17.6 11.62 392.0

80.5 160.3 0.240 0.324 0.004 0.029

Planting-Harvest

81 28.4

21.3

16.2 11.4 392.1

134.3

217.7 0.136 0.184 0.044 0.071

209


stages for Koronivia.

Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

8 Jul

6 Emergence 13 0 0 0 0 0 0 0 0 1

3 Aug


4721 0 0 0 0 0 0 0 0 2

20 Sep

80 Harvest 20183 0 0 0 0 0 0 0 0 2

Table 2.10A shows the environmental and stress factors for Koronivia.


Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


26 26.5

19.6

26.0 11.15 389.0

107.2

0.000 0.000 0.000 0.000 0.000


49 27.8

20.7

25.0 11.61 390.1

142.2

0.000 0.000 0.000 0.000 0.000

Planting-Harvest

81 27.3

20.3

25.3 11.42 390.0

253.3

0.000 0.000 0.000 0.000 0.000

210

Table 2.11A shows the non-potential simulated crop and soil status at main

development stages for Koronivia.

Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

8 Jul

6 Emergence 11 0.01 0 5 42.9 0 0 0 0 1

1 Aug


768 1.17 0 286 37.3 0.22 0.15 0 0 2

20 Sep

80 Harvest 6459 1.93 0 1384 21.4 0.47 0.12 0 0 2

Table 2.12A shows the environmental and stress factors for Koronivia.


Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


24 26.5

19.4

26.1 11.14 389.9

94.2 83.3 0.119 0.219 0.100 0.139


51 27.8

20.8

25.0 11.59 390.1

155.2

194.0 0.426 0.464 0.080 0.124

Planting-Harvest

81 27.3

20.3

25.3 11.42 390.0

253.3

301.1 0.304 0.357 0.080 0.119

211


stages for Nacocolevu.

Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

9 Jul

7 Emergence 13 0.02 0 0 0 0 0 0 0 1

6 Aug


1006 1.51 0 0 0 0 0 0 0 2

20 Sep

80 Harvest 9969 2.11 0 0 0 0 0 0 0 2

Table 2.14A shows the environmental and stress factors for Nacocolevu.


Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth

Emergence- Begin Tuber

28 28.1

19.4

9.4 11.08 389.9

146.8

83.3 0.000 0.000 0.000 0.000


46 30.1

18.5

14.4 11.59 390.1

256.3

194.0 0.000 0.000 0.000 0.000

Planting-Harvest

81 29.4

18.9

12.3 11.36 390.0

253.3

475.6 0.000 0.000 0.000 0.000

212


development stages for Nacocolevu.

Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

9 Jul

7 Emergence 11 0.01 0 5 42.9 0 0 0 0 1

4 Aug


481 0.73 0 238 49.6 0 0.07 0 0 2

20 Sep

80 Harvest 5984 0.77 0 1112 18.6 0 0 0 0 2

Table 2.16A shows the environmental and stress factors for Nacocolevu.


Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


26 28.1

19.3

9.1 11.07 389.9

138.7

63.7 0.000 0.000 0.046 0.074


48 30.1

18.6

14.3 11.57 390.1

264.4

172.4 0.000 0.000 0.000 0.000

Planting-Harvest

81 29.4

18.9

12.3 11.36 390.0

475.6

253.7 0.000 0.000 0.015 0.024

213

Appendix 3: Future climate simulations

Appendix 3 provides the model output for potential and non-potential future climate

simulations. The model simulation was conducted for Banisogosogo, Koronivia and

Nacocolevu using Desiree. PCCSP A1B and A2 emission scenarios were used.

Likewise, it also shows the optimisation results for planting date, row spacing,

fertiliser application, irrigation application and planting depth.

3.0A Future climate simulation

3.1A Results

3.1.1A Banisogosogo climate simulation


stages for 2030 Medium emission scenario.

Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

10 Jul

8 Emergence 12 0.02 0 0 0 0 0 0 0 1

28 Aug


7129 10.54 0 0 0 0 0 0 0 2

20 Sep

80 Harvest 9955 8.44 0 0 0 0 0 0 0 2

214

Table 3.1A shows the environmental and stress factors for 2030 medium emission

scenario.


Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


49 29.4

22.2

15.2 11.27 490.0

93.7 0.000 0.000 0.000 0.000 0.000


24 30.1

23.2

18.4 11.80 490.0

56.0 0.000 0.000 0.000 0.000 0.000

Planting-Harvest

81 29.6

22.5

16.2 11.40 490.0

151.8

0.000 0.000 0.000 0.000 0.000


stages for 2030 high emissions.

Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

10 Jul

8 Emergence 12 0.02 0 0 0 0 0 0 0 1

25 Aug


6516 9.68 0 0 0 0 0 0 0 2

20 Sep

80 Harvest 10701 8.44 0 0 0 0 0 0 0 2

215



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


46 29.3

22.0

14.9 11.25 520.0

95.3 0.000 0.000 0.000 0.000 0.000


27 29.8

22.8

18.6 11.77 520.0

57.0 0.000 0.000 0.000 0.000 0.000

Planting-Harvest

81 29.4

22.2

16.2 11.40 520.0

154.4

0.000 0.000 0.000 0.000 0.000

3.1.2A Banisogosogo non-potential simulation


stages for 2030 medium emissions.

Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

11 Jul

9 Emergence 10 0.01 0 4 42.8 0 0 0 0 1

26 Aug


3397 5.15 0 1272 37.4 0 0.10 0 0 2

20 Sep

80 Harvest 6214 6.68 0 1825 29.4 0.20 0.28 0 0 2

216

Table 3.5A shows the environmental and stress factors. Development Stage Environment Stress

Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


46 29.5

22.2

15.0 11.26 490.0

93.7 130.6 0.000 0.000 0.069 0.099


26 30.0

23.1

18.6 11.78 490.0

56.0 106.5 0.081 0.188 0.234 0.272

Planting-Harvest

81 29.6

22.5

16.2 11.40 490.0

151.8

247.6 0.026 0.060 0.114 0.144



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

10 Jul

8 Emergence 10 0.01 0 4 0 0 0 0 0 1

22 Aug


3093 4.71 0 1251 0 0 0.10 0 0 2

20 Sep

80 Harvest 6928 6.82 0 1898 0 0.14 0.27 0 0 2

217



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


43 29.2

22.0

14.7 11.23 520.0

95.3 114.9 0.000 0.000 0.074 0.100


30 29.8

22.7

18.5 11.75 520.0

57.0 126.1 0.044 0.135 0.223 0.262

Planting-Harvest

81 29.4

22.3

16.2 11.40 520.0

154.4

250.7 0.016 0.050 0.122 0.150


stages for 2055 medium emission.

Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

12 Jul

10 Emergence 10 0.01 0 4 42.8 0 0 0 0 1

12 Sep


4333 7.87 0 1696 31.8 0.07 0.17 0 0 2

20 Sep

80 Harvest 5487 7.25 0 1808 33.0 0.18 0.38 0 0 2

218



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


62 30.2

23.0

15.8 11.39 520.0

126.0

192.6 0.034 0.066 0.129 0.160


9 31.6

24.6

20.4 11.92 520.0

29.0 41.9 0.074 0.183 0.356 0.388

Planting-Harvest

81 30.3

23.2

16.2 11.40 520.0

157.1

246.2 0.034 0.071 0.138 0.166


development stages for 2055 high emissions.

Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

11 Jul

10 Emergence 10 0.01 0 4 42.8 0 0 0 0 1

6 Sep


5741 8.49 0 1824 32.3 0.3 0.14 0 0 2

20 Sep

80 Harvest 6195 7.26 0 1913 30.9 0.13 0.41 0 0 2

219



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


57 30.0

22.8

15.5 11.34 992.1

125.3

167.7 0.011 0.031 0.121 0.156


15 30.7

23.6

19.4 11.87 992.0

29.7 69.3 0.034 0.125 0.377 0.408

Planting-Harvest

81 30.1

23 16.2 11.40 992.0

157.1

247.7 0.014 0.034 0.155 0.185

3.1.3A Koronivia potential simulation



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

9 Jul

7 Emergence 12 0.02 0 0 0 0 0 0 0 1

13 Aug


8200 12.12 0 0 0 0 0 0 0 2

20 Sep

80 Harvest 16385 8.61 0 0 0 0 0 0 0 5

220



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


35 28.0

21.0

26 11.2 490.0

140.7

0.000 0.000 0.000 0.000 0.000


39 29.1

22.0

24.8 11.68 490.0

141.1

0.000 0.000 0.000 0.000 0.000

Planting-Harvest

81 28.5

21.5

25.3 11.42 490.0

286.2

0.000 0.000 0.000 0.000 0.000



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

9 Jul

7 Emergence 12 0.02 0 0 0 0 0 0 0 1

11 Aug


7807 11.59 0 0 0 0 0 0 0 2

20 Sep

80 Harvest 17503 8.66 0 0 0 0 0 0 0 5

221



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


33 27.8

20.7

26.1 11.19 520.0

140.1

0.000 0.000 0.000 0.000 0.000


41 28.8

21.9

24.8 11.66 520.0

147.6

0.000 0.000 0.000 0.000 0.000

Planting-Harvest

81 28.3

21.3

25.3 11.42 520.0

291.3

0.000 0.000 0.000 0.000 0.000



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

10 Jul

8 Emergence 12 0.02 0 0 0 0 0 0 0 1

20 Aug


10701 15.49 0 0 0 0 0 0 0 2

20 Sep

80 Harvest 13341 8.48 0 0 0 0 0 0 0 5

222



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


41 28.9

21.7

25.8 11.25 520.0

145.7

0.000 0.000 0.000 0.000 0.000


32 29.8

23.0

24.8 11.73 520.0

146.1

0.000 0.000 0.000 0.000 0.000

Planting-Harvest

81 29.2

22.2

25.3 11.42 520.0

296.4

0.000 0.000 0.000 0.000 0.000



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

10 Jul

8 Emergence 12 0.02 0 0 0 0 0 0 0 1

18 Aug


12533 17.92 0 0 0 0 0 0 0 2

20 Sep

80 Harvest 16209 8.54 0 0 0 0 0 0 0 5

223



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


39 28.6

21.6

25.8 11.24 990.0

145.7

0.000 0.000 0.000 0.000 0.000


34 29.6

22.6

24.9 11.72 990.2

146.1

0.000 0.000 0.000 0.000 0.000

Planting-Harvest

81 29.0

22 25.3 11.42 990.0

296.4

0.000 0.000 0.000 0.000 0.000

3.1.4A Koronivia non-potential yield


development stages for 2030 medium emissions.

Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

9 Jul

7 Emergence 11 0.01 0 5 42.8 0 0 0 0 1

8 Aug


2311 3.54 0 929 40.2 0.13 0.11 0 0 2

20 Sep

80 Harvest 5501 3.23 0 1409 25.6 0.43 0.10 0 0 5

224



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


30 27.8

20.8

26.1 11.18 490.0

136.3

125.7 0.070 0.134 0.073 0.113


44 29.1

22.0

24.9 11.64 490.2

145.5

174.6 0.376 0.421 0.052 0.099

Planting-Harvest

81 28.5

21.5

25.3 11.42 490.0

286.2

326.1 0.230 0.278 0.055 0.096



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

9 Jul

7 Emergence 11 0.01 0 5 42.8 0 0 0 0 1

9 Aug


2263 3.46 0 922 40.7 0.14 0.12 0 0 2

20 Sep

811 Harvest 5777 3.1 0 1406 24.3 0.41 0.12 0 0 5

225



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


29 27.5

20.6

26.1 11.17 520.0

138.7

119.3 0.071 0.138 0.075 0.114


45 28.9

21.8

24.9 11.64 520.2

148.1

183.6 0.352 0.397 0.072 0.118

Planting-Harvest

81 28.3

21.3

25.3 11.42 520.0

291.3

328.8 0.221 0.270 0.067 0.107



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

10 Jul

8 Emergence 10 0.01 0 4 42.8 0 0 0 0 1

16 Aug


2857 4.27 0 1027 35.9 0.21 0.12 0 0 2

20 Sep

80 Harvest 5210 4.07 0 1421 27.3 0.37 0.18 0 0 5

226



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


37 28.7

21.9

25.7 11.23 550.0

145.7

155.9 0.124 0.185 0.075 0.116


36 29.9

22.7

25.0 11.70 550.2

146.1

147.7 0.351 0.384 0.137 0.180

Planting-Harvest

81 29.2

22.2

25.3 11.42 550.0

296.4

330.9 0.212 0.255 0.095 0.133



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

10 Jul

8 Emergence 10 0.01 0 4 42.8 0 0 0 0 1

14 Aug


3122 4.68 0 1096 35.1 0.15 0.15 0 0 2

20 Sep

80 Harvest 5668 3.56 0 1406 24.8 0.37 0.22 0 0 5

227



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


35 28.5

21.6

25.9 11.21 989.9

145.7

151.6 0.068 0.131 0.104 0.143


38 29.6

22.5

24.9 11.69 990.9

146.1

151.6 0.356 0.385 0.179 0.120

Planting-Harvest

81 29.0

22.0

25.3 11.42 990.0

296.4

330.6 0.197 0.237 0.129 0.165



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

12 Jul

10 Emergence 11 0.01 0 5 42.9 0 0 0 0 1

25 Aug


3327 4.93 0 1110 33.4 0.33 0.16 0 0 2

20 Sep

80 Harvest 4894 5.16 0 1416 28.9 0.24 0.33 0 0 5

228



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


44 29.9

22.7

25.8 11.29 700.0

116.6

162.8 0.239 0.305 0.109 0.150


27 30.9

24.2

24.6 11.71 700.0

135.9

127.1 0.227 0.260 0.292 0.328

Planting-Harvest

81 30.2

23.2

25.3 11.42 700.0

298.9

328.8 0.206 0.252 0.157 0.191



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

13 Jul

11 Emergence 11 0.01 0 5 42.9 0 0 0 0 1

2 Sep


4631 6.91 0 1483 32.0 0.31 0.2 0 0 2

20 Sep

80 Harvest 4826 5.59 0 1385 28.7 0.18 0.49 0 0 5

229



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


51 30.4

23.2

25.6 11.34 950.0

180.4

187.7 0.242 0.306 0.150 0.189


19 31.1

24.6

24.6 11.83 950.0

39.2 97.9 0.322 0.166 0.456 0.483

Planting-Harvest

81 30.5

23.5

25.3 11.42 950.0

309.0

332.2 0.181 0.232 0.201 0.232

3.1.5A Nacocolevu potential yield



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

10 Jul

8 Emergence 12 0.02 0 0 0 0 0 0 0 1

12 Aug


2131 3.22 0 0 0 0 0 0 0 2

20 Sep

80 Harvest 12080 6.76 0 0 0 0 0 0 0 5

230



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


33 29.7

20.4

10.5 11.12 490.0

199.9

0.000 0.000 0.000 0.000 0.000


40 31.4

19.8

14.2 11.64 490.0

248.4

0.000 0.000 0.000 0.000 0.000

Planting-Harvest

81 30.6

20.1

12.3 11.36 490.0

537.4

0.000 0.000 0.000 0.000 0.000



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

10 Jul

8 Emergence 12 0.02 0 0 0 0 0 0 0 1

11 Aug


2021 3.06 0 0 0 0 0 0 0 2

20 Sep

80 Harvest 12626 6.57 0 0 0 0 0 0 0 5

231



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


32 29.5

20.3

10.3 11.11 520.0

187.8

0.000 0.000 0.000 0.000 0.000


41 31.2

19.5

14.2 11.63 520.2

268.4

0.000 0.000 0.000 0.000 0.000

Planting-Harvest

81 30.4

19.9

12.3 11.36 520.0

546.9

0.000 0.000 0.000 0.000 0.000



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

11 Jul

9 Emergence 12 0.02 0 0 0 0 0 0 0 1

18 Aug


3536 5.37 0 0 0 0 0 0 0 2

20 Sep

80 Harvest 10778 7.72 0 0 0 0 0 0 0 5

232



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


38 30.6

20.6

11.6 11.16 550.0

227.8

0.000 0.000 0.000 0.000 0.000


34 32.2

20.9

13.7 11.69 550.0

228.3

0.000 0.000 0.000 0.000 0.000

Planting-Harvest

81 31.3

20.8

12.3 11.36 550.0

556.5

0.000 0.000 0.000 0.000 0.000



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

10 Jul

8 Emergence 12 0.02 0 0 0 0 0 0 0 1

15 Aug


4140 6.28 0 0 0 0 0 0 0 2

20 Sep

80 Harvest 13675 7.96 0 0 0 0 0 0 0 5

233



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


36 30.2

20.7

11.0 11.14 989.9

211.3

0.000 0.000 0.000 0.000 0.000


37 32.0

20.5

14.0 11.66 990.2

252.8

0.000 0.000 0.000 0.000 0.000

Planting-Harvest

81 31.1

20.6

12.3 11.36 990.0

556.5

0.000 0.000 0.000 0.000 0.000



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

12 Jul

10 Emergence 11 0.02 0 0 0 0 0 0 0 1

3 Aug


7931 11.64 0 0 0 0 0 0 0 2

20 Sep

80 Harvest 8435 8.32 0 0 0 0 0 0 0 5

234



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


53 32.0

21.6

13.2 11.28 700.0

352.1

0.000 0.000 0.000 0.000 0.000


18 33.5

22.1

11.2 11.82 700.0

106.0

0.000 0.000 0.000 0.000 0.000

Planting-Harvest

81 32.3

21.8

12.3 11.36 700.0

561.2

0.000 0.000 0.000 0.000 0.000



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

13 Jul

11 Emergence 11 0.02 0 0 0 0 0 0 0 1

9 Sep


10035 14.46 0 0 0 0 0 0 0 2

20 Sep

80 Harvest 8810 9.23 0 0 0 0 0 0 0 5

235



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


58 32.1

21.9

13.1 11.33 950.0

408.8

0.000 0.000 0.000 0.000 0.000


12 35.0

23.0

11.0 11.87 950.0

64.1 0.000 0.000 0.000 0.000 0.000

Planting-Harvest

81 32.6

22.1

12.3 11.36 950.0

580.2

0.000 0.000 0.000 0.000 0.000

3.1.6A Nacocolevu non-potential yield



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

10 Jul

8 Emergence 11 0.01 0 5 42.8 0 0 0 0 1

11 Aug


1164 1.76 0 551 47.4 0 0.07 0 0 2

20 Sep

80 Harvest 7115 2.85 0 1330 18.7 0 0.18 0 0 5

236



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


32 29.7

20.5

10.3 11.11 490.0

184.5

90.2 0.000 0.000 0.047 0.071


41 31.4

19.7

14.2 11.63 490.0

263.7

138.6 0.000 0.000 0.141 0.177

Planting-Harvest

81 30.6

20.1

12.3 11.36 490.0

537.4

249.0 0.000 0.000 0.090 0.118



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

10 Jul

8 Emergence 11 0.01 0 5 42.8 0 0 0 0 1

10 Aug


1073 1.64 0 500 46.6 0 0.07 0 0 2

20 Sep

80 Harvest 7501 2.50 0 1298 17.3 0 0.21 0 0 5

237



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


31 29.4

20.4

10.1 11.11 520.0

184.0

85.6 0.000 0.000 0.047 0.072


42 31.2

19.4

14.3 11.62 520.0

272.2

142.4 0.000 0.000 0.171 0.204

Planting-Harvest

81 30.4

19.9

12.3 11.36 520.0

546.9

248.1 0.000 0.000 0.107 0.133



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

11 Jul

9 Emergence 11 0.01 0 5 42.9 0 0 0 0 1

16 Aug


1933 2.95 0 819 42.4 0 0.07 0 0 2

20 Sep

80 Harvest 5703 4.88 0 1302 22.8 0 0.33 0 0 5

238



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


36 30.5

20.7

11.3 11.15 550.0

215.2

110.2 0.000 0.000 0.044 0.064


36 32.2

20.8

13.9 11.67 550.0

240.9

118.7 0.000 0.000 0.285 0.321

Planting-Harvest

81 31.3

20.8

12.3 11.36 550.0

556.5

251.2 0.000 0.000 0.146 0.171



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

11 Jul

9 Emergence 11 0.01 0 5 42.9 0 0 0 0 1

15 Aug


2257 3.44 0 869 38.5 0 0.08 0 0 2

20 Sep

80 Harvest 6099 4.01 0 1273 20.9 0 0.37 0 0 5

239



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


35 30.2

20.6

11.0 11.14 989.9

203.2

103.5 0.000 0.000 0.045 0.073


37 32.0

20.5

14.0 11.66 990.2

252.8

119.2 0.000 0.000 0.333 0.366

Planting-Harvest

81 31.1

20.6

12.3 11.36 990.0

556.5

245.0 0.000 0.000 0.172 0.199



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

13 Jul

10 Emergence 10 0.01 0 4 42.8 0 0 0 0 1

1 Sep


3365 5.07 0 1129 33.5 0 0.10 0 0 2

20 Sep

80 Harvest 4497 5.52 0 1276 28.4 0 0.41 0 0 5

240



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


50 32.0

21.7

13.3 11.27 700.0

345.3

174.4 0.000 0.000 0.068 0.094


20 33.1

22.1

11.4 11.80 700.0

112.1

54.0 0.000 0.000 0.372 0.403

Planting-Harvest

81 32.3

21.8

12.3 11.36 700.0

561.2

255.5 0.000 0.000 0.134 0.158



Date Crop

Age Growth Stage

Biomass

(kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

13 Jul

11 Emergence 10 0.01 0 4 42.8 0 0 0 0 1

3 Sep


3469 5.22 0 1123 32.4 0 0.13 0 0 2

20 Sep

80 Harvest 4337 5.58 0 1239 28.6 0 0.48 0 0 5

241



Time Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


52 32.2

20.0

13.3 11.28 950.0

363.3

177.8 0.000 0.000 0.095 0.123


18 33.8

22.4

11.2 11.82 950.0

109.6

49.7 0.000 0.000 0.443 0.470

Planting-Harvest

81 32.6

22.1

12.3 11.36 950.0

580.2

254.7 0.000 0.000 0.160 0.184

3.2A Optimisation treatments

3.2.1A Banisogosogo optimisation

Table 3.56A shows the yield at different planting time for Banisogosogo under A1B and

A2 under non-potential simulations.

Yield (kg/ha)

Planting Time 2030 A1B Emission

2030 A2 Emission

2055 A1B Emission

2055 A2 Emission

January 2 0 0 0 0

February 1 0 0 0 0

March 1 0 0 0 0

April 1 1 0 0 0

May 1 1020 1688 0 245

June 1 1931 2571 271 855

July 2 (default run)

2116 2469 757 1423

August 1 1249 1812 168 606

September 1 501 574 234 398

242

October 1 1511 1763 361 836

November 1 0 0 0 0

December 1 103 0 0 0

Table 3.57A shows the yield at different row spacing for Banisogosogo under A1B and

A2 under non-potential simulations.

Yield (kg/ha)

Row Spacing ( cm)

Plant Population (m2)

2030 A1B Emission

2030 A2 Emission

2055 A1B Emission

2055 A2 Emission

75(default run) 5 1261 2713 148 300

30 11 1229 2768 188 302

40 8 1246 2803 148 301

50 6 1256 2713 148 300

80 4 1171 2713 148 300

100 3 1174 2712 147 299

Table 3.58A shows the yield under different irrigation amount for Banisogosogo under

A1B and A2 non-potential simulations.

Yield (kg/ha)

Irrigation ( mm)

2030 A1B Emission 2030 A2 Emission 2055 A1B Emission 2055 A2 Emission

Default run (1.6 mm and 4.0 mm)

1261 2713 148 300

1.6 1194 1822 82 301

4.0 1209 1867 54 241

6.4 1138 1869 38 187

8.0 1135 1722 23 182

9.6 1124 1579 23 184

12.0 1063 1576 23 186

14.4 1127 1585 24 184

16.0 1129 1580 24 184

243

Table 3.59A shows the planting depth with corresponding yield for Banisogosogo under

A1B and A2 non-potential simulations.

Yield (kg/ha)

Planting Depth ( cm) 2030 A1B Emission 2030 A2 Emission 2055 A1B Emission 2055 A2 Emission

1.5 1261 1166 1166 300

2 977 1135 1135 140

4 235 1196 1196 0

6 6 1166 720 0

8 0 32 32 0

10 0 0 0 0

Table 3.60A shows the application of fertiliser (banded beneath surface at different

depth) and corresponding yield for Banisogosogo 2030 A1B emission.


Default run (80 N kg/ha in 2 splits at 5 cm)

1261


60 1260 1260 1260 1251 1240 1233

120 1223 1223 1223 1212 1198 1192

180 1234 1234 1234 1233 1224 1217

240 1212 1212 1212 1209 1138 1139

300 1410 1410 1410 1395 1175 1123

.

244

Table 3.61A shows the optimum fertiliser and irrigation management for Banisogosogo

2030 A1B emission.



300 at 0 cm Default value 1475

300at 2 cm Default value 1475


Table 3.62A shows the application of fertiliser (banded beneath surface) and

corresponding yield for Banisogosogo 2030 A2 emission.



1166


60 1759 1759 1759 1742 1726 1720

120 1867 1867 1867 1843 1821 1812

180 1930 1930 1930 1906 1877 1731

240 1937 1937 1937 2021 1768 1763

300 2265 2265 2265 2092 1867 1792


2030 A2 Emission.




300at 2 cm Default value 2221


245


depth) and corresponding yield for Banisogosogo 2055 A1B Emission.



1166


60 119 119 119 119 119 119

120 96 96 96 96 97 97

180 78 78 78 80 80 80

240 62 62 62 65 37 36

300 33 33 33 21 12 12


2055 A1B emission.


Fertiliser Application Irrigation Application ( mm) Yield (kg/ha)

Default value Default value 82


depth) and corresponding yield for Banisogosogo 2055 A2 emission.



300


60 304 304 304 303 303 303

120 300 300 300 300 301 301

180 294 294 294 296 238 237

240 224 224 224 229 226 178

300 197 197 197 167 165 133

246


2055 A2 emission.



60 kg/ha at 0 cm 1.6 292

60 kg/ha at 2 cm 1.6 292

60 kg/ha at 4 cm 1.6 292

3.2.2A Koronivia Optimisation Treatments

Table 3.68A shows the yield at different planting time for Koronivia for 2030-2090

under A1B and A2 emission scenario.

Yield (kg/ha)

Planting Time

2030 A1B Emission

2030 A2 Emission

2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

2090 A2 Emission

January 2 0 0 0 0 0 0

February 1 0 0 0 0 0 0

March 1 33 292 0 0 0 0

April 1 1655 2245 864 1524 0 0

May 1 3334 3914 2213 3191 833 583

June 1 3457 4249 2587 3923 1123 897


4268 5368 3205 4799 1911 1570

August 1 6692 7827 3590 5961 1117 745

September 1 3692 4255 2515 3609 1485 1128

October 1 2953 3710 1324 2230 235 32

November 1 334 502 36 193 0 0

December 1 139 196 4 62 0 0

247

Table 3.69A shows yield at different row spacing for Koronivia for 2030-2090 under


Yield (kg/ha)

Row Spacing ( cm)

Plant Population per m2

2030 A1B Emission

2030 A2 Emission

2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

2090 A2 Emission

75 (default run)

5 2369 2713 1578 2290 617 233

30 11 2358 2768 1601 2186 667 291

40 8 2403 2803 1607 2148 625 238

50 6 2362 2713 1615 2159 621 234

80 4 2342 2718 1665 2331 611 188

100 3 2451 2712 1687 2268 561 186

Table 3.70A shows the yield under different irrigation application and irrigation

amount for Koronivia for 2030-2090 under A1B and A2 emission scenario.

Yield (kg/ha)


2369 2713 1578 2990 617 233

Irrigation ( mm) 2030 A1B Emission

2030 A2 Emission

2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

2090 A2 Emission

1.6 2949 3330 1874 2159 721 251

4.0 3180 3394 1896 2240 687 213

6.4 3428 3855 2005 2376 609 189

8.0 3485 3866 1885 2354 607 165

9.6 3568 4024 1870 2240 521 121

12.0 3263 4139 1886 2245 406 86

14.4 3576 4548 1882 2314 341 71

16.0 3885 4911 1908 2335 337 59

248

Table 3.71A shows the planting depth with corresponding yield for Koronivia for 2030-

2090 under A1B and A2 emission scenario.

Yield (kg/ha)


2030 A1B Emission

2030 A2 Emission

2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

2090 A2 Emission

1.5 (default run)

2369 2713 1578 2290 617 233

2 2255 2581 1503 2052 241 25

4 2067 2415 1077 1503 0 0

6 1804 2183 181 579 0 0

8 122 450 0 0 0 0

10 0 0 0 0 0 0


depth) and corresponding yield for Koronivia under 2030 A1B emission.



2369


60 2860 2860 2860 2844 2849 2842

120 2881 2881 2881 2867 2855 2852

180 2820 2820 2820 2774 2770 2767

240 2751 2751 2751 2750 2742 2739

300 2914 2914 2914 2878 2876 2890

249

Table 3.73A shows the optimum fertiliser and irrigation management for Koronivia

under 2030 A1B emission.



300 kg/ha at 0 cm 16.0 3750

300 kg/ha at 2 cm 16.0 3449

300 kg/ha at 4 cm 16.0 2933


depth) and corresponding yield for Koronivia under 2030 A2 emission.



2713


60 3217 3217 3217 3194 3176 3166

120 3341 3341 3341 3302 3270 3255

180 3249 3249 3249 3227 3202 3224

240 3248 3248 3248 3148 3127 3115

300 3400 3400 3400 3281 3259 3282


under 2030 A2 emission.



300 kg/ha at 0 cm 16 4342

300 kg/ha at 2 cm 16 4047

300 kg/ha at 4 cm 16 3245

250





1578


60 2014 2014 2014 2010 2003 2001

120 1903 1903 1903 1897 1890 1883

180 1838 1838 1838 1840 1831 1830

240 1816 1816 1816 1813 1849 1848

300 1921 1921 1921 1948 1826 1817





60 kg/ha at 0 cm 6.4 1932

60kg/ha at 2 cm 6.4 1927

60kg/ha at 4 cm 6.4 1811

251





2990


60 2307 2307 2307 2299 2290 2287

120 2185 2185 2185 2170 2159 2154

180 2122 2122 2122 2118 2113 2111

240 2073 2073 2073 2128 2118 2115

300 2114 2114 2114 2232 2227 2253





Default run Default run 2100





617


60 772 772 772 772 773 773

120 761 761 761 761 761 760

180 746 746 746 754 750 754

240 705 705 705 756 756 756

300 693 693 693 749 754 755

252





60 kg/ha at 8 cm 1.6 779

60 kg/ha at 10 cm 1.6 779





233


60 351 351 351 351 350 351

120 245 245 245 245 295 295

180 296 296 296 296 296 296

240 243 243 243 244 244 244

300 240 240 240 241 242 242





60 kg/ha at 0 cm 1.6 401

60 kg/ha at 2 cm 1.6 401

60 kg/ha at 4 cm 1.6 401

60 kg/ha at 6 cm 1.6 357

60 kg/ha at 8 cm 1.6 357

253

3.2.3A Nacocolevu Optimisation Treatments

Table 3.84A shows the yield at different planting time for Nacocolevu for 2030-2090


Yield (kg/ha)

Planting Time

2030 A1B Emission

2030 A2 Emission

2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

2090 A2 Emission

January 2 0 0 0 0 0 0

February 1 0 0 0 0 0 0

March 1 52 179 0 0 0 0

April 1 3635 4369 1094 1826 2 0

May 1 5670 5876 4887 5593 1416 971

June 1 7066 6900 8141 9385 4729 3421


7476 7523 6222 7580 3792 3634

August 1 8178 8522 6090 7547 3399 3205

September 1 5081 5674 3385 4547 1573 1072

October 1 2207 2796 1199 2006 432 288

November 1 1081 1448 158 595 0 0

December 1 0 0 0 0 0 0

254

Table 3.85A shows the yield at different row spacing for Nacocolevu for 2030-2090 A1B

and A2 emission scenario.

Yield (kg/ha)

Row Spacing ( cm)


2030 A1B

2030 A2 Emission

2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

2090 A2 Emission

75 (default run)

5 4908 5521 1993 2949 523 314

30 11 4385 4856 2002 3031 604 315

40 8 4598 4932 1960 3145 600 314

50 6 4833 5365 2013 2902 602 315

80 4 4484 5609 2121 2947 526 252

100 3 4694 6417 2182 2947 522 251


amount for Nacocolevu for 2030-2090 A1B and A2 emission scenario.

Yield (kg/ha)


4908 5521 1993 2949 523 314

Irrigation ( mm)

2030 A1B Emission

2030 A2 Emission

2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

2090 A2 Emission

1.6 5102 5483 2332 2957 678 318

4.0 5077 5440 2386 2914 674 318

6.4 5125 5407 2390 2951 672 317

8.0 5107 5326 2458 3020 677 317

9.6 4988 5430 2375 3012 680 318

12.0 4985 5396 2375 3075 676 319

14.4 4967 5352 2349 2997 678 321

16.0 5043 5368 2375 3131 675 320

255

Table 3.87A shows the planting depth with corresponding yield for Nacocolevu for

2030-2090 A1B and A2 emission scenario.

Yield (kg/ha)


2030 A1B Emission

2030 A2 Emission

2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

2090 A2 Emission

1.5 (default run)

4908 5521 1993 2949 523 314

2 3525 4793 1768 2394 341 157

4 1996 2284 929 1431 3 0

6 1255 1570 363 777 0 0

8 449 883 0 12 0 0


depth) and corresponding yield for Nacocolevu under 2030 A1B emission scenario.



4908


60 3030 3030 3030 3147 3300 3470

120 4598 4598 4598 4793 4836 4838

180 4902 4902 4902 4924 4803 4707

240 4799 4799 4799 4578 4726 4715

300 4690 4690 4690 4867 4997 4984

256

Table 3.89A shows the optimum fertiliser and irrigation management for Nacocolevu

under 2030 A1B emission scenario.



300 kg/ha at 8 cm 6.4 5818


depth) and corresponding yield for Nacocolevu under 2030 A2 emission scenario.



5611


60 2966 2966 2966 3103 3255 3364

120 4589 4589 4589 4554 4844 4911

180 5107 5107 5107 5146 5204 5054

240 5138 5138 5138 4943 5189 5259

300 5332 5332 5332 5398 5404 5473


under 2030 A2 emission scenario.



Default run Default run 5549

257





2014


60 3374 3374 3374 3397 3424 3431

120 3342 3342 3342 3110 3036 3021

180 2652 2652 2652 2704 2703 2686

240 2656 2656 2656 2442 2490 2552

300 2491 2491 2491 2469 2579 2596





60 at 6 cm 8 3325





2988


60 3447 3447 3447 3556 3784 3808

120 3902 3902 3902 3758 3850 3737

180 3653 3653 3653 3282 3304 3323

240 3259 3259 3259 3243 3355 3356

300 3110 3110 3110 3048 3087 3075

258


under 2055 A2 emission scenario.



60 kg/ha at 10 cm 16 mm 3751





519


60 1144 1144 1144 1118 1117 1107

120 1086 1086 1086 1083 1078 1074

180 962 962 962 966 965 867

240 851 851 851 850 830 831

300 846 846 846 827 734 683





60 kg/ha at 0 cm 9.6 974

60 kg/ha at 2 cm 9.6 974

60 kg/ha at 4 cm 9.6 974

259





315


60 749 749 749 741 631 629

120 728 728 728 729 726 640

180 539 539 539 450 461 461

240 454 454 454 453 452 432

300 448 448 448 373 355 333

Table 3.99A shows the optimum fertiliser and irrigation management.



60 kg/ha at 0 cm 16 612

60 kg/ha at 2 cm 16 612

60 kg/ha at 4 cm 16 612

260

Appendix 4: Simulation of other potato varieties

Appendix 4 provides the model output for potential and non-potential for current and

future climate simulations using Sebago and Russet Burbank varieties. The model

simulation was conducted only for Banisogosogo using PCCSP A1B and A2

emission scenarios were used. Likewise, it also shows the optimisation results for

planting date, row spacing, fertiliser application, irrigation application and planting

depth.

4A Cultivar selection

4.1A Sebago current climate simulations

4.1.1 A Banisogosogo potential simulation



Date Crop

Age Growth Stage

Biomass (kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

9 Jul

7 Emergence 12 0.02 0 0 0 0 0 0 0 1

4 Aug


1601 2.43 0 0 0 0 0 0 0 2

20 Sep

80 Harvest 12433 6.14 0 0 0 0 0 0 0 2


Development Stage Environment Stress Time

Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth Photosynth Growth


26 27.6 20.5 13.9 11.12 392.0 18.2 0.000 0.000 0.000 0.000 0.000


48 28.9 21.8 17.4 11.61 392.1 64.5 0.000 0.000 0.000 0.000 0.000

Planting-Harvest

81 28.4 21.3 16.2 11.4 392.1 134.3 0.000 0.000 0.000 0.000 0.000

261

4.1.2 A Banisogosogo non-potential simulation



Date Crop

Age Growth Stage

Biomass (kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

9 Jul

7 Emergence 11 0.01 0 5 42.8 0 0 0 0 1

2 Aug


251 0.37 0 89 35.5 0 0.16 0 0 2

20 Sep

80 Harvest 2898 0.22 0 520 17.9 0.21 0.03 0 0 2



e Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


24 27.6

20.4

13.6 11.11 392.0

18.2 37.7 0.000 0.000 0.117 0.155


50 28.8

21.7

17.4 11.59 392.1

115.7

138.7 0.132 0.202 0.009 0.032

Planting-Harvest

80 28.4

21.3

16.2 11.4 392.1

134.3

184.5 0.082 0.125 0.040 0.066

4.2A Russet Burbank current climate simulation

4.2.1 A Banisogosogo potential simulation



Date Crop

Age Growth Stage

Biomass (kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

9 Jul

7 Emergence 12 0.02 0 0 0 0 0 0 0 1

24 Aug


5589 8.33 0 0 0 0 0 0 0 2

262

20 Sep

80 Harvest 9316 7.10 0 0 0 0 0 0 0 2


Development Stage Environment Stress Time

Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen Photo synth

Growth Photosynth Growth


46 28.2 21.0 14.7 11.24 392.0 84.3 0.000 0.000 0.000 0.000 0.000


48 28.8 21.8 18.6 11.76 392.0 49.6 0.000 0.000 0.000 0.000 0.000

Planting-Harvest

81 28.4 21.3 16.2 11.4 392.1 134..3 0.000 0.000 0.000 0.000 0.000

4.2.2 A Banisogosogo non-potential simulation

Table 4.6A shows non potential simulated crop and soil status at main development


Date Crop

Age Growth Stage

Biomass (kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

9 Jul

7 Emergence 11 0.01 0 5 42.8 0 0 0 0 1

21 Aug


2559 3.89 0 1130 44.2 0 0.10 0 0 2

20 Sep

80 Harvest 6239 5.55 0 1878 30.1 0.19 0.07 0 0 2



e Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


43 28.2

21.0

14.6 11.22 392.0

84.1 109.2 0.000 0.000 0.079 0.104


31 28.7

21.7

18.4 11.74 392.0

49.8 123.6 0.083 0.183 0.021 0.065

Planting-Harvest

80 28.4

21.4

16.2 11.4 392.1

134.3

241.0 0.032 0.070 0.050 0.080

263

4.3A Future climate Sebago simulations

4.3.1A Banisogosogo potential 2030 medium emission scenario


stages.

Date Crop

Age Growth Stage

Biomass (kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

10 Jul

8 Emergence 12 0.02 0 0 0 0 0 0 0 1

12 Aug


2957 4.51 0 0 0 0 0 0 0 2

20 Sep

80 Harvest 9613 6.04 0 0 0 0 0 0 0 2



e Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


33 29.1

21.8

13.9 11.17 490.0

84.8 0.0 0.000 0.000 0.000 0.000


40 30.1

23.1

18.2 11.67 490.0

65.0 0.0 0.000 0.000 0.000 0.000

Planting-Harvest

81 29.6

22.5

16.2 11.40 490.0

151.8

0.0 0.000 0.000 0.000 0.000

4.3.2 A Banisogosogo potential 2030 high emission scenario


stages.

Date Crop

Age Growth Stage

Biomass (kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

10 Jul

8 Emergence 12 0.02 0 0 0 0 0 0 0 1

10 Aug


2628 4.00 0 0 0 0 0 0 0 2

264

20 Sep

80 Harvest 10236 6.09 0 0 0 0 0 0 0 2

Table 4.11A shows environmental and stress factors.


e Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


31 28.8

21.7

13.9 11.15 520.0

86.2 0.0 0.000 0.000 0.000 0.000


42 29.9

22.7

18.0 11.65 520.0

66.1 0.0 0.000 0.000 0.000 0.000

Planting-Harvest

81 29.4

22.3

16.2 11.40 520.0

154.4

0.0 0.000 0.000 0.000 0.000

4.3.3 A Banisogosogo potential 2055 medium emission scenario


stages.

Date Crop

Age Growth Stage

Biomass (kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

11 Jul

9 Emergence 12 0.02 0 0 0 0 0 0 0 1

24 Aug


6068 9.05 0 0 0 0 0 0 0 2

20 Sep

80 Harvest 7949 6.42 0 0 0 0 0 0 0 2



e Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


44 30.2

22.9

14.9 11.25 550.0

97.0 0.0 0.000 0.000 0.000 0.000


28 30.7

23.7

18.6 11.76 550.0

58.0 0.0 0.000 0.000 0.000 0.000

Planting- 81 30. 23. 16.2 11.40 550. 157. 0.0 0.000 0.000 0.000 0.000

265

Harvest 3 2 0 1

4.3.4 A Banisogosogo potential 2055 high emission scenario


stages.

Date Crop

Age Growth Stage

Biomass (kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

11 Jul

9 Emergence

12 0.02

0 0 0 0 0 0 0 1

22 Aug


7361 10.89

0 0 0 0 0 0 0 2

20 Sep

80 Harvest 9945 6.53

0 0 0 0 0 0 0 2

Table 4.15A show the environmental and stress factors.


e Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


42 30.0

22.7

14.5 11.23 992.1

97.0 0.0 0.000 0.000 0.000 0.000


30 30.5

23.4

18.7 11.75 992.1

58.0 0.0 0.000 0.000 0.000 0.000

Planting-Harvest

81 30.1

23.0

16.2 11.40 992.0

157.1

0.0 0.000 0.000 0.000 0.000

4.3.5A Banisogosogo non potential 2030 medium emission scenario


development stages.

Date Crop

Age Growth Stage

Biomass (kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

11 Jul

9 Emergence 10 0.01 0 4 42.8 0 0 0 0 1

10 Aug


580 0.88 0 272 46.9 0 0.13 0 0 2

266

20 Sep

80 Harvest 4814 2.19 0 1175 24.4 0.31 0.03 0 0 2



e Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


30 29.1

21.9

13.6 11.16 490.0

84.8 50.6 0.000 0.000 0.099 0.134


42 30.1

22.9

18.0 11.65 490.0

91.0 65.0 0.206 0.300 0.000 0.029

Planting-Harvest

81 29.6

22.5

16.0 11.40 490.0

151.8

226.5 0.107 0.156 0.037 0.065

4.3.6 A Banisogosogo non-potential 2030 high emission scenario


development stages.

Date Crop

Age Growth Stage

Biomass (kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

10 Jul

8 Emergence 10 0.01 0 4 42.8 0 0 0 0 1

8 Aug


514 0.78 0 234 45.5 0 0.15 0 0 2

20 Sep

80 Harvest 5587 2.22 0 1260 22.5 0.28 0.03 0 0 2



e Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


29 28.8

21.6

13.9 11.14 520.0

59.8 55.9 0.000 0.000 0.108 0.146

Begin Tuber-

44 29.9

22.7

17.6 11.64 520.0

92.6 163.5 0.183 0.272 0.000 0.032

267

Maturity Planting-Harvest

81 29.4

22.3

16.0 11.40 520.0

154.4

229.0 0.100 0.148 0.039 0.070

4.3.7A Banisogosogo non-potential 2055 medium emission scenario


development stages.

Date Crop

Age Growth Stage

Biomass (kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

12 Jul

10 Emergence 10 0.01 0 4 42.8 0 0 0 0 1

23 Aug


2632 4.01 0 1049 39.9 0.01 0.10 0 0 2

20 Sep

80 Harvest 4660 3.93 0 1545 33.2 0.30 0.05 0 0 2



e Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


42 30.2

22.9

14.6 11.24 550.0

97.0 113.3 0.000 0.006 0.069 0.097


29 30.7

23.6

185 11.76 550.0

58.0 108.3 0.194 0.292 0.001 0.051

Planting-Harvest

81 30.3

23.2

16.0 11.40 550.0

157.1

233.3 0.070 0.108 0.036 0.069

268

4.3.8A Banisogosogo non-potential 2055 high emission scenario


development stages.

Date Crop

Age Growth Stage

Biomass (kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

11 Jul

9 Emergence 10 0.01 0 4 42.8 0 0 0 0 1

18 Aug


2383 3.64 0 1007 42.3 0 0.11 0 0 2

20 Sep

80 Harvest 5868 4.16 0 1722 29.4 0.21 0.05 0 0 2



e Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


38 29.9

22.6

14.3 11.21 992.1

89.2 92.5 0.000 0.003 0.082 0.109


34 30.5

23.4

18.3 11.72 992.1

65.9 129.8 0.128 0.206 0.000 0.046

Planting-Harvest

81 30.1

23.0

16.0 11.40 992.0

157.1

233.1 0.054 0.088 0.038 0.070

4.3.9 A Banisogosogo non-potential 2090 medium emission scenario


development stages.

Date Crop

Age Growth Stage

Biomass (kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

14 Jul

12 Emergence 10 0.01 0 4 42.8 0 0 0 0 1

14 Sep


4588 6.82 0 1491 32.5 0.13 0.18 0 0 2

20 Sep

80 Harvest 4349 5.96 0 1487 34.2 0.34 0.36 0 0 2

269



e Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


62 31.3

24.1

15.7 11.41 700.0

120.9

184.4 0.074 0.131 0.144 0.175


7 32.5

25.4

21.5 11.94 700.0

28.1 29.4 0.244 0.292 0.337 0.370

Planting-Harvest

81 31.3

24.2

16.0 11.40 700.0

158.5

228.5 0.076 0.125 0.140 0.166

4.4A Optimisation treatment

4.4.1A Banisogosogo optimisation treatment

Table 4.26A shows the yield at different planting time for Banisogosogo under A1B and


Yield (kg/ha)

Planting Time

2030 A1B Emission

2030 A2 Emission

2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

January 2 0 6 0 0 0

February 1 0 0 0 0 0

March 1 0 47 0 0 0

April 1 1442 1982 124 562 0

May 1 4684 5978 1906 3105 0

June 1 5004 6085 1933 2897 155


4356 5158 2840 4015 597

August 1 4027 4744 1766 3039 125

September 1 871 1017 514 655 191

October 1 2004 2774 1480 2024 271

November 1 56 150 0 1 0

December 1 3 36 0 0 0

270

Table 4.27A shows the yield at different row spacing for Banisogosogo under A1B and


Yield (kg/ha)

Row Spacing ( cm)


2030 A1B Emission

2030 A2 Emission

2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

75 (default run)

5 2947 3703 1485 2566 17

30 11 2988 4081 1492 2464 18

40 8 2964 1005 1493 2479 18

50 6 2949 3793 1487 2555 17

80 4 2956 3697 1486 2599 17

100 3 2984 3711 1488 2603 10

Table 4.28A shows the yield under different irrigation amount for Banisogosogo under


Yield (kg/ha)


2947 3703 1485 2566 17

Irrigation ( mm)

2030 A1B Emission

2030 A2 Emission

2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

1.6 4036 5060 1673 2656 20

4.0 4813 5784 1511 2587 12

6.4 5251 6465 1425 2275 3

8.0 4954 6189 1346 2163 1

9.6 4577 6087 1225 2041 1

12.0 4198 5838 1220 1986 1

14.4 3958 5447 1257 1959 1

271

16.0 4157 5625 1220 1972 1

Table 4.29A shows the Planting Depth with Corresponding Yield for Banisogosogo

under A1B and A2 emission scenario.

Yield (kg/ha)


2030 A1B Emission

2030 A2 Emission

2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

1.5 (default run)

2947 3703 1485 2566 17

2 2389 3229 1194 1805 0

4 1249 1537 369 772 0

6 565 934 0 31 0

8 1 103 0 0 0

10 0 0 0 0 0


depth) and corresponding yield for Banisogosogo 2030 A1B emission scenario.



2947


60 3566 3566 3566 3593 3624 3666

120 3799 3799 3799 3841 3868 3872

180 3827 3827 3827 3957 4051 4060

240 3949 3949 3949 4084 4156 4177

300 3947 3947 3947 4147 4239 4256


2030 A1B Emission.

Fertiliser Application (kg/ha at 10 cm) Irrigation Application ( mm) Yield (kg/ha) 300 kg/ha 6.4 4641

272


depth) and corresponding yield Banisogosogo 2030 A2 emission scenario.



3703


60 4652 4652 4652 4685 4527 4543

120 4587 4587 4587 4720 4847 4840

180 4732 4732 4732 4877 4914 5086

240 4807 4807 4807 4942 5221 5237

300 4868 4868 4868 5155 5306 5367


2030 A2 Emission Scenario.

Fertiliser Application (kg/ha at 10 cm) Irrigation Application ( mm)

Yield (kg/ha)

300 kg/ha 6.4 5749





1485


60 1651 1651 1651 1702 1699 1692

120 1913 1913 1913 1843 1717 1712

180 1965 1965 1965 1999 1922 1936

240 2182 2182 2182 2112 2026 2012

300 2241 2241 2241 2164 1971 1789

273


2055 A1B emission scenario.


300 kg/ha at 0 cm 1.6 2211

300 kg/ha at 2 cm 1.6 2211

300 kg/ha at 4 cm 1.6 2211


depth) and corresponding yield for Banisogosogo 2055 A2 emission scenario.



2566


60 2487 2487 2487 2480 2478 2498

120 2946 2946 2946 2950 2814 2661

180 3000 3000 3000 3072 3064 2905

240 3042 3042 3042 3082 2983 2919

300 3089 3089 3089 3035 3010 2900


2055 A2 emission scenario.

Fertiliser Application (kg/ha at x cm) Irrigation Application ( mm)

Yield (kg/ha)

300 at 0 cm 1.6 3130

300 at 2 cm 1.6 3130

300 at 4 cm 1.6 3130

.

274





17


60 38 38 38 29 29 29

120 19 19 19 20 20 20

180 11 11 11 11 12 12

240 10 10 10 10 11 11

300 5 5 5 6 5 5


2090 A1B emission scenario.

Fertiliser Application (kg/ha at x cm) Irrigation Application ( mm)

Yield (kg/ha)

60 at 0 cm 1.6 991

60 at 2 cm 1.6 991

60 at 4 cm 1.6 991

275

4.5A Russet Burbank future climate simulation

4.5.1 A Banisogosogo non-potential 2030 high emission scenario


development stages.

Date Crop

Age Growth Stage

Biomass (kg\ha)

LAI Leaf Number

Crop (kg\ha)

%N Water Stress

Nitrogen Stress

P1 Stress

P2 Stress

RSTG

1 Jun

0 Start Sim 0 0 0 0 0 0 0 0 0 5

2 Jul

0 Sowing 0 0 0 0 0 0 0 0 0 7

10 Jul

8 Emergence 10 0.01 0 4 42.8 0 0 0 0 1

14 Sep


6153 8.96 0 1966 32.0 0.03 0.17 0 0 2

20 Sep

80 Harvest 5835 7.66 0 1967 33.7 0.18 0.38 0 0 2



e Span Days


T max (oC)

T min (oC)


Photop (day) hr

CO2 ppm

Rain ( mm)

Evapo Trans

Water Nitrogen

Photo synth

Growth

Photosynth

Growth


66 29.3

22.2

15.5 11.39 520.0

84.8 125.0 0.005 0.029 0.133 0.163


7 30.6

23.5

21.5 11.94 520.0

91.0 27.4 0.088 0.156 0.357 0.389

Planting-Harvest

81 29.4

22.3

16.0 11.40 520.0

151.8

154.4 0.011 0.037 0.139 0.167

276

4.6A Optimisation treatment

4.6.1A Banisogosogo optimisation treatment

Table 4.42A shows the yield at different planting time for Banisogosogo for Russet

Burbank variety for 2030 A2 emission scenario.

Yield (kg/ha)

Planting Time

2030 A1B Emission

2030 A2 Emission

2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

2090 A2 Emission

January 2 0 0 0 0 0 0

February 1

0 0 0 0 0 0

March 1 0 0 0 0 0 0

April 1 0 0 0 0 0 0

May 1 0 0 0 0 0 0

June 1 22 165 0 0 0 0


342 843 0 16 0 0

August 1 29 109 0 0 0 0

September 1

210 318 68 124 8 2

October 1 54 566 0 0 0 0

November 1

0 0 0 0 0 0

December 1

0 0 0 0 0 0

277

Table 4.43A shows the yield at different row spacing under 2030 A2 emission scenario

for Russet Burbank variety for Banisogosogo.

Yield (kg/ha)

Row Spacing ( cm)


2030 A1B Emission

2030 A2 Emission

2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

2090 A2 Emission

75 (default run)

5 0 37 0 0 0 0

30 11 1 38 1 1 1 1

40 8 1 38 1 1 1 1

50 6 1 37 1 1 1 1

80 4 0 36 0 0 0 0

100 3 0 22 0 0 0 0


amount for Russet Burbank for Banisogosogo under 2030 A2 emission scenario.

Yield (kg/ha)


0 37 0 0 0 0

Irrigation ( mm) 2030 A1B Emission

2030 A2 Emission

2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

2090 A2 Emission

1.6 1 88 0 0 0 0

4.0 0 50 0 0 0 0

6.4 0 30 0 0 0 0

8.0 0 18 0 0 0 0

9.6 0 18 0 0 0 0

12.0 0 31 0 0 0 0

14.4 0 31 0 0 0 0

16.0 0 31 0 0 0 0

278

Table 4.45A shows the planting depth with corresponding yield for Russet Burbank for

Banisogosogo under 2030 A2 emission scenario.

Yield (kg/ha)


2030 A1B Emission

2030 A2 Emission

2055 A1B Emission

2055 A2 Emission

2090 A1B Emission

2090 A2 Emission

1.5 (default run)

0 37 0 0 0 0

2 0 12 0 0 0 0

4 0 0 0 0 0 0

6 0 0 0 0 0 0

8 0 0 0 0 0 0

10 0 0 0 0 0 0


depth) and corresponding yield for Banisogosogo under 2030 A1B emission.



0


60 7 7 7 7 7 7

120 3 3 3 3 3 3

180 1 1 1 1 1 1

240 0 0 0 0 0 0

300 0 0 0 0 0 0

279





60 kg/ha at 0-8 cm 1.6 6

60 kg/ha at 10 cm 1.6 7


depth) and corresponding yield for Banisogogsogo under 2030 A2emission.



37


60 150 150 150 151 151 151

120 104 104 104 104 104 87

180 85 85 85 85 48 48

240 46 46 46 29 16 16

300 15 15 15 7 4 2





60 kg/ha at 6 cm 1.6 151

60 kg/ha at 8 cm 1.6 151

60 kg/ha at 10 cm 1.6 151

Date post:	11-Oct-2018
Category:	Documents
Upload:	lydat
View:	213 times
Download:	0 times