AGRICOLA - nbri.org.na · sonal and 80 % or more of the total annual rainfall of 373 ± 158,6 mm...

AGRICOLA2011

Number 21

ENQUIRIES TO:MINISTRY OF AGRICULTURE, WATER AND FORESTRY

DIRECTORATE: RESEARCH AND TRAININGPRIVATE BAG 13184

WINDHOEKNAMIBIA

Compiled byPaul van der Merwe

AGRICOLA 2011 3

PUBLISHED BY:MINISTRY OF AGRICULTURE, WATER AND FORESTRY

DIRECTORATE: RESEARCH AND TRAININGPRIVATE BAG 13184

WINDHOEKNAMIBIA

Copyright © MAWF 2011

All rights reserved. No part of this book may be reproduced in any form, or by any microfilm, electronic or mechanical means, including information storage and retrieval devices or

systems, without prior written permission from the publisher.

COVER PHOTOGRAPH:Tiaan van der Merwe

ISSN: 1015-2334

LAYOUT AND GRAPHICS BY:Publish Pro, Windhoek

[email protected]

PRINTED BY:John Meinert Printing

Windhoek

TABLE OF CONTENTS

1. Foreword – Mr. A.N. Ndishishi, Permanent Secretary: Ministry of Agriculture, Water and Forestry ......................... 5

2. Ecological Dynamics of Central Namibia’s Savannas: Part 1 – Grass Ecology – Axel Rothauge ............................................ 7

3. Ecological Dynamics of Central Namibia’s Savannas: Part 2 – Bush Ecology – Axel Rothauge ............................................. 14

4. Breed Preferences, Production Performance and Management of Dairy Cattle Among Selected Smallholder Dairy Farmers of Zimbabwe – Lucia N. Marius, E.V. Imbayarwo-Chikosi, B.T. Hanyani-Mlambo and C. Mutisi .................. 26

5. Diet Selection of Free Ranging Horses in the Highland Savanna of Namibia: A Case Study at Seeis Farm – Lucia N. Marius and Axel Rothauge ........................................................................................................................................ 34

6. Effective Communication of Climate Change by Extension Agents – F.N. Mwazi and J. Ndokosho .................................. 39

7. Accelerating Landscape Incision and the Downward Spiralling Rain Use Efficiency of Namibian Rangelands – Hugh Pringle, Ibo Zimmerman and Ken Tinley ...................................................................................................................... 43

AGRICOLA 2011 5

As we come to the close of yet another year, the time has also come to introduce the 2011 Agricola. We all appreciate the important role that agriculture plays, both economically as well as sustaining the livelihood of thousands of people in Namibia and it is therefore good to reflect on a couple of agricultural hot points via this publication.

Agriculture is a very wide field and covers an extensive range of topics, many of which overlap or are influenced by others. It is therefore important to obtain and discuss information that is relevant to conditions that exist today and to understand how different factors influence one another under Namibian conditions. In any field it is important to receive information the sooner the better to be able to use it to its greatest effect, and this is as true of the sphere of agriculture as of any other domain.

The Ministry of Agriculture, Water and Forestry has made it one of its priorities to ensure that a wide range of pertinent information is accessible to all involved in agriculture. Although many agricultural publications exist worldwide, the Agricola is unique in that it provides mainly scientific information obtained and generated in Namibia, and under

Namibia’s unique conditions. By supporting the process of information dissemination, we believe that our natural resources, which in this country are very fragile, can be used to optimal effect without being depleted. This, in turn, will sustain and improve the livelihood of our people.

To enable the agricultural sector to be proactive, the dissemination of research findings and recommendations must be accurate and timeous. The Ministry of Agriculture, Water and Forestry focuses on identifying appropriate technical messages, adapting them to the various farming conditions and then circulating them to the largest possible number of farmers. It is also, though, an important objective of agricultural research to not only inform our farmers, but to disseminate the research results to as wide an audience as is possible, because it is only then that true research discussions can begin.

Increasing productivity is the major challenge to our farmers if they are to not only maintain, but also improve the competitiveness of their produce at home and on international markets. One does not have to stress how important the agricultural sector is to any economy – and when one considers the extremely arid nature of this country and its variable rainfall patterns, it is not surprising that the perseverance and innovative abilities of Namibian farmers is imperative and must expand to meet the even increasing demand. The basic essentials are the production of enough food for household consumption and also having access to markets that provide favourable returns for any surplus products. The objective of the sector must be to increase productivity by achieving better yields from crops, as well as by increasing livestock off take. However, despite the important role agriculture plays in supplying food to the majority of the people, our agricultural productivity remains regrettably low and much improvement still has to be made. If Namibia wants to maintain its status as a food-producing country, as well as its export capacity, then agricultural research is imperative and it must expand its capacity to meet the demand for information.

I would like to wish a productive year to all those who are involved in making the best use of our natural resources and at the same time conserving them. May the information offered in this issue of the Agricola be of benefit to all and aid the passage of our country towards the vision that we have set ourselves – Vision 2030.

A.N. NdishishiPermanent Secretary

FOREWORD

MR. ANDREW NATANGWE NDISHISHI

Permanent Secretary,Ministry of Agriculture, Water and Forestry,

Private Bag 13184, Windhoek, Namibia

AGRICOLA 2011 7

ABSTRACT

The ecological processes of the three dominant savannas found in central Namibia, namely the Highland, Thornbush and Camelthorn savannas, are explained with a state-and-transition model that incorporates aspects of Clementsian succession as well as Walter’s two-layer theories. A savanna can persist either in a grassy or in a woody (bushy, bush-encroached) state. Within each, climax and pioneer states with higher and lesser production potential exist. The first part of this article describes the grassy states. Transitional forces and events that shift savannas between and within states are described in detail, as well as the opportunities and risks these changes bear for the land manager. A transition that should be avoided at all cost is that from pioneer grassy state to desertification, because desertification is believed to be irreversible in practice. A critical state occurs when the grassy is changed to a bushy state, but this is discussed in greater detail in Part 2 of this article.

INTRODuCTION

A savanna is a tropical or near-tropical rangeland that consists of a continuous herbaceous layer and an open, discontinuous woody layer. It is subject to pronounced seasonal fluctuations in terms of productivity and nutritive value (Skarpe, 1991). The balance between the herbaceous and the woody layer is determined by ecological processes such as climate (mainly rainfall and temperature), soil fertility, incidence of naturally-ignited fires, animal activity, etc., as well as by human factors such as anthropological fires, fire exclusion, selective defoliation of the various plant components, etc. When the grass – bush balance is disrupted, and there is an excessive densification of the woody component at the expense of the grasses, bush encroachment has occurred (Smit, Richter & Aucamp, 1999). Bush encroachment drastically reduces the grass-based carrying capacity of a rangeland (Adams & Werner, 1990). It has affected most of the savannas and dry woodlands (Bester, 1998) that cover 84 % of Namibia’s land area (Mendelsohn, Jarvis, Roberts & Robertson, 2009). It is a problem of national significance and has severe economic implications, causing beef producers an estimated loss of N$ 700 million per year in foregone income because of the drastically reduced production potential of cattle ranches (De Klerk, 2004). Bush encroachment has always existed in Namibia on a small scale, e.g. around oft-frequented

ECOLOGICAL DYNAMICS OF CENTRAL NAMIBIA’S SAVANNAS: PART 1 – GRASS ECOLOGY

AxEl RoTHAuGE

AGRA (Co-operative) Limited, Private Bag 12011, Windhoek, Namibia [email protected]

water holes (Vedder, 1934), but the inundation of whole landscapes by bush is a relatively recent phenomenon. Accordingly, a lot of research has yet to be done as our current knowledge of savanna ecological dynamics is far from complete (Ward, 2005). It is believed that these ecological processes are very similar in the three dominant savanna types found in central Namibia, i.e. the Highland, Thronbush and Camelthorn savannas (Giess, 1971).

When Westoby, Walker and Noy-Meir (1989) devised the state-and-transition model to better explain ecological dynamics in an environment in constant disequilibrium, it caused a paradigm shift in the perception of how semi-arid southern African savannas work and function (e.g. Milton & Hoffman, 1994; Rothauge, 2000; Smit, 2003). The state-and-transition theory includes elements of Clements’ (1928) dated classical vegetation succession theory and Walter’s (1971) two-layer model of competition between herbaceous and woody components of savanna vegetation. The proposed ecological model of central Namibia’s savannas is based mainly on these theories and models, applying in principle to the Highland savanna, the Thornbush savanna to its north and the Camelthorn savanna to its east. This series of papers investigates the ecological processes of central Namibia’s savannas (Joubert, Rothauge & Smit, 2008), firstly the ecology of grassy states and then that of woody states, so that the major problem of landscape-level bush encroachment can be better understood and dealt with.

ECOLOGICAL ChARACTERISTICS OF ThE hIGhLAND SAVANNA OF CENTRAL NAMIBIA

Most of the research for this paper was performed on the farms Neudamm and Krumhuk in the Highland savanna of central Namibia, which will be briefly characterised below. It is one of the smaller vegetation types, occupying only 45 000 km2 or 5,5 % of Namibia’s land area (calculated from Coetzee, 1998). It is probably the vegetation type that has the longest been exposed to commercial ranching; since the late nineteenth century (Rawlinson, 1994). Highland savanna occurs at altitudes of 1 350 m to 2 200 m above sea level and with some extremely steep slopes of more than 30º. The terrain is highly broken and undulating. Typical mountain soils are lithic leptosols, overlying base material of sandstone and metamorphic schists. They are shallow, contain a lot of gravel, extremely low in organic matter content, leached (pH = 5,5), relatively infertile (e.g.

8 AGRICOLA 2011 AGRICOLA 2011 9

phosphorus content < 25 ppm) and erodible because of their low clay content (< 8 %). A surface layer of quartzitic pebbles that reaches a cover of almost 100 % on steep slopes is a notable feature and influences vegetation dynamics by serving as a mulch that improves soil moisture content (Joubert, 1997).

Typically for this sub-tropical area, precipitation is sea-sonal and 80 % or more of the total annual rainfall of 373 ± 158,6 mm occurs from January to March (Rothauge, 2008). Rainfall is highly variable with a coefficient of variation of more than 40 % and decreases from north-east to south-west. Due to the high evaporation rate, the annual water deficit exceeds 2 000 mm (Mendelsohn et al., 2009). In summer, average maximum temperatures (about 29 ºC) are cooler than in lower-lying neighbouring savannas, while winters are fairly cold (average minimum temperature is 3 ºC, but –9 ºC is recorded regularly) and frost occurs about 20 nights per year (Mendelsohn et al., 2009). The early on-set of low night-time temperatures results in a very short vegetative growing period of 40 to 60 days (Coetzee, 1998).

Giess (1971) described the Highland savanna as characterised by woody species including Acacia hereroensis, Acacia hebeclada, Acacia reficiens, Albizia anthelmintheca, Dombeya rotundifolia, Euclea undulata, Ozoroa crassinervia, Rhus marlothii and Tarchonanthus camphoratus. Notably, Acacia mellifera is missing from this list of characteristic species, despite it being the dominant woody species in large parts of the Highland savanna today, and recognised as the major species involved in bush thickening (Bester, Van Eck, Kölling, Van Rooyen & Prinsloo, 2002). Climax grass species include Anthephora pubescens, Brachiaria nigropedata, Cymbopogon spp., Digitaria eriantha and Heteropogon contortus, Sub-climax grasses such as Eragrostis nindensis and Schmidtia pappophoroides are usually the most abundant (e.g. Kellner, 1986; Joubert, 1997). Kellner (1986) described the Highland savanna as mostly an Acacia hereroensis – Eragrostis nindensis alliance but, due to the topographical diversity, it is quite heterogeneous.

The Highland savanna has a high degree of animal and plant biodiversity and endemism in comparison to other regions of Namibia (Barnard, 1998). The higher altitude biomes (e.g. the Auas mountains) support a number of reptile, insect and plant species with a severely restricted range. Despite this, only 0,2 % of the Highland savanna is managed as government-protected conservation areas (Barnard, 1998). Due to the undulating terrain and the proximity of mountain refuges, large ungulate herbivores (e.g. kudu, gemsbok, hartebeest, warthog) and their associated predators (e.g. jackal, caracal, cheetah, leopard) have always been common and occur freely. Their movement is little impeded by obstructions associated with human utilisation of the range, such as fences.

The predominant land use in the Highland savanna is ranching for profitable beef production on privately owned farmland. Livestock is free-ranging within large camps,

dependent on natural vegetation and receives only small inputs of nutritional supplements, labour and management (Pagot, 1992). Commercial farms are typically 5 000 hectares to 10 000 hectares in extent and sub-divided into camps to facilitate planned rotational grazing and livestock management. According to the International Development Consultancy (IDC; 2005), commercial ranchers maintain a fairly static stocking rate of about 15 ha per Large Stock Unit (LSU). Communally farmed areas constitute less than 15 % of the Highland savanna and are less sub-divided. While also focussing on beef production, much more use is made of sheep and goats. These domestic livestock species are stocked at a combined rate of about 4,5 ha/LSU (IDC, 2005). Communal production is predominantly for subsistence.

Irrespective of land ownership, livestock ranching activities have not been successful in maintaining the Highland savanna rangeland in a productive state. Degradation of soil and rangeland is widespread (Bester et al., 2002). This, together with a decline in the comparative profitability of meat production and the high income potential of tourism and ecotourism (Elkan, Van der Linde & Sherbourne, 1993), prompted a switch to game ranching on privately-owned guest and hunting farms. It was initially thought that conservancies, in which groups of ranches are partially managed together to enhance the sustainable utilisation of game animals, would reduce the negative impact of game ranching on the vegetation, but there is growing concern that this is not the case.

ECOLOGICAL STATES AND TRANSITIONS BETWEEN STATES: ThE BASIC MODEL

Typically, central Namibian savannas occur in two basic states that can be sub-divided further according to the advancement of degradation (Figure 1). A savanna can either be in a grassy state or in a bush-encroached (woody, bushy) state. In the grassy state, the continuous herbaceous and the discontinuous woody layer of the savanna are still in balance, giving the impression of an open rangeland in which the woody component occurs at “natural density”. In the bushy state, the density of the woody component has increased far beyond its probable “natural density” and is more continuous while the herbaceous layer is breaking up and discontinuous.

In-between these two basic states is a transitional state, not shown in Figure 1, that of bush seedling establishment. In savannas in the transitional state, the invasive woody species (in this case Acacia mellifera) proliferates and starts to invade the landscape. This is a crucial phase in the ecology of the savanna as intervention at this stage can nip bush encroachment in the bud.

Various degradation forces and events shift the condition of the savanna from climax to pioneer condition and within the woody state from vigorous to the fully mature, senescent condition. These transitions pose a threat to the land user, as they diminish the potential of the rangeland to

sustain animal production. Rehabilitative transitions work in the opposite direction to improve the condition of the savanna. They present opportunities for the land user to restore the rangeland’s production potential. Eventually, a full, comprehensive model emerges (Figure 2) that offers land managers a multitude of opportunities to manipulate the savanna to suit their objectives, within the constraints

Figure 1. The two basic states of a semi-arid savanna in central Namibia.

in climax condition

GRASSY STATE

in pioneer condition

in a vigorous condition

BUSHY STATE

in senescent condition

degradation

rehabilitation

degr

adat

ion

degr

adat

ion

of local ecological conditions. On a single ranch, or within a community’s core grazing area, the savanna can be in different states in different places and different forces may be at work pushing the savanna in different directions; thus creating a mosaic of heterogeneous patches and a plethora of conditions and transitions within the rangeland.

DESERTIFICATION

10: hot fire, active grass re-seeding followed by establishment rest

Climax grassy state

Pioneer grassy state

Vigorous bushy state

SeneScent buShy State

Bush seedling estaBlishment

9: drought, fire without active management

11: a

rtific

ial b

ush

cont

rol,

exce

ssive

gra

zing6: e

xclusio

n of fire

and browsing

5: episodic hot fire, unchanged grazing management

3: rainfall, n

o fire

3: rainfall, no fire

4: episodic hot fire, seed-set rest

1: e

xces

sive

gra

zing

, d

roug

ht

2: a

ctiv

e re

-see

ding

, es

tabl

ishm

ent r

est

7: n

atur

al p

rogr

essi

on

8: ra

infa

ll, c

old

fire

Figure 2. Complete schematic presentation of the five states of the Highland savanna and the eleven transitions between states (red = degradation and green = rehabilitation transitions).


ThE DETAILED ECOLOGICAL MODEL

1. The climax grassy state

The grassy state in good condition is dominated by mesophytic climax perennial grasses that can achieve a long-term average canopy cover in excess of 75 % (Rothauge, 2007) or a basal cover of up to 12 % (Joubert, 1997). More than 70 species of perennial grasses occur, of which about half are climax or sub-climax species (Van Eck, 2007). Depending on the topographical unit, the typically dominant grasses include Schmidtia pappaphoroides, Eragrostis nindensis, and Anthephora pubescens (Joubert, 1997; Rothauge, 2007). In dry years, annual grasses are almost absent, but in wet years, palatable annual grasses such as Melinis repens subsp. grandiflora and Enneapogon cenchroides form a significant part of the sward due to favourable conditions for germination (Rothauge, 2007). The grass-based carrying capacity varies with rainfall, ranging between 5 to 20 ha/LSU (Rothauge, 2007), although the average over the last 20 years was around 13 ha/LSU (Mendelsohn et al., 2009). Such a grass sward of palatable, perennial grasses is able to support highly productive beef cattle production systems.

A wide variety of forbs may contribute up to 15 % of the herbaceous layer (Rothauge, 2007). These include many leguminous species and perennials with fleshy roots of the Aizoon, Aptosimum, Chascanum, Indigofera and Lotononis genera. Their value to domestic livestock lies mainly in the contribution of small amounts of highly nutritious feed at critical times of the year, especially during the hot-dry season (Rothauge, 2006a).

Woody vegetation in the climax grassy state seldom exceeds a canopy cover of 10 % and is dominated by 2 m to 3 m high single-stemmed Acacia hereroensis (Joubert, 1997; Rothauge, 2007). This species appears to have a greater ability than other Acacia species to survive fire, regenerate through coppicing and tolerate harsh frost (Rothauge, 2006b). Woody species co-dominating the climax state are Rhus marlothii and Tarchonanthus camphoratus, which are also fire-tolerant (Rothauge, 2006b). Other species such as Boscia albitrunca and Albizia anthelmintheca form a small, but consistent component of this state on stony soil while Acacia erioloba is common only in the deep loamy soils of river valleys (Rothauge, 2007). Acacia mellifera is typically rare in the climax state on a landscape scale because its immature cohort is susceptible to frost, fire and competition by the vigorous, perennial grass sward.

Natural fires maintain the Highland savanna in a climax grassy state. Early-season dry thunderstorms often ignite fires naturally in high-lying areas,

because lightning has a propensity to strike mountain tops. Mountain fires are difficult to contain due to the rugged terrain. If natural fires follow on wet years, they are very hot (fierce) because of the accumulation of grass fuel and often ravage extensive areas. They either destroy woody plants outright, or burn off their above-ground parts (top-kill), forcing them to coppice again. Coppiced bushes are weakened and revert to an immature state in which they do not produce seeds. Their canopy is now within easy reach of browsers. Many coppiced bushes eventually succumb to the after-effects of the fire within 2 to 5 years (Rothauge, 2006b), or are killed off by next season’s fire.

Perennial grasses are not seriously affected by hot, early-season fires that follow on wet years. Copious or late rains during the preceding season result in surplus soil moisture carried over to the next season. This moisture induces an early-season, reserve-driven growth flush of perennial grasses that makes the tuft base green, moist and more tolerant of an early-season fire. Mountainous areas of the Highland savanna are thus more often in State 1 (climax state) than hilly or flat terrain, with appreciably less woody cover and an excellent sward of perennial grasses, and represent vegetation in a fire climax state.

2. Transition 1 towards a pioneer grassy state

When the climax grassy state degrades to the pioneer state, it does so along a continuum of changes in especially the herbaceous component (Joubert, 1997). It mirrors a classical retrogressive Clementsian succession. Transfor-mation of the species composition of the grass sward to-wards less desirable perennial species and later, annual grass species is the first sign of deterioration of the range-land. In the adjacent semi-arid camelthorn savanna to the east, this transformation occurs when long-term stocking rates exceed 30 ha/LSU or 15 kg cow mass per hectare (Figure 3) (Rothauge, 2006a). Transition towards a pio-neer grassy state decreases the sustainability of ranching

enterprises and increases farming risk and environmental variability. Mesophytic climax grasses decline in vigour and abundance while xerophytic perennial species fill the void. Valuable climax grasses that are most preferred by grazing herbivores such as Brachiaria nigropedata and Anthephora pubescens decline dramatically in response to increased grazing pressure. Preferred species such as Schmidtia pappophoroides and Melinis repens repens decline more gradually (Joubert, 1997; Rothauge, 2006a). Their place in the grass sward is taken by less preferred sub- climax perennial species such as Stipagrostis uniplumis and Eragrostis rigidior, followed later by unpalatable peren-nials such as Aristida stipitata and annual pioneer grasses such as Enneapogon cenchroides, Eragrostis cylindriflora and Chloris virgata. Such grazing pressure may arise when ranchers overstock parts of their farm to compensate for other parts that are already in a bush-thickened state with a low grazing capacity, or that have suffered an unplanned setback to their herbaceous production caused by a wild-fire, drought, locust or termite activity or poor grazing planning.

Typical most-preferred perennial grass species in central Namibia are Anthephora pubescens, Brachiaria nigropedata, various perennial Panicum and Digitaria species and, in north-central Namibia, Urochloa oligotricha. Typical preferred grass species are Schmidtia pappophoroides, Melinis repens repens, Eragrostis lehmanniana and, south-central Namibia, Centropaudia glauca and Stipagrostis obtusa. Typical least preferred grass species are Cenchrus ciliaris, Stipagrostis uniplumis, Eragrostis rigidior and Fingerhutia africana. Unpreferred species typically include the perennial Aristida species and many Eragrostis species, e.g. Eragrostis pallens.

3. The pioneer grassy state

In the grassy pioneer state, the savanna remains grassy and open, but is now dominated by annual pioneer grasses such

as Aristida stipoides, Enneapogon cenchroides, Eragrostis cylindriflora, Eragrostis porosa, Pogonarthria fleckii and Tragus racemosus. Small and low-yielding xeric perennial grasses such as Aristida congesta, Michrochloa caffra, Monelytrum luederitzianum and some resilient, grazing-tolerant, high-yielding bulky sub-climax species like Stipagrostis uniplumis are also still present (Joubert, 1997; Rothauge, 2007). Palatable annuals such as Melinis repens subsp. grandiflora may occur opportunistically in good rainy seasons, but do not dominate. While productivity in exceptional rainfall years may rival that of the grassy climax state, productivity in dry years is extremely low since few annual grasses germinate and grow, eventually leading to smaller cattle size and declining fertility (Figure 4). Grass basal cover can be as low as 0,5 % (Joubert, 1997). Poor cover results in capping of the soil surface by the harsh impact of raindrops during thunderstorms and sheet erosion, caused by the overland flow of rainwater rather than its infiltration (Rothauge, 2007).

Woody species still form a relatively small component of the degraded savanna with Acacia mellifera perhaps forming a larger proportion than before, since transitions towards encroachment may have occurred at microsite level. However, the abundance of dwarf xerophytic karroid shrubs like Eriocephalus luederitzianum, Leucosphaera bainesii, Salsola spp. and Monachne spp. increases noticeably to about 15 % (Rothauge, 2007). These small shrubs indicate aridification of the ecosystem. High browsing pressure may prevent further degradation towards a bushy state, despite excellent rains.

The degraded pioneer grassy state represents a hazardous situation for rangeland managers since in years of exceptional rainfall, bush colonisation is initiated and the transition towards landscape-level bush encroachment may easily occur. Even worse is the likelihood of further degradation towards desertification if this state remains self-perpetuating due to the export of rainwater and soil

Figure 3. Transformation of the grass sward of the Camelthorn savanna due to increasing the long-term stocking rate of beef cattle (Rothauge, 2006a).

Figure 4. Response of individual cattle to an increase in the long-term stocking rate of the camelthorn savanna (Rothauge, 2006a). ICP: inter-calving period, BCS: body condition score.

120

100

80

60

40

20

0

480

460

440

420

400

380

360low low-medium medium-high high

Stocking rate (kg cattle mass per ha)

Mass ICP BCS x 10 Calving rate

Em

pty

body

mas

s (k

g), I

CP

(day

s)

Cal

ving

rate

(%),

BC

S (x

10)

Com

para

tive

abun

danc

e an

d pr

oduc

tivity

Stocking rate (kg cattle mass per ha)15 25 35 45

least-preferred species

preferred species

most preferred species

unpreferred species

annual grass species

variability; risksustainability


nutrients, thus preventing succession towards a climax grassy state, even in years of high rainfall. If large areas have become degraded, seeds of large-seeded climax grasses such as Brachiaria nigropedata, Anthephora pubescens and Schmidtia pappophoroides are likely to be absent since these species are more prone to local extinction than small-seeded species such as Stipagrostis uniplumis, Eragrostis rigidior and the annual grasses (Owen-Smith & Danckwerts, 1997); further reducing the likelihood of a succession towards the climax grassy state. Such desertification is much more difficult to address than degradation.

4. Continued degradation towards desertification

Continued degradation of the pioneer grassy state due to severe grazing combined with a drought may severely affect basic ecological processes and cause desertification (Skarpe, 1991). Desertification results in the export of nutrients and moisture from the landscape and loss of ecosystem services. The ground water table is lowered due to reduced infiltration of rainwater into the soil due to poor permeability caused by soil capping, overland flow of rainwater and soil erosion (both sheet and gully erosion). The landscape aridifies despite normal rainfall; the so-called “man-made drought” is brought about. Increased exposure of the soil surface to the elements, especially the hard impact of raindrops and the sterilising effect of solar radiation, cause the top layer of soil to heat up. High soil temperature destroys soil microflora and plant seeds, which can no longer germinate in these hostile conditions. Weakened grasses attract grass-eating harvester termites, which compete with herbivorous ungulates for growing grasses. The landscape rapidly slides towards a state that cannot easily be restored and may well result in a permanent reduction of the production potential of the land.

Desertification is considered by Hoffman & Ashwell (2001) to be a serious problem in southern Africa. It represents degradation on a higher level. It is much more difficult in terms of time, knowledge and means to rectify than transformation of the grass sward or bush encroachment; and may not be reversible at all.

5. Transition 2 back towards the climax grassy state

The transition from the degraded pioneer grassy state back to the climax state is no fait accompli brought about by resting and de-stocking only. It requires active management such as overseeding with desired climax grass species followed by establishment rest from grazing, breaking the soil cap (e.g. by using animal impact or a pit plough/”kapploeg”) and slowing the overland flow of rainwater (e.g. by erecting brush barriers or ripping on the contour) in addition to destocking grazing herbivores. However, exclusion of livestock from Namibian farmland is not economically feasible, thus potentially preventing this favourable transition. If the pioneer state occurs on pebble mulch, the transition towards a climax state may be facilitated since the mulch improves infiltration and

reduces evaporation (Joubert, 1997); improving overall soil moisture conditions for mesophytic climax grasses to establish. However, this transition has not been researched and therefore it is difficult to determine whether this transition occurs, if at all. Indications from a similar savanna in Texas are that recovery from a pioneer towards a climax grassy state is reversible, intermittent and may take as long as 25 years (Fuhlendorf, Briske & Smeins, 2001).

Hypothetically, a fire with subsequent lenient grazing may initiate the transition towards the climax state by mineralisation of plant nutrients and clearing the seed bank of the dominant annual grasses and herbs. This allows the few perennial grass tufts that may still be present to expand and set seed. Long and efficient rest periods between grazing events aimed at seed production and subsequent seedling establishment of the mesophytic climax grasses would be essential for these grasses to establish, aided (obviously) by good rainfall. If these perennial grasses are locally extinct, active reseeding is required. The re-introduced grass seeds would have to be protected from insect and bird predation by encasing them in, for example, dung slurry or cakes. The emerging, highly palatable grasses would also have to be protected from over-utilisation by packed thorn branches, ring-fencing, etc. Bare ground offers opportunistic alien invasive plants such as various cacti and the grass Pennisetum setaceum, already present in the Highland savanna (Joubert & Cunningham, 2002), a foothold to invade degraded rangelands and are another factor that may prevent recovery towards the climax grassy state.

6. Transition 3 towards the establishment of bush seedlings

This transition towards establishment of Acacia mellifera seedlings is easily overlooked in the field because the establishing woody seedlings are unobtrusive and hidden within the grass sward. It represents a crucial ecological turning point that will be discussed in detail in the next part of this article.

IN SuMMARY

In summary, central Namibian savannas that are in the climax grassy state have great potential for beef cattle and game ranching. The biggest danger is that these climax states may degrade to a much less productive pioneer state because of excessive grazing and drought, or a combina-tion thereof. A critical contributing factor may be the re-duction of the cover of the topsoil, inducing aridification through sheet wash. Once in the pioneer grassy state, perpetuation of over-utilisation may induce desertification; an apparently permanent and highly unproductive condi-tion. Rehabilitation of a degraded grassy savanna may re-quire much more than mere resting and de-stocking, e.g. re-seeding with perennial grasses, breaking the soil cap, treating the symptoms of erosion, etc.

REFERENCES

ADAMS, F. & WERNER, W., 1990. The land Issue in Namibia: An Enquiry. Namibia Institute for Social and Economic Research, University of Namibia, Windhoek, Namibia.

BARNARD, P. (ed.), 1998. Biological Diversity in Namibia: A Country Study. Namibian National Biodiversity Task Force, Directorate of Environmental Affairs, Windhoek, Namibia.

BESTER, F.V., 1998. Major problem: bush species and densities in Namibia. Agricola 10: 1–3.

BESTER, F.V., VAN ECK, J.A.J., KÖLLING, H., VAN ROOYEN, B. & PRINSLOO, R., 2002. Bush encroachment with reference to the occurrence, die-back and regeneration of Acacia mellifera subsp. detinens in Namibia. Proc. 6th Nam. Rangelands Forum ch. 10, 18–19 June 2002, Ogongo Agric. College, Namibia.

CLEMENTS, F.E., 1928. Plant Succession and Indicators: A Definitive Edition of Plant Succession and Plant Indicators. H.W. Wilson Co., New York, USA.

COETZEE, M.E., 1998. Preliminary Agro-Ecological Zones. Addendum to Agricola 10.

DE KLERK, J.N., 2004. Bush Encroachment in Namibia. Report on Phase 1 of the Bush Encroachment Research, Monitoring and Management Project, Ministry of Environment and Tourism, Windhoek, Namibia.

ELKAN, W., VAN DER LINDE, E. & SHERBOURNE, R., 1993. Namibian agriculture and economy-wide policies. In: I. GOLDIN (ed.) Economic Reform, Trade and Agricultural Development ch. 5. St. Martin’s Press/OECD, Paris, France.

FUHLENDORF, S.D. BRISKE, D.D. & SMEINS, F.E., 2001. Herbaceous vegetation change in variable rangeland environments: The relative contribution of grazing and climatic variability. Applied Vegetation Science 4: 177–188.

GIESS, W., 1971. A preliminary vegetation map of South West Africa. Dinteria 4: 31–45.

HOFFMAN, T. & ASHWELL, A., 2001. Nature Divided: land Degradation in South Africa. Univ. Cape Town Press, Cape Town, South Africa.

INTERNATIONAL DEVELOPMENT CONSULTANCY, 2005. Study on land Productivity and Economic Farming units. Consultancy report prepared for Ministry of Agriculture, Water and Forestry (MAWF) in cooperation with GTZ, MAWF, Windhoek, Namibia.

JOUBERT, D.F., 1997. Grazing gradients in the Highland savanna. Dinteria 25: 69–86.

JOUBERT, D.F. & CUNNINGHAM, P.L., 2002. The distribution and invasive potential of Fountain Grass Pennisetum setaceum in Namibia. Dinteria 27: 37–47.

JOUBERT, D.F., ROTHAUGE, A. & SMIT, G.N., 2008. A conceptual model of vegetation dynamics in the semiarid Highland savanna of Namibia, with particular reference to bush thickening by Acacia mellifera. Journal of Arid Environments 72(12): 2201–2210.

KELLNER, K., 1986. ŉ Plantekologiese Studie van die Daan Viljoen-Wildtuin en Gedeeltes van die Plase Claratal en Neudamm. M.Sc. thesis, Potchefstroom University for Christian Higher Education, Potchefstroom, South Africa.

MENDELSOHN, J., JARVIS, A., ROBERTS, C. & ROBERTSON, T., 2009. Atlas of Namibia: A Portrait of the land and its People (3rd ed.). Sunbird Publishers, Cape Town, South Africa.

MILTON, S.J. & HOFFMAN, M.T., 1994. The application of state-and-transition models to rangeland research and management in arid succulent and semi-arid grassy Karoo, South Africa. African Journal of Range and Forage Science 11: 18–26.

OWEN-SMITH, N., & DANCKWERTS, J.E., 1997. Herbivory. In: R.M. COWLING, D.M. RICHARDSON & S.M. PIERCE (eds.). Vegetation of Southern Africa, ch. 17. Cambridge University Press, Cambridge, UK.

PAGOT, J., 1992. Animal Production in the Tropics and Subtropics. Macmillan, London, UK.

RAWLINSON, J., 1994. The Meat Industry of Namibia 1835 to 1994, ch. 8. Gamsberg Macmillan, Windhoek, Namibia.

ROTHAUGE, A., 2000. New ecological perceptions of arid rangelands. Agricola 11: 49–56.

ROTHAUGE, A., 2006a. The effect of frame size and stocking rate on diet selection of cattle and rangeland condition in the camelthorn savanna of east-central Namibia. Ph.D. thesis, University of Namibia, Windhoek, Namibia.

ROTHAUGE, A., 2006b. After-effects of a hot fire in 2001/2002. Unpublished report, Neudamm Agricultural College, Windhoek, Namibia.

ROTHAUGE, A., 2007. A decade of monitoring grazed rangeland transects at Neudamm. Unpublished report, Neudamm Agricultural College, Windhoek, Namibia.

ROTHAUGE, A., 2008. A century of rainfall statistics at Neudamm. Unpublished report, Neudamm Agricultural College, Windhoek, Namibia.

SKARPE, C., 1991. Impact of grazing in savanna ecosystems. Ambio 20: 351–356.

SMIT, G.N., 2003. The importance of ecosystem dynamics in managing the bush encroachment problem in southern Africa. Proceedings of the 7th International Rangelands Congress, pp. 14–22, 26 July to 1 August 2003, Durban, South Africa.

SMIT, G.N., RICHTER, C.G.F. & AUCAMP, A.J., 1999. Bush encroachment: an approach to understanding and managing the problem. In: N.M. TAINTON (ed.) Veld Management in South Africa ch. 10. University of Natal Press, Pietermaritzburg, South Africa.

VAN ECK, J., 2007. M.A.N. Müller: Grasses of Namibia. Ministry of Agriculture, Water & Forestry, Windhoek, Namibia.

VEDDER, H., 1934. Das Alte Südwestafrika: Südwestafrikas Geschicht bis zum Tode Mahareros 1890 (The old South West Africa: South West Africa’s History until the Death of Maharero 1890). Martin Warneck Verlag, Berlin, Germany.

WALTER, H., 1971. Ecology of Tropical and Subtropical Vegetation. Oliver and Boyd, Edinburgh, UK.

WARD, D., 2005. Do we understand the causes of bush encroachment in African savannas? Journal of Range and Forage Science 22: 101–106.

WESTOBY, M., WALKER, B.H. & NOY-MEIR, I., 1989. Opportunistic management for rangelands not at equilibrium. Journal of Range Management 42: 266–274.


ABSTRACT

Bush encroachment by Acacia mellifera is initiated by a wet spell of sufficient duration (three years) to allow this woody species to set viable seeds, followed by successful establishment of the seedlings. It is a rare, sporadic event in Namibia and in the arid zones of southern Africa. Bush encroachment follows a “sleep, then leap” mode of action rather than a steady and continuous “creep” mode. Bush seedling establishment is facilitated hugely by the exclusion of fierce veld fires, a grass sward weakened by grazing and by reduced browsing pressure. Hence, the most obvious manner to prevent bush encroachment from occurring in the first place is to set a fierce fire whenever rainfall was above the average for three years in a row, but only in those areas that are threatened by bush encroachment. Since the seed of Acacia mellifera is not long-lived, not mobile and not distributed in the dung of herbivores, there is little need to burn areas that do not contain a large proportion of mature Acacia mellifera bushes.

If this sporadic opportunity to prevent bush encroachment was missed and the savanna changes from a grassy to a bushy state, fire is no longer an effective means of control. Chemical and mechanical control mechanisms are the best option to thin (rather than eradicate) the population of invasive Acacia mellifera, failing which they will mature over many decades to old age. Only when senescent do they again become vulnerable to stressors and natural control. Control presents an ideal opportunity to improve the grass sward and thus recover the production potential of Namibia’s savanna rangelands, but has to be followed by aftercare.

SuMMARY OF ThE ECOLOGICAL MODEL

The savannas found in central Namibia, the Highland, Thornbush and Camelthorn savanna (Giess, 1971), can occur in a grassy or a bush-encroached (bushy, woody) state (Figure 1). Within these two major states, climax and pioneer states exist (Joubert, Rothauge & Smit, 2008). Various forces and events change the state of a savanna (a “transition”). The degraded grassy (pioneer) state can degrade further to full-scale desertification which is, for all practical purposes, irreversible. A crucial stage for the land manager is when a grassy turns to a bushy state. In such a case the grass – bush balance is severely disrupted and the characteristics of the savanna change completely

(Rothauge, 2007a). Bush encroachment is a problem of national significance and is further elucidated in this, the second part of the article.

In the savannas of central Namibia, the major invasive woody species is Acacia mellifera supsp. detinens (Bester, 1998). Over these vast areas, the abundance and density of Acacia mellifera bushes have exploded until far exceeding its probable natural density. Unfortunately, there is no scientific norm of what this plant’s “natural density” should be. De Klerk (2004), based on recommendations by Prof G.N. Smit, theorised that the probable “natural density” of woody plants (expressed in evapo-transpiration tree equivalents, ETTE, per hectare) in Namibian savannas is twice the average annual rainfall received (in millimetre). An ETTE is the amount of water required by a woody plant of 1,5 m height. The number of ETTE/ha gives an indication of how much soil moisture is extracted from the soil by woody plants (Smit, 1989). If an area receives on average 350 mm per year, the probable natural density of all woody plant species is about 700 ETTE/ha, i.e. 700 bushes of 1,5 m height per hectare. However, considering that the statistical variability of rainfall in Namibia exceeds 40 %, bush density in a 350 mm rainfall area should not be uniform, but vary from as little as 420 ETTE/ha to, in patches, 980 ETTE/ha to achieve a diverse mosaic effect. In bush-encroached rangelands, the density of Acacia mellifera alone can often reach 6 000 and sometimes even 16 000 bush per hectare (Bester, 1998). Such thorny thickets form monostands that are impenetrable to large herbivores and starve the herbaceous component of water.

ThE TRANSITIONAL STATE OF BuSh SEEDLING ESTABLIShMENT

Bush seedlings are easily overlooked and therefore, this state is not often recognised in the field. In this transient state, the still-dominant grass sward conceals a high number of Acacia mellifera saplings which are still below grass-emergent height. At superficial inspection, this state is virtually indistinguishable from the grassy states, masking the progression towards a bushy state. A savanna may persist in this state for a variable length of time, depending on the competitiveness of the grass sward, rainfall, soil fertility, herbivory and especially the frequency and severity of fire. Even after six years, Acacia mellifera saplings may be no taller than year-old seedlings and are only distinguishable by their branches and thicker

stems that bear the abscission scars of previous years’ growth (Rothauge, 2007b), so-called gullivers (Bond & Van Wilgen, 1996). As the seedlings mature, they become more easily visible and the savanna apparently changes “quickly” to the bush-encroached state. However, it had been in transition already for some years.

The transient state of bush seedling establishment represents a critical time period for management intervention to prevent the initiation of bush encroachment, because Acacia mellifera seeds and saplings are extremely prone to fire, frost, predation and competition. These events, if utilised properly, could affect the restorative Transitions 4 and 5 back towards the grassy states. In their absence, Transition 6 is initiated and a bushy state will develop.

TRANSITION 3 TOWARDS ThE ESTABLIShMENT OF BuSh SEEDLINGS

The transition from the grassy state towards establishment of Acacia mellifera seedlings needs three consecutive years of above-average rainfall to be initiated and is facilitated by secondary conditions, especially by a grass sward that has been weakened by over- or selective grazing, the complete suppression of fire and reduced browsing pressure. Successful recruitment of Acacia mellifera in the Highland savanna is rare and probably occurred on only five occasions in the past 95 years (Figure 2). This is in

ECOLOGICAL DYNAMICS OF CENTRAL NAMIBIA’S SAVANNAS: PART 2 – BuSh ECOLOGY

AxEl RoTHAuGE

AGRA (Co-operative) Limited, Private Bag 12011, Windhoek, [email protected]

DESERTIFICATION

10: hot fire, active grass re-seeding followed by establishment rest

Climax grassy state

Pioneer grassy state

Vigorous bushy state

SeneScent buShy State

Bush seedling estaBlishment

9: drought, fire without active management

11: a

rtific

ial b

ush

cont

rol,

exce

ssive

gra

zing6: e

xclusio

n of fire

and browsing

5: episodic hot fire, unchanged grazing management

3: rainfall, n

o fire

3: rainfall, no fire

4: episodic hot fire, seed-set rest

1: e

xces

sive

gra

zing

, d

roug

ht

2: a

ctiv

e re

-see

ding

, es

tabl

ishm

ent r

est

7: n

atur

al p

rogr

essi

on

8: ra

infa

ll, c

old

fire

Figure 1. Complete schematic presentation of the five states of the Highland savanna and the eleven transitions between states (red = degradation and green = rehabilitation transitions).

agreement with findings in another semi-arid savanna in eastern Australia, where six widespread Dodonaea attenuata establishment events were estimated to have occurred in 97 years (Harrington, 1991). In the past century, there were only five occasions of three consecutive years of above-average rainfall at Neudamm, viz. 1917 to 1921, in the early fifties, in the mid fifties, in the late seventies and in the late eighties/early nineties (Figure 2, circled in green). During these wet spells, Acacia mellifera could have been established only if fires were totally excluded, the grass sward was weakened and browsing pressure was greatly reduced. These secondary conditions would probably only have been satisfied from the 1970s onward, when many farmers acquired mechanised fire fighting equipment. It is thus postulated that much of the landscape-level Acacia mellifera encroachment of the Highland savanna occurred sporadically in the late seventies and again in the late eighties/early nineties; 30 and 15 years ago, respectively.

Why are three successive, above-average rainfall years required to initiate bush encroachment by Acacia mellifera in the central savannas of Namibia? A first good rainy season is required for this woody species to produce viable seed. The bush flowers and sets seed before the advent of the rains, using carbohydrate energy reserves accumulated in, and carried over from the previous rainy season. Donaldson (1967) noted that Acacia mellifera fruited profusely following a season of “copious” rain in the bush-encroached Molopo area of South Africa, but


that only 2 % of trees fruited in drier years. At Krumhuk, during the period 1998-2004, copious fruiting was only observed in the two seasons following the exceptional 1999/2000 rainy season, whereas no fruiting occurred in the preceding seasons of below-average rainfall (Joubert, 2007), because flower and especially seed production requires greater energy reserves than can be accumulated in years of below-average rainfall. Flowering “privileged” trees growing along road verges, in gardens, near rivers and dams or in other microhabitats where moisture accumulates sometimes create the impression that a far larger proportion of trees are reproducing than is actually the case, but they are not representative of conditions in the wider landscape.

The second good rainy season is required for the successful germination of recently-formed seeds. Ripe seeds drop from the tree by late November to early December, before the advent of the main rains. Only if the dropped seeds receive good early rains and good follow-up rains will they germinate and establish themselves, before they are depleted by predation or unsuccessful establishment. The chance that rains will be both early and regular is better when the rainy season is above, rather than below average (Figure 3) (Rothauge, 2008). With a good moisture regime, seeds germinate easily and achieve more than 90 % germination (Joubert, 2007). Ease of germination rapidly depletes the seed bank and if establishment is unsuccessful, few seeds remain in reserve. Seeds that have not germinated do not survive long because their testa is soft, offering little protection against seed predators or digestion in the rumen of herbivores (Donaldson, 1967; Smit, 2003). The seeds are relatively small and contain less endosperm than those of large-seeded Acacia species, such as Acacia erioloba, reducing the long-term survival of seeds. Thus, Acacia mellifera seed banks are ephemeral and seedlings can only emerge after seed production in the

same year (Donaldson, 1967; Joubert, 2007). At Neudamm, seedling establishment on 143 plots of 3,14 m2 each was only observed in 2001 following the exceptional 1999/2000 rainy season and not in the seven rainy seasons thereafter (Figure 4) (Rothauge, 2007b).

A third consecutive good rainy season is required to enable the transition to bush-encroached rangeland to succeed by ensuring the best possible survival of seedlings that emerged in the previous season. After poor rainy seasons, sapling mortality is high (Figure 4). Sapling survival is greatly improved if fires are prevented, by heavy stocking of grazers that diminishes the competitive ability of perennial grasses (Teague & Smit, 1992) and by reducing browsing pressure. In the total absence of fires, seedlings of Acacia mellifera establish themselves even in prime grass swards (Rothauge, 2007b), suggesting that bush thickening is inevitable in the absence of fire. Concerning grazing pressure, this transition in the adjacent camelthorn savanna followed upon prolonged periods of stocking more than 45 kg cow mass per hectare (equivalent to 10 ha/LSU) (Rothauge, 2006a). With regard to browsing pressure, sapling density at Neudamm was reduced drastically when they were exposed to small ruminants compared to cattle (Rothauge, 2007b). Suppression of Acacia mellifera establishment was even better when they were exposed to indigenous Damara sheep and Boer goats because they browse much more than other breeds like the Dorper or Karakul (Kamupingene, Mukuahima, Rothauge & Abate, 2005). In elevated areas, severe winter frosts may also curtail the establishment of Acacia mellifera seedlings.

The successful recruitment of Acacia mellifera leads to an unstable transitional state that can either proceed to bush thickening (Transition 6) or revert to the grassy state (Transitions 4 and 5) (Figure 1), representing a crucial juncture for vegetation dynamics and rangeland

management. If not interrupted, Transitions 3 and 6 form a continuous progression from grass- to bush-dominated savanna.

1. The role of fire in preventing bush encroachment

The drivers of the restorative transitions 4 and 5, from the transient state of bush seedling establishment towards a grassy state, are fire, frost, seed and seedling predation, and direct competition with the grass sward, in decreasing order of importance.

Fire has long been known to be an important driver of vegetation dynamics in savannas (Trollope, 1984; Teague

& Smit, 1992; Smit, 2003). It may control bush thickening through its effect on mature trees, fruits, seeds (Trollope, 1974; Trollope, 1980; Sweet, 1982; Trollope, 1982; Trollope, 1984; Hodgkinson & Harrington, 1985; Harrington & Driver, 1995; Skowno, Midgely, Bond & Balfour, 1999; Higgins, Bond & Trollope, 2000; Roques, O’Connor & Watkinson, 2001), and seedlings and saplings (Frost & Robertson, 1987; Trollope, Hobson, Danckwerts & Van Niekerk, 1989). Fire restricts the recruitment of woody seedlings into adult size classes (Higgins et al., 2000). In the Highland savanna, a fire is far more effective at the seedling and sapling stage of Acacia mellifera because they are more fire-prone than more mature growth stages. An effective fire will take the savanna from the transient state

Figure 2. Annual rainfall (mm) received at Neudamm since the 1913/14 season (Rothauge, 2008). Green ellipses indicate favourable rainfall periods with three or more consecutive years of rainfall above the long-term average.

Figure 3. Intra-seasonal rainfall distribution (lines) and total rainfall (bars) at Neudamm from 1913 to 2008, distinguishing between top (wettest) and bottom (driest) quartile (Rothauge, 2008).

Figure 4. Emergence, establishment and survival of Acacia mellifera seedlings (line) at Neudamm in response to rainfall (bars) (Rothauge, 2007b).

1 000

900

800

700

600

500

400

300

200

100

0

Rai

nfal

l (m

m/y

ear)

70

60

50

40

30

20

10

0

Intra

-sea

sona

l rai

nfal

l dis

tribu

tion

(mm

/10

day-

perio

d)

4,0

3,5

3,0

2,5

2,0

1,5

1,0

0,5

0

Den

sity

of A

caci

a m

ellif

era

seed

lings

/m2

700

600

500

400

300

200

100

0

Tota

l rai

nfal

l (m

m/y

ear)

S1

S3

O2

N1

N3

D2 J1 J3 F2 M1

M3

A2

M1

M3 J2 J1 J3 A2

10 day-periods from September to August 1913–2008

Wettest 25 % of seasons

Average 50 % of seasons

Driest 25 % of seasons

800

700

600

500

400

300

200

100

0

Rai

nfal

l (m

m/y

ear)

Seasonal rainfall

Seedlings

Long-term average rainfall

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

200

7


of bush seedling establishment back to the grassy states through Transitions 4 and 5. Trollope’s (1982; 1984) general recommendation of two tons of fuel per hectare to control mature trees is a prudent measure of what would constitute an effective fire, but it is reasonable to expect that a lower fuel load might suffice to control seedlings and saplings.

Fires occur much less frequently today than previously, since farmers are reluctant to accept the risks of burning. Fire is a known danger to ranching because it destroys fodder reserves accumulated as standing hay on the veld and because of its short-term destructive effects on organic life forms in the top layer of soil (Trollope, 1982; 1984). The economic returns after burning may only be realised many years later (Hodgkinson & Harrington, 1985; Harrington & Driver, 1995). However, the lack of fierce fires is one of the main reasons for Transition 6 towards a bushy state. In the absence of fierce fires, bush encroachment is inevitable in Namibia’s savanna rangelands. Like in other semi-arid savannas of the world, bush recruitment is a continuous process, independent of herbaceous biomass and density but regulated mainly by fire (Brown & Archer, 1999). The best stage to apply such a fire would be while the savanna is still in the transient state of bush seedling establishment, stifling the initiation of bush thickening.

Fire as a management tool to prevent the transition of a savanna towards the bushy state should emulate nature. Fire should only be used sporadically every 15 to 20 years (Bond, 1997) once environmental conditions (wet spells of at least three consecutive years of above-average rainfall) favour Acacia mellifera seedling establishment. Successful establishment of Acacia mellifera seedlings can easily be confirmed by monitoring sites close to adult bushes. Quantitative confirmation of impending recruitment should trigger an appropriate response, namely a hot fire that kills most of the bush seed and saplings. Conditions for a hot fire, primarily the accumulation of more than two ton of herbaceous dry matter per hectare to act as fuel, are very likely to occur during the same wet spell that favours bush seedling establishment. Wet spells usually result in more fodder being produced than can be utilised by domestic livestock and thus afford the land user the luxury to burn some off in the interest of manipulating the grass – bush balance of the savanna. If used with this intent, a controlled fire can maintain a savanna in the grassy state and prevent sporadic bush encroachment. It may also initiate the successful establishment of woody climax species such as Acacia erioloba, whose seeds need scarring by amongst others, fire to germinate.

Whenever fire is used to prevent Acacia mellifera colonisation, it must be managed correctly to maximise its advantages and minimise its danger (Trollope, 1999):

• Infested rangeland should only be burned if there isenough combustible herbaceous material to support a really hot fire that is sure to kill most bush saplings (more than two ton of herbaceous dry matter per hectare).

• The veld should be burned in late winter when thebushes break their dormancy and are more sensitive to fire than when they are still dormant (too early) or have already completed sprouting (too late), but before the first rains to ensure that the grass fuel is still dry.

• At that stage,perennialgrassesmayhave started theearly-season green flush, typical of wet spells, enabling them to survive hot fires better.

• The veld shouldbeburnedearly in thedaywhen thebase of grass tufts is still wet from dew and when wind speed is usually low.

• Backfiresburningintothewind,areslower,easiertocontrol and release most of their heat at ground level where they do severe damage to grass tufts. Head fires burning with the wind, are hotter and most of their heat is carried upwards. They ignite wood fuel more effectively, but are more difficult to control because they travel fast. Trollope (1999) recommends the use of head fires to control bush.

• Theveldshouldonlybeburnedifithasbeenprotectedby the necessary precautions, such as fire breaks and if burning activities adhere to the fire management guidelines spelt out in the Forest Act of 2001.

• The veld should not be burned when conditions areunsuitable, e.g. insufficient fuel load, high wind speed, insufficient manpower on standby, etc.

• Burned veld should be rested from grazing until theperennial grasses have recovered their vigour by setting seed. With an active green flush due to inter-seasonal soil moisture transfer during a wet spell, this may happen within 60 days of a late dry season fire (Rothauge, 2006b).

Well-timed fierce fires maintain savannas at a high level of productivity and biodiversity whereas most ill-timed “cool” fires do more damage than good to rangeland.

2. Other drivers that prevent bush encroachment

The second important driver that can prevent a transition towards bush encroachment is frost. At higher altitudes in the Highland savanna, successive nights with sub-zero temperatures occur regularly. Severe frost after en masse seedling recruitment may kill bush seedlings, even those protected by a dense grass sward and revert the vegetation to a grassy state without seedlings.

The third-most important driver preventing bush en-croachment is predation of the seeds, seedlings and sap-lings of Acacia mellifera. Fungi, invertebrates (especially arthropods), birds and small mammals (especially rodents) are major pre-dispersal predators of Acacia seeds (Fagg & Stewart, 1994) and may facilitate the transition to the grassy state. Many studies have reported on the effect of bruchid weevils on the dynamics of various Acacia spe-cies (e.g. Hoffman, Cowling, Douie & Pierce, 1989; Miller, 1994a; Vir, 1996; Okello & Young, 2000). Acacia seeds are typically attacked by seed-eating Bruchidae while still at-tached to the adult tree (Abdullah & Abulfatih, 1995) and infested seeds either do not germinate (Vir, 1996) or their

viability is drastically reduced (Hoffman et al., 1989; Miller, 1994a; Okello & Young, 2000). The effect of seed predation is larger in dry than in wet years. At Krumhuk, following good rains in 2001, an average of 200 seeds were counted per Acacia mellifera tree. Seventy five percent of these were not viable, having mostly been infected by fungi. A further 12,5 % of seeds were infested with bruchid weevil larvae, leaving 25 viable seeds per tree to germinate. In contrast, following the poor rains in 2002, only four seeds were counted per tree of which three (75 %) were infested with bruchid weevil larvae, leaving only one seed per tree to germinate (Joubert, 2007). Small mammals, birds and invertebrates are important post-dispersal seed predators (Kerley, 1991; Miller, 1994b; Ostfeld, Manson & Canham, 1997; Weltzin, Archer & Heitschmidt, 1997; Linzey & Washok, 2000) while rodents and ants are noted for their post-dispersal predation of Acacia seeds (Walters, Milton, Somers & Midgley, 2005). Mice typically remove only the endosperm and leave an empty testa. Since Acacia mellifera seeds germinate with ease, seed banks are depleted on a seasonal basis by predation (Meik, Jeo, Mendelson & Jenks, 2002). In years of exceptional fruit production, seed banks might be too large for seed predators to reduce substantially and the transition towards a bushy state will continue without interruption.

Seed of dehiscent Acacia species such as Acacia mellifera is typically wind dispersed (Miller, 1994a; Okello & Young, 2000) and is routinely destroyed in the gut of animals. In laboratory conditions, Donaldson (1967) found that only 2 % of Acacia mellifera seeds ingested by cattle germinated and it is highly unlikely that seeds of Acacia mellifera are dispersed by ungulates. In contrast, the seeds of indehiscent Acacia species require some form of gastric or ruminant treatment, even scarring by fire to enhance germination (Miller, 1994a; Okello & Young, 2000). Animals that consume Acacia pods include birds, ungulates, rodents, termites and ants (Miller, 1994b; Barnes, 2001).

Small mammals, especially lagomorphs (hares and rabbits) are important seedling and sapling predators in savanna and other ecosystems (Ostfeld et al., 1997; Weltzin et al., 1997), thus regulating vegetation structure. Lagomorphs crop saplings, leaving the branch cut off cleanly at a characteristic diagonal angle. Up to 58 % of the damage to Acacia mellifera saplings at Neudamm was caused by lagomorphs and 3 % were infested with aphids (Rothauge, 2007b). Baboons and warthogs dislodge saplings while digging; similar to the effect of prairie dogs on Prosopis saplings in north-central Texas (Weltzin et al., 1997). At Neudamm, dense stands of Acacia mellifera saplings underneath and close to existing thickets are routinely destroyed by helmeted guinea fowl that scrape under the trees in search of seeds and grubs (Rothauge, 2007b). In the wild, mega-browsers such as elephant contain the establishment of Acacia saplings by uprooting them for feeding (Eltringham, 1979; Skinner & Smithers, 1990).

The least-important driver of Transitions 4 and 5 towards the grassy state is competition by the grass sward. A dense and vigorous grass sward may out-compete woody seedlings (Walter, 1971; Walker, 1981; Smit & Rethman, 1992) in accordance with Walter’s (1971) two-layer hypothesis and might be expected to reduce fruiting success as well. The survival of woody seedlings and saplings is determined by the amount of excess moisture available after the fibrous roots of perennial grasses have removed their share (Davis, Wrage & Reich, 1998). Therefore, Transition 5 to the pioneer grassy state is less likely to occur than Transition 4 to the climax grassy state as annual grasses have much smaller root systems than the climax perennials (Wolfson & Tainton, 1999). However, Kraaij & Ward (2006) show that the competitive effects of the grass sward of a semi-arid savanna may be negligible; rather, grass has the more important role of providing fuel for a fire.

TRANSITION 6 FROM BuSh SEEDLING ESTABLIShMENT TO A VIGOROuS BuShY STATE

This transition is a continuation of Transition 3, if not interrupted. It has a better chance of succeeding in semi-arid than in mesic savannas, as semi-arid savannas rarely have sufficient fuel to produce a fire that might control Acacia mellifera saplings and they might escape the fire-prone stage more easily. If the transition occurred from the pioneer grassy state, there would be even less grass cover and hence, insufficient fuel for a fire. Observations in semi-arid Botswana (Skarpe, 1990) suggest that the absence of grass allows Acacia mellifera gullivers to grow rapidly beyond the fire-prone stage. A poor grass cover thus facilitates this transition. Thus, the crucial stage of onset of bush encroachment is seed production and seedling survival, rather than the release of already existing gullivers from competition by grasses through lack of fire at a later stage, when the transition towards a bush-thickened state is a fait accompli.

Ungulate browsers effectively maintain woody species at a much lower height than if they had not been browsed (Belsky, 1984) and thus control seedlings, saplings and gullivers (O’Connor, 1996), but cannot eradicate mature bush. Boer goats utilise Acacia mellifera at all growth stages, including fruits, seedlings and adults, but it is not their preferred forage species (Rothauge & Engelbrecht, 2000; Rothauge, Abate, Kavendji & Nghikembua, 2003). Thus, there is concern that the browsing pressure which would be required to control Acacia mellifera infestations will have adverse effects on other more desirable fodder bushes and the herbaceous layer (Zimmermann & Mwazi, 2002). Goat pressure in conjunction with controlled burning was able to control bush thickening by Acacia karoo in South Africa (Trollope, 1980), but had little effect on Dichrostachys cinerea (Zimmermann & Mwazi, 2002), another encroaching species. Ungulate browsers by themselves are probably ineffective in preventing a transition towards bush thickening. Rather, they regulate vegetation structure (Teague & Smit, 1992) and should be used in aftercare following other bush control measures (De Klerk, 2004).


ThE VIGOROuS BuShY STATE

After the gullivers of Acacia mellifera have emerged above grass height, the shrubs exert such an influence on the environment that a transition back to a grassy state becomes virtually impossible without some form of active management intervention. This is largely due to the species’ competitive ability to successfully extract water from the soil (Smit, 2003; De Klerk, 2004). In this state, Acacia mellifera remains dominant irrespective of grazing pressure or rainfall and is highly unlikely to be removed by fire. Unfortunately for Namibian ranchers that engage in extensive animal production with mainly grazing livestock, the vigorous bushy state is a very stable state that requires active intervention to change.

The increase in bush density and cover causes a substantial decline in grass production, mainly through competition for soil moisture (Van Vegten, 1983; Skarpe, 1990; Smit, 2003; De Klerk, 2004). A dense stand of Acacia mellifera dries out the soil, lowers the groundwater table and severely restricts the amount of water for use by other rangeland plants. Acacias are C3-plants that use comparatively more water during photosynthesis and transpire more water in hot climates than C4-plants, like tropical grasses. De Klerk (2004) reports that a 2,5 m high Acacia mellifera plant with a canopy spread of 6 m2 uses up to 64,8 litres of water during an eight hour day in the growing season, equivalent to 38,9 litres of water per day used by an ETTE. A thicket of 5 000 ETTE/ha would thus transpire the equivalent of 19,5 mm of rainfall per day for its own purposes. This would deplete the season’s total rainfall in less than three weeks and allow virtually no rainwater to remain for use by other plants. The amount of water transpired by Acacia mellifera is four times higher than that transpired by macrophyllous savanna trees such as Terminalia sericea or Boscia albitrunca. One Acacia mellifera plant transpires

as much water as 188 tufts of Anthephora pubescens, 559 tufts of Schmidtia pappophoroides or 864 tufts of Eragrostis lehmanniana (calculated from De Klerk, 2004). Similarly, a Prosopis-infested mesic rangeland in the south-western USA transpired three-and-a-half times more water than adjoining, dense grassland (Qi et al., 1998). After successful establishment, Acacia mellifera has its water-absorbing infrastructure firmly in place. Its shallow lateral roots can extend at least seven times further horizontally than the height of the bush (Figure 5). Its superior water absorption capacity enables its roots to keep on extracting water from soil so dry that grasses are already wilting (Hipondoka, Aranibar, Chirara, Lihavha & Macko, 2003; Smit, 2003). Little wonder that natural springs and fountains all over Namibia’s bush-encroached rangelands dry up spontaneously, irrespective of trends in precipitation (Bockmühl, 2009; Christian, De Klerk, Bockmühl, Van der Merwe & Mostert, 2010).

Initially, the bushy state consists of a fairly homogenous stand of similar-sized shrubs, mostly below one metre in height, with little grass cover (Joubert, 2007). The homogeneity of the stands attests to the episodic nature of the recruitment event, rather than it being a continuous, gradual process. It reflects the long periods of “sleep” with few vegetation dynamic events, followed sporadically by a “leap” mode of sudden and considerable change. A few larger parent trees of around 3 m to 4 m high, from which the shrubs originated, may be scattered through the thicket. These parent trees probably reflect the original “natural density” of Acacia mellifera in the savanna. Shrub densities around parent trees can reach three shrubs per square metre and almost 100 % canopy cover, but are usually lower than this (Joubert, 2007). On a landscape level, shrub densities of 4 000 to 8 000 shrubs per hectare are common in bush-encroached parts of the Highland savanna and

12 000 shrubs per hectare are not unusual (Bester, 1998; De Klerk, 2004).

Shrubs limit their own growth rate by intense density-dependent inter-shrub competition (Smit, 2003) and may remain immature (not reproductive) for decades until some event initiates self-thinning; enabling the survivors to grow out to maturity. Grasses are out-competed for light, soil nutrients and water under the canopy. The soil underneath bush thickets is often bare, but erosion is limited due to the dense, shallow network of Acacia mellifera roots and the high (woody) canopy cover. Increaser grasses, predominantly annuals of the genera Aristida, Eragrostis, Enneapogon and Tragus, dominate the grass layer. Their yield is low and erratic, depending on the rainfall. Animal biodiversity in these thickets is considerably lower than in the grassy state (Barnard, 1998), even though some animal species utilise thickets as refugia. Livestock production on vigorously bush-encroached range is severely limited by the lack and inaccessibility of palatable grasses. Wood harvesting for electricity generation (Von Oertzen, 2007), firewood and especially charcoal production is increasingly a strategy used by rangeland managers to generate income and return the vegetation to a grassy state. Charcoal production becomes an option as trees mature and stem diameter increases beyond 20 mm (Bester & Reed, 1997).

Transitions to the vigorous bushy state occur close to parent trees and already existing thickets, because seed dispersal is inefficient and seeds do not travel far from the parent plant (Donaldson, 1967; Joubert, 2007). Of 143 plots of 3,14 m2 each monitored at Neudamm, the 10 % of plots most heavily infested with Acacia mellifera seedlings were located only 3,2 ± 3,4 m from the closest presumed parent tree, whereas the 10 % least infected plots were 18,7 ± 18,04 m removed from the closest presumed parent tree (Rothauge, 2007b). Shrub growth is encouraged by good rainfall. Frost and fires only affect the edges of a thicket. Unmanaged “cold” fires may even hasten shrub growth by weakening competing grasses.

TRANSITION 7 TOWARDS A SENESCENT BuShY STATE

Transition 7, towards a senescent thicket of Acacia mellifera, is a progressive succession that takes decades. It is a gradual and inevitable transition, since fierce fires that control the bush are highly unlikely due to the sparse grass sward and the density of the thicket. Self-thinning continues within the thicket due to intra-specific competition for, primarily, soil moisture (Skarpe, 1990; 1991; Teague & Smit, 1992), but canopy cover remains much the same (Joubert, 2007). As the tree canopy gets taller, allowing light to filter into the sub-canopy habitat, broad-leaved shrubs such as Boscia albitrunca, Rhus marlothii, Tarchonanthus camphoratus and Ziziphus mucronata that normally form a component of the grassy state, germinate and grow in the protection of the tall thicket (Joubert, 2007), which acts as a “nursery” to them. Birds that are attracted to the thicket transport the seeds of these fleshy-fruited species to this location.

ThE SENESCENT BuShY STATE

This state is characterised by mature and senescing trees of around 4 m high, often with an understory of immature broad-leaved shrubs and trees that was established during Transition 7. It is not known at what age senescence sets in. The density of trees has typically been reduced to about 2 500 trees per hectare although canopy cover tends to remain high, even 100 % (Joubert, 2007) but slowly declines as trees senesce due to drought stress, old age, fungal pathogens (Holz & Schreuder, 1989a) or combinations of these. Fungi of the Cytospora and Phoma genera, especially Phoma glomerata, attack the red heartwood and sapwood at the base of the trunk and upper taproot of mature Acacia mellifera trees, progressively weakening the tree until an external stressor, e.g. competition from other trees in the thicket or a drought, eventually kills the infected tree (Holz & Schreuder, 1989a, b; Holz & Bester, 2007). The fungi appear to be more active in wet spells than during dry periods. Tens of thousands of hectares of bush-encroached savanna have been cleared in this manner in Namibia.

Gradually, as the canopy cover shrinks and dead trees and branches fall and decompose, conditions favourable for herbaceous plants begin to develop. Initially, broad-leaved forbs, especially of the Amaranthus family, grow in the filtered, nutrient-rich sub-canopy habitat, eventually followed by shade-tolerant grass species commonly associated with savanna trees and the nutrient-rich soil below their canopy, e.g. Cenchrus ciliaris, Eragrostis scopelophila and Eragrostis lehmanniana (Rothauge, 2007a). Kraaij & Ward (2006) show that the elevated nitrogen levels found under Acacia trees in savannas (Smit & Swart, 1994; Rothauge, Smit & Abate, 2003) give grasses a competitive edge over woody seedlings and forbs. Acacia mellifera is a deciduous legume, enriching the soil with rhizomataeous nitrogen, dropped leaves and imported nutrients derived from bird droppings and animals resting in the shade of the canopy. The sites of fallen, decaying tree skeletons, rotting gradually over a period of 5 to 20 years (Milton & Dean, 1996), are often dominated by Cenchrus ciliaris (Rothauge, 2007a), a nitrogen-loving sub-climax grass. Grass cover and the grass-based carrying capacity is variable, but better than in the vigorous bushy state and possibly even in the pioneer grassy state. Bush senescence is thus a rare opportunity to recover the original production potential of Namibia’s central savanna ecosystems.

The build-up of grasses, herbs and forbs under the open thicket canopy, especially on the edges, provides fuel for fires that may be fierce enough to kill the senescing trees. Mature or senescing Acacia mellifera are less able to resprout than young shrubs (Meyer, Ward, Moustakas & Wiegand, 2005). Thus fires are much more effective at removing trees in this state than in the vigorous bushy state. This presents another window of opportunity to land managers to force a transition back to the grassy state (Transitions 9 and 10 to pioneer and climax grassy states, respectively) failing which, this state may cyclically remain in a bushy state (Transition 8 back to bushy state).

Figure 5. The extensive lateral root system of Acacia mellifera (picture courtesy of Prof. G.N. Smit).


TRANSITION 8 BACk TO ThE VIGOROuS BuShY STATE FROM ThE SENESCENT BuShY STATE

After two to three successive high rainfall years, new individuals from the ephemeral seed bank replace the senescent trees and recruitment occurs from the existing thicket of Acacia mellifera trees, similar to Transition 3. Establishment is close to, but not directly under the senescing parent trees, as seeds are not distributed far (Joubert, 2007). It is facilitated by reduced browsing pressure that could check coppice re-growth and seedlings, overgrazing that removes the accumulating grass fuel load and injudicious burning with “cold” fires that damage the grass sward, but not mature trees. Transition 8 may occur at any time between State 4 and 5, provided that the rainfall favours fruiting and seedling establishment and gaps between canopies are present, and is equivalent to the combined Transitions 3 and 6. Transition 8 changes thicket structure from a homogenous thicket of even-aged and even-sized shrubs to a more diversified, patchwork thicket consisting of different cohorts of shrubs of different age and size.

TRANSITION 9 AND 10 BACk TO ThE GRASSY STATE FROM ThE SENESCENT BuShY STATE

A senescent bush thicket may revert back to the pioneer grassy state by Transition 9 or to the climax grassy state by Transition 10. Transition 9 to the pioneer grassy state is likely if below-average rainfall conditions (drought) and excessive grazing pressure prevail, or if a fire clears the senescent bush, but is not followed by active management interventions that facilitate the establishment of perennial sub-climax and climax grass species (e.g. climax grass re-seeding). Ranchers that experience dense stands of Acacia mellifera dying, but do not change the grazing management that assisted bush thickening in the first place, risk this transition, and increase the probability of a return to a bushy state in the following high rainfall period. Bush clearing may also initiate this unfavourable transition if the underlying cause of poor grazing management is not addressed, if no wooden litter is left on the ground as mulch and if there is too little browsing pressure to check re-growth of harvested trees and bushes.

Transition 10 to the climax grassy state may require a hot fire, but certainly requires artificial re-seeding if perennial, climax grass seeds are not naturally available in adjacent areas. The palatable perennial grasses preferred by domestic livestock tend to become locally extinct under intense grazing pressure and subsequent bush encroachment. Grass re-seeding has to be followed by protection from grazing to allow the re-introduced grasses to establish themselves successfully. After re-seeding, a dense and very productive grass sward develops rapidly, benefiting from soil enrichment by leguminous Acacia mellifera plants, the competitive release from woody plants and protection from grazing by woody mulch (Smit & Rethman, 1992). This release effect may last for several years (Smit, 2003).

The precondition for Transition 10 is lenient grazing pressure that allows the desired perennial grass species targeted rest periods to recover from grazing, set seed and establish successfully (Smit, 2003; Rothauge, 2007c). Browsing pressure may be necessary to control re-sprouting trees and bushes. If this transition occurs during a two to three year above-average rainfall spell, there is a danger of reverting to the transitional state with bush seedlings. In this case, a fierce fire will be required to kill as many of the senescing trees, coppicing bushes and emerging woody seedlings as possible. In Namibia, Transitions 9 and 10 are currently being forced by wood harvesting for firewood or charcoal production without changing the underlying conditions that caused bush encroachment in the first place.

TRANSITION 11 BACk TO ThE PIONEER GRASSY STATE FROM ThE VIGOROuS BuShY STATE

Modern technology enables Namibian ranchers to de-bush encroached rangeland using mechanical or chemical means or a combination thereof. De Klerk (2004) describes these control measures in great detail. They result in an open savanna as the invasive species, Acacia mellifera, is in many instances eradicated completely. This creates a new imbalance as the woody component of the savanna is now under-represented, opening a window of opportunity for other woody opportunists to exploit and fill the void rapidly. Many ranches that were cleared of Acacia mellifera are now invaded by a variety of non-thorny, broad-leaved shrubs such as Laggera decurrens in weak grass swards, Grewia flava on sandy aeolian soil, Catophractes alexandrii on limy sub-soil, Tarchonanthus camphoratus on shallow mountain soil and Phaeoptilum spinosum on soils that receive extra moisture (e.g. low-lying areas) or become saturated easily (brackish soils). Smit (2003) therefore recommends that not all Acacia mellifera bushes should be eradicated but that the largest trees are left standing to control re-colonisation by the same, or a different woody species. Besides competitive control, the large individuals contribute to a more diversified habitat, appreciated by game animals and with increased plant and animal biodiversity. Of course, the large survivors are also a potential source of seeds and renewed infestation. However, chemical and mechanical control mechanisms present the land user with an opportunity to restore degraded savanna and recover its potential production capacity, if those factors that caused bush encroachment in the first place, are also addressed. Otherwise, the encroachment cycle will merely repeat itself after a short delay.

Would Transition 11 occur naturally? The main possible drivers of such a transition would be fire and mega-herbivores. In nature, fire is highly unlikely to control an established, vigorous bush thicket. Even a fierce fire will not penetrate a thicket for lack of grass fuel inside the thicket. At best, a fierce fire may kill trees on the boundary, thus reducing the size of the thicket. Edges of thicket and less dense stands will be opened up, creating a mosaic effect and presenting a small window of opportunity for land users.

Mega-herbivores such as elephant and black rhino may have contributed substantially to the fragmentation of bush thickets in the past due to their destructive feeding and resting habits, thereby facilitating Transitions 9, 10 and 11 from the bushy back to the grassy state. These mega-herbivores used to occur in large numbers in the savannas of Namibia in general (Skinner & Smithers, 1990) and in the Highland savanna in particular, according to the writings of early big game-hunters (e.g. Anderson, 1856) and prominent historians (e.g. Vedder, 1934). Elephants are migratory, congregate in big herds and are estimated to have contributed up to 60 % of the animal biomass on pristine African savannas (Du Toit, 2005), thus exerting huge pressure on savanna vegetation. They are mixed feeders and feed very destructively on woody vegetation (Eltringham, 1979; Skinner & Smithers, 1990). A mature elephant requires about 200 kg of feed per day, at least half of which is derived from the woody component of a savanna. It destroys about as much woody vegetation again as it consumes and may affect about two tons of woody vegetation per week. Elephants also uproot Acacia saplings to feed on them. Although elephants do not prefer Acacia mellifera to the same extent as other invasive woody species such as mopane (Colophospermum mopane), they opened up and fragmented thickets, increased their edges and exposed them to the effect of wildfires, thus contributed to limiting the extent of bush encroachment. Similarly, the habit of black rhino to lie up in shady places and to have middens on the edge of thickets would have opened up such thickets and exposed them to the effect of wildfires, although probably on a smaller scale than elephants. However, re-introducing elephant into bush-encroached areas will result in the destruction of valuable, tall fodder trees in the short term, long before the positive effects on bush thickets permeates the landscape in the long term. They probably played a bigger role in preventing bush encroachment in the first place, rather than restoring an encroached landscape.

IN SuMMARY

Bush encroachment by Acacia mellifera probably follows a wet spell of three successive, above-average rainfall years which allows the parent plant to accumulate enough energy reserves to flower and set ripe seeds. Once these are shed, they require regular rainfall to establish successfully. These favourable circumstances occur only sporadically in Namibia. At the same time, seedlings can only establish if fierce fires are prevented, if the grass sward is weakened by incorrect grazing and browsing pressure is low. Once bush gullivers have outgrown the fire-prone stage, they grow and mature inexorably towards a bush-encroached landscape that does not clear itself, unless the bush has become senescent. Prior to old age, active mechanical or chemical intervention to clear or thin the thickets of invasive Acacia mellifera is required. Bush control has to be followed by active aftercare that prevents a return to bush thickening at the first opportunity and facilitates the establishment of desirable perennial grasses. A bush-encroached savanna is a very stable state and is highly unlikely to return to a grassy state on its own within the productive lifetime of a Namibian

farmer. Fire and increased browsing pressure are valuable tools in caring for a landscape in which Acacia mellifera was controlled. Desirable perennial grasses become locally extinct in bush-encroached landscapes and have to be re-introduced artificially, whereafter they need enough time to establish themselves before grazing commences. Once the grass sward has recovered, it needs to be tended and renewed transition towards woody seedling establishment needs to be checked, if a recurrence of the whole bush encroachment cycle and problem is to be avoided.

Custodians of Namibia’s savanna rangelands cannot allow these to swing from one ecological extreme (allowing bush encroachment to happen – to another – wiping out every individual of the invasive species) without being at least aware of the drastic changes this will force in their means of production, the land and its fragile cover of savanna vegetation. The sporadic nature of bush encroachment and the longevity of the invasive species make it difficult to predict its life cycle and reaction to transforming events. Even if Acacia mellifera can be managed successfully, the next major invasive species, Dichrostachys cinerea (sickle bush, “sekelbos”), promises to be an even more difficult species to manage. Dichrostachys cinerea is a major problem in low-frost areas of north-central Namibia. It has hard seeds with an impermeable testa (Bell & Van Staden, 1993) and thus survives ingestion by ungulates. Its persistent seed bank and ability to form root suckers enable it to invade landscapes continuously, rather than sporadically, presenting an even more formidable challenge to Namibian land users. In addition, the effects of global climate change are expected to enhance woody growth; making a thorny situation even more difficult.

ACkNOWLEDGEMENTS

I gratefully acknowledge my many colleagues in Namibia, South Africa and elsewhere who have stimulated and challenged my thinking on savanna dynamics and encouraged me to publish this series of papers as a waypoint in understanding our savanna environment. I am especially indebted to Dave Joubert of the Polytechnic of Namibia for his invaluable contributions to this article. Dave, also the senior author of the international baseline article on this topic (Joubert et al., 2008) is currently busy with PhD research into the bush-encroachment model. I also thank the numerous students of Neudamm College and the Faculty of Agriculture & Natural Resources who participated in field experiments and monitoring.

REFERENCES

ABDULLAH, M.A.R. & ABULFATIH, H.A., 1995. Predation of Acacia seeds by bruchid beetles and its relation to altitudinal gradient in south-western Saudi Arabia. Journal of Arid Environments 29: 99–105.

ANDERSON, C.J., 1856. lake Ngami; or, Explorations and Dis-coveries during four years wanderings in the wilds of South Western Africa. Hurst & Blacket Publishers, London.

BARNARD, P. (ed.), 1998. Biological Diversity in Namibia: A Country Study. Namibian National Biodiversity Task Force, Directorate of Environmental Affairs, Windhoek, Namibia.


BARNES, M.E., 2001. Seed predation, germination and seedling establishment of Acacia erioloba in northern Botswana. Journal of Arid Environments 49: 541–554.

BELL, W.E. & VAN STADEN, J.,1993. Seed structure and germination of Dichrostachys cinerea. South African Journal of Botany 59: 9–13.

BELSKY, A.J., 1984. Role of small browsing mammals in preventing woodland regeneration in the Serengeti National Park, Tanzania. African Journal of Ecology 22: 271–279.

BESTER, F.V., 1998. Major problem: bush species and densities in Namibia. Agricola 10: 1–3.

BESTER, F.V. & REED, E., 1997. Charcoal production in the Karstveld: facts to ponder upon. Spotlight on Agriculture pamphlet series no. 2, Ministry Agriculture, Water & Rural Development, Windhoek, Namibia.

BOCKMŰHL, F., 2009. The Platveld Aquifer study. Presentation at 13th Namibian Rangelands Forum, 27 to 29 October 2009, Neudamm Agricultural College, Windhoek, Namibia.

BOND, W.J., 1997. Fire. In: R.M. COWLING, D.M. RICHARDSON & S.M. PIERCE (eds.) Vegetation of Southern Africa ch. 18, Cambridge University Press, Cambridge, United Kingdom.

BOND, W.J. & VAN WILGEN, B.W., 1996. Fire and Plants. Chapman & Hall, London.

BROWN, J.R. & ARCHER, S., 1999. Shrub invasion of grassland: recruitment is continuous and not regulated by herbaceous biomass or density. Ecology 80: 2 385–2 396.

CHRISTIAN, C., DE KLERK, J.N., BOCKMÜHL, F., VAN DER MERWE, B. & MOSTERT, A., 2010. Desktop Study on the Effect of Bush Encroachment on Groundwater Resources in Namibia. Report for the Namibia Agricultural Union, Windhoek, Namibia.

DAVIS, M.A., WRAGE, K.J. & REICH, P.B., 1998. Competition between tree seedlings and herbaceous vegetation: support for a theory of resource supply and demand. Journal of Ecology 86: 652–661.

DE KLERK, J.N., 2004. Bush Encroachment in Namibia. Report on Phase 1 of the Bush Encroachment Research, Monitoring and Management Project, Ministry of Environment and Tourism, Windhoek, Namibia.

DONALDSON, C.H., 1967. Bush encroachment with special reference to the Blackthorn problem of the Molopo area. Department of Agricultural Technical Services, Armoedsvlakte Research Station, Vryburg, South Africa.

DU TOIT, J.P., 2005. Working within constraints: managing African savannas for animal production and biodiversity. Pastoral Systems in Marginal Environments, Proc. Satellite Workshop 20th Int. Grassl. Congress pp. 81–89, 3 to 6 July 2005, Glasgow, Scotland.

ELTRINGHAM, S.K., 1979. The Ecology and Conservation of large African Mammals. Macmillan Press, London, UK.

FAGG, C.W. & STEWART, J.L., 1994. The value of Acacia and Prosopis in arid and semi-arid environments. Journal of Arid Environments 27: 3–25.

FROST, P.G.H. & ROBERTSON, F., 1987. The ecological effects of fire in savannas. In: B.H. WALKER (ed.) Determinants of Tropical Savannas pp. 93–140. ICSU Press, Miami, USA.

GIESS, W., 1971. A preliminary vegetation map of South West Africa. Dinteria 4: 31–45.

HARRINGTON, G.N., 1991. Effects of soil moisture on shrub seedling survival in a semi-arid grassland. Ecology 72: 1 138– 1 149.

HARRINGTON, G.N. & DRIVER, M.A., 1995. The effects of fire and ants on the seed-bank of a shrub in a semi-arid grassland. Australian Journal of Ecology 20: 538–547.

HIGGINS, S.I., BOND, W.J. & TROLLOPE, W.S.W., 2000. Fire, resprouting and variability: a recipe for grass-tree coexistence in savanna. Journal of Ecology 88: 213–229.

HIPONDOKA, M.H.T., ARANIBAR, J.N., CHIRARA, C., LIHAVHA, M. & MACKO, RSA, 2003. Vertical distribution of grass and tree roots in arid ecosystems of southern Africa: niche differentiation or competition? Journal of Arid Environments 54: 319–325.

HODGKINSON, K.C. & HARRINGTON, G.N., 1985. The case for prescribed burning to control shrubs in eastern semi-arid woodlands. Australian Rangeland Journal 7: 64–74.

HOFFMAN, M.T., COWLING, R.M., DOUIE, C. & PIERCE, S.M., 1989. Seed predation and germination of Acacia erioloba in the Kuiseb River Valley, Namib Desert. South African Journal of Botany 55: 103–106.

HOLZ, G., BESTER, F.V., 2007. A dieback/decline disease of blackthorn (Acacia mellifera subsp. detinens): Monitoring the growth of plants in stands from 1986 to 2005. Agricola Special Edition, 8–58.

HOLZ, G. & SCHREUDER, W., 1989a. Dieback of blackthorn (Acacia mellifera subsp. detinens) in South West Africa. Agricola 7: 32–36.

HOLZ, G. & SCHREUDER, W., 1989b. Soil moisture stress and infection of blackthorn (Acacia mellifera subsp. detinens) by fungi associated with blackthorn dieback. Agricola 7: 48–54.

JOUBERT, D.F., 2007. The population dynamics of Acacia mellifera in relation to climate and microsite factors. Unpublished report, Polytechnic of Namibia, Windhoek, Namibia.

JOUBERT, D.F., ROTHAUGE, A. & SMIT, G.N., 2008. A conceptual model of vegetation dynamics in the semiarid Highland savanna of Namibia, with particular reference to bush thickening by Acacia mellifera. Journal of Arid Environments 72(12): 2 201– 2 210.

KAMUPINGENE, G.T., MUKUAHIMA, G., ROTHAUGE, A. & ABATE, A.L., 2005. Comparative diet selection of the Damara, Dorper and Karakul sheep breeds at Neudamm Farm. Proc. 9th

Nam. Rangeland Forum p. L9, 16 to 18 August 2005, Sandveld Research Station, Gobabis district, Namibia.

KERLEY, G.I.H., 1991. Seed removal by rodents, birds and ants in the semi-arid Karoo, South Africa. Journal of Arid Environments 20: 63–69.

KRAAIJ, T. & WARD, D., 2006. Effects of rain, nitrogen, fire and grazing on tree recruitment and early survival in bush-encroached savanna. Plant Ecology 186: 235–246.

LINZEY, A.V. & WASHOK, K.A., 2000. Seed removal by ants, birds and rodents in a woodland savanna habitat in Zimbabwe. African Zoology 35: 295–299.

MEIK, J.M., JEO, R.M., MENDELSON, J.R. & JENKS, K.E., 2002. Effects of bush encroachment on an assemblage of diurnal lizard species in central Namibia. Biological Conservation 106: 29–36

MEYER, K.M., WARD, D., MOUSTAKAS, A. & WIEGAND, K., 2005. Big is not better: small Acacia mellifera shrubs are more vital after fire. African Journal of Ecology 43: 131–136.

MILLER, M.F., 1994a. The costs and benefits of Acacia seed consumption by ungulates. oikos 71: 181–187.

MILLER, M.F., 1994b. Seed predation by nocturnal rodents in an African savanna ecosystem. South African Journal of Ecology 29: 262–266.

MILTON, S.J. & DEAN, W.R.J., 1996. Rates of wood and dung disintegration in arid South African rangelands. African Journal of Range and Forage Science 13: 89–93.

O’CONNOR, T.G., 1996. Individual, population and community response of woody plants to browsing in African savannas. Bulletin of the Grassland Society of South Africa 7 (Suppl. 1): 14–18.

OKELLO, B.D. & YOUNG, T.P., 2000. Effects of fire, bruchid beetles and soil type on germination and seedling establishment of Acacia drepanolobium. African Journal of Range and Forage Science 17: 46–51.

OSTFELD, R.S., MANSON, R.H. & CANHAM, C.D., 1997. Effects of rodents on survival of tree seeds and seedlings invading old fields. Ecology 78: 1 531–1 542.

QI, J., MORAN, M.S., GOODRICH, D.C., MARSETT, R., SCOTT, R., CHEHBOUNI, A., SCHAEFFER, S., SCHIELDGE, J., WILLIAMS, D., KEEFER, T., COOPER, D., HIPPS, L., EICHINGER, W. & NI, W., 1998. Estimation of evapotranspiration over the San Pedro riparian area with remote sensing and in situ measurements. Special Symposium on Hydrology paper 1.13, 11 to 16 January 1998, American Meteorological Society, Phoenix, Arizona, USA.

ROQUES, K.G., O’CONNOR, T.G. & WATKINSON, A.R., 2001. Dynamics of shrub encroachment in an African Savanna: relative influences of fire, herbivory, rainfall and density dependence. Journal of Applied Ecology 38: 268–280.

ROTHAUGE, A., 2006a. The effect of frame size and stocking rate on diet selection of cattle and rangeland condition in the camelthorn savanna of east-central Namibia. Ph.D. thesis, University of Namibia, Windhoek, Namibia.

ROTHAUGE, A., 2006b. After-effects of a hot fire in 2001/2002. Unpublished report, Neudamm Agricultural College, Windhoek, Namibia.

ROTHAUGE, A., 2007a. A decade of monitoring grazed rangeland transects at Neudamm. Unpublished report, Neudamm Agricultural College, Windhoek, Namibia.

ROTHAUGE, A., 2007b. Emergence, establishment and survival of Acacia mellifera under normal farming conditions. Unpublished report, Neudamm Agricultural College, Windhoek, Namibia.

ROTHAUGE, A., 2007c. Some principles of sustainable rangeland management in Namibia. Agricola 17: 7–15.

ROTHAUGE, A., 2008. A century of rainfall statistics at Neudamm. Unpublished report, Neudamm Agricultural College, Windhoek, Namibia.

ROTHAUGE, A., ABATE, A.L. KAVENDJI G. & NGHIKEMBUA, M.L., 2003. Forage preference of Boer goats in the Highland savanna of Namibia during the rainy season. I: Diet selection. Agricola 13: 35–42.

ROTHAUGE, A. & ENGELBRECHT, J., 2000. Forage preference of Boer goats in the Highland savanna of Namibia during the dry season. Agricola 11: 57–60.

ROTHAUGE, A., SMIT, G.N. & ABATE, A .L., 2003. The effect of the sub-habitat below savanna trees on soil, grass and its utilization by cattle. Proc. 7th International Rangeland Congress pp. 106–109, 26 July to 1 August 2003, Durban, South Africa.

SKARPE, C., 1990. Shrub layer dynamics under different herbivore densities in an arid savanna, Botswana. Journal of Applied Ecology. 27: 873–885.

SKARPE, C., 1991. Impact of grazing in savanna ecosystems. Ambio 20: 351–356.

SKINNER, J.D. & SMITHERS, R.H.N., 1990. The Mammals of the Southern African Subregion. Univ. Pretoria, Pretoria, South Africa.

SKOWNO, A.L., MIDGELY, J.J., BOND, W.J. & BALFOUR, D., 1999. Secondary succession in Acacia nilotica (L.) savanna in the Hluhluwe Game Reserve, South Africa. Plant Ecology 145: 1–9.

SMIT, G.N., 1989. Quantitative description of woody plant communities: Part I. An approach. African Journal of Range and Forage Science 6: 186–191.

SMIT, G.N., 2003. The importance of ecosystem dynamics in managing the bush encroachment problem in southern Africa. Proceedings of the 7th International Rangelands Congress, pp. 14–22, 26 July to 1 August 2003, Durban, South Africa.

SMIT, G.N. & RETHMAN, N.F.G., 1992. Inter-related floristic changes associated with different long-term grazing treatments in sourish mixed bushveld. Journal of the Grassland Society of South Africa 9: 76–82.

SMIT, G.N. & SWART, J.S., 1994. Influence of leguminous and non-leguminous woody plants on the herbaceous layer and soil under varying competition regimes in Mixed Bushveld. African Journal of Range and Forage Science 11: 27–33.

SWEET, R.J., 1982. Bush control with fire in Acacia nigrescens/Combretum apiculatum savanna in Botswana. Proceedings of the Grassland Society of Southern Africa 17: 25–28.

TEAGUE, W.R. & SMIT, G.N., 1992. Relations between woody and herbaceous components and the effect of bush-clearing in southern African savannas. Journal of the Grassland Society of South Africa 9: 60–71.

TROLLOPE, W.S.W., 1974. Role of fire in preventing bush encroachment in the Eastern Cape. Proceedings of the Grassland Society of Southern Africa 9: 67–72.

TROLLOPE, W.S.W., 1980. Controlling bush encroachment with fire in the savanna areas of South Africa. Proceedings of the Grassland Society of Southern Africa 15: 173–177.

TROLLOPE, W.S.W., 1982. Ecological effects of fire in South African savannas. In: B.J. HUNTLEY & B.H. WALKER, (eds.) Ecology of Tropical Savannas pp. 292-306. Ecological Studies Vol. 42, Springer Verlag, New York, USA.

TROLLOPE, W.S.W., 1984. Fire in Savanna. In: P. DE V. BOOYSEN &. N.M. TAINTON (eds.) Ecological Effects of Fire in South African Ecosystems Savannas pp. 199–218. Springer Verlag, New York, USA.

TROLLOPE, W.S.W., 1999. Veld burning. In: N.M. TAINTON (ed.) Veld Management in South Africa ch. 9, University of Natal Press, Pietermaritzburg, South Africa.

TROLLOPE, W.S.W., HOBSON, F.O., DANCKWERTS, J.E. & VAN NIEKERK, J.P., 1989. Encroachment and control of undesirable plants. In: J.E. DANCKWERTS & W.R. TEAGUE (eds.) Veld Management in the Eastern Cape pp. 73–89. Department of Agriculture and Water Supply, Pretoria, South Africa.

VAN VEGTEN, J.A., 1983. Thornbush invasion in a savanna ecosystem in eastern Botswana. Vegetatio 56: 3–7.

VEDDER, H., 1934. Das Alte Südwestafrika: Südwestafrikas Geschicht bis zum Tode Mahareros 1890 (The old South West Africa: South West Africa’s History until the Death of Maharero 1890). Martin Warneck Verlag, Berlin, Germany.

VIR, S., 1996. Bruchid infestation of leguminous trees in the Thar desert. Tropical Science 36: 11–13.

VON OERTZEN, D., 2007. Turning Namibian invader bush into electricity: the CBEND Project. Proc. 12th Agricultural Scientific Society of Namibia (Agrisson) Congress (CD), Neudamm Agricultural College, Windhoek, Namibia.

WALKER, B.H., 1981. Is succession a viable concept in African savanna ecosystems? In: H.H. SHUGART, D.C. WEST & D.B. BOTKIN (eds) Forest Succession: Concepts and Applications pp 431–447. Springer Verlag, New York, USA.

WALTER, H., 1971. Ecology of Tropical and Subtropical Vegetation. Oliver and Boyd, Edinburgh, UK.

WALTERS, M., MILTON, S.J., SOMERS, M.J. & MIDGLEY, J.J., 2005. Post-dispersal fate of Acacia seeds in an African savannas. South African Journal of Wildlife Research 35: 191–199.

WELTZIN, J.F., ARCHER, S. & HEITSCHMIDT, R.K., 1997. Small mammal regulation of vegetation structure in a temperate savanna. Ecology 78: 751–763.

WOLFSON, M.M. & TAINTON, N.M., 1999. The morphology and physiology of the major forage plants. Grasses. In: N.M. TAINTON (ed.) Veld Management in South Africa ch. 3. University of Natal Press, Pietermaritzburg, South Africa.

ZIMMERMANN, I & MWAZI, F.N., 2002. The effect of a prescribed burn, followed up with browsing pressure, on rangeland condition in the Mountain savanna and Karstveld of Namibia. Dinteria 27: 49–57.


BREED PREFERENCES, PRODuCTION PERFORMANCE AND MANAGEMENT OF DAIRY CATTLE AMONG SELECTED

SMALLhOLDER DAIRY FARMERS OF ZIMBABWEluCIA N. MARIuS1, E.V. IMBAYARWo-CHIKoSI2, B.T. HANYANI-MlAMBo2 and C. MuTISI2

1Directorate of Research and Training, Ministry of Agriculture, Water and Forestry, Private Bag 13184, Windhoek, Namibia2Department of Animal Science, University of Zimbabwe, P.O. Box 167, Mt. Pleasant, Harare, Zimbabwe

ABSTRACT

Smallholder dairy farming plays a pivotal role in improving local diets, income and in sustaining rural livelihoods. The objective of the study was to identify breed preferences, production performance and management of dairy breeds. A structured questionnaire was administered to 109 smallholder dairy farmers in the Guruve, Marirangwe and Nharira-Lancashire schemes. Purposive sampling was used to select the farmers. Data were analysed using SAS version 9.1.3. There was a significant difference (P < 0,05) in the breed distribution among the three study areas. In total, there were 53,6 %, 29,7 % and 16,8 % of farmers who kept crossbreds, beef and dairy breeds, respectively. There was significant difference in the proportion of farmers who selected breeds on the basis of milk yield and growth rate between at least two of the three schemes. Average milk production per cow per day was 3,08 ± 1,52 litres in Guruve, 2,76 ± 1,90 litres in Marirangwe and 2,64 ± 2,13 litres in Nharira-Lancashire schemes. The low milk production could be attributed to low-input feed resources, the use of inappropriate breeds and breed combinations. Farmers did not change their management approach on the basis of the breeds (indigenous, crossbred or exotic breeds), all breeds were treated the same. There is potential for increasing milk production from the smallholder dairy schemes, if fodder production and dairy breed constraints are improved.

INTRODuCTION

Dairying has been envisaged as a means to improve on the nutritional status and income generation for poor African families. This has led to the implementation of many developmental projects in favour of dairying (Ndambi et al., 2007). In Zimbabwe, dairying is mainly undertaken along the main watershed covering Natural Regions I, II, III and IV, where annual average rainfall ranges between 500 mm to over 1000 mm (Mupeta, 2000; Mapiye et al., 2007). The dairy industry consists of two sectors: the large-scale commercial and the smallholder dairy sectors that vary with scale of production. The large-scale commercial dairy sector originated in 1912 and has large farms with high producing exotic breeds and their crosses. In the past, this sector produced 60 % to 70 % of marketed milk for the country (Hanyani-Mlambo, 1998). The predominant dairy cattle breeds are the Holstein-Friesian, followed by Jersey and Red Dane breeds. In smallholder dairy farming, each farmer on average has 2 to 10 indigenous, exotic or

cross-bred dairy animals (Mandibaya et al., 1999). Milk yields are usually low, ranging from 2 to 5 litres per cow a day. This sector used to contribute 1 % to 2 % of national milk production (Ngongoni et al., 2006). The production problems in this sector are mainly a lack of finances to meet overhead costs, the use of inappropriate breeds, a poor feed resource base and inadequate managerial skills. The development of the smallholder dairy sector was initiated in 1983 under the Dairy Development Programme (DDP) by the Agricultural and Rural Development Authority (ARDA), a parastatal mandated to spearhead commercial agricultural and rural development projects. On initiation, the programme was funded by the Norwegian Agency for Development (NORAD). Support was also granted from Africa Now (UK), the Danish International Development Agency (DANIDA), Heifer Project International (HP) and the Government of Zimbabwe through the Public Sector Investment Programme (PSIP) (ARDA, 1997). The development of market-oriented smallholder dairy was meant to complement the large scale commercial dairying by extending the milk production base to the rural areas where the then Dairy Marketing Board (now the Dairy-board Zimbabwe Limited) found distribution of milk and milk by-products difficult (Mupeta, 1996; Mandibaya et al., 1999). The DDP initiated and implemented 20 to 30 smallholder dairy projects throughout the country, operating in various stages of development in five provinces (Mupunga and Dube, 1992; Mutukumira et al., 1996). Each scheme has a milk collection centre equipped with storage facilities. Milk was delivered to processors located in major towns. Milk was also produced for home consumption with surplus sold locally through milk collection centres. In addition, DDP provided services which included assisting the smallholder dairy farmers with acquisition of cattle, access to agricultural inputs for dairy and advice on animal management. However, despite these various efforts, established smallholder dairy enterprises were still characterised by low productivity (Hanyani-Mlambo, 1998; Munangi, 2007). This research intended to identify preferred cattle breeds and milk production performance in the smallholder dairy sector in three schemes and the criteria used by farmers in selecting dairy breeds.

Milk production of smallholder dairying has remained relatively low (Hanyani-Mlambo, 1998; Munangi, 2007). Current literature indicates that the causes are: use of inappropriate cattle breeds; shortage of fodder; limited fodder production and poor disease control measures

(Ngongoni et al., 2006; Munangi, 2007; Chinogaramombe et al., 2008). There was lack of information as to which breeds are most ideal for a smallholder dairy farming setup. Insufficient knowledge on the farming objectives, and poor extension advice had led farmers to shift from one breed to another. Breeding is not well defined and herds of individual households mix freely with other herds, particularly in the communal areas where there are no fences. Inferior bulls are rarely castrated; sometimes leading to production of progeny of inferior quality. Currently, the breeds used for milk production in smallholder dairy farms in Zimbabwe include the indigenous Mashona, Tuli, Nkone, and exotic breeds of predominantly the Red Dane, Holstein-Friesian, Jersey and crossbreds of indigenous cows and exotic bulls (Mutukumira et al., 1996; Smith et al., 2002; Munangi, 2007). However, farmers’ breed preferences and criteria used for selection and specific management of different breeds under low-input systems are yet to be explored more extensively in smallholder dairying communities of Zimbabwe. It is therefore important to investigate, and understand smallholder farmers’ views towards the performance of the various breeds they own in terms of milk production. The information generated by this study will be useful in exploring the possibilities for improvement and developing guidelines for recommendations and future research. This research was intended to identify preferred cattle breeds and milk production performance in the smallholder dairy sector in three schemes and the criteria used by farmers in selecting cattle breeds for dairy production.

MATERIALS AND METhODS

Study area

The study was conducted in Guruve, Marirangwe and Nharira-Lancashire smallholder dairy schemes. The centres were selected based on smallholder dairy schemes still operational when the study commenced, and on agro-ecological regions. Marirangwe smallholder scheme is in the Seke district, Mashonaland East Province, in agro-ecological region IIa and IIb. The average rainfall ranges from 600 mm to 1 000 mm per annum, with an average temperature of 29 °C. The major agricultural enterprise is maize and livestock production. The average land holding per farmer is about 100 hectares. The dairy herds were composed mainly of Red Dane, Holstein-Friesian, Jerseys and indigenous Mashona cattle (Mupeta, 1996; Smith et al., 2002). The Nharira-Lancashire smallholders scheme is located 170 km south east of Harare. These farms are located in agro-ecological Region III which is characterised by infrequent rainfall, ranging from 650 mm to 800 mm per annum. The temperature ranges between 10 °C to 30 °C (Mandibaya et al., 1999). The soil in this area is of granite origin with a weak to strong acidity. The vegetation mainly comprises bare ground with some scattered trees and tufts of grass (Mutukumira et al., 1996; Hanyani-Mlambo et al., 1998). Guruve is located about 151 km northeast of Harare in the Mashonaland Central Province, ecological Region II. It has three main sub-centres: Karoe farm, Guruve and Gota. All three the dairy schemes had a milk collection

centre where farmers delivered their milk. The centres were responsible for buying, processing and marketing the milk. The milk and its products were sold in areas around the centres (Mupeta, 2000).

Data collection

Structured questionnaires (pre-tested through interviews), were used for data collection. The data were collected between October 2009 and December 2009. After design of the questionnaire, 30 questionnaires were pre-tested at Dowa and Wedza schemes. The pre-tested questionnaires were reviewed and amended accordingly for the actual data collection. Those questions which were not clear to the farmers were restructured and rephrased. This process took approximately two months as from August to October 2009. The target population for the study was defined as consisting primarily of all smallholder dairy farmers in Guruve, Marirangwe and Nharira-Lancashire. A purposive sampling survey of 52 households in Guruve, 32 households in Marirangwe and 25 in Nharira-Lancashire smallholder dairy schemes was carried out. A total of 109 households which were willing to participate and owned cattle in all three sites, were interviewed. The number of respondents obtained in each dairy scheme was based on those farmers who were still active at the time when the study was carried-out. Personal interviews with the farmers were conducted in which responses from the farmers were entered onto the questionnaire by the investigator. Farmers were not given the questionnaires to fill-in. An assistant was hired in each study area to assist with the administration of the questionnaires.

Statistical analysis

Descriptive statistical analyses were carried out using SAS software version 9.13 (SAS, 2004). A Chi-square test was carried out to test for association between the breed preference and district. The Kruskal-Wallis test of the non-parametric one way analysis of variance was used to compare probability distribution of the criteria used by farmers to choose the breeds across the three districts within the schemes. This test uses the chi-square values to compare the probability distributions among the three districts, but these values approximate the Kruskal-Wallis statistic (H-statistic) when sample size is greater than five (McClave et al., 1997).

RESuLTS AND DISCuSSION

Breed preference

There was a significant difference (P < 0,05) in the breed distribution among the three study areas, as set out in Table 1. About 8,61 % and 6,70 % of farmers kept dairy breeds in Marirangwe and Guruve respectively. There were only 1,44 % of farmers who kept dairy breeds in Nharira-Lancashire. About 20,1 % of farmers in Guruve kept beef breeds, as compared to 6,71 % and 2,88 % in Marirangwe and Nharira-Lancashire, respectively. There were 22,49 % of farmers who kept crossbreds in Guruve, 16,75 % in Marirangwe and 14,36 % in Nharira-Lancashire.


the three dairy schemes had a higher proportion of farmers who selected their breeds on the basis of milk yield and growth rate of the breeds (Table 2). Farmers in Guruve tended to choose breeds on the basis of milk yield as is reflected by the relatively higher proportion of farmers with the Holstein-Friesian breed among their stock (Table 1), whilst a relatively higher proportion of farmers in Nharira-Lancashire chose breeds on the basis of their growth rate. This could probably explain the high proportion of farmers with exotic beef breed crosses in Nharira-Lancashire (Table 1). The use of milk yield as selection criterion and widespread use of the Holstein-Friesian breed among smallholder dairy farmers in Guruve, support earlier reports by Bebe et al. (2003). They stated that the Holstein-Friesian was the most preferred breed by smallholder farmers for their high milk yield. According to Ndebele et al. (2007), most smallholder dairy farmers were of the perception that the high producing, but disease prone and feed demanding exotic animals were the best. Besides milk yield and growth rate, there were no significant (P > 0,05) differences in the proportion of farmers who chose their breeds on the basis of other selection criteria between at least two of the three schemes. Table 3 below shows the mean milk (litres) yield per cow per day in the three smallholder dairy schemes.

Descriptive statistics of milk production

Table 3. Mean milk (litres) yield per cow per day in the three smallholder dairy schemes

Dairy scheme N

Mean milk yield

(litres)

Std. dev. Min. Max.

Guruve 53 3,08 1,52 0,67 8,00

Marirangwe 28 2,76 1,90 0,40 9,00

Nharira-Lancashire 23 2,64 2,13 0,63 10,00

The average daily milk production for Guruve, Marirangwe and Nharira-Lancashire schemes were 8,98 ± 6,11, 9,54 ± 5,12 and 9,54 ± 5,12 litres per herd respectively. The respective average numbers of milking cows in these areas were 3,3 4,1 and 3,8. Based on the numbers of milking cows per herd, daily milk production per cow per day was 3,08 ± 1,52, 2,76 ± 1,90 and 2,64 ± 2,13 litres in Guruve, Marirangwe and Nharira-Lancashire respectively (Table 3). Milk productions per herd and per cow in all three the schemes were somewhat similar. This was probably because during the wet season, animals gained weight and milk production was high, and in the dry season the yield and body condition of the cows declined. These results were consistent with literature which concluded that on average, milk production for crossbreds and indigenous cows in smallholder dairy was between 8 and 4 litres per cow per day, respectively, whilst, it was more than 10 litres per cow per day (300-day lactation) for purebred exotic cows (Hanyani-Mlambo et al., 1998). Ongadi et al. (2007) reported 5,40 ± 0,78 litres of milk per cow per day in smallholder dairy in free-grazing conditions in Kenya (Vihinga district). This low milk production in smallholder

dairy farming has been attributed to poor quality and inadequate quantities of feed (Bebe et al., 2008). Table 4 below shows mean milk yield per cow by sex of household head (HH) per scheme.

Gender roles in smallholder dairying

Table 4. Mean milk yield per cow by sex of household head (HH) per scheme

Dairy scheme

Sex HH N

Mean milk yield

(litres)

Std. dev. Min. Max.

Guruve Male 40 3,08 1,51 0,67 8,00

Female 13 3,10 1,62 1,00 6,00

Marirangwe Male 26 2,53 1,51 0,40 6,00

Female 2 5,75 4,60 2,50 9,00Nharira-Lancashire Male 19 2,38 1,99 0,63 10,00

Female 4 3,85 2,62 1,40 7,00

Most of the households (83,5 %) in the three smallholder dairy projects were male-headed while 16,5 % were women-headed. Milk production obtained from women-headed households tended to be higher than that obtained from male-headed households in all three the schemes (Table 4). The results indicate that women were probably more patient and their involvement in dairying had an impact on milk production. This was consistent with results from studies by Felleke (1995) who found that in Ethiopia, most duties related to small-scale dairying, were carried out by women. Similarly, Turkish women and Indian women of the middle income high caste families in the Ahmedabad and Udaipur districts of India were responsible for milking, feeding cows as well as selling the milk (Tangka et al., 2000). Table 5 below shows breed combination obtained in Guruve, Marirangwe and Nharira-Lancashire dairy schemes.

Breed combination

Table 5. Breed combination among farmers in Guruve, Marirangwe and Nharira-Lancashire

Breed combinations and their crosses

LS Mean milk yield

(litres)

Standard error

Mashona, Holstein-Friesian and Brahman 6,00 1,79

Mashona, Holstein-Friesian and Red Dane 5,50 2,53

Mashona, Holstein-Friesian, Red Dane and Brahman 4,50 2,53

Mashona and Hereford -6,00 2,53Holstein-Friesian, Afrikaner and Hereford -8,00 2,53

Mashona and Sussex -9,13 2,53

In total, 53,6 % of farmers kept crossbreds, 29,7 % kept beef breeds and 16,8% kept exotic dairy breeds (Table 1). The abundance of crossbreds amongst the three districts proves that they are preferred by farmers, probably because they were tolerant to heat and disease in arid agro-ecological regions. At least, Marirangwe had a higher proportion of farmers with dairy breeds; Red Dane crosses being one of the dominant breeds found in the area. This could be because the breed was easily accessed from the neighbouring commercial Red Dane farm. The results are in agreement with those reported in an earlier study by Chinogaramombe et al. (2008) which indicated that the dominance of a breed in an area could be attributed to the fact that it was available within the vicinity of smallholder dairy farmers. The finding that there were more dairy breeds in Guruve and Marirangwe as compared to Nharira-Lancashire, could be attributed to agro-ecological regions and preference of high genotypes with high milk production potential. Nharira-Lancashire had a higher proportion of farmers who kept beef breeds and their crosses. Table 1 below shows the percentage of farmers using the different breeds and crossbreds

Table 1. Proportion of farmers keeping the different breeds and crossbreds

Guruve Marirangwe NhariraDairy breeds

Red Dane 1,44 3,35 0,48

Jersey 1,91 1,91 0

Holstein-Friesian 3,35 2,87 0,96

Ayrshire 0 0,48 0

Total 6,70 8,61 1,44

Beef breeds

Mashona 13,88 3,83 0,48

Brahman 3,35 0,96 0,48

Afrikaner 1,91 0 0

Hereford 0,48 0 0,48

Sussex 0 0,48 0,48

Tuli 0 0,48 0,48

Simmental 0 0 0

Nguni 0,48 0,48 0,48

Nkone 0 0,48 0

Total 20,1 6,71 2,88

Crosses

Mashona x Beef breeds 6,70 6,22 1,44

Mashona x Dairy breeds 10,53 3,35 0Exotic dairy breeds x Exotic dairy breeds 1,91 2,87 0,48

Exotic dairy breeds x Exotic beef breeds 2,87 3,83 1,91

Exotic beef breeds x Exotic beef breeds 0,48 0,48 10,53

Total 22,49 16,75 14,36

X2 = 138,36 P < 0,05 (P = 0,001)

These results contrast earlier findings by Mutukumira et al. (1996), which indicated a predominance of dairy breeds, particularly the Holstein-Friesian, Jerseys and the Red Dane in Nharira-Lancashire. Exotic dairy breeds were initially given to smallholders in Nharira-Lancashire by the Heifer International Project as support to develop smallholder dairying in the project area (Chinogaramombe et al., 2008). The disappearance of the exotic dairy breeds could be attributed to the specific type of area, namely an agro-ecological region III, which is characterised by infrequent rainfall and poor vegetation available for grazing. Farming World (1998); Imbayarwo-Chikosi (2009); Masunda (2009) outlined that regardless of high milk yield of the exotic dairy breeds; they require high feeding maintenance and do not thrive well in poor pastures. Breed adaptation led smallholder farmers to a variety of farming objectives such as beef production, and they lost focus on dairying (Tabbaa and Al-Atiyat, 2009). The results are consistent with those reported in the literature by Millogo et al. (2008) in Burkina Faso; in which case there was an interest in increasing milk yield through crossbreeding with imported breeds but problems with feeding, watering and climate adaptation are more common with imported breeds than with local cow breeds. Table 2 below shows the proportion of farmers per district who selected their breeds based on a specific criterion.

Table 2. Proportion of farmers selecting breeds on specific criterion by district

CriteriaProportion of farmers (%) X2

(or H) valueGuruve Marirangwe Nharira-

LancashireMilk yields 58,97 27,10 23,36 12,92*Fat yields 50,00 50,00 0,00 0,96Body weight 47,37 10,53 42,11 2,11Growth rate 35,71 14,29 50,00 9,01*Fertility 0,00 100,00 0,00 9,10Disease tolerant 39,02 36,59 24,39 1,83

Feeding behaviour 11,11 55,56 33,33 0,48

Draft power 64,67 12,50 20,83 5,82Breeding 33,33 66,67 0,00 0,00Beef production 70,00 0,00 30,00 0,00

Drought resistant 25,00 45,00 30,00 5,35

* Significantly different at P < 0,05 NB: the chi-square value approximates the Kruskal-Wallis statistic (H-statistic) when N > 5

Selection criteria

There were significant differences (P < 0,05) in the proportion of farmers who selected breeds on the basis of milk yield and growth rate between at least two of the three districts (Table 2). This implies that at least one of


whilst 4 % of farmers in Nharira-Lancashire did not control parasites, since they did not consider them to be a big threat to productivity. However, all the farmers had access to dipping and spraying facilities, although the frequency of dipping varied across the three schemes (Figure 1). The majority of farmers in Marirangwe (87,10 %) dipped cattle only when they had high tick loads, whilst the majority of farmers in Guruve and Nharira-Lancashire dipped their cattle once a week and twice a week, respectively. Medicines were available to the households through local veterinary extension assistants, but most of the households did not vaccinate cattle, due to a lack of money (Ngongoni et al., 2006). Of the farmers who vaccinated, vaccinations were carried out as shown in Figure 1 and the diseases that were vaccinated against, are presented in Table 9.

Table 8. Frequency of occurrence of disease and parasite events experienced among the farmers in the three schemes

DiseaseProportion of farmers (%)

Guruve Marirangwe Nharira-Lancashire

Tuberculosis and Contagious abortion 7,50 9,70 16,00*

Blackquarter (Black-leg, Quarter evil) 11,30 6,50 36,00*

Lumpy skin disease 47,20 32,30 30,00*Tick-borne disease 73,60 61,30 88,00*Mastitis 26,40 19,40 28,00Eye infections 5,70 3,20 8,00Abscess 3,80 0,00 8,00*Foot and Mouth Disease 1,90 3,20 0,00

Anthrax 1,90 0,00 4,00Rabies 0,00 0,00 0,00Scours (diarrhoea) 1,90 0,00 0,00Pneumonia 0,00 3,30 0,00

*Significant at P < 0,05

Vaccination

Vaccinations for specific diseases were not carried out (X2: P > 0,05) on the basis of breed combinations of the farmer; this implies that the farmers did not vaccinate according to whether they had certain breeds or not. All breeds, exotic, indigenous and their crosses, received the same vaccinations. A higher proportion of farmers in Guruve (35,8 %) and Marirangwe (32,3 %) vaccinated their cattle only when there was an outbreak of disease of any kind in the area, whilst in Marirangwe and Nharira-Lancashire dairy schemes 35.5 % and 36,0 % of farmers respectively, vaccinated annually (Figure 1). The majority of famers vaccinated against Lumpy skin disease in Guruve (47,2 %) and Marirangwe (74,2 %). About 72 % of farmers vaccinated against Quarter evil in Nharira-Lancashire (Table 9). The frequency of vaccination varied depending on the type of disease and whether it is a notifiable disease

such as Anthrax, Foot and Mouth Disease and Brucellosis; diseases which are normally vaccinated by law or the government through veterinary services. The sector has faced many disease challenges, including Brucellosis, due to reduced veterinary services delivery following the economic depression that has affected the country since the year 2000 (Matope et al., 2010).

Table 9. Diseases that were vaccinated against in the three schemes

Disease

Proportion of farmers vaccinating (%)


Tuberculosis and Contagious abortion 3,80 12,90 16,00

Blackquarter (Black-leg, Quarter evil) 28,30 29,00 72,00

Lumpy skin disease 47,20 74,20 12,00Foot and Mouth Disease 5,70 9,70 16,00

Anthrax 7,50 12,90 20,00

NB: Multiple responses

Figure 1. Cattle vaccination in Guruve, Marirangwe and Nharira-Lancashire.

Mating

Most farmers in Marirangwe (67,7 %) and Nharira-Lancashire (80 %) monitored the mating of the cattle whilst the majority of farmers in Guruve (50,9 %) did not. However, most of the farmers in the three schemes had their own bulls (Guruve 39,6 %; Marirangwe 58,1 %; Nharira-Lancashire 84 %). Of the farmers monitored and who could identify the bulls mated to their cows, 22,6 %, 19,4 % and 60 % of the farmers in Guruve, Marirangwe and Nharira-Lancashire respectively, were mating bulls to their related

40

35

30

25

20

15

10

5

0

Pro

porti

on o

f far

mer

s (%

)

When there is an outbreak

Twice per year

Annually Three times a

year

No vaccination

GuruveMarirangweNharira-Lancashire

35,8

32,3

28

13,2

29

20

35,536

16,9

3,2

8

11,3

8

0

Frequency of vaccination

A combination of Mashona, Holstein-Friesian and Brahman breeds and their crosses produced a higher average milk yield per cow. The Mashona and Sussex breeds and their crosses had the lowest average milk yield (Table 5). In herds where indigenous Mashona cattle were crossed with beef breeds, milk yields were observed to be much lower as compared to when they were crossed with exotic dairy breeds (Table 5). This was probably because exotic dairy breeds have a greater genetic potential for milk production whilst beef breeds have a dominant heritable trait for growth which favours beef production. Holstein-Friesians have a unique genetic ability to adapt to a diversity of agro-ecological regions, although when compared to indigenous breeds, they are not as tolerant to heat and disease in arid agro-ecological regions (Imbayarwo-Chikosi, 2009). Although crossbred cattle were well adapted to marginal production conditions, indications are that they have poor dairy characteristics (Smith et al., 1994). In Bangladesh, average daily milk yield of Holstein x indigenous breeds and Jersey x indigenous crossbreds were 5,5 ± 0,1 and 3,8 ± 0,1 kg, respectively (Nahar et al., 1992). Table 6 below shows the quantity of supplementary feed given to lactating cows during milking time in the three schemes.

Effect of supplementation on milk yield

Table 6. Supplementation of lactating cows during milking

Quantity of supplementary feed LS Mean (kg) Standard error

1 kg to 5 kg 3,75 2,53Ad libitum 3,25 1,79Do not feed 2,25 2,53

Households that supplemented feeding during milking, produced more milk that those that did not supplement as shown in Table 6. This was probably because some of the feeds available to the animals were mainly of poor quality forage such as maize stover and grass hay. Secondly, some farmers were not supplementing, probably because towards the end of the dry season (October to December) which co-incided with data collection of the study, their crop reserves were already depleted. Concentrates such as dairy meal are expensive and they are probably not readily available in the area. In smallholder dairying, during the wet season, animals gained weight and milk production was high, and in the dry season the yield and body condition of the cows declined (Ranjhan, 1999). In Addis Ababa, Khalili et al. (1992) demonstrated significant increases in milk yield of crossbred cows fed hay or oat–vetch hay and supplemented with increasing levels of concentrate. Milk production in the three dairy schemes was significantly (P < 0,05) influenced by breed combination and the interaction of feed quantities and breed combination. District and quantities fed had no effect (P > 0,05) on the milk production. Breed combination contributed 46,72 % of the observed variation in milk production per cow in the three districts, whilst the interaction of breed combination and feed quantities fed to cows was low at 4 %. In Ethiopia, Abraha et al. (2009) reported higher milk yields in crossbreds than indigenous

cows; however crossbred cows under environmental stress and challenge of high risk of diseases coupled with poor feeding strategy, produce milk yields below their genetic potential. Table 7 below shows the utilisation of the different conserved feeds by farmers in the three schemes.

Feeding management

Table 7. Proportion of farmers who conserved cattle feeds in the three schemes

FeedProportion of farmers feeding (%)


Natural pasture 17,30 16,10 100Crop residues 75,00 32,30 52,00Silage 13,50 41,90 20,00Grass hay 36,50 35,50 56,00Mineral blocks 7,70 6,50 16,00

NB: Multiple responses

The majority of farmers (75 %) in Guruve used crop residues as a supplementary feed for their dairy cows as outlined in Table 7. This was probably because maize is one of the common crops grown in the area. In Marirangwe, 41,9 % of farmers used silage as compared to the other two schemes whilst 56 % of famers in Nharira-Lancashire used grass hay for their lactating cows (Table 7). Farmers in Marirangwe used more silage, probably because they have learnt this from their neighbouring commercial farms. Farmers in Nharira-Lancashire tended to use more grass, hay and natural pasture for their dairy stock. This could be due to the agro-ecological Region III which is characterised by poor rainfall and high temperatures that typify the region. The vegetation comprises mainly bare ground with some scattered trees and tufts of grass (Mutukumira et al., 1996; Hanyani-Mlambo et al., 1998). In Nharira-Lancashire, more farmers tended to supplement with mineral blocks because their pasture is low in minerals that lack in their livestock diet. These findings are similar to those reported in the literature by Mupeta (2000); Mapiye et al. ( 2006); Muchenje et al. (2007) concluded that natural pasture and crop residues are the primary feeds available to dairy cattle in the smallholder sector of Zimbabwe (Mupeta, 2000). Table 8 below shows the frequency of occurrence of disease events in the three schemes.

health management

The probability distribution of farmers who had problems with tuberculosis, contagious abortion and tick-borne diseases in Nharira-Lancashire was significantly higher (H: P < 0,05) than in the other districts. The proportion of farmers who had problems with lumpy skin disease was significantly higher (H: P < 0,05) in Guruve than in the other districts. There were no significant differences (H: P > 0,05) in the proportion of farmers among the three schemes with respect to all the other diseases. All farmers in Guruve and Marirangwe controlled parasites in one way or another


management practices in the Gwayi smallholder farming area of South-Western Zimbabwe. liv. Res. Rrl. Dev. 19, 12.

NGONGONI, N.T., MAPIYE, C., MWALE, M. & MUPETA, B., 2006. Factors affecting milk production in the smallholder dairy sector of Zimbabwe, liv. Res. for Rrl. Dev. 18, 5.

ONGADI, P.M., WAKHUNGU, J.W.R., WAHOME, G. & OKITOI, L.O., 2007. Characterization of grade dairy cattle owning households in mixed small scale farming systems of Vihiga, Kenya. liv. Res. Rrl. Dev. 19, 3.

RANJHAN, S.K., 1999. Dairy feeding systems, In: Smallholder dairying in the tropics by International Livestock Research Institute (ILRI), Nairobi, Kenya. pp. 117–132.

SAS., 2004. SAS Institute Inc, SAS/STAT User’s Guide. SAS Institute Inc, Cary, NC, USA.

SMITH, T., MOYO, S., RICHARDS, J.I. & MORTON, J.F., 2002. The role of indigenous and cross-bred cattle for smallholder dairy production in Zimbabwe. Harare, Zimbabwe.

SMITH, T., MOYO, S., BEFFA, M.L. & NDLOVU, K., 1994. Crossbreeding of indigenous cattle for milk production. In: Proceedings of the workshop on Integrated Livestock/crop Production Systems in the small-scale and communal farming systems in Zimbabwe. Ed. Mutisi, C., Gomez, M., Madsen, J. and Hvelplund, T. Depart. Anim. Sci. The Danish Royal Veterinary and Agricultural university (RVAu) and Danish National Instit. of Anim. Sci. (NIAS).

TABBAA, M. & AL-ATIYAT, R., 2009. Breeding objectives, selection criteria and factors influencing them for goat breeds in Jordan. Small Rum. Res. 84, 8–15.

TANGKA, F.K., JABBAR, M.A. & SHAPIRO, B.I., 2000. Gender roles and child nutrition in livestock production systems in developing countries: A critical review. Socio-economic and Policy Research Working Paper 27. (Int. Liv. Res. Instit. ), ILRI. Nairobi, Kenya. p. 64.p

cows. Ngongoni et al. (2006) stated that exotic bulls were not readily available and when available, the price was as high at US$ 750 to US$ 1000 per bull. Chinogaramombe et al. (2008) reported that about 70 % of the farmers practiced uncontrolled breeding. Therefore, with all these breeding challenges, farmers were forced to use breeding bulls available within their vicinity.

CONCLuSION

The abundance of crossbreds among the three schemes shows the preference of the farmers for them. In Guruve, farmers favoured crossbreds between indigenous and dairy breeds (i.e. Holstein-Friesian, Jerseys and Red Dane) whilst Nharira-Lancashire favoured crossbreds between indigenous and beef breeds (such as Brahman, Afrikaner, Simmental, Sussex, etc). The Red Dane breed was favoured in Marirangwe, which is explained by the high proportion of farmers who kept the breed as compared to other two schemes. Farmers in Guruve chose breeds on the basis of milk yield, as was reflected by the relatively higher proportion of famers with the Holstein-Friesian breed among their stock, whilst a relatively higher proportion of farmers in Nharira-Lancashire chose beef breeds on the basis of their growth rate. Milk production in the three areas was low. Farmers did not change their management on the basis of different breeds kept; all crossbreds, exotic and indigenous breeds were treated the same and management has remained suboptimal. Smallholder dairy farmers did not make use of a systematic mating system and natural service was the most common mating method used.

ACkNOWLEDGEMENTS

The author wishes to thank supervisors of the Department of Animal Science, Department of Economics and Extension, University of Zimbabwe for guidance and unweaving support throughout the study, ICART/EU and Zimbabwe National Association of Dairy Farmers for funding the study; and to the Ministry of Agriculture, Water and Forestry, Directorate of Research and Training of Namibia for the opportunity to acquire scientific knowledge through post graduate training.

REFERENCES

ABRAHA, S., BELIHU, K., BEKANA, M. & LOBAGO, M.F., 2009. Milk yield and reproductive performance of dairy cattle under smallholder management system in North-eastern Amhara Region, Ethiopia. Trop. Anim. Hlth. Prod. 41, 1597–1604.

ARDA, 1997. Agricultural and Rural Development Authority. Annual Report 1995/96.

BEBE, B.O., UDO, H.M., ROWLAND, J.G.J. & THORPE, W., 2003. Smallholder dairy systems in the Kenya highlands: Breed preferences and breeding practices. liv. Prod. Sci. 82, 117–127.

BEBE, B.O., UDO, H.M. & THORPE, W., 2008. Characteristics of feeding and breeding practices for intensification of smallholder dairy systems in the Kenya highlands. liv. Res. Rrl. Dev. 20, 2.

CHINOGARAMOMBE, G.N., MUCHENJE, C., MAPIYE, V.C., NDHOVU, T., CHIMONYO, M. & MUSEMWA, L., 2008. Challenges for improving small-holders dairy production in the semi-arid areas of Zimbabwe. liv. Res. Rrl. Dev. 20, 3.

FARMING WORLD., 1998. Stud Breeders Directory in Association with the Zimbabwe Herd Book. 24, 6. Mt Pleasant, Harare, Zimbabwe.

FELLEKE, G., 1995. Women and dairy development in Ethiopia: A case study of introduction of cross-bred Dairy cattle into the Mixed Farming System. In: Proceedings of the Workshop on the Regional Exchange Network for Market oriented Dairy Development (Dec. 1995). Compiled by M. Matanda. University of Zimbabwe.

HANYANI-MLAMBO, B.T., 1998. Socio-economic Aspects of Smallholder Dairying in Zimbabwe. liv. Res. Rrl. Dev. 10, 2.

IMBAYARWO-CHIKOSI, V.E., 2009. Dairy Cattle Genetic and Breeding Module. Department of Animal Science, University of Zimbabwe, Mt Pleasant, Harare.

KHALILI, H., VARVIKKO, T. & CROSSE, S., 1992. The effects of forage type and level of concentrate supplementation on food intake, diet digestibility and milk production of crossbred cows (Bos taurus X Bos indicus). Anim. Prod. 54, 183–189.

MANDIBAYA, W., MUTISI, C. & HAMUDIKUWANDA, H., 1999. Calf rearing systems in smallholder dairy farming areas of Zimbabwe: A diagnostic study of the Nharira-Lancashire area. AJAS. 12, 68–76.

MAPIYE, C., FOTI, R., MWALE, M., CHIKUMBA, N., POSHIWA X., CHIVURAISE, C. & MUPANGWA, J.F., 2006. Constraints to Adoption of Forage and Browse Legumes by Smallholder Dairy Farmers in Zimbabwe. liv. Res. Rrl. Dev. 18, 12.

MASUNDA, B., 2009. Dairy Production Module. Department of Animal Science, University of Zimbabwe, Mt Pleasant, Harare.

MATOPE, G., BHEBHE, E., MUMA, J.B., LUND, A. & SKJERVE, E., 2010. Herd-level factors for Brucella seropositivity in cattle reared in smallholder dairy farms of Zimbabwe. Prev. Vet. Med. 94, 213–221.

MCCLAVE, J.T., DIETRICH, F. H. & SINCICH, T., 1997. Statistics. 7th ed. Prentice Hall Inc. Publ, Upper Saddle River, New Jersey.

MILLOGO, V., OUÉDRAOGO, G.A., AGENÄS, S. & SVENNERSTEN-SJAUNJA, K., 2008. Survey on dairy cattle milk production and milk quality problems in per-urban areas in Burkina Faso. African J. Agric. Res. 3, 215–224.

MUCHENJE, V., CHIMEDZA-GRAHAM, R., SIKHOSANA, J.L.N., ASSAN, N., DZAMA, K. & CHIMONYO, M., 2007. Milk yield of Jersey x Nguni and Tuli F1 and F2 cows reared under small holder farming conditions. South African J. Anim. Sci. 8, 7–10.

MUNANGI, W., 2007. Strategies for improving the contribution of Smallholder Dairy Sector to milk supply in Zimbabwe. Zimbabwe National Dairy Symposium. Harare, Zimbabwe.

MUPETA, B., 1996. Development and Opportunities for market oriented smallholder dairy production in Zimbabwe: Nharira/Lancashire case study. Paper prepared for proceedings of the Workshop on the Regional Exchange Network for Market oriented Dairy Development.

MUPETA, B., 2000. Studies on the potential of ram press sunflower cake (Helianthus anus) as a source of protein for moderate milk production in crossbred (Bos Taurus x Bos indicus) cows in the smallholder dairy sector of Zimbabwe. Ph.D. (Agric) thesis. University of Zimbabwe, Harare.

MUPUNGA, E.G. & DUBE, D.M.J., 1992. Smallholder dairy development programme in resettled and communal areas in Zimbabwe. Agricultural Development Authority, Dairy Development Programme, Harare, Zimbabwe.

MUTUKUMIRA, A.N., DUBE, D.M.J., MUPUNGA, E.G. & FERESU, S., 1996. Smallholder milk production, milk handling and utilization: A case study from the Nharira/Lancashire farming area, Zimbabwe. liv. Res. Rrl. Dev. 8, 1.

NAHAR, T.N., ISLAM, M. & HASHATH, M.A., 1992. A comparative study on the performance of F1 crossbred cows under rural conditions. Asian-Aust. J. Anim. Sci. 5, 435–338.

NDAMBI, O.A., HEMME, T. & LATACZ-LOHMANN, U., 2007. Dairying in Africa-Status and recent developments. liv. Res. Rrl. Dev. 19, 8.

NDEBELE, J.J., MUCHENJE, V., MAPIYE, C., CHIMONYO, M., MUSEMWA, L. & NDLOVU, T., 2007. Cattle breeding


on whether it struck, or fell under, a plant canopy or not. Thereafter the botanical abundance of all plant species encountered was recorded.

Methodology of the carrying capacity

Before the horses were introduced into the experimental paddock, the herbaceous yield was determined by clipping 1 m2 evenly spaced quadrants along the transect. This was done by counting 40 steps of the observer’s pace between each quadrant to be cut. About 45 quadrants were harvested along the diagonal transect of the experimental paddock from one opposite corner of the paddock to the other. The quadrants were placed at random. This was achieved by throwing the quadrant backwards over the observer’s head so that the operator did not see where it fell and therefore could not select the patch to be sampled. At harvesting, the plant material was clipped at ground level using a pair of scissors. These plant materials were sorted into two bags, respectively into palatable and unpalatable plant species. The carrying capacity of the camp was calculated, expressed in kilogram per animal biomass.

Methodology of the diet selection observations

Firstly, an adaptation period of about 3 to 4 days allowed the observer to familiarise herself with the existing vegetation in the paddock i.e. grass species and shrubs. During the adaptation period the observer was enabled to get accustomed to the horse herd. The bite counts method was used in this experiment to quantify the relative number of different classes of plant species selected. The bite counting method was originally developed for goats (Narjisse, 1991), but was found to be applicable to horses too (Aganga, lETSO, & AGANGA, 2000). It is believed that animals are more actively feeding during morning and afternoon hours because those are the times of the day when it is cool (Narjisse, 1991).

In a herd of 22, six horses are randomly selected for each treatment and each horse is observed for a period of 10 minutes. The observation distance was three metres or less. During the ten minutes of observation, all bites taken were counted, and all plant parts or organs utilised were recorded. Should the horse have interrupted his feeding session in the ten minute observation period, the stopwatch was stopped and recording only resumed when the horse restarted feeding. In such a case, observation continued until the full 10 minute observation period of feeding has been reached. This procedure was repeated on three consecutive days, both in the morning and afternoon per treatment.

From these data analysed, the dietary abundance of every forage species was calculated as a percentage based on the frequency of its occurrence in the diet.

Sample collection and Nutrient analysis

After diet selection observation, all samples from every utilised forage plant species collected, (either by hand-clipping or by using a pair of scissors in a manner imitating

the observed selection pattern of horses during bite counting, were collected. These samples were immediately sealed in small plastic bags to retain their field moisture content (although some plants were already dry), weighed and oven-dried, grounded through a 1 mm sieve and subjected to standard chemical analysis to determine their nutritive value. This is presuming to indicate the nutritive value of the selected diet.

Nutrient analysis was done using the proximate system of analysis. Each single sample was used to determine dry matter by heating the sample to extract moisture from the sample at a temperature of about 105 ºC. Ash was determined by burning the samples in a muffle furnace at 500 ºC to 600 ºC for 2 to 3 hours. Crude protein was determined by the Kjeldahl method. Determination of Ether Extract (lipid/fat) was done by using the Sox-let fat extraction apparatus for a period of 12 hours. Crude fiber was obtained by boiling the sample in weak acid and weak alkaline. Lastly, the Nitrogen-Free-Extract was calculated by subtracting the sum of all the other fractions above (NFE % = 100 – % Ash – % CF – % EE – % CP).

RESuLTS AND DISCuSSION

Botanical survey

The botanical composition was determined before the herd of horses was introduced into the paddock. The following table shows all plant species encountered

Table 1. Botanical composition of the experimental paddock

Species Strikes Frequency (%)Acacia mellifera 60 12,2Anthizoma angustifolia 2 0,4Aristida congesta 3 0,6Aristida effusa 22 4,5Aristida meridionalis 3 0,6Boscia albintruca 3 0,6Cenchrus ciliaris 36 7,3Cyperus spp. 1 0,2Enneapogon cenchroides 77 15,7Eragrostis echinochloidea 1 0,2Eragrostis rotifer 91 18,5Fallen grasses 9 1,8Grewia flava 3 0,6Herbs (unknown) 6 1,2Melinis repens 7 1,4Pogonarthria fleckii 23 4,7Schmidtia pappophoroides 9 1,8Stipagrostis uniplumis 119 24,2Tribulus terrestris 4 0,8Unidentified grasses 9 1,8urochloa brachyura 3 0,6Total 491 100

ABSTRACT

The diet selection of free ranging horses was determined at Seeis farm during June, 2004. Direct animal observation technique, technically known as the bite count method, was used and the bites taken from various plant species were compared to the percentage occurrence of those species available in a natural range as determined by systematic step point sampling. Data were analysed using Microsoft Excel version 1997. The horses ate grass species, herbs, bushes, shrubs, bark of the tree, and fallen grass materials. The principal forage was Stipagrostis uniplumis (24,7 %) and Eragrostis rotifer (18,7 %). The most preferred forage was the fallen grass material at 30,1 %. There are differences among means in the chemical composition of different forage species, utilized with regard to dry matter, ash, crude fibre and crude protein. All grass species tend to have little fat content that ranges from 1 % to 1,3 %. Acacia mellifera leaves have the highest fat content among forage species utilized by the horses.

INTRODuCTION

Livestock production contributes about 10 % to 15 % to the Gross Domestic Product, depending on annual rainfall. According to the National Planning Agricultural Census (1997), the horse population in 1994/95 was about 19 886. In Namibia, horses are used for recreation (i.e. racing, sport and riding), driving carts, hunting, pulling and transporting heavy equipment, and for ranching purposes. Horse breeders breed mares and sell the offspring that can be trained for horseracing, shows as well as for special cultural events. The season in which animals graze the given area, is important. Diet will tend to change a little with the changing of the season, because often animals eat what is most nutritious or available during the specific season (Tainton, 1988). Rangeland is a very heterogeneous pasture with a multistratified distribution of forage resources, subject to important quantitative and qualities variations which depend on the season (Tainton, 1988). The quantity of forage eaten each day depends on the time spent grazing, the rate of biting, and the size of each bite (Minson, 1990). Nutrient intake in both quality and quantity of herbage selected are virtually impossible to measure directly; therefore chemical analysis has to be done on the various plant samples so that the nutrient content can be determined (Forbes, 1995). Diet selection research has been done by many others on sheep, goats

DIET SELECTION OF FREE RANGING hORSES IN ThE hIGhLAND SAVANNA OF NAMIBIA:

A CASE STuDY AT SEEIS FARMluCIA N. MARIuS and AxEl RoTHAuGE

Ministry of Agriculture, Water and Forestry, Private Bag 13184, Windhoek, Namibia

and cattle in the highland savanna of Namibia (Kavendjii, 1999; Kamupingene, 2000 and Rothauge, 2006), but there is little or no information available on diet selection of a horse. Feeding and nutrition are important aspects in any farm animal. The ability of herbage to satisfy animal requirements for growth, and or production, maintenance and reproduction depends largely on the nutritive value (chemical composition) of the herbage (Tainton, 1988). The information on diet selection will guide farmers with proper range management plans, for example to know the most preferred forage species for the particular animal species. The main objectives of the study is to investigate what a horse really eats in an extensive grazing condition, identify principal and preferred forages species of the vegetation and to analyse the selected specific plant parts consumed and to know what nutrients are lacking in the horse’s diet, to enable the farmer to supplement them.

MATERIALS AND METhODS

Trial site

The experimental research was conducted at Seeis farm, situated about 47 km east of Windhoek. The farm is privately owned by Dr Wolfgang Späth and it started operating in 1994. The farm is 2 000 hectares in size (17 camps/paddocks), and there are about 133 horses that are being kept there. The annual rainfall is about 300 mm per annum. The area is part of the Highland savannah and the overall carrying capacity for highland savannah ranges from 8 to 10 ha/LSU in good years and 18 ha/LSU to more than 20 ha/LSU in bad years (Giess, 1971). The data was collected at the end of the growing season (June).

Methodology of the botanical composition and ground cover

The experiment was conducted in an un-grazed paddock. Immediately before a treatment horse herd was allowed to graze in the experimental paddock, its botanical composition was determined by a systematically placed step-point sampling method. A three-metre-long iron rod was placed along the paddock’s diagonal transect from one corner of the paddock to the opposite one. During the determination of the abundance of plants along the treatment camp’s diagonal transect, the canopy cover of the soil was determined by letting the three metre iron rod fall freely onto the surface, classifying the exact point of impact of the falling rod either as “bare” or “covered”, depending


Table 5. Tests of between subjects’ effects, to show whether there are differences among species, time (1st to 4th), horses and day (morning versus afternoon)

Source df F Significance PSpecies 16 6,870 0,000 < 0,01Time 1 0,617 0,433 not significantHorse 5 0,923 0,446 not significantDay 3 4,377 0,005 < 0,01

According to the table above, the statistical analysis confirmed that there are highly significant differences (P < 0,01) among species consumed, unlike among horses and time used for the observation.

Chemical composition of plant parts utilised by horses

Table 6 shows the average mean values of forages obtained during chemical analysis using the proximate system of analysis.

Chemical analysis

As indicated in Table 6, there are differences among means in the chemical composition of different forage species utilised, with regard to Dry Matter, Ash, Crude Fibre and Crude Protein. All grass species tend to have little

fat content that ranges from 1 % to 1,3 %. Acacia mellifera leaves have the highest fat content among forage species utilised by the horses. A mature 700 kg lactating mare with a foal requires 13 % to 14 % Crude Protein (National Research Council, 1989). Knowing what a horse requires per day, and how much chemical composition is available in all forage species utilised, will not enable one to determine how much is lacking in the diet. This is because, for example, the amount of Crude Fibre consumed per day was not measured.

Carrying capacity

The carrying capacity estimations indicated that the experimental paddock is capable of keeping 16 Large Stock Unit (LSU) for a period of one year without veld deterioration or loss of animal condition. The paddock has a potential of sustaining 95 mares plus foals for a period of 60 days, but this farmer allocated only 22 mares with foals for 60 days, resulting in 23,2 % paddock utilisation. According to Meissner, Hofmeyr, Van Rensburg, & Pienaar, 1983, a 700 kg mare with a foal need 1,65 LSU, assuming that the mare consumes about 3 % of its body mass per day. The following calculation indicates the procedure for the calculation of the carrying capacity of the experimental paddock:

Table 6. Chemical composition of forage species utilised by horses

Species DM % ASH % FAT % CF % CP % NFE %Aristida effusa 98,2a 6,6ghi 1,0b 37,6ef 5,4efgh 49,5bc

Angustifolia anthizoma 93,7cdef 14,2c 1,4b 31,0h 10,2cd 42,3d

Aristida congesta 96,2b 3,4j 1,1b 33,4gh 5,9efgh 56,4a

Acacia mellifera bark 92,0fg 10,2a 1,3b 70,2b 13,2b 5,2i

Acacia mellifera leaves 94,0ced 25,0e 3,1a 19,0j 9,2d 43,8d

Boscia albatrica 92,9ef 6,4ghi 0,5b 74,8a 16,5a 1,9j

Cenchrus ciliaris 95,4bc 8,9ef 1,1b 36,3fg 6,8ef 47,0c

Cysperus spp. 95,1bcd 20,0b 1,4b 38,9ef 5,3fgh 34,8f

Enneapogon cenchroides 96,3b 18,8b 1,5b 49,9c 6,3efgh 23,6h

Eragrostis echinochloidea 96,1b 7,6fg 1,0b 46,9cd 5,1fgh 38,8e

Eragrostis rotifer 95,1bcd 6,1hi 1,1b 45,9d 5,7efgh 41,4d

Grewia flava 93,4def 12,8d 1,1b 25,1i 11,2c 49,8b

Melinis repens 94,8bcd 7,3gh 1,4b 47,0cd 6,5efg 38,0e

Pogonathria fleckii 95,3cb 5,4i 1,3b 43,8d 5,8efgh 43,7d

Schmidhia pappophoroides 90,7g 7,6fg 1,3b 38,8ef 4,6h 47,7bc

Stipagrostis uniplumis 94,8bcd 5,8hi 0,8b 39,9ef 6,5efg 47,1c

Tribulus terrestris 95,0bcd 12,3d 1,5b 50,0c 9,0d 28,9g

urochloa brachyura 96,1b 12,3d 0,7b 32,1h 7,2e 47,8bc

Mean ± s.e. 94,8 ± 0,53 10,1 ± 0,46 1,2 ± 0,35 41,9 ± 1,12 7,6 ± 0,53 39,2 ± 0,76Palatable 94,8bcd 6,1hi 1,2b 37,3ef 6,6efg 48,9bc

Unpalatable 95,8b 6,1hi 1,1b 40,3e 4,8gh 47,9bc

abcdefg Within the column, means with similar superscripts do not differ (P < 0,05)

The experimental paddock was dominated by Stipagrostis uniplumis, Eragrostis rotifer and Enneapogon cenchroides grass species, as seen in the Table 1. Trees, shrubs and herbs contributed the least to the botanical composition of the experimental site.

Table 2. Ground cover and bare area percentage

Ground surface Strikes % Covered/bareBare 144 22,7Covered ground 491 77,3Total 635 100 %

The frequencies of species occurrence are depicted in Table 1 with the general ground cover, partitioned into bare patches and covered area, reported in Table 2. Various grass species, herbs, shrubs and bushes contributed to the covered portion. It was observed that 77,3 % of the paddock/camp was covered in forage and 22,7 % consisted of bare patches. The used paddock had enough vegetation available for grazing.

Diet selection

Table 3 represents the total number of bites per forage species utilised, during the feeding of horses over the four days of observation in both morning and afternoon sessions.

Table 3. Total number of bites per forage species and their parts utilised by horses

Species Bites Frequency (%)

Plant part utilised

Angustifolia anthizoma 109 1,09 LeavesAristida congesta 28 0,3 Inflores.Aristida effusa 73 0,7 Inflores.Acacia mellifera (bark) 102 1,02 BarkAcacia mellifera (fallen leaves) 82 0,82 Fallen

leavesBoscia albintruca 105 1,05 BarkCenchrus ciliaris 506 5,1 Inflores.Cyperus spp. 3 0,03 Inflores.Enneapogon cenchroides 754 7,6 Inflores.Eragrostis echinochloidea 0 0 Inflores.Eragrostis rotifer 1870 18,7 Inflores.Fallen grasses 3093 30,1 Inflores.Grewia flava 10 0,1 LeavesMelinis repens 12 0,12 Inflores.Pogonarthria fleckii 115 1,2 Inflores.Schmidtia pappophoroides 34 0,3 Inflores.Stipagrostis uniplumis 2466 24,7 Inflores.

Tribulus terrestris 524 5,3 Whole plant

urochloa brachyura 94 0,9 Inflores.Total 9980 100%

Inflores. = Inflorescence

There is a significant difference among the species that were consumed more when compared to those less utilised.Among all forage species utilised, fallen grass materials obtained the highest bite frequency (30,1 %). Stipagrostis uniplumis and Eragrostis rotifer followed with 24,7 % and 18,7 %, respectively. The principal species are those forage species that contributed the most to the animal’s diet, namely; S. uniplumis and E. rotifer. In Table 3, fallen grass material was consumed for 30,1 % more than they had occurred (1,8 % from Table 1) in the paddock. This means fallen materials were the preferred forage by the horses.

Dietary preferences

The diet preference ratio is determined by dividing dietary abundance (%) over the botanical abundance (%).

Table 4. Comparison of dietary and botanical abundance of different forages

SpeciesBotanical

abundance (%)

Diet abundance

(%)

Dietary preference

ratio Angustifolia anthizoma 0,42 1,09 2,60

Aristida congesta 0,63 0,28 0,45Aristida effusa 4,62 0,73 0,16Aristida meridionalis 0,63 0 0Acacia mellifera 12,61 1,84 0,15Boscia albintruca 0,63 1,05 1,67Cenchrus ciliaris 7,56 5,07 0,67Cyperus spp. 0,21 0,03 0,14Enneapogon cenchloides 16,18 7,56 0,47

Eragrostis echinochloidea 0,21 0 0

Eragrostis rotifer 19,12 18,74 0,98Fallen grasses 1,89 30,99 16,40Grewia flava 0,63 0,10 0,16Melinis repens 1,47 0,12 0,08Pogonathria fleckii 4,83 1,15 0,24Schmidtia pappophoroides 1,89 0,34 0,18

Stipagrostis uniplumis 25,00 24,71 0,99

Tribulus terrestris 0,84 5,25 6,25urochloa branchyura 0,63 0,94 1,49Total 100 100 33,08

Key: DPR > 1,0 = preferred forage species

According to Table 4, fallen grass materials are more preferred with a dietary ratio of 16,40 followed by T. Terrestris (6,25). Also, A. angustifolia and U. brachyura were consumed more than they appeared in the experimental paddock with 2,60 and 1,49, respectively.


ABSTRACT

Climate change has often been discussed at conferences, and in meetings and workshops, but it is not known whether society is aware of its continuing vulnerability to this global phenomenon. Agriculture is one of the sectors most affected by climate change, yet only a number of scientists understand the impact of climate change on agriculture. Not even agricultural extension agents, who are responsible for disseminating agricultural information to farmers and communities, understand it clearly. Within the uncertainty that climate change brings, the success of farming activities depends on the extension officer’s understanding of and effective communication about climate change, since they are the agents of change. This article suggests guidelines for effective communication about climate change by extension agents.

INTRODuCTION

Over the past decades, climate change has emerged as one of the most intensely researched and discussed environmental issues ever around the globe. Many climate change studies and assessments point to more and frequent weather disasters to come, with unprecedented consequences on the global population. However, this information is only known and well understood by a small number of scientists and those that interact with them. While climate change has been discussed broadly in workshops, meetings or at conferences, the question remains whether a significant number of the public is aware of their vulnerability to climate change. So far, it is clear that climate change is likely to have major impacts on farming activities in Namibia, with negative consequences on food security, income generation and livelihoods.

In the light of the above, this article attempts to outline some guidelines for effective communication by extension agents in raising awareness and promoting climate change issues in relation to agricultural activities in Namibia. Therefore, it is vital for the extension agents to have a common understanding of climate change; how to communicate about it and its impact on many sectors of our economy, particularly in agriculture. Only once the agents fully understand climate change and its effects on agricultural activities, will they strive to initiate innovative farming practices which will enhance agricultural productivity and farming income, despite global warming. The scientific

EFFECTIVE COMMuNICATION OF CLIMATE ChANGE BY ExTENSION AGENTS

F.N. MWAZI and J. NDoKoSHo

Africa Adaptation Project Namibia (AAP-NAM), Ministry of Environment and Tourism, Windhoek, [email protected] or [email protected]

evidence leaves little room for doubt that our climate is changing and that agriculture will be affected. Hence communication strategies or awareness programmes around climate change need to be put in place in order to ensure that communities or farmers are kept informed and understand this global issue. This will allow them to adopt better adaptation and mitigation mechanisms to cope with the uncertainties of a changing climate.

COMMuNICATING CLIMATE ChANGE ISSuES

The measurable increase in average global temperatures, termed “global warming” is linked to increases in “greenhouse” gases in the earth’s atmosphere (Justus & Fletcher, 2006). When communicating about global warming, the key issues that need to be understood are climate change and climate variability, as well as two complementary issues, namely adaptation and mitigation. When defining climate change one has to understand the difference between weather and climate first. Weather is the current state of the atmosphere on a day-to-day basis for a given area or region, (IPCC and WMO, 2010). Climate, on the other hand, is the average weather of the area over a long period of time; at least over 30 years (IPCC, 2007 & IPCC and WMO, 2010).

Climate change refers to any change in climate over time, whether due to natural variability or as a result of human activity (this is called anthropogenic climate change) (IPCC, 2007). Also, it refers to a statistically significant variation in either the mean state of the climate or in its variability, persisting for an extended period (typically decades or longer) (IPCC and WMO, 2010). In simplified terms, it refers to any long-term significant change in the average weather that a given area experiences. Most scientists believe that climate change is caused by human activities which include the burning of fossil fuels (coal, oil, and natural gas), driving cars, generating electricity, factories, deforestation, or waste disposal. Historically the wealthy countries have been the biggest contributors to greenhouse gas emissions.

Climate variability refers to variations in the mean state and other statistics (such as standard deviations, the occurrence of extremes, etc.) of the climate on all spatial and temporal scales beyond that of individual weather events (IPCC and WMO, 2010). Variability may be due to natural internal processes within the climate system (internal variability),

Step 1: DM grass yield = 10 000 m² x 3485,3 g = 774,51 kg/ha 45 m²

Step 2: Yield after estimated loss from trampling, insects and termites, is 35 % (774,51 kg x 0,65) = 503,43 kg

Step 3: Estimated utilisation 50 % (503,43 x 0,5) = 251,72 kg DM/haTherefore, 500 ha would yield = 125 860,05 kg DM available

Step 4: A 700 kg mare with foal needs 1,65 LSU (Meissner et al., 1983), 1,65 LSU x 13,5 kg/LSU (3 % x 450 kg) = 22,28 kg/mare and foal/dayTherefore, this 500 ha paddock offers forage for 125 860,05kg22,28 kg/mare/day

Step 5: Attempted grazing period 2 months (60 days)5 650,28 mare-days 60 days

22 mares + foals for 60 days, 22/95 x 100 = 23,2 % utilisation

The camp was underutilised. Thus, the recommendation was to put more mares into the camp, or to prolong the grazing period.

Mare and foal on 500 ha for 1 year:

700 kg x 3 % = 21 kg/day x 365 days = 7 665 kg DM/year

125 860,05 kg available on 500 ha paddock 7 665 kg= 16 mares on 500 ha

Carrying capacity on conventional terms:

500 ha 16 mares

125,72 kg DM/ha 10,95 kg DM/365 days

CONCLuSION

In arid and semi-arid areas, horses eat a wide variety of feeds. Horses graze; eat standing hay or fallen grasses, herbs, shrubs and the bark of trees. Horses have teeth and lips that permit them to graze close to the ground i.e. they are able to pick up preferred fallen grass material and herbs from the ground. Therefore, horses are grazers as well as browsers. The principal forage is Stipagristis uniplumis (climax) and Eragrostis rotifer (sub-climax) palatable, perennial grass species.

REFERENCE

AGANGA, A.A., lETSO, M. & AGANGA, O.A., 2000. A research report in the feeding behaviour of domestic donkeys: Journal of Botswana case study. Applied Animal Behaviour 60:235–239.

FORBES, J.M., 1995. Voluntary Food Intake and Diet Selection in Farm Animals, CAB International, UK. pp. 1–11.

GIESS, W., 1971. A preliminary vegetation map of South West Africa. Dinteria No. 4: 32–41.

KAMUPINGENE, G.T., 2000. Research Project: Comparative diet selection of three grazing sheep breeds at Neudamm farm, University of Namibia, Windhoek, Namibia.

KAVENDJII, G.K., 1999. Diet selection of goats in Namibia during the rainy season. Research Project Report, University of Namibia, Windhoek, Namibia.

MEISSNER, H.H., HOFMEYR, H.S., VAN RENSBURG, W.J.J. & PIENAAR, J.P., 1983. Classification of livestock for realistic prediction of substitution values in terms of a biologically defined large Stock unit. Division Agricultural Information. Pretoria, RSA No. 175: 34.

MINSON, D.J., 1990. Ruminant Production and Forage Nutrients, In: Forage in Ruminant Nutrition, ch.1 Academic Press, INC California. USA: 36.

NARJISSE, H., 1991. Feeding Behavior of Goats on Rangelands. In: Goat Nutrition, Morand-Fehr P. (ed) . Ch. 2 Pudoc. Wageningen, The Netherlands: 13–16.

National Planning Agricultural Census, 1997. Central Statistics Office, Technical Report. No. 66: 26.

National Research Council, 1989. Nutrient Requirement of Horses. Sixth Revised Edition. National Academy Press. Washington, DC., USA.

TAINTON, N.M., 1988. Veld and Pasture Management in South Africa. Interpak Natal Pietermaritzbury, RSA: 53–65.

= 5 650,28 mare-days

= 94,71 (95) mares for 60 days

= 1 LSU/30,45 ha (700 kg)

= 22,98 kg animal biomass/ha


that enhance a person’s or community’s resilience to climate change (reforestation or alternative livelihoods that conserve forest resources, water harvesting; and sustainable agriculture) can also be undertaken.

EFFECTIVE COMMuNICATION OF CLIMATE ChANGE

Communication involves imparting knowledge with the intent of raising awareness and promoting understanding. Therefore, getting the right message across about climate change by the extension agents to various communities including farmers, will make them: a) understand why climate change issues are so important in their daily activities; b) put together adjustment mechanisms to cope with this climate change phenomenon; c) reduce agricultural production losses related to climate change; and in return become less vulnerable to climate change and d) put together clear adaptation strategies aimed at enhancing adaptive capacity, generate income and improve livelihoods. When extension agents plan to convey climate change messages to the communities or farmers, the goal should focus on creating a community that understands climate change and thus being able to make reliable choices. In doing so, effective communication of climate change will be realised.

Figure 1 below shows the information flow from researchers, agents and farmers or communities. It is clear that extension agents play a vital role in disseminating relevant information to the communities, as they are the middle-man in the communication process. They also receive information from the communities and scientists. For example, a farmer faces agricultural problems (e.g. stalk-borer in a maize field) and then informs an extension agent about it, who will then later discuss the problem with a researcher in the area for further investigation and to determine the cause of and solution to the problem. Once a scientist has found the cause of and the solution to the problem, he informs the extension agent about the outcome. Thereafter an agent is bound to disseminate the

findings to the farmers, but only after having converted such information to an understandable simple language for the farmers or communities. The outcomes of the findings by the scientist are rarely implemented directly. The findings need to be transformed into a practical oriented approach by the extension agents. Thus, extension officers are the agents of change in fostering rural agricultural developments. Such a process in certain circumstances may require translation of the outcome in the local languages of a community. Therefore, the sooner extension and other service providers become familiar with climate change, the earlier the integration of climate change into agricultural developmental goals.

RuLES OF COMMuNICATING CLIMATE ChANGE

As climate change is a global problem with wide-ranging impacts, it is essential that climate change messages are communicated successfully to many different groups, including town residents, farmers and communities. The aim of a climate change communications campaign is to change public attitudes and behaviour (Table 1). According to GTZ (2009), specific rules to communicate about climate change include:

• Avoid alarmism – base your statements on soundscientific findings.

• Stress the importance both of interpreting climatechange and managing uncertainty – use possibility ranges (several plausible and reasonable futures: that is the most important lesson for every decision maker to learn).

• Provideabackgroundofbasicclimatechangescienceto help decision makers interpret the information.

• Be transparent and precise (and when discussinguncertainty, make it clear what the major sources of uncertainty are – the emission scenarios rather than the models).

• Beexactabouttimescales(asealevelriseofonemetreby 2100 or by 2030 makes a big difference).

• Get support from experts, as they can answer morecritical questions and thereby increase credibility.

• Beawareofandtransparentabouttheconflictyouarein: on the one hand you might be aware of your own uncertainty and possess inadequate knowledge; on the other hand you want to convince people.

• Try to use neutral language and avoid value-ladenstatements.

kEY COMMuNICATION SkILLS

The key communication skills are: listening skills, verbal or speaking skills and writing skills. Therefore, an exceptional listener and communicator who effectively conveys information verbally and in writing, will always get the right message across and will be well understood by the audience. To achieve the expected outcomes, a communication plan needs to be put in place to address questions such as: what change do you want to bring about using communication (objectives), which individuals or

Figure 1. Dissemination link of information flow to different stakeholders.

or to variations in natural or anthropogenic external forces (external variability).

Adaptation is the adjustment in natural or human systems in response to actual or expected climatic changes or their effects, which moderates harm or exploits beneficial opportunities (IPCC, 2007). Adaptation involves undertaking action to deal with the negative effects of climate change. Therefore, adjustment mechanisms need to be put in place which include - but which are not limited to - improved cropping and farming systems, introducing drought tolerant crop varieties and the use of heat tolerant livestock breeds, the building of raised infrastructure in flood areas, and rainwater harvesting in order to adapt to climate change.

Mitigation refers to an anthropogenic intervention to reduce the anthropogenic forcing of the climate system (IPCC, 2007). Therefore, mitigation is lessening emissions of greenhouse gasses. Measures must be put in place to reduce the impact of global warming and its effects on human health and the environment. Such measures include, but are not limited to, the use of gas-fired power generation; efficient lighting and the use of renewable energy sources; improved cooking stoves, reduction of charcoal use; and appropriate aforestation or re-forestation.

Apart from these issues, the main question a crop farmer or community will ask is: How will climate change affect my crops and livestock? To respond to this question, one needs to know the climate variables that affect agricultural production. In general, one will think of declining rainfall and rising temperatures, though there are other variables which could be mentioned such as wind speed, hours of sunshine, and humidity. In this article the examples of changing temperature and rainfall will be discussed.

In Namibia, maximum temperatures have been increasing over the past 40 years to exceed 35 °C in many places, whereas minimum temperatures (below 5 °C) have become a less frequent occurrence, suggesting an overall warming (MET, 2010). It is also predicted with a high degree of certainty that Namibia will become hotter throughout the forthcoming years (with a predicted increase in temperatures of between 1 °C and 3,5 °C in summer and 1 °C to 4 °C in winter for the period 2046 to 2065). Rainfall is predicted to decrease by 10 % in the northern and southern regions and by 20 % in the central regions by 2050 (MET, 2010). The most consistent changes will be an increase in late summer rainfall over major parts of the country, and a decrease in winter rainfall in the south and the west of the country. But, how will the increase in temperature and shifting rainfall patterns affect maize (Zea mays) production? Our response to this question is provided by means of the example below.

Example: Maize thrives in conditions of 500 mm to 900 mm of rain during the growing season and tempera-tures from 21 ºC to 30 ºC (MAWF, 1997 & Mwazi, 2006). With the summer rainfall coming only late in the season

for the past 40 years, planting dates also needed to shift from early to late planting in the season. Failure to do so is likely to cause maize to suffer shortages of water at the beginning of the growing season. Planting late during the growing season may also result in maize being sub-merged by floods due to extreme precipitation received late during a given season. Hence, understanding planting dates and growing period is crucial and needs to be inves-tigated regularly to keep up with the pace of the changing climate patterns. Should the rainfall be below 500 mm (300 mm for example), it will not be enough for the maize to complete its life cycle to reach maturity in that particular season. This is mainly because its requirement (500 mm and above) is higher than what the land can offer (300 mm). As a result, the land becomes marginally suitable (based on rainfall variable only), perpetuating adverse effects to farmers due to reduced yield driven by climate changes.

Since maximum temperatures for the past 40 years have been exceeding 35 °C and minimum temperatures of 5 °C or below have become uncommon, the agricultural sector is likely to be impacted upon with these changes. Such changes in temperature have been taking place in areas formerly ranked suitable (based on temperature ranking only namely a maximum of 30 °C) for maize production. This means that the area becomes moderately suitable as temperatures exceed 30 °C and moves up to a 35 ºC to 40 °C range. It means maize will be stressed by losing too much water due to high evapotranspiration as a result of high temperatures. All these are examples of how an increase in temperature and shifting rainfall patterns as a result of climate change, may affect maize production and farmers. Therefore communicating about climate change should focus on real world scenarios in order to shed more light on what will happen to crops when a change in temperature or a shift in rainfall takes place. From this perspective, it is not helpful to use highly scientific words which may be difficult for farmers and communities to understand. The extension agents should use simple language and provide understandable scenarios when communicating about climate change, as well as work closely with climate modellers to predict the changes so that farmers can get the best advice.

BARRIERS TO GOOD COMMuNICATION

One of the challenges of public education is that awareness and knowledge do not always translate into action. Simply knowing about the effects of climate change is not enough for some people. They need to understand that climate change affects them directly, and that they can do something about it and they need to be motivated to take action. In order to overcome these barriers, communication geared at public awareness and education should provide people with comprehensive information about a subject so that they can better understand it. Thus, the audience can be encouraged to change specific practices or behaviour, for example, a reduction in harmful practices such as deforestation that leads to flooding and land degradation and water and electricity wastage. An increase in practices


groups do you want to influence (target audiences), what do you want to say (key messages) and who or what are the most effective messengers or champions (community leaders; political leaders or farmers themselves).

TARGET AuDIENCES AND RELATED PERSuASIVE MESSAGES

It is always important to segment the target audiences during communication. Consequently, developing a profile of the audience is needed to answer questions such as: how do they prefer to get information (written, audiovisual, face-to-face, etc.)? What is the age range of your audience? Are they mostly men or women? And how do they make a living?

Once the audience segments have been determined, develop messages to address them. A good message addresses a particular objective and should be specific. It communicates clearly to that particular audience, links to something they care about and should be believable and backed up by facts or evidence. Messages about climate change should convey a sense of urgency and emphasise the benefits of making the changes being advocated. Therefore, the messages should show that these changes will build resilience, sustain livelihoods and reduce vulnerability. At the end, request feedback from the communities or farmers which could assist in improving and enhancing the message for the future.

CONCLuSION

To ensure effective communication about climate change, make sure that you understand the issues and concepts before trying to communicate them to others. Speak in plain language; do not use technical, climate change jargon. Keep your messages clear, accurate and simple. Link climate change with other environmental and social issues that might be familiar to people, so that they can understand how the issues are connected. Show the history of climate change, if any, through visuals such as

videos, maps, satellite images and pictures to emphasise the importance of this global phenomenon. Last but not least, encourage the audience to integrate climate change into their development goals in order to remain focused, or to take climate change into consideration during the implementation stage of their projects or daily activities.

ACkNOWLEDGEMENT

A word of appreciations goes to Dr. Fred Sikabongo (Deputy Director: Environmental Impact Assessment, Ministry of Environment & Tourism) and Mr. Servaas van den Bosch (Freelance Journalist) for editing the article for language and clarifications.

REFERENCE

GTZ., 2009. Climate Change Information for Effective Adaptation, A Practitioner’s Manual. GTZ, Eschborn.

IPCC & WMO., 2010. Intergovernmental Panel on Climate Change (IPCC) and World Meteorological Organization (WMO). http://www.wmo.int/pages/prog/wcp/ccl/faqs.html. Accessed on 16 December 2010.

IPCC., 2007. Summary of Policymakers. In: Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Parry, M.L., Canziani, O.F., Palutikof, J.P., Van der Linden, P.J. & Hanson, C.E., Eds., Cambridge University Press, Cambridge, UK, 7–22.

JUSTUS, J.R. & FLETCHER, S.R., 2006. CRS Report RL33602 for Congress, Global Climate Change: Major Scientific and Policy Issues.

MAWF., 1997. Extension Staff Handbook – Crop Production, Ministry of Agriculture, Water and Forestry, Windhoek, Namibia.

MET., 2010. Climate Change Vulnerability and Adaptation Assessment for Namibia’s Biodiversity and Protected Area System. Directorate of Parks & Wildlife Management, Ministry of Environment and Tourism (MET), Windhoek, Namibia.

MWAZI, F.N., 2006. Spatial analysis of land suitability to support alternative land uses, Excelsior Resettlement Project, Oshikoto region, Namibia. M.Sc. thesis. ITC, Now Faculty of Geo-information Science and Earth Observation, University of Twente, Enschede, The Netherlands. Available at: http://www.itc.nl/library/papers_2006/msc/nrm/nyambe.pdf.

Table 1. Shows a shift in attitude to climate change issues if communication is effective

Present situation Expected outcomes

Communities/farmers lack knowledge on causes of climate change and do not understand what needs to be done to tackle it.

Communities/farmers clearly understand climate change and what is causing it.

Communities/farmers think that climate change will not affect them.

Communities/farmers understand the impact climate change may have on their daily activities.

Communities/farmers do not include climate change as an important matter when making decisions.

Communities/farmers include climate change when making their decisions and embrace the positive changes that result.

Communities/farmers think climate change is a depressing and negative issue.

Communities/farmers feel empowered and positive about tackling climate change.

ACCELERATING LANDSCAPE INCISION AND ThE DOWNWARD SPIRALLING RAIN uSE EFFICIENCY OF NAMIBIAN RANGELANDS

HuGH PRINGlE1, IBo ZIMMERMANN2 and KEN TINlEY3

1Ecosystem Management Understanding (EMU) Project TM, P.O. Box 8522, Alice Springs, NT 0871, Australia [email protected]

2School of Natural Resources and Tourism, Polytechnic of Namibia, Private Bag 13388, Windhoek, [email protected]

346A Hope St., White Gum Valley, Western Australia 6162, Australia [email protected]

ABSTRACT

In response to rapidly degrading rangelands there is an urgent need to precondition the land for extremes of weather conditions, which will both mitigate locally against climate change and offer better rain use efficiency and better primary and secondary productivity. Mainstream principles of rangeland management tend to overlook levels of ecological organisation above the “veld type” as well as dehydration caused by landscape incision and the impacts of infrastructure that initiate and accelerate many erosion processes. Dongas erode soil, but far more ecologically significant is their dehydrating impact on affected surrounding landscapes and their sub-catchments. Prior to incision by dongas, the affected landscapes were usually the most productive in their wider catchment context, staying green longer than adjacent run-off and run-through landscape elements and responding rapidly to local rainfalls. Examples are presented from Farm Krumhuk, approximately 20 km south of Windhoek.

INTRODuCTION

Rangelands globally face many challenges, not least of which are increasing costs and declining real prices on produce. Climate change is another challenge which appears certain and predictable at a global and continental level, but little real progress has been made to adapt to likely changes for most regions.

Whatever the impacts of climate change, we argue that preconditioning the land for extreme weather conditions will not only mitigate locally against climate change, but also offer better rain-use efficiency and better primary and secondary productivity. In other words, this work should be done, irrespective of climate change scenarios. Such preconditioning aims to get rainfall into the ground as locally as possible and minimise run-off. For both extremes of weather, the conventional strategy is to have high ground cover at a fine scale to capture raindrops into the soil locally, minimise run-off and protect against soil erosion. We argue that it is equally important to restore base levels, where water is held back in the landscape, at a drainage ecosystem level. The purpose of these inter-dependent strategies are to withstand major rainfall events and make best use of small falls of rain in prolonged dry periods. (Base levels are

the lowest part of a drainage system beyond which erosion cannot occur. When base levels are incised, for instance when a sandy sill of a wetland is breached by animal paths, then a new phase of erosion is initiated upslope).

While standards of grazing management can certainly be improved and systems such as Savory’s Holistic Management (Savory and Butterfield, 1999) and Riaan Dames’ Fodder Flow Grazing Management Strategy (Dames, 2009) offer great opportunities, as well as the older Acocks model (Acocks, 1964), they are fundamentally captured in the traditional, local focus (perhaps obsession) with “veld type” dynamics. What all of these models share is a focus on biologically strategic rest (mainly for grasses), an important but inadequate approach to be truly “holistic”. That lack of holism is in the sense of levels of ecological organisation above the “veld type” (or “ecological site” in the USA), which is partly our focus in this article. Of particular concern is the increasing incision of catchments and their landscape sequences from valley floor to upland headwaters. These processes do not start and stop in a veld type, but transcend them and thus require a higher level of appraisal than conventional in situ veld management.

Two key additional issues need to be considered along with conventional veld management and like all else, interact as factors determining habitat quality for both livestock and wildlife locally. These are i) landscape incision and dehydration and ii) the impacts of infrastructure that initiate and accelerate many of the former degradation processes. We see these three strands of physical land management as prerequisites to ecosystem (“ecologically”) sustainable land management, whatever the land use, culture or location (Pringle and Tinley, 2003; Pringle et al., 2003; Shamathe et al., 2009).

LANDSCAPE INCISION, INITIATED BELOW AND ACCELERATED ABOVE

Gully or “donga” erosion is often blamed on poor ground cover in the hinterland catchment above. This is not generally correct; increased run-off certainly causes sheeting and extension of erosion cells (Pickup, 1985), but dongas generally need a “nick point”; a cut in the landscape to get started. This may be a track graded below the land surface by just a few centimetres, or a cattle or wildlife


pathway. In Mozambique’s Gorongoza National Park, it was the hippo that caused the spread of dongas into the Urema Lake wetlands system (Tinley, 1977).

Easily overlooked, but fundamental to understanding incision-based dehydration of landscapes, are the following key sequences of processes:

• Waterusuallyflowsdownslopewithgravity.• Erosionheadcutsprogressupslopeatallscales,micro

to macro, cutting into adjacent intact soil surfaces.• Detachedsoil isentrainedby theflowandalsomoves

downslope.

There are exceptions, but they will confuse the key message here. What is critical for land managers and their technical supporters is that any donga has come from and been initiated downslope. Increased run off in headwaters acts as an accelerator pedal, not an initiator of donga development. Thus any restoration needs to address:

• Stabilisationoftheaggressiveheadwardcutting“head”;• calmingthehinterland;and• restoringasbestaspossibletheinitial“nickpoint”.

Most importantly, once a head is initiated, it needs to be identified and halted before it splits, as heads will always retreat back into areas from which they receive strong flow. Thus dongas may incise upslope as a simple head initially, but if any tributary flows are encountered on this course, the head will split and cut back into them too. Thus dongas draw exceedingly more water into their channel (“canal”) system at the expense of the surrounding “robbed” landscape. Dongas erode soil, but far more ecologically significant is their dehydrating impact on affected surrounding landscapes and their sub-catchments (Pringle and Tinley, 2003; Tinley, 2001; Tinley, 1982).

Dongas siphon landscapes dry and lead to lower and more intermittent soil moisture balance regimes

Dongas are usually discussed in terms of soil erosion, which is correct and obvious. What is not often appreciated is the wider impact dongas have on the distribution of soil moisture in time and space. Dongas drain away surface water that has not infiltrated the soil, exporting this critical resource from the landscape in which it fell as rain. Dongas effectively “pull the plug out of the bath” leading to shorter periods of positive soil moisture balance. This in turn degrades the rain use efficiency of affected landscapes and their sub-catchments with subsequent rainfalls. Dehydration becomes a positively reinforced downward spiral; the more water is lost, the faster and wider the incision process and so on. Once erosion patches enlarge and link up reaching a critical dimension of increased runoff volume, this results in ongoing self-generated expansion of eroded terrain that no amount of rest will heal without intervention at key sites.

It is imperative that dongas be seen as desiccators of affected landscape complexes and not just as a soil erosion problem. Prior to incision by dongas, the affected landscapes were usually the most productive in their wider catchment context, staying green longer than adjacent run-off and run-through landscape elements and responding rapidly to local rainfalls.

Donga heads not only lower the local base level and thus drain surface water passively; as flows fall over the rim of the gully head, they accelerate. This acceleration acts as a strong physical “pulling” force within the surface water above; the heads of dongas really do “suck” landscapes dry. Thus as donga heads cut back and split upslope; they are effectively becoming increasingly efficient at sucking suites of landscapes dry.

“Spikier” soil moisture balances favour different plant species and vegetation

The export of surface water is particularly profound and problematic in seasonally inundated landscape elements such as upland valley floors, valley-side floodouts, floodplains and swamps. The incision of the ecological and commercially critical landscapes reduces the time they are waterlogged and thus species sensitive to water logging are no longer drowned. Many of these key landscapes were once open hydromorphic (water loving) grasslands and sedgelands (according to farmers of the Auas Oanob Conservancy and various historical accounts and photographs). However, they are now bush encroached because bush seedlings in these previously seasonally inundated landscapes are no longer drowned effectively before they can grow to a size where they can survive water logging during parts of the rainy season, and also because of the opposite dry season conditions of compact cement-like clay or gilgai-cracked desiccated soils that are inimical to scrub seedling survival (Tinley, 1982; 1977). This component of the wider Namibian bush encroachment story is largely unknown and overlooked in the most recent major review (De Klerk, 2004). It is emphasised that this cause of bush encroachment is specific to seasonally inundated landscapes within wider catchment systems in which other factors are usually predominant (Joubert, Zimmermann & Graz, 2008; De Klerk, 2004).

Not only can bush species survive when seasonally inundated surfaces are “unplugged”, hydromorphic grasses become more stressed and lose their habitat-specific competitive advantage over more xeromorphic species in a positive feedback loop resulting in successively greater landscape dehydration and increasing xeromorphic vegetation (grass as well as bush species). Overgrazing might accelerate the process through reduced local infiltration as well as consequent faster landscape incision due to greater run-off.

These major changes also have cascading ecological effects, including favouring browsers (e.g. goats, impala and kudu)

over ecotonal or grassland favouring grazers (e.g. hippo, roan and sable antelope) or mixed feeders (e.g. cattle and eland) (Tinley, 1977). As a general rule, the species favoured by landscape dehydration are more generalist and may be of lesser commercial and/or conservation value.

The major changes are also potentially catastrophic for landholders, both commercial and traditional, as they try to adapt to an increasingly inefficient rain use landscape that is more drought and flood prone. In a climate change context, these changes are likely to make adaptation ever more challenging at a local level. This issue is too important to remain overlooked or treated as a side issue. We must stop the decline and start rehydrating the rangelands (Shamathe, Zimmermann & Pringle, 2008a) for a complex variety of interlinked reasons.

INFRASTRuCTuRE CAuSES MANY OF ThESE INCISION PROBLEMS

As previously discussed, dongas almost always start at some local landscape incision (“nick point”) receiving concentrated flow from above. The most common causes of dongas across southern African and Australian rangelands include:

• Oldwagon tracks cut into the landscape, particularlywhere they traverse narrow valley floor gaps in mountainous terrain or follow valley floors (Cooke and Reeves, 1976).

• Modernaccess tracks thatarealignedtosomeextent(not necessarily directly) up and downslope and have been cut (perhaps only a few centimetres) into the landscape in construction or maintenance (e.g. with a grader blade) or been used when the landscape is still wet following rain.

• Fencelinesrunningtosomeextentupanddownslopethat have concentrated animal traffic along them and thus are prone to incisions of animal pathways.

• Watering points in areas of concentrated flow whichtypically have animal pathways radiating out from them (Pringle and Landsberg, 2004); the worst of these are where they are located at the “key line”, where flow should switch from tributary to distributary as flows emerge from uplands into flatter country (Pringle et al., 2006). Once incised by a donga, the water flow no longer spreads out at the keyline, but instead gushes down the donga.

• Main road culverts, which are usually set below thesurrounding landscape level to facilitate the rapid flow of water below and not over the road surface.

• Excavations with steep slopes (e.g. “borrow” pits forroad building).

• Anyothersourceoflandscapeincisionlikelytoreceivesubstantial run-on.

A keyline (see fourth point above) is the change in slope from uplands to flatter ground. In uplands drainage channels usually come together (the tributary phase), straighten, accelerate and are effective at eroding disturbed soil. Below the keyline channels should start to split up (the distributary phase), slow down and deposition should dominate over erosion. In severely damaged drainage ecosystems, the distributary phase is replaced by channels that keep the water in them like downpipes on a roof, drying out previously hydrated landscapes of whole drainage systems (Pringle & Tinley, 2003).

In the past, wells and even bores were generally sunk where underground water supplies were shallowest and had the best quality and supply. These were often (even usually) in positions in the landscape least suited to intense traffic of animals and humans. However, with reticulation technology (e.g. plastic pipes), it is no longer necessary for watering points to remain in these highly unsuitable locations. Even dams can be fenced and the water reticulated to hydrologically “quieter” and more stable landscape positions. This cannot be done at once, but through a triage process, a long term programme can be affordable and effective. Fence lines causing problems can also be prioritised and gradually relocated or removed.

With machinery, “track creeks” are also repairable by flattening out windrows to allow natural cross flows. Where the “creeks” have been incised too deep and there is not enough material in windrows to flatten them out effectively, strategically placed small banks can be installed to encourage harvested water to return to its natural course.

Floodways at landscape level are far more ecologically appropriate than culverts, but engineers (and the public) often want non-stop traffic flow along main roads. The heads cutting back upslope from road culverts should be stabilised to allow water to flow over them with no further erosion (if the culvert cannot be removed or replaced at natural landscape level). This can be done by flattening down the head’s “cliff” and then armouring it with stones (preferably limestone which sticks together when wet) or geotextiles. This should be a standard practice in main roads management, but is very rarely observed. Flows from culverts should be checked with a short solid bank as close to the road as is allowable to “take the hit” and then spread the water more serenely back out into the landscape.


CASE STuDIES ON FARM kRuMhuk

Landscape incision and dehydration, and their initiation by infrastructure, are captured in photographs from Farm Krumhuk, in the Auas-Oanob Conservancy nearby Windhoek. A sketch of the broader study area appears in Figure 1. Satellite images from Google Earth show

the broader area where two case studies are located (Figure 2). Figure 3 focuses on the case study area around Vlagte Dam. The coordinates for the dam appear on Google Earth as 22º 46' 03" S, 17º 05' 05" E. Photographs of this case study appear in Figures 4 to 8. The other case study of water capture by a track is illustrated by photographs appearing in Figures 9 to 13.

Figure 1. Tracing from Google Earth satellite images of main landscape features in the northern part of Farm Krumhuk, south of Windhoek.

Figure 2. In this example on Farm Krumhuk, a creek flowing out of the mountains (outside and to the left of the Google image) floods out when it meets flat terrain and then splits its sheet flow because of subtle slopes in the plain below. Some goes north-east to a dam near the homestead and some flows down to the vlagte (seasonally inundated wetland) blocked by Vlagte Dam. We know the floodout fan splits because of the distinct donga heads eating back towards the creek’s floodout point from different directions. The “track thief” may well have pulled water towards it and an animal path would have been enough to start the donga head upslope (left). Wherever you see sections of track broader than others on Google Earth (which is free), it is likely they are cut down and cutting back laterally (sideways) to drain landscapes. Should either of the donga heads in the case study of the track creek (Figures 9 to 13) reach the floodout point, there will be a downpipe from the mountains to the nearest major channel and that creek’s contribution to soil moisture balance of the plain will be totally lost (as has already happened with other creeks).

Figure 3. The donga head cutting back towards the Vlagte Dam on Farm Krumhuk has cut back from a more deeply cut down channel that turns sharply west. This deeper channel also ends upslope in gully heads indicating that it is not a natural drainage feature either. The concentrated overflow from Vlagte Dam has spilled over into the incised channel and set off the current donga head cutting back to the dam’s spillway. The old, unchanneled water course is parallel to the donga, as indicated by the line of trees.


Figure 4. A major donga is cutting up towards the Vlagte Dam overflow alongside a cut off natural watercourse on Farm Krumhuk. The donga is draining a broad seasonally inundated plain. Younger and smaller Acacia karoo trees higher in the channel indicate this is a donga, not a natural creek.

Figure 5. A cattle path helps the head of the donga actively pull water towards it. Water flowing over the head’s “waterfall” accelerates, thus exerting a sucking force on the water behind. This is how donga heads actively drain landscapes upslope of them. Thus dongas are a major ecological issue, not just a soil conservation problem.

Figure 6. Water actively sucked into the donga quickly finds its way out of the system, thus reducing the landscape’s ability to turn rainfall into grass and leading to soils drying out more quickly.

Figure 7. Once the storms were over, it was clear that the donga head had eaten its way back into the healthy ground and well up to five metres in places. Well after the rain stopped, the waterfalls continued sucking the plain dry, with the central, lowest waterfall stopping last, leaving no more surface water upslope.

Figure 8. Young Acacia karoo plants establish above the rim of the donga head where they are no longer drowned at seedling size. Unlike their counterparts in the most recently formed part of the donga below them, these bushes are doomed, as the donga head will, in the next big storm, erode the soil that now supports them. A larger plant can be seen in the foreground. This plant dates back to an earlier season than the two smaller ones in the background. It may survive until the next major head retreat. Then the two smaller ones will in turn be “hanging on the edge”.

Figure 9. This is an example of how poorly located and constructed access tracks can drain landscapes as effectively as any true donga. In this case a creek from the mountains in the background has flooded out in a fan when it has hit flat ground, depositing rich alluvium in a highly productive system. The track is cut into the land and has started micro-terraces (broad, small water fall heads) cutting back into the floodout fan. The track starves areas of the fan below it.


Figure 10. The waterfall face of these micro-terraces may be less than an inch high, but is very effective in draining areas upslope as the water accelerates over the edge.

Figure 11. As the micro-terrace faces cut up-slope they often become shallower and get broken up by stable mounds making a pattern of stable islands and stripping rills out of a once water-soaked fan system.

Figure 12. The track is capturing almost all of the sheet flow from the left side and the lower, right hand side of the photograph is now effectively dehydrated by the “track creek”.

Figure 13. A head is cutting back up the track towards the area of the previous photographs. This head is creating a new, deeper base level below the natural landscape and thus helping the existing track “sucking” the landscape above dry and when it gets back to the area above, it will probably set off a new series of micro-terrace faces to accelerate the drying of this fan and virtually cut off flow to the area of the fan downslope.

ADDRESSING LANDSCAPE INCISION TO REhYDRATE RANGELANDS

Soil conservationists have been addressing donga systems for several decades and the techniques they have developed generally work well, but are now often cost prohibitive in terms of the price of concrete to build various check structures or weirs. Alternative, less costly approaches using machinery to build banks have also been successful (but again, not always) (Bastin, 1991). Bush packing has also been successful, such as seen/demonstrated/observed in another part of the Auas-Oanob Conservancy (Kauatjirue et al., 2010; Shamathe et al., 2008b). At Londolozi in South Africa, trees were used to fill dongas; these have vanished and have returned to being productive grasslands (Ken Tinley, personal communication). There are many other approaches to rehydrating incised and dehydrating rangelands and local creativity and innovation is often required to make use of what is at hand.

What is not common, is a systems view to be implemented so that a carefully planned and sequenced set of interventions can be implemented to address the causes as well as the symptoms, and thus turn the system back to better functioning (Whisenant, 1999). Too much attention has been paid to techniques and not enough to systems in restoration planning. Through the Ecosystem Management Understanding (EMU) ProjectTM, we have developed a sequence of key steps in planning repair of an incised and dehydrating rangeland system (Tinley and Pringle, 2006). Key to this approach is to map key landscape patterns and processes (particularly hydrological ones), incisions and their heads and implicated infrastructure, before planning interventions that are sequenced and support each other. What is primarily important, is identifying what has to be achieved at each intervention point, not the technique used. This allows for creativity and innovation to overcome resource limitations.

CONCLuSION

Downward spiralling rain use efficiency is being driven by landscape incision processes globally in rangelands with effective overland flow, and Namibia is no exception. In fact, this issue is possibly the most critical overlooked environmental problem in the country and it affects many aspects of Namibia’s human ecosystems from wildlife conservation and traditional land use to commercial farming. Without being recognised and addressed strategically, the future is one of increasingly long droughts and damaging floods, as well as far greater vulnerability to climate change. Most distinctive and drought buffering landscapes and plants and animals that are specifically adapted to them, will be lost and replaced by bare earth and/or bush encroachment. This issue is too important to be ignored, as the cost of not addressing it will be extremely high for wildlife and human inhabitants of Namibia.

ACkNOWLEGEMENTS

We thank the members of the Auas-Oanob Conservancy for their keen participation in exploring and addressing the dehydration of their landscape.

REFERENCES

ACOCKS, J.P.H., 1964. Karoo vegetation in relation to the development of deserts. In: Davis, D.H.S. (Ed). Ecological studies of southern Africa, pp.100–112. The Hague: Dr. Junk Publishers.

BASTIN, G., 1991. “Rangeland reclamation on Artatinga Station, central Australia.” Australian Journal of Soil and Water Conservation, 4: 18–25.

COOKE, R.U. & REEVES, R.W., (eds.) 1976. Arroyos and Environmental Change in the American South-West. Oxford: Clarendon Press.

DAMES, C.J., 2009. The controlled fodder flow grazing management strategy and “grass fed beef production”: A sustainable, proven, environmentally friendly, extensive animal production model


for the semi-arid and arid environments of southern Africa. Proceedings of the 13th Namibian Rangeland Forum, p. 19. Windhoek, 27 to 29 October 2009.

DE KLERK, J.N., 2004. Bush Encroachment in Namibia. Windhoek: Ministry of Environment and Tourism, Government of Namibia.

JOUBERT, D.F., ZIMMERMAN, I. & GRAZ, P., 2008. A decision support system for bush encroachment. Windhoek: Polytechnic of Namibia.

KAUATJIRUE, J., SHAMATHE, K., PRINGLE, H.J.R. & ZIMMERMANN, I., 2010. Restoration of a gully system in the Highland Savanna of Namibia. In: Schmiedel, U. & Jürgens, N., (Eds). Biodiversity in Southern Africa. Volume 2: Patterns and processes at regional scale, pp. 255–259. Göttingen & Windhoek: Klaus Hess Publishers.

PICKUP, G., 1985. “The erosion cell – a geomorphic approach to landscape classification in range assessment.” Australian Rangeland Journal, 7: 114–121.

PRINGLE, H.J.R. & TINLEY, K.L., 2003. “Are we overlooking critical geomorphic determinants of landscape change in Australian rangelands?” Ecological Management and Restoration, 4(3): 180–186.

PRINGLE, H.J.R., TINLEY, K.L., BRANDIS, T., HOPKINS, A.J.M., LEWIS, M. & TAYLOR, L., 2003. “The Gascoyne-Murchison Strategy: A people-centred approach to conservation in arid Australia.” In Allsop, N., Palmer, A.R., Milton, S.J., Kirkman, K.P., Kerley, G.I.H., Hurt C.R. & Brown C.J., (eds.), Rangelands in the New millenium. Proceedings of the Seventh International Rangelands Congress, pp. 213–223. Durban, South Africa: Document Transformation Technologies.

PRINGLE, H.J.R. & LANDSBERG, J., 2004. Predicting the distribution of livestock grazing pressure in rangelands. Austral Ecology, 29(1): 31–39.

PRINGLE, H.J.R., WATSON, I.W. & TINLEY, K.L., 2006. Landscape improvement, or ongoing degradation: Reconciling apparent contradictions from the arid rangelands of Western Australia. landscape Ecology, 21: 1267–1279.

SAVORY, A. & BUTTERFIELD, J., 1999. Holistic Management. USA: Island Press.

SHAMATHE, K., PRINGLE, H.J.R. & ZIMMERMAN, I., 2008a. “Restoring rain use efficiency to an incised upland valley system in Namibia using filters and Ecosystem Management Understanding (EMU) principles.” Multifunctional grasslands in a changing world, Proceedings of the xxIst International Grasslands Congress/ VIIIth International Rangelands Congress, p. 783. Guangzhou, China: Guandong People’s Publishing House.

SHAMATE, K., ZIMMERNANN, I., PRINGLE, H.J.R., RUSCH, E.A. & RUSCH, I.B., 2008b. Restoration of a gully system in a fertile valley. Spotlight on Agriculture. Windhoek, Namibia: Ministry of Agriculture, Water and Forestry.

SHAMATHE, K., KAUTJIRURE, J., PRINGLE, H.J.R. & ZIMMERMAN, I., 2009. Restoring efficiency of rain in a gully system of the Auas-Oanob Conservancy. Proceedings of the 13th Congress of the Agricultural Society of Namibia (AGRISSON), pp. 202–208. Oshakati, Namibia: AGRISSON.

TINLEY, K.L., 1977. Framework of the Gorongoza Ecosystem, Mozambique. D.Sc. thesis (Ecology and Wildlife Management), University of Pretoria.

TINLEY, K., 2001. Scrub encroachment of productive grasslands: soil moisture balance. Proceedings of the Northern Australia Beef Industry Conference, pp. 11–16. Kununurra, Western Australia.

TINLEY, K.L., 1982. The influence of soil moisture balance on ecosystem patterns in southern Africa. In Huntley, B.J. & Walker, B.H., (eds.), Ecological Studies, Volume 42: Ecology of Tropical Savannas, pp. 175–192. New York: Springer-Verlag.

TINLEY, K. & PRINGLE, H., 2006. Key Principles and Steps in Catchment Repair in Arid Rangelands. Range Management Newsletter, 7(2): 4–5.

WHISENANT, S.G., 1999. Repairing wildlands: a process-oriented, landscape-scale approach. London: Cambridge University Press.

Date post:	30-Apr-2020
Category:	Documents
Upload:	others
View:	4 times
Download:	0 times

AGRICOLA - nbri.org.na · sonal and 80 % or more of the total annual rainfall of 373 ± 158,6 mm...

Documents