International Workshop
“Long-Term Studies
on Biodiversity”
November, 23rd - 24th 2010
The CNPq-PELD/Brazil LTER experience
Francisco Barbosa & Diego Pujoni
ICB – UFMG
The MCT-CNPq/PELD initiative
• Originally proposed as a subprogram of the IntegratedEcology Program (PIE) by CNPq, 1996
• PIE: approved by the Forum of the Graduate Programs in Ecology and presented to CNPq/CAPES in May 5th 1996
subprograms: building capacity in Ecology
LTER (PELD) program
• Goals
define/implement policies for development of ecology in Brazil
develop research and information network
reinforce internacional cooperation
reinforce capacity building and institutional development
N
Forests
Cerrado (Savannas)
Caatinga (Dry Forest)
Pantanal (Wetlands)
Sandplains and Mangroves
Source: IBGE - Brasil
1
2
3
4
569
7 8
Brasil LTER research sites
1- Tropical Rain Forest - Manaus
2- South Pantanal Wetland System
3- Central Plateau Cerrado
4- Atlantic Forest and middle Rio Doce lake System
5- Coastal Sandplain System of Rio de Janeiro
6- High Paraná River Floodplain System
7- Taim Wetland System
8- Patos Lagoon Estuary and Adjacent Coast
9- Araucária Forest System
Research Themes
• Patterns and control of primary productivity
• Dynamic (fluxes) of nutrients
• Conservation of biological diversity
• Population dynamics and organization of
communities and ecosystems
• Patterns and frequency of natural
disturbances and human impacts
ILTER
network
46
countries
North America: 3 Europe: 25 Africa: 5
Central/South America: 5 South-East Asia: 8
The middle Rio Doce basin LTER program
working hypothesis
• i) the present biodiversity experiences a
process of loss and/or alteration at an
unknown level although perceptible
• ii) the remaining areas contribute for the
maintenance of significant portion of this
diversity
• iii) the present biodiversity still encounters
conditions of persistence at a long run
• Iv) exotic species (bio-invasions)!
Main Threats
Deforestation
Mining
Siderurgy
Pollution
(untreated sewage)
Exotic species
The Middle Rio Doce Lake System
State Park of Rio Doce
Doce River
The middle Rio Doce Climate
Studied Lakes Carioca
Águas Claras
Amarela
Dom Helvécio
Palmeirinha
Jacaré
Gambazinho
Temperatura (ºC) - DH
2002 2003 2004 2005 2006 2007 2008 2009
J FMAMJ J ASOND J FMAMJ J ASOND J FMAMJ J ASOND J FMAMJ J A SOND J FMAMJ J ASOND J FMAMJ J ASOND J FMAMJ J ASOND J FMAM
Pro
fund
ida
de (m
)
0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
32
22
24
26
28
30
32
Temperatura no lago Dom Helvécio, durante o período de janeiro de 2002 a
maio de 2009
Oxigênio Dissolvido (mgl-1
) - DH
2002 2003 2004 2005 2006 2007 2008 2009
J F M A M J J A S ON D J F MA M J J A S O N D J F M A M J J A S O N D J F M A M J J A S ON D J F M A M J J A S ON D J F M A M J J A S O N D J F M A M J J A S O N D J F M A M
Pro
fun
did
ad
e (m
)
0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
32
0
5
10
15
Concentrações de oxigênio dissolvido no lago Dom Helvécio durante o período de
janeiro de 2002 a maio de 2009.
Water temperature and disolved
oxygen patterns
New insigths (e.g.) resistence eggs
(an Atlas is in preparation)
Some possibilities with - PELD
Identification of processes and patterns
The fish fauna of Lake CariocaNATIVES INTRODUCED
jeju Traíra
Tuvira Piabanha
Bagre Cumbaca
Piau Lambarí
Tucunaré Piranha
Peixe pedra Tamboatá
Apaiarí
Local extinctions at Lake Carioca - 2010
Fonte
Sunaga & Verani,
1991 Godinho et al., 1994 Latini et al., 2004 Vasconcellos et al., 2005 Presente estudo
Ano de amostragem 1983 1985 1987 1992 2002 2002-2003 2006-2010
Espécies Nativas
1 Hoplias gr. malabaricus -
2 Brycon sp. -
3 Leporinus cf. steindachneri -
4 Rhamdia quelen -
5
Hoplerythrinus
unitaeniatus -
6 Gymnotus sp. -
7 Trachelyopterus striatulus -
8 Astyanax aff. bimaculatus -
9 Steindachnerina elegans -
10 Astyanax fasciatus -
11 Moenkhausia doceana -
12 Oligosarcus solitarius -
13 Geophagus brasiliensis -
14 Cichlasoma sp. -
15 Loricariidae gen. sp. -
16 Astyanax sp. -
Espécies Introduzidas
17 Cichla kelberi
18 Pygocentrus nattereri
19 Hoplosternum littorale
20 Clarias gariepinus
21 Astronotus sp.
22 Callichthys callichthys
TOTAL GERAL 11 11 7 10 - 14 15
TOTAL NATIVAS 11 10 6 8 - 8 9
TOTAL
INTRODUZIDAS 0 1 1 2 4 6 6
Legend: shadow = presence; open areas = absence; - = no information.
• Principal Components: Relationship between variables
Statistical Methods
• Regressions
• Linear Discriminant Analysis: Best discriminant
variables
• Species Accumulation Curves (Ugland et al.,2003; Kindt & Coe, 2005):
Average and SD of richness for k samples. Standardize
different sampling frequencies
• Mantel Correlation: Spatial statistics
• Time Series Analysis: Seazonal Auto-Regressive
Integrated Moving Average Models (SARIMA)
Modeling the richness and diversity in time
Dynamic Regression.
Challenges• How to handle missing values?
• How to handle non-normal distributions?
Estimation by:
-Multiple imputation (Rubin, 1987; Van Buuren & Groothuis-Oudshoorn, 2009)
-Maximum likelihood (Rubin, 1976)
-Non parametric bootstrap confidence intervals (Efron & Tibshirani, 1993)
-Other robust estimators (i.e. Lognormal) (Pennington, 1983; Myers &
Pepin, 1990; Fletcher, 2005)
Diurnal sampling (sporadic)
Low costs, but Insufficient data to represent plankton diversity
Likely to render underestimates due to the concentration of
samplings under specific environmental conditions of a
particular chosen day.
Weekly, bi-weekly, and monthly sampling
More realistic analysis of plankton diversity since include wider
range of environmental conditions and responses of this
community to these changes.
Costs of sampling are much higher, time and staff
consuming.
Remarks
Seasonal sampling (dry/rainy periods)
Allow for a realistic plankton diversity since cover the two most
tropical representative environmental conditions -
stratification and mixing periods.
Allow for detection of the extreme conditions and
comparatively at lower costs.
Annuall sampling
Time between samplings too long to assess diversity
To allow for comparisons samplings would represent the
same phase of the hydrological cycles (subjectivity?)
Remarks
Some results:
Richness paterns
-3 -2 -1 0 1 2 3
-3-2
-10
12
3
PC1 (22%)
PC
2 (
11%
)
Prof
pH
Cond
Turb
OD
Temp
Redox
Alca
Cloro
PTOT
PO4
NTOTNH4
NO3
NO2
SiO4
RiqFito
RiqZoo
Carioca
Dom Helvécio
GambazinhoJacaré
PCA of environmental variables sampling monthly at four
lakes in the middle Rio Doce basin between 2001 and 2006
-4 -2 0 2 4
-4-2
02
4
LDA1
LD
A2 Cond
TurbTempAlca
Cloro
NO2
SiO4
Carioca
Dom Helvécio
GambazinhoJacaré
Linear discriminant analysis of environmental variables
01
02
03
04
05
0
Phytoplankton Sample Richness
Time
Phyto
pla
nto
n R
ichness
Carioca
Dom Helvécio
Gambazinho
Jacaré
Jan/01 Jul/01 Jan/02 Jul/02 Jan/03 Jul/03 Jan/04 Jul/04 Jan/05 Jul/05 Jan/06 Jul/06 Jan/07 Jul/07 Jan/08 Jul/08
05
10
15
20
25
30
Zooplankton Sample Richness
Time
Zoopla
nkto
n R
ichness
Carioca
Dom Helvécio
Gambazinho
Jacaré
Jan/01 Jul/01 Jan/02 Jul/02 Jan/03 Jul/03 Jan/04 Jul/04 Jan/05 Jul/05 Jan/06 Jul/06 Jan/07 Jul/07 Jan/08 Jul/08
Species accumulation curves
0 20 40 60 80
02
04
06
08
01
00
12
0
Dom Helvécio
0 20 40 60 80
02
04
06
08
01
00
12
0
Ric
hn
ess
Águas Claras
0 20 40 60 80
02
04
06
08
01
00
12
0
Dom Helvécio
0 20 40 60 80
02
04
06
08
01
00
12
0
Ric
hn
ess
Águas Claras
0 20 40 60 80
05
01
00
15
02
00
25
0
Samples
Ric
hn
ess
Phytoplankton-18 lakes
A space/temporal comparison of phytoplankton
richness at the middle Rio Doce basin
0 20 40 60 80
02
04
06
08
01
00
12
0
Samples
Ric
hn
ess
Zooplankton-18 lakes
A space/temporal comparison of zooplankton
richness at the middle Rio Doce basin
Comparison of Carlson’s trophic state index
between PERD and surrounding lakes
In Out
43
44
45
46
Ca
rlso
n T
SI M
ed
ian
Oligotrophic
Mesotrophic
10000 bootstrap
samples
Comparison of phytoplankton and zooplankton
richness between PERD and surrounding lakes
Final remarks
• Plankton community richness is not negatively
affected by land-use changes in the lakes
surroundings (Intermediate disturbance
hypothesis – Conan, 1983 and Plankton paradox,
Hutchinson, 1961)
• Investment in new statistical approaches in order
to understand local and regional patterns of
specie´s richness and diversity distribution
• Investment in mathematical modelling in order to
test new hypothesis
Conclusions
LTER data provide the possibility of comparisons
between distinct time scales
These findings are particularly important for tropical
developing countries considering the possibility of
reducing time/costs necessary to guaranteeing
samplings, identification, counting, and interpretation of
the recorded results.
Short/intermediate time scales studies must consider an
intensification of sampling!