Transcript An analysis of taxa congruence and the question of spatial scale

a PhD research project
Biodiversity: An Analysis of Taxa Congruence and
the Question of Spatial Scale; and how this can
contribute to Strategic Conservation Planning in
Uganda
Herbert Tushabe BSc., MSc. (MUK)
The University of Copenhagen, Denmark
& Makerere University, Kampala, Uganda
Supervisor:
Prof. Jon Fjeldså
(Principal Supervisor)
Zoological Museum,
University of Copenhagen
Location in Africa
Location in Africa
Location in Africa
Protected Areas
Objectives of the Study
Main Objective

to develop and test methods for reserve selection and
zonation
based
on
taxa
congruence
and
complementarity analyses.
Specific objectives
To test whether selection of one or two taxa for
conservation will effectively conserve other taxa - using
data from Uganda’s Important Bird Areas (IBAs);
Uganda’s Important Bird Areas
Criteria used, developed by BirdLife International:
Sites with globally threatened species
Sites with restricted-range species
Sites with biome-restricted assemblages
Sites with congregations of species, e.g. waterbirds.
Congregations are considered for both global and
sub-regional populations.
IBAs range from <1 to
4,000 km2
Specific Objectives Cont’d
To test the usefulness of data collected by
various levels of sampling effort in analysis of
congruence;
To find the minimum set of Uganda’s IBAs one
would need to effectively protect other taxa; and
To assess the usefulness of congruence and
complementarity analysis in designing a system
for protected areas; or for effective biodiversity
conservation in existing ones.
Taxa Congruence: A Summary
Conservation costs would be minimal, and efforts more effective, if the theory
of congruence was true.
This theory proposes that:
conserving one or groups of several taxa in an ecosystem effectively
conserves the rest and their species of conservation concern. In
summary, areas that are species-rich for one or more taxa are rich for
others; and rare species are nested within species-rich areas.
Arguments (for and against):
•Identification of priority areas in light of huge gaps in data; and fragmented
information
•Costs of inventories (resources, time), therefore use surrogates
•Criteria used in defining ‘hotspots’: absolute spp. richness, weighting, habitat
loss
•Spatial scaling, taxa preferences, sampling effort differences
•Local, national, continental, global, biological commonalities
Study Approach
The study is involving two levels of analysis:
a
practical test of congruence, complementarity and priority
analysis using existing and field-collected data

the
This involves analysis of the importance of Uganda’s 30 important bird areas (IBAs) that were
identified using internationally developed criteria.
use of larger scale modelled data.
 This involves use of prediction models already developed by the National Biodiversity Data Bank
in Uganda, based on species distributions and environmental parameters associated with their
habitats.
 The Zoological Museum at the University of Copenhagen is currently employing the WORLDMAP
software that uses interactive modelling to identify conservation priority areas, and some of
Uganda’s IBAs have already been identified.
 Results obtained by both models will be compared, as they are done at various spatial scales to
determine levels of efficiency. Results obtained by modelling will be compared for efficiency with
those obtained by use of extensive field work carried out in the IBAs, more especially as the field
work will point out ‘negative’ records that may have been predicted.
1. Congruence and Complementarity
Analyses

Data have been collected for the following taxa in 30 IBAs:






Vascular plants
Dragonflies
Butterflies
Birds
Analysis will evaluate the extent to which taxa overlap using
various measures such as species richness; rarity and
weighting (by developing a scoring system).
examine how areas complement each other in conservation
of biodiversity – using the selected taxa as surrogates.
Determine the minimum set required to conserve
biodiversity in IBAs.
Sample Results

Three sites are considered here:



Bwindi Impenetrable NP (an IBA, forest ecosystem, 331km2)
Lutembe Bay (an IBA, wetland ecosystem, 8km2)
Sango Bay (outside IBA, savanna ecosystem, 6km2)
Area and species totals:
Bwindi
Lutembe
Sango Bay
Area (sq. km.) Plants
Butterflies Birds
331
324
310
348
8
34
89
223
6
38
17
135
Species Accumulation
Lutembe: Species Accumulation and Rarefaction Curves for
Dragonflies
Lutembe: Species Accumulation and Rarefaction Curves for
Plants
160
16
140
14
12
Number of Spp
100
80
60
10
8
6
40
4
20
2
Sampling Hours
Sampling Points
Rarefaction: Finite est.
Rarefaction: Infinite est.
Accumulation
Accumulation
Rarefaction: Finite est.
Rarefaction: Infinite est.
27
25
23
21
19
17
15
13
11
9
7
5
3
235
222
209
196
183
170
157
144
131
118
105
92
79
66
53
40
27
14
1
0
0
1
Number of Spp
120
Accumulation Cont’d
Lutembe: Species Accumulation and Randomised
Curves for Birds
Species Accumulation and Rarefaction Curves for
Butterflies
80
180
70
160
140
60
Number of Spp
40
30
100
80
60
20
40
10
20
Rarefaction: Infinite est.
Accumulation
Observed
Rarefaction: Finite est.
Rarefaction: Infinite est.
27
25
23
21
19
17
15
13
11
TSC Hours
Sampling Hours
Rarefaction: Finite est.
9
7
5
3
29
27
25
23
21
19
17
15
13
11
9
7
5
3
1
0
0
1
Number of Spp
120
50
Results Cont’d
Correlation coefficients (Spearman’s)
Plants
Plants
Butterflies
Birds
Butterflies
Birds
1
0.968900011
1
0.906587633 0.9828221
1
High Correlation Coefficients
Correlation coefficients after correcting for area
Plants
Plants
Butterflies
Birds
Butterflies
1
-0.098816773
1
0.896968382 -0.528566
Birds
1
Here birds are good predictors for plants but poor for butterflies
2. Congruence using species prediction modelling
Rainfall
Vegetation
Other Parameters that were considered:
Human Population Density
Ecological Zones
Land Use/Land Cover
Altitude
Prediction models examples
Blue-spotted Wood Dove
Northern Wheatear
Results of this model have been used to produce a bird atlas for Uganda that is soon to be published:
CARSWELL, M., POMEROY, D., REYNOLDS, J. and TUSHABE, H. (in press). The Bird Atlas of Uganda. British Ornithologists’
Union/ British Ornithologists’ Club.
Prediction Modelling and Congruence
Analyses will be carried out to determine the
extent of congruence of predicted species, and
to determine whether the rare or other species
of conservation concern (such as Red Datalisted species) are captured within the IBAs
and other protected areas.
 Also, in comparison with larger-scale modelled
data, determine the extent to which congruence
analyses are affected by spatial scale.

Analysis Tools
Analysis Tools for Congruence:
A computer programme, EstimateS (Colwell, 1994-99), will be used. This calculates the following
estimators:
Chao 1 (estimates true number of species in an assemblage based on number of rare spp in a sample
Chao2 estimates the distribution of species among samples, using presence/absence data
ACE (Abundance-based Coverage Estimator) developed by Chao & Lee (1992, 1994) estimates
species richness based on abundance data (10 or fewer individuals in a sample)
ICE (Incidence-based Coverage Estimator) developed by Chao & Lee (1992, 1994) estimates
species richness based on incidence data (species in 10 or fewer sampling units)
Other estimators calculated include:
Jackknife 1; Jackknife 2; Bootstrap; Michaelis-Menten; as well as Alpha, Shannon and Simpson
diversity indices
These various estimators/indices have been tested by Colwell and Coddington (1994). In their analyses,
the Chao2 and Jackknife 2 yielded the best results
Analysis Tools for Complementarity:
EstimateS calculates the Chao Estimator of Shared Species between sites, the Jaccard Index of
Similarity as well as the Morisita-Horn Index. These can be used to evaluate the complementarity of the
IBAs for species conservation.
Expected Results


Overall:

to help understand the extent of overlap of taxa in areas considered to be important for the
conservation of species for one taxon.

to assess the extent to which conservationists can rely on results of the survey of one or few
taxa that would act as surrogates for others, thereby saving resources and time in bd
assessments for conservation planning.

show how smaller networks of reserves based on ideas of complementality can be more
efficient in the conservation of biodiversity than larger areas that are difficult to manage.
Application to Conservation:

recommend conservation measures in areas selected as critical for biodiversity

scientific methods will be used for zonation of existing PAs to identify areas where
conservation efforts can be intensified using hotspots identified

identification of most serious gaps based on complementarity analysis with pre-selection of
areas which are already well protected