Work

Analysis of the relation between ungulate species richness in East Africa and climate and remotely sensed productivity indices

Year: 2003
Type: book
Abstract: The aim of the study was to evaluate whether climatic or remotely sensed indices are a better predictor of ungulate species richness.Our hypothesis is that the remotely sensed index the normalized difference vegetation index (NDVI) integrates the influence of climatic and environmental factors, it should be a better predictor of species richness than climatic variables.The study was conducted in East Africa, which has high ungulate species diversity.Out of 95 ungulate species found in Africa, 55 are found within this sub-region.The vegetation is highly heterogeneous, with great variability in productivity that is influenced by the spatial and seasonal distribution of climatic variables, including precipitation and temperature.Species richness was regressed against each environmental variable, using linear, polynomial, logarithmic and exponential models.A test of the equality of two population regression coefficients was used to compare the models.The results show that the strongest correlates of species richness were rainfall and NDVI, with 69% and 55% of the variance explained.A unimodal pattern was exhibited for ungulate species richness with both rainfall and NDVI.The combination of rainfall and potential evapotranspiration (PET) best predicted the ungulate species richness.The climatic model was more predictive than NDVI (r 2 of 0.76 vs. 0.55, t = 2.22, P = 0.028) in areas with a rainfall range between 200 and 1300 mm, and (r 2 of 0.80 vs. 0.67, t = 1.32,P = 0.10) for the drier part of the rainfall gradient (<1000 mm) the difference was not significant.We concluded that NDVI was not a better predictor of species richness than integrated climatic indices.Several factors related to the reflectance of vegetation cover; soil-vegetation spectral mix and soil type /moisture limit the performance of NDVI in predicting both the primary productivity and species richness.Improvements in spectral, spatial and temporal resolutions, including information on surface radiation, ecosystem process variables (leaf area, leaf duration and primary productivity) and land cover characteristics should significantly improve the potential use of remotely sensed data in biodiversity studies.
Language: English
Open Access status: green
OpenAlex ID: W2128918658
FWCI: 0.667
Cites: 72
Cited by: 6
Related to: 9