Camera traps offer promising new approach for fine-scale spatial and temporal analyses
Environmental stochasticity is a hallmark of climate change, with cumulative effects on various environmental patterns such as rainfall and wildfires. Fine-scale analyses of species distributions has become increasingly beneficial in understanding the dynamics of this ever-changing relationship between organisms and the environment.
The Normalized Difference Vegetation Index (NDVI)—a common benchmark used to quantify vegetative greenness—has been successfully used to measure how climate change affects large-scale productivity, using data from 8x8 km2 patches. However, generating NDVI maps requires costly multispectral sensors and produces data far too coarse to conduct fine-scale spatial analyses.
On a global scale, climate change is expected to reduce vegetative quality. The Serengeti National Park (SNP) is one of the last remaining naturally-grazed ecosystems on earth, providing excellent insight into the biotic and abiotic dynamics of a true grazing ecosystem. Local and global-scale changes in climate parameters are contributing to the increasing environmental stochasticity observed within the park, with direct effects on the high diversity of herbivore species present in the Serengeti ecosystem.
Forage data was sourced from a 202-camera grid occupying 11252 km of the SNP in Tanzania. At each site present in the grid each month for the years 2011-2014, images were sampled at approximately two-week intervals. Two ways in which the quality and productivity of available vegetation can be assessed using camera trap data were via greenness—or color—and height, respectively. A one-way ANOVA showed that the true difference in mean proportion of green pixel values between each categorical greenness level was significant (p < 0.001). Recent burns were also found to have significant impacts on vegetation characteristics (p < 0.001).
By using both algorithmic and observational approaches to quantify forage attributes, we were able to demonstrate the efficacy of camera trap data as an accessible alternative to traditional NDVI indices for conducting fine-scale ecological analyses. These findings will allow for the implementation of informed policy choices to improve land management practices in the Serengeti National Park and beyond. Moreover, the results show that an standardized observational approach can be as effective as a quantitative approach with respect to assessment of fine-scale ecological analysis.