Community inferences from single-species data

Background/Question/Methods Data are often collected for a single species within an ecological community, so quantitative tools for drawing inferences about the unobserved portions of the community from single-species data are valuable. I will present and examine a method for estimating community dimension (the number of strongly interacting species or groups) from time-series data on a single species.

Results/Conclusions The dynamics of one species can be strongly affected by environmental stochasticity acting not only on itself, but also on other species with which it interacts. By fully accounting for the effects of stochasticity on populations embedded in a community, our approach gives better estimates of community dimension than commonly used methods. Using a combination of time series data and simulations, we have found that failing to properly account for stochasticity when attempting to relate population dynamics to attributes of the community can lead to conclusions that are significantly flawed.