Herrera-Lopera, J. M., E. Pineda, M. Solé & C. A. Cultid-Medina

Sample size effects, Hill numbers, and trophic niches in anurans. pp. 193-202 plus Supplementary document 1 and Supplementary document 2.

Abstract. Niche breadth and niche overlap measurements have been widely used together to assess resource dynamics in biological communities. However, the estimations of these measurements are susceptible to bias due to sample size. Alternatives such as standardization of niche breadth have been implemented to try and mitigate this bias, but standardization does not solve the underlying problem. In anuran trophic ecology, sampling is usually incomplete, which constrains comparisons. A solution to this problem could lie in implementing an analysis using Hill’s series (also called Hill numbers) for comparisons under the same sample coverage; however, this approach has not yet been tested in anuran trophic ecology research. The objectives of this work were to evaluate the bias resulting from sample size for the most frequently used measurements of niche breadth and niche overlap in anuran trophic ecology vis-à-vis overlap measurements derived from Hill’s series and to provide a least biased protocol for anuran trophic ecology evaluations. We built data matrices with similar features to those expected for anuran assemblages and quantified the bias of each measurement for different sample sizes. We found that Hill’s series measurements were less biased and more informative than traditional ones. We provide an analysis guideline based on Hill’s series that facilitates direct comparisons between predator species regarding their consumed and shared prey communities.

Key words. Amphibia, trophic ecology, bias estimation, data simulation, measuring diversity, effective number of prey animals, niche analysis protocol, trophic interactions.