Verifying Small Sample Bootstrap Estimates of Species Interactions
Qualms about bootstrap confidence intervals have led to the development of new techniques with the added complexities of bias estimation and acceleration constants. These complications make the current theory appear almost Ptolemaic to the uninformed. Furthermore, bootstrap estimates based on small samples must be interpreted with caution. It seems prudent and advisable to look at other analyses to, at least partially, verify bootstrap results in small samples. Here, we analyze using the bootstrap and traditional nonparametric and parametric tests the results of a small scale perturbation experiment in ecological research to assess the importance of the interspecific interactions of competition, prédation, and mutualism.
Unable to display preview. Download preview PDF.
- Culver, D.C., Fong, D.W. and Jernigan, R.W.(1990) Assessing species interactions in cave communities, preliminary report.Google Scholar