Null models for everyone: A two-step approach to teaching null model analysis of biological community structure

Since being introduced by Connor and Simberloff in response to Diamond’s assembly rules, null model analysis has been a controversial tool in community ecology. Despite being commonly used in the primary literature, null model analysis has not featured prominently in general textbooks. Complexity of approaches along with difficulty in interpreting results may reduce classroom application of this technique. Yet, readily available software makes this set of tools accessible to everyone from practitioners to educators. This exercise describes a hands-on approach that can be cheaply implemented with students as a stand-alone exercise or as a prelude to use of full-blown simulations in follow up lab sessions. Along with a detailed hands-on approach, we have provided a brief introduction to a computer-based approach. This paper includes a link to 294 data sets extracted from Patterson’s Nestedness software and reformatted for use in EcoSim. These exercises demonstrate how to compare the observed number of checkerboard patterns in real community data to a frequency distribution of checkerboard patterns in randomized communities. The hands-on simulation uses fake data as a starting point and dozens of randomizations to produce a histogram. Computer-based simulations generate thousands of simulations to rapidly analyze real data sets. The results provide a springboard for discussion of the underpinnings of inferential science as applied to ecological communities.