loading page

Differentiating diverse diets: a novel approach to analyzing cafeteria feeding trials
  • +1
  • Michael Cramer,
  • Amy Le Blanc,
  • David Flagel,
  • Samantha Driscoll
Michael Cramer
University of Notre Dame
Author Profile
Amy Le Blanc
University of Notre Dame
Author Profile
David Flagel
University of Notre Dame
Author Profile
Samantha Driscoll
Syracuse University
Author Profile

Abstract

Foraging decisions are essential for survival, thus investigation of food preferences is a major goal in behavioral ecology. Cafeteria trials, in which multiple options are displayed to foragers simultaneously, can quantify these preferences. However, statistical analyses of these data are problematic. The aim of this research is to determine the efficacy of using Shannon Diversity (Hs) and Evenness (J) to quantify diet preferences and apply this approach to field data collected in a model system. Shannon Diversity (Hs) provides a single value to represent a forager’s diet, based on proportions of each food type offered. Shannon Evenness (J’) assesses the relative quantity of each food type consumed, and is an intuitive way to address foraging preferences. Null models inspected how values of Hs and J’ change with diet breadth. This approach was tested for a cafeteria trial of two common chipmunks [Eastern (Tamias striatus) and Least (Neotamias minimus)] with five different seed types [Red (Acer rubrum) and Sugar (A. saccharum) Maple, Balsam Fir (Abies balsamea), Pin Cherry (Prunus pensylvanica), and Beaked Hazelnut (Corylus cornuta)]. Hypothesis tests determined if observed diets were statistically different from those predicted based on equal preferences. Values of Hs increased with diet breadth; although mean J’ was unaffected by diet breadth, variability increased. Chipmunks displayed significant seed preferences: Tamias preferred Sugar Maple whereas Neotamias preferred Red Maple. Moreover, each species avoided the seed type preferred by the other chipmunk. Null models indicate that Hs accurately estimates real changes in diet breadth. The fact that J’ did not significantly change with increasing diet breadth also supports the utility of this approach to determine non-random diet preferences. Use of diversity indices to quantify feeding choices can be extended to other contexts (e.g., comparison of suites of foragers, and altering feeding context by changing amounts of each resource offered).