The Matrix revisited: How to analyze the structure of behavior

L. Hemerik¹, T. Bukovinszky², R. Gols², J.C. van Lenteren² and L.E.M. Vet^2,3

¹Biometris, Mathematical and Statistical Methods, Wageningen University, Wageningen, The Netherlands
²Laboratory of Entomology, Department of Plant Sciences, Wageningen University, Wageningen, The Netherlands
³Department of Multitrophic interactions, Netherlands Institute of Ecology, Heteren, The Netherlands

Several methods are available for the analysis of first order behavioral transition matrices in biological data. These methods analyze different aspects of transition matrices, but a comprehensive framework for their use is lacking. We analyzed parasitoid foraging behavior in different environments.

The resulting complex data set was analysed using the following stepwise procedure. With a test we detected the abrupt changes in the behavioral records of all tested individuals. This served as a criterion for splitting the behavioral records into two parts. For both parts Mantel’s test was used to detect differences between first order transition matrices, whereas the iterative proportional fitting method was used to find behavioral flows that positively deviate from random transitions. In addition, we detected hidden repetitive sequences with the package Theme™ (PatternVision Ltd, Iceland).

We discuss the results from the example in a biological context and the comprehensive use of the different methods. We however, want to stress that an analysis in which the timing of behaviors is considered should always accompany an analysis like this one.

Paper presented at Measuring Behavior 2005 , 5^th International Conference on Methods and Techniques in Behavioral Research, 30 August - 2 September 2005, Wageningen, The Netherlands.

© 2005 Noldus Information Technology bv