Agonistic behavior, occurring during intraspecific conflicts, refers to the spectrum of conflict-related behavior including not only aggressive behavior, but also the elements indicative of escape, defence, submission and other behavioral elements. Using the model of agonistic behavior of singly-housed male mice (outbred strain ICR) on dyadic interactions with non-aggressive group-housed partners in neutral cages, we have learned that isolates might be divided into three categories: a) aggressive isolates exhibiting at least one attack towards the opponent, b) timid mice exhibiting a lot of defensive-escape behavior but no attack, c) sociable mice exhibiting a majority of sociable behavioral elements, ambivalent locomotor activities and no defensive-escape or aggressive behavior.

For collection of behavioral data from videotaped 4-min agonistic interactions we are using the system The Observer 3.1 (Noldus Information Technology b.v., The Netherlands) which enables us to use a computer keyboard for recording, coding and analysis of not only frequencies but also time parameters of events observed from the videotape. Five time parameters can be evaluated: latency (amount of time elapsed from the start of the observation to a specific event), total duration (which can also be expressed as a percentage of the total observed time), mean duration, minimum duration (smallest duration of an element bout) and maximum duration (largest duration of an element bout). Recommendation is to transfer the summary data gathered with The Observer into Microsoft Excel for statistical analyses. While doing so we were facing to the problem of allocation of experimental animals into three individual categories, and furthermore, of evaluation of our ethopharmacological experiments in which we need to compare behavioral changes elicited by several different treatments: e.g. vehicle as a control + usually at least 2 or 3 drug treatments. The projects of experiments are randomized; thus, each animal passes all types of treatment during the experiment. The differences between the occurrence of acts in control and experimental interactions are evaluated by Wilcoxon matched-pairs signed-ranks test (pertinently by Mann-Whitney U test in the case of experiment with the repeated drug treatment requiring two independent singly-housed mouse samples).

For the above-mentioned reasons an add-on program was developed enabling statistical analysis of pharmacologically induced behavioral changes in three differential categories of mice classified according to their behavior exhibited in the control agonistic interactions. It works under the Excel 97 and was prepared using macros in Visual Basic for Applications to make the program user-friendly. The operator just has to insert data informing about the sequences of experimental interaction types belonging to individual singly-housed mice during the whole experiment. The program counts the means and medians for each 'Observer parameter', and processes statistical evaluation for the levels of significance p<0.05 and p<0.01. From Excel the final data can be transferred to graphical programs or stored in a database. The present presentation demonstrates utilization of the described method for analysis of pro-aggressive effects of new anticonvulsants gabapentin and lamotrigine.

This work was supported by the Czech Ministry of Education Project: CEZ:J07/98:141100001

© 2000 Noldus Information Technology b.v.