DURATION OF BEHAVIOR PATTERNS DURING OVIPOSITION BY THE BRUCHID BEETLE, CALLOSOBRUCHUS-MACULATUS

The behaviour of insects between making contact with a potential ovipo sition site and the act of oviposition itself determines the condition s in which their progeny begin development and, in many cases, whether a crop or commodity is damaged. In this paper the behaviour of female cowpea seed beetles, Callosobruchus maculatus (Fabricius) (Coleoptera : Bruchidae), during this period has been examined. The duration of di screte behaviour patterns within this period and the transitional freq uencies between them have been used to compare the behaviour of beetle s on two different hosts. In a 30-min period, beetles laid more eggs o n mung than adzuki beans and rejected mung as an oviposition site much less frequently. A log-linear contingency table analysis of the trans itional frequencies indicated a highly significant interaction between successive behaviours and between the second of the two behaviours an d the host seed, indicating that the sequence of behaviours is partial ly determined by the host seed. Grooming, departure from the seed and loops in behavioural sequences occur more often on adzuki than mung an d are indicative of a less acceptable host. Discriminant analysis show ed that the total time spent exhibiting different behaviour patterns w as significantly different on the two hosts. Mann-Whitney tests reveal ed small but significant differences in the mean durations of only thr ee behaviour patterns on the two seed types. Measurement of the durati on of each behavioural pattern confirms what can be deduced about host acceptability by observing the simple occurrence of different behavio ural patterns and the determination of the frequency of transitions be tween them. Differences in acceptability were predominantly reflected in the frequency of certain behaviour patterns and of particular trans itions from one pattern to another, not in the mean duration of each e xhibition of a defined behaviour pattern. It is argued that the utilit y of measuring the duration of behavioural elements could only be demo nstrated because behavioural patterns or categories were precisely def ined.