Ecological and Behavioural Relationships of Some Native Cockroaches (Dictyoptera: Blattidae).

Abstract

Emphasis in the debate on community organisation has shifted from competition to coexistence in attempt to understand the processes involved. The extent of overlap and nature of interactions among related species are of key interest in this regard. This project was undertaken to obtain a functional view, from ecological and behavioural studies, of habitat use by three endemic forest-dwelling cockroach species (Celatoblatta undulivitta, C.vulgaris & C.subcorticaria), which often congregate in local areas of moderately dense population. A secondary objective was to contribute to the largely unexplored field of invertebrate dynamics in indigenous New Zealand forest, using a common and generalist insect model. The resource categories represented by seasonal occurrence (time), dispersion (space) and food selection were examined in a mixed Kunzia ericoides-Nothofagus solandri forest at a 750m2 site where the species coexist. Phenological data were obtained by live-sampling ~5000 cockroaches over a two-year period, using unbaited shelters arranged on the ground. Information on food sources was obtained from crop contents, and from nocturnal observations. Diel activity and the agonistic content of social interactions were examined in captivity to indicate the behavioural organisation of activities. Data were obtained with an activity recorder, and by scan-sampling of behaviour under various regimens of species grouping, density, and seasonal time. Time-lapse photography was used to examine the dispersions involved. Results show well-defined and synchronous cycles of seasonal abundance with summer peaks indicating a dispersal phase. Diel regimes of locomotory activity and social behaviour reconcile nocturnal activities, on vegetation and in the litter, with localized and limited availability of daytime refuge sites among bark and dry debris. Differences between species in the use of shelter sites by day and in their nocturnal dispersion, especially when densities were high, are interpreted as evidence of habitat partitioning within this context. Significant levels of separation in frequencies of co-occurrence, combined with social segregation in captivity, also suggest active dissociation of the two more abundant species. Aggressive interspecies discrimination by captive males, however, was apparently the result of misdirected sexual bebaviour. Variation in crop contents and behavioural characteristics are considered to indicate divergence in slightly different microhabitats. The limited availability and temporary nature of refuge sites, in conjunction with seasonal population movements, and the rapid colonisation of new sites by transient individuals, suggest a dynamic pattern of habitat use, with alternating periods of aggregation and dispersion, irrespective of species. These changes occur locally on an annual scale, and probably over a wider area in the longer term as shelter sites are destroyed or subject to successional change in the Plant community. comparisons with published distribution records and habitat types for each species suggest that the co-occurrence of species is correlated with elements in the vegetation, while the activities of at least one of the species may depend on which other species are present.