Diversity of butterflies and day-flying moths in urban habitat fragments, south-western Australia

Abstract

This study adapted and developed methods of assessing and modelling biodiversity of butterflies and day-flying moths in habitat fragments, and determined those factors affecting their presence, abundance and species richness in a sample of 46 isolated urban remnants in south-west Western Australia. The specific objectives were to: (i) assess the effectiveness of transect–based sampling to quantify the species richness of habitat fragments; (ii) examine patterns of species richness in habitat fragments and quantify the detectability of each species recorded; (iii) review and rationalize the methods used to fit species–area–habitat models; and (iv) model species incidence, abundance and total richness of butterflies in urban habitat fragments and determine implications and priorities for their conservation.These objectives were achieved and the principal findings of the research are: (i) The transect method provides an accurate assessment of butterfly species richness in isolates provided that the level of sampling (proportion of area surveyed) is adequate, that sufficient surveys are conducted during the flight season to ensure high levels of detectability, and that surveys are conducted at appropriate times and during suitable weather conditions. Although randomly placed transects are preferable, logistic constraints often dictate the use of existing pathways, roadsides or management tracks – which requires the use of longer transects but is more practical in urban remnants.(i) The transect method provides an accurate assessment of butterfly species richness in isolates provided that the level of sampling (proportion of area surveyed) is adequate, that sufficient surveys are conducted during the flight season to ensure high levels of detectability, and that surveys are conducted at appropriate times and during suitable weather conditions. Although randomly placed transects are preferable, logistic constraints often dictate the use of existing pathways, roadsides or management tracks – which requires the use of longer transects but is more practical in urban remnants.(iii) Almost a century of fitting species–area curves has failed to produce agreement on which function is the best model of the relationship. Many of the proposed functions are identical, special cases of others or have arisen from transcription errors. Empirical comparison of these functions requires methods suited to the distribution of species number such as the generalized linear model, method of maximum likelihood and the information-theoretic approach, and proper attention to covariates and their interactions.(iv) Site area and vegetation condition were the dominant determinants of the presence, abundance and total species richness of resident butterflies and day-active moths in 46 urban habitat fragments in south-west Western Australia. Larger sites with more high quality (undisturbed) vegetation favoured 16 of 20 native species and only one benefited from disturbance. A further nine species not sufficiently widespread or abundant to enable individual analysis were collectively more prevalent in larger sites. Resource quality and quantity dominated the patterns of site occupancy, and increased site connectivity did not favour any species – results consistent with habitat resources, not metapopulation effects, determining current distribution patterns. As expected, the presence of non-resident species was unaffected by site area. The total number of resident species at each site reflected the collective responses of the individual species: increasing with area and declining with vegetation disturbance. The effects of area and vegetation quality were not simply additive: disturbance had a far greater impact on small remnants. This interaction is inconsistent with the area per se hypothesis: in the absence of disturbance there was no evidence of a species–area effect.This study is the first comprehensive, quantitative assessment of the distribution and ecology of butterflies and day-flying moths in Australian urban habitat fragments and provides a baseline against which future changes in species distributions may be measured. The results have important implications for the conservation of butterflies and day-flying moths in the region. Maintenance of vegetation quality is of paramount importance and is vital in smaller remnants. Large remnants, being less susceptible to local extinctions, will be essential for the persistence of many species. Many functions have been proposed to model the species–area relationship but empirical comparisons have been hindered by methodological problems – this study conducted a re-examination of the relationship and presents an appropriate framework to compare functions. This study is also one of few to demonstrate and quantify the importance of interactions in explaining patterns of species richness and should stimulate future research into the importance of these effects.