An Ecological Classification of British Butterflies: Ecological Attributes and Biotope Occupancy
- Cite this article as:
- Shreeve, T., Dennis, R., Roy, D. et al. Journal of Insect Conservation (2001) 5: 145. doi:10.1023/A:1017556113534
We classify British butterflies using 136 non-biotope associated binary state ecological attributes describing all stages of butterfly life-cycles. Using cluster analysis we identify two groups of woodland species, a group occurring in tall open grassland, another group associated with short sward herb-rich grassland, and a ruderal group. Principal Component and Factor analyses (4 factor solution) are used to identify ecological attributes that determine species groupings. No single attribute or attribute type is responsible for the groupings, which are also insensitive to hostplant type. We use presence/absence data from Butterfly Monitoring Scheme transects in southern Britain to test our classification. On the basis of adult occurrences, similarities within two of the four groups identified from PCA are greater than between groups. Exclusivity between species pairs is also more frequent between groups than within groups. Species' ranges, distributions, biotope range, dispersal ability and recent decline in abundances differ between groups identified by their factor loadings. Ruderal species have large ranges, abundances, extensive mobilities and show little recent decline. The group associated with short sward grassland have the lowest mobilities, and the smallest distributions within their geographic ranges. True woodland species have the smallest biotope range, and the species associated with open areas have the second smallest decline in their distributions. Our ecological classification identifies characteristics of species that determine their habitat requirements and could serve to predict the response of species groups to environmental change on the basis of their ecological attributes. Our method may be of use in identifying the relative importance of ecological attributes of less-well studied taxa and be applicable in less well known geographic regions.