The effects of visual apparency on bias in butterfly recording and monitoring

A basic assumption in conservation is that comparable data are available for species to facilitate risk assessment of extinction. However, the capacity for carrying out cross-species comparisons for abundances and distributions among butterflies depends on the absence of bias in recording and monitoring or the existence of appropriate techniques for removing bias. Here, we investigate potential bias in cross-species comparisons for the apparency of butterfly adults (wing colour, size and behaviour) in three pivotal UK datasets: dates of discovery, Butterfly Monitoring Scheme (BMS) transect measures of abundance and Butterflies for the New Millennium (BNM) national recording scheme. Bias is found in all three datasets. Discovery date is affected by wing colour and size, BNM recording by adult behaviour and wing colour and BMS monitoring by adult behaviour. The nature and degree of bias differs between uncorrected cross-species comparisons and those with the application of phylogenetic contrasts. The findings urge caution in using these datasets for cross-species analysis without improvements and standardisation of recording and monitoring and the development of techniques to adjust for biases, in particular the use of suitable comparative techniques. The latter requires the construction of a molecular phylogeny for butterflies. (c) 2005 Elsevier Ltd. All rights reserved.