05 January 2006

Bias in butterfly monitoring

I read an interesting paper in the journal Biological Conservation regarding bias involved in detecting various species of butterflies in standardized monitoring schemes. This bias, of course, can lead to misleading comparisons of populations of different species. The study, which took place in Great Britain, found several biases.
  • The larger and more apparent butterfly species are, and the longer flight periods and wider distributions they have, the earlier is their first discovery date.
  • Abundance measures on the national project Butterflies for the New Millennium are higher for larger butterflies and lower for butterflies that are active on higher vegetation substrates (versus, say, ones that fly closer to the ground). Distribution on this project is positively correlated with visual apparency (ease of detection).
  • The Butterfly Monitoring Scheme uses transect data to track changes in annual abundances. Abundance increases with dorsal basking postures (open versus closed wings) and use of lower vegetation substrates.
These biases don't really affect the ability to assess changes in a single species, but make a difference if comparing populations of different species, and importantly points out that we need to understand the limitations of various survey schemes and their data sets. Some of these same types of bias are no doubt applicable to odonata, birds, and other taxa as well.

Dennis, R.L.H., T.G. Shreeve, N.J.B. Isaac, D.B. Roy, P.B. Hardy, R. Fox and J. Asher. 2006. The effects of visual apparency on bias in butterfly recording and monitoring. Biological Conservation 128:486-492.