How reliable are Malaise traps for biomonitoring? – A bivariate species abundance model evaluation using alpine Chironomidae (Diptera)
Article first published online: 11 DEC 2012
© 2012 The Royal Entomological Society
Insect Conservation and Diversity
Volume 6, Issue 5, pages 561–571, September 2013
How to Cite
Diserud, O. H., Stur, E., Aagaard, K. (2013), How reliable are Malaise traps for biomonitoring? – A bivariate species abundance model evaluation using alpine Chironomidae (Diptera). Insect Conservation and Diversity, 6: 561–571. doi: 10.1111/icad.12012
- Issue published online: 2 SEP 2013
- Article first published online: 11 DEC 2012
- Manuscript Accepted: 24 OCT 2012
- Norwegian Research Council. Grant Numbers: 182272/S30, 185109
- bivariate Poisson-lognormal;
- community structure;
- non-biting midges;
- species diversity
- In this study, the potential of Malaise traps to collect representative portions of an insect community was investigated. To do so, the complete catch (nearly 22 000 specimens) of male Chironomidae (Diptera) from five parallel Malaise traps along an alpine stream was identified and assigned to 108 different species. The traps were run for 4 weeks in June and July, 2008.
- The similarity in community composition between parallel samples, that is, from different traps the same week, was evaluated by fitting a bivariate Poisson-lognormal species abundance model. The estimated correlation in this bivariate distribution was used as a measure of similarity since this approach is utilising all the available species abundance information and accounts for the sampling process.
- Estimated similarities showed non-significant differences in chironomid community structure between parallel samples. The five Malaise traps sampled equally representative portions of the Chironomidae community present at the site, so the traps were found to be very reliable in the monitoring of Chironomidae community structure.
- Application of the bivariate correlation as a similarity measure offers advantages over traditional measures because it takes account of the complete species abundance distributions. This approach provides an approximately unbiased estimate of similarity despite varying sample sizes and detection/non-detection of species that are present, but rare.