How reliable are Malaise traps for biomonitoring? – A bivariate species abundance model evaluation using alpine Chironomidae (Diptera)

Abstract

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.

Choice of DNA extraction method affects DNA metabarcoding of unsorted invertebrate bulk samples

“…3d). Correlation close to zero indicates that the communities are independent in structure, while a negative correlation show that dominant species in one community tend to be rare in the other and vice versa (Diserud et al 2013). An obvious reason for the low consistency in the lentic subsamples is the lower relative and total number of target specimens compared to the lotic kick samples and subsequent inferior target specimen homogenisation of the lentic samples.…”

“…Partitioned pair-wise betadiversity values (β sim and β sne , Baselga 2010) were calculated using the R-package betapart (Baselga et al 2018, R Core Team 2016 to see how much of the variation in species richness was due to nestedness (β sne , subsamples with less species include a subset of the species in subsamples with more species) and turnover (β sim , change in species composition). To study repeatability of the DNA extraction methods, similarity in invertebrate community composition was analysed by fitting the bivariate Poisson-lognormal species abundance distribution to pairs of subsamples (Engen et al 2002, Engen et al 2011, Diserud et al 2013, using the R-package poilog Engen 2015, R Core Team 2016). Species abundances from a community were assumed to follow a lognormal distribution and the sampling process to be described by Poisson sampling.…”

Stable baselines of temporal turnover underlie high beta diversity in tropical arthropod communities

“…Our results also reinforce conclusions based on studies that examined only selected arthropod taxa. For examples, short‐term Malaise trap studies were found to reliably monitor dipteran communities along Norwegian streams (Diserud, Stur, & Aagaard, ) and to reveal diversity patterns in Mesoamerican ichneumonid faunas that corresponded with well‐documented trends (Veijalainen et al., ). However, a major increase in sampling effort will be needed to comprehensively document biodiversity.…”

Investigating suburban micromoth diversity using DNA barcoding of malaise trap samples

“…Hosking 1979;Noyes 1989). Moreover, Malaise traps that sample the same communities in the same time period produce only insignificant differences in the monitored species composition (Diserud et al 2013). Thus, Malaise traps are well-suited for comparative studies of the flying arthropod fauna at a large geographical scale.…”