The role of the electrofishing equipment type and the operator in assessing fish assemblages in a non-wadeable lowland river

The wider the spatial and temporal scales of a fish survey are the higher the chance that the methodology will vary, most probably in the configuration of the sampling equipment or the composition of the crew. It is important to know how these changes affect data quality. The objective of this study was to assess the effect of the differences both in the electrofisher equipment type (low power, battery powered vs. high power, generator supplied units) and in the operators when assessing fish assemblages in a non-wadeable lowland river in Hungary. We found that compared to the superior effect of “natural” spatio-temporal heterogeneity, pure methodological factors accounted for a low or moderate (<15%) part of the variance in fish data. The most commonly measured assemblage level variables, such as rarefied species richness, similarity based assemblage composition (i.e. presence/absence) and relative abundance data were the most insensitive to changing the equipment type and/or the operator. However, the shared effect of methodological and spatio-temporal factors was important in shaping CPUE data and mean fish size of some species, suggesting that the sensitivity of these variables to methodological variations can vary in space and time. We concluded that in systems with high spatio-temporal heterogeneity, the relative importance of the herein investigated methodological bias is likely to remain within an acceptable range. Thence, in studies examining large-scale ecological patterns over wide range of habitats and/or large areas, some flexibility in the methodology may be a reasonable compromise in favor of sampling more sites and increasing effort. Nevertheless, if any change in the methodology is indispensable, at least its effect on sample quality should be assessed.

► We evaluated the effects of electrofisher and operator on fish assemblage data.
► Majority of variance in data were related to spatio-temporal heterogeneity.
► Pure methodological factors represented a moderate part (<15%) of variance.
► Shared effect of methodology was also important in some assemblage attributes.
► Rarefied species richness and relative abundance data were the most insensitive.