Combining taxon-by-trait and taxon-by-site matrices for analysing trait patterns of macroinvertebrate communities: A rejoinder to Monaghan & Soares (2014)

Abstract

Summary: Monaghan and Soares (2014) suggested that combining traits with log-transformed abundance of taxa may cause anomalies in analyses of stream macroinvertebrate communities. While they addressed an important issue in stream ecology, here we present an opposite view. To identify the causes of these contrasting opinions, we carefully examined the examples provided by Monaghan and Soares (2014) and demonstrated how traits can be weighted by the presence, abundance and log-transformed abundance of the taxa in a meaningful way. We found that Monaghan and Soares (2014), following other authors, use the term 'weighting' differently from classical papers of stream ecology. The general practice is to calculate the sum of trait values multiplied by the abundance of each taxon and divide it by the total invertebrate abundance to get a community-level trait value. In contrast, Monaghan and Soares (2014) did not perform the final division and consequently did not get a standardised community-level trait value. It follows that the term 'weighting' is used with different meanings in stream ecology, and ecologists should keep these differences in mind. We agree with Monaghan and Soares (2014) that the addition of log-transformed data is equivalent to multiplication on an arithmetic scale. However, we disagree that this provides an inconsistent scaling that confounds quantitative analyses. Using example data sets, we illustrate how trait-based data analysis can be performed in community ecology in a meaningful way. © 2014 John Wiley & Sons Ltd.