Macroinvertebrate traits distinguish unregulated rivers subject to water abstraction

Can biological traits of stream invertebrates help disentangle the effects of multiple stressors in an agricultural catchment?

“…Water abstraction can increase the frequency and duration of low flows, enhance the effects of natural droughts and generate artificial droughts (Finn, Boulton & Chessman, 2009;Brooks, Chessman & Haeusler, 2011). see Richards et al, 1997;Dol edec et al, 2006;Mellado-D ıaz, Su arez Alonso & Vidal-Abarca Guti errez, 2008), but the effects of water abstraction on trait composition have been studied much less often (Brooks et al, 2011) than the effects of natural flow variations and droughts (Williams, 1996;Bêche et al, 2006;Bêche & Resh, 2007;Bonada, Rieradevall & Prat, 2007). see Richards et al, 1997;Dol edec et al, 2006;Mellado-D ıaz, Su arez Alonso & Vidal-Abarca Guti errez, 2008), but the effects of water abstraction on trait composition have been studied much less often (Brooks et al, 2011) than the effects of natural flow variations and droughts (Williams, 1996;Bêche et al, 2006;Bêche & Resh, 2007;Bonada, Rieradevall & Prat, 2007).…”

Predicting reference assemblages for freshwater bioassessment with limiting environmental difference analysis

“…The accuracy of the LEDA models for the ACT, GL, and Y T data sets, measured as the average BC similarity between observed and modeled assemblages for unimpacted validation samples, was mostly within or above the range of 0.4 to 0.6 achieved in previous LEDA modeling for various biotic groups (Chessman et al 2008a, Brooks et al 2011. For all 3 data sets, accuracy rose significantly and substantially as the SP increased.…”

“…LEDA modeling (Chessman et al 2008a, Brooks et al 2011 creates biological reference data for comparison with observed data for any sample of interest (hereafter a target sample) as a simple average of reference-site samples that are environmentally matched to the target sample. Reference samples that are not matched to the target sample make no contribution to the prediction for that sample, although they may contribute to the prediction for other samples.…”

A proposed unified terminology of species traits in stream ecology

“…For instance, Brooks et al (2011) used traits of macroinvertebrate families with ordinal-scale states (sensu Poff et al 2006), such as swimming ability and occurrence in drift, to characterize river communities. They coded these states by integer values (e.g., swimming ability: none [code = 1], weak [2], and strong [3]) and then 'standardized' these values by division using the maximum value for the given trait (here 3).…”

“…For the same reason, the Euclidean distance measure is incompatible with ordinal-scale data (Podani 2000, Podani andSchmera 2006). The solutions to this problem are: 1) the use of analytical methods developed for ordinal scale (Podani 2000(Podani , 2005, 2) the reduction of ordinal variables to nominal ones (as made by Poff et al 2006), or 3) the expansion of the ordinal scale to an interval (or ratio) scale (as done in some analyses by Brooks et al 2011). However, the expansion of the data scale should be justified and explained clearly to avoid mathematically nonsensical statements (i.e., standardization of an ordinal scale variable by its maximum).…”