Hierarchical multi‐taxa models inform riparian vs. hydrologic restoration of urban streams in a permeable landscape

Abstract

The degradation of streams caused by urbanization tends to follow predictable patterns; however, there is a growing appreciation for heterogeneity in stream response to urbanization due to the local geoclimatic context. Furthermore, there is building evidence that streams in mildly sloped, permeable landscapes respond uncharacteristically to urban stress calling for a more nuanced approach to restoration. We evaluated the relative influence of local-scale riparian characteristics and catchment-scale imperviousness on the macroinvertebrate assemblages of streams in the flat, permeable urban landscape of Perth, Western Australia. Using a hierarchical multi-taxa model, we predicted the outcomes of stylized stream restoration strategies to increase the riparian integrity at the local scale or decrease the influences of imperviousness at the catchment scale. In the urban streams of Perth, we show that local-scale riparian restoration can influence the structure of macroinvertebrate assemblages to a greater degree than managing the influences of catchment-scale imperviousness. We also observed an interaction between the effect of riparian integrity and imperviousness such that the effect of increased riparian integrity was enhanced at lower levels of catchment imperviousness. This study represents one of few conducted in flat, permeable landscapes and the first aimed at informing urban stream restoration in Perth, adding to the growing appreciation for heterogeneity of the Urban Stream Syndrome and its importance for urban stream restoration.

Large‐scale environmental flow results in mixed outcomes with short‐term benefits for a semi‐arid floodplain plant community

“…Furthermore, these metrics ignored taxa identification and so did not capture the variety of responses within functional groups. These disadvantages of community summary metrics have been noted in the ecological literature (Gwinn et al., ; Olden, ; Olden, Joy, & Death, ) with multi‐taxa models proposed as a solution. However, multi‐taxa models are rarely applied to plant research (but see Chen, Kery, Plattner, Ma, & Gardner, ) and default methods of determining community patterns include modelling community summary metrics (e.g.…”

“…functional groups) to reduce variation in plant data and to increase the ease of interpreting complex patterns can mask true changes in assemblage structure because taxonomic identity is not retained. For example, a few dominant taxa can disproportionately influence the value of summary metrics, masking potential responses by rare taxa and failing to capture the diversity of responses among taxa (Bent & Forney, ; Gwinn et al., ).…”

Going across taxa in functional ecology: Review and perspectives of an emerging field

“…We ran the same analysis either by keeping or removing the outliers (articles with >15 taxa, e.g. Gwinn et al, 2018), and replacing the number of taxonomic ranks (study taxonomic range) by the mean pairwise phylogenetic distance between taxa (MPD; Webb et al, 2002) for articles with at least two taxa. We also ran two sensitivity analyses to test whether (i) the allocation of traits into groups of trait types affected the results and (ii) the trade‐off would be maintained if we removed trait types from the dataset.…”

Demonstrating the need to simultaneously implement all water sensitive design methods for aquatic ecosystem health

“…In Australasia ecologists in research institutions and networks are using bioindicators, for this purpose, and specifically using macroinvertebrate indices appropriate to location and waterway type. Australian examples of such institutes include the CRC Water Sensitive Cities (2019) ; the Waterway Ecosystem Research Group at the University of Melbourne ( Walsh and Webb, 2013 ; Walsh et al., 2016 ); and collaborative researchers in Perth ( Gwinn et al., 2018 ). Walsh et al.…”

“…In Australasia ecologists in research institutions and networks are using bioindicators, for this purpose, and specifically using macroinvertebrate indices appropriate to location and waterway type. Australian examples of such institutes include the CRC Water Sensitive Cities 2019 and collaborative researchers in Perth (Gwinn et al, 2018). Walsh et al (2015) reported on the as yet little-changed biological and water quality responses of a comparative catchment experiment where they retrofitted 289 stormwater retention systems to the treatment catchment.…”