Abstract
Swan and Brown (2017) recently addressed the effects of restoration on stream communities under the meta-community framework. Using a combination of headwater and mainstem streams, Swan and Brown (2017) evaluated how position within a stream network affected the outcome of restoration on invertebrate communities. Ostensibly, their hypotheses were partially supported as restoration had stronger effects in headwater streams: invertebrate taxonomic richness was increased and temporal variability decreased in restored reaches; however, these results were not consistent upon closer scrutiny for both the original paper (Swan and Brown 2017) and the later erratum (Swan and Brown 2018). Here, I provide a secondary analysis of the data, with hypotheses and interpretations in the context of stream, metacommunity, and restoration ecology. I did not find any effects of restoration on local diversity, spatial dissimilarity, or temporal variability, let alone differential effects of restoration between headwaters and mainstems; these results are contrary Swan and Brown (2017, 2018), who reported that restoration increased taxonomic richness, increased spatial dissimilarity, and decreased temporal variability in restored headwater streams. I demonstrate further that the statistical tests conducted by Swan and Brown (2017, 2018) were invalid and, therefore, recommend the use of the results presented here. More broadly, I suggest that river and stream restoration will likely have greater success if a regional approach is taken to designing and implementing restoration projects.
Footnotes
The manuscript has been updated following comments provided by the recommender and two reviewers at PCI ecology. Changes include: (1) Incorporating effect size calculations to allow for easier comparisons among models, following suggested revisions by a reviewer at PCI Ecology (2) Brief description of the study and sampling design (3) Brief description of the effect size metric and how it is calculated, and then describing for which response variables and model sets effect sizes were calculated to maintain consistency and logical comparison among studies.