Fig 1.
Location of study sites and Boardman River dams for riffle macroinvertebrate survey, 2011–2016, Boardman River, Michigan, USA.
See Table 1 for full names and locations of sites. The Brown Bridge, Boardman, and Sabin dams were removed in 2013, 2017, and 2018, respectively. Reservoirs are pictured for later-removed dams that were removed after the study period. Maps generated by J. Betts in ArcMap, using publicly available shapefiles from the Michigan Geographical Framework (https://www.michigan.gov/som/0,4669,7-192-78943_78944---,00.html). 1 June 2013 [28].
Table 1.
Description of study sites in riffle macroinvertebrate survey, 2011–2016, Boardman River, Michigan, USA.
Fig 2.
Benthic macroinvertebrate data, summed from six Surber samples (0.093 m2) at each of 11 sites, and averaged at upstream, downstream, and formerly impounded (new channel) sites of the Brown Bridge Dam (removed in between 2012–2013) along the Boardman River, Michigan, USA.
Variables measured were: Invertebrates per square meter (A), taxon richness (family level, sometimes higher classification for non-insects) (B) and fraction of insects in orders Ephemeroptera, Plecoptera, and Trichoptera (%EPT) (C). Rectangles show mean +/- SE with mean dividing each rectangle. Points indicate measured values. Dashed line indicates time of dam removal. Due to large numbers of New Zealand mud snails (Potamopyrgus antipodarum) at all sites in later years, this taxon was excluded from summary variable calculations.
Table 2.
A) ANOVA tables of linear models of summary variables Invertebrate density, Taxon richness, and EPT percent as predicted by different site categorizations. B) Post hoc tests (Tukey HSD) for differences in EPT percent between strongly impacted (S) and weakly impacted (W) sites by year.
Fig 3.
Non-metric multi-dimensional scaling (NMDS) ordination of stream macroinvertebrate community composition in the Boardman River, Michigan, (USA).
3-axis solution displayed (Stress: 0.17; Linear Fit R2 = 0.83). Ordination hulls with distinct coloration highlight site categories, Site × Years displayed with the last 2 digits of the year of sampling. Chironomidae and Simuliidae were strongly associated with ordination axis 1. There was a weaker association of Lepidostomidae with ordination axis 2 and of Baetidae and Pleuroceridae with ordination axis 3.
Table 3.
A) ANOVA tables of linear models for five functional feeding groups (FFG) according to Bouchard (2004) as predicted by different site categorizations. B) Post hoc tests for differences in site category comparisons for significant FFG groups (CG and SC).
Fig 4.
Selected macroinvertebrate taxa whose abundances were sensitive to position above or below dam, timing of dam removal (dashed vertical line), or development of new channel sites following reconstruction of Boardman River channel following dam removal.
Abundance data from 2011–2016, Boardman River, Michigan (USA). Taxa grouped by conditions indicated. Top row: Bivalvia (unknown taxa at more specific levels) were Upstream indicators. Middle row: Helicopsychidae, Philopotamidae, and Brachycentridae were taxa indicative of Downstream-before sites compared to both Upstream and Downstream-after sites, potentially representing groups most affected by dam removal. Bottom row: Baetidae, Chironomidae, and Gordiidae were taxa indicative of sites in New Channel areas. Points show observed abundances. Vertical lines indicate mean +/- SE. Right-hand Y axis scale applies to third (last) group in second and third rows. Site groupings are offset on the X-axis for clarity.
Table 4.
Taxa that are significant (p<0.05) or marginally significant (p<0.1) indicators for each site category, with their functional feeding group (FFG) [34].
Table 5.
Density (individuals/m2) of New Zealand mud snails (Potamopyrgus antipodarum-Hydrobiidae) from 2011–2016, Boardman River, Michigan (USA).