Examining effects of a 100-year flashflood on in-stream riparian characteristic and macroinvertebrate meta-community dynamics in fourteen 1st-3rd order headwater streams

Background/Question/Methods

In 2009, a 100-year flash flood scoured 14 neighboring small 1^st-3^rd-order headwater streams in Zoar Valley, New York USA. Habitat features, macroinvertebrate composition, and watershed variables of these streams were previously characterized in 2006, and results were reported in Hydrobiologia. Streams initially were marked as having distinct heterogeneous habitat patches, whereas post-flood, streams were barren without substrate and course woody debris. To assess the long-term influence of this disturbance within these streams, macroinvertebrate composition and habitat features were assessed for the years 2010 through 2013, mirroring efforts from 2006. In-stream and riparian features were quantitated using the Qualitative Habitat Evaluation Index (QHEI), watershed characteristics were verified via ground-truthing and aerial photography, while biota was sampled using replicate Surber-net samples in each stream, and identified to genus and species when possible. Particular attention is given to how to the meta-community dynamics of these streams have transformed in terms of environmental (niche-based or neutral) and/or spatial organization (patch dynamics) of macroinvertebrate composition. In 2006, a significant environmental structuring of macroinvertebrates r=0.594, p < 0.001) suggested either a niche-based “species sorting” or mass effects framework controlling community organization. NMDS biota ordinations supported this, both 3^rd- and 2^nd –orders clustered while diverse 1^st-orders diverged. 

Results/Conclusions

Before disturbance, eight streams had excellent QHEI rankings. In 2010, initially post-disturbance, no streams received such ranking. By 2012 and 2013, overall QHEI scores increased, but 11/14 streams remained lower than before, with highly incised 3^rd order streams remaining most decremented. Spearman’s rho revealed an increasing-trend for the correlations of environmental conditions between year 2006 and across years 2011-2013, with 2013 revealing the strongest relationship (r=0.997, p<0.001), suggesting a post-flood recovery. A similar recovery-trend trend was revealed by Spearman’s rho for all biota across the 2011-2013 period, when compared with 2006. Again, the 2013 biotic correlation was the strongest (r=0.724, p<0.000), perhaps suggesting the start of return for macroinvertebrate communities. However, no significant environmental structuring of invertebrates has returned since flooding (also supported by NMDS ordinations). In contrast, weak, yet significant correlations emerged between the distance and biota matrices for 2012 (r=0.271, p=0.010) and 2013 (r=0.357, p=0.001), suggesting an influence of dispersal on the distribution of taxa. This type of structuring suggests a switch to homogenous patch dynamics framework, where habitat characteristics are less critical in site selection due to greater dispersal influence. Perhaps with another year of data, distinctions between operating frameworks and alternate community states will become clearer.