Reconstructing historic stonefly species distributions across an interglacial suture zone

Jason L. Robinson, Yong Cao, R. Edward DeWalt, Tari Tweddale, Leon C. Hinz Jr.

Research output: Contribution to conferenceOtherpeer-review


Background/Question/Methods: Aquatic insect species distributions are often poorly known by researchers and even more poorly reported in the scientific literature, where "state lists" are a typical format for distributional data. We obtained and taxonomically verified over 20,000 unique specimen records for 145 species of stoneflies (Insecta: Plecoptera) across IL, IN, OH, MI and WI. These specimens were obtained from university or institutional collection loans and additional collections made by the authors. We used environmental predictors based on temperature, precipitation, pre-European settlement vegetation, hydrology, elevation and geology variables (summarized at the scale of USGS HUC12 watersheds) to build Maxent models of the distribution of 79 species in 8691 watersheds across the five state region. Since some watersheds have been more intensely sampled than others, we eliminated species records from 100 of the most heavily sampled watersheds and compared distribution models built from this reduced dataset to evaluate whether predictions are robust to unequal sampling effort across watersheds. Results/Conclusions: The study area includes glaciated and unglaciated areas of varying species richness and environmental characteristics. Previous authors have used areas within this region as examples of “suture zones”, where contact or hybridization occurs between recently diverged lineages. We used a Getis-Ord local G statistic to identify “hotspots” of high species richness, consistent with several proposed glacial refugia (Appalachian, Driftless Areas of WI/IL), and “coolspots” of low species richness corresponding to patterns of habitat quality as a function of glacial processes (e.g., fine sediments deposited in till plains). Since many (66) species had too few unique watershed occurrences for distribution modeling, we use interpolation and hotspot methods to address whether our subset of 79 modeled species is likely to reconstruct the true regional pattern for all 145 species. Finally, we discuss the significance of the post-glacial recolonization patterns implied by these results (under presettlement environmental conditions) to the potential effects of climate change on species distribution in contemporary landscapes with greatly altered land uses.
Original languageEnglish (US)
StatePublished - 2013


  • INHS


Dive into the research topics of 'Reconstructing historic stonefly species distributions across an interglacial suture zone'. Together they form a unique fingerprint.

Cite this