Stoneflies (Insecta: Plecoptera) are one of many faunal groups that reflect the historical geography of Ohio. The presence and distribution of stoneflies in Ohio demonstrate not only the results of the terraforming effects of Quaternary glaciation, but also the various invasion routes available in preglacial epochs. For example, the preglacial (and pre-Ohio River) Teays River drainage, originated in western North Carolina and provided access to Ohio, Indiana and Illinois (Hansen 1995, King 1983, Ver Steeg 1946) Whereas, this extensive drainage is buried under 500 feet or more of glacial till from central Ohio westward, at least a few of the stoneflies which colonized Ohio using this route may have found refuge in the Western Allegheny and Appalachian Plateaus of eastern and southeastern Ohio during the glacial epochs. Others recolonized from refugia in the Cumberland Plateau, Southern Appalachian Mountains, and possibly the Ozark Mountains (Pessino et al. 2014, Ross et al. 1967). The series of glacial events flattened most of northwestern and western Ohio, down to the Cincinnati area, creating lake, till, and drift plains, bogs, and fens. In northwestern Ohio the Black Swamp (a.k.a. Great Black Swamp), a wooded wetland complex, was formed atop lake plains of ancient glacial Lake Maumee (Kaatz 1955). This area was not drained until the second half of the 19^th century. The sum of these historical events, in conjunction with more recent natural and human-caused factors, in large part, explains Ohio’s stonefly fauna today.

Properly maintained natural history (museum) collections provide a permanent record of life on Earth (Mehrhoff 1997). The use of information technology, coupled with data standards (unique identifiers, georeferencing, and data sharing formats), has recently improved access and manipulation of the information. The specimens and their labels place a species in space and time, making natural history collections useful not only for such typical purposes as systematics research, but also as a source of verifiable data to examine range changes over time, to study the effects of environmental degradation, and to predict the extent and severity of invasions of exotic species. We may also extract from these data life history information, habitat requirements, understand the imperilment of species at multiple scales, plan for restoration activities, and examine relationships of distributions to landscape and species trait constraints.

Given that stoneflies are one of the most sensitive indicators of change in habitat and water quality (Stewart and Stark 2002), they are important targets for digitization of museum specimen records and ecological analyses based on those records. Much work to this affect has already occurred in Illinois. Favret and DeWalt (2002) and DeWalt et al. (2005) amassed 5117 records for Illinois, demonstrating that 28% of the original fauna had been extirpated from the state, that every region of the state experienced losses, and that the data were sufficient to build state level conservation statuses for each species. Another direct result of compiling these large data sets was the Cao et al. (2013) predictions of pre-European settlement distribution and richness patterns of Illinois stoneflies at the USGS HUC12 watershed scale . Other studies in the USA that have benefited from accumulating stonefly museum data include DeWalt et al. (2012) for Ohio, DeWalt and Grubbs (2011) for Indiana, and Grubbs et al. (2013b) for Michigan. In Europe, Bojková et al. (2012) in the Czech Republic used 170 fixed sites to examine changes in the assemblage from the middle 20^th century. Additionally, RWB and colleagues are working on an atlas of stoneflies for Nevada, USA with a greatly expanded species list, distributional maps, specimen images, and a comprehensive database slated for publication in spring, 2017.

Prior to DeWalt et al. (2012), Ohio’s stonefly fauna had been studied in a piecemeal fashion. Walker (1947) provided a southeastern Ohio treatment, including a few records from the southwestern and northwestern corners of the state. His list contained 30 species, the identities of some being questionable and the majority unverifiable due to loss of the specimens. Later, Gaufin (1956) published on southwestern Ohio, bringing to 53 the number of species known from the state. His specimens were mainly larvae, but his material exists in various collections, especially at the Monte L. Bean Museum at Brigham Young University (BYUC) and in the Illinois Natural History Survey Insect Collection (INHS). Tkac (1979) conducted a more comprehensive study across the northeastern quarter of the state, but producing only 54 species. His dissertation included the first illustrated taxonomic key to Ohio stonefly larvae and adults. Relatively few of Tkac's specimens have been located and Dr. Ben Foote (pers. comm.) confirms that they are not at Kent State University where the degree was conferred. Late in the current study it was suggested that specimens may reside in the United States National Museum (USNM), but no formal records indicate such a donation ever took place. Many additional studies of a narrower scope have been published, either documenting the stonefly fauna of single streams, as taxonomic revisions, or as short updates to the known fauna. All known works have been documented and discrepancies in name usage have been reconciled in this document.

A much needed update of the Ohio fauna was begun in the 1980s and continued through the 1990s, conducted by RWB, SMC, BJA, and Ralph F. Kirchner (Wheeling, West Virginia). These efforts did not result in publication, but their thousands of specimens form the basis of this work. Beginning in 2005, RED and SAG borrowed material from individuals and institutions, identified the specimens, digitized the label data for 4,080 vials and pins of stoneflies, and georeferenced all locations, resulting in DeWalt et al. (2012). Subsequently, Grubbs et al. (2013b) discussed the distribution of some uncommon and rare species occurring in Ohio, but reported no additional species. Since then, a large collection of additional Ohio stoneflies was donated to the INHS by the Ohio Biological Survey. In addition, many more Ohio Environmental Protection Agency (OEPA) records were made available that dramatically improved the coverage of several species and underrepresented drainages.

Other specimens that improved our coverage include a substantial number of records from Edge of Appalachia Preserve (Adams County, Ohio Brush Creek drainage) collected by RED and specimens collected by Gary A. Coovert since 2004 from Crane Hollow Nature Preserve (Hocking County, Queer Creek drainage). Both locations added new locations for several rare species and confirmed the presence of another. All total, 7,723 specimen records now exist for Ohio stoneflies. This dramatic increase in specimens makes an update desirable, provides an opportunity to present a complete historical accounting of stonefly research conducted in Ohio, explore some relationships of species richness to drainage characteristics, add range maps, conduct analyses of stream widths used by species, and present an analysis of the succession of adult presence throughout the year. None of these analyses were present in DeWalt et al. (2012), though some distribution maps for rare species were provided in Grubbs et al. (2013b).

This publication is volume II in a series of atlases of aquatic insects inhabiting Ohio and complements volume I on caddisflies (Armitage et al. 2011). Future volumes will provide information on Ohio mayflies, aquatic beetles, crane flies, and aquatic and semiaquatic Heteroptera.

Digitization of specimen data. Data presented in this work represents a combination of verified specimens, specimen data from the OEPA, and trusted literature. We verified identifications of many of the most difficult to identify species among the OEPA specimens, strongly supporting their inclusion in this study. The specimen data source and number of records (# of vials or pins) are provided for each institution and colleague who provided specimens/data. The methodology for preparing specimens is available in DeWalt et al. (2012). We associated most specimens with their database record using a paper catalog number—a unique identifier. Unfortunately, this was not the case for OEPA specimens, the Western Kentucky University material, and literature sources. Specimen data were gathered in accordance with iDigBio (2014a) wet collection protocols. All data will be shared with the Global Biodiversity Information Facility (GBIF) and with iDigBio (2014b).

Most location labels printed prior to 2000 did not contain geographic coordinates. We georeferenced these locations using Acme Mapper 2.1 (Acme Mapper 2016, datum WGS-84). In the USA, this program provides topographic, satellite, and road map coverages that ensure the greatest possibility of finding complex locations. In addition, where collectors provided coordinates they were projected to verify that the coordinates matched verbal descriptions (correct county, distance and direction from locality, road crossing). Where they did not match, coordinates were corrected or recorded with lower precision in the database. We used a decimal degree format, most often to five significant figures, to improve the usability of the data by others. Estimated precision is presented as a radius in meters. Maps were exported from an ArcView 9.3 (ESRI) project file using a WGS-84 projection, overlaid on United States Geological Survey Hierarchical Unit Code eight (USGS HUC8, 42 drainages) scale drainages with outlines of the 88 Ohio counties. A map was constructed with all unique locations, and individual maps for each species.

Succession of species. Adults of stonefly species succeed each other as they emerge throughout the year (Stewart and Stark 2002). This is most clearly demonstrated from single site studies (Ernst and Stewart 1985), but regional data may also be used successfully for this type of analysis if latitudinal differences in the data are ignored. Our data are not derived from emergence traps; accordingly, they reflect presence rather than emergence. Adult stoneflies often live one or two weeks past their date of emergence (DeWalt and Stewart 1995). Hence, the succession of adults presented in contains a bias for the presence of adults collected after peak emergence. We have used adult records in the data set to build a table that depicts adult presence throughout the year on a weekly basis. Records for each species were examined and cells in an Excel spreadsheet were shaded corresponding to the intensity of emergence: dark gray when one or more collecting events (site/date combinations) in a week contained ≥3 adults; medium gray when collecting events contained ≤2 adults; and light gray where no adults were present, but when we assumed from larval records and our experience that adults would be available. All outlying dates of emergence were recorded and the species ordered chronologically to display the sequence of emerging species.

Species richness vs. county and watershed relationships. All georeferenced specimen records were associated with HUC8 coverage in GIS and the drainage numbers and names were returned to the data. The total species richness and number of unique locations within a HUC8 drainage were compiled. A map depicting of the number of species vs. HUC8 drainage was constructed so that drainages with similar species tallies were similarly color-coded. Scatterplots were constructed of species richness versus HUC8 area in km² and the number of unique locations within a HUC8 to determine if these variables were important to species richness. Deviations from trend lines produced from simple linear regression analyses were noted. Ohio counties, of which there are 88, are geopolitical units for local government (Anonymous 2016). In an effort to determine if there were areas not well sampled across the state, the number of total records were tallied for each county. A histogram was produced that depicts the number of stonefly records for each county. Those counties with high and low richness were examined for where they occurred within the state.

Distribution of species in stream size/type categories. Stoneflies live in a wide range of waterbody sizes, even in large lakes. Drainage area and perhaps the number of links (tributaries) are the best measures of stream size and may often be recovered from Geographic Information Systems data layers. However, these data sets often lack data for the smallest streams. To account for this streams were categorize by stream wetted width (1=seep, 2=1-2 m wide stream, 3=3-10 m wide, 4=11-30 m wide, 5=31-60 m wide, 6=>61 m wide, 7=large lake (Lake Erie specifically). These estimates were made from Acme Mapper (2016) satellite coverages using the scale provided by the program. A histogram of the frequency of site/date events within each stream width or lake category was constructed for each species for all sites that could be georeferenced to a stream or lake (91.2% of 7,723 records).

Access to the data. All specimen data used in this study are archived as a Darwin Core Archive file supported by Pensoft's Integrated Publishing Toolkit (DeWalt et al. 2016b). This data set contains some duplication in the form of literature records that may also be available as specimen data with unique identifiers, but we included in order to provide a complete record.

A total of 7,797 records were gathered from 21 institutional, government, personal collection sources, and from literature sources (Table 1). Most specimens (>5000) from physical collections were examined by RED & SAG. A total of 2769 unique locations have been georeferenced and mapped (Fig. 1).

Figure 1.  

Ohio stonefly collection records, county boundaries, and HUC8 drainages.

At least 53 papers have appeared in print that reference Ohio stoneflies (Suppl. material 1). These include faunal lists and analyses of species richness patterns for the state as a whole or a subset (DeWalt et al. 2012, Gaufin 1956, Grubbs et al. 2013b, Tkac 1979, Walker 1947), records of taxa from a single stream (Beckett 1987, Tkac and Foote 1978, Robertson 1984, Robertson 1979, Fishbeck 1987), discussion of morphological features or genetic diversity for one or more species (Clark 1934, Yasick et al. 2007, Yasick et al. 2015), or included records of Ohio species from descriptions or revisionary or other works (Baumann 1974, Frison 1942, Fullington and Stewart 1980, Grubbs 2006, Grubbs 2015, Grubbs and DeWalt 2008, Grubbs and DeWalt 2012, Grubbs and Stark 2001, Grubbs et al. 2014, Grubbs et al. 2013c, Kondratieff 2004, Kondratieff and Kirchner 1993, Kondratieff and Kirchner 2009, Kondratieff et al. 1988, Nelson 2000, Ricker 1952, Ricker and Ross 1968, Ricker and Ross 1969, Ross and Ricker 1964, Ross and Ricker 1971, Ross and Yamamoto 1967, Ross et al. 1967, Stark 1986, Stark 1989, Stark 2000, Stark 2004, Stark and Baumann 1978, Stark and Baumann 2004, Stark and Gaufin 1974, Stark and Gaufin 1976, Stark and Kondratieff 2010, Stark and Kondratieff 2012, Stark et al. 1988, Stewart 2000, Surdick 2004, Szczytko and Kondratieff 2015, Szczytko and Stewart 1978, Szczytko and Stewart 1981, Young et al. 1989, Zwick 1971).

Species present and those dismissed from the state tally

Species richness vs. watershed and county relationships

Stonefly species richness varied tremendously with HUC8 affiliation (Fig. 2). The Lower Scioto River drainage supported 72 stonefly species, 18 more than the next richest drainage. The richest drainages were in northeast, south-central, and southern Ohio. These areas are heavily forested, have the highest slopes (DeWalt et al. 2012), and were part of or adjacent to the unglaciated, Western Allegheny Plateau of Ohio . The drainages with the lowest richness were mostly found in the northwestern quarter of Ohio, which was the most glaciated area of Ohio and site of the Great Black Swamp during the post-glacial period. Eight western drainages supported five or fewer species with three drainages, the Upper Wabash, Ottawa-Stony, and St. Mary's supporting only one or two species (Fig. 2). Dominated by glacial lake plain topography, these drainages have low slope values, fine-grained sediments, and now, approximately 90% coverage in row crop agriculture (DeWalt et al. 2012). Historically, they would not have supported many stonefly species, and with the agriculturally modified landscape, few remain.

Figure 2.  

Stonefly species richness for 41 Ohio USGS HUC8 watersheds. Watershed color coded by similar richness. Watershed names for some species poor and species rich drainages provided.

Surface area of HUC8 drainages appears to be an unimportant predictor of stonefly species richness (Fig. 3). One point is well above the line-of-best-fit, that of the Lower Scioto drainage. It is the richest, despite not being the largest, HUC8 drainage. Many relatively small HUC8s have high richness, while many intermediate sized drainages support only a few stonefly species. The number of unique locations sampled within a watershed appears to be a much stronger predictor of stonefly species richness (Fig. 4). Again, the Lower Scioto drainage exceeds predictions. Conversely, the Upper Scioto, the Upper Greater Miami, and Little Muskingum drainages all fall below the line-of-best-fit. These drainages are either largely agricultural, have high industrialization, or have large human populations in them, all conditions that would lead to lower than expected stonefly richness.

Figure 3.  

Stonefly species richness vs. HUC8 surface area (km²). Simple linear regression equation, R², and line-of-best-fit provided. Lower Scioto watershed point indicated.

Figure 4.  

Stonefly species richness vs. number of HUC8 unique locations. Simple linear regression equation and R² provided. Names of HUC8s with greatest deviation from line-of-best-fit provided.

At least one stonefly record is available for each of Ohio's 88 counties (Fig. 5). Hocking County in south-central Ohio has more stonefly records than any other county by nearly a factor of two. It is the most important county contributing to the richness of the Lower Scioto drainage (59 of 72 spp., next has 44 spp.). Because Hocking County has never been glaciated, it maintains a rugged topography with deep ravines composed of Pennsylvanian and Mississippian age sandstones and shales, respectively (Hansen 1975). These ravines and the creation of Ohio State Forests in 1915 protected streams from logging and farming, preserving much of the rich native stonefly fauna of the area. Protected areas in the county include Hocking Hills State Park, Hocking Hills State Forest, and the small but species-rich Crane Hollow Nature Preserve. Other species rich counties are located in northeastern, south-central, and southern Ohio. Those counties with the lowest diversity are generally northwestern, again their diversity suffering from historically flat terrain, lake plain topography, sluggish streams, and the contemporary dominance of agricultural land use (DeWalt et al. 2012).

Figure 5.  

Stonefly species richness for 88 Ohio counties (only every other name presented). Regions of the state with richest and poorest totals presented.

Succession of adult presence