The Poweshiek skipperling Oarisma poweshiek, Lepidoptera: Hesperiidae is a historically common prairie butterfly with a range extending throughout prairie systems of the upper midwestern United States and southern Manitoba, Canada. Rapid, range-wide declines have reduced the number of verified Poweshiek skipperling locations to one in Manitoba prairie, one in Wisconsin prairie, and four in prairie fens in Michigan. Our objective was to investigate parameter suites with the potential to be biologically relevant to Poweshiek skipperling occupancy with the goal of informing conservation efforts. At 18 prairie fens categorized as occupied (n = 9) or unoccupied (n = 9), we collected information on plant biodiversity, water chemistry, soil chemistry, site geometry, and surrounding current and historical land cover at three spatial scales. To address the complexity of these systems, we used multiresponse permutation procedures and nonmetric multidimensional scaling to explore associations between variable groups thought to be relevant to Poweshiek skipperling (conditions for suspected larval host plants, system integrity, and agricultural influence) and occupancy categories. We used indicator species analysis to understand the relationships between plant biodiversity and Poweshiek skipperling occupancy at whole- and intrafen scales. Multiresponse permutation procedures analysis suggested that conditions for suspected larval host plants differed between occupied and unoccupied prairie fens. At the whole-fen scale, we identified 14 plant species associated with Poweshiek-occupied sites, including two purported larval host plants, Muhlenbergia richardsonis and Schizachyrium scoparium. At the intrafen scale, we identified 52 species associated with unoccupied Poweshiek sites, including many weedy species and those tolerant of inundated conditions. Our results can inform the evaluation of potentially suitable habitat for introduction and reintroduction efforts.

The Poweshiek skipperling Oarisma poweshiek Lepidoptera: Hesperiidae is a historically common butterfly of prairie systems in the upper midwestern United States and southern Canada (Figure 1). Distribution-wide declines in the past 20 y have prompted the United States and Canada to classify the Poweshiek skipperling as federally endangered (U.S. Endangered Species Act [ESA 1973, as amended]; SARA 2002; COSEWIC 2014). In 2016, surveyors observed Poweshiek skipperling individuals in Michigan (n = 104; unpublished data) and Wisconsin (n = 2) (P. Delphey, personal communication). Surveyors observed five individuals in Manitoba, Canada in 2016 (R. Westwood, unpublished data). There are 10 locations in Michigan with documented Poweshiek skipperling occupancy within the past 20 y; each of these locations is within prairie fen natural communities. From 2014 to 2016, surveyors observed Poweshiek skipperling at four of these locations. Despite the historical abundance of this butterfly, very little is known about the ecology of the species in an unaltered landscape.

Figure 1

Poweshiek skipperling Oarisma poweshiek nectaring on a black-eyed Susan (Rudbeckia hirta) in Oakland County, Michigan, 2013 (photo courtesy David Cuthrell).

Figure 1

Poweshiek skipperling Oarisma poweshiek nectaring on a black-eyed Susan (Rudbeckia hirta) in Oakland County, Michigan, 2013 (photo courtesy David Cuthrell).

Close modal

Poweshiek skipperling is univoltine, having a single time per year when adults emerge from pupae to nectar, mate, and lay eggs. This period, or flight, lasts up to 4 wk across the range, with individual adults believed to live 10–14 d in the wild. During the flight, Poweshiek skipperling mate and females are reported to lay fertilized eggs on thin, threadlike graminoids, on or near suspected larval host plants (McAlpine 1972; Cuthrell and Slaughter 2012). Suspected larval host plants reported in the literature for the U.S. populations include prairie dropseed Sporobolus heterolepis; Michigan special concern, mat muhly Muhlenbergia richardsonis; Michigan threatened, little bluestem Schizachyrium scoparium, elliptic spikerush Eleocharis elliptica, undetermined sedges in the Carex genus (potentially Carex buxbaumii), porcupine grass Stipa spartea, big bluestem Andropogon gerardii, and sideoats grama Bouteloua curtipendula (Holzman 1972; McCabe 1974; Borkin 1995; Selby 2005; Cuthrell and Slaughter 2012; Pointon 2015). Additional in situ observations are needed to understand and verify use and preferences of larvae for various potential host plants both in Michigan and throughout the range (Selby 2005; Cuthrell and Slaughter 2012).

Several factors have been put forth as possible causes for the decline of the Poweshiek skipperling. Potential stressors include habitat destruction, degradation, and fragmentation (Ricketts 2001; Selby 2005; Donald and Evans 2006; Davis et al. 2007; Weibull et al. 2008; Cuthrell and Slaughter 2012), pesticides (Dover et al. 1990; Selby 2005; Godfray et al. 2014), invasive species (Kost et al. 2007), imprudent burning practices (Selby 2005; Swengel et al. 2010), pathogens (Selby 2005; Werren et al. 2008; Nice et al. 2009; Poweshiek Working Group 2011), climate change (Selby 2005; Hoving et al. 2013), and genetic bottlenecks (Saarinen et al. 2016). Many of these proposed stressors are directly or indirectly related to environmental factors, such as plant biodiversity, hydrogeochemisty, site geometry, geography, and surrounding land cover.

Throughout the Poweshiek skipperling's historical range, the species occupied a range of prairie habitats that included fens, grassy lake and stream margins, moist meadows, and wet-mesic to dry tallgrass prairies (Selby 2005). In Michigan, all detected populations have been found within prairie fens (Michigan Natural Features Inventory 2014). Prairie fens are at-risk wetland natural communities found throughout the upper Midwest. A tallgrass prairie component helps distinguish these communities from other fens (Kost et al. 2007). The relationships between moist soils, tallgrass prairie flora, and historical juxtaposition with upland prairies or savannas possibly produced habitat for Poweshiek skipperling similar to other natural communities throughout their historical range. These similarities, along with post-European conversion of prairies to agriculture, have likely left prairie fens as the only natural community in Michigan in which Poweshiek skipperling can survive (Samson and Knopf 1994; Landis et al. 2012). However, it is not understood to what extent Poweshiek skipperling might have used other prairie systems in Michigan before European colonization.

Prairie fens are highly diverse wetlands found in the formerly glaciated Midwest region of North America. These natural communities are calcareous, groundwater fed, and dominated by sedge (Cyperaceae) species, and also include a prominent tallgrass prairie biotic component (Kost et al. 2007). Because of their unique ecology and hydrogeology, these systems provide habitat for many rare and sensitive biota, including the Mitchell's satyr Neonympha mitchellii mitchellii, Eastern Massasauga rattlesnake Sistrurus catenatus catenatus, small white lady's-slipper Cypripedium candidum, and Poweshiek skipperling. Prairie fens are comprised of four vegetation zones: sedge meadow, inundated flat, wooded fen, and calcareous groundwater seepage area (Spieles et al. 1999; Kost et al. 2007). Poweshiek skipperling are often found on peat domes that incorporate sedge meadow and tallgrass prairie flora (Figure 2). Prairie fen ecosystems are vulnerable to surrounding land use, invasive species, nutrient loading, agricultural runoff, and hydrological alterations (Spieles et al. 1999).

Figure 2

Poweshiek skipperling Oarisma poweshiek-occupied patch of prairie fen in Oakland County, Michigan, 2014. Blooming Dasiphora fruticosa provide nectar sources in this area. Examples of invasive species (cattail; Typha spp.) and shrub encroachment on the margins (dogwood; Cornus spp.) are visible (photo courtesy Clint Pogue).

Figure 2

Poweshiek skipperling Oarisma poweshiek-occupied patch of prairie fen in Oakland County, Michigan, 2014. Blooming Dasiphora fruticosa provide nectar sources in this area. Examples of invasive species (cattail; Typha spp.) and shrub encroachment on the margins (dogwood; Cornus spp.) are visible (photo courtesy Clint Pogue).

Close modal

Despite the decline of Poweshiek skipperling, few peer-reviewed articles have explored the ecology, habitat requirements, and factors potentially driving the decline of this species. Those that have explored the biology of this species call for more intensive work into the investigation of potential reasons for decline (Swengel and Swengel 1999, 2014; Landis et al. 2012; Dearborn and Westwood 2014; Pogue et al. 2016; Saarinen et al. 2016). The historical ecology of Poweshiek skipperling is poorly understood. The first detection of Poweshiek skipperling within Michigan was not until 1892 (Belitz et al. 2018), and much of the information regarding the pre-European settlement ecosystems (e.g., full distribution of ecosystems inhabited, plant associations, geochemistry) in which Poweshiek skipperling occurred does not exist.

The reduction of Poweshiek skipperling populations over the last 20 y, recent status change to federally endangered, ongoing conversion of habitat, and decreasing survey counts call for assessments of Poweshiek skipperling status, including an investigation of potentially defining habitat requirements. The objectives of our study were to 1) assess the characteristics of Poweshiek skipperling habitat in Michigan by examining biologically relevant parameter suites and 2) identify potential indicator plant species within and across prairie fens that are associated with occupied or unoccupied Poweshiek skipperling sites. Our findings have the potential to inform conservation management, reintroduction efforts, and further studies of this declining species.

Site classification

As detailed in Pogue et al. (2016), we completed modified Pollard–Yates surveys (Pollard and Yates 1993) targeting Poweshiek skipperling in prairie fens from 2014 through 2016. We used results from these surveys to determine the Poweshiek skipperling status at each site (Michigan Natural Features Inventory 2014). Occupied prairie fens have documented Poweshiek skipperling occurrence records within the past 20 y that have been validated by a species expert per the Natural Heritage established methodology. Presumed unoccupied prairie fens have had no previous documentation of presence. We a priori selected nine occupied sites and nine unoccupied sites to compare for the study. We selected unoccupied sites by identifying sites with similar size, geography, and biotic composition (including presence of potential larval hosts) as those of occupied sites. We also included proximity to occupied sites and ranking in a Poweshiek skipperling habitat suitability model in our selection process (Pogue et al. 2016; Figure 3). We completed additional repeated surveys 2014 through 2016 to confirm current nonoccupancy at these locations (Pogue et al. 2016).

Figure 3

Distribution of prairie fens (N = 18) from study examining local- and landscape-level factors contributing to Poweshiek skipperling Oarisma poweshiek occupancy in Michigan. Green circles (n = 9) indicate that the site was occupied within the past 20 y and was verified by a species expert. Red circles (n = 9) indicate that the site has been visited at least twice to confirm nonoccupancy of Poweshiek skipperling. Poweshiek skipperling occupancy data were collected in 1995–2015.

Figure 3

Distribution of prairie fens (N = 18) from study examining local- and landscape-level factors contributing to Poweshiek skipperling Oarisma poweshiek occupancy in Michigan. Green circles (n = 9) indicate that the site was occupied within the past 20 y and was verified by a species expert. Red circles (n = 9) indicate that the site has been visited at least twice to confirm nonoccupancy of Poweshiek skipperling. Poweshiek skipperling occupancy data were collected in 1995–2015.

Close modal

Plant diversity and hydrogeochemistry

During 2012–2015, we completed species-level vegetation surveys and collected geochemical samples at 18 prairie fens. We used an area-proportional random design to determine our vegetation sampling locations within each prairie fen for both spring and summer sampling periods as detailed in Hackett et al. (2016). To set sampling locations within the prairie fens, we used ArcMap 10.2.2 and prairie fen polygons retrieved from the Michigan Natural Features Inventory Natural Heritage Database (ESRI 2014; Michigan Natural Features Inventory 2014). We sampled a minimum of 20 quadrat locations within each prairie fen regardless of site dimensions. We expanded the range of prairie fen sizes used by Hackett et al. (2016) to match Poweshiek skipperling-occupied prairie fens ranging from 5.6 ha to 91.7 ha.

Within each 1-m2 quadrat, we counted and identified all species and recorded a Daubenmire cover value for each species (Daubenmire 1959). When we encountered a species new to this study, we collected a voucher specimen. Voucher specimens were deposited in the Central Michigan University Herbarium.

Using species and cover data from two sample periods in a single growing season, we calculated species occurrences, exotic relative abundance (ERA), mean coefficient of conservatism (C̄), and adjusted Floristic Quality Index (FQIadj). We used species occurrence data to calculate Sorensen's similarity index between occupied and unoccupied prairie fens (Magurran 2004). We sampled spring and summer seasons to capture the most diversity possible (Hackett et al. 2016). The phenology of these systems does not naturally clump into two distinct seasons and the cutoff date between the two periods was often resultant of project logistics, so we did not analyze these data on a per-season basis.

Exotic relative abundance is the percentage of the area sampled that is occupied by exotic species and serves as an index of the influence invasive species have had on the plant communities. We calculated ERA by converting the Daubenmire cover class of an individual exotic species to the mean percent coverage, divided that value by the total area sampled, and summed for all exotic species (Daubenmire 1959). The coefficient of conservatism reflects the likelihood that a species existed at a site before European colonization and gives insight into the level of human disturbance (Herman et al. 2001). This coefficient can range from 0 to 10, with higher values indicating that a species is less likely to be found in anthropogenically disturbed areas. These values are assigned on a state or local level. For Michigan, the C for each species was determined by using the methodology proposed by Swink and Wilhelm (1994) and Wilhelm and Masters (1995). We gathered these coefficients from University of Michigan's Michigan Flora (Reznicek et al. 2011). This is especially relevant in our study as the Poweshiek skipperling larvae are potentially reliant on state threatened and special concern grass species, including mat muhly and prairie dropseed. Adjusted FQI measures how the flora differs from its likely historical composition. In the FQI scale, a score of less than 20 meant that the habitat is of insignificant quality; a score greater than 35 was considered floristically important, and a score greater than 50 gave a site a rare rating and it was considered extremely important to the biodiversity of the state (Herman et al. 2001). Adjusted FQI (FQIadj) is a manipulation of FQI that accounts for exotic species, but it is often strongly correlated with FQI. We calculated FQIadj and used this metric for analysis instead of FQI to account for effects of exotic, and potentially invasive, species. The FQIadj uses C and standardizes it to the number of species found at the site, including nonnative species ():

We collected soil and water samples for geochemical analyses according to Hackett (2013). We sampled the soil at five randomly selected quadrats in each wetland during the spring season and deposited them at Michigan State University Soil Lab to be analyzed for mean calcium, pH, phosphorus, total nitrogen, nitrate-nitrogen, and ammonium-nitrogen. We collected water samples at five randomly selected quadrats in each wetland during the summer season and submitted these samples to the Center for Applied Research and Technology at Central Michigan University to be analyzed for mean ammonium, nitrates–nitrites, and soluble reactive phosphorus to be used in our analysis (USEPA 1994).

Geography and geometry

To explore the influence of edge effects, dispersal potential, beta diversity, and current and historical surrounding land cover on Poweshiek skipperling presence, we used ArcMap 10.2.2 to calculate mean distance to the nearest three prairie fens and proportions of current and historical surrounding land-cover classes at 100-m, 1,000-m, and watershed spatial scales (ESRI 2014; Timoney 1999). We calculated the watersheds for each prairie fen using the Watershed tool in ArcMap 10.2.2 (ESRI 2014). We determined current land cover using National Oceanic and Atmospheric Association's (NOAA 2014) Coastal Change Analysis Program (Regional Land Cover Database based on the National Land Cover Database of 2010; overall accuracy of 84.8%) to determine the percentage of surrounding land covers within each spatial scale. The Coastal Change Analysis Program data set contains 24 land-cover classes (Herold 2014). We combined the land classes into two land-cover categories (natural land cover and crop land cover), which we predicted to be biologically relevant to Poweshiek skipperling. We predicted these two categories to be biologically relevant because a greater proportion of natural land cover could suggest a more intact system and a greater proportion of crop land cover could suggest an increased exposure to pesticides, which has been put forth as a potential reason for the decline of Poweshiek skipperling (Runquist and Heimpel 2017). We combined all classes of forest, wetland, grassland, and scrub–shrub into the natural land-cover category and crops were retained as one category. To calculate proportions, we created buffers at each scale that did not include the prairie fen area. We then used the “Tabulate Area” function in ArcMap 10.2.2 to calculate land-cover proportions within each buffer scale. We determined historical land-cover estimates using Michigan Resource Information System land-cover class estimates for Michigan circa 1800 (Comer and Albert 1997). We calculated surrounding proportions of wet prairie land-cover class at 100-m, 1,000-m, and watershed scales using the same methodology as the current land cover.

Selection and delineation of biologically relevant subsets

We created three potentially biologically relevant subsets of variables to use in analyses of prairie fens occupied and unoccupied by Poweshiek skipperling. We used these variable sets to describe conditions that could be critical to Poweshiek's life cycle and survival. The three variable sets were 1) conditions for suspected larval host plants (mat muhly and prairie dropseed), 2) system integrity, and 3) agricultural influence (Table 1).

Table 1

Untransformed mean (± 2 SE) of 20 variables collected from 18 Michigan prairie fens from 2012 through 2015. Prairie fens are grouped into two mutually exclusive categories: occupied (n = 9) and unoccupied (n = 9). Occupied fens have documentation of Poweshiek skipperling Oarisma poweshiek occupancy within the past 20 y. Unoccupied fens have been surveyed for Poweshiek skipperling and determined to not be occupied by the species. Each parameter is included in one of three biologically relevant parameter subsets for analysis in determining Poweshiek skipperling occupancy of Michigan prairie fens: conditions for larval hosts, system integrity, and agricultural influence. Parameter subsets were used in nonmetric multidimensional scaling and multiresponse permutation procedure analyses to visualize and determine levels of grouping between occupied vs. unoccupied prairie fens.

Untransformed mean (± 2 SE) of 20 variables collected from 18 Michigan prairie fens from 2012 through 2015. Prairie fens are grouped into two mutually exclusive categories: occupied (n = 9) and unoccupied (n = 9). Occupied fens have documentation of Poweshiek skipperling Oarisma poweshiek occupancy within the past 20 y. Unoccupied fens have been surveyed for Poweshiek skipperling and determined to not be occupied by the species. Each parameter is included in one of three biologically relevant parameter subsets for analysis in determining Poweshiek skipperling occupancy of Michigan prairie fens: conditions for larval hosts, system integrity, and agricultural influence. Parameter subsets were used in nonmetric multidimensional scaling and multiresponse permutation procedure analyses to visualize and determine levels of grouping between occupied vs. unoccupied prairie fens.
Untransformed mean (± 2 SE) of 20 variables collected from 18 Michigan prairie fens from 2012 through 2015. Prairie fens are grouped into two mutually exclusive categories: occupied (n = 9) and unoccupied (n = 9). Occupied fens have documentation of Poweshiek skipperling Oarisma poweshiek occupancy within the past 20 y. Unoccupied fens have been surveyed for Poweshiek skipperling and determined to not be occupied by the species. Each parameter is included in one of three biologically relevant parameter subsets for analysis in determining Poweshiek skipperling occupancy of Michigan prairie fens: conditions for larval hosts, system integrity, and agricultural influence. Parameter subsets were used in nonmetric multidimensional scaling and multiresponse permutation procedure analyses to visualize and determine levels of grouping between occupied vs. unoccupied prairie fens.

Larval hosts

We assembled variables to describe the potential conditions for mat muhly and prairie dropseed, referencing multiple oviposition events documented on mat muhly in Michigan prairie fens (Pointon 2015), and there has been documented larval feeding on prairie dropseed in Wisconsin mesic prairies (Borkin 1995). The potential reliance upon delicately structured graminoids could limit Poweshiek skipperling success at the larval stage, which is the longest stage of the life cycle. This variable subset consisted of soil calcium concentration, soil pH, mean coefficient of conservatism, and historical proportions of wet prairie surrounding prairie fens at the watershed and 100-m scales. We included soil calcium, as these species are known to grow in calcareous environments in Michigan (Higman and Penskar 1999; Penksar and Higman 1999). Mat muhly and prairie dropseed also grow in slightly alkaline soils, so we included soil pH (Higman and Penskar 1999; Penksar and Higman 1999). We included mean coefficient of conservatism as these species are threatened and of special concern in Michigan (Reznicek et al. 2011). To incorporate connectivity and dispersal of these species, we included historical proportion of wet prairie. We selected the watershed and 100-m scales to address the local and landscape scales while mitigating spatial autocorrelation.

System integrity

We created the system integrity data set to address habitat quality, anthropogenic influence, and fragmentation. Rare species such as Poweshiek skipperling, and the grasses upon which it potentially relies, often require high-quality habitats with less human disturbance. Saarinen et al. (2016) suggested that fragmentation has contributed to reduction of genetic diversity and overall survivability of the species but noted that fragmentation alone was not likely to explain the declines observed in recent years. We wanted to examine this further to investigate if fragmentation was associated with Poweshiek skipperling occupancy in prairie fens over the past 20 y. We selected mean distance to the nearest three prairie fens, ERA, FQI of the fen, and the proportion of surrounding natural land cover within 100 m and 1,000 m for this partition. Mean distance to nearest three fens and the proportion of surrounding natural land cover were used as indicators of human disturbance and fragmentation at the landscape scale. Exotic relative abundance serves as an indicator of human degradation and invasive species within the fen, whereas FQI provides an index of the quality and composition of the plant community at each fen.

Agricultural influence

We created the agricultural influence data set to address the potential impacts of pesticide application, fragmentation, and nutrient loading upon the species. It has been documented that the neonicotinoid class of pesticides has negatively affected other insects and has been proposed as a potential cause for decline of the Poweshiek skipperling (Poweshiek Working Group 2011; Godfray et al. 2014). We used soil concentrations of ammonium, nitrates, nitrogen, phosphorus, and potassium, as well as water concentrations of soluble reactive phosphorous, nitrates, and ammonia in this set of variables, as these are commonly found as components of fertilizers and elevated nutrient levels could indicate increased agricultural pressure on the wetlands. To characterize the surrounding land use, we included the proportion of surrounding row crop agriculture at the 100-m and 1,000-m scales to address local and landscape scales. For the larger scale, we selected 1,000 m instead of watershed-scale buffer to more directly address the potential for wind-borne pesticides that might not have been as well indicated at the watershed scale because of the irregular shape of the fens and the watersheds.

Characteristics of occupied and unoccupied prairie fens

Multiresponse permutation procedures

We used multiresponse permutation procedures to determine the probability that occupied and unoccupied prairie fens were different groups based on conditions for larval host plants, variables representing system integrity, and indicators of agricultural influence. We used Bray–Curtis distance matrices in our calculations, as this has been demonstrated to be effective when examining community data (McArdle and Anderson 2001). Multiresponse permutation procedure provides the significance of delta, A, and t-test statistics. Significance of delta, represented by a P-value, gives the probability that the two groups are different because of chance. We a priori determined a significance of delta cutoff at 0.10 to minimize type II errors (Watała 2007). We completed this analysis using PC-ORD 6.21 (McCune and Mefford 2011).

Nonmetric multidimensional scaling

To visualize how sites grouped on the basis of conditions for larval hosts, system integrity, and agricultural influence, we reduced the dimensionality using nonmetric multidimensional scaling (NMDS). We performed NMDS using the Bray–Curtis distance measure. We set the maximum number of axes to three to minimize stress while allowing for more simple visualization of patterns. To visualize the patterns, we examined NMDS ordinations in two dimensions with all possible permutations (i.e., NMDS1 vs. NMDS2; NMDS1 vs. NMDS3; NMDS2 vs. NMDS3). We plotted each of the variables from each partition (e.g., system integrity) and each of the sites (n = 18) on the NMDS axes. To perform and visualize the NMDS, we used the “metaMDS” and “ordiplot” functions, respectively, within the “vegan” package in R statistical software (Oksanen et al. 2015).

Plant indicator species analysis

We used indicator species analysis to identify plant species that characterize Poweshiek skipperling habitat at whole-fen and intrafen scales (Hill et al. 1975; Dufrêne and Legendre 1997). Indicator species analysis compares species occurrence or abundance data between two or more user-defined groups to identify group-characteristic species. We extracted species occurrences from the plant quadrat data for occupied and unoccupied fens. To identify indicator species at the whole-fen scale, we compared species occurrence data from nine occupied and nine unoccupied prairie fens.

At the intrafen scale, we examined only those prairie fens with Poweshiek skipperling populations. We categorized each quadrat by its proximity to Poweshiek skipperling observations. We used Poweshiek skipperling occurrence data from 2009 to 2015 to calculate 30-m buffers around each observation. We considered quadrat locations within these buffers to represent occupied Poweshiek skipperling habitat. We identified 70 quadrats within the buffer zones. One prairie fen did not have any quadrat locations within the Poweshiek skipperling buffer zone.

We used the “multipatt” function from the “indicspecies” package in R programming language to compute this analysis (De Caceres and Legendre 2009; R Core Team 2016). We selected to use the “IndVal.g” index for both analyses as it is best suited for the occurrence data and includes the sample size correction (De Cáceres et al. 2010). This analysis assigns an indicator value for each species. Perfect indicators (index value = 1) are exclusive to one category and occur frequently within the category, whereas species found exactly equally among sites are poor indicators and have an index value of 0. Dufrêne and Legendre (1997) suggest a cutoff of 0.25 for indicator values. The significance of the indicator value is calculated through a permutation test and is reported as a P-value. Because of our large sample size (n = 588 quadrats), we set an a priori alpha cutoff at 0.05. For each indicator species, we obtained coefficients of conservatism and wetness, along with wetness indices from Michigan Flora Online (Reznicek et al. 2011).

In 2016, we detected 104 Poweshiek skipperling individuals across four sites categorized as occupied and failed to detect Poweshiek in five sites categorized as occupied where they previously occurred. We did not observe Poweshiek skipperling at any site categorized as unoccupied. Poweshiek skipperling counts in 2016 were similar to those from 2015 (Pogue et al. 2016; Table 2).

Table 2

Poweshiek skipperling Oarisma poweshiek survey results from 2014 through 2016 at locations with at least one positive detection of Poweshiek skipperling within the past 20 y in Michigan. Bottom row indicates the last year Poweshiek skipperling were detected at each site.

Poweshiek skipperling Oarisma poweshiek survey results from 2014 through 2016 at locations with at least one positive detection of Poweshiek skipperling within the past 20 y in Michigan. Bottom row indicates the last year Poweshiek skipperling were detected at each site.
Poweshiek skipperling Oarisma poweshiek survey results from 2014 through 2016 at locations with at least one positive detection of Poweshiek skipperling within the past 20 y in Michigan. Bottom row indicates the last year Poweshiek skipperling were detected at each site.

Across the 18 sites from 2013 through 2015, we sampled a total of 851 quadrats for plant biodiversity, which resulted in 380 observed species and 1,507 vouchered specimens (Hackett et al. 2016; http://midwestherbaria.org/portal/projects/index.php?pid=113). We documented 213 species in common between occupied and unoccupied prairie fens. We found 54 species unique to occupied and 66 species unique to unoccupied prairie fens. We calculated a Sørensen similarity index of 0.780 between groups of occupied and unoccupied sites.

Characteristics of occupied and unoccupied prairie fens

Multiresponse permutation procedures

Multiresponse permutation procedure results indicated that the occupied and unoccupied prairie fens grouped separately on the basis of conditions for mat muhly and prairie dropseed (P = 0.093, A = 0.065, t = −1.383). Occupied and unoccupied prairie fens did not group separately on the basis of variables used as indicators of system integrity (P = 0.461, A = −0.011, t = 0.250). Also, occupied and unoccupied prairie fens did not group separately for agricultural influence (P = 0.244, A = 0.06453, t = −0.536; Table 3).

Table 3

Test statistics from multiresponse permutation procedure used to compare Poweshiek skipperling Oarisma poweshiek occupied and unoccupied prairie fens in Michigan. Data were collected in 2012–2015.

Test statistics from multiresponse permutation procedure used to compare Poweshiek skipperling Oarisma poweshiek occupied and unoccupied prairie fens in Michigan. Data were collected in 2012–2015.
Test statistics from multiresponse permutation procedure used to compare Poweshiek skipperling Oarisma poweshiek occupied and unoccupied prairie fens in Michigan. Data were collected in 2012–2015.

Nonmetric multidimensional scaling

Visualization of the data sets in a two-dimensional space did not reveal any clustering of sites by occupancy status for any of the parameter sets or permutations of the NMDS-generated axes. Minimum stress values were 0.013 for mat muhly and prairie dropseed conditions, 0.038 for system integrity, and 0.087 for agricultural influence. In the mat muhly and prairie dropseed NMDS, proportion of historical wet prairie in the watershed was the strongest factor (loading = 0.57 along NMDS2; Figure 4; Table 4). Proportion of historical wet prairie within the 100-m buffer was a weaker factor, with the opposite sign (NMDS2 loading = −0.18). In the system integrity NMDS, ERA (NMDS1 loading = −0.23), proportion of natural land cover within the 100-m buffer (NMDS2 loading = 0.19), and mean distance to nearest three prairie fens were the strongest factors (NMDS1 loading = −0.18; Figure 5; Table 4). Also, mean distance to nearest three prairie fens was the strongest factor for NMDS3; however, the loading value was lower than those on other axes (NMDS3 loading = 0.14). In the agricultural influence NMDS, proportions of crops within the 100-m buffer was the dominant factor for all three NMDS axes (loadings = −0.39, 0.27, and −0.15, respectively; Figure 6; Table 4). Other, weaker factors include soluble reactive phosphorus (NMDS1 and NMDS3 loadings = 0.16 and −0.14, respectively), nitrates in water (NMDS1 loading = −0.23), ammonia in water (NMDS2 loading = −0.15), and proportions of crops within 1,000-m (NMDS2 loading = −0.19).

Figure 4

Nonmetric multidimensional scaling (NMDS) ordinations of 18 prairie fen sites in Michigan based on conditions for mat muhly Muhlenbergia richardsonis and prairie dropseed Sporobolus heterolepis. Poweshiek skipperling Oarisma poweshiek-occupied prairie fens are represented by plus signs and unoccupied fens are represented by open circles. The numbers indicate the variable loadings on the axes. Soil calcium = 1, soil pH = 2, mean coefficient of conservatism = 3, proportion of historical wet prairie within the watershed = 4, proportion of historical wet prairie within 100-m buffer = 5. Axes vary and include all permutations of two-dimensional ordinations (i.e., NMDS1 vs. NMDS2; NMDS1 vs. NMDS 3; NMDS2 vs. NMDS3). Data were collected in 2012–2015.

Figure 4

Nonmetric multidimensional scaling (NMDS) ordinations of 18 prairie fen sites in Michigan based on conditions for mat muhly Muhlenbergia richardsonis and prairie dropseed Sporobolus heterolepis. Poweshiek skipperling Oarisma poweshiek-occupied prairie fens are represented by plus signs and unoccupied fens are represented by open circles. The numbers indicate the variable loadings on the axes. Soil calcium = 1, soil pH = 2, mean coefficient of conservatism = 3, proportion of historical wet prairie within the watershed = 4, proportion of historical wet prairie within 100-m buffer = 5. Axes vary and include all permutations of two-dimensional ordinations (i.e., NMDS1 vs. NMDS2; NMDS1 vs. NMDS 3; NMDS2 vs. NMDS3). Data were collected in 2012–2015.

Close modal
Table 4

Nonmetric multidimensional scaling parameter loadings for three axes in three data sets that are biologically relevant to Poweshiek skipperling Oarisma poweshiek. Data were collected in 2012–2015 from prairie fens in Michigan.

Nonmetric multidimensional scaling parameter loadings for three axes in three data sets that are biologically relevant to Poweshiek skipperling Oarisma poweshiek. Data were collected in 2012–2015 from prairie fens in Michigan.
Nonmetric multidimensional scaling parameter loadings for three axes in three data sets that are biologically relevant to Poweshiek skipperling Oarisma poweshiek. Data were collected in 2012–2015 from prairie fens in Michigan.
Figure 5

Nonmetric multidimensional scaling (NMDS) ordinations of 18 prairie fen sites in Michigan based on system integrity. Poweshiek skipperling Oarisma poweshiek-occupied prairie fens are represented by plus signs and unoccupied fens are represented by open circles. The numbers indicate the variable loadings on the axes. Proportion of natural land cover within 100-m buffer = 1, proportion of natural land cover within 1,000-m buffer = 2, adjusted floristic quality index = 3, exotic relative abundance = 4, mean distance to nearest three prairie fens = 5. Axes vary and include all permutations of two-dimensional ordinations (i.e., NMDS1 vs. NMDS2; NMDS1 vs. NMDS 3; NMDS2 vs. NMDS3). Data were collected in 2012–2015.

Figure 5

Nonmetric multidimensional scaling (NMDS) ordinations of 18 prairie fen sites in Michigan based on system integrity. Poweshiek skipperling Oarisma poweshiek-occupied prairie fens are represented by plus signs and unoccupied fens are represented by open circles. The numbers indicate the variable loadings on the axes. Proportion of natural land cover within 100-m buffer = 1, proportion of natural land cover within 1,000-m buffer = 2, adjusted floristic quality index = 3, exotic relative abundance = 4, mean distance to nearest three prairie fens = 5. Axes vary and include all permutations of two-dimensional ordinations (i.e., NMDS1 vs. NMDS2; NMDS1 vs. NMDS 3; NMDS2 vs. NMDS3). Data were collected in 2012–2015.

Close modal
Figure 6

Nonmetric multidimensional scaling (NMDS) ordinations of 18 prairie fen sites in Michigan based on agricultural influence. Poweshiek skipperling Oarisma poweshiek-occupied prairie fens are represented by plus signs and unoccupied fens are represented by open circles. The numbers indicate the variable loadings on the axes. Soil nitrogen = 1, soil phosphorus = 2, soil potassium = 3, soil nitrates = 4, soil ammonium nitrates = 5, soluble reactive phosphorus in water = 6, nitrates in water = 7, ammonia in water = 8, proportion of crops within 100-m buffer = 9, proportion of crops within 1,000-m buffer=10. Axes vary and include all permutations of two-dimensional ordinations (i.e., NMDS1 vs. NMDS2; NMDS1 vs. NMDS 3; NMDS2 vs. NMDS3). Data were collected in 2012–2015.

Figure 6

Nonmetric multidimensional scaling (NMDS) ordinations of 18 prairie fen sites in Michigan based on agricultural influence. Poweshiek skipperling Oarisma poweshiek-occupied prairie fens are represented by plus signs and unoccupied fens are represented by open circles. The numbers indicate the variable loadings on the axes. Soil nitrogen = 1, soil phosphorus = 2, soil potassium = 3, soil nitrates = 4, soil ammonium nitrates = 5, soluble reactive phosphorus in water = 6, nitrates in water = 7, ammonia in water = 8, proportion of crops within 100-m buffer = 9, proportion of crops within 1,000-m buffer=10. Axes vary and include all permutations of two-dimensional ordinations (i.e., NMDS1 vs. NMDS2; NMDS1 vs. NMDS 3; NMDS2 vs. NMDS3). Data were collected in 2012–2015.

Close modal

Plant indicator species analysis

At the whole-fen scale, we found 14 indicator species that characterized occupied fens (Table 5). Indicator values for statistically significant species (P ≤ 0.05) ranged from 0.745 to 0.934. No species were identified as characteristic of unoccupied fens.

Table 5

Whole-fen-level indicator species that characterize Poweshiek skipperling Oarisma poweshiek-occupied prairie fens in Michigan, 2012–2015. * indicates a nonnative species and therefore, no coefficient of conservatism in Michigan.

Whole-fen-level indicator species that characterize Poweshiek skipperling Oarisma poweshiek-occupied prairie fens in Michigan, 2012–2015. * indicates a nonnative species and therefore, no coefficient of conservatism in Michigan.
Whole-fen-level indicator species that characterize Poweshiek skipperling Oarisma poweshiek-occupied prairie fens in Michigan, 2012–2015. * indicates a nonnative species and therefore, no coefficient of conservatism in Michigan.

No species were identified to be characteristic of Poweshiek skipperling-occupied areas within prairie fens. We identified 52 indicator species that characterized non-Poweshiek skipperling areas within occupied prairie fens (Table 6). Indicator values for statistically significant species (P ≤ 0.05) ranged from 0.760 to 1.

Table 6

Intrafen indicator species that characterize areas not occupied by Poweshiek skipperling Oarisma poweshiek within prairie fens in Michigan, 2012–2015.

Intrafen indicator species that characterize areas not occupied by Poweshiek skipperling Oarisma poweshiek within prairie fens in Michigan, 2012–2015.
Intrafen indicator species that characterize areas not occupied by Poweshiek skipperling Oarisma poweshiek within prairie fens in Michigan, 2012–2015.

To better understand if habitat variables are associated with patterns of occurrence in Poweshiek skipperling in Michigan, we explored patterns in parameters potentially relevant to the species' life cycle and environmental conditions between occupied and unoccupied sites. Overall, we found similar patterns between occupied and unoccupied fens in the three suites of variables examined as designed in this study (larval hosts, system integrity, and agricultural influence). Although multiresponse permutation procedures indicated potential differences between occupied and unoccupied fens in variables characterizing conditions for larval host plants, this difference, although significant (P = 0.093), was not large (t = −1.38), nor supported by patterns observed in NMDS analysis. We did not find significant differences in multiresponse permutation procedures comparisons of the system integrity and agricultural influence variable sets. This was consistent with our NMDS analyses.

The lack of separation in our multivariate analyses of habitat conditions at occupied and unoccupied sites in Michigan could be due to several reasons: 1) inadequate sample size; 2) remaining sites represent low-quality habitats for Poweshiek skipperling; 3) the species is responding to environmental/habitat variables that we did not measure; or 4) factors other than habitat are driving occupancy patterns and population declines. Given the extreme rarity of Poweshiek skipperling, we only had a small number of occupied fens available for study. Knowing the limited distribution of the species, we attempted to detect any previously undetected populations that would have increased our sample size and likelihood of pattern detection (Pogue et al. 2016). Despite comprehensive and expanded surveys in 16 previously unsurveyed areas presumed to have habitat characteristics beneficial to Poweshiek skipperling, no new populations were detected. Similarly, the intensive survey effort required to assess plant diversity and other conditions, as well as confirm sites as unoccupied, further limited the sample size. The small sample size could have reduced our ability to detect differences between occupied and unoccupied sites. Further, to be a valid comparison, our unoccupied sites needed to possess potential habitat (e.g., known or potential host plants present), which inherently made the unoccupied sites similar to occupied fens.

Although prairie fens in Michigan represent the remaining “stronghold” for the species across its range, we cannot assume that these are indicative of high-quality conditions for Poweshiek skipperling. As Landis et al. (2012) suggested, Poweshiek skipperling occupancy of fens in Michigan could be because they are the only communities remaining that possess the prairie plant species necessary for survival (e.g., host plants). For example, prairie fens historically occurred within landscapes having substantial areas of adjacent oak barrens containing prairie plant communities. Poweshiek skipperlings may have used these fen–barrens complexes, rather than fens exclusively. Unfortunately, the ecology of the species under pre-European settlement conditions is unknown. Additional habitat or environmental factors have the potential to influence the occurrence of Poweshiek skipperling such as the size, distribution, and spatial configuration of fen zones (i.e., sedge meadow, inundated flats, wooded, and groundwater seeps) and structural categories (e.g., herbaceous vs. woody). More research is needed to explore if and how the spatial complexities of prairie fens are affecting the species' occurrence among and within fens.

Insect–plant codependence in the form of flower pollination, larval hosts, and adult insect nectar sources is well documented (Jermy 1984; Fisher 1998; Hamilton 2004; Davis et al. 2008). Additionally, plant communities often contribute to defining habitat for species, help indicate biotic and abiotic processes, and are used in many studies as a proxy for system integrity (Tilman et al. 1997; Young 2000; Lopez and Fennessy 2002; Noss 2005; Ruiz-Jaén and Aide 2005; Hackett at al. 2016). We further examined potential plant and habitat associations using indicator species analyses to determine if particular plant species were associated with occupied and unoccupied fens and occupied and unoccupied areas within wetlands occupied by Poweshiek skipperling. Indicator species analysis at the whole-fen level indicated that 14 species were associated with occupied prairie fens (Table 5). Two of these species, mat muhly and little bluestem, have been documented to support larval feeding of Poweshiek skipperling or oviposition by female adults (Borkin 1995; Pointon 2015). This result supports previous predictions or observations that these are among the larval host species, but further work is needed to confirm them as larval host plants and determine the characteristics that make them suitable for this species. If larval host plants are found to be a limiting factor in Poweshiek skipperling success, the presence and amount of these grasses would need to be considered when making management decisions and selecting sites for introduction, as they could indicate suitable habitat. Another indicator species, sticky tofieldia Triantha glutinosa, has been confirmed as an adult nectar source and could also be useful in identifying potentially suitable habitat (Cuthrell and Slaughter 2012). Although Poweshiek skipperling require nectar, availability of nectar sources does not appear to be lacking, with many individuals of known nectar species observed relative to the small number of Poweshiek skipperling detected.

We found 52 plant species associated with unoccupied portions of occupied fens but could not determine indicator species for occupied areas. This aligns with field observations that some areas in prairie fens resemble occupied patches (e.g., open areas; abundance of nectar species; presence of larval host plants) but do not contain Poweshiek skipperling (unpublished data). Surveys for Poweshiek skipperling could be streamlined by excluding areas characterized by these plant species. Further work examining the structure and distribution of these nonsuitable areas could provide insight into fine-scale habitat requirements of this species. Coupled with habitat investigations, observational studies of Poweshiek skipperling behavior, including butterfly mating patterns, could give clues as to how this species uses prairie fens and identify undetected key habitat and plant associations. Although our indicator species analyses suggested some association between Poweshiek skipperling occupancy and plant assemblages, we calculated a Sørensen similarity index of 0.780 between occupied and unoccupied fens, which indicates that at the whole-fen level, the occupied and unoccupied fens share many of the same species. More work is needed to better understand the species' interactions with fen plant communities, such as determining the plant species used for oviposition, larval feeding, and adult nectaring, and investigating the influence of plant composition, structure, and spatial configuration on adult movement and behavior within prairie fens.

Our results could indicate that factors other than habitat are driving Poweshiek skipperling distribution and population declines in the study area. In addition to habitat loss and degradation, a variety of factors has been posited as reasons for the recent sharp decline in populations, including contamination, climate change, disease, and low levels of genetic diversity (Landis et al. 2012; Poweshiek Skipperling Working Group 2011; Saarinen 2016). Ongoing research is being conducted to evaluate the influence of these potential factors on the occurrence of Poweshiek skipperling, such as associations between occupancy and climatic and landscape (e.g., agricultural practices) factors over time (Belitz et al. 2019).

With the dramatic reduction of Poweshiek skipperling populations over the past 20 y, determining the habitat requirements as well as the cause or causes for the decline are essential to developing strategies for conservation of the species. We view this study as the first step in a long-term research effort to better understand the ecology of Poweshiek skipperling and inform conservation strategies. Additional studies are needed to increase knowledge about the species' biology, examine habitat associations, investigate potential associations between range-wide distribution changes, and investigate large-scale environmental and land-use patterns. Future research potential includes identifying potential stressors such as contaminants, intrafen zonal changes, genetic bottlenecks, and climate change that could be contributing to the decline of the species. Additionally, refined understanding of habitat structure including stem density, percent cover, litter composition, and vegetation height in these areas might reveal requirements or additional characteristics of Poweshiek skipperling habitat within prairie fens. This understanding could help the evaluation of potentially suitable habitat for population augmentation and reintroduction efforts. Insights from the ecology of Poweshiek skipperling have the potential to inform other studies looking at drivers of diversity in the vegetation zones of prairie fens, other rare and threatened animals, ecosystem threats, and overall pollinator declines.

Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any supplemental material. Queries should be directed to the corresponding author for the article.

Data S1. Interfen plant incidence data resultant from quadrat-level surveys for 18 prairie fens in Michigan. Data were collected from 2012 to 2015. Prairie fens are grouped into two mutually exclusive categories: occupied (“op”; n = 9) and unoccupied (“nop”; n = 9). Occupied fens have documentation of Poweshiek skipperling Oarisma poweshiek occupancy within the past 20 y. Unoccupied fens have been surveyed for Poweshiek skipperling and determined to not be occupied by the species.

Found at DOI: https://doi.org/10.3996/122018-JFWM-117.S11 (22 KB XLSX).

Data S2. Intrafen plant incidence data resultant from quadrat-level surveys for eight prairie fens occupied by Poweshiek skipperling Oarisma poweshiek in Michigan. Occupied fens have documentation of Poweshiek skipperling occupancy within the past 20 y. Data were collected from 2012 to 2015. Data are grouped into two mutually exclusive categories: occupied portions (delineated by “_1” at the end of the identifier; n = 8) and unoccupied portions (delineated by “_0” at the end of the identifier; n = 8). We categorized each quadrat by its proximity to Poweshiek skipperling observations. We used Poweshiek skipperling occurrence data from 2009 to 2015 to calculate 30-m buffers around each observation. We considered quadrat locations within these buffers to represent occupied portions of Poweshiek skipperling habitat. We identified 70 quadrats within the buffer zones.

Found at DOI: https://doi.org/10.3996/122018-JFWM-117.S2 (26 KB XLSX).

Data S3. Prairie fen-level biotic and abiotic factors for 18 prairie fens in Michigan. Data were collected from 2012 to -2015. Prairie fens are grouped into two mutually exclusive categories: occupied (“op”; n = 9) and unoccupied (“nop”; n = 9). Occupied fens have documentation of Poweshiek skipperling Oarisma poweshiek occupancy within the past 20 y. Unoccupied fens have been surveyed for Poweshiek skipperling and determined to not be occupied by the species.

Found at DOI: https://doi.org/10.3996/122018-JFWM-117.S3 (24 KB XLSX).

Reference S1. Borkin SS. 1995. 1994 Ecological studies of the Poweshiek skipper (Oarisma poweshiek) in Wisconsin. Milwaukee Public Museum, Milwaukee, Wisconsin.

Found at DOI: https://doi.org/10.3996/122018-JFWM-117.S4 (1.33 MB PDF).

Reference S2. Cuthrell DL, Slaughter SL. 2012. Special animal abstract for Oarisma poweshiek (Poweshiek skipperling). Michigan Natural Features Inventory, Lansing, Michigan.

Found at DOI: https://doi.org/10.3996/122018-JFWM-117.S5 (1.28 MB PDF); also available at http://mnfi.anr.msu.edu/abstracts/zoology/Oarisma_poweshiek.pdf

Reference S3. Hackett RA. 2013. Prairie fen plant biodiversity: the influence of site- and landscape-level factors on plant diversity. Master's thesis. Mount Pleasant: Central Michigan University.

Found at DOI: https://doi.org/10.3996/122018-JFWM-117.S6 (2.29 MB PDF).

Reference S4. Herold N. 2014. C-CAP Land Cover Classification Scheme. Silver Spring, Maryland: National Oceanic and Atmospheric Administration.

Found at DOI: https://doi.org/10.3996/122018-JFWM-117.S7 (1.16 MB PDF); also available at https://www.canr.msu.edu/uploads/234/62936/CLIMATE_CHANGE.pdf

Reference S5. Higman PJ, PenskarMR. 1999. Special plant abstract for Sporobolus heterolepis (prairie dropseed). Lansing, Michigan: Michigan Natural Features Inventory.

Found at DOI: https://doi.org/10.3996/122018-JFWM-117.S8 (214 KB PDF); also available at https://mnfi.anr.msu.edu/abstracts/botany/Sporobolus_heterolepis.pdf

Reference S6. Hoving CL, Lee YM, Badra PJ, Klatt BJ. 2013. Changing climate, changing wildlife—a vulnerability assessment of 400 species of greatest conservation need and game species in Michigan. Michigan Department of Natural Resources Wildlife Division Report No. 3564.

Found at DOI: https://doi.org/10.3996/122018-JFWM-117.S9 (867 KB PDF); also available at https://www.canr.msu.edu/uploads/234/62936/CLIMATE_CHANGE.pdf

Reference S7. Kost MA, Albert DA, Cohen JG, Slaughter BS, Schillo RK, Weber CR, Chapman KA. 2007. Natural communities of Michigan: classification and description. Michigan Natural Features Inventory, Report No. 2007-21, Lansing, Michigan.

Found at DOI: https://doi.org/10.3996/122018-JFWM-117.S10 (1.57 MB PDF); also available at https://mnfi.anr.msu.edu/reports/MNFI-Report-2007-21.pdf

Reference S8. McCabe TL. 1974. Skippers (Hesperioidea) of North Dakota. Master's thesis. Fargo: North Dakota State University.

Found at DOI: https://doi.org/10.3996/122018-JFWM-117.S11 (2.9 MB PDF).

Reference S9. Penskar MR, Higman PJ. 1999. Special plant abstract for Muhlenbergia richardsonis (mat muhly). Michigan Natural Features Inventory, Lansing, Michigan.

Found at DOI: https://doi.org/10.3996/122018-JFWM-117.S12 (178 KB PDF); also available at https://mnfi.anr.msu.edu/abstracts/botany/Muhlenbergia_richardsonis.pdf

Reference S10. Poweshiek Working Group. 2011. Minutes. Dupont.J, editor. Winnipeg, Manitoba: Poweshiek skipperling workshop.

Found at DOI: https://doi.org/10.3996/122018-JFWM-117.S13 (1.82 MB PDF); also available at http://www.poweshiekskipper.org/project/Images/Final_POSK_Workshop_2011.pdf

Reference S11. Runquist E, Heimpel GE. 2017. Potential causes of declines in Minnesota's prairie butterflies with a focus on insecticidal control of the soybean aphid. Prepared for College of Food, Agricultural and Natural Resource Sciences–University of Minnesota.

Found at DOI: https://doi.org/10.3996/122018-JFWM-117.S14 (1.62 MB PDF); also available: https://mitppc.dl.umn.edu/sites/g/files/pua746/f/media/mitppc_soybean.final.pdf

Reference S12. Selby G. 2005. Status assessment and conservation guidelines: Poweshiek skipperling (Oarisma poweshiek (Parker) [Lepidoptera: Hesperiidae]). Bloomington, Minnesota: Twin Cities Field Office, U.S. Fish and Wildlife Service.

Found at DOI: https://doi.org/10.3996/122018-JFWM-117.S15 (5.83 MB PDF); also available at http://www.fws.gov/midwest/es/soc/insects/posk_sa.pdf

Reference S13. [SARA] Species at Risk Act, Revised Statutes of Canada (2002, c. 29)

Found at DOI: https://doi.org/10.3996/122018-JFWM-117.S16 (918 KB PDF); also available at http://laws-lois.justice.gc.ca/PDF/S-15.3.pdf

Reference S14. Spieles JB, Comer PJ, Albert DA, Kost MA. 1999. Natural community abstract for prairie fen. Michigan Natural Features Inventory, Lansing, Michigan.

Found at DOI: https://doi.org/10.3996/122018-JFWM-117.S17 (2.85 MB PDF); also available at http://mnfi.anr.msu.edu/abstracts/ecology/prairie_fen.pdf

Reference S15. [USEPA] U.S. Environmental Protection Agency. 1994. National Water Quality Inventory: 1992 Report to Congress. Office of Water. Washington, D.C.: U.S. Environmental Protection Agency.

Found at DOI: https://doi.org/10.3996/122018-JFWM-117.S18 (41.12 MB PDF).

The funding for this project was provided by the U.S. Fish and Wildlife Service, Great Lakes Restoration Initiative through the U.S. Fish and Wildlife Service's Endangered Species Program, Central Michigan University Biology Department, Sigma Xi, and Prairie Biotics Research, Inc. We thank Michael Belitz, who reviewed this heavily to make sure all comments were compatible with ongoing and in-prep efforts. We also thank The Nature Conservancy for supporting field researchers through housing and review of methodologies. Funding also came from the National Science Foundation award DBI-1730526 to develop educational materials for biodiversity data literacy (https://www.biodiversityliteracy.com/poweshiek-skipperling). Additionally, we express deep gratitude to the Associate Editor and the anonymous reviewers for their constructive comments. We also thank the landowners, including the Michigan Nature Association, Oakland County, Springfield Township, Michigan Department of Natural Resources, and private landowners for allowing us access to their properties for survey activities. Finally, we are grateful to the members of the Partnership for Poweshiek Skipperling Conservation for their continued long-term dedication to this species. This paper is Contribution No. 129 of the Central Michigan University Institute for Great Lakes Research.

Any use of trade, product, website, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the U.S. Government.

Belitz
MW,
Hendrick
L,
Monfils
MJ,
Cuthrell
DL,
Marshall
CJ,
Kawahara
AY,
Cobb
NS,
Zaspel
JM,
Horton
AM,
Huber
SL,
Warren
AD,
Forthaus
GA,
Monfils
AK.
2018
.
Aggregated occurrence records of the federally endangered Poweshiek skipperling (Oarisma poweshiek). Biodiversity Data Journal 6:e29081
.
Belitz
MW,
Monfils
MJ,
Cuthrell
DL,
Monfils
AK.
2019
.
Life history and ecology of the endangered Poweshiek skipperling Oarisma poweshiek in Michigan prairie fens
.
Journal of Insect Conservation
23
:
635
.
Borkin
SS.
1995
.
1994 Ecological studies of the Poweshiek skipper (Oarisma poweshiek) in Wisconsin
.
Milwaukee Public Museum
,
Milwaukee, WI
(
see Supplemental Material, Reference S1
).
Comer
P,
Albert
D.
1997
.
Michigan Nature Features Inventory (retrieved from vegetation circa 1800 maps
: .
[COSEWIC] Committee on the Status of Endangered Wildlife in Canada
.
2014
.
Poweshiek skipperling
. .
Cuthrell
DL,
Slaughter
SL.
2012
.
Special animal abstract for Oarisma poweshiek (Poweshiek skipperling)
.
Michigan Natural Features Inventory
,
Lansing, Michigan
(
see Supplemental Material, Reference S2
).
Daubenmire
R.
1959
.
A canopy-coverage method of vegetational analysis
.
Northwest Science
33
:
43
64
.
Davis
JD,
Debinski
DM,
Danielson
BJ.
2007
.
Local and landscape effects on the butterfly community in fragmented Midwest USA prairie habitats
.
Landscape Ecology
22
:
1341
1354
.
Davis
JD,
Hendrix
SD,
Debinski
DM,
Hemsley
CJ.
2008
.
Butterfly, bee and forb community composition and cross-taxon incongruence in tallgrass prairie fragments
.
Journal of Insect Conservation
12
:
69
79
.
De Caceres
M,
Legendre
P.
2009
.
Associations between species and groups of sites: indices and statistical inference
.
Ecology
90
:
3566
3574
.
De Cáceres,
Legendre
P,
Moretti
M.
2010
.
Improving indicator species analysis by combining groups of sites
.
Oikos
119
:
1674
1684
.
Dearborn
K,
Westwood
R.
2014
.
Predicting adult emergence of Dakota skipper and Poweshiek skipperling (Lepidoptera: Hesperiidae) in Canada
.
Journal of Insect Conservation
,
18
:
875
884
.
Donald
PF,
Evans
AD.
2006
.
Habitat connectivity and matrix restoration: the wider implications of agri-environment schemes
.
Journal of Applied Ecology
43
:
209
218
.
Dover
J,
Sotherton
N,
Gobbett
K.
1990
.
Reduced pesticide inputs on cereal field margins: the effects on butterfly abundance
.
Ecological Entomology
15
:
17
24
.
Dufrêne
M,
Legendre
P.
1997
.
Species assemblages and indicator species: the need for a flexible asymmetrical approach
.
Ecological Monographs
67
:
345
366
.
ESRI
.
2014
.
ArcGIS Desktop: Release 10.2.2
.
Redlands, California
:
Environmental Systems Research Institute
.
Fisher
BL.
1998
.
Insect behavior and ecology in conservation: preserving functional species interactions
.
Annals of the Entomological Society of America
91
:
155
158
.
Godfray
HCJ,
Blacquière
T,
Field
LM,
Hails
RS,
Petrokofsky
G,
Potts
SG,
Raine
NE,
Vanbergen
AJ,
McLean
AR.
2014
.
A restatement of the natural science evidence base concerning neonicotinoid insecticides and insect pollinators
.
Proceedings of the Royal Society B: Biological Sciences
281
:
20140558
.
Hackett
RA.
2013
.
Prairie fen plant biodiversity: the influence of site- and landscape- level factors on plant diversity. Master's thesis
.
Mount Pleasant
:
Central Michigan University (see Supplemental Material, Reference S3)
.
Hackett
RA,
Monfils
MJ,
Monfils
AK.
2016
.
Evaluating a sampling protocol for assessing plant diversity in prairie fens
.
Wetlands Ecology and Management
24
:
609
622
.
Hamilton
KGA.
2004
.
Beyond ecology: bugs reveal the deep roots of grassland ecoregions
.
Proceedings of the 19th North American Prairie Conference
.
Herman
DK,
Masters
LA,
Penskar
MR,
Reznicek
AA,
Wilhelm
GS,
Brodovich
WW,
Gardinier
KP.
2001
.
Floristic quality assessment with wetland categories and examples of computer application for the state of Michigan. 2nd edition
.
Lansing, Michigan
:
Michigan Department of Natural Resources Wildlife Edition
. .
Herold
N.
2014
.
C-CAP land cover classification scheme
.
Silver Spring, Maryland
:
National Oceanic and Atmospheric Administration (see Supplemental Material, Reference S4)
.
Higman
PJ,
Penskar
MR.
1999
.
Special plant abstract for Sporobolus heterolepis (prairie dropseed)
.
Lansing, Michigan
:
Michigan Natural Features Inventory (see Supplemental Material, Reference S5)
.
Hill
MO,
Bunce
RGH,
Shaw
MW.
1975
.
Indicator species analysis, a divisive polythetic method of classification, and its application to a survey of native pinewoods in Scotland
.
Journal of Ecology
63
:
597
613
.
Holzman
RW.
1972
.
Eastern range extension for Oarisma powesheik Parker (Lepidoptera: Hesperiidae)
.
Great Lakes Entomologist
.
5
:
111
114
.
Hoving
CL,
Lee
YM,
Badra
PJ,
Klatt
BJ.
2013
.
Changing climate, changing wildlife: a vulnerability assessment of 400 species of greatest conservation need and game species in Michigan
.
Michigan Department of Natural Resources Wildlife Division Report No. 3564
(
see Supplemental Material, Reference S6
).
Jermy
T.
1984
.
Evolution of insect/host plant relationships
.
American Naturalist
124
:
609
630
Kost
MA,
Albert
DA,
Cohen
JG,
Slaughter
BS,
Schillo
RK,
Weber
CR,
Chapman
KA.
2007
.
Natural communities of Michigan: classification and description
.
Lansing, Michigan
:
Michigan Natural Features Inventory, Report No. 2007-21
(
see Supplemental Material, Reference S7
).
Landis
DA,
Fiedler
AK,
Hamm
CA,
Cuthrell
DL,
Schools
EH,
Pearsall
DR,
Herbert
ME,
Doran
PJ.
2012
.
Insect conservation in Michigan prairie fen: addressing the challenge of global change
.
Journal of Insect Conservation
16
:
131
142
.
Lopez
RD,
Fennessy
MS.
2002
.
Testing the floristic quality assessment index as an indicator of wetland condition
.
Ecological Applications
12
:
487
497
.
Magurran
AE.
2004
.
Measuring biological diversity
.
Oxford, UK
:
Blackwell
.
McAlpine
WS.
1972
.
Observations on life history of Oarisma powesheik (Parker) 1870
.
Journal of Research on the Lepidoptera
11
:
83
93
.
McArdle
BH,
Anderson
MJ.
2001
.
Fitting multivariate models to community data: a comment on distance-based redundancy analysis
.
Ecology
82
:
290
297
.
McCabe,
TL.
1974
.
Skippers (Hesperioidea) of North Dakota. Master's thesis
.
Fargo
:
North Dakota State University (see Supplemental Material, Reference S8)
.
McCune
B,
Mefford
MJ.
2011
.
PC-ORD. Multivariate analysis of ecological data
.
Version 6.2.1
.
Gleneden Beach, Oregon
:
MjM Software
.
Michigan Natural Features Inventory
.
2014
.
Biotics 5—Michigan's Natural Heritage Database
.
Lansing, Michigan
.
[NOAA] National Oceanic and Atmospheric Administration Coastal Services Center
.
2014
.
Eastern Great Lakes 2010 Coastal Change Analysis Program Accuracy Assessment [Internet]
. .
Nice
CC,
Gompert
Z,
Forister
ML,
Fordyce
JA.
2009
.
An unseen foe in arthropod conservation efforts: the case of Wolbachia infections in the Karner blue butterfly
.
Biological Conservation
142
:
3137
3146
.
Noss
RD.
2005
.
Indicators for monitoring biodiversity: a hierarchical approach
.
Conservation Biology
4
:
355
364
.
Oksanen
J,
Blanchet
FG,
Kindt
R,
Legendre
P,
Minchin
PR,
O'Hara
RB,
Simpson
GL,
Solymos
P,
Stevens
MHH,
Wagner
H.
2015
.
vegan: community ecology package. R package version 2.3-1
.
Penskar
MR,
Higman
PJ.
1999
.
Special plant abstract for Muhlenbergia richardsonis (mat muhly)
.
Lansing, Michigan
:
Michigan Natural Features Inventory (see Supplemental Material, Reference S9)
.
Pogue
CD,
Monfils
MJ,
Cuthrell
DL,
Heumann
BW,
Monfils
AK.
2016
.
Habitat suitability modeling of the federally endangered Poweshiek skipperling (Oarisma poweshiek) in Michigan
.
Journal of Fish and Wildlife Management
7
:
359
368
.
Pointon
H.
2015
.
Larval host plant selection and daily behavior of Poweshiek skipperling (Oarisma poweshiek) in Michigan. Undergraduate thesis
.
Kalamazoo, Michigan
:
Kalamazoo College
. .
Pollard
E,
Yates
TJ.
1993
.
Monitoring butterflies for ecology and conservation
.
London
:
Chapman & Hall
,.
Poweshiek Working Group Participants
.
2011
.
Minutes. Dupont.J, editor
.
Winnipeg, Manitoba
:
Poweshiek skipperling workshop
.
see Supplemental Material, Reference S10)
.
R Core Team
.
2016
.
R: a language and environment for statistical computing
.
Vienna
:
R Foundation for Statistical Computing
.
Available: https://www.R-project.org/ (November 2018)
Reznicek
AA,
Voss
EG,
Walters
BS.
2011
.
Michigan Flora Online
.
Ann Arbor
:
University of Michigan
.
Available: http://michiganflora.net/. (June 2019)
.
Ricketts
TH.
2001
.
The matrix matters: effective isolation in fragmented landscapes
.
American Naturalist
158
:
87
99
.
Ruiz-Jaen
MC,
Aide
TM.
2005
.
Restoration success: how is it being measured?
Restoration Ecology
13
:
569
577
.
Runquist
E,
Heimpel
GE.
2017
.
Minnesota Invasive Terrestrial Plants and Pests Center
.
Minneapolis
:
University of Minnesota
.
Potential causes of declines in Minnesota's prairie butterflies with a focus on insecticidal control of the soybean aphid (see Supplemental Material, Reference S11)
.
Saarinen
EV,
Reilly
PF,
Austin
JD.
2016
.
Conservation genetics of an endangered grassland butterfly (Oarisma poweshiek) reveals historically high gene flow despite recent and rapid range loss
.
Insect Conservation and Diversity
9
:
517
528
.
Samson
F,
Knopf
F.
1994
.
Prairie Conservation in North America
.
BioScience
44
:
418
421
.
Selby
G.
2005
.
Status assessment and conservation guidelines: Poweshiek skipperling (Oarisma poweshiek (Parker) (Lepidoptera: Hesperiidae)
.
Bloomington, Minnesota
:
Twin Cities Field Office, U.S. Fish and Wildlife Service (see Supplemental Material, Reference S12)
.
[SARA] Species at Risk Act, Revised Statutes of Canada
.
2002
.
c. 29 see Supplemental Material, Reference S13).
Spieles
JB,
Comer
PJ,
Albert
DA,
Kost
MA.
1999
.
Natural community abstract for prairie fen
.
Lansing, Michigan
:
Michigan Natural Features Inventory (see Supplemental Material, Reference S14)
.
Swengel
SR,
Schlicht
D,
Olsen
F,
Swengel
AB.
2010
.
Declines of prairie butterflies in the midwestern USA
.
Journal of Insect Conservation
15
(
1–2
):
327
339
.
Swengel
AB,
Swengel
SR.
1999
.
Observations of prairie skippers (Oarisma poweshiek, Hesperia dacotae, H. Ottoe, H. leonardus pawnee, and Atrytone arogos iowa (Lepidoptera: Hesperiidae)
.
Great Lakes Entomologist
32
:
267
292
.
Swink
F,
Wilhelm
G.
1994
.
Plants of the Chicago region. 4th edition
.
Indianapolis
:
Indiana Academy of Science
.
Tilman
D,
Knops
J,
Wedin
D,
Reich
P,
Ritchie
M,
Siemann
E.
1997
.
The influence of functional diversity and composition on ecosystem processes
.
Science
277
:
1300
1302
.
Timoney
K.
1999
.
The habitat of nesting whooping cranes
.
Biological Conservation
89
:
189
197
.
[ESA] U.S. Endangered Species Act of 1973, as amended, Pub. L. No. 93-205, 87 Stat. 884 (Dec. 28, 1973)
.
[USEPA] U.S. Environmental Protection Agency
.
1994
.
National Water Quality Inventory: 1992 Report to Congress. Office of Water
.
Washington, D.C
.:
U.S. Environmental Protection Agency (see Supplemental Material, Reference S15
).
Watała
C.
2007
.
Sample size and significance—somewhere between statistical power and judgment prostration
.
Archives of Medical Science
3
:
5
13
.
Weibull
AC,
Bengtsson
J,
Nohlgren
E.
2008
.
Diversity of butterflies in the agricultural landscape: the role of farming system and landscape heterogeneity
.
Ecography
23
:
743
750
.
Werren
JH,
Baldo
L,
Clark
ME.
2008
.
Wolbachia: master manipulators of invertebrate biology
.
Nature Reviews Microbiology
6
:
741
751
.
Wilhelm
GS,
Masters
LA.
1995
.
Floristic quality assessment in the Chicago region and application computer programs
.
Lisle, Illinois
:
Morton Arboretum
.
Young
TP.
2000
.
Restoration ecology and conservation biology
.
Biological Conservation
92
:
73
83
.

Author notes

Citation: Pogue CD, Monfils MJ, Cuthrell DL, Hackett RA, Zionce RA, Monfils AK. 2019. Local- and landscape-level variables related to Poweshiek skipperling presence in Michigan, USA prairie fens. Journal of Fish and Wildlife Management 10(2):375–390; e1944-687X. https://doi.org/10.3996/122018-JFWM-117

The findings and conclusions in this article are those of the author(s) and do not necessarily represent the views of the U.S. Fish and Wildlife Service.

Supplemental Material