Endangered and Threatened Wildlife and Plants; Threatened Status for Dakota Skipper and Endangered Status for Poweshiek Skipperling

DEPARTMENT OF THE INTERIOR

Fish and Wildlife Service

50 CFR Part 17

[FWS-R3-ES-2013-0043; 4500030113]

RIN 1018-AY01

Endangered and Threatened Wildlife and Plants; Threatened Status for Dakota Skipper and Endangered Status for Poweshiek Skipperling

AGENCY:

Fish and Wildlife Service, Interior.

ACTION:

Proposed rule.

SUMMARY:

We, the U.S. Fish and Wildlife Service, propose to list the Dakota skipper as a threatened species and the Poweshiek skipperling as an endangered species under the Endangered Species Act of 1973, as amended. If we finalize this rule as proposed, it would extend the Act's protections to the Dakota skipper and the Poweshiek skipperling. The effect of this regulation is to add the Dakota skipper and the Poweshiek skipperling to the List of Endangered and Threatened Wildlife. We also propose a special rule under section 4(d) of the Act that outlines the prohibitions necessary and advisable for the conservation of the Dakota skipper, if it is listed as a threatened species.

DATES:

Written Comments:

We will accept comments received or postmarked on or before December 23, 2013. Comments submitted electronically using the Federal eRulemaking Portal (see

ADDRESSES

section, below) must be received by 11:59 p.m. Eastern Time on the closing date. We must receive requests for public hearings, in writing, at the address shown in the

ADDRESSES

section by December 9, 2013.

Public Informational Meetings:

To better inform the public of the implications of the proposed listing and to answer any questions regarding this proposed rule, we plan to hold five public informational meetings. We have scheduled informational meetings regarding the proposed rule in the following locations:

(1) Minot, North Dakota, on November 5, 2013, at the Souris Valley Suites, 800 37th Avenue SW;

(2) Milbank, South Dakota, on November 6, 2013, at the Milbank Chamber of Commerce, 1001 East 4th Avenue;

(3) Milford, Iowa, on November 7, 2013, at the Iowa Lakeside Laboratory, 1838 Highway 86;

(4) Holly, Michigan, on November 13, 2013, at the Rose Pioneer Elementary School, 7110 Milford Road; and

(5) Berlin, Wisconsin, on November 14, 2013, at the Berlin Public Library, 121 West Park Avenue.

Except for the meeting in Berlin, Wisconsin, each informational meeting will be from 5:30 p.m. to 8:00 p.m.; the meeting in Berlin, Wisconsin will be from 4:30 p.m. to 7:00 p.m.

ADDRESSES:

You may submit comments by one of the following methods:

(1)

Electronically:

Go to the Federal eRulemaking Portal:

http://www.regulations.gov.

In the Search box, enter FWS-R3-ES-2013-0043, which is the docket number for this rulemaking. Then, in the Search panel on the left side of the screen, under the Document Type heading, click on the Proposed Rules link to locate this document. You may submit a comment by clicking on “Comment Now!”

(2)

By hard copy:

Submit by U.S. mail or hand-delivery to: Public Comments Processing, Attn: FWS-R3-ES-2013-0043; Division of Policy and Directives Management; U.S. Fish and Wildlife Service; 4401 N. Fairfax Drive, MS 2042-PDM; Arlington, VA 22203.

We request that you send comments only by the methods described above. We will post all comments on

http://www.regulations.gov.

This generally means that we will post any personal information you provide us (see the Public Comments section below for more information).

FOR FURTHER INFORMATION CONTACT:

Pete Fasbender, Field Supervisor, U.S. Fish and Wildlife Service, Twin Cities Ecological Services Office, 4101 American Boulevard East, Bloomington, Minnesota, 55425, by telephone (612) 725-3548 or by facsimile (612) 725-3609. Persons who use a telecommunications device for the deaf (TDD) may call the Federal Information Relay Service (FIRS) at 800-877-8339.

SUPPLEMENTARY INFORMATION:

Executive Summary

Why we need to publish a rule.

Under the Endangered Species Act (Act), if a species is determined to be an endangered or threatened species throughout all or a significant portion of its range, we are required to promptly publish a proposal in the

Federal Register

and make a determination on our proposal within one year. Listing a species as an endangered or threatened species can only be completed by issuing a rule. A species may warrant protection through listing under the Act if it meets the definition of an endangered or threatened species throughout all or a significant portion of its range.

This rule consists of:

• A proposed rule to list the Poweshiek skipperling as an endangered species;

• A proposed rule to list the Dakota skipper as threatened species; and

• A proposed special rule under section 4(d) of the Act that outlines the prohibitions necessary and advisable for the conservation of the Dakota skipper.

Elsewhere in today's

Federal Register

, we propose to designate critical habitat for the Dakota skipper and Poweshiek skipperling under the Act.

The basis for our action.

Under the Act, we can determine that a species is an endangered or threatened species based on any of five factors: (A) The present or threatened destruction, modification, or curtailment of its habitat or range; (B) overutilization for commercial, recreational, scientific, or educational purposes; (C) disease or predation; (D) the inadequacy of existing regulatory mechanisms; or (E) other natural or manmade factors affecting its continued existence. Furthermore, whenever a species is listed as a threatened species, we may issue regulations that are necessary and advisable for the conservation of that species under section 4(d) of the Act.

We have determined the threats to both species include:

• Habitat loss and degradation of native prairies and prairie fens, resulting from conversion to agriculture or other development; ecological succession and encroachment of invasive species and woody vegetation primarily due to lack of management; past and present fire, haying, or grazing management that degrades or eliminates native prairie grasses and flowering forbs; flooding; and groundwater depletion, alteration, and contamination.

• Other natural or manmade factors, including loss of genetic diversity, small size and isolation of sites, indiscriminate use of herbicides such that it reduces or eliminates nectar sources, climate conditions such as drought, and other unknown stressors.

• Existing regulatory mechanisms are inadequate to mitigate these threats to both species.

We will seek peer review.

We are seeking comments from knowledgeable individuals with scientific expertise to review our analysis of the best available science and application of that science and to provide any additional scientific information to improve this proposed rule. Because we will consider all comments and information received during the comment period, our final

determinations may differ from this proposal.

Information Requested

We intend that any final action resulting from this proposed rule will be based on the best scientific and commercial data available and be as accurate and as effective as possible. Therefore, we request comments or information from other concerned government agencies, the scientific community, industry, or any other interested party concerning this proposed rule. We particularly seek comments concerning:

(1) The species' biology, range, and population trends, including:

(a) Habitat requirements for feeding, breeding, and sheltering;

(b) Genetics and taxonomy;

(c) Historical and current range including distribution patterns;

(d) Historical and current population levels, and current and projected trends; and

(e) Past and ongoing conservation measures for the species, its habitat, or both.

(2) The factors that are the basis for making a listing determination for a species under section 4(a) of the Act (16 U.S.C. 1531

et seq.

), which are:

(a) The present or threatened destruction, modification, or curtailment of its habitat or range;

(b) Overutilization for commercial, recreational, scientific, or educational purposes;

(c) Disease or predation;

(d) The inadequacy of existing regulatory mechanisms; or

(e) Other natural or manmade factors affecting its continued existence.

(3) Biological, commercial trade, or other relevant data concerning any threats (or lack thereof) to these species and existing regulations that may be addressing those threats;

(4) Additional information concerning the historical and current status, range, distribution, and population size of these species, including the locations of any additional populations;

(5) Any information on the biological or ecological requirements of these species and ongoing conservation measures for these species and their habitat;

(6) Specific information on the amount and distribution of the Dakota skipper and Poweshiek skipperling and their habitat; and

(7) Our approach to determining the status of each species at each site, and our definitions of “present,” “unknown,” “possibly extirpated,” and “extirpated” as described under

Species Status,

below.

(8) Suitability of the proposed 4(d) rule for the conservation, recovery, and management of the Dakota skipper.

(9) Whether it would be appropriate to allow routine livestock grazing activities on lands inhabited by Dakota skipper in any additional counties. The proposed 4(d) rule would allow routine livestock grazing activities on lands inhabited by the Dakota skipper in counties where the species does not primarily occur in relatively flat and moist (wet-mesic or mesic) prairie habitats. Wet-mesic or mesic habitats in which the Dakota skipper occurs are typically hayed after July 15 and not grazed. We are seeking comments on whether or not grazing may be implemented in these habitats in a manner that would allow for the persistence of the Dakota skipper.

(10) Any information on Tribal regulations or Tribal conservation efforts that may affect either the Dakota skipper or Poweshiek skipperling and their habitat.

Please include sufficient information with your submission (such as scientific journal articles or other publications) to allow us to verify any scientific or commercial information you include.

Please note that submissions merely stating support for or opposition to the action under consideration without providing supporting information, although noted, will not be considered in making a determination, as section 4(b)(1)(A) of the Act directs that determinations as to whether any species is a threatened or endangered species must be made “solely on the basis of the best scientific and commercial data available.”

You may submit your comments and materials concerning this proposed rule by one of the methods listed in the

ADDRESSES

section. We request that you send comments only by the methods described in the

ADDRESSES

section.

If you submit information via

http://www.regulations.gov,

your entire submission—including any personal identifying information—will be posted on the Web site. If your submission is made via a hardcopy that includes personal identifying information, you may request at the top of your document that we withhold this information from public review. However, we cannot guarantee that we will be able to do so. We will post all hardcopy submissions on

http://www.regulations.gov.

Please include sufficient information with your comments (such as scientific journal articles or other publications) to allow us to verify any scientific or commercial information you include.

Comments and materials we receive, as well as supporting documentation we used in preparing this proposed rule, will be available for public inspection on

http://www.regulations.gov,

or by appointment, during normal business hours, at the U.S. Fish and Wildlife Service, Twin Cities Ecological Services Office (see

FOR FURTHER INFORMATION CONTACT

).

Previous Federal Actions

The U.S. Fish and Wildlife Service (Service) initiated proceedings to list the Dakota skipper as a threatened species in 1978 (43 FR 28938), but withdrew the proposed rulemaking after Congress amended the Endangered Species Act in 1979 (45 FR 58171). The Dakota skipper was designated a category 2 candidate species in the May 22, 1984, Notice of Review (49 FR 21664) and remained a category 2 species (January 6, 1989, 54 FR 572; November 21, 1991, 56 FR 58830; and November 15, 1994, 59 FR 59020). A category 2 candidate was defined as a species for which information in the Service's possession indicates that listing was possibly appropriate, but for which sufficient information on biological vulnerability and threats was not currently available to support a proposal for listing under the Act.

On January 21, 1994, the Service received a petition from the Biodiversity Legal Foundation to list the Dakota skipper as an endangered or threatened species and to designate critical habitat. We made a 90-day finding that the petition presented substantial information to indicate that the requested action may be warranted; the finding was published in the

Federal Register

on July 28, 1994 (59 FR 38424). On February 27, 1995, we announced a 12-month finding in which we determined that the species should remain as a category 2 candidate, that timely appropriate prairie management and protection may eliminate the need to list the species, and that researchers indicated that more surveys, particularly in Minnesota, Iowa, and North Dakota, were needed (60 FR 10535).

In a December 5, 1996 (61 FR 64481) decision, the Service discontinued the practice of maintaining a list of species regarded as “category-2 candidates.” Instead, the Service would keep a single list of candidate species—species for which the Service has on file sufficient information to support issuance of a proposed listing rule.

In 2002, the Service reviewed the status of the Dakota skipper and determined that it met the definition of a candidate species. The Dakota skipper was assigned a listing priority number of 11 on June 13, 2002 (67 FR 40657).

The Dakota skipper remained a candidate species with a listing priority number of 11 in subsequent notices, including May 4, 2004 (69 FR 24876), May 11, 2005 (70 FR 24870), and September 12, 2006 (71 FR 53756). The Service changed the listing priority from 11 to 8 on December 6, 2007 (72 FR 69034), and the Dakota skipper remained a candidate species with a listing priority number of 8 in subsequent notices, including December 10, 2008 (73 FR 75176), November 9, 2009 (74 FR 57804), November 10, 2010 (75 FR 69222), and October 26, 2011 (76 FR 66370).

On May 12, 2003, the Service received a petition from the Biodiversity Conservation Alliance and five others to list the Dakota skipper as endangered or threatened and to designate critical habitat. The Service agreed with the petitioners, by virtue of having made it a candidate in 2002, that the Dakota skipper warranted listing as threatened or endangered under the Act. The petition did not contain evidence supporting emergency listing or changing the listing priority number; therefore, the Service took no further action on the petition.

On July 12, 2011, the Service filed a proposed settlement agreement with the Center for Biological Diversity in a consolidated case in the U.S. District Court for the District of Columbia. The settlement agreement was approved by the court on September 9, 2011. As part of this settlement agreement, the Service agreed to complete a proposed listing rule or not warranted finding for the Dakota skipper by September 30, 2013.

The Service identified the Poweshiek skipperling (

Oarisma poweshiek

) as a candidate species, with a listing priority number of 2, in a notice of review published in the

Federal Register

on October 26, 2011 (76 FR 66370).

Status Assessments for Dakota Skipper and Poweshiek Skipperling

Background

Dakota Skipper

Species Description

The Dakota skipper (

Hesperia dacotae

) is a member of the skipper family Hesperiidae and was first described in 1911 from collections taken at Volga, South Dakota, and Grinnell, Iowa (Skinner 1911 in Royer and Marrone 1992a, p. 1). The family Hesperiidae includes 3 other subfamilies, and the genus

Hesperia

contains 18 species (Miller and Brown 1981, p. 31; Ferris 1989 in Royer and Marrone 1992a, p. 1). Dakota skipper is the accepted common name for

H. dacotae.

The Dakota skipper is a small to medium-sized butterfly with a wingspan of 2.4-3.2 centimeters (cm) (0.9-1.3 inches (in)) and hooked antennae (Royer and Marrone 1992a, p. 3). Like other Hesperiidae species, Dakota skippers have a faster and more powerful flight than most butterflies because of a thick, well-muscled thorax (Scott 1986, p. 415).

Adult Dakota skippers have variable markings. The dorsal surface of adult male wings ranges in color from tawny-orange to brown and has a prominent mark on the forewing; the ventral surface is dusty yellow-orange (Royer and Marrone 1992a, p. 3). The dorsal surface of adult females is darker brown with diffused tawny orange spots and a few diffused white spots restricted to the margin of the forewing; the ventral surfaces are dusty gray-brown with a faint white spotband across the middle of the wing (Royer and Marrone 1992a, p. 3). Adult Dakota skippers may be confused with the Ottoe skipper (

H. ottoe

), which is somewhat larger with slightly longer wings (Royer and Marrone 1992a, p. 3). Dakota skipper pupae are reddish-brown, and the larvae are light brown with a black collar and dark brown head (McCabe 1981, p. 181).

General Life History

Dakota skippers are univoltine (having a single flight per year), with an adult flight period that may occur from the middle of June through the end of July (McCabe 1979, p. 6; McCabe 1981, p. 180; Dana 1991, p. 1; Royer and Marrone 1992a, p. 26; Skadsen 1997, p. 3; Swengel and Swengel 1999, p. 282). The actual flight period varies somewhat across the range of each species and can also vary significantly from year-to-year, depending on weather patterns. Females emerge slightly later than males (Dana 1991, p. 1), and the observed sex ratio of Dakota skippers was roughly equal during peak flight periods (Dana 1991, p. 15; Swengel and Swengel 1999, pp. 274, 283).

The Dakota skipper flight period in a locality lasts two to four weeks, and mating occurs throughout this period (Braker 1985, p. 46; McCabe and Post 1977a, p. 38; McCabe and Post 1977b, p. 36; McCabe 1979, p. 6; McCabe 1981, p. 180; Dana 1991, p. 15; Swengel and Swengel 1999, p. 282). Adult male Dakota skippers exhibit perching behavior (perch on tall plants to search for females), but occasionally appear to patrol in search of mating opportunities (Royer and Marrone 1992a, p. 25).

Dakota skippers lay eggs on broadleaf plants (McCabe 1981, p. 180) and grasses (Dana 1991, p. 17), although larvae feed only on grasses. Potential lifetime fecundity is between 180 and 250 eggs per female Dakota skipper; realized fecundity depends upon longevity (Dana 1991, p. 26). Female Dakota skippers lay eggs daily in diminishing numbers as they age (Dana 1991, pp. 25-26). Dana (1991, p. 32) estimated the potential adult life span of Dakota skipper to be 3 weeks and the average life span (or residence on site before death or emigration) to be 3 to 10 days on one Minnesota prairie.

Dakota skippers overwinter as larvae and complete one generation per year. Dakota skipper eggs hatch after incubating for 7-20 days; therefore, hatching is likely completed before the end of July. After hatching, Dakota skipper larvae crawl to the bases of grass plants where they form shelters at or below the ground surface with silk, fastened together with plant tissue (Dana 1991, p. 16). They construct 2-3 successively larger shelters as they grow (Dana 1991, p. 16). The larvae emerge from their shelters at night to forage (McCabe 1979, p. 6; McCabe 1981, p. 181; Royer and Marrone 1992a, p. 25) and appear to clip blades of grass and bring them back to their shelters to consume (Dana 2012a, pers. comm.).

Dakota skippers have six or seven larval stages (instars) (Dana 1991, pp. 14-15) and overwinter (diapause) in ground-level or subsurface shelters during either the fourth or fifth instar (McCabe 1979, p. 6; McCabe 1981, pp. 180, 189; Dana 1991, p. 15; Royer and Marrone 1992a, pp. 25-26). In the spring, larvae resume feeding and undergo two additional molts before they pupate. During the last two instars, larvae shift from buried shelters to horizontal shelters at the soil surface (Dana 1991, p. 16).

Food and Water

Nectar and water sources for adult Dakota skippers vary regionally and include purple coneflower (

Echinacea angustifolia

), bluebell bellflower (

Campanula rotundifolia

), white prairie clover (

Dalea candida

), upright prairie coneflower (

Ratibida columnifera

), fleabanes (

Erigeron spp.

), blanketflowers (

Gaillardia spp.

), black-eyed Susan (

Rudbeckia hirta

), groundplum milkvetch (

Astragalus crassicarpus

), and yellow sundrops (

Calylophus serrulatus

) (McCabe and Post 1977b, p. 36; Royer and Marrone 1992a, p. 21). Plant species likely vary in their value as nectar sources due to the amount of nectar available during the adult flight period (Dana 1991, p. 48). Swengel and

Swengel (1999, pp. 280-281) observed nectaring at 25 plant species, but 85 percent of the nectaring was at the following three plants, in declining order of frequency: Purple coneflower, blanketflower, and groundplum milkvetch. Dana (1991, p. 21) reported the use of 25 nectar species in Minnesota with purple coneflower most frequented; McCabe (1979, p. 42, McCabe 1981, p. 187) observed Dakota skippers using eight nectar plants. In addition to nutrition, the nectar of flowering forbs provides water for Dakota skipper, which is necessary to avoid desiccation during flight activity (Dana 1991, p. 47; Dana 2013, pers. comm.).

Dakota skipper larvae feed only on several native grass species; little bluestem (

Schizachyrium scoparium

) is a frequent food source of the larvae (Dana 1991, p. 17; Royer and Marrone 1992a, p. 25), although they have been found on

Panicum

spp.,

Poa

spp., and other native grasses (Royer and Marrone 1992a, p. 25). Seasonal senescence patterns of grasses relative to the larval period of Dakota skippers are likely important in determining the suitability of grass species as larval host plants. Large leaf blades, leaf hairs, and the distance from larval ground shelters to palatable leaf parts preclude the value of big bluestem and Indian grass as larval food plants (Dana 1991, p. 46).

Dispersal

Dakota skipper are not known to disperse widely; the species was evaluated among 291 butterfly species in Canada as having relatively low mobility. Experts estimated Dakota skipper to have a mean mobility of 3.5 (standard deviation = 0.7) on a scale of 0 (sedentary) to 10 (highly mobile) (Burke

et al.

2011, p. 2279; Fitzsimmons 2012, pers. comm.). Dakota skippers may be incapable of moving greater than 1 kilometer (km) (0.6 miles (mi)) between patches of prairie habitat separated by structurally similar habitats (

e.g.,

crop fields, grass-dominated fields or pasture, but not necessarily native prairie) (Cochrane and Delphey 2002, p. 6). Royer and Marrone (1992a, p. 25) concluded that Dakota skippers are not inclined to disperse, although they did not describe individual ranges or dispersal distances. McCabe (1979, p. 9; 1981, p. 186) found that concentrated activity areas for Dakota skippers shift annually in response to local nectar sources and disturbance.

In a mark-recapture study, average adult movements of Dakota skipper were less than 300 meters (m) (984 feet (ft)) over 3-7 days; marked adults crossed less than 200 m (656 ft) of unsuitable habitat between two prairie patches and moved along ridges more frequently than across valleys (Dana 1991, pp. 38-40). Dana (1997, p. 5) later observed reduced movement rates across a small valley with roads and crop fields compared with movements in adjacent widespread prairie habitat. Skadsen (1999, p. 2) reported possible movement of Dakota skippers in 1998 from a known population at least 800 m (2625 ft) away to a site with an unusually heavy growth of purple coneflower; he had not found Dakota skippers in three previous years when coneflower production was sparse. The two sites were connected by native vegetation of varying quality, interspersed by a few asphalt and gravel roads (Skadsen 2001, pers. comm.).

In summary, dispersal of Dakota skipper is very limited due in part to its short adult life span and single annual flight. Therefore, the species' extirpation from a site is likely permanent unless it is within about 1 km (0.6 mi) of a site that generates a sufficient number of emigrants or is artificially reintroduced to a site; however, the capability to propagate the Dakota skipper is currently lacking.

Habitat

Dakota skippers are obligate residents of undisturbed (remnant, untilled) high-quality prairie, ranging from wet-mesic tallgrass prairie to dry-mesic mixed-grass prairie (Royer and Marrone 1992a, pp. 8, 21). High-quality prairie contains a high diversity of native plant species, including flowering herbaceous plants (forbs). Royer and Marrone (1992a, p. 21) categorized Dakota skipper habitat into two main types that were once intermixed on a landscape scale, but are now mostly segregated. The first, referred to as “Type A” by Royer

et al.

(2008, pp. 14-16), is low wet-mesic prairie that occurs on near-shore glacial lake deposits. Type A Dakota skipper habitat is dominated by bluestem grasses, with three other plant species almost always present and blooming during Dakota skipper's flight period: Wood lily (

Lilium philadelphicum

), bluebell bellflower, and mountain deathcamas (smooth camas;

Zigadenus elegans

) (McCabe 1981, p. 190). This habitat type has a high water table and is subject to intermittent flooding in the spring, but provides “sufficient relief to provide segments of non-inundated habitat during the spring larval growth period within any single season” (Royer

et al.

2008, p. 15). Common forbs in bloom during the late season in Type A habitat include Rocky Mountain blazing star (

Liatris ligulistylis

), Canada goldenrod (

Solidago canadensis

), strict blue-eyed grass (

Sisyrinchium montanum

), common goldstar (

Hypoxis hirsuta

), and black-eyed Susan (Lenz 1999a, p. 6). Type A habitats also contain small patches of dry-mesic prairie inhabited by Dakota skippers. Common forb species in these dry-mesic areas include stiff sunflower (

Helianthus pauciflorus Nutt. ssp. pauciflorus

), and candle anenome (

Anemone cylindrica

), although purple coneflower was rare in these habitats (Lenz 1999a, pp. 6-11). Dakota skipper inhabits Type A habitat in north-central North Dakota, southeast North Dakota, and Manitoba.

The second Dakota skipper habitat type, referred to as “Type B” by Royer

et al.

(2008, p. 14), occurs on rolling terrain over gravelly glacial moraine deposits and is dominated by bluestems and needle grasses (

Heterostipa spp.

). As with Type A habitat, bluebell bellflower and wood lily are also present in Type B habitats, but Type B habitats also support more extensive stands of purple coneflower, upright prairie coneflower, and common gaillardia (

Gaillardia aristata

) (Royer and Marrone 1992a, p. 22). Both Type A and Type B prairies may contain slightly depressional (low topographical areas that allow for the collection of surface water) wetlands with extensive flat areas and slightly convex hummocks, which are dryer than the wet areas (Lenz 1999b, pp. 4, 8).

In northeastern South Dakota, Dakota skippers inhabit primarily Type B habitats with abundant purple coneflower, but they also occur in nearby Type A habitats in some areas (Skadsen 1997, p. 4). All Type A habitats occupied by Dakota skipper in South Dakota are near hill prairie (Type B) habitats that are managed with fall haying (Skadsen 2006b, p. 2).

Little bluestem and porcupine grass are the predominant grass species in Dakota skipper habitat in South Dakota (Skadsen 2006b, p. 2). Dry-mesic prairies suitable for Dakota skippers in South Dakota typically include little bluestem, side oats grama, porcupine grass, needle-and-thread grass (

H. comata

), and prairie dropseed, and a high diversity and abundance of forbs, including purple coneflower, purple prairie clover (

Dalea purpurea

), white prairie clover, yellow sundrops, prairie groundsel (

Packera plattensis

), groundplum milkvetch, eastern pasqueflower (

Pulsatilla patens

), old man's whiskers (prairie smoke,

Geum triflorum

), western silver aster (

Symphyotrichum sericeum

), dotted blazing star (

Liatris punctata

), tall blazing star (

L. asper

), meadow zizia

(

Zizia aptera

), blanket flower (

Gaillardia sp.),

prairie sagewort (

Artemisia frigida

), and leadplant (

Amorpha canescens

) (Skadsen 2006b, pp. 1-2). Purple coneflower occurs at all sites where the Dakota skipper has been recorded in South Dakota, although it is absent at some sites where Dakota skipper is abundant in other states (Skadsen 2006b, p. 2).

In Minnesota, Dakota skippers inhabit Type B habitats. Dana (1997, p. 8) described typical habitat in Minnesota as dry-mesic prairie dominated by mid-height grasses with an abundance of nectar sources including purple coneflower and prairie milkvetch (

Astragalus laxmannii Jacq. var. robustior

). Southern dry prairies in Minnesota are described as having sparse shrub cover (less than 5 percent) composed primarily of leadplant, with prairie rose (

Rosa arkansana

), wormwood sage, or smooth sumac (

Rhus glabra

) present and few, if any, trees (Minnesota DNR 2012a). Dana (1991, p. 21) never encountered Dakota skippers in wet or wet-mesic prairies in Minnesota, despite abundance of suitable plants and the frequent use of these habitats by similar skipper species. In systematic surveys at twelve Minnesota sites, Swengel and Swengel (1999, pp. 278-279) found that Dakota skippers were significantly more abundant on dry prairie than on either wet-mesic prairie. In Manitoba, Dakota skippers inhabit Type A habitats, occupy the slightly higher, drier areas of wet-mesic prairie where nectar sources are more abundant (Webster 2003, p. 7). Occupied habitats in Saskatchewan are similar to the drier upland dry-mesic mixed-grass prairie hillside habitats in Manitoba, which is dominated by bluestems and needlegrass. The Dakota skipper was most common on ridgetops and hillsides near purple coneflower (Webster 2003, p. 8).

In North Dakota, an association of bluestems (

Schizachyrium scoparium, Andropogon gerardii

) and needlegrasses, typically invaded by Kentucky bluegrass (

Poa pratensis

), typifies dry-mesic Dakota skipper habitat in the rolling terrain of river valleys and the Missouri Coteau (Royer and Marrone 1992a, p. 22). These prairies, located on the western edge of the species' known range, typically contain wood lily, bluebell bellflower, coneflowers, and other asters as nectar sources; in some areas, mountain deathcamas also occurs (Royer and Marrone 1992a, p. 22). The location of larval food plants rarely seems to affect Dakota skipper distribution within habitats because these warm-season grasses are usually dominant and evenly dispersed (Swengel 1994, p. 6), although invasion by smooth brome grass (

Bromus inermis

) and other invasive species may displace or extirpate native larval food plants (Culliney 2005, p. 134, Bahm

et al.

2011, p. 240, LaBar and Schultz 2012, p. 177).

Two key factors, soils unsuitable for agriculture and steep topography, have allowed remnant native prairie habitats inhabited by Dakota skippers to persist (Royer and Marrone 1992a, p. 22). McCabe (1979, pp. 17-18; 1981, p. 192) and Royer

et al.

(2008, p. 16) have linked the historical distribution of Dakota skippers to surface geological features and soils that are glacial in origin and, possibly, regional precipitation-evaporation ratios (ratio of evaporation occurring naturally in one location over a given area compared to the amount of precipitation, such as rain and snow, falling over the same area). Soil types typical of Dakota skipper sites were described as sandy loams, loamy sand, or loams (Lord 1988 in Royer

et al.

2008, pp. 3, 10). Additional edaphic (soil) features, such as soil moisture, compaction, surface temperature, pH, and humidity, may be contributing factors in larval survival and, thus, important limiting factors for Dakota skipper populations (Royer

et al.

2008, p. 2). For example, edaphic parameters measured in sites throughout the range of Dakota skipper included a bulk density (an indicator of soil compaction) that ranged from 0.9g/cm

3

to 1.3 g/cm

3

and mean soil pH that ranged from 6.3 to 6.7 with high micro-scale variation (variation on a small scale) (Royer

et al.

2008, p. 10). Soil texture ranged from 4 to 12 percent clay, 53 to 74 percent sand, and 14 to 39 percent silt (Royer

et al.

2008, p. 12). Seasonal soil temperatures, measured at three depths (20, 40, and 60 cm (8, 16, and 24 in)) were the same at all depths within a site; Minnesota sites generally had higher soil temperatures at all depths than sites in North Dakota or South Dakota (Royer

et al.

2008, p. 11).

Dakota skipper larvae are particularly vulnerable to desiccation (drying out) during dry summer months and require “vertical water distribution” (movement of shallow groundwater to the soil surface) in the soils or wet low areas to provide relief from high summer temperatures (Royer

et al.

2008, pp. 2, 16). Humidity may also be essential for larval survival during winter months since the larvae cannot take in water during that time and depend on humid air to minimize water loss through respiration (Dana 2013, pers. comm.). Royer (2008, pp. 14-15) measured microclimalogical levels (climate in a small space, such as at or near the soil surface) within “larval nesting zones” (between the soil surface and 2 cm deep) throughout the range of Dakota skippers, and found an acceptable rangewide seasonal (summer) mean temperature range of 18 to 21°C (64 to 70 °F), rangewide seasonal mean dew point ranging from 14 to 17 °C (57 to 63 °F), and rangewide seasonal mean relative humidity between 73 and 85 percent.

Species Occupancy

We generally consider the Dakota skipper or Poweshiek skipperling to be “present” at sites where the species was detected during the most recent survey, if the survey was conducted in 2002 or more recently and there is no evidence to suggest the species is now extirpated from the site, (

e.g.,

no destruction or obvious and significant degradation of the species' habitat), with the exception of the following five sites. We consider the species to be present at one Poweshiek skipperling site in Michigan where the species was observed at the site in 1996 and no further surveys have been conducted. This site, however, still has suitable habitat for the species according to species experts in the State and at least one other species of prairie fen dependent butterfly is present (Hosler 2013, pers. comm.). Therefore, the Poweshiek skipperling is most likely still present at this site. We also consider the species to be present at one Dakota skipper site (Frenchman's Bluff Preserve in Minnesota) where the most recent survey was from 1993. At this site, no evidence suggests the species is not still present because, based on a species-expert review of the site, the habitat and management is still conducive to the species. Additional sites where we consider Dakota skipper to be present include two sites in Minnesota with 1996 records (Bluestem Prairie and Buffalo River State Park) and one site with a 1998 record (an unnamed site in North Dakota). Although no survey for the species has taken place at Bluestem Prairie since 1996, a 2012 assessment of the habitat at the site indicates that this site is a high-quality prairie that contains the native prairie flora conducive to the Dakota skipper (Selby 2012, p. 9). The site at Buffalo River State park, which adjoins Bluestem Prairie, has not been surveyed since 1996 but recent habitat assessments show that it still contains prairie habitats with the native prairie flora conducive to the species (MN DNR 2013, unpubl.). Furthermore, the species expert in Minnesota supports that the species is most likely still present at these sites. Little information is known about the one unnamed site in North

Dakota; however, the best information we have indicates that the habitat is still suitable for the species, and the North Dakota species expert supports that the species is likely present.

We assigned a status of “unknown” if the species was found in 1993 or more recently, but not in the most recent one to two sequential survey year(s) since 1993 and there is no evidence to suggest the species is now extirpated from the site (

e.g.,

no destruction or obvious and significant degradation of the species' habitat). We considered a species is to be “possibly extirpated” at sites where it was detected at least once prior to 1993, but not in the most recent one to two sequential survey years(s). A species is also considered “possibly extirpated” at sites where it was found prior to 1993 and no surveys have been conducted in 1993 or more recently. At least three sequential years of negative surveys were necessary for us to consider the species “extirpated” from a site, because of the difficulty of detecting these species, as explained further in this section. A species is also considered “extirpated” at sites where habitat for the species is no longer present.

When determining whether the species occupancy is unknown, possibly extirpated, or extirpated at a particular site, we used the survey year 1993 as a cut-off date, because most known sites (more than 75 percent of known Poweshiek skipperling sites and over 89 percent of known Dakota skipper sites) have been surveyed at least once since 1993 and survey data more than 20 years old may not reflect the current status of a species or its habitat at a site (for example, due to habitat loss from secondary succession of woody vegetation or a change in plant communities due to invasive species). Although it cannot be presumed that the species is absent at sites not surveyed since 1993, the likelihood of occupancy of these sites should be considered differently than sites with more recent survey data (

e.g.,

due to woody vegetation succession over time). When analyzing survey results, we disregarded negative surveys conducted outside of the species' flight period or under unsuitable conditions (

e.g.,

high wind speeds).

After we applied these standards to initially ascertain the status of the species, we asked species experts and Service personnel to help verify, modify, or correct species' occupancy at each site (particularly for sites with questionable habitat quality or those that have not been surveyed recently). In most cases, we used the status confirmed during expert review, unless we received additional information (

e.g.,

additional survey or habitat data provided after the expert reviews) that suggests a different status at a particular site.

Timing of surveys is based on initial field checks of nectar plant blooms and sightings of butterfly species with synchronous emergence (sightings of butterfly species that emerge at the same time as Dakota skipper and Poweshiek skipperling), and, more recently, emergence estimated by a degree-day emergence model using high and low daily temperature data from weather stations near the survey sites (Selby, undated, unpublished dissertation). Surveys are conducted during flight periods when the species' abundance is expected to be at levels at which the species can be detected. However, as with many rare species, detection probabilities are imperfect and some uncertainty remains between non-detection and true absence (Gross

et al.

2007, pp. 192, 197-198; Pellet 2008, pp. 155-156). Three sequential years of negative surveys is sufficient to capture variable detection probabilities, since each survey year typically encompasses more than one visit (

e.g.,

the average number of visits per Dakota skipper site per year ranges from 1 to 11) and the probability of false absence after 5-6 visits drops below 5 percent for studied butterfly species with varying average detection probabilities (Pellet 2008, p. 159). Therefore, the site is considered “extirpated” if there are three sequential years of negative surveys.

It cannot be presumed that the species is not persisting at a site only because there have not been recent surveys. At several sites, the species has persisted for longer than 20 years; for example, Dakota skipper was first recorded at Scarlet Fawn Prairie in South Dakota in 1985 and has had positive detections every survey since that date—the most recent detection was in 2012. The year 1993 was chosen based on habitat-related inferences, specifically, the estimated time for prairie habitat to degrade to non-habitat due to woody encroachment and invasive species. For example, native prairies with previous light-grazing management that were subsequently left idle transitioned from mixed grass to a mix of woody vegetation and mixed grass in 13 years and it was predicted that these idle prairies would be completely lost due to woody succession in a 30-year timeframe (Penfound 1964, pp. 260-261). The time for succession of idle prairie depends on numerous factors, such as the size of the site, edge effects (the changes that occur on the boundary of two habitat types), and the plant composition of adjacent areas.

This approach is the most objective way to evaluate the data range-wide. Most sites have been surveyed over multiple years, although the frequency and type of surveys varied among sites and years. In several cases, species experts provided input on occupancy based on their familiarity with the habitat quality and stressors to populations at particular sites.

To summarize, there are few sites with relatively older data where we consider the species to still be present. In general, most sites with a present status have had a positive detection in 2002, or more recently with a few exceptions. At one Poweshiek skipperling site, the species was observed at the site in 1996, and no further surveys have been conducted. The remaining Poweshiek skipperling sites where the species is considered present have had detections in 2012, except one site where the species was detected in 2011 and no further surveys have occurred. Likewise, at four Dakota skipper sites we consider the species to be present with the most recent record from 2001 or earlier including one site where the most recent survey was from 1993, two sites with 1996 records, and one site with a 1998 record. No evidence suggests that the species is not still present at these sites because the best information indicates that the site's habitat is still conducive to the butterfly, and, therefore, the species may still be present there. We also consider Dakota skipper to be present at the following sites: 20 sites in Canada that were surveyed only once in 2002; 1 additional site with a 2002 detection of the species and a favorable habitat assessment in 2012; 1 site with a 2003 detection; 1 site with a 2005 detection; 2 sites with a 2006 detection; 25 sites in Canada that were surveyed only once in 2007; 1 additional site with a 2007 detection; 7 sites with a positive detection in 2008; 2 sites with a positive detection in 2009; and 27 sites with positive detections in 2012.

Population Distribution and Occupancy Status

Once found in native prairies in five states and two Canadian provinces, the Dakota skipper and its habitat have undergone dramatic declines; the species is now limited to native prairie remnants in three states and two Canadian provinces. The Dakota skipper is presumed extirpated from Illinois and Iowa and no longer occurs east of western Minnesota—an approximately 690-kilometer (km) (430-mile) reduction of its range. Populations persist in

western Minnesota, northeastern South Dakota, North Dakota, southern Manitoba, and southeastern Saskatchewan. Royer and Marrone (1992a, p. 5) stated that Dakota skippers may also occur in far eastern Montana and southeastern Saskatchewan, in habitats similar to those occupied by the species in northwestern North Dakota. The Dakota skipper was subsequently found in Saskatchewan in 2001 after 40 years of searching (Hooper 2002, pers. comm.), but Royer (2002, pers. comm.) no longer thinks that the species occurs in Montana.

From its earliest identification, the Dakota skipper was considered rare (Royer and Marrone 1992a, p. 1), although considerable destruction of its habitat likely occurred even before the species was first described in 1911. Habitat destruction and degradation has greatly fragmented Dakota skipper's range from its core through its northern and western fringes (McCabe 1981, p. 179; Royer and Marrone 1992a, p. 28; Schlicht and Saunders 1994, p. 1; Royer 1997, p. 2; Schlicht 1997a, p. 2; Schlicht 1997b, p. 2; Skadsen 1997, pp. 25-26; Skadsen 1999, p. 15; Swengel and Swengel 1999, p. 267). The historical distribution of Dakota skippers may never be precisely known because “much of tallgrass prairie was extirpated prior to extensive ecological study” (Steinauer and Collins 1994, p. 42), such as butterfly surveys. Destruction of tallgrass and mixed-grass prairie began in 1830 (Samson and Knopf 1994, p. 418), but significant documentation of the ecosystem's butterfly fauna did not begin until about 1960. Therefore, most of the species' decline probably went unrecorded. Based on records of vouchered specimens, however, we know that Dakota skipper range has contracted northward out of Illinois and Iowa. The species was last recorded in Illinois in 1888 (McCabe 1981, p. 191) and in Iowa in 1992 (Orwig and Schlicht 1999, p. 6). Britten and Glasford's (2002, pp. 363, 372) genetic analyses support the presumption that this species formerly had a relatively continuous distribution; the small genetic divergence (genetic distance) among seven sites in Minnesota and South Dakota indicate that populations there were once connected. Dakota skipper dispersal is very limited due in part to its short adult life span and single annual flight. Therefore, the species' extirpation from a site is likely permanent unless it is within about 1 km (0.62 mi) of a site that generates a sufficient number of emigrants or is artificially reintroduced to a site.

The Dakota skipper's range once comprised native prairie in five states and Canada, extending from Illinois to Saskatchewan; it now occurs only in native prairie remnants in portions of three states and two Canadian provinces. Of the 259 historically documented sites, there are 91 sites where we consider the Dakota skipper to be present, 81 sites with unknown status, 40 possibly extirpated sites, and 47 that are considered extirpated (Table 1). Approximately half (45 of 91) of the sites where the species is considered to be present are located in Canada, mostly within three isolated complexes, and were observed in either 2002 or 2007 with no subsequent surveys. The remaining 46 sites where the species is considered to be present are about equally distributed among Minnesota (14 sites), North Dakota (18 sites), and South Dakota (14 sites). Researchers made positive detections of the species in 27 of these sites in 2012. Other sites with a present status with relatively older positive detections and no subsequent surveys for the species include 2 sites with positive detections in 1996, one site with a positive detection in 1998, one site with a positive detection in 2002, one site with a positive detection in 2003, one site with a positive detection in 2005, 2 sites with a positive detection in 2006, one site with a positive detection in 2007, 7 sites with a positive detection in 2008, and 2 sites with a positive detection in 2009. At several of these sites, the habitat has been assessed more recently than they were surveyed for the species. The distribution and status of Dakota skipper in each state of known historical or extant occurrence are described in detail below.

Table 1—Number of Historically Documented Dakota Skipper Sites Within Each State and the Number of Sites Where the Species Is Thought To Be Present, Unknown, Possibly Extirpated, or Extirpated

State

Present

Unknown

Possibly

extirpated

Extirpated

Total

Percent of total number of historical sites by state

Illinois

1

1

0.4

Iowa

3

3

1

Minnesota

14

22

18

12

66

26

North Dakota

18

13

10

13

54

21

South Dakota

14

46

10

15

85

33

Manitoba

31

0

2

3

36

14

Saskatchewan

14

0

0

0

14

5

Total Number of Historically Documented Sites

91

81

40

47

259

Percent of the Total Number of Historical Sites by Occupancy

35

31

16

18

100

Illinois

Dakota skippers are considered to be extirpated from Illinois. The species was last recorded near Chicago in 1888 (McCabe 1981, p. 191).

Iowa

There are three historical records of Dakota skippers in three counties in Iowa (Dickinson, Poweshiek, and Woodbury), but the species is presumed extirpated from the State (Schlicht and Orwig 1998, pp. 84-85; Selby 2004a, pp. 1, 5; Selby 2012, pers. comm.; Nekola and Schlicht 2007, p. 9). The species was last seen at Cayler Prairie (Dickinson County) in 1992, but surveys of this site in 2000, 2004, 2005, and 2007 were negative, so we presume it to be extirpated from that site (Schlicht and Orwig 1998, p. 85; Selby 2004a, p. 5; Selby 2006a, p. 5; Selby 2008, p. 6).

The species was not observed at eight sites surveyed between 1988-1997 (Swengel and Swengel 1999, pp. 288-289), at eight sites surveyed in 2004 (Selby 2004a, p. 5), nor during extensive surveys at 32 sites in 2007 (Selby 2008, p. 6).

Minnesota

Minnesota historically contained about 26 percent of the sites where the Dakota skipper has been recorded (Table 1) (Service 2013, unpubl. geodatabase). Since the earliest known record (1965) of the species in Minnesota, 66 sites have been recorded in the State, but recent surveys indicate that the species is declining in the State (Service 2013, unpubl. geodatabase). Of the 66 known locations of Dakota skipper in Minnesota; the species is extirpated or possibly extirpated from 30 of those sites and the status is unknown at 22 others (Service 2013, unpubl. geodatabase). Dakota skipper is considered to be present at 14 sites in Minnesota in 6 counties: Clay, Lincoln, Murray, Norman, Pipestone, and Pope, although 2 of those sites have not been surveyed since 1996 and 1 site has not been surveyed since 1993.

McCabe (1981, p. 187) observed very stable population numbers in Minnesota prairies that he visited repeatedly from 1968-1979. On dry-mesic prairie in Lincoln County, Minnesota, Dana (Dana 1997, pp. 3-5) also observed stable numbers into the thousands during his intensive studies from 1978 to 1983. Schlicht (1997a, p. 13) and Reiser (1997, p. 16) reported more variable numbers on the same sites in 1995-1996, and based on these more recent observations, Dana (1997, pp. 3-5) suggested that populations could experience significant size fluctuations between years. At Hole-in-the-Mountain preserve, Minnesota, Dana (1991, pp. 36-37) found peak abundance of approximately 1,000 Dakota skippers over about 40 ha (98 ac); he estimated that 2,000-3,000 individuals may have been alive at various times during the flight period and that only one-third to one-half of adults were alive simultaneously. Where they occur, these high adult densities persist for only about a week to 10 days during the single annual flight period (Selby and Glenn-Lewin 1989, pp. 24-28).

The percentage of sites surveyed each year in Minnesota with positive detections remained relatively stable from 1985 to 2005, with an average detection rate of 67 percent for all survey years with more than one site surveyed (excluding sites newly discovered in the first year it was discovered), an average of 70 percent detection rate for survey years with 5 or more sites surveyed and an average of 66 percent detection rate for survey years with 10 or more sites surveyed. One exception to the high detection rates was 1994; only 26 percent (5 of 19 sites) of sites surveyed in 1994 resulted in positive detections. Recent surveys of the species resulted in significantly lower than average positive detections. The percent of sites surveyed each year with positive detections has recently decreased from 70 percent (7 of 10 sites) in 2005, to 47 percent (8 of 17 sites) in 2007, to 56 percent (10 of 18 sites) in 2008, to 6 percent (1 of 18 sites) in 2012 (for years with greater than 10 sites surveyed, see Figure 1). Only one individual was detected in Minnesota during 2012 surveys, which included 18 sites with previous records and 23 prairie remnants without previous records for the species (Dana 2012c, pers. comm.; Runquist 2012a, pers. comm.; Olsen 2012, pers. comm.). The cause for this sharp decline is unknown.

EP24OC13.088

The Dakota skipper is presumed extirpated at 12 sites in Minnesota; at 7 of these sites the species has not been observed since 1984 or earlier. Four sites at which the species is now presumed to be extirpated have had fairly recent positive observations. The species was last observed at Prairie Waterfowl Production Area (WPA) in Big Stone County in 2000 (Skadsen 2000, p. 1), for example, but was not found in 2008 (Selby 2009a, p. i), 2010, and 2012 (Service 2013, unpubl. geodatabase). Dakota skippers were observed at the Glacial Lakes WPA in 2001 (Schlicht 2001b, p. 18), but the species was not observed in 2003, 2004, and 2005 (Selby 2006b, p. Appendix A xii); the species is now considered to be extirpated at that site (Service 2013, unpubl. geodatabase). The last observation of Dakota skipper at the Big Stone National Wildlife Refuge (NWR) in Lac Qui Parle County was in 2000, and it was not observed during surveys in 2009, 2011, or 2012 (Skadsen 2012a, p. 5). Dakota skippers were observed at Chippewa Prairie in 1995, but not in 1996, 2005, and 2012 (Service 2013, unpubl. geodatabase). Of the 18 sites where the species is possibly extirpated, 10 have not been surveyed since the species was last seen in 1988 or earlier. Dakota skippers at two of the sites where the species is possibly extirpated have not been observed since 1991 (Service 2013, unpubl. geodatabase). The remaining 6 sites had positive observations prior to 1993, were surveyed once more recently, and had a negative observation (Service 2013, unpubl. geodatabase).

The status of Dakota skipper is unknown at 22 sites; Dakota skipper have not been observed at 11 of these sites since the mid- to late 1990s, despite one or two years of survey effort at several sites. The remaining 11 sites with unknown status have had positive observations in 2007 or more recently, but are given this designation due to a subsequent negative survey. For example, Dakota skipper was documented at the Gens Prairie in Murray County and Woodstock Prairie in Pipestone County in 2007, but the species was not observed during surveys in 2008 (Selby 2009a, p. Appendix 5 li, xxxiii and Appendix 4 xlix).

In 2007 and 2008, the Minnesota DNR carried out a broad survey effort to assess the status of Dakota skipper and other prairie butterflies in the State after experts noted significant declines in these species in west-central Minnesota beginning in 2003 (Selby 2006b, p. 30). Researchers surveyed 17 and 19 sites with previous Dakota skipper records in 2007 and 2008, respectively; Dakota skipper was found at 8 sites each year and at 1 site where it had not previously been recorded (Selby 2009a, p. 6). The surveys confirmed Dakota skipper's extirpation from one site in Cottonwood County, where it was last recorded in 1970.

A parallel study in 2007 (Dana 2008), consisted of more intensive work at a few sites thought to contain some of the State's most viable populations of Dakota skipper. Among these sites was The Nature Conservancy's Hole-in-the-Mountain preserve in Lincoln County, which was the only Minnesota population rated as secure in 2002 (Cochrane and Delphey 2002, p. 16). The 2007 surveys indicated that the site still supported a substantial population, but that it may have decreased in size since earlier studies were conducted (Dana 1991, p. 36; Dana 2008, p. 18). Dakota skippers were not detected during the 2012 flight period (Runquist

2012, pp. 13-14, 18-20; Runquist 2012a, pers. comm.); therefore, we consider the status of the species at the Hole-in-the-Mountain preserve to be unknown.

Relatively important populations of Dakota skipper in Minnesota may still occur at the Prairie Coteau, Felton Prairie, and Glacial Lakes complexes, but the 2012 survey results raised concern for the species' status at Prairie Coteau. The number of Dakota skippers encountered per 100 m (328 ft) of transect at Prairie Coteau State Natural Area (SNA) were 1.7 in 1990 and 1.1 in 2007 (Dana 2008, p. 19). No Dakota skippers were observed at Prairie Coteau SNA during the 2012 flight period (Runquist 2012, pp. 9-10); therefore, we consider the status of the species to be unknown at that site. Selby (2009b, Appendix 4, p. iv) recorded 14 Dakota skippers during a 5-hour survey in 2007 at the Felton Prairie SNA. During a one-hour survey in 2008, nine Dakota skippers were recorded and with little indication of any substantial change since the previous year (Selby 2009b, Appendix 5, p. iv); Felton Prairie has not been resurveyed since 2008 (Service 2013, unpubl. geodatabase). The number of Dakota skippers recorded during recent surveys at Glacial Lakes State Park has been low despite good habitat conditions. An apparently widespread population was present as recently as 2001 when Skadsen (2001, p. 24) found Dakota skippers along almost all of 25 mi (40 km) of transect in and around the park—he recorded as many as 31 Dakota skippers along one transect (Skadsen 2001, p. 24). Selby (2009a, p. l and liv) surveyed the same areas in 2007 and 2008, describing habitat at survey sites as good to excellent, but recorded only eight Dakota skippers during about seven hours of surveys in and around the park (Selby 2009a, p. 1 and liv). Glacial Lakes State Park surveys conducted in 2012 were outside of the Dakota skipper flight period (Runquist 2012a, pers. comm.).

In summary, the Dakota skipper is now considered to be extirpated or possibly extirpated from at least 30 of the 66 sites in Minnesota, which historically contained approximately 26 percent of all known historical Dakota skipper locations rangewide (Table 1). The species is considered to be present and unknown at 14 and 22 sites, respectively. However, only one individual male was detected in the State during 2012 surveys, which included 18 sites with previous records; 2012 surveys for undiscovered populations were also carried out on 23 prairie remnants without previous records for the species. Similar surveys of prairie remnants with no previous documentation of Dakota skipper were completed in Minnesota in 2007 and 2008. Based on these surveys, the likelihood that significant undiscovered Dakota skipper populations occur in Minnesota is low.

North Dakota

North Dakota historically contained approximately 21 percent of all known historical locations of Dakota skippers rangewide (Table 1); the State contained 54 historical sites distributed among 18 counties (Service 2013, unpubl. geodatabase). The Dakota skipper is currently present at 18 sites in 5 North Dakota counties, of these, 13 occur within the Towner-Karlsruhe complex in McHenry County, 1 is within the Sheyenne National Grasslands complex in Ransom County, 2 in northern McKenzie County, 1 site is in Wells County, and 1 site in McLean County. Of the 18 sites where we consider the Dakota skipper to be present, 15 sites had positive observations of the species in 2012 and the remaining 3 sites had positive observations between 1998 and 2003. The status of the species is unknown at 13 sites; 10 of these sites have not had positive records since the mid- to late 1990s and the other 3 sites had positive records between 2001 and 2003. The Dakota skipper is presumed extirpated from 13 sites and 4 counties, primarily due to heavy grazing, weed control, and other disturbances (

e.g.,

bulldozing at Killdeer Mountain to reduce aspen growth, Royer 1997). The species is possibly extirpated from 10 additional sites and 3 additional counties. Researcher surveyed 25 sites, believed to possibly have Dakota skipper populations, in 2012; of these sites, 23 had previous records of the species (Royer and Royer 2012a, entire). Thirteen of the 25 surveyed sites had Dakota skipper present (Royer and Royer 2012a, pp. 3-4; Royer and Royer 2012b, pp. 2-3). One new site was found in 2012 (Royer and Royer 2012a, p. 33), adjacent to a site with previous records but with different land-ownership, so the researcher considered it a new site. Another new site was found in North Dakota in 2012, in Wells County, where two observations were made—possibly the same individual (HDR, Inc. 2012, pp. 21-23). At sites with Dakota skipper, lower average encounter frequencies were observed across the State in 2012 (state average = 9.4 encounters per hour) than during the 1996-1997 statewide surveys (state average = 17.4 encounters per hour) (Royer and Royer 2012b, p. 5; Royer and Royer 2012a. pp. 7-8).

Of the Dakota skipper populations in North Dakota, none may be secure, although the Towner-Karlsruhe complex was considered to be the stronghold for the species in the State in 2002 (Cochrane and Delphey 2002, p. 17) and most of the sites where the species is currently present are still occupied by “viable populations” (Royer 2012a, pers. comm.). All of the habitat where the species is present in the Towner-Karlsruhe complex is Type A (wet-mesic) habitat (Royer and Marrone 1992a, p. 21-22; Royer

et al.

2008, pp. 14-16). Five sites within the Towner-Karlsruhe complex are owned by the North Dakota State Land Department, and the remaining seven sites with extant populations are privately owned. Some Towner-Karlsruhe sites are linked by highway rights-of-way that contain native prairie vegetation and by other prairie remnants (Royer and Royer 2012a, p. 18). In 2002, none of these sites were described as secure (Cochrane and Delphey 2002, pp. 66-67) since each is subject to private or State management options that could extirpate Dakota skipper from the site. In 1999, it was estimated that about 30 percent of the Towner-Karlsruhe area still contained native prairie (Lenz 1999b, p. 2); more recent observations indicate that several native prairie sites have been invaded to varying extents by nonnative species, such as leafy spurge, Kentucky bluegrass, and alfalfa (

Medicago sativa

), and several are subject to intense grazing or early haying (Royer and Royer 2012b, pp. 5-6, 7-10, 13-16, 18-19, 22-23; Royer 2012, in litt.).

Dakota skipper populations in the Sheyenne National Grasslands complex have experienced intensive grazing, leafy spurge (

Euphorbia esula

) invasion, and the effects of herbicides used to control leafy spurge and grasshoppers (Royer 1997, pp. 15 and 27). For example, McCabe (1979, p. 36) cited the McLeod Prairie in the Sheyenne Grasslands in southeastern North Dakota as the best site for Dakota skippers in North Dakota. Since then, however, leafy spurge invasion has significantly modified the habitat and the Dakota skipper is now extirpated from the site (Royer 1997, p. 14). Swengel and Swengel (1999, p. 286) did not find Dakota skippers at eight survey sites in the Sheyenne grasslands during 1988-1997, although Royer did observe a few isolated Dakota skippers in the Sheyenne National Grasslands during this period (e.g., Royer 1997, pp. 14-15). Dakota skippers were recorded at one new site (Gregor) in the Sheyenne National Grasslands in 2001 (Spomer 2004, pp. 14-15). The status of Dakota skipper at the Gregor site is currently

unknown, since the species was not observed during the 2002 survey (Royer and Royer 2012a, pp. 3-4). Orwig (1996, p. 3) suggested that Brown's Ranch in Ransom County, owned by The Nature Conservancy, had potential to support a metapopulation (groups of local populations interconnected by dispersal habitat) in the Sheyenne River watershed. More recently, however, Spomer (2004, p. 36) found that the population there was not doing well, and Royer failed to find the species in 2012 (Royer and Royer 2012a, p. 3). Therefore, the status of the species at the Brown Ranch site is unknown. Royer (1997, pp. 15 and 27) claimed that, throughout the Sheyenne Grasslands, both public and private lands have been so heavily grazed and altered by grasshopper and leafy spurge control that extirpation of Dakota skippers from the area is almost certain to occur. The population at Venlo Prairie, for example, deteriorated from good/fair in 2001 to poor in 2003 due to intense grazing and disappearance of flowers (Spomer 2004, pp. 9, 12); the species is now considered to be extirpated at that site.

In 2002, experts ranked all sites outside of the two complexes discussed above as threatened or vulnerable; most were small and isolated populations threatened by conversion and invasive species (Cochrane and Delphey 2002, pp. 66-67). Most of these sites are now considered extirpated or possibly extirpated. Today, only 4 sites outside of the Towner-Karsruhe Complex and Sheyenne National Grasslands complexes are thought to have extant (present) Dakota skipper populations, including Garrison Training Center in McLean County. In addition to the Towner-Karsruhe Habitat Complex sites in McHenry County, only 2 of the 25 sites surveyed by Royer in 2012, both in northern McKenzie County, may have “viable populations” (Royer 2012b, pers. comm.), although only one individual was observed at each site in 2012 (Royer and Royer 2012b, pp. 16-17).

In summary, North Dakota contains approximately 21 percent (N= 53) of all known historical locations of the species rangewide; however, the current occupancy status of the Dakota skipper is unknown at 12 sites, and it is considered to be extirpated or possibly extirpated from at least 23 of the 53 known sites in the state (Table 1). The species is considered to be present at only 18 sites in the State. North-central North Dakota may hold hope for the species' long-term conservation. Dakota skipper was detected at 13 of the 25 sites surveyed during 2012 (23 of the sites had previous Dakota skipper records); average encounter frequencies observed across the State in 2012 (9.4 encounters per hour), however, were lower than during the 1996-1997 statewide surveys (ND state average = 17.4 encounters per hour).

Although only a small fraction of all grassland in North Dakota has been surveyed for Dakota skippers, a significant proportion of the un-surveyed area is likely not suitable for Dakota skipper. The species was never detected at approximately 135 additional locations in North Dakota that were surveyed for the species from 1991-2012 (USFWS 2013, unpubl. geodatabase). Many of these sites have been surveyed multiple times over multiple years (USFWS 2013, unpubl. geodatabase). Surveys for the Dakota skipper are typically conducted only in areas where floristic characteristics are indicative of their presence. New potential sites surveyed are generally focused on prairie habitat that appear suitable for the species and have a good potential of finding the species, in other words, sites are not randomly selected across the landscape. Therefore, these sites have a higher likelihood of detecting the species than at sites randomly selected across the landscape. Based on these surveys, the likelihood that significant numbers of undiscovered Dakota skipper populations occur in North Dakota is low. Moreover, data available from the numerous sites that have been surveyed are likely to be representative of areas that have not been surveyed—that is, population trends and the nature and extent of stressors that may impact the populations in un-surveyed areas can reasonably be inferred by analyzing data collected from the sites that have been surveyed.

South Dakota

South Dakota historically contained approximately 33 percent of all known locations of Dakota skippers rangewide (Table 1). Since the earliest known record of Dakota skipper (1905) of the species in South Dakota, 85 sites have been documented across 11 counties in the State, but recent surveys indicate that the species is declining in the State (Service 2013, unpubl. geodatabase). Of the 85 historical sites, Dakota skipper is presumed extirpated from 15 sites and 2 counties (Brown and Moody), and is possibly extirpated from 10 additional sites. Dakota skipper is considered present at 14 sites and the status of the species is unknown at 46 sites. Twenty-six sites in South Dakota with previous Dakota skipper records were surveyed in 2012; the species was detected at 9 of those sites (Service 2013, unpubl. geodatabase). Eight additional sites within the species' historical range were surveyed during the 2012 flight period, which resulted in the discovery of two new nearby Dakota skipper sites (Service 2013, unpubl. geodatabase; Skadsen 2012a, pers. comm.). The proportion of positive surveys at known sites has fluctuated over time; however, the 2012 surveys had the lowest positive detection rate (35 percent) for the last 16 years (since 1996), much less than comparable survey years (years with 10 or more sites surveyed) in South Dakota.

While there are some sites with earlier records, most South Dakota sites were initially documented during extensive surveys conducted during 1996 to 1998. Forty-eight locations without previous records were surveyed during 2002-2004, which resulted in the discovery of 20 new Dakota skipper sites in northeastern South Dakota (Skadsen 2003, p. 8; Skadsen 2004, pp. 3-6), but due to more recent negative surveys, the occupancy of the species is currently unknown or extirpated at many of these sites (Skadsen 2011, p. 5; Skadsen 2012, pp. 4-5; Skadsen, 2012, pers. comm.; Skadsen 2003, p. 10; Skadsen 2004, p. 2;; Skadsen 2006a, p. 2, 10; Skadsen 2006b, p. 5; Skadsen 2007, p. 3; Skadsen 2008, p. 3, 12; Skadsen 2009, p. 3). Additional survey effort resulted in the discovery of nine new sites between 2005 and 2012, with a maximum of three new sites discovered in 2006 (Skadsen 2010a, p. 6; Skadsen 2012, pp. 4-5; Skadsen 2012, pers. comm.; Skadsen 2005, pp. 5-6, Skadsen 2006a, p. 12; Skadsen 2006b, p. 5; Skadsen 2007, p. 3; Skadsen 2008, p. 9; Skadsen 2009, p. 2). Eight additional sites without previous documentation of the species were surveyed in 2012, which resulted in the discovery of two nearby sites (Service 2013, unpubl. geodatabase). To summarize, new sites have been discovered in South Dakota during most survey years since 2002, however, the number of new sites discovered each year has been low recently; 2 or 3 new sites have been discovered each survey year since 2005 (3 sites in 2005, 2 sites in 2006, 2 sites in 2007, zero sites in 2010, and 2 sites in 2012). The rate that known sites are becoming extirpated is higher than the rate of new discovery—the occupancy of the species at many sites is now unknown or extirpated due to more recent negative surveys.

The species has never been documented in Clark County, but because few surveys have been conducted there, the county may contain undiscovered populations (Skadsen 2006b, p. 1). Skadsen (2012b,

pers. comm.) doubts the existence of public lands with suitable Dakota skipper habitat in Clark County and has not received permission to survey a few possible suitable locations that are privately owned.

Although only a small fraction of all grassland in eastern South Dakota has been surveyed for Dakota skippers (

e.g.,

Dakota skipper surveys have been conducted on less than approximately 30,000 acres (12,140 ha) in South Dakota within the species range (Service 2013, unpubl. geodatabase)), a significant proportion of the un-surveyed area is likely not suitable for the Dakota skipper. For example, there is an estimated 1,620,549 acres (ac) (655,813 hectares (ha)) of unbroken (untilled) grasslands (excluding Conservation Reserve Program (CRP) grasslands, which generally do not provide habitat for the Dakota skipper (Larson 2013, pers. comm.)) in the 9 counties where the Dakota skipper is considered be present or to have unknown occupancy in South Dakota (HAPET 2012, unpubl. data). Additional areas of unbroken prairie were estimated in three other counties where the species may have occurred historically (HAPET 2012, unpubl. data). While these lands represent unbroken grassland in South Dakota, the models used to identify unbroken grassland are not able to identify plant species, plant species composition, floristic quality, or presence of invasive species (Loesch 2013 pers. comm.). Therefore, these unbroken grasslands may not contain the specific native prairie plants that the Dakota skipper requires (as discussed in detail in the Background section of this proposed rule) and, therefore, may not equate to suitable habitat for the species.

The species was never detected at approximately 73 additional locations in South Dakota that were surveyed from 1991 through 2012 (USFWS 2013, unpubl. geodatabase). Several of these sites have been surveyed multiple times in one year or during multiple years (USFWS 2013, unpubl. geodatabase). Surveys for Dakota skipper are typically conducted only in areas where floristic characteristics are indicative of their presence. For example, in South Dakota, Skadsen (1997, p. 2) selected for surveys dry-mesic prairie that supported purple coneflower and wet-mesic prairie that supported wood lily and mountain deathcamas based on searches for these sites by car and reports from resource managers. New potential sites surveyed are generally focused on prairie habitat that appear suitable for the species and have a good potential of finding the species, in other words, sites are not randomly selected across the landscape. Therefore, these sites have a higher likelihood of detecting the species than at sites randomly selected across the landscape. Based on these surveys, the likelihood that significant undiscovered Dakota skipper populations occur in South Dakota is low. Moreover, data available from the numerous sites that have been surveyed are likely to be representative of areas that have not been surveyed—that is, population trends and the nature and extent of stressors that may impact the populations in un-surveyed areas can reasonably be inferred by analyzing data collected from the sites that have been surveyed.

Since there is little long-term quantitative data for sites in South Dakota, we examined presence-absence (non-detection) data over time. The percent of sites surveyed each year with positive detections of the species remained relatively stable from 1985 to 2010, with an average positive detection rate of 63 percent for all survey years with more than one site surveyed (excluding new sites for the first year of discovery), an average positive detection rate of 60 percent for survey years with at least 5 sites surveyed, and an average positive detection rate of 71 percent for survey years with at least 10 sites surveyed. One exception to the high detection rates was during the 1991 survey year when none (0 of 7 sites) of the sites surveyed in 1991 resulted in positive detections of the species, excluding 3 new sites that were discovered that year. Another exception was in 1996, when 2 of the 8 sites with previous records surveyed had a positive detection; however, 6 new sites were discovered that year. The detection rate remained relatively stable until 2010, when the percent of sites with positive detections fell from 89 percent (8 of 9 sites) in 2010, to 46 percent (5 of 11 sites) in 2011, and 35 percent (9 of 26 sites) in 2012 (Figure 2). These types of fluctuations had been observed in prior years; therefore, it is difficult to determine a clear trend in the data using positive detections—the last two survey years may fall within the normal range of variation.

EP24OC13.089

The Outer Coteau des Prairies sub-section of the North Central Glaciated Plains section of Bailey's Eco-regions is thought to be a stronghold for Dakota skipper, since nearly 40 percent of the total documented Dakota skipper sites are within that subsection (83 of the 259 documented sites—Service 2013, unpubl. geodatabase). Most of these Outer Coteau des Prairie sites are in South Dakota; 73 of the 85 Dakota skipper sites in South Dakota are within the Outer Coteau des Prairies subsection (Service 2013, unpubl. geodatabase). Dakota skipper is considered to be present at only 10 of those 73 sites—the species status is unknown at 41 of those sites, possibly extirpated at 8 sites, and extirpated at the remaining 13 sites within that ecoregion subsection in South Dakota (Service 2013, unpubl. geodatabase).

In summary, South Dakota historically contained approximately 33 percent of all known locations of the species rangewide. The current occupancy status of the Dakota skipper is unknown at 46 sites and it is considered to be extirpated or possibly extirpated from at least 25 of the 85 known sites in the State, although large areas of grasslands remain in South Dakota and substantial additional populations of Dakota skipper would be expected to be found if more surveys were conducted. Furthermore, downward trends and threats impacting populations at known sites are also likely occurring at potentially undiscovered sites. The species is considered to be present at 14 of the 85 documented sites in the State. Twenty-six sites in South Dakota with previous Dakota skipper records were surveyed in 2012; the species was detected at nine of those sites; eight sites with no previous records for the species were surveyed during the 2012 flight period, which resulted in the discovery of two nearby sites. The proportion of positive surveys at known sites has fluctuated over time; however, the 2012 surveys had the lowest positive detection rate (35 percent) for the last 16 years (since 1996)—much less than comparable survey years in South Dakota.

Manitoba

Manitoba historically contained approximately 14 percent (N = 36) of the known locations of the Dakota skipper rangewide. The Dakota skipper is considered present at 1 isolated site and 30 sites split between 2 distinct complexes, 14 sites near Griswold and 16 sites along Lake Manitoba. The 14 sites near Griswold are located approximately 200 km (124 mi) southwest of the populations along Lake Manitoba (at 16 sites) and about 125 km (78 mi) northeast of the nearest population in Saskatchewan (Webster 2003, pp. 5-6; Webster 2007, p. 4). The species is presumed extirpated or possibly extirpated from five sites in Manitoba, including from the Tallgrass Prairie Preserve, where it has not been found in the seven most recent survey years (Webster 2003, p. 5; Westwood

et al.

2012, p. 1; Westwood 2007, pers. comm.; Hamel

et al.

2013, pp. 8-16)—(the later surveys were focused on Poweshiek skipperlings, but other species were recorded) and one site that was converted to a flaxseed field (Webster 2003, p. 7). Population

estimates and trends at these sites have not been examined quantitatively; however, the population appears to be stable at two sites with repeated survey years. Numbers observed during searches at a site near Griswold in 2007 did not appear to change appreciably since 2002 surveys, when the population was estimated (non-quantitatively) to be approximately 750 individuals (Webster 2003, p. 5; Webster 2007, p. 4). A total of 273 adults were observed during a 3.3-hour survey at the second site, where the population was estimated non-quantitatively to be about 2,000 individuals (Webster 2007, p. 4).

Dakota skipper was first recorded near Miniota in 1944 and then at two additional sites in the early 1990s. In 2002, the species was observed at 19 sites near Lundar, within about 25 km (16 mi) east of Lake Manitoba (Webster 2003, p. 4); however, most of these sites have not been surveyed since. In 2007, researchers surveyed 16 sites for the Dakota skipper near Griswold, Manitoba (Webster 2007, p. 4) and found Dakota skippers at 14 of the 16 sites; 12 of these represent new sites for the species in Manitoba (Webster 2007, p. 4). Several additional areas were examined for potential Dakota skipper habitat in 2007, including areas east of Hwy 21, within the Lauder Sandhills Wildlife Management Area, north of Oak Lake and near Tilston, Sinclair, Cromer, and Brandon, as well as other locations. Most of the areas examined were under row crop agriculture, were heavily grazed, were dry scrub prairies or were otherwise habitats unsuitable for Dakota skipper (Webster 2007, p. 6). The areas near Brandon and the high ground within the wetland complexes near Oak Lake may still contain suitable habitat (Webster 2007, p. 6).

The nearest known extant (present) population of Dakota skippers in Manitoba is approximately 120 km (75 mi) from the closest extant (present) population in North Dakota and about 200 km (125 mi) from the closest Saskatchewan population. Britten and Glasford (2002, pp. 367, 372) suggested that Manitoba populations are genetically distinct from a group of populations in Minnesota and South Dakota, although populations in additional intervening locations should be sampled to confirm this hypothesis (Runquist 2012b, pers. comm.).

Saskatchewan

Saskatchewan historically contained approximately 5 percent (N= 14) of all known records of Dakota skippers rangewide. In Saskatchewan, the Dakota skipper is restricted to undisturbed or lightly grazed, steep, south-facing hills near the Souris River (Webster 2007, p. ii). The Dakota skipper was first recorded south of Oxbow, Saskatchewan, in 2001 where three males were collected (Hooper 2003, p. 124) on an ungrazed knoll within a patch of mixed-grass prairie that was approximately one ha (2 ac) in extent. Dakota skippers were found at three additional sites during 2002 surveys (Webster 2003, pp. 6-7). In 2007, researchers surveyed 16 sites in southeastern Saskatchewan and found Dakota skippers at 10 of these sites (including Oxbow); 8 of these represent new sites for the species in Saskatchewan (Webster 2007, p. i). During 2007 surveys, which were conducted late in the flight period, only a few individuals were observed at each site where the species was present (Webster 2007, p. ii). Nine of these sites where the species was found in 2007 were surveyed along an approximate 50-km (31-mi) stretch of steep hillsides along the ridgeline north of Souris River; distances between sites range from 1 to 28 km (0.8 mi to 17 mi). We consider Dakota skipper to be present at all 14 sites in Saskatchewan, although 3 of those sites have not been surveyed since 2002. The nearest known extant population of Dakota skippers in Saskatchewan is approximately 111 km (69 mi) from the closest extant (present) population in North Dakota and 200 km (125 mi) from the closest Manitoba population.

Poweshiek skipperling

Species Description

The Poweshiek skipperling (

Oarisma poweshiek

) is a member of the skipper family, Hesperiidae, and was first described by Parker (1870, pp. 271-272). Parker (1870, pp. 271-272) provided the original description of this species from his type series collected near Grinnell, Iowa. It was named for the county in which it was found (Poweshiek County), but it was misspelled, Powesheik, in the original description. This spelling was retained by most early authorities (Lindsey 1922, p. 61; Holland 1931, p. 360). Miller and Brown (1981, p. 31) used the corrected spelling, Poweshiek, but then Miller and Ferris (1989, p. 31) changed it back in their supplement. Current usage is mixed, with many authorities retaining the original spelling (e.g., Miller 1992, p. 20), while others have opted for the corrected spelling (Layberry

et al.

1998, p. 48; Opler

et al.

1998, p. 363; Glassberg 1999, p. 167; Brock and Kaufman 2003, p. 306). Layberry

et al.

(1998, p. 48) state “. . . since it is a clear case of an original incorrect spelling it can be corrected [rule 32(c)ii of the International Code of Zoological Nomenclature].”

Poweshiek skipperlings are small and slender-bodied, with a wingspan generally ranging from 2.3 to 3.0 cm (0.9 to 1.2 in). The size of Poweshiek skipperlings appears to vary somewhat across their range (Royer and Marrone 1992b, p. 3). North Dakota and South Dakota specimens tend to be slightly smaller than the 2.9 to 3.2 cm (1.1 to 1.3 in) range given by Parker (1870) for the type specimens from Grinnell, Iowa (Royer and Marrone 1992b, p. 3). A sample of Richland County, North Dakota, specimens from Royer's collection had an average wingspan of 2.8 cm (1.1 in) for males and 3.0 cm (1.2 in) for females. South Dakota specimens in Marrone's collection had an average wingspan of 2.6 cm (1.0 in) for males and 2.7 cm (1.1 in) for females. The upper wing surface is dark brown with a band of orange along the leading edge of the forewing. Ground color of the lower surface is also dark brown, but the veins of all but the anal third of the hindwing are outlined in hoary white, giving an overall white appearance to the undersurface.

The Poweshiek skipperling is most easily confused with the Garita skipperling (

Oarisma garita

), which can be distinguished from Poweshiek skipperling by their smaller size, quicker flight, and overall golden-bronze color (Royer and Marrone 1992b, p. 3). Another distinguishing feature is the color of the anal area of the ventral hindwing (orange in Garita; dark brown in Poweshiek). The Garita skipperling generally occurs west of Poweshiek skipperling range, although there are records of both species from two counties in southeastern North Dakota and two counties in northwestern Minnesota (Montana State University—Big Sky Institute 2012, Butterflies of North America

http://www.butterfliesandmoths.org/

Accessed 5/14/12; Minnesota Department of Natural Resources (DNR) 2012, Rare features database. Accessed 5/14/12).

McAlpine (1972, pp. 85-92) described Poweshiek skipperling eggs as pale yellowish green, mushroom shaped with a flattened bottom, a slightly depressed micropyle (pore in the egg's membrane through which the sperm enter) and smooth surfaced. They were 0.8 millimeters (mm) (0.01 in) long, 0.7 mm (0.03 in) wide and 0.5 mm (0.02 in) high. The overall color of the head and body of the larvae is pale grass green, with a distinctive darker green mid-dorsal stripe and seven cream-colored stripes on each side. First instars were 1.8 mm (0.07 in) at hatching, and the

lone 7th instar survivor was 23.6 mm (1.0 in) near the end of that stage. McAlpine did not have any observations past the 7th instar (the stage between successive molts, the first instar being between hatching and the first molt) (McAlpine 1972, pp. 85-93).

General Life History

Poweshiek skipperlings lay their eggs near the tips of leaf blades and overwinter as larvae on the host plants (Bureau of Endangered Resources in Swengel and Swengel 1999, p. 285, Borkin 2000a, p. 7). McAlpine (1972, pp. 85-92) described the various life-history stages of Poweshiek skipperling. McAlpine (1972, pp. 85-93) observed hatching of larvae Poweshiek skipperling after about nine days. McAlpine's records were incomplete, and he did not have any observations past the 7th instar, but he believed that there should have been one or two additional instars, followed by the chrysalis (pupa) and then the imago (adult) stages (McAlpine 1972, pp. 85-93). After hatching, Poweshiek skipperling larvae crawl to the base of grasses, but unlike Dakota skippers, Poweshiek skipperling do not form shelters underground (McAlpine 1972, pp. 88-92; Borkin 1995a, p. 9; Borkin 2008, pers. comm.). Poweshiek skipperling are not known to form shelters, instead the larvae overwinter up on the blades of grasses and on the stem near the base of the plant (Borkin 2008, pers. comm.; Dana 2008, pers. comm.). Borkin (2008, pers. comm.) observed larvae moving to the tips of grass blades to feed on the outer and thinner edges of the blades, with later movement down and among blades.

Food and Water

For the Poweshiek skipperling, preferred nectar plants vary across its geographic range. Smooth ox-eye (

Heliopsis helianthoides

) and purple coneflower were noted as the favored nectar plants in Iowa, Minnesota, and North Dakota (Swengel and Swengel 1999, p. 280). Other nectar species used, in descending order of number of observations, were stiff tickseed (

Coreopsis palmata

), black-eyed Susan, and palespike lobelia (

Lobelia spicata

) (Swengel and Swengel 1999, p. 280). On drier prairie habitats in Iowa and Minnesota, purple coneflower is used almost exclusively, and the emergence of the adults corresponds closely to the early maturity of this species' disk florets (Selby 2005, p. 5). On the wetter prairie habitats of Canada and the fen habitats of Michigan, favored nectar plants are black-eyed Susan, palespike lobelia, sticky tofieldia (

Triantha glutinosa

), and shrubby cinquefoil (

Dasiphora fruticosa ssp. floribunda

) (Nielsen 1970, p. 46; Holzman 1972, p. 111; Catling and Lafontaine 1986, p. 65; Bess 1988, p. 13; Summerville and Clampitt 1999, p. 231). In addition to nutrition, the nectar of flowering forbs provides water for Poweshiek skipperling, which is necessary to avoid desiccation during flight activity (Dana 2013, pers. comm.).

Until recently, the larval food plant was presumed to be elliptic spikerush (

Eleocharis elliptica

) or sedges, but this was based on limited observations, primarily from the Michigan populations (e.g, Holzman 1972, p. 113). More recent observations show that the preferred larval food plant for some populations of Poweshiek skipperling is prairie dropseed (

Sporobolus heterolepis

) (Borkin 1995b, p. 6); larvae have also been observed feeding on little bluestem (

Schizachyrium scoparium

) (Borkin 1995b, pp. 5-6) and sideoats grama (

Bouteloua curtipendula

) (Dana 2005a, pers. comm.). Poweshiek skipperling have been observed laying eggs (ovipositing) on mat muhly (

Muhlenbergia richardsonis

) (Cuthrell 2012a, pers. comm.), a grass in Michigan's prairie fens (Penskar and Higman 1999, p. 1).

In southwestern Minnesota dry hill prairies, Poweshiek skipperling oviposition was observed on prairie dropseed, little bluestem, big bluestem (

Andropogon gerardii

), porcupine grass (

Hesperostipa spartea

), and a couple unidentified species; a larva was observed feeding on sideoats grama (Dana 2005a, pers. comm.). Poweshiek skipperlings were observed to oviposit on big bluestem in Wisconsin (Borkin 2012a, pers. comm.), although indiscriminate oviposition on unsuitable larval plants has been observed during high summer temperatures (Borkin 1995a, p. 6). Dana (2005b, pers. comm.) noted that larvae and ovipositing females prefer grasses with “very fine, threadlike structures” and hypothesized that Poweshiek skipperling lack a specific host and may adapt to acceptable plant species at a site.

Dispersal

Poweshiek skipperlings are also not known to disperse widely; the species was evaluated among 291 butterfly species in Canada as having relatively low mobility; experts estimated Poweshiek skipperling to have a mean mobility of 2 (standard deviation = 1.4) on a scale of 0 (sedentary) to 10 (highly mobile) (Burke

et al.

2011, p. 2279; Fitzsimmons 2012, pers. comm.). A maximum dispersal distance of 1.6 km (1.0 mi) is estimated to be a reasonable and likely distance for male Poweshiek skipperling to travel between patches of prairie habitat separated by structurally similar habitats (

e.g.,

perennial grasslands but not necessarily native prairie). The species, however, will not likely disperse across habitat that is not structurally similar to native prairies, such as certain types of row crops or anywhere not dominated by grasses (Westwood 2012a and 2012b, pers. comm; Dana 2012b, pers. comm.). In Manitoba, Poweshiek skipperling have been observed avoiding dispersal over short distances, even to suitable habitat, if a barrier such as a road exists between suitable prairie habitat or nectar sources (Westwood

et al.

2012, p.18). Since experts estimated Dakota skippers to have a mean mobility of 3.5 (standard deviation = 0.7) on a scale of 0 (sedentary) to 10 (highly mobile), which is higher than the estimate for Poweshiek skipperling (mean mobility of 2) (Burke

et al.

2011, p. 2279; Fitzsimmons 2012, pers. comm.), a more conservative estimated dispersal distance would be that of the Dakota skipper, approximately 1 km (0.6 mi) (Cochrane and Delphey 2002, p. 6).

In summary, dispersal of Poweshiek skipperling is very limited due in part to its short adult life span and single annual flight. Therefore, the species' extirpation from a site is likely permanent unless it is within about 1 km (0.6 mi) of a site that generates a sufficient number of emigrants or is artificially reintroduced to a site; however, the capability to propagate the Poweshiek skipperling is currently lacking.

Habitat

Poweshiek skipperling habitats include prairie fens, grassy lake and stream margins, moist meadows, and wet-mesic to dry tallgrass prairie. McCabe and Post (McCabe and Post 1977a, p. 38) describe the species' habitat in North Dakota as “. . . high dry prairie and low, moist prairie stretches as well as old fields and meadows.” Royer and Marrone (1992b, p. 12) describe Poweshiek skipperling habitat in North Dakota and South Dakota as moist ground in undisturbed native tallgrass prairies. Poweshiek skipperling habitat throughout Iowa and Minnesota is described as both “high dry” and “low wet” prairie (McCabe and Post 1977a, p. 38). The only documented Illinois record was associated with high rolling prairie (Dodge 1872, p. 218); the only documented Indiana record was from marshy lakeshores and wetlands

(Blatchley 1891, p. 398; Shull 1987, p. 29).

Southern dry prairies in Minnesota are described as having sparse shrub cover (less than 5 percent) composed primarily of leadplant, with prairie rose, wormwood sage, or smooth sumac present and few, if any, trees (Minnesota DNR 2012a, p. 1). Southern mesic prairies also have sparse shrubs (5-25 percent cover) consisting of leadplant and prairie rose with occasional wolfberry (

Symphoricarpos occidentalis

) and few, if any, trees (Minnesota DNR 2012b, p. 1).

The disjunct populations of Poweshiek skipperlings in Michigan have more narrowly defined habitat preferences, variously described as wet marshy meadows (Holzman 1972, p. 114), bog fen meadows or carrs (Shuey 1985, p. 181), sedge fens (Bess 1988, p. 13), and prairie fens (Michigan Natural Features Inventory 2011, unpubl. data; Michigan Natural Features Inventory 2012, unpubl. data); prairie fen is the currently accepted name for this habitat type. Bess (1988, p. 13) found the species primarily in the drier portions of Liberty Fen, Jackson County, dominated by “low sedges” and an abundance of nectar sources. Summerville and Clampitt (1999, p. 231) noted that the population was concentrated in areas dominated by spikerush and that only 10-15 percent of the fen area was occupied despite the abundance of nectar sources throughout. Poweshiek skipperling have been described as occupying peat domes within larger prairie fen complexes in areas either dominated by mat muhly or prairie dropseed (Cuthrell 2013a, pers. comm.). A few prairie fens in Michigan also contain other rare butterflies, such as Mitchell's satyr and swamp metalmark (Cuthrell 2013a, pers. comm.).

Poweshiek skipperling populations in Wisconsin are also disjunct from the population to the west and are associated with areas that contain intermixed wet-mesic, and dry-mesic prairie habitats (Borkin 1995b, p. 6). The dry-mesic habitats contain “extensive patches of prairie dropseed and little bluestem grasses” (Borkin 1995b, p. 7). Survival in wetter areas, which tend to burn cooler and less completely, coupled with low recolonization rates, or the disproportionate loss of wet versus dry prairie could give the false impression that the wet areas were their preferred habitat (Borkin 1995b, p. 7). Like Dakota skipper, Poweshiek skipperling larvae may be vulnerable to desiccation during dry summer months (Borkin 2012a, pers. comm.) and require movement of shallow groundwater to the soil surface or wet low areas to provide relief from high summer temperatures or dry conditions (Royer

et al.

2008, pp. 2, 16; Borkin 2012a, pers. comm.). Humidity may also be an essential factor to larval survival during winter months since the larvae cannot take in water during that time and depend on humid air to minimize water loss through respiration (Dana 2013, pers. comm.). Royer (2008, pp. 14-15) measured microclimalogical (climate in a small space, such as at or near the soil surface) levels within “larval nesting zones” (between the soil surface and 2 cm deep) at six known Poweshiek skipperling sites, and found an acceptable rangewide seasonal (summer) mean temperature range of 18 to 21 °C (64 to 70 °F), rangewide seasonal mean dew point ranging from 14 to 17 °C (57 to 63 °F), and rangewide seasonal mean relative humidity between 73 and 85 percent

Canadian populations of Poweshiek skipperlings are restricted to a single 2,300-ha (5,683-ac) area in southeastern Manitoba (COSEWIC 2003, p. 5). The wet to mesic tallgrass prairie in this area is characterized by low relief (1-2 m (3-7 ft)), with alternating lower, wetter areas and higher, drier prairie; Poweshiek skipperlings tend to be concentrated on or near the edge of the higher, drier prairie (COSEWIC 2003, p. 8). Spikerush is frequent in the wetter areas, and prairie dropseed, black-eyed Susan, and palespike lobelia are frequent in the drier areas (COSEWIC 2003, pp. 7-8).

Prairie fen habitat soils in Michigan are described as saturated organic soils (sedge peat and wood peat) and marl, a calcium carbonate (CaCO

3

) precipitate (MINFI Web site accessed August 3, 2012). In other states, soil textures in Poweshiek skipperling habitats are classified as loam, sandy loam, or loamy sand (Royer

et al.

2008, pp. 3, 10); soils in moraine deposits are described as gravelly, except the deposits associated with glacial lakes.

Population Distribution and Occupancy

The Poweshiek skipperling is historically known from eight states, ranging widely over the native wet-mesic to dry tallgrass prairies from eastern North and South Dakota (Royer and Marrone 1992b, pp. 4-5) through Iowa (Nekola and Schlicht 2007, p. 7) and Minnesota (Minnesota DNR, Division of Ecological Resources, unpubl. data), with occurrences also documented in northern Illinois (Dodge 1872, p. 218), Indiana (Blatchley 1891, p. 898), Michigan (Holzman 1972, p. 111; McAlpine 1972, p. 83), and Wisconsin (Borkin 2011, in litt.; Selby 2010, p. 22). The relatively recent discovery of Poweshiek skipperling populations in the Canadian province of Manitoba further extends its known historical northern distribution (Westwood 2010, pp. 7-22; Dupont 2010, pers. comm.). Additional historical accounts of Poweshiek skipperling from the States of Montana, Colorado, and Nebraska are likely misidentifications of its western congener, the Garita skipperling.

Once common and abundant throughout native prairies in eight states and at least one Canadian province, the Poweshiek skipperling and its habitat have experienced significant declines. The species is considered to be present at a few native prairie remnants in two states and one location in Manitoba, Canada. The species is presumed extirpated from Illinois and Indiana, and the status of the species is uncertain in four of the six states with relatively recent records (within the last 20 years). The historical distribution of Poweshiek skipperling may never be precisely known because “much of tallgrass prairie was extirpated prior to extensive ecological study” (Steinauer and Collins 1994, p. 42), such as butterfly surveys. Destruction of tallgrass and mixed-grass prairie began in 1830 (Sampson and Knopf 1994, p. 418), but significant documentation of the ecosystem's butterfly fauna did not begin until about 1960. Therefore, most of the decline of the Poweshiek skipperling probably went unrecorded. Poweshiek skipperling dispersal is very limited due in part to its short adult life span and single annual flight. Therefore, the species' extirpation from a site is likely permanent unless it is within about 1 km (0.6 mi) of a site that generates a sufficient number of emigrants or is artificially reintroduced to a site.

Recent survey data indicate that Poweshiek skipperling has declined to zero or to undetectable levels at 87 percent of sites where it has ever been recorded. Until about 2003, Poweshiek skipperling was regarded as the most frequently and reliably encountered prairie-obligate skipper butterfly in Minnesota, which contains nearly 50 percent of all known Poweshiek skipperling locations rangewide. Numbers and distribution dropped dramatically in subsequent years, however, and the species has not been seen in Minnesota since 2007. In Iowa, the Poweshiek skipperling was found at 2 of 33 sites with previous records surveyed in 2007; the species was last observed at one site in 2008. Iowa contains about 14 percent of documented sites rangewide. Unidentified threats to the species have acted to extirpate or sharply diminish

populations at all or the vast majority of sites in Iowa and Minnesota (Dana 2008, p. 16; Selby 2010, p. 7).

South Dakota historically contained about 24 percent of the rangewide sites with documented presence of Poweshiek skipperling, although recent surveys in that State also suggest an emergent and mysterious decline. The species was last observed in South Dakota in 2008, at three sites. North Dakota historically contained about six percent of the rangewide sites with documented presence of Poweshiek skipperling; the species was last observed in North Dakota in 2001. Survey efforts in North Dakota have been minimal between 1998 and 2011, but surveys conducted in 1997 documented more than 10 Poweshiek skipperlings at 1 site; 6 individuals were counted at 1 site, and 0 were detected at 6 other sites. Surveys conducted during the 2012 flight season resulted in zero detections of the species.

Seven Michigan sites were recently ranked as having good or better “viability”, a habitat-based element occurrence rank assigned by the Michigan Natural Features Inventory (2011); however, the number of individuals observed at a few of those sites has declined in recent years and the species is presumed extirpated from one of those sites. Currently, four of the ten extant occurrences of Poweshiek skipperling in Michigan are considered to have good or better viability (Michigan Natural Features Inventory (2011, unpubl. data). Each of those faces threats of at least low to moderate magnitude, and the State contains only about 6 percent of all known historical Poweshiek skipperling records. There is one population of Poweshiek skipperling in Wisconsin with fairly consistent numbers observed over the last 5 years (17 to 63 individuals counted, no consistent measure of effort) and one population in Manitoba with fairly consistent numbers (typically hundreds of individuals observed each year). To summarize, of the 296 documented sites, there are 14 sites where we consider the Poweshiek skipperling to be present, 131 sites with unknown status, 98 possibly extirpated sites, and 53 where we consider the species to be extirpated (Table 2). The distribution and status of Poweshiek skipperling in each state of known historical or extant occurrence are described in detail below.

Table 2—Number of Historically Documented Poweshiek Skipperling Sites Within Each State and the Number of Sites Where the Species Is Thought To Be Present, Unknown, Possibly Extirpated, or Extirpated

State

Present

Unknown

Possibly extirpated

Extirpated

Total

Percent of the total number of historical sites by state

Illinois

3

3

1

Indiana

1

1

0.3

Iowa

4

24

13

41

14

Michigan

10

1

6

17

6

Minnesota

67

68

7

142

48

North Dakota

10

6

1

17

6

South Dakota

48

22

70

24

Wisconsin

3

1

4

1

Manitoba

1

1

0.3

Total Number of Historically Documented Sites

14

131

98

53

296

Percent of the Total Number of Historical Sites by Occupancy

5%

44%

33%

18%

Illinois

The Poweshiek skipperling historically occurred in Illinois, although only one historical occurrence is supported (Table 2). In the early 1870s, Dodge (1872, p. 218) reported abundant Poweshiek skipperling occupying “the high rolling prairie that forms the divide between the Illinois and Rock rivers” in Bureau County, Illinois. In addition to Bureau County, the Web site

Butterflies and Moths of North America

lists Poweshiek skipperling historical occurrences for Lake and Mason Counties, which were submitted to the Web site before the date field was required, so a default date of January 1, 1950, was assigned, which is outside of the typical flight period (

http://www.butterfliesandmoths.org/species/Oarisma-poweshiek;

accessed August 16, 2012). The Web site maintains a verifiable database on species occurrences, but there is no accessible supporting data for the Lake and Mason Counties records (Lundh 2012, pers. comm.). Poweshiek skipperling is, therefore, presumed to be extirpated from Illinois.

Indiana

There is one supported historical occurrence of Poweshiek skipperlings in Indiana (Table 2). Blatchley (1891, p. 898) reported small numbers of Poweshiek skipperlings near Whiting, Indiana; Shull (1987, p. 49) expressed confidence that this record is authentic. The Poweshiek skipperling is considered extirpated from Indiana.

Iowa

Iowa historically contained approximately 14 percent (N= 41) of all known records of Poweshiek skipperlings rangewide (Table 2). The Poweshiek skipperling was historically known to occur at 38 sites in 13 counties in Iowa (Nekola 1995, p. 8; Saunders 1995, pp. 27-28; Selby 2005, p. 18; Nekola and Schlicht 2007, p. 7; Selby 2010, p. 6); however, this number may vary slightly (up to 41 sites) depending on how one divides sites along the Little Sioux River in the Freda-Cayler area (Selby 2012a, pers. comm.). Early reports from Parker (1870, p. 271) described Poweshiek skipperling as abundant on a prairie slope at Grinnell, Iowa, while Lindsey (1917, p. 352; 1920, p. 320) noted additional rare occurrences in Story, Dickinson,

Poweshiek, and Woodbury Counties, Iowa—among these, habitat has long since been destroyed in all but Dickinson County.

In 1993-1994, 65 sites were surveyed in 17 counties where Dakota skipper or Poweshiek skipperling had been previously recorded or where prairie and butterfly surveys or infra-red photography suggested the presence of Poweshiek skipperling habitat (Saunders 1995, pp. 7-8). Among the 65 sites surveyed, Poweshiek skipperlings were found at 29 sites in 10 counties (Saunders 1995, p. 27). In 2000, Poweshiek skipperlings were found at six sites surveyed in and near Cayler Prairie and Freda Haffner Kettlehole state preserves in Dickinson County (Selby 2000, p. 19). Followup surveys of this complex in 2004, 2005, and 2007, however, produced no confirmed sightings (Selby 2010, p. 6). Extensive surveys were conducted in 2007, and included 32 of the 38 sites in the State with post-1990 records (Selby 2008, pp. 4, 6). Poweshiek skipperlings were found at 2 of the 38 sites surveyed—Hoffman Prairie State Preserve in Cerro Gordo County and Highway 60 Railroad Prairie in Osceola County (Selby 2008, pp. 6-7). Five of the six sites not included in the 2007 surveys had very little quality prairie (Selby 2012a, pers. comm.). Supplementary surveys conducted further west along U.S. Highway 18 in Hancock County also produced no confirmed sightings (Selby 2010, p. 7). No surveys were conducted at previously known Poweshiek skipperling sites in the State during the 2012 flight season.

The Poweshiek skipperling is presumed extirpated or possibly extirpated from all but four of the known sites in Iowa. The status of the Poweshiek skipperling is unknown at four sites: Highway 60 Railroad Prairie, Floete Prairie in Dickinson County, Florenceville Prairie, and Hayden Prairie in Howard County. There have been no surveys at Highway 60 Railroad Prairie since the species was observed there in 2007 (Selby 2012a, pers. comm.). The last observation of Poweshiek skipperling at Floete Prairie was in 1994 and the habitat “did not appear to be very good quality” in 2007, although the site was not surveyed for butterflies that year (Selby 2012a, pers. comm.) or in subsequent years. The Poweshiek skipperling was last observed at the Florenceville Prairie in 1994 (Saunders 1995, p. 27), but not during the 2007 survey year (Selby 2010, pp. 8-11). The species was last observed at Hayden Prairie in 2005, but not during surveys conducted in 2007 (Selby 2010, p. 10). Four Poweshiek skipperlings were found at Hoffman Prairie in Cerro Gordo County in 2008 (Selby 2009b, p. 3), but none were found during surveys in 2009 (Selby 2009b, p. 7) and 2010 (Selby 2010, p. 7). We initially assigned an unknown status to Hoffman Prairie site because the species had not been seen in the last two survey years; however, Selby believes that the species may be extirpated from this site (Selby 2012a, pers. comm.), so we have assigned a status of extirpated to this site.

To summarize, Poweshiek skipperling was historically documented in 41 sites in Iowa. The species occupancy is unknown at 4 of those sites and the species is considered to be extirpated or possibly extirpated at 13 and 24 sites, respectively (Table 2). The species is not considered to be present at any of the sites in Iowa.

Michigan

Michigan historically contained approximately 6 percent (N=17) of all known records of Poweshiek skipperlings rangewide (Table 2). Poweshiek skipperling has been historically documented at 17 sites in 6 counties in Michigan. The species was first recorded in Michigan in 1893 at Lamberton Lake near Grand Rapids in Kent County (Holzman 1972, p. 111) and then at nearby Button Lake Fen (also known as Emerald Lake Fen) in 1944 (McAlpine 1972, p. 83). Shrubs have invaded both sites, however, and no Poweshiek skipperlings have been found at either of these two western Michigan sites since 1944 and 1968, respectively (Michigan Natural Features Inventory 2011, unpubl. data). Holzman (1972, p. 111) documented Poweshiek skipperling in Oakland County in 1970, and the species has since been found at a total of 15 locations in eastern Michigan.

The Poweshiek skipperling is currently considered to be present at ten sites (Table 2) in four counties in Michigan: Jackson, Lenawee, Oakland, and Washtenaw. The species has been observed very recently (2007-2012) at most of those sites, except at the Liberty Bowl Fen in Jackson County, which has not been surveyed since one individual was observed in 1996. The status of the species is unknown at one site; Bullard Lake in Livingston County, where Poweshiek skipperling were last seen in 2007, but not in subsequent surveys in 2008 and 2009 (Cuthrell 2012a, pers. comm.). The species is presumed extirpated from six sites including the only two sites in Kent County and three sites in Oakland County; Rattalee Road, Fenton Road, and Rattalee Lake Fen (Call C Burr Preserve) fens. The species has not been observed at the Rattalee Road and Fenton Road sites since 1970 and 1973, respectively (Michigan Natural Features Inventory 2011, unpubl. data). Four Poweshiek skipperlings were seen in 2009 at the Rattalee Lake Fen (Calla C Burr Preserve), but none were observed during surveys conducted in 2010, 2011, and 2012 (Cuthrell 2012a, pers. comm.; Michigan Natural Features Inventory 2011, unpubl. data). The Michigan Natural Features Inventory (MNFI) also considers the two sites in Kent County to be extirpated due to habitat loss and destruction, Lamberton Lake and Button Lake (also known as Emerald Lake); the species has not been observed at either site since 1968 and 1944, respectively. The species is presumed to be extirpated at Whalen Lake Fen in Livingston County, where the species has not been observed since 1998 despite three subsequent years of surveys (Michigan Natural Features Inventory 2011, unpubl. data).

Four of Michigan's ten extant (present) Poweshiek skipperling occurrences are considered to have at least good viability (Michigan Natural Features Inventory 2011, unpubl. data). Three of these sites (Buckthorn Lake (also known as Big Valley), Brandt Road Fen (also known as Holly Fen) and Long Lake Fen) are within 20 km (12 mi) of one another in Oakland County; all with relatively large numbers (61-389) of the species recorded in 2010-2012 surveys (Michigan Natural Features Inventory 2011, unpubl. data; Cuthrell 2012a, pers. comm.). The largest extant (present) Poweshiek skipperling population in Michigan is at Long Lake Fen, where 225 individuals (1.3/hr.) were counted during 2012 surveys, down from 389 individuals (2.2/hr.) observed in the previous survey year with similar sampling effort. Long Lake Fen is likely the largest population of Poweshiek skipperling in the United States, and is subjected to intense development pressure. The fourth site, Grand River Fen (also known as Liberty Fen) in Jackson County, is approximately 100 km (62 mi) from the other three sites. In 2010, researchers counted 54 (0.3/hr.) Poweshiek skipperling at Grand River Fen, and 114 (0.6/hr.) in 2011 (Michigan Natural Features Inventory 2011, unpubl. data; Cuthrell 2012a, pers. comm.). This number fell to 14 (0.1/hr.) in 2012 (Cuthrell, 2012a, pers. comm.; 2012b, pers. comm.).

Small populations, immediate threats that have significant impacts on the species, or both limit the viability of the remaining five sites where we consider Poweshiek skipperling to still be present

in Michigan. In 2010, eight (0.1/hr.) Poweshiek skipperling were recorded at Park Lydon in Washtenaw County; 12 individuals were counted in 2011 (0.1/hr.), and 22 were counted in 2012 (0.2/hr.) (Cuthrell 2012a, pers. comm.). Two individuals (0.02/hr.) were recorded at Goose Creek Grasslands (also known as Little Goose Lake Fen) in Lenawee County in 2010, nine (0.07/hr.) were seen in 2011 (Cuthrell 2012a, pers. comm.; Cuthrell 2012b, pers. comm.). Only one Poweshiek skipperling was seen during a 15-minute 3-person survey in 2007 at the Snyder Lake site. Fourteen individuals were observed during 2008 surveys at Halstead Lake Fen (Michigan Natural Features Inventory 2011, unpubl. data), and 18 were observed in 2012 (Cuthrell 2012a, pers. comm.); neither survey year had units of effort associated with the counts at this site. One individual was counted at Bullard Lake fen in 2007, but the species was not observed in the two most recent survey years (2008 and 2009); therefore, the status is unknown at that site. We have only one year of data from Liberty Bowl Fen, where the species was recorded in 1996. The Eaton Road Fen is thought to be fairly viable, where 15-20 individuals were observed on multiple occasions in 2005 and a high of 68 individuals were observed in 2011 (Cuthrell 2013b, pers. comm.). The Eaton Road site is approximately 1 mi (0.6 km) from the Long Lake Fen site and is considered a sub-site within Long Lake Fen (Cuthrell 2013b, pers. comm.), but we consider it to be a separate site for the purposes of this rule.

To summarize, Poweshiek skipperling was historically documented in 17 sites in Michigan (Table 2). The species is considered to be present at 10 of the sites. The occupancy is unknown at 1 site, and the species is considered to be extirpated at 6 sites.

Minnesota

Minnesota historically contained approximately 48 percent (N=142) of all known records of Poweshiek skipperlings rangewide (Table 2). There are approximately 189 historical Poweshiek skipperling occurrence records in 32 counties in Minnesota [Minnesota Natural Heritage Inventory (MN NHI) database accessed June 19, 2013, plus additional surveys]. Clusters of records occur within five general areas from the State's southwest corner to near the Canadian border in the north. Based on the proximity of some occurrences to one another (

e.g.,

overlapping or occurrences in close proximity to one another in one general location), there appear to be approximately 142 distinct historical site records in the State (Dana 2012d, pers. comm; Service 2013, unpubl. geodatabase). Poweshiek skipperling are presumed extirpated or possibly extirpated from at least 75 of these known sites. The status of the species is unknown at 67 sites, although 31 of those locations have not been surveyed since 2003, and the species has undergone a sharp decline in the State since then.

Until about 2003, the Poweshiek skipperling was regarded as “the most frequently and reliably encountered prairie-obligate skipper in Minnesota” (Dana 2008, p. 1). Signs of the species' decline in Minnesota were noted in 2003 when Selby (2005, p. 20) found sharply lower numbers in and near Glacial Lakes State Park (Selby 2005, p. 20) compared to those observed in 2001 (Skadsen 2001, pp. 22-24). For example, numbers recorded along four transects that were surveyed in both years decreased from 104 to 2 individuals (Selby 2006b, Appendix 2, p. ii). In 2004 and 2005, Selby (2006b, Appendix 2, p. 2) did not record a single Poweshiek skipperling on any of these transects in and around the park during 11 separate surveys.

An extensive survey effort was conducted in 2007 and 2008 throughout most of the species' known range in the State (Selby 2009a, entire). Sites with previous Poweshiek skipperling records that were considered to have the greatest conservation importance to the species (large, high-quality prairie remnants) were surveyed, as well as sites with no previous records that appeared likely to support the species (Selby 2009a, p. 2). In 2007, 70 sites in 15 counties were surveyed, including 26 sites with previous Poweshiek skipperling records (Selby 2009a, pp. 1, 6). In 2008, 58 sites were surveyed in 13 counties, including 22 sites with prior records (Selby 2009a, pp. 1, 6). A total of 34 sites with previous Poweshiek skipperling records were surveyed in both years combined. Poweshiek skipperling presence was recorded on only three of the 70 surveyed sites in 2007; each of these three sites had just one confirmed individual (Selby 2009a, p. 1). The 2008 surveys documented no Poweshiek skipperling records on any of the 58 sites surveyed (Selby 2009a, p. 1).

An extensive survey effort was also completed in 1993 and 1994 (Schlicht and Saunders 1994, entire; Schlicht and Saunders 1995, entire). During those surveys, Poweshiek skipperlings were found in 11 of 19 sites on which it had been previously recorded and in 13 new sites, for a total of 25 of 63 surveyed prairie sites; the species was present at 30 and 39 percent of the sites in 1993 and 1994, respectively (Schlicht and Saunders 1995, pp. 5-7). These results contrast sharply with those from the surveys conducted in 2007 and 2008, when the species was found at four and zero percent of sites, respectively. Although the species was apparently more common in 1993 and 1994, numbers of Poweshiek skipperling found during surveys were typically low. Large numbers were observed at only three sites (Schlicht and Saunders 1995, p. 4). At one of these sites, Glynn Prairie, 25 Poweshiek skipperling were recorded during a 50-minute survey in July 1993 (Schlicht and Saunders 1995, data sheet); no Poweshiek skipperling were observed at this site during the 2007 survey despite good survey conditions (Selby 2009a, p. xxxv).

In 2007, multiple transect surveys were conducted in four sites with previously well-documented Poweshiek skipperling populations—transects totaling 52,985 m (33 mi) were surveyed without observing a single Poweshiek skipperling (Dana 2008, p. 5). About half of these transects (totaling 20,959 m (13 mi)) were in Prairie Coteau Scientific and Natural Area (SNA), where in 1990 Selby recorded 116 Poweshiek skipperlings during the flight peak (Selby and Glenn-Lewin 1990, pp. 19-20) along a total of about 6,250 m (4 mi) of transects (Dana 2008, p. 16). No Poweshiek skipperling were observed during surveys of the Prairie Coteau SNA in 2012 (Runquist 2012, pp. 9-10).

Additional surveys were conducted in 2012, however, Poweshiek skipperling were not observed at any of the 18 sites with relatively recent records (Runquist 2012, pp. 4-25; Selby 2012, p. 2; Selby 2013, p. 2; Dana 2012c, pers. comm.; Runquist 2012a, pers. comm.; Olsen 2012a, pers. comm.). Fifteen additional prairie sites with potential habitat or records of other skippers were surveyed in 2012, but no Poweshiek skipperling were observed (Runquist 2012, pp. 4-25; Selby 2012, p. 2; Selby 2013, p. 2; Dana 2012c, pers. comm.; Runquist 2012a, pers. comm.; Olsen 2012a, pers. comm.).

Nearly half (approximately 48 percent) of all documented Poweshiek skipperling sites rangewide are in Minnesota, thus the apparent collapse of large numbers of Poweshiek skipperling populations across the State may pose a significant challenge for the long-term existence of this species. Although the possibility remains that the species is extant at some sites where recent (2007, 2008, or 2012) surveys were negative, it seems unlikely that it is present at those sites in any significant numbers. Extensive surveys in 1993 and 1994 documented the species at about 35

percent of all surveyed sites, whereas the 2007 effort found them at only about 2 percent of all sites surveyed; no Poweshiek skipperling were detected despite widespread and robust survey efforts involving multiple observers in 2008 or 2012 (Dana 2008, p. 8; Selby 2009a, p. 1; Dana 2012c, pers. comm.; Runquist 2012a, pers. comm.; Olsen 2012, pers. comm.; Runquist 2012, pp. 4-25; Selby 2012, p. 2, 2013, p. 2).

To summarize, Poweshiek skipperling was historically documented in approximately 142 sites in Minnesota (Table 2). The species is not considered to be present at any of these sites (Table 2). The occupancy is unknown at 67 sites, and the species is considered to be extirpated or possibly extirpated at 7 and 68 sites, respectively (Table 2).

North Dakota

North Dakota historically contained approximately 6 percent (N=17) of all known records of Poweshiek skipperlings rangewide (Table 2). Poweshiek skipperlings have been historically documented at 17 sites (Table 2) in 7 North Dakota counties (Selby 2010, p. 18; Service 2013, unpubl. geodatabase): Cass, Dickey, LaMoure, Ransom, Richland, and Sargent in the southeastern corner of the State and Grand Forks County in the Northeast. Poweshiek skipperling are now considered extirpated or possibly extirpated from seven sites and four counties (Cass, Dickey, LaMoure, and Grand Forks) in North Dakota. The status of the species is unknown at 10 sites, where the species was last observed between 1996 and 2001, but not during the most recent 1-2 year(s) surveyed. The status of the species is also unknown at one site where the species was observed in 1996 with no recent surveys for the species, but the habitat was recently rated as poor (Service 2013, unpubl. geodatabase). Four sites with fairly recent Poweshiek skipperling records were surveyed in 2012; Poweshiek skipperling were not found at any of those sites (Royer and Royer 2012b, pp. 21-24; Royer and Royer 2012a, p. 6). One additional site was surveyed, which had the potential for Poweshiek skipperling presence because of its proximity to a known site for the species; however, no Poweshiek skipperling were found (Royer and Royer 2012b, pp. 18-19; Royer and Royer 2012a, p. 6; Royer 2012b, pers. comm.).

The Poweshiek skipperling was known from seven North Dakota sites across six counties in the 1990s; however, only two of those sites were considered to have extant populations at that time; three records were represented by incomplete or ambiguous locality data and the species was assumed to be extirpated at one site (Royer and Marrone 1992b, pp. 8-11). Surveys conducted in the State after 1992 documented additional populations, but the most recent surveys at these sites were mostly negative. Orwig discovered eight new populations of Poweshiek skipperling (six in Richland County and two in Sargent County) during three years of survey work (1995-1997) in southeast North Dakota (Orwig 1995, pp. 3-4; Orwig 1996, pp. 4-6, 9-12; Orwig 1997, p. 2). The species was found at two of the eight sites surveyed in 1997 (Orwig 1997, p. 2) and at two additional sites in 1996 (Spomer 2004, p. 11).

Once abundant at several known sites in North Dakota, Poweshiek skipperlings have experienced a dramatic decline over the last few decades. In 1977, McCabe and Post (1977a, p. 38), for example, found Poweshiek skipperling to be abundant at McLeod Prairie in Ransom County, stating that they could “be collected two at a time on the blossoms of Long-headed coneflower…” In six years of subsequent monitoring (1986-1991), however, Royer failed to find a single Poweshiek skipperling at the site after it was converted to a cattle-loading area (Royer and Marrone 1992b, p. 10). Royer and Marrone (1992b, pp. 10-11) assumed the species had been extirpated at this site. Similarly, the number of Poweshiek skipperlings recorded during surveys at the West Prairie Church site along the boundary of Cass and Richland counties, fell from hundreds in 1986, to four in 1990 and zero in 1991 and 2012 (Royer and Marrone 1992b, p. 8; Royer and Royer 2012b, p. 21). Poweshiek skipperlings are unlikely to persist at this small and isolated site (Royer and Royer 2012b, p. 21; Royer 2012c, pers. comm.).

The last observation of a live Poweshiek skipperling in North Dakota was in 2001, at a new site discovered by Spomer (2001, p. 9) in Ransom County. Poweshiek skipperlings were not found in subsequent surveys at this site in 2002, 2003, and 2012 (Spomer 2001, p. 2; Spomer 2002, p. 3; Spomer 2004 p. 36; Selby 2010, p. 18; Royer and Royer 2012b, p. 22), although the 2012 survey may have been conducted too late in the year to detect the species at that site (Royer 2012b, pers. comm; Royer 2012d, pers. comm.). Therefore, the status of the species at this site is unknown.

To summarize, Poweshiek skipperling was historically documented in 17 sites in North Dakota (Table 2). The species is not considered to be present at any of these sites (Table 2). The occupancy is unknown at 10 sites, and the species is considered to be extirpated or possibly extirpated at 1 and 6 sites, respectively (Table 2).

South Dakota

South Dakota historically contained approximately 24 percent (N=70) of all known records of Poweshiek skipperlings rangewide (Table 2). The Poweshiek skipperling has been historically documented at approximately 70 sites (Table 2) across 10 counties in South Dakota (Selby 2010, p. 19). Based on expert review and additional survey and habitat information, the status of the species was determined to be unknown at 48 sites and presumed extirpated at the remaining 22 sites (Table 2); at least 8 of the extirpated sites have been destroyed by conversion, gravel mining, loss of native vegetation, flooding, or heavy grazing (Skadsen 2012c, pers. comm.).

The Poweshiek skipperling was not detected at any site that was surveyed between 2009 and 2012: 6 sites in 2009, 10 sites in 2010, 1 sites in 2011, and 10 sites in 2012 (Skadsen 2009, p. 12; Skadsen 2011, p. 5; Skadsen 2010, pers. comm.; Skadsen 2012a, pers. comm.; Skadsen 2012, p. 3). The 2009 to 2012 results are in marked contrast to surveys conducted in 2002 when the species was recorded at 23 of 24 sites surveyed (Skadsen 2003, pp. 11-45). Cool and wet weather may have depressed butterfly populations, in general, in eastern South Dakota and west-central Minnesota in 2009 as it apparently did in 2004 (Skadsen 2004, p. 2; Skadsen 2009, p. 2).

Wisconsin

Wisconsin historically contained approximately 1 percent (N=4) of all known records of Poweshiek skipperlings rangewide (Table 2). Naturalists reported Poweshiek skipperling to be common to abundant on prairies in southeastern Wisconsin in the late 1800s (

e.g.,

in Milwaukee and Racine Counties), although exact localities are unknown (Borkin 2011, in litt.; Selby 2010, p. 22). By 1989, however, the species was listed as State endangered (Borkin 2011, in litt.). The Poweshiek skipperling is considered to be present at three sites in Wisconsin (Table 2); two sites are within the Southern Unit of the Kettle Moraine State Forest in Waukesha County. The third site, Puchyan Prairie State Natural Area (SNA), is approximately 100 km (62 mi) to the northwest of the Kettle Moraine State Forest in Green Lake County. The status of the species is unknown at another site within the

Kettle Moraine State Forest. An additional 2010 record of a butterfly was incorrectly identified as a Poweshiek skipperling at Melendy's Prairie Unit of the Scuppernong Prairie SNA (Borkin 2012b, pers. comm.).

The two occurrences of Poweshiek skipperling in the Kettle Moraine State Forest inhabit small areas that were once part of a larger prairie complex, which was fragmented by conversion to agriculture, other human development, and encroachment of woody vegetation (Borkin 2011, in litt.). The larger of the two populations at Kettle Moraine State Forest inhabits a 6-ha (15-ac) prairie remnant on Scuppernong Prairie SNA, which had record counts exceeding 100 individuals in 1994, 1995, 1998, and 1999 (Borkin 1995a, p. 10; Borkin 1996, p. 7; Borkin 2000b, p. 4; Borkin 2011, in litt.). Four were found in 2007 (Borkin 2008, in litt., p. 1), although these data were collected during a single transect survey that may have been early in the flight season and are, therefore, not comparable to other survey years (Borkin 2012a, pers. comm.). A maximum count of 42, 17, 63, and 45 were counted in 2009, 2010, 2011, and 2012, respectively (Borkin 2011a, pers. comm.; Borkin 2012c, pers. comm.). There was some concern that a controlled burn in late March of 2012 may correlate with lower numbers observed during the 2012 flight (Borkin 2012a, pers. comm.); however, this difference is within the range of variation observed over the previous four years (Wisconsin DNR 2012, in litt.).

After brush was cleared from the area in 2002, a small number of Poweshiek skipperlings were discovered the following year in a small isolated prairie remnant patch at a second site in the Kettle Moraine State Forest, (Borkin in litt. 2008). Once the intervening woody growth was removed, individuals presumably dispersed from the Scuppernong SNA remnant prairie to a small habitat patch about 200 ft (61 m) away (Borkin 2012a, pers. comm.). Surveys at each habitat patch have consistently yielded counts of less than 10 (Borkin 2008, in litt.), with a combined high count of 11 to 15 individuals in 2011. A total of six individuals, with a high single day count of three, were observed in eight surveys during 2012 (Borkin 2012c, pers. comm.; Borkin 2012a, pers. comm.).

The status of the Poweshiek skipperling is unknown at a third and much larger fragment of Kettle Moraine State Forest, the Kettle Moraine Low Prairie SNA, which is adjacent to the Wilton Road site. The Kettle Moraine Low Prairie SNA was overgrown by shrubs including willows (

Salix spp.

), quaking aspen (

Populus tremuloides

), and glossy buckthorn

(Frangula alnus

) and has been managed with a series of controlled burns, in addition to a 1975 wild fire (Borkin 2011, in litt; Borkin 2012a, pers. comm.; Wisconsin DNR 2012, in litt). The highest number recorded at the Kettle Moraine Low Prairie SNA was 28 on July 8, 1995 (Borkin 2012a, pers. comm.). Preliminary attempts in 2000 to 2003 to augment the population with adults from Scuppernong SNA and captive-reared larvae were not successful (Borkin 2012a, pers. comm.). A single Poweshiek skipperling was sighted there on July 2, 2004, but none were found in surveys conducted in 2007-2009 and 2011-2012 (Borkin 2011b, pers. comm.; Borkin 2012a and 2012c, pers. comm.). Two Poweshiek skipperlings were recorded in 2010 at this site (Wisconsin DNR2012, in litt.); however, there were no photographs or voucher specimens to confirm the sighting. This site was surveyed less intensively than Scuppernong Prairie, because of the species' relatively low density and abundance at Kettle Moraine Low Prairie SNA (Borkin 2012a, pers. comm.). Extensive brush cutting, additional burns, and restoration of the hydrology have been undertaken in recent years (Borkin 2012a, pers. comm.).

Poweshiek skipperlings are present at a third site in Wisconsin, Puchyan Prairie SNA, in Green Lake County, although this population is small and declining (Borkin 2009, pers. comm.). The Poweshiek skipperling was first discovered at Puchyan Prairie in 1995, and 6 to 30 individuals have been recorded in subsequent surveys (Borkin 2008, in litt.; Swengel 2012, pers. comm). In 2012, Swengel (2012, pers. comm.) found a maximum of three individuals, despite several hours of searching over three days.

Additional sites in eight counties (Crawford, Grant, Iowa, Jefferson, Monroe, Rock, Sauk, and Walworth) have been surveyed in an attempt to find undiscovered Poweshiek skipperling populations. Four of the eight sites surveyed in 1998 and 1999 seemed to have adequate host plants, nectar resources, and size typical of Poweshiek skipperling habitat, but Poweshiek skipperling were not present at any of the sites (Borkin 2000b, pp. 5-7).

To summarize, Poweshiek skipperling was historically documented in 4 sites in Wisconsin (Table 2). The species is considered to be present at three sites and the occupancy is unknown one site (Table 2).

Manitoba

Manitoba historically contained less than 1 percent (N=1) of all known records of Poweshiek skipperlings rangewide (Table 2); however, multiple Poweshiek skipperling historical records occur in one general location—a complex of several nearby small sites within the Tallgrass Prairie Preserve—in far southern Manitoba, near the United States border. Poweshiek skipperlings were first recorded in Canada near Vita, Manitoba, in 1985 at each of seven prairies surveyed, and populations were described as abundant but localized (Catling and Lafontaine 1986, p. 63). Poweshiek skipperlings were found at 15 of 18 locations surveyed within the same area in 2002 (COSEWIC 2003, p. 5).

The Poweshiek skipperling is currently present at one location in Canada, The Nature Conservancy's Tall Grass Prairie Preserve near Vita, Manitoba (Westwood 2010, p. 2; Westwood

et al.

2012, p. 1; Hamel

et al.

2013, p. 1). Poweshiek skipperlings were historically moderately common in areas of the preserve (Klassen

et al.

1989, p. 27). In 2002, Webster (2003, p. 5) counted approximately 150 individuals, and in 2006, approximately 126 individuals were sighted across 10 sites (Westwood 2010, p. 3). Surveys of 10 sites in 2008 and 2009 yielded 281 and 79 Poweshiek skipperlings, respectively (Dupont 2010, pers. comm.). Poweshiek skipperling numbers in the preserve declined sharply after a 647-ha (1,600-ac) wildfire in fall 2009 burned much of the species' habitat, including areas that likely contained the largest and highest density populations (Westwood 2010, p. 2); surveys of comparable effort to the 2008 and 2009 surveys yielded only 13 Poweshiek skipperlings on the preserve in 2010 (Westwood 2010, pp. 7-22). Surveys of 45 sites within the Tall Grass Prairie Reserve during 2011 resulted in 13 sites with positive sightings, 9 of which were new sites (Westwood

et al.

2012, p. 11; Dupont 2011, pers. comm.). The average number of Poweshiek skipperlings found at each site ranged from 10 to 15 per hour. These numbers are up considerably from 2010, but not as high as observed in 2008 (Dupont 2011, pers. comm.). In 2012, a total of 50 individuals were observed, which was “low when compared to historic densities” (Hamel

et al.

2013, p. 17). The preserve has detailed management recommendations to facilitate recovery of the Poweshiek skipperling (Westwood 2010, p. 5).

Following an assessment and status report completed in 2003 under the Committee on the Status of Endangered Wildlife in Canada (COSEWIC), the Poweshiek skipperling was listed under the Species at Risk Act as Threatened in Canada in July 2005 (COSEWIC 2003). A recovery strategy is now in place for the species in Canada (Environment Canada 2012), which includes critical habitat designations within and adjacent to the Tall Grass Prairie Preserve (Environment Canada 2012, p. ii).

Summary of Factors Affecting the Species

Section 4 of the Act (16 U.S.C. 1533), and its implementing regulations at 50 CFR Part 424, set forth the procedures for adding species to the Federal Lists of Endangered and Threatened Wildlife and Plants. Under section 4(a)(1) of the Act, we may list a species based on any of the following five factors: (A) The present or threatened destruction, modification, or curtailment of its habitat or range; (B) overutilization for commercial, recreational, scientific, or educational purposes; (C) disease or predation; (D) the inadequacy of existing regulatory mechanisms; and (E) other natural or manmade factors affecting its continued existence. Listing actions may be warranted based on any of the above threat factors, singly or in combination. Each of these factors is discussed below.

We evaluated the level of impact to the population at each site of stressors at 170 Dakota skipper sites where the occupancy status of the site is considered to be present or unknown, as defined in the Background section of this rule. These 170 sites are found across the current range of the species in Minnesota, North Dakota, and South Dakota. Two Dakota skipper sites with an unknown or present occupancy were not evaluated. To determine the levels of impact to the population at each site, we used the best available and most recent information for each site, including reports, discussions with site managers, information from natural heritage databases, etc. (Service 2012, unpubl. data; Service 2013, unpubl. geodatabase). We only evaluated a stressor to the population at any one site if we had sufficient information to determine if the level of impact was high, medium, or low as defined for each stressor below; therefore, the number of sites evaluated varies with each stressor.

We evaluated the level of impact to the species from stressors at 68 Poweshiek skipperling sites where the occupancy status of the site is considered to be present or unknown, as defined in the Background section of this proposed rule. Although we did not evaluate every stressor at all 145 Poweshiek skipperling sites with present or unknown occupancy, the 68 sites that were evaluated are representative of all those sites in terms of geography (sites in Iowa, Michigan, Minnesota, North Dakota, South Dakota, and Wisconsin were evaluated), ownership, and management. To determine the levels of impact to the population at each site, we used the best available and most recent information, including reports, discussions with site managers, and information from natural heritage databases (Service 2012, unpubl. data; Service 2013, unpubl. geodatabase). We only evaluated a particular stressor at any one site if we had sufficient information to determine if the level of impact was high, medium, or low (as defined below); therefore, the number of sites evaluated varies with each stressor.

Factor A. The Present or Threatened Destruction, Modification, or Curtailment of Its Habitat or Range

Habitat quality is a powerful determinant of extinction probability in butterflies such as the Dakota skipper and Poweshiek skipperling (Thomas

et al.

2001, p. 1795). Among butterfly species in the United Kingdom, for example, equilibrium density of butterflies at sites with optimum habitat are from 25 to more than 200 times greater than those for occupied sites with suboptimal, yet suitable, habitat (Thomas 1984, cited in Thomas

et al.

2001, p. 1794). Consistently good habitat quality is especially important for Dakota skipper and Poweshiek skipperling isolated populations, which would not be naturally recolonized if they were extirpated. Protection or restoration of habitat quality at these isolated sites is critical to the survival of both species, although stochastic events still pose some risk, especially for smaller populations and at small sites.

The Poweshiek skipperling and Dakota skipper depend on a diversity of native plants endemic to tallgrass prairies and, for the Poweshiek skipperling in Michigan, prairie fens. When nonnative or woody plant species become dominant, Poweshiek skipperlings and Dakota skippers decline due to insufficient sources of larval food and nectar for adults. For example, at Wike Waterfowl Production Area in Roberts County, South Dakota, the extirpation of Poweshiek skipperling is attributed to the deterioration of native vegetation, in particular, the loss of nectar sources for adult butterflies due to invasive species encroachment (Skadsen 2009, p. 9).

Destruction of native tallgrass and mixed-grass prairie began in 1830 (Samson and Knopf 1994, pp. 418-419). Extant populations of Dakota skipper and Poweshiek skipperling are restricted to native prairie remnants and prairie fens; native prairies have been reduced by 85 to 99.9 percent of their former area throughout the historical range of both species (Samson and Knopf 1994, pp. 418-419). Degradation and destruction of habitat occurs in many ways, including but not limited to: conversion of native prairie to cropland or development; ecological succession to woody vegetation; encroachment of invasive species; past and present fire, haying, or grazing management that degraded or destroyed the species' habitats; flooding; and, groundwater depletion, alteration, and contamination, which are discussed in further detail below.

We evaluated the level of impact to the population at each site of several habitat-related stressors at 170 Dakota skipper sites where the occupancy status of the site is considered to be present or unknown, as defined in the Background section of this proposed rule (Table 3). These 170 sites are found across the current range of the species in Minnesota, North Dakota, and South Dakota. Two sites with an unknown or present occupancy were not evaluated. To determine the levels of impact to the population at each site, we used the best available and most recent information for each site, including reports, discussions with site managers, information from natural heritage databases, etc. (Service 2012, unpubl. data; Service 2013, unpubl. geodatabase). We only evaluated a stressor to the population at any one site if we had sufficient information to determine if the level of impact was high, medium, or low as defined for each stressor below. Similarly, the level of impact to the population was evaluated at 68 Poweshiek skipperling sites with present or unknown status (Table 4). Although we did not evaluate Factor A stressors at all 145 Poweshiek skipperling sites with present or unknown occupancy, the 68 sites that were evaluated are representative of all the present or unknown Poweshiek skipperling sites in terms of geography (range of the species,

i.e.,

sites in Iowa, Michigan, Minnesota, North Dakota, South Dakota, and Wisconsin were evaluated), ownership, and management. Many sites for both species (59 sites for Dakota skipper and 32 sites for Poweshiek skipperling) experience at least two habitat-related

stressors at a medium or high level of impact (Tables 3 and 4).

Table 3—Number of Dakota Skipper Sites With Each Level of Impact and the Total Number of Sites That Were Rated for Each Type of Stressor—A Total of 170 Dakota Skipper Sites With Either Present or Unknown Status Were Examined; Only Sites With Sufficient Data for a Particular Stressor Were Rated as High, Medium, or Low (Service 2012 Unpubl. Data; Service 2013, Unpubl. Data)

Stressor

High level of impact

Medium level of

impact

Low level of impact

Total

number of

rated sites

Destruction & Conversion (Agricultural & Nonagricultural Development)

3

87

60

150

Wind Development

1

0

8

9

Flooding

0

6

6

12

Invasive Species

13

31

18

62

Fire

9

4

6

19

Grazing

10

29

14

53

Haying & Mowing

2

11

27

40

Lack of Management

10

5

3

18

Size/Isolation

50

35

58

143

Herbicide and/or Pesticide Use

5

2

9

16

Table 4—Number of Poweshiek Skipperling Sites With Each Level of Impact and the Total Number of Sites That Were Rated for Each Type of Stressor—A Total of 68 Poweshiek Skipperling Sites With Either Present or Unknown Status Were Examined; Only Sites With Sufficient Data for a Particular Stressor Were Rated as High, Medium, or Low (Service 2012 Unpubl. Data; Service 2013, Unpubl. Data)

Stressor

High level of impact

Medium level of

impact

Low level of impact

Total

number of

rated sites

Destruction & Conversion (Agricultural & Nonagricultural Development)

1

13

40

54

Wind Development

0

0

6

6

Flooding/Hydrology

2

4

14

20

Invasive Species

9

30

12

51

Fire

7

3

14

24

Grazing

7

14

2

23

Haying & Mowing

0

3

7

10

Lack of Management

5

6

2

13

Size/Isolation

25

24

19

68

Herbicide and/or Pesticide Use

3

1

6

10

Destruction and Conversion of Prairies

Destruction and Conversion of Prairies to Agricultural Land

Conversion of prairie for agriculture may have been the most influential factor in the decline of the Poweshiek skipperling and Dakota skipper since Euro-American settlement, but the threat of such conversion to extant populations is not well known and may now be secondary to other threats. By 1994, tallgrass prairie had declined by 99.9 percent in Illinois, Iowa, Indiana, North Dakota, Wisconsin, and Manitoba; and by 99.6 percent in Minnesota; and 85 percent in South Dakota (Samson and Knof 1994, p. 419). Samson and Knof (1994, p. 419) did not provide a figure for the decline of tallgrass prairie in Saskatchewan, but mention an 81.3 percent decline in mixed grasses from historical levels. By 1994, mixed-grass prairie had declined from historical levels by 99.9 percent in Manitoba and 71.9 percent in North Dakota (Samson and Knof 1994, p. 419). Destruction of tallgrass and mixed-grass prairie began in 1830, but significant documentation of the ecosystem's butterfly fauna did not begin until about 1960. Therefore, most of the decline of the Dakota skipper and Poweshiek skipperling probably went unrecorded.

Since about 1980, observers have documented the extinction of several populations of the Dakota skipper and Poweshiek skipperling due to habitat conversion to agricultural use in the United States and Canada. For example, four Dakota skipper sites in North Dakota were converted to irrigated potato fields, and one in South Dakota was converted for crop production (Royer and Marrone 1992a, p. 17). The Fannystelle site in Manitoba, where the Dakota skipper was last recorded in 1991, was subsequently converted for row-crop agriculture (Webster 2003, p. 7). In North Dakota, further conversion is a threat to Dakota skippers in the important Towner-Karlsruhe complex (Royer and Royer 1998, p. 22; Lenz 1999b, p. 13), where the flat topography and high water table facilitate conversion to irrigated crop production. Populations of Dakota skipper in Manitoba typically occupy flat terrain that may be vulnerable to conversion to cropland, although soil conditions may be unsuitable for row crops at some of these sites (Webster 2003, p. 10). Similarly, conversion of native prairie to cropland continues to be a threat to Poweshiek skipperling habitat throughout its range (Royer and Marrone 1992b, p. 17).

The Dakota skipper, and until recently, the Poweshiek skipperling, have largely persisted in areas that are relatively unsuitable for row crop agriculture because of their steep terrain (

e.g.,

in the Prairie Coteau of South Dakota) or where soils are too wet or rocky for row-crop agriculture (McCabe 1981, pp. 189-190, Webster 2003, p. 10). Densely spaced, large glacial rocks, for example, may have deterred cultivation

at the Chippewa Prairie in Minnesota and “spared Chippewa Prairie in Minnesota from the plow” (Dana 2012, pers. comm.). In areas where Poweshiek skipperling and Dakota skipper habitat persists but is adjacent to agriculture, added nutrients from agricultural runoff affects groundwater and additional nutrients in the system contribute to the dominance of invasive plants (Fiedler and Landis 2012, p. 51: Michigan Natural Features Inventory 2012, p. 4).

In summary, conversion for agriculture on lands suitable for such purposes is a current, ongoing stressor of high level of impact to the Poweshiek skipperling and Dakota skipper populations in areas where such lands still remain. Advances in technology may also increase the potential of conversions in areas that are currently unsuitable for agriculture.

We rated the level of impact to the populations of the stressor posed by habitat destruction or conversion for both agriculture and nonagricultural purposes (except for conversion for wind energy development, which was analyzed separately) at 150 Dakota skipper and 54 Poweshiek skipperling sites with present or unknown status (see Tables 3 and 4) where we had sufficient information to evaluate the stressor. In our evaluation of this stressor, we combined agricultural and nonagricultural impacts—our analyses are discussed below (see

Destruction and Conversion of Prairies due to Nonagricultural Development

).

Destruction and Conversion of Prairies to Nonagricultural Development

Conversion of prairie for nonagricultural land uses, such as energy development, gravel mining, transportation, and housing are stressors to both Poweshiek skipperling and Dakota skipper populations. For example, a site where the Dakota skipper and Poweshiek skipperling were recorded in 1997 (Skadsen 1997, pp. 15-16, B-1) in the Bitter Lake area of Day County, South Dakota, is now a gravel pit, and the species' habitat no longer exists there (Skadsen 2003, pp. 47-48).

Almost all prairie remnants with Poweshiek skipperling and Dakota skipper populations are associated with gravelly glacial till soils (Service 2013, unpubl. geodatabase); therefore, gravel mining is a potential stressor to populations at a large number of sites. Gravel mining is a stressor to Poweshiek skipperling and Dakota skipper populations at several sites in Minnesota (Dana 1997, p. 15). For example, gravel mining is a threat in at least three of the five sites that comprise the Felton Prairie complex (Cochrane and Delphey 2002, pp. 16-17); however, the Clay County Stewardship Plan (Felton Prairie Stewardship Committee 2002) may have reduced the likelihood of the gravel mining stressor to populations at this complex. On at least seven sites in Minnesota, Dakota skippers inhabit northern dry prairie plant communities, which are generally impacted by gravel mining due to the predominance of gravel soils (Minnesota DNR 2006, p. 221). Gravel mines are considered a stressor with a high level of impact to populations of both species because, where it occurs, the habitat is completely destroyed.

Energy development (oil, gas, and wind) and associated roads and facilities result in the loss or fragmentation of suitable prairie habitat (Reuber 2011, pers. comm.). Much of the Dakota skipper's range and some of the Poweshiek skipperling's range overlaps with major areas of oil and gas development, which have been increasing rapidly in parts of both species' ranges. North Dakota, for example, is now one of the top two oil-producing states in the United States, and new development is occurring rapidly (MacPherson 2012, p. 1; North Dakota Petroleum Council 2012, p. 1). The number of drilling permits in North Dakota nearly doubled between 2007 and 2008, from 494 permits issued in 2007 to 946 in 2008 (North Dakota Petroleum Council 2009, p. 2). Permits dropped to 627 in 2009 (North Dakota Petroleum Council 2010, p. 2), but increased dramatically to 1,676 in 2010 (Ogden 2011, p. 1). While much of the oil activity is currently occurring in areas of native prairie overlaying the Bakken and Three Forks formations to the west of known locations for both species, mineral exploration has occurred in all but one county in North Dakota (North Dakota Petroleum Council 2012, p. 1). McKenzie County falls in the center of this development and McHenry County is also within these formations (Mueller 2013, pers. comm.). The oil development on the Bakken formation in North Dakota, for example, is a future stressor to Dakota skipper populations in McKenzie County (Royer and Royer 2012b, p. 16). Oil company officials anticipate that production will continue to expand at record levels (MacPherson 2012, p. 1; MacPherson 2010, entire).

Native prairie habitat would be destroyed in the footprint of an oil and gas well pad, but the pads are relatively small. However, each oil and gas well pad requires new road construction, and evidence suggests that Poweshiek skipperlings may avoid crossing roads (Westwood

et al.

2012, p. 18). Oil and gas development can double the density of roads on range lands (Naugle

et al.

2009, pp. 11, 46), increase pipelines, and increase the number of gravel pits to accommodate the increased road construction (Mueller 2013, pers. comm.). In areas with ranching, tillage agriculture, and oil and gas development, 70 percent of the developed land was within 100 m (109 yards (yd)), and 85 percent of the developed land was within 200 m (218 yd), of a human structure (Naugle

et al.

2009, p. 11). Researchers estimated that in those areas, every square km (0.39 square miles) of land may be both bounded by a road and bisected by a power line (Naugle

et al.

2009, p. 11). The habitat fragmentation associated with oil and gas development may amplify other threats to both species, such as the effects of population isolation and the impacts of stochastic events.

Energy development has additional undesirable and potentially significant cumulative impacts on wildlife. Catastrophic events, such as oil and brine spills, could cause direct mortality of Dakota skipper or Poweshiek skipperling larvae that are in shelters at or below the soil surface. Such spills may also cause the loss of larval host and nectar plants in the spill path. Additional plants may be lost during spill response, particularly if the response involves burning. No such spills are known to have occurred in the region, however, and the likelihood of spills occurring on the small fraction of land that remains native tallgrass prairie in North Dakota (less than one percent according to Samsom and Knoff 1994, p. 419) is low.

Wind energy turbines and associated infrastructure (

e.g.,

maintenance roads) are likely stressors to Dakota skipper and Poweshiek skipperling populations, particularly on private land in South Dakota (Skadsen 2002, p. 39; Skadsen 2003, p. 47; Skadsen 2012d, pers. comm.). Similar to oil and gas development, wind development would destroy native prairie habitat in the footprint of the structure, add access roads and other infrastructure that may further fragment prairies, and could be catalysts for the spread of invasive species. Further, it is unknown if the noise and flicker effects associated with wind turbines may impact Dakota skipper or Poweshiek skipperling populations beyond direct impacts from the turbines and/or infrastructure. Other wildlife species, such as birds, have shown significant avoidance of grasslands where wind development has occurred (Pruett

et al.

2009, p. 1256;

Shaffer

et al.

2012, p.). Wind development was assessed at nine Dakota skipper sites and six Poweshiek skipperling sites where we had sufficient information. The level of threat was considered to be low at most sites because although the site may be in an area with the potential for wind development, there are no specific plans or proposals to develop wind power on the site. Wind development is considered a stressor of high level of impact to populations at sites where development is proposed and there are no actions or plans to mitigate impacts to the species. For example, a wind facility was recently proposed at a Dakota skipper site in South Dakota (Skadsen 2012d, pers. comm.), which poses a high-level threat for the species at that site because there are no plans to mitigate impacts of habitat destruction. Although wind power development currently poses a high level of impact to the population at only one site, the extent of this threat will likely increase in the future, due to the high demand for wind energy and the number of Dakota skipper and Poweshiek skipperling sites that are conducive to wind development (

e.g.,

Skadsen 2003, pp. 47-48). Furthermore, power distribution lines may be developed in order to accommodate the added power of wind farms, for instance, a new power line is currently being planned in the Prairie Coteau in South Dakota for that purpose (Mueller 2013, pers. comm.).

Housing construction has likely contributed to the loss of at least two Poweshiek skipperling populations in Michigan, and the largest extant population in Michigan is located in an area under intense development pressure (Michigan Natural Features Inventory 2011, unpubl. data). Residential wells and drainage disrupt prairie fen hydrology by reducing water levels and thus, facilitating rapid growth of woody vegetation. In addition, nutrients added to the groundwater from leaking septic tanks contribute to the dominance of invasive plants, such as narrow-leaved cattail (

Typha angustifolia

) and red canary grass (

Phalaris arundinacea

) (Michigan Natural Features Inventory 2012, p. 4).

Road construction impacts Poweshiek skipperling and Dakota skipper habitat because it increases the demand for gravel, and as a result of routine maintenance (

e.g.,

broadca

This text is long and has been trimmed here. Open the source document for the complete record.