Federal Register: PROPOSED RULES Endangered and Threatened Wildlife and Plants: [50 CFR Part 17]

DEPARTMENT OF THE INTERIOR

CFR Citation: 50 CFR Part 17

FWS ID: [FWS-R6-ES-2008-0013; 1111 FY07 MO-B2]

NOTICE: PROPOSED RULES

ACTION: Endangered and Threatened Wildlife and Plants:

DOCUMENT ACTION: Notice of a 12-month petition finding.

SUBJECT CATEGORY: Endangered and Threatened Wildlife and Plants; 12-Month Finding on a Petition To List the Gunnison's Prairie Dog as Threatened or Endangered

DATES: This finding was made on February 5, 2008.

DOCUMENT SUMMARY: We, the U.S. Fish and Wildlife Service (Service), announce our 12month finding on a petition to list the Gunnison's prairie dog (Cynomys gunnisoni) as an endangered or threatened species under the Endangered Species Act of 1973, as amended (Act). After a thorough review of all available scientific and commercial information, we find that the species is not threatened or endangered throughout all of its range, but that the portion of the current range of the species located in central and southcentral Colorado and northcentral New Mexico (the northeastern portion of the range) represents a significant portion of the range where the Gunnison's prairie dog is warranted for listing under the Act. Currently, listing is precluded by higher priority actions to amend the Lists of Endangered and Threatened Wildlife and Plants. We have assigned a listing priority number (LPN) of 2 to this species, because threats have a high magnitude, and are imminent. We will develop a proposed rule to list the Gunnison's prairie dog in the northeastern (montane) portion of its range as our priorities allow.

SUMMARY: 12-month Finding on a Petition to List the Gunnisons Prairie Dog as Threatened or Endangered,

SUPPLEMENTAL INFORMATION

Background

Section 4(b)(3)(B) of the Act (16 U.S.C. 1531 et seq.), requires that, for any petition containing substantial scientific and commercial information that listing may be warranted, we make a finding within 12 months of the date of receipt of the petition on whether the petitioned action is(a) not warranted, (b) warranted, or (c) warranted, but that immediate proposal of a regulation implementing the petitioned action is precluded by other pending proposals to determine whether species are threatened or endangered, and whether expeditious progress is being made to add or remove qualified species from the Lists of Endangered and Threatened Wildlife and Plants. Section 4(b)(3)(C) of the Act requires that we treat a petition for which the requested action is found to be warranted but precluded as though resubmitted on the date of such finding; that is, requiring a subsequent finding to be made within 12 months. We must publish these 12month findings in the Federal Register.

Previous Federal Actions

On February 23, 2004, we received a petition from Forest Guardians and 73 other organizations and individuals requesting that the Gunnison's prairie dog (found in Arizona, Colorado, New Mexico, and Utah) be listed as threatened or endangered.

On July 29, 2004, we received a 60day notice of intent to sue for failure to complete a finding. On December 7, 2004, an amended complaint for failure to complete a finding for this and other species was filed. We reached a settlement agreement with the plaintiffs, and on February 7, 2006, we published a 90day finding in the Federal Register (71 FR 6241) determining that the petition did not present substantial scientific information indicating that listing the Gunnison's prairie dog species may be warranted.

On August 17, 2006, Forest Guardians and eight other organizations and individuals provided written notice of their intent to sue regarding the determination in the 90day finding. On December 13, 2006, the plaintiffs filed a complaint challenging the finding. On June 29, 2007, we reached a settlement agreement with the plaintiffs for submittal to the Federal Register of a 12month finding by February 1, 2008. The court adopted the terms and conditions of the agreement on July 2, 2007.

On August 28, 2007, we published a notice initiating the 12month finding and opening a 60day public comment period on the Gunnison's prairie dog (72 FR 49245).

Species Information

A description of the Gunnison's prairie dog is included in the 90 day petition finding (71 FR 6241; February 7, 2006) and in a concise review of the published information by Underwood (2007, pp. 613). In addition, we used data in the Western Association of Fish and Wildlife Agencies' (WAFWA) Gunnison's Prairie Dog Conservation Assessment (Seglund et al. 2005) to complete much of our analysis in this finding.

The Gunnison's prairie dog has sometimes been divided into two subspecies: Cynomys gunnisoni gunnisoni and C. g. zuniensis (Hollister 1916, p. 29). We currently regard the Gunnison's prairie dog as a single species because the most recent published analyses (Goodwin 1995, pp. 100, 101, 110; Pizzimenti 1975, pp. 11, 15, 63) do not support subspecies designation. Unpublished research (Hafner 2004, p. 6; Hafner et al. 2005, p. 2) indicates that the distribution of mitochondrial DNA (deoxyribonucleic acid) haplotype lineages supports past geographic isolation, followed by limited mixing in regions coincident with the recognized borders of the two purported subspecies. Although this analysis will likely be substantiated through additional research, it is still preliminary and needs to be verified before we can use it as evidence for subspecies designation. For the same reasons, although Gunnison's prairie dogs in montane habitat may be ``markedly separate'' from those in prairie habitat, we are not proposing listing the montane prairie dogs as a distinct population segment (DPS) under our Policy Regarding the Recognition of Distinct Vertebrate Population Segments Under the Endangered Species Act (61 FR 4722; February 7, 1996). We anticipate that future funding may become available for genetic, taxonomic, and range research to determine whether subspecies or DPS status is valid.

Gunnison's prairie dogs are a colonial species, historically occurring in large colonies over large areas. Colonial behavior offers an effective defense mechanism by aiding in the detection of predators, but it also can play an important role in the transmission of disease (Antolin et al. 2002, p. 19; Biggins and Kosoy 2001, p. 911). Complexes of Gunnison's prairie dog colonies (metapopulations) expand or contract over time depending upon various natural factors (such as reproduction, food availability, and disease) and humancaused factors (such as chemical control and shooting). To substantially augment depleted populations or replace populations without human intervention, a metapopulation structure is required across the landscape so that migration between colonies is possible (Gilpin and Soule 1986, p. 24; Clark et al. 1982, pp. 574575; Lomolino and Smith 2001, p. 938). Habitat

Gunnison's prairie dog habitat includes level to gently sloping grasslands and semidesert and montane shrublands, at elevations from 6,000 to 12,000 feet (1,830 to 3,660 meters) (Bailey 1932, p. 125; Findley et al. 1975, p. 133; Fitzgerald et al. 1994, p. 183; Pizzimenti and Hoffman 1973, p. 1; Wagner and Drickamer 2002, p. 4). Grasses are the most important food item, with forbs, sedges, and shrubs also occasionally used (Pizzimenti and Hoffman 1973, p. 3; Shalaway and Slobodchikoff 1988, p. 840).

Gunnison's prairie dog range can be considered to occur in two separate range portionshigher elevations in the northeast part of the range and lower elevations elsewhere (Bailey 1932, pp. 125127; Pizzimenti and Hoffman 1973, pp. 12; Hall 1981, p. 7; Knowles 2002, p. 4). We refer to these areas as montane and prairie, respectively, throughout the document to differentiate them; however, we recognize that these terms are an oversimplification of the actual habitats present, and describe them in more detail below.

In Figure 1, we provide a map illustrating the division of the general range of the species into the northeastern (montane) and southwestern (prairie) portions. The outer boundary in Figure 1 is referenced from maps depicting the species' gross range (Hollister 1916, p. 24; Pizzimenti and Hoffman 1973, p. 2; Pizzimenti 1975, p. 4; Hall 1981, p. 415; Knowles 2002, p. 6), and from maps of the species' range in Arizona (Hoffmeister 1986, p. 194), Colorado (Armstrong 1972, p. 139; Fitzgerald et al. 1994, p. 185), New Mexico (Findley et al. 1975, p. 133), and Utah (Durrant 1952, p. 106). An approximate boundary dividing the montane and prairie range portions was established from several maps that recognize discrete range portions for each of the two purported subspecies,
[[Page 6662]]
Cynomys gunnisoni gunnisoni and C. g. zuniensis (Hollister 1916, p. 24; Armstrong 1972, p. 139; Pizzimenti and Hoffman 1973, p. 2; Pizzimenti 1975, p. 4; Hall 1981). Maps that depict the geographic variation in Gunnison's prairie dog mitochondrial DNA in southern Colorado and northern New Mexico (Hafner 2004, p. 6; Hafner et al. 2005, p. 2) were used to improve the resolution of the montane and prairie boundary in this region, as these maps provide a boundary based on genetic differences between Gunnison's prairie dogs in the two range portions. Lastly, we used topographic maps to adjust the boundary on a finer scale along the mountain ranges and ridges of southern Colorado and northern New Mexico, because geography partly separates the Gunnison's prairie dog populations and allows limited overlap between the two range portions (Knowles 2002, p. 3; Hafner et al. 2005, p. 1).

In summary, the maps we used to delineate the montane and prairie range portions vary in their age, projection, scale, and accuracy, and depict boundaries based on geography, morphological traits of Gunnison's prairie dog populations, and genetic characteristics from Hafner's work (Hafner 2004, p. 6; Hafner et al. 2005, p. 2). They contribute to the best available information used to establish the montane and prairie portions of the species' range for further analysis.
BILLING CODE 431055P
[[Page 6663]]
[GRAPHIC] [TIFF OMITTED] TP05FE08.000
BILLING CODE 431055C
[[Page 6664]]

Montane Habitat

The northeastern range (central and southcentral Colorado, and northcentral New Mexico) consists primarily of higher elevation, cooler and more mesic plateaus, benches, and intermountain valleys. We call this portion ``montane'' for ease of reference, and it comprises approximately 40 percent of the total potential habitat within the current range. Gunnison's prairie dogs occupy grassshrub areas in low valleys and mountain meadows within this habitat (Seglund et al. 2005, p. 12). The Gunnison's prairie dogs in this portion of the range are limited by pronounced physiographic barriers (Pizzimenti and Hoffman 1973, p. 1), including the Uncompahgre Plateau and San Juan mountains in Colorado and Utah, and the Sangre de Cristo, San Juan, and Jemez mountain ranges in New Mexico.

Prairie Habitat

The southwestern range (southeastern Utah, southwestern Colorado, northwestern New Mexico, and northeastern Arizona) consists primarily of lower elevation, warmer and more xeric plains and plateaus (Bailey 1932, pp. 125127; Pizzimenti and Hoffman 1973, pp. 12; Hall 1981, p. 7; Knowles 2002, p. 4). We call this portion ``prairie'' for ease of reference, and it comprises approximately 60 percent of total potential habitat within the current range. Gunnison's prairie dogs occupy shortgrass and midgrass prairies within this habitat (Seglund et al. 2005, p. 12).

Distribution, Abundance, and Trends

The current distribution of the species includes northeastern Arizona; central, southcentral, and southwestern Colorado; north central and northwestern New Mexico; and extreme southeastern Utah (Bailey 1932, pp. 125127; Pizzimenti and Hoffman 1973, pp. 12; Hall 1981, p. 7; Knowles 2002, p. 4) (see Figure 1 above). Limited overlap occurs in the ranges of Gunnison's prairie dogs and blacktailed prairie dogs (Cynomys ludovicianus) in New Mexico (Goodwin 1995, p. 101; Sager 1996, p. 1), and Gunnison's prairie dogs and whitetailed prairie dogs (Cynomys leucurus) in Colorado (Knowles 2002, p. 5), but we have no evidence that interbreeding is occurring. Currently, 27 percent of potential Gunnison's prairie dog habitat occurs in Arizona, 25 percent in Colorado, 45 percent in New Mexico, and 3 percent in Utah (Seglund et al. 2005, p. 83). We used the data in Seglund et al. (2005, pp. 82, 8587) to calculate that approximately 22 percent of the potential habitat occurs on private lands, 12 percent on State lands, 17 percent on Federal lands, and 49 percent on Tribal lands/Bureau of Indian Affairs (BIA). The Tribal lands habitat occurs mostly in Arizona and New Mexico; a large amount of potential habitat is on Navajo lands (Cole, p. 1).

Most estimates of prairie dog populations in the available literature are expressed in terms of area (acres (ac) or hectares (ha)) of occupied habitat rather than in numbers of individuals, most likely because counting individuals is feasible only for small areas (Biggins et al. 2006, p. 94). Also, the number of animals present in a locality has been observed to vary with habitat, season, colony age, precipitation, forage, predation, disease, chemical control, shooting, and other factors (Knowles 2002, pp. 78); density of individuals typically ranges from 2 to 23 per ac (5 to 57 per ha) (Fitzgerald et al. 1994, p. 184). Most prairie dog surveys do not result in a density estimate because of the associated effort and cost. Estimates of Gunnison's prairie dog occupied habitat provide one of the best available and most reasonable means of evaluating the status of the species across its range.

Obtaining estimates of occupied area is itself timeconsuming and costly. Ground or aerial mapping of colonies over a predicted habitat range of 23 million ac (9.5 million ha) in 4 States would be required to determine a rangewide estimate of the area occupied by the Gunnison's prairie dog (Seglund et al. 2005, pp. 1719). Recent attempts at less expensive aerial surveys (for example, air photo interpretation) have been limited in their effectiveness when applied to Gunnison's prairie dogs (Johnson et al. 2006, p. 3; Seglund et al. 2005, pp. 2324). Whether surveying is performed from the air or on the ground, it is often difficult to accurately and consistently discern colony boundaries (thus introducing error in the area measurements). Older studies did not benefit from technologies such as global positioning systems and geographic information systems (GIS) in mapping colonies. Accuracy suffers when studies are performed over the longer time intervals necessary to visit large range portions, because colony area, location, and persistence on the landscape often change relatively quickly (Wagner et al. 2006, p. 335).

In summary, we recognize that different methodologies were used at different times and in different locales to derive the various historical occupied area estimates we obtained for review. These estimates contribute to the best available information, and we consider them comparable for determining longterm population trends, while acknowledging potential error margins on the scale of an order of magnitude.

Since our 90day finding in 2006, all States within the range of the species have applied occupancy modeling methodology to investigate the habitat occupied by Gunnison's prairie dogs. This is a newer technique that yields estimates of the percentage of random plots occupied across the habitat range under consideration (MacKenzie et al. 2002, pp. 22482249; MacKenzie et al. 2003, pp. 22002201). These estimates are statistically based and, therefore, are considered more objective (Andelt et al. 2006, pp. 12; Colorado Division of Wildlife (CDOW) 2007, p. 19; WAFWA 2007, p. 4).

A drawback is that estimates of percent occupancy by Gunnison's prairie dogs are not directly comparable to estimates of occupied acres (including most historic estimates), because when a random plot is visited, only detection or nondetection (not acres occupied) is recorded by the observers. If mapping is not performed during a site visit, no information about colony or complex size or location is obtained.

The positive aspects of this method are statistical rigor, precision estimates, largescale application in a single season, and trend analysis if performed over subsequent years. In addition, the results of individual surveys can be interpreted separately to assess prairie dog occupancy and document trends within in specific areas of concern. Although only a single year (2007) of occupancy modeling results are available (with the exception of Colorado data from 2005 and 2007), we used these estimates, along with estimates of occupied areas, to assess the status and trends of the Gunnison's prairie dog in each of the four States.

Historical Estimates of Abundance

Historical estimates of Gunnison's prairie dog occupied habitat in Arizona and New Mexico are available from Federal records of early poisoning efforts, such as by the Bureau of Land Management (BLM) and the U.S. Forest Service (USFS). In 1916, approximately 6.6 million ac (2.7 million ha) of Gunnison's prairie dog occupied habitat occurred in Arizona, and 11 million ac (4.4 million ha) occurred in New Mexico (Oakes 2000, pp. 169171). In our 90day finding in 2006 (71 FR 6241, February 7, 2006), we calculated historical estimates (circa 1916) for Colorado (6 million ac (2.4 million ha)) and Utah (700,000 ac (284,000 ha)) from prairie dog information in various
[[Page 6665]]
publications and reports, because data were not available for these States. By summation, based on the best available information, our rangewide estimate for historic (circa 1916) Gunnison's prairie dog occupied habitat was approximately 24 million ac (9.7 million ha).

In 1961, an estimated 445,000 ac (180,000 ha) of habitat was occupied by Gunnison's prairie dog in Arizona; 116,000 ac (47,000 ha) in Colorado; 355,000 ac (144,000 ha) in New Mexico; and 100,000 ac (41,000 ha) in Utah (Bureau of Sport Fisheries and Wildlife 1961, pp. 1, 5). By summation, the rangewide estimate for Gunnison's prairie dog occupied habitat in 1961 was approximately 1 million ac (405,000 ha). These data suggest that, from 1916 to 1961, Gunnison's prairie dog populations decreased by approximately 93 percent in Arizona, 98 percent in Colorado, 97 percent in New Mexico, and 86 percent in Utah, or by approximately 95 percent rangewide. However, historic declines may not support a conclusive inference that current populations continue to decline.

In summary, empirical data on acres occupied indicate that, between 1916 and 1961, habitat occupied by the Gunnison's prairie dog throughout its range declined from approximately 24,000,000 ac (9,700,000 ha) to approximately 1,016,000 ac (406,400 ha).

Statewide Estimates of Abundance

As indicated above, estimates of percent occupancy arrived at through recent occupancy modeling (presence or absence at a random plot) do not equate to acres occupied. The method currently used by States to assess the Gunnison's prairie dog's status, in conjunction with both historic and recent mapping efforts, provides empirical data on percent occupancy of potential habitat. This data is useful as a grossscale comparison to historical estimates of acres occupied. Both types of data are valid and represent the best available science.

Full occupancy of surveyed habitat would not directly equate to 100 percent of available habitat, but it would provide a gross approximation of occupancy at a larger geographic scale. For the purposes of interpreting the percent occupancy numbers in this document, current State survey efforts utilize a scale from 1 to 100, indicating the percentage of occupied cells surveyed. Because we do not have historical data on percent of habitat occupied or on occupancy rates, we use the current percentage of occupied habitat to compare between habitats that currently appear to have a functional metapopulation structure (prairie) and that do not (montane). For example, the following paragraphs illustrate that Gunnison's prairie dog occupancy in plots sampled in montane habitat is estimated to be approximately 3.6 percent as compared to approximately 18.3 percent in plots sampled in prairie habitat in Colorado. Of the total montane habitat, approximately 85 percent occurs in Colorado.

Arizona

In 2007, occupied habitat on nonTribal lands in Arizona comprised approximately 108,570 ac (40,500 ha) (Underwood 2007, p. 30). No comprehensive data are available from Tribal lands in Arizona, which include 50 percent of the Statewide potential habitat. Therefore, the 2007 estimate for Arizona (Underwood 2007, p. 30) is likely substantially less than what actually exists. Due to a lack of any Tribal estimates since 1961, recent population trends on Tribal lands statewide are unknown, but may have increased over the 1961 estimate of 435,419 ac (176,207 ha). We are unaware of any disproportionate adverse effects to the species on Tribal lands during this interval, and we assume that habitat trends may have followed a similar pattern as on nonTribal lands. All habitat within Arizona is considered prairie. Colorado

The Colorado Department of Agriculture (CDA 1990, p. A3) solicited questionnaire responses from farmers and ranchers from which they extrapolated a 1990 estimate of 1,553,000 ac (621,200 ha) of occupied habitat for all 3 species of prairie dogs found in Colorado (Gunnison's, whitetailed, and blacktailed). Based on species occurrence by county, Seglund et al. (2005, p. 26) estimated that 438,876 ac (177,607 ha) were occupied by Gunnison's prairie dogs.

From 2002 to 2005, the Colorado Division of Wildlife (CDOW) interviewed field personnel from CDOW, the Service, the USFS, and the BLM regarding the habitat occupied by Gunnison's prairie dogs in the State. Colonies were mapped on 1:50,000 scale U.S. Geological Survey county sheets and were designated as ``active'' (known to have prairie dogs inhabiting the colony within the last 3 years); ``inactive'' (prairie dogs occurred in the area but have not been present in more than 3 years); or ``unknown'' (prairie dogs were known to occur historically, but current status was unknown). From this effort, CDOW estimated 182,237 ac (72,895 ha) of active colonies; 9,042 ac (3,617 ha) of inactive colonies; and 171,970 ac (68,788 ha) of colonies in unknown status within Colorado (CDOW 2007, p. 3). These data suggest an increase over the historical 1961 estimate of 115,650 ac (46,802 ha) of occupied habitat in Colorado. We have no way of estimating what percent of this difference may be due to different mapping techniques. We believe that the difference is mostly due to an actual increase in prairie dogs, likely within the prairie portion of the range, because data from the montane portion of the range indicate significantly reduced occupancy rates (see additional analysis below). We used area estimates from 2002 to 2005 to compute a Statewide occupancy estimate of 2.1 percent (known active colony area divided by area of potential habitat) (CDOW 2007). However, the occupancy modeling studies performed in 2005 and 2007 in Colorado, including both prairie and montane portions of the range, yielded Statewide occupancy estimates of 7.5 and 8.6 percent, respectively (Andelt et al. 2006, p. 15; CDOW 2007, p. 19), and these estimates are considered more reliable.

Montane and Prairie Habitat in Colorado

Within Colorado, CDOW has designated individual population areas to identify where Gunnison's prairie dogs exist and where management activities should be focused. The montane portion of the species' range in Colorado is composed of the Gunnison, San Luis Valley, South Park, and Southeast population areas. By using CDOW (2007, p. 28) estimates of potential habitat, we determined that the montane range portion in Colorado comprises about 80 percent (6.9 million of 8.5 million ac (2.8 million of 3.4 million ha)) of the available Gunnison's prairie dog habitat in the State. However, the montane range portion only contains about 40 percent (73,861 of 182,237 ac (29,544 of 72,894 ha)) of the available Gunnison's prairie dog habitat occupied in the State, based on our calculations using CDOW mapped area data (CDOW 2007, p. 3).

The La PlataArchuleta and Southwest population areas, in the prairie portion of Colorado's Gunnison's prairie dog habitat, comprise about 20 percent of the Gunnison's prairie dog habitat and contain about 60 percent of habitat occupied in the State (CDOW 2007, pp. 3, 19). The higher proportion of occupied habitat in the smaller prairie portion of the State indicates that Gunnison's prairie dogs are more abundant in the prairie habitat area.

[[Page 6666]]

The 2005 occupancy modeling studies also indicate a higher proportion of occupancy (16 percent) in the prairie portion of the range in Colorado, and a lower proportion of occupancy (3.2 percent) in the montane portion of the species' range in Colorado (Andelt et al. 2006, p. 17; CDOW 2007, p. 19). When the study was repeated over the same plots in 2007, occupancy was again found to be higher (18.3 percent) in the prairie portion and lower (3.6 percent) in the montane range portion in Colorado (CDOW 2007, p. 19).

New Mexico

We have no current information on occupied habitat in New Mexico. The best available science is from Bodenchuck (1981 p. 1), who solicited questionnaire responses from agricultural producers in 1981. Respondents reported 107,574 ac (43,567 ha) of Gunnison's prairie dog occupied habitat. Bodenchuck (1981, p. 8) extrapolated a Statewide total of 348,000 ac (141,000 ha) of occupied habitat for the species. Oakes (2000, p. 216) questioned this extrapolation because of possibly faulty assumptions used to derive it. Knowles (2002, p. 22) estimated that 75,000 ac (30,000 ha) of occupied habitat existed in 1982. New Mexico Department of Game and Fish used Digital Orthophoto Quarter Quadrangles to estimate a minimum of 9,108 ac (3,689 ha) of occupied habitat Statewide in 2004 (Seglund et al. 2005, p. 23). However, this method appears to be hampered by inaccurate detection of disturbances, time elapsed since photography, time elapsed since ground mapping, temporal changes in prairie dog towns, and other factors (Seglund et al. 2005, p. 33). While these estimates have limited accuracy, general use in assessing Statewide occupied habitat indicates that Gunnison's prairie dogs appeared to be decreasing between 1961 and 2004. Montane and Prairie Habitat in New Mexico

New Mexico also includes both montane and prairie habitat. The montane habitat is geographically connected to the montane portion of the Gunnison's prairie dog habitat in Colorado. It comprises about 17 percent of the Gunnison's prairie dog habitat in New Mexico; we do not have accurate data on total acres in New Mexico, and therefore do not provide an acre estimate for the montane portion. We have no data on the percent occupancy in this habitat.

The prairie habitat in New Mexico comprises about 83 percent of the habitat; we do not have accurate data on total acres in New Mexico, and therefore do not provide an acre estimate for the prairie portion. We have no data on the percent occupancy in this habitat.

Utah

The Utah Division of Wildlife Resources estimated that 22,000 ac (8,906 ha) of occupied Gunnison's prairie dog habitat existed in Utah in 1968 (Seglund et al. 2005, p. 35). Knowles (2002, p. 21) estimated a minimum of 3,678 ac (1,490 ha) of occupied habitat Statewide. The Statewide trend in occupied habitat appears to have decreased from 100,000 ac (40,500 ha) in 1961 (Bureau of Sport Fisheries and Wildlife 1961, p. 5), to 40,000 ac (16,000 ha) in 2007 (Lupis et al. 2007, p. 3). The Gunnison's prairie dog occupancy in Utah was estimated to be 15.7 percent in 2007 (Lupis et al. 2007, p. 3). We consider all Gunnison's prairie dog habitat in Utah as prairie.

Summary of Statewide Estimates of Abundance

We have empirical data on Gunnison's prairie dog occupancy that indicate a large decline in rangewide occupied acres. We also have recent empirical data that indicates percent occupancy within two separate portions of the range is significantly different.

Data on acres occupied indicate that between 1916 and the present, habitat occupied by Gunnison's prairie dogs throughout its range declined from approximately 24,000,000 ac (9,700,000 ha) to between 340,000 and 500,000 ac (136,000200,000 ha). This represents a rangewide decline of greater than 95 percent.

Summary of Factors Affecting the Species Rangewide

Section 4 of the Act (16 U.S.C. 1533) and implementing regulations at 50 CFR 424, set forth procedures for adding species to the Federal Lists of Endangered and Threatened Wildlife and Plants. In making this finding, we summarize below information regarding the status and threats to the Gunnison's prairie dog in relation to the five factors provided in section 4(a)(1) of the Act.

In making this 12month finding, we have considered all scientific and commercial information received or acquired between the time of the initial petition (February 23, 2004) and the end of the most recent public comment period (October 29, 2007), and additional scientific information from ongoing species surveys and studies as they became available.

Under section (4) of the Act, we may determine a species to be endangered or threatened on the basis of any of the following five factors: (A) Present or threatened destruction, modification, or curtailment of habitat or range; (B) overutilization for commercial, recreational, scientific, or educational purposes; (C) disease or predation; (D) inadequacy of existing regulatory mechanisms; or (E) other natural or manmade factors affecting its continued existence. We evaluated whether threats to the Gunnison's prairie dog may affect its survival. Our evaluation of threats, based on information provided in the petition, available in our files, and available in published and unpublished studies and reports, is presented below.
A. The Present or Threatened Destruction, Modification, or Curtailment of its Habitat or Range

Agricultural land conversions historically had a significant impact on Gunnison's prairie dog habitat (Knowles 2002, p. 12). Gunnison's prairie dogs have been displaced from some of the more productive valley bottomlands in Colorado and New Mexico (Longhurst 1944, p. 36). Agriculture currently impacts 2,063,930 ac (834,243 ha), or less than three percent, of the Gunnison's prairie dog range (Seglund et al. 2005, p. 43). Seglund et al. (2005, p. 41) indicate agriculture is not a major rangewide threat because of the small percentage of the range affected, but also because agriculture provides highly productive forage in place of the native arid landscape. Current adverse impacts relate to secondary actions at a local scale, such as prairie dog control (for example, poisoning, shooting) in areas where prairie dogs occupy lands used for agriculture, particularly private lands. We assess shooting under Factor C, poisoning under Factor E, and both in Factor D.

Urbanization also has caused habitat loss for Gunnison's prairie dog. Seglund et al. (2005, p. 41) determined that urbanization affects 577,438 ac (233,681 ha) within the range of the species (less than two percent of the range). However, it appears this analysis considered only the direct effects of habitat loss. Urbanization also exerts indirect effects (for example, poisoning and shooting of prairie dogs), extending a human ``disturbance zone'' outward from the actual development footprint.

Lowerdensity suburban development occurring in the southern Rocky Mountains is scattered and results in a fragmenting of habitats. In Colorado, urban development on the west slope of the Rocky Mountains (montane habitat) is occurring rapidly; 38 percent of Gunnison's prairie dog range is predicted to be impacted by low urban development (less than 40 units per ac;
[[Page 6667]]
99 per ha), 6 percent by moderate development (40 to 80 units per ac; 99 to 198 per ha), and 5 percent by high development (fewer than 80 units per ac) between 2000 and 2020 (CDOW 2007, p. 28). We do not have information on the extent of development projected to occur in the other States within the species' range (Utah, Arizona, and New Mexico). Potential threats to Gunnison's prairie dog populations due to urban and suburban development exist, but have not been quantified, in the four cities of Santa Fe and Albuquerque, New Mexico; Flagstaff, Arizona; and Gunnison, Colorado (CDOW 2007, p. 4). In some areas, Gunnison's prairie dogs threatened by urban development have been captured and relocated to preserves or other nearby habitats, mitigating effects to overall population numbers, but not to area of habitat.

Although urban and suburban development exert adverse impacts on Gunnison's prairie dog populations at a local scale, they likely affect less than three percent of the species' range; low density development appears to be compatible with continued use by prairie dogs, due to the offsets provided by lawns and pastures that provide high quality forage (Seglund et al. 2005, p. 41).

Noxious weeds can increase in the presence of livestock overgrazing, and a relationship likely exists between overgrazing, Bromus tectorum (cheat grass) proliferation, and increased fire frequency and intensity (Seglund et al. 2005, p. 43). However, we have no data that quantifies these factors or their correlation with effects to Gunnison's prairie dog populations. The impact of overgrazing on prairie dog populations is contradictory. Some reports have noted that species density is positively correlated with the number of native plants (Slobodichikoff et al. 1988, p. 406), and that grazing has decreased forage availability (Seglund et al. 2005, p. 42). Other reports have concluded that prairie dog density is positively correlated with an increase in grazing, which simulates the shortgrass type of prairie environment preferred by prairie dogs (Fagerstone and Ramey 1996, p. 88; Marsh 1984, p. 203, Slobodchikoff et al. 1988, p. 406). Considering the conflicting conclusions of published literature, and the lack of largescale population decreases due to habitat alterations from livestock grazing, we find this is not a significant threat to the Gunnison's prairie dog.

Numerous land parcels within the Gunnison's prairie dog range are leased for oil and gas development (Seglund et al. 2005, pp. 36, 42). However, no information is available that quantifies the amount of occupied habitat affected. In a study of whitetailed prairie dogs, Menkens and Anderson (1985, p. 13) concluded that any impact from seismic testing is negligible. However, we acknowledge that oil and gas development is rapidly occurring (Seglund et al. 2005, p. 44), and that this potential threat should be considered more closely when more accurate data are available.

Roadrelated Gunnison's prairie dog mortality exists in proximity to specific population areas. Roads may be increasing due to oil and gas development. However, no studies quantify road mortality of Gunnison's prairie dogs. We have no data indicating that roads are currently threatening the species rangewide, and we conclude that prairie dog populations are able to recover from individual losses due to road mortality.

Conservation principles indicate that smaller, more isolated populations are more vulnerable to extirpation (Barnes 1993, p. 34; Cully 1993, p. 43; Fitzgerald 1970, p. 78; Gilpin and Soule 1986, pp. 3031; Miller et al. 1994, p. 151; Mulhern and Knowles 1995, p. 21; Wilcox and Murphy 1985, p. 883; Wuerthner 1997, p. 464). Lomolino et al. (2003, p. 116) found that persistence of Gunnison's prairie dog colonies increased significantly with larger colony size and decreased isolation. However, we found no studies or data that specifically assess the magnitude of the threats discussed under Factor A (agriculture land conversions, urbanization, grazing, roads, and oil and gas leasing) and resulting fragmentation throughout the range of Gunnison's prairie dog habitat.

Summary of Factor A

After assessing the best available science on the magnitude and extent of the effects of agricultural land conversion, urbanization, grazing, roads, oil and gas development, and fragmentation of habitat, we find that the destruction, modification, and curtailment of Gunnison's prairie dog's habitat or range are not significant threats. Agriculture, urbanization, roads, and oil and gas development each currently affect a small percentage of Gunnison's prairie dog habitat. Effects of livestock grazing, while widespread, have not resulted in measurable population declines. However, we need more information on the impacts of fragmentation and isolation with regard to persistence of prairie dog populations and on the magnitude of the potential threat posed by increasing oil and gas development.
B. Overutilization for Commercial, Recreational, Scientific, or Educational Purposes

Gunnison's prairie dogs have been historically subjected to recreational shooting and shooting as a form of pest management on ranch and agricultural land; these practices continue under current State regulations (see Factor D. Inadequacy of Existing Regulatory Mechanisms). Prairie dogs are especially vulnerable to shooting due to their colonial behavior, which facilitates easy access to many individuals at once (Seglund et al. 2005, p. 48). Most field studies on the effects of shooting prairie dogs have been carried out on black tailed prairie dogs, but we consider the results relevant to Gunnison's prairie dogs (CDOW 2007, p. 41). Shooting effects include population reduction and alteration of behavior, such as decreased foraging rates and increased vigilance, which reduce individual prairie dog vigor and result in lower reproductive output (Knowles 1988, p. 54; Reeve and Vosburgh in press, p. 5; Vosburgh 1996, pp. 3233; Vosburgh and Irby 1998, p. 368; Pauli and Buskirk 2007, pp. 12231224).

Recreational shooting can reduce prairie dog population density at specific sites (Knowles 2002, p. 14; Miller et al. 1993, p. 91; Vosburgh 1996, pp. 1314; Vosburgh and Irby 1998, pp. 366367). Local extirpation of colonies may have occurred in isolated circumstances in the past (Knowles 1988, p. 54). However, increased population growth rates or recovery from very low numbers following shooting also have been reported (Knowles 1988, p. 54; Reeve and Vosburgh in press, p. 7). Recent studies of the effects of shooting on blacktailed prairie dogs appear to contradict the idea that populations quickly rebound from shooting. Reproductive output on colonies subjected to shooting decreased by 82 percent, while control colonies maintained a stable reproductive rate over the same period (Pauli and Buskirk 2007, p. 1228). Therefore, blacktailed prairie dogs do not appear to rebound quickly from shooting.

The International Union for the Conservation of Nature/Species Survival Commission (IUCN/SSC) Conservation Breeding Specialist Group evaluated the effects of shooting mortality on population viability of Gunnison's prairie dogs (CDOW 2007, p. 124). Simulations were run with a shooting closure in place from March 1 through June 14 each year (approximating State closures) and without any closures. Having the closure in place resulted in positive population growth and [[Page 6668]]
negligible risk of extinction, except in scenarios with the highest levels (20 percent) of shootingbased mortality. Simulations run without the seasonal shooting closure in place suggest that when initial population sizes are smaller (less than 250 individuals) and shooting mortality is high (20 percent), a decrease in growth rate and an increase in population extinction risk exist (CDOW 2007, pp. 135 137). Colorado, Utah, and Arizona (outside Tribal lands) have implemented seasonal closures on prairie dog shooting. In Arizona and New Mexico, the Navajo Nation monitors this threat but currently implements no closures on shooting because it finds the level of shooting to be low on its lands (Cole 2007, p. 4).

Summary of Factor B

We have determined that shooting continues to be a threat to the Gunnison's prairie dog throughout all of its range and contributes to the decline of the species when combined with the effects of disease (see Factor C below). However, this threat is being monitored and managed in all States and the Navajo Nation, and modeling results suggest seasonal shooting closures implemented in Colorado and Arizona will likely reduce populationlevel losses. Therefore, we have determined that overutilization for commercial, recreational, scientific, or educational purposes is not a significant threat to the Gunnison's prairie dog.

C. Disease or Predation

While prairie dogs are prey to numerous species, including coyotes, badgers, blackfooted ferrets, and various raptor species, there is no information available to indicate that predation has an overall adverse effect on the species. Blackfooted ferrets have been reintroduced into two locations in Arizona, including the Aubrey Valley, where Gunnison's prairie dog populations appear to be stable.

The Gunnison's prairie dog is, however, affected by sylvatic plague, which occurs in regular outbreaks and causes population declines and extirpations. Plague is an exotic disease foreign to the evolutionary history of North American species (Barnes 1982, p. 238; Barnes 1993, p. 29; Biggins and Kosoy 2001, p. 907). This fleaborne disease, caused by infection with the bacterium Yersinia pestis, is shared by humans and other vertebrate animals. Rodents are the primary vertebrate hosts of Y. pestis, but other mammals can be infected. Y. pestis is transmitted to mammals by bites of infected fleas, direct contact with infected animals, and rarely by inhalation of infectious respiratory droplets from another animal (Gage et al. 1995, pp. 695 696). Plague was first observed in wild rodents (termed sylvatic plague) in North America near San Francisco, California, in 1908 and was detected in Gunnison's prairie dogs in the 1930s (Eskey and Hass 1940, p. 6). Plague has subsequently spread so that it now encompasses the entire range of the species (Centers for Disease Control 1998, p. 1; Cully 1989, p. 49; Girard et al. 2004, p. 8408). Therefore, it has only been present within the species' range for approximately 70 years, allowing very little time for any resistance to evolve (Biggins and Kosoy 2001, p. 913). Once established in an area, plague becomes persistent and periodically erupts, with the potential to eventually extirpate or nearly extirpate entire colonies (Barnes 1982, p. 255; Barnes 1993, p. 28; Cully 1989, p. 51; Cully et al. 1997, p. 711; Fitzgerald 1993, pp. 5253). The term ``enzootic'' describes plague existing at a less severe level, sometimes referred to as a ``maintenance'' condition, that is present continuously throughout a species' habitat; the term ``epizootic'' describes a severe plague outbreak or amplification transmission cycle (Gage et al. 1995, p. 696).

Prairie dogs are highly susceptible to plague, and this susceptibility is thought to be a function of high population densities, abundant flea vectors, and uniformly low resistance (Biggins and Kosoy 2001, p. 913). Gunnison's prairie dogs can experience mortality rates of greater than 99 percent during epizootics, and eradication of populations can occur within one active season (Lechleitner et al. 1962, pp. 190192; Lechleitner et al. 1968, p. 736; Rayor 1985, p. 194; Cully 1989, p. 49).

Oral vaccination through consumption of vaccineladen baits could reduce mortality from plague. Mencher et al. (2004, pp. 55045505) report protection against plague in blacktailed prairie dogs, elicited through voluntary consumption of a vaccineladen bait in the laboratory. The vaccine has been shown to be safe in numerous animals including blackfooted ferrets, raccoons, skunks, bobcats, cats, dogs, and sheep. However, future field trials are required to test the efficacy on the Gunnison's prairie dog.

Recovery rates of Gunnison's and Utah prairie dog colonies studied 2 years postepizootic found that Gunnison's prairie dog colonies experienced 100 percent mortality and remained depopulated throughout the study due to the lack of available immigrants (Turner 2001, p. 14). Partial or complete recovery following population reductions due to plague have been reported for both whitetailed and blacktailed prairie dogs (Cully 1993, pp. 4041), but little to no recovery has been noted in montane Gunnison's prairie dog colony dieoffs, even after long periods of time (Capodice and Harrell 2003, pp. 57; Cully et al. 1997, p. 717; Lechleitner et al. 1968, p. 734). Possible long term consequences of continued plague infection in Gunnison's prairie dog populations may be:
(1) local extirpation of colonies;
(2) reduced colony size;
(3) increased variance in local population sizes, and
(4) increased intercolony distances (CDOW 2007, p. 43).

The factors that influence interspecific (between species) transmission of plague from mammalian or avian reservoirs (for example, coyotes, raptors, corvids) into prairie dog populations are unclear, but seem to be primarily through fleas that could increase in moister climates (Parmenter et al. 1999, p. 818; Rayor 1985, p. 195). However, interspecific transmission does not seem to be a significant factor creating plague epizootics. Plague is now considered enzootic throughout the range of the Gunnison's prairie dog.

The primary factor influencing plague enzootics in Gunnison's prairie dogs is thought to be abundance of fleas within their own colonies. This appears to be correlated with seasonal moisture in specific habitat areas. Plague outbreaks may be triggered by climatic conditions, such as mild winters and moist springs (Parmenter et al. 1999, p. 818; Rayor 1985, p. 195). Enscore et al. (2002, p. 191) found a close relationship between human plague cases in the southwestern United States and high amounts of late spring (February to March) precipitation (timelagged 1 and 2 years) and maximum daily summer temperature values in the moderately high range (85 to 90 [deg]F; 29 to 32 [deg]C).

Girard et al. (2004, p. 8408) postulated that when resistant hosts of plague encounter a susceptible species that is plague na[iuml]ve and has a high population density, an epizootic occurs. During epizootic phases, declines in abundance of susceptible species like prairie dogs are observed (Hanson et al. 2007, p. 790). The rapid dispersal of the pathogen through an area can be followed by an enzootic phase, a slower transmission cycle that disperses through the lowerdensity, more resistant hosts remaining from the first cycle. This establishes the disease in stable reservoirs for future emergence
[[Page 6669]]
(Girard et al. 2004, p. 8413; Gage and Kosoy 2005, pp. 506509).

Enzootic infection is generally considered characteristic of a stable rodentflea infectious cycle where host rodents are relatively resistant to the disease. However, Hanson et al. (2007, p. 792) found that an unexpectedly high percentage of blacktailed prairie dog colonies in Montana tested positive for plague. They speculate that, under some conditions, blacktailed prairie dogs, rather than acting as resistant hosts, may serve as enzootic hosts or carriers of the pathogen. Plague antibody titers (concentrations in blood) have been found in small numbers of Gunnison's prairie dogs in New Mexico, indicating individual exposure to plague and subsequent recovery (Cully et al. 1997, p. 717; Cully and Williams 2001, p. 898). Plague appears to have had little effect on a Gunnison's prairie dog population in Aubrey Valley, Arizona (Wagner and Van Andelt 2007, p. 2). However, little evidence of resistance to plague has been found in any species of prairie dog at this time.

In conducting a Population Viability Analysis on Gunnison's prairie dogs, the IUCN/SSC Conservation Breeding Specialist Group (CDOW 2007, p. 123) hypothesized that in an enzootic scenario, plague operates at a relatively low level each year, thereby increasing average annual rates of mortality above what would occur in a more benign nonenzootic scenario.

Gunnison's prairie dog populations are more susceptible to decline from plague than whitetailed prairie dog populations and are at least as, if not more, susceptible than blacktailed prairie dog populations (Antolin et al. 2002, p. 14; Cully 1989, p. 51; Cully and Williams 2001, p. 899; Hubbard and Schmitt 1983, p. 51; Knowles 2002, p. 13; Ruffner 1980, p. 20; Torres 1973, p. 31; Turner 2001, p. iii). Gunnison's prairie dogs commonly forage outside of their home territory, a characteristic that may play a significant role in the susceptibility of the species to plague. The Gunnison's prairie dog may be more susceptible to plague than the blacktailed prairie dog because of the Gunnison's less exclusive territorial behavior (many mix relatively freely throughout adjacent territories) and thereby contribute to the communicability of plague (Hoogland 1999, p. 8).

The Gunnison's prairie dog is also likely more susceptible to plague than the whitetailed prairie dog because the Gunnison's typically occurs at higher densities and is less widely dispersed on the landscape, allowing for more frequent transmission of the disease from one individual to another (Antolin et al. 2002, p. 19; Cully 1989, p. 49; Cully and Williams 2001, p. 901; Turner 2001, p. 31). Biggins (2003, p. 6) speculated that if transmission rates for plague are at least partly dependent on host density, prairie dog populations on good quality sites may undergo both larger declines and more rapid recoveries than those on poor sites.

Available literature is inconclusive regarding whether isolation or density of a colony affects the number and frequency of plague outbreaks. Lomolino et al. (2003, p. 118) and others (Cully and Williams 2001, p. 901; Miller et al. 1993, pp. 8990) suggested that isolation and fragmentation may provide some protection to prairie dogs from plague by lessening the likelihood of disease transmission. However, this theory no longer applies when plague is enzootic throughout the range of Gunnison's prairie dog (as it appears to be), in which case isolation of colonies reduces the chance of
recolonization after extirpation (Wagner and Drickamer 2002, p. 16; Lomolino and Smith 2001, pp. 942943). In areas where Gunnison's prairie dog colonies are located close to each other (less than 6 miles (mi) (10 kilometers (km) apart), intercolony dispersal of plague is likely through infected prairie dogs (Girard et al. 2004, p. 8412). For colonies separated by long distances or unsuitable habitats, infection may occur due to longdistance dispersal of plagueinfected fleas by domestic dogs, coyotes, raptors, or other predators and scavengers (Barnes 1993, p. 34), or plague may already persist as enzootic throughout Gunnison's prairie dog range.

The impacts of plague outbreaks, which lead to the loss of prairie dog colonies of all sizes (Roach et al. 2001, p. 956), are magnified by isolation of colonies. Colony growth after an epizootic is mainly the result of recolonization by intercolony dispersers (Antolin et al. 2002, p. 17). Wagner et al. (2006, pp. 334335) studied cycles of extirpation and recolonization in Gunnison's prairie dogs in Arizona, including a large number of colonies over a large geographic area, and found a significant relationship between the persistence of colonies and the persistence of their nearest neighboring colony. Increased isolation decreases the likelihood of recolonization following a plague outbreak if the distance between the infected colony and the next nearest colony is beyond the dispersal capabilities of the species. For example, Lechleitner et al. (1962, pp. 195, 197) documented a 1959 plague outbreak in a Gunnison's prairie dog colony in Colorado that killed all members of the colony. Prior to the outbreak, this colony had been continuously occupied for 20 years, despite several poisoning attempts. Two years after the plague outbreak, the colony still had not been recolonized, likely because it was isolated from other colonies by 8 mi (13 km) (Lechleitner et al. 1962, p. 187).

Research is underway on the efficacy of insecticides in protecting various prairie dog species from plague. Biggins and Godby (2005, p. 2) hypothesized that if enzootic plague is affecting populations of prairie dogs, an ambitious effort to remove the disease should result in increased survival rates of prairie dogs. Fleas in Utah prairie dog burrows were effectively controlled by annual treatments of the insecticide deltamethrin; fleas were reduced 96 to 98 percent within one month of treatment (Biggins and Godby 2005, p. 5). Studies of the effects of flea control on blacktailed and whitetailed prairie dogs have shown similar results (Biggins 2007). At this time, chemical dusting of individual prairie dog burrows is labor intensive and expensive.

All recent, major Gunnison's prairie dog colony declines documented in published literature have been attributed to plague epizootics. However, the magnitude of the plague threat appears to be different in the montane and prairie portions of the Gunnison's prairie dog range. Population declines in prairie habitat are less dramatic than those in montane habitat; partial recovery or establishment of new colonies have been documented following plague in the prairie range portion, but are rare or absent following plague outbreaks in the montane range.

We reviewed literature on the status of Gunnison's prairie dog populations within the two portions of the range and, specifically, all published and unpublished literature on the effects of plague on prairie dogs. While some studies were not recent, summarizing them below provides background on the responses of Gunnison's prairie dog populations to plague in each portion of the range.

Effects of Plague in Montane Habitat

Several wellstudied colonies within the montane portion of the Gunnison's prairie dog range have been documented as being extirpated, or nearly so, due to plague. The South Park, Colorado, population area included estimated occupied habitat of 915,000 ac (371,000 ha) in 1945; 74,000 ac (30,000 ha) in 1948; and 42 ac (17 ha) in 2002 (CDOW [[Page 6670]]
2007). This decline was largely due to plague and affected a substantial portion of the species' extant occupied habitat in Colorado (at least 15 percent). A plague event in Saguache County, Colorado, that progressed across seven colonies in 2 years left only scattered individuals surviving in two colonies (Lechleitner et al. 1968, p. 734). In Gunnison, Saguache, and Montrose Counties, Colorado, plague also was responsible for a decline from 15,569 ac (6,228 ha) of occupied habitat in 1980, to 770 ac (308 ha) in 2002 (note that Montrose County is in the Southwest population area in prairie habitat) (Capodice and Harrell 2003, pp. 57). A complete dieoff of a colony due to plague in Chubbs Park, Chaffee County, Colorado, occurred in 1959 (Lechleitner et al. 1962, p. 185). In August 1958, the population was stable and healthy, but in 1959 an epizootic spread 2 mi (3 km) within 3 months; prairie dogs continued to be absent from the area in 1960 and 1961, and we have no recent information on the existence of prairie dogs in that location. Plague resulted in the complete loss, over a 2year period, of a colony in South Park, Colorado (Fitzgerald 1970, pp. 6869).

Approximately 1,000 to 1,500 Gunnison's prairie dogs were killed by an outbreak of plague in a 148ac (60ha) colony in Curecanti National Recreation Area near Gunnison, Colorado, in 1981 (Rayor 1985, p. 194). A few animals survived the disease and Gunnison's prairie dogs were again abundant in the area in 1986 (Cully 1989, p. 49). In 2002, 252 ac (102 ha) of habitat in the Recreation Area were occupied by Gunnison's prairie dog colonies (Capodice and Harrell 2003, p. 23), but the current estimate is 12 ac (4.8 ha) (Childers 2007, p. 2). Colonies within the Recreation Area experienced six plague epidemics between 1971 and 2007. Of the 9 historic Gunnison's prairie dog colonies, 3 are currently active, and 2 act as source populations for the main prairie dog concentration area (Childers 2007, p. 1). If the source colonies die off due to plague, repopulation may not be possible because any other Gunnison's prairie dog populations remaining will be separated by distance (more than 6 mi (10 km)) and impassable geographical features such as rivers and mountains (Lomolino et al. 2003, p. 116; Pizzimenti and Hoffman 1973, p. 1).

Recently, plague has been implicated in the loss of several large colonies on BLM land within the Gunnison population area (CDOW 2007, p. 4). A large colony southeast of Gunnison, Colorado, that was very active in 2005, was totally devoid of prairie dogs in 2006 and 2007. Four other large colonies in the same vicinity were active in 2006, but by 2007, no prairie dog activity was observed. Plague is the suspected cause of these extirpations, because of the complete elimination of the prairie dogs with no sign of poisoning (CDOW 2007, p. 4).

Fitzgerald (1993, p. 52) expressed concern about the status of the Gunnison's prairie dog in Colorado, indicating that plague had eliminated many populations, including almost all of the populations in South Park. He also suggested that populations appeared to be in poor condition in the San Luis Valley, and were extirpated from the extreme upper Arkansas River Valley, as well as Jefferson, Douglas, and Lake Counties. These areas comprise most of the Gunnison's prairie dog montane habitat in Colorado.

During 1984 through 1987, a plague event reduced the population of Gunnison's prairie dogs in the Moreno Valley of New Mexico from more than 100,000 individuals to between 250 and 500, a decline of greater than 99 percent (Cully et al. 1997, pp. 708711). Although the growth rate of the remaining population increased following the epizootic, another plague event swept through the area in 1988, and the population in July 1996 was still ``a fraction'' of what it had been in 1984 (Cully et al. 1997, p. 718).

Occupancy modeling performed in Colorado in 2005 indicated a lower proportion of occupancy in the montane portion of the species' range within Colorado (3.2 percent) than in the prairie portion within Colorado (16.0 percent) (Andelt et al. 2006, p. 17; CDOW 2007, p. 19). When the study was repeated over the same plots in 2007, occupancy was again found to be lower (3.6 percent) in the montane range portion in Colorado than in the prairie portion (18.3 percent) (CDOW 2007, p. 19). The only recent threat responsible for whole population declines and extirpations, as documented in the studies cited in this section, is plague.

The frequency of plague epizootics appears to be high in montane habitat due to moister environmental conditions that are conducive to greater flea densities. The impact of plague epizootics in montane habitat is great because the small, isolated populations cannot recolonize. Within the South Park, Gunnison, and Southeast montane population areas in Colorado, no prairie dog complexes that approach a size considered sustainable exist, and only a few small complexes exist within the San Luis Valley population area (CDOW 2007, pp. 117). Without a metapopulation structure, an overall decline in persistence takes place (Lomolino and Smith 2001, p. 942).

The landscape status in the montane portion of Gunnison's prairie dog range is characterized by fewer, smaller colonies that are isolated, and few to no complexes or metapopulation structure. Isolation of populations is related to the montane geography in this portion of the range. Gunnison's prairie dogs occupy low valleys and mountain meadows within this habitat (Seglund et al. 2005, p. 12), likely because the short growing season at elevations higher than 10,000 ft (3,048 m) limits forage (Andelt et al. 2006, p. 17). In addition, mountain topography minimizes the zone of contact between populations (Knowles 2002, p. 3). At least four mountain ranges within the montane portion of the range act as barriers to Gunnison's prairie dog dispersal (Pizzimenti and Hoffman 1973, p. 1). These factors make the prairie dogs in this habitat highly susceptible to plaguerelated declines, and we have no evidence of longterm recovery from plague in the montane habitat area.

Effects of Plague in Prairie Habitat

The Southwest and the La PlataArchuleta populations in Colorado are within the prairie portion of Gunnison prairie dog range. The Southwest population comprises the largest population of Gunnison's prairie dogs in Colorado, with an estimated 88,267 ac (35,307 ha) of active colonies. Currently, prairie dogs can be found in nearly any habitat suitable for occupation, although densities are low to very low in native rangeland areas. Plague may be a problem in this area, because periodic dieoffs not associated with poisoning or other control measures have been noted by local farmers and ranchers in the past. However, unlike populations in montane habitat within Colorado, these populations appear to rebound from periodic epizootics (CDOW 2007, p. 16).

Populations in the La PlataArchuleta population area appear to undergo plague outbreaks every 4 to 7 years, which may be limiting some populations (CDOW 2007, p. 7). Occupancy modeling in 2005 and 2007 documented Gunnison's prairie dog occupancy of 17.6 percent and 27.0 percent, respectively, in the Southern Ute Reservation (part of the La PlataArchuleta population area), and 15.6 percent and 16.3 percent in the Southwest area (CDOW 2007, p. 19). The persistence of these populations, while undergoing repeated plague outbreaks, is likely due to their proximity to other populations within the prairie portion of the species' range
[[Page 6671]]

and immigration from those populations.

In Arizona, from 1987 to 2001, an estimated 68 percent reduction in the number of active Gunnison's prairie dog colonies occurred, primarily due to outbreaks of plague (Underwood 2007, p. 18; Wagner and Drickamer 2002, p. 15). However, in the area known as the Coconino Plateau, the area occupied by Gunnison's prairie dogs increased from 2,126 ac (860 ha) in 1992 to 40,942 ac (16,569 ha) in 2007 (Van Pelt 2007, p. 3), suggesting the species can withstand large plague epizootics through colony expansion or recolonization from nearby colonies. In addition, the Aubrey Valley Complex (in northwestern Arizona, the westernmost part of the species' range) has remained stable since at least 1974, despite the presence of plague, and the size of this complex increased from approximately 30,000 ac (12,000 ha) in 1997 (Underwood 2007, p. 23), to 40,000 ac (16,800 ha) in 2005 (Van Pelt 2005, p. 2), to 47,785 ac (19,338 ha) in 2007 (Van Pelt 2007, p. 2). Gunnison's prairie dogs at this site had significantly higher levels of antigens associated with diseasecausing pathogens such as plague, the same immune response expected if the prairie dogs had been vaccinated against plague (Wagner and Van Andel 2007, p. 2).

Of 293 colonies surveyed within Gunnison's prairie dog range in Arizona outside of Tribal lands, 57 (19 percent) experienced dieoffs during the summers of 2000 and 2001 (Wagner and Drickamer 2002, p. 13). Plague was confirmed as the causative agent for 15 of these 57 colonies but is thought to be the likely cause for them all, because it is the only disease that causes outbreaks with high mortality in prairie dogs (Barnes 1993, p. 34; Wagner and Drickamer 2002, p. 13). During surveys, they also identified the approximate boundaries of two previous plague outbreaks (Wagner and Drickamer 2002, p. 14).

An outbreak occurred over approximately 1,120 square mi (2,900 square km) west of the town of Dilkon, Arizona, on the Navajo Indian Reservation. This outbreak probably occurred in 1995 or 1996 (Wagner and Drickamer 2004, p. 14). Previous surveys in the area documented 45 colonies on 8,649 ac (3,500 ha). Reexamination of these colonies in 2000 and 2001 showed that all but two colonies were inactive. At most of the inactive colonies, burrow entrances were completely closed, and only mounds indicated where they formerly occurred.

An outbreak occurred east of the town of Seligman, Arizona, across approximately 425 square mi (1,100 square km) around 1996. The Arizona Game and Fish Department conducted surveys in this area between 1990 and 1994, and identified 47 active colonies that covered approximately 8,649 ac (3,500 ha). In 1996, dieoffs were observed in this area, and the U.S. Centers for Disease Control and Prevention confirmed plague as the cause. Although prairie dog numbers were increasing again in 1998, surveys in 2001 indicated that only 11 of the 47 colonies were active. Possibly another, undocumented, plague outbreak occurred in 1999 or 2000, again reducing the number of individuals (Underwood 2007, p. 19). Despite this persistent plague activity, Gunnison's prairie dogs are becoming reestablished in some areas within the boundaries of the Seligman outbreak (Wagner and Drickamer 2002, pp. 1415). This apparent resiliency is most likely due to immigration from other colonies in the prairie portion of the species' range.

Plague cycles have been observed in Gunnison's prairie dogs in Utah, and populations have been known to die off and then recover (Lupis et al. 2007, p. 32). Because plague testing has not been conducted on Gunnison's prairie dogs in Utah, declines cannot definitively be attributed to the disease (Seglund et al. 2005, p. 52). Plague is anticipated to be an ongoing threat to Gunnison's prairie dog populations in Utah at both a localized, and a wid

FOR FURTHER INFORMATION CONTACT Al Pfister, Field Supervisor, U.S. Fish and Wildlife Service, Western Colorado Field Office (see ADDRESSES). If you use a telecommunications device for the deaf (TDD), call the Federal Information Relay Service (FIRS) at 8008778339.