Endangered and Threatened Wildlife and Plants; 12-Month Finding on a Petition To List the Sonoran Population of the Desert Tortoise as Endangered or Threatened

FederalRegulations

Ask Donna

How this section applies to your facts.

Federal Register › Vol. 75 › 75 FR 78094

This text was captured on Aug 14, 2026. It is a snapshot, not a live feed, so check the official code before relying on it.

Text

DEPARTMENT OF THE INTERIOR Fish and Wildlife Service 50 CFR Part 17 [Docket No. FWS-R2-ES-2009-0032; MO 92210-0-008] Endangered and Threatened Wildlife and Plants; 12-Month Finding on a Petition To List the Sonoran Population of the Desert Tortoise as Endangered or Threatened AGENCY:

Fish and Wildlife Service, Interior.

ACTION:

Notice of 12-month petition finding.

SUMMARY:

We, the U.S. Fish and Wildlife Service, announce a 12-month finding on a petition to list the Sonoran population of the desert tortoise ( Gopherus agassizii ) as endangered or threatened and to designate critical habitat under the Endangered Species Act of 1973, as amended (Act). After review of all available scientific and commercial information, we find that listing the Sonoran population of the desert tortoise is warranted. Currently, however, listing the Sonoran population of the desert tortoise is precluded by higher priority actions to amend the Lists of Endangered and Threatened Wildlife and Plants. Upon publication of this 12-month petition finding, we will add the Sonoran population of the desert tortoise to our candidate species list. We will develop a proposed rule to list the Sonoran population of the desert tortoise as our priorities allow. We will make any determination on critical habitat during development of the proposed listing rule. In any interim period we will address the status of the candidate taxon through our annual Candidate Notice of Review (CNOR).

DATES:

The finding announced in this document was made on December 14, 2010.

ADDRESSES:

This finding is available on the Internet at http://www.regulations.gov at Docket Number FWS-R2-ES-2009-0032. Supporting documentation we used in preparing this finding is available for public inspection, by appointment, during normal business hours at the U.S. Fish and Wildlife Service, Arizona Ecological Services Office, 2321 West Royal Palm Road, Suite 103, Phoenix, Arizona 85021

ADDRESSES:

FOR FURTHER INFORMATION CONTACT:

Steven L. Spangle, Field Supervisor Arizona Ecological Services Office (see ADDRESSES ); by telephone at (602) 242-0210; or by facsimile at (602) 242-2513. If you use a telecommunications device for the deaf (TDD), please call the Federal Information Relay Service (FIRS) at 800-877-8339.

SUPPLEMENTARY INFORMATION:

Background

Section 4(b)(3)(B) of the Endangered Species Act of 1973, as amended (Act) (16 U.S.C. 1531 et seq. ), requires that, for any petition to revise the Federal Lists of Endangered and Threatened Wildlife and Plants that contains substantial scientific or commercial information that listing the species may be warranted, we make a finding within 12 months of the date of receipt of the petition. In this finding, we determine that the petitioned action is: (a) Not warranted, (b) warranted, or (c) warranted, but the immediate proposal of a regulation implementing the petitioned action is precluded by other pending proposals to determine whether species are threatened or endangered, and expeditious progress is being made to add or remove qualified species from the Federal 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 12-month findings in the Federal Register

d species from the Federal 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 12-month findings in the Federal Register.

Previous Federal Actions

On October 15, 2008, we received a petition dated October 9, 2008, from WildEarth Guardians and Western Watersheds Project (petitioners) requesting that the Sonoran population of the desert tortoise be listed under the Act as a distinct population segment (DPS), as threatened or endangered rangewide (in the United States and Mexico), and critical habitat be designated. The petition contained detailed information on the natural history, biology, current status, and distribution of the Sonoran population of the desert tortoise. It also contained information on what the petitioners reported as potential threats to the Sonoran population of the desert tortoise, such as livestock grazing, urbanization and development, mining, international border patrol activities, illegal collection, inadequacy of existing regulations, altered fire regimes, off-highway vehicle use, drought, and climate change. We acknowledged the receipt of the petition in a letter to the WildEarth Guardians and Western Watersheds Project, dated November 26, 2008. In that letter we also stated that we had reviewed the information presented in the petition and determined that issuing an emergency regulation temporarily listing the species as per section 4(b)(7) of the Act was not warranted. We also stated that we intended to make our finding on whether the petition presented substantial information that the requested action may be warranted, to the maximum extent practicable within 90 days of receipt of the petition, according to the provisions of section 4(b)(3) of the Act

ng an emergency regulation temporarily listing the species as per section 4(b)(7) of the Act was not warranted. We also stated that we intended to make our finding on whether the petition presented substantial information that the requested action may be warranted, to the maximum extent practicable within 90 days of receipt of the petition, according to the provisions of section 4(b)(3) of the Act.

On August 28, 2009, we made our 90-day finding that the petition presented substantial scientific information indicating that listing the Sonoran population of the desert tortoise ( Gopherus agassizii ) may be warranted. The finding and notice of our initiation of a status review was published in the Federal Register on August 28, 2009 (74 FR 44335).

On April 10, 2010, a stipulated settlement agreement ( WildEarth Guardians and Western Watersheds Project v. Salazar, 10-cv-86-ACT-RHS (D. NM)) was filed. In this agreement, we agreed to submit a 12-month finding to the Federal Register on or before December 5, 2010. The stipulated settlement agreement was signed and adopted by the District Court of New Mexico on April 15, 2010.

This notice constitutes our 12-month finding for the petition to list the Sonoran population of the desert tortoise as threatened or endangered.

Other Federal Actions

Throughout this finding, we use “Mojave” to describe desert tortoise populations north and west of the Colorado River, as well as any reference to the biotic community known as the “Mojave Desert” or “Mojave desertscrub.” These uses are consistent with the previous and current spelling of the common name in Federal actions that have addressed this population. We use “Mohave” in the geographic context to remain consistent with its reference by the U.S. Board of Geographic Names ( e.g., Mohave County)

olorado River, as well as any reference to the biotic community known as the “Mojave Desert” or “Mojave desertscrub.” These uses are consistent with the previous and current spelling of the common name in Federal actions that have addressed this population. We use “Mohave” in the geographic context to remain consistent with its reference by the U.S. Board of Geographic Names ( e.g., Mohave County). In addition, while the Sonoran population of the desert tortoise is not currently formally recognized as a unique taxonomic entity, for ease of reference, we refer to the Sonoran population of the desert tortoise as the “Sonoran desert tortoise” in this finding.

On December 30, 1982, we published a notice of review which determined the desert tortoise throughout its range in the United States and Mexico to be a Category 2 Candidate species (47 FR

On April 2, 1990, we issued a final rule designating the Mojave population of the desert tortoise (occurring north and west of the Colorado River) as a threatened species under the Act (55 FR 12178; see final rule for a summary of previous actions regarding the Mojave population of the desert tortoise). Currently, the Mojave population of the desert tortoise is recognized as a distinct population segment under the Act. As part of that rulemaking, we designated any desert tortoise from the Sonoran population as threatened when observed outside of its known range, due to similarity of appearance under section 4(a) of the Act.

On December 5, 1996, we published a rule that discontinued the practice of keeping a list of Category 2 Candidate species (61 FR 64481). Since that time, the Sonoran desert tortoise observed inside its known range has had no Federal Endangered Species Act status

noran population as threatened when observed outside of its known range, due to similarity of appearance under section 4(a) of the Act.

For a detailed account of previous Federal actions that pertained to the desert tortoise in the United States, please review the following Federal Register documents: “Proposed Endangered Status and Critical Habitat for the Beaver Dam Slope Population of the Desert Tortoise” (43 FR 37662, August 23, 1978); “Requirement to withdraw or supplement proposals to determine various U.S. taxa of plants and wildlife as Endangered or Threatened or to determine Critical Habitat for such species” (44 FR 12382, March 6, 1979); “Reproposal of Critical Habitat for the Illinois mud turtle and Beaver Dam Slope population of the desert tortoise” (44 FR 70680, December 7, 1979); “Listing as Threatened With Critical Habitat for the Beaver Dam Slope Population of the Desert Tortoise in Utah” (45 FR 55654, August 20, 1980); “Review of Vertebrate Wildlife for Listing as Endangered or Threatened Species” (47 FR 58454, December 30, 1982); “Notice of Findings on Four Petitions, and Review of One Species” (50 FR 13054, April 2, 1985); “Review of Vertebrate Wildlife” (50 FR 37958, September 15, 1985); “Finding on Desert Tortoise Petition” (50 FR 49868, December 5, 1985); “Findings on Pending Petitions and Description of Progress of Listing Actions” (53 FR 25511, July 7, 1988); “Findings on Pending Petitions and Description of Progress of Listing Actions” (53 FR 52746, December 29, 1988); “Emergency Determination of Endangered Status for the Mojave Population of the Desert Tortoise” (54 FR 32326, August 4, 1989); “Desert Tortoise” (54 FR 42270, October 13, 1989); “Determination of Threatened Status for the Mojave Population of the Desert Tortois

FR 25511, July 7, 1988); “Findings on Pending Petitions and Description of Progress of Listing Actions” (53 FR 52746, December 29, 1988); “Emergency Determination of Endangered Status for the Mojave Population of the Desert Tortoise” (54 FR 32326, August 4, 1989); “Desert Tortoise” (54 FR 42270, October 13, 1989); “Determination of Threatened Status for the Mojave Population of the Desert Tortoise” (55 FR 12178, April 2, 1990); “Finding on a Petition to List the Sonoran Desert Tortoise as Threatened or Endangered” (56 FR 29453, June 27, 1991); “Proposed Determination of Critical Habitat for the Mojave Population of the Desert Tortoise” (58 FR 45748, August 30, 1993); “Determination of Critical Habitat for the Mojave Population of the Desert Tortoise” (59 FR 5820, February 8, 1994); “Determination of Critical Habitat for the Mojave Population of the Desert Tortoise” (59 FR 9032, February 24, 1994); “Notice of Final Decision on Identification of Candidates for Listing as Endangered or Threatened” (61 FR 64481, December 5, 1996); and “90-Day Finding on a Petition To List the Sonoran Population of the Desert Tortoise ( Gopherus agassizii ) as a Distinct Population Segment (DPS) with Critical Habitat” (74 FR 44335, August 28, 2009).

Species Information

Taxonomy

The desert tortoise is in the genus Gopherus (Rafinesque 1832), or gopher tortoises, and is a member of the Testudinidae family, or terrestrial tortoises. The North American tortoises formerly comprised two genera, Gopherus and Xerobates, with the latter including X. agassizii, the desert tortoise (Crumly 1994, pp. 7-8). Scientific nomenclature assigned to the desert tortoise has undergone a series of changes since its initial description by Cooper (1863) as X. agassizii (Barrett and Johnson 1990, p. 5); the currently recognized scientific name for the desert tortoise is Gopherus agassizii. Further information is available in Barrett and Johnson (1990, p

ng X. agassizii, the desert tortoise (Crumly 1994, pp. 7-8). Scientific nomenclature assigned to the desert tortoise has undergone a series of changes since its initial description by Cooper (1863) as X. agassizii (Barrett and Johnson 1990, p. 5); the currently recognized scientific name for the desert tortoise is Gopherus agassizii. Further information is available in Barrett and Johnson (1990, p. 5) or in the detailed account of desert tortoise phylogeny (evolutionary development) and systematics (taxonomic classification) by Crumly (1994, pp. 7-32). The desert tortoise is known in Mexico with the common names of “tortuga del monte,” “Galápago de desierto,” or the “xtamóosni” (Rorabaugh 2008, p. 35).

Physical Description of Sonoran Desert Tortoises

Adult Sonoran desert tortoises range in total carapace (straight-line top shell) length from 8 to 15 inches (in) (20 to 38 centimeters (cm)), with a relatively high domed shell (AGFD 2001, p. 1; Brennan and Holycross 2006, p. 54). The record length for a Sonoran desert tortoise is 19.4 in (49 cm) total carapace length (Jackson and Wilkinson-Trotter 1980, p. 430). The carapace is usually brownish with a definite pattern and prominent growth lines (AGFD 2001, p. 1). The plastron (bottom shell) is yellowish and is not hinged (AGFD 2001, p. 1; Brennan and Holycross 2006, p. 54). The hind limbs are very stocky and elephantine; forelimbs are flattened for digging and covered with large conical scales (AGFD 2001, p. 1; Brennan and Holycross 2006, p. 54). Male Sonoran desert tortoises are differentiated from females by having elongated gular (throat) shields, chin glands visible on each side of the lower jaw (most evident during the breeding season), a concave plastron, and larger overall size (AGFD 2001, p. 1)

lephantine; forelimbs are flattened for digging and covered with large conical scales (AGFD 2001, p. 1; Brennan and Holycross 2006, p. 54). Male Sonoran desert tortoises are differentiated from females by having elongated gular (throat) shields, chin glands visible on each side of the lower jaw (most evident during the breeding season), a concave plastron, and larger overall size (AGFD 2001, p. 1).

Distribution

The desert tortoise includes portions of southern California, southern Nevada, southwestern Utah, and the western, northwestern, and southern portions of Arizona in the United States, and also includes the Mexican State of Sonora into the northern portion of Sinaloa. One-third of the geographic range of the desert tortoise occurs in northwestern Mexico (Bury et al. 2002, p. 86). The specific distribution of desert tortoise is influenced by habitat and climatic characteristics (vegetation community for food), soil and substrate characteristics (for shelter), and precipitation pattern (for water availability) within the appropriate elevation range.

The distribution of the Sonoran desert tortoise in the United States is considered to be entirely within Arizona and comprises approximately 26.8 million acres (ac) (10.8 million hectares (ha)); east and south of the Colorado River (Barrett and Johnson 1990, pp. 4-5; Lamb et al. 1989, p. 84). Sonoran desert tortoise distribution in Arizona is limited to the northeast by the limits of the Sonoran Desert. The Arizona portion of their range constitutes approximately 52 percent of their total distribution

rizona and comprises approximately 26.8 million acres (ac) (10.8 million hectares (ha)); east and south of the Colorado River (Barrett and Johnson 1990, pp. 4-5; Lamb et al. 1989, p. 84). Sonoran desert tortoise distribution in Arizona is limited to the northeast by the limits of the Sonoran Desert. The Arizona portion of their range constitutes approximately 52 percent of their total distribution. In Arizona, the Sonoran desert tortoise occurs primarily on Federal land but also occurs on a variety of non-federal lands as well as on ten Native American reservations: (1) Fort Mojave Indian Tribe; (2) Colorado River Indian Tribe; (3) Hualapai Tribe; (4) Fort McDowell Yavapai Nation; (5) Salt River Pima-Maricopa Indian Community; (6) Gila River Indian Community; (7) Ak Chin; (8) Tohono O'odham Nation; (9) Pasqua Yaqui Tribe; and, (10) San Carlos Apache Tribe (AIDTT 2000, p. 40).

In Mexico, where 48 percent of their range occurs, the distribution of the Sonoran desert tortoise extends from the international border of Sonora and Arizona, south to the vicinity of Guaymas, and north of the Río Yaqui (the southern and southeastern-most border of their distribution), in southern Sonora (Germano et al. 1994, p. 77; Fritts and Jennings 1994, p. 51; Bury et al. 2002, p. 88; Van Devender 2002a, p. 5; Edwards et al. 2009, pp. 7-8). This includes approximately the western half of the State of Sonora from the Gulf of California coast east roughly to the transition to unsuitable woodland and conifer forest areas in the higher elevations of the Sierra Madre Occidental. In 30 timed searches conducted August to September 1983, and beyond the known distribution of Sonoran desert tortoises in Sonora, Mexico, Fritts and Jennings (1994, p. 52) found several patterns in Sonoran desert tortoise distribution. First, most Sonoran desert tortoises in the eastern and northern extent of their distribution in Mexico occur below the 2,600 foot (ft) (790 meters (m)) elevation contour (Fritts and Jennings 1994, p. 52)

September 1983, and beyond the known distribution of Sonoran desert tortoises in Sonora, Mexico, Fritts and Jennings (1994, p. 52) found several patterns in Sonoran desert tortoise distribution. First, most Sonoran desert tortoises in the eastern and northern extent of their distribution in Mexico occur below the 2,600 foot (ft) (790 meters (m)) elevation contour (Fritts and Jennings 1994, p. 52). Second, populations may be the densest and the least patchy between elevations of 900 and 1,600 ft (270 and 490 m) (Fritts and Jennings 1994, p. 52). They were also not found in habitat in Mexico that received an average of 3.9 in (10 cm) or less of rain annually (Fritts and Jennings 1994, p. 53).

One question about the distribution of the Sonoran desert tortoise concerns the origin of a small number of tortoises that have been found in far southeastern Cochise County, Arizona, an area generally considered well east of the known distribution. There is some evidence that these tortoises may represent a naturally occurring population based on the presence of suitable habitat (Rorabaugh 2009, pers. comm.), similar animal communities (Rosen 2009, pers. comm.), and historic and current observations of tortoises in the area (Hulse and Middendorf 1979, p. 546; Radke 2009, pers. comm.; Van Devender et al. 1976, pp. 300-303). However, these observations have traditionally been discounted as released pets rather than a natural population (AIDTT 2000, p. 3; Germano et al. 1994, p. 81). Also, recent genetic analysis of a Sonoran desert tortoise collected from this area in 2009 indicated it was most closely related to tortoises in the Phoenix, Arizona, area and is likely, therefore, a “released or escaped captive” tortoise (Edwards 2010, pers. comm.). We recognize there is a fair amount of uncertainty regarding the origin of this population

2000, p. 3; Germano et al. 1994, p. 81). Also, recent genetic analysis of a Sonoran desert tortoise collected from this area in 2009 indicated it was most closely related to tortoises in the Phoenix, Arizona, area and is likely, therefore, a “released or escaped captive” tortoise (Edwards 2010, pers. comm.). We recognize there is a fair amount of uncertainty regarding the origin of this population. However, because Sonoran desert tortoises are infrequently documented from this area and recent genetic testing indicated that observations represent released captives, we conclude that desert tortoises from this area do not represent a naturally-occurring, disjunct population. Consequently, we will not evaluate potential threats to the tortoises in this area of Cochise County in this finding.

Habitat

Sonoran desert tortoises are most closely associated with the Arizona Upland and Lower Colorado River subdivisions of Sonoran desertscrub and Mojave desertscrub vegetation types. They occur most commonly on rocky (predominantly granitic rock), steep slopes and bajadas (lower mountain slopes often formed by the coalescing of several alluvial fans (fan-shaped deposits at the ends of canyons formed when fast flowing streams slow and widen)) and in paloverde-mixed cacti associations (Ortenburger and Ortenburger 1927, p. 120; Burge 1979, p. 49; 1980, p. 48). Sonoran desert tortoise density has been observed to be higher in the Arizona Upland subdivision of the Sonoran desertscrub than in the Lower Colorado subdivision of the Sonoran desertscrub or in Mojave desertscrub (Berry 1984, p. 434; AIDTT 2000, p. 4; Boarman and Kristan 2008, p. 19). In addition to the use of vegetation to meet energy and nutritional needs, the Sonoran desert tortoise uses vegetation for predator avoidance, thermal protection, and in social behaviors (Avery and Neibergs 1997, p. 13; Grandmaison et al. in press, p. 3)

Colorado subdivision of the Sonoran desertscrub or in Mojave desertscrub (Berry 1984, p. 434; AIDTT 2000, p. 4; Boarman and Kristan 2008, p. 19). In addition to the use of vegetation to meet energy and nutritional needs, the Sonoran desert tortoise uses vegetation for predator avoidance, thermal protection, and in social behaviors (Avery and Neibergs 1997, p. 13; Grandmaison et al. in press, p. 3). An important attribute of Sonoran desert tortoise habitat is the presence of cryptogamic crusts (soil crusts with unique, microscopic association of flora and fauna) (Bowker et al. 2008, p. 2309). These occur on the surface of Sonoran Desert soils and assist with nitrogen-fixing to enhance soil fertility, improve water infiltration into soils, and prevent or lessen effects from wind and water erosion, all of which help to sustain vegetation vital to the Sonoran desert tortoise (DeFalco 1995, p. 22; DeFalco et al. 2001, pp. 1, 9).

Sonoran desert tortoises rarely occur in oak woodland habitat. However, one such population occurs at approximately 5,000-ft (1,500-m) elevation in Chiminea Canyon in the Rincon Mountains of Pima County, Arizona (Van Devender 2002a, p. 23), and they are also known from similar elevation in the Atascosa and Pajarito Mountains in south-central Arizona. Zylstra and Steidl (2008, p. 747) found that habitat selection by Sonoran desert tortoises was most closely associated with topographic (degree of steepness of slope) and geomorphologic (rock type and structure) influences rather than by vegetation type. Specifically, Zylstra and Steidl (2008, p. 747) found that the likelihood of observing Sonoran desert tortoises increased with increasing slope, with a strong association to aspect (the direction to which a slope faces), with east-facing slopes preferred over north-facing slopes. However, the season of use may affect which slope-aspects (the direction a particular slope faces) Sonoran desert tortoises are likely to use based on their needs at that time (Zylstra and Steidl 2008, p. 752)

oran desert tortoises increased with increasing slope, with a strong association to aspect (the direction to which a slope faces), with east-facing slopes preferred over north-facing slopes. However, the season of use may affect which slope-aspects (the direction a particular slope faces) Sonoran desert tortoises are likely to use based on their needs at that time (Zylstra and Steidl 2008, p. 752). Specifically, Sonoran desert tortoises have different thermoregulatory and physiological needs based upon their seasonal behaviors, such as hibernation or seeking temporary shelter during the tortoise's surface-active seasons.

In addition to steep, rocky slopes and bajadas, Sonoran desert tortoises also use inter-mountain valleys as part of their home ranges and for dispersal at all age classes (Averill-Murray and Averill-Murray 2002, p. 16). In the Ironwood National Forest, Averill-Murray and Averill-Murray (2005, p. 65) found tortoises or their signs (such as scat (droppings) and burrows) on 92 percent of transects in boulder habitat, on 71 percent of transects that included incised washes (dry stream beds that flow in response to precipitation), and on 25 percent of transects that had neither boulder habitat nor incised washes. Sonoran desert tortoises were found up to one mile (mi) (1.6 kilometers (km)) away from the nearest slope, indicating that they occur in low densities in inter-mountain valleys. Averill-Murray and Averill-Murray (2005, p. 65) stated that maintaining these areas “may be important for long-term population viability.” Washes might also be selectively chosen by reproductive female Sonoran desert tortoises as all eggs and hatchling desert tortoises observed by Barrett (1990, p. 205) occurred there

st slope, indicating that they occur in low densities in inter-mountain valleys. Averill-Murray and Averill-Murray (2005, p. 65) stated that maintaining these areas “may be important for long-term population viability.” Washes might also be selectively chosen by reproductive female Sonoran desert tortoises as all eggs and hatchling desert tortoises observed by Barrett (1990, p. 205) occurred there. Sonoran desert tortoises on the 40-square-mile (sq mi) (64-square-kilometer (sq km)) Florence Military Reservation in Pinal County, Arizona, primarily use xeroriparian habitat (a habitat association with plant species tolerant to hyper-arid conditions) along washes, with caliche caves (caves formed along steep banks of washes within cemented, sedimentary rock formations of calcium carbonate) within washes being an important component to occupied habitat (Lutz et al. 2005, p. 22; Riedle et al. 2008, p. 418). Another frequently Larrea tridentata ) and white bursage ( Ambrosia dumosa ) during all periods of the year; somewhat atypical for Sonoran desert tortoises in other portions of its range (Lutz et al. 2005; p. 22; Grandmaison et al. in press, p. 4). In this habitat, Sonoran desert tortoises often used packrat middens (organic debris piles constructed for nesting purposes which often are comprised of wood material, cactus pads, etc.) as shelter sites, especially those with suitable canopy cover, an absence of cattle activity, and proximity to roads and washes (Lutz et al. 2005, p. 22; Grandmaison et al. in press, p. 2).

et al. in press, p. 2).

Sonoran desert tortoises in Arizona generally occur within elevations from 510 to 5,300 ft (155 to 1,615 m) (Barrett and Johnson 1990, p. 7; AGFD 2001, p. 4). According to the AGFD's Heritage Data Management system, 95 percent of Sonoran desert tortoise observations in Arizona have occurred at an elevation of 904 to 4,198 ft (275 to 1279 m) (Zylstra and Steidl 2009, p. 8). However, one example of an extreme exception was a Sonoran desert tortoise observed at 7,808 ft (2,379 m) in a ponderosa pine-dominated coniferous community in the Rincon Mountain District of Saguaro National Park in Pima County, Arizona (Aslan et al. 2003, p. 57). The nearest road was 8.6 mi (13.9 km) away by trail and nearly 2,000 ft (610 m) lower in elevation from the observed location of the tortoise, which strongly dismisses any notion that human activity was responsible for its location at such a high elevation (Aslan et al. 2003, p. 57).

Sonoran desert tortoises in Mexico are generally found at lower elevations, ranging from approximately 1,000 to 1,640 ft (305 to 500 m) in elevation in rocky outcrops in desertscrub and foothills thornscrub habitat (Bury et al. 2002, p. 89). As in Sonoran desertscrub habitat in Arizona, Sonoran desert tortoises in Mexico often use shrubs as temporary shelter sites, and species such as mesquite ( Prosopis spp.) and ironwood ( Olneya tesota ) may play important roles in the natural history of Sonoran desert tortoises in Mexico (Bury et al. 2002, p. 100). Sonoran desert tortoises in Mexico have not been documented in flatter areas between mountain ranges (Bury et al. 2002, p. 89), although we presume they use these areas to some extent for dispersal much like they do in similar inter-mountain basins of Arizona. With the exception of the El Pinacate Desert Bioreserve in northwestern Sonora, Sonoran desert tortoises have not been documented using the extremely arid Lower Colorado subdivision of the Sonoran Desert in Mexico (Bury et al. 2002, p. 89)

(Bury et al. 2002, p. 89), although we presume they use these areas to some extent for dispersal much like they do in similar inter-mountain basins of Arizona. With the exception of the El Pinacate Desert Bioreserve in northwestern Sonora, Sonoran desert tortoises have not been documented using the extremely arid Lower Colorado subdivision of the Sonoran Desert in Mexico (Bury et al. 2002, p. 89). However, based on their presence in El Pinacate and the general lack of surveys in Mexico, the Sonoran desert tortoise may potentially be found in this habitat in northwestern Sonora in low densities. The extent of Sonoran desert tortoise distribution in northeastern Sonora, an area characterized as a transitional zone of foothills thornscrub, tropical deciduous forest, and Madrean oak woodland, is poorly understood (Bury et al. 2002, p. 89).

Burrow Use

Adequate shelter, often in the form of constructed burrows, is one of the most important habitat features for the Sonoran desert tortoise. Burrows are constructed under rocks and boulders, beneath vegetation, on semi-open slopes, within the sidewalls of washes, or by using rocky crevices which may or may not be altered by the tortoise (Burge 1979, p. 44; 1980, pp. 44-45; Barrett 1990, p. 205; Averill-Murray et al. 2002a, pp. 136-137, Grandmaison et al. in press, p. 14). Sonoran desert tortoises construct burrows in a variety of soil types including silt, silt with loose gravel, diatomite (a light-colored porous rock composed of the shells of diatoms) and diatomaceous marl (a crumbly mixture of clays, calcium and magnesium carbonates, with remnants of shells), and well-lithified (process whereby loose particles are converted into rock) volcanic ash, as observed in the lower San Pedro River Valley of Arizona (Bailey et al. 1995, pp. 363-364)

g silt, silt with loose gravel, diatomite (a light-colored porous rock composed of the shells of diatoms) and diatomaceous marl (a crumbly mixture of clays, calcium and magnesium carbonates, with remnants of shells), and well-lithified (process whereby loose particles are converted into rock) volcanic ash, as observed in the lower San Pedro River Valley of Arizona (Bailey et al. 1995, pp. 363-364). Burrows are used for thermoregulation, nesting, and protection from predators, and the lack of suitable conditions for constructing burrows may be a limiting factor in Sonoran desert tortoise populations (Barrett and Humphrey 1986, p. 262; Bailey et al. 1995, p. 366; Zylstra and Steidl 2008, p. 752). In fact, Sonoran desert tortoise population densities appear to be highly correlated with available burrows, or potential burrow sites (Averill-Murray and Klug 2000, p. 69; Averill-Murray et al. 2002b, p. 126). Sonoran desert tortoises often use a group of relatively closely-located burrows as focal areas of activity in their home range. In doing so, they establish circular or slightly linear movement patterns, and may temporarily move on to another such cluster of burrows within the same active season (Bulova 1994, p. 140; Averill-Murray and Klug 2000, p. 62; Lutz et al. 2005, p. 21).

Burrows influence a variety of Sonoran desert tortoise behaviors and physiological characteristics. During the winter dormancy period (colder, winter months of inactivity), female Sonoran desert tortoises typically use more shallow burrows that are more susceptible to variation in ambient temperatures and consequently females emerge earlier in the spring (as early as late February) than do males, who often remain dormant until the commencement of the summer monsoon (AIDTT 2000, p. 7; Ernst and Lovich 2009, p. 547). Averill-Murray and Klug (2000, p. 66) and Bailey et al. (1995, p

an desert tortoises typically use more shallow burrows that are more susceptible to variation in ambient temperatures and consequently females emerge earlier in the spring (as early as late February) than do males, who often remain dormant until the commencement of the summer monsoon (AIDTT 2000, p. 7; Ernst and Lovich 2009, p. 547). Averill-Murray and Klug (2000, p. 66) and Bailey et al. (1995, p. 367) suggest that shallow burrows may account for responsiveness of females to warming periods in early spring for additional foraging opportunities to increase energy reserves for egg development, as shallower burrows are more reflective of ground-surface temperatures. Alternatively, cool, less variable temperatures in deeper burrows selected by male Sonoran desert tortoises may enhance sperm development and viability, as cooler temperatures allow more sperm production (Bailey et al. 1995, p. 367).

The season may influence the locations and dimensions of burrows used by Sonoran desert tortoises in order to meet their behavioral and physiological needs (Barrett 1990, p. 205; Bailey et al. 1995, pp. 363, 366). Finally, particularly in hatchling and juvenile size classes, the burrow microclimate can affect the rate of water loss in desert tortoises, which results in behaviors (drinking pooled rain, withdrawing into their shell, seeking long, deep burrows) to avoid lethal dehydration in relatively hot, dry seasons (Wilson et al. 2001, p. 158; Bulova 2002, pp. 184-186).

Other forms of shelter used by Sonoran desert tortoise include packrat middens, which are often shared with other native reptiles, including other tortoises (Averill-Murray et al. 2002a, pp. 136-137; Lutz et al. 2005, p. 22; Grandmaison et al. in press, p. 2). These shelter types provide less insulation than earthen burrows and are therefore used for shorter duration, especially during the months with extremely hot or cold temperatures. This was the most commonly used shelter site at Florence Military Reservation

ve reptiles, including other tortoises (Averill-Murray et al. 2002a, pp. 136-137; Lutz et al. 2005, p. 22; Grandmaison et al. in press, p. 2). These shelter types provide less insulation than earthen burrows and are therefore used for shorter duration, especially during the months with extremely hot or cold temperatures. This was the most commonly used shelter site at Florence Military Reservation.

Seasonal Behavior and Long-Distance Movements

The Sonoran desert tortoise is diurnal (active during daylight hours) but sometimes emerge at night in response to rainfall (Ernst and Lovich 2009, p. 544). Sonoran desert tortoises may be surface-active every month of the year, but in the winter, surface activity is likely a response to thermoregulatory needs or movements between burrows (Averill-Murray and Klug 2000, p. 66). et al. 2010, p. 11). Precipitation amounts and timing vary among the populations of desert tortoise. The lowest amount of rainfall (usually during the winter) occurs in the northwestern-most portion of the species' range, and gradually increases and becomes seasonally bimodal pattern (rains in winter and summer) to the south into the southern-most extent of the species range in northern Sinaloa, Mexico (Germano et al. 1994, p. 76). Sonoran desert tortoise surface activity largely mimics the warm-season precipitation pattern (Averill-Murray et al. 2002a, p. 139; Van Devender 2002a, p. 7). Like the Arizona populations, Sonoran desert tortoises in Mexico seem to be most active in late summer (Ernst and Lovich 2009, p. 544). Sonoran desert tortoises are approximately half as active during the spring as they are in the summer, with females typically becoming surface active to forage in late March, while males typically emerge (but are not necessarily active) in late April (Averill-Murray et al. 2002a, p. 138)

ions, Sonoran desert tortoises in Mexico seem to be most active in late summer (Ernst and Lovich 2009, p. 544). Sonoran desert tortoises are approximately half as active during the spring as they are in the summer, with females typically becoming surface active to forage in late March, while males typically emerge (but are not necessarily active) in late April (Averill-Murray et al. 2002a, p. 138).

The summer monsoon (occurring typically from late June through September), characterized by both excessive heat and frequent thunderstorms, is the peak activity season for the Sonoran desert tortoise (Averill-Murray et al. 2002a, pp. 139-140). During this period, new growth of perennial plants is initiated and annual plants germinate, providing forage for tortoises (Averill-Murray et al. 2002a, p. 140). The onset of the summer monsoon triggers Sonoran desert tortoises to drink, flush their bladders, and rehydrate, establishing a positive water and energy balance, and spurring reproductive behaviors (AIDTT 2000, p. 7). Sonoran desert tortoises have been observed to seek out rocks with surface depressions during summer months to drink puddled water from monsoon storm events (Oftedal 2007, p. 23). Surface activity in Sonoran desert tortoises begins to wane as early as late September and ends by mid-December as they prepare for hibernation. Temperature and photoperiod (the duration of daylight) are likely the cues used by Sonoran desert tortoises to commence hibernation (Bailey et al. 1995, p. 367; Averill-Murray et al. 2002a, p. 147). Periods of hibernation (typically from mid-November through mid-February) in Sonoran desert tortoises appear to vary greatly among populations and among years but appear to correlate with seasonal temperatures (Bailey et al. 1995, p. 367; Averill-Murray and Klug 2000, p. 66).

The behavior and ecology of hatchling Sonoran desert tortoises is poorly understood because their small size makes them very difficult to observe in the wild

om mid-November through mid-February) in Sonoran desert tortoises appear to vary greatly among populations and among years but appear to correlate with seasonal temperatures (Bailey et al. 1995, p. 367; Averill-Murray and Klug 2000, p. 66).

The behavior and ecology of hatchling Sonoran desert tortoises is poorly understood because their small size makes them very difficult to observe in the wild. Their scat is small, inconspicuous, and ephemeral, and burrows used by individuals in this size class resemble those of other terrestrial vertebrates in Sonoran desert tortoise habitat (Germano et al. 2002, pp. 271-272). This size class is thought to be the most vulnerable, experiencing the highest mortality rates (Morafka 1994, p. 161).

Home range sizes of Sonoran desert tortoises vary with precipitation levels, contracting during wet years and expanding during dry years in response to the availability of forage plants (Averill-Murray and Klug 2000, p. 67). The home range of Sonoran desert tortoises may be as small as 6.4 ac (2.6 ha) but can vary widely, with males having larger home ranges than females (Barrett 1990, p. 203; Averill-Murray and Klug 2000, pp. 55-61; Averill-Murray et al. 2002a, pp. 150-151). In the lower San Pedro River Valley, Meyer (1993, p. 99) found Sonoran desert tortoise home ranges varied between 45 and 640 ac (18 and 258 ha) in size. Sonoran desert tortoises are known to exhibit high fidelity to their home ranges, with exception to dispersal movements when they move to new areas (Zylstra and Swann 2009, p. vi). They likely habituate to specific attributes of their home range, including the location of mates, water catchments, mineral licks, and burrow sites (Berry 1986a, p. 113).

Sonoran desert tortoises are known to make long-distance movements between populations in adjacent mountain ranges. In an extreme example, Edwards et al. (2004, p

ts when they move to new areas (Zylstra and Swann 2009, p. vi). They likely habituate to specific attributes of their home range, including the location of mates, water catchments, mineral licks, and burrow sites (Berry 1986a, p. 113).

Sonoran desert tortoises are known to make long-distance movements between populations in adjacent mountain ranges. In an extreme example, Edwards et al. (2004, p. 494) tracked an adult female Sonoran desert tortoise moving 20 mi (32 km) between the Rincon and Santa Rita mountains of southern Arizona (also see Zylstra and Swann 2009, p. 10). During this long-distance movement, this tortoise encountered several barriers to movement that required human intervention to overcome such as fence lines, railroad tracks, an interstate highway, and several captures (including a temporary adoption) by humans (Edwards et al. 2004, p. 494). In another example, in the San Pedro Valley of southern Arizona, a sub-adult Sonoran desert tortoise was captured and marked in 1992. It was recaptured in 2005 approximately 14 mi (23 km) from its original point of capture (Meyer et al. 2010, p. 18). Dispersal distances of hatchling Sonoran desert tortoises are not well understood, but are likely shorter than those of adults because of the complex habitat of boulders and vegetation (where they occur) may inhibit long-distance movements (Van Devender 2002a, p. 14).

Gibbons (1986, p. 104) suspected that long-distance movements by turtles can be explained by: (1) Nest site selection; (2) seasonal migration; (3) departure from unfavorable habitat conditions; or (4) movement by males in search of females. Averill-Murray and Klug (2000, p. 68) suggested that long-distance movements may be interpreted as random wanderings, infrequent travels to known sources of biological needs, explorations, adaptations for genetic exchange, or for dispersal to other suitable areas

ite selection; (2) seasonal migration; (3) departure from unfavorable habitat conditions; or (4) movement by males in search of females. Averill-Murray and Klug (2000, p. 68) suggested that long-distance movements may be interpreted as random wanderings, infrequent travels to known sources of biological needs, explorations, adaptations for genetic exchange, or for dispersal to other suitable areas. Precipitation may influence the likelihood of long-distance movements, especially in individuals approaching reproductive age in populations that experience above-average precipitation for a 2- to 3-year period (AIDTT 2000, p. 8). Averill-Murray and Klug (2000, p. ii) stated, “A large cohort of young tortoises that experiences a relatively wet and productive environment, with high survival, may provide the stock for dispersal between populations as they approach sexual maturity, in addition to replacing aging adults within the local population.” Long-distance movements by Sonoran desert tortoises observed by Averill-Murray and Klug (2000, p. 69) suggest the potential for metapopulation (interrelated population dynamics between regionally proximal populations) relationships between local populations inhabiting regional hillsides. Habitat features may also influence the Sonoran desert tortoises' ability to make long-distance movements. Dispersal of Sonoran desert tortoises between populations might be less likely through sparse desertscrub in very hot, dry river valleys in the Lower Colorado River subdivision of Sonoran desertscrub. Van Devender (2002a, p. 16) suggested that populations occurring in the Eagletail, Maricopa, Sand Tank, and similarly situated mountain ranges might have existed in isolation for decades, if not centuries.

ail, Maricopa, Sand Tank, and similarly situated mountain ranges might have existed in isolation for decades, if not centuries.

There are no data to evaluate long-distance movements in populations that occur in Mexico. Although Sonoran desert tortoises in Mexico are known to occupy slopes, arroyos, and bajadas, they are infrequently observed using valley bottoms (Fritts and Jennings 1994, p. 52). Sonoran desert tortoise populations in Mexico have been poorly studied, but we presume individuals make similar long-distance movements between populations.

Longevity

Estimates of longevity in wild Sonoran desert tortoises vary considerably from 30 years to over 100 years (Germano 1992, pp. 369-370; 1994, p. 176; Zylstra and Swann 2009, p. vii). Using a growth equation to extrapolate longevity in Sonoran desert tortoises, Germano et al. (2002, p. 271) estimated that the average oldest ages attained for Sonoran desert tortoises is 62.2 years in females and 64.4 years in males; however, they admitted that correlating age with size is problematic in turtles. Zylstra and Swann (2009, p. vii) suspected that Sonoran desert tortoises may reach 80 to 100 years of age in the wild. Sonoran desert tortoises have been shown to live longer in the wild than those from the Mojave population.

Bladder Physiology

The bladder in the Sonoran desert tortoise is unique and serves an important function in its survival. Sonoran desert tortoises are capable of drinking large amounts of water when it is available, and may even construct water catchments by digging earthen depressions, likely as an adaptation to the infrequent and unpredictable nature of rainfall events throughout their range (Ernst and Lovich 2009, p. 546)

er in the Sonoran desert tortoise is unique and serves an important function in its survival. Sonoran desert tortoises are capable of drinking large amounts of water when it is available, and may even construct water catchments by digging earthen depressions, likely as an adaptation to the infrequent and unpredictable nature of rainfall events throughout their range (Ernst and Lovich 2009, p. 546). The bladder of Sonoran desert tortoises is a large and bilobed (divided into two lobes) organ critical for withstanding the effects of seasonal and short-term drought because of its ability to store water, dilute excess dietary salts and metabolic wastes, and reabsorb water into the bloodstream (Averill-Murray et al. 2002a, p. 146; Ernst and Lovich 2009, p. 545). In seasonal or short-term drought conditions, the concentration of urine in Sonoran desert tortoises allows them to forage on dried vegetation by reducing the dehydration effects of such forage types (Averill-Murray et al. 2002a, p. 146; Ernst and Lovich 2009, p. 545). Water serves an important role in flushing salts from the body of Sonoran desert tortoises and resetting the electrolytic balance, preparing the Sonoran desert tortoise for the next dry period (Averill-Murray et al. 2002a, pp. 140, 146).

Diet, Foraging Behavior, and Potassium Excretion Potential

The Sonoran desert tortoise is an herbivore, and has been documented to eat 199 different species of plants, including herbs (55.3 percent), grasses (17.6 percent), woody plants (22.1 percent), and succulents (5 percent) (Ogden 1993, pp. 1-8; Van Devender et al. 2002; pp. 175-176; Brennan and Holycross 2006, p. 54; Oftedal 2007, p. 21; Ernst and Lovich 2009, p. 562; Meyer et al. 2010, pp. 28-29, 44-48)

onoran desert tortoise is an herbivore, and has been documented to eat 199 different species of plants, including herbs (55.3 percent), grasses (17.6 percent), woody plants (22.1 percent), and succulents (5 percent) (Ogden 1993, pp. 1-8; Van Devender et al. 2002; pp. 175-176; Brennan and Holycross 2006, p. 54; Oftedal 2007, p. 21; Ernst and Lovich 2009, p. 562; Meyer et al. 2010, pp. 28-29, 44-48). Of the numerous nonnative plant species that have become established throughout the range of the Sonoran desert tortoise, only red brome ( Bromus rubens ) and redstem filaree ( Erodium cicutarium ) are frequently eaten and considered relatively important nonnative species in the diets of Sonoran desert tortoises (Van Devender et al. 2002, p. 183). However, physical injury to Mojave desert tortoises resulting from consuming nonnative grass species (i.e., red brome and cheatgrass ( Bromus tectorum )) has been documented, and sharp seeds have been found lodged between the tortoises' upper and lower jaw. This injury may adversely affect their foraging ability or become a source for infection (Medica and Eckert 2007, p. 447). Though this study focused on Mojave desert tortoises, this may affect all desert tortoises wherever these plant species occur (i.e., within the Sonoran Desert in Arizona).

Significant differences in the nutritional quality of native versus nonnative forage for desert tortoises were not found by Hazard et al. (2010, pp. 139-145). Nagy et al. (1998, pp. 260, 263) compared the nutritional values of native and nonnative grasses (native: Indian ricegrass ( Achnatherum ( Oryzopsis ) hymenoides ); nonnative: Mediterrean grass ( Schismus barbatus )) and forbs (native: desert dandilion ( Malacothrix glabrata ); nonnative: redstem filaree), finding that the two grasses possessed similar nutritional value. The dry matter and energy digestibility of the two grasses were much lower than those of the forbs, providing little nitrogen, and tortoises lost more water than they gained while processing grasses

ve: Mediterrean grass ( Schismus barbatus )) and forbs (native: desert dandilion ( Malacothrix glabrata ); nonnative: redstem filaree), finding that the two grasses possessed similar nutritional value. The dry matter and energy digestibility of the two grasses were much lower than those of the forbs, providing little nitrogen, and tortoises lost more water than they gained while processing grasses. The native forb was more readily digestible than the nonnative forb as dried mass, but the inverse was true as fresh mass (Nagy et al. 1998, p. 263). However, the native forbs provide significantly more nitrogen and water than the nonnative forbs, which is important in maintaining a positive water balance. Results of these feeding trials suggest that the proliferation of nonnative grasses leading to the exclusion of forbs places desert tortoises at a nutritional disadvantage. If, instead of eating to obtain a given volume of food, tortoises consume just enough food to satisfy their energy needs (as commonly noted in other vertebrate groups), then the native forbs provide the best nutrition. Nagy et al. (1998, p. 260) concluded that the life stage of the plant and the plant type (forb or grass) were important predictors of nutritional quality versus a plant being native or nonnative to a particular region. In summary, research has shown that forbs are more valuable to Sonoran desert tortoise nutrition than grasses, and that native forbs are more valuable than nonnative forbs in a dried state, which may be important in periods of drought.

Diets of Sonoran desert tortoises vary among populations in response to seasonal availability of plant species and in response to precipitation amounts (Martin and Van Devender 2002, p. 31). In years of low winter rainfall, Sonoran desert tortoises are less selective in plant species consumed because there are fewer options, but in years of high winter rainfall, Sonoran desert tortoises have exhibited highly selective foraging habits (Oftedal 2002, pp. 205-206)

nse to seasonal availability of plant species and in response to precipitation amounts (Martin and Van Devender 2002, p. 31). In years of low winter rainfall, Sonoran desert tortoises are less selective in plant species consumed because there are fewer options, but in years of high winter rainfall, Sonoran desert tortoises have exhibited highly selective foraging habits (Oftedal 2002, pp. 205-206). During years when monsoon rains are light or irregular, Sonoran desert tortoises consume dried plant material (Averill-Murray et al. 2002a, p. 140). Within Saguaro National Park in southern Arizona, Sonoran desert tortoises frequently ate annual legumes in the spring (high in water content, low in potassium), and annual and perennial grasses (supplemented by prickly pear fruit ( Opuntia engelmannii )) during the monsoon when ponding water can replenish water reserves (Oftedal 2007, p. 17). In most years, Sonoran desert tortoises consume enough calories during the summer monsoon to fuel growth and store fat for the next year (Van Devender 2002a, p. 10).

Desert tortoises are uniquely vulnerable to changes in their potassium levels (Oftedal 2002, p. 208). Because potassium cannot be easily stored in the body, excess potassium must be excreted to avoid toxicological effects (Oftedal 2002, p. 208). Therefore, Sonoran desert tortoises that must forage on plants with high potassium content must also flush their bladders more frequently and therefore risk a net loss in metabolic water levels and subsequent dehydration (Oftedal 2002, p. 209).

The potassium excretion potential (PEP) is an index of water, nitrogen, and potassium levels in a plant that relates to a desert tortoise's ability to efficiently excrete potassium. PEP is a critical consideration for determining the value or risk of particular forage species during times of drought or major perturbations to habitat, and for comparing potential effects of forage competition between tortoises and livestock

otential (PEP) is an index of water, nitrogen, and potassium levels in a plant that relates to a desert tortoise's ability to efficiently excrete potassium. PEP is a critical consideration for determining the value or risk of particular forage species during times of drought or major perturbations to habitat, and for comparing potential effects of forage competition between tortoises and livestock. A positive PEP value for a plant species (preferred by tortoises) means there is more water and nitrogen

In addition to herbivory, Sonoran desert tortoises are also geophagous; in other words, they consume bones, stones, and soil for additional nutrient and mineral supplements, for mechanical assistance in grinding plant matter in the stomach, or to expel parasites in the intestinal tract (Sokol 1971, p. 70; Marlow and Tollestrup 1982, p. 475; Esque and Peters 1994, pp. 108-109; Stitt and Davis 2003, p. 57; Walde et al. 2007b, p. 148). Sonoran desert tortoises are highly attracted to sites with exposed calcium carbonate and have been observed congregating at these sites year after year eating these soils (Meyer et al. 2010, p. 11). Soil condition and quality are important to the Sonoran desert tortoise, not only for nutrients derived from eating soil, but also production and maintenance of vegetation that is consumed by tortoises (Avery and Neibergs 1997, p. 13).

Desert tortoises have been observed eating scat from black-tailed jack rabbits, wood rats, collared peccaries, and even desert tortoises. This behavior could possibly aid in the transfer of gut microflora such as bacteria or fungi or it could be used as a source of additional nutrients (Walde et al. 2005, p. 77-78). Bostick (1990, p. 149) asserted that desert tortoises feed “primarily on dung” although this claim was refuted in the literature (Boarman 2002, pp. 27, 35, 38). Infrequent observations of sand, bird feathers, arthropod parts, and snake and lizard skins have also been made during fecal analyses of desert tortoises (Ernst and Lovich 2009, p

urce of additional nutrients (Walde et al. 2005, p. 77-78). Bostick (1990, p. 149) asserted that desert tortoises feed “primarily on dung” although this claim was refuted in the literature (Boarman 2002, pp. 27, 35, 38). Infrequent observations of sand, bird feathers, arthropod parts, and snake and lizard skins have also been made during fecal analyses of desert tortoises (Ernst and Lovich 2009, p. 560).

Reproduction

The Sonoran desert tortoise breeding season begins with the summer monsoon when male-male combat over receptive females can be observed, such as at sites with exposed calcium carbonate soils, where tortoise densities may be higher (discussed above) (Meyer et al. 2010, p. 11). Sexual maturity and first reproduction in female Sonoran desert tortoises occurs from 12 to 22 years of age, or at 8.7 in (22 cm) in midline carapace length, and reproductive activity is highly influenced by winter and spring precipitation (Averill-Murray and Klug 2000, p. 69; Averill-Murray et al. 2002b, p. 119; Bury et al. 2002, p. 100; Germano et al. 2002, p. 265). Females may store sperm for up to two years, meaning that one season's mating produces the following season's clutch of eggs (Palmer et al. 1998, pp. 704-705; Averill-Murray et al. 2002a, p. 141). Female Sonoran desert tortoises may lay one clutch of 1-12 eggs per year, usually around the onset of the summer rainy season, although they may not produce a clutch every year (Averill-Murray 2002b, p. 295). Eggs hatch in September and October (Van Devender 2002a, pp. 10-11; Averill-Murray 2002b, p. 295). The average clutch size is 3.8 to 5.7 eggs, and in contrast to Mojave Desert tortoises, clutch size is not positively correlated with female body size (Mueller et al. 1998, p. 313; Averill-Murray 2002b, p. 299; Averill-Murray et al. 2002b, p. 119). Late oviposition (deposition of eggs) dates recorded on the Sugarloaf study site in central Arizona in 1998 and 1999 suggest that eggs and hatchlings may occasionally overwinter in nests (Averill-Murray 2002b, p. 299)

ave Desert tortoises, clutch size is not positively correlated with female body size (Mueller et al. 1998, p. 313; Averill-Murray 2002b, p. 299; Averill-Murray et al. 2002b, p. 119). Late oviposition (deposition of eggs) dates recorded on the Sugarloaf study site in central Arizona in 1998 and 1999 suggest that eggs and hatchlings may occasionally overwinter in nests (Averill-Murray 2002b, p. 299). Female desert tortoises have been known to urinate on their nest sites before and after nesting; this may be to aid in digging the nest, and might make it more difficult to dig up the nest after the soil dries, or possibly to hydrate soils in contact with eggs as the rigid-shelled eggs of desert tortoises have been shown to uptake moisture from the soil faster than it evaporates from the shell exposed to air (Patterson 1971, p. 199; Spotila et al. 1994, p. 112). Female Sonoran desert tortoises that survive to reproductive age are believed to produce as many as 85 eggs over the course of their lives, with perhaps two or three of those hatchlings surviving to reproductive age (Van Devender 2002a, p. 11).

Desert tortoises exhibit environmental sex determination, which means that incubation temperatures during embryonic development determine the sex of the tortoises. Higher incubation temperatures produce more females and lower temperatures produce more males (Spotila et al. 1994, pp. 109-111; Rostal et al. 2002, p. 313). Incubation temperatures at or below 86.9 degrees Fahrenheit (° F) (30.5 degrees Celsius (° C)) result in the production of all male desert tortoises, whereas temperatures of 90.5 °F (32.5 °C) result in all females, and eggs incubated at the “pivotal” temperature of 88.3 °F (31.3 ° C) develop a 1:1 sex ratio (Rostal et al. 2002, p. 313).

Predation

As adults, Sonoran desert tortoises are relatively protected from natural predation because of their hard shells

(30.5 degrees Celsius (° C)) result in the production of all male desert tortoises, whereas temperatures of 90.5 °F (32.5 °C) result in all females, and eggs incubated at the “pivotal” temperature of 88.3 °F (31.3 ° C) develop a 1:1 sex ratio (Rostal et al. 2002, p. 313).

Predation

As adults, Sonoran desert tortoises are relatively protected from natural predation because of their hard shells. Mountain lions ( Felis concolor ) appear to be the only natural predator in the Sonoran Desert with the jaw strength required to puncture or crack the shells of adult Sonoran desert tortoises. However, mountain lion predation is not known to contribute to elevated mortality rates within monitored Sonoran desert tortoise populations (AIDTT 2000, p. 8; Meyer et al. 2010, p. 18; Riedle et al. 2010, p. 165). Dickenson et al. (2001, p. 254) recorded 14 Sonoran desert tortoise mortalities in the Little Shipp Wash and Harcuvar monitoring plots from 1990-1994, five of which were attributed to mountain lion predation. Javelina ( Tayassu tajacu ) predation on Sonoran desert tortoises was suspected in the San Pedro Valley of southern Arizona (Meyer et al. 2010, p. 18). Other mammalian predators may include badgers ( Taxidea taxus ), ring-tailed cats ( Bassiriscus astutus ), bobcats ( Felis rufus ), skunks ( Spilogale gracilis, Mephitis mephitis, M. macroura, Conepatus mesoleucus ), kit foxes ( Vulpes macrotis ), gray foxes ( Urocyon cinereoargenteus ), coyotes ( Canis latrans ), and domestic dogs ( Canis familiaris ) (Boarman 2002, p. 17; Ernst and Lovich 2009, p. 563).

on cinereoargenteus ), coyotes ( Canis latrans ), and domestic dogs ( Canis familiaris ) (Boarman 2002, p. 17; Ernst and Lovich 2009, p. 563).

Both golden eagles ( Aquila chrysaetos ) and common ravens ( Corvus corvax ) have been documented to prey upon all size classes of Mojave desert tortoises in California (Berry 1985, pp. 1, 6-10). Such predation might also occur on Sonoran desert tortoises. The greater roadrunner ( Geococcyx californianus ) is also a suspected predator on juvenile Mojave desert tortoises, based upon one field observation of roadrunner tracks next to a freshly killed individual (Berry 1985, p. 11); such predation might also occur on Sonoran desert tortoises. However, because avian predators rely exclusively on their vision to detect prey, we expect lower rates of avian predation on Sonoran desert tortoises occupying Arizona upland Sonoran desertscrub because the dense, complex habitat structure likely limits birds' ability to detect tortoises. Habitat-associated protection from avian predation may be less pronounced where Sonoran desert tortoises occur in the sparser vegetation of the Lower Colorado River subdivision of Sonoran desertscrub.

Sonoran desert tortoises are most vulnerable to predation while in their eggs or as hatchlings and juveniles predominantly because of their size and undeveloped, softened shells (which do not adequately harden until approximately 7 years of age) which provide little protection and are easily compromised. Higher mortality rates in the hatchling and juvenile age classes may also be partially due to their higher metabolic rates, which necessitates longer periods of surface activity to obtain suitable amounts of forage. Longer surface activity may cause greater risk of detection by predators (Morafka 1994, p. 163). Nest predation levels may be high in some populations. Seventy-five percent of Sonoran desert tortoise nests suffered predation over a two-year period at the Sugarloaf study plot in Maricopa County, Arizona (Averill-Murray 2002b, p. 298)

periods of surface activity to obtain suitable amounts of forage. Longer surface activity may cause greater risk of detection by predators (Morafka 1994, p. 163). Nest predation levels may be high in some populations. Seventy-five percent of Sonoran desert tortoise nests suffered predation over a two-year period at the Sugarloaf study plot in Maricopa County, Arizona (Averill-Murray 2002b, p. 298). Gila monsters ( Heloderma suspectum ) are a primary predator on tortoise eggs, and female Sonoran desert tortoises in the process of oviposition will actively defend the burrow and aggressively pursue Gila monsters in attempting to drive them away (Barrett and Humphrey 1986, p. 262). Coachwhips ( Coluber flagellum ) and gophersnakes ( Pituophis catenifer ) have been reported consuming juvenile Sonoran desert tortoises (Amarello et al. 2004, p. 178; Ernst and Lovich 2009, p. 563). Presumably, other snake species such as common kingsnakes ( Lampropeltis getula ) with generalized prey preferences consume eggs or hatchling Sonoran desert tortoises, but we did not find other examples in the literature.

For more detailed information on all aspects of Sonoran desert tortoise biology, see Barrett and Johnson (1990, pp. 1-95) and Bury and Germano (1994, pp. 1-212).

Monitoring and Population Status

Monitoring and Statistical Analyses

We are unaware of any structured, long-term monitoring program for Sonoran desert tortoises in Mexico; therefore, we are unable to assess the current status or population trends in that part of the range. Therefore, we discuss only Arizona studies in this section.

Twenty-eight individual Sonoran desert tortoise populations in Arizona have been studied since the mid-1970s but few populations have been studied for more than a few years (Averill-Murray 2000, p. 1; Averill-Murray et al. 2002b, p. 109). Monitoring plots (also referred to as “plots”) have varied from 0.2 to 1.5 sq mi (0.3 to 2.4 sq km) in size (Averill-Murray 2000, p. 4). Beginning in 1987, AGFD and the U.S

t individual Sonoran desert tortoise populations in Arizona have been studied since the mid-1970s but few populations have been studied for more than a few years (Averill-Murray 2000, p. 1; Averill-Murray et al. 2002b, p. 109). Monitoring plots (also referred to as “plots”) have varied from 0.2 to 1.5 sq mi (0.3 to 2.4 sq km) in size (Averill-Murray 2000, p. 4). Beginning in 1987, AGFD and the U.S. Bureau of Land Management (BLM) have established and maintained 17 plots in Arizona as long-term monitoring plots and have surveyed them in a somewhat irregular, but repeated fashion. Each plot has been surveyed between two and nine times during this timeframe, with 11 to 86 person-days (cumulative days spent by researchers working on plots) spent during each survey (AGFD 2010, p. 1). These long-term monitoring plots are located in six counties within Arizona, and their locations were chosen to represent Sonoran desert tortoise distribution in the State.

General monitoring objectives for the 17 plots are to document abundance, density, and changes of Sonoran desert tortoise populations across the State using capture-recapture methods (Averill-Murray 2000, p. 3). Records of demographic characteristics of each population, including sex ratios and age/size structure as well as individual health and signs of disease within each population were also recorded during monitoring activities (Averill-Murray 2000, p. 3). Monitoring protocols used from 1987 to 2000 are summarized in Averill-Murray (2000, pp. 3-7).

ing protocols used from 1987 to 2000 are summarized in Averill-Murray (2000, pp. 3-7).

The Sonoran desert tortoise is a difficult species to monitor in the wild because of its slow movement and camouflaged appearance, especially in the smaller hatchling and juvenile age classes. These factors can significantly hamper a surveyor's ability to detect them in the field (Zylstra et al. 2010, p. 1311). In addition, Arizona Upland subdivision of Sonoran desertscrub (where Sonoran desert tortoise population densities are the highest) is complex, often with many large boulders, somewhat dense vegetation, and challenging topographic relief. Drought and emigration also affect the reliability of data from Sonoran desert tortoise population monitoring because the tortoises may be inactive (in their burrows) or have left the population (dispersed). In these cases the absence of observations might be mistaken as mortality. Also, Sonoran desert tortoises can occur in low densities with little surface activity both seasonally and daily (Zylstra et al. 2010, p. 1311). Alone or in combination, these factors, in addition to a relatively short sampling period for such a long-lived species, make subtle population trends difficult to distinguish and overall population trend analysis problematic.

Low detectability may have been responsible for long periods between recaptures of marked desert tortoises in an 18-year desert tortoise study from 1980 to 1997 in the San Pedro Valley, Arizona. For example, a sub-adult Sonoran desert tortoise was captured and marked in 1992, and was not encountered again until 2005, when it was incidentally observed approximately 14 mi (22.5 km) from its original point of capture, 8 years after the conclusion of the study (Meyer et al. 2010, p. 18)

ed desert tortoises in an 18-year desert tortoise study from 1980 to 1997 in the San Pedro Valley, Arizona. For example, a sub-adult Sonoran desert tortoise was captured and marked in 1992, and was not encountered again until 2005, when it was incidentally observed approximately 14 mi (22.5 km) from its original point of capture, 8 years after the conclusion of the study (Meyer et al. 2010, p. 18). Within the entire duration of this study, approximately 30 percent of 577 marked Sonoran desert tortoises were never recaptured, with only 15 total carcasses found, indicating potential emigration, long-term burrow use, or difficulties in detecting individuals in complex landscapes (Meyer et al. 2010, p. 20). The amount of time between recaptures of Sonoran desert tortoises can be significant; durations between recaptures of some individuals in the San Pedro Valley study were as high as 18 years (Meyer et al. 2010, p. 20).

Several authors have investigated how detectability may bias results of Mojave desert tortoise monitoring. For example, Anderson et al. (2001, p. 583) studied the degree to which field observers can meet the assumptions underlying line-transect sampling to monitor populations of desert tortoises in Mojave desertscrub. They found that when all Mojave desert tortoises are not detected along the centerline of the transect route (which routinely occurs), biases in sampling data result (Anderson et al. 2001, p. 583). Anderson et al. (2001, p. 593-596) noted that surveyor numbers and level of experience contribute to the reliability of line transect methods. Freilich and LaRue (1998, p. 594) experimentally tested the effect of personnel experience on Mojave desert tortoise survey outcomes in Mojave desertscrub. They found that observers consistently overestimated the number of desert tortoise burrows (falsely assigning other animal burrows as those made by desert tortoises), and found fewer desert tortoises and scat than were actually placed on test plots

lich and LaRue (1998, p. 594) experimentally tested the effect of personnel experience on Mojave desert tortoise survey outcomes in Mojave desertscrub. They found that observers consistently overestimated the number of desert tortoise burrows (falsely assigning other animal burrows as those made by desert tortoises), and found fewer desert tortoises and scat than were actually placed on test plots. Their results indicated that experience played a relatively small role in detecting Mojave desert tortoises (Freilich and LaRue 1998, pp. 593-594). In an effort to increase detections, some investigators have tested the use of tortoise detection dogs in Mojave desert tortoise monitoring projects (Cablk and Heaton 2006, p. 1926; Heaton et al. 2008, pp. 476-477; Nussear et al. 2008, pp. 109-111). Because Sonoran desertscrub is more dense and complex than Mojave desertscrub, detection is even more difficult in Sonoran desert tortoise monitoring. Zylstra and Steidl (2009, p. 16) found that line transect methods are

The seasonal timing of surveys and fluctuating influence of precipitation on Sonoran desert tortoise surface activity also create problems with monitoring populations and interpreting results. Sonoran desert tortoises often become inactive, residing in their burrows, during periods of seasonal or short-term drought. For example, in a multi-year mark and recapture study of Mojave desert tortoises in Joshua Tree National Park, Freilich et al. (2000, pp. 1487-1488) found that in years of below-normal precipitation, desert tortoise home ranges decreased, individual captures decreased, and the effort required to find each tortoise nearly doubled; indicating the significant influence of precipitation on the possible discrepancy between the number of tortoises that can be observed versus the number of tortoises that actually occur within a monitoring plot.

In an attempt to improve monitoring protocols to account for such complicating factors described above, Averill-Murray (2000, pp

the effort required to find each tortoise nearly doubled; indicating the significant influence of precipitation on the possible discrepancy between the number of tortoises that can be observed versus the number of tortoises that actually occur within a monitoring plot.

In an attempt to improve monitoring protocols to account for such complicating factors described above, Averill-Murray (2000, pp. 7-13) critiqued the original protocols used for long-term monitoring plots of Sonoran desert tortoise populations in Arizona. This work became the basis for several changes in monitoring protocols, beginning in 2000. Although line transect methods have not been implemented on Arizona's Sonoran desert tortoise long-term monitoring plots, the capture-recapture methods currently used likely violate assumptions about equal detection probability (all animals having the same probability of being captured during every sampling occasion) (Zylstra and Steidl 2009, p. 9).

While monitoring of Sonoran desert tortoise populations in Arizona has been ongoing for several decades, attempts to quantify temporal trends in abundance have been hampered by the data limitations discussed above (Zylstra and Steidl 2009, p. 5; Zylstra et al. 2010, pp. 1311-1317). Effective monitoring is largely dictated by the objective of the monitoring, whether that objective is to detect changes in distribution, abundance, density, or survival. In addition, using existing plot data to establish rangewide trends in Sonoran desert tortoise populations is generally problematic because the current set of monitoring plots does not represent a random sample from the species' entire range in Arizona (Averill-Murray and Klug 2000, p. 25). Despite the history and effort dedicated to monitoring Sonoran desert tortoise populations in Arizona since 1987, there are limitations of these data with respect to interpreting rangewide trends of the Sonoran desert tortoise. Averill-Murray (2000, pp

current set of monitoring plots does not represent a random sample from the species' entire range in Arizona (Averill-Murray and Klug 2000, p. 25). Despite the history and effort dedicated to monitoring Sonoran desert tortoise populations in Arizona since 1987, there are limitations of these data with respect to interpreting rangewide trends of the Sonoran desert tortoise. Averill-Murray (2000, pp. 12-13) identified problems with extrapolating the results of the plot monitoring data to making range-wide assessments outside of the plots. We elaborate on these problems in our assessment of Boarman and Kristan (2008) below.

Boarman and Kristan (2008, pp. 3-12) analyzed mark and recapture data from the 17 Sonoran desert tortoise long-term monitoring plots throughout Arizona that were surveyed on the average of once every 4 years from 1987 to 2006. Boarman and Kristan (2008, p. ii) concluded that the Sonoran population of the desert tortoise in Arizona experienced statistically significant declines, at an annual rate of 3.52 percent over the 20-year period; equating to a cumulative 51 percent decline in overall numbers during this timeframe.

We received several comments from the public in response to our 90-day finding that addressed the Boarman and Kristan (2008) report (AGFD 2010, pp. 4-6; Carothers et al. 2010, pp. 5, 8-12; Ogden 2009, pp. 3-12, Smith 2010, pp. 4-5). Commenters criticized the method and manner with which Boarman and Kristan (2008) used statistical tests, as well as the conclusions they made. Significant concerns were noted with respect to the type of statistical tests used by Boarman and Kristan (2008) because data were extrapolated beyond the statistical tests' ability to avoid inherent biases (AGFD 2010, p. 4). Problems associated with the statistical confidence intervals for monitoring plot data used by Boarman and Kristan (2008) were also identified (Ogden 2009, pp. 2-3). Also, monitoring plot data used in Boarman and Kristan (2008, p

e of statistical tests used by Boarman and Kristan (2008) because data were extrapolated beyond the statistical tests' ability to avoid inherent biases (AGFD 2010, p. 4). Problems associated with the statistical confidence intervals for monitoring plot data used by Boarman and Kristan (2008) were also identified (Ogden 2009, pp. 2-3). Also, monitoring plot data used in Boarman and Kristan (2008, p. 20) were not designed to compare population trends among individual plots (Ogden 2009, p. 2). Carothers et al. (2010, pp. 8-12) identified numerous additional problems with the statistical analysis provided by Boarman and Kristan (2008). Collectively, based upon comments received from the public as well as our internal review, the number and magnitude of potential problems associated with Boarman and Kristan's (2008) statistical analysis call into question the validity of their conclusions. After careful review of the report and the questions raised by reviewers of the report, we decided that the conclusions pertaining to overall Sonoran desert tortoise population trends do not represent the best available information and, therefore, we did not use the report in this finding. However, other information in the Boarman and Kristan (2008) report was used in our analysis of the status of and threats to the Sonoran desert tortoise and is cited in this finding. For a more detailed analysis of the Boarman and Kristan (2008) report, see our “Review of Boarman and Kristan (2008)” provided at http://www.regulations.gov (Docket Number FWS-R2-ES-2009-0032).

Survivorship and Population Densities in Arizona

Viable populations in turtles usually require that both juvenile and adult size classes have high survivorship (Averill-Murray and Klug 2000, p. 70). Data on the recruitment of juveniles into Sonoran desert tortoise populations, and their survivorship, are generally lacking due to the difficulty detecting juveniles in the field (AGFD 2010, p. 3). Data on juvenile and adult survivorship in Sonoran desert tortoises require long-term, repeated population monitoring, which in turn, requires long-term, reliable funding sources. Consequently, these data are conspicuously rare or absent for most Sonoran desert tortoise monitoring plots making population viability estimates for Sonoran desert tortoise populations within Arizona problematic at best. As expected for a long-lived species, survivorship in Sonoran desert tortoises (using data generated from a few long-term monitoring plots in Arizona) is generally high for adults but potentially lower for juveniles and hatchlings (Zylstra and Steidl 2009, p. 7). Where enough data from long-term monitoring plots or independent studies exist, survivorship has been calculated for adults in the following plots or study areas: Sugarloaf Mountain (96-98 percent), Florence Military Reservation (88-97 percent), Little Shipp Wash (94-97 percent), Granite Hills (94-97 percent), and Eagletail Mountains (94-97 percent) (AGFD 2010, p. 2; Riedle et al. 2010, p. 165).

Densities of Sonoran desert tortoises among populations vary considerably. In 2000, the density of Sonoran desert tortoises, as determined by surveys on long-term monitoring plots and other monitoring plots during the 1990s, varied from 15 to 150 individuals per square mile (2.6 sq km) (AIDTT 2000, pp. 5-6; Averill-Murray and Klug 2000, p. i). In the San Pedro Valley of southern Arizona, the average density of the Sonoran desert tortoise population was 38 individuals per square mile (Meyer et al. 2010, p. 17). Stager et al. (2010, p

by surveys on long-term monitoring plots and other monitoring plots during the 1990s, varied from 15 to 150 individuals per square mile (2.6 sq km) (AIDTT 2000, pp. 5-6; Averill-Murray and Klug 2000, p. i). In the San Pedro Valley of southern Arizona, the average density of the Sonoran desert tortoise population was 38 individuals per square mile (Meyer et al. 2010, p. 17). Stager et al. (2010, p. 37) suspect that Sonoran desert tortoise populations in Mohave County, Arizona may be naturally lower due to limited burrowing habitat available to them to survive cold winters and hot summers.

Periodic, Localized Declines in Arizona Populations

There are no records of actual extirpations of Sonoran desert tortoises from any of the monitored populations. However, periodic, localized, and sometimes substantial declines have been documented in at least five of 17 monitored populations (Hart et al. 1992, p. 60; Averill-Murray et al. 2002b, p. 124; AGFD 2010, p. 4). Because of their life history, Sonoran desert tortoise populations may be slow to rebound from declines (Howland and Rorabaugh 2002, p. 340). The AGFD (2010, p. 4) suggested that observed declines in certain plots demonstrate localized, stochastic events and are not indicative of population trends as a whole across the distribution of the Sonoran desert tortoise. Sonoran desert tortoise populations are particularly vulnerable to elevated mortality of adults. Sustaining the adult, reproductive age class within Sonoran desert tortoise populations is important because mortality rates of juveniles are high and because it takes a long time for a Sonoran desert tortoise to reach sexual maturity (Howland and Rorabaugh 2002, p. 339). The relatively higher visibility of adult Sonoran desert tortoises leaves them more vulnerable to human impacts like collecting or shooting, and their tendency to move longer distances make them more susceptible to road mortality (Howland and Rorabaugh 2002, p. 340)

re high and because it takes a long time for a Sonoran desert tortoise to reach sexual maturity (Howland and Rorabaugh 2002, p. 339). The relatively higher visibility of adult Sonoran desert tortoises leaves them more vulnerable to human impacts like collecting or shooting, and their tendency to move longer distances make them more susceptible to road mortality (Howland and Rorabaugh 2002, p. 340).

The largest population decline noted at any Sonoran desert tortoise monitoring plot was observed on the Maricopa Mountains plot, where substantially more tortoise carcasses were found than live tortoises in successive years from 1987 through 1991 (Hart et al. 1992, p. 54; Averill-Murray et al. 2002b, p. 124). Regional drought from 1984-1992 was a suspected cause of the die-off of Sonoran desert tortoises in the Maricopa Mountains (Hart et al. 1992, p. 60; Averill-Murray et al. 2002b, p. 124). However, in 1987, the estimated density of Sonoran desert tortoises on the Maricopa Mountains plot was uncharacteristically high at 146 tortoises per square mile (2.6 sq km), suggesting that the population may have been in the process of naturally correcting to carrying capacity (the state at which a population level is commensurate with available resources) (AGFD 2010, p. 3). Since 1991, the Sonoran desert tortoise population on the Maricopa Mountains plot has experienced relatively high survivorship and shown evidence of reproduction. No additional carcasses have been documented, indicating the population may be stable, if not returning to the previous 1987 level (AGFD 2010, p. 3).

The AGFD (2010, p. 3) and Hart et al. (1992, p. 120) confirm Sonoran desert tortoise populations declined from initial population estimates (as demonstrated by density estimates and relative carcass numbers) on three additional plots (Hualapai Foothills, San Pedro Valley, and East Bajada), suspecting that drought conditions may have played a role in the observed declines on these plots (Ogden 2009, pp. 12-13)

10, p. 3) and Hart et al. (1992, p. 120) confirm Sonoran desert tortoise populations declined from initial population estimates (as demonstrated by density estimates and relative carcass numbers) on three additional plots (Hualapai Foothills, San Pedro Valley, and East Bajada), suspecting that drought conditions may have played a role in the observed declines on these plots (Ogden 2009, pp. 12-13). An observed decline on the Tortilla Mountains plot in 2001 may have been an artifact of low surface activity in response to below-average precipitation, because an increase in carcasses was not detected (AGFD 2010, p. 3).

For detailed information on monitoring and survey results from the previous three decades for the Sonoran desert tortoise in Arizona, see the following reports: Schneider (1981), Shields and Woodman (1987), Wirt (1988), Woodman and Shields (1988), Holm (1989), Shields et al. (1990), SWCA (1990a; 1990b; 1990c), Hart et al. (1992), Murray and Schwalbe (1993; 1997), Woodman et al. (1993; 1994; 1995; 1996; 1998; 1999a; 1999b; 2000; 2001; 2002; 2003; 2004; 2005; 2006; 2007; 2008; 2009), AIDTT (2000, pp. 5-6), Averill-Murray (2000, pp. 3-7), Averill-Murray and Klug (2000, pp. 3-25), Averill-Murray et al. (2002b, pp. 110-112), Walker and Wood (2002), Young et al. (2002), and Zylstra and Swann (2009).

It should be noted that an average generation time for a Sonoran desert tortoise is 12-15 years and that monitoring of Sonoran desert tortoise populations has only occurred for about 30 years, representing approximately two generations. Many threats described below have been potentially acting on Sonoran desert tortoise populations for many decades, longer than populations have been studied. Below, we discuss the effects of various threats to individual Sonoran desert tortoises. However, due to limitations in monitoring data, we are unable to discern how Sonoran desert tortoise populations may have responded to these threats over time, or identify any long-term, historical trends in tortoise populations

tortoise populations for many decades, longer than populations have been studied. Below, we discuss the effects of various threats to individual Sonoran desert tortoises. However, due to limitations in monitoring data, we are unable to discern how Sonoran desert tortoise populations may have responded to these threats over time, or identify any long-term, historical trends in tortoise populations. We have not observed any extirpations among monitored populations.

Distinct Population Segment

We consider a species for listing under the Act if available information indicates such an action might be warranted. “Species” is defined by the Act as including any subspecies of fish or wildlife or plants, and any distinct population segment (DPS) of any species of vertebrate fish or wildlife that interbreeds when mature (16 U.S.C. 1532(16)). We, along with the National Marine Fisheries Service (now the National Oceanic and Atmospheric Administration—Fisheries), developed the Policy Regarding the Recognition of Distinct Vertebrate Population Segments (61 FR 4722; February 7, 1996), to help us in determining what constitutes a DPS. The policy identifies three elements that are to be considered regarding the status of a possible DPS. These elements include: (1) The discreteness of the population segment in relation to the remainder of the taxon (group of similar biological organisms); (2) the significance of the population segment to the taxon to which it belongs; and (3) the population segment's conservation status in relation to the Act's standards for listing ( i.e., whether the population segment, when treated as if it were a species, is endangered or threatened) (61 FR 4722, February 7, 1996). The first two elements are used to determine if a population segment constitutes a valid DPS. If it does, then the third element is used to consider whether such DPS warrants listing

n segment's conservation status in relation to the Act's standards for listing ( i.e., whether the population segment, when treated as if it were a species, is endangered or threatened) (61 FR 4722, February 7, 1996). The first two elements are used to determine if a population segment constitutes a valid DPS. If it does, then the third element is used to consider whether such DPS warrants listing. In this section, we will consider the first two criteria (discreteness and significance) to determine if the Sonoran desert tortoise is a valid DPS ( i.e., a valid listable entity). Our policy further recognizes it may be appropriate to assign different classifications ( i.e., threatened or endangered) to different DPSs of the same vertebrate taxon (61 FR 4722).

Discreteness

Under the DPS policy, a population segment of a vertebrate species may be considered discrete if it satisfies either one of the following two conditions:

(1) It is markedly separated from other populations of the same taxon as a consequence of physical, physiological, ecological, or behavioral factors. Quantitative measures of genetic or morphological discontinuity (separation based on genetic or morphological characters) may provide evidence of this separation.

(2) It is delimited by international governmental boundaries within which significant differences in control of exploitation, management of habitat, conservation status, or regulatory mechanisms exist that are significant in light of section 4(a)(1)(D) of the Act.

Based upon available information, the international boundary between Mexico and the United States is not considered for delineation of discreteness because

co and the United States is not considered for delineation of discreteness because

The Colorado River and Río Yaqui are two perennial rivers that form biogeographical barriers (a natural barrier that prevents the migration of species) to movement of tortoises between the Mojave and Sonoran desert tortoise populations, and between the Sonoran and Sinaloan desert tortoise populations, respectively. The Colorado River, separating California and Arizona, comprises the northern and western boundaries of the Sonoran desert tortoise population as identified in the April 2, 1990, final rule designating the Mojave population of the desert tortoise (occurring north and west of the Colorado River) as a threatened species under the Act (55 FR 12178; see final rule for a summary of previous actions regarding the Mojave population of the desert tortoise). The eastern boundary is the extent of the range of the Sonoran desert tortoise where desert habitats end and grassland, chaparral, and mountain habitats begin, which are areas that do not contain desert tortoises. The southern boundary of the Sonoran desert tortoise DPS, as considered in this finding, is the Río Yaqui in southern Sonora, Mexico; south and east of there, desert tortoises are considered Sinaloan populations. Potential threats to the Sinoloan desert tortoise are not evaluated as part of this finding.

In view of this biogeographical isolation, significant ecological divergence has occurred between the Mojave and Sonoran populations of desert tortoise, largely due to significant differences in geology, vegetation types, and precipitation cycles where the populations are distributed. Desert tortoises in the Mojave population are most dense in the intermountain valleys that have soil types favorable to the construction of large, deep burrows (Bury et al. 1994, pp. 66-70). However, Sonoran desert tortoises reach maximum densities in the rocky bajadas and hillsides of higher slopes, with reduced densities in the intermountain valleys (Berry 1984, p

he populations are distributed. Desert tortoises in the Mojave population are most dense in the intermountain valleys that have soil types favorable to the construction of large, deep burrows (Bury et al. 1994, pp. 66-70). However, Sonoran desert tortoises reach maximum densities in the rocky bajadas and hillsides of higher slopes, with reduced densities in the intermountain valleys (Berry 1984, p. 434; AIDTT 2000; p. 4; Van Devender 2002a, p. 7; Brennan and Holycross 2006, p. 54; Zylstra and Steidl 2008, p. 747). At the southern end of the DPS, Edwards et al. (2009, pp. 7-8) suggested that Sinaloan population of desert tortoise uses Sinaloan thornscrub and tropical deciduous forest habitats. These different habitat types with differing soils and vegetation communities are created by higher precipitation levels. However, some level of gradation may occur in the vegetative transition zone between Plains of Sonora subdivision of Sonoran desertscrub and Sinaloan thornscrub habitats of central Sonora such as in the vicinity of the Río Yaqui (Edwards et al. 2009, p. 8).

In addition to habitat differences, morphological differences have also been documented among the three populations of desert tortoise. Several morphological differences in carapace size and shape have been documented between the Mojave, Sonoran, and Sinaloan populations of desert tortoise: The carapace of the Mojave desert tortoise is the widest and tallest of the three, the Sinaloan desert tortoise carapace is the most narrow and least domed, and the carapace of the Sonoran desert tortoise is intermediate between the two in those dimensions (Germano 1993, pp. 324-325; AGFD 2001, p. 1). Using eight independent shell measurements, Weinstein and Berry (1987, pp

ons of desert tortoise: The carapace of the Mojave desert tortoise is the widest and tallest of the three, the Sinaloan desert tortoise carapace is the most narrow and least domed, and the carapace of the Sonoran desert tortoise is intermediate between the two in those dimensions (Germano 1993, pp. 324-325; AGFD 2001, p. 1). Using eight independent shell measurements, Weinstein and Berry (1987, pp. 26-28) documented three distinct phenotypes (physical appearances) in desert tortoise populations within the United States based on morphometric (body measurement) analyses: The “California” phenotype (Mojave population), “Beaver Dam Slope” phenotype (Mojave form in Arizona), and the “Sonoran type” (Sonoran population). Desert tortoises from southern Sonora and northern Sinaloa in Mexico were not studied as part of this effort.

Differences in reproduction strategies between the Sonoran and Mojave populations of desert tortoises also occur. Mojave desert tortoises lay up to three clutches of eggs per year with larger clutch sizes (more eggs), earlier in the year (April to mid-July) (Wallis et al. 1999, p. 405) while those in the Sonoran population lay one clutch per year of smaller size, later in the year (June through August) (Averill-Murray et al. 2002a, p. 141). These differences led Averill-Murray (2002b, pp. 119-122) to the conclusion that Sonoran desert tortoises invest all reproductive effort into a single clutch which hatches at the peak of forage and water availability and abundance owing to late-summer rainfall. Whereas desert tortoises in the Mojave population (maturing at smaller body sizes) (Berry et al. 2002a, p. 259) have higher clutch numbers to offset higher mortality from greater variability in environmental conditions.

population (maturing at smaller body sizes) (Berry et al. 2002a, p. 259) have higher clutch numbers to offset higher mortality from greater variability in environmental conditions.

The Mojave, Sonoran, and Sinaloan populations of the desert tortoise have been found to have significantly differentiated genotypes (genetic characteristics) (Lamb and McLuckie 2002, p. 74; Van Devender 2002a, p. 24). Genetic distances, expressed as percent sequence divergence (an estimate of percent difference in the genetic code), are substantial among the three populations of desert tortoise. Divergence is 5.1-5.6 percent between the Sonoran and Mojave populations, 4.2 percent between the Sonoran and Sinaloan populations, and 5.1 percent between the Sinaloan and Mojave populations (Lamb and McLuckie 2002, pp. 74, 77). Considering geographic distribution, genealogical depth, and a suite of other characteristics, the Mojave, Sonoran, and Sinaloan populations of desert tortoise are considered to be ecologically significant units (populations or groups of populations historically isolated from one another, and thus representing deep phylogenetic (evolutionary development of species over time) subdivisions within species) (Lamb and McLuckie 2002, pp. 81-82). According to mitochondrial DNA markers, the Sonoran and Mojave populations appear to have diverged some 5 million years ago (Lamb et al. 1989, p. 83; Lamb and McLuckie 2002, p. 76).

McCord (2002, p. 62) presented three possible causes of the significant genetic differentiation between Sonoran and Mojave desert tortoises. First, genetic differentiation between Sonoran and Mojave desert tortoises may have been the result of differences in rainfall patterns between the winter-dominated rainfall pattern of the Mojave Desert and the summer-dominated rainfall pattern of the Sonoran desert

(2002, p. 62) presented three possible causes of the significant genetic differentiation between Sonoran and Mojave desert tortoises. First, genetic differentiation between Sonoran and Mojave desert tortoises may have been the result of differences in rainfall patterns between the winter-dominated rainfall pattern of the Mojave Desert and the summer-dominated rainfall pattern of the Sonoran desert. Second, genetic differentiation between Sonoran and Mojave desert tortoises may have occurred because the Sonoran desert tortoises may be represented as a relict population (remnant survivor from the past) of the tropical deciduous forest-evolved population of the Sinaloan population (based upon their general absence in valley bottoms due to heavy flooding during summer rains, a phenomenon generally absent in the Mojave Desert). Last, genetic differences between Sonoran and Mojave desert tortoises may have resulted from their mutual competition with the Bolson tortoise ( Gopherus flavomarginatus ), another desert tortoise species which was widely distributed throughout Arizona in the Pleistocene, but which never occurred in California. The competing Bolson tortoise population may have acted as a wedge between the Sonoran and Mojave populations, driving them even farther apart, in a process known as competitive displacement.

To explore the evolutionary track the three desert tortoise populations may have taken and the extent of their current genetic differentiation on the landscape, Edwards et al. (2009, p. 8) collected genetic samples from desert et al. 2009, p. 8). This confirms the similar genetic relationships of Sonoran desert tortoises throughout the DPS. Genetic samples from the Ciudad Obregón region, southward, showed clear genetic distinction and supported prior evidence for a third distinct population of desert tortoise, referred to as the Sinaloan population (Edwards et al. 2009, p. 8)

p. 8) collected genetic samples from desert et al. 2009, p. 8). This confirms the similar genetic relationships of Sonoran desert tortoises throughout the DPS. Genetic samples from the Ciudad Obregón region, southward, showed clear genetic distinction and supported prior evidence for a third distinct population of desert tortoise, referred to as the Sinaloan population (Edwards et al. 2009, p. 8). The southern limits of desert tortoise distribution in northern Sinaloa are likely influenced by the growth of disease-causing bacteria and fungi present in the soil of burrows, exacerbated by the hot, humid, and wet conditions during tropical summer rainy seasons (Van Devender 2002b, p. 43).

Evaluation of Discreteness

Some biological similarities do exist among the three populations of desert tortoise (Mojave, Sonoran, and Sinaloan). For example, some overlap in habitat use occurs. It is well known that Sonoran desert tortoises generally occur on steep, rocky slopes and bajadas in contrast to the Mojave desert tortoise, which occurs primarily along the valley bottoms. But to a lesser extent, Sonoran desert tortoises also use valley bottoms and Mojave desert tortoises also use steep slopes and mountain bajadas (Gardner and Brodie 2000, p. 51; Averill-Murray and Averill-Murray 2002, p. 16; Lutz et al. 2005, p. 22; Grandmaison et al. in press, p. 4; Riedle et al. 2008, p. 418). However, there are many more numerous and convincing data in the scientific literature to support the discreteness of the three recognized populations of Gopherus agassizii, including differences in their ecology, behavior, morphology, physiology, and genetics (Weinstein and Berry 1987, pp. 26-28; Germano 1993, pp. 324-325; Germano et al. 1994, p. 82; AGFD 2001, p. 1; Averill-Murray 2002b, pp. 299-300; Berry et al. 2002a, p. 259; Lamb and McLuckie 2002, pp. 74, 77; McCord 2002, p. 62; Van Devender 2002a, pp. 24-25; Van Devender 2002b, p. 45; Zylstra and Steidl 2008, p. 747; Edwards et al. 2009, p. 8)

ology, behavior, morphology, physiology, and genetics (Weinstein and Berry 1987, pp. 26-28; Germano 1993, pp. 324-325; Germano et al. 1994, p. 82; AGFD 2001, p. 1; Averill-Murray 2002b, pp. 299-300; Berry et al. 2002a, p. 259; Lamb and McLuckie 2002, pp. 74, 77; McCord 2002, p. 62; Van Devender 2002a, pp. 24-25; Van Devender 2002b, p. 45; Zylstra and Steidl 2008, p. 747; Edwards et al. 2009, p. 8).

We have reviewed the best available commercial and scientific information and find that the Sonoran population of the desert tortoise as it occurs east and south of the Colorado River, south to the Río Yaqui, in Sonora, Mexico, is discrete, under the Service's DPS policy, from the Mojave and Sinaloan desert tortoise populations. We base this conclusion on ecological (habitat use), physiological (reproductive characteristics), morphological (shell dimensions), and behavioral (seasonal activity patterns) differences that are further supported by analysis of genetic differences that concluded significant divergence has occurred among the three populations.

Significance

If a population segment is considered discrete under one or more of the conditions described in the Service's DPS policy, its biological and ecological significance will be considered in light of Congressional guidance that the authority to list DPSs be used “sparingly” while encouraging the conservation of genetic diversity. In making this determination, we consider available scientific evidence of the discrete population segment's importance to the taxon to which it belongs. Since precise circumstances are likely to vary considerably from case to case, the DPS policy does not describe all the classes of information that might be used in determining the biological and ecological importance of a discrete population. However, the DPS policy describes four possible classes of information that provide evidence of a population segment's biological and ecological importance to the taxon to which it belongs

likely to vary considerably from case to case, the DPS policy does not describe all the classes of information that might be used in determining the biological and ecological importance of a discrete population. However, the DPS policy describes four possible classes of information that provide evidence of a population segment's biological and ecological importance to the taxon to which it belongs. As specified in the DPS policy (61 FR 4722), this consideration of the population segment's significance may include, but is not limited to, the following:

(1) Persistence of the discrete population segment in an ecological setting unusual or unique to the taxon;

(2) Evidence that loss of the discrete population segment would result in a significant gap in the range of a taxon;

(3) Evidence that the discrete population segment represents the only surviving natural occurrence of a taxon that may be more abundant elsewhere as an introduced population outside its historic range; or

(4) Evidence that the discrete population segment differs markedly from other populations of the species in its genetic characteristics.

A population segment needs to satisfy only one of these conditions to be considered significant. Furthermore, other information may be used as appropriate to provide evidence for significance.

The current range of the Sonoran desert tortoise, as described in the discussion above pertaining to discreteness, represents several hundred miles of occupied habitat spanning across an international border. This population segment is confined by two large perennial rivers: The Colorado River in its northern periphery (separating the Mojave and Sonoran populations), and the Río Yaqui at its southern periphery (separating the Sonoran and Sinaloan populations)

in the discussion above pertaining to discreteness, represents several hundred miles of occupied habitat spanning across an international border. This population segment is confined by two large perennial rivers: The Colorado River in its northern periphery (separating the Mojave and Sonoran populations), and the Río Yaqui at its southern periphery (separating the Sonoran and Sinaloan populations). These two rivers represent significant biogeographical barriers to genetic exchange between adjacent population segments and, therefore, preclude recolonization of this expanse of habitat from adjacent populations, should the Sonoran population of the desert tortoise become extirpated. Thus, the loss of the Sonoran desert tortoise would constitute a significant gap of several hundred miles in the range between the Mojave and Sinaloan populations of desert tortoises, and may constitute as much as 40 percent of the total range occupied by desert tortoises as a whole, rangewide, which affirms its significance to the entire species.

In addition, our evaluation of discreetness above found extensive scientific support concluding that the Sonoran desert tortoise differs significantly in its behavior (reproduction, seasonal activity), ecology (habitat use and burrow construction), morphology (physical characteristics), and genetics from either the Sinaloan or the Mojave populations. Because of these distinctions, the loss of the Sonoran desert tortoise population would result in the permanent loss of a unique biological entity and would diminish the natural variation within the species as a whole.

Evaluation of Significance

We have reviewed the best available commercial and scientific data, and based on that review, we find that the Sonoran desert tortoise is significant to the continued existence of the taxon

of the Sonoran desert tortoise population would result in the permanent loss of a unique biological entity and would diminish the natural variation within the species as a whole.

Evaluation of Significance

We have reviewed the best available commercial and scientific data, and based on that review, we find that the Sonoran desert tortoise is significant to the continued existence of the taxon. We base this conclusion on: (1) The large geographic range of the Sonoran population, which is significant (approximately 40 percent) to the taxon as a whole; (2) a gap of several hundred miles that would result from the loss of the Sonoran population, which would effectively bisect the species' range; and

Determination of Distinct Population Segment

Based on our review of the best commercial and scientific information available, the Sonoran population of desert tortoise is discrete from the Mojave and Sinaloan populations and significant to the species as a whole. As a result, we have determined that the Sonoran population of desert tortoise qualifies as a DPS and a listable entity under the Act.

In the August 23, 2009, 90-day finding (74 FR 44335), we discussed a local population of Mojave-genotype (genotype: genetic code) desert tortoises (that also share Mojave phenotype (the physically-expressed genetic code) and habitat-use characteristics with the Mojave desert tortoise population) occurring within the delineated Sonoran population in the Black Mountains area of western Mohave County, Arizona. This population is isolated from the threatened Mojave DPS that occurs north and west of the Colorado River. The exact geographic extent of this Mojave-genotype in Arizona is currently undefined and we expect there is interbreeding between desert tortoises with the Mojave and Sonoran genotype along the periphery of this population in the Black Mountains. Therefore, we include this population of desert tortoises as part of our status assessment for the Sonoran desert tortoise in this finding

lorado River. The exact geographic extent of this Mojave-genotype in Arizona is currently undefined and we expect there is interbreeding between desert tortoises with the Mojave and Sonoran genotype along the periphery of this population in the Black Mountains. Therefore, we include this population of desert tortoises as part of our status assessment for the Sonoran desert tortoise in this finding.

Distinct Population Segment Five-Factor Analysis

Section 4 of the Act (16 U.S.C. 1533) and implementing regulations (50 CFR part 424) set forth procedures for adding species to, removing species from, or reclassifying species on the Federal Lists of Endangered and Threatened Wildlife and Plants. Under section 4(a)(1) of the Act, a species may be determined to be endangered or threatened 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;

(D) The inadequacy of existing regulatory mechanisms; or

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

In making this finding, information pertaining to the Sonoran desert tortoise in relation to the five factors provided in section 4(a)(1) of the Act is discussed below.

In considering what factors might constitute threats to a species, we must look beyond the exposure of the species to a particular factor to evaluate whether the species may respond to that factor in a way that causes actual impacts to the species. If there is exposure to a factor and the species responds negatively, the factor may be a threat and, during the status review, we attempt to determine how significant a threat it is. The threat is significant if it drives, or contributes to, the risk of extinction of the species such that the species warrants listing as endangered or threatened as those terms are defined in the Act

o the species. If there is exposure to a factor and the species responds negatively, the factor may be a threat and, during the status review, we attempt to determine how significant a threat it is. The threat is significant if it drives, or contributes to, the risk of extinction of the species such that the species warrants listing as endangered or threatened as those terms are defined in the Act. However, the identification of factors that could impact a species negatively may not be sufficient to compel a finding that the species warrants listing. The information must include evidence sufficient to suggest that these factors are operative threats that act on the species to the point that the species may meet the definition of endangered or threatened under the Act.

In our review of the best scientific and commercial data available, we found numerous threats are impacting Sonoran desert tortoises or their habitat throughout their range. Some of these threats occurred historically, some are current, and some will continue into the foreseeable future. As described in detail below, these threats include nonnative plant species and altered fire regimes, urban and agricultural development, barriers to dispersal and genetic exchange, off-highway vehicles, roads and highways, ironwood and mesquite tree harvest, improper livestock grazing, undocumented human immigration, illegal collection, effects from field research and manipulation, predation from feral dogs, human depredation and vandalism, drought, and climate change. The effect of habitat disturbances on Sonoran desert tortoises may differ among age classes, but may be most significant to hatchlings or juveniles (Tracy et al. 2006b, pp. 271-272).

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

search and manipulation, predation from feral dogs, human depredation and vandalism, drought, and climate change. The effect of habitat disturbances on Sonoran desert tortoises may differ among age classes, but may be most significant to hatchlings or juveniles (Tracy et al. 2006b, pp. 271-272).

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

Nonnative Plant Species and Altered Fire Regimes

The most significant modification to Sonoran desert tortoise habitat is associated with the ongoing invasion of nonnative plants in Mojave and Sonoran desertscrub habitats, permanently altering these ecosystems and causing a change in the frequency, duration, intensity, and magnitude of wildfires in a region that largely evolved in the absence of invasive nonnative plants. These ecosystem-level changes cause both direct and indirect effects on the Sonoran desert tortoise and its habitat.

Much of the available research on the effects of nonnative plant species invasions and wildfire used in our analysis has focused on Mojave desertscrub habitats, largely due to the presence of the Mojave desert tortoise, which is already listed as endangered. However, Brooks and Matchett (2006, p. 158) suggest that research from the Mojave Desert is applicable to the Sonoran Desert when stating, “Both (Mojave and Sonoran deserts) occur at elevations above the hyperarid shrublands, are often positioned on the lower slopes of mountain ranges, and possess moderate woody plant cover.” Therefore, we used the information available from research on Mojave Desert habitats in our assessment of the effects of nonnative plants in the Sonoran Desert.

Nonnative perennial plants like buffelgrass, fountain grass, and Lehmann lovegrass were historically introduced to the Sonoran Desert of Arizona as livestock forage and to prevent soil erosion

es, and possess moderate woody plant cover.” Therefore, we used the information available from research on Mojave Desert habitats in our assessment of the effects of nonnative plants in the Sonoran Desert.

Nonnative perennial plants like buffelgrass, fountain grass, and Lehmann lovegrass were historically introduced to the Sonoran Desert of Arizona as livestock forage and to prevent soil erosion. For example, buffelgrass was included in the nonnative plant species recommended for release by the Tucson Plant Materials Center of the Soil Conservation Service until at least 1987 (Bahr 1991, p. 156). These nonnative plant species subsequently became common and widespread in Sonoran desertscrub in Arizona (Brooks and Pyke 2001, p. 5). They have since colonized new areas, often taking advantage of disturbed soils, such as those resulting from construction associated with roadways, power lines, and railroad tracks (Bahre 1991, p. 155; D'Antonio and Vitousek 1992, p. 65). Construction and maintenance of roads and highways can also significantly enhance the likelihood of nonnative plant invasions by increasing nitrogen deposition in the soil, the dispersal potential of nonnative seeds, and adjacent soil moisture (Brooks 2007, pp. 153-154). Roadside ditches along highways are particularly important dispersal corridors for nonnative plant species such as red brome and buffelgrass (Esque et al. 2002, p. 313).

Mechanisms that allow the spread of nonnative species generally pertain to ground disturbance, but the plants may also be spread by other mechanisms. For example, Smith et al. (2000, pp. 79-80), and Brooks and Esque (2002, p. 337) both found that elevated atmospheric carbon dioxide levels, predicted as a result of climate change (discussed in

ample, Smith et al. (2000, pp. 79-80), and Brooks and Esque (2002, p. 337) both found that elevated atmospheric carbon dioxide levels, predicted as a result of climate change (discussed in

The prevalence of nonnative grasses in many areas of Sonoran desertscrub habitats has resulted in high amounts of flammable fuels in interspaces between native plants that would otherwise be free of vegetation. This situation serves to promote the ignition and carrying of wildfire (Brooks 1999, p. 13). In our review of the best scientific data available, red brome, splitgrass (or Mediterranean grass, Schismus spp.), and buffelgrass were considered the nonnative plant species that pose the greatest concern to the Sonoran desert tortoise and its habitat, because they are thoroughly integrated into some areas of the desertscrub communities, and serve to promote and carry wildfire (Bahre 1991, p. 155; D'Antonio and Vitousek 1992, pp. 65, 75; Brooks 1999, p. 13; Brooks and Pyke 2001, p. 5; Brooks and Esque 2002, p. 337; Esque et al. 2002, p. 313; Van Devender 2002a, p. 16; Brooks and Matchett 2006, p. 148; DeFalco 2007a, p. 1; Zouhar et al. 2008, p. 157; Abella 2010, p. 1249; AGFD 2010, p. 13). Red brome is known to carry wildfire in Sonoran desertscrub habitat north of Tucson, natal grass is known to carry wildfire in desert grassland habitat south of Tucson to Nogales, Arizona, and buffelgrass is known to carry wildfire in Sonoran desertscrub and foothills thornscrub south of the international border to central Sonora (Esque et al. 2002, p. 316). Other nonnative plant species identified in the literature as present in Sonoran and Mojave desertscrub communities include Saharan (or Asian) mustard ( Brassica tournefortii ), thistles (genera Centaurea and Cirsium ), crimson fountaingrass ( Pennisetum setaceum ), natal grass ( Melinus repens ), and Lehmann lovegrass ( Eragrostis lehmanniana ) (Brooks 2001, p. 4; Brooks and Pyke 2001, pp. 3, 5)

Other nonnative plant species identified in the literature as present in Sonoran and Mojave desertscrub communities include Saharan (or Asian) mustard ( Brassica tournefortii ), thistles (genera Centaurea and Cirsium ), crimson fountaingrass ( Pennisetum setaceum ), natal grass ( Melinus repens ), and Lehmann lovegrass ( Eragrostis lehmanniana ) (Brooks 2001, p. 4; Brooks and Pyke 2001, pp. 3, 5).

We are not aware of any good estimates of the number of acres of desertscrub that have been invaded by nonnative plant species, but Thomas and Guertin (2007, Appendices I and II) calculated the number of records by county for many known invasive, nonnative plants in Arizona that are harmful to Sonoran desert tortoise habitat. These data illustrate general locations where certain nonnative species are most common and describe which nonnative species are the most reported in each area. Thomas and Guertin (2007, Appendices I and II) reported the following for Arizona as of 2007 (relative number of reports of densities being “extremely high,” “high,” “moderate,” and “occurs,” all within the distribution of the Sonoran desert tortoise):

(1) Buffelgrass is the most-reported nonnative plant species in Arizona, at 16.3 percent of total reports with 6,287 reports (p. 3); it reaches extremely high densities in Maricopa and Pima Counties, with high densities in Pinal and Yuma Counties and moderate densities in Santa Cruz and La Paz Counties, but it also occurs in Yavapai, Gila, and Cochise Counties (A-I, p. 60);

(2) Schismus spp. is one of the top 20 invasive plant species, at 2.4 percent of total reports, with 919 reports (p. 3); it reaches high densities in Maricopa, Pinal, and Pima Counties, with moderate densities in Mohave, Yavapai, Gila, La Paz, and Yuma Counties, but it also occurs in Santa Cruz County (A-I, p. 69);

ounties, but it also occurs in Yavapai, Gila, and Cochise Counties (A-I, p. 60);

(3) Red brome is one of the top 20 invasive plant species, at 3 percent of total reports, with 1,152 reports (p. 3); it reaches high densities in Yavapai, Gila, Pinal, and Pima Counties, with moderate densities in Mohave and Maricopa Counties, but it also occurs in La Paz and Yuma Counties (A-I, p. 24);

(4) Saharan mustard is one of the top 20 invasive plant species, at 3.3 percent of total reports, with 1,261 reports (p. 3); it reaches high densities in Maricopa, Pinal, Pima, La Paz, and Yuma Counties, with moderate densities in Mohave, Yavapai, and Gila Counties, but it also occurs in Cochise County (A-I, p. 21);

(5) Centaurea spp. had a total of 3-318 reports (depending on species) (p. 9) and reaches high densities in Pima County, with moderate densities in Mohave, Yavapai, Gila, Pinal, and Cochise Counties (A-I, pp. 15, 28-30);

(6) Bull thistle ( Cirsium vulgare ) is one of the top 20 invasive plant species, at 3.1 percent of total reports, with 1,195 reports (p. 3); it reaches moderate densities in Yavapai and Gila Counties (A-I, p. 35);

(7) Crimson fountaingrass is one of the top 20 invasive plant species, at 2.6 percent of total reports, with 999 reports (p. 3); it reaches high densities in Pima County, with moderate densities in Yavapai, Gila, La Paz, Santa Cruz, and Maricopa Counties (A-I, p. 61); and

ercent of total reports, with 1,195 reports (p. 3); it reaches moderate densities in Yavapai and Gila Counties (A-I, p. 35);

(8) Lehman lovegrass is one of the top 20 invasive plant species, at 2.5 percent of total reports, with 980 reports (p. 3); it reaches high densities in Pima and Cochise Counties, with moderate densities in Yavapai, Gila, Santa Cruz, Maricopa, and Pinal Counties, but also occurs in La Paz County (A-I, p. 45).

No spatial data were provided for natal grass, but there were 191 observations (Thomas and Guertin 2007, p. 10).

Buffelgrass has widely invaded Arizona and northern Mexico since its introduction in 1939 (Stevens and Fehmi 2009, p. 379). While buffelgrass invasions are occurring and are poised to seriously impact the southwestern United States, the species has already exacted significant tolls on Sonoran desertscrub communities in Sonora, Mexico, because its expansion continues to be facilitated through intentional plantings and cultivation. Consequently, the clearing of Sonoran desertscrub and Sinaloan thornscrub in Sonora to plant pastures of buffelgrass for livestock grazing creates a near monoculture (area covered by a single plant species) that is highly prone to wildfires, and therefore represents a substantial threat to the Sonoran desert tortoise in Mexico (Bury et al. 2002, p. 104; Walker and Pavlakovich-Kochi 2003, p. 14; Van Devender and Reina 2005, pp. 160-161; University of Arizona 2010, p. 2). Buffelgrass has been planted in Sonora's desertscrub lands since the 1950s and at least 5.5 million ac (2.2 million ha) of potential Sonoran desert tortoise habitat has already been converted into a near monoculture of buffelgrass (Stoleson et al. 2005, p. 62)

et al. 2002, p. 104; Walker and Pavlakovich-Kochi 2003, p. 14; Van Devender and Reina 2005, pp. 160-161; University of Arizona 2010, p. 2). Buffelgrass has been planted in Sonora's desertscrub lands since the 1950s and at least 5.5 million ac (2.2 million ha) of potential Sonoran desert tortoise habitat has already been converted into a near monoculture of buffelgrass (Stoleson et al. 2005, p. 62). Buffelgrass has become established in both the lower valley habitats and into the granite boulder-strewn areas of adjacent foothills, and has altered historical fire regimes, regionally converting large areas of Sonoran desertscrub into habitat resembling the African savannah (Bury et al. 2002, p. 104).

In Arizona, the Southern Arizona Buffelgrass Coordination Center

Brooks and Minnich (2006, p. 9) stated that southwestern desert ecosystems likely evolved in a fire regime best described by “low intensity, patchy burns and long fire return intervals.” Wildfire capable of carrying itself in Sonoran desertscrub is a recent phenomenon in evolutionary and geological contexts and only became apparent recently in the Sonoran Desert (Brooks and Pyke 2001, p. 5; Esque et al. 2002, p. 312; Zouhar et al. 2008, pp. 155, 160). From 1937 to 1986, only 1 percent of all lightning-caused fires in the Rincon Mountains area of southern Arizona occurred in desertscrub habitat; 5.6 percent occurred in desert grassland habitat (Bahre 1991, p. 126). While historical wildfires in desertscrub habitat were exceptionally rare, after successive years of above-average levels of precipitation, enough native fuels can develop to carry wildfire in desertscrub communities, such as happened south of Florence, Arizona in 1979 (Bahre 1991, p. 141; Brooks and Esque 2002, p. 336; Brooks and Minnich 2006, p. 9). While increased precipitation enhances plant growth and subsequently increases the likelihood for wildfire starts in desertscrub habitat, drought can have an inverse effect with respect to certain nonnative plant species

carry wildfire in desertscrub communities, such as happened south of Florence, Arizona in 1979 (Bahre 1991, p. 141; Brooks and Esque 2002, p. 336; Brooks and Minnich 2006, p. 9). While increased precipitation enhances plant growth and subsequently increases the likelihood for wildfire starts in desertscrub habitat, drought can have an inverse effect with respect to certain nonnative plant species. Red brome, for example, is sensitive to drought conditions and, therefore, might contribute to reduced fuel loads and decreased fire frequency during long-term drought (Brooks and Esque 2002, p. 337), which might help to minimize the likelihood of wildfires in areas where red brome has formed a monoculture. Smith et al. (2000, p. 79) noted, “This shift in species composition in favor of exotic annual grasses, driven by global [climate] change, has the potential to accelerate the fire cycle, reduce biodiversity and alter ecosystem function in the deserts of western North America.”

Wildfire ignitions in the Sonoran Desert region historically resulted from lightning but ignitions are now more common from human sources such as burning trash, parking vehicles over dry vegetation, fireworks, discarded cigarettes, and accidental starts from backcountry recreationists (Esque et al. 2002, p. 313). Human-caused wildfires in desertscrub habitat are most common near urban developments, major roadways, and in areas where off-highway vehicle use is unregulated, while lightning-caused wildfire in desertscrub is typically located in more remote wilderness areas (Brooks 1999, p. 13). In central Sonora, ranchers intentionally set fires to maintain the vigor of buffelgrass for livestock forage (Esque et al. 2002, p. 313)

s in desertscrub habitat are most common near urban developments, major roadways, and in areas where off-highway vehicle use is unregulated, while lightning-caused wildfire in desertscrub is typically located in more remote wilderness areas (Brooks 1999, p. 13). In central Sonora, ranchers intentionally set fires to maintain the vigor of buffelgrass for livestock forage (Esque et al. 2002, p. 313).

Numerous wildfires, varying in size, have occurred in recent times in many areas throughout the Sonoran Desert including the: (1) Pusch Ridge Fire of 1987 on the southern slopes of the Santa Catalina Mountains; (2) Skyline (1992) and Rock Peak (1993) fires in the San Tan Mountains; (3) Mother's Day Fire of 1994 on the eastern slope of the Rincon Mountains (Esque et al. 2002, p. 323; 2003, p. 104); and (4) Cave Creek Complex fire of 2005 northeast of Cave Creek, Arizona, which burned 248,310 ac (100,487 ha) of Sonoran desert tortoise habitat; the largest wildfire ever recorded in Sonoran desertscrub in the United States.

The BLM has kept records of wildfire in Sonoran desert tortoise habitat. From 1990 to 2008, there have been 61 wildfires, affecting 21,977 ac (8,894 ha) in Category I Sonoran desert tortoise habitat; 285 wildfires, affecting 33,364 ac (13,502 ha) in Category II Sonoran desert tortoise habitat; and 508 wildfires, affecting 109,460 ac (44, 297 ha) in Category III Sonoran desert tortoise habitat (USBLM 2010, p. 9). In total, during the 1990-2008 period, 164,801 ac (66,693 ha) of categorized and uncategorized Sonoran desert tortoise habitat has burned on BLM lands (USBLM 2010, p. 9)

toise habitat; 285 wildfires, affecting 33,364 ac (13,502 ha) in Category II Sonoran desert tortoise habitat; and 508 wildfires, affecting 109,460 ac (44, 297 ha) in Category III Sonoran desert tortoise habitat (USBLM 2010, p. 9). In total, during the 1990-2008 period, 164,801 ac (66,693 ha) of categorized and uncategorized Sonoran desert tortoise habitat has burned on BLM lands (USBLM 2010, p. 9). Combining the known area of habitat affected by fire on both BLM and other lands, an estimated 1.5 percent of habitat in Arizona has been adversely affected due to wildfire in recent years; rangewide this is estimated to be 0.8 percent, although total acreage data on wildfires in Mexico are unknown and the total percentage of affected habitat is likely higher because of the higher incidence of buffelgrass and lessened capacity to fight wildfire in Sonora, Mexico. The total area reported as burned is a relatively small proportion of BLM lands and has not likely been a significant impact to most Sonoran desert tortoise populations in Arizona so far. As the invasion of nonnative plants continues to expand, the high number of fire starts has a greater potential of creating larger and more destructive wildfires, especially where they occur in remote, inaccessible areas as a result of lightning strikes.

Indirect effects of wildfires on Sonoran desert tortoises are variable and can be significant, including habitat changes such as altered nutrient availability and quality, loss of perennial plant species that are important as temporary cover from predators, loss of thermal refugia, altered tortoise behavior, shifts in biotic community, pronounced desert tortoise emigration from burned habitat, and lower growth and reproductive output (Esque et al. 2003, p. 107; DeFalco 2006, p. 5; McLuckie et al. 2007, p. 8)

changes such as altered nutrient availability and quality, loss of perennial plant species that are important as temporary cover from predators, loss of thermal refugia, altered tortoise behavior, shifts in biotic community, pronounced desert tortoise emigration from burned habitat, and lower growth and reproductive output (Esque et al. 2003, p. 107; DeFalco 2006, p. 5; McLuckie et al. 2007, p. 8). While a single fire in an area may or may not produce long-term reductions in plant cover or biomass, repeated wildfires in a given area are capable of ecosystem type-conversion from native desertscrub to nonnative annual grassland, and render the area unsuitable for desert tortoises (Brooks and Esque 2002, p. 336). Increased frequency in wildfires caused by nonnative plant species invasion increases light intensity at ground level and soil nutrient availability, and reduces competition from native perennial plants. These changes further promote dominance by nonnative plant species (Brooks and D'Antonio 2003, p. 29). Wildfire in desertscrub habitats can reduce native and nonnative seed banks (Brooks and Draper 2006, p. 2). In Mojave desertscrub, the effects of fire are most pronounced under shrubs, where fire can kill seed banks and reduce annual grass diversity, due to higher burn intensity (Brooks 2002a, p. 1; 2002b, p. 1088). Microhabitat associated with shrubs in Sonoran desert tortoise habitat is an important source of temporary shelter and provides foraging opportunities while tortoises are thermoregulating.

Fires associated with nonnative plant species have already affected Sonoran desert tortoise populations in Arizona. The AGFD (2010, p. 13) reported results from an unpublished study after the Edge Complex Fire of 2005 in the Four Peaks area on the Tonto National Forest, which indicated higher numbers of Sonoran desert tortoises (or their scat were observed in unburned versus burned habitat), but they acknowledged that the study was preliminary and very limited in scope (AGFD 2010, p. 13)

populations in Arizona. The AGFD (2010, p. 13) reported results from an unpublished study after the Edge Complex Fire of 2005 in the Four Peaks area on the Tonto National Forest, which indicated higher numbers of Sonoran desert tortoises (or their scat were observed in unburned versus burned habitat), but they acknowledged that the study was preliminary and very limited in scope (AGFD 2010, p. 13).

In Sonora, Mexico, 5.5 million ac (2.2 million ha), representing an estimated 22 percent of Sonoran desert habitat in Mexico, or 11 percent rangewide, has been planted to bufflegrass. This figure still does not account for the land area et al. 2005, p. 62). Combining the current and predicted number of acres converted to buffelgrass in Mexico, 34 percent of the Sonoran desert tortoises' habitat is lost or at risk across its range. In the area of El Batamote, 29 mi (47 km) north of Hermosillo, Sonora, buffelgrass has invaded Sonoran desert tortoise habitat in the adjacent foothills, which has led to wildfires that burned so hot that the soil was scorched and the bedrock cracked (Esque et al. 2002, p. 321).

In addition to impacts from fire, Franklin and Molina-Freaner (in press, p. 1) found that these large-scale conversions from desertscrub to grasslands in Sonora have reduced plant species richness by half, and reduced tree and shrub cover by 78 percent, vastly affecting the ability of Sonoran desert habitat to meet the species' thermoregulatory needs (that is, using vegetation as cover to regulate body temperature)

acts from fire, Franklin and Molina-Freaner (in press, p. 1) found that these large-scale conversions from desertscrub to grasslands in Sonora have reduced plant species richness by half, and reduced tree and shrub cover by 78 percent, vastly affecting the ability of Sonoran desert habitat to meet the species' thermoregulatory needs (that is, using vegetation as cover to regulate body temperature). These changes have resulted in substantial changes in primary productivity (creation of organic nutrients and the lowest level of the food chain, the plant community) and vegetation structure (the physical structure of plant sizes and shapes as a mosaic on the landscape) which can affect the forage base and habitat suitability for Sonoran desert tortoises, as well as lessened the feasibility of restoring native plant communities in Sonora without aggressive land management (Franklin and Molina-Freaner, in press, p. 1). Dense stands of buffelgrass have also been shown to physically disrupt tortoise movements in the closely related Texas tortoises ( Gopherus berlandieri ) (Fujii and Forstner 2010, p. 61), so this may also be true for Sonoran desert tortoises. The grass can become so thick that the tortoises cannot walk through it, and the grass may be too tall for the tortoises to walk on top of it.

In addition to damaging Mojave and Sonoran desertscrub habitat, wildfires can directly injure and kill Sonoran desert tortoises. Wildfire may kill a desert tortoise by incineration, by elevating body temperature, by poisoning from smoke inhalation, or by asphyxiation (Brooks et al. 1999, p. 40; Brooks and Esque 2002, p. 335; McLuckie et al. 2007, p. 7). Survival rates of Sonoran desert tortoises may be contingent upon several factors, including soil type, substrate, vegetation, tortoise activity during fire, whether tortoises are active and above ground or in shelter during a fire, weather, fire behavior, and shelter depth (McLuckie et al. 2007, p. 8)

yxiation (Brooks et al. 1999, p. 40; Brooks and Esque 2002, p. 335; McLuckie et al. 2007, p. 7). Survival rates of Sonoran desert tortoises may be contingent upon several factors, including soil type, substrate, vegetation, tortoise activity during fire, whether tortoises are active and above ground or in shelter during a fire, weather, fire behavior, and shelter depth (McLuckie et al. 2007, p. 8). The desert tortoise is most vulnerable to the direct effects of wildfire when they are surface active and away from primary cover sites such as burrows, caliche caves, and rock shelters, because these structures reduce direct exposure to heat and smoke (Brooks and Esque 2002, p. 335). Gravid (with fertilized eggs) female Sonoran desert tortoises may be more likely to perish from wildfire than other tortoises because peak wildfire season in Sonoran desertscrub occurs during the months of May and June. This is when reproductive females are actively foraging on spring growth to compensate for energy used in egg development; (Esque et al. 2002, pp. 323-324; 2003, p. 106).

Sonoran desert tortoises that survive the wildfire itself may struggle to survive in post-burned Sonoran desertscrub habitat due to: (1) A reduction in forage and shade structure, such as packrat ( Neotoma sp.) middens and shrubs; and (2) increased visibility to predators (which may be further increased in intermountain valleys where temporary shade, predator avoidance, and available forage are particularly important in long-distance movements in these dispersal corridors) (Esque et al. 2002, pp. 325-326).

orridors) (Esque et al. 2002, pp. 325-326).

The effects on Sonoran desert tortoises of one particular fire were studied in some detail. Within Saguaro National Park, the Mother's Day Fire of 1994 burned 340 ac (138 ha) of Arizona Upland Sonoran desertscrub habitat that was occupied by Sonoran desert tortoises, killing an estimated 11 percent of the tortoise population (Esque et al. 2003, p. 105). To assess how Sonoran desert tortoises used burned versus unburned habitat following this fire, transmitters were attached to 12 tortoises, 6 each in burned and unburned habitat within or adjacent to the Mother's Day Fire footprint. Surprisingly, no differences were observed in movement or activity patterns between tortoises in burned and unburned areas, nor were long-term effects of the fire on surviving tortoises noted over the 6-year study period (Zylstra and Swann 2009, p. 7). These results indicate that different tortoise populations may respond differently to wildfires and that numerous variables and factors are at work.

One of the principal reasons that nonnative plants pose a significant impact to Sonoran desert tortoise habitat is because few, if any, reasonable methods currently exist to control the ongoing invasion of these plants or to remediate areas where they have become established. Mechanical removal is one option that has been implemented on a small scale in some areas, but is extremely labor intensive and not practical for treating large areas. Prescribed fire has been proposed as an alternative means to control nonnative plant species invasions, but also carries obvious inherent risks to habitat and to Sonoran desert tortoises (Brooks 2006, p. 31).

vious inherent risks to habitat and to Sonoran desert tortoises (Brooks 2006, p. 31).

It is also important to note the limitations of Sonoran desert habitat with respect to post-disturbance (for example, after fires) regeneration (ability for native vegetation to recover). Desertscrub regions receive low annual precipitation totals, and the plant communities have correspondingly low growth rates. Based on the type of disturbance, recovery time estimates range from 40 years to centuries (Abella 2010, pp. 1271, 1273). Combined, these factors result in slow, post-disturbance recovery periods and it may take a long time before any area becomes suitable for Sonoran desert tortoises to recolonize, if at all. The presence of nonnative species such as buffelgrass, cheatgrass, or red brome in disturbed Mojave or Sonoran desertscrub may further limit post-disturbance recovery, delay recovery, or prevent recovery altogether (Brown and Minnich 1986, p. 411; Brooks 1999, p. 18).

In our review of the best available information, we have documented that nonnative plant species pose a significant threat to the Sonoran desert tortoise and its habitat, both in Arizona and Sonora, by promoting and carrying wildfire in an ecosystem that evolved in its absence. Wildfires that are facilitated by nonnative plant species invasions may have direct and indirect adverse effects on tortoises and tortoise populations. The threat from nonnative plant species to the Sonoran desert tortoise occurs throughout the species' range and is expected to increase over time with the expansion of nonnative plants. There is currently no viable solution to curbing this continued expansion across the landscape. This threat also acts synergistically with other threats discussed in this finding.

Urban Development and Agriculture

Human population growth results in the disturbance or loss of Sonoran desertscrub or the conversion of land for urban and agricultural development

the expansion of nonnative plants. There is currently no viable solution to curbing this continued expansion across the landscape. This threat also acts synergistically with other threats discussed in this finding.

Urban Development and Agriculture

Human population growth results in the disturbance or loss of Sonoran desertscrub or the conversion of land for urban and agricultural development. Arizona increased its population by 394 percent from 1960 to 2000, and was second only to Nevada as the fastest growing State during this timeframe (Social Science Data Analysis Network (SSDAN) 2000, p. 1). Since 1990, Arizona's population has grown by 44 et al. 2006, p. 7).

Urban expansion and human population growth trends in Arizona are expected to continue into the future. Maricopa-Pima-Pinal county areas of Arizona are expected to grow by as much as 71 percent in the next 15 years, creating rural-urban edge effects across millions of acres of public lands currently supporting Sonoran desert tortoise populations (AIDTT 2000, p. 10; BLM files—Lands Livability Initiative). In another projection, the population in Arizona is expected to more than double within the next 20 years compared to the 2000 population estimate (U.S. Census Bureau 2005, p. 1). Many cities and towns within the distribution of the Sonoran desert tortoise have already experienced substantial growth during the 8-year time span, 2000-2008: City of Avondale (118.3 percent); City of Buckeye (392.5 percent); Bullhead City (20.3 percent), Town of Carefree (30.5 percent); Casa Grande (56 percent); Town of Cave Creek (44.2 percent); City of Chandler (37.5 percent); City of Coolidge (24.9 percent); City of El Mirage (195.6 percent); City of Eloy (22.3 percent); City of Florence (20.3 percent); Town of Fountain Hills (23.2 percent); City of Gilbert (84.5 percent); City of Goodyear (203 percent); City of Kingman (32.2 percent); Lake Havasu City (33.3 percent); City of Litchfield Park (34.2 percent); City of Mammoth (45 percent); Town of Marana (139.9 perce

nt); City of Coolidge (24.9 percent); City of El Mirage (195.6 percent); City of Eloy (22.3 percent); City of Florence (20.3 percent); Town of Fountain Hills (23.2 percent); City of Gilbert (84.5 percent); City of Goodyear (203 percent); City of Kingman (32.2 percent); Lake Havasu City (33.3 percent); City of Litchfield Park (34.2 percent); City of Mammoth (45 percent); Town of Marana (139.9 percent); City of Maricopa (2,508 percent); Town of Oro Valley (32.5 percent); Town of Queen Creek (544.5 percent); Town of Saguarita (507.3 percent); City of San Luis (58.5 percent); City of Somerton (63.2 percent); City of Surprise (187.3 percent); City of Tolleson (43.2 percent); and, Town of Youngtown (62.2 percent) (U.S. Census Bureau 2008, pp. 1-4).

This population growth has spurred a significant increase in urbanization and development in these areas. Regional development is predicted to be extreme in certain areas within the distribution of the Sonoran desert tortoise in Arizona. In particular, a wide swath from the international border in Nogales, through Tucson, Phoenix, and north into Yavapai County (called the Sun Corridor “Megapolitan”) is predicted to have 8 million people by 2030, an 82.5 percent increase from 2000 (Gammage et al. 2008, pp. 15, 22-23). If build-out occurs as expected, it will encompass a significant proportion of the Sonoran desert tortoise distribution in Arizona, and will in effect permanently isolate Sonoran desert tortoise populations that occur on either side of the Interstate 19, Interstate 10, and Interstate 17 corridors.

terstate 19, Interstate 10, and Interstate 17 corridors.

The land area permanently altered by human activities from urban development and agriculture has grown to 13 percent of all land in the western United States, Lue et al. (2008, p. 1130). Lue et al. (2008, p. 1133) concluded that in low-productivity habitat, such as desertscrub habitats, slight human disturbances can have pronounced effects. Significant urban development occurs within intermountain valleys, within or adjacent to occupied Sonoran desert tortoise habitat, which increases the likelihood of effects along the rural-urban interface, and may also inhibit movement of individuals between populations on nearby hillsides or mountain ranges. Disturbances to Sonoran desert tortoise habitat on the landscape can take many forms and cover extreme distances. Roads, canals, pipelines, and railroad tracks are examples of linear habitat destruction. We discuss the potential effects of linear disturbances below in the section titled, “Development as a Barrier.”

Development pressure across Arizona has slowed due to the recent economic downturn and decline in the housing market. However, development will likely continue in the future, although perhaps at a slower pace than in the earlier part of this century. We also recognize that economic trends are difficult to predict into the future. The most recent draft Pinal County Comprehensive Plan (February 2009) acknowledges that the county is in the middle of the Sun Corridor Megapolitan and proposes four shorter-term growth areas in defining where development will likely occur, or be encouraged to develop, over the next decade, but does not discourage growth outside of these areas (Pinal County Comprehensive Plan 2009, p. 109). These four growth areas (Gateway/Superstition Vistas, West Pinal, Red Rock, and Tri-Communities) fall completely within the range of the Sonoran desert tortoise

oposes four shorter-term growth areas in defining where development will likely occur, or be encouraged to develop, over the next decade, but does not discourage growth outside of these areas (Pinal County Comprehensive Plan 2009, p. 109). These four growth areas (Gateway/Superstition Vistas, West Pinal, Red Rock, and Tri-Communities) fall completely within the range of the Sonoran desert tortoise. The Gateway/Superstition Vistas growth area alone encompasses 176,000 ac (71,225 ha), or 275 sq mi (712 sq km), of State Trust land, and it is anticipated that 800,000 to more than 1 million people will one day live in this development (Pinal County Comprehensive Plan 2009, p. 115). The loss of 176,000 ac (71,225 ha) constitutes a loss of 0.7 percent of Sonoran desert tortoise habitat in Arizona; rangewide, 0.34 percent. The Pinal County Comprehensive Plan (2009, p. 117) identifies many miles of new freeways and principal arterials in the analysis area at build-out, which the plan acknowledges may take over a half century to realize (Pinal County Comprehensive Plan 2009, p. 115). The effect of roads on Sonoran desert tortoises is discussed below.

Additionally, the Maricopa County Comprehensive Plan calls for growth areas to the south and east of Chandler and Mesa, Arizona, which are within the range of the Sonoran desert tortoise (Maricopa County Comprehensive Plan 2002 (revised), p. 92). City comprehensive plans within the range of the Sonoran desert tortoise also call for future growth areas. For example, the City of Eloy has designated six such areas encompassing 15,520 ac (6,281 ha), mostly along the Interstate 10 corridor (City of Eloy General Plan 2004, pp. 7-6 through 7-10). The loss of 15,520 ac (6,281 ha) constitutes a loss of 0.06 percent of their habitat in Arizona; rangewide, 0.03 percent

plans within the range of the Sonoran desert tortoise also call for future growth areas. For example, the City of Eloy has designated six such areas encompassing 15,520 ac (6,281 ha), mostly along the Interstate 10 corridor (City of Eloy General Plan 2004, pp. 7-6 through 7-10). The loss of 15,520 ac (6,281 ha) constitutes a loss of 0.06 percent of their habitat in Arizona; rangewide, 0.03 percent. While much of this area has already been impacted by development or irrigated agriculture, any remaining dispersal habitat for the Sonoran desert tortoise will likely be negatively affected as development and its associated infrastructure progress into these areas.

Much of the past and projected development within the range of the Sonoran desert tortoise in central and southwestern Arizona has occurred and is expected to continue as a conversion from agricultural uses to municipal uses. Land traditionally used for agriculture is not occupied by Sonoran desert tortoises, but has a comparatively minor effect on adjacent Sonoran desert tortoises. When these lands are converted to municipal uses, the effect to adjacent Sonoran desert tortoise populations increases human access, and use of adjacent undeveloped land increases as a result of development of these former agricultural areas.

The human population of Sonora, Mexico, doubled in size from 1970 (1.1 million) to 2000 (2.2 million) (Stoleson et al. 2005, p. 54). The population of Sonora is expected to increase by 23 percent, to 2.7 million people, in 2020 (Stoleson et al. 2005, p. 54). In discussing threats to Sonoran desert tortoise populations adjacent to, and stemming from, urbanization in Sonora, Mexico, Fritts and Jennings (1994, p. 53)

ng threats to Sonoran desert tortoise populations adjacent to, and stemming from, urbanization in Sonora, Mexico, Fritts and Jennings (1994, p. 53)

Urban development has been identified as a concern for Sonoran desert tortoise conservation in several areas within Arizona because of the associated increase in human-based threats to populations in close proximity. Averill-Murray and Swann (2002, p. 1) stated that urban development adjacent to the Saguaro National Park in Pima County threatens the Sonoran desert tortoise via several mechanisms including harassment and predation by feral or off-leash domestic dogs, illegal releases of captive Sonoran desert tortoises and exotic species that may transmit diseases to wild Sonoran desert tortoises, elevated mortality on roads, and illegal collection for pets. Averill-Murray and Swann (2002, p. 7) stated that mid- to large-scale development projects on the bajadas and foothills of the Rincon, Santa Rita, Santa Catalina, Tortolita, and Tucson Mountains has likely led to area-wide decreases in Sonoran desert tortoise populations. However, no population estimates for Sonoran desert tortoises before development of these areas exist, and, therefore, population responses to development of these areas cannot be ascertained.

In addition to the Tucson metropolitan area, urban encroachment on Sonoran desert tortoise habitat occurs adjacent to the greater Phoenix metropolitan area, in the area around South Mountain and adjacent to the Superstition Mountains (AGFD 2010, p. 7). Sonoran desert tortoises are known or suspected to still occur in 12 of the 16 Maricopa County and City of Phoenix urban mountain parks and reserves. The four parks where no tortoise sign has been found in recent years are completely surrounded by urban development (AGFD 2010, p. 7). Urban development has occurred adjacent to five monitoring plots, but only the Hualapai Foothills plot is completely surrounded by developed lands (AGFD 2010, p. 7)

o still occur in 12 of the 16 Maricopa County and City of Phoenix urban mountain parks and reserves. The four parks where no tortoise sign has been found in recent years are completely surrounded by urban development (AGFD 2010, p. 7). Urban development has occurred adjacent to five monitoring plots, but only the Hualapai Foothills plot is completely surrounded by developed lands (AGFD 2010, p. 7). A development consisting of 48,000 single family homes, south of the Colorado River in western Mohave County, is also currently being planned (THS 2009, p. 4; Mardian 2010, p. 1).

Because less area is being used currently for agriculture in the United States, habitat loss due to agricultural development is more of a historical issue. However, impacts to Sonoran desert tortoise dispersal habitat within valley floors from historical agricultural use and wood harvesting are still evident. The vegetation and soils of many valleys in the Sonoran Desert were shaped by the periodic flooding of dynamic wash systems, which partially recharged a shallow, fluctuating groundwater table. Because of agricultural development, these valleys no longer experience these defining processes and there has been a permanent loss of meso- and xeroriparian habitat which are known to be corridors for movement by Sonoran desert tortoises (Jackson and Comus 1999, pp. 233, 249; Lutz et al. 2005, p. 22; Riedle et al. 2008, p. 418).

Agriculture in Sonora, Mexico, has shifted from small-scale, local markets toward large-scale agro-industry, with Sonora producing 40 percent of the country's total wheat crop (Stoleson et al. 2005, p. 59). While agriculture in Sonora is largely constrained to valleys (along the Rio Sonora), many types of habitat used by Sonoran desert tortoises have been cleared for agriculture, including Sonoran desertscrub, thornscrub, and tropical deciduous forest (Stoleson et al. 2005, p. 60)

ge-scale agro-industry, with Sonora producing 40 percent of the country's total wheat crop (Stoleson et al. 2005, p. 59). While agriculture in Sonora is largely constrained to valleys (along the Rio Sonora), many types of habitat used by Sonoran desert tortoises have been cleared for agriculture, including Sonoran desertscrub, thornscrub, and tropical deciduous forest (Stoleson et al. 2005, p. 60). In 1994, the total irrigated acreage in Sonora was 128,000 ac; in 2004 that figure rose to 530,509 ac (214,689 ha), an increase of 314 percent (AQUASTAT 2007, p. 2

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

This is a copy of a public record, reproduced as it was published. It is not legal advice, and it may not be the version a court would rely on. Check the official source before you cite it.