Introduction

The U.S. Department of the Army (Army) plans to commence military activity and training exercises within the National Training Center (NTC) and the Fort Irwin Western Training Area (WTA) in San Bernardino County, California. If approved, the NTC will establish training areas to test the combat readiness of brigade-sized units (1,000–5,000 soldiers and 1,000–1,500 vehicles) in a realistic battlefield environment. Starting in 2025 (app. 1), as many as 10 brigade-level training events may happen with force-on-force and live-fire scenarios to prepare units for combat and security missions. In addition to Army units stationed at Fort Irwin, joint military branches (Marine Corps, Navy, and Air Force), Army Reserve, National Guard, Special Operations Forces, multinational partnerships, and regular and transitional law enforcement units also train at the NTC. Planned actions in the WTA would likely change land use patterns in areas that were previously undisturbed and affect desert tortoises and their habitats such that translocation of the federally and California State listed Mojave Desert tortoises (Gopherus agassizii; hereafter tortoise) would be required by the U.S. Fish and Wildlife Service (USFWS) before training ensues to minimize tortoise mortality (U.S. Fish and Wildlife Service, 1994, 2011, 2020, 2022b). Minimization of land used for military activities in the WTA will be considered throughout the translocation process. This study was done in cooperation with the U.S. Department of Defense, U.S. Army Garrison Fort Irwin, California, and U.S. Geological Survey.

The Army’s plan for military activities in the WTA follows previously completed actions from the Fort Irwin Military Land Withdrawal Act of 2001 (Congress Public Law 107-107, div. B, title XXIX, December 28, 2001, 115 United States Statues at Large 1336), authorizing the NTC to expand its training activities into approximately 577 km² of military lands previously designated as “critical habitat for tortoises” (U.S. Fish and Wildlife Service, 2011). These lands included NTC expansion areas that had enough tortoises to warrant translocation, the Southern Expansion Area (SEA), and the Superior Valley (referred to as “Western Expansion Area,” or “WEA,” during the 2005 translocation; now referenced as the Western Training Area; 285.6 km²). To prepare for the NTC’s first large-scale tortoise translocation in 2005, available information documenting tortoise translocation was used to formulate the NTC translocation plan and support translocation as an adequate solution for successful relocation of displaced tortoises. Much of the available information focused on several short-term translocation success metrics (movement patterns and survivorship), with further investigation required to evaluate the effects of translocation at several temporal scales to evaluate long-term (15–30 years) success given the long lifespan of desert tortoises (Tasse, 1989; Dickinson and Fa, 2000; Fischer and Lindenmayer, 2000; U.S. Fish and Wildlife Service, 2020). Desert tortoises generally are sexually mature when they reach greater than 180 millimeters (mm) in carapace length (Turner and others, 1986, 1987). The time it takes for tortoises to reach this size ranges between 15 and 21 years, depending on environmental conditions and other habitat characteristics that affect resource availability during their development (Turner and others, 1987; Tracy and Tracy, 1995; Medica and others, 2012). Long-term monitoring (25 years) is implemented to fully understand the success of translocation.

Previous plans for the translocation of tortoises from NTC expansion areas (Esque and others, 2005, 2009) sought to fill gaps in translocation knowledge, documented actions, and procedures to: (1) provide safe, humane, and successful translocation of tortoises with minimal effect to resident (animals living within the recipient sites before translocation) and reference (telemetered animals living outside of the translocation sites, but whose movements are predicted not to overlap with translocated or resident animals) tortoises at sites where translocated animals are released; (2) study tortoises affected by translocation to increase understanding of the ecology, conservation, and management of desert tortoises (U.S. Fish and Wildlife Service, 1994, 2011, 2022b; Tracy and others, 2004); and (3) define measures of success for translocation and provide metrics to evaluate success during multiple time scales.

There were large-scale translocations of approximately 650 adult tortoises from the Southern Expansion Area (SEA) at the NTC to public lands just to the south (the Southern Expansion Translocation Area) in 2008. The NTC’s 2005 Translocation Plan described conservation science activities sponsored by the NTC to be enacted during the next approximately 20 years (Esque and others, 2005). In 2011, all conservation science activities supported by the NTC for tortoises associated with the SEA and WEA (now WTA) as described in the 2005 and 2009 Translocation Plans were discontinued prior to completing the WEA translocation (Esque and others, 2005; Esque and others, 2009); however, extramural funding from the National Science Foundation supported limited activities for approximately 4 years and provided some post-translocation short-term information (table 1). The NTC translocations were followed by additional translocations for planned military training activities in the region (such as Marine Corps Air Ground Combat Center, 2016; U.S. Fish and Wildlife Service, 2017).

Research and monitoring of tortoises, habitats, and translocation activities that were done in association with the 2008 NTC translocation substantially contributed to knowledge of tortoise ecology, regional landscape conditions, and related effects of translocation, with numerous studies supported financially and logistically by the NTC (table 1). Various surveying and analytical methods used to detect tortoise presence have informed regional tortoise density estimates, population trends, and habitat models used to predict potential areas of tortoise habitat (Karl, 2002; Aycrigg and others, 2004; Berry and others, 2006; Heaton and others, 2008a, 2008b; Nussear and others, 2008, 2009; Harless and others, 2010; Allison and McLuckie, 2018; Carter and others, 2020; Parandhaman and others, 2022; Averill-Murray and Allison, 2023; Kissel and others, 2023; Zylstra and others, 2023). Identification, assessment, and protection of suitable tortoise habitat has become critical to tortoise conservation because enduring tortoise population declines have been documented in four of five federally designated Mojave Desert recovery units (Allison and McLuckie, 2018; Zylstra and others, 2023). Human development and habitat fragmentation have contributed to tortoise population declines and affected the demographic viability of tortoise populations vital to the survival of the species (Averill-Murray and Hagerty, 2014; Allison and McLuckie, 2018; Carter and others, 2020; Hromada and others, 2020; Averill-Murray and others, 2021; Averill-Murray and Allison, 2023). One aspect of supporting demographically viable tortoise populations is identifying tortoise genetic units on the landscape to allow evaluations of the costs and benefits to genetic diversity, which can be a driver of healthy tortoise populations when moving tortoises among sites (Averill-Murray and Hagerty, 2014; Scott and others, 2020). Initial genetic integration of translocated tortoises into release areas was low when paternity of young tortoises was analyzed based on genetics (Mulder and others, 2017); however, further longer-term reproductive investigation is needed.

During and after the 2008 NTC SEA translocation, research was done to better understand movement and space use (through habitat and among burrows), disease transmission risks, stress levels, and gene flow after translocation (Latch and others, 2011; Drake and others, 2012; Aiello and others, 2014, 2018; Averill-Murray and Hagerty, 2014; Bowen and others, 2015; Farnsworth and others, 2015; Sah and others, 2016; Mulder and others, 2017; Mack and Berry, 2023). The stress response of translocated tortoises was assessed by quantifying and comparing values of the reptilian hormone corticosterone for translocated, resident, and reference tortoises. Results indicated that translocation did not elicit a detectable physiological stress response from tortoises but rather patterns varied by sex, activity season, and year (Drake and others, 2012). Additional post-translocation studies corroborated findings on space-use and site fidelity from previous years, in which translocated tortoises dispersed greater distances (1.5 times more than reference tortoises) and had lower site fidelity in the first year after translocation when compared to resident and reference populations (Hinderle and others, 2015). Translocated tortoises are likely to disperse shorter distances and have a higher likelihood of survival when the recipient sites have an abundance of tortoise burrows, a variety of soil substrate textures that provide opportunities for burrow construction, and plentiful washes on the landscape for travel corridors and foraging, although translocated tortoises generally visit fewer burrows than resident tortoises (Mack and others, 2015; Sah and others, 2016; Nafus and others, 2017a). Threats to desert tortoises, including proximity to urban areas and predation by mammalian carnivores, were documented after the 2008 SEA translocation; however, these threats were not unique to the NTC or translocation activities and instead were documented throughout the Mojave Desert in relation to prolonged drought conditions and subsidized predators in proximity to urbanized areas (Esque and others, 2010; Cypher and others, 2018; Emblidge and others, 2015). Other documented threats included roads, litter, climate, and ravens (Corvus corax; Walde and others, 2007; McIntyre and others, 2010; Mack and Berry, 2023). Tortoise disease, particularly Upper Respiratory Tract Disease (URTD) and Testudinid herpesvirus 2 (TeHV2), has been further chronicled, with the pathogenicity of suspected causative agents Mycoplasma agassizii and Mycoplasma testudineum confirmed, refined antibody and pathogen presence laboratory tests developed, and transmission patterns in captive and wild populations studied (Jacobson and others, 2012; Aiello and others, 2014, 2018; Berry and others, 2015).

Project Area: Site Descriptions

Western Training Area

The WTA (286 km²) is in the southwest corner of the Fort Irwin NTC (fig. 1). The WTA is bounded by the geographical designations of 3908200 and 3890200 northing and 492500 and 516500 easting Universal Transverse Mercator (UTM) lines. The WTA borders the Naval Air Weapons Station-China Lake (3908200 northing UTM; 492500 easting UTM) to the north, the Paradise Range and Lane Mountain to the south, and Superior Dry Lake to the west. The WTA is comprised of broad flat valleys with many sandy washes interspersed by low gentle hills and rocky outcrops (northwest corner of the WTA). Most of the area is internally drained by the Superior and Goldstone Basins. The elevation within the WTA ranges from 814 to 1,382 meters (m).

The WTA includes two contiguous areas of restricted access that are not considered further in this translocation plan. Excluding the conservation or restricted access areas, the WTA encompasses 254 km² (fig. 1). The first restricted access area is known as the “East Paradise Conservation Area” that is 18 km² and was designated as a “BLM Area of Critical Environmental Concern” (Bureau of Land Management, 2005) for the conservation of the Mojave Desert tortoise, endemic Lane Mountain milkvetch (Astragalus jaegerianus), and the Mojave ground squirrel (Xerospermophilus mohavensis). The East Paradise Conservation Area is fenced with tortoise exclusionary fencing to the northeast, allowing tortoises from the southwest to access this area but not the rest of the WTA. The second restricted access area, Brinkman Wash Restricted Area, was designated by the Army for foot traffic only and is 14 km².

Western Training Area Translocation Site

The WTATS was delineated through discussions among the BLM, NTC, U.S. Geological Survey (USGS), and USFWS by reviewing suitable translocation sites for tortoises and using subsequent analyses by USGS (see the “Modeling Habitat Site Selection for Recipient and Reference Sites” section). We evaluated approximately 5,585 km² of lands, mostly west, south, and southeast of the WTA in San Bernardino County, California (fig. 1). During our evaluation, we reduced the footprint of this larger landscape (used and referenced as the study area) to include habitats most appropriate for translocated tortoises, which is now referred to as the “WTATS” (fig. 1). The WTATS covers approximately 3,296 km² of mostly public lands north of Barstow and Hinkley, California, and is bounded on the north by the Naval Air Weapons Station China Lake (3917952 and 3849256 northing UTM lines), to the south by the 3849332 northing UTM line, to the east by the 458197 easting UTM lines, and to the west by 571068 easting UTM line within the Soda Mountains. The eastern side of the WTATS incorporates habitats where NTC previously translocated tortoises from its SEA (Esque and others, 2005) in 2008. The WTATS includes two BLM designated Wilderness Areas: (1) Grass Valley and (2) Black Mountain, and there are two recreation areas and public campgrounds at Rainbow Basin and Owl Canyon, which collectively comprise 210 km² (fig. 1). The WTATS is a mosaic of property ownership and management, with public lands managed by the BLM, which administers the greatest amount of land (approximately 2,145 km²; 65 percent; fig. 1) and oversees a large network of roads and trails (including off-highway vehicles [OHVs]) in the region. Holdings by the NTC (referred to as “Fort Irwin mitigation parcels” or “Irwin mitigation parcels” that are 2.59 km²; approximately 320 km²; 9.7 percent), the State of California lands (approximately 93 km²; 2.8 percent), and non-Federal or private property (approximately 742 km²; 22.5 percent) represent the remaining ownership and are largely concentrated in the southern WTATS (fig. 1). The WTATS is more topographically diverse than the WTA and comprised of large broad valleys, rugged volcanic and granitic mountains, and gentle hills made up of diverse sedimentary parent materials. The region encompasses a large network of washes that drain into the Superior and Harper Valley Basins and associated dry lakes. The elevation in this area ranges from 516 to 1,250 m.

Site Selection Guidance from the Bureau of Land Management

Due to a complex network of property ownership, management, and landscape use throughout the West Mojave management area, staff at BLM Barstow and Ridgecrest Field Offices provided recommendations for habitats and areas that should be excluded from consideration as potential recipient sites within the WTATS study area. These recommended avoidance areas included (1) areas south of Interstate 15 (I–15), Interstate 40 (I–40), and State Route 58, (2) areas east and south of a primary transmission utility corridor and access road, (3) BLM designated wilderness (Grass Valley and Black Mountain wildernesses), and (4) targeted areas south and southwest of Fossil Bed Road that have highly intense recreation activities and other landscape concerns (fig. 2). In later discussions, the BLM indicated that any translocated tortoises that moved into designated Wilderness Area habitats from their recipient sites would not be removed by the BLM; however, such a scenario is unlikely because proposed recipient sites, and their calculated dispersal range buffers (6.5 kilometers [km]; U.S. Fish and Wildlife Service, 2020), are not expected to extend into Wilderness Areas and are separated from Wilderness Areas by natural geographic barriers. The BLM Barstow Field Office also provided information regarding where BLM and their partners are focusing route restoration efforts (areas between and east of the Wilderness Areas), including the suggestion that these areas receive higher priority as recipient sites due to in-progress and anticipated improvements in habitat conditions.

Baseline Tortoise Investigations (2020–22)

We performed baseline tortoise and habitat investigations in the WTA and the WTATS after the boundaries and habitat considerations were identified. All baseline activities pertaining to tortoises and their habitats were authorized under a USFWS Federal permit (number TE-63428D-0, -1), a CDFW Memorandum of Understanding (MOU; signed “March 30, 2020”), and a BLM MOU (signed “March 31, 2020”). A subset of tortoises located during the tortoise surveys (described in the next section) had radio transmitters attached to locate them during further investigations (telemetered tortoises). General methods and handling protocols done from April 2020 to November 2022 were approved by the USGS-Western Ecological Research Center Animal Care and Use Committee (U.S. Geological Survey, 2020). Tortoise surveys were done using tortoise surveying, capturing, handling, monitoring methods, and applications as described in the “Tortoise Clearance Protocols for the Western Training Area” section of this plan.

Tortoise Survey Plots

Proposed training areas (in the WTA) and recipient sites (in the WTATS) were surveyed to estimate tortoise density and abundance and to document habitat characteristics. Survey plots (300 by 300 m) were randomly distributed on public lands for tortoise surveys during spring and fall 2020, 2021, and 2022 by following protocols similar to USFWS (2022a; fig. 3). All tortoise sign was recorded, including live tortoises (tortoises greater than 180-mm carapace length [hereafter adults] and tortoises less than or equal to 180-mm carapace length [hereafter juveniles]), carcasses, and burrows. Plots were at least 50 m from BLM designated roads and excluded from non-federally owned parcels, campgrounds, dry lake beds (when possible), and designated Wilderness Areas. Survey transects were spaced at 10-m intervals beginning in the southwestern corner of each plot. A total of 1,408 plots were surveyed in the project area from spring 2020 to spring 2022.

Baseline Tortoise Health Assessments

The 2020–22 plot surveys (fig. 3) and monitoring efforts for telemetered tortoises throughout the project area included observations of 783 tortoises, 41 of which were from the 2008 NTC translocation efforts (fig. 4). Of the tortoises observed, 86 percent were adult tortoises, with a consistent two male to one female sex ratio among years. Most tortoise encounters were when tortoises were in burrows or under vegetation. The most frequently used vegetative cover species throughout the study area were Larrea tridentata (creosote bush), Ambrosia dumosa (burro bush), and Atriplex polycarpa (desert saltbush).

From 2020 to 2021, in the WTATS and WTA, telemetered tortoises were evaluated for clinical health conditions with physical assessments, including body condition scoring (BCS) and tissue collection (blood samples and oral swabs) whenever possible, following USFWS guidance (U.S. Fish and Wildlife Service, 2019). Clinical health conditions of tortoises were characterized by examining each animals posture, respiration, face (with specific attention to the eyes, periocular tissue, nares, mouth, tongue, and oral mucosa), skin, and shell for signs of disease, abnormalities, damage, or discoloration. Health assessors also searched for any discharge from the cloaca, eyes, nares, and mouth or evidence of ulceration, erythema, swelling, or discharge on the skin (U.S. Fish and Wildlife Service, 2019). The overall condition and fat stores with respect to skeletal features of the head and limbs of animals were characterized through assignment of numerical body condition scores, first through categorization as “under,” “adequate,” or “over” condition, and then by numerical values (under: 1–3, adequate: 4–6, over: 7–9) to provide a precise and repeatable measurement (U.S. Fish and Wildlife Service, 2019). Ectoparasites observed on tortoises (including Ornithodoros spp. [ticks]) were counted, placed in cryogenic vials, and stored on wet ice while in the field and later flash frozen with dry ice or placed directly into ultracold storage (−70 degrees Celsius [°C]).

Immediately after the physical assessment, tissues were collected from each animal (when possible, including blood and oral swabs). Whole blood was extracted (0.3–2.0 milliliters [mL] separated into aliquot samples, when appropriate) via subcarapacial venipuncture (Hernandez-Divers and others, 2002) using a 3.81 centimeter (cm), 23-gauge needle, and 3-mL syringe coated in sodium heparin to prevent coagulation. Whole blood was either placed directly onto a Whatman FTA card (GE Healthcare Life Science, Marlborough, Massachusetts; less than 0.01-mL blood), or into a cryogenic vial with Invitrogen RNAlater (ThermoFisher Scientific, Vilnius, Lithuania) mixed at 2 parts solution to 1 part blood for future ribonucleic acid (RNA) extraction and gene expression analysis or into a BD Microtainer tube with lithium heparin (Becton Dickinson and Company, Franklin Lakes, New Jersey) for subsequent centrifugation to separate plasma. Sloughed epithelial cells from mouths of tortoises were collected using oral swabs that were rotated slowly across surfaces of the tongue and oral mucosa. All samples were stored on wet ice for no more than 4 hours and then transferred to an ultracold freezer (−70 °C). Blood plasma and oral swab samples were sent to labs for enzyme-linked immunosorbent assay (ELISA) testing for acquired antibodies and quantitative polymerase chain reaction (qPCR) testing for pathogen presence of Mycoplasma agassizii (Myag) or Mycoplasma testudineum (Myte)—both causative agents of URTD in desert tortoises—and TeHV2 (Origgi and others, 2002; Johnson and others, 2005; Wendland and others, 2007; Jacobson and others, 2012; Burgess and others, 2021). ELISA lab results were reported as negative (antibody titer less than 32), suspect (antibody titer greater than or equal to 32 and less than 64), or positive (antibody titer greater than or equal to 64), whereas qPCR lab results were reported as negative, positive, or equivocal (inconclusive) based on cycle threshold values.

Health assessments were done on 393 telemetered tortoises throughout the NTC project area during 2020–22. Most tortoises examined were classified as “clinically normal” and described as adequately conditioned (BCS 4 or BCS 5); however, some were documented as under-conditioned (BCS 3) for muscle and fat reserves in 2022. Most tortoises presented with recessed eyes, likely related to temporary dehydration states that corresponded to the limited rainfall since 2020. A few tortoises showed notable health characteristics, including abnormal beaks, periocular swelling and redness, conjunctival swelling, mucoid ocular discharge, occluded and eroded nares, nasal discharge, active skin lesions, and active shell trauma, although these animals represented less than 6 percent of the assessed population. Tissue samples assessed during 2020–21 yielded positive laboratory results from within the WTATS either for antibodies specific to Myag and Myte (via ELISA testing; n=4 or 3.3 percent of assessed population) or pathogen presence (via qPCR testing; n=6 or 6.7 percent of assessed population) of Myag and Myte.

Baseline Tortoise Mortalities

There were mortalities of study and incidental tortoises in the WTATS and WTA study areas after initial encounters (n=37 or 5 percent of encountered tortoises) during 2020–22. A higher tortoise mortality rate was observed in 2022 (n=14 or 8.7 percent of encountered tortoises) than previous years, likely related to prolonged drought conditions in the southwestern North America (Williams and others, 2022). More male than female tortoise mortalities were recorded (four male, two female, and one unknown); however, the proportion was consistent with observed regional 2:1 sex ratio for the population. Predator controls on ravens (egg-oiling, removal, and so on) were implemented in the Superior-Cronese Critical Habitat Unit and surrounding areas, which have been effective for reducing raven reproduction rates and predation rates on sensitive species like the desert tortoise and sage-grouse (Shields and others, 2019; Xiong, 2020; Holcomb and others, 2021; Sanchez and others, 2021). Suitable site models have been developed to assist managers in identifying areas of predator concern (Shields and others, 2019; Xiong, 2020; Currylow and others, 2021). Preliminary results and observations do not indicate recent high die-off areas in the project area from predation, disease, or climate variability.

Modeling Habitat Site Selection for Recipient and Reference Sites

Guidelines for translocating Mojave Desert tortoises are available as USFWS recovery objectives and in updated translocation protocols (U.S. Fish and Wildlife Service, 1994, 2011, 2020). These guidelines propose that (1) translocated tortoises be placed into recipient sites of suitable tortoise habitat that support all tortoise life stages with no foreseeable habitat development or other effects (for example, increased OHV recreation activity and solar energy development); (2) contain a depleted tortoise population without evidence of a disease outbreak; (3) avoid private land not secured for conservation/mitigation and access limitations; (4) provide recipient sites that have a minimum tortoise dispersal range of 6.5 km (lacking barriers) and no closer than 6.5 km to major unfenced roads or human development; and (5) do not overlap with designated sites where reference tortoises live (reference sites) so that translocation success can be measured by comparing response variables in animals among sites where environmental conditions vary. Based on the guidelines provided by USFWS and consultations with local and regional partners (see the “Site Selection Guidance from the Bureau of Land Management” section), a model was created to inform site selection for recipient and reference sites related to the NTC translocation activities using these guidelines (as much as possible).

The model identifies suitable sites for tortoise translocation into the WTATS. The model uses previous studies on desert tortoise ecology (for example, resource selection, habitat suitability, predators [raven nests], and environmental effects) and knowledge from expert biologists (BLM, NTC, USFWS, and USGS) to define model parameters. Parameters included geospatial and environmental data considered important to the survival and health of tortoise populations, such as habitat suitability, precipitation, raven threats, and several anthropogenic factors (for example, roads, land use, land ownership). The model can be used to run hypothetical scenarios, based on user selected values, that permit investigation of the relative costs and benefits of a variety of potential management actions and scenarios that are not limited to the NTC translocation.

Technological Framework

Variables used for prioritizing potential recipient and reference sites included biological and anthropogenic factors likely to affect desert tortoise populations. The variables, relations between variables, and variable weights used to evaluate the potential of a site were analyzed using a form of ordered weighted averaging (Yager, 1988) to produce a variety of models for evaluation in this translocation plan. Each model consisted of a series of factors thought to have a positive effect on tortoise population success and a second set that were considered to decrease the effectiveness of translocation. The areas proposed for desert tortoise translocation have a weighted value equal to or greater than the mean model value (app. 2). The following seven criteria were selected for our analyses to evaluate suitable translocation sites.

Recipient Site Selection Model Criteria

1. (1) Land ownership—Parcels purchased as mitigation by the NTC were approved by the military and considered as potential recipient sites. Privately held lands, non-Federal lands, State lands, and wilderness/conservation areas were not considered as potential recipient sites.

2. (2) Habitat suitability—Since the previous translocation effort involving the expansion of the NTC (Esque and others, 2005; Heaton and others, 2008b), a desert tortoise habitat model was developed (Nussear and others, 2009; Parandhaman and others, 2022) using desert tortoise presence data (1970–2008) and environmental data (including surface roughness, slope, aspect, bulk density, rockiness, soil depth, precipitation, annual plant potential, and perennial plant cover) to analyze and develop a probability of habitat potential to identify areas of suitable desert tortoise habitat throughout the Mojave Desert and parts of the Sonoran Desert. Nussear and others (2009) is the primary model in use to delineate Mojave Desert tortoise habitat throughout its range since its publication. We modified the model for our use in ranking potential habitat suitability by converting the original 1-km² raster cell size habitat model to the 2.59-km² cell size for our analysis using an area weighted average.

3. (3) Distance to roads—OHV activity and large networks of roads reduce numbers of tortoises and decrease habitat quality (Custer and others, 2017; Averill-Murray and Allison, 2023). We developed a roads layer using BLM roads dataset (consisting of designated OHV roads, dirt roads on public land, and access roads), TIGER/Line 2019 (U.S. Census Bureau, 2019; consisting of access roads and paved roads) and archived USGS Global Positioning System (GPS) tracks (consisting of designated OHV roads and dirt roads that may not have been present on BLM road file) from previous work in the area.

4. (4) Raven nest site density—Common ravens (Corvus corax) have long been considered one of the important threats to tortoise populations throughout the Mojave Desert (Berry, 1986; U.S. Fish and Wildlife Service, 1994; Tracy and others, 2004; Holcomb and others, 2021). We used a raven nest site density model created by Xiong (2020) to predict nesting sites on anthropogenic and natural areas that are associated with evidence of tortoise predation.

5. (5) Connectivity—Tortoise populations may be isolated by a variety of factors, including habitat loss, degradation, and fragmentation, which can result in reduced population-level connectivity and decreased gene flow (Hand and others, 2014; Haddad and others, 2015, Dutcher and others, 2020, Hromada and others, 2020, 2023). The connectivity model incorporated here (Gray and others, 2019) uses tortoise movement data to estimate connectivity across the landscape via a circuit-theoretic approach.

6. (6) Precipitation—Precipitation is essential for tortoise hydration, supports the growth of tortoise herbivorous forage, and must be balanced with food intake for positive nitrogen and energy balances (Medica and others, 1975; Nagy, 1988; Peterson, 1996; Esque and others, 2014). Average winter precipitation was taken from the Parameter Elevation Regression on Independent Slopes Model dataset at 800-m resolution between the months November and February from 2013 to 2018 (Daly and others, 2008; Xiong, 2020; Zylstra and others, 2023). The layer was rescaled with a cubic spline resampling method to a common resolution of 250 m.

7. (7) Terrestrial development index—The terrestrial development index (TDI) was derived from the surface disturbance footprint of terrestrial development for the western United States. This footprint includes urban areas, roads, highways, and agriculture, among other disturbances (Carr and Leinwand, 2020).

Relative Weighting Criteria

We used a form of ordered weighted averaging (Yager, 1988) to multiple scenarios for site selection using input raster layers (criteria) that could be considered beneficial or detrimental to tortoise translocation in a specified area (Malczewski, 2006). In our model, positive effects included desert tortoise habitat suitability (Nussear and others, 2009), desert tortoise movement potential (Gray and others, 2019), and average winter precipitation. Negative effects included raven nest density (considering anthropogenic and natural nest densities; Xiong, 2020), distance to roads (including paved or dirt public and BLM designated trails and primitive roads and private roads), and TDI (a measure of the cumulative anthropogenic effects within a 1-km window; Carr and others, 2017; Carr and Leinwand, 2020; Carter and others, 2020). Each of the criteria were rescaled from 0 to 1 for analysis. We calculated variance inflation factors (VIF) for the set of criterions used in the model to exclude highly correlated variables (VIF is greater than 3) through a stepwise procedure using the vifstep() function in the R package usdm (Naimi and others, 2014; R Core Team, 2022).

Each of the layers were manipulated in two ways. First, layers were standardized to a range between 0 and 1 and then weighted within that range to indicate the relative effect/weight (in other words, importance; w_i in equations [eqs.] 1 and 2) in the overall model. For example, habitat suitability may have a large effect (w_i=1), TDI as a small effect (w_i=0.2), and precipitation as another large effect (w_i=0.8) on the overall model. Next, the layers were scaled (q_i in eqs. 1 and 2) by parameters that adjusted the values of the raster via a beta probability distribution value between 0 and 1. Linear or nonlinear scaling of each raster can be implemented by changing the two shape parameters (α, β) of the beta probability distribution (via the pBETA() function in the R package fitODBOD [v1.5.0]; Mahendran and Wijekoon, 2019; R Core Team, 2022), where the shape parameters were allowed to vary between 1 and 5 (fig. 5). For example, increasing the scale on lower bound values such that α=4 increases the probability of lower values in the weighted raster, resulting in higher values having less consideration because the upper bound remains unscaled (β=1; fig. 5B). Each of the weighted layers was then multiplied by the respective weighting curve. The positive and negative effects on desert tortoise populations were each summed and scaled from 0 to 1, and then negative effects were subtracted from the positive effects (p_i and k_i in eq. 3), yielding a final weighted layer used as the model for consideration.

Positive effect

$$p_i={\displaystyle {\sum }_{i=1}^n{q}_i}\times w_i$$

(1)

where

p_i

is the score of positive effect of each weighted raster,

n

is the total sum of rasters available,

i

is the available raster starting at the first raster to the n^th,

w_i

is the associated weight for each raster, and

q_i

is the probabilistic weighting function based on a vector of probability density function.

Negative effect

$$p_i={\displaystyle {\sum }_{i=1}^n{q}_i}\times w_i$$

(2)

where

k_i

is the score of negative effect of each weighted raster,

n

is the total sum of rasters available,

i

is the available raster starting at the first raster to the n^th,

w_i

is the associated weight for each raster, and

q_i

is the probabilistic weighting function based on a vector of probability density function.

Suitability probability

$S={\displaystyle {\sum }_{i=1}^n\frac{p_i}{\max \left(p_i\right)}}-{\displaystyle {\sum }_{i=1}^n\frac{k_i}{\max \left(k_i\right)}}$

(3)

where

S

is suitability probability,

n

is the total sum of rasters available,

i

is the available raster starting at the first raster to the n^th,

p_i

is the score of positive effect raster,

max(p_i)

is the maximum score of positive effect rasters,

k_i

is the score of negative effect raster, and

max(k_i)

is the maximum score of negative effect rasters.

Site Visitation

Potential recipient and reference sites were visited by authors of this report and USGS staff members extensively from spring 2020 to fall 2022. Representative digital photographs were recorded at the center of each proposed recipient and reference site or grouped sites in fall 2022 (app. 3). Field crew members visited each site on BLM designated open routes. By visiting the sites, we concluded that some were unsuitable for tortoise translocation because of excessive OHV use or other anthropogenic effects (for example, private property, radio tower access, and utility corridors). Those sites were disqualified as potential translocation sites. Selected recipient and reference areas are described later in the text and were characterized by typical desert tortoise habitat in mixed shrub communities mostly dominated by Larrea tridentata and Ambrosia dumosa (Nussear and Tuberville, 2014). The BLM, NTC, and non-government organizations have cooperated to make substantial investments in habitat restoration throughout large parts of the general site and in reducing road incursions, leaving access on a network of designated roads.

Translocation Site 1

• Recipient site R1—Easily accessible from a designated two-track dirt road (fig. 6; app. 3—site photographs). There was no evidence of recent (since 2020) unauthorized OHV use at the site center; however, several designated BLM roads and established campsites are in the general area closer to the dry lake and east of the recipient site. The center of this recipient site is on a gentle hill that slopes into a wide, flat, and open expanse to the northeast. Medium-sized rolling hills of moderate slope are to the southwest. The soil is soft, sandy loam topped with gravel composite and suitable for tortoise burrows. Small mammal burrows were present in high density throughout the site. Vegetation consisted primarily of Larrea tridentata, Ambrosia dumosa, and Ephedra californica (and Ephedra nevadensis (California and Nevada jointfir; respectively) as well as several other less dominant species, such as Acamptopappus sphaerocephalus (rayless goldenhead) and Thamnosma montana (turpentine broom).

• Recipient sites R2a and R2b—Accessible from designated two-track dirt roads and are 1.5–3 km south of a dry lake (fig. 6; app. 3—site photographs). Like R1, designated BLM roads and established campsites are concentrated further north from R2a and R2b, toward the dry lake. Both sites are generally in flat areas that gently slope down in their northern sections toward the dry lake. At these sites, sandy soil is interspersed by rocks. The dominant vegetation is comprised of Larrea tridentata, Ephedra californica, Ephedra nevadensis, and multiple Atriplex species (saltbushes).

• Recipient sites R3a and R3b—Are from 3 to 6 km south of the southern fenced WTA border and off two-track dirt roads just west of the graded Copper City Road (fig. 6; app. 3—site photographs). There are more trafficked designated two-track dirt roads that skirt the boundaries of R3a and R3b. These sites are comprised of low hills with sandy soil containing some gravel. Vegetation is dominated by Larrea tridentata and Ambrosia dumosa. Yucca brevifolia (Western Joshua trees) are present but more numerous in the southern region of R3a at higher elevation.

Translocation Site 2

• Recipient sites R7a and R7b—Easily accessible from a dirt BLM road from the north and west with only moderate OHV use noted (fig. 6; app. 3—site photographs). The sites were approximately 3 km from a major paved tortoise-fenced road (Fort Irwin Road). Private properties with trailers are east of the recipient sites, but within the translocation site on the northeast boundary, just off Fort Irwin Road. A private property with dozens of trailers in the recipient site area is closer to Fort Irwin Road. The site is surrounded by mountains with moderate eastward facing slopes. Soil is characterized as sandy, gravelly loam. The shrub community is dominated by small Larrea tridentata (most less than or equal to 1 m tall) and Ambrosia dumosa. Annual vegetation from the previous year was present on the landscape.

• Recipient sites R8a and R8b—Are within 1.5–4 km of the WTA to the north (fig. 6; app. 3—site photographs). Between these sites is two-track Paradise Valley Road, which connects Fort Irwin Road to the gated southern edge entrance to the WTA. The sites are moderately sloped from mountains to the west and east, consisting of semi-rocky and sandy soil with outcrops of fine-grained consolidated sediments in the north. R8a and R8b contain the densest and tallest vegetation (Larrea tridentata and Ambrosia dumosa) of all the recipient sites.

Translocation Site 3

• Recipient sites R4a and R4b—Accessible via an unmarked two-track road, approximately 3 km from a primary transmission utility corridor to the south and approximately 300 m up a gentle slope (fig. 6; app. 3—site photographs). Although there are marked BLM roads south of these sites, minimal to no OHV disturbance was observed in these recipient areas. These sites are east of the Alvord Mountain Range and west of a plateau with a radio tower, which is approximately 60 m from the site center. The sites are typified by low gravelly and sandy hills with outcrops of fine-grained consolidated sediments and several moderately deep (2–5 m) washes. R4a and R4b are dominated by mixed Larrea tridentata and Lycium cooperi (peach thorn) as well as Ambrosia dumosa and Senna armata (desert senna). Vegetation at this site was sparser than most other recipient sites.

• Recipient sites R5a and R5b—Accessible from a two-track road off graded Manix Trail Road, which is used by the NTC to transport military equipment to and from the southern NTC border and the I–15 (fig. 6; app. 3—site photographs). There was only one OHV trail running through R5a, with other trails ending just to the southwest of these sites. R5a and R5b are west of the Alvord Mountain Range and northeast of Coyote Lake (approximately 4 km) and are characterized by low hills. Soil is mostly sandy, littered with surface rocks near the bajada to the south and east, and dense volcanic gravel covers the hillsides. Vegetation primarily is Larrea tridentata, Ambrosia dumosa, and Senna armata.

• Recipient sites R6a and R6b—Are just south of the Alvord Mountain Range and north of a primary utility transmission corridor (fig. 6; app. 3). Additionally, the Old Spanish National Historic Trail is marked on the west side of the sites. The sites are on low hills and generally slope down to the south. The soil is very sandy with relatively sparse vegetation on the southern end of the site. Dominant vegetation included Larrea tridentata, Ambrosia dumosa, and Senna armata, but vegetation is the sparsest of all the recipient sites.

Reference Sites C1 and C2

The potential reference sites (C1 and C2) are north of the Interstate Highway 15 corridor. Reference site C1 is north of Barstow, Reference site C2 is east of Barstow, and they are separated by Fort Irwin Road and south of all the Recipient Sites (fig. 6; app. 3—site photographs). However, the reference sites also stretch northwest, north, and northeast from Barstow and areas within them are as far or farther (approximately 26 km at furthest point) from the cities than the release sites (fig. 6; app. 3—site photographs). C1 contains the more private land holdings to the south, but also Black Mountain Wilderness (BLM), BLM recreation areas (Rainbow Basin natural area, Owl Canyon campground) to the southeast, and two graded dirt roads (Fossil Bed Road and Copper City Road). C1 has variable terrain, soil, and vegetation, areas with larger hills and canyons, and rockier soils and denser Larrea tridentata, Ambrosia dumosa, and Yucca brevifolia in the north. The southern part of C1 has smaller rolling hills, sandier soil, and sparser vegetation. C2 is bordered by the tortoise-fenced I–15 highway to the south and has more private properties and motorized recreation areas in the south and west. C2 also encompasses the Calico Mountains and is southwest of Coyote Lake (unsuitable for tortoises; fig. 7). In C2, soil is coarse, sandy loam with a mixed shrub Larrea tridentata and Ambrosia dumosa community among large hills and canyons, turning to medium grade slopes to the north and south.

Tortoise Density Estimates

Populations of reptiles, such as desert tortoises, are most efficiently surveyed with spatially structured transects or spatially unstructured area searches (Allison and McLuckie, 2018; Mitchell and others, 2021b; Royle and Turner, 2022; Zylstra and others, 2023). To produce reptile population density and abundance estimates, detection data from transect surveys are typically analyzed with distance sampling models, whereas detection data from area/plot searches are typically analyzed with nonspatial capture-recapture models. However, many reptiles have characteristics that present challenges when attempting to use those models to estimate density and abundance. For example, conventional line-distance sampling models assume that detection of individuals of the focal species that are directly on the transect line (that is, distance equals 0 from the line) are all found (that is, perfect detection). However, some species, such as desert tortoises, violate these assumptions because a portion of individuals likely are in burrows and not visible to observers (detectors) when a given transect is surveyed (Allison and McLuckie, 2018). In contrast, spatial capture-recapture (SCR) models overcome many of those issues by incorporating the spatiotemporal information about survey effort and the locations where individual animals were detected in estimations. These data are accommodated in SCR models with a spatially explicit observation submodel and an ecological submodel that describe animal distribution (density) as a realized Poisson point process (Efford, 2004; Borchers and Efford, 2008; Royle and others, 2014). All spatial information collected on Global Positioning Systems and used in mapping and analyses use the North American Datum of 1983 (NAD 83).

Spatial Capture-Recapture and Survey Results

Movements and Detection Rates

The number of survey occasions ranged from 28 to 55 days, depending on season, year, and study area (table 4). Mean search effort per grid cell ranged from 247 meters per cell (m/cell) during spring 2022 at WTA to 1,625 m/cell during spring 2020 at the WTATS-West. The average number of tortoises detected during a given survey in a certain study area was 117 individuals (range: 52–180), and an average of 6 individual tortoises were detected per day. The average number of recaptures obtained during a given survey in a certain study area was 108 (range: 6–266), whereas the average number of spatial recaptures (in other words, tortoise detected in greater than 1 grid cell) obtained during a given survey in a certain study area was 43 (range: 2–143).

Table 4.    

Study area specific detection results and survey design metrics for ground-based surveys of adult Mojave Desert tortoises at Fort Irwin, California.

[m, meter; WTATS, Western Training Area Translocation Site]

Table 4.    Study area specific detection results and survey design metrics for ground-based surveys of adult Mojave Desert tortoises at Fort Irwin, California.

Year	Season	Occasions (days)	Tortoises detected	Recaptures	Spatial recaptures	Cell spacing (m)	Mean search effort (m/cell)
Western Training Area
2021	Spring	55	52	36	10	200	1,117
2021	Fall	52	107	175	92	374	1,274
2022	Spring	28	111	22	11	320	247
WTATS-East
2021	Fall	52	57	58	10	374	1,304
2022	Spring	28	131	123	24	320	729
WTATS-West
2020	Spring	40	96	266	143	320	1,625
2020	Fall	31	122	154	64	374	1,153
2021	Spring	55	180	6	2	320	1,180
2021	Fall	52	156	152	51	374	732
2022	Spring	28	153	85	22	320	318

Male tortoises tended to have substantially larger σ estimates and, therefore, larger range sizes than females, whereas females tended to have higher λ₀ estimates and, therefore, higher detection probabilities than males. Mean σ across all 7 of the analyses in which sex-varying σ was present in the top-ranked model were 211.14 m and 149.74 m for males and females, respectively. Assuming home ranges were bivariate normally distributed (in other words, approximately circular; Royle and others, 2014; Sun and others, 2015), those σ values corresponded to average seasonal range sizes of 0.84 km² and 0.42 km² for males and females, respectively. Among the four analyses in which sex-varying λ₀ was present in the top-ranked model, mean λ₀ was 0.09 and 0.16 for males and females, respectively. In contrast, among the six analyses in which the top-ranked model indicated that λ₀ did not differ between sexes, the mean λ₀ was 0.15. Across all 10 analyses, each area was estimated to have a male-biased sex ratio within each season by year, ranging from 53 percent to 71 percent males versus from 29 percent to 47 percent females. For the NTC across all areas, seasons, and years, the mean sex ratio was 64 percent males versus 36 percent females.

Tortoise density spatially varying as a function of habitat suitability was included in the top-ranked model for three of the analyses, and all three of those estimated relations were positive (tables 5–7; fig. 8). Among the seven analyses in which a density-habitat relationship was not present in the top-ranked model, a competing model (∆AIC_c less than 2) contained that relationship in six of those analyses, indicating that was supported for nearly all areas within each season across years. However, in one case (WTA during spring 2021), the competing model with the density-habitat suitability relationship had a coefficient estimate with 95-percent confidence interval (CI) that overlapped zero, a nominal change in model log-likelihood relative to the top-ranked model, and the same model weight as the top-ranked model, all of which indicated that the habitat suitability covariate was uninformative for that particular dataset (Arnold, 2010). The completely null model (in other words, spatially random density, λ₀ and σ shared between sexes) was the top-ranked model for three of the analyses, which were also the three datasets with the fewest total number of recaptures (WTA spring 2021, WTA spring 2022, and WTATS-West spring 2021).

Table 5.    

Spatial capture-recapture model selection results for the Western Training Area (WTA) study area in each season by year combination from surveys of Mojave Desert tortoises at Fort Irwin, California (2020–22).

[Estimated model parameters were density (D), baseline detection rate (λ₀), and spatial scale of detection (σ). We considered models in which tortoise density was spatially random (approximately [~] 1) or density varied spatially as a log-linear function of habitat suitability that estimated from a previous analysis (habitat; Nussear and others, 2009), and we allowed λ₀ and σ to differ between sexes (Sex) or be shared between sexes (about 1). Abbreviations: AIC_c Akaike's Information Criterion corrected for small sample size; ∆AIC_c, difference between AIC_c value of the model and AIC_c value of the top-ranked model]

Table 5.    Spatial capture-recapture model selection results for the Western Training Area (WTA) study area in each season by year combination from surveys of Mojave Desert tortoises at Fort Irwin, California (2020–22).

Model	Number of model parameters	Log-likelihood of model	AICc	∆AICc	Model weight
Spring 2021
D~1 λ0~1 σ~1	4	−349.92	711.17	0	0.35
D~habitat λ0~1 σ~1	5	−351.17	711.21	0.04	0.35
D~1 λ0~Sex σ~1	5	−350.77	712.87	1.7	0.15
D~1 λ0~1 σ~Sex	5	−351.11	713.55	2.39	0.11
D~1 λ0~Sex σ~Sex	6	−350.73	715.37	4.21	0.04
Fall 2021
D~1 λ0~1 σ~Sex	5	−860.66	1,731.93	0	0.48
D~habitat λ0~1 σ~Sex	6	−860.1	1,733.06	1.13	0.27
D~1 λ0~Sex σ~Sex	6	−860.47	1,733.79	1.86	0.19
D~1 λ0~1 σ~1	4	−864.4	1,737.21	5.27	0.03
D~1 λ0~Sex σ~1	5	−863.75	1,738.1	6.17	0.02
Spring 2022
D~1 λ0~1 σ~1	4	−542.61	1,093.6	0	0.33
D~habitat λ0~1 σ~1	5	−541.76	1,094.11	0.51	0.26
D~1 λ0~Sex σ~1	5	−542.28	1,095.15	1.55	0.15
D~1 λ0~Sex σ~Sex	6	−541.32	1,095.46	1.86	0.13
D~1 λ0~1 σ~Sex	5	−542.51	1,095.6	2	0.12

Table 6.    

Spatial capture-recapture model selection results for the Western Training Area Translocation Site (WTATS)-East study area in each season by year combination from surveys of Mojave Desert tortoises at Fort Irwin, California (2021–22).

[Estimated model parameters were density (D), baseline detection rate (λ₀), and spatial scale of detection (σ). We considered models in which tortoise density was spatially random (approximately [~] 1) or density varied spatially as a log-linear function of habitat suitability estimated from a previous analysis (habitat; Nussear and others, 2009), and we allowed λ₀ and σ to differ between sexes (Sex) or be shared between sexes (~1). Abbreviations: AIC_c Akaike's Information Criterion corrected for small sample size; ∆AIC_c, difference between AIC_c value of the model and AIC_c value of the top-ranked model]

Table 6.    Spatial capture-recapture model selection results for the Western Training Area Translocation Site (WTATS)-East study area in each season by year combination from surveys of Mojave Desert tortoises at Fort Irwin, California (2021–22).

Model	Number of model parameters	Log-likelihood of model	AICc	∆AICc	Model weight
Fall 2021
D~habitat λ0~1 σ~Sex	6	−385.3	784.29	0	0.82
D~1 λ0~1 σ~Sex	5	−388.93	789.04	4.75	0.08
D~1 λ0~1 σ~1	4	−390.43	789.63	5.34	0.06
D~1 λ0~Sex σ~Sex	6	−388.85	791.38	7.09	0.02
D~1 λ0~Sex σ~1	5	−390.2	791.57	7.28	0.02
Spring 2022
D~habitat λ0~Sex σ~Sex	7	−920.97	1,856.87	0	0.95
D~1 λ0~Sex σ~Sex	6	−925.14	1,862.96	6.09	0.05
D~1 λ0~1 σ~Sex	5	−932.2	1,874.88	18.01	0
D~1 λ0~1 σ~1	4	−944.99	1,898.32	41.45	0
D~1 λ0~Sex σ~1	5	−944.81	1,900.12	43.25	0

Table 7.    

Spatial capture-recapture model selection results for the Western Training Area Translocation Site (WTATS)-West study area in each season by year combination from surveys of Mojave Desert tortoises at Fort Irwin, California (2020–22).

[Estimated model parameters were density (D), baseline detection rate (λ₀), and spatial scale of detection (σ). We considered models in which tortoise density was spatially random (approximately [~] 1) or density varied spatially as a log-linear function of habitat suitability that was estimated from a previous habitat suitability analysis (habitat; Nussear and others, 2009), and we allowed λ₀ and σ to differ between sexes (Sex) or be shared between sexes (~1). Abbreviations: AIC_c Akaike's Information Criterion corrected for small sample size; ∆AIC_c, difference between AIC_c value of the model and AIC_c value of the top-ranked model]

Table 7.    Spatial capture-recapture model selection results for the Western Training Area Translocation Site (WTATS)-West study area in each season by year combination from surveys of Mojave Desert tortoises at Fort Irwin, California (2020–22).

Model	Number of model parameters	Log-likelihood of model	AICc	∆AICc	Model weight
Spring 2020
D~habitat λ0~Sex σ~Sex	7	−1,124.47	2,264.24	0	0.75
D~1 λ0~Sex σ~Sex	6	−1,127.21	2,267.39	3.15	0.16
D~1 λ0~1 σ~Sex	5	−1,129.25	2,269.18	4.94	0.06
D~1 λ0~1 σ~1	4	−1,131.37	2,271.18	6.94	0.02
D~1 λ0~Sex σ~1	5	−1,131.35	2,273.39	9.15	0.01
Fall 2020
D~1 λ0~Sex σ~Sex	6	−924.7	1,862.13	0	0.71
D~habitat λ0~Sex σ~Sex	7	−924.49	1,863.97	1.84	0.29
D~1 λ0~1 σ~Sex	5	−931.92	1,874.36	12.23	0
D~1 λ0~1 σ~1	4	−946.7	1,901.74	39.61	0
D~1 λ0~Sex σ~1	5	−946.67	1,903.86	41.72	0
Spring 2021
D~1 λ0~1 σ~1	4	−856.95	1,722.13	0	0.41
D~1 λ0~1 σ~Sex	5	−856.72	1,723.78	1.65	0.18
D~1 λ0~Sex σ~1	5	−856.75	1,723.84	1.71	0.17
D~habitat λ0~1 σ~1	5	−856.75	1,723.85	1.72	0.17
D~1 λ0~Sex σ~Sex	6	−856.67	1,725.83	3.7	0.06
Fall 2021
D~1 λ0~Sex σ~Sex	6	−1,161.56	2,335.68	0	0.49
D~1 λ0~Sex σ~1	5	−1,163.41	2,337.22	1.54	0.23
D~habitat λ0~Sex σ~Sex	7	−1,161.51	2,337.77	2.09	0.17
D~1 λ0~1 σ~Sex	5	−1,164.19	2,338.78	3.1	0.1
D~1 λ0~1 σ~1	4	−1,168.21	2,344.68	9	0.01
Spring 2022
D~1 λ0~1 σ~Sex	5	−961.09	1,932.59	0	0.55
D~1 λ0~Sex σ~Sex	6	−960.97	1,934.51	1.92	0.21
D~habitat λ0~1 σ~Sex	6	−960.97	1,934.52	1.93	0.21
D~1 λ0~Sex σ~1	5	−964.42	1,939.24	6.65	0.02
D~1 λ0~1 σ~1	4	−966.29	1,940.85	8.26	0.01

Density Estimates for the Western Training Area Translocation Site and Western Training Area

Accounting for the spatiotemporally varying survey effort (meters searched/grid cell/occasion) resulted in, on average, 16-percent increases in mean density estimates. Point estimates of mean density ranged from 0.27 to 1.85 adult tortoises per square kilometer (tortoises/km²), with an average for NTC across all 10 area-by-season-by-year estimates of 0.95 adult tortoises/km² (fig. 9). All the SCR density estimates are within the range of densities predicted for the Superior-Cronese Tortoise Conservation Area (range: 0.24–3.99) for a similar timeframe (2020; Zylstra and others, 2023). Study area-specific mean densities, averaged across seasons and years, were 1.08, 0.51, and 0.95 adult tortoises/km² at WTA, WTATS-East, and WTATS-West, respectively. Density estimates were generally lower during the fall season than the spring season, differing by as much as 105 percent between seasons within a study area, and density estimate precision (coefficient of variation; CV) ranged from 0.10 to 0.19, with a mean of 0.14 across all 10 area-by-season-by-year estimates.

• Post-hoc analyses—Estimated density increased over time in all three study areas such that the derived average seasonal population growth rates across the duration of sampling were 1.52 (95-percent CI: 1.19, 1.77), 1.32 (95-percent CI: 1.07, 1.64), and 1.55 (95-percent CI: 1.48, 1.63) at WTA, WTATS-East, and WTATS-West, respectively. Results from our Gamma GLMs indicated a strong positive relationship between tortoise density and number of tortoises detected (β=0.29; 95-percent CI: 0.1, 0.48; p=0.003), whereas strong negative relationships existed between tortoise density and numbers of recaptures and spatial recaptures (β_Recaps=−0.47; 95-percent CI: −0.65, −0.29; p<0.0001; β_SpatRecaps=−0.42; 95-percent CI: −0.64, −0.20; p=0.0002). Density estimates were invariant to the number of survey occasions and the study area (state space) sizes (β_Occasion=−0.06; 95-percent CI: −0.29, 0.16; p=0.57; β_Area=−0.002; 95-percent CI: −0.26, 0.25; p=0.99).

Predicted Densities for Western Training Area Translocation Sites

We predicted mean spatial tortoise densities for each translocation site by converting the Habitat Suitability Index raster (Nussear and others, 2009) into spatially explicit densities using coefficient estimates from the top-ranked SCR-Habitat Suitability Index models for season by year combinations. The SCR model-predicted density surfaces were created by using the ArcMap Raster Calculator function and using the following conversion equation for log-linear relations:

Density (adults/km

²

) =

exp(β

_Density

±β

_Covariate

×Covariate Raster)×100

(5)

where

β_Density

is the coefficient estimate for the density parameter, and

β_Covariate

is the coefficient estimate for the covariate parameter.

1. a. Spring 2020 WTATS-West:

   exp(−8.3568271+3.0753677דHabitat Suitability Index”)×100

2. b. Fall 2021 WTATS-East:

   exp(−8.5557512+3.4784546דHabitat Suitability Index”)×100

3. c. Spring 2022 WTATS-East:

   exp(−7.0513422+2.3790902דHabitat Suitability Index”)×100

New rasters of the mean cell values (densities) among the above three predicted density surfaces were produced for each Translocation Site (TS1, TS2, and TS3) using ArcMap’s Cell Statistics function (fig. 10). We clipped the resulting mean values raster to each translocation site to obtain site-specific means, SEs, and 95-percent CIs.

• • TS1 (355 cells): Mean=0.47 adults/km²; SE=0.0114; 95-percent CI=0.46–0.48

• • TS2 (350 cells): Mean=0.43 adults/km²; SE=0.0126; 95-percent CI=0.42–0.44

• • TS3 (178 cells): Mean=0.41 adults/km²; SE=0.0220; 95-percent CI=0.39–0.43

Post-Translocation Density Estimates for Western Training Area Translocation Sites

The mean estimated adult tortoise density at WTA, averaged among estimates produced during 2021–22, was 1.08 adults/km² (95-percent CI: 0.44–1.73), which corresponds to 273 live adult tortoises (greater than 180 midline carapace length [MCL]) in the WTA (95-percent CI: 111–438 adults) to be translocated to the WTATS. The density of tortoises in the translocation sites will not exceed 1 standard deviation above the mean density for that area (U.S. Fish and Wildlife Service, 2020). The estimated threshold for the Superior-Cronese is a density of 3.9 adult tortoises/km² and was calculated from the USFWS range wide monitoring program (Allison and McLuckie, 2018; U.S. Fish and Wildlife Service, 2020). Based on field data from 2020 to 2022, the estimated densities for WTATS-East, WTATS-West, combined, and within each translocation site are below the tortoise density threshold. Translocation into each site is estimated to increase local densities while not exceeding the threshold (table 8).

Table 8.    

Adult tortoise density estimates before and after translocation (includes release and surrounding areas to which tortoises are expected to disperse).

[Post-translocation density and abundance are estimated based on a range of potential translocated animals to each site (from 0 to 438 animals; the upper confidence interval of estimated number of adult tortoises). Estimated density and number of adult tortoises are based on data collected from April 11, 2020, to September 12, 2022. Abbreviations: km², square kilometer; #, number; WTA, Western Training Area; WTATS, Western Training Area Translocation Site; —, not applicable]

Table 8.    Adult tortoise density estimates before and after translocation (includes release and surrounding areas to which tortoises are expected to disperse).

Area	Area type	Area size (km2)	Mean density during 2020–22 (# adult/km2)	Estimated # of adult tortoises	Post-translocation density (# tortoises/km2)1	Estimated post-translocation # of tortoises
WTA
WTA	Western Training Area	253	1.08	111–438	0	0
			(0.44–1.73)
WTATS-West
TS1	Translocation Site	330	0.47	164	0.46–1.83	164–603
			(0.46–0.48)
TS2	Translocation Site	159	0.43	64	0.42–3.16	64–503
			(0.42–0.44)
WTATS-East
TS3	Translocation Site	293	0.41	—	0.39–1.92	123–562
			(0.39–0.43)	123

¹

Post-translocation density estimates and abundances for the WTA are based on clearance of all tortoises from the area.

Tortoise Clearance Protocols for the Western Training Area

The procedures in this section and the three major sections “Tortoise Disposition Plan and Translocation Package,” “Translocation of Tortoises from the Western Training Area,” and “Post-Translocation Monitoring: Short and Long-Term Success Criteria” describe recommendations from the USFWS translocation guidance (U.S. Fish and Wildlife Service, 2020). The procedures are meant to inform the implementation of the U.S. Fish and Wildlife Service guidelines as USGS understands them, based on the scientific expertise, data, and experience of the USGS. Recommendations are not made by the USGS. Any dates are based on our understanding of the NTC’s proposed project timeline.

Tortoise clearance protocols include all activities to prepare for and implement the removal of tortoises from harm’s way as a result of a project in tortoise habitat, including (1) although not holding tortoises in pens is preferred, if required due to injury, disease, or seasonal timing of discovery, outdoor predator-proof pens may be necessary as temporary housing; (2) fencing the project area boundaries to prevent tortoise movement in and out of the area; (3) doing clearance surveys to find and attach radio transmitters to all tortoises when appropriate in the project area to monitor or place in enclosures; (4) doing health assessments and analyzing samples on all tortoises to be translocated; and (5) translocation of all tortoises from the project area to designated approved release sites and burrow crushing as outlined by the U.S. Fish and Wildlife Service (2009, 2020). Further details on clearance procedures are provided in the next sections.

To comply with timelines and USFWS guidelines, by post-translocation winter year 0 (app. 1), we suggest that (1) complete records of all tortoises found within the WTA after doing clearance surveys, along with information collected upon encounters (for example, attached unique identifier, radio transmitter, location), be collated; (2) health screenings be completed for all tortoises in the WTA, as well as for select resident and reference tortoises in the WTATS; (3) translocation release plans and landscape radio frequency plans be written; and (4) all tortoise exclusionary fence work be completed, including at tortoise containment facilities (app. 1). To acquire and compile baseline data on habitats and resident tortoises before translocations, as recommended by U.S. Fish and Wildlife Service (2020), surveys of recipient sites and tortoise monitoring (for example, home range, density, health) may likely continue through post-translocation spring year 0. During this time, tortoise exclusionary fencing around the WTA can be installed, enclosures can be constructed, and tortoises housed in enclosures cared for and monitored. We suggest that an annual meeting with regional project partners and land managers be done in preparation for translocation.

Before translocation of tortoises from the WTA, USFWS, BLM, and CDFW may require the NTC to coordinate with them to finalize the selection of recipient sites within the WTATS and inter-agency agreements regarding translocation. In addition, it would be prudent if the NTC coordinated with any entities doing ongoing research in the project area in preparation for translocation. All activities related to desert tortoises (capture, handling, and translocation) during clearance surveys will be done in accordance with the USFWS 2021 Biological Opinion (U.S. Fish and Wildlife Service, 2021a). Fort Irwin’s USFWS 10(a)(1)(A) permit will address post translocation monitoring. State permits will be obtained before translocation.

Tortoise Health Assessments, Tissue Sampling, and Laboratory Diagnostics

In preparation for translocation, assessments of clinical health conditions and physiological health status in this section will follow methods detailed in a handbook titled “Health Assessment Procedures for the Desert Tortoise (Gopherus agassizii)—A Handbook Pertinent to Translocation” (U.S. Fish and Wildlife Service, 2019). Tortoises will not be eligible for translocation if health conditions show signs that may affect survival, including weakness or lethargy, moderate to severe serous or mild to severe mucoid nasal discharge, or crusts, plaques, or ulcers in the mouth (U.S. Fish and Wildlife Service, 2020). Tortoises not eligible for translocation will be held in containment enclosures and will be cared for with the protocol outlined in a tortoise husbandry plan and may be used during future or other research activities (as approved by USFWS; see the “Tortoise Enclosures” section). Tortoises with improved health may be eligible for translocation (case by case evaluation, approved by USFWS) in alternative suitable sites or in coordination with Federal and State agencies.

Before translocation, health assessments must be completed: (1) one within 1 year of translocation and (2) at least two that are 14–30 days apart, with the last assessment immediately before the translocation date (U.S. Fish and Wildlife Service, 2020). Biological tissue samples, including blood and oral epithelial cells (see next section), must be collected within 1 year of translocation (U.S. Fish and Wildlife Service, 2020). Additional health assessments will be done on a subset of animals in the resident and reference populations, with a target sample subset size estimated as those needed to detect 10-percent prevalence at the 95-percent CI and 5-percent precision (U.S. Fish and Wildlife Service, 2020).

As part of the physical health assessment, general health signs will be described, including the animal's general posture, respiration, face (with specific attention to the eyes, periocular tissue, nares, mouth, tongue, and oral mucosa), skin, and shell for any clinical signs of disease, abnormalities, damage, or discoloration (U.S. Fish and Wildlife Service, 2020). The cloaca, eyes, nares, mouth, and skin will be examined for any evidence of lesions, ulceration, erythema, swelling, or discharge and will be noted. Numerical BCS will be used to assess the overall muscle condition and fat stores with respect to skeletal features of the head and limbs. The BCS scores are first categorized as “under,” “adequate,” or “over” condition, and then numerical values are assigned to provide a precise and repeatable measurement (for example, under: 1–3, adequate: 4–6, over: 7–9; U.S. Fish and Wildlife Service, 2019). Tortoises that are eligible for translocation should have a normal attitude and respiration, have a BCS greater than 4, present no evidence of active lesions (shell and oral) or mucoid discharge (ocular and nasal), and no other health condition that may affect their survival (U.S. Fish and Wildlife Service, 2020; fig. 11).

Immediately after a physical assessment, tissues will be collected from each animal, when applicable (U.S. Fish and Wildlife Service, 2020). Other tissues may be collected, as needed, for associated research or monitoring purposes. Protocol for shipping samples will follow USFWS Health Assessment (U.S. Fish and Wildlife Service, 2019) procedures. Aliquots of plasma will be shipped on dry ice to the Mycoplasma Laboratory at the University of Florida (Gainesville, Florida) and screened for targeted immune responses (antibodies) specific to Mycoplasma agassizii (hereafter Myag) and M. testudineum (Myte), using an ELISA (measuring immunoglobulin M [IgM] and IgY light chains; Wendland and others, 2007; U.S. Fish and Wildlife Service, 2019). Results typically are reported from ELISA as negative (antibody titer less than 32), suspect (antibody titer greater than or equal to 32 and less than 64), or positive (antibody titer greater than or equal to 64). The associated absorbance (A405) values for each ELISA result may also be evaluated to better understand immune responses to Mycoplasma spp. within tortoise populations.

Sloughed epithelial cells from inside the buccal area will be collected using oral swabs (U.S. Fish and Wildlife Service, 2019). One oral swab from each sampling encounter will be shipped on dry ice to the San Diego Zoo Amphibian Disease Laboratory (Escondido, California) to detect and estimate the abundance of Myag, Myte, and TeHV2 deoxyribonucleic acid (DNA) present in the sample using qPCR (Braun and others, 2014; U.S. Fish and Wildlife Service, 2019). Results for all qPCR tests will be reported as negative, positive, or equivocal (inconclusive) based on cycle threshold (Ct) values as indicated by USFWS guidance (U.S. Fish and Wildlife Service, 2019) and experience of the USGS. The USGS suggests requesting Ct values and plasmid counts for each sample evaluated to better understand pathogen presence and pathogen load within tortoise populations. All remaining tissue samples that were collected will be stored in ultracold freezer storage (−70 °C) or other conditions as appropriate.

Priority attention will be given to assessment and sample quality, collection, processing, and care during storage, shipping, and understanding of associated results for all health-related work. All measures needed to reduce disease and pathogen transmission between tortoises and populations will be taken (U.S. Fish and Wildlife Service, 2019). All tortoises that void bladder contents will be re-hydrated using permitted methods such as soaking, nasal-oral uptake, or epicoelomic injections (U.S. Fish and Wildlife Service, 2019; see the “Monitoring of Tortoises via Very High Frequency Telemetry or Similar Technology” section).

Translocation of tortoises will focus on minimizing risk to the tortoise population, especially relative to disease transmission. Prevalence of M. agassizii can be as high as 50–90 percent in healthy populations and reveal no signs of poor body condition indices or signs of URTD that would result in an ineligible status for translocation (Sandmeier and others, 2017, 2018; Weitzman and others, 2017). Translocation of tortoises into recipient sites will maintain levels of M. agassizii and ELISA-positivity for the recipient population based on baseline health assessments (pre-translocation) to maintain disease resilient populations (U.S. Fish and Wildlife Service, 2020).

Tortoise Disposition Plan and Translocation Package

After the tortoise clearance procedure is complete and before translocation, the NTC will coordinate with USFWS, CDFW, and BLM to finalize a tortoise disposition plan and a translocation package for all tortoises found in the WTA (U.S. Fish and Wildlife Service, 2020). Per USFWS, the tortoise disposition plan would include a step-by-step plan describing preparations for tortoises that will be translocated or temporarily housed in enclosures (including juveniles) in addition to highlighting translocation recommendations for each tortoise based on prior health assessments, lab results, and conditions of the habitat in which they were found. The plan will specify locations (UTMs) at which tortoises will be released within a release site (U.S. Fish and Wildlife Service, 2020). Maintaining the area’s tortoise sex ratio (two males to one female) during translocation planning may also sustain population dynamics.

The USFWS will receive the translocation disposition plan at least 15 days before translocation for approval (U.S. Fish and Wildlife Service, 2020), and the translocation package will include but will not be limited to tortoise disposition plans, maps and Geographic Information System (GIS) files of last known locations of tortoises within the WTA and planned release site locations, identification of resident and reference tortoises, health data and photographs of tortoises to be translocated and select resident and reference tortoises, and recipient site survey data.

Tortoises may be found in the WTA after clearance procedures and translocation has been completed and during military training activities. If tortoises are found, Fort Irwin will coordinate the disposition of these animals with USFWS. If possible, these animals can be incorporated into one of the translocation research programs; otherwise, animals are to be moved into enclosure pens or moved to a pre-determined location for tortoises found after the large translocation event, provided environmental conditions as described earlier in the text are suitable for the release of tortoises (U.S. Fish and Wildlife Service, 2020). A formal consultation between the Army and USFWS will be restarted and adaptive management considered if 10 or more tortoises (greater than 180 MCL) have died due to military activities in the WTA after translocation or during translocation within any calendar year within the WTA after translocation or during translocation (U.S. Fish and Wildlife Service, 2021a).

Translocation of Tortoises from the Western Training Area

The procedures in this section are prescribed by requirements in USFWS guidance documents (U.S. Fish and Wildlife Service, 2009, 2020) and the U.S. Fish and Wildlife Service, (2021a) Biological Opinion; recommendations are not made by the USGS. The text in the next section is meant to inform the implementation of the requirements as USGS understands them, based on the scientific expertise, data, and experience of the USGS.

Per published literature and our experience, translocations of tortoises are best done only in the spring (April 1–May 31) or fall (September only) when the weather conditions are suitable for tortoise activity. We suggest that translocation be postponed if average precipitation (collected from weather stations on site with outsource supplementary data as needed; see the “Measurements of Environmental Variables” section) from 2021 to the year of anticipated release is less than one standard deviation from the 30-year average precipitation of the study area. The NTC will consult with climate experts and the USFWS to determine the status of translocation and monitoring activities during a megadrought (Williams and others, 2022). Drought or years with lower-than-average precipitation and annual biomass production have been observed to increase predation rates on tortoises and decrease survival rates range-wide (Longshore and others, 2003; Esque and others, 2010). Initiation of translocation may need to be delayed allowing for prey base populations to recover after drought. Decisions regarding translocation during drought years would be coordinated with USFWS. In accordance with desert tortoise handling permits and regulations, no desert tortoise shall be captured, moved, transported, released, or purposefully caused to leave its burrow for whatever reason when the ambient air temperature is above or anticipated to exceed 95 °F (35 °C) before handling or processing can be completed (U.S. Fish and Wildlife Service, 2020). Tortoises found in burrows during translocation can be “tapped out” by field crews to encourage them to exit (Medica and others, 1986), or they may require careful excavation (Desert Tortoise Council, 1994; U.S. Fish and Wildlife Service, 2020). Multiple visits may be necessary if tortoises are inaccessible, such as within caliche caves. After the removal of tortoises from burrows, burrows may be crushed so they cannot be re-occupied by other tortoises during translocation activities.

Per USFWS guidelines (U.S. Fish and Wildlife Service, 2019, 2020), all tortoises in the WTA that meet translocation criteria will be removed from the site, as described in this paragraph. If additional tortoises that were not previously found during clearance surveys are located, then they will have transmitters attached if they meet translocation criteria (USFWS, 2020). Tortoises are to be transported in vehicles to designated release sites by permitted biologists and released in the same day. During transportation, care must be taken to avoid stressful conditions, such as high temperatures, while waiting for transport, travelling in vehicles, or while waiting at the release site for dispersal. Tortoises in any phase of the translocation should not be left unattended for any period. Juvenile tortoises (less than 300 g or less than 150 mm; Medica and others, 1975) or other individuals that may have been housed in enclosure pens but meet translocation criteria can be translocated within the same season as other tortoises are being translocated from the WTA to the WTATS (U.S. Fish and Wildlife Service, 2019, 2020).

Tortoises are to be transported in clean, protective, and ventilated containers to ensure their safety during translocation. Containers will be sterilized using a 10-percent bleach solution (or diluted) or ready-to-use Rescue (Virox, Mississauga, Ontario, Canada) requiring 1 to 5 minute contact times (disinfection guidance found in U.S. Fish and Wildlife Service, 2019) before being used to translocate other tortoises (U.S. Fish and Wildlife Service, 2019, 2020). The area cleared and total number of tortoises found will be reported to the USFWS and CDFW (see the “Adaptive Management” section; U.S. Fish and Wildlife Service, 2020).

Releases of tortoises will be done when temperatures range from 65 to 85 °F (18–30 °C) and are not forecasted to exceed 90 °F (32 °C) within 3 hours of release or 95 °F (35 °C) within 1 week of release (U.S. Fish and Wildlife Service, 2020). Tortoises will not be released when it will be cooler than 50 °F (10 °C) within 1 week post release (U.S. Fish and Wildlife Service, 2020).

When released, translocated tortoises will be provided drinking water for 15 to 20 minutes and placed into unoccupied-shelter sites, such as a tortoise soil burrow (if available), caliche caves, or in the shade of a shrub (U.S. Fish and Wildlife Service, 2020). Releasing tortoises into unoccupied shelter sites within washes may contribute to increased site fidelity after translocation (Nafus and others, 2017a) and is encouraged. In previous studies, tortoises released into artificially made burrows did not appear to show fidelity to those sites and left immediately to seek out or construct other suitable cover sites nearby (Field, 1999; Nussear and others, 2012). Translocated tortoises rarely returned to burrows into which they were released during translocation; instead, they found or constructed other suitable cover sites. Ambient temperatures at the time of translocation can also affect the success of the release. Tortoises released under similar conditions to those recommended by USFWS are typically able to find suitable shelter without having signs of overheating or thermal duress (Lohoefener and Lohmeier, 1986; Corn, 1991; Field, 1999; Nussear and others, 2012).

Translocated tortoises can move long distances during the first year following translocation (Aiello and others, 2014), possibly moving outside the typical range of radio transmitters used for tortoise tracking (approximately 700–900 m; Esque, 1994; Nussear and others, 2012); therefore, monitoring of all translocated tortoise locations will be most effective if done within 24 hours of release, twice weekly for 2 weeks after release, weekly during the first active season, and twice monthly for the duration of the first year after release. After the first year of translocation, monitoring activities can be reduced to twice per month during active periods (April–October) and once per month during inactive periods (November–March) per tortoise. Any tortoises missing, either because their VHF signals could not be detected or their transmitters were recovered in the field, can be searched for within 24 hours since found missing then once per month thereafter by listening for signals throughout the project area and visiting burrows the tortoises previously used.

Post-Translocation Monitoring—Short and Long-Term Success Criteria

An appropriately designed monitoring program includes (1) standardized criteria for success, (2) hypotheses that are used to critically evaluate if management goals have been met, and (3) additional guidance for adaptive management to inform future actions (Morrison, 2002; Miller and others, 2014; Bell and Herbert, 2017; U.S. Fish and Wildlife Service, 2020). The NTC and USFWS agreed to develop a monitoring program for tortoises translocated from the WTA to better understand short- and long-term tortoise responses to translocation and to contribute information toward the range-wide recovery of the species (U.S. Fish and Wildlife Service, 2021a). Monitoring of short- and long-term success criteria can contribute to tortoise recovery and minimize mortality of desert tortoises, as outlined in the USFWS Recovery Plan (U.S. Fish and Wildlife Service, 2011) and Translocation Guidance (U.S. Fish and Wildlife Service, 2020). Such monitoring also can advance the Department of Defense’s (DOD) contribution to recovery goals as part of the Recovery and Sustainment Partnership Initiative (U.S. Fish and Wildlife Service, 2021a). The scope of the monitoring program for measuring the success metrics of the WTA translocation will be finalized by the NTC and USFWS before translocation. The program will be structured to ensure that there is coordination among all the monitoring activities done under this translocation plan (U.S. Fish and Wildlife Service, 2021a). This translocation plan also aims to revisit and expand upon previous goals and objectives outlined in the 2005 and 2009 NTC Translocation Plans that were not fully implemented during the original translocation in addition to following U.S. Fish and Wildlife Service (2020) success criteria guidelines (table 9; Esque and others, 2005; Esque and others, 2009). The evaluation of success metrics discussed later in the text will allow the NTC to evaluate the success of the proposed translocation and continue progress on the obligations defined for the previous translocation.

Evaluation of change that is ecologically relevant or detectable can be difficult to determine. We provide a meaningful target to detect changes that are within expected ranges in variation based on expert knowledge and current data to inform decision making. As initial guidance, triggers for adaptive management and tortoise responses among study groups can be compared with a 20-percent differential and baseline measurements as necessary. All tortoises involved in the translocation monitoring and tortoises involved in the studies of cooperators are best sampled for multiple parameters (growth, presence or absence of disease, and genetics) to determine study group responses (growth, survival, and movements). Responses that are within 20 percent of each other can be considered within the expected range of variation among groups (Esque and others, 2005; Brand and others, 2016; U.S. Fish and Wildlife Service, 2020). Evaluation of post-translocation data and baseline measurements (such as survival and health condition) may provide information on annual responses to environmental conditions that may be cause for concern. Effect sizes for success metrics are expected to vary, yet little is known regarding desert tortoises. Therefore, a standardized 20-percent difference can be applied but is subject to change as new data become available. Additionally, we hypothesize that a 20-percent differential will be detectable given the sample sizes that will be available. Although tortoise responses measured by various success metrics may differ from one another, and especially with regard to movements, we generally expect translocated tortoises to have similar responses to that of reference animals after they have had as many as 5 years to acclimate to their new environments. If this expectation is the case, we would consider their translocations to be “successful” in the short-term. Metric responses greater than 20 percent among study groups can trigger evaluation among all metrics (growth, survival, and movement) to assess factors that may cause a response difference greater than 20 percent and implement appropriate adaptive management actions. To assess compliance and continuity of translocation plan actions, Fort Irwin will create an advisory group (Fort Irwin, USFWS, CDFW, BLM, and others) that will meet annually to review and advise actions and results related to translocation, share the information gathered, and determine if the monitoring activities remain within the thresholds bounded for each success criterion. In addition, the meeting can facilitate coordination and data dissemination among all stakeholders. A framework can be developed to collect and archive all field data so that the assessments of the long-term goals are accurate and to assure that the data from all activities done under this plan are archived for future use (app. 1).

Sample size is an important consideration for any monitoring plan, and this is especially true when the mortality of research animals is a certainty either by slow attrition or catastrophic events, such as drought and predation (Longshore and others, 2003; Esque and others, 2010). As previously discussed (see the “Monitoring of Tortoises via Very High Frequency Telemetry or Similar Technology” section), all the translocated tortoises can be monitored simultaneously with 75–100 tortoises in the resident and reference tortoise groups. The USGS estimated sample sizes based on power analysis with 80-percent power, 0.10 significance level, and 20-percent difference in response rate among study groups. Sample size was increased to a maximum of 100 animals to account for animals that cannot be relocated, or die. The selected residents and reference population are best distributed across sites so that they represent locations where the translocated tortoises are released and sufficiently represent adult and non-reproductive size classes for meaningful analyses. Translocation studies have used 100–150 tortoise sample sizes previously, including for evaluating success of the previous 2008 NTC translocation (Mack and Berry, 2023). However, sufficient samples of resident and reference animals are needed to compare with post-translocation tortoise response and to evaluate translocation success.

Generally, monitoring plans for large translocations include tracking each tortoise in each study group (for example, translocated, resident, and reference tortoises) and a sample population of resident and reference animals for the first 6 years of the program, followed by an additional 20–30 years of long-term monitoring of a subset population of translocated animals and biennial surveys of the recipient and reference populations (U.S. Fish and Wildlife Service, 2020). Long-term effects of translocation are still not well understood, and long-term monitoring is particularly needed to determine the effectiveness of translocations of long-lived species like the desert tortoise. The monitoring program for this plan includes success metrics from the first translocation (Esque and others, 2005) and USFWS guidelines (U.S. Fish and Wildlife Service, 2020), which consist of five stages during an approximate 30-year period to adequately evaluate success criteria and to better address gaps in knowledge about tortoise translocation (table 9). The success of this translocation will be based on the quantifiable and hypothesis-driven criteria (Tracy and others, 2004; Miller and others, 2014; Bell and Herbert, 2017) that follow. The success metrics described in the next section are designed to measure responses of tortoises in relation to the range of environmental variation they are likely to encounter.

Measurements of Environmental Variables

Metrics used to evaluate success for this translocation plan must be considered relative to the responses among the three study groups (translocated, resident, and reference tortoises). Comparisons must be made among study groups to understand the direct effects of translocation as compared to responses due to other factors (prolonged drought or widespread predation linked to a drought; Esque and others, 2010). Some examples of metrics used to evaluate success are survival, growth, reproduction, and genetic integration rates (see table 9 and sections “Short-Term Metrics: Success Criteria Stages 1-3a” and “Long-Term Metrics: Success Criteria Stages 3b-5” for details).

Translocation of Tortoises and Habitat Quality

The Biological Opinion (U.S. Fish and Wildlife Service, 2021a) for the Recovery and Sustainment Partnership Initiative, Augmentation Strategy, updated Recovery Plan, and 5-Year species review (U.S. Fish and Wildlife Service, 2021a, 2021b, 2011, 2022b) all suggest using translocated tortoises to augment areas with depleted populations. Possible implications of this action must be considered carefully and in consideration of future outcomes (Frazer, 1992). Causes of depleted populations at several locations across the West Mojave are unknown; however, relationships have been hypothesized (U.S. Fish and Wildlife Service, 2021b; Mack and Berry, 2023). If an area selected for translocation has experienced or is actively experiencing population depletion, it is possible that translocated animals as well as the residents and reference population are being subjected to unknown or unquantified stress factors. Therefore, translocations must include monitoring and experimentation to ensure that the effects to the existing population and translocated populations in that area can be identified (Tracy and others, 2004).

Fine-scale measurement of environmental variables, such as precipitation, temperature, and vegetation, are vital to understanding the relation between habitat and tortoise ecology and are best recorded throughout all stages of the monitoring program. Weather stations measure fine-scale and daily changes in temperature, humidity, and precipitation, and rain gauges can be used to determine sporadic precipitation that may not be recorded otherwise due to the distance between publicly available weather stations. Perennial and annual vegetation surveys can be used to quantify habitat quality, available forage, and vegetative cover (see the “Measurements of Environmental Variables” section).

It is difficult to design experiments or observational studies that assess all possible factors related to population fluctuations, particularly if multiple factors are suspected of causing change, such as in the case of Mojave Desert tortoises (Tracy and others, 2006; U.S. Fish and Wildlife Service, 2011). Factors that may be related to population declines include road mortality, development resulting in habitat destruction, predator subsidies, invasive plant species presence on the landscape, wildfire, contaminants, activities related to illegal marijuana growing operations, and climate change, among others (U.S. Fish and Wildlife Service, 2021a). In fact, many populations that have been monitored for decades continue to decline despite years of increased conservation management (Tracy and others, 2004; Allison and McLuckie, 2018; Averill-Murray and others, 2021) and restoration efforts (Esque and others, 2021b). This information indicates that the suite of effects that can cause tortoise populations to decline are still present in those locations (Frazer, 1992; Zylstra and others, 2023), and the short- or long-term success of an experimental release of tortoises may depend on uncovering additional landscape stressors and adapting management actions to mitigate or remove them. Quantifying as many effects to tortoise survival as possible can ensure success after translocation.

Climate, soils, and vegetation in the Mojave Desert ecosystems are interrelated, and characterization of these variables is a critical part of understanding habitat suitability for desert tortoises. For long-lived desert plants and animals, such as the desert tortoise, climatic data are valuable for interpreting ecological patterns (Beatley, 1974). Availability of precipitation is correlated with the growth of juvenile tortoises (Medica and others, 1975). Alternatively, adult tortoises may show adverse responses to prolonged drought (Peterson, 1994; Longshore and others, 2003). Vegetation and climate monitoring, including percent perennial cover, composition and species richness, will be done annually as part of determining quality of tortoise habitat within the WTATS and WTA. The – covers a landscape that varies in topography, substrate, and vegetative communities. To study how different habitat characteristics and conditions affect tortoises, climate and annual and perennial vegetation are best monitored at selected, randomly stratified points throughout the WTATS and WTA.

Climate monitoring stations (Upward Innovations, Inc., DGS-001, East Falmouth, Massachusetts, or similar) are best distributed throughout the WTATS and WTA, assembled according to National Oceanic and Atmospheric Administration standards (National Weather Service, 2023), and outfitted with several climate sensors (rain gauges, thermometers, barometers, anemometers, and pyranometers to measure solar irradiance). Data collected from the stations can be uploaded via satellite and accessible through a web portal and analyzed annually. Cost-efficient rain gauge stations (fig. 12; TruCheck, Edwards Manufacturing Company of Albert Lea, Minnesota, or similar) may be placed randomly on the landscape to collect supplementary precipitation data. Rain gauge data are best recorded once per month (less frequently in cases when no precipitation was recorded by climate stations or no storm cells moved through the area) and emptied. Mineral oil can be added to slow evaporation of precipitation collected in gauges. Wire mesh can be placed inside the top of the rain gauges (such that they do not interfere with the collection of precipitation) to reduce the accumulation of insects. During months of expected below-freezing temperatures (for example, November through February), a small amount of antifreeze can be added to prevent freezing of any precipitation in gauges. Stations are best placed approximately 500–1000 m from roads and in areas with less evidence of human settlement (camps, trailers, and so forth) or disturbance, when possible, to prevent tampering with or damage to the stations.

Annual and perennial vegetation monitoring through field sampling efforts and remote sensing capture localized vegetation information that can be related to tortoise landscape use. Forage availability for tortoises can be sampled by recording each species, phenology, and biomass of available annual and select perennial forage plants within quadrats (1 square meter [m²] in size), while identifying and recording land cover strata (for example, upland, wash, rocky slope, and dry lake) at selected random points (Elzinga and others, 1998) within the WTATS. Sampling is most effective in spring and fall when tortoises are active and annual plant growth is expected. Our experience suggests that a robust design includes recording all species of live forage plants and their phenology within each quadrat. Available biomass can be obtained within a 0.1-m² section of each quadrat by clipping all live forage plants at ground level for collection. Clipped biomass samples are then sorted into monitored plant functional groups (for example, native grasses, invasive grasses, native forbs, and invasive forbs) and weighed, both freshly clipped and dried, to quantify plant water content.

Perennial plants provide essential cover resources for tortoises in the Mojave Desert (Nussear and Tuberville, 2014). Our experience suggests a defensible design to conduct perennial vegetation sampling within the WTATS. Points within each site are randomly stratified and selected for repeated sampling over years. Perennial sampling will be done once per year at independently selected random points and randomly generated transect azimuths (Elzinga and others, 1998). To quantify cover and characterize the communities of perennials present at each site, line-intercept data are recorded along the 50-m transect. Along each transect, observers record the species of each individual shrub that intersects the transect as well as the beginning and end marks (in centimeters [cm]) of that shrub’s canopy along the transect (Elzinga and others, 1998). Height is recorded for each shrub encountered along the transect. Within belt transects, observers record the number of individuals of each perennial species rooted within the belt. Data collection from remotely sensed enhanced vegetation indices (EVI) from the Moderate Resolution Imaging Spectroradiometer satellite-based sensor and Unmanned Aerial Vehicle-based sensors may improve analysis and evaluation of tortoise habitat quality at a finer resolution.

Introducing juvenile desert tortoises into translocation areas as part of the experimental design as biological probes can benefit this endeavor greatly (Nafus and others, 2017a). Juvenile tortoises have greater sensitivity (growth and survival rates and health responses) to disturbances, resulting in larger and more detectable responses to treatment effects than adult tortoises (Drake and others, 2016). In cooperation with the BLM, a defined investment in habitat improvements can be implemented to study habitat quality and effects of restoration efforts in augmented areas.

Roads, Habitat Fragmentation, and Human Effects

The NTC has invested in habitat improvements related to habitat restoration and roads in coordination with the BLM. Further investment in this commitment to intensive management actions could increase the success of this translocation in an area that is open to the public. Research on the effects of roads and other disturbances on tortoises and their habitats is recommended by the Recovery Plan (U.S. Fish and Wildlife Service, 2011). Downward population trends of Mojave Desert tortoises can be directly or indirectly linked to presence of roads, habitat fragmentation, and other assorted anthropogenic threats across the range (Stebbins, 1974; Bury and others, 1977; Boarman, 2002a; Tracy and others, 2004; Custer and others, 2017; Allison and McLuckie, 2018; Dutcher and others, 2020; U.S. Fish and Wildlife Service, 2021b; Averill-Murray and Allison, 2023). Road presence may decrease tortoise populations through various mechanisms, including direct mortality from vehicle collisions, which reduces the number of larger reproductive animals that could contribute to population recruitment (Nafus and others, 2013). Furthermore, roads across the Mojave Desert increase access to desert tortoise habitat and can introduce other human effects, such as introduction and invasion of non-native plants and exposure to predation by feral dogs (U.S. Fish and Wildlife Service, 2021b).

The number of paved and unpaved roads and OHV routes, as well as off-road vehicle use and habitat degradation, have increased in the Mojave Desert, including in the WTATS (Tracy and others, 2004). Human development, including renewable energy, continues to expand throughout the Mojave Desert (Agha and others, 2020). A recent study indicated that 60–70 percent of tortoise habitat had human development within 1 km, and 43 percent of undeveloped tortoise habitat was outside of current Federal, State, or local habitat protections (Carter and others, 2020). Cumulatively, these effects will encourage the success of the translocations; however, well-coordinated conservation management actions could be taken to increase the likelihood of translocation success.

Many metrics can be used to evaluate landscape patterns. Spatial pattern analysis may consider area, density, or edge, shape, core area, isolation or proximity, contrast, contagion or interspersion, connectivity, and diversity (McGarigal and others, 2002). In addition, linear network pattern analysis may be useful with development of a variety of other metrics (Forman, 1995). Measurement of road density may be used as a surrogate for fragmentation. Road density can be measured by the number of miles or kilometers of roads and trails per unit area.

A long-term analysis from the Mojave Desert Tortoise Recovery Office found that no tortoise populations increased in areas with road density greater than 0.75 km/km² (Averill-Murray and Allison, 2023); Averill-Murray and Allison (2023) recommended management actions to reduce tortoise declines relative to road density, including increasing law enforcement, public outreach, and tortoise-exclusion fencing, as well as setting limits for road density (through communication and efforts between BLM and NTC). The DOD will fund BLM contact park rangers to patrol focal areas, described in the USFWS Biological Opinion (U.S. Fish and Wildlife Service, 2021a), to increase BLM presence and monitor for illegal activities (U.S. Fish and Wildlife Service, 2021a). Mean patch size, number of patches, edge density, and landscape shape index related to road networks can also be measured and may be correlated with changes in the composition of native perennial plant communities and relative presence of exotic and native annual plants, which may affect tortoise diets (Oftedal and others, 2002; Drake and others, 2016). Designating closed OHV wash zones throughout the area may reduce effects to suitable tortoise habitat. Current Federal OHV policy and regulations have shown positive effects of reasonable compliance on sensitive habitats when open versus closed routes were clearly marked (Custer and others, 2017), but other areas have experienced low compliance with road closures (Ouren and others, 2007).

As part of the Recovery and Sustainment Partnership Initiative and Recovery Plan Strategic Element 2 (U.S. Fish and Wildlife Service, 2021a), the NTC will work with the USFWS to protect existing tortoise habitat by providing funding for the acquisition and conservation of private inholdings on the recipient site landscape (WTATS), particularly in areas where rapid reduction of fragmented habitat and management may be possible. The Recovery Plan (U.S. Fish and Wildlife Service, 2011) recommends management by working to connect blocks of desert tortoise habitat to maintain gene flow between populations, with priority conservation areas chosen based on minimum connection of habitat needed for desert tortoise long-term success (Averill-Murray and Hagerty, 2014; Averill-Murray and others, 2021). Habitat restoration activities focused on unauthorized routes may reduce human use of these areas. Activities such as developing seed sources of native plant materials and removal of invasive plant species may contribute to long-term success of tortoise populations and other wildlife (Olwell and Riibe, 2016; Esque and others, 2021a). Reducing the use of unauthorized routes through fencing would allow for future restoration efforts and reducing tortoise mortalities from vehicle collisions.

As part of long-term success metrics, the NTC can monitor translocated, resident, and reference tortoises throughout the WTATS, which has varying levels of habitat disturbance, including on BLM and DOD lands; therefore, data will be available to study the effects of roads and habitat fragmentation on tortoise populations. The DOD will support the BLM to reduce route density (U.S. Fish and Wildlife Service, 2021a) within the translocation areas to achieve a level of greater than 0.75 km/km² as per Averill-Murray and Allison (2023), including through road closures and restoration.

Reporting and Data Storage

As required by the Biological Opinion (U.S. Fish and Wildlife Service, 2021a) and recovery permit (number TE-63428D-0, -1), the Army must provide electronic annual and comprehensive reports for all permitted activities by January 3, after each year the recovery permit is in effect. Reporting is an important aspect of the work to benefit desert tortoise recovery. Submission of the Annual Summary Report form (FWS Form 3-2530 or similar) and comprehensive project report will be provided to USFWS and will summarize all the desert tortoise, habitat, maps, health results, environmental data, and any additional information (for example, relevant GIS layers, master data sheets, photographs, notes) required by the USFWS recovery permit (number TE-63428D-0, -1). Datasheets and electronic data collection used in the field will be developed in coordination with USFWS and entered in a USFWS/BLM-provided master database (U.S. Fish and Wildlife Service, 2020). All desert tortoise, habitat, and environmental data may be archived in a standardized data repository such as ServCat or ScienceBase or an approved repository from USFWS, such that data collected will be open, machine-readable, secure, and accessible. Health data collection will conform to the current translocation health assessment guidance (U.S. Fish and Wildlife Service, 2020). All injuries and mortalities discovered during monitoring will be reported to the USFWS within 24 hours, per permit requirements (U.S. Fish and Wildlife Service, 2021a). The report must include the tortoise identification (ID), date, time, location of the carcass (UTMs), a photograph, cause of death (if known), and any other pertinent information (for example, sex, size, date, and UTMs of last known live location; U.S. Fish and Wildlife Service, 2021a).

After the completion of the short-term and long-term post-translocation monitoring periods, final reports are to be completed to assess the overall success of the translocation and monitoring program. The final reports will summarize translocation monitoring activities and other compliance-related reporting as specified in the Biological Opinion (U.S. Fish and Wildlife Service, 2021a). Completed recovery permit reports can include discussion of any adaptive management used throughout the monitoring period, with an assessment of the success of each adaptive management strategy. Reporting timelines and report content will be coordinated with USFWS guidance to ensure appropriate content is included per permit requirements (U.S. Fish and Wildlife Service, 2020).

Adaptive Management

This translocation plan describes procedures to plan, implement, and research translocation of tortoises by the NTC; however, adaptive management measures will be implemented during the translocation and monitoring processes after identifying concerns, immediately addressing issues in the field, and consulting with all involved agencies. Evidence of translocation project-related disturbance or increased risks to desert tortoises may necessitate discussions with the USFWS to outline these adaptive measures in translocation and monitoring procedures in additional or edited project documentation. Annual meetings between the NTC and USFWS, along with the post-translocation year 5 interim report, to assess compliance and implementation of the translocation plan may also drive remedial management actions for future study years. Altered management may be enacted in relation to the results of short-term and long-term tortoise translocation success metrics outlined in this section.

If there are concerns in the field regarding immediate threats to one or more tortoises, field staff can make adaptive management decisions in the best interest of the tortoise(s) through (1) coordination in the field; (2) phone calls to agency representatives within 24 hours to describe the actions taken and results of the actions; and (3) a brief email report from field staff that describes the adaptive management actions taken, along with reasons for and results of these actions. If there are concerns in the field that do not pose an immediate threat to one or more tortoises, a designated field representative can notify the agencies of proposed adaptive management decisions via email, and field personnel can wait as much as 1 week for consultation or additional direction and response from agency personnel before actions are taken. These adaptive measures can be taken by the NTC in addition to following minimum mitigation requirements outlined in the Biological Opinion (U.S. Fish and Wildlife Service, 2021a).

Triggers to implement adaptive management may include translocation activities that are projected to or have put the well-being of a tortoise at risk (for example, tortoises pacing a tortoise exclusionary fence, or a tortoise found within an active training area). After consultation with agency representatives, adaptive management measures may include, but are not limited to, the following:

Summary

The U.S. Department of the Army (Army) plans to initiate military training exercises in a realistic battle environment within the National Training Center (NTC) and the Fort Irwin Western Training Area (WTA) in San Bernardino County, California, in 2025. Such large and intensive training activities would disturb desert tortoises and their habitats to the extent that translocation of tortoises would likely be required by the U.S. Fish and Wildlife Service (USFWS) before training ensues to minimize tortoise mortality.

The Fort Irwin Military Land Withdrawal Act of 2001 authorized the NTC to expand its training activities into Southern Expansion Area (SEA) and the Western Training Area (formerly known as the “Western Expansion Area,” or “WEA,” during the 2008 translocation). Translocation of tortoises from the SEA was completed in 2011. Planning for the SEA translocation sought to (1) provide safe, humane, and successful translocation of tortoises with minimal effect to resident tortoises and reference tortoises; (2) study tortoises affected by translocation to increase understanding of the ecology, conservation, and management of desert tortoises; and (3) define measures of success for translocation and provide metrics to evaluate successes. All activities supported by the NTC, as described in the 2008 Translocation Plan, were discontinued before completing the WEA (now WTA) expansion. Although monitoring for the 2008 NTC translocation ended prematurely, at least 37 peer-reviewed products on tortoise ecology, regional landscape conditions, and effects of translocation were published, thus partially meeting the original goals by increasing knowledge on tortoises, habitats, and translocation.

The WTA expansion and Western Training Area Translocation Site (WTATS) are the focus of this translocation plan. This plan describes how to monitor desert tortoise fitness compared with environmental variables to determine success metrics of translocation and identify thresholds for agency decision-making. This plan builds on the previous NTC translocation activities and includes results of recent (April 2020–November 2022) baseline surveys of tortoises and their habits in the WTA and WTATS, release areas, and reference areas. We identify tortoise locations in the WTA, evaluate habitat risks, provide choices for translocation and reference sites throughout the WTATS, and identify short- and long-term monitoring metrics for future surveys to evaluate the status of translocated tortoises in comparison with resident and reference populations.

The WTA (286 square kilometers [km²]) borders the Naval Air Weapons Station-China Lake (NAWSCL) to the north, the Paradise Range and Lane Mountain to the south, Superior Dry Lake to the west, and is in the southwest corner of the Fort Irwin NTC. Habitats are comprised of typical Mojave Desert shrubland with interior draining dry lakes. In the WTA, there is restricted access in the “East Paradise Conservation Area” and “Brinkman Wash Restricted Area.”

The WTATS (approximately 3,296 km²) is bounded on the north by the NAWSCL, WTA, and the NTC and is north of Barstow and Hinkley, in San Bernardino County, California.

Most of the WTATS is on public lands managed by the Bureau of Land Management (BLM), interspersed by private inholdings, and properties owned by the NTC. In addition to Mojave Desert shrublands, the WTATS has greater topographic diversity than the WTA by the addition of rugged volcanic and granitic mountains.

The staff at BLM Barstow and Ridgecrest Field Offices provided recommendations for areas to exclude as potential tortoise recipient sites in the WTATS due to land use conflicts, and they also indicated that translocated tortoises that move into designated Wilderness Areas from their recipient sites do not need to be removed.

Baseline tortoise and habitat investigations were performed in the WTA and the WTATS in 2020–22. A subset of tortoises had radio transmitters attached to their carapace so they could be located during future investigations. We established random tortoise survey plots (300 by 300 meters[m]) across the WTA and WTATS to estimate tortoise density and abundance following protocols adapted from the USFWS. There were 1,408 plots surveyed in the project area, and 783 tortoises were found. The 2020–22 plot surveys and telemetered tortoises included 783 tortoises, with 86 percent adults and a two male to one female sex ratio. Baseline tortoise health assessments were completed throughout the WTA and WTATS for telemetered tortoises. Tortoises were examined for clinical health conditions by physical assessment and tissue collection, following USFWS guidance. Collected samples were sent to labs to test for presence of tortoise diseases. Most health assessments classified the tortoises as “clinically normal” and “adequately conditioned.” Tissue samples from the WTATS yielded positive laboratory results either for antibodies specific to Mycoplasma agassizii and Mycoplasma testudineum (via enzyme-linked immunosorbent assay [ELISA] testing; n=4 or 3.3-percent of assessed population) or pathogen presence (via quantitative polymerase chain [qPCR] testing; n=6 or 6.7-percent of assessed population). There was 5-percent mortality among all marked tortoises averaged across the WTA and WTATS in 2020–22. Mortality analyses indicate only moderate losses from predation, disease, or climate variability in the project area, especially considering recent drought conditions.

One of the primary objectives of this translocation plan was to provide landscape analyses to inform the best locations for recipient and reference sites associated with the plan. Based on USFWS guidelines and consultations with partners, a model was created to inform site selection for recipient and reference sites related to the NTC translocation activities. The models identify suitable translocation sites for tortoises in the WTATS. Model parameters included biological, geospatial, and environmental data important to the survival and health of tortoise populations. The final seven modeling criteria for recipient site selection included: land ownership, desert tortoise habitat suitability, distance to roads, common raven nest site density, desert tortoise habitat connectivity, precipitation, and a terrestrial disturbance index.

Scenarios were developed using a form of ordered weighting averaging by manipulating the weight and scale of each input raster layer (criteria) that could be considered beneficial or detrimental to tortoise translocation in the WTATS. Spatial models providing desert tortoise habitat suitability, desert tortoise movement potential, and average winter precipitation were positive parameters, and negative effects included raven nest densities, distance to roads, and a total disturbance index. Potential scenarios for desert tortoise translocation success were based on the combination of five variations of the raster scaling. These scenarios were discussed with agency personnel with administrative responsibilities for the study area, who provided feedback in relation to agency policies.

In addition to combined modeling outputs, candidate sites for references sites were parcels owned by the NTC, are outside excluded habitats, have a suitability value greater than or equal to the mean model value (for example, greater than or equal to 0.39), and were considered as potential recipient sites for translocated tortoises from the WTA. Eight potential recipient sites and two potential reference areas were identified through the combined analyses. The recipient sites were ranked in order from highest to lowest priority. A 6.5-kilometer (km) movement buffer was created from the centroid of each selected Fort Irwin-owned land parcel (recipient sites), resulting in three potential translocation sites for translocated tortoises (TS1, TS2, TS3; may vary depending on exact release site of translocated tortoises).

Site visitations were made by authors of this report and other USGS staff members extensively from spring 2020 to fall 2022 for evaluation of recipient and reference sites. Recipient and reference sites or grouped sites were photographed at their centroids and digitally archived in fall 2022. We found some sites were unsuitable for tortoise translocation because of excessive off-highway vehicle (OHV) use or other anthropogenic effects and were disqualified as potential translocation sites. Selected recipient and reference areas were characterized by typical desert tortoise habitat in mixed shrub communities and represented by a Larrea tridentata (creosote bush) and Ambrosia dumosa (burro bush) plant association. The BLM, NTC, and non-government organizations have cooperated to make substantial investments in habitat restoration throughout large parts of the general site by reducing road incursions, leaving access on a network of designated roads. This translocation provides detailed descriptions of each translocation, and reference site.

Seasonal tortoise densities were estimated using spatial capture-recapture (SCR) models in a spatially explicit search area-encounter approach. We subset tortoise detections and surveyors’ search tracks by year and season, which we assigned to discretized grids of “effective detectors.” Season-specific grid cell spacings were based on mean seasonal range size estimates generated from tortoise telemetry using the 95-percent autocorrelated kernel density estimates.

A Poisson observation model was used for the SCR detection process, and a hazard half-normal detection function was used that described the rate of decay in baseline detection rate at an individual’s activity center (λ₀) as a function of distance between the activity center and grid cell in which the individual was detected, represented by the spatial scale parameter (σ). We accounted for spatially and temporally varying survey effort by specifying hazard-based effort effects and modeled a two-class sex effect on the λ₀ and σ parameters.

The spatial distribution of home range (activity) centers was described using two separate ecological point process models in each parameter estimation area or state space (S). To ensure that S was large enough to contain all individuals that had a non-negligible probability of detection, the discretized grid cells were buffered by 3–5×σ_Season to define the spatial extent of S. We adjusted the spatial extent of each S to accommodate landscape barriers, such as tortoise exclusionary fencing, and improve accuracy of SCR model parameter estimates, which resulted in the larger NTC being divided into three smaller study areas (WTA, WTATS-West, and WTATS-East), among which natural tortoise movement was not possible.

We fit SCR models with a homogeneous Poisson point process ecological model, which assumed that individual tortoise activity centers were randomly distributed throughout each S, as well as an inhomogeneous Poisson point process ecological model, which allowed the number and spatial distribution of tortoise activity centers to vary as a log-linear function of an existing desert tortoise Habitat Suitability Index (HSI).

Each study area by-season by-year dataset was analyzed separately, fitting the same suite of SCR models to each of the 10 datasets. Parameter estimates were produced from the top-ranked SCR model for each area by-season by-year analysis. We also derived estimates of average seasonal study-area specific population growth rates using an exponential growth equation. Additionally, to investigate potential sources of bias in density estimates relative to the characteristics of survey results, we fit four separate Gamma generalized linear models (GLMs) with the SCR estimated densities as the response variable and the total numbers of detected tortoises, recaptures, spatial recaptures, survey occasions, and S sizes as the predictor variables.

On average, we detected 117 tortoises per day per study area (range: 52–180); 108 recaptures per day per study area (range: 6–266), and 43 spatial recaptures per day per study area (that is, tortoise detected in greater than 1 grid cell; range: 2–143). The mean sex ratio was 64 percent males versus 36 percent females across all areas, seasons, and years. Tortoise density spatially varying as a function of HSI was included in the top-ranked model for three of the analyses, and all those estimated relationships were positive. Point estimates of mean density ranged from 0.27 to 1.85 adult tortoises per square kilometer (tortoises/km²), with an average for the NTC across all 10 area-by-season-by-year estimates of 0.95 adult tortoises/km². Study area-specific mean densities, averaged across seasons and years, were 1.08, 0.51, and 0.95 adult tortoises/km² at WTA, WTATS-East, and WTATS-West, respectively. Density estimate precision (coefficient of variation [CV]) ranged from 0.10 to 0.19, with a mean of 0.14 across all 10 area-by-season-by-year estimates. Estimated density increased over time in all three study areas such that the derived average seasonal population growth rates were 1.52 (95-percent confidence interval [CI]: 1.19, 1.77), 1.32 (95-percent CI: 1.07, 1.64), and 1.55 (95-percent CI: 1.48, 1.63) at WTA, WTATS-East, and WTATS-West, respectively. Gamma GLMs indicated a strong positive relationship between tortoise density and number of tortoises detected (β=0.29; 95- percent CI: 0.1, 0.48; p=0.003), whereas strong negative relationships existed between tortoise density and numbers of recaptures and spatial recaptures (β_Recaps=−0.47; 95 percent CI: −0.65, −0.29; p<0.0001; β_SpatRecaps=−0.42; 95-percent CI: −0.64, −0.20; p=0.0002). Density estimates were invariant to the number of survey occasions and the study area (state space) sizes (β_Occasion=−0.06; 95-percent CI: −0.29, 0.16; p=0.57; β_Area=−0.002; 95-percent CI: −0.26, 0.25; p=0.99).

We predicted mean spatial tortoise densities for each translocation site by converting the HSI raster into spatially explicit densities using coefficient estimates from the top-ranked SCR-HSI models. The following translocation site-specific means, SEs, and 95-percent CIs were produced: TS1 (355 cells): Mean=0.47 adults/km²; SE=0.0114; 95-percent CI=0.46–0.48; TS2 (350 cells): Mean=0.43 adults/km²; SE=0.0126; 95-percent CI=0.42–0.44; and TS3 (178 cells): Mean=0.41 adults/km²; SE=0.0220; 95-percent CI=0.39–0.43.

The mean estimated adult tortoise density at WTA, averaged among estimates produced during 2021–22, was 1.08 adults/km² (95-percent CI: 0.44–1.73), which corresponds to 273 live adult tortoises (greater than 180 midline carapace length) in the WTA (95-percent CI: 111–438 adults) to be translocated to the WTATS. Based on field data, the estimated densities for WTATS-East, WTATS-West, combined and within each translocation site, are below the tortoise density threshold of 3.9 adult tortoises/km². Translocation into each site is estimated to increase local densities while not exceeding the threshold.

Desert tortoise clearance procedures follow guidelines provided by the USFWS and are meant to inform the implementation of the USFWS guidelines as USGS understands them, based on the scientific expertise, data, and experience of the USGS. Recommendations are not made by the USGS. Any dates are based on our understanding of the NTC’s proposed project timeline. Tortoise clearance protocols include but are not limited to many activities related to structures required for temporarily housing tortoises in advance of clearances, habitat clearance surveys, marking and measuring tortoises, a monitoring program for tortoises via telemetry or GPS, tortoise health assessments, and plans for fencing tortoises.

If holding tortoises in enclosures can be avoided, that is preferred; however, USFWS guidelines or recent husbandry guidance can be used for construction or modification of outdoor predator-proof tortoise enclosures to temporarily house tortoises too small for radio telemetry, individuals with conditions that warrant husbandry or veterinary care, or individuals deemed unsuitable for translocation for other reasons. Tortoises housed in temporary facilities require annual health assessments, and it may be necessary to arrange veterinary visitation. Plans for temporary holding facilities and husbandry are to be approved by USFWS before clearance surveys.

The USFWS requires that areas within the WTA and connected to high intensity training areas be searched with 100-percent coverage to locate and remove tortoises during tortoise clearance surveys according to USFWS protocols. Tortoises are to be removed from the WTA before the commencement of military activities. The WTA will be surveyed two complete times, and if tortoises are found during the second survey, a third pass may be required. If juvenile tortoises or tortoise nests are located, intensive focal searches using concentric circles will be included in the survey. Permits for handling desert tortoises are required by the California Department of Fish and Wildlife, BLM, and USFWS before surveys.

Tortoises will be marked and measured as outlined in the NTC Federal recovery permit issued by USFWS. Besides health data, observer, date and time, tortoise number, and Universal Transverse Mercator (UTM) location (using Global Positioning System) will be recorded, as well as maximum carapace length, plastron length, and total mass. Tortoises will be externally marked with a unique number via a small piece of paper glued to the carapace using clear epoxy; the numbers assigned to tortoises will be coordinated with the USFWS.

Very high frequency radio transmitters will be attached to the carapace of eligible tortoises using standardized techniques. The mass of the transmitter must be less than 10 percent of the tortoise body mass. Tortoises that are too small for radios will be housed in temporary facilities until the translocation. Data loggers using a Global Positioning System also will be attached to eligible tortoises for future study of home range and space use.

Health assessments, tissue sampling, and laboratory diagnostics will follow USFWS guidelines provided in this report. Tortoises are eligible for translocation if they do not show any known signs of poor health, including lethargy or weakness, mucoid nasal discharge, aberrations around the mouth discovered during the health assessments or analyzed tissue samples. Tortoises that do not meet these criteria will be assigned to the temporary housing structures. After husbandry, and with approval from Federal and State agencies, tortoises showing improved health may be translocated or moved to alternative sites. Physical health assessments will describe the animal’s posture, condition of facial features and signs of disease abnormalities, discoloration and damage to the skin or shell, including the eyes, nares, and cloaca. A body condition score can be ranked overall and then assigned a numerical value, as per USFWS protocols. Tissue samples, including plasma, can be collected for ELISA testing of Mycoplasma agassizii and M. testudineum. Sloughed epithelial cells can be scraped from inside the mouth with oral swabs to collect deoxyribonucleic acid and test for the Mycoplasma spp. and Testudinid herpesvirus 2 (TeHV2) using qPCR analyses. All tissue can be sent to laboratories for analysis using USFWS standard protocols.

The USFWS requires exclusionary perimeter fencing around areas where tortoises are confined or from which they are to be excluded to prevent disease transmission of captives and protect tortoises from harm from features such as intensive military training areas or highways. Many areas of concern regarding fencing have already been addressed in the WTA and WTATS. Fencing can also prevent tortoises from attempting to return to areas from which they were translocated. All construction, inspection, maintenance, and repair of desert tortoise exclusionary fencing will be specified in appropriate planning, as specified by USFWS. The USFWS stipulates that newly constructed exclusionary fencing will be monitored during the active tortoise season and during excessive heat periods to ensure tortoises are not trapped. Exclusionary fences will be monitored after high rainfall periods, and damages such as debris accumulation or gaps will be repaired within 48 hours of discovery.

The NTC will develop a tortoise disposition plan and a translocation package for all tortoises in the WTA that includes the last known locations of tortoises, planned release site locations, and the records for all resident and reference tortoises as health records and photographs. Fort Irwin will coordinate the disposition of tortoises found in the WTA after planning the clearance and translocation with the USFWS. Excessive mortality of tortoises during translocation may result in formal consultation with USFWS.

Translocation of tortoises from the WTA will follow USFWS guidelines and State and Federal permits. Procedures are prescribed by requirements in USFWS guidance documents and the U.S. Fish and Wildlife Service Biological Opinion; recommendations are not made by the USGS. The best window for successful translocations of tortoises is during tortoise active seasons in the spring or fall. If precipitation does not meet the criteria described in the plan, translocations may be postponed, and such decisions can be coordinated through USFWS. Climate and weather patterns can also exacerbate other complications that can coincide with translocation, such as increased predation during drought periods. Excessive tortoise predation may also require consultation with USFWS. Decisions on translocation during drought years would be coordinated with the USFWS.

Desert tortoise translocations will be done in accordance with all USFWS guidelines regarding the capture, transport, and release of desert tortoises. Tortoises sequestered in burrows may be “tapped out” carefully. If these tortoises are not successfully extracted after initial visits, they may be carefully excavated, according to USFWS guidelines. Burrows will be crushed after excavation of tortoises to render them unusable. Release and post-translocation monitoring of tortoises will follow USFWS guidelines.

This plan provides metrics that assist the NTC in evaluating the success of the proposed translocation. Tortoises monitored for translocation are best sampled for multiple parameters of health and genetics to understand responses in growth, survival, and movements. Response variables within 20 percent of reference points are acceptable, and we hypothesize that a 20-percent differential will be detectable with adequate sample sizes. Metric responses greater than 20 percent among study groups can trigger evaluation of all study metrics to develop appropriate adaptive management actions. We generally expect translocated tortoises have similar responses to reference tortoises after having up to 5 years to acclimate. If correct, translocation would be considered “successful” in the short-term. Sample sizes were estimated using analyses with 80-percent power, 0.10 significance level, and 20-percent difference in response rate among study groups. A maximum of 100 translocated animals is required during monitoring. Sufficient samples of resident and reference animals are similarly required.

Monitoring for large translocations can include tracking tortoises in study groups for the first 6 years of the program, followed by an additional 20–30 years of long-term monitoring of a subset study animals. Long-term monitoring is particularly important to determine the effectiveness of translocations for the desert tortoise. The monitoring program for this plan includes success metrics from the previous Fort Irwin translocation and USFWS guidelines.

Measuring environmental variables can be important in evaluating translocation success in comparison with survival, growth, reproduction, and genetic integration rates in relation to variation in environmental variables that include precipitation, temperature, vegetation parameters, and roads and other human disturbance factors.

The NTC and USFWS agreed to use translocated tortoises to augment areas with depleted populations. Including properly designed monitoring and experimentation in translocation activities ensures that implications of translocations to all tortoise involved are considered in relation to short- or long-term success of an experimental release of tortoises. Importantly, such studies can uncover landscape stressors that require adaptive management actions to mitigate or remove them. In cooperation with the BLM, habitat improvements such as road closures and restoration can be implemented to study habitat quality and effects of restoration efforts in augmented areas.

Quantifying as many effects to tortoise survival as possible can increase the likelihood of informing future translocations. In this translocation plan, we propose quantifying fine-scale environmental variables, such as precipitation, temperature, vegetation, and physiography and soils (for example, adequate representation of wash, upland, friable soils, and cover sites in the habitat) and are best recorded throughout all stages of the monitoring program. Additional factors we propose to monitor include road mortality, development resulting in habitat destruction, predator subsidies, invasive plant species presence on the landscape, wildfire, contaminants, activities related to illegal marijuana growing operations, renewable energy development, and climate change.

In coordination with the BLM, the NTC has invested in habitat improvements, such as road restoration. We propose that spatial pattern and linear network pattern analyses may be useful, along with development of a variety of other metrics. As part of long-term success metrics, the NTC can monitor translocated, resident, and reference tortoises throughout the WTATS, which has various levels of habitat disturbance on BLM and DOD lands; therefore, data will be available to study the effects of roads and habitat fragmentation on tortoise populations.

The DOD plans to fund BLM contact park rangers to patrol focal areas. The NTC will work with the USFWS to protect existing tortoise habitat by providing funding for the acquisition and conservation of private inholdings on the recipient site landscape (WTATS), particularly in areas where rapid reduction of fragmented habitat and management may be possible.

Short-term monitoring will be completed during the first 6 years of the translocation project and will include tortoise movements, site fidelity, home range, egg production and nest success and recruitment, growth rates, disease, stress, and survival rates. The analysis of animal movements provides a quantitative measure that can be used to relate desert tortoise population status to variation in their habitats. Disturbances, social interactions, landscape features, sex and age, season, weather or availability of forage, water, and shelter can affect tortoise movements. Locational and movement data can be used to analyze home range and site fidelity through telemetry or GPS data-loggers. Monitoring translocated tortoises, in addition to 75–100 resident and 75–100 reference tortoises through VHF telemetry or GPS data-loggers is among the best ways to monitor tortoise movements. A less than 20-percent difference in movement between translocated and recipient populations by year 4 of monitoring post-translocation would meet success criteria for this metric.

Important components of tortoise population recruitment can be measured by monitoring successful egg production, survival of hatchling tortoises, and growth into larger size classes. Nest success is a variable that can be used to measure the success of translocated populations assimilating into the recipient population and to predict their potential effect on recipient site demographic patterns. In addition, reproductive success may indicate if physiological stressors are affecting tortoises at an ecological level.

The second component of measuring reproductive success is identifying the proportion of eggs that produce hatchling tortoises emerging from nests. To assess tortoise reproduction and nest success in the release areas post-translocation, we suggest the NTC monitor 20 female tortoises from each study group for egg production by doing X-rays on each female in the group. Detected nests can be fenced and monitored by camera traps to follow nest activity, including egg-laying, potential predation or other disturbances, and emergence of neonates from the nest. Egg production, nest success, survival, and size class recruitment among tortoise study groups that do not differ by more than 20 percent from the reference tortoises would indicate translocation success based on this metric.

Tortoise growth rates can be measured by recording dimensions of the shell and the mass of animals over time and comparing result among study groups. After accounting for age, sex, and variation among sites, the translocation would be considered successful if growth rates of individual tortoises do not vary by more than 20 percent among study groups after 3 years post-translocation. If growth rates vary more than 20 percent during the first 3 years post-translocation, potential causes for differences may be investigated.

Disease and stress can affect survival rates, and monitoring for all three of these issues is planned. Tissue samples collected from tortoises during routine health assessments will be screened for the pathogens that cause Upper Respiratory Tract Disease (URTD) and TeHV2.

Survival rates can be quantified by monitoring marked individuals and comparing survival with tortoise health and environmental data. Disease levels, stress levels, predation, and survival varying less than 20 percent among study groups indicate translocation success for these metrics. If disease, stress, or survival rates for translocated animals vary by more than 20 percent from those of residents or controls in similar conditions, the apparent causes can be investigated so that adaptive management of the translocation program can potentially mitigate any identified problems.

Comparing mortality rates of translocated tortoises among study groups in similar habitats with different predation pressures can elucidate underlying causes. Predation can trigger management actions if greater than 2 percent of the study populations are lost to predation in a season. Permit terms require all tortoise mortalities to be reported immediately to the Desert Tortoise Recovery coordinator. Such planning may avoid an ineffective crisis response in the event of high predation incidence.

Long-term measures of success can be monitored through surveys of demography, reproduction and recruitment, genetic integration, and survivorship and disease. Long-term monitoring is valuable because variability of demography, genetics, and disease prevalence can exceed short-term monitoring periods. Short-term measures of success last for 6 years after translocation, whereas long-term metrics last 7–30 years after translocation.

Demography is the study of how population characteristics vary through time and across space. Having information about population demographics is fundamental to species management, and the Recovery Plan emphasizes this need by calling for analysis of key vital rates through long-term, range-wide demographic monitoring. Demographic parameters of interest that are included in this report include population densities and size, growth, range, size class distributions, and vital statistics, such as generation time, reproductive rates, recruitment rates, and survival and mortality rates. Tortoise demographics have been monitored in two primary ways: (1) permanent study plots (1–2.6 km² in area) and (2) distance sampling. We suggest monitoring a subset of 100 tortoises per study group that is stratified by sex and age-class ratio using findings from baseline demographics.

Guidelines for monitoring long-term reproductive output and juvenile recruitment are provided by the USFWS. We revisit field methods and research design for determining tortoise reproduction and recruitment success. The NTC can augment information collected on tortoise reproduction during the short-term post-translocation monitoring period by continuing to monitor 20 translocated females, 20 resident females, and 20 control females.

Genetic integration can be evaluated as a metric of success by the presence of juvenile tortoises of mixed parental ancestry among the translocated and resident tortoises. Monitoring genetics of desert tortoise populations can inform translocation research by quantifying if translocated tortoises become reproductively integrated with resident tortoises, resulting in increased population growth and inform whether there are increased conservation benefits to the activity. Previous work at Fort Irwin provided short-term genetic results; however, that study was completed 4 years post-translocation; therefore, it may not represent the full potential for genetic contributions during the long-term integration of translocated tortoises. Quantifying genetic integration in desert tortoise populations that are not manipulated can also be important for comparison. Tissue samples taken from all the study groups, including before and after recruitment, samples collected during short-term monitoring, and previous translocations can be useful in genetic analysis. Few data are available on relatedness among tortoises within small geographic areas at this time.

By monitoring a part of all the tortoises among study groups, including the tortoises previously sampled in the WTA and WTATS during pre-translocation activities, long-term survivorship and disease can be quantified. When disturbances, such as translocation, are introduced into populations, monitoring survivorship and disease may be especially important toward providing basic information for understanding population demography and health. The Recovery Plan recommends long-term research on survivorship and epidemiology and factors that contribute to mortality of desert tortoises, in addition to research on the long-term effect of translocation on population dynamics.

Reporting for all permitted activities under the translocation are required by January 3 of each year of the translocation plan. Reporting will include all the data concerning desert tortoises, habitat, maps, health results, environmental data, and any additional information required by the USFWS recovery permit. Digital data will be entered into a USFWS/BLM-provided master database. Archived data will be deposited in a standardized data repository approved by the USFWS.

Adaptive management measures will be implemented during the translocation and subsequent monitoring activities for the life of the project. Adaptive management triggers may include activities that are projected to or have put the well-being of one or more tortoises at risk. Incidents that may affect the health or well-being of one or more desert tortoises can be addressed immediately and followed up through official communications. If there are concerns in the field that do not pose an immediate threat to one or more tortoises, a designated field representative may notify the agencies of proposed adaptive management decisions within 1 week.