The life history of Gopherus agassizii is typical of long-lived chelonians and has been reviewed by Berry and Murphy (2019). Tortoises require 17–20 years to reach sexual maturity at a straight-line carapace length (CL) of 18 cm or more (Woodbury and Hardy 1948, Turner et al. 1987, Medica et al. 2012). Variation in years is dependent on desert region, frequency of droughts, and quality of available forage. In the northern part of the geographic range, females smaller than 20.9 cm were not reproducing (Mueller et al. 1998).  Maximum lifespan was estimated by Turner et al. (1987) at 75 years, but few live beyond 50 yrs in the wild (Germano 1992). Generation time was estimated to be 20–32 years (Turner et al. 1987, USFWS 1994). Based on data from three desert regions, mean sizes of females ranges from 21.4 to 23.1 cm, whereas the mean sizes of males ranged from 24.3 to 24.9 cm; the largest desert tortoises on record, a male, reached 38.1 cm carapace length (Stebbins 2003), whereas a female was 37.4 cm, but these animals were exceptions (Berry and Murphy 2019).  

Mature females may lay clutches of one to 10 eggs in up to three clutches per year in spring and early summer; in some years, some females do not lay eggs (Rostal et al. 1994, Henen 1997, Mueller et al. 1998, Wallis et al. 1999, McLuckie and Fridell 2002, Ennen et al. 2012, Lovich et al. 2015). Annual fecundity ranges from 0 to 16 eggs (Mueller et al. 1998, Lovich et al. 2015). Several factors may affect egg production: site, year, size of female, size and number of eggs, and available water and protein from precipitation and forage in the year preceding egg laying, as well as the year eggs are laid (Henen 1997). 

Incubation times for eggs range from 67 to 104 days (Burge 1977, McLuckie and Fridell 2002, Ennen et al. 2012). Hatching success varies and appears to depend on year, location of the nest, and whether it is the first or second clutch. Eggs may be infertile or broken during laying (e.g., 12%; Turner et al. 1987). Many nests are destroyed by predators before hatching and the loss of eggs (and nests) varies by year (Turner et al. 1987); they estimated an average loss of 37.1% of nests in a multi-year study. Hatching success in intact nests, undisturbed by predators, has been shown to vary from 73 to 100% (McLuckie and Fridell 2002, Rostal et al. 2002, Bjurlin and Bissonette 2004, Ennen et al. 2012).

Desert tortoises inhabit desert scrub habitats, including saltbush, creosote bush, Joshua Trees and Mojave yuccas, and microphyll woodlands with ironwood, palo verde, desert willow, and smoke trees (Berry and Murphy 2019). In the northeastern part of their geographic range, they occur in an ecotone between the Mojave and Great Basin deserts with sand sagebrush and junipers. Actual occurrences tend to be in valleys, alluvial fans, bajadas, and ephemeral stream channels, although tortoises can be found in low sand dunes and on steep slopes of mesas and cliffs (Berry and Murphy 2019).

Desert tortoises are herbivorous and selective in their choice of plant species (Jennings 1993, Oftedal 2002, Oftedal et al. 2002, Jennings and Berry 2015). They primarily eat forbs when available. In years of abundant precipitation, they are selective feeders and prefer specific species of annuals and herbaceous perennials in the legume, mallow, borage, aster, four o’clock, and cactus families (as well as other families). Although they eat grasses, a diet solely of grasses is deficient in nutrients and is likely to inhibit growth and survival, especially in neonate, juveniles, and immature tortoises (Hazard et al. 2009, 2010; Drake et al. 2016). The quality and quantity of preferred plant foods has diminished because of continuing invasion of non-native annual grasses and forbs and increased fire associated with the highly combustible non-native grasses (D’Antonio and Vitousek 1992, Brooks and Berry 2006, Brooks and Matchett 2006, Berry et al. 2014b). 

Annual survival and mortality of adults is dependent on sex, size of the tortoise, frequency and severity of droughts, numbers and types of anthropogenic uses, location, and decade of study. In a multi-year study in the eastern Mojave Desert, annual survivorship of juveniles increased with size, ranging from 0.767 when <6.0 cm to 0.861 when 6.0 to 17.9 cm (Turner et al. 1987). When tortoises reach breeding age at an estimated 18.0 cm, survival rates were 0.87 to 0.944. Freilich et al. (2000) reported an annual survival of 0.883 for adults at Joshua Tree National Park. In a study in the Colorado Desert, Agha et al. (2015) estimated adult survival at a wind-turbine energy site (0.96) and an adjacent area (0.92). At two sites in the eastern Mojave Desert, Longshore et al. (2003) reported annual survival of adults of 0.985 and 0.829, with the lower survival rate at a site affected by drought.

Woodbury and Hardy (1948) estimated that 1% of adults died per year in a population mostly comprised of adults. In the northeastern Mojave Desert, Turner et al. (1984) reported mortality rates of 18.4% in a year of drought and 4.4% in a normal year. In the western Mojave Desert, death rates were lowest at a protected Research Natural Area (2.8%/yr) and highest in critical habitat (20.4%/yr). At Joshua Tree National Park, the mortality rate was 11.7% (Freilich et al. 2000), and in Red Rock State Park, 67% (Berry et al. 2008). In a demographic study of tortoises at 21 sites in the central Mojave Desert, mortality rates of adults ranged from 1.9 to 95.2% (Berry et al. 2006).

Turner et al. (1987) predicted an annual rate of population increase of ca. 2% in a model based on a tortoise subpopulation in the eastern Mojave Desert between 1977 and 1985. By 2000, this subpopulation had declined precipitously, apparently due to disease (see Christopher et al. 2003). Freilich et al. (2002) estimated the recruitment rate of young tortoises into the adult subpopulation at 0.092 in a plot in Joshua Tree National Park. This number of tortoises on this plot was thought to be stable between 1991 and 1995, but later declined (Lovich et al. 2014).

The Desert Tortoise was previously considered to be a single wide-ranging species, Gopherus agassizii, inhabiting the Mojave and Sonoran Desert regions of the southwestern United States and northwestern Mexico from southern California and Arizona through Sonora and into northern Sinaloa (Stebbins 1966, 2003; Iverson 1992). The species was found to be polytypic by Murphy et al. (2011), who split the morphologically and genetically distinct Sonoran Desert populations as Gopherus morafkai, the Sonoran Desert Tortoise. Further analysis demonstrated that G. morafkai was also polytypic and split further to separate and describe the Sinaloan Thornscrub Tortoise further to the south as Gopherus evgoodei (Edwards et al. 2016). 

Geographically restricted G. agassizii, the Mojave or Agassiz’s Desert Tortoise, is endemic to the United States, inhabiting southeastern California, southern Nevada, southwestern Utah, and extreme northwestern Arizona west and north of the Colorado River (TTWG 2017, Berry and Murphy 2019). The Sonoran Desert Tortoise, G. morafkai, occurs in both the United States and Mexico, inhabiting Arizona south and east of the Colorado River, Sonora (including Isla Tiburón), and extreme northern Sinaloa (Murphy et al. 2011, TTWG 2017). The Sinaloan Thornscrub Tortoise, G. evgoodei, is endemic to Mexico and occurs in southern Sonora, northern Sinaloa, and extreme southwestern Chihuahua (Edwards et al. 2016, TTWG 2017).

Within its geographic range, G. agassizii occurs in the Mojave Desert, the western Sonoran or Colorado Desert, the ecotone of the Mojave with the Great Basin Desert, and ecotones with vegetation types typical of higher elevations on the lower slopes of the Sierra Nevada, Transverse, Peninsular and desert mountain ranges (USFWS 1994). McLuckie et al. (1999) identified a subpopulation of G. agassizii east of the Colorado River in the Black Mountains of northwestern Arizona in which morphometric and mtDNA characteristics of the majority of the subpopulation were typically Mojavean; however, elements typical of tortoises in the Sonoran Desert were also evident. Edwards et al. (2015), using new genetic techniques, examined this and other nearby tortoise subpopulations, and identified hybrids (F2) in three mountain ranges near the Colorado River in Arizona. The two Gopherus species come in contact in limited places where Mojave Desert habitats meet Sonoran Desert habitats. The two species likely maintain largely independent taxonomic identities due to ecological niche partitioning (Inman et al. 2019). The species has been recorded at elevations of up to 1,570 m asl (Rautenstrauch and O’Farrell 1998); however, tortoises may be found in unusual places, often transported by humans or other animals (e.g., the type specimen of Xerobates lepidocephalus [Ottley and Velázquez-Solis 1989] from southern Baja California, Mexico, is actually an introduced Gopherus agassizii [Murphy et al. 2011]).

Conservation Measures taken:
The first legal conservation measures for Gopherus agassizii came from the State of California in 1939 (California Department of Fish and Game Code 1939–1981). Additional protective regulations followed until G. agassizii was listed as threatened under the California Endangered Species Act in 1989 (California Dept. of Fish and Wildlife 2016). Federal legislation to protect G. agassizii first occurred in 1980 and was restricted to the Beaver Dam Slope population in Utah (USFWS 1980). In 1989–1990, G. agassizii was federally listed as threatened (USDI 1990 and references therein). The only population of G. agassizii that is not protected by the Endangered Species Act of 1973, as amended, is in the northwest corner of Arizona (Edwards et al. 2015). Recovery efforts have been underway since 1990. The U.S. Fish and Wildlife Service (USFWS 1994) published the first Recovery Plan in 1994, coupled with designations of critical habitat units by the U.S. Department of the Interior (USFWS 1994); this was followed by a revised Recovery Plan in 2011 (USFWS 2011), and regional Recovery Implementation Teams established in 2012. These teams are chaired by an employee of the USFWS Desert Tortoise Recovery Office, and are composed of federal, state, and county employees from the range of the desert tortoise, including representatives from local and national conservation and other stakeholder organizations.

The species is included in CITES Appendix II as part of Testudinidae spp., requiring that any commercial international trade be documented not to be detrimental to the survival of wild populations. CITES Trade records generally show very low levels of international exports of live animals; the vast majority of live traded Desert Tortoises are personal pets moving in-country with their owners, and many of the records in fact concern seizures of illegally transported specimens (CITES UNEP-WCMC trade database).

Conservation and recovery efforts began in the early 1970s, long before efforts of the federal actions by the USFWS in 1989–90. The Desert Tortoise Council formed in 1974-75 out of an interim recovery effort involving the four Southwestern states. This non-profit corporation was and continues to be dedicated to preserving representative populations of desert tortoises; educating the public; holding annual introductory workshops; and annual symposia to bring together representatives from government agencies, academia, and the public to learn and discuss important topics aimed at recovery of tortoise populations. The Desert Tortoise Council was instrumental in providing critical materials for federal and state listings of the species. The Desert Tortoise Preserve Committee, Inc., was formed in 1974 to establish protected areas for G. agassizii. This non-profit organization is a land trust and mitigation bank, a source of education, and research. They were instrumental in establishing the Desert Tortoise Research Natural Area and increasing its size.

Two preserves or protected areas exist with moderately high degrees of protection. One is the 100 km² (and increasing) Desert Tortoise Research Natural Area, which was formally designated by the U.S. Congress in 1980. It is fenced, with no vehicle access, livestock grazing, mining, or surface disturbances other than a few limited natural trails and a kiosk. The Natural Area is for wild tortoises only and populations are allowed to fluctuate naturally with no augmentation. Population density of adults throughout the Natural Area in 2011-12 was 10.2/km² (Berry et al. 2014a). The second preserve is Red Cliffs Desert Reserve in Utah (251 km²). The Red Cliffs National Conservation Area provides additional protection for federal lands within the Reserve. Several paved roads, fenced and unfenced, run through the Reserve and recreation occurs throughout (e.g., hiking, horseback riding, mountain biking). The next and lower level of protection could be described as occurring within National Parks, State Parks, and National Recreation Areas such as Joshua Tree and Death Valley National Parks, Mojave National Preserve, Red Rock Canyon, Anza-Borrego, and Red Rocks State Parks, Lake Mead National Recreation Area, and the Beaver Dam Wash National Conservation Area. These parks and recreation areas have very high visitor use, unfenced paved roads, and some illegal collecting and release of captive tortoises of one or more species.

Twelve critical habitat units, the basis for Tortoise Conservation Areas (term defined in USFWS 2011), were designated by the USFWS (1994), and have far less protection than either the Desert Tortoise Research Natural Area or the Red Cliffs Desert Reserve and are subject to multiple land uses that fragment and degrade habitat and create vulnerabilities and risks to the tortoises (e.g., invasive non-native grasses and other non-native species; highways; roads; utility poles, towers, and electrical transmission lines; gas lines and fibreoptic cables; recreational vehicle use; shooting; domestic and feral dogs; cattle grazing and feral burros; mining; military installations; fire that causes degradation of habitat).

Seventeen monitored subpopulations in the 12 critical habitat units are contained within five recovery units which cover a total of 25,678 km². The following information for each recovery unit and the 17 Tortoise Conservation Areas reports area (km²), and density of breeding adults per km² in 2014. Western Mojave Recovery Unit: Fremont-Kramer (2,347 km², 2.6/km²), Ord-Rodman (852 km², 3.6/km²), Superior-Cronese (3,094 km², 2.4/km²); Colorado Desert Recovery Unit: Chocolate Mountains Aerial Gunnery Range (713 km², 7.2/km²), Chuckwalla (2,818 km², 3.3/km²), Chemehuevi (3,763 km², 2.8/km²), Fenner (1,782 km², 4.8/km²), Joshua Tree (1,152 km², 3.7/km²), Pinto Mountain (508 km², 3.4/km²), Piute Valley (927 km², 5.3/km²); Eastern Mojave Recovery Unit: El Dorado Valley (999 km², 1.5/km²), Ivanpah Valley (2,447 km², 2.3/km²); Northeastern Mojave Recovery Unit: Beaver Dam Slope (750 km², 6.2/km²), Coyote Spring (960 km², 4.0/km²), Gold Butte (1,607 km², 2.7/km²), Mormon Mesa (844 km², 6.4/km²); Upper Virgin River Recovery Unit: Red Cliffs Desert Reserve (115 km², 15.3/km²) (USFWS 2015; Allison and McLuckie 2018). The overall decline in tortoise populations in critical habitats (Tortoise Conservation Areas) between 2004 and 2014 was 32.2% (USFWS 2015). Four of the five recovery units are in a state of decline, with 11 of the 17 subpopulations registering declines in adult tortoises ranging from 26.6 to 64.7% during the 10 years (USFWS 2015). Most of the increasing subpopulations were in Nevada. Population densities for adults ranged from 1.5 to 7.2/km² in declining populations as of 2014 (USFWS 2015).

Extensive research has been published in peer-reviewed journals on many aspects of natural history, general ecology, physiological ecology, reproduction, health and diseases, population attributes, causes of death, movements and home range, predators, head-starting, translocation, and many other topics, making G. agassizii likely the most well-researched non-marine turtle species (Lovich and Ennen 2013b). Over 400 journal articles were published as of 2018, most between 1990 and 2018, as well as hundreds of reports (see three annotated bibliographies covering almost 160 years: Hohman et al. 1980, Grover and DeFalco 1995, Berry et al. 2016). Some information has been integrated into recovery programs, but many of the recovery measures recommended in the first Recovery Plan (USFWS 1994) have not been implemented as of 2020.

Economic relevance: The approximate cost of USD 100 million to develop and implement the first and second Recovery Plans is significant within the regulatory, scientific and local economic sectors involved and much remains to be implemented (USFWS 1994, 2011; Averill-Murray et al. 2012).


Conservation Measures needed:
The USFWS (1994) published recommended regulations for the areas that were designated as critical habitat. They described activities to be prohibited (e.g., all vehicle activity off designated roads; all competitive and organized recreational vehicle events on designated roads; habitat destructive military manoeuvres, clearing for agriculture, landfills and other surface disturbances; domestic livestock grazing, grazing by feral burros and horses; vegetation harvest; collection of biological specimens or vegetation harvest except by permit; dumping and littering; and deposition of captive or displace desert tortoises except under authorized translocation research projects; uncontrolled dogs out of vehicles; discharge of firearms except for hunting of game between September and February. There were many other recommended management actions but few of these recommendations were adopted when critical habitat units were officially described (USFWS 1994), and others have only been partially implemented by 2020. There were also recommendations for monitoring and research. In the second recovery plan, the USFWS (2011) identified and ranked (Darst et al. 2013) priority actions for recovering the Desert Tortoise and established regional Recovery Implementation Teams to implement these recovery actions. These Recovery Implementation Teams identify local, regional, and range-wide actions by submitting proposals to team members for discussion and prioritization. Ultimately the proposals are submitted to range-wide Management Oversight Groups composed of state, federal, and county government agencies for review, discussion, and potential sources of funding. Some projects are successfully funded and implemented, while many recommended in 1994 remain unfulfilled.


 In association with the following standardized categories of Conservation Actions Needed, we provide the following notes:
1.1. Land/water protection -> Site/area protection
    a. Better protection of Critical habitats could ensure that populations of tortoises become stable and/or increase. Examples of protective measures included in recovery measures for the tortoise are exclusion fencing and culverts along highways and roads; reduction in populations of hyper-predators such as the Common Ravens; control and removal of newly introduced and previously existing non-native plants; and control of recreational vehicle use.

2.1. Land/water management -> Site/area management
    a. The first recovery plan identified site-specific or critical habitat-specific measures to ensure protection of habitat and reduction of deaths of tortoises from anthropogenic sources (USFWS 1994). Most of these recommendations are still relevant. The Recovery Implementation Teams have provided recommendations similar to those in the first recovery plan. Many of these measures remain to be implemented. For example, in the State of California where most desert tortoise habitat and populations occur, acquisition of private land would be beneficial, because a substantial portion of habitat is in multiple private ownership. Both the USFWS and State of California recommend that developers of tortoise habitat acquire replacement habitat for habitat lost to development, and such actions have been occurring for ~20 years. Another topic and critical area that would benefit from protection is the population and hybrid zone with G. morafkai east of the Colorado River in Arizona (Edwards et al. 2015). This small population is not protected under the federal Endangered Species Act (Edwards et al. 2015).

2.2. Land/water management -> Invasive/problematic species control
  a. Non-native grasses (e.g., Schismus arabicus, S. barbatus, Bromus tectorum, B. madritensis rubens) and forbs (e.g., Brassica tournefortii, Hirschfeldia incana) present serious and severe problems to tortoises because tortoises are selective in the choice of forage (Jennings and Berry 2015). The non-native annuals contribute to changes in forage availability, habitat structure, and increases in fire (D’Antonio and Vitousek 1992). These non-native species thrive under disturbance and spread via roads, livestock, military maneuvers, and disturbances created by recreational vehicle use off-road (e.g., D’Antonio and Vitousek 1992, Brooks and Berry 2006, Brooks et al. 2006, Brooks and Matchett 2006). The grasses are highly combustible and fire-prone in wildlands that did not evolve with short-term fire cycles (D’Antonio and Vitousek 1992). The grasses also compete with native annuals used as forage by the tortoises, and the species of grasses contain little nutrition, require water to metabolize, cause weight loss in the tortoises, and can become embedded in the jaws (Medica and Eckert 2007; Hazard et al. 2009, 2010; Drake et al. 2016). Similarly, Brassica tournefortii competes with native species used for forage and often occurs in dense stands, inhibiting movements of tortoises (Berry et al. 2014b).

3.2. Species management -> Species recovery
  a. Species management and recovery are guided by the Recovery Plan and the U.S. Fish and Wildlife Service. On-the-ground management is by the administering agency, e.g., U.S. Bureau of Land Management, National Park Service, Department of Defense, States (for state land), and private owners. That being said, much can be done by implementing actions recommended in the first Recovery Plan (USFWS 1994) and by restoring degraded habitat (e.g., Abella and Berry 2016); controlling recreation vehicle use off-road and reducing fragmentation of habitat; limiting spread of invasive, non-native grasses and forbs; controlling hyper-predation in common ravens (USFWS 2008) and coyotes; preventing dogs and dog packs from running loose in the desert; and acquiring habitat.

4.3. Education & awareness -> Awareness & communications
  a. See Conservation Actions in Place. Expansion of on-going programs to prevent take or shooting in the wild and release of captive tortoises of several species.


In association with the following standardized categories of Research Needed, we provide the following notes:

While research on some topics is desirable, more is known about G. agassizii than most other reptiles (Lovich and Ennen 2013b, Berry et al. 2016 and references therein). Instead, implementation of previously identified actions to protect populations and habitat is more critical, specifically actions that will reduce deaths and loss or degradation of habitat. 

1.1. Research -> Taxonomy
  a. Genetic relationships between and within populations: human-mediated translocations of tortoises have occurred for decades, some authorized, some not (see Murphy et al. 2007). One recent question is the source of tortoises in Anza Borrego Desert State Park in the Colorado Desert of California. One might expect that the source would be tortoises occurring in the Colorado Desert, but instead tortoises have genotypes typical of the southwestern Mojave Desert population (Manning and Edwards 2019). More information on nearby tortoises (e.g. Lovich et al. 2020) occurring on the east-facing slopes of the Peninsular Range north of the Park may shed light on whether this is a naturally occurring population or a source that came from human-mediated translocations.
  b. Translocation of thousands of tortoises has occurred in the last >20 years. Yet the only information available as to whether these translocated tortoises have been assimilated into the recipient or existing resident populations is research by Mulder et al. (2017) on assimilation of translocated males into the population of resident tortoises. Much more needs to be done on following males and females over a 10- to 20-year period to determine if and when adult males are assimilated into resident populations.

1.2. Research -> Population size, distribution & trends
  a. More information on current population attributes such as size-age class structure, recruitment of juveniles into adult populations, sex ratios of adult tortoises, and causes and contributors to death is highly desirable. Landscape sampling undertaken and managed by the U.S. Fish and Wildlife Service’s Desert Tortoise Recovery Office has provided valuable region-wide information on adult densities but not on other essential population attributes (i.e., Allison and McLuckie 2018). Resurvey of long-term, mark-recapture tortoise plots has been spotty for the past 20 years while support has increased for line-distance sampling representatively and on a landscape scale (see USFWS 2015, Allison and McLuckie 2018). Nonetheless, it is clear (USFWS 2011) that species recovery cannot be assumed based on patterns of adult counts alone, and active work to describe vital rates across the range will be an important part of assuring tortoise populations reflect healthy population dynamics or determining regional and size-specific recovery needs.

1.3. Research -> Life history & ecology
  a. More information is needed on survival of neonate, juvenile, and immature size classes (first 12 to 15 years of life) and causes of mortality in the wild. Frequent input of new data on causes of and contributors to mortality for all size classes is essential for improving management of the species and for achieving upward trends.

1.5. Research -> Threats
  a. The USFWS developed a model to identify major threats to the species (Darst et al. 2013); the information in this model is based on published research only, and not on the hundreds of reports and manuscripts available in Annual Reports to the USFWS on research permits. The model is outdated and needs major revisions to more accurately reflect available information and more recent priorities. In addition, support could be provided to speed up publication of important research projects that will lead to more protective management actions.   

3.1. Monitoring -> Population trends
  a. Monitoring is especially needed on population attributes in critical habitat, near highways, and in critical habitat near urban areas.

3.4. Monitoring -> Habitat trends
  a. Monitoring is especially needed on wildfires, non-native plants, seed beds, and recovery of preferred forage plants.