H. Giant Kangaroo Rat (Dipodomys ingens)

Endangered Species Recovery Program
Home \| News \| Publications \| Species profiles \| Data and maps \| About \| Staff \| Links \| Department of Biological Sciences \| CSU Stanislaus

Recovery Plan for Upland Species of the San Joaquin Valley, California
Contents . Introduction . Species accounts . Recovery . Stepdown . Implementation . References . Appendix

H. Giant Kangaroo Rat (Dipodomys ingens)

1. Description and Taxonomy

Taxonomy.-- Dipodomys ingens was described as Perodipus ingens by Merriam (1904a), who listed the type locality as Painted Rock, 20 miles SE Simmler, Carrizo Plain, San Luis Obispo County, California. The type locality was amended to 41 kilometers (25 miles) SE of Simmler by Williams and Kilburn (1991). The genus name Perodipus was used for several years to include all the kangaroo rats with five toes on the hind feet. Grinnell (1921) relegated Perodipus to a synonym of Dipodomys. This taxonomy has been sustained in the latest taxonomic review of the family Heteromyidae (Williams et al. 1993a).

Description.-- The giant kangaroo rat is adapted for bipedal locomotion (two-footed hopping) (Eisenberg 1963). The hind limbs are large compared to the size of the forelimbs; the neck is short; and the head is large and flattened. The tail is longer than the combined head and body length and has a dorsal crest of long hairs towards the end of the tail, terminating in a large tuft (Figure 37). Large, fur-lined cheek pouches open on each side of the mouth. The pouches extend as deep invaginated pockets of skin folded inward along the sides of the head (Grinnell 1922).

Identification.-- Giant kangaroo rats are distinguished from the coexisting species, San Joaquin kangaroo rat (D. nitratoides) and Heermanns kangaroo rat (D. heermanni), by size and number of toes on the hind foot. The hind feet of adult giant kangaroo rats each have five toes and are longer than 47 millimeters (1.85 inches) (Best 1993). The giant kangaroo rat is the largest of more than 20 species in the genus (Grinnell 1922, Hall 1981, Best 1993). Grinnell (1932a) reported a mean mass of 157.0 grams (5.54 ounces) for 15 adult males and 151.4 grams (5.34 ounces) for 7 adult females. Adult Heermanns kangaroo rats average 65 to 80 grams (2.29 to 2.82 ounces), with maximum weights not exceeding about 90 grams (3.17 ounces) (Williams 1992); the hind foot also has five toes but individuals feet usually measure less than 45 millimeters (1.77 inches) (Best 1993). Average weight of San Joaquin kangaroo rats is less than 45 grams (1.59 ounces), and they have four toes on each hind foot. Length of the hind foot does not exceed 39 millimeters (1.54 inches) (Grinnell 1922).

Figure 37. Illustration of the giant kangaroo rat (drawing by Jodi Sears, based on photo © by D.F. Williams).

2. Historical and Current Distribution

Historical Distribution.-- Up until the 1950s colonies of giant kangaroo rats were spread over hundreds of thousands of acres of continuous habitat in the western San Joaquin Valley, Carrizo Plain, and Cuyama Valley (Grinnell 1932a, Shaw 1934, Hawbecker 1944, 1951). The historical distribution of giant kangaroo rats encompassed a narrow band of gently sloping ground along the western edge of the San Joaquin Valley, California, from the base of the Tehachapi Mountains in the south, to a point about 16 kilometers (10 miles) south of Los Banos, Merced County in the north; the Carrizo and Elkhorn Plains and San Juan Creek watershed west of the Temblor Mountains, which form the western boundary of the southern San Joaquin Valley; the upper Cuyama Valley next to and nearly contiguous with the Carrizo Plain; and scattered colonies on steeper slopes and ridge tops in the Ciervo, Kettleman, Panoche, and Tumey Hills, and in the Panoche Valley (Figure 38). Within this circumscribed geographic range were abut 701,916 to 755,844 hectares (1,734,465 to 1,867,723 acres), which included different estimates of the amount of nonhabitat depending on different assumptions. The most liberal estimate of historical habitat was about 631,724 hectares (1,561,017 acres; Williams 1992).

Figure 38. Distributional records of the giant kangaroo rat (Dipodomys ingens)

Current Distribution.-- The species population is currently fragmented into six major geographic units: A) the Panoche Region in western Fresno and Eastern San Benito Counties; B) Kettleman Hills in Kings County; C) San Juan Creek Valley in San Luis Obispo County D) western Kern County in the area of the Lokern, Elk Hills, and other uplands around McKittrick, Taft, and Maricopa; E) Carrizo Plain Natural Area in eastern San Luis Obispo County; and F) Cuyama Valley in Santa Barbara and San Luis Obispo Counties (Figure 39; Williams 1980, 1992, OFarrell et al. 1987a, Williams et al. 1995). These major units are fragmented into more than 100 smaller populations, many of which are isolated by several miles of barriers such as steep terrain with plant communities unsuitable as habitat, or agricultural, industrial, or urban land without habitat for this species. Extant habitat was last estimated to be 11,145 hectares (27,540 acres), about 1.8 percent of historical habitat (Williams 1992).

Figure 39. Locations of extant populations of giant kangaroo rats (Dipodomys ingens)

Within the area of currently occupied habitat, populations of giant kangaroo rats have expanded and declined with changing weather patterns since 1979. At their peak in 1992 to 1993, there probably were about 6 to 10 times more individuals than at their low point in spring of 1991, when a majority of the 11,145 hectares (27,540 acres) probably was uninhabited and most of the rest was inhabited by less than 10 percent of peak numbers (Williams 1992, Williams et al. 1993b, Williams et al. 1995, Allred et al. in press, Williams and Nelson in press, D.F. Williams unpubl. data).

3. Life History and Habitat

Food and Foraging.-- Giant kangaroo rats are primarily seed eaters, but also eat green plants and insects. They cut the ripening heads of grasses and forbs and cure them in small surface pits located on the area over their burrow system (Shaw 1934, Williams et al. 1993b). They also gather individual seeds scattered over the grounds surface and mixed in the upper layer of soil. Surface pits are uniform in diameter and depth (about 2.5 centimeters, 1 inch), placed vertically in firm soil, and filled with seed pods. After placing seeds and seed heads in pits, the animal covers them with a layer of loose, dry dirt. Pits are filled with the contents of the cheek pouches after a single trip to harvest seeds. Before being moved underground, the seeds, including filaree and peppergrass (Lepidium nitidum), are sun-dried which prevents molding (Shaw 1934).

Individuals in many populations of D. ingens also make large stacks of seed heads on the surfaces of their burrow systems (Hawbecker 1944, Williams et al. 1993b). The material is cured, then stored underground. Amounts cached in surface stacks may not correspond with annual herbaceous productivity. No stacks were found in 1990, a year with no seed production, and 1991, a year with the second highest plant productivity between 1987 and 1994 (Williams and Nelson in press).

Grinnell (1932a, p. 313) examined three nursing females who had their cheek pouches "literally crammed with green stuff", and speculated that green foliage might be an important part of the diet during lactation. Other individuals, including a young female and adult males, were captured with foliage and fruits of peppergrass and foliage of filaree in their cheek pouches (Grinnell 1932a). In captivity, giant kangaroo rats have been maintained for periods from 2 weeks to more than 2 years on a diet of air-dried seeds, consisting primarily of millet, oat, and sunflower, occasionally supplemented with green plants. Of the green plants, captives preferred forbs to annual grasses, and usually ignored the blades of perennial grasses (Williams and Kilburn 1991). Shaw (1934) found a live insect of the bee and wasp fmily in the cheek pouch of a giant kangaroo rat. Eisenberg (1963) kept a giant kangaroo rat in captivity on a diet that included seeds, lettuce, and mealworm (darkling beetle) larvae (Tenebrio sp.).

Giant kangaroo rats forage on the surface from around sunset to near sunrise, though most activity takes place in the first 2 hours after dark. Foraging activity is greatest in the spring as seeds of annual plants ripen. Typically, plants such as peppergrass ripen first, and early caches, mostly in pits instead of stacks, consist of pieces of the seed-bearing stalks of this and other early-ripening species. The ability to transport large quantities of seeds and other food in cheek pouches and their highly developed caching behaviors, coupled with relatively high longevity of adults with established burrow systems, probably allow giant kangaroo rats to endure severe drought for 1 or 2 years without great risk of population extinction (Williams et al. 1993b, D.F. Williams unpubl. data).

Reproduction and Demography.-- Results of studies conducted between 1987 and 1995 in colonies on the Elkhorn and Carrizo Plain indicated that giant kangaroo rats have an adaptable reproductive pattern that is affected by both population density and availability of food. During times of relatively high density, females have a short, winter reproductive season with only one litter produced and there is no breeding by young-of-the-year. This was true both in years of high plant productivity and drought. In contrast, populations at low densities continue to breed into summer during drought. In 1990, a year of severe drought and no seed production, most females appeared not to reproduce; the few that bred apparently failed to raise young. In most years, females were reproductive between December and March or April, but in colonies with low densities, reproduction extended into August or September (Williams et al. 1993b, Williams and Nelson in press, Endangered Species Recovery Program unpubl. data). Mating strategies are being studied on the Carrizo Plain by Dr. Jan Randall. Initial results indicate that mating strategies are flexible and may be responding to the age of males, proximity of females, and changes in sex ratios (Hekkala 1995).

Giant kangaroo rats can breed the year of their birth when environmental and social conditions permit (sufficient food and space). At the Soda Lake colony, juvenile females had their first litters at an estimated mean age of 5 months. Some females had two to three litters per year. This relatively high rate of reproduction probably was promoted by high plant productivity and low population density (Williams and Nelson in press).

Little information is available on age-specific litter size. The mean of known embryo counts and litter sizes is 3.75, probably a value higher than the number born (Williams and Kilburn 1991, D.F. Williams unpubl. data). Dr. Jan Randalls research showed that gestation was 30 to 35 days (Hekkala 1995). During a post-drought January through May breeding season, 44 percent of the litters contained two young. One female had a litter of three, the remaining 39 percent had a litter of one.

The major time for dispersal of giant kangaroo rats seems to be following maturation of young, about 11 to 12 weeks after birth. However in years of high density, when most or all burrow systems are occupied, most young appear to remain in their natal burrows until opportunity to disperse arises or they finally are driven off by the mother or one of the siblings. Under these circumstances, death or dispersal of the resident does not leave a burrow system vacant for long. Williams and Nelson (in press) found on a study site at Soda Lake, San Luis Obispo County that more females than males dispersed although males more often moved longer distances. Females had a nearly 60 percent greater survival rate than males. Dispersal of adults with established burrow systems was occasionally detected; one adult male moved more than 120 meters (131.2 yards) from his established home to take up a new residence in a new burrow system he constructed (Williams et al. 1993b, Williams and Nelson n press, Williams and Tordoff 1988).

Estimated home range size ranges from about 60 to 350 square meters (71.8 to 418.6 square yards). There is no significant difference in size of home range between sexes. The core area of the territory, located over the burrow system (precinct) is the most intensely used location in the home range (Braun 1985). Grinnell (1932a) and Shaw (1934) suggested that territories were occupied by a single animal. More recent studies indicate that multiple individuals may live in precincts. These appeared to be family groups of females and offspring of different ages (Randall 1997).

Estimates of density, employing both trapping and counts of precincts ranged from 1 to 110 animals per hectare (1 to 44 animals per acre) (Grinnell 1932a, Braun 1985, Williams 1992). Changes in density generally coincide with amount of rainfall and herbaceous plant productivity, though numbers in populations studied in 1989 remained high despite drought and low plant productivity (Figure 40). Large seed caches made in spring 1988 probably carried individuals through 1989 and 1990 during drought (Williams et al. 1993b, Williams and Nelson in press, D.F. Williams unpubl. data). The population on the Elkhorn Plain typically was at much higher density than other populations recently studied, and fluctuated less than populations elsewhere, suggesting that the habitat on this part of the Elkhorn Plain is some of the best remaining.

Figure 40. Numbers of giant kangaroo rats captured during August censuses, Elkhorn Plain

Population Genetics.-- Partial results of on-going studies of population genetics of giant kangaroo rats provide guidance for designing a recovery strategy. The northern populations in Fresno and San Benito Counties are highly differentiated genetically from the southern populations on the Carrizo Plain Natural Area.

The genetic structure of the Carrizo Plain population differs from northern populations in that it has effectively acted as one large population, though the genetic data strongly suggest that the inhabited areas there have gone through episodes of substantial expansion and contraction in size (Mosquin et al. in press). This is consistent with recent observations from population censuses (Williams 1992, Williams et al. 1993b, Williams and Nelson in press, Allred et al. in press, D.F. Williams unpubl. data).

In the north, the population along the edge of the Valley at the eastern base of Monocline Ridge (San Joaquin Valley population) is substantially differentiated genetically from the other large population in the southeastern end of Panoche Valley (Figure 41). These two populations show little evidence of gene flow between them, and the San Joaquin Valley population is closer genetically to the Carrizo Plain population than any other of the semi-isolated northern populations. Clearly, this represents the remnant of the historical population that was distributed along the western edge of the Valley between Merced and Kern Counties. The two large, northern populations (San Joaquin Valley and Panoche Valley) appear to have been the sources of the small, semi-isolated populations on ridge-tops in the Ciervo and Tumey Hills. These latter populations are differentiated from both of the large populations, and from each other. They appear to have played the major role in gene flow between the Panoche Valley (Figure 41, see area B) and San Joaquin Valley populations. Interpopulation movements appear to have been achieved over relatively long periods in a stepping-stone manner between small populations on these ridge tops. Though small, they contain a significant proportion of the rare and unique genes of the northern population (Mosquin et al. in press).

Figure 41. Distribution of extant colonies of giant kangaroo rats (Dipodomys ingens) in their northern geographic range

The genetic studies show that effective population size (number of successfully-breeding individuals) in the north is smaller than current population size, indicating there has been a large increase in the northern population size very recently. This is consistent with the increase measured after the end of the drought in 1991 (Williams et al. 1995). In the south, estimated effective population size is slightly greater than current population size, indicating that current and hstorical population sizes are approximately the same (Mosquin et al. in press).

The genetic structure of giant kangaroo rat populations also shows that the effective dispersal distance of giant kangaroo rats (i.e., dispersal of genes) is much greater than predicted on the basis of capture-recapture and behavioral studies. Results from trapping of kangaroo rats show most movements are less than 100 meters (330 feet) and rarely as much as 1 kilometer (0.62 mile) (Jones 1988, 1989, Williams and Nelson in press). The genetic data suggest that effective distances are several times greater than 1 kilometer (0.62 mile). There are too few data, and analyses are too incomplete to make a precise estimate, but they do suggest effective dispersal over several kilometers and through highly inhospitable habitat in the northern population (Mosquin et al. in press).

Behavior and Species Interactions.-- Little direct evidence exists on aggression by giant kangaroo rats, but they seem to be much more aggressive than the two co-occurring species. Wherever giant kangaroo rats were found by Grinnell (1932a), they dominated the community to the exclusion of other rodent species. Hawbecker (1944, 1951) and Tappe (1941) corroborated Grinnells observations, finding that giant kangaroo rats excluded all other nocturnal rodents from areas where they occurred.

Braun (1983), however, found that a population of giant kangaroo rats on the Carrizo Plain, San Luis Obispo County, did not exclude other species of rodents to the extent reported by others. Braun (1983) believed that the lack of exclusivity supported the hypothesis that this population was living in suboptimal habitat.

The giant kangaroo rat, by its relative abundance and burrowing activity, is a keystone species in grassland and shrub communities (Schiffman 1994, Goldingay et al. 1997). When abundant locally, giant kangaroo rats are a significant prey item for many species, including San Joaquin kit foxes (an umbrella species), American badgers (Taxidea taxus), coyotes (Canis latrans), long-tailed weasels (Mustela frenata), burrowing owls (Athene cunicularia), barn owls (Tyto alba), great horned owls (Bubo virginianus), and short-eared owls (Asio flammeus). Snakes seen within giant kangaroo rat colonies included the coachwhip (Masti- cophis flagellum), gopher snake (Pituophis melano- leucus), common king snake (Lampropeltis getulus), and western rattlesnake (Crotalis viridis; Williams 1992). Giant kangaroo rat burrows also are used by blunt-nosed leopard lizards and San Joaquin antelope squirrels. On the Carrizo Plain Natural Area, the endangered California jewelflower grows primarily on the burrow systems of giant kangaroo rats (Cypher 1994a). In spring, precincts show as distinct, evenly-spaced, dark green patches because of the more lush growth of herbaceous plants compared to intervening spaces (Grinnell 1932a). Measurements of plant productivity on and off precincts over an 8-year period show that when rainfall was sufficient to promote growth and fruiting of plants, the net productivity of herbaceous plants was two to five times greater on precincts than surrounding ground (Hawbecker 1944, Williams et al. 1993b, Williams and Nelson in press). Further, growth of herbaceous plants on precincts contained about 4 percent more protein than plants from surrounding ground. These differences were attributed directly to the presence and activities of the giant kangaroo rats (Williams et al. 1993b).

Activity Cycles.-- Giant kangaroo rats are active all year and in all types of weather. They do not migrate or become dormant or torpid. Although primarily nocturnal, giant kangaroo rats have been seen above ground during daylight, including midday in the hottest part of the year (Williams et al. 1993b, Williams and Tordoff 1988). Giant kangaroo rats typically emerge from their burrows soon after sunset and are active for about 2 hours (time of first emergence to time of last disappearance). There usually is no second period of activity before dawn. Animals are above ground only forabout 15 minutes per night. Activity patterns appear to be unaffected by distance from the home burrow, snow, rain, wind, moonlight, or season (Braun 1985).

Habitat and Community Associations.-- Historically, giant kangaroo rats were believed to inhabit annual grassland communities with few or no shrubs, well-drained, sandy-loam soils located on gentle slopes (less than 11 percent) in areas with about 16 centimeters (6.3 inches) or less of annual precipitation, and free from flooding in winter (Grinnell 1932a, Shaw 1934, Hawbecker 1951). However, more recent studies in remaining fragments of historical habitat found that giant kangaroo rats inhabited both grassland and shrub communities on a variety of soil types and on slopes up to about 22 percent and 868 meters (2,850 feet) above sea level. This broader concept of habitat requirements probably reflects the fact that most remaining populations are on poorer and marginal habitats compared to the habitats of the large, historical populations in areas now cultivated. Yet these studies demonstrated that the preferred habitat of giant kangaroo rats still was annual grassland communities on gentle slopes of generally less than 10 percent, with friable, sandy- loam soils. Few plots in flat areas were inhabited, probably because of periodic flooding during heavy rainfall (Williams 1992, Williams et al. 1995, Allred et al. in press).

Below about 400 meters (1,300 feet) at Panoche Creek in western Fresno County and in the Lokern, Buena Vista Valley, and Elk Hills regions of the southern San Joaquin Valley, giant kangaroo rats are found in annual grassland and saltbush scrub. Scattered common and spiny saltbushes characterize areas where giant kangaroo rats are associated with shrubs. The most common herbaceous plants are red brome, annual fescue, and red-stemmed filaree (Williams 1992).

Upper Sonoran subshrub scrub associations support relatively large populations of giant kangaroo rats at elevations above about 400 meters (1,300 feet). In the southern portion of the extant geographic range of giant kangaroo rats, these communities are characterized by open stands of the dominant shrub, California ephedra. Annual grasses and forbs, particularly red-stemmed filaree, peppergrass, and Arabian grass dominate areas between shrubs. Giant kangaroo rats are most numerous where annual grasses and forbs predominate, with scattered ephedra bushes and fewer shrubs such as Anderson desert thorn (Lycium andersonii), eastwoodia (Eastwoodia elegans), and pale-leaf goldenbush Isocoma acradenia var. bracteosa) (Williams 1992).

Above about 600 meters (2,000 feet) in elevation, eastwoodia, California buckwheat, winter fat (Krascheninnikovia lanata), and chaparral yucca (Yucca whipplei) are more common on steep slopes (greater than about 5 to 6 percent) and sandy ridgetops. Cheesebush (Hymenoclea salsola) and matchweed are common only in arroyos. Only satellite colonies of giant kangaroo rats or scattered individuals are found in these latter associations. In the northern portion of the geographic range of giant kangaroo rats, Anderson desert thorn is absent; otherwise, the woody shrubs comprising the ephedra community are the same or closely-related species (Williams 1992, Williams et al. 1995).

4. Reasons for Decline and Threats to Survival

Reasons for Decline.-- Until the late 1960s and early 1970s, little land within the historical range of the giant kangaroo rat had been permanently cultivated and irrigated or otherwise developed. Completion of the San Luis Unit of the Central Valley Project and the California Aqueduct of the State Water Project resulted in rapid cultivation and irrigation of natural communities that had provided habitat for giant kangaroo rats along the west side of the San Joaquin Valley (Williams 1992, Williams and Germano 1993). Between about 1970 and 1979, almost all the natural communities on the western floor and gentle western slopes of the Tulare Basin were developed for irrigated agricultur, restricting occurrence of most species of the San Joaquin saltbush and Valley Grassland communities, including the giant kangaroo rat. This rapid habitat loss was the main reason for its listing as endangered. At the time of its listing, relatively little of its extant habitat was publicly owned or protected from possible destruction.

Use of rodenticide-treated grain to control ground squirrels and kangaroo rats also may have contributed to the decline of giant kangaroo rats in some areas. From the 1960s into the early 1980s rodenticides such as Compound 1080 were often broadcast over broad areas by airplane. Today, there are large areas in the Sunflower Valley (western corners of Kings and Kern Counties), Kettleman and Tent Hills in Kings County, and the eastern foothills of the Panoche Hills, Fresno County, that show characteristic features of giant kangaroo rat precincts, but are unoccupied by kangaroo rats. Williams (1992) believed that populations in these areas may have been eliminated by use of rodenticides.

Based on remarks by Grinnell (1932a) and Shaw (1934), giant kangaroo rats can survive in areas that have been grazed to a point where almost no plant material remains. It is not known, however, if they could survive indefinitely if those grazing intensities were sustained.

Destruction of natural communities to develop the infrastructure for petroleum exploration and extraction also has reduced habitat for giant kangaroo rats and contributed to their decline, especially in the area around Coalinga, Fresno County, and in the oil fields of western Kern County. The small cities and towns along the western edge of the San Joaquin Valley between Coalinga and Maricopa also have developed on what was once habitat for giant kangaroo rats. These developments, plus mineral extraction, roads and highways, energy and communications infrastructures, and agriculturally related industrial developments collectively have contributed to the endangerment of the giant kangaroo rat, but were not as important as loss of habitat by cultivation.

Threats to Survival.-- Since listing as endangered (USFWS 1987), conversion of habitat for giant kangaroo rats has slowed substantially, because most tillable land has already been cultivated and because of a lack of water for irrigation. However, urban and industrial developments, petroleum and mineral exploration and extraction, new energy and water conveyance facilities, and construction of communication and transportation infrastructures continue to destroy habitat for giant kangaroo rats and increase the threats to the species by reducing and further fragmenting populations. Though many of these recent and future losses will be mitigated for by protecting habitat elsewhere, they still result in additional loss and fragmentation of habitat. Habitat degradation due to lack of appropriate habitat management on conservation lands, especially lack of grazing or fire to control density of vegetation (including shrubs) may be a threat to giant kangaroo rats (Williams and Germano 1993).

Though 60 population monitoring plots, range-wide, for giant kangaroo rats were established in 1995 by the Endangered Species Recovery Program (Williams and Kelly in litt. 1994a), there are no funds obligated to carry out a monitoring program in the future. Regular monitoring is important to any endangered species management program. Without monitoring, the effects of management prescriptions cannot be properly evaluated or altered in response to changes in populations due to both management actions and environmental variation. Perhaps no active management program is needed for giant kangaroo rat habitat, but that cannot be determined until after several years of range-wide monitoring and evaluation of effects of different land uses on populations.

The sale of Naval Petroleum Reserve #1 in Elk Hills to private interests (Henry 1995a, 1995b) could represent a threat to one of the three largest regional populations of giant kangaroo rats if rates of exploration and production are increased. The giant kangaroo rat population in western Kern County is isolated from all others, and though at times is fairly widspread, it seems especially sensitive to variable precipitation patterns, declining to only a few small areas during drought and after periods of heavy rainfall. Thus, its vulnerability to extinction by random catastrophic events (e.g., drought, flooding, fire) seems relatively high (B.L. Cypher pers. comm., T. Kato pers. comm., L. Spiegel pers. comm., Endangered Species Recovery Program unpubl. observ.). Any factor that would reduce substantially the amount of protected habitat in that region would pose a major threat to the population. The greatest value of the Naval Petroleum Reserves in California to giant kangaroo rats is the large extent of habitat of varying quality and its connectivity to adjacent habitat in the Lokern area. The publicly-owned portion of the Naval Petroleum Reserves in California ensures that giant kangaroo rat habitat will be protected during and after extraction of petroleum deposits.

Land in western Fresno County at the edge of irrigated ground provides an important area for recovery of the northern population of giant kangaroo rats (Williams et al. 1995) (Figure 39). The extant population on natural lands along the border of cultivated ground is split into two segments (Figure 41, see area A). One occupies only a narrow band about 6.44 kilometers (4 miles) long and from about 200 meters (660 feet) to 320 meters (1,050 feet) wide. The other, separated by only a few hundred meters, occupies about 250 hectares (617 acres) in an oval pattern about 2,400 by 1,200 meters (1.5 by 0.75 miles; Williams et al. 1995). Together, they support about 27 percent of the entire northern population in times of high population numbers, and probably more than 50 percent in times of lowest population numbers. This population represents the "up-slope" remnant of a formerly huge colony that stretched among the gentle slopes of the western edge of the Valley from around the alluvial fan of Laguna Seca Creek in Merced County, southward to Coalinga, a distance of about 97 kilometers (60 miles). During population irruptions it also is the "connector" population to small, scattered populations in the Ciervo and Tumey Hills, and along Panoche and Silver Creeks (Figure 41, see area C). The narrow band of habitat for this population is bisected lengthwise and degraded in quality by roads, power lines, and pipelines. Moderate levels of livestock grazing on this property probably have maintained nearly optimum conditions for giant kangaroo rats in what is only mediocre-quality habitat in comparison to historical habitat, but among the better-quality habitat remaining. Any additional loss or degradation of habitat from construction of permanent roads and energy conveyance facilities or cultivation could pose a substantial threat to the entire northern population.

Habitat for three of the six regional populations of giant kangaroo rats include no public or conservation lands (Figure 39). These are the populations in Cuyama Valley (about 194 hectares, 480 acres), Kettleman Hills (about 1 hectare, 2.47 acres), and San Juan Creek Valley (estimate unavailable because of lack of access to private land; Williams 1992). All are small and vulnerable to extinction from demographic and random catastrophic events (e.g., drought, flooding, fire), and inappropriate land uses that would degrade or destroy habitat.

5. Conservation Efforts

Designation as State (1980; Table 1) and federally (USFWS 1987) endangered has resulted in substantial habitat protection for giant kangaroo rats. Most significant has been protection on the U.S. Department of Energy Naval Petroleum Reserves in California in western Kern County (OFarrell and Kato 1987, OFarrell et al. 1987a, 1987b), and on USBLM-administered Federal properties (USBLM 1987, 1993). Acquisition of private property in the jointly managed Carrizo Plain Natural Area by the State of California, U.S. Government, and The Nature Conservancy (Table 2) has significantly reduced threats to the species from dryland cultivation and illegal use of rodenticides. It also has allowed for control of livestock grazing on this land by the change in onership from private to public. Other significant acquisitions that have benefited conservation of giant kangaroo rats have been the land exchanges and purchases within western Fresno and eastern San Benito Counties by the USBLM, and compensation, donation, and acquisition of parcels in the Lokern area of western Kern County by the California Energy Commission, CDFG, and The Nature Conservancy (Table 2).

Substantial progress in understanding the current distribution, habitat associations, demography, and population genetics of giant kangaroo rats has been achieved by a series of research projects, mainly supported by USFWS section-6 funds and money from the Endangered Species Tax Checkoff Program and Environmental License Plate Program administered by the CDFGs Bird and Mammal Conservation Program (R. Schlorff pers. comm.). Additional funding and logistic support for research on giant kangaroo rats has been provided by the Bureau of Reclamation, USBLM, USFWS, and The Nature Conservancy. This research has been summarized in a series of reports and publications (Williams 1980, Williams 1992, Williams et al. 1993b, 1995, Allred et al. in press, Mosquin et al. in press, Williams and Nelson in press, Williams and Tordoff 1988). Additionally, substantial information on distribution, habitat, and population fluctuation has been provided by the U.S. Department of Energy through EG&G Energy Measurements for research conducted at the Naval Petroleum Reserves in California in western Kern County (OFarrell and Kato 1987, OFarrell et al. 1987b, EG&G Energy Measurements 1995a,b), and for the southern San Joaquin Valley (Anderson et al. 1991) and the Carrizo Plain Natural Area (Kakiba-Russell et al. 1991) by the California Energy Commission.

U.S. Environmental Protection Agency County bulletins governing use of rodenticides have greatly reduced the risk of significant mortality to giant kangaroo rat populations by State and county rodent-control activities. The California Environmental Protection Agency, California Department of Food and Agriculture, county agricultural departments, CDFG, and U.S. Environmental Protection Agency collaborated with the Service in the development of County Bulletins that both are efficacious and acceptable to land owners (R.A. Marovich pers. comm.).

6. Recovery Strategy

Recovery of giant kangaroo rats can be achieved when the three largest populations (western Kern County, Carrizo Plain Natural Area, and the Panoche Region) and the populations in the Kettleman Hills, San Juan Creek Valley and Cuyama Valley are protected and managed appropriately. Because the giant kangaroo rat is a keystone species, protection of the above areas will benefit many other listed species that share the same habitat types.

Information on reproductive rates and survivorship still is insufficient to adequately model population viability, though measured population growth strongly suggests that reproductive capacity of giant kangaroo rats is ample to rapidly rebuild depleted population numbers and to expand into newly available habitat. The principal factor in recovery of giant kangaroo rats is protection of existing habitat and key local populations within the three regional populations.

Current understanding of demographics, distribution (Williams 1992, Williams et al. 1993b, 1995, Allred et al. in press, Williams and Nelson in press), and population genetics (Mosquin et al. in press) of giant kangaroo rats is sufficient to presume that the species is not threatened by inbreeding, low reproductive rates, etc., though some small, isolated populations are at risk from these factors. Population responses to environmental variation seen during the last 16 years (Williams 1980, 1992, Williams et al. 1993b, Williams and Nelson in press, D.F. Williams unpubl. data) suggest that random catastrophic events (e.g., drought, flooding, prolonged rainfall) poses the greatest risk to long-term survival of the species. Protection from random catastrophic events requires both relatively large habitat areas with varying topography and haitat conditions, and land uses that provide optimum habitat conditions.

Recovery Actions.-- Though substantial habitat for giant kangaroo rats is now in public ownership, recovering giant kangaroo rats requires additional habitat protection. Key to protection is an adequate understanding of compatible land uses and management prescriptions that provide optimum habitat conditions for giant kangaroo rats (Williams and Germano 1993). Several other listed species, including the California jewelflower, San Joaquin woolly-threads, blunt-nosed leopard lizard, San Joaquin antelope squirrel, and San Joaquin kit fox, seem to require the same or similar habitat conditions, so there is unlikely to be conflicts in habitat management prescriptions for most of the listed species where they coexist. Land acquisition, purchase of conservation easements, or other incentive mechanisms that will ensure that suitable habitat will be maintained in perpetuity also are needed to protect key local populations. Some existing public lands could be inhabited or support larger populations if suitably restored. Yet, available data are insufficient to know the types and amounts of compatible land uses or appropriate forms of habitat restoration and management. Recovery actions to protect habitat for giant kangaroo rats follow:

Of highest priority for habitat protection is proper land use and management on publicly-owned and conservation lands in the Carrizo Plain Natural Area, Naval Petroleum Reserves in California, Lokern Natural Area, and Ciervo-Panoche Natural Area. Where populations of giant kangaroo rats and associated, listed species appear to be robust, land use should not be changed when ownership or conservation status of parcels changes unless there are compelling reasons to do so. For land already in public and conservation ownership, historical uses that maintained habitat for giant kangaroo rats, such as livestock grazing, should be reestablished where appropriate.
Of equal priority is supporting research on habitat management and restoration, focusing on effects of livestock grazing on habitat quality, and habitat restoration on retired farmland, especially abandoned dryland farms.
Second in priority for habitat protection is the protection of additional land supporting key populations by acquisition of title, conservation easement, or other mechanisms. Areas to be protected are prioritized, as follows:
1. 1. Land in the Lokern Area of western Kern County. The goal is to protect 90 percent of the existing natural land bounded on the east by natural lands just east of the California Aqueduct, on the south by Occidental of Elk Hills, on the west by State Highway 33, and on the north by Lokern Road;
  2. Land in the Naval Petroleum Reserves in California of western Kern County. The goal is to maintain in a natural state (i.e., grassland and saltbush scrub communities) 90 percent of the existing natural land in Occidental of Elk Hills, and 80 percent of the natural land in Naval Petroleum Reserve in California No. 2, including all in the Buena Vista/McKittrick Valley between Elk Hills Road on the southeast and State Highway 33 on the northwest;
2. Existing natural land providing habitat for giant kangaroo rats in western Fresno and easternSan Benito Counties. The goal is to protect all existing natural land on the Silver Creek Ranch, and existing habitat for this species along the eastern bases of Monocline Ridge and the Tumey Hills, between Arroyo Ciervo on the south and Panoche Creek on the north;
3. Acquire and restore habitat on periodically farmed land with no or Class-3 irrigation water rights immediately east of occupied natural habitat along the strip described in 3,b, and west of Interstate Highway 5;
4. Other natural land occupied by giant kangaroo rats in western Kern County. The goal is to protect 80 percent of existing habitat for giant kangaroo rats;
5. Land occupied by giant kangaroo rats in the Cuyama Valley, Santa Barbara County;
6. Land occupied by giant kangaroo rats in the Kettleman Hills, Kings County;
7. Land occupied by giant kangaroo rats in the San Juan Creek Valley, San Luis Obispo County.

The above areas described in items e through g are important to the continued existence and recovery of other species, though it is not known if giant kangaroo rat populations have sufficient habitat in those areas to maintain viability indefinitely. Their keystone role in the ecosystem, however, makes it important to try to maintain these giant kangaroo rat populations.

A long-term program to periodically monitor populations range-wide is important to understanding population responses to random catastrophic events (e.g., drought, flooding, fire) and differing land uses, response to adaptive management, and to measure progress toward recovery. This program would measure responses of populations, key elements of their plant community, environmental variation, and soil erosion or formation to variation in climate and land uses (Williams and Kelly in litt. 1994a). Monitoring should be conducted annually for at least a 10-year period, and periodically thereafter at 5-year intervals.