Desert tortoise (Gopherus agassizii) populations have experienced precipitous declines resulting from the cumulative impact of habitat loss, and human and disease-related mortality. Evaluation of hematologic and biochemical responses of desert tortoises to physiologic and environmental factors can facilitate the assessment of stress and disease in tortoises and contribute to management decisions and population recovery. The goal of this study was to obtain and analyze clinical laboratory data from free-ranging desert tortoises at three sites in the Mojave Desert (California, USA) between October 1990 and October 1995, to establish reference intervals, and to develop guidelines for the interpretation of laboratory data under a variety of environmental and physiologic conditions. Body weight, carapace length, and venous blood samples for a complete blood count and clinical chemistry profile were obtained from 98 clinically healthy adult desert tortoises of both sexes at the Desert Tortoise Research Natural area (western Mojave), Goffs (eastern Mojave) and Ivanpah Valley (northeastern Mojave). Samples were obtained four times per year, in winter (February/March), spring (May/June), summer (July/August), and fall (October). Years of near-, above-and below-average rainfall were represented in the 5 yr period. Minimum, maximum and median values, and central 95 percentiles were used as reference intervals and measures of central tendency for tortoises at each site and/or season. Data were analyzed using repeated measures analysis of variance for significant (P Ͻ 0.01) variation on the basis of sex, site, season, and interactions between these variables. Significant sex differences were observed for packed cell volume, hemoglobin concentration, aspartate transaminase activity, and cholesterol, triglyceride, calcium, and phosphorus concentrations. Marked seasonal variation was observed in most parameters in conjunction with reproductive cycle, hibernation, or seasonal rainfall. Year-to-year differences and long-term alterations primarily reflected winter rainfall amounts. Site differences were minimal, and largely reflected geographic differences in precipitation patterns, such that results from these studies can be applied to other tortoise populations in environments with known rainfall and forage availability patterns.

How do female desert tortoises (Gopherus agassizii) reproduce every year despite variability in winter rainfall and food availability? To answer this question, I measured energy budgets of individual female desert tortoises from July 1987 to July 1989. Females produced eggs in years with low levels of winter annual plants by relaxing their control of energy and water homeostasis. They tolerated large deficits and surpluses in their body dry‐matter composition (both nonlipid and lipid) on a seasonal, annual, and longer time scale. They could increase body energy content (lipid and nonlipid energy) before winter and use this reserve (especially nonlipid energy) the following spring to produce eggs. Females used high‐protein foods and rainwater, when available, to achieve energy surpluses that helped them survive periods of low resource availability (e.g., during hibernation and droughts). Adjusting seasonal and annual field metabolic rates (FMR) and food requirements to levels of food availability, they still managed to produce eggs, even in a drought year. Egg production in 1988 (mean ± 1 sd: 3.56 ± 1.94 eggs, N = 9) did not differ from that in 1989 (3.00 ± 2.69 eggs, N = 9); both were lower than during 1983–1987 (6.75 ± 3.05 eggs, N = 100). Energy per se did not limit egg production in 1988 and 1989, but the availability of nonlipid energy (probably protein) limited egg production in 1988 and was limiting in spring 1989. Yet, water was the primary resource limiting egg production in 1989. Females forgoing egg production in 1989 accumulated body nonlipid energy and lost less total body water than did females producing eggs. Females that produced eggs in 1989 forfeited body nonlipid energy. In 1988 and 1989, the paucity of new annual plants in spring contributed to a lower egg production. The amount of annuals that germinate in summer can affect egg production, because females stored nonlipid energy during summer 1988 when eating these annuals, and allocated this energy to eggs in the following spring (1989). Tortoises stored lipids during summer when consuming dry annuals. These lipids are critical for surviving the winter, but females forfeited body water and nonlipid dry matter to digest the dry annuals. Reproductive effort (RE) was higher during the drought year (July 1988–July 1989: RE = 26.1%) than during the wetter year (July 1987–July 1988: RE = 13.0%) because tortoises reduced FMR by 70–90% in 1989. Compared to two other chelonians, RE of desert tortoises was consistent with the K‐selected trend of lower RE for larger, long‐lived, and late‐maturing species. Forfeiting body condition to produce only a few eggs, even under a great environmental stress, was consistent with a life history strategy called bet hedging.

We studied egg production in two Californian populations of desert tortoises, (Gopherus agassizii) in 1992 and 1993. One population inhabited the Desert Tortoise Research Natural Area (DTNA) in the western Mojave Desert, where most of the rain falls in the winter. The second population lived near Goffs, in the eastern Mojave, where annual precipitation is divided more evenly between winter and summer. Due to El Niio conditions, heavy winter rains fell at both sites in both years (1991-1992 and 1992-1993). Consequently, the biomass of spring annuals and annual egg production by tortoises were high in both years at both sites. There were no differences in reproductive output between years so we pooled data for both years to examine the relationship between egg-laying parameters (clutch size and frequency, annual egg production, egg size, etc.) and female size. Variation in annual egg production was due mainly to variation in clutch size, not clutch frequency. Annual egg production per female was lower at DTNA than at Goffs, because some adult females at DTNA did not produce eggs in some years. Females that did lay eggs produced the same number of eggs per year at both sites, even though females at Goffs were smaller (midline carapace length = 214 mm) than females at DTNA (MCL = 234 mm). Despite correction for these body size differences, the eggs produced at Goffs were smaller in all dimensions than eggs produced at DTNA. Smaller eggs and presumably smaller neonates may be related to the greater predictability of summer rain and consequent greater food supply for emergent hatchlings at Goffs. For adult females, food supply probably limits reproduction only during drought years. How can individual females vary their annual reproductive output? Our more extensive data for DTNA tortoises showed that larger females produced larger clutch sizes. In addition, by statistically removing the effects of body size we showed that larger clutches contained smaller eggs. Moreover, larger females produced eggs earlier in the year giving them a better opportunity to produce a second clutch that year. Thus, timing of first clutch was important Still, much of the variation in reproductive output was not explained. Other characteristics of individuals (e.g, age, genetics, physiological maturity, home range quality, or forage selection) may explain some of the variation in reproductive output Rainfall in the Mojave Desert is unpredictable and varies greatly with time and location. While the long-term average annual precipitation is about 150 mm, annual precipitation ranges from as little as three millimeters to as much as 400 mm. Rain often falls primarily in the winter, promoting a flush of annual plants in the spring (Beatley, 1974), but seasonal rainfall patterns usually vary considerably from year to year. Rainfall also differs between the eastern and western regions of the Mojave Desert (Nagy and Medica, 1986; Peterson, 1996a). Rainfall from July through September, when many desert tor