The interrelations of black- tailed jackrabbits and the desertshrub vegetation on which they were feeding were studied in Curlew Valley, Northern Utah. The vegetation was described as a threecornered continuum, the corners being types dominated respectively by Artemisia tridentata, Atriplex ~ onfertifolia, and Sarcobatus vermiculatus. Jackrabbit diet was studied by microscopic analysis of plant fragments in stomachs from shot animals. The method was inaccurate, apparently because the ratio of identifiable tissues to all ingested tissues was very low, and varied between plant taxa, and seasonally. This problem seems intractable for desert shrub vegetation. The diet was similar to that reported by other workers on this species, with perennial grasses and forbs most important in sprlng and summer, shrubs in autumn and win ter. Features new to this vegetat ion were large percentages of Halogeton glomeratus, particularly in autumn and winter, and intense selection for Kochia americana. Attempts to explain the foods chosen ln terms of t heir nutrient contents were partically successful. Diet selection by large generalist herbivores was conceptualized as optimization of nutrient intake, mediated by long-delay learning, and constrained by food availability only at very low levels of availaoility. Spatial variation in jackrabbit diets confirmed this "cut-offll response to ava i 1 all i 1 i ty . Percentage utilization was estimated indirectly as jackrabbit density, times yearly food consumption per jackrabbit, times yearround percentage of each taxon in the diet, div i ded by available biomass of each taxon. Less abundant plants were more intensely used, which is expected if consumption does not vary continuously with availability. Perennial grasses, Kochia americana and possibly Grayia spinosa seemed to be under damaging pressure at high jackrabbit densities. Kochia had almost disappeared from outside a sheep- and jackrabbitproof exclosure since the 1950 1 s. In other exc1osures, the presence or absence of jackrabbits seemed to make no difference to the rate of vegetation recovery over 5-7 years after exclusion of sheep. Jackrabbit use of a crested wheatgrass seeding was concentrated ln a 300 m band around its edge.

Population biology of the black-tailed jackrabbit population on a 250-square-mile area in Curlew Valley, northern Utah, was studied from 1962-70. During this period the fall population density index increased from 40.0 in 1962, to 60.6 in 1963, decreased progressively to a low of 21.2 in 1967, then increased the following 3 years to a high of 185.0 in 1970. Breeding was synchronous with four conception periods each year; in some years a fifth conception period was evident. The first conception period occurred about the last half of January; other periods followed at 40-day intervals indicating a 40-day gestation period and postpartum estrus with subsequent conception. Over the 9 years of study, the mean percentages of females breeding during the five conception periods were 88, 99, 100, 70, and 11 percent, respectively. The mean number of ova ovulated per breeding female for the five periods was 1.9, 5.1, 6.4, 4.9, and 3.6, respectively. During the period of decreasing density, 1963-67, the yearly mean number of ova ovulated per female surviving the breeding season ranged from 13.2-19.3, but varied independently of density. During the 3 consecutive years of density increase, 1968-70, however, the number of ova ovulated per female decreased progressively from 19.8 in 1968 to 14.2 in 1970. Mortality rates of the total population from October-March remained relatively constant (mean: 63 percent) during the years of population decline, but dropped to 33 percent during the first year of population increase (1968). March-October mortality of adults decreased to 9 percent during the first year of population increase from a previous mean of 73 percent, and juvenile mortality from parturition to fall census, decreased from a mean of 68 percent to 38 percent. The effects of variations in mortality rates on population density have overshadowed the effects of the relatively less extreme variations in natality rates. As a result the pattern of density change was almost entirely a result of changes in mortality rates. Changes in mortality rates of adults and juveniles were well correlated with the coyote/rabbit ratio on the study area. Exceptions occurred with juvenile mortality rates at the relatively high rabbit densities observed in 1969-70. During these two years, juvenile mortality rates from parturition to fall census (61 and 68 percent, respectively) were greater than could be accounted for by coyote predation. The factor or factors responsible for the increased juvenile mortality are not known. Observed annual density changes were described with the mathematical model: Nt+4 = Nt (1 - 37.8 - 988x1) (1 + 11.2 - 1130x2 - 0.0581x3 + 42000x22 + 0.00115x32) where Nt is the number of animals at the end of October, Nt+1 is the number of animals at the end of the following October, x1 is the coyote/rabbit ratio from October-March, x2 is the coyote/rabbit ratio from March-October, and x3 is the mean number of rabbits per square mile from March-October. The model accounts for 99 percent of the observed change in rabbit density from 1968-70.