PubMedDate not supplied
The relationship between fecal corticosterone concentrations and characteristics of the environment and population demography were studied in adult male gerbils (Rhombomys opimus Licht.) at the southern border of Kyzylkum desert (Reserve "Ecocentre Dzeiran", Bukhara region, Republic Uzbekistan) in spring and fall seasons from 1999 to 2004. We extracted hormones from air-dried fecal samples and analyzed their concentrations by radioimmunoassay (Gerlinskaya et al., 1993). An analysis for year-specific relationships between hormone concentrations and environmental variables of temperature and precipitation using Pearson's r statistic revealed that corticosterone concentrations correlated positively with total precipitation in January and February and negatively with precipitation during March and April. There was also a significant negative relationship between fecal corticosterone and the number of hot days in March (>20 degrees C). Demographic variables that characterized population densities (percent of burrow systems occupied, mean and maximum number of burrow systems/1 ha, number of females in the burrow system) correlated positively with corticosterone concentrations in feces in the beginning of spring, but these relationships were small compared with mean concentrations of corticosterone for the entire spring season that were strongly and positively correlated with number of gerbils, including all pups emerged, in burrow systems owned by one male (within its home range). In contrast, correlation coefficients of corticosterone concentrations with characteristics of feeding resources in the spring were low and negative. In the long-term perspective (interannual comparison), mortality among adult males was highly negatively correlated with mean corticosterone concentrations in the beginning of spring, which is within the period of maximum reproductive effort and potential stress. Body mass was independent of corticosterone concentrations in males in either the beginning of spring, or during the whole spring. In the fall, mean concentrations of fecal corticosterone in males was positively correlated with the number of days from June to October with mean daily temperatures exceeding 30 degrees C, and with percent of burrow systems where at least one adult, > or = 1 year old gerbil had survived. Mortality from fall to spring of the next year and the fall body mass did not correlate with concentrations of corticosterone in feces collected in the fall. When we analyzed corticosterone concentrations in spring seasons of all years combined using a stepwise regression analysis of a sampling of individual males (we analyzed residuals after withdrawal of year effect) on a set of variables representing habitat resources, distances between nearest neighbor males, and variables representing group demography we found low R2 values not exceeding 0.17. Within the six-year period, concentrations of corticosterone in the spring related negatively with abundance of annual herbs and positively with number of females in a male's social group. When only years of high density were analyzed, fecal corticosterone concentrations in males in the spring were again negatively determined by abundance of herbs, as well as by the nearest neighbor distance, and positively determined by the number of females within a male's home range. At the beginning of spring the only determinants were distance to the nearest neighbor male and number of females. In years of low density none of the variables were found to affect corticosterone levels during the whole spring, while in the beginning of spring only partial regression coefficients of abundance of herbs were negative and significant. Stepwise logistic regression analysis revealed positive dependence (P = 0.05) of disappearance of adult males during summer drought on concentrations of fecal corticosterone in the spring, but only when burrow systems with at least one adult (male or female) surviving after the summer were considered. Our results provide evidence ground for the assumption that in a desert rodent with non-regular population fluctuations such as the great gerbil, density may be more suppressed by external factors and not by density dependent mortality mediated by stress. Density dependent increases of stress caused by intense reproductive effort occurred when feeding and climatic conditions were favorable to compensate for negative effects on survival. However, in individual gerbils mortality mediated by stress can take place because we found higher stress in the beginning of spring in males, which did not occur in the population after the summer drought.