The Rocky Mountain elk (Cervus elaphus) of the Greater Yellowstone Ecosystem (GYE) support biodiversity and ecosystem functioning and provide important cultural and economic benefits. Calf recruitment has declined in some GYE elk populations and reduced pregnancy rates are thought to be a contributing factor. Previous studies suggest that elk pregnancy is limited by reduction in summer forage quality as a result of drought and warmer temperatures, or by increasing predation risk due to gray wolf recovery. The reproductive disease brucellosis and changes in population age structure may also play a role. Identifying the relative contributions of these factors has been challenging because most studies are limited to one or a few populations.
We use archived blood sera, collected during routine monitoring over a 20 year period, to broadly assess environmental influences on pregnancy of GYE elk. We identified seasonal ranges using GPS telemetry obtained from a collared subsample of individuals. We then used these to sample climate variables, vegetation productivity as estimated by integrated NDVI, spring green-up and fall senescence indices, and wolf density for over 1400 elk at 25 sites throughout the GYE. By modelling elk pregnancy probability as a multivariate logistic regression in a Bayesian framework, we elucidate which candidate factors best explain elk pregnancy region-wide.
Results/Conclusions
Climate, Spring-green up and absolute productivity metrics were not consistent predictors of elk pregnancy status. Inclusion of forage senescence date improved on the base model that included age and brucella status alone, indicating a sensitivity to changes in phenology and highlighting the importance of late summer/early fall forage availability as elk enter the rut. Individual level factors such as age and brucellosis seroprevalence are strong predictors of elk pregnancy. Probability of pregnancy is a quadratic function of age due to elk senescence, with young and very old elk having the least probability of pregnancy. Given this, there is potential for further reduction in pregnancy rates, if sustained low recruitment results in an elk population that consists of a greater proportion of older individuals. We did not find significant evidence for a wolf effect on pregnancy, indicating either that wolf:elk ratios are a poor proxy for wolf predation risk or that these effects are not operating strongly in the study area. In light of these results, we discuss implications for ecology and management in the greater Yellowstone ecosystem.