The impact of climate change on the dynamics of populations has been well documented and is widespread. However, weather variability influences populations both directly
and indirectly, and is mediated by species interactions. This complexity may impede proper climate impact assessments. Hence, predicting the consequences of climate change may require
including processes that occur both with time lags and across trophic levels. Based on our current
understanding of the mechanisms linking local climate and trophic interactions in tundra ecosystems, we used a state-space formulation of a mediation model that allowed for assessing the relative contribution of direct and indirect environmental (weather and trophic) effects on reindeer
Rangi fer tarandus reproductive success. Our study showed that the mediator effect of body condition caused delayed but predictable effects of weather, plant productivity, and reindeer densities on reproductive success. Furthermore, these predictors also affected reproductive success
directly and with the same sign, suggesting that direct and indirect effects pulled in the same
direction with respect to their combined total effect on reproductive success. Hence, poor weather
conditions not only affect calf production negatively the same year, but also increase the likelihood of poor reproductive success the subsequent year. The results support the expectation that
calf slaughter mass (as a proxy for herd body condition) is an important indicator of the state of
reindeer herds with respect to their production potential and resilience to weather events and climate change. Finally, the model framework employed in the present study can be further
developed as a potential vehicle for near-term forecasting, and thereby constitutes a useful
tool for adaptive management.
Mediation · Arctic · Heterogeneous effects · Ungulates · Body condition · Reproductive
success · SEM · Structural Equation Model
