An exploration of direct and indirect drivers of herbivore reproductive performance in arid and semi arid rangelands by means of structural equation models

Abstract

Climate variability affects both animal and plant populations. Understanding how this variability is modulated by topography, vegetation and population densities and how this impact on ungulate demography is an important step in order to forecast consequences of alternative management scenarios or the effects of climate change. Structural equation modeling (SEMs) is a tool particularly suited to this kind of situations, and allowed us to study the direct and indirect effects of climate, topography, structural and functional aspects of vegetation and population density upon reproductive performance of Patagonian sheep flocks, measured at paddock scale. The application of SEMs in conjunction with information criteria and related techniques for model selection, model averaging and multi-model inference revealed that despite considerable model uncertainty, those paddocks towards the East, with a greater spatial variability in July temperatures and greater primary production during fall-winter (estimated by remote sensing) showed greater lambing rates. Paddocks with higher proportion of meadows and with more intense forage consumption, were also associated with a better reproductive performance. Our results not only provide quantitative hypothesis about the controls at the landscape level of herbivore performance but also provide the basis to devise better management alternatives.