Functional analysis of Normalized Difference Vegetation Index curves reveals overwinter mule deer survival is driven by both spring and autumn phenology

Large herbivore populations respond strongly to remotely sensed measures of primary productivity. Whereas most studies in seasonal environments have focused on the effects of spring plant phenology on juvenile survival, recent studies demonstrated that autumn nutrition also plays a crucial role. We tested for both direct and indirect (through body mass) effects of spring and autumn phenology on winter survival of 2315 mule deer fawns across a wide range of environmental conditions in Idaho, USA. We first performed a functional analysis that identified spring and autumn as the key periods for structuring the among-population and among-year variation of primary production (approximated from 1 km Advanced Very High Resolution Radiometer Normalized Difference Vegetation Index (NDVI)) along the growing season. A path analysis showed that early winter precipitation and direct and indirect effects of spring and autumn NDVI functional components accounted for 45% of observed variation in overwinter survival. The effect size of autumn phenology on body mass was about twice that of spring phenology, while direct effects of phenology on survival were similar between spring and autumn. We demonstrate that the effects of plant phenology vary across ecosystems, and that in semi-arid systems, autumn may be more important than spring for overwinter survival.