Plant phenology, the timing of important life events such as growth or reproduction, dictates when critical resources may be available for higher trophic levels within a food web. In aridlands, plant growth is controlled by stochastic rain events, leading many individuals to fail to grow or reproduce in some seasons or years. This extreme intra- and interannual variation in resource availability cascades throughout arid food webs, however most studies employ qualitative descriptions of “good” or “bad” resource years for consumer populations. Here, we leverage two decades of monthly plant phenology and seasonal biomass datasets to estimate monthly species-specific biomass, or Phenomass, across three semi-arid biomes. We then create monthly-resolution Forage, Flower, and Fruit Availability Indices (AI). These AI are related to the abundance of three consumer taxa at the same sites. The Forage AI is related to herbivorous grasshoppers; the Flower AI to pollen-dependent bees; and the Fruit AI to granivorous rodents. We identify a core community of plants species and functional groups that consistently contribute to each AI, as well as those species which create periodic pulses in resource availability. We also employ lag analyses to determine the legacy effects of previous-season or previous-year resource availability on consumer populations.
Results/Conclusions
Data used were part of the Sevilleta Long-Term Ecological Research program. Data from three sites, a Chihuahuan Desert grassland, a Chihuahuan Desert shrubland, and a mixed Shortgrass Plains grassland, were used. Our initial results show high seasonal correlation (85-95%) between estimated plant Phenomass and measured biomass. We also find high monthly correlation (70+%) between monthly estimated Phenomass and satellite greenness (MODIS NDVI), a proxy for plant production. Granivorous species within the small mammal community experienced increases in abundance in the years following pulses in monsoon-season Fruit AI at each site. Grass-specialist grasshopper species abundance was affected by current and previous season grass Forage AI. Bee populations were the most dynamic of the animal taxa considered. Weak correlations were found between generalist-bee abundance and previous year Flower AI. Overall, we found that a relatively simplistic phenology dataset could be used to accurately estimate monthly species-specific biomass and was useful in quantifying resource availability for three consumer taxa, giving us a better understanding of food web dynamics in aridlands.