Disentangling the effect of multiple ecological processes on recruitment in an invaded annual grassland using a trait based approach

Loralee Larios, Botany and Plant Sciences, University of California-Riverside, Riverside, CA, Lachlan Charles, University of California, Riverside and John L. Maron, Division of Biological Sciences, The University of Montana, Missoula, MT

Background/Question/Methods
In temperate systems small mammals play an important role in mediating plant community composition directly via their foraging, or by mediating other assembly filters. However, their dynamics have yet to be fully integrated in trait-based approaches evaluating community assembly. Disentangling these dynamics is of particular importance in invaded communities as local filters may disproportionately affect native and non-native species. Here we aimed to disentangle the effects of local environment and small mammal herbivory on recruitment of native and non-native species in an annual California grassland (Browns Valley, CA). At eight sites, we set up a split-plot experiment where we manipulated the presence of small mammals and within these larger plots, we established subplots containing a factorial experiment of three fixed factors: Seed addition (seed added or control), origin of seed (native, nonnative) and background environment (litter removed or intact). We measured recruitment and reproduction for the seeded species and assessed whether plant traits predicted these demographic stages.
Results/Conclusions
We observed that non-native species recruited and reproduced at greater densities than native species. The recruitment of native species was sensitive to local filters such that their probability of recruitment was the lowest in the absence of small mammals and when litter was intact; conversely non-native recruitment was not impacted by these conditions. Seed size best explained the probability of recruitment and recruitment density with recruitment increasing with increasing seed size. The number of reproductive stems and seed output was best explained by leaf nitrogen content. As with recruitment, native reproduction was sensitive to the presence of litter, such that the number of reproductive stems and total seed output was less than when litter was removed. These results highlight the context dependency of assembly filters, as we observed that local assembly filters more strongly impacted native than non-native species and that the pertinent traits and filters differed with the demographic stage with small mammals being important at the recruitment stage.