INTERACTIVE EFFECTS OF SEED SIZE AND ENVIRONMENTAL CONDITIONS ON THE OUTCOME OF THE INTERACTION BETWEEN RODENTS AND A THREATENED ATACAMA DESSERT SHRUB
Loayza, A.P.1, Luna, C.A.1 and Calviño-Cancela, M.2
1Departamento de Biología, Universidad de La Serena, La Serena, Chile
2Departamento de Biología y Ecología Animal, Universidad de Vigo, Vigo, España
Contact: Andrea Loayza, firstname.lastname@example.org
Spatial and temporal variations in the biotic and abiotic conditions of an environment can lead to context dependence in the strength and outcome of species interactions. Myrcianthes coquimbensis is a narrow endemic shrub of the Atacama Desert whose seeds are solely consumed by three species of scatter-hoarding rodents. These rodents act both as seed predators and dispersers, depending partly on seed size. Along its distribution range, strong interannual variability in rainfall leads to changes in M. coquimbensis crop sizes; wet years are characterized by fruit abundance, whereas in dry years fruits are scarce. In this study we parameterized an individual-based model -using three years of field data, as well as cafeteria experiments- to determine how context-dependence in terms of fruit availability and seed size interact to alter the sign and strength of the interaction between rodents and M. coquimbensis. We found that the quality of seed dispersal services provided by the three rodent species were strongly context-dependent, but were also determined by seed size. During both wet and dry years, large seeds manipulated by rodents had lower recruitment probabilities than those that were ignored and left under the parent plant. In contrast, during wet and dry years, small seeds manipulated by all rodent species had higher recruitment probabilities than small seeds that were ignored under the parent plant. Differences in recruitment probabilities were partly explained by the habitats where seeds arrived. In wet years, seeds had a higher recruitment probability under conspecific plants than in rock cavities, whereas the opposite was true in dry years. Finally, we also found an interaction between habitat and seed size on the probability of recruitment; in wet years the probability of recruitment in rock cavities decreased with smaller seeds, in contrast, in dry years this probability increased with larger seeds. In summary, our results provide new evidence that the strength and sign of the interaction between a plant and its frugivores can vary with time, with the identities of the interacting species, and with the interaction of time and seed size.