Extrinsic factors rather than seed traits mediate strong spatial variation in seed predation

Abstract

Seed predation is a key demographic process for plant population dynamics and community structure. The strength of seed predation for a given plant species often varies among populations depending on both plant traits (e.g. seed size) and extrinsic factors (e.g. seed predator abundance). Because both types of factors tend to covary spatially, identifying their relative contribution to changes in seed predation through observational approaches is challenging. During two consecutive years, we reciprocally translocated seeds of Iberian pear (Pyrus bourgaeana) from five localities in SW Spain to estimate changes in interaction strength between the tree and its seed-eaters (mostly rodents). We also measured relevant seed traits, the abundance of granivore rodents, and the density of fruiting-shrubs in the same five populations. Both seed traits and extrinsic factors varied among the five studied populations. As predicted, we found significant spatial covariance between some extrinsic factors (i.e. rodent abundance and fruiting-shrub density) as well as between extrinsic factors and plant traits (seed size). Seed removal likelihood varied highly among offering localities, with differences up to 11.6-fold. However, and despite differences in seed traits among populations, no evidence of seed predator preference for seeds from any locality was found in any of the five local experiments. Our study indicated that extrinsic factors (e.g. seed predator abundance) rather than measured seed traits mediated among-population variation in seed predation. Selection pressures exerted by seed predators on seed traits could thus be less pervasive than previously thought. Our experimental approach can help to better understand the potential of different functional animal groups (pollinators, herbivores, granivores, frugivores) to exert selection pressures on plant traits.