A challenge to restoration practice is that outcomes may hinge upon legacies of activities that happened decades or centuries in the past. For example, agricultural abandonment has left a large portion of the terrestrial biosphere available for restoration, but restoration in these systems may be affected by both the long-lived legacy of agricultural use and the nature of present-day management imposed (e.g., seed sowing, timber thinning). Despite the potential for past agricultural land use to interact with contemporary management to alter consumer and plant communities, there are no studies that couple consumer manipulations, seed addition, and large-scale variation in past land use. Working in the longleaf pine ecosystem, we used 27 experimental landscapes that spanned post-agricultural and non-agricultural land use, seed additions of 12 plant species, and exclosures that manipulated consumer access to evaluate the hypothesis that rodent consumers limit plant restoration and that the weedy habitats created by the restoration of post-agricultural communities would provide high-quality habitat for rodents, leading to greater numbers of rodents in post-agricultural communities and greater consumer pressure on plants of restoration concern.
At our study sites, cotton mice (Peromyscus gossypinus) and oldfield mice (P. polionotus) were the dominant small-mammal consumers across three years of sampling (96% of the individuals ). Contrary to our hypothesis, we found that P. gossypinus was significantly more abundant in non-agricultural habitats, not post-agricultural habitats, following canopy thinning. The abundance of P. polionotus was significantly greater in habitats that had been thinned, but there was no difference due to past land use. In contrast to abundance of P. gossypinus, the data from experimental seed additions and exclosures support the prediction that greater rodent abundance leads to greater consumer pressure on plant establishment: seed additions and excluding rodents both significantly increased the richness of sown plant species, but this increase was greatest in post-agricultural habitats. These results suggest that consumer communities can play an important role in shaping the resulting plant community and that seed limitation is a primary determinant of plant establishment in these systems. However, our results also caution that past land use may only affect some members of the consumer community, and linkages between consumer abundance and plant establishment may be difficult to predict.