How livestock and flooding mediate the ecological integrity of working forests in Amazon River floodplains

The contribution of working forests to tropical conservation and development depends upon the maintenance of ecological integrity under ongoing land use. Assessment of ecological integrity requires an understanding of the structure, composition, and function and major drivers that govern their variability. Working forests in tropical river floodplains provide many goods and services, yet the data on the ecological processes that sustain these services is scant. In flooded forests of riverside Amazonian communities, we established 46 0.1-ha plots varying in flood duration, use by cattle and water buffalo, and time since agricultural abandonment (30-90 yr). We monitored three aspects of ecological integrity (stand structure, species composition, and dynamics of trees and seedlings) to evaluate the impacts of different trajectories of livestock activity (alleviation, stasis, and intensification) over nine years. Negative effects of livestock intensification were solely evident in the forest understory, and plots alleviated from past heavy disturbance increased in seedling density but had higher abundance of thorny species than plots maintaining low activity. Stand structure, dynamics, and tree species composition were strongly influenced by the natural pulse of seasonal floods, such that the defining characteristics of integrity were dependent upon flood duration (3-200 d). Forests with prolonged floods >= 140 d had not only lower species richness but also lower rates of recruitment and species turnover relative to forests with short floods < 70 d. Overall, the combined effects of livestock intensification and prolonged flooding hindered forest regeneration, but overall forest integrity was largely related to the hydrological regime and age. Given this disjunction between factors mediating canopy and understory integrity, we present a subset of metrics for regeneration and recruitment to distinguish forest condition by livestock trajectory. Although our study design includes confounded factors that preclude a definitive assessment of the major drivers of ecological change, we provide much-needed data on the regrowth of a critical but poorly studied ecosystem. In addition to its emphasis on the dynamics of tropical wetland forests undergoing anthropogenic and environmental change, our case study is an important example for how to assess of ecological integrity in working forests of tropical ecosystems.