Diagnosis and conditioning factors of the Cerrado vegetation resilience in abandoned pastures

Abstract

Restoration of degraded ecosystems, over the world, has been mostly based on the reintroduction of plant species, and less frequently in techniques to accelerate the natural regeneration. However, once the ecosystem is resilient, recovery may occur naturally, without the need for intervention and, therefore, without costs. Quantify resilience - which includes estimating the time needed for the ecosystem to recover its attributes previous to disturbance and quantify the proportion of components that may be recovered - are challenges for both science and practice of restoration. The objective of this research is to develop tools to quantify the resilience of the cerrado vegetation in abandoned pastures and to verify if Cerrado resilience is predictable on the basis of biotic and abiotic factors that influence ecosystem recovery. The research will be based on sampling the vegetation in natural regeneration in a number of sites of abandoned pastures (30), in two occasions separated by a two-years interval, to quantify the rate of recovery, using native ecosystems (six replicates) as reference (the target). The rate of recovery will be obtained for variables related to community structure (density, cover, basal area) and plant species richness. These rates will be used as response variables in Generalized Linear Models (GLMs) using as predictor variables (factors influencing resilience) the land use history, environmental factors (soil and climate) and landscape (cover of native vegetation, distance from seed sources) in each study site. In addition to elucidating pending questions in the ecology of ecosystems, this research has the potential to provide practical application tools to assist the decision-making, particularly on the need or not of restoration interventions. (AU)