Early Regeneration of Tropical Dry Forest from Abandoned Pastures: Contrasting Chronosequence and Dynamic Approaches

ABSTRACT Old field secondary succession of tropical dry forests (TDFs) is poorly understood, particularly regarding the dynamics of seedlings, saplings, and sprouts (regenerative communities). We used chronosequence and dynamic approaches to: (1) document successional trajectories of regenerative communities during the first dozen years of regeneration in abandoned pastures at Chamela, Mexico; (2) test the usefulness of chronosequences to predict the dynamics of regenerative communities along time; and (3) assess the influence of surrounding forest matrix, stand density, and understory light availability (in the rainy season) as driving factors of such dynamics. More than 1000 plants and 95 species of shrubs and trees 10–100 cm tall were monitored between 2004 and 2007 in nine abandoned pastures (0–12 yr since abandonment) and two old‐growth forest (OGF) sites; gain and loss rates of plants, species, and plant cover were obtained. Chronosequence predicted a rapid and asymptotic increase of plant density, species density, and plant cover toward the OGF values. Such prediction did not match with dynamic data that showed negative or neutral net community rates of change, independently of fallow age. Recruitment and species gain rates increased with the amount surrounding forest matrix. No other effect of the explored factors was detected. Strong rainfall shortenings could be responsible for the high loss and low gain rates of plants and species recorded in most sites. We highlight the critical role of supra‐annual rainfall variability on the dynamics of TDF regenerative communities and the poor predictive value of chronosequences in forest systems subjected to strong environmental temporal variation. Abstract in Spanish is available at http://www.blackwell‐synergy.com/loi/btp .