Sucessão florestal em cronossequência na floresta atlântica : capacidade de resiliência e influência do meio

Abstract

Primary forests are becoming scarcer, while the amount of secondary forests increases. Especially in the tropics, regenerating areas become potential forests for biodiversity conservation and for ecosystem services maintenance. However, there is not a unified model of the natural recovery of these forests, if progressive, converging to mature forests, or retrogressive, becoming floristic and structurally different from mature forests, especially in urban areas with frequent disturbances. We aimed at evaluating if after abandonment, there is convergence from secondary to mature forest using a chronosequence in a highly disturbed urban landscape. The study was conducted with the canopy and sub-canopy woody plant assemblage (trees, palm trees and lianas), in the State Park of Dois Irmãos, Recife, Pernambuco, Brazil, in five locations: two in a mature area and three in a regenerating area. We determined land use histories using historical aerial photographs, which allowed to determining the chronosequence. The chronosequence consisted of two locations of mature forests (M1 and M2), one location in late regeneration stage (R1, aging between 34 and 46 years) and two locations in initial regeneration stage, one less impacted (R2) and the other more impacted (R3), both under 30 years of abandonment). We used circular sample plots of 254 m2 (canopy) and 78 m2 (sub-canopy), were all individuals with DBH ≥ 10 cm (canopy) and ≥ 5 cm (sub-canopy) were identified and measured. Luminosity and slope information of each sample plot was gathered. Data were analyzed for richness, through rarefaction, similarity (NMDS and ANOSIM), indicator species, physiognomic indicators (Kruskal-Wallis followed by Dunn's test), abiotic factors (Kruskal-Wallis followed by Dunn's test and ANOVA followed by Tukey's test) and the influence of abiotic factors (RDA followed by Monte Carlo) for both canopy and sub-canopy. Throughout the chronosequence, there was a significant increase in richness, floristic similarity (greater similarity between areas with closer ages), greater similarity of the sub-canopy with older areas canopy, and physiognomy (height, density and basal area). We observed that late successional indicator species only occurred in the mature forest (M1 and M2) and the late regeneration area (R1). Among the initial regenerating stages, the one that was more impacted diverged in richness and physiognomy from the less impacted having the same age. This pattern confirms the convergence of regenerating forest to mature forest, especially in sub-canopy that showed no differences in richness and physiognomy between M1, M2 and R1. The abiotic factors of luminosity and slope poorly explained succession, 9.5% for the canopy and 8.7% for the sub-canopy, although these percentages were significant (p <0.05). Therefore, the best predictor of species richness and diversity in this chronosequence was the age since abandonment. The impact intensity in the regenerationg area may lead to restoration necessity due to the regeneration being delayed in such conditions. Thus, we conclude that the sub-canocopy can converge in approximately 40 years, while the canopy will require more time to converge, even though it already presents this tendency.