Litterfall, vegetation structure and tree composition as indicators of functional recovery in passive and active tropical cloud forest restoration

Abstract

Litterfall is an indicator of ecosystem function and its temporal dynamics can be used to evaluate self-organizing ecosystems on a recovery trajectory following restoration. Few studies have evaluated the recovery trajectories of forest litterfall by simultaneously monitoring forest restoration strategies and reference ecosystems. The general objective of our study was to determine the functional recovery of an abandoned pasture under passive and active restoration, and in secondary (40-year-old) and mature (120-year-old) forest, by analyzing litterfall and its components (leaves, flowers, fruits, woody parts) over a period of five consecutive years. We determined the vegetation structure and tree species composition of these four conditions and compared 1) production of litterfall, leaves, flowers and fruits, 2) leaf litter nutrient inputs (C and N) and, 3) recovery by tree species in leaf litterfall. In five years, litterfall increased from 2.6 to 7.8 Mg ha−1 and from 3.5 to 9.1 Mg ha−1 in the passive and active restoration treatments, respectively, while it increased from 6.0 to 8.6 Mg ha−1 in the secondary forest. In the mature forest, litterfall varied around 10.0 Mg ha−1. The reproductive component increased significantly in restoration (0.3 to 1.5 Mg ha−1) but remained around 0.6 Mg ha−1 in secondary and 1.3 Mg ha−1 in mature forest. Secondary and mature forests both presented correlations to monthly precipitation and maximum and minimum temperatures, indicating strong seasonality. However, litterfall production in the passive and active treatments was continuous throughout the year. Basal area and tree density were higher under active compared to passive restoration. Although the dominant tree species were similar under passive and active restoration, active restoration presented a higher forest species recovery, while the dominance of exotic grass patches persisted under passive restoration. The results suggest that litterfall production can be a useful and accurate indicator with which to evaluate the recovery of ecosystem function, while flower and fruit component can indicate reproductive recovery. Although litterfall production increased rapidly after five years, it did not present a seasonal dynamic. This is probably due to the fact that the species composition still differs from that of the reference systems. Evaluations of cloud forest restoration success should therefore include temporal assessments of vegetation structure and biodiversity recovery in relation to the reference forests in order to establish additional restoration techniques, particularly in the case of passive restoration strategies.