Fragmentation, Edge Effects and Regeneration of Tropical Dry Dipterocarp Forest in Thailand

Abstract

The fragmentation of contiguous old-growth forests is among the most severe consequences of deforestation. Southeast Asia has suffered disproportionately and in particular its tropical dry forests (TDFs). The aim of this study was to investigate the impacts of land use changes on a dry dipterocarp forest (DDF) fragment at the Sakaerat Environmental Research Station (SERS) in northeastern Thailand.

The research focused on the natural regeneration potential of DDF following different disturbance histories, the independent contribution of disturbance history on contemporary tree communities, the presence of edge effects in DDFs, and the importance of the width of secondary forests for providing old-growth forest habitat conditions and for buffering the old-growth forest from adverse edge effects.

According to the literature on global TDF biomes, I hypothesized that: (i) DDF has the potential to naturally recover within 20-30 years, (ii) disturbance history is a dominant driver of modern tree communities, (iii) edge effects in TDFs differ from tropical wet forest types, and (iv) the potential of secondary forest as complementary old-growth forest like habitat increases with increasing buffer width.

To assess the direction of regeneration, a tree community inventory was carried out within 59, 20 x 20 m plots, across three secondary forest study sites based on varying disturbance histories and three reference old-growth study sites with a similar range of soil types. Floristic, structural, and diversity parameters were determined by means of Analysis of Variance (ANOVA) and Importance Values Indices (IVIs). The independent influence of disturbance history on the variation in tree community composition was explored using (partial) Redundancy Analysis (RDA) based on Hellinger transformed abundance data by means of Variation Partitioning.

Then in order to investigate the presence of edge effects and the importance of secondary forest buffer width, nine transects, three within each of three types of buffer categories, were extended from old-growth forest into the surrounding matrix: (i) old-growth forest adjacent to agriculture (no buffer), (ii) old-growth forest adjacent to 30-year old secondary forest (< 320 m width) and then surrounded by agriculture (narrow buffer), and (iii) old-growth forest adjacent to extensive 30-year old secondary forest (> 320 m width; wide buffer). Floristic, structural, and diversity features were sampled in a total of 84, 20 x 20 m plots and microclimate recorded at sampling stations to a maximum distance of 512 m in each direction from the edge. Edge effects with respect to biotic and abiotic parameters were quantified using different statistical approaches and the main drivers of variation were identified applying partial RDAs on Hellinger transformed basal area data.

Both the disturbance history and the soil type accounted for variation among secondary forest study sites. Disturbance history amounted to at least 14.2 % and up to 78.3 % of the overall explained variation of 36.5 %, larger than the independent variation explained by environment and space, respectively. Continuous disturbance from cattle grazing was identified as the major independent discriminator (17.7 %). Differences were mainly demonstrated by the relative importance of the dominant Shorea DDF indicator species and significant variation in tree structural features.

No tree community related edge effects could be detected across the 'no buffer' category. The width of the secondary forest buffers proved critical for expanding the old-growth forest like habitat. The secondary forest of the 'wide buffer' transects closely resembled old-growth forest conditions whereas those of the 'narrow buffer' category revealed significant differences in composition and tree density. The results of the pRDA on ten edge-affected environmental variables showed that only distance effects and the 'no buffer' category accounted for significant variation.

Chronic disturbance clearly impaired the recovery of DDF while natural regeneration after low-intensity disturbance and in close proximity to seed sources allowed for rapid recovery within few decades following disturbance. In contrast to wet forest types, edge effects were not apparent, presumably due to the structural heterogeneity of DDFs. The importance of the buffer width suggested a negative correlation between proximity to contemporary clear-cut edges and regeneration dynamics. The general presence of endangered DDF species in old-growth as well as secondary forest emphasized the importance of the reserve as a repository for the conservation of biological diversity. Further research on the consequences of clear-cut edges on animal-plant interactions and the ecological quality of secondary forest as habitat is required for sound management of reserves.