J.B. Hall

Predicting Species Response to Disturbance from Size Class Distributions of Adults and Saplings in a Jamaican Tropical Dry Forest

The diversity of tropical dry forests is poorly described and their regeneration ecology not well understood, however they are under severe threat of conversion and degradation. The Hellshire Hills constitute a dry limestone forest reserve on the south coast of Jamaica that is of high conservation value. In order to describe the structure and composition of this forest and assess the extent to which the population structures of its tree species do characterize their regeneration ecologies, pre-disturbance structure, floristics and seedling populations were compared with post-disturbance species responses in twelve 15xa0m × 15xa0m permanent sample plots which were laid out in a blocked design in April 1998, giving a total sample area of 0.27xa0ha. These plots were subjected to disturbance in April 1999 (cutting) with each of four blocks being assigned with two randomly allocated treatment plots (partially and clear cut) and one control plot (uncut). A total of 1278 trees (≥2xa0cm DBH) and 7863 seedlings and saplings (0–2xa0mm and 2–20xa0mm root collar diameter (RCD) respectively), comprising 60 and 52 species, respectively, were sampled in the plots prior to disturbance. The species-area curve for trees reached a maximum at 0.20xa0ha, and abundance was widely distributed amongst the species (26 had importance values greater than 1%); four species were notably codominant (with importance values between 7 and 8%). The forest stand structure had a reverse J-shaped curve for tree and for seedling/sapling size-class distributions, which indicated that the forest as a whole was probably regenerating adequately. From an analysis based on adult and sapling size-class distributions (SCDs), 21 species with 15 or more individuals were classified into 3 groups. Many of the species (15 of the 21), had flat adult SCDs that deviate from the whole-community reverse J-shaped SCD. However, sapling SCDs for 6 of the 15 species were strongly positive indicating the potential for their populations to be sustained by recruitment from the saplings present. No general association was found between these SCD species groupings and the actual ability of the species to recover from disturbance. Analysis of post-disturbance response revealed that for only 9 of the 21 species did adult SCDs provide adequate prediction, but for an additional 6 of the species information on sapling SCDs improved the accuracy of prediction if the ‘release’ of saplings or smaller individuals predominated recovery. However in this forest, recovery following disturbance which left stem and roots in place is predominantly by coppice regrowth, and there were no significant correlations found between adult SCDs and the species’ ability to coppice.