Twenty years of forest monitoring in Cinnamon Bay Watershed, St. John, U.S. Virgin Islands

Weaver, Peter

St. John, and probably the Cinnamon Bay watershed, has a history of human use dating to 800 B.C. The most notable impacts, however, occurred from 1730 to 1780 when sugar cane and cotton production peaked on the island. As agriculture was abandoned, the island regenerated in secondary forest, and in 1956, the Virgin Islands National Park was created. From 1983 to 2003, a period that included Hurricanes Hugo in 1989 and Marilyn in 1995, and a severe drought in 1994-95, 16 plots, stratified by elevation and topography, were monitored in the Cinnamon Bay watershed. In all years, from 55 to 60 percent of the stems were in height classes between 4 and 8 m, and 75 percent of the stems were in dbh (diameter at 1.3 m above the ground) classes between 4 and 10 cm. Stem density was greatest on the summit, followed by ridges, then slopes, and lowest in valleys. Sixty-five percent of the original stems survived after 20 years with an average dbh growth rate of0.07±0.00 cm yr-1. Tree species showed a preference for particular sites according to elevation and topography within watershed. Total aboveground biomass on all plots, averaging 138.7 t ha-1 in 1983, declined by nearly 7 percent by 2003. In 1983, biomass was greatest on the summit and least on ridges; by 2003, the quantities had converged except on the summit plot. Total aboveground productivity was 10.64 t ha-1 yr-1. Standing herbivory (leaves) was estimated at 4.5 percent, and the herbivory rate at 4.6 percent per year. The standing crop of litter was 9.33 t ha-1. Hurricanes had a major impact on forest structure and species composition. Impacted trees (snapped, uprooted, or standing dead) totaled 210 ha-1 10 months after Hurricane Hugo, and 288 ha-1 19 months later. The proportion of impacted stems differed by elevation, topography, aspect, and slope. Tree species with >20 individuals showed a difference in the proportion of impacted stems, ranging from as low as 0.6 percent for Pimenta racemosa (Mill.) J.W. Moore to as high as 22.8 percent for Nectandra coriaceae (Sw) Griseb. In conclusion, the structure, species composition, and forest dynamics within the Cinnamon Bay watershed vary in time and space, and short-term observations characterize only a fragment of the watershed's continuously changing vegetational history. Monitoring of forest structure and dynamics should continue.