Human or natural disturbance: Landscape-scale dynamics of the tropical forests of Puerto Rico

Increasingly, ecologists are recognizing that human disturbance has played an important role in tropical forest history and that many assumptions conc erning the relative importance of natural processes warrant re-examination. To assess the historical role of broad-scale human vs. natural disturbance on an intensively studied tropical forest we undertook a landscape-level a nalysis of forest dynamics in the Luquillo Experimental Forest (LEF; 10 871 ha) in eastern Puerto Rico. Using aerial photographs (1936 and 1989), GIS, a model of topographic exposure to hurricane winds, and historical data, w e sought to: (1) document historical changes in extent, cover, and type of forest vegetation; (2) evaluate the distribution of land use and hurricane impacts; (3) assess the contributions of these processes in controlling cur rent vegetation patterns; and (4) relate these results to ongoing ecologica l, conservation, and natural resource discussions. With >1000 m of relief in the LEE the broad vegetation zones of Tabonuco (< 600 m above sea level), Colorado (600-900 m), Dwarf (>900 m), and Palm fore st are determined by environmental gradients. However, over the past 60-100 years, forest extent, cover, and type have been transformed: in 1936, 40% of the LEF was unforested or secondary forest and <50% had continuous canop y (>80% cover); in 1989, >97% was continuous forest. Secondary forest and a gricultural lands in 1936 were replaced largely by Tabonuco and Colorado fo rest, which increased from 8% and 28% (1936) to 26% and 45% (1989). These broad-scale vegetation dynamics are best explained by a gradient of h uman land use, intense at low elevations and decreasing on steep, high terr ain, which peaked historically around 1900, followed by a gradual decline i n agriculture. GIS analysis and historical sources suggest that essentially all of the LEF was affected by human activity and that Tabonuco forest, wh ich is the focus of LTER research, has been most substantially altered and is largely of secondary origin. Rapid reforestation following agricultural decline has obscured much of the past land use and confirms the resiliency of some tropical forests to intensive human disturbance. Impacts of earlier hurricanes (e.g., in 1928 and 1932), although not evident in the broad for est pattern in 1936, may be significant in explaining the distributions of Colorado and Palm forest. Damage from Hurricane Hugo in 1989 indicates that natural disturbance is increasingly important as land use declines and for est cover and height increase. However, this study and post-Hugo studies em phasize that land use legacies are long-lasting and need to be considered i n modern ecological studies and natural resource management. The subtle, al though persistent, effects of historical human activities may have profound consequences for modern forest ecosystems in the tropics.