Patterns in Reproductive Litterfall along an Elevation Gradient in a Tropical Forest Daniel Keck Abstract The ecological impacts of global climate change have become increasingly well-documented recently. Tropical ecosystems may be particularly sensitive to small climatic changes due to their relatively constant, aseasonal climates. This study sought to investigate the effects of climate change on reproductive litterfall production in a tropical forest. The study site was situated along a 330 m elevation gradient in a humid Puerto Rican forest. The elevation gradient and the corresponding 1°C temperature gradient were used as a proxy for potential climate change. Litterfall was collected for eight years at 12 sites along the elevation gradient spanning four distinct forest types. The litter was then sorted into four fractions, dried, weighed for mass, and analyzed for carbon and nutrient content. This study focuses on the fraction of the litter containing reproductive structures including fruits, flowers, and seeds. Differences in magnitude as well as temporal patterns of reproductive litterfall support the classification of the cloud forests as distinct forest types. Overall, the forests displayed a significant negative relationship between elevation and reproductive litterfall mass. Reproductive litterfall production decreased significantly during the course of the study in all of the forest types, but the temporal patterns in each forest type varied dramatically. These results suggest that small climatic changes may significantly alter reproductive patterns in tropical forests. Changes in reproduction are likely to have significant impacts on ecosystem function. It is likely that any climatic changes that do occur will differentially affect the forest types described in this study.Introduction The vast majority of climate change research has focused on mid and high latitudes where warming is predicted to be the greatest (Walther et al. 2002). However, recent modeling studies suggest that humid, tropical ecosystems may be particularly sensitive to even very small changes in climate because of their current lack of significant drought stress and near constant temperatures throughout the year (Loope and Giambelluca 1998, Wang et al. 2003). These small climatic changes have the potential to alter carbon and nutrient allocation patterns (Silver 1998). Using elevation gradients to investigate the effects of differences in microclimate is a common approach, especially in conjunction with litterfall measurements (Reiners and Lang 1987, Waide et al. 1998). There are a number of well-established relationships between temperature, elevation, litterfall production, and litterfall nutrient content. Litterfall production is positively correlated with temperature (Silver 1998). In the tropics, litterfall production tends to be negatively correlated with elevation, which follows from the previous correlation since elevation and temperature are negatively correlated (Waide et al. 1998). Litterfall nitrogen concentrations are also positively correlated with temperature (Vitousek and Sanford 1986). Allocation of carbon and nutrients in plants has been the subject of extensive research, but the vast majority of work has focused on carbon allocation to vegetative structures and has neglected reproductive structures as carbon and nutrient sinks. The small amount of work that has focused on reproductive allocation has investigated the effects of elevated carbon dioxide levels (He and Bazzaz 2003). Research into the reproductive allocation of wild species has remained especially limited (Amthor 2001, Saxe et al. 2001). Reproduction is critical to ecosystem function and deserves much more attention than it has received. Island tropical montane cloud forests, like my study site in Puerto Rico, are especially valuable for investigating the effects of climate change due to their extreme sensitivity to climatic variation (Loope and Giambelluca 1998). My research focuses on how carbon and nutrient allocation to reproduction in a mixed tropical forest vary along an elevation gradient. I hypothesize that differences in elevation, and consequently climate, will lead to differences in resource allocation, specifically carbon and nutrient allocation to reproductive structures. This question and hypothesis reflect the broader goal of the project, which is to develop a more complete understanding of the ecological impacts of climate change.Methods Study Site The study sites were located along a 330 m elevation gradient (635-968 meters above sea level) in the upper elevations of the Luquillo Experimental Forest (LEF), Puerto Rico (18° 19’ N 65° 45’W). A network of twelve 30 m x 10 m permanent plots were used along the gradient where all trees ≥ 2.5 cm diameter at breast height (dbh) were tagged and identified to species. The forest type changes along the elevation gradient with colorado forest at 500-750 meters above sea level (masl), palm forest at 500-1000 masl, and cloud forest at 750-1050 masl. A significant difference in tree height and a small difference in tree species composition were noted between the short and tall cloud forests. Mean annual rainfall at all sites is approximately 3700 mm, and temperatures decrease from 20oC in the lower elevation sites to 19oC in the upper elevations. Cloud water deposition adds an estimated 500 mm/yr in precipitation (Brown et al. 1986, Weaver 1995). Litterfall Collection and Analyses Litterfall was sampled every two weeks from 1994 to 2004 from five, perforated plastic, mesh-lined baskets (0.14 or 0.17 m2) per plot, distributed in a stratified random manner to assure plot coverage. Baskets were suspended above the ground on short (approximately 50 cm) PVC poles to facilitate drainage. Litter was dried at 65oC and sorted into fruits and flowers, wood (< 10 cm diameter), leaves, and miscellaneous material. The fruits and flowers fraction contains all reproductive structures including fruits, flowers, seeds, and buds. The wood fraction contains small twigs, bark and woody stems. The leaf fraction contains all leaves except palm fronds and ferns; because palm and fern fragments and leaflets could not be identified with confidence they were included in the miscellaneous (unrecognizable organic matter) fraction. After sorting, samples were redried at 65oC and weighed to determine mass. Samples were then bulked by litter
View Full Document
Unlocking...