Landscape Patterns of Wood Density and Aboveground Biomass Along a Tropical Elevation Gradient in Costa Rica

Abstract:

This research sought to understand how tree wood density and taxonomic diversity relate to topography and three-dimensional vegetation structure in the tropical montane forest of Braulio Carrillo National Park in Costa Rica. The study utilized forest inventory and botanical data from twenty 1-ha plots ranging from 55 m to 2800 m above sea level and remote sensing data from an airborne lidar sensor (NASA’s Land, Vegetation, and Ice Sensor [LVIS]) to quantify variations in forest structure. There is growing evidence that ecosystem structure may help to control the functional variations across landscapes. This study relates patterns of tree functional wood density and alpha diversity to three-dimensional structure using remote sensing observations of forest structure. We were able to test the effect of the gradient on wood density measured from collected tree cores and on the subsequent aboveground biomass estimations. We sought to determine if there was a significant pattern of wood density across the altitudinal gradient, which has implications for conservation of both ecosystem services and biodiversity. We also wanted to determine how many random individuals could be sampled to accurately estimate aboveground biomass in a one-hectare plot.

Our results indicate that there is a strong relationship between LVIS-derived forest 3D-structure and alpha diversity, likely controlled by variations in abiotic factors and topography along the elevation. Using spatial analysis with the aid of remote sensing data, we found patterns along the environmental gradients defining species composition and forest structure. Wood density values were found to vary significantly from database values for the same species. This variation in tree growth has repercussions on overall forest structure, and subsequent carbon estimates extrapolated from field measurements. Because these wood density values are directly tied to biomass estimates, it is possible that carbon storage has been overestimated along this gradient using prior methods. Wood density exhibited a non-linear pattern with increasing elevation, analogous to a similar study in Peru. We also determined that sampling the wood density of just ten percent of random individuals has the potential to reasonably estimate aboveground biomass at the one-hectare scale.