Regional analysis of forest biomass at the rainforest/sclerophyll boundary in northern Queensland, Australia

Abstract

We present a regional analysis of the tree biomass in the Rainforest/Wet Sclerophyll ecotone of the tropical regions of Northern Queensland, Australia. Diameter at breast height (DBH) was measured for all trees larger than 2 cm girth at breast height on a total of 3237 20×20 m plots. DBH was converted to biomass using allometric equations. For all plots, the environmental conditions were characterized by 10 local attributes and, in addition, the geographic information system for the region was queried for a further 52 variables (climate, topographic, soil parent material). By evaluating the predictive capability of all variables using generalized linear and additive models, we ranked the relative importance of environmental controls on biomass. There is a clear signal in an overall very noisy data set. Spatial patterns of precipitation and to a lesser degree temperature are the strongest controls of biomass. Soil parent material and soil depth play a lesser role at the considered scale. Topographic soil moisture surrogates are the weakest predictors, possibly due to the relatively coarse pixel size of one ha of the digital terrain model. The best predictor is the mapped forest type (air-photo interpretation) indicating a large amount of systematic, unaccounted variability - most likely fire history or unaccounted rainfall pattern. The variability within regional climate zones is large compared to the variability within zones that can be readily identified from airphotos.