Assessing aboveground carbon density with field observations and biophysical and spectral predictors in the northmost Neotropical dry forest

Abstract

The tropical dry forest ecosystem is the most widespread terrestrial tropical vegetation in Mexico and is highly threatened by anthropic disturbance and climate change. Accurate estimates of aboveground biomass and corresponding carbon stocks can influence forest management strategies and help direct or evaluate the effectiveness of REDD+ programs. Here, we assess the aboveground carbon density estimated with field observations and biophysical and spectral predictors in the northmost tropical dry forest of the Americas occurring in the state of Sonora in northwestern Mexico. Our top candidate model with biophysical predictors (tree and structural richness, slope, cation exchange capacity and soil depth) showed the best fit and lower prediction error (pR² = 0.4, RMSE = 0.458) of the spatial distribution of aboveground carbon density in this forest. The effect of structural richness and soil depth was stronger; therefore, these appear to be the most important drivers of aboveground carbon spatial variation across the region. The total aboveground carbon storage predicted with this model in the entire region was 19 305 499.5 Mg C ha⁻¹ (mean = 11.8, sd = 6), with higher aboveground carbon density estimated toward more tropical latitudes. A comprehensive assessment of aboveground carbon density in the tropical dry forest requires a synergistic approach combining field observations and biophysical drivers in lieu of more advanced remote sensing techniques, such as LiDAR that are still not available or validated in many tropical regions.