Aboveground biomass of naturally regenerated and replanted semi-tropical shrublands derived from aerial imagery

摘要

Rapid assessment of plant size and population densities is important for estimating biomass over large areas, but it has often been limited by methods requiring intensive labor and resources. In this study, we demonstrate how shrub biomass can be estimated from fine-grained aerial photographs for a large area (23,000 ha) located in the Lower Rio Grande Valley, Texas, USA. Over the past 30 years, refuge land management has included the replanting of native shrubs to promote the restoration of wildlife habitat and carbon sequestration. To assess shrub regrowth, we developed a method to estimate individual shrub canopy areas from digital aerial imagery that was used to calculate biomass from allometric equations. The accuracy of the automated delineation of individual canopies was 79 % when compared to that of hand-digitized shrub canopies. When applied to photographs across the refuge, we found higher shrub densities for older naturally regenerated sites (174 individuals ha(-1)) compared to those of younger replanted sites (156 individuals ha(-1)). In contrast, naturally regenerated sites had less biomass (3.43 Mg ha(-1)) than replanted sites (4.78 Mg ha(-1)) indicating that shrubland restored for habitat conservation has the potential to sequester more carbon in a shorter period. There was an inverse relationship between aridity and aboveground shrub biomass for replanted sites in the drier west (p < 0.05). We found a difference in predicted biomass among shrub species in replanted sites that was also associated with climate (p < 0.05). We conclude that the canopy of individual shrubs detected from remote sensing can be used to estimate and monitor vegetation biomass over large areas across environmental gradients.