Thu, Aug 18, 2022: 9:15 AM-9:30 AM
516A
Background/Question/MethodsOver the last couple of decades, the capacity of metabolic ecology theory (MET) to explain the fundamental scaling allometric relationships in vascular plants has been challenged on its lack of universal fit to empirical data. Despite MET undergoing a revision a decade after its original inception, the lack of empirical fit remained, and thus the subsequent work to predict wider ecosystem level variables using the theory is also disputable. We therefore re-evaluate the fundamental MET relationships, culminating in amended universal relationships between tree diameters and their height and biomass (generalized MET models). These amended relationships should therefore underpin wider ecological relationships. Most predominantly, in an age when lidar altimeters onboard planes and satellites are enabling massive measurements of tree heights globally, these universal relationships will permit direct quantification of carbon fluxes and other ecological variables. We fitted our generalized MET models using a N = 4,004 dataset, independently compiled for pantropical models that are widely in REDD+ assessments of forest carbon.
Results/ConclusionsWe compared our generalized MET models against the scaling MET relationships, and the original pantropical models. The results strikingly showed that the generalized MET relationships outperformed the original pantropical models fitted from that same datasets, improving its root mean squared errors from 3.822 to 3.803 m in the height model, and from 1.404 to 1.304 Mg in the biomass model. Moreover, the generalized version rectified the bias of the original MET, reducing the mean differences from 30.8 to 6.5 cm in height prediction, and from 30 to 8 kg in biomass prediction.The results therefore demonstrate the ability of the generalized MET models to perform better than or comparably to the best available statistically fitted models. As such the revaluation of MET not only improves conceptually upon the original, but also solves the previous issues of empirical inferiority. The generalized MET relationships thus give a renewed avenue of development for a series of statistically powerful theory based ecological models. The integration of these amended theory-based relationships with remote sensing based models (namely lidar) are therefore a priority for the monitoring of ecosystem variables at scale.
Results/ConclusionsWe compared our generalized MET models against the scaling MET relationships, and the original pantropical models. The results strikingly showed that the generalized MET relationships outperformed the original pantropical models fitted from that same datasets, improving its root mean squared errors from 3.822 to 3.803 m in the height model, and from 1.404 to 1.304 Mg in the biomass model. Moreover, the generalized version rectified the bias of the original MET, reducing the mean differences from 30.8 to 6.5 cm in height prediction, and from 30 to 8 kg in biomass prediction.The results therefore demonstrate the ability of the generalized MET models to perform better than or comparably to the best available statistically fitted models. As such the revaluation of MET not only improves conceptually upon the original, but also solves the previous issues of empirical inferiority. The generalized MET relationships thus give a renewed avenue of development for a series of statistically powerful theory based ecological models. The integration of these amended theory-based relationships with remote sensing based models (namely lidar) are therefore a priority for the monitoring of ecosystem variables at scale.