Allometric models generalisation adding more variables to estimate aboveground biomass for algarrobo(Neltuma flexuosa)across environmental gradients in Argentinian Monte Desert woodlandsOA

Dryland forests provide essential contributions to human communities,including timber production and carbon sequestration.Estimating forest biomass and productivity is an essential input for forest management planning,but it is difficult in dryland forests due to the variability in architecture between one-and multi-stemmed in-dividuals.In consequence,the models developed for biomass estimation have the disadvantage of being site-specific,limiting their application in large areas of dryland forests.We created a set of allometric models that allow us to estimate the aboveground biomass(AGB;kg)of Neltuma flexuosa in woodlands throughout the Monte Desert without using destructive biomass assays.To account for the species’characteristic multi-stemmed ar-chitecture,the models focused on two levels:the stem level and the individual level.At the stem-level model,the main predictor variable was basal diameter(BD;cm).Individual-level models use equivalent basal diameter(EBD;cm),crown diameter(C;m),and/or tree height(H;m)as predictors.In both cases,we compared the benefits of general and site-specific models in order to account for the climatic gradient of the study sites.The results proved that stem-level models must include site-specific parameters in addition to the BD to accurately predict AGB.Among the individual-level models,the highest predictive power was provided by the full model,including all predictors,with a coefficient of determination R^(2)=0.94 and corrected Akaike information criterion(ΔAICc)>23.This model can be used as a general predictive tool for all the Monte woodlands.In contrast,simplified individual-level models use fewer variables,allowing less measuring fieldwork but reducing the predictive power.In this case,the models benefit significantly from site-specific parameters,making their use for multiple woodlands less appropriate.