Identifying drivers of carbon using structural equation modeling in a subtropical urban forest

Background/Question/Methods

Understanding variables that are associated with higher or lower aboveground carbon in forest ecosystems will aid in managing our forest resources for reducing carbon emissions through sequestration in plant biomass. We tested the effect of land use (as a proxy of anthropogenic influence on forest ecosystems), tree density, vegetation cover, species diversity, composition, and richness on aboveground carbon using data from urban forests in San Juan Puerto Rico. These variables interact with each other; therefore, they have both direct and indirect effects, via intermediate variable, on aboveground carbon. We teased out these relationships using structural equation modeling.

Results/Conclusions

Percentage tree cover had a significant positive effect on species composition, species diversity, basal area, and aboveground carbon. Species composition and basal area also affected aboveground carbon positively; however, species diversity did not have a statistically significant influence on carbon. Land use also had significant direct and indirect effects on the aboveground. The direct effect of land use was lower than its indirect effect via tree density. Our study indicated that land use, tree density, and species composition are important predictors of aboveground carbon and these issues should be addressed while managing our forest resources, especially in urban areas.