Aboveground biomass allometric models for a private semi-natural forest in Nigeria

Abstract

Private forests with conservation priority such as Abayomi Farm Estate (AFE) Emerald forest reserve, Nigeria can significantly contribute to the global carbon cycle while enhancing sustainable livelihoods. However, little consideration is given to accounting for their biomass pools and carbon sequestration. This study, therefore, developed models for estimating aboveground biomass in the private semi-natural forest. Four (4) temporary sample plots (TSPs) of 50 x 50 m were systematically sampled with a complete, non-destructive enumeration of 176 individual tree species with a diameter at breast height (DBH) > 10 cm. Aboveground biomass models were developed using the enumerated parameters covering a wide range of DBH and total height (H), as well as wood density (WD) as predictor variables. The models were developed for the two most-abundant, native tree species and all species combined in the forest. The models were evaluated using different indices such as coefficients of determination (R²), root mean square error (RMSE). Selected models were cross-validated. The species-specific biomass models with double predictors proved more accurate and reliable for estimating aboveground biomass in the forest than the DBH-only allometry, with their adjusted R² as high as 95 % and RMSE < 0.23. Mixed-species allometry fitted by all the three predictors (DBH, H and WD) was the most suitable, depicting the added relevance of wood density and sample size in biomass modelling. It recorded RMSE and adjusted R² of 0.22 and 97 %, respectively. Overall, all the models provided good estimates and could be used for assessing the carbon storage in the forest estate.