Allometric equations for mangrove biomass estimation: A global mixed-effects analysis
| Authors: | Charles A. Price, Benjamin Branoff, Todd A. Schroeder, Skip J. Van Bloem, Monica Papeș, Humfredo Marcano-Vega |
| Year: | 2026 |
| Type: | Scientific Journal |
| Station: | Southern Research Station |
| DOI: | https://doi.org/10.1016/j.ecoinf.2026.103991 |
| Source: | Ecological Informatics |
Abstract
Accurate estimation of mangrove biomass is a key component of the blue carbon economy, and allometric equations are a cornerstone of methodological approaches to quantify carbon from plot to landscape scales. A perennial question facing investigators is whether global allometric equations can be applied to their region of interest, or whether trees need to be harvested to provide accurate local estimates. Further, it is not always clear which combinations of tree measurements provide the most accurate estimates. We examined 16 different allometric equations for four tree measurements within a global data set of 710 trees for three common trans-Atlantic mangroves species. To determine the best performing models, we used mixed effects with site location and tree species modeled as random effects, and combinations of tree height, diameter at breast height (DBH), canopy diameter, and wood density as fixed effects resulting in 112 different models. The best performing model was a fixed effects model with three of four parameters (R2=0.974, RMdSE=2.495) although many mixed effects models were comparable, and in general models with more parameters performed better even after penalizing model complexity. Variance partitioning and the conditional mode of univariate models indicate that site and species identity contribute relatively little to overall variance. These results provide qualified support for the use of global equations when high-quality local equations are not available. For example, a DBH only fixed effects model adds only 514 grams of median error relative to the best model (R2=0.949, RMdSE=3.009). Our results can also help subsequent investigators prioritize data collection efforts, by highlighting data gaps, and variable combinations that yield the most accurate biomass estimation versus the time required to measure them.