Development of soil organic carbon quantification model and comparison based on CHN analyser, Loss on Ignition, and Walkley-Black methods for mangrove soils in Madagascar

Abstract

Mangrove ecosystems have a high capacity to sequester and store atmospheric carbon dioxide (CO₂) compared to other tropical forests. Most of this CO₂ is stored as organic carbon in the mangrove soils (50–90%). While quantifying carbon stocks in mangrove trees is generally a straightforward process, quantifying carbon stocks in mangrove soils can be difficult and expensive due to the lack of analytical equipment or the high costs associated with soil organic carbon (SOC) analysers, respectively. The large number of samples coupled with repeated soil sampling required to quantify soil carbon sequestration/burial rates can further inflate these costs. The development of SOC models to quantify carbon content can help resolve these issues and increase the ease of quantifying soil carbon stocks or sequestration rates. We compared different methods to quantify SOC stocks from multiple locations in Madagascar mangroves. A total of 1327 soil samples from 66 plots sampled from the west coast of Madagascar were consecutively analysed with the Walkley-Black (WB), Loss on Ignition (LOI₄₀₀, LOI₅₅₀) and CHN analyser methods. These four available methods resulted in significant differences (p < 0.001) of SOC values, varying from 1.93 ± 0.04% with CHN, 4.49 ± 0.08% with LOI_400, 4.54 ± 0.09% with WB, and 7.33 ± 0.11% with LOI₅₅₀. As the CHN method is widely regarded as the most accurate method thus, CHN values were used to calibrate and estimate LOI_400, LOI₅₅₀ and WB results with linear regression functions. After calibration, the coefficient of regressions resulted in conversion factors of 0.415 for WB, 0.253 for LOI₄₀₀ and 0.157 for LOI₅₅₀. This study demonstrated the reliability of locally available methods for SOC analysis for mangroves and developed calibration coefficients that can be used to increase the accuracy of SOC methods when CHN analyzers are not available for blue carbon stock assessment.