Soil texture prevails over vegetation change in determining soil organic carbon storage across savanna-forest boundaries

1. Tropical savannas play a unique and significant role in the global carbon cycle, yet they are increasingly being targeted for carbon sequestration through afforestation, based on the assumption that increased tree cover will substantially boost carbon stocks in both tree biomass and soils. However, the response of soil organic carbon (SOC) to afforestation in these ecosystems remains highly uncertain, and the underlying reasons for this variability are not well understood. 2. The dynamic boundaries between savannas and forests provide an ideal setting for studying the potential impacts of afforestation on soil carbon storage and stability in savannas. Here, we sampled surface soils (0-15 cm) across six savanna–forest boundaries in Hluhluwe iMfolozi Park, South Africa, and analyzed patterns of SOC stocks, including their particulate and mineral-associated organic matter (POM and MAOM) pools, and pyrogenic carbon (PyC) stocks along transects from savannas into forests. 3. Soils from savannas had comparable SOC, POM, MAOM, and PyC stocks, as well as similar proportions of POM and MAOM relative to SOC, to those from forests. Savanna soils had a significantly higher proportion of PyC relative to SOC compared to forest soils (p = 0.008). Despite the limited influence of vegetation type (savannas vs. forests) on all measured carbon pools, substantial variation in these pools was observed across transects, largely driven by differences in soil clay and silt content across the landscape. 4. Synthesis. These results contradict our expectation that savanna-to-forest transitions would increase POM pools, thereby leading to higher SOC stocks in forests compared to savannas. Instead, MAOM contributed most to SOC and was primarily driven by soil clay and silt content in this system. These findings highlight that soil texture can override the influence of vegetation on soil carbon formation and persistence, suggesting that increasing tree cover alone cannot be simply linked to soil carbon gains during savanna-to-forest transitions.