Soil Organic Matter Dynamics in the Ericaceous and Afroalpine Belts of the Bale Mountains, Ethiopia: Influence of Vegetation, Fire, and Topographic Factors

Soil organic matter (SOM) dynamics in tropical montane ecosystems remain poorly understood, particularly regarding the relative importance of particulate versus mineral-associated fractions under varying disturbance regimes. This study investigated SOM fraction distribution across the Ericaceous and Afroalpine belts of Bale Mountains National Park, Ethiopia, an Andosol-dominated landscape subject to recurrent fire. Using a stratified sampling design (n = 30 plots) across four vegetation classes (Ericaceous belt, fragmented Ericaceous belt, herbaceous and heathland, and giant Lobelia areas), three fire history categories (<10, 10–25, and >25 years since fire), and three topographic positions (northern slopes, southern slopes, and central plateau), we quantified coarse particulate organic matter (cPOM: 149–2000 μm), fine particulate organic matter (fPOM: 53–149 μm), and mineral-associated organic matter (MAOM: <53 μm). Particulate fractions dominated the SOM pool, with cPOM and fPOM together accounting for >99% of measured organic carbon. Multivariate ordination revealed a primary gradient (PC1, 61.7%) contrasting particulate-dominated soils in less disturbed areas with relatively MAOM-enriched soils in fire-impacted and fragmented zones. A global comparison reveals a profound stability gap: the Bale Mountains utilize <2% of the mineral stabilization potential of comparable Andosols, demonstrating that extreme fire frequency (<25 yr return interval) overrides even the most reactive mineralogy. We critically evaluate whether standard size-based fractionation adequately captures mineral-associated carbon in volcanic soils and discuss methodological limitations. These results provide baseline data for conservation planning in this biodiversity hotspot and underscore the need for fire management strategies that balance ecological integrity with carbon storage objectives.