Tracing the trail of eroded fertile soils during a high intensity rainfall event: A fingerprinting study in war-torn tropical mountains

The depletion of fertile topsoil presents a critical challenge in tropical mountain agroecosystems. Impacts are intensified during heavy storm events that strip unprotected topsoils and pose risks to downstream water ecosystems. To better understand such dynamics, we investigated an agricultural mountainous catchment located on the Democratic Republic of the Congo shore of Lake Kivu. This area is characterised by weak governance systems exacerbated by protracted violent conflicts, which have prevented the application of remediation practices to address extensive land use changes, including deforestation and monoculture proliferation, notably banana plantations. Additionally, the spread of Banana Xanthomonas wilt (BXW) has further augmented land use impacts. Herein the complete diseased mat uprooting (CDMU) method employed to combat the disease has resulted in significant erosion and sediment export rates. Following the recovery of banana plantations from CDMU, sediment export rates continued to increase without a clear responsible party but more as a legacy of past and present land degradation activities and embedded soil erosion and connectivity process pathways. To analyse the impact of high-intensity storm events coupled with the aftermath of conflict instability and the effects of BXW on erosion dynamics, the sediment fingerprinting technique was implemented. We collected 60 sediment source samples across a 42 km² catchment, predominantly cultivated since the mid-20th century. Samples spanned four main land uses/land covers: seasonal crops, banana plantations, bare or degraded soil, and channel banks. Additionally, mixture/target samples were collected at the summit of six fluvial terraces located at different altitudes along the vertical profile of the river after a high-intensity storm event. These terraces serve as natural indicators of sediment accumulation under different flow intensities, with higher terraces representing sediment deposition during higher discharge. Furthermore, sediment samples were collected via a gravity core in delta deposits at the river's inflow into the lake, allowing us to identify the main sources contributing to fine sediment export to the lake. This methodology enabled us to assess how sediment provenance has changed in response to varying discharge levels while elucidating the primary sources responsible for fine sediment inputs to the lake and the subsequent water quality decrease. Our analysis indicates increased contributions occurred from channel banks and degraded areas as discharge increased. Despite steep channel banks and degraded soils being primary sediment sources on the terraces, agricultural and banana sources predominantly contributed to the fine sediments exported to the lake. These findings underscore the hazards of high-intensity storm events in these fragile ecosystems and highlight the role of current banana plantation and cropland management in degrading land and water quality. Understanding the primary factors driving land degradation and sediment export is crucial for preventing further ecosystem degradation and mitigating heightened instability in conflict-affected areas. This research suggests that remediation practices should be incentivised to create buffer structures in upslope agricultural areas and protect the fertile sediments from banana plantations from being exported to the lake if socio-cultural and economic hurdles to implementation can be overcome.