Landscape Patterns and Modelling of Soil-Vegetation Relationship and Related Ecosystems Services to Support Landscape Conservation in the Mo River Basin (Togo)

Abstract

In order to support integrated landscapes and restoration efforts, this research focused on the assessment and monitoring of the spatio-temporal land use/cover change (LUCC) and degradation in the Mo River Basin (a subunit of the Volta basin of about 1,490 km2 in Central Togo). Field measurements, legacy and ancillary data were subjected to sequential multivariate methods, correlation analyses, geostatistics and modelling to analyse landscape conditions. First, along a gradient of land protection regime, data from extensive soil sampling and forest inventory were used to analyse soil organic carbon (SOC) and total nitrogen (TN) storage up to 30 cm depth, and the interactions between vegetation-soil conditions. Next, Landsat images of 1972, 1987, 2000 and 2014 combined with most updated global topographic and soil databases were used to analyse the landscape changes and its impacts on SOC, TN, soil loss potential and landscape patterns. Finally, the Landscape Management and Planning Tool adapted for the Mo basin (LAMPT_Mo), a spatially explicit model based on the Revised Universal Soil Loss Equation (RUSLE), was used to model the historical soil loss, and evaluate the efficiency of some land management scenarios. Different databases and field characterisation were used for model calibration and validation. The results showed that SOC and TN varied significantly according to land use/cover types, soil depths, topographical positions and land protection regime. With forests and woodlands exhibiting highest amounts of nutrients, mean TN varied from 0.06 to 0.16 % in the topsoil (0 – 10 cm) and 0.04 to 0.09 % in the subsoil (10 – 30 cm). Similarly, SOC ranged from 1.81 % in farmlands to 3.58 % in forests in the topsoil while woodlands had highest SOC in the subsoil (2.23 %). The river basin is made up of four and three vegetation types in unprotected and protected areas, respectively. The synergized effects of land protection status, soil conditions, landform, and human disturbances drive these vegetation patterns. From the historical analyses, natural lands dominated the basin,

though their area constantly decreased since 1972. Contemporary LUC (in 2014) is dominated by savannahs/shrubs (53 %), woodlands (27 %) and forests (11 %). Non- cultivated and vegetation regrowth areas were the most dominant of the LUCC trajectories to whom SOC, TN and soil loss potential were responsive. Trajectories of land cover decline induced soil quality deterioration while correlation analyses showed soil loss to be more landform-driven than LUCC. Simulations using LAMPT_Mo yielded values of net soil loss (NSL) far higher than the tolerable limits for the Tropics. NSL markedly changed over time with about 26, 23, 27 and 44 Mg ha-1y-1, for 1972, 1987, 2000 and 2014, respectively. Steep slopes (≥ 15 ᴼ), poorly covered lands, and riversides (distances ≤ 100 m) are critical areas of sediment source. Some intervention measures such as controlling erosion hotspots through LUC protective measures could help reducing NSL up to 70 %, to the tolerable limits for the Tropics. The combination of methods and approaches used for the monitoring and assessment of landscape change and degradation enabled to capture the different spatial aspects of the problem of land degradation in the Mo basin. The study demonstrated that important appropriate conservation measures would be necessary for the catchment rehabilitation, protection and sustainable resource use.