Modelling and mapping erosion in smallholder agro‐ecosystems, Tanzania

Abstract

The West Usambara Highlands in north‐eastern Tanzania have many smallholder agro‐ecosystems with unknown composition, management, and vulnerability to erosion. Their specific locations and spatial extent are difficult to trace by satellite images or remote sensing imagery alone. To address these limitations, we combined ground soil surveys, geographic information system, and erosion modelling to (a) locate and map smallholder agro‐ecosystems, (b) determine their biophysical characteristics, and (c) model their soil losses. Land resource information was collected from 301 random 0.1‐ha plots sampled from a total area of 200 km2. Annual soil losses were estimated using the universal soil loss equation. The study located six dominant agro‐ecosystems with the following spatial extent: maize‐bean (24.9%), maize‐bean‐agroforestry (31.2%), maize‐bean‐agroforestry‐high value trees (18.9%), tree farms (7.0%), forests (15.6%), and grazing lands (2.3%). Agroforestry and other tree‐based agro‐ecosystems dominate the area due to historical land use change and later institutional interventions. This study finds combined use of soil surveys, geographic information system, and modelling to be reliable in locating, mapping, and assessing soil losses in smallholder agro‐ecosystems. The agro‐ecosystems differ significantly (p < 0.05) in slope, vegetation cover, soil conditions, and soil losses. Soil loss in the maize‐bean agro‐ecosystem (28.3 t ha−1 yr−1) was 18 times higher compared with natural forests (1.57 t ha−1 yr−1) due to lower soil cover and inefficient conservation and cultivation practices. Our results show that adoption of soil conservation measures and improved vegetation cover technologies across the agro‐ecosystems reduces soil losses by 37% and increases organic carbon levels by 16%.

Influence of biophysical aspects on the performance of sustainable land management measures in the Usambara highlands of Tanzania

“…Although social and economic factors affecting land degradation in the West Usambara Highlands are known (Tenge et al, 2004), perceptions of farmers, development workers and policy makers towards it are not fully understood. Further, though soil erosion is the major form of soil degradation affecting the West Usambara Highlands, its magnitude in smallholder agro-ecosystems of the area is unclear (Wickama et al, 2018). Smallholder agro-ecosystems do not have a unique spectral property from which one can detect their locations, composition, and biophysical characteristics by spatial/remote sensing tools and eventually model their soil loss situations.…”