PROJECT ASSESSMENTS - Science for Agriculture and Rural Development in Low-income Countries - page 158

out in the catchment with 120 households. The review in the Tanzanian site
consisted of group discussions and transect walks. A total of 104 households were
interviewed and several fields were visited during the transect walks.
Cluster 2: Identification and calibration of indicators , to (i) identify the main
indicators that farmers use to quantify erosion and to (ii) attach semi-quantitative values
to the erosion indicators, using scientific measurements. In group meetings, in semi-
structured interviews with 120 households, discussions with key-informants and during
transect walks farmers listed known erosion indicators and assessed the indicators in
their fields and their causes. Indicators were categorized into current indicators (those
that are observable immediately after a rainfall event) and past indicators (resulting
from long-term erosion) leading to a consensus list of erosion indicators for the study
area. Field measurements were done to link measured soil loss and yield loss to the
presence of the most important indicators. On 9 combinations of soil type and slope
class runoff plots were installed to relate the sheet-rill erosion indicators to actual
soil loss. Five erosion indicators were identified within about 25 farmers' fields to
estimate crop yield gaps.
Cluster 3: Construction of a tool for participatory soil erosion mapping . In this
cluster farmers were assisted to produce a soil erosion map based on the indicators
identified before. A participatory mapping exercise was facilitated resulting in a
detailed topographic map of the area (with property delineation, and infrastructure).
A field-by-field survey of soil erosion and crop production levels was carried out
during 2 seasons using the previously compiled consensus list of indicators, resulting in
a soil erosion map for the catchment, according to farmers' knowledge and perceptions.
The method was evaluated and described.
Cluster 4: Field surveys and modelling exercises to assess the degree of soil
erosion using scientific methods. The actual erosion was assessed in the field following
the guidelines of the Assessment of Current Erosion Damage method (ACED). The
method was applied along transects (Kenya case) and on field level (Tanzania case),
and resulted in both cases in an erosion map for the catchment, indicating the spatial
distribution of erosion in a (semi-) quantitative way.
Two different models were used to estimate spatial patterns of soil erosion. The
Morgan, Morgan and Finney (MMF) model is an empirical model developed to
estimate mean annual soil loss from field-sized areas on hill slopes. The model was
selected for its simplicity and relative low data requirements. The second model
used is the LISEM model, a model based on physical-chemical laws and equations
that predict erosion patters within a catchment for a single rainfall event. The
LISEM model was calibrated and validated with data on soil and water loss at the
outlet of the study catchments. For this purpose spatial data on climate, soils and
crops were collected as well as data on runoff and soil loss at the outlets of the
catchments. Flumes were constructed and equipped with automatic sampler equip-
ment for discharge and sediment load.