Improving predictive efficiency of Comparative Performance Analysis (CPA) of Agroecosystems for Sustainable Land Management (SLM)
by Open Science Repository Agriculture
Maize is the main food crop in the hilly areas of Nepal. A Comparative Performance Analysis (CPA) study was undertaken to identify maize yield constraints and quantify their impacts on yield in Makuwanpur District (Central (Nepal). For 68 fields, yield ranged from 200 to 4900 kg/ha with a mean of 2020kg/ha, and a standard deviation of 1120kg/ha. Data on land and management aspects were collected during a field survey. Data acquisition methods included direct measurement/observation of biophysical aspects (recorded on relevee sheet), interviews of land users, visit to a nearby research station and extension offices. The dataset included land and management parameters. The production model identified a terrain unit (Upland plateau & Highland hills versus all others), a management parameter (Quantity of applied urea) and lodging as the three major yield constraints that explained 65% of the encountered yield variability. The impact of each identified constraint on yield (yield-gap of 1928 kg/ha) was 61%, 23% and 16%, respectively. Results of this CPA study were in line with the farmers` perception of constraints faced to explain differences between actual and expected yields. CPA was further explored to find significant interactions of urea applied with other land and management aspects to improve the predictive power of the earlier model. Use of specific constraint-management functional relationships enhances the standard CPA method and provides opportunities to identify by constraint specific management requirements. This method could have a wide range of application in the context of diverse agroecosystems and NRM constraints.Keywords:
Land Use Systems, CPA, SLM, quantified production function model, stepwise regression.
Full textImproving predictive efficiency of Comparative Performance Analysis (CPA) of Agroecosystems for Sustainable Land Management (SLM)
Click here to post comments
Return to journal abstracts.