Collecting the Dirt on Soils : Advancements in Plot-Level Soil Testing and Implications for Agricultural Statistics

Carletto, Calogero ; Aynekulu, Ermias ; Gourlay, Sydney ; Shepherd, Keith

World Bank, Washington, DC

关键词: LAND PRODUCTIVITY;    HOUSEHOLD SURVEYS;    SOIL SPECTROSCOPY;    SOIL FERTILITY;    LOCAL KNOWLEDGE;   

DOI  :  10.1596/1813-9450-8057
RP-ID  :  WPS8057

学科分类：社会科学、人文和艺术（综合）

来源: World Bank Open Knowledge Repository

【 摘 要 】

Much of the current analysis onagricultural productivity is hampered by the lack ofconsistent, high quality data on soil health and how it ischanging under past and current management. Historically,plot-level statistics derived from household surveys haverelied on subjective farmer assessments of soil quality or,more recently, publicly available geospatial data. TheLiving Standards Measurement Study of the World Bankimplemented a methodological study in Ethiopia, whichresulted in an unprecedented data set encompassing a seriesof subjective indicators of soil quality as well as spectralsoil analysis results on plot-specific soil samples for1,677 households. The goals of the study, which wascompleted in partnership with the World Agroforestry Centreand the Central Statistical Agency of Ethiopia, weretwofold: (1) evaluate the feasibility of integrating a soilsurvey into household socioeconomic data collectionoperations, and (2) evaluate local knowledge of farmers inassessing their soil quality.Although a costlier methodthan subjective assessment, the integration of spectral soilanalysis in household surveys has potential for scale-up. Inthis study, the first large scale study of its kind,enumerators spent approximately 40 minutes per plotcollecting soil samples, not a particularly prohibitivefigure given the proper timeline and budget. The correlationbetween subjective indicators of soil quality and key soilproperties, such as organic carbon, is weak at best.Evidence suggests that farmers are better able todistinguish between soil qualities in areas with greatervariation in soil properties. Descriptive analysis showsthat geospatial data, while positively correlated withlaboratory results and offering significant improvementsover subject assessment, fail to capture the level ofvariation observed on the ground. The results of this studygive promise that soil spectroscopy could be introduced intohousehold panel surveys in smallholder agriculturalcontexts, such as Ethiopia, as a rapid and cost-effectivesoil analysis technique with valuable outcomes. Reductionsin uncertainties in assessing soil quality and, hence,improvements in smallholder agricultural statistics, enablebetter decision-making.