EVALUATING SOIL MAPS FOR PREDICTION OF SOIL-WATER PROPERTIES

Conventional soil survey stratifies a region into mapping classes and characterizes each by a representative soil profile within it. The eff icacy of the procedure for predicting particle-size fractions, bulk de nsity, water retention, and available water capacity (AWC) of the soil at previously unvisited sites on the Plain of Languedoc in southern F rance is evaluated for three scales of survey (1/10 000, 1/25 000 and 1/100 000) and is compared to that of prediction from stratified rando m and simple random samples. Data from 85 soil profiles on a random tr ansect were used for evaluation. Classification partitioned the variat ion of the measured properties, except for AWC, well at the 1/10 000 a nd 1/25 000 scales, whereas classification at the 1/100 000 scale was less effective. At the 1/10 000 and 1/25 000 scales both classificatio n and stratified random sampling were better for prediction than simpl e random sampling for the same total sample. On average the representa tive profiles proved substantially better predictors than the stratifi ed random samples, but in most situations where soil stratification pe rformed well efficiencies of the two predictors were similar. In essen ce, the more successful the classification was the more difficult it w as to improve prediction by selecting representatives instead of sampl ing randomly within classes. These results confirmed statistically tha t the soil surveyor can exercise intuition and judgement to classify a nd select representatives.