A new approach to the interpretation of soil tests for phosphate response by grazed pasture

Abstract

Five soil test methods (Truog, Olsen, Bray 1 and 2, and Egner et al.) using two depths of soil, and volumes of soil as well as weights for Truog and Olsen, were compared as methods of predicting response to phosphate by grazed clover - grass pastures. Sixty-two field experiments were run for four years on seven soil types representing three major soil groups of New Zealand (yellow-brown loams, yellow-brown pumice soils, and yellow-grey earths). Response to phosphate was measured in spring and autumnwinter by mowing, but for most of the year the experimental areas were included in the surrounding paddock and managed and grazed accordingly. Soil samples (0 – 38,0 –75 mm) were collected each season. Correlation coefficients between relative yield (RY) and soil tests within soil types were extremely inconsistent between times of measurement of response. For example, within a soil type, the R value for one method ranged from 0.85 to -0.32 over the 8 measurements of response. At anyone time of measuring response, however, there was good correlation between methods (R>0.85). Using the mean values of RY and the soil tests from individual sites within a soil type did not improve the level of correlation. Overall correlations using all measurements of RY and their corresponding soil test values within soil groups gave low correlation coefficients (<0.5). An alternative approach has been to relate the soil test values to a probable minimum yield (PMY). The plots of RY against soil test values within soil types or groups did not show the wide random scatter expected from the low correlation coefficients but there was a well defined pattern. At low soil tests, the range of RY was wide (100–40%) but this narrowed progressively, as the test values increased, to a range of 100 - 90%. There were few, if any, low RY at high soil test values. PMY has been defined as the RY below which the observed RY will fall on only 20% of occasions at a given soil test value. PMY was related to the soil tests by the expression: PMY = c + d loge soil test. It represents pastures under soil and plant conditions at which they are very close to their maximum sensitivity to lack of phosphate. The present evidence is that these conditions occur generally over the slow growth period from summer to late winter and when pastures have a high content of white clover. The possibility of overestimating the phosphate status of a site and depriving pasture of phosphate is lessened by using the soil test to predict PMY, rather than using the average RY derived from the regression of RY on the soil test. This approach ensures better growth of pasture during the time when adequate pasture is critical to total annual animal performance. It also ensures a phosphate environment in which white clover can compete more strongly against grasses and maintain the essential input of nitrogen. The best prediction of PMY over the three soil groups was given by the Olsen test (0–75 mm) using a volume of soil.

Keywords:

• available phosphorus
• clover
• grass
• pasture
• soil testing
• field experiments
• phosphate fertiliser
• seasonal response
• correlation