Spatial data calibration for site-specific phosphorus management

Accounting for spatial variability of soil properties commonly requires intensive soil sampling, which inevitably involves a high cost. Geo-spatial statistical tools enable characterization of spatial variability and development of sampling strategies from limited data. This study outlines a simple approach of using classical and geo-spatial statistics to understand the spatial variability of soil Phosphorus (P) and discusses its relevance to sampling strategy and variable rate P application. The Bray (I) extractable-P data, obtained from a previous study, was first explored using descriptive statistics, box plot and normal quantile plot analyses. Spatial description of the data was performed using qualitative (data posting) and quantitative (variography) methods. Information derived from the fitted semivariogram was used to perform data interpolation (kriging). A management zone concept was used to delineate the Bray P test values. Results showed that Bray P exhibited a strong spatial dependence with 94% of its variability explained. The spatial conelation length was 177 m. Spatial attributes of the data appeared to justify the sampling design employed with regard to sample size, spacing and arrangement. To facilitate variable rate P application, three management zones were established so as to receive low, moderate and high P rates, respectively.