Contrasting Covariance Patterns of Soil-Test Phosphorus and Potassium in Routine Agricultural Soil Records

Atav V., Kayalı E., Gürbüz M. A., Kardeş T. A., Demirci P., Körükçü B. C.

Research in agricultural sciences, cilt.57, sa.3, ss.290-300, 2026 (Scopus, TRDizin)

Özet

Routine soil analysis reports are widely generated, yet they are commonly interpreted one

indicator at a time. When such records are compiled into databases, it becomes important to

determine whether different soil-test indicators reflect the same general soil condition gradient or

require context-specific interpretation. In this study, a cleaned analytical database was compiled

from routine soil analysis records covering 2023–2026 from northwestern Türkiye to examine the

joint behavior of saturation percentage, pH, electrical conductivity, organic matter, available

phosphorus, and available potassium. The extraction workflow yielded 1,450 rows from the

uppermost reported soil layer, and a complete-case subset of 905 observations was retained for

the main multivariate analyses. Descriptive review, Pearson and Spearman correlation analyses,

targeted phosphorus pairwise analysis, principal component analysis, and grouped pH × organic

matter summaries were used to evaluate co-variation. The results did not support a single fully

coherent soil-property pattern. Saturation, pH, organic matter, and available potassium showed a

comparatively regular shared structure, with the strongest overall pairwise association observed

between saturation and pH (Pearson r = 0.586, p < .001). Available phosphorus was only partly

aligned with that broader pattern and showed its clearest association with potassium (Pearson r =

0.534, p < .001), while its links with pH, saturation, and electrical conductivity were weak. Grouped

pH × organic matter summaries further showed that potassium increased more regularly across

contextual classes, whereas phosphorus behaved more unevenly. Overall, available potassium

fitted the broader multivariate structure more consistently, whereas available phosphorus required

direct contextual interpretation rather than being treated as a simple component of a general

fertility axis.