Assessing linkage between soil phosphorus forms in contrasting tillage systems by path analysis

•Path analysis improve the understanding of soil P forms and availability in NT and CT.•CT increase moderate labile organic P contribution to Resin-P due to mineralization.•Moderate inorganic P was directly linked to Resin-P in NT.•Resin-P was explained by P that comes from a weak HCl-P extraction, mainly in NT.•Residual organic and inorganic P fraction were not related to any fraction of P. Path analysis applied to sequential chemical fractionation of Hedley may improve our understanding on the linkage between P forms and its availability in soils. In this work, we assessed the role of Hedley-P fractions in buffering Resin-P (a plant-available soil P index) in a very clayey Oxisol (720gkg−1 clay) and the validity of the postulated causal models for two long-term (23-yr) tillage systems (conventional-CT and no-till-NT) by path analysis. The model that accounted for the path from the less labile organic and inorganic P fractions to more labile ones, and from these fractions to the Resin-P showed the highest p value in NT (p=0.36) and CT (p=0.05), showing that the proposed models are a plausible representation of the tested causal relationships. These models explained 75 and 93% of Resin-P (U=0.25 and 0.07) in CT and NT systems, respectively. The buffering flux of organic fractions was more pronounced in NT. However, the organic P pool has a higher direct contribution to buffer Resin-P in CT (94%) than in NT (35%), due to higher mineralization of organic P forms with moderate lability caused by soil disturbance. On the other hand, in the long-term NT, moderate inorganic P showed a high contribution to directly buffer Resin-P (40%). Although inorganic P associated with Ca is a very small fraction of P in strongly weathered soils, the path analysis showed that this fraction was a direct source of P in both soil tillage systems, but it was more important source to buffer Resin-P in NT (16.7%) than in CT (1.9%) due to the higher P content and path coefficient of this fraction in NT. Residual organic and inorganic P fraction were not related to any fraction of P, indicating that these fractions were neither a sink nor a source of P in both tillage systems, or that they become a temporary source and sink at the same time in the long-term experiment. The path analysis showed to be an important tool to interpret the results obtained in sequential chemical fractionation of P, improving our understanding of the soil P dynamics in contrasting tillage systems.