Effects of cropping system and fertilization regime on soil phosphorous are mediated by rhizosphere-microbial processes in a semi-arid agroecosystem

In semi-arid regions, soil phosphorus (P) dynamics in cereal-legume intercropping are not yet fully elucidated, particularly in relation to integrated application of fertilizers. To this aim, we investigate the effects of different fertilizers on various P fractions in relation to the rhizosphere-microbial processes in a cowpea/maize intercropping system. Field experiments were conducted during two consecutive years (2016–2017) in a split-plot design by establishing cowpea/maize alone or intercropped onto the main plot, while the sub-plot was treated with four types of fertilization, i.e. no fertilizer addition (control), organic amendment (compost), mineral fertilizers (NPK) and multi-nutrient enriched compost (NPKEC). Our results showed that NPKEC fertilizer increased NaHCO3-Pi by 69% in maize, 62% in cowpea and 93% in intercropped plots compared to control plots. Similarly, a significant increase in the NaHCO3-Po fraction was also recorded with NPKEC treatment in all cropping systems. In case of moderately labile P, NPKEC fertilizer caused the highest increase of NaOH-Po (12.87 ± 0.50 mg P kg−1 soil) and NaOH-Pi (22.29 ± 0.83 mg P kg−1 soil) fractions in intercropped plots. Except for intercropping, NPK application caused an increase in the non-available P fraction (HCl-Pi), while the use of NPKEC decreased the HCl-Pi concentration in all cropping systems, suggesting stronger merits both for intercropping and NPKEC. Surprisingly, maize exhibited substantially higher phosphatases activity compared to cowpea in monoculture amended with compost, implying distinct crop strategies for adaptation under low P conditions. Based on the multi-factor analysis, the close association of NaHCO3–P with P solubilizing bacteria, root carboxylates and pH indicated that rhizosphere processes are the strongest predictors of immediately available P. Since alkaline phosphatase (ALP) is a P-degrading enzyme of microbial origin, rhizosphere related ALP association may have originated from root-associated microflora promoting P mobilization. Furthermore, the strong association of microbial biomass P (MBP) and acid phosphates (ACP) with NaOH–P fraction indicated moderately available P cycle in soil was mainly driven by microbial-related processes. Factor analysis map and two-way ANOVA confirmed that fertilization regime had a stronger effect on all tested variables compared to cropping system. Altogether, our results suggest that a combination of microbial-rhizosphere processes