Relative contributions of abiotic properties, soil microbes, and management practices to soil health in intensively managed rice paddies

Preserving the health of soils worldwide is the grounds on which we can solidify human progress. Soil health describes the capacity of soils to function and depends on both abiotic properties and soil biota. The degree to which soil properties or soil microbes dictate soil health, as well as whether the contributions of soil properties and soil biota to soil health overlap, remain uncertain. We address here the relative contributions of abiotic properties, PLFAs and management practices in determining soil health, which we assessed via three ecosystem processes. We specifically evaluated how the abiotic variables (0–10 cm depth) and common phospholipid fatty acid (PLFA) markers determined soil health over a network of 36 intensively managed smallholder farms. The experimental design was a block design with three blocks (i.e. three villages) and one manipulation, whether the farms undergone an intermediate fallow period. We abstracted soil health in relation to three ecosystem processes (proxies of ecosystem functioning), crop yield (kg/mu), water stable aggregation (WSA, %) and decomposition, assessed using the in situ 30-day weight loss of green and rooibos tea bags, assayed twice in November 2022, and August 2023. We subsequently questioned the degree to which these three proxies of soil health depended on biotic and abiotic properties of the soils. Across the 36 rice farms, key soil properties exhibited substantial variation, with pH ranging from 3.94 to 7.37, sand content from 9.06 % to 53.50 %, soil organic carbon from 0.53 to 3.39 g/kg, phosphorus from 7.77 to 57.68 mg/kg. We observed that PLFA markers were predominantly driven by soil nutrient availability of C (Soil organic carbon), N (Total nitrogen and NH₄⁺-N) and P, as well as soil pH and sand content. The soil properties accounted for 4.6 % of the variability in PLFA markers and were strongly associated with them. Briefly, high nutrient availability of C, N and P would increase the abundance of soil microbes. In addition, abiotic soil properties (pH, Olsen-Phosphorus, ammonium and soil sand content) and PLFA markers (arbuscular mycorrhizal fungi, fungi; bacteria; total microbial biomass) made comparable (13.3 and 11.7 %, respectively) contributions (i.e. variation partitioning) to the functioning of the ecosystems. We thus propose that maintaining soil health requires the simultaneous consideration of both abiotic and biotic variables.