Plant growth-promoting microorganisms enhance maize performance under salinity stress: a comprehensive meta-analysis of morphological, physiological, and biochemical responses

Abstract

Salinity stress severely limits global maize ( Zea mays L.) productivity by disrupting plant growth, physiology, and yield formation. Plant growth-promoting microorganisms (PGPMs), including bacteria and mycorrhizal fungi, have been widely investigated as sustainable tools for improving crop performance under salinity stress conditions. This quantitative systematic review and meta-analysis, conducted under PRISMA guidelines, integrated data from 107 peer-reviewed publications to generate pooled effect-size estimates quantifying the effects of inoculation with plant growth-promoting microorganisms (PGPMs) on maize under varying salinity levels. Random-effects modelling indicated that PGPM inoculation was associated with significant improvements in key morphological traits, including grain weight, shoot and root biomass, and leaf area, with the most consistent effects observed under moderate salinity stress conditions (approximately 25–40% increases across biomass-related traits). Physiological responses were consistent with improved plant water status and membrane stability, including increased relative water content (~ 15–30%) and reduced electrolyte leakage (~ 20–45% under stress conditions), alongside improved K⁺/Na⁺ ratios. Biochemical responses included significant increases in osmolytes and antioxidant metabolites, such as glycine betaine (GB) (~ 30–35%), soluble sugars (~ 40–70% under stress), flavonoids and phenolics (~ 25–40%), and ascorbic acid (~ 30%), whereas proline responses were context-dependent and varied according to stress intensity. Subgroup analyses showed that arbuscular mycorrhizal fungi (AMF) and 1-aminocyclopropane-1-carboxylate (ACC) deaminase-producing microorganisms were consistently associated with enhanced stress-related responses, including increased mycorrhizal colonisation and modulation of ethylene-related processes. Substantial between-study heterogeneity was detected (I 2 > 50% for multiple traits), reflecting variation in microbial taxa, maize genotypes, and experimental conditions. Assessment of publication bias and robustness using funnel plot inspection, trim-and-fill analysis, and leave-one-out sensitivity analyses indicated that pooled estimates were not substantially affected by potential small-study effects or individual influential studies .