STORAGE EFFECTS ON INDIGENOUS SOIL MICROBIAL COMMUNITIES AND PGPR EFFICACY

We tested the hypothesis that storage-induced changes in the compositi on of soil microbial communities altered the efficacy of three Bacillu s and three Pseudomonas plant growth promoting rhizobacteria (PGPR) st rains on spruce seedlings. Forest soils were collected from near Macke nzie, Salmon Arm and Williams Lake, B.C. Soil microbial population siz es were evaluated using dilution plating. Carbon substrate utilization patterns were determined on fresh forest soils and after 32 weeks of storage at 4 degrees C and -10 degrees C using the Biolog system. Smal l aliquots of fresh and stored soils comprising <2% of seedling growth medium were also used in PGPR assays. The population sizes of viable soil microorganisms, particularly bacteria and mycorrhizal fungi, decl ined 10-1000 fold during soil storage at both temperatures. Difference s in substrate utilization patterns-between soils and after storage tr eatments were also detected by Shannon's diversity index (H'), princip al component analysis and cluster analysis of BIOLOG data. The three f resh soils formed a distinct cluster when plotted on ordination axes o r by Euclidean distance. Williams Lake and Salmon Arm soils maintained a greater degree of substrate utilization diversity when stored froze n, but the opposite was true for Mackenzie soil. In general, the six P GPR strains had similar effects on spruce seedling growth within soil x storage treatments. Pseudomonas root colonization did not differ sig nificantly between strains (Bacillus colonization was not evaluated), forest soils and storage treatments, but PGPR efficacy depended on the origin and treatment of forest soil. For example, spruce growth promo tion was greatest in fresh Williams Lake and Salmon Arm soil, but PGPR efficacy in Mackenzie soil was greatest after storage at -10 degrees C. Our results support the hypothesis that storage-induced changes in soil microflora contribute to variability in PGPR efficacy, but we wer e unable to identify specific characteristics of these forest soils th at related directly to plant growth response variability. (C) 1998 Els evier Science Ltd. All rights reserved.