Thermophilic actinomycetes and bacilli were recovered from mushroom compost by conventional dilution plating and sedimentation chamber-Andersen sampler methods. Excessive growth of thermophilic bacilli on dilution plates accounted for the poor recovery and limited diversity of actinomycete colonies, and this result was largely unaffected by the use of modffied extraction procedures and diluents. Assessment of the actinomycete population was more successfully achieved by applying the sedimentation chamber method, by using selective media, or both. Background resistance of the compost microflora to selective agents (kanamycin, novobiocin, tetracycline, thiostrepton, and NaCI) was extremely varied, but both actinomycetes and bacilli were particularly sensitive to tetracycline. The selective isolation of Thermoactinomyces spp. and Thermomonospora chromogena by novobiocin and kanamycin, respectively, was shown to be reproducible, and the use of high concentrations of kanamycin resulted in the isolation of a novel group of unidentified thermophilic actinomycetes. Comparison of nonselective nutrient media demonstrated that the nutrient-rich protoplast regeneration medium R5 was surprisingly efficient for actinomycete recovery. This medium was found to be particularly appropriate for the recovery of Saccharomonospora viridis BD125, introduced as spores into both sterile and fresh samples of mushroom compost. This stable pigmented variant of the S. viridis strains indigenous to compost was released at concentrations of up to 107 spores g of compost-' in order to provide information for future experiments on the release and recovery of genetically manipulated strains. The detection limits for this strain were in the region of 102 g-' from sterilized compost but only 105 g-1 from nonsterile compost. These figures correspond to mean recovery efficiencies of approximately 70% (sterilized compost) and 53% (fresh compost) of viable spores released. Further improvements in the detection and recovery of S. viridis strains released into compost should be achieved by the introduction of selectable markers developed from this information on the antibiotic resistance profile of the indigenous compost microflora.

A new medium, which we propose to call R8, was developed for the isolation and enumeration of the thermophilic actinomycete, Saccharomonospora viridis. This organism has been implicated in a range of hypersensitivity pneumonitides, including farmer's lung, but is generally isolated in small numbers from contaminated environments. Recovery of S. viridis from moldy hay and mushroom compost on R8 medium was compared with recovery on conventional media. S. viridis was isolated from both substrates but in highest numbers and most consistently on the R8 medium. The selectivity of this medium was best observed when the sedimentation chamber method was used for hay samples. Here S. viridis accounted for up to 80% of the total number of actinomycetes recovered on R8 and could not be recovered on rifampin selective medium under the same conditions. R8 was also found to be an efficient recovery medium for a range of thermophiic actinomycetes from mushroom compost and for another allergenic species, Faenia rectivirgula, from moldy hay. Contamination of isolation plates by thermophilic bacilli was reduced on R8 compared with the activity on half-strength tryptone soy agar, supplemented with 0.2% casein hydrolysate, and this, together with specific improvements in S. viridis growth, accounts for the selective effect. It is possible that the occurrence of S. viridis and its role as a causative agent of hypersensitivity pnuemonitis have been underestimated by the use of suboptimal recovery protocols. It is hoped that use of R8 in conjunction with dilution plate techniques will generate information on the ecology of S. viridis and contribute to health risk assessment studies.

Survival of Bacillus subtiZis strain 168 containing plasmid pAB224, which carries a gene for tetracycline resistance, was studied in mushroom compost under mesophilic and thermophilic conditions. Stable populations of B. subtiZis were maintained as spores in both sterile and fresh mushroom compost incubated at 37 "C. At 65 "C, the introduced B. subtiZis populations declined during incubation but spores were still detectable after 28 d. Survival at the higher temperature was greater in fresh than in sterile compost. There was no apparent loss of plasmid pAB224 or plasmid-determined phenotype from the introduced B. sub& population at either incubation temperature. The frequency of tetracycline resistance in the indigenous Bacillus population was very low (lo+), but some tetracycline-resistant isolates contained plasmid DNA. Four plasmid DNA profiles were found associated with five Bacillus phenotypes, and some evidence for homology with pAB224 was found. However, pAB224 was found to be a suitable marker for release studies because it was easily recovered, readily distinguished from indigenous plasmids on agarose gels, and was maintained in compost-grown B. subtiZis 168 in the absence of any selective pressure.