Density, activity, and diversity of bacteria indigenous to a karstic aquifer.

The microbial ecology of karstic ground water is largely unknown. The density, activity, and diversity of bacteria indigenous to subsurface karstic material in Mammoth Cave National Park, Mammoth Cave, Kentucky were studied using minimally disruptive, on-site procedures. Two sites, located 100 m below the surface and consisting of saturated fine to coarse sand in pooled water, were examined. Samples were taken aseptically using modified, sterile 60-cc syringes. Total cell and total respiring cell densities were determined using an acridine orange/p-iodonitrotetrazolium violet (AO/INT) staining procedure. Cells in selected cores were stained with INT and incubated in the cave for 4 h prior to fixing with glutaraldehyde and subsequent transport to the laboratory. Cells were stained with AO in the laboratory. Low- and high-nutrient media were used to determine viable cell counts. Plates were incubated in the cave for 1 day at ambient temperature prior to transportation to the laboratory in an insulated cooler. Viable cell counts ranged from 1.0 × 106 to 8.1 × 106 cells wet g(-1) of sediment. Total direct counts were 3.9 × 106 and 1.4 × 107 cells wet g(-1) for the Olivia's Dome and the Catherine's Dome sites, respectively. Viable cell counts were highly similar to respiring cell counts at both sites. At the Olivia's Dome site, viable cell counts represented 26-31% of the direct cell counts, while 58% of the total cell count were actively respiring. At the Catherine's Dome site, viable cell counts represented 11-58% of the direct counts, while 53% of the cells were actively respiring. A total of 237 strains recovered from low- and high-nutrient media at both Olivia's and Catherine's Domes, and 10 reference strains were examined for 117 morphological, biochemical, and physiological characteristics. Results were coded in a binary fashion and analyzed using numerical taxonomic techniques. Similarity values were calculated using a simple matching coefficient. Fifty-two clusters, ranging in size from 2 to 13 members, were defined at the 80-85% similarity level with the weighted pair-group mathematical average algorithm (WPGMA). The matrix was examined using the Jaccard coefficient and WPGMA clustering to control for distortion due to negative matches and varying group size. Presumptively identified genera include, Arthrobacter, Brevibacterium, Bacillus, Cornyebacterium, Actinomyces, Aureobacterium, Chromobacterium, and Mycobacterium. Pseudomonas spp. were not recovered. Fifty percent of the clustered operational taxonomic units (OTUs) were not identified. Thirty percent of the clustered OTUs were irregular, asporogenous, Gram-positive rods. The bacterial communities varied between sites, and isolation medium had a strong influence on the strains recovered. The bacterial community in the karstic sediments sampled exhibits a high degree of diversity having no dominant strain or strains.