Date of this Version
J. Environ. Qual. doi:10.2134/jeq2015.06.0280
The inherent spatial heterogeneity and complexity of antibioticresistant bacteria and antibiotic resistance (AR) genes in manureaffected soils makes it difficult to sort out resistance that can be attributed to human antibiotic use from resistance that occurs naturally in the soil. This study characterizes native Nebraska prairie soils that have not been affected by human or food-animal waste products to provide data on background levels of resistance in southeastern Nebraskan soils. Soil samples were collected from 20 sites enumerated on tetracycline and cefotaxime media; screened for tetracycline-, sulfonamide-, b-lactamase–, and macrolide-resistance genes; and characterized for soil physical and chemical parameters. All prairies contained tetracyclineand cefotaxime-resistant bacteria, and 48% of isolates collected were resistant to two or more antibiotics. Most (98%) of the soil samples and all 20 prairies had at least one tetracycline gene. Most frequently detected were tet(D), tet(A) tet(O), tet(L), and tet(B). Sulfonamide genes, which are considered a marker of human or animal activity, were detected in 91% of the samples, despite the lack of human inputs at these sites. No correlations were found between either phenotypic or genotypic resistance and soil physical or chemical parameters. Heterogeneity was observed in AR within and between prairies. Therefore, multiple samples are necessary to overcome heterogeneity and to accurately assess AR. Conclusions regarding AR depend on the gene target measured. To determine the impacts of food-animal antibiotic use on resistance, it is essential that background and/or baseline levels be considered, and where appropriate subtracted out, when evaluating AR in agroecosystems.