Subinhibitory levels of antibiotics can promote the development of antibiotic resistance in bacteria. However, it is unclear whether antibiotic concentrations released into aquatic systems exert adequate pressure to select populations with resistance traits. To examine this issue, 15 mesocosms containing pristine surface water were treated with oxytetracycline (OTC) for 56 days at five levels (0, 5, 20, 50, and 250 microg L(-1)), and six tetracycline-resistance genes (tet(B), tet(L), tet(M), ted(O), tet(Q), and tet(W)), the sum of those genes (tet(R)), "total" 16S-rRNA genes, and transposons (Tn916 and Tn 1545) were monitored using real-time PCR. Absolute water-column resistance-gene abundances did not change at any OTC exposure. However, an increase was observed in the ratio of tet(R) to 16S-rRNA genes in the 250 microg L(-1) OTC units, and an increase in the selection rate of Tc(r) genes (relative to 16S-rRNA genes) was seen when OTC levels were at 20 microg L(-1). Furthermore, tet(M) and Tn916/1545 gene abundances correlated among all treatments (r2 = 0.701, p = 0.05), and there were similar selection patterns of tetR and Tn916/1545 genes relative to the OTC level, suggesting a possible mechanism for retention of specific resistance genes within the systems.

Antibiotic resistance genes (ARGs) are emerging contaminants that are being found at elevated levels in sediments and other aquatic compartments in areas of intensive agricultural and urban activity. However, little quantitative data exist on the migration and attenuation of ARGs in natural ecosystems, which is central to predicting their fate after release into receiving waters. Here we examined the fate of tetracycline-resistance genes in bacterial hosts released in cattle feedlot wastewater using field-scale mesocosms to quantify ARG attenuation rate in the water column and also the migration of ARGs into peripheral biofilms. Feedlot wastewater was added to fifteen cylindrical 11.3-m3 mesocosms (some of which had artificial substrates) simulating five different receiving water conditions (in triplicate), and the abundance of six resistance genes (tet(O), tet(W), tet(M), tet(Q), tet(B), and tet(L)) and 16S-rRNA genes was monitored for 14 days. Mesocosm treatments were varied according to light supply, microbial supplements (via river water additions), and oxytetracycline (OTC) level. First-order water column disappearance coefficients (kd) for the sum of the six genes (tetR) were always higher in sunlight than in the dark (-0.72 d(-1) and -0.51 d(-1), respectively). However, water column kd varied among genes (tet(O) < tet(W) < tet(M) < tet(Q); tet(B) and tet(L) were below detection) and some genes, particularly tet(W), readily migrated into biofilms, suggesting that different genes be considered separately and peripheral compartments be included in future fate models. This work provides the first quantitative field data for modeling ARG fate in aquatic systems.