Differential emergence and diversity of bacterial communities from activated sludge in response to varied cultural conditions using 2,4-dichlorophenoxyacetic acid (2,4-D) were investigated by coupling molecular analyses based on 16S rDNA with functional genes. We employed three different cultural conditions: (1) a culture sequentially fed a high concentration (300 mg/L) of 2,4-D (HS); (2) a culture continuously fed a low concentration (10 mg/L) of 2,4-D (LC); and (3) a serial batch culture in which 1% (v/v) of culture was transferred to a fresh medium containing a high concentration (300 mg/L) of 2,4-D (HB). The HS and LC bioreactors were operated for 3 months and HB was repeatedly transferred for 1 month. The 2,4-D was stably degraded under all the cultural conditions tested. PCR amplification and cloning-based analysis of functional genes using community DNAs from the cultures revealed five different oxygenase genes that may be involved in the initial step of 2,4-D degradation. All five gene-types were present in HS, while one of the five genes, type V (tftA) was not detected in LC. Quantitative PCR analysis showed that in HS, Ralstonia eutropha JMP 134 type-tfdA4 (type I) was the most abundant in copy number (2.0 +/- 0.1 x 10(7) copies/microg DNA) followed by RASC type-tfdA (type II) (1.8 +/- 1.0 x 10(6) copies/microg DNA), putative cadA-like gene (type IV) (2.6 +/- 0.8 x 10(5) copies/microg DNA), cadA gene (type III) (1.3 +/- 1.0 x 10(4) copies/microg DNA), and tftA gene (type V) (3.5 +/- 1.1 x 10(3) copies/microg DNA). Similar results were obtained in LC. In contrast, HB contained only type I and type III genes, and the type I gene was five orders of magnitude greater in copy number than the type III gene. Denaturing gel gradient electrophoresis (DGGE) analysis of PCR, amplified 16S rDNA fragments of bacterial communities in the three different cultures showed low similarity coefficient values (< or =0.35) when compared to the original activated sludge, suggesting that 2,4-D amendment caused a drastic change in the bacterial community. Particularly, HB showed only six bands (16-18 bands in the other cultures) and very low similarity coefficient values when compared to the other communities (0.10 to HS, 0.17 to LC, and 0.0 to original sludge). These results indicated that serial batch culturing (HB) resulted in a phylogenetically limited number of 2,4-D degrading bacteria carrying limited catabolic genes whereas more diverse 2,4-D degraders and catabolic genes were present in HS and LC. Therefore, the approach used for monitoring should be taken into account when one evaluates the population dynamics of contaminant-degrading bacteria at bioremediation sites.