Little is known about the chronic effect of herbicides on the soil microbial community, with most studies focusing on acute impacts. In this study, we investigated the effect of 20 years of atrazine and metolachlor application on the community structure, abundance and function of bacterial groups in the bulk soil of a maize monoculture. Group-specific PCR-DGGE (polymerase chain reaction-denaturing gradient gel electrophoresis) of 16S rRNA genes was used to characterize the composition of the microbial community. Primers specific for the entire bacterial group, as well as for the Acidobacterium group, the actinomycetes, the ammonium oxidizers, and the Type I and II methanotrophs were used in the PCR-DGGE analysis. Cluster analysis of the DGGE patterns obtained with the different primer sets differentiated between the herbicide-treated and the non-treated control soil. In particular the methanotrophic community differed, with three bands (phylotypes) being absent (or showing a lower intensity) in the DGGE patterns of the herbicide-treated soils compared to the patterns of the control soil. The differences in the methanotrophic community structure were not reflected in the methane oxidation capacity, which was similar for the two treatments. The prevalence of methanotrophs as evaluated with real-time PCR analysis also did not differ between the herbicide-treated and non-treated soil. Our results indicate that the long-term use of the herbicides atrazine and metolachlor resulted in an altered soil community structure, in particular for the methanotrophic bacteria. These observed changes did not cause a decreased community function (methane oxidation), probably because the total abundance of the methanotrophs in the soil system was preserved.

Agricultural practices, such as mineral nitrogen fertilization, have an impact on the soil's ability to oxidize methane, but little is known about the shifts in the methanotrophic community composition associated with these practices. Therefore, the long-term effect of both mineral (NH4NO3) and organic (manure and GFT-compost) fertilizer applications on the soil methanotrophic community activity and structure were investigated. Both high and low affinity methane oxidation rates were lower in the soil treated with mineral fertilizer compared to the other soils. An enhanced nitrate concentration was observed in the mineral fertilized soil but nitrate did not show a direct affect on the high affinity methane oxidation. In contrast, the low affinity methane oxidation was slowed down by increased nitrate concentrations, which suggests a direct effect of nitrate on low affinity methane oxidation. Denaturing gradient gel electrophoresis (DGGE) analysis of 16S rRNA gene fragments specific for methanotrophs revealed a distinct community between the mineral and organic fertilized soils as extra Type I methanotrophic bands (phylotypes) became visible in the organic fertilized soils. These phylotypes were not visible in the patterns of the added organic fertilizers suggesting an indirect effect of the organic fertilizers on the methanotrophic community. Additionally, a molecular analysis was performed after the low affinity methane oxidation test. The enhanced methane concentrations used in the test enriched certain low affinity methanotrophs in the organic fertilized soils but not in the mineral fertilized soil. Supporting the molecular and functional observations, fatty acids characteristic for methanotrophs were less abundant in the soil treated with mineral fertilizer compared to the soil treated with compost. In conclusion, the function and molecular and chemical composition of the methanotrophic community are all altered in soil fertilized with mineral fertilizer.

In a pear tree orchard planted on loam soil, each plot was treated in April 1998 with either one of the ureas diuron or chlorotoluron, or triazine simazine herbicides applied at 3, 4, and 2 kg AI ha(-1), respectively. Some plots had not been previously treated with one of these herbicides. Other plots had been treated annually during the past 12 years with the same herbicide. One herbicide, and always the same, was thus applied to each plot. In the plots treated for the first time with either diuron, chlorotoluron, or simazine, the soil half-lives of these herbicides in the 0-10 cm surface soil layer were 81, 64, and 59 days, respectively. In the plots treated with the same herbicide for 12 years, the corresponding soil half-lives were 37, 11, and 46 days. Diuron thus produced a moderately enhanced biodegradation, chlorotoluron a high one, and simazine a low but significant one. In another pear tree orchard planted on sandy loam soil, each plot was treated in April 1998 with one of the amide propyzamide (1.25 or 1.0 AI kg ha(-1)) or diflufenican (250 g AI ha(-1)) herbicides. In the plots not previously treated with propyzamide, the propyzamide soil half-life was the same for both doses, i.e., about 30 days. In the plots treated annually for 3 or 14 years with propyzamide, the soil half-life was 12 and 10 days, respectively. In the plots treated for the first time with diflufenican and in those treated annually with diflufenican for 3 years, the diflufenican soil half-life was the same, i.e., 65 days. Propyzamide thus already showed a highly accelerated biodegradation after 3 years of repeated annual applications. Diflufenican, however, did not show enhanced biodegradation after 3 years of repeated annual applications.

By 2015, all herbicide use on public pavements in Flanders (northern region of Belgium) will be phased out. Currently, little is known about weed flora in these pavements or their interactions with different weed control methods. The objectives of this study were to explore the species composition of pavements in relation to various abiotic factors and applied weed control methods. A vegetation survey was conducted on 163 public pavements constructed with small paving elements across Flanders. Botanical composition was determined, and a score for street scene perception was calculated. For each pavement, a set of environmental conditions and technical characteristics was determined, and data on the applied weed control methods were collected. Apart from Musci (mosses), the most important plant species were Poa annua, Sagina procumbens, Conyza canadensis, Taraxacum officinale and Plantago major. Weed species composition and street scene perception were affected by intensity of use, joint width and light intensity. Weed prevention measures may be built into the construction of pavements by manipulating light, joints and use in their design. Repeated use of a single weed control method caused shifts in weed flora. This suggests that more optimal weed control on pavements is likely to be achieved by alternating weed control methods

No abstract