Movement and fate of pesticides in wetland buffer zones: reactive transport and the influence of seasonal environmental changes.

Riparian wetlands are proposed to mitigate diffuse pollution of surface water by pesticides in agricultural landscapes. Biodegradation might play an important role in mitigation in wetlands, especially in case of mobile pesticides, but information on this is scarce. The overall goal of this thesis was to acquire a better understanding of the process of biodegradation of pesticides in riparian wetlands. Emphasis was put on the dynamics of the biodegradation process and the responsible microorganisms as a consequence of seasonally dependent environmental changes on the one hand and the involved reactive transport mechanisms on the other. Isoproturon (IPU) and 2-methoxy-4-chlorophenoxy acetic acid (MCPA) were used as relevant model pesticides mainly because of their low tendency to sorb to soil. A riparian wetland in Bernissem (Belgium) that was restored in 2009 as a controlled flooding basin, was used as a major study subject.

In a first study, it was investigated how seasonally dependent environmental changes affect pesticide (IPU and MCPA) mineralization kinetics in flooded conditions in wetland microcosms, using sediment from the restored riparian wetland in Bernissem. The microcosms were subjected to artificially induced freeze-thaw and dry-wet events before they were spiked with 14C-labeled pesticides and subsequent 14CO2 production was monitored. The mineralization of IPU and MCPA was significantly reduced after exposure to especially freeze-thaw events, as evidenced by lower mineralization rates and longer lag times compared to control microcosms that were not exposed to freeze-thaw or dry-wet events. We expected that the altered mineralization kinetics could be explained by reduced population sizes of pesticide degrading biomass, which we investigated using a most probable number (MPN) approach. Interestingly, the numbers of IPU and MCPA mineralizers were unexpectedly higher in freeze-thaw and dry-wet cycle exposed microcosms compared to the control microcosms and hence the pesticide mineralization kinetics correlated poorly with cell numbers of pesticide mineralizers. This suggested that the observed effects of seasonally dependent environmental changes were due to other mechanisms than decay of pesticide mineralizers, like catabolite repression. In addition, in a second series of microcosms in which the growth of pesticide mineralizing bacteria was stimulated by prior amendment of IPU and MCPA, exposure to a freeze-thaw or dry-wet event only marginally affected the pesticide mineralization kinetics.

Only little information exists about the spatial and temporal variations of pesticide biodegradation potential in wetlands, including variations due to in situ seasonal changes like freeze-thaw events. It was therefore investigated in a second study how IPU and MCPA mineralization potential in the sediment of the restored riparian wetland in Bernissem varied during a two year period that included various freezing events. To that end, wetland sediment samples were repeatedly taken at 30 locations to determine the mineralization kinetics of IPU and MCPA ex situ in a high throughput laboratory assay. The potential for mineralization of both pesticides was clearly present in the wetland’s sediment, up to four years after restoration. However, the mineralization potential varied strongly spatially in the wetland even for replicate samples taken at the same location, which can be related to the occurrence of hotspots of pesticide mineralization potential in the sediment. The differences in pesticide mineralization potential between the locations within the wetland could to a large extent be related to the physicochemical characteristics of the sediment. Moreover, locations that were constantly inundated showed consistently less pesticide mineralization potential compared to locations that were periodically non-inundated, which indicated that the physico-chemical conditions and pesticide mineralization potential might be related to the redox conditions of the sediment. The mineralization potential also differed between sampling time points, with larger lag times for samples taken in winter when the temperatures were lower, particularly in samples from wet locations. The mineralization potential was clearly affected after a (series of) freezing event(s) but recovered during spring. The results further indicate that spatial and temporal variation have to be taken into account to determine the contribution of biodegradation in wetlands and their effectiveness to mitigate pesticide pollution of surface water bodies.

Often when pesticides enter a wetland and a pesticide mineralization potential is present, a lag time precedes the actual mineralization of the pesticide, which we also in particular observed in our wetland microcosms. Because the pesticide is not effectively degraded during the lag time, it might instead diffuse into the anoxic sediment layer. We hypothesized that aerobic degradation of the pesticide in the oxic water column and top layer of the sediment and concomitant growth of the responsible microbiota would subsequently deplete the concentration of the pesticide at the sediment surface and induce desorption and upward diffusion of the pesticide. This mechanism was examined and supported using an experimental design based on laboratory diffusion cells of which the water column was spiked with IPU and the sediment was inoculated with a specialist IPU-degrading Sphingomonas sp. strain SH. The concentration of IPU in the diffusion cells was monitored by HPLC and the growth of Sphingomonas sp. strain SH was verified by quantification of catabolic genes using real-time quantitative PCR. By means of a one-dimensional mathematical model, the interaction of downward and upward diffusion and growth-linked degradation of the pesticide was successfully illustrated using microbial biomass densities and growth rates determined in suspension in batch experiments. Optimization of the model indicated that the effective initial size of the biomass in the diffusion cells and the microbial growth rates are important factors that govern the pesticide degrading process in wetlands. The study suggests that the wetland sediment can provide a storage and buffer of a unique carbon and energy source for pesticide degraders to cope with irregular pesticide pulses in a wetland. It elucidated a mechanism that is likely to occur in surface flow riparian wetlands and is of importance for understanding the fate of mobile pesticides in wetland ecosystems.

Our work contributes to the understanding of how pesticides are biodegraded in riparian wetlands. In general, our findings can be concluded by four statements:

- Mineralization of pesticides can be realized under flooded conditions.

- Pesticide mineralization potential varies spatially in wetlands, but redox conditions might explain why the mineralization potential is lower in locations that are permanently inundated than in locations that are periodically non-indated.

- Seasonal disturbances, especially freezing and thawing, reduce pesticide mineralization kinetics, which has considerable implications for wetland effectiveness.

- The fate of pesticides in wetland sediment is determined by the interaction of biotic (growth-related degradation) and abiotic (diffusion and sorption) processes.