The influence of dissolved organic carbon and pH on the photodegradation of methylmercury in lake waters

Photodegradation rates of methylmercury (MeHg) were measured in water samples from several lakes in the Lake Berthelot region in Quebec in order to investigate the impact of drainage basin logging and associated changes in dissolved organic carbon (DOC) concentrations on MeHg levels in lake water. The lakes were selected on the basis of the amount of logging that had previously taken place in the drainage basin, and on DOC levels which were correspondingly higher in logged lakes due to increased runoff. Lakes DA9 and DF9 had DOC concentrations of 12.3 and 20.9 ppm, respectively and were designated as high DOC for the purposes of this study. Lakes N55 and N70 had DOC concentrations of 6.9 and 6.8 ppm, respectively and were designated as low DOC lakes. Experiments were conducted with both ambient and 2 ppt spiked MeHg levels in the high and low DOC lakes to determine whether the spiked levels reflect photodegradation patterns occurring in nature. The effect of different sized DOC fractions on photodegradation rates were also examined. Upon spiking, the binding of added MeHg reached equilibrium virtually immediately (within minutes) for all size fractions. It was hypothesized that less photodegradation of MeHg would be found in high DOC lakes than in low DOC lakes. In contrast to expected results, no photodegradation was observed in the two lower DOC lakes, but there was photodegradation in the two higher DOC lakes. Hypothesizing that larger size fractions of DOC would result in lower methylmercury (MeHg) photodegradation rates, water samples were fractionated into three DOC size fractions (300 kDa, 30 kDa, and 5 kDa). In the two high DOC lakes in which photodegradation was detected, DOC size fractions had an effect on photodegradation rates, though this effect varied. Average k values (hr-1) for DA9 were -1.25 x 10-1 for the 5 kDa fraction, -1.16 x 10 -1 for the 30 kDa fraction, and -9.21 x 10-2 for the 300 kDa fraction, indicating a decrease in photodegradation rates with larger fractions. For DF9, average k values were -1.69 x 10-1 for the 5 kDa fraction, -6.38 x 10-2 for the 30 kDa fraction, and -1.34 x 10-1 for the 300 kDa fraction, which does not indicate a clear trend in photodegradation rates with respect to DOC size fraction. A comparison of samples from a wetland outflow using ambient MeHg levels demonstrated that 300 kDa DOC fractions showed no photodegradation while 5 kDa DOC showed photodegradation with a k value of -9.12 x 10-2 (hr-1). Overall, watershed logging status was found not to affect MeHg photodegradation rates in the lakes studied. A possible contributor to the variation in photodegradation rates between lakes was postulated to be pH, with H+ ions displacing MeHg from binding sites on DOC and rendering it available for photodegradation. To study the relationship between rates of photodegradation and pH, water samples were adjusted to pH values of approximately 4.5, 6.5, and 7.5 and incubated in sunlight. Rates of photodegradation increased with decreasing pH conditions in all lakes, except for one lake (N55) which also had no detectable photodegradation during the initial field trials. In the remaining three lakes studied, however, lower pH values (4.1-4.6) resulted in greater photodegradation than higher pH values that resulted in little or no photodegradation. It appears that this effect only occurs in lakes with a potential for photodegradation. As results suggested that the size of DOC controls rates of photodegradation, samples were analyzed for the size of DOC and corresponding MeHg content using tangential flow ultrafiltration (TFUF). The efficacy of the TFUF system as a means of both removing microbes as well as fractionating DOC into differing size fractions was evaluated. Results indicated that this is an effective field-portable method. Mass balances for MeHg recovery after filtering large volumes of water were found to be 96% for the 5 kDa filter, 81% for the 30 kDa filter, and 109% for the 300 kDa filter, while flow rates decreased by 26%, 17%, and 50% respectively. Potential artifacts of DOC fractionation using the TFUF were investigated. The amount of DOC passing through the filter was found to decrease with increasing volume of filtrate that passed through the membrane