Proximity to Riparian Wetlands Increases Mercury Burden in Fish in the Upper St. Lawrence River

Round 1

Reviewer 1 Report

Generally, the paper is well researched and clearly written.  The authors appear to have a good handle on the literature specific to this study despite the obscurity of some of the sources.  There are only a few issues to be addressed.

1) The notion of the distances between WL and NW sites is a bit confusing.  I assume the purpose of the distance analysis is to make sure one or both species do not spend equal amounts of time in both habitats thus not being indicative of the Hg content from one habitat or the other.  The use of this methodology could be clearer.

2) One would expect wetlands to have more trace element accumulation than non-wetlands as clays and organic matter can effectively adsorb metal ions.  Were any studies of the sediment trace metal geochemistry done to validate that expectation?  Were any studies of the Hg content in the Yellow Perch and Round Goby food chain done to see if bioaccumulation was in fact happening and thus the larger the fish the more Hg would be present as the older fish fed on the higher (and greater Hg accumulation) food chain species?

3) Is this a public health issue?  Is the level of Hg in these species high enough to be of concern for eating by animals or humans?  What is the level of concern for Hg in freshwater fish?  As we close coal-fired power plants the Hg previously sent in to the environment will be sequestered deeper in the wetland sediments presumably making it less bioavailable.  Will a similar thing happen to Hg as occurred with the decrease in acid rain pH over many decades?  In other words, how much of a bio-environmental threat is this for humans and can this concern, if any, be projected to the other freshwater sites in the US and abroad?

Author Response

Reviewer #1:

Generally, the paper is well researched and clearly written.  The authors appear to have a good handle on the literature specific to this study despite the obscurity of some of the sources.  There are only a few issues to be addressed.

1) The notion of the distances between WL and NW sites is a bit confusing.  I assume the purpose of the distance analysis is to make sure one or both species do not spend equal amounts of time in both habitats thus not being indicative of the Hg content from one habitat or the other.  The use of this methodology could be clearer.

An edit to the text on line 92 was made to address this concern.

2) One would expect wetlands to have more trace element accumulation than non-wetlands as clays and organic matter can effectively adsorb metal ions.  Were any studies of the sediment trace metal geochemistry done to validate that expectation?  Were any studies of the Hg content in the Yellow Perch and Round Goby food chain done to see if bioaccumulation was in fact happening and thus the larger the fish the more Hg would be present as the older fish fed on the higher (and greater Hg accumulation) food chain species?

No, there were no sediment trace metal geochemistry analyses completed to compare the two site types. Although, we agree with the statement that more trace metals may accumulate in wetlands vs non-wetlands. Some geochemistry analyses associated with mercury cycling have been completed in other Upper St. Lawrence River riparian wetlands (Brahmstedt et al. 2019), but there are no data for the non-wetland shoreline sites for comparison.

However, the focus of the study here is mercury mobilization, rather than the general status of mercury in wetlands vs non-wetlands since the bioavailability is what will ultimately lead to public health and wildlife health concerns. We assume the mercury mobilized into the food chain and into yellow perch and round goby is representative of the bioavailability of mercury in wetlands and non-wetlands, and that these two species represent mobilization into the greater food chain since they feed lower in the food chain but may be consumed by higher predators, which may lead to greater Hg biomagnification. We can see this trend somewhat in the yellow perch data across lengths as we see their diet shift in age and increasing length. As far as assessing the food chain at trophic levels below yellow perch and round goby, this was not assessed. However, given that food item consumption is the primary vector and cause of bioaccumulation in organisms, we assume the trends between wetland and non-wetland sites that we see in yellow perch and round goby are representative of that occurring elsewhere in the food chain, particularly the parts of the food chain that are isolated to either wetland or non-wetland sites.

3) Is this a public health issue?  Is the level of Hg in these species high enough to be of concern for eating by animals or humans?  What is the level of concern for Hg in freshwater fish?  As we close coal-fired power plants the Hg previously sent in to the environment will be sequestered deeper in the wetland sediments presumably making it less bioavailable.  Will a similar thing happen to Hg as occurred with the decrease in acid rain pH over many decades?  In other words, how much of a bio-environmental threat is this for humans and can this concern, if any, be projected to the other freshwater sites in the US and abroad?

We agree these are good questions and have addressed most, if not all, with additional text at the end of the discussion that discusses relevance to public health and consideration of mercury as an ongoing threat. (lines 286-292 and 296-303)

Reviewer 2 Report

This article is straightforward, interesting, and reads well. The authors collected fishes (yellow perches and gobies) from eight wetlands and seven non-wetland habitats along the Upper St. Lawrence River and analyzed their mercury contents. They found that mercury levels are higher in fishes from wetlands than in fishes from non-wetland habitats. They conclude that wetlands may mobilize mercury into the food chain more so than non-wetland habitats. However they do not offer a possible explanation for this in the discussion. In fact, the explanation for this phenomenon is given in the introduction from lines 61 to 73. I suggest to add a sentence or two in the discussion restating what has been said in the introduction regarding methylmercury production and bioaccumulation in wetlands. Also, a recent study by Brasso et al. (2020) has yielded mixed evidence that wetlands animals (songbirds in their study) are at greater risk of mercury bioaccumulation than in non-wetlands. It might be good to add a reference to this article to moderate the conclusions. Finally, I'd suggest adding a hint of two regarding what solutions could be implemented to solve this problem. For example, a paper by Ackerman (2020) reports on an experiment involving the construction of open‐ and deep‐water treatment cells at the downstream end of seasonal wetlands to promote naturally occurring MeHg removal processes.

Minor comments:
- l19: "Mercury levels were significantly (p<0.01) higher ..." I believe the authors refer to their ANCOVA results here? If that's the case shouldn't it be p<0.001?
- l137: "Sampled analyzed in wet weight." Please add a verb to the sentence.
- l143: A t-test was used to determine if there was a difference between the two groups in length. Please note that if the normality tests fails, it is usually better to use a non-parametric test instead of the t-test.

References:
- Ackerman, J. T.; Fleck, J. A.; Eagles-Smith, C. A.; Marvin-DiPasquale, M.; Windham-Myers, L.; Herzog, M. P. & McQuillen, H. L. Wetland Management Strategy to Reduce Mercury in Water and Bioaccumulation in Fish Environmental Toxicology and Chemistry, 2019, 38, 2178-2196.
- Brasso, R.; Rittenhouse, K. A. & Winder, V. L. Do songbirds in wetlands show higher mercury bioaccumulation relative to conspecifics in non-wetland habitats? Ecotoxicology (London, England), 2020, 29, 1183-1194.

Author Response

Reviewer #2:

This article is straightforward, interesting, and reads well. The authors collected fishes (yellow perches and gobies) from eight wetlands and seven non-wetland habitats along the Upper St. Lawrence River and analyzed their mercury contents. They found that mercury levels are higher in fishes from wetlands than in fishes from non-wetland habitats. They conclude that wetlands may mobilize mercury into the food chain more so than non-wetland habitats.

1. However they do not offer a possible explanation for this in the discussion. In fact, the explanation for this phenomenon is given in the introduction from lines 61 to 73. I suggest to add a sentence or two in the discussion restating what has been said in the introduction regarding methylmercury production and bioaccumulation in wetlands. Also, a recent study by Brasso et al. (2020) has yielded mixed evidence that wetlands animals (songbirds in their study) are at greater risk of mercury bioaccumulation than in non-wetlands. It might be good to add a reference to this article to moderate the conclusions.

We agree this suggested addition improves the discussion and have added text accordingly to the first paragraph of the discussion, specifically lines 195-205

2. Finally, I'd suggest adding a hint of two regarding what solutions could be implemented to solve this problem. For example, a paper by Ackerman (2020) reports on an experiment involving the construction of open‐ and deep‐water treatment cells at the downstream end of seasonal wetlands to promote naturally occurring MeHg removal processes.

We believe we address some solutions (fish monitoring for consumption advisories, continued research on mobilization areas and areas with water level fluctuations) in the final paragraph. And these have been elaborated on a bit more due to comments from Reviewer 1. As for solutions like that by in the study by Ackerman et al. (2020), they may not be feasible in this system for several reasons. First, the areal extent of these wetlands would be difficult to manage from a remediation standpoint, and there are many wetlands of such size on both sides of the river, adding an additional layer of complication. Implementation of any devices to trap Hg would be challenging and likely unacceptable given that these wetlands contain low levels of Hg arising from non-point sources, as opposed to a wetlands near a point Hg source of contamination.

Minor comments:
- l19: "Mercury levels were significantly (p<0.01) higher ..." I believe the authors refer to their ANCOVA results here? If that's the case shouldn't it be p<0.001?

Corrected in text.

- l137: "Sampled analyzed in wet weight." Please add a verb to the sentence.

Corrected in text.

- line 143: A t-test was used to determine if there was a difference between the two groups in length. Please note that if the normality tests fails, it is usually better to use a non-parametric test instead of the t-test.

Adjusted the text to note that we have conducted a Welch two-sample t-test, which is more robust, in contrast to a Student’s t-test, for example.

References:
- Ackerman, J. T.; Fleck, J. A.; Eagles-Smith, C. A.; Marvin-DiPasquale, M.; Windham-Myers, L.; Herzog, M. P. & McQuillen, H. L. Wetland Management Strategy to Reduce Mercury in Water and Bioaccumulation in Fish Environmental Toxicology and Chemistry, 2019, 38, 2178-2196.
- Brasso, R.; Rittenhouse, K. A. & Winder, V. L. Do songbirds in wetlands show higher mercury bioaccumulation relative to conspecifics in non-wetland habitats? Ecotoxicology (London, England), 2020, 29, 1183-1194.