Assessing the Risk of Internal Loading of Phosphorus from Drinking Reservoir Sediments

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

In this study , the sediments from various reservoirs were collected and P forms and P sorption capacity were analyzed. A internal P loading risk assessment was made based on the obtained results and some proposal were given. Only minor revision was required.

1.Why org-P was not inclued in this study?

2. sediment P release was not carried out in this study and which is the direct evidence of possible water quality changes cause.

3. What is the best management for the reservoir internal P control? 

Comments on the Quality of English Language

The sentence should concise in this paper.

Author Response

In this study, the sediments from various reservoirs were collected and P forms and P sorption capacity were analyzed. A internal P loading risk assessment was made based on the obtained results and some proposal were given. Only minor revision was required.

Comment 1: Why org-P was not included in this study?

Response 1: The standard fractions important for monitoring internal P loading in reservoirs are those that are susceptible to redox and pH changes, hence Fe-P and Ca-P. Although Org-P is important as a source in the water column, as an internal nutrient supply it does not significantly contribute to the legacy storage of P in reservoir sediments. Hence Org-P is not one of the major storage pools of P, the main pools are instead those included within the study. We therefore did not focus on Org-P within this manuscript.

 

Comment 2: Sediment P release was not carried out in this study and which is the direct evidence of possible water quality changes cause.

Response 2: Within the manuscript we investigated Equilibrium Phosphorus Capacity (EPC), which is an indicator of P release (i.e. a higher EPC value suggests a higher risk of internal loading). Hence the focus of the study was on the storage of P and the potential for internal loading as determined by the value of EPC. These are accepted determinants for determining internal loading of the paper and hence what can be determined from studying sediments from a water company perspective. Direct measurements of sediment P release requires in situ flux measurements which are highly costly and problematic to run and hence of little relevance to a water company. This was therefore well beyond the scope of the paper.

 

Comment 3: What is the best management for the reservoir internal P control? 

Response 3:  Thank you for your comment. We have reviewed management options for controlling internal loading of P in lines 312 -383 of the discussion.  However, we have additionally included a summary of these management options in the conclusions (lines 393-400), which reads as follows:

The methods outlined here would complement, and inform on the successes or failures of, potential approaches for managing internal loading of P in reservoir sediments, which may include i) sediment dredging, ii) biomanipulation of fish stocks and establishment of macrophytes to facilitate recovery through nature-based solutions (i.e. bottom-up nutrient control), and iii) prevention of stratification of deeper sites through engineered mixing or aeration, to facilitate a higher redox potential in the benthic region and decrease anoxia-driven sediment release of nutrients.

Reviewer 2 Report

Comments and Suggestions for Authors

The effects of legacy phosphorus stored in aquatic systems remains an important question on reservoir management;  Watson et al. nicely set up the question of internal loading of P from reservoir sediments.  The general approach utilizes examining concentrations of P, including Ca- and Fe-bound fractions, an approach that has been widely used in European and North American studies.  The determination of EPC and isotherms is difficult and these data look really good.  It is interesting that there appears to be excess potential to bind more inorganic P to Fe, suggesting that Fe is a critical buffer to supplying sediment P to the water column, pending on redox status.

The biggest concern regarding this study is the indefinite depth of the sediment used in the study, it is characterized as ~1 kg of sediment.  Presumably, this parcel of sediment includes the upper few mm of truly aerobic sediment, a zone in which Fe, and possibly Mn, are electron acceptors, sulphate reduction depending on water sulphate concentrations, and methanogenesis.  Given the concentrations of Fe in these samples, it is highly likely that some of the sediment has high dissolved Fe(II) and possibly iron sulfide species in the sediment.  The oxidation of Fe(II) to a high surface area iron oxide would increase P adsorption, the oxidation of FeS and subsequent pH decrease is likely less of an issue.  The authors cite Carey and Rydin, a paper in which the vertical distribution of P can be used to identify potential P releases in lakes of varying trophic status.

Overall, the paper is well-written and the analytical procedures appropriate.  In addition, the emphasis is on utilizing straight-forward approaches for very applied questions.  The authors should address the broader meaning of their observations relative to the homogenization of sediments undergoing a range of diagenetic processes.  I don’t know that their analysis is wrong; the data seem to make sense regarding processes in these reservoirs.  However, I could not argue that the conclusions are correct without further consideration of the indefinite sediment depth issue by the authors. 

Author Response

The effects of legacy phosphorus stored in aquatic systems remains an important question on reservoir management; Watson et al. nicely set up the question of internal loading of P from reservoir sediments.  The general approach utilizes examining concentrations of P, including Ca- and Fe-bound fractions, an approach that has been widely used in European and North American studies.  The determination of EPC and isotherms is difficult and these data look really good.  It is interesting that there appears to be excess potential to bind more inorganic P to Fe, suggesting that Fe is a critical buffer to supplying sediment P to the water column, pending on redox status.

The biggest concern regarding this study is the indefinite depth of the sediment used in the study, it is characterized as ~1 kg of sediment.  Presumably, this parcel of sediment includes the upper few mm of truly aerobic sediment, a zone in which Fe, and possibly Mn, are electron acceptors, sulphate reduction depending on water sulphate concentrations, and methanogenesis.  Given the concentrations of Fe in these samples, it is highly likely that some of the sediment has high dissolved Fe(II) and possibly iron sulfide species in the sediment.  The oxidation of Fe(II) to a high surface area iron oxide would increase P adsorption, the oxidation of FeS and subsequent pH decrease is likely less of an issue.  The authors cite Carey and Rydin, a paper in which the vertical distribution of P can be used to identify potential P releases in lakes of varying trophic status.

Overall, the paper is well-written and the analytical procedures appropriate.  In addition, the emphasis is on utilizing straight-forward approaches for very applied questions.  The authors should address the broader meaning of their observations relative to the homogenization of sediments undergoing a range of diagenetic processes.  I don’t know that their analysis is wrong; the data seem to make sense regarding processes in these reservoirs.  However, I could not argue that the conclusions are correct without further consideration of the indefinite sediment depth issue by the authors.

Response: Thank you for your comment. Upon revision of our methods, we can see that this is unclear and appreciate the confusion here. To clarify, approximately 1kg of sediment was taken for transportation purposes - we then took samples for analysis from top 3cm of that mass. To be more precise, 60 cm³ of sediment (with an internal core diameter of 9cm, taken at a 3cm depth) was collected per site.

We use a 3cm depth as the top 3cm of sediment is the homogenizsed depth which is largely the aerobic layer. Within this manuscript, we’re interested in the depth at which you see a diffusion gradient across the water-sediment interface (which drives the release of P). The standard depth at which this is usually studied is 3cm (Perkins et al., 2000). Additionally, and importantly, the top 3cm of the sediment is what’s relevant for water industry management purposes. Oxygen only penetrates over mere millimetres, with nothing significant for redox occurring beyond that (Bryant et al., 2011). Ideally, we would have taken a 15cm sample and analysed in 3cm increments but this wasn’t possible due to sampling restrictions at the sites.

We have now amended this accordingly within the methods section (lines 105 – 107), which now reads as follows:

Between April 2023 and February 2024, 60 cm³ sediment cores (with an internal core diameter of 9cm, taken at a 3cm depth) were collected from eight drinking water supply reservoirs, belonging to four British water companies, using a UWITEC USC 06000 gravity corer.

Reviewer 3 Report

Comments and Suggestions for Authors

In the paper were investigated the sediment and water column P content of eight British reservoirs, by analyzing iron-bound (Fe-P), calcium-bound (Ca-P) and labile P fractions. The aim of this study was to determine the extent of P storage from historical legacy inputs within a range of water-supply reservoirs. Results highlight the importance of studying reservoir sediments to evidence reservoir management. The experimental approach focused on determining the extent of P storage using simple sampling and laboratory techniques and demonstrating how relatively limited sampling can still be informative for management. Findings showed Ca-P to be the most abundant form of sediment P; however, all sites had a significant amount of P bound as Fe-P. In comparison, labile P was at a much lower sediment content, possibly due to release into the water column along diffusional gradients across the sediment-water interface. Findings indicated potential for pH, temperature and redox-mediated processes to induce internal loading. Furthermore, comparing EPC with water-column OP concentrations, it would appear that these reservoirs would all have significant internal P loading. This is highly important when considering lake and reservoir management approaches to reduce the effects of biological water quality risk, e.g., HABs and T&O risk. These findings indicate that reducing external loading of P through catchment management may be ineffective as this will not address internal P loading from sediment.

This valuable work. It could be published after minor revision as below.

 

General Remarks

 

Abstract  

Please do not use an abbreviation in the abstract or provide the full name.

 

Detailed remarks

 

1. Line 107 - full name of UWITEC should be added
2. Table 1 – full names of 4 companies and 8 reservoirs should be added
3. Figure 1,2,3 should be self-explaining. Please add proper information.
4. Table1,2,3 should be self-explaining. Please add proper information.

Author Response

In the paper were investigated the sediment and water column P content of eight British reservoirs, by analyzing iron-bound (Fe-P), calcium-bound (Ca-P) and labile P fractions. The aim of this study was to determine the extent of P storage from historical legacy inputs within a range of water-supply reservoirs. Results highlight the importance of studying reservoir sediments to evidence reservoir management. The experimental approach focused on determining the extent of P storage using simple sampling and laboratory techniques and demonstrating how relatively limited sampling can still be informative for management. Findings showed Ca-P to be the most abundant form of sediment P; however, all sites had a significant amount of P bound as Fe-P. In comparison, labile P was at a much lower sediment content, possibly due to release into the water column along diffusional gradients across the sediment-water interface. Findings indicated potential for pH, temperature and redox-mediated processes to induce internal loading. Furthermore, comparing EPC with water-column OP concentrations, it would appear that these reservoirs would all have significant internal P loading. This is highly important when considering lake and reservoir management approaches to reduce the effects of biological water quality risk, e.g., HABs and T&O risk. These findings indicate that reducing external loading of P through catchment management may be ineffective as this will not address internal P loading from sediment.

This valuable work. It could be published after minor revision as below.

 

General Remarks

 Abstract  

Comment 1: Please do not use an abbreviation in the abstract or provide the full name.

Response 1:  As far as we can see, all abbreviations in the abstract have been defined in full first. If we have indeed missed one, we will amend during the proofing stage if highlighted by the editorial team.

 

Detailed remarks

Comment 2: Line 107 - full name of UWITEC should be added

Response 2: This has now been amended in the methods.

 

Comment 3: Table 1 – full names of 4 companies and 8 reservoirs should be added

Response 3: Unfortunately, we are unable to provide the full names of the water companies involved with this research due to a non-disclosure clause. We are only able to provide data in an anonymised format.

 

Comment 4: Figure 1,2,3 should be self-explaining. Please add proper information.

Response 4: This has now been amended for all.

 

Comment 5: Table1,2,3 should be self-explaining. Please add proper information.

Response 5: This has now been amended for all.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for the clarification.  The 3 cm depth of collection is indeed typical of the orthophosphate pore water gradient in many systems.  This zone likely mostly anaerobic, but may well have the major complement of Fe(III) persisting within it.