Interactive comment on “ Stream restoration and sanitary infrastructure alter sources and fluxes of water , carbon , and nutrients in urban watersheds ”

Journal: HESS Title: Stream restoration and sanitary infrastructure alter sources and fluxes of water, carbon, and nutrients in urban watersheds Author(s): M.J. Pennino et al. MS No.: hess-2015-444 MS Type: Research article To the Editor of HESS: We would like to thank the editor and reviewers for all of their comments and suggestion. We have compiled the editor’s and both of the reviewer’s comments below and we have thoroughly addressed and responded to each comment. Please note that all

We would like to thank the editor and reviewers for all of their comments and suggestion. We have compiled the editor's and both of the reviewer's comments below and we have thoroughly addressed and responded to each comment. Please note that all C7072 HESSD 12, C7072-C7085, 2016 Interactive Comment Full Screen / Esc

Printer-friendly Version
Interactive Discussion Discussion Paper of our responses are in red.

Response to Referee 2 Comments
The topic of how urban streams transfer nutrients is of interest. The experimental setup is quite confusing though, it is not obvious why and what you compare. As suggested by reviewer 1, we have now changed some of the confusing terminology and simplified the abstract, introduction, and the methods. We believe we have made the experimental set up more clear and easy to understand. For example, 1) we are now primarily using the terms restored or urban degraded streams, 2) we clarified the introduction were we list our objectives and we make our second objective more clear (as described above) by stating that we are looking at hydrology and nutrient sources in 1 restored and 3 urban degraded streams "to assess the role of stream restoration and potential pollutant sources, such as leaky sanitary sewers," 3) we simplified our site description section (described more below), and 4) we simplified the methods section on "Comparison of Pre and Post Restoration Hydrologic Response" by putting most of the text in to Supplementary Material. Some terminology is not very specific and confusing: unrestored, sanitary infrastructure etc. How are you assessing the impact of the 'sanitary infrastructure'? We are no longer using the term "unrestored", instead we are using "urban degraded streams." We are also no longer using "sanitary infrastructure" but instead "sewers" or "sewer infrastructure." To assess the impact of sewer infrastructure (also described elsewhere) we used data on carbon quality (fluorescence spectroscopy results), data on 15N-nitrate stable isotopes, data on fluoride and iodide concentrations, and carbon and nitrogen export results.
Do you have enough study sites to derive meaningful and statistically significant con-C7073 HESSD 12, C7072-C7085, 2016 Interactive Comment Full Screen / Esc

Printer-friendly Version
Interactive Discussion Discussion Paper clusions? Similar to our response to the first reviewer, we can say based on our statistical analysis of time series data collected at each of the four sites whether one stream has different hydrologic metrics, nutrient sources, or exports than the other sites. However, because we do not have enough replication of study sites, we cannot say with statistical confidence whether stream restoration or management has a significant effect (except for the before and after hydrologic analysis for the restored stream site). Yet, based on the results of this study we can still suggest the potential influence of leaky sewer infrastructure and stream restoration. Our results also provide new information regarding sources and exports of water, carbon, and nitrogen in urban restored streams, and there are relatively few papers analyzing sources of water, carbon, and nitrogen in urban restored and degraded streams.
In my opinion, lots of the comparison between streams/catchments in terms of the 'sanitary infrastructure' is speculative and does not support your conclusions. We feel that our revisions and the literature added support our conclusions. There has been considerable background work evaluating the importance of leaky sewers in these watersheds. Please see our more detailed and specific responses below.
Below are some more specific comments: Title -should be fluxes of water and nutrients We do not agree with this change because carbon is not typically considered a nutrient and we want to make sure it is clear that carbon is studied in this paper.
Abstract and introduction -the description of what streams you measure is very confusing, do you measure only urban streams? Restored and unrestored? Please rewrite to make it clear what you compare with what. The abstract and introduction have been revised to make it clear that we are measuring only urban streams. We removed the use of the term unrestored and used the term "urban degraded streams" instead.
Methods, page 13154, line 9 -please rewrite 'the entire mainstem of the stream from headwaters to mouth is greater than 95% restored' We made this part a new sentence and changed it to read "Also, about 95% of Minebank Run's mainstem has been restored. . ." Methods, page 13154, most of this information should be in a table This information is in Table 1. We have removed some of the text from this paragraph on page 13154 and summarized some of the data for the four sites.
Methods, page 13155, please consider putting some of this text in supplementary material Most of these methods from this section "Comparison of pre and post restoration hydrologic response" were moved to supplementary material and some of the text was re-written for clarity.
Methods, page 13157 -how long were the NO3N samples stored for before analysis?
In the supplementary material we wrote "All nitrogen species, except samples for stable isotope analysis, were preserved by acidifying to pH 2 with sulfuric acid and stored C7075 Interactive Comment Full Screen / Esc

Printer-friendly Version
Interactive Discussion Discussion Paper frozen in HDPE bottles until analysis." And "All other samples, besides NO3-isotopes samples, were analyzed within 1-2 months." For nitrate isotopes we added this text to the supplementary material: "Samples for nitrate isotope analysis were all analyzed on the same date, resulting in the samples being stored frozen from 7 months to 2 years and 7 months." We could not find any literature to suggest there is any difference in isotope results depending on storage time.
Methods section is far too long, please make it more concise. We have considerably cut down the site description section and the section "Comparison of pre and post restoration hydrologic response." We feel that the rest of the methods section material is necessary.
Methods -I am not sure how useful are your load estimates if the approach does not sample highflows? Even though the annual loads may not fully estimate stormflow contribution, we added text to say ". . .because all four sites are within the same city and receive relatively the same rainfall during storm events, the relative annual loads estimated for the sites are comparable and it is appropriate to draw conclusions among the four study sites." Discussion: A lots of discussion is speculative. Do you have any evidence in support of your hypotheses about leaky sewers, erosion of stream channel etc? There are several sources of evidence to suggest that there are leaky sewers at the streams in this study, which were stated in the results and discussion (but we have now added any new references and added more discussion to the text): 1) The high 15N-NO3 isotope levels and the nitrate isotope mixing model results suggests N wastewater sources at all four sites (Kaushal et al., 2011, Divers et al. 2014. Also during summer baseflow, the 15N-NO3-isotope levels were consistently high along the entire stream length at all four sites, also suggesting the influence of leaky sewers inputs through groundwater recharge (Divers et al. 2014, Hall et al. 2016 2) The fluorescence spectroscopy results indicates there is more labile organic matter and protein-like organic matter in the urban degraded streams as well as the restored stream, suggesting wastewater sources (leaky sewers, since there are no point wastewater discharges) based on the literature (Baker, 2001;Goldman et al., 2012;Li et al., 2015;Yu et al., 2015).

4)
We also added more introductory text on how leaky sewers are a global problem with references from the United States, Europe and Asia: "These techniques and others have been used globally to detect the influence of leaky sewer infrastructure on water quality (Ekklesia et al., 2015;Hall et al., 2016;Risch et al., 2015;Tran et al., 2014;Wolf et al., 2012) and it has been shown that sewer leaks have impacts during baseflow and stormflow (Divers et al., 2013;Divers et al., 2014;Phillips and Chalmers, 2009;Rose, 2007)." 5) The increase in wastewater inputs from wastewater sources (which include C, N, and P) during stormflow (pulsed behavior) has been shown to be related to leaky pipes and sewer overflows (Divers et al., 2014;Phillips and Chalmers, 2009;Kaushal et al., 2011). 6) We have added text saying that "In fact, Baltimore City has detailed records for the dates and locations of sewer overflows through their open data website (data.baltimorecity.gov) and these sewer overflows have occurred within the watersheds in this study." In terms of evidence for erosion of the stream channel, we indicated that this was observed by the authors in the test by stating "(personal observation)." The hydrological metrics and their interpretation is convincing but not the biogechemical part of the study. Perhaps you should tease out more the differences in nutrients. At the moment this aspect is not clear. We added to the discussion section on nutrient sources and exports further interpretations for why the restored stream sometimes behaved differently than the other streams.
1) Specifically, we added details to the discussion on nitrate sources to show the influence of leaky sewers, seasons, and differences in sites: "High 15N-NO3-isotope levels are indicative of nitrate from wastewater sources (Divers et al., 2014;Kaushal et al., 2011). . .. Due to stream restoration at MBR, the neighboring sewer pipes were repaired and stabilized Mayer et al., 2010;US EPA, 2009), likely resulting in less sewer leaks at Minbank Run in and along the restored reach. During summer baseflow, the 15N-NO3-isotope levels were consistently high along each stream length suggesting the influence of leaky sewers inputs through groundwater recharge (Divers et al., 2014;Hall et al., 2016), but during the rainier spring season, C7080 the more urban streams (DRN and PMR) showed a decline in 15N-NO3-isotope levels indicating possible dilution of sewer sourced nitrate from rainwater entering from connected impervious surfaces (Divers et al., 2014). This dilution of wastewater NO3-was not observed at the other sites, potentially due to less connected impervious surfaces at the least urban watershed (RRN) and the reduction of peak discharge due to the reconnected floodplain for the restored stream (MBR) (Boyer and Kieser, 2012;Cendon et al., 2010;Poff et al., 2006).
2) In the section on C sources we provide further references on why the fluorescence results suggest C sources from wastewater: "From studies throughout the globe, it is know that protein-like and more bioavailable or labile organic matter is typically associated with wastewater carbon sources (Baker, 2001;Goldman et al., 2012;Li et al., 2015;Yu et al., 2015)." "As a result, the higher BIX, P/H ratio, and protein-like organic matter in the restored stream MBR, as well as the more urban watersheds (PMR and DRN), is likely due to leaky sewers typically found in older urban watersheds (Hudson et al., 2008;Kaushal et al., 2011) since the watersheds in this study are not influenced by combined sewer overflows or typical point source discharges of wastewater." 3) In the section on C exports we provided further interpretation on why the restored stream had the lowest C exports. "The restored stream also likely had lower C exports due to increased ability to retain and process carbon in transient storage zones, such as pools, in the reconnected floodplain or through hyperheic exchange (Bukaveckas, 2007;Groffman et al., 2005;Mulholland et al., 1997;Pennino et al., 2014), whereas degraded urban streams that are highly eroded can have less transient storage areas to potentially store and process carbon (Kurth et al., 2015;Sudduth et al., 2011a)." 4) We added more interpretation for why the N and P exports were lower for the restored stream. "The higher TN exports in the more urban sites (PMR and DRN) compared to the restored stream (MBR) may be due to various reasons, such as greater N inputs from leaky sewers in the more urban and older watersheds and/or greater N removal through denitrification in the restored stream due its hydrologically connected  (Kaushal et al., 2008), and alluvial wetlands, and greater hyporheic exchange (Bukaveckas, 2007;Harrison et al., 2011;Kaushal et al., 2008;Roley et al., 2012). In fact, the stream restoration at MBR involved some repairs to help stabilize the sewer pipes Mayer et al., 2010;US EPA, 2009) and consequently may have reduced sewer leaks, but detailed research is needed to evaluate the effects of sewer repairs on watershed N inputs." For Phosphorus we added "The lower TP exports in the restored stream may be due to increased hyporeic exchange and floodplain connection, which have been shown to increase P retention (Butturini and Sabater, 1999;Mulholland et al., 1997)." 5) We also added to the interpretation on why the restored stream had less pulsed C and nutrient exports and less exports during higher flows. "The lower proportion of N exports during higher flows for the restored stream (MBR) may be due to the connected floodplain attenuating higher flows, as evidenced by the effective discharge results described above and due to less connected impervious cover (Poff et al., 2006;Smith et al., 2013)." Later in the discussion we added: "The higher C, N, and P exports during baseflow at the restored stream (MBR) and the least urban stream likely corresponds with there being greater groundwater recharge at these sites, due to less impervious surface cover and floodplain reconnection (Boyer and Kieser, 2012;Cendon et al., 2010)." To the following sentence we added new citations: "Dissolved C, N, P, F-, and I-exports in the more urban watersheds could have also been more variable due to runoff from impervious surfaces and/or increased contributions from storm drains (Bernhardt et al., 2008;Hatt et al., 2004) and elsewhere in the stream corridor (i.e. sewage leaks) during storms, as shown in other studies (Divers et al., 2014;Kaushal et al., 2011;Phillips and Chalmers, 2009)." We also added this sentence: "The attenuation of peak discharge due to stream restoration observed at MBR, which reconnected the stream with the floodplain is likely Finally, the paper needs language revision to remove not very scientific expressions as those highlighted above. As mentioned above, we have now revised the language to make the paper more clear.