the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 02 Dec 2024
Physicochemical and urban land-use characteristics associated with resistance to precipitation in estuaries vary across scales
Abstract. Estuaries are subject to frequent stressors, including elevated nutrient loading and extreme hydrologic events, which impact water quality and disrupt ecosystem stability and health. The capacity of an estuary to resist changes in function in response to precipitation events is a key component of maintaining estuarine health in our changing climate. However, generalizable patterns in factors related to estuarine responses to extreme precipitation remain unknown. We investigate physicochemical factors and land-use characteristics that are associated with ecological resistance to precipitation – broadly defined as the magnitude of ecosystem change induced by an event – in five disparate estuaries distributed across the continental United States. Using long-term meteorological and water quality data from the National Estuarine Research Reserve System along with land use/land cover and population data, we examine relationships between the resistance index – a proxy for ecosystem stability calculated using dissolved oxygen – and physicochemical and urban land use characteristics on local-to-continental scales. Contrary to our initial hypothesis, we found that more urbanized estuaries were more resistant to precipitation events, and that water temperature, water column depth, turbidity, nitrogen, and chlorophyll-a were related to resistance on a continental scale. However, these trends interacted with estuarine salinity and varied across individual estuaries; where we found additional relationships of resistance with salinity, turbidity, phosphate concentrations, N:P, and tree cover. Considering emerging stressors from new climatic scenarios and from urbanization, these results are important for representing the impacts of disturbances in large-scale models and for informing management decisions regarding estuarine water quality.
Status: open (extended)
-
RC1: 'Comment on egusphere-2024-3338', Anonymous Referee #1, 20 Feb 2025
reply
General comments:
This paper looks at how major precipitation events influence dissolved oxygen changes—an integrative measure of estuarine resistance to disturbance—in the NERRS estuaries across the US. There are not many comparative estuarine studies that leverage the network of NERRS data, so this paper fills an important research gap in that respect. The synthesis of the input dataset is a feat in and of itself. Additional strengths of the work are the graphical synthesis of complex results (Figure 5) and a detailed discussion.
However, the paper also has major weaknesses that deserve further attention in a revision. The biggest weakness is the attempt to infer drivers of resistance across a very small number of estuaries through bivariate correlations. As the authors acknowledge in lines 395-397, resistance is associated with myriad specific factors, and these factors interact. Potential multicollinearity within the dataset between percent urbanization and other factors such as watershed contributing area, flushing time, tidal influence, etc. makes it challenging to infer how urbanization influences resistance on the basis of a bivariate correlation. Ideally, one would address collinearities in potential predictors and construct a generalized linear model to account for covariates and isolate contributions of factors like urbanization to resistance in order to draw inferences. However, with this small dataset, this is probably not an option.
I instead recommend a few options to address this issue:
- Quantitative approach: Fully characterize the set of potential drivers of resistance hypothesized to be important and how they vary across estuaries (and potentially across sites, being careful to avoid pseudoreplication—inclusion of sites as multiple datapoints when they all correspond to one independent variable like percent urbanization). This should likely include variables related to estuarine circulation, like flushing time and stratification indices, and potentially watershed area. Then conduct a principle components analysis followed by a Varimax rotation to identify how the predictor variables group together and how the estuaries vary with respect to each other and the predictor variables. The Varimax rotation makes each factor more interpretable. Then, plot the factors from the Varimax rotation against resistance, or consider how resistance differs among different groups that emerge from the factor analysis.
- Qualitative approach: Alternatively, eliminate the cross-estuarine correlations from the analysis and take a case-study approach to elucidate how storm events impact resistance in different types of estuaries. For this approach to be successful, the case-study discussions should be tied to hypothesized detailed mechanisms of how storm events would impact resistance in different types of estuaries.
Regardless of which option is chosen, I recommend moving some of the excellent literature review in the Discussion to the Introduction and synthesizing it to present a coherent conceptual model of the different ways in which physicochemical variables and storm events are expected to interactively influence resistance. A conceptual figure would be nice as well. Without this theoretical underpinning to orient readers at the outset, the detailed and understandably messy dataset and results are challenging for readers to wade through.
Another big-picture factor to consider further is that some of the estuaries may be less impacted by precipitation events than by flow events controlled by upstream reservoirs. One factor that should probably be discussed is whether some estuaries may appear more resistant to rain events because upstream reservoirs act as flood-detention buffers. In general, a table and/or text providing a more detailed characterization of the estuaries (beyond land-use) is a needed addition. If a PCA is attempted as suggested in #1 above, it would be ideal to include variables that characterize hydrologic differences among the estuaries or across storms, such as a normalized increase in inflow (if available) and statistics about storm durations, which could be another driver of cross-estuary differences in resistance.
Specific comments:
The abstract should more specifically state the big-picture implications of these results. What matters about them and why?
Line 133-136: Is the 10-km lengthscale, then, the spatial scale expected to drain to the point of interest over the sensitive timescale of days? How the two parts of this sentence link is not entirely clear. If runoff generation length scales are smaller, a larger proximity zone would presumably add noise to statistical relationships.
Line 176: Should mention atmospheric rivers for CA
Line 220-221: Was the tidal cycle accounted for in selecting the window of time for averaging? Averaging over partial tidal cycles may skew results in a way that could produce spurious differences among sites in the same estuary or across estuaries.
Line 231-232: Provide a table of predictor variables
Line 479: N loading and low DO are not always associated. San Francisco Bay is one place where they are not, so this estuary is not the best one to highlight earlier in the paragraph.
Line 532-534: Please rewrite this paragraph with a more coherent narrative. The topical sentence, focused on variation within single estuaries, is not supported by the following sentences, focused on a comparison between two estuaries (CBM and GTM). The topical sentence also focuses on relationships between physicochemical factors and resistance, whereas the supporting sentences focus on the response of physicochemical factors to precipitation.
Technical comments:
Line 87: Reverse the order of the clauses in the previous sentence, since this sentence references the later study, which is in the former part of the sentence (could be confusing!).
Line 105: Add “of” after “understanding.”
Line 205: Replace “infer” with “be related to”
Line 241: Alway --> always
Citation: https://doi.org/10.5194/egusphere-2024-3338-RC1
Viewed
Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 65 | 0 | 0 | 65 | 0 | 0 |
- HTML: 65
- PDF: 0
- XML: 0
- Total: 65
- BibTeX: 0
- EndNote: 0
Viewed (geographical distribution)
Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1