Next Article in Journal
A New Machine Learning Algorithm to Simulate the Outlet Flow in a Reservoir, Based on a Water Balance Model
Previous Article in Journal
Annual and Seasonal Trends in Density, Biomass, and Secondary Production of Chironomid Midges (Diptera: Chironomidae) in the Middle Reaches of the Shinano River, Japan, 2015–2019
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Limnological Review
Volume 25
Issue 3
10.3390/limnolrev25030028
limnolrev-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Total Phosphorus Loadings and Corrective Actions Needed to Restore Water Quality in a Eutrophic Urban Lake in Minnesota, USA: A Case Study

by
Neal D. Mundahl
1,* and
John Howard
2
1
Ecology and Environmental Science Program, Department of Biology, Winona State University, Winona, MN 55987, USA
2
Natural Resources and Sustainability, Engineering Department, City of Winona, Winona, MN 55987, USA
*
Author to whom correspondence should be addressed.
Limnol. Rev. 2025, 25(3), 28; https://doi.org/10.3390/limnolrev25030028
Submission received: 18 April 2025 / Revised: 15 June 2025 / Accepted: 23 June 2025 / Published: 1 July 2025
Browse Figures
Versions Notes

Abstract

Lake Winona, a 129 ha eutrophic urban lake comprised of two interconnected basins, exceeds state water quality standards for total phosphorus. Historical lake nutrient data and traditional watershed modeling for the lake’s two basins highlighted multiple major pathways (e.g., municipal stormwater discharges, watershed runoff, internal loading, and wetland discharges) for total phosphorus (P) loading, with >900 kg P/year estimated entering the water columns of each basin. Updated data sources and newer watershed modeling resulted in significantly different (both higher and lower) P loading estimates for the various P sources, especially watershed runoff and internal loading. Overall, basin-specific loading estimates using the updated model were significantly lower (28–40%) than previous estimates: 680 and 546 kg P/year mobilized in the western and eastern basins, respectively. To achieve state water quality standards (<60 ppm P for the western basin, <40 ppm for the eastern basin), watershed and internal P loading each would need to be reduced by approximately 120 kg P/year across the two basins. Reductions could be achieved by a combination of alum treatments to reduce internal loading, removal of common carp (Cyprinus carpio) to prevent interference with alum treatments and nutrient releases via excretion and defecation, and six engineered structures to intercept P before it enters the lake. The different P reduction projects would cost USD 119 to 7920/kg P removed, totaling USD 5.2 million, or USD 40,310/hectare of lake surface area.

1. Introduction

With the passage of the Clean Water Act in the USA in 1972, water quality standards were established for all surface waters and permits were required for any point sources discharging directly to those waters [1]. Various federal, state, and tribal regulatory agencies were tasked with monitoring surface water quality to assure that standards were being met. If any water body was found to be non-compliant with one or more standards, section 303(d) of the Clean Water Act authorized the Environmental Protection Agency (EPA) to work with its regulatory partners to list the impaired waters and develop Total Maximum Daily Loads (TMDLs) for those waters [2]. By establishing the maximum amount of a pollutant that can enter a body of water from various sources and still allow that body to meet water quality standards, the TMDL was intended to serve as a planning tool to set in motion a variety of activities to restore water quality [2].
Throughout the world, freshwater lakes, especially those in urban and agricultural areas, have experienced the adverse effects of excessive nutrient pollution [3]. Nutrients such as phosphorus (P) and nitrogen (N) stimulate algal productivity, resulting in algal blooms and accelerated (cultural) eutrophication when nutrient inputs and/or internal recycling from sediments are high and sustained. Early efforts to understand the cause of lake eutrophication focused on inputs of N, carbon (C), and P, singly and in combination [3,4,5], but P appears to be the major driver of eutrophication based on whole-lake assessments rather than small-scale mesocosm or bottle experiments [3,6]. P is generally the factor most limiting (i.e., least available due to low abundance) for plant productivity in lakes, especially those that are small, shallow, and oligotrophic to meso-eutrophic [7,8,9,10]. Consequently, mitigating or reversing the cultural eutrophication of lakes in the face of expanding agriculture, increasing urbanization, and an uncertain climate future likely will focus on managing P rather than N or C [3,6].
Management of P to mitigate lake eutrophication typically differs based on lake size, driven largely by management costs. For small lake basins (<a few hundred ha), P concentrations can be reduced significantly and economically through chemical precipitation of P, adding iron, alum, bentonite clay, or other materials to hold P in lake sediments and prevent P release back into the water column [3,11,12,13,14,15,16,17,18]. In some small lake systems, adding nitrate can have a similar effect on P by maintaining high redox conditions at the sediment–water boundary [3]. Such controls in small lakes can produce nearly immediate reductions in P concentrations and algal chlorophyll levels [3,12,13,15,16]. In-lake chemical controls of P are not economically feasible for large lakes, but reduced inputs of P (especially from point sources) have proven successful in reducing algal blooms in systems as large as the North American Great Lakes; however, response times may require 5 to 30 years [3].
Once a lake is listed as impaired due to excessive P, the TMDL process [2] identifies all possible sources of P (e.g., internal cycling, permitted direct discharges, tributary discharges, surface drainage runoff, and atmospheric inputs) and their current contributions to overall lake P loading. Subsequently, the various inputs are modeled to determine how the lake can best meet water quality standards through reduced P loadings from the different sources [2,19,20]. With P reduction goals in place, various P reduction projects or activities can then be explored to determine a feasible (i.e., logistic and economic) approach for restoring lake water quality [2,20].
Managing urban lakes to reduce P loadings can be especially problematic due to the volume of stormwater (with significant P loads from an urban landscape) collected from impervious surfaces within a city and subsequently discharged to those lakes via municipal stormwater systems [21,22,23,24]. Controlling P loading by reducing or preventing stormwater from entering municipal systems, using rain gardens or bioswales to increase infiltration, can be extremely costly on a scale large enough to be effective [24,25]. End-of-pipe engineered solutions for P reduction can be highly effective but also come with high costs [21,22,23,25,26,27]. In addition, both approaches require significant amounts of land in the proper locations to facilitate management structures, which often may not be available in many high-density urban areas [17].
We here present an example of an urban lake with a long history of human impacts [28] that ultimately led to its listing as impaired for excess P. With its hydrology altered by numerous municipal stormwater outfalls emptying into it and a local stream rerouted through it, nutrients from throughout its urban watershed entered the lake and caused it to become highly eutrophic [28]. We examine the original TMDL process for this system, describe a second modeling process with fewer assumptions and more current data, and critically assess the various options recommended for reducing P loadings from a variety of sources. Through this assessment, we hope to highlight and clarify the decision-making processes involved in restoring a lake’s water quality and reveal the types of concerns that may arise as various mitigation options are explored.

2. Study Site: Lake Winona

Lake Winona (Figure 1) is a 129 ha hypereutrophic, dimictic lake located within the City of Winona, Minnesota, USA (44°02′19.54″ N, 91°38′37.24″ W; 197 m above sea level). It occupies a portion (3.2 km long × 0.5 km wide) of a former side channel of the Mississippi River approximately 2 km from the present main channel of the river [28]. During the past 100+ years, the lake has been modified extensively by dredging, filling of fringe wetlands, creation of new shorelines, and separation into two interconnected basins (a 36 ha western basin and 93 ha eastern basin) by an elevated street causeway. A single, 3 m diameter culvert serves as the only connection between basins. The maximum and mean depths are 12.2 m and 4.2 m, respectively [29], with waters below 4 m typically becoming anoxic during late summer (July–September) thermal stratification [28]. The lake has a history of toxic cyanobacterial blooms (especially Aphanizomenon and Microcystis) during late summer and fall [28].
Since 1885, Gilmore Creek has been diverted into and through Lake Winona before emptying into the Mississippi River (Figure 1) [28]. The creek flows into the western lake basin via an artificial channel, County Ditch Number 3, and exits the eastern basin via another artificial channel, County Ditch Number 4 (Figure 1). Thirty-five municipal stormwater outfalls (13 on the western basin and 22 on the eastern basin, including a mix of underground drainage pipelines and open ditches) carry urban runoff waters into the lake during rain events and periods of snowmelt. In addition, runoff from >20 additional stormwater outfalls enters County Ditch Number 3 upstream of the lake and is carried into the western basin. Since 1944, Gilmore Creek also has flowed through Boller Lake prior to entering County Ditch Number 3. Boller Lake is a shallow, 40 ha lake/wetland complex that serves as a settling basin for the eroded soils carried by Gilmore Creek (Figure 1), preventing their delivery to Lake Winona [9].
The City of Winona operates its municipal stormwater system under permit from the Minnesota Pollution Control Agency (MPCA). The city has a Phase II (smaller communities) National Pollutant Discharge Elimination System (NPDES) municipal separate storm sewer system (MS4) general permit that allows it to convey stormwater runoff in the city directly into local surface waters (Lake Winona, Gilmore Creek/Ditch #3, and the Mississippi River) [30]. The permit requires the city to reduce the amount of pollution entering surface and groundwater from the storm sewers “to the maximum extent practicable”, and to develop a stormwater pollution prevention program utilizing best management practices [10].
Since 2014, Lake Winona has been on Minnesota’s impaired waters list due to excessive nutrient levels (i.e., high total P concentrations) [19]. The impairment listing set into motion a series of actions designed to (1) better understand the cause(s) of elevated P concentrations in the lake basins, (2) gather additional data to use in modeling current P loading conditions in both basins, and (3) determine what P sources could be addressed to achieve P load reductions in both basins sufficient to meet or exceed state water quality standards.

3. Methods

3.1. Historical Phosphorus Data

To better assess the lake conditions that led to the state listing for high total P, we sought out publicly available P data for Lake Winona (documented separately for the eastern and western basins). Data from the 1970s and 1980s were contained in a report describing the reclamation of Lake Winona [20], whereas data from the 1990s up to the present were stored on the Minnesota Pollution Control Agency’s website [31,32]. During the 1970s and 1980s, P data were available during ice-free and ice-cover time periods, whereas more recent data were all from the ice-free season. In addition, data during the ice-free seasons included samples collected from both above and below the thermocline (i.e., in oxygenated and oxygen-poor waters, respectively).

3.2. 2016 TMDL for Lake Winona

The MPCA completed a TMDL assessment for Lake Winona in 2016 to model the P problem and better understand potential options for remediation [19]. To simultaneously address region-wide (watershed scale and larger) water quality impairments and to more efficiently address impairment problems that spanned multiple governmental jurisdictions, the study of Lake Winona P impairments was combined with a study of other impaired (by nutrients, bacteria, and total suspended solids) surface waters (rivers and streams) in the southeastern Minnesota region (Mississippi River—Winona watershed), including 169,600 hectares spanning portions of three counties (Wabasha, Winona, and Olmsted) [19]. The Lake Winona portion of the study used standardized modeling software (BATHTUB Version 6.1) [33] paired with a Canfield–Bachmann P sedimentation model [34] to estimate total P loadings from known sources (tributaries, watershed runoff, regulated stormwater sources, atmospheric inputs, and internal recirculation) during a typical year [19]. Rather than using only Lake Winona-specific P data for this TMDL, the process utilized average values for the various loading sources (e.g., urban stormwater runoff and atmospheric inputs) derived from large region-wide or nationwide datasets [33,34]. Long-term average basin hydrology variables (e.g., average monthly precipitation, estimated runoff volumes, and average Gilmore Creek streamflow) were used in combination with estimated P concentrations to produce annual loading estimates (kg P/year) for each P source.
Models were constructed for each basin of Lake Winona separately. The total annual P load for each basin was determined by summing individual loads from each P source. The maximum allowable annual P load that a given basin could receive and not exceed the state impairment standard was determined and compared to the estimated existing load. The difference between the estimated existing load and the maximum allowable annual load was expressed as a percentage of the existing annual load, this value now becoming the % load reduction in P needed to achieve the state water quality standard goal. The total load reduction determined for each basin was allocated among P sources based on the likelihood for achievable reductions. For example, atmospheric inputs were not assigned a % reduction goal due to a lack of local control over such inputs, whereas internal loadings were assigned high % reduction goals due to potential success by using chemical (e.g., alum) control measures.

3.3. Updated Lake Winona P Modeling and Assessment 2018–2020

Soon after completion of the 2016 TMDL study, Winona County and the City of Winona retained two consulting firms to conduct an additional modeling and assessment study of P dynamics for Lake Winona using more current P loading and lake bathymetry data and newer modeling approaches. This reassessment of P loading in Lake Winona, driven by perceived shortcomings in the data and modeling used in the 2016 TMDL (e.g., limited contemporary site-specific data and assumptions and estimates based on region- or nation-wide data rather than local information), was completed in 2020 using updated Lake Winona watershed P data (water and sediment), flow rates, and bathymetry, and newer, more site-specific modeling approaches. This latter study also included specific recommendations for engineered solutions and other approaches to reduce P loading and bring lake phosphorus levels into compliance with state standards.
To generate more contemporary data for use in modeling P dynamics in the Lake Winona watershed and to reduce the use of assumptions based on region-wide or nationwide data, new studies of water quality, lake sediments, flow rates, and lake bathymetry were initiated during 2018. Water at nine sites (two in upper Gilmore Creek, two in Boller Lake, one in Ditch #3, four in Lake Winona) was sampled 10 to 25 times for total P, generally weekly or bi-weekly during March–December, but more frequently during rain events. In addition, two more deepwater sites in Lake Winona were sampled (n = 9–14) during May–November for total P at three locations within the water column: the surface, middle, and bottom. Bottom sediments were collected and sampled for total P from two locations in the western basin of Lake Winona and from four sites in the eastern basin to aid in estimating potential internal P loading. Water flow rates were also measured at one location in upper Gilmore Creek, one site in Ditch #3, and at the lake outlet to the Mississippi River. Finally, bathymetry was updated for both basins of Lake Winona and for Boller Lake to determine lake basin volumes for use in estimating P loading from inflows and internal P release from sediments.
Lake Winona hydrology and water quality modeling of stormwater runoff was carried out using an updated P8 Urban Catchment Model [33]. The model is designed to predict the generation and transport of stormwater runoff and associated pollutants in urban drainages. Specifically, modeling predicts movements of soil particles in runoff across land and impervious surfaces to estimate pollutant loads delivered to a particular water body [20]. Watershed modeling was conducted for various Lake Winona subwatersheds (Gilmore Creek, Boller Lake, and the western and eastern basins of Lake Winona) and calibrated with precipitation and water discharge data for 2018, and multi-site P concentration data collected during the period May–October 2018. Potential P contributions from residential septic systems within the Gilmore Creek watershed were estimated and compared to modeled loadings for Boller Lake [20]. For the P8 model for the eastern basin, construction, industrial, and municipal stormwater discharges were combined with general watershed drainage to create a single, combined direct watershed P source for each basin. This same combination appeared in the P8 model for the western basin, plus two additional sources not separately identified in the TMDL BATHTUB model: Boller Lake and watershed runoff to the Ditch Number 3 section of lower Gilmore Creek.

4. Results

4.1. Historical Phosphorus Data Leading to Lake Impairment Listing

Historical P data from the 1970s and 1980s [28] indicated very high (>150 ppb total P) levels for both basins during both summer and winter (Figure 2A). Based on measurements made in 1994 and later, surface water P levels have been reduced considerably since the 1980s, especially in the eastern basin (Figure 2B). However, bottom (>5 m deep) water P concentrations have increased in the eastern basin. Overall, mean western basin P levels still exceeded the North Central Hardwood Forest Minnesota Shallow Lakes standard of <60 ppb total P, whereas mean eastern basin P levels still exceeded the North Central Hardwood Forest Minnesota General standard of <40 ppb (Figure 2B) [20].

4.2. 2018 Phosphorus Data

Phosphorus data collected from the Lake Winona watershed during 2018 for updated P loading modeling [20] indicate that the total P concentrations remained high in Gilmore Creek, Boller Lake, Country Ditch #3, and Lake Winona. Average P concentrations at all sites exceeded state water quality standards for both basins of Lake Winona (Figure 3A). At the two deepwater sites within the lake (western basin site = 7 m deep, eastern basin site = 12 m deep), surface P concentrations were at or slightly higher than state P standards, but samples taken from the lake bottom and mid-water column had P levels two to five times higher than the surface values (Figure 3B).

4.3. 2016 Lake Winona BATHTUB Model TMDL

Modeling of P loading in Lake Winona estimated that 941 and 895.5 kg P/year were mobilized in the western and eastern basins of the lake, respectively (Table 1). To achieve state water quality standards for the two basins, P loading would need to be reduced by at least 39% (by >368 kg/year) and 31% (by >278 kg/year) in the two basins. Significant reductions in P loading in both basins were expected to come from MS4 stormwater, watershed runoff, and internal (sediment) sources (Table 1), although the proportional reduction assigned to these sources differed between basins. Because the western basin drains into the eastern basin, 75% of the reductions in P loading for the eastern basin were assigned to reductions in P entering from the western basin (Table 1). No strategies or mechanisms were recommended for achieving the proposed reductions.

4.4. 2018–2020 Lake Winona P8 Model TMDL

P8 modeling estimated that 678.9 and 544.8 kg P/year were mobilized in the western and eastern basins of Lake Winona, respectively (Table 2). To achieve state water quality standards for the two basins, P loading would need to be reduced by 14% (by 95.5 kg/year) and 30% (by 162.4 kg/year) in the two basins. Significant reductions in P loading in the western basin were expected to come solely by reducing contributions from Ditch #3, and in the eastern basin by reducing contributions from direct watershed runoff, internal (sediment) sources, and inflows from the western basin (Table 2).

4.5. Differences Between BATHTUB and P8 Models

The results of the 2016 BATHTUB and the 2018–2020 P8 models for P loading in Lake Winona produced dramatically different results. The predicted existing basin-specific loads from the P8 model were 28% (western basin) and 40% (eastern basin) lower than the P loadings estimated for the two basins by the earlier BATHTUB model (Table 1 and Table 2). The direction of the differences between models also differed between the lake basins. For the western basin, the predicted total watershed load (direct + Boller Lake + runoff to Ditch #3) was 19% lower for the P8 model than the total watershed load (watershed + C/I and MS4 stormwater) for the BATHTUB model. Similarly, the predicted internal load for the western basin was 62% less in the P8 model than in the BATHTUB model. In contrast, the predicted watershed load in the eastern basin was 156% higher in the P8 model than from the same source (watershed + C/I and MS4 stormwater) in the BATHTUB model, and internal load predictions were also 707% higher for the eastern basin in the P8 model. Finally, P loading coming from the western basin into the eastern basin was 72% lower in the P8 model.

4.6. Recommended Correction Actions to Reduce P Loading

The initial TMDL study of P loading in Lake Winona did not provide specific recommendations for reducing P loading from the various sources (MS4 stormwater, direct watershed, and internal) targeted for reductions in the TMDL analysis. The only directive from the MPCA was for the City of Winona to submit by 2018 a “schedule” identifying the BMPs that would be implemented during the next 5 years to move toward target P reductions from MS4 stormwater runoff, a timeline for their implementation, an assessment of progress to date, and a strategy (long-term) for achieving and maintaining P load reductions from MS4 stormwater runoff [19]. The cost to achieve the P loading reductions was estimated at USD 3.14 million based on median costs to reduce P in previous projects implemented in the USA [19,35].
Unlike the 2016 TMDL study for Lake Winona, the 2018–2020 follow-up P loading study included recommendations for specific, targeted projects to reduce P loading into the lake from various sources that would allow both lake basins to meet or exceed current state water quality standards (Table 3). These recommended projects included in-lake alum treatments and common carp management to address internal P loading, and two types of engineered structures to deal with focused P loads entering the lake from the watershed. Additionally, this study also presented cost estimates for each targeted project (Table 3). The total cost to achieve all the P loading reductions and maintain those reductions over a 15-year time span was estimated at USD 5.2 million.

5. Discussion and Evaluation of Recommended Corrective Actions for Reducing P Loading

The P loadings predicted by the BATHTUB and P8 TMDL models differed dramatically for several reasons. First, the BATHTUB model relied heavily on only a single Gilmore Creek monitoring site, limited historical Lake Winona data, and comparable P data from regional and national databases for P loading estimation, whereas the P8 model utilized seven in-watershed sites, three of which also had flow measurements (see Figure 1). Next, the BATHTUB model incorporated land-use-specific annual export rates from the literature, whereas the P8 model was calibrated to daily water quality monitoring and P load data that helped to simulate P removal by Boller Lake and other within-watershed BMPs (best management practices). Finally, the BATHTUB model was based on average annual steady-state conditions and incorrect western basin bathymetry, whereas the P8 model used mass-balance data calibrated to daily observations and used correct updated bathymetry for the western basin. Consequently, these differences likely resulted in the BATHTUB model overestimating (1) internal loading in the western basin, (2) the total P flowing from Boller Lake, and (3) the total P flowing from the western basin into the eastern basin [20].

5.1. In-Lake Alum Treatments

In oxygenated lake waters, iron binds to P and causes it to precipitate out and become unavailable for uptake by phytoplankton, algae, and macrophytes [11,36,37]. However, P is released from iron under anoxic conditions or when winds resuspend sediments in shallow lakes [15]. The P that enters the water column can then be taken up quickly (due to its relative scarcity) by photosynthesizing plants or be captured again by iron (if oxygen is present in the water column) and returned back to the sediments [36,37].
Alum (aluminum sulfate) and sodium aluminate treatments are used typically to prevent the release of phosphorus from bottom sediments (especially under anoxic conditions) to reduce or prevent the summer algal blooms often characteristic of nutrient-rich lakes [12,13,14,15,16,18,38,39]. By suppressing P release from sediments, the only P available to growing plants would be new inputs of P from the watershed and in-lake P released during the natural decay and breakdown of organic matter (both dissolved and particulate) [15]. Once a lake is treated with alum or sodium aluminate, the aluminum component binds with P in the sediments, forming a compound that remains intact during anoxia, preventing P from entering the water column [12,16,18,39]. Alum or sodium aluminate in excess of that needed to bind sediment P may be added to a lake to capture additional P that may enter or develop in the system during future months or years [20]. Alternatively, lake inflows could be treated with alum before entering a lake, reducing the need for in-lake alum treatment [25,26,27].
The greatest reductions (111.8 kg P/year; Table 2) in P loading in Lake Winona’s eastern basin were predicted to result from alum treatment of the basin’s bottom sediments [20]. If alum was also to be applied to the western basin (along with sodium aluminate if needed to maintain a proper pH level), up to an additional 54.5 kg P/year could be kept out of circulation within Lake Winona. Environmental consultants have advised that treatments for one or both basins should be split into two applications separated by two or more years, to better capture and sequester both the P released through organic matter decay and additional new P carried into the lake from the watershed [20]. Estimated costs for treating each basin, based on basin bathymetry and the sediment P levels measured in 2018, would be USD 200,000 for the eastern basin and USD 400,000 for the western basin (Table 3). Assuming a 15-year lifespan (for comparative purposes only; some treatments may last longer) for treatment effectiveness, annual costs per kg P sequestered would be USD 119 for the eastern basin and USD 484 for the western basin (Table 3) [20].
A factor that may complicate alum treatments in Lake Winona is the abundance of aquatic macrophytes within the lake [28]. Both basins currently support extensive stands of native coontail (Ceratophyllum demersum L.) and non-native Eurasian watermilfoil (Myriophyllum spicatum L.) throughout lake littoral zones (93% of the western basin and 47% of the eastern basin) during summer and fall [28,29,40]. However, more problematic is non-native curly-leaf pondweed (Potamogeton crispus Linnaeus), a cool-season macrophyte that dominates Lake Winona’s macrophyte community during spring and early summer before senescing in mid-summer [20]. This mid-summer senescence can release large amounts of P into the water column, potentially reducing the effectiveness (both annual and total lifespan) of alum treatment via bypassing the alum floc at the sediment–water interface [38]. The City of Winona has a long history of dealing with curly-leaf pondweed [20,28], and has partnered previously with the Minnesota Department of Natural Resources in attempts at controlling/reducing P. crispus in Lake Winona. Such efforts will need to continue to maximize the effectiveness of any future alum treatments in controlling internal P loading in the lake.

5.2. Common Carp Management

Effective alum treatment that keeps P bound to aluminum and locked in lake bottom sediments requires that those sediments remain relatively undisturbed [38]. Unfortunately, many species of fish disturb bottom sediments when feeding, uprooting aquatic plants and ingesting and sorting through bottom sediments for aquatic invertebrates, algae, and/or organic detritus [41]. These feeding activities can interfere with the alum–P interaction by disrupting the protective chemical boundary layer formed by alum at the sediment surface and reduce the effectiveness and lifespan of the alum treatment [38,42].
Common carp are omnivorous benthivores, feeding on a range of foods (e.g., aquatic macrophytes, algae, benthic invertebrates, and organic detritus) derived from the lake bottom. When adult carp are present at high densities (carp biomass > 50 kg/hectare), their feeding habits could lead to the reduced effectiveness or even failure of alum treatments intended to bind P in the sediments [24,42,43]. Alum treatments to hypolimnion sediments should remain relatively undisturbed, as carp may spend their overnight hours in deep waters but do not feed there [44]. However, intensive carp feeding activities in shallow waters every day [44] can disturb and suspend bottom sediments [24,42,43], interfering with whole-lake alum treatments [42]. Any P released from shallow-water sediments during carp feeding should quickly re-complex with iron in the presence of oxygen and return to the sediments, leaving only a brief time period for P to be picked up and utilized by phytoplankton. Unfortunately, carp themselves may cycle nutrients (both P and N) back into the water column via excretion and defecation [45], providing a continuing source of nutrients to the lake environment during the summer months, especially when carp populations are high.
The common carp population biomass in the eastern and western basins of Lake Winona has been estimated at 116 and 360 kg/hectare, respectively [46], levels that exceed those (50–100 kg/hectare) that can cause significant damage to lake ecosystems [47,48,49,50] and likely also affect lake nutrient levels [45]. Prior to undertaking alum treatments in Lake Winona, it was recommended that the current common carp abundance be assessed and, if necessary, the population reduced to a level that will not potentially counteract alum treatments [20]. In addition, to prevent the carp population from rebounding in future years, further recommendations included documenting carp spawning locations and taking measures (e.g., installation of electric carp barriers and carp capture/removal during spawning migrations) to reduce the chance of carp reproducing successfully [20,28,46]. Depending on the degree of need, carp management activities could cost as much as USD 500,000 for the entire Lake Winona system (the eastern and western basins, Ditch #3, and Boller Lake), but by itself would reduce suspension or excretion (via carp waste products) [45,51] of an estimated 27.3 kg P/year at an annual cost of USD 1232/kg (Table 3) [20]. Annual costs for maintaining the barriers and/or catching carp could vary widely and would be a necessary component to suppress the carp population.

5.3. Stormwater Treatment Pond for Ditch #3/Lower Gilmore Creek

Water entering Lake Winona’s western basin through Ditch #3 is responsible for two-thirds of the annual external P loading for that basin. P sources contributing to this load include outflows from Boller Lake (fed by flows from upper Gilmore Creek and its watershed), stormwater outfalls (n > 20) that discharge directly into Ditch #3, and other general watershed drainage to the ditch. Intercepting and removing a portion of this P load flowing in from Ditch #3 (Table 2) has been recommended as the only action needed to bring P levels in the western basin into compliance with state standards [20].
The most effective management action to reduce P loading from Ditch #3 was identified as construction of a 1.6-hectare stormwater treatment pond immediately upstream from the ditch’s confluence with the western basin (Figure 4), utilizing a portion of an existing fringe wetland (adjacent to Ditch #3) that was partially filled with dredge spoils during the 1950–1953 dredging of Lake Winona [28]. Weirs would be used to regulate inflows from the ditch and outflows from the stormwater treatment basin to facilitate proper retention time for treatment (i.e., settling to remove P). If needed, chemical treatment (e.g., alum or iron) [25,26,27] or various filtration media (e.g., calcite, zeolite, sand, or iron filings) [21,22,23,25,52,53] could be added to the treatment pond to optimize P removal [20]. Such a treatment pond was projected to reduce P loading to the western basin of Lake Winona by 95.5 kg P/year at a cost of USD 1.6 million, or USD 1100/kg P removed over a 15-year period (Table 3) [20]. Implementing this solution would also require regulatory approval for converting a wetland, and incur wetland mitigation banking costs that likely would raise the P reduction cost by USD 100–120/kg P removed.

5.4. Iron-Enhanced Sand Filtration Basins—Eastern Basin

Urban stormwater runoff typically has P concentrations exceeding ambient water quality standards [21,26,54,55], which can lead to eutrophication of the receiving surface waters [54] and compromise efforts to manage in-lake P concentrations [26]. In many jurisdictions, various treatments have been applied to inflows to prevent stormwater P loads from impacting receiving lakes, those treatments ranging from directing stormwaters into temporary retention basins for particulate settling (see Section 5.3 above) to passing stormwaters through various filtration materials and media that can effectively remove significant proportions of the soluble P load [21,22,23,25,26,27,52,53].
Twenty-two municipal stormwater outfalls carry urban runoff into Lake Winona. These outfalls and other surface runoff annually contribute 37% of the P load to the eastern basin of Lake Winona, which represents the largest single source of P to the basin (Table 2). Treating stormwaters at a subset of six of these outfalls via the installation of five engineered stormwater filtration basins has been recommended as an achievable action to reduce the P load to the eastern basin [20]. Although the proportion (10%) of the direct watershed P load removed by these structures may be modest compared to other actions (Table 2), the reduction would be important in achieving the P load reduction required to meet state water quality standards for the basin [20].
Five iron-enhanced sand filtration basins [21,22,53] could be designed and constructed as off-line (parallel to existing subterranean ductworks) structures near the northern shore of the eastern basin of Lake Winona (Figure 5) to filter/absorb significant proportions (30–95%) [21,22,53] of the dissolved P load of stormwater flows, preventing nutrient delivery to the lake [20]. Each basin would pass all stormwater flows through a sand filter amended with some form of elemental iron (e.g., iron filings or steel wool) to bind dissolved P [21] before conveying the now-filtered water on into the lake. Basins would be sized appropriately to accommodate the maximum flows associated with each outfall while allowing retention/contact times sufficient to remove transported P. Each filtration basin would be located on publicly owned, under-utilized parkland in close proximity to the lake.
Collectively, the five filtration basins were expected to remove nearly 21 kg P/year from stormwater runoff, the smallest P reduction of any of the actions recommended for Lake Winona (Table 2). Filtration basins also would be the most expensive of the P reduction actions, costing an estimated USD 2.5 million to construct all five basins, or USD 7920/kg P removed over a 15-year period (Table 3) [20]. Basin maintenance costs would be borne over the lifetime of the structure in the form of City of Winona staff time. Debris, vegetation, and trash would need to be removed, likely on a quarterly basis, to maintain effective P removal.

5.5. Collective Benefits and Costs

The P8 modeling for P reduction in Lake Winona specifically highlighted four P sources (one in the western basin and three in the eastern basin; Table 2) where reductions (268.5 kg P/year) could be accomplished to bring both basins into compliance with state water quality standards. However, the recommended actions described above to achieve P reductions (see Section 5.1, Section 5.2, Section 5.3 and Section 5.4) would produce an estimated annual reduction of 310 kg P/year, 15% more than needed according to the P8 model but >50% less than what the 2016 TMDL model estimated would be needed to achieve compliance with state standards [19].
The overall costs to remove 310 kg P/year from Lake Winona P loading sources for 15 years was estimated at USD 5.2 million in total, or USD 347,000/year (Table 3) [20]. This total cost equates to a per-unit cost of USD 1118/kg P reduced. Nearly 79% of P reduction costs accrue from design and construction (but not maintenance) of the engineered stormwater treatment pond and filtration basins, which were expected to produce <38% of the P reductions. While engineered structures can be effective means for reducing P loading in both urban and rural landscapes [17,55,56], their cost-to-benefit ratios can average 50 times more than those of in-lake or more “natural-based” (e.g., wetland treatments or riparian buffers) landscape options [55,56]. Similar to the cost estimates for Lake Winona, engineered solutions for P reductions often range from USD 1000–5000/kg P removed, whereas reductions attributed to wetlands, riparian buffers, and in-lake actions cost < USD 500/kg P reduced [56].
Urban areas like Winona may have somewhat limited options available for managing P loading to lakes due to the built-up nature of the environment [17]. Natural wetlands for treating stormwaters, space to site constructed wetlands, and natural vegetated buffers frequently are lacking in urban settings, often leading to the use of smaller-scale options like filtration and/or retention basins or end-of-pipe solutions [17,25]. Such smaller-scale P reduction systems can prove successful in some settings [21,25,26], but they also can suffer from highly variable efficiencies and may be unable to reduce P loading significantly [17,21]. Lake Winona benefits from public ownership of the entire lakeshore, allowing for installation of multiple sand filter basins at sites where their success is likely [20]. However, their extremely high cost-to-benefit ratios can be troublesome to decision-makers. It may be far more cost-effective and fiscally responsible to allow P loading from municipal stormwaters to continue without abatement and then use a much less expensive P reduction treatment in-lake (e.g., alum to trap P in the sediments) to treat the P load inflows [17]. For example, the eastern basin of Lake Winona can be treated with alum 11 times (or both basins treated four times) for the same cost as utilizing the five iron-enhanced sand filtration basins.

5.6. Implementing P-Reduction Recommendations

The City of Winona is exploring its options for managing P loadings to Lake Winona. As recommended [20], the city will first assess common carp abundance in Lake Winona before proceeding with any alum treatments. Partial funding has been secured from regional sources to perform initial carp assessments. Depending on the outcome of those initial surveys, the city likely will need to pursue grant funding if more extensive work is needed on the carp population. Work also has continued (as recommended) [20] on managing aquatic macrophytes in Lake Winona. In 2020, a portion (4.8 hectares) of the western basin was chemically treated in spring to suppress curly-leaf pondweed. While successful in killing the active growth in 2020, surveys the following spring revealed no obvious effect on the extent or density of curly-leaf pondweed. No new plant management efforts have been implemented since then.
The city continues to take actions that could reduce the amount of P loading into Lake Winona from the municipal stormwater system [57]. Actions to date have included the following: incentives (small grants) for area residents to install rain gardens on their properties to capture runoff and allow it to seep into the ground; fall and spring street sweeping schedules to collect fallen leaves and other organic matter before that material can enter the stormwater system; public education programs that encourage area residents to keep street-side stormwater grates clear of organic materials (e.g., sticks, grass clippings, and leaves) and control pet wastes; and maintaining a natural vegetative buffer between mowed turfgrass parklands and Lake Winona [57,58]. Such proactive approaches, especially if done at the ideal time (e.g., street sweeping immediately after leaf fall and again before the onset of spring rains), could reduce particulate P inputs and may reduce the need for expensive engineered P reduction filtration basins.
Unfortunately for the city, any actions taken to manage P in-lake (e.g., alum treatments and common carp management) do not qualify as progress toward meeting its NPDES permit goals for its municipal storm sewer system. Consequently, in addition to the on-land activities described above, the city also is seeking external grant funding to explore the use of stormwater treatment ponds and sand filters to manage P loading from the municipal stormwater system, as recommended by their consultants [20]. Despite the high cost/benefit ratios of these engineered solutions, these stand-alone solutions to intercept P after it has entered the stormwater system likely will be more effective and similar in cost-effectiveness to wider-scale implementation of rain gardens during street reconstruction projects to intercept P before it enters the system [20]. Street reconstruction costs were projected to increase by 10–15% if rain gardens or other bio-infiltration methods were incorporated to capture street runoff [20].
A factor that potentially could complicate the P loading issues for Lake Winona is the possibility of invasion by non-native zebra mussels (Dreissena polymorpha). Zebra mussels currently are not present in the lake, although they have been present in the nearby Mississippi River for several decades. In some lakes in North America, invading zebra mussels have caused significant declines in total P concentrations [59,60,61], although thermal stratification, lake depth, and water residence time can variously neutralize effects of zebra mussels on total P [61]. At this time, it is not possible to predict how P concentrations in Lake Winona might be affected if zebra mussels colonize the lake in the future. However, it is possible that zebra mussel invasion could lower P concentrations sufficiently for both lake basins to meet water quality standards without the need for any in-lake or watershed treatments.

5.7. Success in Restoring P-Impaired Lakes in Minnesota

Approximately 20 to 25% of Minnesota’s 12,000 lakes likely are impaired due to excessive nutrient levels, mostly from P, although only 700 lakes officially appear on the impaired waters list for nutrients [62,63]. During the past 20 years, over 60 nutrient-impaired lakes have been restored and delisted [62,64,65], the majority (73%) of those in the Minneapolis–Saint Paul urban area [62]. For these lakes, most restorations and delistings took over a decade to achieve, with some requiring 20 years or more [62,64,65]. Over 90% of restored lakes utilized various urban watershed best management practices (BMPs) such as bioretention/infiltration basins, increased street sweeping, shoreline stabilization/restoration, iron-enhanced sand filters, and wetland restoration/enhancement, whereas >50% used internal BMPs such as alum treatment and carp management [62]. Nearly all lake delistings were accomplished only after using multiple BMPs or other strategies [62,64,65]. Funding for these BMPs generally came from local watershed districts, governmental grants, and stormwater management fees paid by area residents. Consequently, it appears that the standard procedure for reducing P loadings and eventual impairment delisting of the other urban lakes in Minnesota most likely will require employing a variety of in-lake and watershed approaches carried out over several years.

6. Conclusions

Reducing P loading to Lake Winona to achieve state water quality standards is a manageable task that will require a diversity of actions applied to the main P sources. Reductions in internal loading via alum treatments will be the most cost-effective and should produce observable effects quickly, although the presence of common carp and curly-leaf pondweed may complicate the process. However, the city cannot apply in-lake reductions toward meeting its permit goals for stormwater pollution reduction. P reductions from upstream (watershed) sources and from the city’s municipal stormwater system can apply toward the city’s permit goals but will require more expensive engineered retention and filtration basins. External grant funding will be needed to assist the city in reducing P loads to meet state standards, especially for implementing the more costly engineered solutions.

Author Contributions

Conceptualization, N.D.M. and J.H.; methodology, N.D.M.; formal analysis, N.D.M. and J.H.; resources, J.H.; writing—original draft preparation, N.D.M.; writing—review and editing, N.D.M. and J.H. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Data are available from the authors upon reasonable request.

Acknowledgments

We thank the City of Winona for providing details on the TMDL studies and ongoing efforts to reduce P loadings.

Conflicts of Interest

The authors declare no competing interest.

References

  1. United States Environmental Protection Agency. Summary of the Clean Water Act; United States Environmental Protection Agency: Washington, DC, USA, 2024. Available online: https://www.epa.gov/laws-regulations/summary-clean-water-act (accessed on 6 April 2025).
  2. United States Environmental Protection Agency. Identifying Polluted Waters and Developing Plans to Restore Them; United States Environmental Protection Agency: Washington, DC, USA, 2025. Available online: https://www.epa.gov/tmdl (accessed on 6 April 2025).
  3. Schindler, D.W.; Carpenter, S.R.; Chapra, S.C.; Hecky, R.E.; Orihel, D.M. Reducing phosphorus to curb lake eutrophication is a success. Environ. Sci. Technol. 2016, 50, 8923–8929. [Google Scholar] [CrossRef] [PubMed]
  4. Schindler, D.W. The dilemma of controlling cultural eutrophication of lakes. Proc. R. Soc. Lond. Ser. B 2012, 279, 4322–4333. [Google Scholar] [CrossRef] [PubMed]
  5. Bhagowati, B.; Ahamad, K.U. A review on lake eutrophication dynamics and recent developments in lake modeling. Ecohydrol. Hydrobiol. 2019, 19, 155–166. [Google Scholar] [CrossRef]
  6. Wang, H.; Wang, H. Mitigation of lake eutrophication: Loosen nitrogen control and focus on phosphorus abatement. Prog. Nat. Sci. 2009, 19, 1445–1451. [Google Scholar] [CrossRef]
  7. Carpenter, S.R. Phosphorus control is critical to mitigating eutrophication. Proc. Natl. Acad. Sci. USA 2008, 105, 11039–11040. [Google Scholar] [CrossRef]
  8. Oliveira, M.; Machado, A.V. The role of phosphorus on eutrophication: A historical review and future perspectives. Environ. Technol. Rev. 2013, 2, 117–127. [Google Scholar] [CrossRef]
  9. Liang, Z.; Soranno, P.A.; Wagner, T. The role of phosphorus and nitrogen on chlorophyll a: Evidence from hundreds of lakes. Water Res. 2020, 185, 116236. [Google Scholar] [CrossRef]
  10. Qin, B.; Zhou, J.; Gardner, W.S.; Deng, J.; Brookes, J.D. Water depth underpins the relative roles and fates of nitrogen and phosphorus in lakes. Environ. Sci. Technol. 2020, 54, 3191–33198. [Google Scholar] [CrossRef] [PubMed]
  11. Singer, P.C. Control of phosphate by iron. J. Water Pollut. Control Fed. 1972, 44, 663–669. [Google Scholar]
  12. Garrison, P.J.; Knauer, D.R. Long-term evaluation of three alum treated lakes. Lake Reserv. Manag. 1984, 1, 513–517. [Google Scholar] [CrossRef]
  13. Smeltzer, E.; Kirn, R.A.; Fiske, S. Long-term water quality and biological effects of alum treatment of Lake Morey, Vermont. Lake Reserv. Manag. 1999, 15, 173–184. [Google Scholar] [CrossRef]
  14. Rydin, E.; Huser, B.; Welch, E.B. Amount of phosphorus inactivated by alum treatments in Washington lakes. Limnol. Oceanogr. 2000, 45, 226–230. [Google Scholar] [CrossRef]
  15. Steinman, A.D.; Ogdahl, M. Ecological effects after an alum treatment in Spring Lake, Michigan. J. Environ. Qual. 2008, 37, 22–29. [Google Scholar] [CrossRef]
  16. Huser, B.; Brezonik, P.; Newman, R. Effects of alum treatment on water quality and sediment in the Minneapolis Chain of Lake, Minnesota, USA. Lake Reserv. Manag. 2011, 27, 220–228. [Google Scholar] [CrossRef]
  17. Huser, B.J.; Futter, M.; Lee, J.T.; Perniel, M. In-lake measures for phosphorus control: The most feasible and cost-effective solution for long-term management of water quality in urban lakes. Water Res. 2016, 97, 142–152. [Google Scholar] [CrossRef] [PubMed]
  18. Brattebo, S.K.; Welch, E.B.; Gibbons, H.L.; Burghdoff, M.K.; Williams, G.N.; Oden, J.L. Effectiveness of alum in a hypereutrophic lake with substantial external loading. Lake Reserv. Manag. 2017, 33, 108–118. [Google Scholar] [CrossRef]
  19. Minnesota Pollution Control Agency. Mississippi River-Winona Watershed Pollutant Reduction Project (Total Maximum Daily Load Study) for Nutrients, Sediment and Bacteria; Minnesota Pollution Control Agency: Saint Paul, MN, USA, 2016.
  20. WHKS and Barr Engineering. Lake Winona Water Quality Improvement Plan: A Targeted, Prioritized, and Measurable Implementation Plan to Effectively Restore Lake Winona; Barr Engineering: Minneapolis, MN, USA, 2020. [Google Scholar]
  21. Fairbairn, D.J.; Trojan, M.D. Iron-enhanced sand filters: Multi-year urban runoff (stormwater) quality performance. Sci. Total Environ. 2023, 859, 160177. [Google Scholar]
  22. Erickson, A.J.; Gulliver, J.S.; Weiss, P.T. Capturing phosphates with iron enhanced sand filtration. Water Res. 2012, 46, 3032–3042. [Google Scholar] [CrossRef]
  23. Wen, D.; Chang, N.-B.; Wanielista, M.P. Assessing nutrient removal in stormwater runoff for urban farming with iron filings-based green environmental media. Sci. Rep. 2020, 10, 9379. [Google Scholar] [CrossRef]
  24. Huser, B.J.; Bajer, P.G.; Chizinski, C.J.; Sorensen, P.W. Effects of common carp (Cyprinus carpio) on sediment mixing depth and mobile phosphorus mass in the active sediment layer of a shallow lake. Hydrobiologia 2016, 763, 23–33. [Google Scholar] [CrossRef]
  25. Bodus, B.; McDonald, W. End-of-pipe filter to reduce phosphorus concentrations from the effluent of green infrastructure underdrains. J. Environ. Manag. 2025, 375, 124131. [Google Scholar] [CrossRef]
  26. Pilgrim, K.M.; Brezonik, P.L. Treatment of lake inflows with alum for phosphorus removal. Lake Reserv. Manag. 2005, 21, 1–9. [Google Scholar] [CrossRef]
  27. Pilgrim, K.M.; Brezonik, P.L. Evaluation of the potential adverse effects of lake inflow treatment with alum. Lake Reserv. Manag. 2005, 21, 77–87. [Google Scholar] [CrossRef]
  28. Fremling, C.R.; Heins, G.A. A Lake Winona Compendium: Information Concerning the Reclamation of an Urban Winter-Kill Lake at Winona, Minnesota, 2nd ed.; Winona State University: Winona, MN, USA, 1986. [Google Scholar]
  29. Mundahl, N.D.; Hoisington, J. Game fish response to dredging of a eutrophic urban lake in Minnesota. Lake Reserv. Manag. 2020, 37, 170–185. [Google Scholar] [CrossRef]
  30. City of Winona. Stormwater; City of Winona: Winona, MN, USA, 2025. Available online: https://www.cityofwinona.com/341/Stormwater (accessed on 8 April 2025).
  31. Minnesota Pollution Control Agency. Lakes and Streams Water Quality Dashboard, Winona: Northwest Bay; Minnesota Pollution Control Agency: Saint Paul, MN, USA, 2025. Available online: https://webapp.pca.state.mn.us/surface-water/impairment/85-0011-02 (accessed on 9 April 2025).
  32. Minnesota Pollution Control Agency. Lakes and Streams Water Quality Dashboard, Winona: Southeast Bay; Minnesota Pollution Control Agency: Saint Paul, MN, USA, 2025. Available online: https://webapp.pca.state.mn.us/surface-water/impairment/85-0011-01 (accessed on 9 April 2025).
  33. Walker, W.; Walker, J. P8 Urban Catchment Model: Program for Predicting Polluting Particle Passage thru Pits, Puddles, and Ponds. 17 January 2017. p8. Available online: https://www.wwwalker.net/p8/ (accessed on 8 April 2025).
  34. Canfield, D.; Bachman, R. Prediction of total phosphorus concentration, chlorophyll-a, and Secchi depths in natural and artificial lakes. Can. J. Fish. Aquat. Sci. 1981, 38, 414–423. [Google Scholar] [CrossRef]
  35. Forasté, J.A.; Goo, R.; Thrash, J.; Hair, L. Measuring the cost-effectiveness of LID and conventional stormwater management plans using life cycle costs and performance metrics. In Low Impact Development Technology: Implementation and Economics, Proceedings of the 2012 Ohio Stormwater Conference, Toledo, OH, USA, 6–8 June 2015; American Society of Civil Engineers: Reston, VA, USA, 2015. [Google Scholar]
  36. Horne, A.J.; Goldman, C.R. Limnology, 2nd ed.; McGraw-Hill, Inc.: New York, NY, USA, 1994. [Google Scholar]
  37. Wetzel, R.G. Limnology: Lake and River Ecosystems, 3rd ed.; Academic Press: San Diego, CA, USA, 2001. [Google Scholar]
  38. Welch, E.B.; Schrieve, G.D. Alum treatment effectiveness and longevity in shallow lakes. In Nutrient Dynamics and Biological Structure in Shallow Freshwater and Brackish Lakes. Developments in Hydrobiology; Mortensen, E., Jeppesen, E., Søndergaard, M., Nielsen, L.K., Eds.; Springer: Dordrecht, The Netherlands, 1994; Volume 94, pp. 423–431. [Google Scholar]
  39. Pilgrim, K.M.; Huser, B.J.; Brezonik, P.L. A method for comparative evaluation of whole-lake and inflow alum treatment. Water Res. 2007, 41, 1215–1224. [Google Scholar] [CrossRef]
  40. Henkelman, M.L.; Mundahl, N.D. Growth and condition of largemouth bass (Micropterus nigricans) and bluegill (Lepomis macrochirus) in a Minnesota, USA, lake with separate dredged and non-dredged basins. Limnol. Rev. 2024, 24, 235–249. [Google Scholar] [CrossRef]
  41. Diana, J.S. Biology and Ecology of Fishes, 2nd ed.; Cooper Publishing Group: Traverse City, MI, USA, 2004. [Google Scholar]
  42. Gao, J.; Dai, Q.; Fei, Q.; Guo, F.; Miao, T.; Zhang, H.; Liao, M.; Wang, H. Common carp Cyprinus carpio L. weaking the activity of two commonly used water-clearing agents for lake eutrophication control: A mesocosm study. Appl. Ecol. Environ. Res. 2025, 23, 915–924. [Google Scholar] [CrossRef]
  43. Bartodziej, W.; Sorensen, P.W.; Bajer, P.G.; Pilgrim, K.; Blood, S. A Minnesota story: Urban shallow lake management. Lakeline 2017, 37, 23–29. [Google Scholar]
  44. Benito, J.; Benejam, L.; Zamora, L.; García-Berthou, E. Diel cycle and effects of water flow on activity and use of depth by common carp. Trans. Am. Fish. Soc. 2015, 144, 491–501. [Google Scholar] [CrossRef]
  45. Chumchal, M.M.; Drenner, R.W. Interrelationships between phosphorus loading and common carp in the regulation of phytoplankton biomass. Arch. Hydrobiol. 2004, 161, 147–158. [Google Scholar] [CrossRef]
  46. Mundahl, N.D.; Schnaser, A.C.; Kluzak, C.M.; Henkelman, M.L. Common carp (Cyprinus carpio) reinvasion and population expansion in Lake Winona: A modified urban floodplain lake in Minnesota, USA. Water 2024, 16, 2571. [Google Scholar] [CrossRef]
  47. Bajer, P.G.; Beck, M.W.; Cross, T.K.; Koch, J.D.; Bartodziej, W.M.; Sorensen, P.W. Biological invasion by a benthivorous fish reduced the cover and species richness of aquatic plants in most lakes of a large North American ecoregion. Glob. Change Biol. 2016, 22, 3937–3947. [Google Scholar] [CrossRef]
  48. Bajer, P.G. A decade of common carp research and management in Minnesota. Lakeline 2017, 37, 43–48. [Google Scholar]
  49. Vilizzi, L.; Tarkan, A.S.; Copp, G.H. Experimental evidence from causal criteria analysis for the effects of common carp Cyprinus carpio on freshwater ecosystems: A global perspective. Rev. Fish. Sci. Aquac. 2015, 23, 253–290. [Google Scholar] [CrossRef]
  50. Lechelt, J.D.; Bajer, P.G. Elucidating the mechanism underlying the productivity-recruitment hypothesis in the invasive common carp. Aquat. Invasions 2016, 11, 469–482. [Google Scholar] [CrossRef]
  51. Schaus, M.H.; Vanni, M.J.; Wissing, T.E.; Bremigan, M.T.; Garvey, J.E.; Stein, R.A. Nitrogen and phosphorus excretion by detritivorous gizzard shad in a reservoir ecosystem. Limnol. Oceanogr. 1997, 42, 1386–1397. [Google Scholar] [CrossRef]
  52. Reddy, K.R.; Xie, T.; Dastgheibi, S. Nutrients removal from urban stormwater by different filter materials. Water Air Soil Poll. 2013, 225, 1778. [Google Scholar] [CrossRef]
  53. Erickson, A.J.; Gulliver, J.S.; Weiss, P.T.; Huser, B.J. Iron-enhanced sand filtration for stormwater phosphorus removal. In Proceedings of the 17th IAHR APD Congress, Auckland, New Zealand, 22–24 February 2010; Available online: https://www.iahr.org/library/infor?pid=10400 (accessed on 28 March 2025).
  54. Minnesota Pollution Control Agency. Phosphorus in Stormwater—Minnesota Stormwater Manual; Minnesota Pollution Control Agency: Saint Paul, MN, USA, 2023. Available online: https://stormwater.pca.state.mn.us/index.php?title=Phosphorus_in_stormwater (accessed on 3 April 2025).
  55. Kramer, D.B.; Polasky, S.; Starfield, A.; Palik, B.; Westphal, L.; Snyder, S.; Jakes, P.; Hudson, R.; Gustafson, E. A comparison of alternative strategies for cost-effective water quality management in lakes. Environ. Manag. 2006, 38, 411–425. [Google Scholar] [CrossRef] [PubMed]
  56. Hjerppe, T.; Väisänen, S. A practical tool for selecting cost-effective combinations of phosphorus loading mitigation measures in Finnish catchments. Int. J. River Basin Manag. 2015, 13, 363–376. [Google Scholar] [CrossRef]
  57. City of Winona. 2045 Comprehensive Plan: City of Winona; City of Winona: Winona, MN, USA, 2024. Available online: https://www.cityofwinona.com/DocumentCenter/View/5248/Winona-Comprehensive-Plan_JUNE2024_Final_Reduced?bidId= (accessed on 7 April 2025).
  58. Mundahl, N.D.; Howard, J. Rehabilitation and continuing management of an urban lake shoreline in southeastern Minnesota, USA. Land 2024, 13, 2224. [Google Scholar] [CrossRef]
  59. Fahnenstiel, G.L.; Lang, G.A.; Nalepa, T.F.; Johengen, T.H. Effects of zebra mussels (Dreissena polymorpha) colonization on water quality parameters in Saginaw Bay, Lake Huron. J. Great Lakes Res. 1995, 21, 435–448. [Google Scholar] [CrossRef]
  60. Bykova, O.; Laursen, A.; Bostan, V.; Bautista, J.; McCarthy, L. Do zebra mussels (Dreissena polymorpha) alter lake water chemistry in a way that favors Microcystis growth? Sci. Total Environ. 2006, 371, 362–372. [Google Scholar] [CrossRef] [PubMed]
  61. Cha, Y.; Stow, C.A.; Bernhardt, E.S. Impacts of dreissenid mussel invasions on chlorophyll and total phosphorus in 25 lakes in the USA. Freshw. Biol. 2013, 58, 192–206. [Google Scholar] [CrossRef]
  62. Strom, J.; Timm, A.; Anderson, J.; MacLean, S. Twenty years of lake nutrient impairment delistings in Minnesota. Lakeline 2024, 44-1, 9–13. [Google Scholar]
  63. Minnesota Pollution Control Agency. After More Than 20 Years, Restoration of Kohlman Lake Sees Results; Minnesota Pollution Control Agency: Saint Paul, MN, USA, 2023. Available online: https://www.pca.state.mn.us/news-and-stories/kohlman-lake-delisted (accessed on 13 June 2025).
  64. Minnesota Water Pollution Control Agency. Water Pollutant: Phosphorus; Minnesota Pollution Control Agency: Saint Paul, MN, USA, 2025. Available online: https://www.pca.state.mn.us/pollutants-and-contaminants/phosphorus (accessed on 13 June 2025).
  65. Minnesota Water Pollution Control Agency. Reducing Phosphorus Is Key to Helping Metro-Area Lakes; Minnesota Pollution Control Agency: Saint Paul, MN, USA, 2022. Available online: https://www.pca.state.mn.us/news-and-stories/reducing-phosphorus-is-a-key-to-helping-metro-area-lakes (accessed on 13 June 2025).
Limnolrev 25 00028 g001
Figure 1. Aerial view of Lake Winona in Winona, MN, USA, indicating major watershed features. The Mississippi River is located in the upper right of the photograph. Q and C designate the locations where water flow rates and total P concentrations were assessed. Blue arrows indicate the direction of water flow. Modified from [20].
Figure 1. Aerial view of Lake Winona in Winona, MN, USA, indicating major watershed features. The Mississippi River is located in the upper right of the photograph. Q and C designate the locations where water flow rates and total P concentrations were assessed. Blue arrows indicate the direction of water flow. Modified from [20].
Limnolrev 25 00028 g001
Limnolrev 25 00028 g002
Figure 2. Historical (1974–1982) [28] (A) and more recent (B) [31,32] total phosphorus concentrations (mean + standard deviation) in the western and eastern basins of Lake Winona. P levels from surface and bottom (>5 m deep) water samples are shown separately. Horizontal black and red lines indicate state water quality standards for the western and eastern basins, respectively. Numbers above bars are sample sizes.
Figure 2. Historical (1974–1982) [28] (A) and more recent (B) [31,32] total phosphorus concentrations (mean + standard deviation) in the western and eastern basins of Lake Winona. P levels from surface and bottom (>5 m deep) water samples are shown separately. Horizontal black and red lines indicate state water quality standards for the western and eastern basins, respectively. Numbers above bars are sample sizes.
Limnolrev 25 00028 g002
Limnolrev 25 00028 g003
Figure 3. Total phosphorus concentrations (means + one SD) of water samples collected during March–December 2018 from nine sites in the Lake Winona watershed (A) and from three depths at two deepwater sites in Lake Winona (B). Numbers above bars are sample sizes. Horizontal green and red lines indicate state water quality phosphorus standards for the western (<60 ppb) and eastern (<40 ppb) basins of the lake, respectively. Data summarized from [20].
Figure 3. Total phosphorus concentrations (means + one SD) of water samples collected during March–December 2018 from nine sites in the Lake Winona watershed (A) and from three depths at two deepwater sites in Lake Winona (B). Numbers above bars are sample sizes. Horizontal green and red lines indicate state water quality phosphorus standards for the western (<60 ppb) and eastern (<40 ppb) basins of the lake, respectively. Data summarized from [20].
Limnolrev 25 00028 g003
Limnolrev 25 00028 g004
Figure 4. Proposed stormwater retention basin to remove phosphorus from water diverted from County Ditch #3 prior to it entering the western basin of Lake Winona. Modified from [20].
Figure 4. Proposed stormwater retention basin to remove phosphorus from water diverted from County Ditch #3 prior to it entering the western basin of Lake Winona. Modified from [20].
Limnolrev 25 00028 g004
Limnolrev 25 00028 g005
Figure 5. Proposed locations for five iron-enhanced sand filters (indicated by red stars) to remove phosphorus from stormwater runoff before it enters the eastern basin of Lake Winona. Purple lines indicate the city “watershed” that drains into the eastern basin via the stormwater system. Yellow lines indicate underground stormwater lines. Modified from [20].
Figure 5. Proposed locations for five iron-enhanced sand filters (indicated by red stars) to remove phosphorus from stormwater runoff before it enters the eastern basin of Lake Winona. Purple lines indicate the city “watershed” that drains into the eastern basin via the stormwater system. Yellow lines indicate underground stormwater lines. Modified from [20].
Limnolrev 25 00028 g005
Table 1. Total Maximum Daily Loading (TMDL) analysis (BATHTUB model) for total phosphorus (P) in the two basins of Lake Winona and estimated load reductions needed to achieve state water quality goals in surface waters during June–September (western basin: <60 ppm total P; eastern basin: <40 ppm total P) [19]. C/I is stormwater runoff from construction and industrial activities. MS4 is the City of Winona municipal stormwater runoff system.
Table 1. Total Maximum Daily Loading (TMDL) analysis (BATHTUB model) for total phosphorus (P) in the two basins of Lake Winona and estimated load reductions needed to achieve state water quality goals in surface waters during June–September (western basin: <60 ppm total P; eastern basin: <40 ppm total P) [19]. C/I is stormwater runoff from construction and industrial activities. MS4 is the City of Winona municipal stormwater runoff system.
LAKE BasinP SourceExisting Load kg/YearReduction to Achieve Goal
% (kg)
WesternC/I stormwater0.20
MS4 stormwater180.738 (67.9)
Direct watershed570.723 (129.7)
Internal180.794 (170.6)
Atmospheric14.70
Basin total947.039 (368.2)
EasternC/I stormwater0.10
MS4 stormwater89.441 (37.1)
Direct watershed38.732 (12.5)
Internal19.7100 (19.7)
Atmospheric38.70
Western basin714.829 (209.6)
Basin total901.431 (278.9)
Table 2. P8 Urban Catchment Model analysis for total phosphorus (P) in the two basins of Lake Winona and estimated load reductions needed to achieve state water quality goals in surface waters during June–September (western basin: <60 ppm total P; eastern basin: <40 ppm total P) [19]. Watershed sources included direct drainage plus C/I (construction/industrial) and MS4 (municipal) stormwater sources.
Table 2. P8 Urban Catchment Model analysis for total phosphorus (P) in the two basins of Lake Winona and estimated load reductions needed to achieve state water quality goals in surface waters during June–September (western basin: <60 ppm total P; eastern basin: <40 ppm total P) [19]. Watershed sources included direct drainage plus C/I (construction/industrial) and MS4 (municipal) stormwater sources.
Lake BasinP SourceExisting Load Kg/YearReduction to Achieve Goal
% (kg)
WesternDirect watershed202.30
Boller Lake118.80
Runoff to Ditch #3287.133 (95.5)
Internal68.90
Atmospheric1.80
Basin total678.914 (95.5)
EasternDirect watershed200.510 (20.9)
Internal139.380 (111.8)
Atmospheric5.40
Western basin199.615 (29.7)
Basin total544.830 (162.4)
Table 3. Lake Winona phosphorus (P) reduction options and estimated costs (in USD). Annual costs based on a 15-year project life expectancy. Modified from [20].
Table 3. Lake Winona phosphorus (P) reduction options and estimated costs (in USD). Annual costs based on a 15-year project life expectancy. Modified from [20].
P Reduction OptionEstimated Annual P Reduction (kg)Estimated Total CostsAnnual Cost/kg P Removed
Stormwater treatment pond for Ditch #3 inputs95.5USD 1.6 millionUSD 1100
Eastern basin alum treatment111.8USD 200,000USD 119
Western basin alum treatment54.5USD 400,000USD 490
Carp assessment, control27.3USD 500,000USD 1232
Eastern basin sand filters20.9USD 2.5 millionUSD 7990
Totals310.0USD 5.2 millionUSD 1118 average
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Mundahl, N.D.; Howard, J. Total Phosphorus Loadings and Corrective Actions Needed to Restore Water Quality in a Eutrophic Urban Lake in Minnesota, USA: A Case Study. Limnol. Rev. 2025, 25, 28. https://doi.org/10.3390/limnolrev25030028

AMA Style

Mundahl ND, Howard J. Total Phosphorus Loadings and Corrective Actions Needed to Restore Water Quality in a Eutrophic Urban Lake in Minnesota, USA: A Case Study. Limnological Review. 2025; 25(3):28. https://doi.org/10.3390/limnolrev25030028

Chicago/Turabian Style

Mundahl, Neal D., and John Howard. 2025. "Total Phosphorus Loadings and Corrective Actions Needed to Restore Water Quality in a Eutrophic Urban Lake in Minnesota, USA: A Case Study" Limnological Review 25, no. 3: 28. https://doi.org/10.3390/limnolrev25030028

APA Style

Mundahl, N. D., & Howard, J. (2025). Total Phosphorus Loadings and Corrective Actions Needed to Restore Water Quality in a Eutrophic Urban Lake in Minnesota, USA: A Case Study. Limnological Review, 25(3), 28. https://doi.org/10.3390/limnolrev25030028

Article Metrics

Back to TopTop