Introduction

When excess nutrients enter surface waters, they have the potential to degrade water quality and cause the growth of nuisance or harmful algal blooms (Breitburg and others, 2018). Algal blooms create hypoxic or “dead” zones where oxygen levels are too low to support aquatic life, thereby harming or killing aquatic communities and disrupting their functions. Eutrophication also negatively affects local human populations by degrading drinking water quality and reducing aquatic recreation, habitat, and tourism (Femmer, 2012). Eutrophication, hypoxia, and algal blooms persist in the Nation’s waters (U.S. Environmental Protection Agency, 2017). Of the surveyed stream and river miles in the United States, 46 percent were impaired by excess phosphorus and 41 percent were impaired by excess nitrogen. Also, approximately 40 percent of U.S. lakes are in a “most disturbed” condition because of excess phosphorus. These impairments have been linked to increased anthropogenic nutrient inputs, mainly originating from nonpoint agricultural sources, in the late 20th and early 21st centuries (Brown and Froemke, 2012; Stephenson and others, 2022).

A consequence of excessive nutrients in streams and rivers is the formation of hypoxic or dead zones in Bays and Gulfs at the mouths of large nutrient-rich rivers (including the Mississippi River), which have spread in the coastal oceans during the last five decades (Rabalais and others, 2002, 2010). In the Mississippi River Basin, the task of reducing nutrient loads to the Mississippi River has mostly fallen to individual State nutrient reduction strategies aided by Federal funding (U.S. Environmental Protection Agency, 2014). With respect to nonpoint nutrient sources, State nutrient reduction strategies and Federal programs have focused on funding voluntary implementation of conservation practices by individual farmers and landowners (Kirk and others, 2024). To improve the effectiveness of these nutrient reduction strategies, critical source areas, defined as locations that deliver a disproportionate amount of pollutant, have been targeted for conservation practice implementation (Osmond and others, 2012). To this end, the U.S. Department of Agriculture Natural Resources Conservation Service (NRCS) started a conservation landscape initiative in 2010 known as the Mississippi River Basin Healthy Watersheds Initiative (MRBI; Natural Resources Conservation Service, 2023). The MRBI offers financial and technical assistance to agricultural producers in priority drainage basins within the Mississippi River Basin, including the Lower Grand River hydrologic unit (fig. 1; hereafter referred to as the “Lower Grand”). This assistance focused on implementing voluntary conservation practices that avoid, control, and trap nutrients released from working agricultural lands within the Mississippi River drainage basin (Natural Resources Conservation Service, 2023). The primary goal of the MRBI has been to improve water quality while ensuring the economic viability of agricultural lands. Secondary goals include restoration of wetlands and wildlife habitat.

Predicting the ability of management activities to improve water quality before implementation is difficult. First, the effectiveness of agricultural conservation practices in reducing nutrient and sediment loads can vary widely (Liu and others, 2017; Webber and others, 2024), especially at the drainage basin scale (Tomer and Locke, 2011). Second, conservation practices can have compounding or counteracting effects. For example, terraces (NRCS Conservation Practice Standard, code 600) have the potential to moderately reduce nutrient transport to surface waters. However, terraces are often paired with underground outlets (NRCS Conservation Practice Standard, code 620), which have the potential to increase nutrient transport to surface waters and negate the nutrient reduction benefit of the terrace conservation practice (Natural Resources Conservation Service, 2024). Third, the release of legacy nutrients, which are nutrients stored in groundwater and soils, can make the effects of conservation practices on water quality difficult to detect for years or decades after implementation (Van Meter and Basu, 2015). Finally, changing climate may accelerate nutrient losses despite increased conservation practices because of increasing extreme precipitation events that increase the amount of nutrients leached from the soil (Sinha and others, 2017; Lu and others, 2020). Monitoring studies quantifying changes in stream nutrient concentrations resulting from implementation of conservation practices help estimate conservation practice effectiveness at the drainage basin scale.

Previous Investigations

In 2010, the U.S. Geological Survey (USGS) and the Missouri Department of Natural Resources began a cooperative study to evaluate temporal changes in total nitrogen (TN) and total phosphorus (TP) concentrations at selected sites within the Lower Grand (Krempa and Flickinger, 2017). The study objectives were to quantify temporal changes in TN and TP concentrations, then compare those concentration changes among sites and relative to the acres of conservation practices implemented.

At the six study sites within the Lower Grand (hereafter referred to as “MRBI sites”), annual mean TN concentrations (referred to as “flow-weighted concentrations” in Krempa and Flickinger [2017]) from 2011 to 2015 ranged from 0.65 to 2.8 milligrams per liter (mg/L). During the same period, annual mean TP concentrations ranged from 0.09 to 0.40 mg/L. In addition to the annual means, flow-adjusted (FA) concentrations were used to qualitatively assess temporal changes in nutrient sources. Visual inspection of FATN and FATP concentrations between 2011 and 2015 revealed stable FATN and FATP concentrations, which indicated that the sources of stream nutrients had not changed during the study period.

At selected Missouri River tributary sites and Missouri River sites upstream and downstream from the confluence with the Grand River (referred to as “long-term sites”), greater decreases in flow-normalized (FN) TN and TP concentrations were observed at the nearby Missouri River tributary sites than at the Grand River site despite the increase in conservation practices in the Lower Grand from 2011 to 2015. A regression analysis indicated weak relations between the acres of conservation practice acres in the regional drainage basins and annual FNTN and FNTP concentrations.

Purpose and Scope

This report describes temporal changes in stream TN and TP concentrations within the Lower Grand and throughout northwest Missouri and southern Iowa focused on the period from October 1, 2010, through September 30, 2023, or water years 2011–23. Within the Lower Grand, temporal changes in TN and TP concentrations were used to evaluate the effects of the MRBI on water quality. Additionally, temporal changes in stream TN and TP concentrations in the regional drainage basins widened the scope of this report to document broader nutrient trends across northwest Missouri and southern Iowa during this period. This report updates the work completed by Krempa and Flickinger (2017).

Site Description

This study focuses on the Lower Grand drainage basin in Missouri and Iowa (fig. 1). The Lower Grand is a subbasin within the Grand River drainage basin (site 9 [USGS streamgage 06902000; Grand River near Sumner, Missouri]), and the Grand River is a tributary of the Missouri River (sites 8 [USGS streamgage 06818000; Missouri River at St. Joseph, Mo.] and 11 [USGS streamgage 06934500; Missouri River at Hermann, Mo.]; fig. 1, table 1). In 2010, the Lower Grand was designated as an MRBI focus area. As a focus area, the Lower Grand received increased funding for conservation practices and water quality monitoring facilitated through the MRBI and its partners (Natural Resources Conservation Service, 2023). Within the Lower Grand, six water-quality sites were established on tributaries in 2010 and labeled as MRBI sites (Krempa and Flickinger, 2017). Of these six MRBI sites, three were selected for continued monitoring in 2019 (sites 1 [USGS streamgage 06899900; Medicine Creek at Lucerne, Mo.], 2 [USGS streamgage 06900050; Medicine Creek near Laredo, Mo.], and 5 [USGS streamgage 06901500; Locust Creek near Linneus, Mo.]; fig. 1, table 1). The drainage areas of these study sites range from 305 to 1,435 square kilometers, and agriculture is the primary land use within each site’s drainage basin (table 1; Krempa and Flickinger, 2017).

Additional long-term monitoring sites included one site on the Grand River (site 9), two additional Missouri River tributary sites (sites 7 [USGS streamgage 06817700; Nodaway River near Graham, Mo.] and 10 [USGS streamgage 06905500; Chariton River near Prairie Hill, Mo.]) and two Missouri River sites, one upstream (site 8) and one downstream (site 11) from the confluence of the Grand and Missouri Rivers (fig. 1, table 1). All long-term USGS monitoring sites were also included in the previous publication by Krempa and Flickinger (2017). Although the Grand River drainage basin includes the Lower Grand, the Grand River site (site 9) could not be directly linked to long-term water-quality changes in the Lower Grand because no long-term sites existed on the Middle or Upper Grand River. Instead, changes in TN and TP concentrations at this site could only reflect potential effect of the MRBI in the Lower Grand on water quality in the broader Grand River Basin upstream of the USGS streamgage.

The two additional Missouri River tributary sites (sites 7 and 10) were of similar size and land use to the Grand River but were not part of NRCS priority drainage basins. As such, they could act as reference sites to help determine if water-quality changes in the Grand River were unique in the region. Finally, differences in temporal changes of water quality between the Missouri River sites upstream and downstream of the Grand River confluence could potentially indicate if the region was reducing its nutrient export.

Streamflow and Water-Quality Data Collection

USGS streamgages were colocated with water-quality collection sites. Stage data were collected every 15 minutes using nonsubmersible pressure transducers and uploaded to the USGS National Water Information System (NWIS) database (U.S. Geological Survey, 2024). Routine streamflow measurements were made to develop and maintain a stage-streamflow relation for each site, which were used to estimate streamflows from stage data (Turnipseed and Sauer, 2010). Daily mean streamflows were computed from the 15-minute streamflows and stored in NWIS.

Water-quality samples were collected and processed using standard equal-width increment collection methods representative of the entire water column (U.S. Geological Survey, variously dated) and analyzed at the USGS National Water Quality Laboratory according to published USGS laboratory methods in use during the time of sample collection and analysis (Patton and Truitt, 1992, 2000; Fishman, 1993; U.S. Environmental Protection Agency, 1993; Fishman and others, 1994; Patton and Kryskalla, 2003, 2011). Concentrations of nitrate plus nitrite (dissolved or total), ammonia plus organic nitrogen, TN, and TP were obtained from the USGS NWIS database (U.S. Geological Survey, 2024). If a date had more than one sample, only the water-quality sample flagged as a regular sample in the NWIS database (sample type code=9) was kept. At the MRBI sites, TN concentrations were pulled directly from the NWIS database using parameter code 00600. For the long-term sites, TN was defined as the sum of nitrate plus nitrite (dissolved or total) and ammonia plus organic nitrogen (total Kjeldahl nitrogen), and only samples with concentrations for both nitrogen constituents were used. Measured nitrogen and phosphorus concentrations from water-quality samples were paired with daily mean streamflow from colocated streamgages to estimate annual TN and TP concentrations.

Methods for Assessing Temporal Changes in Nutrient Concentrations

The primary study objective was to quantify temporal changes in TN and TP concentrations in the Lower Grand and nearby drainage basins. At the MRBI sites, annual mean TN and TP concentrations were estimated to provide a quantitative metric to track temporal changes in the “actual” nutrient concentrations; however, trends in nutrient concentrations can be obscured by interannual flow variability (Rowland and others, 2021). To improve the trend signal, the effect of flow variability was reduced through additional analysis.

Methods to reduce the effect of flow variability include calculating FA and FN concentrations. As of this publication, the best available method to reduce the effect of interannual flow variability is an extension of the Weighted Regressions on Time, Discharge, and Season (WRTDS) model known as flow normalization (Rowland and others, 2021). The WRTDS model generally requires at least 20 years of data and more than 200 samples collected at the sampling site (Hirsch and others, 2010). Furthermore, the daily time step of the WRTDS model makes it unsuitable for small, flashy rivers that can pass most of the streamflow associated with a storm event in less than one day. These requirements make flow normalization using the WRTDS model acceptable for the long-term sites but not for the MRBI sites.

For the MRBI sites, FA concentrations were used as an alternative to FN concentrations to reduce the effect of streamflow on water-quality changes. FA concentrations are defined as the difference between measured sample concentrations and the predicted concentration using the relation with streamflow (White and others, 2004; Simpson and Haggard, 2018). Statistically, they are the model residuals of a concentration-streamflow model. Because the modeled concentration values were predicted from streamflow alone, the residuals from the streamflow-only model represent changes in concentrations not explained by streamflow variation. These residuals can be plotted over time to investigate changes in nutrient delivery to the stream or river regardless of changes in streamflow (Crawford and others, 1983; Heimann and others, 2011).

Overall, at the MRBI sites (sites 1, 2, and 5), annual mean TN and TP concentrations were used to evaluate the “actual” TN and TP concentrations, and FATN and FATP concentrations were used to evaluate trends in TN and TP sources from water year 2011 through water year 2023. At long-term sites (sites 7–11; fig. 1, table 1), FNTN and FNTP concentrations were used to reduce the effects of streamflow variability on estimated nutrient concentrations and evaluate temporal changes in TN and TP concentrations from water year 2010 through water year 2023. MRBI sites did not have complete datasets during water year 2010; therefore, the analysis started with water year 2011. Although the values produced from each analysis cannot be compared directly, the direction of the trends derived from the values can be compared. The combined results complemented each other and added depth to the analysis in the report. All three methods were used in the previous study (Krempa and Flickinger, 2017) and provide continuity to the analysis of water quality in the region.

Temporal Changes in Annual Mean Nutrient Concentrations at Mississippi River Basin Healthy Watersheds Initiative Sites

Annual mean TN concentrations were similar among the three MRBI sites, ranging from 0.31 to 2.13 mg/L; the minimum value was measured at site 2 and the maximum at site 5 (table 3). When ranked by mean value, annual TN concentrations were highest at site 5 and lowest at site 2. Annual mean TP concentrations were also similar across sites with values ranging from 0.08 to 0.66 mg/L. Site 5 generally had the highest annual mean TP concentrations, while site 1 had the lowest. As with TN, the minimum and maximum annual mean TP concentrations were recorded at sites 2 and 5, respectively. The ranges of annual mean TN and TP concentrations were consistent with those reported in the previous study (Krempa and Flickinger, 2017).

Annual mean TN and TP concentrations varied considerably from year to year, generally increasing as streamflow increased (figs. 2 and 3). This relation was apparent during the dry year of 2018, which coincided with the minimum concentrations, and the following wet year of 2019, which produced near maximum concentrations. From 2011 to 2023, neither annual mean TN concentrations nor annual mean TP concentrations had statistically significant trends (table 4). Slopes in annual mean TN concentrations ranged from −0.004 to 0.024 milligram per liter per year ([mg/L]/yr). Slopes in annual mean TP concentrations ranged from −0.002 to 0.004 (mg/L)/yr. These slopes indicated no meaningful increase or decrease in annual mean TN and TP concentrations at any of the MRBI sites; for example, the −0.004 (mg/L)/yr slope in the annual mean TN concentration at site 2 would equate to a 0.05-mg/L reduction at a site with a mean TN concentration of 1.08 mg/L over the 13 years of the study.

Temporal Changes in Flow-Adjusted Nutrient Concentrations at Mississippi River Basin Healthy Watersheds Initiative Sites

FATN and FATP concentrations showed that small to negligible trends in nutrient sources occurred at most sites (fig. 4, table 5). No sites had a significant trend in FATN concentrations, and only one site (site 2) had a significant trend in FATP concentrations. A positive trend in FATP concentrations was detected at site 2, which indicates that sources of TP were likely increasing at this site. Although not significant, the slopes of the linear regressions for FATN were negative at each site, which indicates some possible reduction in TN sources in the Lower Grand. Taken together, these results hint at a potential for increasing phosphorus transport in recent years (especially at site 2) and the potential for a minor decrease in TN concentrations, but they generally indicate that streamflow variability was the most substantial factor affecting temporal changes in TN and TP concentrations at the MRBI sites.

Temporal Changes in Annual Flow-Normalized Concentrations at Long-Term Sites

Visual inspection indicated temporal changes in annual FNTN concentrations were generally similar across the five long-term sites; however, the timing and strength of trends varied among sites (fig. 5A, B). Across sites 7–11, annual FNTN concentrations ranged from 1.23 to 4.26 mg/L (app. 1, table 1.4). In water year 2000, the sites ranked from highest to lowest annual FNTN concentrations as 7, 8, 11, 9, and 10, and this ranking did not change from water year 2000 to water year 2023 (fig. 5A, B). For water years 2000 through 2023, annual FNTN concentrations decreased at all sites (fig. 5A, B, table 6). During the 2010–23 period, the Missouri River sites (8 and 11) had the greatest percentage reductions in annual FNTN concentration (35 percent and 28 percent at sites 8 and 11, respectively). At sites 7, 8, and 11, the magnitude of change and the percentage of change in annual FNTN concentrations flipped from an increase during the 2000–10 period to a decrease during the 2010–23 period. At site 9, where nutrient trends would have the potential to be affected by the implemented conservation practices in the Lower Grand, a greater decrease in annual FNTN concentration was observed during the 2010–23 period (period containing the implementation of the MRBI) relative to the 2000–10 period (period prior to the MRBI). At site 10, annual FNTN concentrations declined after 2010, but this decline was of a lower magnitude than the decline during the 2000–10 period.

Unlike FNTN concentrations, temporal changes in annual FNTP concentrations varied across the five sites since 2000 (fig. 6A, B). Annual FNTP concentrations ranged from 0.28 to 0.65 mg/L among the sites (app. 1, table 1.5). The order of sites from highest to lowest FNTP concentration also varied over the period (fig. 6A, B). Similar to the FNTN results, site 7 had the highest FNTP concentration over the whole period from 2000 to 2023. During the 2010–23 period, four of the sites indicated a decline in annual FNTP concentration (table 6). The exception was at site 10, which had a 7-percent increase during 2010–23. Sites 8 and 9 had the two greatest percentage reductions since the MRBI began in 2010 (−34 percent and −26 percent at sites 8 and 9, respectively). At all sites, except site 10, the magnitude of change and the percentage of change in annual FNTP concentrations flipped from an increase from 2000 to 2010 to a decrease from 2010 to 2023. This change to declining annual FNTP concentrations indicates that progress is being made in reducing TP across the study area. Even site 10, which had increases in FNTP concentration during 2010–23, did not have a substantial increase in annual FNTP concentration over the entire study period (2000–23). At site 9, annual FNTP concentrations declined at a rate of 2 percent per year from 2010 to 2023, which corresponds to a −26-percent change in FNTP concentration (table 6). This decrease is in direct contrast to the 1.6-percent-per-year increase in annual FNTP concentrations before the MRBI.

Nutrient Dynamics Within the Lower Grand During the Mississippi River Basin Healthy Watersheds Initiative

The analysis of TN and TP concentrations provided two key insights for nutrient transport in the Lower Grand. First, the continued positive relation between annual mean TN and TP concentration and streamflow (fig. 3) further supported the initial conclusion from Krempa and Flickinger (2017) that most of the stream nitrogen and phosphorus are derived from nonpoint sources. Second, the generally stable FATN and FATP concentrations indicated that changes in nutrient delivery have not been detected in the Lower Grand regardless of streamflow conditions. Together, these results indicate that neither the magnitude nor the transport of nutrients in the Lower Grand has changed substantially since 2010.

Although the results indicate that nutrient concentrations in the Lower Grand have not changed substantially, the conservation practices implemented as part of the MRBI may still be reducing nutrients. Additional monitoring could inform whether there is an observable trend in nutrient reduction because of the MRBI. The noise associated with water-quality data because of variations in streamflow can make parsing out the long-term trend or signal in nutrient dynamics difficult without 15–20 years of data (refer to the data requirements for the commonly used WRTDS model [Hirsch and others, 2010]). Additionally, stream nutrient concentrations may respond slowly to changes in nutrient inputs because of a large storage of nutrients within the drainage basin resulting from decades of nutrient surpluses (Van Meter and Basu, 2015; Vero and others, 2018). As of 2024, the water-quality dataset for the MRBI sites includes 13 years of water-quality samples (minimal samples were collected in 2020 because of the global pandemic). Thus, the potential to begin quantifying a long-term trend with 2–7 more years of data exists. Furthermore, continuous sensors were installed at these sites in 2020 to allow for more accurate estimates of annual nutrient concentrations and loads moving forward.

Nutrient Dynamics in the Grand River and Similar Nearby Tributaries During the Mississippi River Basin Healthy Watersheds Initiative

Since 2010, negative trends in annual FNTN concentrations have been detected at all long-term sites (table 6). The Grand River (site 9), which includes the Lower Grand, had greater reductions in FNTN concentrations than the adjacent Chariton River (site 10) but lower reductions than the nearby Nodaway River (site 7). Furthermore, the Missouri River sites (sites 8 and 11) also had substantial declines in annual FNTN concentrations. These results indicate a regional reduction of TN concentrations in northwestern Missouri drainage basins and that the Grand River generally follows this broader trend. For annual FNTP concentration trends, the Grand River had slightly greater reductions than the Nodaway River and much greater reductions than the Chariton River—a site that had an increase in annual FNTP concentration. In 2023, the annual FNTP concentration in the Grand River, which had declined to 0.31 mg/L, was now lower than the annual FNTP concentration in the Chariton River, which had increased to 0.33 mg/L in 2023. Similar to annual FNTN concentrations, the Missouri River sites indicated substantial declines in annual FNTP concentration since 2010, indicating a regional reduction in TP concentrations.

As previously mentioned, the drainage area to site 9 includes the Grand River drainage basin, not only the Lower Grand area included in the MRBI. The nutrient concentration reductions in the long-term trends of site 9 may hint at improved water quality because of the implementation of conservation practices in the basin. However, similar trends in the Nodaway and Missouri Rivers indicated that the long-term reduction of stream TN and TP detected in the Grand River was not limited to the Grand River. Additionally, the stable or increasing stream nutrient concentrations at the MRBI sites indicated that the conservation practices implemented under the MRBI may not have been a major factor in the decline of TN and TP at site 9. It is unknown how much of the stream nutrient concentration reductions at site 9 were due to conservation practices versus broader changes in agricultural practices in northwestern Missouri.

Summary

This study evaluated temporal changes in total nitrogen (TN) and total phosphorus (TP) concentrations at selected sites in the Lower Grand and regional sites across northwestern Missouri for water years 2010 through 2023. At the three Mississippi River Basin Healthy Watersheds Initiative (MRBI) sites in the Lower Grand, annual mean concentrations and flow-adjusted concentrations were used to assess nutrient concentration variability and potential changes in nutrient sources. At the six long-term U.S. Geological Survey (USGS) monitoring sites, flow-normalized (FN) concentrations were estimated using the Weighted Regressions of Time, Discharge, and Season method to evaluate nutrient trends before and during the MRBI period.

At MRBI sites, annual mean TN and TP concentrations were positively related to streamflow and flow-adjusted TN and flow-adjusted TP concentrations were stable to slightly increasing. These results indicate that the source, magnitude, and transport of nutrients in the Lower Grand have not changed substantially since 2010. Continued monitoring may help to more definitively assess long-term trends in nutrient concentrations within the Lower Grand River hydrologic unit. At the Grand River long-term site (site 9; USGS streamgage 06902000; Grand River near Sumner, Missouri), annual FNTN and FNTP concentrations declined from 2010 to 2023, with the rate of this decline increasing since the start of the MRBI. These reductions were greater than those at the adjacent Chariton River (site 10; USGS streamgage 06905500; Chariton River near Prairie Hill, Mo.), which was not part of the MRBI. However, similar reductions in nutrient concentrations were observed at the other long-term sites during this period, including sites 7 (USGS streamgage 06817700; Nodaway River near Graham, Mo.), 8 (USGS streamgage 06818000; Missouri River at St. Joseph, Mo.), and 11 (USGS streamgage 06934500; Missouri River at Hermann, Mo.). The similar reductions in FNTN and FNTP concentrations across long-term sites, combined with the relatively minor changes in nutrient concentrations observed at the MRBI sites, suggest that while conservation practices in the Lower Grand may be contributing to nutrient reductions, broader regional factors may also be affecting nutrient trends in the Grand River. Continued monitoring may help to quantitatively determine trends in stream nutrient concentrations at the MRBI sites.