Abstract
Archived water quality data collected between 1901 and 2019 were used to reconstruct annual averages of various forms of C, N, P, and silicate concentrations and alkalinity in the lower Mississippi River. During this interval the average annual nitrate concentrations doubled pre-dominantly from fertilizer applications and tiling, silicate concentrations decreased by half as diatom sedimentation increased as dams were built, and alkalinity increased 16%. Variances in silicate concentrations were proportional to river discharge before 1980 and concentrations have been stable since then. Average annual temperatures, discharge and alkalinity increased simultaneously around 1980; this suggests that there was greater weathering thereafter and is supported by the positive relationships between variations in alkalinity and variations in nitrate, phosphate, and silicate concentrations. The conversion of forests and grasslands into farmlands and improved drainage resulted in less evapotranspiration, a higher percent of precipitation going into streams and altered soil water bio-geo-chemistries. Field trials demonstrating soil health improvements resulting from more live roots and soil cover and greater biodiversity demonstrate water quality improvements and no effect on farm profitability. Lowering nitrate loading to the coastal waters will reduce summertime hypoxic waters formation offshore, but alkalinity in the river will increase further with climate warming.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
“In every respect, the valley rules the stream”
Hynes (1960)
Introduction
Water quality impairments in the world’s rivers noticeably declined in the last century with chemical and microbial contamination, and pH declines (Meybeck, 2003), including in the United States (US) where 200 years of colonization and population growth resulted in plowed fields replacing forests, impervious surfaces covering 6.4% of the land, and 90,000 dams being built (https://nid.sec.usace.army.mil/ords/f?p=105:22:4482115045565::NO; Turner & Rabalais, 2003; Stets & Striegl, 2012; Homer et al., 2015; Yin et al., 2023). Agricultural land uses influenced nitrate concentrations in the Mississippi River watershed by the early 1900s—a half century before the use of artificial fertilizer (Broussard & Turner, 2009) and waterways became both a recipient of waste and a drinking water source. The initial growth in farmed acreage in the contiguous states plateaued at 1.4–1.5 million km2 between 1960 and 1980, fewer new dams were constructed thereafter, and the growth in nitrogen and phosphorus fertilizer use slowed (Fig. S1). The nitrogen and phosphorus sources in this watershed today are geographically centered in the US midwestern states (Alexander et al., 2008) where intense fertilization and drainage of corn and soybean fields are located (https://nassgeodata.gmu.edu/CropScape/; Jaynes & James, 2007; Figs. S2, S3, and S4).
Some water quality improvements occurred by the late 1900s as sewerage treatment expanded after the 1972 Clean Water Act and related legislation was implemented; bacterial densities and lead concentrations declined, oxygen saturation rose, and pH increased from a monthly low of 5.8 in 1965 to 8.2 in 2019 (Turner, 2021). However, the concentrations of other constituents increased and by the 1980s the nitrogen loading from the Mississippi River caused the formation of the largest hypoxic zone in the western Atlantic Ocean; its size in late July/early August is predicted by the May nitrogen loading from the Mississippi River (Turner et al., 2012; Scavia et al., 2017).
A clearer understanding of the relationships among land, water, atmosphere, climate and a people’s history is coming into focus as this watershed has become dominated by human-centered enterprises. New and future water quality stressors include global scale changes affecting water availability, temperature, storm frequency and intensity, and flooding, etc., driven by rising concentrations of greenhouse gases (GHGs) in the atmosphere (IPCC, 2021). Increasing soil carbon this century is a seen as a significant modulator of global climate changes (Six et al., 2004; Bai & Cotrufo, 2022; Chabbi et al., 2022). The interplay of forces within and between ecosystems are becoming recognized more explicitly, sometimes addressed in parts, and now considered more often as existing within a panoply of conditions, including acknowledgment of unknowns.
Here I ask how water quality changed over the last one hundred and twenty years of agricultural intensification in the Mississippi River watershed and then focus on the last few decades of relatively stable land use and a changing climate. It is the largest watershed on the North American continent, drains 41% of the conterminous US, and brings 80% of the river water entering the Gulf of Mexico from the US, and 91% and 88% of the nitrogen and phosphorus, respectively (Dunn, 1996). The USGS estimate of land cover in the watershed in 2001 is based on land use/land cover classifications systems designed specifically for use with remotely sensed imagery (https://edna.usgs.gov/watersheds/ws_chars.php?title=Mississippi&name=mississippi). These data showed that landcover in the watershed in 2001 was approximately 36.2% cropland and pasture, 24.9% grassland and sedges, 21.9% forest, 6.8% shrubland, 5.3% urban, 4.5% water, and 0.3% barren (bare rock/sand/clay or quarries). The watershed has a robust water quality data base, a capable research community, and consequential economic and social capital. I use archived water quality data collected from 1901 to 2019 from Federal and State agencies, the New Orleans Sewerage and Water Board (NOSWB), and universities to reconstruct a 100+ year record of some indicators of water quality in the lower Mississippi River. The analytical focus is on the alkalinity, and the concentration of various forms of carbon, nitrogen, phosphorus, and silica. It concludes with a discussion of national policies and future possibilities for agriculture and climate change adaptations.
Methods
Water quality archival data
Water quality data are from five locations on the southern end of the Mississippi River at St. Francisville, Baton Rouge, two locations in New Orleans, and at Belle Chasse, LA, located 428, 370, 167, 153, and 122 km, respectively, upstream from the Head-of-Passes where the river divides into three main outlets to the Gulf of Mexico (Fig. 1). The two major metropolitan areas are Baton Rouge and New Orleans (ca. 830 thousand and 1270 thousand people in 2020, respectively). The Belle Chasse station is associated with a ferry crossing, and the New Orleans samples are from river water intake pipes at the Carrollton and Algiers water treatment plants bringing drinking water to the New Orleans area.
The Louisiana Department of Environmental Quality (LaDEQ) sampling occurred at two locations: St. Francisville (LaDEQ stations 9, 55, 318, and 4031), and Belle Chase (LaDEQ stations 51, 52, and 320). The LaDEQ stations were sampled from 1966 to the present.
The United States Geological Survey (USGS) reported water quality data at St. Francisville, LA, in Water Supply Papers (WSP) from 1954 to 1998. These WSP data begin before the LaDEQ data collections started in 1968 and overlap with the LaDEQ data. The USGS provided a flow-averaged estimate of monthly loadings and concentrations for the Mississippi River from 1968 to 2015 using a Loadset formulation (Runkel et al., 2004); these data no longer exist on the web but are in a data repository (Table 1). The monthly estimates of loading in each calendar year were summed to calculate the annual load, and the concentration calculated by dividing loads by discharge. A newer interpretation of the same USGS data are for 1968 through 2021 using a flow-normalized concentration and loads described by Lee (2022; https://www.sciencebase.gov/catalog/item/629e0d14d34ec53d276f6960). This more recently developed flow-normalized method adjusts for variations of concentration with discharge among many years and infrequent sampling using weighted regressions on time, discharge, and season (WRTDS) with Kalman filtering (WRTDS–K) methods (Lee et al., 2017). These estimates use multiple years to compute discharge-concentrations relationships and so there is a reduction in the variation from 1 year to the next (Fig. S5). For this reason, the WRTDS flow-normalized data were not used when trends in the annual variations were being investigated, but were used in the discussion of reductions in nitrate loading to meet the goals of the Hypoxia Action Plan of 2001 (Mississippi River/Gulf of Mexico Watershed Nutrient Task Force, 2001) because that model’s output is used as a metric of success.
The water quality instruments, and analytical methods used for samples collected at Baton Rouge, LA, from 1996 to 2018, are described in Turner et al. (2019). Briefly, water samples were collected at least once monthly, and up to five times per month during spring and summer. Unfiltered water samples were frozen until determination of dissolved forms of nitrogen (N), phosphorus (P), and silicate (DSi) using either a Technicon Autoanalyzer II (USEPA Method 353.2 for ammonia and nitrate/nitrite (DIN), USEPA Method 365.2 for phosphate (DIP), and Technicon Method 186-72W/B for DSi) or a Lachat Quick-Chem 8000 Flow Injection Analyzer using the Lachat Methods approved by USEPA: method 31-107-06-1-B for ammonium, method 31-107-04-1-C for nitrate/nitrite, method 31-115-01-1-H for phosphate (DIP), and method 31-114-27-1-C for DSi. A 5-point standard curve was used and QC standards were analyzed before, during and after each set of samples analyzed. Total nitrogen (TN) and total phosphorus (TP) concentrations were measured using a Technicon Autoanalyzer II or LaChat Quick-Chem after persulfate wet oxidation digestion (Raimbault et al., 1999). The concentration of total carbon (TC) was measured using a Shimadzu® TOC-5000A Analyzer. Total organic carbon (TOC) was measured by acidifying samples with HCl and then sparging before analysis to remove the inorganic carbon (IC). The Coefficient of Determination for the standard curve was > 0.98 for all nutrient analyses. The data are at https://doi.org/10.5061/dryad.x95x69pkm.
Miscellaneous other water quality data used
There are four data sources for nitrate concentrations and one of silicate concentrations that were made in the early 1900s. McHargue & Peter (1921) measured nitrate in one sample from the Mississippi River at Baton Rouge in spring, 1915. Wiebe (1931) collected 15 surface water samples from the river at Baton Rouge in 1930. Wiebe reported minimum and maximum nitrate concentrations of 0.82 and 14.74 µmol l−1, respectively, and the average value was used here (7.78 µmol l−1 nitrate). Dole (1909) report water quality measurements of dissolved silicate and nitrate concentrations for 1905 and 1906 at the New Orleans water intake pipe. The NOSWB (Sewerage and Water Board, 1903) made measurements of nitrate in 1901 at the same intake pipe.
Normalized values of silicate, nitrate, phosphate, and alkalinity
A 3-year running average of the concentrations for nitrate and silicate from the USGS data (1974–2015) was supplemented by samples collected in Baton Rouge from 2016 to 2018, inclusive. The alkalinity data were compared to a similar calculation for alkalinity using the NOSWB alkalinity data and LaDEQ data for 2006 to 2015. The values were then converted to a proportion of the average value over the entire record so that the normalization yielded an average value = 1.0. The monthly average bicarbonate concentration (mg C l−1; µmol l−1 ± 1 SE) was converted to carbon as described by Raymond et al. (2008). They calculated that the inorganic carbon represented in alkalinity titrations is mostly from bicarbonate, with little contribution from carbonate in the well buffered Mississippi River. Raymond et al. (2008) used temperature and pH measurements from the Mississippi River to estimate that on average ~ 93% of total dissolved inorganic carbon (CO2, HCO3−, CO3−2) is in the form of HCO3−, with the remaining 7% being mostly CO2. The bicarbonate flux, therefore, captures the majority of the fluvial export of DIC that is not evaded to the atmosphere.
River discharge
The annual discharges of the Mississippi River at Vicksburg, MS, from 1900 to 2019 are for daily measurements made by the United States Army Corps of Engineers (USACE) but are found in two different report series. I used two sources of the daily river discharge data that have overlapping data records with annual data. The annual discharge data at Vicksburg extends from 1817 to 2021 and are in the Mississippi River Commission (1955). These data were compared to the second dataset derived from the daily discharge records available at the USGS website (https://waterdata.usgs.gov) under the State tab for ‘current conditions,’ then ‘daily discharges,’ and then ‘time series’ tabs. The data are in Turner (2022; https://doi.org/10.5061/dryad.1jwstqjzb) and used to plot discharge versus silicate concentrations.
Suspended sediments
Meade & Moody (2010) report results from isokinetic point sampling at five depths along each of 5 to 8 verticals across the Mississippi River at Tarbert Landing that were paired with sediment-discharge values every 2 weeks from calendar year 1950 to 2007. Daily discharge values of suspended sediments were used to determine annual fluxes. The annual tons were normalized to the highest concentration year (1950). The data were used to test the hypothesis that impoundment influences silicate retention as a result of improvements in light conditions and longer residence time that favors phytoplankton that sink behind dams to sequester the silica in diatom frustrules (Humborg et al., 2000).
Annual temperatures
The United States Environmental Protection Agency website ‘Climate Change Indicators’ https://www.epa.gov/climate-indicators) has the average annual air temperatures in the 48 contiguous States and the deviations from the long-term average from 1901 to 2020 (anomalies). The source data is from weather station and updated in July 2022. Details about the data analysis are in USGCRP (2017).
Summary of data set sources
The data set (Table 1) includes alkalinity, inorganic nitrate, ortho-phosphate (phosphate), silicate (DSi), total phosphorus (TP), total nitrogen (TN), inorganic and organic carbon, suspended sediment, temperature, and discharge. Some samplings resulted in multiple data collections for each month and some months had only one sample. Multiple values for 1 month were combined and an average value for the 12 months made to determine an average and standard error for each year.
Statistics
I used Prism Version 10.0.0 (131), June 13, 2023 software for Mac (GraphPad Software, Boston, Massachusetts USA, www.graphpad.com for statistical analyses where significance P = 0.05 and to compute an annual average and standard error. Regression slopes for different intervals tested the null hypothesis that the slopes were identical (the lines are parallel) by comparing slopes to calculate a P value (two-tailed test) determining if the chance that randomly selected data points had slopes that were different. Log transforms were made for graphing purposes. Data were fit to a spline/Lowess analysis with three knots to identify the inflection point where the discharge rates went from a declining to an increasing rate.
Results
Discharge and alkalinity
The river discharge at Vicksburg, MS, was not correlated with year from 1900 to 1979 but significantly increased with year from 1980 to 2019 [Discharge (km3 year−1) = 4.07*Year − 7521 (R2 = 0.14, F = 5.98, P = 0.02)]. The slopes for before and after 1980 were different (Fig. 2A; F = 3.93, DFn = 1, DFd = 116; P = 0.0498) (Fig. 2A).
The average annual temperature anomaly in the lower 48 States was not correlated with year from 1900 to 1979 but was significantly related with year from 1980 to 2019 [Anomaly (°C) = 0.0256*X − 50.71 (R2 = 0.33, F = 19.7, P < 0.001)]. The slopes for before and after 1980 were different (F = 16, 5, DFn = 1, DFd = 117; P < 0.001) (Fig. 2B).
The simple linear regression of the bicarbonate C concentration (mg l−1) versus year was also not significant for the years 1900 to 1979 but was from 1980 to 2020: Concentration = 0.0849–140.5 (R2 = 0.16, F = 7.17, P = 0.01). The slopes for bicarbonate C concentration for before and after 1980 were different (F = 6.51, DFn = 1, DFd = 108; P = 0.01) (Fig. 2C). Alkalinity increased 16% from 1900 to 1920.
The annual bicarbonate flux (TgC year−1) increased slightly from 1900 to 1979 (Y = 0.0346*year − 55.8; F = 5.36, DFd = 70, P = 0.02), and even more so after 1979 (Y = 0.1477*year − 278.8; F = 12.1, DFd = 38; P = 0.02) and there was a difference in the two slopes (F = 7.32, DFn = 1, DFd = 108; P < 0.01). The averaged bicarbonate C flux was 41% higher in 2019 than in 1980 (Fig. 2D).
Similarities among different observations
There is good agreement, in general, between the annual average annual concentrations of DSi, nitrate, and phosphate calculated from the different data sources for the same analyte when compared to the flow-normalized values calculated by the USGS (Fig. 3). The concentrations of nitrate and silicate (Fig. 3A, B), in particular, had an R2 value of 0.99 and small intercept, but the R2 = 0.29 for annual phosphate concentrations at Baton Rouge, LA. The silicate:nitrate ratios are in agreement with the USGS values, but there were two values from the NOSWB that appear as outliers (Fig. 3D). The annual total phosphate concentrations for samples collected at New Orleans by the LaDEQ and at St. Francisville by the USGS were not significantly related to concentrations calculated using the flow-normalized USGS estimates, whereas the concentrations at Baton Rouge were significantly related to the flow-normalized USGS estimates. The concentrations of total nitrogen at Baton Rouge were 11% lower than the concentrations determined using the flow-normalized USGS data (Fig. 3F).
The concentration of DSi increased slightly from 1901 to 1960, decreased after 1960 and became relatively stable by 1980 at about half of the concentration in the beginning of the 1900s (Fig. 4A). The nitrate concentration increased from the beginning of the twentieth century until the late 1950s, but then declined (Fig. 4B) before rising to a value from 2009 to 2018 that was 9 times greater than the 1901 value at New Orleans. There were no data on phosphate before the 1980s (Fig. 4C). By the late 1980s the phosphate concentration dropped to one third of the value in 1980, and then rose slightly between 1998 and 2019. The ratio of DSi:nitrate changed from 14:1 to 1:1 over the last 100 years and the dissolved nitrate:phosphate molar ratio was above 30:1 in the last decade (Fig. 4D).
The fluxes of nitrate and silicate were proportional to discharge over different intervals (Fig. 5). There were fewer data points for before 1950 than later, but the distinctions are evident. The slope of the nitrate flux versus discharge rose from before 1950, from 1950 to 1979, and then again after 1980 with significant differences between slopes (P = 0.001, F = 12.4; Fig. 5A). The equations for each are: before 1950, Y = 0.061*discharge − 15.46 (F = 7.53, DFd = 5; P < 0.05); from 1950 to 1980, Y = 0.066*discharge − 5.54 (F = 16.7, DFd = 23; P ≤ 0.001); after 1980, Y = 0.125*discharge − 15.5 (F = 123, DFd = 40; P < 0.001). The slope of the silicate flux versus discharge fell from before 1967 to after 1967 from 0.17 to 0.11 Gmoles year−1, respectively, with significant differences between them (P = 0.02, F = 5.6; Fig. 5B). The equations for each are: before 1967, Y = 0.169*discharge − 6.86 (F = 34.7, DFd = 13; P < 0.0001); after 1967, Y = 0.115*discharge − 4.97 (F = 209, DFd = 50; P < 0.001).
The sediment flux before and after 1967 is coincidental with the changes in silicate concentration (Fig. 6). Furthermore, the minimum values for the spline-modeled curves were similar for both—between 1987–1990 and 1989 for suspended sediment flux and silicate concentrations, respectively.
Alkalinity tends to increase each year, but there is also variance around this general upward trend. The normalized values of this variance in alkalinity changes with the normalized values of nitrate, silicate, and phosphate concentration (Fig. 7), but only after 1994, not before 1994. The slope of the normalized values for nitrate vs. alkalinity and silicate vs. alkalinity from 1995 to 2019 were significant (Fig. 7B, F = 50.9, DFd = 22, P < 0.01, R2 = 0.70; Fig. 7D, F = 20.4, DFd = 22, P < 0.01, R2 = 0.46; Fig. 7F = (Fig. 6F, F = 18.8, DFd = 18, P < 0.01, R2 = 0.51). The coefficient of determination (R2) between the normalized values for nitrate and silicate was 0.24 (F = 11.6, P < 0.01; not shown).
Total C, N, P
The concentrations of total nitrogen (Fig. 8A) and total phosphorus (Fig. 8B) were relatively constant after 1990. Before 1990 the TN concentration rose but the concentration of TP did not. The same yearly trend in TN was exhibited by nitrate; a simple linear regression had an R2 of 0.72 (P < 0.001; F = 519; not shown), indicating a common driver of the variance. The average TN concentration from 2003 to 2015 at Baton Rouge was 132 ± 4.9 µmols l−1 N and 4.6 ± 0.2 µmol l−1 P (molar ratio = N:P::28.7:1). The data for the concentration of total carbon (TC) and both inorganic carbon and total organic carbon (TOC) for the same interval indicate, in contrast, increases over the time series (Fig. 8C). The average annual concentration of inorganic carbon (2.39 mmol l−1) was 86.7% of the TC (2.75 mmol l−1) from 2003 to 2015. The average concentration of TOC over the same interval (0.37 mmol l−1) was 13.5% of the TC and increased at a faster rate than the TC concentration at 0.039 mmol l−1 year−1 from 2003 to 2015 (Y = 0.0385*year − 74.4; F = 196.4, DFd = 531; P < 0.0001). The molar ratios of the total amounts for this interval was 524:29:1::C:N:P.
Discussion
The water quality measurements discussed here are from the lower end of the largest river in North America; they are the summed consequences of varying loadings and processing within different channel sizes that have varying relationships within a heterogenous landscape, including lagged responses (Murphy et al., 2014). The channel sizes have changed over the last 100 years with wetland losses, deforestation, urbanization, flood control levee growth, and tiling, etc. (Paul & Meyer, 2001; Allan, 2004; Julian et al., 2015). As a result, the water quality at New Orleans does not represent a homogenous water quality found throughout the basin now or 100 years ago. Compared to large channels, for example, headwater streams have greater total channel length contact with the landscape (Wollheim et al., 2022), greater heterotrophy, and are shaded and shallow (Gardner & Doyle, 2018) so that the benthic to water column dominance is larger than in downstream channels (Reisinger et al., 2015). Water is retained within wetlands that release dissolved organic matter (DOM) and denitrify nitrogen that decreases rapidly with channel size (Alexander et al., 2000; Alvarez-Cobelas et al., 2008). Also, downstream channels have higher DOM concentrations than upstream, but phosphorus and silica concentrations are reduced by up to 50% downstream compared to in headwaters (Finlay et al., 2011). Climate change will increase the number of rarer but large precipitation events that are a significant source of terrestrial DOM to riverine systems (Raymond et al., 2016). The result is that land use, hydrologic alterations, drainage improvements and climate change from the headwaters to the Gulf of Mexico have changed water quality at the lower end of the River but have had different spatial and temporal effects within the watershed.
Discharge
The discharge of the lower end of the Mississippi River varied with changes in global weather patterns from 1826 to 1969, but not afterward when it appears that drainage improvements and climate change are the dominating drivers (Turner, 2022). Schilling & Libra (2003) provide an example of these effects by examining changing stream discharges in eleven HUC8 watersheds in Iowa. They found that converting habitats to agricultural fields replaces deeply rooted perennial plants with shallow-rooted annuals, (principally corn and soybeans today) which reduces evapotranspiration and shunts more water into streams that then increases the total amount of water discharge and the percentage of runoff as baseflow (Schilling & Libra, 2003; Schilling et al., 2008). These have such a significant effect that Xu et al. (2013) concluded that land use change contributed twice as much as climate change to stream discharge increases in 55 unregulated streams in the Midwest from the 1930s to 2010. The reason for the inflection point in discharge around 1980 is coincidental with the acceleration in temperature, but drainage improvements could also have been a factor, as well as an increase in precipitation (Peterson et al., 2013). The records for tiling fields, unfortunately, do not record the drainage intensity or diverse equipment. A recent survey, perhaps the most comprehensive one, showed that 63% of the counties in Iowa had drainage improvements amounting to between 20 and 60% between 1969 and 2017 (Edwards & Thurman, 2022). But the time between estimates are too long to determine if tiling activities distinctively increased around 1980. Data demonstrating that there was a dramatic change in drainage improvements around 1980 is lacking and seems improbable given how slowly drainage improvement policies are implemented (Jaynes & James, 2007).
Alkalinity
Alkalinity is the acid-neutralizing capacity of water and primarily a result of mineral weathering dependent on exposure to carbonic, sulfuric, and nitric acids, and nitric and sulfur oxides (Raymond & Hamilton, 2018). Because the conversion of one mole of ammonium to nitrite and then nitrate produces two moles of H+, an ammonium-based fertilizer becomes a weathering agent for carbonate and silicate minerals when the pH is > 6.5, which it was in the river after 1969 (Turner, 2021). The nitrogen in ammonium fertilizer applied in the US has been comprised of about 90% ammonium since the 1990s (Lu et al., 2018). A rise in alkalinity, therefore, may be accompanied with proportional increases in the concentration of nitrates (Jarvie et al., 1997). Other sources of HCO3− will be added when weatherable soils are exposed a higher pH as a result of the oxidation of sulfur in newly drained soils. Lime is sometimes added to raise soil pH and may dissolve, particularly when nitrogen additions of fertilizer and manure are added, also causing an increase in the concentration of dissolved inorganic carbon (DIC) (Barnes & Raymond, 2009). Respiration of soil organic matter can be stimulated by nutrient additions and will also contribute to alkalinity. The weathered minerals release silicates and nitrates as a result and are an explanation for why variations in alkalinity are directly related to nitrate, phosphate, and silicate concentrations over the last few decades. The flux of bicarbonate C since 1980 now represents about 81% of the total carbon in the river. The increased bicarbonate flux becomes a sink for atmospheric carbon when it dissociates into carbonate and silicate deposition in the ocean. The eventual dissolution of carbonates will take thousands of years, and silicate weathering will take millions of years (Berner et al., 1983), making carbonate and silicate deposition a temporary, but long-lasting carbon sink.
The rise in alkalinity in the Mississippi River over the last 120 years, first described by Raymond & Cole (2003) and Raymond et al. (2008), has since been shown to occur elsewhere. Tank et al. (2012), for example, showed that thawing of ancient permafrost deposits raises HCO3− fluxes as the weathering zone is exposed. Small temperate northeastern US streams (49 to 1492 km2) draining agricultural lands contained 3.3 more dissolved inorganic carbon than in forest-draining streams (Barnes & Raymond, 2009). Raymond et al. (2008) estimated that about 60% of the rise in downstream alkalinity flux in the Mississippi River watershed at the 1980 inflection point was due to tile drainage but not river discharge, which is similar to the 75% that Stets & Striegl (2012) found for the eastern seaboard watersheds. The remaining 40% was due to either precipitation or other factors. The 9% increase in discharge that is not balanced by precipitation is ascribed to drainage improvements (Raymond et al., 2012). The coincidental inflection points of alkalinity and temperature around 1980 are consistent with the observation of an inverse relationship between average annual silicate concentration and latitude in the world’s rivers (Turner et al., 2003).
Organic carbon
The total carbon (TC) delivered by the Mississippi River to the Gulf of Mexico in 2019 was about 20.1 TgC year−1 and the organic carbon was about 4.2 TgC year−1 (20% of the total). Much of the carbon going into the river is not present at the mouth because a significant amount of CO2 gas is evaded from river to atmosphere. Dubois et al. (2010), for example, estimated that CO2 evasion was 130% of the DIC flux in the Mississippi River (samples mostly from 2000 to 2001) when they used a gas diffusion coefficient determined by Wanninkhof (1992) but was 63% higher if they used the gas diffusion coefficient from Raymond & Cole (2003). Butman & Raymond (2011) used a longer water quality record of alkalinity and pH from mostly 1965 to 2000 data to estimate that 26.7 TgC year−1 is emitted as CO2 from the lower Mississippi River. Dubois et al. (2010) used isotopic composition data in the lower Mississippi River to indicate that the respired carbon source in the CO2 going into the atmosphere was from carbonate dissolution in soils. Butman & Raymond (2011) point out that some of the evasion of recently fixed CO2 in the Mississippi River’s streams and rivers may represent carbon that has been shunted into hydrologic networks. CO2 losses from streams and rivers through evasion, therefore, apparently equal or even exceed the annual discharge of total carbon in the lower Mississippi River. This suggests that the evasion of CO2 from the watershed to the atmosphere is greater than the delivery of DIC by the river to the sea.
The 4.2 TgC year−1 of total organic carbon entering the Gulf of Mexico at the river’s end increased 19% over the last 4 decades but is a minor part of changes in soil stocks of carbon. The carbon losses over the last 100 years are indicative of soil losses, but not a significant quantity relative to carbon losses in soils worldwide. Sanderman et al. (2017), for example, estimated that the carbon in the upper 2 m of global soils decreased by 8.1% compared to the historical stocks in 10,000 BC, primarily as a result from grazing and cropland agriculture (133 PgC). They estimated that 28% of the carbon in the soil’s upper 1 m was lost when it converted from grasslands to croplands. Soil carbon stocks in watersheds of the midwestern United States lost even more, perhaps equaling 25 to 50% of the soil organic matter present before cultivation (West et al., 2010). The conversion of grassland to arable land reduces soil organic matter and releases mineral N (Whitmore et al., 1992). Lu et al. (2018) modeled the contribution of intensive and extensive farming to crop production in the five midwestern states known as the ‘western cornbelt’ (ND, SD, NB, MN, and IA). These states produced 47% and 41% of the US corn and soybean harvest, respectively, from 2005 to 2017 (Lu et al., 2018). They found that between 2006 and 2016 that every kilogram of additional grain yield led to a soil carbon loss of 2.3 kg. The large carbon cost per kg gain production achieved by cropland expansion and rotation in the past decade was about 390 times higher than that by crop technology improvement. Lu et al. (2018) also found that 45% of carbon loss occurred in what had been grasslands, followed by 31% in former wetlands, 13% in former cropland converted to other land cover types, and 9% in previously forested lands, while over 59% of carbon gain was found in cropland due to its expansion. Although 14% of the newly expanded cropland was converted from wetlands, it contributed to ∼ one third of carbon loss due to the high soil carbon density in wetlands. The rising concentration of organic carbon in the river over the last four decades (19% of the total) indicates that these soil carbon losses will continue.
Nitrogen
The rise in nitrate concentration and decline in silicate concentration occurring in the two decades after the 1960s has become relatively stable over the last 40 years. The average annual nitrate concentration at St. Francisville, LA, for example, has been about the same from 1992 to 2015. The variability of nitrate and silicate concentrations is now moving coincidentally with variations with alkalinity. The rising nitrate concentrations are often ascribed to changing temporal and spatial applications of fertilizers. This is because fertilization increases loading rates to the land and converting forests to agricultural land and improving drainage, especially by installing subsurface tiles, increases the percent of the applied nitrogen that goes into drainage channels (Randall & Gross, 2001). The N fertilizer application rate increased almost three orders of magnitude from 1950 to 2015: from less than 0.01 gN m−2 year−1 in 1850 to 9.04 g N m−2 year−1 in 2015 (Cao et al., 2018) as fertilizer use climbed from 1.5 to 8.1 TgN year−1 from 1960 to 2014 (Tian et al., 2020). After then use stabilized (Fig. S1). Nitrogen fertilizer use in the Mississippi River basin now accounts for ~ 65% of the total fertilizer application in the continental United States (Tian et al., 2020).
David et al. (2010) investigated how much drainage improvements affected nitrogen yield compared to fertilizer applications by conducting a whole basin analysis of 153 watersheds using county-level data for N inputs and land use. They found that fertilizer inputs were tightly coupled with the fraction of land in row crops, just as Crumpton et al. (2006; Fig. 11) did in Iowa, Hatfield et al. (2009) did in the Raccoon watershed, Iowa, and Broussard & Turner (2009) discovered in their analysis of nitrate yields in farmlands in the northern Mississippi River watershed from 1906 to 1912. Raymond et al. (2012) estimated that 34% of the applied nitrogen in the Mississippi River water is exported and Booth & Johnson (2007) suggested that fertilizer runoff was 59% of loading, and that 10% of the area contributed 75% of the input. Howarth et al. (2012) estimated that 25% of the nitrogen applied in 154 watersheds in Europe and the US was exported to rivers and streams where the nitrogen applied was greater than 1.07 kg N km−2 year−1. Turner & Rabalais (1991) calculated that the nitrate concentration (at that time) could have come from 22% of the fertilizer applied. The increased P loading into P-limited systems would make diatom production more likely (especially in a nitrogen replete environment) as Downing et al. (2016) demonstrated to occur in agricultural fields, but not in reservoirs.
Much of the total nitrogen yield in the basin is a consequence of structural changes in the landscape from building swales and tiling, not from solely (or simply) the nitrogen fertilizer application rates (e.g., Kaspar et al., 2003; Tomer et al., 2003; McIsaac & Hu, 2004; Nangia et al., 2008; Randall & Gross, 2001). Land drainage dries soils, as intended, but reduces denitrification, and moves the leached nitrate to waterways that reduces the potential sediment trapping and denitrification in riparian zones (Burt & Pinay, 2005). Randall & Gross (2001) showed that there was a close correspondence between tile drain water yield and nitrate yield, which is visually apparent when maps of the county-level nitrogen yield and drainage are compared (Fig. S4). This is why McIsaac & Hu (2004) found that the 1945–1961 riverine nitrate flux in an extensively tile drained region in Illinois averaged 6.6 kg N ha−1 year−1, compared to 1.3 to 3.1 kg N ha−1 year−1 for the non-tile drained region, even though the nitrogen application was greater in the non-tile drained region. Arenas Amado et al. (2017) found that tiles “delivered up to 80% of the stream N load while providing only 15–43% of the water” in the 122 km2 Otter Creek watershed in Iowa. Ikenberry et al., (2014) estimated that 97% of the nitrate flow occurred during 50% of the highest flows in a 5-year study of the 5132 ha Walnut Creek watershed in Iowa. They reported that two-thirds of row crop land in Central Iowa and the Minnesota Till Prairies Major Land Resources Area had subsurface tile drainage. Measurements of inorganic C concentrations in the intensively farmed Raccoon River, IA, by Jones & Schilling (2013) are another example of the consequence of drainage on nitrate concentrations. The alkalinity there doubled from 2000 to 2011 compared to in 1931 to 1944 (Jones & Schilling, 2013), during which there was an increase in drainage and nitrate concentrations, but not fertilizer applications (Hatfield et al., 2009).
Research done at the local scale suggests that tile depth and spacing are important factors for water quality restoration. Nangia et al. (2008), for example, used 14 years of field data from the Raccoon River watershed in Iowa to determine what controls variations in the nitrate yields for similar fields under different drainage. They found that a simple rearrangement of the fertilizer application (no reduction in total fertilizer application) resulted in a 21% reduction in nitrate yield. Hofmann et al. (2004) found that subsurface tile drainage spaced 20 m apart optimized corn crop yields.
In summary, Nitrogen yields from field to streams is generally dominated by fertilizer application which has increased in the 1960 and stabilized in the last decades. More nitrate escapes farm fields when tiled and so drainage improvements are a significant multiplier of these loadings.
Silicate and suspended sediment
The silicate concentrations are directly related to discharge, but there is less silicate per discharge now than earlier in the twentieth century. Fortner et al. (2012) also showed a direct relationship of Dsi yield and discharge. Silicate concentrations decreased from the 1900s to 1960s as alkalinity and nitrate concentrations rose, but after 1994 the nitrate and silicate concentrations are directly related to alkalinity concentrations. The Dsi:nitrate molar ratio in the river now is about 1:1, compared to 4:1 in the early 1900s.
Several hypotheses explain changes in the silicate concentration in riverine ecosystems which are distinct from why nitrogen concentrations have increased. Humborg et al. (2000) argued that damming created longer residence times and improved light conditions favoring algal growth, including diatoms that sequester the silica in their frustules. This was observed in the Danube River when silicate concentrations were reduced from 79 to 20 µmol l−1 after the construction of the Iron Gate I dams, and the decline was not a step-function but linear over 17 years (Cociasu et al., 1996). Data in a recent review (Ma et al., 2017) supports this interpretation because their data show a higher percent silicate removal in reservoirs as residence times increase (Fig. S6). The decline in silicate concentrations, coincidental with decreased suspended sediment concentrations, can be ascribed to be primarily due to a longer water turnover time as a result of dam construction. The 50% decline in suspended sediment flux after the early 1950s was when dam construction expanded rapidly to increase hydrologic storage behind them (Meade & Moody, 2010). Other concurrent activities such as channel straightening, dikes, revetments, and soil erosion controls also affected sediment storage in various ways (Belt, 1975; Meade & Moody, 2010). The Missouri-Mississippi River sediment supply separated into two periods in 1967 which Meade and Moody (2010) describe as the going from a transport-limited system to a supply-limited system. The silicate concentrations stabilized about 10 years earlier than the peak in nitrate concentration.
Vegetative cover
Vegetative cover and land use is also important. Struyf et al. (2010) and Fortner et al. (2012), for example, demonstrated that agriculture development led to a 50% reduction of silicate delivery. Struyf et al. (2010) suggested that the initial land disturbance raised Si releases, which later was lowered as crop removal had an effect. Crop harvesting removes a large enough amount of silica to significantly change terrestrial silica cycling (Vandevenne et al., 2012) because the return of silica is greatly reduced; also, the “absence of deep-rooting to shallow-rooted crops prevents vegetation-stimulated mineral weathering” (Schaller et al., 2021). DSI export, therefore, is also partially the result of a dynamic balance between weathering of lithosphere and vegetation, and how vegetation influences weathered products (Cornelis & Delvaux, 2016). The accumulation of Si in plants ranges over two orders of magnitude (Hodson et al., 2005) and mineral weathering increases at higher temperatures and with greater hydrologic throughput. Further, Si is bound in different forms of unequal dissolution potentials. The vegetative influences include plant type, pH, vegetative uptake through rhizo-fungal interactions, phytolith production and uptake by diatoms and sponges.
Phosphate
The records of phosphate concentrations in the river don’t begin until after the major changes in fertilization, dam construction, and discharge increases. We reasonably might expect that soil disturbance from fertilizer additions, agricultural expansion and forest clearance resulted in more phosphorus runoff. The phosphorus in soil is bound to clay particles but there is no direct relationship between suspended sediment and phosphate concentrations in these samples. Furthermore, phosphorus has long been known to accumulate in soils (Bennett et al., 2001), and water treatment expansion after the Clean Water Act (Turner, 2021) makes hindsight predictions of cause-and-effects unreliable.
Summary of major influences on river constituents
A brief summary of the prominent drivers of variations in Mississippi River discharge and the concentration of five inorganic constituents over the last 120 years is in Table 2. Climate change and land use is exerting a strong enough effect on discharge rates to confound the pre-1980 relationship with the NAO. The land use changes include vegetative as well as hydrologic features, principally drainage improvements, that affect alkalinity. The variations in alkalinity have a knock-on effect on soil weathering that releases silicate, nitrate, and phosphate concentrations in riverwater for the last few decades when fertilizer applications and water impoundment construction stabilized. Suspended sediment concentrations declined before 1970 and with proportionality to silicate concentrations that a demonstrably higher trapping efficiency with greater hydrologic storage. The low in silicate concentrations from trapping behind impoundments preceded by about 10 years the dramatic rise in nitrate concentrations which were driven by fertilizer applications and drainage improvements. Before 1950, the changes in alkalinity appear to be driven by less intensive agricultural expansion than after 1950. The temporally sparse record of dissolved organic carbon is consistent with the various observations of soil carbon losses throughout the watershed.
Restoration/resilience
Because soil quality affects riverine carbon sources and sinks, there is interest in restoring soil organic matter to reduce the 11.2 ± 0.4% of the total global emissions of CO2, N2O, and CH4 that are from agriculture (Tubiello et al., 2015; Rumpel et al., 2018). Current global soil stocks contain 2 to 3 times more carbon than the atmosphere (Le Quéré et al., 2018), and grasslands contain about one third of it, almost entirely as roots and organic matter (Bai & Cotrufo, 2022). Sanderman et al. (2017) suggested that perhaps 10 to 30% of the historic loss could be restored, which is more than 75 years of the current global fossil fuel emissions. Fargione et al. (2018) estimated that as much as 21% of the current annual CO2 emissions in the US might be captured each year by restoring soil carbon. The Mississippi’s annual total carbon and organic carbon was 6.8 and 1.4%, respectively, of that amount. The path to restoring soil carbon will be through re-invigorating belowground plant production and storage.
Shifting to more diverse plant rotations will be needed to have a major effect on carbon losses and to sustain soil fertility and farming enterprises. Achieving this restoration does mean that it will result in negative consequences for farmers or consumers. Soil carbon restoration has many benefits important to sustaining agriculture because organic material holds water, minerals, nutrients and organisms giving it soil structure, resistance to erosion and increased soil fertility (Chambers et al., 2016; Jackson et al., 2017). Improving soil health will provide public benefits of improved water quality, flood reduction, enhanced wildlife habitat, reduced air pollutants, and reduced global warming potential (Boody et al., 2005; Jordan et al., 2007; Ryan et al., 2018). Plant rotations will include cover crops, especially deep-rooted perennials (Poeplau & Don, 2015; Jungers et al., 2019; Paustian et al., 2019). Land under tile drainage can be a part of these efforts (Randall & Mulla, 2001; Dinnes et al., 2002). Tile drainage can enter buffer strips before reaching streams, drain into wetlands, or even not be used if row cropped fields are converted to perennials. Planting only shallow-rooted annuals leaves the ground without plant cover for more than half the year and results in more soil erosion (Heathcote et al., 2013). Replacement of annuals with deep-rooted perennials provides continuous living cover and reduces soil erosion (Song et al., 2014). Water quality has improved in some sub-watershed streams of the Mississippi watershed because of soil conservation (Rabotyagov et al., 2014; McIsaac et al., 2016; García et al., 2016).
A signature example of alternative management of perennial grains is provided by Davis et al. (2012), Liebman et al. (2013), and Tomer and Liebmann (2014) who conducted a 7-year field trial of different cropping systems for corn-soybean rotations. Some key findings were that, by using cover crops for 4 years, there was a 50% or more reduction in fossil fuel use, a doubling of employment, and not loss of profits. The diversification of crop coverage with small grains and legumes had a 91% reduction in fertilizer use, 97% reduction in herbicide use, and increased carbon storage. There is much work remaining—half of the US stream and river miles violate water pollution standards (Keiser & Shapiro, 2019). Whole system analyses of land use alternatives for the small and large farms are needed to include not only GHG emissions, but also energy expenditures, wildlife, water quality, and social factors. Boody et al. (2005) found that alternative land management schemes using the same resources could create “improved water quality, healthier fish, increased carbon sequestration, and decreased greenhouse gas emissions, while economic benefits include social capital formation, greater farm profitability, and avoided costs.” On-the-ground experiments at a watershed scale that include social governance (Meyfroidt et al., 2022) are recommended.
Restoring water quality within the watershed means, in part, reducing nitrate and phosphorous runoff that contributes to the hypoxic zone on the continental shelf of the northern Gulf of Mexico. A Hypoxia Action Plan of 2001 (HAP; Mississippi River/Gulf of Mexico Watershed Nutrient Task Force, 2001) established a goal of reducing the size of the hypoxic zone to less than an average of 5000 km2 over 5 years by 2035. The identified mechanism to do this at the time was to reduce nitrate loading in the river which Scavia et al. (2017) estimated to be a 59% reduction in nutrient loading to meet the 5000 km2 goal. But the nitrate loading in the river declined by only 4% since the 2001 agreement (Fig. S5). Clearly the results so far are insufficient, slow to develop and will result from changes in land use, primarily in the agricultural sector.
Conclusion
The relationships between alkalinity, nitrogen, phosphorus, and silica are interwoven so tightly today that it is difficult to conceive of modifying one without affecting the other. Fertilizer applications in the watershed are not always directly proportional to nutrient loading in the Mississippi River; their influence is modified by tile drainage, climate variation, plant choices, and farm management. Conversion of natural habitat to agricultural fields with improved drainage reduces plant cover (and hence evapotranspiration) and more water is shunted into streams. This, in turn, changes the rate of soil accumulation and denitrification as well as alkalinity and the concentrations of alkalinity, nitrate, phosphate, and silicate. The integrated culmination of these land use changes across the watershed over the last 100 years reduced water residence time and nutrient processing in soil but increased concentrations in water bodies; the percent of nitrate exported from soils to water becomes higher as a result. Today agricultural lands are hemorrhaging the carbon needed to build and sustain healthy soils and a significant carbon sink is not being realized. Examples of accommodations for simultaneous water quality improvements while building soil health exist, but more on-the-ground examples are needed to integrate both the biogeochemical factors and practical socio-political-economic aspects of farming.
References
Alexander, R. B., R. A. Smith & G. E. Schwarz, 2000. Effect of stream channel size on the delivery of nitrogen to the Gulf of Mexico. Nature 403: 758–761.
Alexander, R. B., R. A. Smith & G. E. Schwarz, 2008. Differences in phosphorus and nitrogen delivery to the Gulf of Mexico from the Mississippi River basin. Environmental Science and Technology 42: 822–830.
Allan, J. D., 2004. Landscapes and riverscapes: the influence of land use on stream ecosystems. Annual Review Ecology and Systematics 35: 257–284.
Alvarez-Cobelas, M., D. G. Angeler & S. Sánchez-Carrillo, 2008. Export of nitrogen from catchments: a worldwide analysis. Environmental Pollution 156: 261–269.
Arenas Amado, A., K. E. Schilling, C. S. Jones, N. Thomas & L. J. Weber, 2017. Estimation of tile drainage contribution to streamflow and nutrient loads at the watershed scale based on continuously monitored data. Environmental Monitoring and Assessment 189: 426. https://doi.org/10.1007/s10661-017-6139-4.
Bai, Y. & F. Cotrufo, 2022. Grassland soil carbon sequestration: current understanding, challenges, and solutions. Science 377(6606): 603–608. https://doi.org/10.1126/science.abo2380.
Barnes, R. T. & P. A. Raymond, 2009. The contribution of agricultural and urban activities to inorganic carbon fluxes within temperate watersheds. Chemical Geology 266: 318–327. https://doi.org/10.1016/j.chemgeo.2009.06.018.
Belt, C. M., 1975. The 1973 flood and man’s constriction of the Mississippi River. Science 1975(189): 681–684. https://doi.org/10.1126/science.189.4204.681.
Bennett, E. M., S. R. Carpenter & N. F. Caraco, 2001. Human impact on erodible phosphorus and eutrophication: a global perspective. BioScience 51(3): 227–234.
Berner, R. A., A. C. Lasaga & R. M. Garrels, 1983. The carbonate-silicate geochemical cycle and its effect on atmospheric carbon dioxide over the past 100 million years. American Journal Science 283: 641–683.
Boody, G., B. Vondracek, D. A. Andow, M. Krinke, J. Westra, J. Zimmerman & P. Welle, 2005. Multifunctional agriculture in the United States. BioScience 55(1): 27–38.
Booth, M. S. & H. P. Johnson, 2007. Spring nitrate flux in the Mississippi River basin: a landscape model with conservation applications. Environmental Science and Technology 41: 5410–5418.
Broussard, W. & R. E. Turner, 2009. A century of changing land-use and water-quality relationships in the continental U.S. Frontiers in Ecology and the Environment 7: 302–307.
Burt, T. P. & G. Pinay, 2005. Linking hydrology and biogeochemistry in complex landscapes. Progress in Physical Geography 29(3): 297–316.
Butman, D. & P. A. Raymond, 2011. Significant efflux of carbon dioxide from streams and rivers in the United States. Nature Geoscience 4(12): 839–842.
Cao, P., C. Lu & Z. Yu, 2018. Historical nitrogen fertilizer use in agricultural ecosystems of the contiguous United States during 1850–2015: application rate, timing, and fertilizer types. Earth Systems Science Data 10: 969–984. https://doi.org/10.5194/essd-10-969-2018.
Chabbi, A., C. Rumpel, F. Hagedorn, M. Schrumpf & P. C. Baveye, 2022. Editorial: carbon storage in agricultural and forest soils. Frontiers in Environmental Science 10: 848572. https://doi.org/10.3389/fenvs.2022.848572.
Chambers, A., R. Lal & K. Paustian, 2016. Soil carbon sequestration potential of US croplands and grasslands: implementing the 4 per thousand initiative. Journal Soil and Water Conservation 71: 68A-74A.
Cociasu, A., L. Dorogan, C. Humborg & L. Popa, 1996. Long-term ecological changes in Romanian coastal waters of the Black Sea. Marine Pollution Bulletin 32(1): 32–38.
Cornelis, J. T. & B. Delvaux, 2016. Soil processes drive the biological silicon feedback loop. Functional Ecology 30(8): 1298–1310.
Crumpton, W. G., G. A. Stenback, B. A. Miller & M. J. Helmers, 2006. Potential benefits of wetland filters for tile drainage systems: impact on nitrate loads to Mississippi River subbasins. Final Project Report to U.S. Department of Agriculture, Project number: IOW06682.
David, M. B., L. E. Drinkwater & G. F. McIsaac, 2010. Sources of nitrate yields in the Mississippi River basin. Journal of Environmental Quality 39: 1657–1667. https://doi.org/10.2134/jeq2010.0115.
Davis, A. S., J. D. Hill, C. A. Chase, A. M. Johanns & M. Liebman, 2012. Increasing cropping system diversity balances productivity, profitability and environmental health. PLoS ONE 7(10): e47149. https://doi.org/10.1371/journal.pone.0047149.
Dinnes, D. L., D. L. Karlen, D. B. Jaynes, T. C. Kaspar, J. L. Hatfield, T. S. Colvin & C. A. Cambardella, 2002. Nitrogen management strategies to reduce nitrate leaching in tile-drained Midwestern soils. Agronomy Journal 94(1): 153–171.
Dole, R. B., 1909. The quality of surface waters in the United States. Part 1. Analyses of waters east of the one hundredth meridian. Washington, DC. USGS Water Supply Paper 236.
Downing, J. A., C. T. Cherrier & R. W. Fulweiler, 2016. Low ratios of silica to dissolved nitrogen supplied to rivers arise from agriculture not reservoirs. Ecology Letters 19(12): 1414–1418.
Dubois, K. D., D. Lee & J. Veizer, 2010. Isotopic constraints on alkalinity, dissolved organic carbon, and atmospheric carbon dioxide fluxes in the Mississippi River. Journal Geophysical Research 115: G02018. https://doi.org/10.1029/2009JG001102.
Dunn, D. D., 1996. Trends in nutrient inflows to the Gulf of Mexico from streams draining the conterminous United States 1972–1993. U.S. Geological Survey, Water-Resources Investigations Report 96-4113. Prepared in cooperation with the U.S. Environmental Protection Agency, Gulf of Mexico Program, Nutrient Enrichment Issue Committee, U.S. Geological Survey, Austin, TX; 60 pp.
Edwards, E. C. & W. N. Thurman, 2022. The economics of climatic adaptation: agricultural Drainage in the United States. Proceedings NBER/USDA Conference on Economic Perspectives on Water Resources, Climate Change, and Agricultural Sustainability.
Fargione, J. E., S. Bassett, T. Boucher, S. D. Bridgham, R. T. Conant, S. C. Cook-Patton, P. W. Ellis, A. Falcucci, J. W. Fourqurean, T. Gopalakrishna & H. Gu, 2018. Natural climate solutions for the United States. Science Advances 4(11): eaat1869.
Finlay, J. C., J. M. Hood, M. P. Limm, M. E. Power, J. D. Schade & J. R. Welter, 2011. Light-mediated thresholds in stream-water nutrient composition in a river network. Ecology 92: 140–150.
Fortner, S. K., W. B. Lyons, A. E. Carey, M. J. Shipitalo, S. A. Welch & K. A. Welch, 2012. Silicate weathering and CO2 consumption within agricultural landscapes, the Ohio-Tennessee River Basin, USA. Biogeosciences 9: 941–955.
García, A. M., R. B. Alexander, J. G. Arnold, L. Norfleet, M. J. White, D. M. Robertson & G. Schwarz, 2016. Regional effects of agricultural conservation practices on nutrient transport in the Upper Mississippi River Basin. Environmental Science and Technology 50(13): 6991–7000.
Gardner, J. R. & M. W. Doyle, 2018. Sediment–water surface area along rivers: water column versus benthic. Ecosystems 21: 1505–1520. https://doi.org/10.1007/s10021-018-0236-2.
Hatfield, J. L., L. D. McMullen & C. S. Jones, 2009. Nitrate-nitrogen patterns in the Raccoon River Basin related to agricultural practices. Journal Soil Water Conservation 64: 190–199.
Heathcote, A. J., C. T. Filstrup & J. A. Downing, 2013. Watershed sediment losses to lakes accelerating despite agricultural soil conservation efforts. PLoS ONE 8(1): e53554. https://doi.org/10.1371/journal.pone.0053554.
Hodson, M. J., P. J. White, A. Mead & M. R. Broadley, 2005. Phylogenetic variation in the silicon composition of plants. Annals of Botany 96(6): 1027–1046.
Hofmann, B. S., S. M. Brouder & R. F. Turco, 2004. Tile spacing impacts on Zea mays L. yield and drainage water nitrate load. Ecological Engineering 23(4–5): 251–267.
Homer, C., J. Dewitz, L. Yang, S. Jin, P. Danielson, G. Xian, J. Coulston, N. Herold, J. Wickham & K. Megown, 2015. Completion of the 2011 National Land Cover Database for the conterminous United States – representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing 81(5): 345–354.
Howarth, R., D. Swaney, G. Billen, J. Garnier, B. Hong, C. Humborg, P. Johnes, C. M. Mörth & R. Marino, 2012. Nitrogen fluxes from the landscape are controlled by net anthropogenic nitrogen inputs and by climate. Frontiers in Ecology and the Environment 10(1): 37–43.
Humborg, C., D. J. Conley, L. Rahm, F. Wulff, A. Cociasu & V. Ittekkot, 2000. Silicon retention in river basins: far-reaching effects on biogeochemistry and aquatic food webs in coastal marine environments. AMBIO 29(1): 45–50.
Hynes, H. B. N., 1960. The Biology of Polluted Waters, Liverpool University Press, Liverpool:, 202.
Ikenberry, C. D., M. L. Soupir, K. E. Schilling, C. S. Jones & A. Seeman, 2014. Nitrate-nitrogen export: magnitude and patterns from drainage districts to downstream river basins. Journal Environmental Quality 43(6): 2024–2033.
IPCC, 2021. Climate Change, 2021: the physical science basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge.
Jackson, R. B., K. Lajtha, S. E. Crow, G. Hugelius, M. G. Kramer & G. Piñeiro, 2017. The ecology of soil carbon: pools, vulnerabilities, and biotic and abiotic controls. Annual Review of Ecology, Evolution, and Systematics 48(1): 419–445.
Jarvie, H. P., C. Neal, D. V. Leach, G. P. Ryland, W. A. House & A. J. Robson, 1997. Major ion concentrations and the inorganic carbon chemistry of the Humber rivers. Science of the Total Environment 194: 285–302.
Jaynes, D. B. & D. E. James, 2007. The Extent of Farm Drainage in the United States. USDA, Washington, DC. ARS. https://www.ars.usda.gov/ARSUserFiles/50301500/theextentoffarmdrainageintheunitedstates.pdf.
Jones, C. S. & K. E. Schilling, 2013. Carbon export from the Raccoon River, Iowa: patterns, processes, and opportunities. Journal Environmental Quality 42: 155–163.
Jordan, N., G. Boody, W. Broussard, J. D. Glover, D. Keeney, B. H. McCowan, G. McIsaac, M. Muller, H. Murray, J. Neal, C. Pansing, R. E. Turner, K. D. Warner & D. L. Wyse, 2007. Sustainable development of the agricultural bio-economy. Science 316: 1570–1571.
Julian, J. P., N. A. Wilgruber, K. M. de Beurs, P. M. Mayer & R. N. Jawarneh, 2015. Long-term impacts of land cover changes on stream channel loss. Science of the Total Environment 537: 399–410.
Jungers, J. M., L. H. DeHaan, D. J. Mulla, C. C. Sheaffer & D. L. Wyse, 2019. Reduced nitrate leaching in a perennial grain crop compared to maize in the Upper Midwest, USA. Agriculture, Ecosystems and Environment 272: 63–73.
Kaspar, T., D. Jaynes, T. Moorman & T. Parkin, 2003. Reducing nitrate levels in subsurface drain water with organic matter incorporation. Final Report to the American Farm Bureau by the Foundation for Agriculture. National Soil Tilth Laboratory Ames, Iowa.
Keiser, D. A. & J. S. Shapiro, 2019. Consequences of the Clean Water Act and the demand for water quality. The Quarterly Journal Economics 134(1): 349–396.
Le Quéré, C., R. M. Andrew, P. Friedlingstein, S. Sitch, J. Hauck, J. Pongratz, P. A. Pickers, J. I. Korsbakken, G. P. Peters, J. G. Canadell & A. Arneth, 2018. Global carbon budget 2018. Earth System Science Data 10(4): 2141–2194.
Lee, C. 2022. Nutrient loads to the Gulf of Mexico produced by the USGS National Water Quality Network, 1968-2021: U.S. Geological Survey. https://doi.org/10.5066/P9G0EEUE.
Lee, C. J., J. C. Murphy, C. G. Crawford & J. R. Deacon, 2017. Methods for Computing Water-Quality Loads at Sites in the U.S. Geological Survey National Water Quality Network. Open-File Report 2017-1120. https://doi.org/10.3133/ofr20171120.
Liebman, M., M. J. Helmers, L. A. Schulte & C. A. Chase, 2013. Using biodiversity to link agricultural productivity with environmental quality: results from three field experiments in Iowa. Renewable Agriculture and Food Systems 28(2): 115–128.
Lu, C., Z. Yu, H. Tian, D. A. Hennessy, H. Feng, M. Al-Kaisi, Y. Zhou, T. Sauer & R. Arritt, 2018. Increasing carbon footprint of grain crop production in the US Western Corn Belt. Environmental Research Letters 13: 124007. https://doi.org/10.1088/1748-9326/aae9fe.
Ma, N., Z. Song, B. Wang, F. Wang, X. Yang, X. Zhang, Q. Hao & Y. Wu, 2017. Effects of river damming on biogenic silica turnover: implications for biogeochemical carbon and nutrient cycles. Acta Geochimica 36(4): 626–637.
McHargue, J. S. & A. M. Peter, 1921. The removal of mineral plant-food by natural drainage water. Kentucky Agricultural Experiment Station Bulletin 237.
McIsaac, G. F. & X. Hu, 2004. Net N input and riverine N export from Illinois agricultural watersheds with and without extensive tile drainage. Biogeochemistry 70: 251–271.
McIsaac, G. F., M. B. David & G. Z. Gertner, 2016. Illinois River nitrate-nitrogen concentrations and loads, long-term variation and association with watershed nitrogen inputs. Journal Environmental Quality 45(4): 1268–1275. https://doi.org/10.2134/jeq2015.10.0531.
Meade, R. H. & J. A. Moody, 2010. Causes for the decline of suspended-sediment discharge in the Mississippi River system, 1940–2007. Hydrological Processes: An International Journal 24(1): 35–49.
Meybeck, M., 2003. Global analysis of river systems: from earth system controls to Anthropocene syndromes. Philosphical Transactions of the Royal Society B 358(1440): 1935–1955. https://doi.org/10.1098/rstb.2003.1379.
Meyfroidt, P., A. de Bremond, C. M. Ryan, E. Archer, R. Aspinall, A. Chhabra, G. Camara, E. Corbera, R. DeFries, S. Díaz & J. Dong, 2022. Ten facts about land systems for sustainability. Proceedings National Academy of Sciences (USA) 119(7): e2109217118.
Mississippi River Commission, 1955. Annual Maximum Minimum and Mean Discharges of the Mississippi River to 1953, U.S. Army Corps of Engineers, Vicksburg:
Mississippi River/Gulf of Mexico Watershed Nutrient Task Force, Action Plan for Reducing, Mitigating, and Controlling Hypoxia in the Northern Gulf of Mexico (Mississippi River/Gulf of Mexico Watershed Nutrient Task Force, Washington, DC), 2001. [available at https://www.epa.gov/sites/production/files/2015-03/documents/2001_04_04_msbasin_actionplan2001.pdf].
Murphy, J. C., R. M. Hirsch & L. A. Sprague, 2014. Antecedent flow conditions and nitrate concentrations in the Mississippi River Basin. Hydrology and Earth System Sciences 18(3): 967–979.
Nangia, V., P. H. Gowda, D. J. Mulla & G. R. Sands, 2008. Water quality modeling of fertilizer management impacts on nitrate losses in tile drains at the field scale. Journal Environmental Quality 37: 296–307.
Paul, M. J. & J. L. Meyer, 2001. Streams in the urban landscape. Annual Review Ecology and Systematics 32: 333–365.
Paustian, K., E. Larson, J. Kent, E. Marx & A. Swan, 2019. Soil C sequestration as a biological negative emission strategy. Frontiers in Climate 1: 8. https://doi.org/10.3389/fclim.2019.00008.
Peterson, T. C., R. R. Heim Jr., R. Hirsch, D. P. Kaiser, H. Brooks, N. S. Diffenbaugh, R. M. Dole, J. P. Giovannettone, K. Guirgui, T. R. Karl, R. W. Katz, K. Kunkel, D. Lettenmaier, G. J. McCabe, C. J. Paciorek, K. R. Ryberg, S. Schubert, V. B. S. Silva, B. C. Stewart, A. V. Vecchia, G. Villarini, R. S. Vose, J. Walsh, M. Wehner, D. Wolock, K. Wolter, C. A. Woodhouse & D. A. Wuebbles, 2013. Monitoring and understanding changes in heat waves, cold waves, floods, and droughts in the United States: state of knowledge. Bulletin of the American Meteorological Society 94(6): 821–834. https://doi.org/10.1175/BAMS-D-12-00066.1.
Poeplau, C. & A. Don, 2015. Carbon sequestration in agricultural soils via cultivation of cover crops –a meta-analysis. Agriculture, Ecosystems and Environment 200: 33–41.
Rabotyagov, S. S., T. D. Campbell, M. White, J. G. Arnold, J. Atwood, M. L. Norfleet, C. L. Kling, P. W. Gassman, A. Valcu, J. Richardson & R. E. Turner, 2014. Cost-effective targeting of conservation investments to reduce the northern Gulf of Mexico hypoxic zone. Proceedings National Academy of Sciences (USA) 111(52): 18530–18535.
Raimbault, P., W. Pouvesta, F. Diaz, N. Garcia & R. Sempéré, 1999. Wet-oxidation and automated colorimetry for simultaneous determination of organic carbon, nitrogen and phosphorus dissolved in seawater. Marine Chemistry 66: 166–169.
Randall, G. W. & M. J. Gross, 2001. Nitrate losses to surface water through subsurface tile drainage. In Hatfield, J. L. & R. F. Follett (eds), Nitrogen in the Environment: Sources, Problems, and Management Elsevier Sciences B.V., Amsterdam: 145–175.
Randall, G. W. & D. J. Mulla, 2001. Nitrate nitrogen in surface waters as influenced by climatic conditions and agricultural practices. Journal Environmental Quality 30: 337–344.
Raymond, P. A. & J. J. Cole, 2003. Increase in the export of alkalinity from North America’s largest river. Science 301: 88–91.
Raymond, P. A. & S. K. Hamilton, 2018. Anthropogenic influences on riverine fluxes of dissolved inorganic carbon to the ocean. Limnology and Oceanography Letters 3: 143–155.
Raymond, P. A., N. H. Oh, R. E. Turner & W. Broussard, 2008. Anthropogenic enhanced fluxes of water and carbon from the Mississippi River. Nature 451: 449–452.
Raymond, P. A., M. B. David & J. E. Saiers, 2012. The impact of fertilization and hydrology on nitrate fluxes from Mississippi watersheds. Current Opinion in Environmental Sustainability 4(2): 212–218.
Raymond, P. A., J. E. Saiers & W. V. Sobzak, 2016. Hydrological and biogeochemical controls on watershed dissolved organic matter transport: pulse-shunt concept. Ecology 97: 5–16.
Reisinger, A. J., J. L. Tank, E. J. Rosi-Marshall, R. O. Hall & M. A. Baker, 2015. The varying role of water column nutrient uptake along river continua contrasting landscapes. Biogeochemistry 125: 115–131.
Rumpel, C., F. Amiraslani, L. S. Koutika, P. Smith, D. Whitehead & E. Wollenberg, 2018. Put more carbon in soils to meet Paris climate pledges. Nature 564: 33–34.
Runkel, R. L., C. G. Crawford & T. A. Cohn, 2004. Load Estimator (LOADEST): a FORTRAN Program for estimating constituent loads in streams and rivers: US Geological Survey Techniques and Methods. Techniques and Models Book 4. Chapter 5, US Geological Survey, Reston, VA.
Ryan, M. R., T. E. Crews, S. W. Culman, L. R. Dehaan, R. C. Hayes, J. B. Jungers & M. G. Bakker, 2018. Managing for mulifunctionality in perennial grain crops. BioScience 68: 294–304.
Sanderman, J., T. Hengl & G. J. Fiske, 2017. Soil carbon debt of 12,000 years of human land use. Proceedings National Academy of Sciences (USA) 114(36): 9575–9580.
Scavia, D., I. Bertani, D. R. Obenour, R. E. Turner, D. R. Forrest & A. Katkin, 2017. Ensemble modeling informs hypoxia management in the northern Gulf of Mexico. Proceedings National Academy of Sciences (USA) 114: 8823–8828. https://doi.org/10.1073/pnas.1705293114.
Schaller, J., D. Puppe, D. Kaczorek, R. Ellerbrock & M. Sommer, 2021. Silicon cycling in soils revisited. Plants 10(2): 295. https://doi.org/10.3390/plants10020295.
Schilling, K. E. & R. D. Libra, 2003. Increased baseflow in Iowa over the second half of the 20th century. Journal American Water Resources Association 39(4): 851–860.
Schilling, K. E., M. K. Jha, Y.-K. Zhang, P. W. Gassman & C. F. Wolter, 2008. Impact of land use and land cover change on the water balance of a large agricultural watershed: historical effects and future directions. Journal Water Resources Research 44: W00A09. https://doi.org/10.1029/2007wr006644.
Sewerage and Water Board, 1903. Report on Water Purification Investigation and of Plans Proposed for Sewerage and Waterworks Systems, A.W. Hyatt Stationery Manufacturing Co., Ltd, New Orleans:
Six, J., S. M. Ogle, F. Jay Breidt, R. T. Conant, A. R. Mosier & K. Paustian, 2004. The potential to mitigate global warming with no-tillage management is only realized when practised in the long term. Global Change Biology 10(2): 155–160.
Song, Z., K. Müller & H. Wang, 2014. Biogeochemical silicon cycle and carbon sequestration in agricultural ecosystems. Earth-Science Reviews 139: 268–278.
Stets, E. G. & R. G. Striegl, 2012. Carbon export by rivers draining the conterminous United States. Inland Waters 2: 177–184. https://doi.org/10.5268/IW-2.4.510.
Struyf, E., A. Smis, S. Van Damme, J. Garnier, G. Govers, B. Van Wesemael, D. J. Conley, O. Batelaan, E. Frot, W. Clymans & F. Vandevenne, 2010. Historical land use change has lowered terrestrial silica mobilization. Nature Communications 1(1): 1–7.
Tank, S. E., P. A. Raymond, R. G. Striegl, J. W. McClelland, R. M. Holmes, G. J. Fiske & B. J. Peterson, 2012. A land-to-ocean perspective on the magnitude, source and implication of DIC flux from major Arctic rivers to the Arctic Ocean. Global Biogeochemical Cycles 26: GB4018. https://doi.org/10.1029/2011GB004192.
Tian, H., R. Xu, S. Pan, Y. Yao, Z. Bian, W. J. Cai, C. S. Hopkinson, D. Justic, S. Lohrenz, C. Lu & W. Ren, 2020. Long-term trajectory of nitrogen loading and delivery from Mississippi River basin to the Gulf of Mexico. Global Biogeochemical Cycles 34: e2019GB006475. https://doi.org/10.1029/2019GB006475.
Tomer, M. D. & M. Liebman, 2014. Nutrients in soil water under three rotational cropping systems, Iowa, USA. Agriculture, Ecosystems and Environment 180: 105–114.
Tomer, M. D., D. W. Meek, D. B. Jaynes & J. L. Hatfield, 2003. Evaluation of nitrate nitrogen fluxes from a tile-drained watershed in Central Iowa. Journal Environmental Quality 32: 642–653.
Tubiello, F. N., M. Salvatore, A. F. Ferrara, J. House, S. Federici, S. Rossi, R. Biancalani, R. D. Condor Golec, H. Jacobs, A. Flammini & P. Prosperi, 2015. The contribution of agriculture, forestry and other land use activities to global warming, 1990–2012. Global Change Biology 21: 2655–2660.
Turner, R. E., 2021. Declining bacteria, lead, and sulphate, and rising pH and oxygen in the lower Mississippi River. Ambio 50: 1731–1738. https://doi.org/10.1007/s13280-020-01499-2.
Turner, R. E., 2022. The discharge of the Mississippi River and tributaries from 1817 to 2020. PLoS ONE 17(12): e0276513. https://doi.org/10.1371/journal.pone.0276513.
Turner, R. E. & N. N. Rabalais, 1991. Changes in the Mississippi River this century: implications for coastal food webs. BioScience 41: 140–147.
Turner, R. E. & N. N. Rabalais, 2003. Linking landscape and water quality in the Mississippi River Basin for 200 years. BioScience 53: 563–572.
Turner, R. E., N. N. Rabalais, D. Justić & Q. Dortch, 2003. Global patterns of dissolved silicate and nitrogen in large rivers. Biogeochemistry 64: 297–317.
Turner, R. E., N. N. Rabalais & D. Justić, 2012. Predicting summer hypoxia in the northern Gulf of Mexico: redux. Marine Pollution Bulletin 64: 318–323. https://doi.org/10.1016/j.marpolbul.2011.11.008.
Turner, R. E., E. M. Swenson, C. Milan & J. M. Lee, 2019. Spatial variations in chlorophyll a, C, N, and P in a Louisiana estuary from 1994 to 2016. Hydrobiologia 834(1): 131–144. https://doi.org/10.1007/s10750-019-3918-7.
USGCRP (U.S. Global Change Research Program), 2017. Climate Science Special Report: Fourth National Climate Assessment, Volume I. Wuebbles, D. J., D. W. Fahey, K. A. Hibbard, D. J. Dokken, B. C. Stewart, T. K. Maycock (eds). https://science2017.globalchange.gov. https://doi.org/10.7930/J0J964J6.
Vandevenne, F., E. Struyf, W. Clymans & P. Meire, 2012. Agricultural silica harvest: have humans created a new loop in the global silica cycle? Frontiers in Ecology and the Environment 10(5): 243–248.
Wanninkhof, R., 1992. Relationship between wind speed and gas exchange over the ocean. Journal Geophysical Research: Oceans 97(C5): 7373–7382.
West, P. C., H. K. Gibbs, C. Monfreda, J. Wagner, C. C. Barford, S. R. Carpenter & J. A. Foley, 2010. Trading carbon for food: global comparison of carbon stocks vs. crop yields on agricultural land. Proceedings National Academy of Sciences (USA) 107: 19645–19648.
Whitmore, A. P., N. J. Bradbury & P. A. Johnson, 1992. Potential contribution of ploughed grassland to nitrate leaching. Agriculture, Ecosystems and Environment 39: 221–233.
Wiebe, A. H., 1931. Dissolved phosphorus and inorganic nitrogen in the water of the Mississippi River. Science 73: 652.
Wollheim, W. M., T. K. Harms, A. L. Robison, L. E. Koenig, A. M. Helton, C. Song, W. B. Bowden & J. C. Finlay, 2022. Superlinear scaling of riverine biogeochemical function with watershed size. Nature Communications 13: 1230. https://doi.org/10.1038/s41467-022-28630-z.
Xu, X., B. R. Scanlon, K. E. Schilling & A. Sun, 2013. Relative importance of climate and land surface changes on hydrologic changes in the US Midwest since the 1930s: implications for biofuel production. Journal Hydrology 497: 110–120. https://doi.org/10.1016/j.jhydrol.2013.05.041.
Yin, S., G. Gao, Y. Li, Y. J. Xu, R. E. Turner, L. Ran, X. Wang & B. Fu, 2023. Long-term trends of water, sediment and nutrient fluxes from the Mississippi River Basin: impacts of climate change and human activities. Journal Hydrology 616: 128822.
Acknowledgements
Support was received from the NOAA Coastal Ocean Program MULTISTRESS Award No. NA16OP2670 to Louisiana State University, a Gulf of Mexico Research Initiative grant to the Coastal Waters Consortium and NSF Award Number (FAIN) 2103843: Large-scale CoPe Hub: Rising Voices, Changing Coasts: The National Indigenous and Earth Sciences Convergence Hub. The Baton Rouge data and all yearly averages used in graphs are available through the Dryad repository at https://doi.org/10.5061/dryad.x95x69pkm, and https://doi.org/10.5061/dryad.x95x69pkm, respectively.
Author information
Authors and Affiliations
Contributions
RET: Conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Resources; Validation; Visualization; Roles/Writing—original draft; Writing—review & editing.
Corresponding author
Ethics declarations
Conflict of interest
The author has no relevant financial or non-financial interests to disclose.
Additional information
Handling editor: John M. Melack
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Turner, R.E. Water quality at the end of the Mississippi River for 120 years: the agricultural imperative. Hydrobiologia 851, 1219–1239 (2024). https://doi.org/10.1007/s10750-023-05383-4
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10750-023-05383-4