Abstract
Nitrate (NO3−) losses from agricultural fields to groundwater and surface waterways are a major concern that could be further exacerbated by a changing climate. Although individual field-scale studies provide critical information, investigation on the interactive effect of various management practices across different soil types experiencing wide variations in precipitation is necessary to extend our understanding of what approaches may mitigate NO3− losses to the environment. Synthesizing and analyzing large data sets from multiple studies provides an opportunity to investigate the interactive impact of multiple management practices, soil texture, and rainfall. We assembled peer-reviewed field studies from the Midwest United States and analyzed their associated field data to (1) quantify the range of NO3− leaching associated with different agroecosystems and (2) determine the individual and interactive effect of management practices (tillage and amount of nitrogen [N] fertilizer added), cropping systems (crop type and rotation), and precipitation across multiple soil types on NO3− leaching. Our results showed that fertilized potatoes (Solanum tuberosum L.) had the highest NO3− leaching rate among all systems studied (59.3 ± 8.4 kg N ha−1 y−1) while unfertilized perennial systems exhibited the lowest NO3− leaching (6.1 ± 0.9 kg N ha−1 y−1). Our results suggested that corn (Zea mays L.)–soybean (Glycine max [L.] Merr.) rotations can reduce NO3− leaching compared to continuous corn by 25% in clay soils and also reduce the impact of high rainfall on NO3− leaching compared to continuous corn management. Nitrate leaching in sandy soils exhibited a greater sensitivity and amplified response to increasing N fertilizer amount and annual precipitation compared to other soil types. Compared to conventional tillage, no-tillage soil management significantly reduced NO3− leaching in sandy and silty loam soils. While some management practices can curb NO3− leaching losses, more drastic land management change from row crops to perennial systems offered the most benefit. We conclude that a changing climate will make it more challenging for farmers to increase N use efficiency and reduce NO3− leaching, especially on coarse textured soils.
Introduction
Nitrate (NO3−) losses from the agricultural land throughout the Midwest United States are a major contributor to water quality degradation in groundwater and regional surface water bodies (Campbell et al. 2021; McIsaac et al. 2010) and also contribute significantly to the long-term hypoxia problem downstream in the Gulf of Mexico (Donner et al. 2004). The Gulf of Mexico Hypoxia Task Force has set a goal of reducing nitrogen (N) transport to the Gulf by 20% by the year 2025 (Mississippi River/Gulf of Mexico Watershed Nutrient Task Force 2022). Increasing precipitation and an increased frequency of extreme, heavy rainfall events associated with climate change will also amplify the risk for nutrient losses from soils from increased runoff, leaching, and soil erosion (Campbell et al. 2021). Identifying effective and economically viable conservation and agricultural management practices is important to achieve water quality improvement goals for both groundwater and surface water. A portion of water infiltrating agricultural fields moves laterally through tile-drained fields and as baseflow to rivers and streams; therefore, reductions in NO3− loss can also help improve water quality at the watershed scale (Ferin et al. 2020; Motew et al. 2017; Raymond et al. 2012; Van Meter et al. 2016). Nitrate leaching from agricultural fields is a complex process controlled by a number of factors such as crop type, timing, and rate of fertilizer applications or tillage practices along with soil type and amount and distribution of rainfall (Spiess et al. 2020).
Nitrogen loss reduction strategies for the Midwest (ISU 2013; IEPA, IDOA, and University of Illinois Extension 2015) identified the reduction of average N application rate and land-use change through conversion of continuous corn (Zea mays L.) and corn–soybean (Glycine max [L.] Merr.) systems to perennial vegetation as the major strategies to achieve N loss reduction goals (Campbell et al. 2021; Donner and Kucharik 2003). The Iowa Nutrient Reduction Strategy (ISU 2013) and Illinois Nutrient Loss Reduction Strategy efforts (IEPA, IDOA, and University of Illinois Extension 2015) recognized the need for investigating the individual and combined effect of different management practices such as land-use, crop rotations, nutrient applications, tillage, and conservation practices at the landscape scale to combat water quality issues.
Previous research has suggested that adhering to recommended application timing and rates of fertilizer N are common practices suggested to limit NO3− loss from the cropping systems (Helmers et al. 2012). Unfortunately, optimal timing and adhering to recommended N fertilizer application amounts do not eliminate the risk of N leaching losses. An increased rate of NO3− leaching with increasing fertilizer N rate has been noted by many studies (Andraski et al. 2000; Brye et al. 2001; Kucharik and Brye 2003), and NO3− leaching loss varies widely with soil texture (Gaines and Gaines 1994) and precipitation (Pasley et al. 2021). Fine-textured soils exhibit a more amplified crop yield response to high N fertilizer rates than for coarse-textured soils (Cambouris et al. 2016; Tremblay et al. 2012). This response highlights the magnified challenges with reducing NO3− leaching from well-drained sandy soils that typically have higher N fertilizer requirements for crops like corn and potato (Solanum tuberosum L.). That challenge is further exacerbated by increasing rainfall and extreme events associated with climate change.
A better understanding of the effect of crop rotations and tillage management practices on limiting NO3− losses from agriculture is needed to achieve water quality goals. Midwest US cropping systems that contain corn receive some of the highest amounts of N fertilizer (USDA NASS 2020), and also have some of the highest NO3− loss values. Weed and Kanwar (1996) and Dougherty et al. (2020) in Iowa reported lower NO3− loss from corn–soybean rotations compared to continuous corn systems, while Helmers et al. (2012) and Ochsner et al. (2010) reported no significant differences in NO3− leaching between corn–soybean rotation and continuous corn in Iowa and Minnesota, respectively. In contrast, Klocke et al. (1999) reported higher NO3− losses with irrigated corn–soybean rotations in Nebraska. Along with cropping systems, the implementation of tillage—or the adoption of conservation or no-tillage (NT)—may have a direct influence on NO3− leaching from the agricultural fields due to the impact on soil water infiltration. Daryanto et al. (2017) reported a greater amount of NO3− leaching under NT compared to conventional tillage (CT) in fine-textured soil but not in coarse-textured soil. In contrast, Rekha et al. (2011) and Waring et al. (2020) noted reduced NO3− concentration in groundwater with NT compared to CT corn–soybean rotation fields in Iowa. Angle et al. (1993) reported lower NO3− concentrations under NT plots than CT plots in moderately well-drained fields in continuous corn system, and these differences were greater at higher fertilizer N rates. Finally, other researchers have reported no significant difference in NO3− leaching between NT and CT plots in poorly drained soils (Dougherty et al. 2020; Jabro et al. 2019).
The threat of increased precipitation and a higher frequency of extreme rainfall events associated with climate change will likely exacerbate the risk of NO3− leaching from agricultural systems, but the impacts will depend on cropping system, soil texture, tillage, and the timing and amount of N fertilizer applied (Martinez-Feria et al. 2019). Previous work has illustrated that wetter conditions can increase NO3− leaching (Martinez-Feria et al. 2019; Pasley et al. 2021) while drier conditions or drought might reduce denitrification and increase soil NO3− accumulation (Greaver et al. 2016). Alternating wet and dry conditions may promote more tightly coupled nitrification and denitrification processes leading to buildup and flushing of soil NO3− depending on the ratio of transport to reaction rates (Greaver et al. 2016).
The aforementioned results represent a small subset of a much larger number of field studies that have quantified NO3− leaching associated with various agroecosystems and management practices in the Midwest United States. The number of published studies has continued to grow rapidly given the expanding challenge to improve water quality while continuing to increase food production in a changing climate. Although these previous studies provide important information, they often focus on a single question regarding the impact of a specific management practice on a particular soil type at a specific location. Aggregating these data as part of a larger synthesis may allow for new information to emerge that illustrates a more complete story about the impacts of management practices (e.g., tillage, N fertilizer amount, and agroecosystem types), but also how these interact with soil texture and precipitation to affect NO3− leaching losses. This type of information is particularly important to understand the likely effects of increased precipitation on leaching losses.
In this synthesis study, we reviewed the peer-reviewed literature and investigated how soil texture, precipitation amount, and management practices such as N fertilizer rate, cropping systems and type, and tillage practices affect NO3− leaching from common agroecosystems found in the Midwest United States. We addressed the following questions:
What is the range of NO3− leaching associated with a variety of Midwest agroecosystems?
How is NO3− leaching in continuous corn and corn–soybean rotations impacted by tillage practice, rate of N fertilizer applied, and soil textural type?
What is the relationship between NO3− leaching loss and precipitation in corn and corn–soybean rotations, and are there interactive effects with tillage practice, soil type and N fertilizer rate?
Materials and Methods
A search of the peer-reviewed literature was conducted using Web of Science (https://www.webofknowledge.com) and Google Scholar (https://scholar.google.com) online databases to identify studies that reported annual NO3− leaching losses in agroecosystems of the Midwest United States that were measured in the states of North and South Dakota, Nebraska, Kansas, Iowa, Minnesota, Wisconsin, Illinois, Indiana, Ohio, Missouri, and Michigan. Articles were screened for meeting the following criteria that warranted inclusion in our analysis: the study (1) was conducted in one of the aforementioned 12 Midwest states; (2) reported annual cumulative NO3− leaching or NO3− concentration; (3) was conducted as part of a field experiment using lysimeters or analysis of water samples from drainage tiling; and (4) included additional information on soil texture, tillage practices, cropping systems, crop or vegetation type, yield (where appropriate), fertilizer rate, and total precipitation. This yielded 325 articles of which 62 unique field studies were published in the scientific literature from 1975 to 2020, and a total of 1,411 site-years of data observations (see the supplementary materials for a full biography and site list). WebPlotDigitizer version 4.5 (Rohatgi 2017) was used in some cases for data extraction from published figures.
Nitrate leaching loss measurements were collected using different types of lysimeters and through water sampling from fields that were tile-drained (table S1). For studies that provided flow-weighted annual NO3− concentration, total annual NO3− loss was calculated by multiplying the NO3− concentration by volume of water flow (Randall et al. 1997). Soil texture was categorized into fine, medium, and coarse textural classes and referred to as clay, silty loam, and sandy soil. Silty clay loam was grouped with silty loam. A subset of studies (n = 17) only provided growing season precipitation data, and five others did not provide any precipitation data. We gap-filled missing monthly precipitation data for sites from either nearby National Oceanic and Atmospheric Administration (NOAA) weather stations (NOAA 2022) or via the PRISM Climate Group online database (PRISM Climate Group 2022). Annual rainfall and supplemental irrigation totals for each site and year were grouped into the following three categories: low (<81 cm y−1), medium (81 to 100 cm y−1), and high (>100 cm y−1) (table 1). All five studies conducted in potato field received irrigation. Nearly 12% of the site year data observations received irrigation.
Out of 1,411 site year data observations, 844 observations were fertilized, of which around 86% observations received N source from synthetic fertilizer. Urea ammonium nitrate (UAN), ammonium nitrate (NH4NO3−), anhydrous ammonia (NH3), and urea were the major fertilizer sources in the studies we compiled for our analysis. Around 67% of the studies used single application of fertilizer while 33% of the observation used split application. A majority of the studies used injection and broadcasting method, while a few studies used band placement. For some statistical analyses, annual rates of N fertilizer applications were grouped into three categories: low (<80 kg N ha−1 y−1), medium (80 to 170 kg N ha−1 y−1), and high (>170 kg ha−1 y−1) (table 1). This created three groups with a nearly equal number of data points for statistical comparison. For statistical analysis and creating figures 1 to 3 and tables S1, S2, and S3, all site year data from all agroecosystems were used (1,411 site years), and for remaining analyses, we limited data to continuous corn and corn–rotation systems (1,175 site years). To determine statistical significance, we used the R statistical package and carried one- or two-way analysis of variance (ANOVA) followed by Tukey’s honestly significant difference (HSD) test.
Results and Discussion
Annual Soil Nitrate Leaching across Cropping Systems. The fertilized potato received an average rate of 238.8 ± 9.5 kg N ha−1 fertilizer annually and lost the largest amount of soil NO3− among all systems (59.3 ± 8.4 kg N ha−1 y−1; figure 1 and table S2). When analysis was limited to just fertilized systems, fertilized continuous corn lost an average of 11.5 kg N ha−1 y−1 more compared to fertilized corn–soybean rotation, but those differences were not significant (figure 1 and table S2). Nitrate leaching from fertilized corn–alfalfa (Medicago sativa L.) systems, which received an average of 118.6 ± 1.4 kg N ha−1 y−1, was similar to continuous corn and corn–soybean rotations even though they received 72 kg N ha−1 y−1 less N fertilizer annually (table S2). Unfertilized perennial systems, which included poplar, grass, bioenergy crops (switchgrass [Panicum virgatum L.] and miscanthus [Miscanthus giganteus]) had the lowest annual NO3− leaching (6.1 ± 0.9 kg N ha−1 y−1) values among all systems (figure 1, table S2). All systems except for perennial system lost more than 22 kg N ha−1 y−1 even when they were not fertilized (figure 1 and table S2).
Annual Soil Nitrate Leaching by Individual Crop Type. On average, the largest annual NO3− leaching loss came from fertilized potato fields (59.3 ± 8.4 kg N ha−1 y−1) and was significantly different from all other crop types and other perennials measured (p < 0.01; figure 2 and table S3). Second to potato, fertilized alfalfa or clover (Trifolium spp.) also lost 44.3 ± 8.8 kg N ha−1 y−1. Fertilized corn crops lost around 20% more soil NO3− than soybean crops that did not receive any fertilizer (figure 2 and table S3), but those differences were not significant. Nitrate leaching from bioenergy crops (miscanthus and switchgrass), grasses, and poplar trees were all less than 12 kg N ha−1 y−1 even though grasses and bioenergy crops received some modest amounts of N fertilizer (figure 2 and table S3).
Impacts of Soil Texture and Nitrogen Fertilizer Rate on Nitrate Leaching in Continuous Corn and Corn–Soybean Rotations. Annual NO3− leaching was impacted by soil textural type (p < 0.01; table S4). Sandy soils had the largest annual NO3− leaching (57.9 ± 3.8 kg N ha−1 y−1), followed by silty loam (32.6 ± 1.4 kg N ha−1 y−1) and clay (27.6 ± 0.8 kg N ha−1 y−1) (table S4). The average annual N fertilizer application for crops grown on these soils was 120.6, 139.3, and 93.3 kg N ha−1 y−1 for sandy, silty loam, and clay soils, respectively (table S4). Sandy soils exhibited a more substantial increase of NO3− leaching as N fertilizer amounts increased that was not present in silty loam and clay soils (figure 3 and tables S4 and S5). The rate of NO3− leaching increase with increasing N fertilizer was similar for silty loam and clay soil (figure 3). To analyze the interaction between soil texture and N fertilizer rate, annual N fertilizer application rates were divided into three groups: low (<80 kg N ha−1), medium (80 to 170 kg N ha−1), and high (>170 kg N ha−1). There was a significant, interactive effect of soil texture and fertilizer amount on NO3− leaching (p < 0.01). Nitrate leaching generally increased with increased fertilizer N application across all three soil types, but sandy soils had the highest NO3− leaching compared to clay and silty loam soil across all N fertilizer amounts (table 2). Sandy and clay soils had significantly different rates of NO3− leaching between the low and high categories, but silty loam soils did not (table 2). Nitrate leaching from sandy soil was 93%, 64%, and 116% higher than clay soil at <80 kg N ha−1, 80 to 170 kg N ha−1, and >170 kg N ha−1 N fertilization rates, respectively (p < 0.01; table 2). Similarly, NO3− leaching from sandy soils was 54%, 71%, and 100% higher than silty loam soil at <80 kg N ha−1, 80 to 170 kg N ha−1, and >170 kg N ha−1 N fertilization rates, respectively (p < 0.01; table 2). Although soil NO3− leaching increased at high fertilizer N application rates (>170 kg N ha−1) compared to medium fertilizer N application rates (80 to 170 kg N ha−1) in clay and silty loam soil, a significant increase (66% increase, p < 0.01) was only observed on sandy soil (table 2).
When multiple year crop rotations containing corn were analyzed, a significant interaction (at the p < 0.1 level) between soil texture and cropping system (continuous corn or corn–soybean rotation) on NO3− leaching was determined (p = 0.08; table 3). The presence of corn–soybean rotations reduced NO3− leaching by 25% compared to continuous corn in clay soils, but this effect was not found on sandy and silty loam soils (table 3). In continuous corn systems, although sandy soils received 7% less fertilizer N than clay soil and 30% less fertilizer N than silty loam soils, sandy soils lost 65% more soil NO3− than clay soils and 77% more than silty loam soils (table 3). Similarly in the corn–soybean rotation system, although sandy soils received 15% less fertilizer N than clay soil and 33% less fertilizer N than silty loam soils, sandy soils lost 118% more soil NO3− than clay soils and 77% more than silty loam soils (table 3).
Tillage Management Effects on Nitrate Leaching Affected by Soil Texture in Continuous Corn and Corn–Soybean Rotations. On average, NT crop management reduced NO3− leaching by 22% compared to CT (p < 0.01; table 4). The majority of this impact occurred for the highest N fertilizer rates (>170 kg N ha−1) (table S6). The type of cropping systems (continuous corn or corn–soybean rotation) did not have any significant interactive effect with tillage type on NO3− leaching (table S7).
NT management appeared to be most effective in reducing NO3− leaching on sandy and silty loam soils compared to clay soil (figure 4 and table S8). There was significant interaction between tillage and soil texture on NO3− leaching (p < 0.01). NT management reduced annual NO3− leaching by 20% in silty loam and 55% in sandy soil compared to CT (figure 4 and table S8). The average annual fertilizer N application in NT was 21% and 13% less in silty loam and sandy soil, respectively, compared to CT (table S8). Average annual NO3− leaching with NT was not significantly different across textural types, but this was in contrast to the response observed for cropping systems managed with CT (figure 4 and table S8).
Effect of Annual Precipitation on Nitrate Leaching Loss in Continuous Corn and Corn–Soybean Rotations. Increasing NO3− leaching losses were driven by increasing annual precipitation (table 5). Across all observations, the annual NO3− leaching was 22% higher for medium annual precipitation (81 to 100 cm) compared to the lowest annual precipitation category (<81 cm), and leaching was 33% higher for the highest annual precipitation category (>100 cm) compared to medium annual precipitation (81 to 100 cm) regardless of fertilization rate (table 5). Although high precipitation (>100 cm) had higher NO3− leaching compared to medium precipitation (81 to 100 cm) in clay and silty loam soils, those differences were not significant. However, sandy soil lost 48% more soil NO3− at high annual precipitation (>100 cm) compared to medium precipitation (81 to 100 cm) despite receiving 33% less fertilizer N compared to medium precipitation (81 to 100 cm) (table 6). Clay soil lost significantly lower soil NO3− at low precipitation (<81 cm) compared to medium precipitation (81 to 100 cm), but there were no significant differences in NO3− leaching between low (<81 cm) and medium precipitation (81 to 100 cm) in silty loam and sandy soil (table 6). Although the overall effect is that increasing precipitation caused higher NO3− leaching, there was no significant interaction between precipitation and N fertilizer rate on NO3− leaching (p = 0.15; table 7). The highest average NO3− leaching (46.7 ± 2.3 kg N ha−1 y−1) was observed with the combination of highest precipitation (>100 cm) and highest amount of annual average fertilizer (>170 kg N ha−1 y−1) (table 7).
Although NT has lower annual NO3− leaching values compared to CT at all three levels of annual precipitation amounts, there was no significant interaction between precipitation and tillage management on NO3− leaching (p = 0.13; table 8). The lowest average NO3− leaching (15.9 ± 2.0 kg N ha−1 y−1) was observed with the combination of lowest precipitation (<81 cm) and NT (table 8), while the highest average NO3− leaching (44.0 ± 1.8 kg N ha−1 y−1) was observed with the combination of highest precipitation (>100 cm) and CT (table 8).
Corn–soybean rotations had 34% less NO3− leaching compared to continuous corn system at high precipitation (>100 cm) values, but no significant differences were detected at low (<81 cm) and medium (81 to 100 cm) precipitation values (figure 5 and table S9). In continuous corn, there was no significant difference in NO3− leaching between low precipitation years (<81 cm) and medium (81 to 100 cm) annual precipitation, but soil NO3− leaching was 66% higher at high (>100 cm) precipitation years compared to low (<81 cm) and medium precipitation (81 to 100 cm) years. For the corn–soybean rotation systems, there was no significant difference in NO3− leaching between low and medium annual precipitation (81 to 100 cm), but was 21% higher at high (>100 cm) precipitation years compared to medium precipitation (81 to 100 cm) (figure 5 and table S9).
Effects of Nitrogen Fertilizer Rate on Nitrate Leaching and Yield in Corn. Quadratic model fits through all observations for corn grown either in continuous rotation or in rotation with soybean (n = 797) illustrated an expected increasing yield response to fertilizer, with a plateau after a N fertilizer rate of 250 kg N ha−1 (p < 0.01; figure 6). The corresponding NO3− leaching response exhibited a nonlinear increase with increasing N fertilizer amounts with no observed limit or plateau (figure 6). It should be noted that the average annual NO3− leaching at zero fertilizer N was 29.2 kg N ha−1 and the corresponding average annual corn yield was 5.0 Mg ha−1. Based on model responses, the increase of fertilizer N from 0 to 150 kg N ha−1 increased the NO3− leaching by 3.1 kg N ha−1 and corn yield by 4.1 Mg ha−1 while the additional 150 kg application of fertilizer N (i.e., from 150 to 300 kg N ha−1) increased the NO3− leaching by 11.95 kg N ha−1 and corn yield by only 1.2 Mg ha−1 (figure 6 and table S10). This illustrates how much more sensitive NO3− leaching losses are to fertilizer increases compared to corn yield gains.
Perennial Agroecosystems Can Help Reduce Nitrate Leaching from Agricultural Landscapes. The Midwest United States is a major producer of corn and soybean in the world and NO3− leaching to groundwater and transport to the Gulf of Mexico remains a difficult problem to solve. The Gulf of Mexico Hypoxia Task Force has set a goal of significant reductions in N and phosphorus (P) loads (45%) to reduce the size of the hypoxic zone to 5,000 km2 by the year 2035, with an intermediate goal of a 20% reduction in N and P loading by 2025 (Mississippi River/Gulf of Mexico Watershed Nutrient Task Force 2022). To do so, it is important to identify cropping systems and management practices that can help achieve this goal and overcome the presence of legacy stores of N in soils and groundwater that have built up given the extent of corn grown on the landscape each year (Campbell et al. 2021; Van Meter et al. 2016). Continuous corn and corn–soybean cropping systems account for about 72% of cultivated land in the top 12 corn-producing states in the United States (USDA NASS 2020). Outside of potato crops, which are grown on a smaller number of hectares in the Midwest, continuous corn and corn–soybean rotations exhibited the highest rates of NO3− leaching in this study, and even in unfertilized cases in this study, nearly 30 kg N ha−1 is lost each year. This illustrated the presence of carry-over and legacy N that is present in these soils given long-term historical management (Campbell et al. 2021; Van Meter et al. 2016). Helmers et al. (2012) noted that greater than 10 mg L−1 NO3−-N concentration (the US Environmental Protection Agency threshold for safe drinking water) in subsurface drainage water was found when recommended N fertilizer application rates for corn were applied, and concentrations were 5 mg L−1 even with no fertilizer applications in corn–soybean rotation and continuous corn. This illustrates the difficulty in obtaining NO3− concentration less than the drinking standard even with recommended N management practices are employed.
Our meta-analysis, however, showed very clearly how alternative cropping systems such as perennials (grasses, trees, and bioenergy crops) can help meet water quality goals in the future and increase the likelihood that landscapes are more resilient to increased intensity of rainfall associated with a changing climate (Campbell et al. 2021; Liebman et al. 2013; Schulte et al. 2017). These systems had significantly lower N losses from leaching compared to traditional row crops at a magnitude that, for average recharge rates in the Midwest, would help push NO3−-N concentrations in groundwater below the 10 mg L−1 safe drinking water threshold. These results also follow other previous research by Campbell et al. (2021), Masarik et al. (2014), Brye et al. (2001), and Hussain et al. (2019). Other strategies that include perennials such as inclusion of legumes in cool season grass pastures (Jackson 2020), wheatgrass systems (Jungers et al. 2019), filter strips (Zhou et al. 2010), and managed intensive grazing systems (Gilker and Weil 2017) also have been shown to reduce NO3− leaching. Perennial systems also help increase resilience of landscapes to climate change and the increased risk of soil erosion and loss of nutrients to waterways (Campbell et al. 2021; Liebman et al. 2013; Schulte et al. 2017). However, with a small number of studies cited as part of our analysis, alfalfa, which is an N-fixer, received an average of 120 kg N ha−1 from manure additions and therefore contributed to higher than expected NO3− leaching loss.
Corn–Soybean Rotations Reduce Nitrate Losses Compared to Continuous Corn in a Changing Climate. This study suggests that corn–soybean rotations can reduce NO3− leaching compared to continuous corn, although the effect was limited to clay soils. Furthermore, NO3− leaching in general increased in these systems as annual precipitation increased, which suggests that trends toward wetter conditions with a changing climate will exacerbate the NO3− leaching problem. Weed and Kanwar (1996) and Dougherty et al. (2020) also reported lower NO3− loss from corn–soybean rotation compared to continuous corn systems in clay soils of Iowa. This is generally attributed to continuous corn receiving N fertilizer every year while corn–soybean rotations only have N applied every other year. Bakhsh et al. (2005) also reported lower NO3− leaching in corn–soybean rotation compared to continuous corn applied with swine manure, but no significant difference in NO3− leaching when mineral fertilizer was applied. Helmers et al. (2012) in clay soil of Iowa and Ochsner et al. (2017) in silty loam soil in Minnesota reported no significant difference in NO3− leaching between corn–soybean rotation and continuous corn. In contrast to our results, Klocke et al. (1999) reported higher NO3− leaching with irrigated corn–soybean rotation in silt loam soils in a semiarid climate of Nebraska. One potential reason for higher or comparable NO3− leaching rates from corn–soybean rotations in relation to continuous corn is faster decomposition of soybean litter as reported by Hall et al. (2019). However, given the type of weather after Midwest soybean harvest, slower decay rates generally occur, and the nutrients released from decomposed litter is made available for uptake by corn the following year (Needelman et al. 1999).
Corn–soybean rotations reduced the impact of high rainfall on NO3− leaching compared to continuous corn. Greater susceptibility of continuous corn to high rainfall compared to corn–soybean rotation is likely due to a higher concentration of soil NO3− available for leaching after addition of N fertilizer. The timing of spring N fertilizer additions in corn often coincide with high soil moisture levels after snowmelt and low rates of evapotranspiration, and at a time of year when extreme rainfall events are also possible as the severe weather season commences with warming temperatures in April and May across the Midwest. Besides lower rates of NO3− leaching in corn–soybean rotations, previous research also supports increased corn yields in corn–soybean rotations compared to continuous corn. Corn rotated with soybean has been shown to produce 21% higher yields than continuous corn despite receiving 25% less N fertilizer input (Helmers et al. 2012). Corn yields are also higher in corn–soybean rotation compared to continuous corn when using a comparable fertilizer rate (Poffenbarger et al. 2017). Recommended rates of N fertilizer for continuous corn are generally higher than those for corn–soybean, where recommendations usually account for a N credit from the N-fixing soybean part of the rotation. For example, recommended N application rates in Iowa for corn in a corn–soybean rotation are 112 to 168 kg N ha−1 compared with 168 to 224 kg N ha−1 for corn in a continuous corn system. At the recommended rate, continuous corn lost more soil NO3− compared to corn–soybean rotation (Helmers et al. 2012). These results suggest that there is greater opportunity of reducing NO3− leaching with corn–soybean rotation by reducing the amount of N fertilizer applied to corn in corn–soybean rotation to achieve equivalent yield to continuous corn.
No-Tillage Crop Management Can Reduce Nitrate Leaching Compared to Conventional Tillage. Compared to CT, NT was effective in reducing NO3− leaching in sandy and silty loam soils, but the impact was not present in clay soils. A meta-analysis by Daryanto et al. (2017) reported less leachate NO3− concentration with NT compared to CT in coarse and medium textured soil but increased leachate NO3− concentration in fine textured soil. A greater fraction of water-stable aggregates in NT compared to CT soil likely reduces the NO3− leaching from NT soil compared to CT soil (Bruce et al. 1990). Higher residual NO3− in the surface 20 cm with NT compared to CT has been reported (Canisares et al. 2021; Randall and Iragavarapu 1995), indicating better N retention capacity with NT soil management. NT supports increased retention of NO3− in the soil matrix (Gicheru et al. 2004; Luo et al. 2010; Shrestha et al. 2019; Zhu and Fox 2003). This is attributed to increased soil organic matter (SOM), which immobilizes the available soil NO3− more efficiently than in CT soils (Gicheru et al. 2004; Luo et al. 2010). More significant SOM decomposition with tillage has been reported in sandy soil compared to clay soils; this indicates higher N mineralization that increases the likelihood of N leaching. The increased SOM decomposition in tilled soil was reported by Abdalla et al. (2016) in a meta-analysis as shown by larger differences in soil carbon dioxide (CO2) respiration between CT and NT in sandy soil compared to clay soil. Other studies have reported that soil C in the sand fraction is more easily decomposable and depleted upon soil disturbance (Six et al. 1999), and that the chemical composition of SOM in silt and clay fractions is not drastically influenced by changes in soil management and land use (Christensen 1996). The increased N mineralization coupled with increased drainage in CT in sandy soils increases the likelihood of NO3− leaching (Bruce et al. 1990). However, in loamy soil, Hess et al. (2020) reported decreased drainage in CT soil likely due to breaking up of soil aggregates, worm channels, and other preferential flow paths leading to potentially fewer macropores. These previous findings suggest that sandy soils would have a more robust reduction in NO3− leaching associated with NT management compared to more fine-textured soils. Research by Kucharik and Brye (2003) and Masarik et al. (2014) also supports this conclusion, as they reported no significant difference in cumulative NO3− leaching between NT and CT over eight years in a well-drained silty loam soil in southern Wisconsin.
Increased Precipitation Increases the Risk of Nitrogen Leaching Loss and Exacerbates Groundwater Quality Issues. Increasing precipitation escalated the rate of N losses from corn and corn–soybean cropping systems, with soil texture playing a secondary but important role in leaching response. Sandy soils exhibited a substantial increase in NO3− leaching with increase in N fertilizer application (table 2) and high precipitation (>100 cm) (table 6) that generally was not present in silty loam and clay soils. Poor yield response of sandy soil to high N fertilizer compared to clay soil (Tremblay et al. 2012) may also indicate increased risk of N losses associated with increasing precipitation. This suggests that farmers growing potato and corn, which have the highest recommend fertilizer rates on sandy soils, will be increasingly challenged by increasing precipitation and an increasing frequency of heavy rainfall events—both associated with a changing climate—to keep N fertilizer in the root zone and out of groundwater. However, given the results in this study, it is likely that increasing precipitation will challenge all Midwest farmers across all soil types given the likelihood of increased leaching and N loss. Interestingly, no significant difference in NO3− leaching was observed in silty loam with the change in annual precipitation amounts. The variable response of different soil textures to increased precipitation suggests the need for soil texture-specific nutrient management strategies to reduce the risk of NO3− loss in a changing climate. The anticipated increase in rainfall intensity and change in rainfall patterns may exacerbate the risk of NO3− leaching. The accumulation of N in the soil during low rainfall until a following year when normal or above normal precipitation returns, can cause an increased flush of N out of the system, sometimes referred to as a “weather whiplash” effect; the 2012 drought followed by the wet spring of 2013 is often used as a key example of this phenomenon on water quality problems (Loecke et al. 2017).
Continuous corn systems appear more prone to NO3− leaching in higher precipitation (>100 cm) years compared to corn–soybean rotation, but significant differences were not observed in medium (81 to 100 cm) and low precipitation (<81 cm) years. Pasley et al. (2021) also reported a higher risk of NO3− leaching from continuous corn compared to corn–soybean rotation. Both corn–soybean and continuous corn see increases in leaching with increasing rainfall >81 cm, but corn–soybean rotations have significantly less N leaching than continuous corn in the highest precipitation category. The average rate of N leaching for continuous corn at highest rainfall is one of the highest rates of leaching (for any categorical analysis) outside of potato. Finally, both CT and NT show significant increases in leaching with increasing rainfall for comparable annual fertilizer applied to the landscape. Therefore, we conclude that while NT can help reduce leaching, it can’t eliminate the risk of increasing NO3− leaching with increasing rainfall.
Tradeoffs between Increasing Yields and Water Quality. Our analysis illustrated the magnified tradeoff between small increases in corn yield but large increases in NO3− leaching associated with increasing N fertilizer amounts, especially above 150 kg N ha−1 y−1. The high variability in NO3− leaching response to N rate showed the critical role of management, soil, and climatic factor on the effect of N rate on NO3− leaching. Increased rates of NO3− leaching with increasing fertilizer rate have also been reported in other studies (Andraski et al. 2000; Kucharik and Brye 2003; Pasley et al. 2021). The associated NO3− leaching curve has been a key missing piece in guidance provided to farmers concerning recommended amounts of N fertilizer applications to Midwest crops, especially corn, and protection of water quality, and will be even more valuable as the frequency of heavy rainfall events continues to escalate in a changing climate. Unfortunately, adhering to N fertilizer recommendations does not eliminate the risk of N leaching from agricultural systems. The yield response to fertilizer and the economics of maximum return to N, or determination of the fertilizer rate that maximizes profit given current cost of fertilizer the price of commodity crops, is what currently guides recommendations and decision-making in the field (Morris et al. 2018; Zhao et al. 2017). Nitrate leaching response curves can be used to help guide policy decision-making charged with protecting and improving water quality. An improved understanding and quantification of how much NO3− leaching can be reduced with less N fertilizer can be used to develop incentive programs (or reassign subsidies) for farmers to offset a reduction in crop yield that could result from lower N inputs. Given the amount of spread in data points illustrated in this meta-analysis, response functions will need to be fine-tuned with numerical models and other decision support tools to account for large-scale variability in climate, soils, land-use history, and current management.
The picture is very clear that additional N fertilizer may increase yield above an optimum recommended amount but the economic return does not, and those small yield increases are often much smaller than what a farmer experiences from year-to-year due to weather variability and other pressures that reduce yield (e.g., pests, disease, weeds, etc.). At some level of N fertilizer and beyond the yield response plateau, the amount of N loss continues to increase significantly and challenges water quality improvement. As climate change continues to increase rainfall and extreme events, the likelihood of increased nutrient losses from agricultural fields will continue to escalate (Campbell et al. 2021; Motew et al. 2018).
Shortcomings with this Meta-Analysis. Nitrate leaching data was collected using a variety of observational methods—including a variety of different lysimeters and water sampling—that can lead to increased uncertainty with results and may lead to some debate as to what type of measurement approaches are most accurate. Some measurement methods are preferred in a certain type of soil and region; for example, analysis of water samples from drainage water is common in fields with clay soil and with tile or subsurface drain. Carry-over N effects from previous years suggested that unfertilized treatments are not decoupled from previous land management history and may not produce results that are significantly different from their companion treatments that receive fertilizer. The method of fertilizer application (e.g., split versus single broadcast) may also cause variations in the amount of N loss. Finally, although field-scale studies provide critical information, when they are pooled together it is clear that large spatial variability in NO3− leaching is present across the Midwest US region, and that is likely due to additional factors (soil, weather, and management) that were not analyzed as part of this study. Crop rotations can reduce inorganic N fertilizer needs and at the same time reduce the amount of N available for leaching, both of which are important to farmers in the western Corn Belt.
Summary and Conclusions
This study confirmed that increased precipitation leads to an increased rate of NO3− leaching losses from agricultural fields in the Midwest that are dominated by continuous corn and corn–soybean rotations. Therefore, climate change will continue to exacerbate water quality problems through an increased frequency of heavy rainfall events and more precipitation in general. Clearly, the introduction of perennial agroecosystems across the landscape is the best management option to increase resilience to a changing climate and protect soil and water resources from further degradation and meet water quality goals. While we identified that changes in tillage practices from CT to NT, expansion of crop rotations (corn–soybean), elimination of continuous corn, and reducing N fertilizer applied will help reduce the amount of N leaching losses from Midwest agroecosystems, those reductions will fall short of the magnitude needed to significantly improve water quality across agricultural regions, or shrink the Gulf of Mexico Dead Zone. Furthermore, specialty crops grown on sandy soils like potato that have high N inputs and rapid recharge will pose an even greater challenge to farmers in regions with elevated rainfall where the likelihood of leaching losses is highest. At this critical time in recent history when we’ve seen increasing precipitation coupled with an expansion of corn cropland at the expense of Conservation Reserve Program land to support corn ethanol, we are trending in the wrong direction. Hopefully, this study and others that follow can help clarify the increased risk of NO3− leaching that comes with row crops, and that applying even more N fertilizer to the landscape puts our drinking and surface water at even higher risk of degradation in a changing climate.
Supplemental Material
The supplementary material for this article is available in the online journal at https://doi.org/10.2489/jswc.2023.00048.
Acknowledgements
This work was supported by a grant from the Wisconsin Department of Natural Resources (DNR) and as part of the National Science Foundation Innovations at the Nexus of Food, Energy, and Water Systems (INFEWS) program (award number 1855996). We thank Brian Austin and Bruce Rheineck at the Wisconsin DNR for spearheading efforts to study nitrate leaching and groundwater quality in Wisconsin and securing financial support for this work.
- Received March 28, 2022.
- Revision received August 8, 2022.
- Accepted October 7, 2022.
- © 2023 by the Soil and Water Conservation Society