|
Effects of Shallow Surface Drainage Ditches With Controlled Subsurface Drainage Management on Crop Yields in North Carolina
Mitchell L. Watkins1, Chad Poole1,*, Mohamed A. Youssef1, Hossam Moursi1, Rachel Vann2, Ron Heiniger2
Published in Journal of the ASABE 67(2): 349-361 (doi: 10.13031/ja.15537). Copyright 2024 American Society of Agricultural and Biological Engineers.
1 Biological and Agricultural Engineering, North Carolina State University, Raleigh, North Carolina, USA.
2 Crop and Soil Sciences, North Carolina State University, Raleigh, North Carolina, USA.
* Correspondence: chad_poole@ncsu.edu
Submitted for review on 15 January 2023 as manuscript number NRES 15537; approved for publication as a Research Article and as part of the “Advances in Drainage: Selected Works from the 11th International Drainage Symposium” Collection by Community Editor Dr. Zhiming Qi of the Natural Resources & Environmental Systems Community of ASABE on 7 November 2023.
Highlights
Abstract. Agricultural drainage in the coastal areas of North Carolina (NC) is commonly achieved through large trapezoidal-shaped ditches. The coastal region of NC has limited topographic relief (slopes < 1%) with poorly drained soils that can cause substantial issues with surface water ponding during high-intensity or long-duration precipitation events without some form of surface drainage. Installation of large free flowing surface ditches (FD) with field crowning improves the drainage intensity but can create negative consequences such as over drainage and side slope scouring within the ditch. Large open ditches remove tillable land from production and serve as a primary transport pathway for pollutants. An alternative drainage design (SD) has been implemented that decreases the size of the surface ditches, limiting their drainage effect to only surface water and potentially improving equipment trafficability. The smaller ditches, installed with precision grade equipment, are placed on a grade sufficient to direct surface flow while keeping soil movement to a minimum. Lateral subsurface drainage tiles are installed to provide subsurface drainage and are connected to a main tile line operated with an outlet control structure for controlled drainage (CD). This study evaluates the crop yield and water table effects of the SD system compared to FD over nine crop seasons from 2014-2022. The SD treatment increased yield in eight of the nine crop seasons overall, four of five corn (Zea mays L.) crops, and all four soybean (Glycine max L.) crops. Overall, SD increased corn yields by 0.4 Mg/ha or 4% (0.7 Mg/ha or 6.6% with the exclusion of 2016) and soybean yields by 0.5 Mg/ha (14.3%). The effects of SD on crop yield and water table show that the system can be utilized to improve crop health and provide better management of cropland for producers.
There is a growing demand worldwide for food and fiber, which underscores the need for increasing the efficiency of crop production systems, while minimizing negative impacts on downstream surface water bodies. In the United States, more than 40 million hectares of land are artificially drained (USDA-NASS, 2022), out of which 751 thousand hectares are in North Carolina (37% of the state’s cropland). Enhancing drainage of this land improves field trafficability and helps protect crops from excess water conditions in the root zone (Skaggs et al., 2012). Improving drainage artificially is achieved by improving surface and/or subsurface drainage, and the methods for drainage improvement tend to vary with geographic locations.
Subsurface drainage is typically achieved by installing subsurface pipe drains called “tiles” or digging open ditches to remove excess water from the soil profile. Modern subsurface tiles consist of flexible plastic pipes with engineered slits allowing for water entry with minimal sediment (Ghane, 2022). Research has shown that artificial drainage increases crop yield, but it also accelerates the loss of nutrients, particularly nitrate-nitrogen, via drainage water (Jaynes et al., 1999; Randall and Gross, 2008; Poole et al., 2018; Helmers et al., 2022; Moursi et al., 2023). Excessive nutrient losses from drained cropland to surface water bodies have been identified as the leading cause of hypoxic conditions of major aquatic ecosystems, including the Gulf of Mexico and the Chesapeake Bay (Rabalais and Turner, 2019; Zhang et al., 2021). Therefore, drainage system design and management should consider both crop yield and water quality benefits.
Surface drainage is also necessary in geographic locations such as the coastal region of North Carolina (NC), where the landscape is extremely flat and precipitation intensity is greater than the designed removal rate of most subsurface drainage systems (Carter, 1999). Improved drainage in eastern NC is achieved by field ditches, typically 1 to 1.5 m deep and up to 1.5 m wide, which facilitate both surface and subsurface drainage (fig. 1). These field ditches are usually spaced 100 m apart in a parallel pattern and discharge into a main drainage canal.
Ditches remove surface water during periods of high-intensity rainfall, which can cause significant damage to crops. However, ditch spacing is governed in many cases by equipment size and thus may not provide adequate subsurface drainage intensity to minimize plant stress due to excess water conditions. It is a common practice to enhance surface drainage by land forming (e.g., surface crowning) to compensate for the low subsurface drainage intensity provided by widely spaced field ditches (Carter, 1999; Fangmeier et al., 2006). The relatively steep side-slopes and lack of vegetative cover cause excessive erosion and scouring issues with the relatively deep field ditches (fig. 2). Subsequently, field ditches usually require yearly maintenance and cleaning, which increases the operation cost and the negative water quality impacts of the drainage systems. A viable alternative to traditional open-ditch drainage is needed because of the low subsurface drainage intensity, the cost of seasonal maintenance, and the cost of associated surface land forming.
 |
| Figure 1. Excavation for maintenance of a conventional surface drainage ditch in North Carolina (NC). |
 |
| Figure 2. Conventional surface drainage ditch in North Carolina (NC) following fourteen years of production and cleaning. |
The need to limit crop yield losses due to temporal precipitation variations and the need to limit nutrient loading to receiving water bodies have led to the development of controlled drainage (CD). The need for drainage may vary depending on the weather conditions during the growing season. Controlled drainage is implemented by installing a control gate or riser at the outlet point of the drainage system to restrict drainage outflow and elevate the water table. Controlled drainage stores more water in the soil profile, which contributes to meeting crop water demand and subsequently increases crop yield during relatively dry growing seasons (Skaggs et al., 2012; Youssef et al., 2023). Some studies have shown that CD has a positive effect on crop yield, with an increase in yield of up to 10% in corn and soybeans (Ghane et al., 2012; Poole et al., 2013; Wesstrom and Messing, 2007). However, other studies suggest CD provides no significant difference or a slight reduction in yield (Fausey, 2005; Helmers et al., 2012). Differences in yields from CD are commonly reflective of differences in wet and dry growing seasons (Youssef et al., 2023). The yield benefits of CD are not as consistent as its water quality benefits (Youssef et al., 2023). Controlled drainage raises the water table and increases saturated conditions in the soil profile during fallow periods, which promotes the denitrification of nitrate-N (Poole et al., 2018; Liu et al., 2019). It has also been documented that CD reduces drainage outflow and therefore decreases the total N loading to downstream water bodies (Evans et al., 1995; Gilliam et al., 1979; Skaggs et al., 2010; Helmers et al., 2022; Youssef et al., 2018). Surface ditches can also be managed in CD using control structures placed downstream of the ditches in the main outlet canal or in the outlets of most parallel drainage ditch systems (Poole et al., 2018).
Many growers in eastern North Carolina are increasingly installing subsurface tile as a direct replacement for the surface ditches. It is thought that the trafficability hazards and scouring of sediments can be resolved by filling in the surface ditches while retaining the yield benefits of a proper drainage system. In addition, more land will be available for cropping. However, there are multiple issues regarding the sole utilization of subsurface tile for drainage. Also, relying solely on subsurface tiling alone has yet to be proven effective in limiting surface ponding that occurs following high precipitation events on lands with limited topographic relief in North Carolina. The total drainage outflow of only a subsurface tile system compared to the conventional surface system depends on the drainage intensity of both the surface and subsurface designs. The effectiveness of subsurface tile drainage is dependent on the ability of water to move through the soil profile (Robinson and Rycroft, 1999). Drainage systems should be designed to provide sufficient drainage that eliminates excess moisture stress on crop roots during high precipitation periods while not excessively draining the soil profile so that excess nutrient loss and potential drought stress would occur (Ghane et al., 2021; Skaggs, 2007; Skaggs and Chescheir, 2003). If a system is overdesigned with a higher drainage intensity, the installation price will increase, as will the potential for crop stress due to dry conditions. Similarly, if CD is not implemented, then the soil profile can become drier at a quicker rate, which exacerbates drought stress on crops. Typical subsurface drainage design coefficients range between 6 mm to 20 mm per day in the southeast United States (Skaggs et al., 2006). This drainage coefficient range is not high enough to handle large storms commonly occurring during the summer months in eastern NC without additional surface drainage. Eastern NC has limited natural topographic relief; as a result, the incorporation of precise surface grading is essential for proper surface water drainage.
A design that utilizes much smaller surface ditches in combination with subsurface drainage tile provides a potential solution to some common issues with traditional open ditch drainage practices. The proposed system would include a subsurface tile system designed to outlet at a single location chosen based on the field contours. This would allow the operator to manage the entire field with one CD structure. In addition to the subsurface tile, shallow surface ditches would be installed solely for surface drainage during large storm events when rainfall intensity exceeds the infiltration rate and natural topography is not enough to remove ponding water. This will reduce field trafficability hazards as well as decrease bank scouring. Additional land can be cultivated because the profile size of the surface ditches will be greatly reduced. It is critical that these shallow systems be installed on precision grade (fig. 3) for adequate surface drainage.
 |
| Figure 3. Installation of shallow surface furrow with precision grade drainage plow. |
Placing these surface ditches on a specified grade with Real Time Kinematic (RTK) equipment will promote consistent outflow. In addition to the surface ditches, RTK equipment can also be utilized to create a small crown on the soil surface with land-leveling between the shallow ditches to direct surface flow to the ditches. The smaller ditch dimensions and precision installation should promote bank stabilization and lower yearly maintenance (fig. 4). This work focuses on the yield benefits of this new design. The objectives of this paper are to (1) quantify differences in corn and soybean grain yield between SD and FD over nine growing seasons with varying weather conditions in Eastern NC and (2) investigate the effects of SD and FD on the water table to determine if changes in the water table can explain observed differences in grain yield.
 |
| Figure 4. Shallow surface furrow after fourteen years of installation. |
Materials and Methods
Site Description
The research site is located in Beaufort County near the town of Bath, North Carolina, USA (35°28'1.90" N, 76°44'1.38" W; fig. 5). The site is nearly flat (0%-2% surface slope) and has a Portsmouth Sandy Loam soil (Typic Umbraquult; fine-loamy, siliceous, thermic), which is classified as poorly drained. The site was originally set up in 2008 and then upgraded in 2014 to assess the crop yield and water quality benefits of water table management practices including controlled drainage and subirrigation (Poole et al., 2013, 2018). Two treatments (namely conventional drainage and shallow drainage) are included in this study. This study was conducted over five growing seasons of corn (2014, 2016, 2018, 2020, and 2022) and four growing seasons of soybean (2015, 2017, 2019, and 2021).
Conventional Drainage
The conventional drainage treatment, also referred to as free drainage (FD), was implemented on a 3.3 ha field area that is drained by a combination of parallel open ditches (three) and subsurface drain tiles (six). The open ditches are 1.07 m deep, 1.5 m wide, 183 m long, and 60 m apart. Two subsurface drain tiles were installed between the open ditches at a 20 m spacing and 1.1 m depth (fig. 5). The subsurface drain tiles discharge into the open ditches near the ditch outlet. Thus, each open ditch receives additional discharge from two adjacent subsurface tiles. The combined drainage system at the site was designed to meet drainage needs while facilitating smooth field operations with modern equipment (Poole et al., 2013). In the FD treatment, drain outlets remained open year-round, allowing the drainage system to drain the field according to the design drainage rate. The FD treatment was surface leveled without the use of laser-grade equipment. Leveling included a land plane and box blade to smooth the surface. The treatment was also crowned to facilitate surface water movement towards the ditches. The resulting crown is approximately 7.6 to 15 cm (Poole et al., 2013). Each of the three open ditches, along with the two associated subsurface tiles, was considered a replicate of the FD treatment. A fourth open ditch located on the eastern side of the treatment and a subsurface tile on the western side of the treatment (neither are shown in the layout) with similar management to the FD treatment allow the treatment to be hydraulically isolated from other treatments.
 |
| Figure 5. Satellite image of the drainage water management showing site layout, treatment boundaries, open-ditch and subsurface tile locations, water table monitoring wells, and flow direction. |
Shallow Drainage
The SD treatment was implemented on a 9.1 ha field area. Like FD, the SD treatment is drained by a combination of open ditches and subsurface tile drains. However, the surface ditches of the SD treatment are much shallower (30 cm) than the ditches of the FD treatment. The main function of the shallow ditches was to timely remove water from the field via overland flow and minimize water ponding following large storm events. The SD treatment had five shallow ditches 1.2 m wide, 382 m long, and 60 m apart (fig. 5). The SD treatment utilizes subsurface tiles to provide subsurface drainage of infiltrated water into the soil profile. The subsurface drainage system consists of lateral drains (10 cm diameter) installed at an average depth of 1.1 m and a spacing of 20 m. The subsurface laterals were installed perpendicular to the shallow ditches on 0.1% grade (fig. 5). The laterals discharge into a 20.3 cm diameter main that was installed directly below the westernmost shallow ditch. All SD surface ditches are the same except for the westernmost ditch, which receives the outflow from the subsurface drainage main line. An outlet riser was installed at the exit point of the westernmost ditch (SDout) to enable water table management (e.g., controlled drainage) at the SD treatment. The riser is equipped with two gates that can be manually adjusted to switch between FD and CD with different management depths as shown in figure 6. The CD gate height in the SD treatment was set to 45 cm for all growing seasons until 2022, when it was raised to 30 cm. The only time that SD treatment was managed in FD mode was during a very wet period in the June 2016 growing season. This was a management decision to lower the water table and reduce ponded conditions.
Water Table Monitoring
Two 2.10 m deep water table monitoring wells, equipped with Onset U20-001-04 (ONSET COMPUTER CORPORATION, BOURNE, MA) pressure transducers, were installed in each of the FD and SD treatments. Occasionally, the water table depth exceeded the installation depth of the monitoring wells. These periods were excluded from the statistical analysis. The SD treatment had periods of missing water table data in 2016 and 2022 due to malfunctioning water level loggers. A correlation equation, developed from the FD and SD data in other periods with similar water table depths, was used to estimate the missing water table depth data for the SD treatment. Water table data for 2014 and 2015 were not available, so these years were excluded from water table analysis.
Weather Data
Precipitation was monitored on-site at two locations by HOBO Rain Gauge Tipping Bucket Data Loggers (Onset Computer Corporation, Bourne, MA) as well as two manual precipitation gauges. Weather data were obtained from two local weather stations operated by the National Weather Service and provided through the NC Climactic Office (Aurora 6 N; 35.387°N, 76.776°W and Washington WWTP 4w; 35.555°N, 77.072°W) when onsite data were not available. The weather stations were located 16.6 km and 37.5 km from the research site. The 30-year normal precipitation was calculated using daily precipitation data obtained from the same weather stations.
 |
| Figure 6. (a) Shallow surface outlet ditch (SDout) with buried subsurface drainage main outlet structure. (b) Riser with manually adjustable gates to switch the shallow treatment between free drainage and controlled drainage modes. |
Daily grass reference evapotranspiration (ETo) was estimated by the FAO-56 Penman-Monteith equation (Allan et al., 1998) using data obtained from the local weather station (Pamlico Aquaculture Field Lab; 35.362°N, 76.716°W). The daily ETo values were multiplied by crop coefficients (Kc, table 1) at various growth stages obtained from the NC Irrigation Guide to estimate the crop-specific evapotranspiration estimates (ETc; NRCS, 2010).
| Table 1. Corn and soybean crop coefficients (Kc) during stage of growth determined by percentage of growing season. | ||||||||||
| % Growing Season Kc Factors | ||||||||||
| Crop | 10% | 20% | 30% | 40% | 50% | 60% | 70% | 80% | 90% | 100% |
| Corn | 0.25 | 0.35 | 0.69 | 1.03 | 1.20 | 1.20 | 1.20 | 1.15 | 0.87 | 0.60 |
| Soybean | 0.25 | 0.42 | 0.76 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 0.74 | 0.45 |
Crop Management
Data for this experiment were collected over nine growing seasons from 2014 to 2022. Corn (Zea mays L.) and soybean (Glycine max L.) were rotated each year such that corn was planted in even years and soybean planted in odd years. A total of five corn growing seasons and four soybean growing seasons were monitored. Corn was planted in May and harvested in September in all years except for 2020, when harvest occurred in October. The corn hybrids planted at the research site were (113–118) day relative maturity hybrids. Soybeans were planted in June, except in 2015, when they were planted in early July and harvested in November of each year. Soybean planted in the four growing seasons of this study were determinate varieties. Hybrids, varieties, fertilizer, and tillage practices varied, but all factors remained constant across treatments within a given year (table 2). All practices used in this experiment are typical for crop production in eastern North Carolina.
Corn Growth Stages
Corn growth stages were estimated using the growing degree days (GDD), which were calculated using equation 1.
GDD = ((Tmin + Tmax)/2) –Tbase (1)
where
GDD = growing degree days,
Tmin= minimum daily temperature (°C)
Tmax=maximum daily temperature (°C)
Tbase =base temperature (°C).
The allowed minimum daily temperature and base value were set to 10°C. If the daily minimum temperature fell below 10°C, then it would be replaced with 10°C. Likewise, the maximum allowed temperature was set to 30°. If the daily temperature exceeded this limit, it was replaced with 30°C.
| Table 2. Crop data with planting and harvest dates. | ||||
| Year | Crop | Variety | Plant Date | Harvest Date |
| 2014 | Corn | Pioneer 1615 | 7-May | 18-Sep |
| 2015 | Soybean | Pioneer P53T18X | 2-Jul | 20-Nov |
| 2016 | Corn | Pioneer 1197 | 8-May | 15-Sep |
| 2017 | Soybean | Southern States 5615 NR2 | 20-Jun | 27-Nov |
| 2018 | Corn | Pioneer 1870 | 5-May | 22-Sep |
| 2019 | Soybean | Syngenta S56-B7X | 5-Jun | 28-Nov |
| 2020 | Corn | Pioneer 1870 | 11-May | 3-Oct |
| 2021 | Soybean | MorSoy 6808 RXT | 19-Jun | 11-Nov |
| 2022 | Corn | DEKALB DKC 68-69 | 5-May | 18-Sep |
The relationship between the accumulated GDD and the different stages of corn growth was obtained from the North Carolina State Climate Office (NCSCO, 2022) (table 3). Growth stages were grouped into three distinct stages: rapid growth (V10-V16), silking (VT-R1), and grain fill (R2-R6).
| Table 3. Estimated growth stages of corn from calculated growing degree day (GDD). | ||
| Stage | Description | GDD |
| V10 - V16 | Rapid Growth | 740 - 1,139 |
| VT - R1 | Silking | 1,400 – 1,659 |
| R2 - R6 | Grain Fill | 1,660 – 2,450 |
Soybean Growth Stages
Soybean growth stages were determined using SoyStage software (https://soystage.uark.edu/), which predicts the growth stages based on the emergence date of the soybean, latitude, and average daily temperatures (Santos et al., 2019). The study site is outside of the software’s calculator zone, which introduces uncertainty in predicted dates of different growth stages (Morris et al., 2021). We used a model prediction of soybean growth stages for a location (35.46, -87.508) that had the same latitude as the research site near Bath, NC. Soybean planted in 2015, 2017, and 2019 was maturity Group V, while soybean Group VI was planted in 2021. Maturity group V soybean develops faster than Group VI; however, the planting date and growing conditions can impact the length of growth stages and time to maturity. Soybean growth stages were grouped into vegetative growth (Planting-V15), flowering (R1-R2), pod development (R3-R4), and seed fill (R5-R7). Emergence date was selected by using a predetermined date in the SoyStage software that was closest to seven days after the planting (DAP) date.
Yield Measurement
Yield maps generated by the harvester equipped with the Global Positioning System (GPS) were clipped in ArcMap such that yield comparison between the two treatments could be done over equal areas of .81 ha. The northern and southern ends of the FD treatment and the southern end of the SD treatment were clipped by 24.4 m to remove the effects of turns, overlaps, and field end rows. This was also done to minimize the effects of varying soil conditions on the north end of the SD treatment. The moisture content of harvested grain was adjusted back to a standardized 15.5% for corn and 13% for soybeans for yield analysis.
Statistical Analysis
Crop yields of the two treatments were compared by performing a Two-Way analysis of variance (ANOVA) with the proc GLIMMIX procedure in SAS software (SAS, 2021). For corn, the analysis was performed twice, once including and once excluding the corn yield in 2016. In the statistical analysis, crop yield was the dependent variable, and both year and treatment were the independent variables. Growing season water table depths were statistically analyzed using the paired t-Test in SAS software (SAS, 2021). The 10% significance level was used to evaluate statistical differences.
Results and Discussion
Corn Yield
The average corn yields for the FD and SD treatments over the five growing seasons were 11.2 Mg/ha and 11.6 Mg/ha, respectively. Compared to FD, SD significantly increased corn yield (p = 0.0004) by 0.4 Mg/ha, or 4%. The effects of year and the interaction between treatment and year factors were both statistically significant (p < 0.0001). This suggests that the effect of the SD treatment on crop yield may fluctuate from year to year, depending on weather conditions. The largest increase in corn yield in SD occurred in 2018, when both treatments had the lowest yield among the five seasons (table 4). The smallest difference in corn yields between the two treatments occurred in 2022, when corn yield in the two treatments was highest among the five seasons.
In 2016, corn in the SD treatment experienced stress due to water ponding caused by a series of large precipitation events, totaling 236 mm, which occurred from 29 May to 8 June. The gate assembly at the ditch outlet of the SD treatment was not timely lowered to quickly remove the excess water and lower the water table following the large precipitation events. As a result, the SD treatment had a lower corn yield in 2016 than the FD treatment by 0.7 Mg/ha. This yield loss could have been reduced or completely avoided if the management of the system had been more responsive to precipitation in 2016 (Youssef et al., 2023). The corn yield comparison was repeated, excluding the 2016 yield data, to demonstrate the potential effect of more active management of the outlet of the SD treatment on crop yield response. Excluding 2016 yield data, the SD treatment significantly increased (p < .0001) corn yield by 0.7 Mg/ha, or 6.6% on average. The year effect proved to be statistically significant (p < 0.0001); however, the treatment-year interaction was not statistically significant (p = 0.24). This suggests that the SD effect on corn yield would have remained consistent over different years if the management delay in 2016 had been avoided.
Soybean Yield
Soybean data were collected over four growing seasons from 2015 to 2021. The SD treatment had significantly higher soybean yield (p < .0001) than the FD treatment, with an average increase of 0.5 Mg/ha, or 13.7%. The effects of year and treatment-year interaction were both statistically significant (p < 0.0035). This suggests that the treatment effect on soybean yield may vary from year to year. During the four growing seasons, the SD never produced a lower soybean yield than the FD treatment. The soybean yield difference between the two treatments was the smallest in 2017 and the greatest in 2021 (table 4).
| Table 4. Measured corn and soybean yield for the free drainage (FD) and shallow drainage (SD) treatments at the experimental site in Bath, North Carolina, during 2014-2022. | |||||
| SD[a] | FD[a] | Yield Increase | |||
| Year | (Mg/ha) | (Mg/ha) | Mg/ha | % | |
| Corn | 2014 | 12.93 AB | 11.97 C | 0.96 | 8.0 |
| 2016[b] | 10.14 E | 10.86 D | -0.72 | -6.6 | |
| 2018 | 10.02 E | 9.25 F | 0.77 | 8.3 | |
| 2020 | 12.02 C | 11.20 D | 0.82 | 7.3 | |
| 2022 | 13.09 A | 12.73 B | 0.36 | 2.9 | |
| AVG | 11.64 A | 11.20 B | 0.44 | 4.0 | |
| Soybean | 2015 | 2.57 F | 2.21 G | 0.35 | 16.0 |
| 2017 | 3.77 BC | 3.61 D | 0.16 | 4.4 | |
| 2019 | 3.66 CD | 3.16 E | 0.50 | 15.8 | |
| 2021 | 4.71 A | 3.90 B | 0.81 | 20.8 | |
| AVG | 3.68 A | 3.22 B | 0.46 | 14.3 | |
[a] Letters A, B, C, D, and E indicate statistically significant differences among crop yields within years or between the average treatment yields for a specific crop. [b] Ponded surface water from delayed SDout weir height adjustment caused excessive wet stress (29 May to 8 June). | |||||
Precipitation
Precipitation amounts over the nine-year data set averaged 1353 mm, with a notable year-to-year variation in precipitation (SD = 246 mm; table 5). For corn, the amount of precipitation during the growing season showed considerable year-to-year variation, with a standard deviation of 154 mm. Specifically, the 2018 growing season was exceptionally wet, recording a total seasonal precipitation of 863 mm, substantially surpassing the 30-year average of 548 mm. Conversely, conditions were drier in 2014, with a total seasonal precipitation of 449 mm, well below the long-term average.
Precipitation during soybean growing seasons ranged from 457 mm in 2021 to 643 mm in 2015, with a less year-to-year variation (SD = 82 mm) compared to corn. Throughout the growing seasons of both crops (May to October), June consistently stood out as the wettest month, while October was consistently the driest. Notably, September exhibited the highest month-to-month variation in precipitation (SD = 71 mm), whereas July had the lowest variability in precipitation (SD = 38 mm).
| Table 5. Monthly precipitation measured at the site in Bath, NC, with 30-year normal (1991-2021) and seasonal amounts for main crop growing periods. | |||||||||||
| Precipitation (mm) | |||||||||||
| 2014[a] | 2015 | 2016 | 2017 | 2018 | 2019 | 2020[b] | 2021[c] | 2022 | AVG±SD[d] | 30-Year Normal | |
| Jan | 70 | 125 | 85 | 118 | 125 | 89 | 41 | 133 | 83 | 97±31 | 105 |
| Feb | 43 | 103 | 156 | 50 | 24 | 89 | 145 | 208 | 29 | 94±64 | 102 |
| Mar | 116 | 98 | 84 | 113 | 133 | 50 | 69 | 97 | 136 | 100±29 | 98 |
| Apr | 131 | 119 | 59 | 164 | 157 | 111 | 74 | 28 | 63 | 101±47 | 110 |
| May | 34 | 101 | 161 | 151 | 139 | 48 | 172 | 62 | 190 | 118±58 | 123 |
| Jun | 82 | 213 | 204 | 94 | 254 | 73 | 124 | 209 | 149 | 156±66 | 114 |
| Jul | 134 | 120 | 123 | 164 | 219 | 79 | 120 | 136 | 134 | 137±38 | 106 |
| Aug | 99 | 77 | 132 | 179 | 97 | 118 | 148 | 141 | 69 | 118±36 | 106 |
| Sep | 144 | 151 | 300 | 116 | 155 | 124 | 112 | 44 | 209 | 151±71 | 98 |
| Oct | 18 | 154 | 230 | 78 | 51 | 78 | 62 | 30 | 8 | 79±71 | 75 |
| Nov | 89 | 194 | 14 | 49 | 152 | 93 | 161 | 19 | 69 | 93±64 | 87 |
| Dec | 86 | 125 | 79 | 122 | 173 | 63 | 155 | 75 | 133 | 112±38 | 105 |
| Seasonal[e] | 449 | - | 747 | - | 863 | - | 638 | - | 619 | 663±154 | 548 |
| Seasonal[f] | - | 643 | - | 613 | - | 554 | - | 457 | - | 567±82 | 587 |
| Total | 1045 | 1580 | 1628 | 1396 | 1679 | 1013 | 1383 | 1182 | 1272 | 1353±246 | 1230 |
[a] Values from local weather stations. [b] Values from local weather stations (DOY 260-366). [c] Values from local weather stations (DOY 1-105). [d] Avg±SD= Monthly average of precipitation ± Standard deviation. [e] Plant to harvest dates (Corn). [f] Plant to harvest dates (Soybean). | |||||||||||
Water Table Response to Shallow Drainage and Free Drainage Treatments
A shallower water table facilitates a greater contribution of groundwater to satisfy ET demand through upward flux. For all years, the average water table in the SD treatment had on overage a shallower water table depth during the growing season than the FD treatment (table 6). When WTD during all growing seasons were grouped together and analyzed with a paired t-test for means, the SD water table was significantly higher (13 cm closer to the surface) than the FD treatment (p = 0.01). When WTD data were grouped by crop and analyzed with a paired t-test for means, the SD had a significantly higher average growing season water table than the FD for corn (p = 0.02), but not for soybeans (p = 0.2). Over the four corn growing seasons with available WTD data, the SD averaged a water table that was 10.8 cm shallower than the FD treatment (table 6). On average, over the three soybean growing seasons with available WTD data, the SD treatment had 15.2 cm shallower water table than the FD treatment (table 6).
The SD treatment increased water ponding on the soil surface in a few instances. The water table in the SD treatment rose above the soil surface for two hours on July 4 in 2017, three hours on 22 June in 2021, and ten hours on 4 August in 2021. The FD treatment had periods where the water table rose to the top 1 cm of the soil but never went above the soil surface for an extended time. The short periods of ponding conditions that occurred with the SD treatment may have influenced crop yield for that treatment, but the only prolonged period of stress occurred in 2016. Management decisions to lower the gate height during high or constant precipitation events could have a positive effect on the system’s ability to avoid excess moisture stress if implemented in time.
| Table 6. Average water table depth of FD and SD treatments during the growing season from 2016 to 2022. | ||||
| Water Table Depth (cm) | ||||
| Year | FD | SD | Difference | |
| Corn | 2016 | 114.3 | 104.4 | 9.9[a] |
| 2018 | 81.1 | 72.4 | 8.7[a] | |
| 2020 | 97.7 | 79.6 | 18.1[a] | |
| 2022 | 133.2 | 126.8 | 6.3[a] | |
| AVG | 106.6 | 95.8 | 10.8[a] | |
| Soybean | 2017 | 107.8 | 105.8 | 2.1[a] |
| 2019[b] | 159.2 | 128.9 | 30.3[a] | |
| 2021[b] | 129.0 | 115.7 | 13.3[a] | |
| AVG | 132.0 | 116.8 | 15.2 | |
[a]????Significant difference. [b]????WTD fell deeper than the installation depth of the monitoring sensors for a period(s) of the growing season. | ||||
Corn Yield as Affected by Weather Conditions and the Shallow Drainage System
Precipitation exceeded ETc in all corn growing seasons, except for 2020 (table 7, fig. 7). However, the timing of precipitation does not always align with crop water demands, as exemplified by the 2018 growing season, where the total precipitation significantly exceeded ETc, yet the corn yield was the lowest among all corn seasons. In this study, growing seasons characterized by extended periods of little to no precipitation resulted in the lowest crop yields. This was observed in 2016 and 2018, when corn experienced extended dry spells during the critical silking (pollination) stage, resulting in the lowest yields during the study for both treatments (fig. 7). In 2018, there was a 23-day period between 27 June and 19 July, with insufficient precipitation to meet the ETc demand. During this period, only 12 mm of precipitation occurred in two events (5 July and 16 July), while the ETc demand totaled 137 mm, creating a deficit of 125 mm. This 23-day deficit coincided with the tasseling and silking stages when corn had the highest water demand, and adequate soil moisture was crucial for high corn yields (Evans and Fausey, 1999; Sudar et al., 1979).
| Table 7. Estimated crop-specific evapotranspiration and collected rainfall for corn and soybean growing seasons. | ||||
| Estimated Growing Season ETc (mm) | ||||
| Year | Precipitation | ETc | Difference | |
| Corn | 2014 | 449 | 448 | 1 |
| 2016 | 747 | 538 | 209 | |
| 2018 | 863 | 541 | 322 | |
| 2020 | 638 | 670 | -32 | |
| 2022 | 619 | 575 | 44 | |
| AVG | 663 | 560 | 103 | |
| Soybean | 2015 | 643 | 393 | 250 |
| 2017 | 613 | 441 | 173 | |
| 2019 | 554 | 657 | -102 | |
| 2021 | 457 | 448 | 9 | |
| AVG | 567 | 515 | 52 | |
In contrast, the growing seasons with the highest corn yield were those in which precipitation was well-distributed throughout the growing season, particularly during critical growth stages of pollination (e.g., 2014, 2020, and 2022). The corn yield in 2022 averaged the highest statistically. Compared to the same 23-day period in 2018, corn in 2022 had an estimated ETc of 100 mm and received 114 mm of precipitation (fig. 7).
Poole et al. (2013) attributed the positive impact of controlled drainage on yield to the increased water storage in the soil profile because it restricted the loss of drainage water during critical growth periods. Controlled drainage has been shown to reduce drain flow by up to 52% during dry years and up to 28% in wet years in other studies (Helmers et al., 2022). Conserved drainage water can be utilized by the crop to meet ET demands, and therefore alleviating crop stress due to water deficit conditions. Insufficient soil water in the plant root zone can reduce corn yields by 25% if occurred prior to silking, by 50% if they occurred during silking, and by 21% if they occurred after silking (Denmead and Shaw, 1960).
The observed yield increases by the SD treatment in this study are likely attributed to the management of drainage water, which raised the water table and increased soil water availability in the root zone. The shallow surface ditches and controlled subsurface tiles were able to remove excess water in the top 30 to 45 cm of the soil profile in a timely manner (apart from a single period in 2016). The CD allowed for the management of drainage outflow, which slowed the water table drawdown and increased soil water availability for transpiration during critical periods. For example, the SD treatment retained more drainage water during critical growth periods for corn in 2018, when multiple storm events in late June through July raised the water table (fig. 8). These storm events did not generate drainage outflow from the SD treatment as the shallow system was managed in CD, and subsurface outflow could not occur unless the water table depth rose above 45 cm from the soil surface. Thus, the potential pathways for the received precipitation were soil storage, deep or lateral seepage, and plant uptake. Since both treatments have the same major soil types and subsurface drainage intensity, seepage differences between the two treatments are likely small. Therefore, the primary sink for this water was ET. Utilizing CD enabled the SD system to retain drainage water for corn water uptake and subsequently increase corn yield by 8.3% and 7.3% in 2018 and 2020, respectively. It should be mentioned that the yield benefits of this treatment or similar practices (e.g., controlled drainage) on reduction of crop stress due to dry conditions that occur over an extended period are limited because additional irrigation water would be needed under these conditions (Youssef et al., 2023).
The only year in this study where the SD treatment yielded less than the FD occurred in 2016 (fig. 10). In 2016, the research site received 239 mm of precipitation in 10 days between 29 May and 7 June. Prior to this precipitation, the SD WTD was 33 and the FD WTD was 58 cm. The precipitation raised the water table in the FD treatment to a depth of 4.9 cm below the ground surface, with an average depth of 10.1 cm during the 10-day period. The SD treatment’s water
 |
| Figure 7. Graph illustrating the comparison between weekly precipitation and estimated crop evapotranspiration (ETc) demand across the five corn growing seasons, with key growth stages integrated into the plot. |
 |
| Figure 8. The WTD for FD and SD in 2018 with estimated crop growth stages. |
table rose above the ground surface and ponding conditions existed from 30 May to 4 June. On 4 June, the control gate was opened to drain the field, and as a result, the water table quickly fell below the ground surface. During this period, the SD WTD averaged 1.2 cm below ground level, creating fully saturated conditions in the root zone. Corn was between the V4 and V5 growth stages, which is a critical growth period for root development and ear formation. Yield potential was likely impacted as the number of rows and kernels per row started to form during the V5 stage (Strachan, 2004). It has been reported that exposing roots to saturated conditions for more than 3 days during this critical growth stage can significantly increase root mortality (McDaniel et al., 2016). The FD treatment’s water table never fully rose to the surface, leaving some roots not impacted by the anaerobic conditions caused by the high water table. The negative yield impact of the SD in 2016 means that the impact of the wet stress during V4-V5 outweighed the potential positive impact of increasing soil water storage during silking and grain fill (4 July-10 Aug) when ETC was greater than precipitation for an extended time period (fig. 7). Conversely, in 2022, the SD treatment produced higher final corn yields due to water conserved with CD during pollination, followed by a period when ETC was greater than precipitation during grain fill (14 July-10 Aug) (fig. 7). This demonstrates the importance of managing controlled drainage to minimize both wet and dry stress at critical periods throughout the entire growing season.
 |
| Figure 9. The WTD for FD and SD in 2019 with estimated crop growth stages. |
 |
| Figure 10. The WTD for FD and SD in 2016 with estimated crop growth stages and ponded conditions. |
Soybean Yield as Affected by Weather Conditions and the Shallow Drainage System
Precipitation amounts during all growing seasons of soybean were greater than ETc, except in 2019 (table 7, figs. 9 and 11). Soybean is most susceptible to yield reductions when ETc demand exceeds available water during pod formation and pod filling, stages R3-R6 (Doss et al., 1974; Sionit and Kramer, 1977). Similar to corn, precipitation deficits during critical soybean growth stages were associated with lower yields in certain years, like 2015 and 2019. Conversely, years with well distributed precipitation resulted in higher yields, as seen in 2017 and 2021. The percent difference in soybean yield was the lowest between treatments in 2017 (table 4). In 2017, the site received several timely and intense precipitation events during the critical growth periods that plants in both treatments utilized to satisfy estimated ETc demand (fig. 11). The end of pod filling in 2017 was a relatively dry period when soil water conservation by SD may have contributed to soybean yield increase, but yield differences between treatments were limited due to earlier precipitation during pod development. In contrast, 2021 experienced a precipitation deficit during R3-R6 (not shown in figure). In 2021, soybean in the SD treatment was likely able to utilize more soil water during these stages due to the raised water table and restricted drainage outflow.
 |
| Figure 11. Comparison of weekly precipitation and estimated ETc demand during the four growing seasons of soybean. |
Effect of Precision Grading on Yield
In addition to the factors of rainfall, evapotranspiration, and water table depth that influence crop yield, it is crucial to consider the effects of agricultural practices such as precision grading. Surface crowning and shallow surface ditches were installed with precision grade equipment, which minimized the movement of topsoil to the field crown. The precision grading smooths surface irregularities and subsequently distributes surface water more uniformly. The lines separating replications indicate the center of the surface crown between ditches. The effect of conventional ditches on crop yield can be seen in the spatial yield maps (fig. 12), with areas of lower grain yield in the center of each replicate of the FD treatment. This area is directly adjacent to the open ditches. The continuous depletion of the water table near the ditch banks, loss of planted area, and reduction of topsoil depth near the traditional ditches in the FD treatment likely contribute to lower grain production in these areas. A small effect on the crop yield of the shallow surface ditches can be seen with the 2022 yield map; however, the negative impact on the spatial crop yield is much less than that of the FD
ditches. The spatial yield impact of the SD ditches appeared to be very small in the 2021 soybean crop and almost non-existent compared to the FD ditch impact.
Summary and Conclusions
Shallow surface ditches with controlled subsurface tile (SD) were experimentally evaluated to investigate the effects of this drainage system on corn and soybean yields in eastern North Carolina. Compared to conventional open ditch drainage (referred to as free drainage, FD), the SD system significantly increased crop yields in eight of nine growing seasons under a corn and soybean rotation. The increase in crop yield with SD varied with weather conditions during the crop growing seasons. The greatest benefit of SD and associated controlled drainage (CD) occurred when there was stored soil water near the crop root zone and the site received minimal precipitation during critical growth periods.
Installation of shallow surface ditches on a controlled grade and with precision surface crowning between ditches proved to be an effective way to remove excessive surface water from the field site. With the smaller profile of the surface ditches and the subsurface drainage system being maintained in intense CD, the SD system was able to increase crop yield by limiting both wet and dry stresses associated with unmanaged drainage ditches. The lack of proper riser board management causing wet stress in conventional ditch CD systems is a common issue that can be significantly reduced with the addition of shallow surface ditches. The shallow ditches effectively disconnect the surface drainage system from the subsurface drainage system. In doing so, it alleviates excessively wet conditions due to a lack of management during high intensity rainfall events while simultaneously conserving subsurface drainage water. This design shows promise in improving crop yield and reducing riser board management while retaining the water quantity benefits of the CD practice in subsurface drainage systems.
 |
| Figure 12. Spatial yield map of 2021 soybean and 2022 corn crops with replications to show spatial impact of surface ditches on soybean yield. |
Further research is needed to determine the crop yield benefits that could be potentially achieved with active management of the subsurface CD outlet during extremely wet periods. Active management could minimize periods of wet stress on the crop roots and contribute to higher yields. There is also a need to investigate the effects of SD with CD on drainage and nutrient loading to determine if the system can minimize the impacts of artificial drainage on downstream water quality.
Acknowledgments
We would like to express our gratitude to the USDA-NIFA Postdoctoral Fellowship for their support through the project entitled 'Development and Promotion of Drainage Water Management Systems Adapted to Maximize Crop Production in Response to Climate Variability' (2016-2018). Additionally, we appreciate the funding provided by the NC Corn Growers Association for the period 2015-2017. BIBLIOGRAPHY \l 1033
References
Allan, Richard & Pereira, L. & Smith, Martin. (1998). Crop evapotranspiration-Guidelines for computing crop water requirements-FAO Irrigation and drainage paper 56.
Carter, C. E. (1999). Surface drainage. In R. Skaggs, & J. v. Schilfgaarde (Eds.), Agricultural drainage (Vol. 38, pp. 1023-1048). Madison, WI: SSSA. https://doi.org/10.2134/agronmonogr38.c32
Denmead, O. T., & Shaw, R. H. (1960). The effects of soil moisture stress at different stages of growth on the development and yield of corn. Agron. J., 52(5), 272-274. https://doi.org/10.2134/agronj1960.00021962005200050010x
dos Santos, C., Salmerón, M., & Purcell, L. C. (2019). Soybean phenology prediction tool for the US midsouth. Agric. Environ. Lett., 4(1), 190036. https://doi.org/10.2134/ael2019.09.0036
Doss, B. D., Pearson, R. W., & Rogers, H. T. (1974). Effect of soil water stress at various growth stages on soybean yield. Agron. J., 66(2), 297-299. https://doi.org/10.2134/agronj1974.00021962006600020032x
Evans, R. O., & Fausey, N. R. (1999). Effects of inadequate drainage on crop growth and yield. In R. W. Skaggs, & J. Schilfgaarde (Eds.), Agricultural Drainage (Vol. 38, pp. 13-54). Madison, WI: SSSA. https://doi.org/10.2134/agronmonogr38.c2
Evans, R. O., Skaggs, R. W., & Gilliam, J. W. (1995). Controlled versus conventional drainage effects on water quality. J. Irrig. Drain. Eng., 121(4), 271-276. https://doi.org/10.1061/(ASCE)0733-9437(1995)121:4(271)
Fangmeier, D. D., Elliot, W. J., Workman, S. R., Huffman, R. L., & Schwab, G. O. (2006). Soil and water conservation engineering (5th ed.). Clifton Park, NY: Thomson Delmar Learning.
Fausey, N. R. (2005). Drainage management for humid regions. Int. Agric. Eng. J., 14(4), 209-214.
Ghane, E. (2022). Choosing between sock-wrapped and sand-slot pipes. Extension Bulletin E3467. Michigan State University Extension. Retrieved from https://www.canr.msu.edu/field_crops/uploads/files/E3467_Knitted-Sock_Bulletin_1.pdf
Ghane, E., Askar, M. H., & Skaggs, R. W. (2021). Design drainage rates to optimize crop production for subsurface-drained fields. Agric. Water Manag., 257, 107045. https://doi.org/10.1016/j.agwat.2021.107045
Ghane, E., Fausey, N. R., Shedekar, V. S., Piepho, H. P., Shang, Y., & Brown, L. C. (2012). Crop yield evaluation under controlled drainage in Ohio, United States. J. Soil Water Conserv., 67(6), 465-473. https://doi.org/10.2489/jswc.67.6.465
Gilliam, J. W., Skaggs, R. W., & Weed, S. B. (1979). Drainage control to diminish nitrate loss from agricultural fields. J. Environ. Qual., 8(1), 137-142. https://doi.org/10.2134/jeq1979.00472425000800010030x
Helmers, M. J., Abendroth, L., Reinhart, B., Chighladze, G., Pease, L., Bowling, L.,... Strock, J. (2022). Impact of controlled drainage on subsurface drain flow and nitrate load: A synthesis of studies across the U.S. Midwest and Southeast. Agric. Water Manag., 259, 107265. https://doi.org/10.1016/j.agwat.2021.107265
Helmers, M., Christianson, R., Brenneman, G., Lockett, D., & Pederson, C. (2012). Water table, drainage, and yield response to drainage water management in southeast Iowa. J. Soil Water Conserv., 67(6), 495-501. https://doi.org/10.2489/jswc.67.6.495
Jaynes, D. B., Hatfield, J. L., & Meek, D. W. (1999). Water quality in walnut creek watershed: Herbicides and nitrate in surface waters. J. Environ. Qual., 28(1), 45-59. https://doi.org/10.2134/jeq1999.00472425002800010005x
Liu, Y., Youssef, M. A., Chescheir, G. M., Appelboom, T. W., Poole, C. A., Arellano, C., & Skaggs, R. W. (2019). Effect of controlled drainage on nitrogen fate and transport for a subsurface drained grass field receiving liquid swine lagoon effluent. Agric. Water Manag., 217, 440-451. https://doi.org/10.1016/j.agwat.2019.02.018
McDaniel, V., Wayne Skaggs, R., & Negm, L. M. (2016). Injury and recovery of maize roots affected by flooding. Appl. Eng. Agric., 32(5), 627-638. https://doi.org/10.13031/aea.32.11633
Morris, T. C., Vann, R. A., Heitman, J., Collins, G. D., & Heiniger, R. W. (2021). Maximizing soybean yield by understanding planting date, maturity group, and seeding rate interactions in North Carolina. Crop Sci., 61(6), 4365-4382. https://doi.org/10.1002/csc2.20603
Moursi, H., Youssef, M. A., Poole, C. A., Castro-Bolinaga, C. F., Chescheir, G. M., & Richardson, R. J. (2023). Drainage water recycling reduced nitrogen, phosphorus, and sediment losses from a drained agricultural field in eastern North Carolina, USA. Agricultural Water Management, 279, 108179.
NCSCO. (2022). Corn development table. North Carolina state climate office climate dashboard for corn growers. Retrieved from https://products.climate.ncsu.edu/ag/corn/about/
NRCS. (2010). North Carolina irrigation guide. 210-vi-NEH 652, IG Amend. NC1. In National Engineering Handbook 652. Retrieved from https://efotg.sc.egov.usda.gov/references/public/NC/NC_Irrigation_Guide_Apr_2010.pdf
Poole, C. A., Skaggs, R. W., Cheschier, G. M., Youssef, M. A., & Crozier, C. R. (2013). Effects of drainage water management on crop yields in North Carolina. J. Soil Water Conserv., 68(6), 429-437. https://doi.org/10.2489/jswc.68.6.429
Poole, C. A., Skaggs, R. W., Youssef, M. A., Chescheir, G. M., & Crozier, C. R. (2018). Effect of drainage water management on nitrate nitrogen loss to tile drains in North Carolina. Trans. ASABE, 61(1), 233-244. https://doi.org/10.13031/trans.12296
Rabalais, N.N., & Turner, R.E. (2019). Gulf of Mexico Hypoxia: Past, Present, and Future. Limnology and Oceanography Bulletin, 28.
Randall, G. W., & Goss, M. J. (2008). Chapter 6 - Nitrate losses to surface water through subsurface, tile drainage. In J. L. Hatfield, & R. F. Follett (Eds.), Nitrogen in the environment: Sources, problems, and management (2nd ed., pp. 145-175). San Diego: Academic Press. https://doi.org/10.1016/B978-0-12-374347-3.00006-8
Robinson, M., & Rycroft, D. W. (1999). The impact of drainage on streamflow. In R. W. Skaggs, & J. Schilfgaarde (Eds.), Agricultural drainage (Vol. 38, pp. 767-800). Madison, WI: SSSA. https://doi.org/10.2134/agronmonogr38.c23
SAS. (2021). SAS OnDemand for Academics. Ver. 3.81. SAS Institute.
Sionit, N., & Kramer, P. J. (1977). Effect of water stress during different stages of growth of soybean. Agron. J., 69(2), 274-278. https://doi.org/10.2134/agronj1977.00021962006900020018x
Skaggs, R. W. (2007). Criteria for calculating drain spacing and depth. Trans. ASABE, 50(5), 1657-1662. https://doi.org/10.13031/2013.23971
Skaggs, R. W., & Chescheir III, G. M. (2003). Effects of subsurface drain depth on nitrogen losses from drained lands. Trans. ASAE, 46(2), 237-244. https://doi.org/10.13031/2013.12974
Skaggs, R. W., Fausey, N. R., & Evans, R. O. (2012). Drainage water management. J. Soil Water Conserv., 67(6), 167A-172A. https://doi.org/10.2489/jswc.67.6.167A
Skaggs, R. W., Youssef, M. A., & Chescheir, G. M. (2006). Drainage design coefficients for eastern United States. Agric. Water Manag., 86(1), 40-49. https://doi.org/10.1016/j.agwat.2006.06.007
Skaggs, R. W., Youssef, M. A., Gilliam, J. W., & Evans, R. O. (2010). Effect of controlled drainage on water and nitrogen balances in drained lands. Trans. ASABE, 53(6), 1843-1850. https://doi.org/10.13031/2013.35810
Strachan, S. D. (2004). Corn grain yield in relation to stress during ear development. Pioneer, a DuPont Company. Retrieved from https://www.pioneer.com/us/agronomy/ear_development.html
Sudar, R. A., Saxton, K. E., & Spomer, R. G. (1979). A predictive model of water stress in corn and soybeans. Paper no. 79-2004. St. Joseph, MI: ASAE. https://doi.org/10.13031/2013.34206
USDA-NASS. (2022). 2021 State Agriculture Overview. Washington, DC: USDA-NASS. Retrieved from https://www.nass.usda.gov/Quick_Stats/Ag_Overview/stateOverview.php?state=NORTH%20CAROLINA
Wesström, I., & Messing, I. (2007). Effects of controlled drainage on N and P losses and N dynamics in a loamy sand with spring crops. Agric. Water Manag., 87(3), 229-240. https://doi.org/10.1016/j.agwat.2006.07.005
Youssef, M. A., Abdelbaki, A. M., Negm, L. M., Skaggs, R. W., Thorp, K. R., & Jaynes, D. B. (2018). DRAINMOD-simulated performance of controlled drainage across the U.S. Midwest. Agric. Water Manag., 197, 54-66. https://doi.org/10.1016/j.agwat.2017.11.012
Youssef, M. A., Strock, J., Bagheri, E., Reinhart, B. D., Abendroth, L. J., Chighladze, G.,... Pease, L. (2023). Impact of controlled drainage on corn yield under varying precipitation patterns: A synthesis of studies across the U.S. Midwest and Southeast. Agric. Water Manag., 275, 107993. https://doi.org/10.1016/j.agwat.2022.107993
Zhang, Q., Fisher, T.R., Trentacoste, E.M, Buchanan, C., Gustafson, A.B., Karrh, R., Murphy, R.R., Keisman, J., Wu, C., Tian, R., Testa, J.M., Tango, P.J., (2021). Nutrient limitation of phytoplankton in Chesapeake Bay: Development of an empirical approach for water-quality management. Water Research, 188, 116407.