Use of Narrow Rows in Sprinkler-Irrigated Corn Systems to Increase Grain Yields, Aboveground Biomass, and Water and Nitrogen Use Efficiencies

Abstract

Narrow rows and optimum nitrogen applications are effective best management practices (BMPs) to enhance crop yield in an economically viable way. In a set of four studies, we aimed to compare the traditional method of planting (TMP) in wider rows (76.2 cm) with a lower plant population (84,600 seeds ha⁻¹) against a new BMP of planting in narrow rows (38.1 cm) with a higher plant population (158,000 seeds ha⁻¹). Implementation of the BMP resulted in 29.9 Mg ha⁻¹ of dry matter (DM) silage, which was 42.5% higher than the 21.0 Mg ha⁻¹ observed with the TMP. The p-values for the BMP versus TMP silage comparisons were p < 0.05, p < 0.05 and p < 0.001 for studies 2, 3, and 4, respectively, showing the significantly higher production with the BMP. Silage production water use efficiency (WUE) and agronomic nitrogen use efficiency (AE) were also higher with the BMP in studies 2, 3 and 4. The average harvested grain DM of the three BMP studies (9.9 Mg ha⁻¹) was 9.5% higher than the 9.0 Mg ha⁻¹ harvested grain DM with the TMP. The BMP of narrow rows with higher plant populations increases silage and grain production in sprinkler-irrigated systems.

1. Introduction

With over 90 million acres planted in the United States, corn (Zea mays L.) may well be the most-commonly planted crop in the country. Several climatic, physiological, and management factors affect the yields of this popular crop. There has been increasing interest in exploring the potential of planting corn in narrow rows to increase yields. The available literature thus far shows that yield responses to the use of narrow rows have been inconsistent.

1.1. Narrow Rows—Positive Responses

Several researchers have found that narrow rows can be used as an alternative management practice to increase yields. In Minnesota, Porter et al. [1] found that for each year of a 3 year study, corn yields ranged from 7.2 to 8.5% higher with 50.8 and 25.4 cm rows than 76.2 cm rows. They studied the effect of plant populations ranging from 54,360 to 98,840 plants ha⁻¹ and found no interaction between the yield responses due to row spacing and plant populations. Yields increased at populations of 86,480 plants per hectare or higher at Lamberton in 1992 and Waseca and Lamberton in 1994 [1]. Johnson and Hoverstad [2] conducted weed management studies in Minnesota and found significantly higher grain yields with 51 cm rows compared to corn grown in 76 cm rows.

In Iowa, Licht et al. [3] studied the effects of row spacing (50.8 and 76.2 cm) and plant population (74,130 to 103,780 plants ha⁻¹) across 22 site-years from 2009 to 2018. They found that in 11 of the 22 site-years, corn grain yields in 50.8 cm rows were higher by 0.3 to 1.19 Mg ha⁻¹. The increase they observed, which ranged from 2.7 to 10%, has also been seen in other studies. Additionally, the dry corn yields of the 50.8 cm rows were lower than those of the 76.2 cm rows in only 2 of the site-years. They concluded that in low-yield sites, there was not a difference, while in high-yield sites, the 50.8 cm rows produced yields that were 0.5 to 0.63 Mg ha⁻¹ higher, so farmers should use 50.8 cm rows in sites with yields greater than 14.75 Mg ha⁻¹.

Crozier et al. [4] studied the effects of plant population, row spacing, and nitrogen (N) management on corn yields grown in the Coastal Plain, Piedmont, and Mountain regions of North Carolina. Yields of corn, averaged across all N rates, grown in narrow rows with N application during sidedress (11.7 Mg ha⁻¹) were higher than yields with wider rows fertilized at planting or sidedress (11.0 Mg ha⁻¹) [4]. In Balcare, Argentina, Barbieri et al. [5] studied the effects of row spacing (35 and 70 cm) and N rates (0 and 120 or 140 kg ha⁻¹) on corn grain yields, finding narrow rows significantly increased yields.

In Indiana, Nielsen [6] found 38.1 cm rows increased yields by 2.7% when compared to 76.2 cm rows, though an increase in stalk breakage was observed in narrow rows. Nielsen [6] also found a significant interaction between hybrids, spacing, yields and location. Widdiecombe and Thelen [7] conducted studies about the effects of row width and plant density in six locations of the northern Corn Belt, reporting that corn grain yield increased by 2 and 4% with 56 cm and 38 cm rows, respectively, compared to 76 cm rows.

Bernhard and Below [8,9] studied the effects of plant populations ranging from 94,000 to 139,000 plants ha⁻¹ planted with 51 and 76 cm row spacing in Illinois. They found a significant interaction between hybrid and row spacing; however, when averaged across hybrids and all row spacings, hybrid yields were an average of 0.8 Mg ha⁻¹ higher with 51 cm row spacing. They found the highest grain yield of 18.5 Mg ha⁻¹ with 109,000 corn plants ha⁻¹ in a 51 cm row. In Michigan, Thelen [10] found that in narrow (38.1 cm) rows, there was an increase in yield of 0.57 Mg ha⁻¹ in coarse-textured soils and 0.28 Mg ha⁻¹ in finer-textured clay loam soils when compared to corn grown in 76.2 cm rows in these differently textured soils. Nelson et al. [11] studied the effects of hybrids (10 hybrids), row spacing (38.1 and 76.2 cm rows) and plant population density (74,130 and 88,950 seeds ha⁻¹) on corn grain yields in Missouri, finding that when averaged over hybrid and seeding rate, average grain yields across all treatments were 2.7% higher in 38.1 cm rows.

In Minnesota, Sharratt and McWilliams [12] studied the effect of hybrids and narrow (38, 57, and 76 cm) rows on grain yields and water use efficiency, finding the grain yield of corn grown in narrow rows equaled in one year, and surpassed in the second year, the yields of the 76 cm rows (wider rows). In the second year, the water use of the 38 cm row was higher than that of the 56 and 78 cm rows. Even when the narrow (38 cm) rows used significantly more water, the agronomic efficiency per unit of yield was not significantly different during the second year.

Few studies have been conducted for narrow rows under irrigation. Shapiro and Wortmann [13] studied the effect of N rate (0 to 252 kg N ha⁻¹), row spacing (51 and 76 cm), and plant density (61,800; 74,160; and 86,520 plants ha⁻¹) in eastern Nebraska. This 3 year study under irrigation found that narrow rows increased grain yield by 4%. Optimum N rates for yield were not affected by row spacing, and increasing plant density above 61,800 plants ha⁻¹ did not increase grain yield [13]. When averaging three plant densities and four N rates, the aboveground biomass increased in two of the three years with 51 cm rows compared to the 76 cm rows, increasing from 13.97 to 14.58 Mg ha⁻¹ in 1996 and from 11.84 to 12.75 Mg ha⁻¹ in 1997 (p < 0.088). At the highest plant density (86,520 plants ha⁻¹), the narrow rows had significantly higher grain yield (p < 0.05) in 1997 for all N rates (e.g., 8.51 vs. 7.12 Mg ha⁻¹ at an N rate of 252 kg N ha⁻¹). They found that the applied N had an agronomic efficiency of 5.8 to 22.1, 4.8 to 14.6, and 3.8 to 9.6 kg grain yield increase per kg of N applied for the 84, 168 and 252 kg N application rates, respectively.

1.2. Twin Rows—Positive Responses

Karlen and Camp [14] found that twin-row production increased yields in sandy soils in the Atlantic Coastal Plain. They also studied the effect of water management. Irrigation nearly doubled yields in 1980 and 1981, increasing grain production by 150 and 161%, respectively. These years experienced periods of drought during the growing season, while 1982 (a year which saw much higher precipitation) saw an increase in yields of only 8% with irrigation. They found that grain yields were not significantly different, although the computer-based scheduling method used less irrigation than the other irrigation methods studied. Karlen and Camp [14] found that when averaged across water management systems, twin rows had an average of 0.64 Mg ha⁻¹ higher yield than that of single rows with 96 cm spacing.

Mackey et al. [15] studied the effects of row spacing (38.1 cm rows, twin 20.3 cm rows on 76.2 cm centers, and 76.2 cm rows), plant populations (74,130; 86,485; 98,840; and 111,195 seeds ha⁻¹), and hybrids (3 hybrids) for corn grown in Lexington and Hodgenville, Kentucky. They found that in 2011, corn grain yields at Hodgenville with twin 20.3 cm rows on 76.2 cm centers had 6.7% higher yields (13.1 Mg ha⁻¹) than those obtained with 76.2 cm rows (12.3 Mg ha⁻¹) or with 38.1 cm rows (12.8 Mg ha⁻¹). Williams et al. [16] conducted a three-year study about the effects of weed management and row spacing on yields in North Carolina and concluded that twin rows may be a good management option to increase yields. Liang et al. [17] studied the effects of row spacing (twin rows [TR] 20 cm apart on 75 cm centers; narrow rows [NR] of 50 cm; narrow twin rows [NTR] 20 cm apart on 50 cm centers; and conventional rows [CR] 75 cm apart) and plant population of an early-maturing (93 days) corn grown at Rakuno Gakuen University, located in Ebetsu, Japan. They found that the grain yields were highest for NTR, followed by TR, NR, and CR across all planting densities, and that total biomass production had the same response as the grain production.

1.3. Narrow and Twin Rows—Negative Responses

1.3.1. Narrow Rows

In South Africa, Haarhoff and Swanepoel [18] conducted research with 40,000 to 80,000 plants ha⁻¹ planted at 0.5, 0.76 and 1 m row spacings to evaluate the effects of row spacing on yields under conservation agriculture. Higher yields were achieved at high plant populations and wide row spacings, and they concluded larger plant populations have higher water use and extract more water from the soils than smaller populations, reducing the plant-available water.

1.3.2. Twin Rows

Balkcom et al. [19] conducted four field studies and reported that a twin-row configuration did not have much impact on corn yields, and plant density had the greatest effect on yields. In these studies, the twin rows were spaced 19 cm apart on 76 cm centers and compared to single rows spaced 76 cm apart in two sites in Alabama and two sites in Florida. They reported the minimal corn yield increases with twin rows may not warrant changing to this configuration in dryland, but noted that if you are using these systems there will not be yield decreases. The results showed that twin rows did not provide a consistent benefit to yields, and they suggested that soil moisture was a limiting factor in achieving higher yields with this row configuration.

Novacek et al. [20] reported that grain yields with twin-row irrigated maize were similar to single-row production, concluding that since there were few interactions between hybrids, plant population, and row spacing, twin-row production has little potential to increase maize grain yields in the western Corn Belt. Robles et al. [21] studied the effects of twin rows, hybrids (3 hybrids), and plant density (from 69,000 to 105,000 plants ha⁻¹) on corn grain yields for three years in west-central Indiana. They found that twin rows did not produce higher dry corn grain yields than single rows at any hybrid combination or plant density. Kratochvil and Taylor [22] studied the effect of row spacing (twin rows 19.1 cm apart on 76.2 cm centers; and 76.2 cm rows), hybrids (28 hybrids), and populations on corn grain yield in Maryland, and found across all the studied hybrids, plant populations, and locations in Maryland, that twin rows did not increase the corn yields or provide any advantage in the Delmarva region.

1.4. Silage Responses to Narrow Rows

Some studies have found a positive correlation between narrow rows or twin rows and higher corn silage production [23,24,25,26,27]. Yilmaz et al. [23] compared twin 20 cm rows on 55 cm centers, with narrow (50 cm) rows and conventional (75 cm) rows in the early 2000s in Turkey. Twin rows produced 10.2% more dry silage biomass than conventional rows and 5.9% more dry matter silage than narrow rows [23]. In contrast, Cox et al. [24] found in the early 2000s that the DM silage production with narrow rows (17.6 Mg ha⁻¹ ) was higher than that of twin-row (17.2 Mg ha⁻¹) and conventional production systems (16.6 Mg ha⁻¹) in the northeastern USA. Additionally, in studies conducted in the 1990s, Cox et al. found higher DM silage production with narrow rows than conventional (wider) rows in this region [25]. Dry matter silage production was higher with narrow rows in 1994 and 1996 but not in 1995 [25], and averaged 4.2% higher across the three years. Cox et al. [25] reported higher net income with the narrow (38.1 cm) rows compared to the wider (76.2 cm) rows. Cox and Cherney [26] also reported higher DM silage production with narrow (38 cm) rows than the conventional (76 cm) row spacing, regardless of plant density and nitrogen levels, in the 1990s. In studies conducted in Michigan, Widdicombe and Thelen [27] found an increase in silage biomass of 1 Mg ha⁻¹ with narrow (38 cm) rows compared to wider (76 cm) rows, similar to the increases reported in other research in the northeastern U.S. [24]. However, Baron et al. [28] observed mixed results in their irrigated studies, finding higher forage production with narrow rows over wider rows with one hybrid, but not the other. Furthermore, research in Alberta, Canada has found planting in narrow rows did not affect silage production in irrigated forage systems [29].

1.5. Research Gaps

There is a need to conduct new studies to assess the effects of narrow rows under irrigated conditions in the dry West. Lee [30] reported that in the U.S, narrowing the width of corn rows usually does not contribute to increased yields south of 43° N latitude; in other words, for corn grown in the central and southern United States, wide rows are better for maximum yields. However, results from more recent research, including studies with twin rows in Kentucky [15] and North Carolina [16], which found an advantage to this configuration, as well as from a Licht et al. [3] study in Iowa that found 50.8 cm rows increased yields compared to 76.2 cm rows in 11 of 22 site-years in Iowa (and decreased in only 2 site-years), suggest that this topic should be revisited. Additionally, we could not find literature on narrow row studies conducted in irrigated systems of Colorado and how narrow row systems affect irrigation water use efficiencies and agronomic efficiency or silage production, supporting the need for research in this area.

Most narrow row studies have focused on assessing their effects on grain yields. For irrigated systems of the West, research on the effects of narrow rows on grain corn as well as silage or aboveground total biomass production is needed so farmers could evaluate the potential to use narrow rows for silage or grain production. Most of the research on silage production with narrow rows has been conducted in the northeastern U.S. (e.g., New York) and Illinois, finding positive results for silage production when using narrow rows. However, mixed results have been reported in some studies (e.g., irrigated systems in Alberta, Canada). There is a need for additional research to assess the effects of narrow rows in irrigated silage systems in the western U.S., such as those under sprinkler irrigation in Colorado. Our goals were to assess the effects of nitrogen rates and narrow rows on corn grain and silage production as well as the effects on irrigation water application efficiencies and agronomic nitrogen use efficiencies. Our hypothesis was that narrow rows with a higher plant population would increase silage, harvest grain production, water application efficiencies, and agronomic nitrogen use efficiencies.

2. Materials and Methods

Four different studies were conducted from 2018 to 2020 (

Table 1. Summary of the main treatments and management factors in the four independent studies comparing the traditional method of planting (TMP) in 76.2 cm rows with a lower plant population to the best management practice (BMP) of planting in 38.1 cm (narrow) rows with a higher plant population *.

	Studies
Factor	Study 1	Study 2	Study 3	Study 4
Planting density (seed ha−1)	84,600 seeds ha−1 with BMP and 84,600 seeds ha−1 with TMP	158,000 seeds ha−1 with BMP and 84,600 seeds ha−1 with TMP	158,000 seeds ha−1 with BMP and 84,600 seeds ha−1 with TMP	158,000 seeds ha−1 with BMP and 84,600 seeds ha−1 with TMP
Variety	Channel 193-53 STXRIB	Channel 193-53 STXRIB	Channel 192-10 STXRIB	Channel 192-10 STXRIB
N rates (kg ha−1)	0, 202, 246 and 314	0 and 202	0 and 202	157
Plot size	4.6 m × 14.6 m	4.6 m × 14.6 m	4.6 m × 14.6 m	20 m × 37.5 m
Experimental design	Randomized block	Randomized block	Randomized block	Adjacent strips(paired)
Tillage operations	roto-tilled to a depth of 15 cm prior to planting	roto-tilled to a depth of 15 cm prior to planting	roto-tilled to a depth of 15 cm prior to planting	disced, mulched, and cultivated to a depth of 7.5 cm prior to planting
Previous crop	Corn	Corn	Corn	Winter wheat
Irrigation	Valley® linear-move sprinkler	Valley® linear-move sprinkler	Zimmatic® linear-move sprinkler	Zimmatic® linear-move sprinkler
Harvesting	Harvesting was done by hand when grain water content was at 15.5%, collecting ears by hand from an area of 11.6 m2	Harvesting was done by hand when grain water content was at 15.5%, collecting ears by hand from an area of 11.6 m2	Harvesting was done by hand when grain water content was at 15.5%, collecting ears by hand from an area of 11.6 m2	Harvesting was done by machine when grain water content was at 15.5%, encompassing an area of 23.2 m2

* In study 1, we compared the TMP to a BMP with the same plant population.

). All of these studies focused on assessing the effects of narrow rows on silage production and harvested grain yields by comparing a new best management practice (BMP) with narrow crop row spacing of 38.1 cm with higher plant populations to the traditional method of planting (TMP) in 76.2 cm rows with lower plant populations. The first study, which was conducted in 2018, compared the BMP of planting in 38.1 cm (narrow) rows to the TMP of planting in 76.2 cm (wide) rows using the same plant population at four different N fertilizer rates (0, 202, 246 and 314 kg urea N ha⁻¹). The second study (2019) and the third and fourth studies (2020) all compared 38.1 cm rows with higher plant populations to the traditional row spacing of 76.2 cm (Table 1). These studies were conducted on a Fort Collins clay loam soil (fine-loamy, mixed, mesic Aridic Haplustalfs) with a 1 to 2% slope at the Agricultural Research Development and Education Center (ARDEC) (40°39′6″ N, 104°59′57″ W, 1535 m above sea level) near Fort Collins, CO. The region is semiarid with a mean daily temperature of 9.3 °C and mean annual precipitation of 272 mm (2001–2019). A summary of the factors for studies 1, 2, 3, and 4 is provided in Table 1.

Study 1 was established in 2018, incorporating the treatments of crop row spacing (38.1 and 76.2 cm row spacing) and N fertilizer application (0, 202, 246 and 314 kg urea N ha⁻¹) in a 2 × 2 factorial design with three replicates (blocks) on plots measuring 4.6 m × 14.6 m. These 4.6 m × 14.6 m plots were labelled as the small plots (blocked studies; 2018 [study 1], 2019 [study 2] and 2020 [study 3]). Each spring, granular urea (46-0-0) was surface applied to the fertilized plots with a handheld fertilizer spreader after planting but before germination. Right after urea application, the study site was irrigated to incorporate the applied urea. Since in 2018 (study 1) no significant differences were observed from 202 kg N ha⁻¹ to 314 kg N ha⁻¹ for either the BMP or TMP with the same plant population, in 2019 (study 2) and 2020 (study 3) the experiment was repeated in those same plots, except with 2 levels of crop row spacing (38.1 and 76.2 cm) and 2 levels of N application (0 and 202 kg urea N ha⁻¹) (Table 1). This response to N rate from study 1 agrees with findings from Villacis et al. [31] that reported profit-maximizing N rates ranging from 162 to 197 kg N ha⁻¹ for continuous no-till corn planted at 76.2 cm row spacing.

Prior to 2018, the area used for studies 1, 2, and 3 was planted to corn (during 2016 and 2017) without any N fertilizer applied. In the spring of each year from 2016 to 2020, the study area was roto-tilled to a depth of 15 cm prior to planting corn. One month after tillage, corn was planted with a John Deere MaxEmerge XP planter. In 2018 (study 1), the corn variety Channel 193-53 STXRIB was planted at densities of 84,600 seeds ha⁻¹ in the 38.1 cm rows and 84,600 seeds ha⁻¹ in the 76.2 cm rows. The same variety was planted in 2019 (study 2), at densities of 158,000 seeds ha⁻¹ (83% higher) in the narrow (38.1 cm) rows and 86,400 seeds ha⁻¹ in the standard (76.2 cm) rows (Table 1). Since the corn variety Channel 193-53 STXRIB was not available during 2020 due to the COVID-19 pandemic, a new variety, Channel 192-10 STXRIB, was planted in the small plots (study 3; blocked study) with the same row spacing, nitrogen rates and planting densities used in 2019 (Table 1).

A second study location was added in 2020 (study 4) in a neighboring field with the same soil type, and consisted of two adjacent strips, with one of the strips planted with the BMP of using narrower (38.1 cm) spacing with higher plant density, and the other strip planted with the TMP of standard (76.2 cm) row spacing with lower plant density (Table 1). Each strip was 20 m in width by 150 m in length. Four paired locations were sampled along the 150 m, at approximately 18.8, 56.3, 93.8 and 131.3 m from the north side of the plots, and were labeled as the large, paired strip plots. The longer strips allow us to harvest the study area using a small plot combine to simulate harvesting by mechanized farm equipment for two different row widths.

In the spring of 2020, the area for study 4 (stripped study) was disced, mulched, and cultivated to a depth of 7.5 cm; in the fall of 2019 it was disced to a depth of 7.5 cm and subsoiled to a depth of 25 cm following a crop of winter wheat that was planted in the fall of 2018 and harvested in the summer of 2019 (Table 1). The same variety used for study 3, Channel 192-10 STXRIB, was planted in strip rows at 158,000 seeds ha⁻¹ in the 38.1 cm rows (83% higher in the stripped 2020 study) and 84,600 seeds ha⁻¹ in the 76.2 cm rows. The strips were fertilized with urea U (46-0-0) that was surface-applied with a Barber drop spreader that was 3 m in width, delivering urea at 157 kg urea N ha⁻¹ after planting but before germination, and the site was irrigated to incorporate the applied urea.

During 2020, planting, urea fertilization and irrigation were done the same day at the small plots (study 3) and large, paired strip plots (study 4). For both the small plots (blocked studies) and for the large, paired strip plots, a Valley^® linear-move sprinkler irrigated the study areas from 2016 to 2019, but that was replaced in 2020 with a Zimmatic^® linear-move sprinkler, which irrigated the corn as needed (Table 1). For determining irrigation, we used an Irrometer^® solar powered wireless mesh network to collect real-time soil moisture and temperature data in our different research fields. This IRROmesh^® Monitoring System is connected to WaterMark^® sensors at the 22.9 and 45.7 cm soil depths and measures soil tension in centibars. We use this information to estimate how much water the plants are using and how much irrigation will be needed weekly to maintain available soil water at 50% or greater. All corn plots received the same amount of water per irrigation regardless of row spacing or nitrogen treatment.

Silage and harvested grain yields were measured from samples that were collected during two distinct field operations each year: (i) sampling for silage (collecting samples at physiological maturity, R6), in which 15 randomly selected plants were cut by hand just above ground level; and (ii) a harvest sampling performed after standing plants had field-dried to approximately 15.5% water content, in which grain was harvested from all plants in a specified area of two adjacent rows, each one 7.62 m in length. Biomass samples were collected by hand to assess silage production at physiological maturity (R6-black layer) at the small plots (blocked studies: studies 1, 2, and 3) and large, paired strip plots (study 4) at approximately 146 days after planting [DAP]). Sampling at harvesting was done by hand for the small plots (studies 1, 2, and 3) and by grain harvesting machine (study 4) in the large, paired strip plots at approximately 173 DAP.

After the biomass samples were collected, ears were manually removed from plants, oven dried, and then shelled in a corn sheller, while the stalks and leaves were shredded together, subsampled, and oven dried. The shelled cobs were also retained and oven dried. For each plot in the blocked studies (studies 1, 2 and 3), the harvest sampling was performed by collecting ears by hand from an area of 11.6 m² (Table 1). Those ears were air-dried and mechanically threshed, and the grain was then subsampled and oven dried. For the stripped study (study 4), the harvest sampling was performed with a Massey Ferguson 8XP plot combine that collected grain separately from four distinct locations within each strip along the study’s length, each encompassing an area of 23.2 m² (Table 1). In study 4, using the mechanized equipment allowed us to harvest a larger area (double the area) than that harvested for the small plots (studies 1, 2 and 3), simulating a larger farming harvesting operation. Harvested grain was subsampled and oven dried. (All oven-dried samples were left in a 60 °C oven for at least 48 h.) In both study locations, the biomass samples were collected outside the harvest sample areas. For the biomass sampling at physiological maturity (R6), yields were separately calculated for biomass grain, stalks plus leaves, and cobs (data not shown). The total aboveground silage biomass for biomass sampled at physiological maturity (R6) was calculated by summing all the compartments. In the present work, we define this total aboveground corn biomass production at physiological maturity (R6) as aboveground silage production. Yield of harvest grain was also calculated for the samples collected by hand in the small plots (blocked studies) and by machine in the large, paired strip plots.

Water use efficiency (WUE) was calculated for each study by dividing the aboveground silage or harvested grain by the total water (irrigation plus precipitation) input during the growing season. The WUE was calculated with the following formula: WUE = kg of dried yield at a given plot ÷ (mm of total applied irrigation water during the growing season + mm of total precipitation during the growing season). The agronomic nitrogen use efficiency (AE) was calculated for the fertilized study. The AE was calculated with the following formula: AE = (kg of dried yield at a given N-fertilized plot) ÷ given N fertilizer rate).

For studies 1, 2, and 3, treatment effects on these responses within each year were determined by analysis of variance, using function aov in R (R version 4.1.0), while treatment effects in the stripped study (study 4) were determined using a paired t-test, using function t.test in R.

3. Results

Nitrogen fertilizer significantly increased the dry weight biomass (

Table 2. Differences in total biomass silage and harvested grain (HG) mean dry weight yields (Mg ha⁻¹) of corn grown in the small plots (study 1) during 2018 under different row spacing systems and nitrogen rates ^¶.

		Row Spacing Level β
Plant Compartment	N Fertilizer Level (kg ha−1)	38.1 cm Rows	76.2 cm Rows	Mean Over Spacing Levels ¥
Silage (Mg ha−1)	N = 0 N = 202	16.7 23.8	17.2 20.6	17.0 B ¥, b ¥ 22.2 A, a
	N = 246	23.6	23.6	23.6 A, a
	N = 314	23.5	23.3	23.4 A, a
	Mean over N levels ¥	22.3 A ¥, a ¥	21.2 Aa
HG (Mg ha−1)	N = 0	9.0	6.0	7.2 B ¥, b ¥
	N = 202	10.8	9.8	10.3 A, a
	N = 246	9.8	9.2	9.4 A, ab
	N = 314	11.3	9.3	10.3 A, a
	Mean over N levels ¥	10.4 A ¥, a ¥	8.6 B, b

^¶ Silage was collected at physiological maturity (R6-black layer). Harvested grain (HG) was collected when grain water content was at 15.5%. ^β In 2018, planting densities were 84,600 ha⁻¹ in the 38.1 cm rows and 84,600 ha⁻¹ in the 76.2 cm rows. ^¥ Within a plant compartment, treatments with different letters are significantly different; capital letters indicate differences are significant at α = 0.10, and lowercase letters indicate differences at α = 0.05.

) of silage (whole aboveground plant biomass: cobs, stalks, leaves and grain) (p < 0.10) sampled at physiological maturity (R6), as well as the dry grain at harvest (p < 0.10). On average, the application of 202, 246 and 314 kg N ha⁻¹ significantly increased the production of silage and grain when compared to zero application of N fertilizer, as shown in Table 2. On average, yields for all the applied nitrogen treatments were significantly higher than for the zero-nitrogen-fertilizer plots for all the plant compartments shown in Table 2, except for the harvested grain of the 246 kg N ha⁻¹ treatment. However, since the 246 kg N ha⁻¹ treatment was also not significantly different from the 202 and 314 kg N ha⁻¹ treatments (that were significantly different from zero N), all silage and harvested grain responded significantly to N fertilizer applications over the control (zero fertilizer application). There was not a significant interaction between row spacing and N fertilizer application. Nitrogen fertilizer increased the silage production, as well as the dry grain at harvest, by 35.7 and 39.0%, respectively, compared to the non-fertilized plots.

The BMP of narrow rows with the same plant population significantly increased the dry grain at harvest (p < 0.085; Table 2). The dry grain yield of corn increased an average of 20.9% across all nitrogen rates (0, 202, 246 and 314 kg N ha⁻¹) with the BMP of using narrow (38.1 cm) rows, which was higher than the TMP of using wider (76.2 cm) rows. For the N application rate of 202 kg N ha⁻¹, the yield observed with the BMP of planting in narrow (38.1 cm) rows (10.8 Mg ha⁻¹) was 10.2% higher than that observed with the TMP (9.8 Mg ha⁻¹; Table 2).

In study 1, there was not a significant difference in WUE (p = 0.31) and AE (p = 0.12) of silage production when the BMP was compared to the TMP. However, since the grain production was significantly higher with the BMP of narrow rows with the same plant population, both AE (p < 0.05) and WUE (p < 0.01) for corn grain production were significantly higher with the BMP when compared to the TMP with wider rows.

In study 2, we observed the same response with the 0 and 202 kg N ha⁻¹ in 2019 as we did in 2018 (

Table 3. Differences in total biomass silage and harvested grain (HG) mean dry weight yields (Mg ha⁻¹) of corn grown in the small plots (study 2) during 2019 under different row spacing systems and nitrogen rates ^¶.

	0 Applied N		202 kg ha−1 Applied N		Spacing Levels §		Nitrogen Levels *
Plant Compartment	38.1 cm β ROWS	76.2 cm β Rows	38.1 cm β Rows	76.2 cm β Rows	38.1 cm β Rows ¥	76.2 cm β Rows ¥	0 Applied N ¥	202 kg ha−1 Applied N ¥
Silage (Mg ha−1)	13.9	10.1	26.4	20.6	20.1 ¥	15.3 ¥	12.4 ‡	24.1 ‡
HG (Mg ha−1)	6.2	4.5	9.5	8.1	7.9 ¥	5.9 ¥	5.4 ‡	8.9 ‡

^¶ Plant compartments were collected at physiological maturity (R6-black layer), except harvested grain, which was collected when grain water content was at 15.5%. ^β In 2019, planting densities were 158,000 seeds ha⁻¹ in the 38.1 cm rows and 84,600 seeds ha⁻¹ in the 76.2 cm rows. ^¥,^‡ Within a plant compartment, treatments with different symbols are different at ^¥ 0.05 > p ≥ 0.01; ^‡ p < 0.001. ^§ Mean over spacing levels; * Mean over N levels.

). Additionally, there was also no interaction between nitrogen rates (0 and 202 kg N ha⁻¹) and row spacing (38.1 and 76.2 cm) treatments in study 2 during 2019 (Table 3). In study 2 in 2019, the N fertilizer increased the silage production as well as the dry grain at harvest, by 94.3, and 64.8%, respectively, which were higher than the non-fertilized plots (p < 0.01, average across row treatments; Table 3).

The BMP with the 202 kg N ha⁻¹ treatment for the narrow (38.1 cm) rows with a higher plant population increased the silage dry weight biomass (p < 0.05) as well as the dry grain at harvest (p < 0.05) by 28.3, and 18.0%, respectively, when compared to the TMP of wider (76.2 cm) rows with a lower plant population (Table 3). A greater response was observed for the BMP with the narrow (38.1 cm) rows without nitrogen fertilizer (0 kg N ha⁻¹) versus the TMP with wider (76.2 cm) rows without nitrogen fertilizer (0 kg N ha⁻¹); the BMP had an increase in the dry weight biomass of corn silage (whole aboveground biomass at R6-black layer), as well as the dry grain at harvest, of 37.6 and 37.8%, respectively, with the non-fertilized treatment.

In study 2, the silage had higher WUE (p < 0.05) and AE (p < 0.10) with the BMP with narrow (38.1 cm), fertilized rows and a higher plant population than the TMP with wider (76.2 cm), fertilized rows and a lower plant population. Additionally, since the grain production was significantly higher with the BMP with narrow (38.1 cm) rows, both AE (p < 0.05) and WUE (p < 0.05) for corn grain production were significantly higher with the BMP with narrow (38.1 cm), fertilized rows than the TMP of wider (76.2 cm), fertilized rows with a lower plant population.

The same response to nitrogen fertilizer application observed in 2018 and 2019 was also observed in study 3 (small plots) in 2020, and there was also no interaction between nitrogen rates and row spacing treatments (

Table 4. Differences in total biomass silage and harvested grain (HG) mean dry weight yields (Mg ha⁻¹) of corn grown in the small plots (study 3) during 2020 under different row spacing systems and nitrogen rates ^¶.

	0 Applied N		202 kg ha−1 Applied N		Spacing Levels §		Nitrogen Levels *
Plant Compartment	38.1 cm β Rows	76.2 cm β Rows	38.1 cm β Rows	76.2 cm β Rows	38.1 cm β Rows ¥	76.2 cm β Rows ¥	0 Applied N ¥	202 kg ha−1 Applied N ¥
Silage (Mg ha−1)	18.2	14.8	28.1	21.8	23.2 ¥	18.3 ¥	16.5 ‡	25.0 ‡
HG (Mg ha−1)	5.3	5.3	10.6	10.5	7.9	7.9	5.4 ‡	10.5 ‡

^¶ Plant compartments were collected at physiological maturity (R6-black layer), except harvested grain, which was collected when grain water content was at 15.5%. ^β In 2020, planting densities were 158,000 seeds ha⁻¹ in the 38.1 cm rows and 84,600 seeds ha⁻¹ in the 76.2 cm rows. ^{¥, ‡} Within a plant compartment, treatments with different symbols are different at ^¥ 0.05 > p ≥ 0.01; ^‡ p < 0.001. ^§ Mean over spacing levels; * Mean over N levels.

). The nitrogen fertilizer increased the silage dry weight biomass (whole aboveground plant biomass [cobs, stalk, leaves and grain at R6]), as well as the dry grain at harvest, by 51.5 and 94.4%, respectively (p <0.01, average across row treatments; Table 4).

In study 3, the BMP of narrow (38.1 cm) rows with a higher plant population increased the silage (whole aboveground plant biomass [cobs, stalks, leaves and grain at R6]) production (p < 0.05) by 28.6% compared to the TMP with wider (76.2 cm) rows with a lower plant population, under the 202 kg N ha⁻¹ treatment (Table 4). The same response in silage production was observed for the average with the BMP with narrow (38.1 cm) rows with a higher plant population across nitrogen rates (0 and 202 kg N ha⁻¹) when it was compared to the average for the TMP with the wider (76.2 cm) rows with a lower plant population across nitrogen rates (0 and 202 kg N ha⁻¹) at R6 (black layer). The dry weight of the harvested grain with the BMP with narrow (38.1 cm), fertilized rows and a higher plant population was on average 1.3% higher, but not significantly different when compared to the TMP with wider (76.2 cm), fertilized rows and a lower plant population. In study 3, the silage had a higher WUE (p < 0.05) and AE (p < 0.05) with the BMP of narrow (38.1 cm), fertilized rows with a higher plant population than the TMP with wider (76.2 cm), fertilized rows and a lower plant population. However, since there were no significant differences in harvested grain production, both AE (p = 0.84) and WUE (p = 0.90) for corn grain production were not different in study 3.

In study 4, the large, paired strip plots had the same response from the silage that was observed in study 2 (2019) and study 3 (2020) (

Table 5. Differences in total biomass silage and harvested grain (HG) mean dry weight yields (Mg ha⁻¹) of corn grown in the large, paired strip plots during 2020 under different row spacing systems (study 4) ^¶.

	38.1 cm Rows β			76.2 cm Rows β
Plant Compartment	Mean Yield (Mg ha−1)	S.D. Yield (Mg ha−1)	n	Mean yield (Mg ha−1)	S.D. Yield (Mg ha−1)	n	Paired-t p value
Silage	35.2	1.5	4	20.6	1.7	4	<0.001
HG	9.7	0.5	4	8.9	0.9	4	0.16

^¶ Plant compartments were collected at physiological maturity (R6-black layer), except harvested grain, which was collected when grain water content was at 15.5%. ^β In 2020, planting densities were 158,000 seeds ha⁻¹ in the 38.1 cm rows and 84,600 seeds ha⁻¹ in the 76.2 cm rows.

). The BMP with narrow (38.1 cm) rows with a higher plant population increased the silage production (p < 0.001) by 70.9% (Table 5) compared to the TMP of wider (76.2 cm) rows with a lower plant population. Although the harvested grain was 9.0% higher with the BMP of narrow (38.1 cm) rows with a higher plant population than the TMP of wider (76.2 cm) rows with a lower plant population, the statistical analysis shows that it was not significant at the p-value of p < 0.10, and that it was close to being significant with a p-value of p = 0.16. For the silage production in study 4, the BMP of planting in narrow (38.1 cm), fertilized rows with a higher plant population, had a higher WUE (p < 0.001) and AE (p < 0.001) than the TMP with wider (76.2 cm) rows with a lower plant population. For the harvested grain, the WUE and AE of the BMP were not significant at p < 0.10, but were close to a significant value (p = 0.16) when compared to the TMP.

4. Discussion

In study 1, across all N rates, harvested grain yields with the BMP of using narrow (38.1 cm) rows were significantly higher (20.9%) than the harvested grain yields observed for the TMP with the wider (76.2 cm) rows. This response of increased harvested grain with the BMP was greater than the 6.4% increase in yields with narrow rows observed by Crozier et al. [4] across all N rates. These findings are also in agreement with one of the two irrigated studies [13] that we found that concluded that optimum N rates for harvested corn grain yield were not affected by row spacing. Our increase of 1.1 Mg ha⁻¹ was similar to the increase of 1.0 Mg ha⁻¹ observed in the late 1990s by Shapiro and Wortmann [13] for the N rate of 254 kg N ha⁻¹. We also noticed in study 1 that although nitrogen rates did not increase silage production, the 202 kg N ha⁻¹ rate with narrow rows increased biomass silage production by 15.5% when compared to the wider rows (Table 2). After considering this response with the 202 kg N ha⁻¹ in study 1, we tested the BMP with narrow rows and a higher plant population against the TMP with wider rows and a lower plant population in 2019 and 2020 (Table 1).In studies 2, 3, and 4, the narrow (38.1 cm) rows with an average of 86.8% higher plant population increased the silage production by an average of 42.5% compared to the conventional, wider (76.2 cm) row spacing with a lower plant population, and this increase was significant. Additionally, the harvested grain production for studies 2, 3, and 4 was 18.0% (p < 0.05), 1.3% (n.s.), and 9.0% (p = 0.16) higher, respectively, than harvested grain with the conventional, wider (76.2 cm) rows, for an average increase across the three studies of 9.5%. We propose that since in all three studies the silage production was significantly higher with the BMP of planting in narrow rows with a higher plant population, these agronomic practices can significantly increase silage production with sprinkler-irrigated systems. We also propose that since the average harvested grain production for three studies was 9.5% higher (in study 2, was significantly higher, and in study 4, had a p-value of p = 0.16), this BMP could also increase the average harvested grain production with these sprinkler-irrigated systems.

It has been suggested that narrowing the width of corn rows usually does not contribute to higher yields in areas south of 43° N latitude in the United States [23]. Our results support the hypothesis that silage and/or harvested grain yields of sprinkler-irrigated corn systems south of 43° N latitude in the United States can be increased with a BMP of using narrow rows with a higher plant population. Since corn for silage could be harvested until just prior to R6 (black layer), we propose that the increases in total biomass production at R6 suggest higher silage production at earlier growth stages with this BMP.

Since the value of a metric ton of silage and/or the value of the 10% higher corn production will be higher than the cost of the additional seed needed with the BMP, our results suggest that this BMP will be economically viable and have a positive economic impact on sprinkler-irrigated corn grain and/or silage production. Furthermore, no additional water or N fertilizer was used with the narrow row system, which increased the AE and WUE of the applied water and nitrogen with this BMP.

Our positive results with narrow rows in irrigated systems agree with the other two irrigated studies with narrow rows that we found in the literature. Karlen and Camp [14] reported yields with twin-row production in irrigated systems in sandy soils in the Atlantic Coastal Plain that were 0.64 Mg ha⁻¹ higher than single row production. Shapiro and Wortmann [13] studied irrigated corn systems closer to Colorado (e.g., eastern Nebraska) from 1996 to 1998 and found production increased by 4% with narrow rows.

Our observation of increased agronomic nitrogen use efficiency agrees with findings from Shapiro and Wortmann [13] that nitrogen use efficiency increases with narrow rows. The benefits we found with narrow rows in Colorado in our 2018 to 2020 studies are supported by responses that have been observed in Nebraska [13] and the Atlantic Coastal Plain [14]. We propose that by irrigating to meet the water needs of a larger plant population, yields could potentially increase even in areas south of 43° N latitude, suggesting that this could be a potential BMP to increase silage and/or grain yields at harvest. Although we did not sample the plots at R4 (or other periods closer to R6 such as R5.5 [50% kernel milk] or R5.8 [20% kernel milk] when we could also harvest the crop for silage), the higher biomass production at R6 suggests that total biomass production at R4, R5.5 and R5.8 would also be higher, and thus there is potential to use this system for silage production.

Our results show that the use of narrow rows in center-pivot-irrigated systems for corn production is beneficial because the irrigation water is delivered above the canopy, so the plant does not interfere with the movement of water. We suggest that when using narrow rows with an irrigated furrow system, the plants will interfere with the movement of water, and new approaches need to be studied, such as two 38.1 cm rows followed by a furrow. Our studies with narrow rows showing a 10% increase are in the upper limit of previous increases in yields reported by other researchers. Some of the increases in yields reported in the literature for non-irrigated sites include increases of 7.2 to 8.5% in Minnesota [1]; 2.7 to 10% in Iowa [3]; 6.4% in the Coastal Plain, Piedmont, and Mountain regions of North Carolina [4]; 2.7% in Indiana [6]; 2 to 4% in the northern Corn Belt [7]; and 2.7% in Missouri [11]. Mackey et al. [15] reported 6.7% higher yields with twin 8 inch rows on 76.2 cm centers than systems with 76.2 cm rows or 38.1 cm rows.

• Although additional research is needed to verify our 2018 to 2020 results in other soils and other sites, we propose that this BMP with narrow rows is a viable practice for sprinkler-irrigated systems in Colorado and can potentially increase silage and harvested grain yields compared to the TMP. Our average increase in dry matter silage production with the BMP of planting in narrow rows with a higher plant populations was 8.9 Mg ha⁻¹ (42.5%), which is higher than the 0 to 10.2% increase that has been observed in other studies [23,24,25,26,27,28,29]. Although there is a need for additional research on silage production with sprinkler-irrigated systems, our preliminary assessment of cost of seed vs. the value of the increased silage produced with narrow rows suggests that this will be an economically viable practice, which is consistent with findings from research conducted in the northeastern U.S. [25].
• While additional assessment of the biomass produced at R4, R5.5 and R5.8 is needed, we propose that the response observed at R6 of higher biomass production was already achieved by R4, R5.5 and R5.8 and that this should be studied further.
• p-values for comparisons of WUE with the BMP versus the TMP were p < 0.05, p < 0.05 and p < 0.001 for studies 2, 3, and 4, respectively. P-values for comparisons of AE with the BMP versus the TMP were p < 0.10, p < 0.05 and p < 0.001 for studies 2, 3, and 4, respectively. We propose that this agronomic BMP with narrow rows provides a soil and water conservation advantage, potentially reducing the nitrogen losses to the environment and reducing the potential for water leaching below the root zone, since significantly higher total silage biomass production will most likely use more water and nitrogen.
• We propose that if farmers use this agronomic BMP of narrow rows with 42.5% higher biomass silage production and do not harvest the silage but instead just harvest the increased production of grain (on average, 9.5% higher) at harvest, the higher quantity of stalks and leaves left in the field with this new BMP will likely increase the potential for carbon sequestration. This should be studied further in long-term studies on narrow rows with tillage and no-till practices under sprinkler-irrigated systems. Additionally, since there was significant aboveground total production of biomass at R6, although not sampled, we infer that belowground root production was also higher. If this is the case, the BMP with narrow rows and higher planting densities that increased aboveground production may be a best management practice to increase carbon sequestration potential, aboveground biomass production, grain yield, and water and nitrogen use efficiencies, making it an effective practice for providing ecosystem services.

5. Conclusions

All three studies with the BMP of planting in narrow rows with a higher plant population increased silage production compared to the TMP of using wider row spacing with a lower plant population (146 days after planting [DAP]). The average grain production in these three BMP studies with narrow rows and higher plant populations was higher by 9.5% compared to the TMP of using wider row spacing with a lower plant population, with one study being significantly higher (p < 0.10), a second study with a p-value of p = 0.16, and a third study that had a higher average of 1.3% (not significant). Water use efficiencies and agronomic nitrogen use efficiencies for silage production were significantly higher in these three studies with the BMP of planting in narrow rows with higher plant populations.

These studies suggest that a BMP of using narrow rows with a higher plant population is a potential management practice to increase yields, economic returns for farmers growing silage and/or grain corn, and the efficiency of water and nitrogen inputs in sprinkler-irrigated systems of the western United States. These studies suggest that this planting method may have potential as a best management practice that could be considered by NRCS to improve soil and water conservation in sprinkler-irrigated corn systems. Additionally, the significantly higher total silage biomass production in the three studies with narrow rows and higher plant populations suggest that this will be an agronomic practice that could potentially contribute to carbon sequestration.