Effects of Different Cereal Rye Seeding Rates, Cotton Seeding Rates, Planter Type, and Working Speeds on No-Till Cotton

Abstract

No-till cotton producers are focused on saving resources by reducing planting rates, while maintaining yields. A 3-year field experiment was conducted in Shorter, Alabama, USA, to evaluate cotton planted into a roll/crimped cereal rye cover crop seeded at rates of 50 and 101 kg ha⁻¹. Cotton planter performance was also compared between a mechanical planter and an electronic planter at speeds of 5.6 and 11.2 km h⁻¹ along with low and high cotton planting rates (90,193 and 180,387 seeds ha⁻¹). Results indicate that cereal rye seeding rates did not affect its biomass. The emergence rate index (ERI) was influenced by the planter type at the 5.6 km h⁻¹ speed with a higher ERI (9.70% day⁻¹) for the mechanical planter compared to a lower ERI (9.05% day⁻¹) for the electronic planter. The cotton population was proportional to planting rates generating 66,650 and 114,178 plants ha⁻¹ at low and high rates. Standard deviation (STD) of cotton plant spacing had a lower STD for the electronic planter compared to the mechanical planter, but did not affect the cotton yield. The seed cotton yield was not dependent on the cover crop seeding rate, planter type, and speed, but differed among years. Drought in 2019 caused a yield reduction (1844 kg ha⁻¹) compared to higher yields of 3981 kg ha⁻¹ in 2018 and 4152 kg ha⁻¹ in 2020.

1. Introduction

Producers are under pressure from various sources to reduce environmental impacts. Many are adopting cover crops and no-till growing systems to meet these requests and maintain profit margins. Cotton producers in the Southern United States use cover crops to sustain cotton growth during periods of extreme weather, such as drought or intensive rainfalls that have detrimental effects on crops yields and soil health. Cover crop benefits include the reduction in soil erosion and water runoff, weed suppression, carbon sequestration, increased soil organic matter, and better soil water storage and infiltration [1,2,3,4]. However, added cost is associated with planting and maintaining cover crops. One way to decrease the production cost is to reduce the seeding rates for both cover crops and cotton. Another way is to reduce the time of cotton seed planting by increasing the operating speed, as planting at faster speeds can help producers get their crops planted within recommended optimum planting dates for a specific region [5,6]. In addition, using advanced planting technology for controlling seed planting uniformity plays an important role in modern conservation agriculture [7]. Cereal rye (Secale cereale L.) is one of the most popular cover crops that has been widely accepted in the Southern United States, including Alabama, because it generates large amounts of biomass, up to 7840 kg ha⁻¹ [8]. Research has also shown that rye does increase biomass production when supplemented with nutrients but did not increase biomass with higher seeding rates [9]. Another study resulted in more height and less tillering at higher seeding rates of cereal rye and the inverse for lower seeding rates of rye, with no differences in aboveground biomass production between any seeding rates [10]. This illustrates the potential in achieving adequate rye biomass at lower seeding rates, thus saving costs.

The cotton seeding rate can also influence cotton yield and profits to producers. According to [11], average cottonseed cost ranges from 15% to 20% of the total cotton production cost. Results from a field experiment [12] indicated that cotton yield was at a similar sustainable level when plant populations were between 74,000 and 111,000 plants ha⁻¹. Field research conducted by [13] comparing five different cotton seeding rates (from a very low rate of 8500 seeds ha⁻¹ to a high rate of 119,000 seeds ha⁻¹) stated that when cotton seed cost increased, farmers apply lower seeding rates to decrease seed cost. On the other hand, when the cotton lint price increases, producers tend to increase cotton planting rates.

Technological advancements in agricultural machinery, including planters, gives producers the ability to adjust on the fly and increase data availability throughout the growing season. Technology has also increased the accuracy of seed placement at faster workings speeds [6]. However, technology comes at a high cost with purchasing electronic equipment or retrofitting existing equipment along with reoccurring subscription costs for some services. Producers looking to invest in this technology need to know if they will obtain a return on investment and if planting performance will be improved compared to traditional mechanical planters. Research from [14] investigated differences in a traditional seed meter and an electronic seed meter. Results showed no significant difference between them in regard to seeding singulation over four operating speeds (6.1, 7.1, 8.2, and 9.5 km h⁻¹) when planting corn. It was concluded that seed meter performance was more influenced by the seeding rate than the operating speed in terms of percent skipped seeds and coefficient of variation [14]. An experiment was also conducted with soybean seeding rates using a high-speed planter under speeds ranging from 8–16 km h⁻¹ and seeding rates of 222,000 and 321,000 seeds ha⁻¹ [6]. Results showed that planting speed did not affect soybean emergence but some decreases in plant spacing uniformity were observed. Also, higher planting speeds showed no damaging effects on soybean yield. Research from [15] also investigated different planting speed of soybeans (7.9–15.5 kg h⁻¹) using both a mechanical drive and an electronically controlled planter and found that higher speeds showed increased plant spacing, variability, and population but no adverse effects on yield. Overall, planting at high speeds seems possible, but research is limited on experimenting with seeding rates, planting speeds, and planter types in a no-till cotton production system.

Therefore, the objectives of this study are to determine the effects of two seeding rates for a cereal rye cover crop on biomass production and cotton planting quality, to evaluate the performance of two planter types at two working speeds, and to assess the influence of two cotton seeding rates on planter performance and cotton yield.

2. Materials and Methods

2.1. Site Description

Field operations were initiated in fall of 2017 in central Alabama, at the Auburn University E.V. Smith Research and Extension Center, USA, in Shorter, Alabama (32.39° N, −85.92° W), by a planting cereal rye (var. Wrens Abruzzi) cover crop in mid-November of 2017, 2018, and 2019. Rye was planted at two rates of 50 kg ha⁻¹ and 101 kg ha⁻¹ on a Compass loamy sand (thermic Plinthic Paleudults) soil series. The rate of 101 kg ha⁻¹ was recommended by the USDA-ARS-NSDL [16]. Cereal rye was fertilized at rate 33.6 kg N ha⁻¹, as per recommendation for cover crop application rate in cotton production in Alabama [17]. Field activities for each of the three growing seasons are given in

Table 1. Field activities during the three growing seasons 2018–2020.

Field Activities	Growing Season
	2018	2019	2020
Cereal rye cover crop planted 1	15 November 2017	6 November 2018	6 November 2019
Applied nitrogen to cover crop	20 February 2018	8 February 2019	27 February 2020
Biomass and plant heights collected	12 April 2018	17 April 2019	17 April 2020
Rolled/crimped cover crops	19 April 2018	23 April 2019	21 April 2020
Cover crop burndown	20 April 2018	24 April 2019	21 April 2020
Planted cotton	11 May 2018	22 May 2019	14 May 2020
Cotton emergence collection started	18 May 2018	28 May 2019	21 May 2020
Cotton emergence collection completed	4 June 2018	13 June 2019	8 June 2020
Applied nitrogen to cotton	28 June 2018	3 July 2019	1 July 2020
Counted final cotton population	13 July 2018	25 July 2019	20 July 2020
Collected uniformity readings	18 July 2018	25 July 2019	23 July 2020
Harvesting cotton	6 October 2018	3 October 2019	22 October 2020

¹ Growing season starts by planting the cover crop in the fall of the previous calendar year.

.

2.2. Experimental Design

The experiment was a split-plot design with six main blocks randomly assigned to the whole experimental area. Three replicates of each rye rate were randomly assigned to the blocks (six main plots). Within each rye rate block, eight treatments were randomly assigned in a 2 × 2 × 2 factorial (eight subplots). Subplot treatment effects included two cotton seeding rates, two planter types, and two cotton planting rates as outlined below.

Treatment descriptions:

• Cereal Rye Seeding Rates (CovRT)
  • Low rate of 50 kg ha⁻¹
  • High rate of 101 kg ha⁻¹
• Two seed cotton planting rates (CotRT)
  • Low rate of 90,193 seeds ha⁻¹
  • High rate of 180,387 seeds ha⁻¹
• Two different planter types (PltType):
  • Mechanical
  • Electronic
• Two cotton planting speeds (PltSpeed)
  • 5.6 km h⁻¹
  • 11.2 km h⁻¹

Two commercially available planters were selected for comparison. (Figure 1). The first one was a mechanical planter with a wheel-driven John Deere MaxEmerge Plus (Deere and Company, Moline, IL, USA) with manually configured spring operated downforce (Figure 1a). The second one was an electronic planter John Deere MaxEmerge XP 1700 (Deere and Company, Moline, IL, USA) outfitted with Precision Planting (Precision Planting, Tremont, IL, USA) vSet meters and Delta-Force automated downforce control (Figure 1b).

Both planters included vacuum metering devices and were mounted via a 3-point hitch and equipped with Dawn model 1572 (Dawn Equipment Company, Sycamore, IL, USA) row cleaners with row-spacing of 0.91 m. Phytogen 330 was the cotton variety for all three seasons planted according to specified rates (Corteva Agrisciences, Indianapolis, IN, USA). Cotton was fertilized prior to flowering (sidedress) with 67.3 kg N ha⁻¹ liquid urea ammonium nitrate (UAN) for all plots as recommended by [18].

To determine each of the planter’s performance in terms of its uniformity of planting cotton seeds, the distance between emerged plants was measured in four random 1 m lengths of row, four locations per plot. From these measurements, the variability of plant spacing was determined by calculating the standard deviation (SD) using the procedure described in [19], and SD data were analyzed for each treatment. The lowest SD indicated best uniformity and the highest SD indicated poor uniformity.

Experimental design with treatment assignment (total of 48 subplots) is depicted in Figure 2. A randomization process was performed according to the procedure described by [20] for which eight random numbers with sequence from 1 to 8 and corresponding ranks from the highest (treatment 8) to the lowest (treatment 1) random number were used to assign the treatment to each plot (9.1 m long and 3.7 m wide) within each block.

Data collection for cereal rye biomass was performed by randomly tossing a wire frame with dimensions of 0.5 m × 0.5 m for each subplot. Biomass was then cut at ground level and placed in paper sacks to dry in a 55 °C oven for 72 h. Samples were then weighed, and results were expressed in kg ha⁻¹ of dry mass. Rye plant heights were collected six times per plot by placing a wooden measuring ruler at ground level and taking the height measurement to the top of the seed head.

During the three growing seasons, rolling/crimping was performed when cereal rye was at the milk growth stage (Zadoks #77) [21]. An experimental 3.7 m wide straight bar roller, developed at the USDA-ARS National Soil Dynamics Laboratory (NSDL) in Auburn, Alabama, was mounted on a John Deere 7730 tractor (Deere and Company, Moline, IL, USA) with a 77-kW diesel engine (on the tractor’s rear three-point hitch) and used for rolling treatments. After rolling/crimping, burndown herbicide (Roundup^TM Weather Max, Bayer CropScience, St. Louis, MO, USA) was applied separately using a John Deere 6700 self-propelled sprayer (Deere and Company, Moline, IL, USA) at a rate of 1.6 L ha⁻¹ (glyphosate active ingredient) diluted with water spraying 140 L ha⁻¹ as per manufacture recommendation.

Cotton emergence assessment began at first visible cotton emergence. Plants were then counted at two random locations per row along a 1.5 m long measuring pole on the middle two rows of each four-row plot (four readings per plot). Stand counts were collected two times per week until no new plant emergence was noticed, usually about 3 weeks after the first collection date.

To compare cotton plant emergence rates across cotton seed planting rates and different cover crops, the emergence rate index (ERI) in % per day was calculated using Equation (1) according to the procedure described by [22,23]

$$ERI=\sum _{n=first}^{last}{\displaystyle \frac{[\%n-\%\left(n-1\right)]}{n}}$$

(1)

where: %n is the percent plants emerged on day n, %(n − 1) is the percent plants emerged on day n − 1, n is the number of days after planting, first is the number of days after planting that the first plant emerged (first counting day), and last is the number of days after planting when emergence was considered complete (last counting day).

Higher ERI values are an indication of faster emergence of cotton plants and lower ERI values indicate a slower emergence of cotton plants. Final cotton populations were collected mid-growing season by using the average of four 1.5 m random lengths among the middle two rows of each plot, expressed in number of plants per hectare.

Cotton population (final stand) were collected each year at mid-growing season by averaging counted plants within four 1.5 m random lengths among the middle two rows of each plot and calculating the number of plants per hectare. Cotton was harvested in the first week of October in 2018 and 2019, and the fourth week of October in 2020 with a Case IH 2555 four-row cotton picker (Case IH, Racine, WI, USA) equipped with a weighing basket system. The weight of seed cotton from each plot was expressed in kg per hectare.

2.3. Statistical Analysis

Cover crops, cotton planting rates, and years were considered fixed effects; however replications and interactions of replications with covers were random effects [20]. Where differences in each year for dependent variables were significant, and when interactions between cover crops, cotton seeding rates, and years occurred, data were analyzed separately. Data were subjected to analysis of variance using the GLIMMIX procedure in SAS software v 9.4 [24], with a proper error term of rep and main plot used with split-plot design. Analysis for different cover crops/cotton seeding rates t-test grouping for the least-squares means was performed at alpha (α) = 0.1. Based on our objectives, only two-way interactions were analyzed and reported.

3. Results and Discussion

3.1. Plant Height and Biomass of Cereal Rye Cover Crop

Overall, the cover crop seeding rate had a significant effect on the rye height (p-value = 0.0280;

Table 2. ANOVA results for variables height and biomass with respect to seeding rate treatments of cereal rye cover crop.

Variable	Effect	Num DF	F-Value	p-Value
Height	CovRT	1	4.99	0.0280 *
	YEAR	2	0.67	0.5598
	YEAR*CovRT	2	2.41	0.0962
Biomass	CovRT	1	0.02	0.8752
	YEAR	2	1.17	0.3979
	YEAR*CovRT	2	0.57	0.5661

* Numbers in bold format show a statistically significant difference for variables and interactions.

) with an average height of 161.4 cm and 164.6 cm for the 50 and 101 kg ha⁻¹ seeding rates, respectively. Also, there were interactions between the cover crop seeding rates and the year (p-value = 0.0962). In contrast, no difference in cereal rye height occurred among the years (p-value = 0.5598).

Similarly, for cereal rye biomass, the growing season (year) and cover crop seeding rate did not affect the biomass amount of the cereal rye cover crop, generating a 3-year average of 5496 kg ha⁻¹ and 5526 kg ha⁻¹ for 50 and 101 kg ha⁻¹ seeding rates, respectively (

Table 3. Cereal rye cover crop production during the 2018–2020 growing seasons.

Year	Rye Height (cm)			Biomass (kg ha−1)
	Cover Crop Planting Rate (kg ha−1)
	50	101	p-Value	50	101	p-Value
2018	161.6	161.4	0.9343	5607	5884	0.3618
2019	159.1 b *	166.4 a	0.0038 ***	5589	5482	0.7238
2020	163.6	166.0	0.3301	5291	5213	0.7964
2018–2020	161.4 B **	164.6 A	0.0280	5496	5526	0.8752

* Similar lowercase letters in row (2019) indicate no statistical differences in the cereal rye plant height Tukey’s HSD at a 0.1 significance level. ** Similar uppercase letters in the last row (2018–2020) indicate no statistical differences in the cereal rye plant height between cover crop planting rates. *** Numbers in bold format show a statistically significant difference for variables.

). No interactions between the cover crop seeding rate and the year were present. These cereal rye biomass results agree with different field studies. According to [25], the cereal rye seeding rate did not generate proportional biomass, indicating that a higher cereal rye seeding rate does not generate higher biomass production. In another study, Boyd et al. [10] stated that at a low rye seeding rate, there was an increased generation of multiple stems (tillering), which compensated for overall plant growth and biomass production. Research conducted by [26] reported that no significant effect in biomass production was observed from increased seeding rates for cereal rye rates. They indicated that in higher plant population, because of competition, it can reduce the individual plant size, which can be translated into similar amounts of cereal rye biomass. Overall, biomass from the cereal rye cover crop was at a comparable production level obtained from a field experiment in Alabama [27], which reported that cereal rye cover crop production was ranging from 5832 kg ha⁻¹ and 6891 kg ha⁻¹ with a respective plant height between 157 cm and 159 cm.

3.2. Emergence Rate Index (ERI)

The year (growing season) for cotton emergence rate index (ERI) was significantly different among growing seasons (p-value < 0.0001;

Table 4. ANOVA table for the emergence rate index for cover crop rate (CovRT), cotton rate (CotRT), and Year for the 2018–2020 growing seasons.

Variable	Effect	Num DF	F-Value	p-Value
Emergence rate index	CovRT	1	17.70	0.0521 *
	CotRT	1	9.52	0.0027
	YEAR	2	169.29	<0.0001
	CovRT*YEAR	2	11.39	<0.0001
	CotRT*YEAR	2	1.87	0.1597
	PltType*YEAR	2	30.45	<0.0001
	PltSpeed*YEAR	2	2.99	0.0551
	PltType*PltSpeed	1	8.53	0.0044

* Numbers in bold format show a statistically significant difference for variables and interactions.

). Similarly, cotton ERI values were significantly influenced by both the cotton seeding rate and rye seeding rate with respective p-values of 0.0521 and 0.0277. Average ERI values were 8.7, 11.2, and 7.9 in 2018, 2019, and 2020, respectively. A higher ERI value of 9.7% day⁻¹ was reported for cotton in the 101 kg ha⁻¹ cereal rye planting rate treatments, whereas for the 50 kg ha⁻¹ rye rate, the cotton ERI value of 8.9 was lower. In addition, a higher ERI average value of 9.5% day⁻¹ was obtained for the low cotton seeding rate compared with a lower ERI of 9.1% day⁻¹ for the high cotton seeding rate.

Moreover, significant interactions occurred between the cover crop planting rate by year, cotton rate by year, planter type by year, and planter speed by year with p-values ranging from <0.0001 to 0.0425 (Table 4). Due to year and interactions of year and variables such as CovRT, PltType, and PltSpeed being significant, ERI data were again reanalyzed separately by each year. ANOVA results from each growing season are shown in

Table 5. Emergence rate index (ERI) in % day⁻¹ for cover crop rate (CovRT), cotton seeding rate (CotRT), planter type (PltType), and planter speed (PltSpeed) during the three growing seasons.

Effect		Emergence Rate Index (ERI) % per Day
		2018	2019	2020	2018–2020
CovRT	50 kg ha−1	8.03 b *	10.68 b	8.03 a	8.91 b
	101 kg ha−1	9.43 a	11.79 a	7.75 a	9.66 a
	p-value	0.0436 **	0.0569	0.3065	0.0521
CotRT	Low	8.85 a	11.38 a	8.34 a	9.52 a
	High	8.60 a	11.09 a	7.44 b	9.04 b
	p-value	0.4241	0.3141	0.0002	0.0027
PltType	Mechanical	9.37 a	11.63 a	7.14 b	9.37 a
	Electronic	8.09 b	10.84 b	8.63 a	9.19 a
	p-value	0.0002	0.0089	<0.0001	0.2178
PltSpeed	5.6 km h−1	8.90 a	11.50 a	7.72 a	9.37 a
	11.2 km h−1	8.55 a	10.97 b	8.06 a	9.19 a
	p-value	0.2540	0.0644	0.1160	0.2396
PltType*PltSpeed
Mechanical	5.6 km h−1	9.96 a	12.17 a	6.95 a	9.70 a
	11.2 km h−1	8.77 b	11.08 b	7.33 a	9.06 b
Electronic	5.6 km h−1	7.84 b	10.84 b	8.48 a	9.05 b
	11.2 km h−1	8.33 b	10.85 b	8.78 a	9.32 ab
p-value		0.0096	0.0578	0.8556	0.0044

* Similar lowercase letters in columns 3 (2018), 4 (2019), 5 (2020), and 6 (2018–2020) indicate no statistical differences using Tukeys HSD at a 0.1 significance level. ** Numbers in bold format show a statistically significant difference for variables and interactions.

. For significant interactions that occurred among dependent variables, only two-way interactions between years and other dependent variables (cover and cotton seeding rates, planter type and speed) were evaluated.

In 2018, cover crop seeding rates and planter type did affect cotton ERI. In contrast, cotton seeding rate and planter speed did not influence ERI. A lower ERI of 8.03 was associated with the 50 kg ha⁻¹ cereal rye seeding rate compared with a higher ERI of 9.43 at 101 kg ha⁻¹. Conversely, a higher ERI of 9.37 was obtained for the mechanical planter type compared to a lower ERI of 8.09% day⁻¹ for the electronic planter type. The average ERI value across the cotton seeding rate and planter speed was the same (8.73% day⁻¹).

In the 2019 growing season, the cover crop seeding rate, planter type, and planter speed affected the ERI, whereas the cotton planting rate did not influence the ERI, with an average ERI of 11.24% day⁻¹. At the 50 kg ha⁻¹ seeding rate for cereal rye, the ERI was lower compared to a higher ERI at the 101 kg ha⁻¹ cereal rye seeding rate. A higher ERI was obtained with the mechanical planter type compared to a lower ERI for the electronic planter type. Also, at the 5.6 km h⁻¹ planter’s speed, the ERI was higher when compared with a lower ERI at 11.2 km h⁻¹. Overall, ERI values in 2019 were similar to these reported by Kornecki [28] with ERI values between 10.1 and 11.0 for cotton planted into a cereal rye cover crop. These results indicate no restriction in cotton emergence from the rolled down and crimped cereal rye cover crop residue.

In 2020, the cotton seeding rate and planter type influenced the ERI; in contrast, the cover crop seeding rate and planter speed did not. A higher ERI was associated with the low cotton seeding rate compared to a lower ERI at the high cotton seeding rate. The ERI value for the mechanical type planter was lower compared to a higher ERI for the electronic planter type.

Significant interaction for ERI was detected between planter type and planter speed (PltType*PltSpeed; p-value = 0.0044). The mechanical planter type at 5.6 km h⁻¹ had a significantly higher ERI of 9.70% day⁻¹ compared to a lower ERI of 9.06 and 9.05% day⁻¹ for the mechanical type at 5.6 km h⁻¹ and the electronic type planter at 5.6 km h⁻¹, respectively. The ERI for the electronic planter type at a speed of 11.2 km h⁻¹ had 9.32%, but was no different than for the mechanical type planter at both speeds and the electronic planter type at 5.6 km h⁻¹.

3.3. Cotton Population

Across the three growing seasons, the cotton population varied between cotton seeding rates (CotRT; p-value < 0.0001), planter types (PltType; p-value = 0.0017), planter speeds (PltSpeed; p-value = 0.0115), cover crop seeding rate (CovRT; p-value = 0.0825), and among the growing seasons (p-value = 0.0208). There were significant interactions between year and cotton seeding rates (p-value < 0.0001) and between year and planter type (p-value = 0.0008) influencing the plant cotton population. ANOVA results with respect to cotton population is presented in

Table 6. ANOVA table for cotton population with respect to cover crop rate (CovRT), cotton seeding rate (CotRT), planter type (PltType), planter speed (PltSpeed) and year.

Variable	Effect	Num DF	F-Value	p-Value
Cotton Population	CovRT	1	10.64	0.0825 *
	CotRT	1	678.94	<0.0001
	PltType	1	10.46	0.0017
	PltSpeed	1	6.65	0.0115
	Year	2	4.04	0.0208
	CotRT*YEAR	2	20.05	<0.0001
	PltType*YEAR	2	7.69	0.0008
	CotRT*PltType	1	6.94	0.0099
	PltType*PltSpeed	1	4.48	0.0371

* Numbers in bold format show a statistically significant difference for variables and interactions.

. Because of these differences and interactions, data were reanalyzed separately by year with results presented in

Table 7. Cotton plant population (plants ha⁻¹) during the three growing seasonsfor cover crop rate (CovRT), cotton rate (CotRT), planter type (PltType), and planter speed (PltSpeed).

Effect		Cotton Plant Population (Plants ha−1)
		2018	2019	2020	2018–2020
CovRT	50 kg ha−1	88,210 a *	96,949 a	83,505 a	87,289 b
	101 kg ha−1	92,767 a	90,152 a	90,899 a	93,538 a
	p-value	0.3219	0.1314	0.1953	0.0825 **
CotRT	Low	65,728 b	63,264 b	70,957 b	66,650 b
	High	115,249 a	123,838 a	103,448 a	114,178 a
	p-value	<0.0001	<0.0001	<0.0001	<0.0001
PltType	Mechanical	97,099 a	97,696 a	85,298 a	93,364 a
	Electronic	83,878 b	89,406 b	89,107 a	87,464 b
	p-value	0.0007	0.0053	0.1714	0.0017
PltSpeed	5.6 km h−1	94,335 a	96,427 a	87,538 a	92,767 a
	11.2 km h−1	86,642 b	90,675 b	86,866 a	88,061 b
	p-value	0.0360	0.0450	0.8062	0.0115

* Similar lowercase letters in columns 3 (2018), 4 (2019), 5 (2020), and 6 (2018–2020) indicate no statistical differences using Tukeys HSD at a 0.1 significance level. ** Numbers in bold format show a statistically significant difference for variables.

.

Overall, a higher cotton population of 114,178 plants ha⁻¹ averaged over other treatments and growing seasons were associated with the high cotton seeding rate compared with 66,650 plants ha⁻¹ at the low cotton seeding rate. For planter type, a higher cotton population of 93,364 plants ha⁻¹ was obtained with the mechanical planter type compared to a lower cotton population of 87,464 plants ha⁻¹ for the new planter type. Likewise, for the cereal rye planting rate at 101 kg ha⁻¹, the cotton plant population was higher with 93,538 plants ha⁻¹ compared to the lower cotton population of 87,289 plants ha⁻¹ at the 50 kg ha⁻¹ seeding rate. Conversely, the higher cotton population of 92,767 plants ha⁻¹ was obtained for the speed of 5.6 km h⁻¹ compared to the lower cotton plant population of 88,061 plants ha⁻¹ at the higher planter’s working speed of 11.2 km h⁻¹.

In the 2018 growing season, cover crop seeding rates did not influence the cotton plant population. In contrast, the cotton seeding rate had a significant effect on the cotton population, generating a lower cotton population of 65,728 plants ha⁻¹ at a low cotton planting rate compared to a higher cotton population of 115,249 plants ha⁻¹ at the high cotton seeding rate. Similarly, the mechanical planter type generated a higher cotton population of 97,099 plants ha⁻¹ compared to a lower cotton population with the electronic type planter (83,878 plants ha⁻¹). Also, the planter speed affected the cotton population generating a higher cotton population at 5.6 km h⁻¹ compared to a lower cotton population at 11.2 km h⁻¹. Significant interaction between PltType*PltSpeed occurred having a higher cotton population with the mechanical planter type operating at 5.6 km h⁻¹ compared to a lower cotton population at 11.2 km h⁻¹ and the electronic planter type operating at both speeds (

Table 8. Significant interactions related to the cotton population between the planter type and the planter speed (PltType*PltSpeed) in 2018; between the planter type and the cotton seeding rate (PltType*CotRT) in 2019; and between the planter type and the rye cover crop planting rate (PltType*CovRT) in 2020.

2018			2019			2020
PltType*PltSpeed			PltType*CotRT			PltType*CovRT
Mechanical	5.6 km h−1	105,016 a *	Mechanical	Low	69,911 b	Mechanical	50	77,978 b
	11.2 km h−1	89,181 b		High	125,481 a		101	89,032 ab
Electronic	5.6 km h−1	83,654 b	Electronic	Low	56,616 c	Electronic	50	92,617 a
	11.2 km h−1	84,102 b		High	122,195 a		101	89,181 ab
p-value		0.0272 **	p-value		0.0786	p-value		0.0125

* Similar letters in columns 3 (2018), 6 (2019), and 9 (2020) indicate no statistical differences using Tukey’s HSD at a 0.1 significance level. ** Numbers in bold format show a statistically significant difference for variables and interactions.

). It appears that at the lower planter’s speed, there is better control for seeds discharge compared to the higher speed. This might be related to more skips due to not enough time for proper seed placing in the planting unit at a higher speed.

The same trend in cotton population continued in the 2019 growing season. The cover crop seeding rate did not influence the cotton population. Intuitively, the cotton seeding rate had a significant effect on the cotton population having a higher cotton population at a higher planting rate compared to a lower cotton population at the low cotton seeding rate. Significant interactions between the cotton seeding rate and planter type (PltType*CotRT) indicated that both planter types at the high seeding rate generated a higher cotton population than at the low cotton seeding rate for both planter types. Although, a 19% lower cotton population was indicated for the electronic planter at the low cotton seeding rate compared to the mechanical planter.

In 2020, only the cotton seeding rate influenced cotton population, whereas the cover crop seeding rate, planter type, and planter speed did not. A lower cotton population of 70,957 plants ha⁻¹ was associated with the low cotton seeding rate compared to a higher cotton population of 103,448 plants ha⁻¹ at the high cotton seeding rate.

Across growing seasons, significant interactions existed between the cotton seeding rate and planter type (Table 8; CotRT*PltType p-value = 0.0099). Specifically, at the high cotton seeding rate with both planter types, the cotton population was significantly higher (114,726 plants ha⁻¹ for the mechanical type and 113,630 plants ha⁻¹ for the electronic planter type) without difference between planters. In comparison, lower cotton populations at the low cotton seeding rate generated 72,002 plants ha⁻¹ for the mechanical type and a significantly lower population of 61,297 for the electronic planter type.

Similarly, a significant interaction was detected between the planter type and the planter speed (PltType*PltSpeed p-value = 0.0371). The mechanical planter type at 5.6 km h⁻¹ generated a significantly higher cotton population of 97,647 plants ha⁻¹ compared to a lower cotton population of 89,082 plants ha⁻¹ at 11.2 km h⁻¹. In addition, the electronic planter generated 87,887 plants ha⁻¹ and 87,040 plants ha⁻¹ at 5.6 and 11.2 km h⁻¹, respectively, without statistical differences among planter’s speeds.

3.4. Seed Cotton Yield

Seed cotton yield was significantly different among growing seasons (YEAR; p-value < 0.0001). In addition, across all growing seasons, significant difference in seed cotton yield were found between cotton planting rates (CotRT; p-value = 0.0001). Likewise, there were significant interactions between the growing season and the seed cotton planting rate (CotRT*YEAR) with p-value = 0.0045 (

Table 9. ANOVA table for seed cotton yield with respect to cover crop rate (CovRT), cotton seeding rate (CotRT), planter type (PltType), planter speed (PltSpeed) and Year.

Variable	Effect	Num DF	F-Value	p-Value
Seed cotton yield	CovRT	1	0.38	0.6007
	CotRT	1	16.35	0.0001 *
	PltType	1	0.95	0.3315
	PltSpeed	1	0.17	0.6853
	YEAR	2	575.00	<0.0001
	CotRT*YEAR	2	5.73	0.0045
	CovRT*PltType	1	3.36	0.0700

* Numbers in bold format show a statistically significant difference for variables and interactions.

). Furthermore, significant interaction occurred between the cover crop planting rate and the planter type (CovRT*PltType; p-value = 0.0700).

Because seed cotton yield was significantly different among growing seasons, yield data were reanalyzed separately by each year and results are shown in

Table 10. Seed cotton yield (kg ha⁻¹) during the three growing seasons for cover crop rate (CovRT), cotton rate (CotRT), planter type (PltType), and planter speed (PltSpeed).

Effect		Seed Cotton Yield kg ha−1
		2018	2019	2020	2018–2020
CovRT	50 kg ha−1	4032 a *	1857 a	4056 a	3307 a
	101 kg ha−1	3929 a	1832 a	4248 a	3345 a
	p-value	0.3093	0.4718	0.4053	0.2570
CotRT	Low	3860 b	1845 a	3896 b	3200 b
	High	4101 a	1844 a	4407 a	3451 a
	p-value	0.0037 ***	0.9503	0.0035	<0.0001
PltType	Mechanical	4015 a	1824 a	4047 a	3295 a
	Electronic	3946 a	1865 a	4256 a	3356 a
	p-value	0.3755	0.1684	0.2022	0.3315
PltSpeed	5.6 km h−1	4066 a	1830 a	4120 a	3338 a
	11.2 km h−1	3896 b	1895 a	4184 a	3313 a
	p-value	0.0341	0.3087	0.6897	0.6853
Overall average yield		3981 B **	1844 C	4152 A	p-value < 0.0001

* Similar lowercase letters in columns 3 (2018), 4 (2019), 5 (2020), and 6 (2018–2020) indicate no statistical differences using Tukeys HSD at a 0.1 significance level. ** Similar uppercase letters in rows indicate no statistical difference at a 0.1 significance level. *** Numbers in bold format show a statistically significant difference for variables and interactions.

. Overall, a higher seed cotton yield was generated in 2020 (4152 kg ha⁻¹), followed by a lower yield in 2018 (3981 kg ha⁻¹), and the lowest significantly reduced yield of 1844 kg ha⁻¹ was observed in 2019. In addition, across all growing seasons, significant differences in seed cotton yield were found between cotton planting rates generating 3449 kg ha⁻¹ at the higher cotton planting rate compared to 3202 kg ha⁻¹ for the low cotton planting rate.

In 2018, the cotton seeding rate and planter speed affected the seed cotton yield. A higher cotton yield was obtained with the high seeding rate (4101 kg ha⁻¹) compared to a lower yield (3860 kg ha⁻¹) at the low seeding rate. Conversely, a higher yield (4066 kg ha⁻¹) was generated at a lower speed of 5.6 km h⁻¹ compared to a lower yield (3896 kg ha⁻¹) at a planter’s speed of 11.2 km h⁻¹. The cover crop rate and planter type did not influence the cotton yield.

In 2019, none of the cover crop, cotton planting rates, nor planter type and planting speed had an impact on the seed cotton yield. The yield was significantly decreased and generated only 1844 kg ha⁻¹, which was 46% of the seed cotton yield obtained in 2018 and 44% in 2020. The main reason for this decline was a hot summer and a drought period in 2019 compared to two other seasons. Specifically, the rainfall amount from cotton planting to cotton harvest was only 204 mm, which was 43% of the total rainfall amount (477 mm) that occurred in 2018 and 33% of rainfall fallen in 2020 (628 mm) during the same growing period. Also, the maximum average ambient temp of 33.8 °C was higher than in 2018 and 2020 (

Table 11. Weather data with ambient temperatures and precipitation during the three growing seasons.

Activities	Growing Season
	2018			2019			2020
Weather data for specific periods with respect to the rye cover crop	Planting and termination dates for the cereal rye cover crop
	15 Nov 2017–19 Apr 2018			15 Nov 2018–23 Apr 2019			6 Nov 2019–22 Apr 2020
	Number of days: 156			Number of days: 158			Number of days: 168
	Rainfall (mm)	Temp °C		Rainfall (mm)	Temp °C		Rainfall (mm)	Temp °C
		Max	Min		Max	Min		Max	Min
From planting to rolling rye	424	17.9	5.0	661	18.7	6.4	918	19.6	6.9
Weather data for specific periods with respect to cotton planting and harvesting	Planting and harvesting dates of cotton
	11 May 2018–6 Oct 2018			22 May 2019–7 Oct 2019			14 May 2020–22 Oct 2020
	Number of days: 148			Number of days: 137			Number of days: 162
	Rainfall (mm)	Temp °C		Rainfall (mm)	Temp °C		Rainfall (mm)	Temp °C
		Max	Min		Max	Min		Max	Min
From planting to harvesting	477	31.9	21.4	204	33.8	20.8	628	30.4	19.9

). Physiological effects on cotton during drought stress include lower photosynthesis rates, decreased plant growth, fruit shed, early boll maturation, reduced lint quality, and lower yields [29,30].

In 2020, only the cotton seeding rate affected the cotton yield of 4407 kg ha⁻¹ at the high cotton seeding rate compared to a significantly lower yield of 3896 kg ha⁻¹ at the low cotton seeding rate. In contrast, the cover crop planting rate, planter type, and planter’s speed did not have any effect on cotton yield. Generally, a higher cotton yield in 2020 was associated with optimum rainfall amounts during cotton plant development (Table 11).

3.5. Cotton Seed Planting Uniformity

A lower standard deviation (SD) indicates a better seed spacing uniformity. Since the spacing between cotton plants is substantially different at low and high cotton seeding rates, standard deviation of spacing was analyzed separately by low and high cotton seeding rates with results presented in

Table 12. ANOVA table for standard deviation of spacing between cotton plants in row for cover crop rate (CovRT), planter type (PltType), planter speed (PltSpeed), and Year.

Variable	Effect	Num DF	F-Value	p-Value
STD of Planting Uniformity at LOW seeding rate	CovRT	1	0.75	0.4778
	PltType	1	3.50	0.0681 *
	PltSpeed	1	1.79	0.1882
	Year	2	1.19	0.3142
	Year*PltSpeed	2	2.70	0.0784
STD of Planting Uniformity at HIGH seeding rate	CovRT	1	3.46	0.2037
	PltType	1	6.51	0.0803
	PltSpeed	1	5.63	0.0269
	Year	2	2.14	0.1292

* Numbers in bold format show a statistically significant difference for the effect (variables and interactions).

. Th planter type showed significant influence at both rates with Year*PltSpeed (p-value = 0.0784) indicating a slight influence for the low rate. For the high speed, planter speed was significant at p-value = 0.0269.

Plant spacing standard deviation was not impacted by the cover crop seeding rate for either low or high cotton rates. For the low rate in 2018 (

Table 13. ANOVA table of standard deviation (SD) with respect to spacing between cotton plants at LOW cotton seeding rate for cover crop rate (CovRt), planter type (PltType), and planter speed (PltSpeed) at each growing season.

Low Cotton Rate		SD of Cotton Plant Spacing (cm)
		2018	2019	2020	2018–2020
CovRT	50 kg ha−1	9.79 a *	8.91 a	8.81 a	9.17 a
	101 kg ha−1	8.63 a	7.83 a	9.68 a	8.72 a
	p-value	0.2105	0.2446	0.3473	0.4778
PltType	Mechanical	10.28 a	8.24 a	9.78 a	9.43 a
	Electronic	8.14 b	8.50 a	8.71 a	8.45 b
	p-value	0.0231 ***	0.7697	0.2441	0.0681
PltSpeed	5.6 km h−1	8.10 b	8.04 a	9.63 a	8.59 a
	11.2 km h−1	10.32 a	8.70 a	8.86 a	9.29 a
	p-value	0.0187	0.4769	0.4035	0.1882
Average STD		9.21 A **	8.37 A	9.25 A	8.94

* Similar letters in columns 3 (2018), 4 (2019), 5 (2020), and 6 (2018–2020) for each treatment (i.e., cover crop planting rate: CovRT) indicate no statistical differences using Tukeys HSD at a 0.1 significance level. ** Similar uppercase letters in the last row indicate no significant difference at a 0.1 significance level. *** Numbers in bold format show a statistically significant difference for variables.

), the mechanical planter experienced increased standard deviations for plant spacing compared to the electronic planter (p-value = 0.0231). However, no other significance was detected between the two planter types for the 2019 or 2020 season. Across all three seasons, the mechanical planter performed with 11.6% higher standard deviation compared to the electronic planter at the low cotton seeding rate.

For the low cotton seeding rate including both planter types for 2018, the higher speed showed significantly higher variation with 27.4% more standard deviation compared to the slower speed. However, no other differences were noticed between speeds for 2019, 2020, or across all years.

For the planter type in the high cotton rate, no differences existed between planters for any of the seasons. However, pooled over all seasons, a higher SD for the mechanical planter was observed compared to the electronic planter by 10.8% (p-value = 0.0803). The high speed for the high cotton rate (

Table 14. ANOVA table of standard deviation (SD) with respect to spacing between cotton plants at HIGH cotton seeding rate for cover crop rate (CovRt), planter type (PltType), and planter speed (PltSpeed) at each growing season and across years.

High Cotton Rate		SD of Cotton Plant Spacing (cm)
		2018	2019	2020	2018–2020
CovRT	50 kg ha−1	6.32 a *	5.79 a	6.33 a	6.14 a
	101 kg ha−1	5.85 a	4.88 a	5.81 a	5.51 a
	p-value	0.4271	0.1293	0.3845	0.2037
PltType	Mechanical	6.56 a	5.37 a	6.46 a	6.13 a
	Electronic	5.60 a	5.30 a	5.67 a	5.53 b
	p-value	0.1091	0.9043	0.1856	0.0803 ***
PltSpeed	5.6 km h−1	5.68 a	5.10 a	5.54 b	5.44 b
	11.2 km h−1	6.48 a	5.57 a	6.60 a	6.22 a
	p-value	0.1793	0.4293	0.0782	0.0269
Average STD		6.08 A **	5.33 A	6.07 A	5.83

* Similar letters in columns 3 (2018), 4 (2019), 5 (2020), and 6 (2018–2020) for each treatment (i.e., cover crop planting rate: CovRT) indicate no statistical differences using Tukeys HSD at a 0.1 significance level. ** Similar uppercase letters in the last row indicate no significant difference at a 0.1 significance level. *** Numbers in bold format show a statistically significant difference for variables.

) noticed a 19.1% increase in standard deviation compared to the slower speed. Also, across all years, the higher speed showed a 10.5% increase in SD compared to the lower speed. These results are consisted with a study conducted by Virk [14] confirming that their coefficient of variation had a significant increasing trend with increased meter speeds while planting corn. These results suggest that the planter seed spacing variation was more at the higher cotton rate and higher speed compared to the lower rate and lower cotton rate. This is expected as the planter seeding plate must rotate faster and place more seeds under these conditions compared to the lower cotton seeding rate.

4. Conclusions

The cereal rye cover crop seeding rate did not have an adverse effect on the cotton population planted with either planter. The biomass production for the rye was similar across both seeding rates, indicating potential cost savings with the lower rate and no loss of effectiveness. The cotton seeding rate effected the yield in two of the three seasons, with the higher rate being significantly greater than the lower rate. The mechanical planter seemed to work as well as the electronic planter in terms of planting at higher seed populations while maintain similar cotton yields. However, the mechanical planter also produced higher ERI values in two out the three seasons, but overall, no difference in the ERI existed for three years between planters. The mechanical planter indicated similar SDs as the electronic planter in two of the three seasons for the low cotton rate and for each season at the high rate. In 2018, a yield reduction of only 4.2% was noticed at the higher planting speed, but no differences between planting speeds were observed in 2019 and 2020. Therefore, seeding at higher speeds is a feasible option without concerns of yield degradation. Also, seeding a rye cover crop and cotton at a lower rate using a less expensive mechanical cotton planter are the main factors of lowering cost. Such an arrangement will reduce seeds cost, equipment cost, and labor cost, while maintaining cotton yields. Based on results from this research and considering optimal weather and soil moisture conditions, saving resources without sacrificing seed cotton yield can be a way to ensure sustainable and profitable no-till cotton production.