Recurring Rolling/Crimping Effects on Termination Effectiveness of Iron Clay Pea and Pearl Millet Warm-Season Cover Crops

Abstract

Summer cover crop utilization by no-till vegetable farms is essential for continuous soil protection, especially in the southern United States where intense storms are likely to occur in hot and humid summer months. A field experiment was conducted at the National Soil Dynamics Laboratory in Auburn, AL, USA, between the summers of 2015 and 2017 to determine the effectiveness of an experimental roller/crimper in mechanically terminating summer cover crops. Iron clay peas (Vigna unguiculata, L.) planted on a sandy loam and pearl millet (Penninsetum glaucum, L.) planted on clay soil were selected to determine termination rate effectiveness in single, double, and triple rolling/crimping over the same area. Overall, termination rates for both cover crops were higher for rolling three times (71%) compared to rolling once (55%) or twice (63%). However, cover crop termination was inhibited due to rainfalls on the experimental area during the three-week evaluation period. In 2016, drought conditions and high temperatures (32.6 °C) caused biomass reduction, especially for pearl millet, of over 31% to 39% compared to 2017 and 2015. Rolling provided higher soil-water conservation compared with the non-rolled control due the cover crop mulch layer blocking sunlight, keeping the soil surface cooler and preventing water evaporation. Recurrent rolling did not cause soil compaction above the 2.0 MPa level that inhibits root growth, but changes in soil strength were dependent on the soil moisture content.

1. Introduction

Cover crops are a key component of conservation agriculture, and they must be managed appropriately to optimize their benefits, which include increased water infiltration and reduced soil erosion, runoff, and soil compaction [1,2]. Warm-season cover crops provide similar benefits as cool-season cover crops, but fast-growing summer cover crops that generate large amounts of biomass provide increased potential for weed suppression; otherwise, weeds can spread rapidly and compete with main crops [3]. Droughts are also more likely to occur in summer months due to increased temperatures and sporadic rainfall, especially in the southeastern US; therefore, it is important to select cover crop species that are drought tolerant and can produce optimum biomass. In addition, warm-season cover crops provide a significant contribution to biodiversity, creating ecological balance, helping to alleviate plant diseases and destructive insects. Beneficial insects, birds and amphibians feed on warm-season cover crops and reproduce during the summer growing season [3,4].

Rollers/crimpers have been used in conservation farming systems to mechanically terminate cover crops at an appropriate growth stage by flattening and crimping plant tissue at equal intervals with crimping bars mounted evenly around the roller’s cylindrical drum. After cover crops have been crimped, plant desiccation results in dry residue that forms a thick mulch that covers the soil surface. In agroecosystems where water for crop production is in short supply, flattened cover crops can be left on the soil surface and used as a mulch to conserve water by shading and cooling the soil surface. This reduces the evaporation of water from the soil surface. Mulching is a technique that involves covering the soil with a layer of organic material (cover crop residue). Mulching has several benefits including retaining moisture by preventing water evaporation, suppressing weeds by blocking sunlight and not allowing weed seeds to germinate, improving soil health by adding nutrients and organic matter to the soil, preventing soil erosion by reducing the impact of raindrops on the soil surface, and regulating soil temperature by keeping the soil cooler in hot weather conditions in Alabama. Benefits from this soil coverage include reduced soil erosion, reduced weed germination and growth, and soil water conservation for the following cash crop that can be planted directly into cover crop residue [5,6,7]. In Alabama, rolling/crimping should be performed at least 21 days before planting a cash crop into rolled residue since this period is needed to prevent the cover crop from competing for soil moisture and nutrients [6,8]. Results from a field study [6] indicated that due to accelerated rye senescence, rye termination rates above 90% were sufficient to plant a cash crop. To speed up the cover crop termination process, producers utilize herbicides to supplement rolling/crimping. However, in organic vegetable production, commercial herbicides cannot be used. Thus, terminating cover crops mechanically must be as efficient as chemical termination and might require multiple passes with roller/crimper over the same cover crop area to obtain the recommended cover crop termination rate above 90%.

Moreover, the farming community has been questioning if recurrent rolling/crimping over the same cover crop could cause soil compaction issues. Previous research has shown that multiple passing of agricultural equipment over the same area of a field causes soil compaction and contributes to significant soil degradation such as soil structure deformation by decreasing porosity and increasing bulk density [9]. Research conducted by [10] has shown that the soil quality deteriorated either with a greater number of passes by a lighter tractor or by a heavy tractor with fewer passes. According to [11], conventional tillage operations are known to have the most negative effect on causing soil compaction, which is associated with serious global environmental problems due to the degradation of the soil structure, which lowers water infiltration and hinders plant growth.

Field research that was conducted by [12] compared the long-term effects of conventional and conservation tillage practices on silty loam soil properties in Germany. These researchers concluded that the recurrent deep impact of tillage tools in conventional tillage plots caused the permanent destruction of freshly developed soil aggregates. Consequently, a relatively weak soil structure in the tilled soil layers proved unable to support an applied single-wheel 2.5-ton dynamic load. Conversely, in the conservation tillage plots, soil was able to adequately handle a 2.5-ton wheel load due to a strong aggregate system and did not lead to structural soil degradation. The effects of conventional, minimum, and zero-tillage on soil water, soil organic matter, and soil compaction were evaluated by the authors of [13]. They found that the zero-tillage treatment conserved much more soil water than conventional and minimum-tillage treatments, especially in dry years. Additionally, for conventional tillage, a soil compaction restrictive layer was detected in the subsoil compared to the topsoil. Field experiments conducted by [14] indicated that the soil strength limit for unrestricted root growth for cotton is 2.0 MPa, particularly for higher soil moisture content, where the difference in soil resistance for root penetration between compacted soil (with traffic) and uncompacted soil (no traffic) is low.

To address farm needs associated with managing summer cover crops in no-till systems, a field experiment was initiated at the NSDL in Auburn, Alabama, to evaluate the effectiveness of a two-stage experimental roller/crimper in terminating two warm-season cover crops. For this field experiment, the legume, iron clay pea (Vigna unguiculata, L.), and grass, pearl millet (Penninsetum glaucum, L.), were selected. Iron clay pea has been used in Alabama as a popular summer legume cover crop due to it having excellent soil erosion and weed control. It also provides good protection from soil compaction, and it is a heat- and drought-resistant plant during hot summer months but does not perform well in excessively wet soil [15,16]. In addition, this legume can produce 110 to 160 kg of nitrogen ha⁻¹ [17,18]. Pearl millet is a warm-season annual bunchgrass with a height ranging from 1.8 m to 2.1 m that scavenges nitrogen, protects from soil erosion, and suppresses weeds in the summer. This plant is well suited to grow on many different soils, including clay soils, as it can break up compacted soil and develops a deep root system that can survive under water shortage [19,20,21].

According to [22], pearl millet has the highest biomass growth potential compared with other types of millet. The objective of this study was to determine the effects of number of rolling passes (rolling once, twice and three times over the same area) with a two-stage roller/crimper on iron clay pea and pearl millet warm-season cover crop termination, volumetric soil moisture content (VMC), and soil compaction.

2. Materials and Methods

A three-year experiment, carried out over the 2015–2017 growing seasons, was initiated at the National Soil Dynamics Laboratory (NSDL), Auburn, Alabama USA, (latitude north: 32.6; longitude west: −85.5), which has a subtropic climate. The effectiveness of a patented two-stage roller/crimper prototype [23] (Figure 1) in terminating iron clay peas and pearl millet with recurrent passes over the same cover crop area was evaluated. The experimental layout was a randomized complete block design with treatments randomly assigned to each block with four replications, where rolling treatments were considered fixed effects and blocks (replications) were considered random effects.

Each experimental plot area was 465 m² (0.0465 ha) and both cover crops were broadcast in each growing season to the entire plot area with 5.9 kg of seed for iron clay pea and 3.2 kg of seed for pearl millet utilizing a chest spreader for all three seasons. Planting rates per hectare were 127 kg/ha for iron clay peas and 68 kg/ha for pearl millet. Then, seeds were tilled in (incorporated) and packed in with a steel mesh (basket) roller to provide a level surface and good seed-to-soil contact. A total of 16 plots with dimensions of 6.1 m long by 1.8 m wide were comprised in the field experiment.

Field testing of the two-stage roller/crimper was conducted on the soil bins at NSDL. The two-stage roller/crimper [23] comprises two drums, with the first being a smooth drum that flattens the cover crop on the soil surface, and the second drum, with equally spaced crimping bars on the drum’s circumference, crimps stems and flattens the cover crop. The crimping principle is based on applying vertical force generated by the weight of the drum to the cover crop against a firm soil surface. To increase the crimping effectiveness, the second drum assembly with crimping bars is also preloaded with two compression springs on each end of the drum. When the roller/crimper moves forward, springs are compressed and decompressed (due to advancing from one crimping bar to the next), and they release their kinetic energy as a downward force (perpendicular) to the laying down cover crop, causing injury (crushing tissue) to the cover crop at equal intervals.

Identical experimental designs were employed for each cover crop on two adjacent soil bins (specialized testing areas specific to the NSDL constrained by metal I-beams), Iron clay peas were planted on a sandy loam soil (Hiwassee sandy loam: a clayey kaolinitic thermic (oxidic) Typic Rhodudults with 73% sand, 11% silt, and 16% clay). The sandy loam soil was designated to grow iron clay pea cover crop as this summer legume grows best on sandy soils and full sunlight [15,17,18]. Pearl millet was planted on a clay soil (Davidson clay: a clayey kaolinitic thermic (oxidic) Rhodic Paleudults with 25% sand, 31% silt, and 44% clay). Clay soil was chosen to grow pearl millet as it can grow very well on many different soils including clay soils [20,21].

Field activities for each growing season are presented in

Table 1. Field activities for treatment application and data collections during the 2015–2017 summer growing seasons.

Field Activity	Summer Growing Season
	2015	2016	2017
Iron clay pea cover crop planted	3 June 2015	21 June 2016	9 June 2017
Biomass, height collection, and termination of iron clay pea cover crop and soil strength measurements	28 August 2015	20 September 2016	18 September 2017
Data collection: termination rates, VMC for iron clay peas: Week 1 after rolling	4 September 2015	28 September 2016	26 September 2017
Data collection: termination rates, VMC for iron clay peas: Week 2 after rolling	11 September 2015	4 October 2016	3 October 2017
Data collection: termination rates, VMC for iron clay peas: Week 3 after rolling	18 September 2015	11 October 2016	10 October 2017
Pearl millet cover crop planted	3 June 2015	21 June 2016	9 June 2017
Biomass, height collection and termination of Pearl millet cover crop and soil strength measurements	14 August 2015	8 September 2016	18 August 2017
Data collection: termination rates, VMC for pearl millet: Week 1 after rolling	21 August 2015	14 September 2016	23 August 2017
Data collection: termination rates, VMC for pearl millet: Week 2 after rolling	28 August 2015	22 September 2016	1 September 2017
Data collection: termination rates, VMC for pearl millet: Week 3 after rolling	4 September 2015	29 September 2016	7 September 2017

.

As depicted in Figure 1, the two-stage patented roller/crimper prototype [23] was a 3-point hitch mounted on the John Deere 4710 4-wheel drive tractor (34.5 kW).

Rolling/crimping treatments included rolling once, twice and three times operated at a speed of 5.0 km h⁻¹, completed on the same day. Cover crop termination rates were assessed at zero (day of rolling/crimping treatment application), one, two, and three weeks after rolling and were compared to untreated cover crops (control). To evaluate cover crop termination rates, a handheld chlorophyll meter (SPAD 502 light sensor-based, Konica-Minolta; Ramsey, NJ, USA) was used to measure chlorophyll activity or “greenness” of cover crops. Termination rates were evaluated on a scale of 0% (no plant injury symptoms: green plant) to 100% (complete plant death: zero greenness).

Three chlorophyll measurements were taken of the leaves in each plot. The first measurement was collected close to the centerline of the plot (lengthwise) approximately 0.5 m from the plot edge, the second was obtained about 2.0 m from the first one, and the third one was taken also 2.0 m from the second one (about 0.5 m from the end of the plot). The leaf selection was based on a representative sample chosen randomly in proximity to the plot centerline. The average value from three samples was reported for each plot. Termination rate measurements were taken at the same time going through the 16 plots, completing 48 measurements. All 48 measurements required approximately one hour to complete.

Quantitative values for plant condition at 100% termination rate were determined with 10 measurements of a randomly selected cover crop leaf (10 leaves) using the chlorophyll meter SPAD 502 along with visual observations confirming that the plant tissue was dead. Termination assessment using SPAD 502 started with 10 leaves of each cover crop when it was green (0% termination). Then, every day, as cover crops changed in appearance (color and moisture), these measurements were repeated. When no changes in the chlorophyll meter’s readings were observed, the termination rate was assumed to be 100%; Figure 2a,b show 0% (green) and 100% (dead) for iron clay peas, and Figure 3a,b show 0% (green) and 100% (dead) for pearl millet, respectively. For chlorophyll meter values expressing termination rates between 0% and 100%, a graphical representation of chlorophyll data was performed in Excel to establish the best-representing equation relationship for that graph.

Results indicate the best fit for chlorophyll meter data for both cover crops were linear relationships. Based on instrument readings and knowing full greenness (0% termination rate) and full desiccation (100% termination rate), to calculate termination rates for summer cover crops between 0% and 100%, a linear relationship between instrument readings and termination rates was established and used to obtain the plant termination rate (%):

Iron clay peas termination rate (%) = −1.70 × instrument reading * + 106.4

(1)

* Instrument reading range: 3.96–62.1

Pearl millet termination rate (%) = −1.84 × instrument reading ** + 105.0

(2)

** Instrument reading range: 2.90–56.7

For the cover crop termination assessment and for volumetric soil moisture content (VMC) measurement, three measurements per plot were completed. Having 4 replications per treatment, a total of 12 measurements for each treatment were obtained. Then, the average value from three measurements for termination rate and VMC were reported for each plot.

On the day of rolling, the cover crop plant height was measured at eight randomly selected locations within each plot, and these values were averaged for each plot area (6.1 m × 1.8 m). The biomass was collected from one representative location in each plot using a 0.25 m² area (0.5 × 0.5 m) steel wire frame. The cover crop biomass samples were oven-dried for 72 h at 55 °C using an electric oven (Model No. SC-350, Grieve Corporation, Round Lake, IL, USA). Termination rates were evaluated 0, 7, 14, and 21 (week 0, 1, 2, and 3) days after rolling (DAR). Soil volumetric moisture content (VMC) was measured 0, 7, 14, and 21 DAR using a portable time domain reflectometry (TDR) moisture meter with 12 cm-long rods (Spectrum Technologies, Plainfield, IL, USA).

Soil strength was measured with a FieldScout SC900 soil compaction meter (Spectrum Technologies, Aurora, IL, USA). This handheld meter was pushed into the ground and automatically collected soil strength measurements every 2.5 cm increment with a built-in ultrasonic depth sensor. In each plot, the soil strength was measured three times before and after rolling treatments for both the center of the plot and on the wheel driving path.

Analysis of variance (ANOVA) was performed on termination rates and VMC using SAS software release 9.2 [24]. Treatment means were separated by the Fisher’s protected LSD test at an α = 0.10 probability level. Where significant interactions between treatments and years occurred, or differences between years were significant, data were analyzed and presented separately; otherwise, data were combined.

3. Results and Discussion

3.1. Height and Biomass for Iron Clay Peas and Pearl Millet

Cover crop height and biomass data obtained at each growing season are presented in

Table 2. Cover crop height and biomass at three growing seasons.

Cover Crop	Parameters	Growing Season			p-Value	LSD	Average 2015–2017
		2015	2016	2017
Iron Clay Pea	Biomass (kg ha−1)	7916 a *	6427 b	7036 b	0.0064	746	7126
	Height (cm)	52.8 a	45.4 c	47.9 b	<0.0001	1.8235	48.7
Pearl Millet	Biomass (kg ha−1)	14,038 a	8497 b	12,230 a	0.0003	2111.2	11,588
	Height (cm)	190.7 b	176.6 c	222.2 a	<0.0001	6.1492	196.5

* Same lower-case letters in each row (by year) indicate no significant differences among means.

. There were significant differences in cover crop biomass production and height for both iron clay peas and pearl millet among three growing seasons (p-value < 0.0001). These differences are most likely associated with different weather conditions in each growing season. Results indicate that a proportional relationship does not always exist between the length of the plant and its weight. For example, research from [25] presents results from the evaluation of nine different pearl millet varieties in which plant height at maturity does not coincide with the highest biomass production. Similar results for [26] also show variation between sorghum height and dry matter yield. In fact, shorter plants can develop stems that have a larger diameter than the tallest plants. Weather data with rainfall amounts and temperatures from cover crop planting to rolling treatment application along with the three-week cover crop termination evaluation period are shown in

Table 3. Rainfall amounts and average maximum and minimum ambient temperatures from planting to rolling and during three-week cover crop termination evaluation period. Data were obtained in 2022 from MEDIUS, Auburn University [27], an on-site weather station for NSDL soil bins).

Weather Data for Specific Periods Concerning Iron Clay Pea Cover Crop Rolling/Crimping Date	Growing Season
	2015			2016			2017
	Rolling/Crimping Dates for Iron Clay Pea Cover Crop
	28 August 2015			20 September 2016			18 September 2017
	Rainfall (mm)	Temp °C		Rainfall (mm)	Temp °C		Rainfall (mm)	Temp °C
		Max	Min		Max	Min		Max	Min
From planting date to rolling/crimping date	347	31.4	21.1	205	34.1	21.9	457	30.6	21.1
Between rolling and 7 DAR	65	34.1	21.8	0	35.7	21.4	54.9	34.9	21.1
Between 7 days and 14 DAR	20	34.2	22.0	0	31.2	16.1	0	32.0	19.7
Between 14 days and 21 DAR	0	29.8	16.7	0	30.9	16.0	121.7	30.4	19.1
Weather Data for Specific Periods with Respect to Pearl Millet Cover Crop Rolling/Crimping Date	Rolling/Crimping Dates for Pearl Millet Cover Crop
	14 August 2015			8 September 2016			18 August 2017
	Rainfall (mm)	Temp °C		Rainfall (mm)	Temp °C		Rainfall (mm)	Temp °C
		Max	Min		Max	Min		Max	Min
From planting date to rolling/crimping date	299	31.9	21.4	203	34.2	21.8	310	30.1	21.1
Between rolling and 7 DAR	61.6	34.1	22.5	1.2	35.9	21.3	1.8	36.4	23.1
Between 7 days and 14 DAR	0	33.8	21.0	1.1	34.8	21.8	56.6	32.1	21.5
Between 14 days and 21 DAR	130.6	34.1	21.8	0	34.5	21.2	3.7	30.5	17.9

.

The highest biomass for iron clay peas was detected in 2015, generating 7916 kg ha⁻¹, followed by 7036 kg ha⁻¹ in 2017 and 6427 kg ha⁻¹ in 2016 with respective plant heights of 52.8 cm, 47.9 cm, and 45.4 cm. The biomass amounts for iron clay peas were related to rainfall amounts between planting and termination dates, which, in 2015, from planting to termination (86 days) received 347 mm of rainfall. In 2017, 457 mm of precipitation was received during the 101 days between planting and termination. The least amount of rainfall occurred in 2016 with 205 mm between planting and termination (91 days), but only 2 mm of this rainfall occurred 26 days before termination (average temperature of 28 °C), suppressing iron clay pea growth. The 2016 season not only experienced rainfall deficiency, but the average maximum temperature of 34.1 °C caused increased soil evaporation and plant evapotranspiration. Biomass production for pearl millet showed similar trends compared to the iron clay peas with the highest production reported in 2015 at 14,038 kg ha⁻¹ (height 190.7 cm), followed by 12,230 kg ha⁻¹ (height 222.2 cm) in 2017, and the lowest biomass of 8497 kg ha⁻¹ (height 176.7 cm) was obtained in 2016. These differences were associated with different weather and rainfall amounts. Specifically in 2015, the experimental area received 299 mm of rainfall during the 72-day period between planting to cover crop termination. In 2017, 310 mm of precipitation was received during the 70 days from planting pearl millet to its termination. In contrast, 2016 recorded lower precipitation of only 203 mm during 79 days between planting to termination, which slowed down pearl millet biomass production for this season.

This experiment was conducted under natural weather conditions in Alabama. The experimental area planted with cover crops was not irrigated, and cover crop growth was considered as dryland field testing conditions. Cover crop maturity is dependent solely on the weather conditions such as rainfall and temperature. In hot summer months, rainfalls can provide positive development of cover crops, but too much rainfall can be detrimental for plant development, inhibiting growth and causing the inability to reach the required maturity for mechanical termination within the same period during each growing season. Therefore, variation in the number of days from planting to reaching full maturity for mechanical termination does not affect mechanical termination using rolling/crimping technology.

3.2. Cover Crop Termination Rates

Initial data analysis of variance showed that there were significant differences in termination rates among years, weeks, and treatments (p-values < 0.0001) with significant interactions between treatments and weeks as well as treatments and years (p-values < 0.0001); therefore, cover crop termination data were again analyzed separately by year and week for assigned rolling treatments. Statistical results for ANOVA are presented in

Table 4. ANOVA results for cover crop termination rates.

Source	DF	ICP Termination		Pearl Millet Termination
		F-Value	Pr > F	F-Value	Pr > F
YEAR	2	34.86	<0.0001	10.32	<0.0001
TRT	3	469.89	<0.0001	855.49	<0.0001
WEEK	3	790.46	<0.0001	1020.87	<0.0001
REP	3	0.79	0.4993	0.60	0.6191
REP*TRT	9	1.07	0.3878	2.40	0.0150
WEEK*TRT	9	54.18	<0.0001	90.26	<0.0001
YEAR*TRT	6	7.27	<0.0001	6.40	<0.0001

.

3.2.1. Iron Clay Peas

ANOVA results (Table 4) indicate that there were significant differences in termination rates of iron clay peas for variable year, rolling treatment (TRT), and week with p-values < 0.0001. In addition, significant interactions between variables WEEK*TRT and YEAR*TRT occurred with p-values < 0.0001; therefore, data were analyzed separately for each week within each year. Termination results for iron clay peas for each growing season and each week of evaluation within the growing season with respect to rolling treatments are presented in

Table 5. Iron clay peas termination rates (%) by rolling treatment during 2015, 2016, and 2017 growing seasons.

Treatment	Growing Season
Number of Rolling Passes	2015				2016				2017
	DAR				DAR				DAR
	0	7	14	21	0	7	14	21	0	7	14	21
No-rolled	5.7	13.9 d *	21.8 c	22.8 c	15.7	33.6 d	30.6 c	62.8 c	9.3	13.8 d	11.0 d	33.6 d
One	8.3	73.0 c	53.4 b	73.8 b	17.8	52.6 c	67.8 b	77.7 b	9.4	55.7 c	63.3 c	59.0 c
Two	5.7	87.9 b	69.1 a	77.7 b	10.6	75.3 b	86.6 a	83.3 b	10.8	69.5 b	76.7 b	77.3 b
Three	8.9	99.5 a	77.8 a	86.3 a	19.4	93.5 a	95.3 a	95.2 a	11.9	95.2 a	89.1 a	90.9 a
p-value	0.4859	<0.0001	<0.0001	<0.0001	0.3485	<0.0001	<0.0001	0.0021	0.8497	<0.0001	<0.0001	<0.0001
LSD	N/S	5.3848	10.529	7.4372	N/S	5.5363	10.432	10.399	N/S	7.7832	7.8511	12.498

* Same lower-case letters in each column (each week) indicate no significant differences among means for each week and year separately (N/S = no significant differences in column).

.

• The 2015 Growing Season

In the 2015 growing season, day 0 (rolling treatment applications) average optimal plant health was reduced by 7.2% (7.2% termination). At 7 DAR, there were significant differences among rolling treatments with the lowest for no-rolled iron clay peas, then higher termination for rolling once (73%) and rolling twice (87.9%). The highest termination was observed for rolling three times, reaching a 99.5% termination rate. At 14 DAR, significant differences occurred between no rolling and rolling treatments, with rolling once at 53.4% and average rates of 73.5% for rolling once and twice, although no differences in termination between these treatments were present.

The main reason why termination rates of iron clay peas were lower at 14 DAR compared to 7 days was that two rainfall events occurred between the day of rolling and 7 DAR with a total rainfall amount of 65 mm. Rainfall events, totaling 20 mm, also occurred between 7 and 14 DAR that allowed for the iron clay peas plants initially injured by the roller’s crimping bars to recover, especially for these stems where crimping was ineffective due to variations in the soil surface. At 21 DAR, no difference in termination rates between rolling once and twice was observed (74–78%), and these rates were higher than for no-rolled cover crop (23.0%) but significantly lower than rolling three times rolling treatment (86.3%).

2.

The 2016 Growing Season

In 2016, at the time of rolling (week 0), the average optimal plant health was reduced by 15.9% (15.9% termination) across all plots. At 7 DAR, there were significant differences in termination rates with the lowest recorded for rolling once (52.6%), followed by a rate of 75.3% for rolling twice, and the highest rate of 93.5% was obtained for rolling three times; the untreated control had 33.6% termination. Similarly, at 14 DAR, significant differences occurred among rolling treatments with the lower rate for rolling once (67.8%) then a higher rate (86.6%) for rolling two times and the highest rate (95.3%) for rolling three times; the control was 30.6%. At 21 DAR, no differences were found between rolling once and rolling twice with respective values of 77.7% and 83.3%, compared to higher rates of 95.2% for rolling three times.

3.

The 2017 Growing Season

In 2017, at week 0, the average termination for iron clay peas measured 10.4% prior to rolling. At 7 DAR, there were significant differences among rolling treatments generating termination rates of 55.7%, 69.5%, and 95.2% for rolling once, twice and three times, respectively. The termination rate for the control was 13.8%. Similarly, at 14 and 21 DAR, there were significant differences among rolling treatments; however, the termination rates did not increase because of 121.7 mm of precipitation that occurred between 14 and 21 days, allowing plants to recover and continue growing, especially due to the vining nature of this cover crop. Iron clay peas have stems that are trailing or creeping along the ground, forming an interlocking thick mat on the soil surface. When used in a mixture, stems climb and interlock on stalks of other species (such as on such as millets, sorghum, and corn) and loop around different plants. Termination rates were 63.3% (14 days) and 59% (21 days) for rolling once and rolling twice, which measured 76.7% (14 days) and 77.3% (21 days). Rolling three times resulted in rates of 89.1 and 90.9% at 14 and 21 DAR, respectively.

3.2.2. Pearl Millet

Based on ANOVA results presented in Table 4, significant differences in termination rates for pearl millet were present for variable year, week, and rolling treatment (TRT) with p-values < 0.0001. Moreover, there were significant interactions between variables WEEK*TRT and YEAR*TRT with p-values < 0.0001. Because of these significant differences and interactions, data were analyzed separately for each week within each year separately. The pearl millet termination rate means for rolling treatments are presented in

Table 6. Pearl millet termination rates (%) by rolling treatment during 2015, 2016, and 2017 growing seasons.

Treatment	Growing Season
Number of Rolling Passes	2015				2016				2017
	DAR				DAR				DAR
	0	7	14	21	0	7	14	21	0	7	14	21
No-rolled	8.0	10.9 b *	15.3 b	18.7 b	17.5	18.8 d	20.5 d	20.9 c	6.7	12.3 d	13.9 c	20.7 c
One	7.2	76.6 a	90.4 a	93.4 a	16.7	60.0 c	63.9 c	79.4 b	12.4	70.9 c	76.3 b	74.4 b
Two	7.2	80.3 a	94.0 a	95.0 a	21.7	72.8 a	73.3 b	86.7 a	7.7	82.7 b	83.0 ab	80.8 ab
Three	8.6	84.0 a	97.7 a	98.2 a	21.6	75.6 a	84.4 a	90.9 a	6.8	89.9 a	89.7 a	86.7 a
p-value	0.8864	<0.0001	<0.0001	<0.0001	0.6673	<0.0001	<0.0001	<0.0001	0.8297	<0.0001	<0.0001	<0.0001
LSD	N/S	9.7262	7.782	7.7649	N/S	8.6335	7.5736	6.4978	N/S	4.4406	8.1306	6.5851

* Same lower-case letters in each column (each week) indicate no significant differences among means for each week and year separately (N/S = no significant differences in column).

.

• The 2015 Growing Season

In 2015, at the day of termination (day 0), there were no significant differences in termination rates among rolling treatments (average 7.7%) with the control averaging 8.0%. At 7 DAR, a significant difference was reported between the control (10.9%) and rolling treatments; however, there was no significant difference among rolling treatments generating 76.6%, 80.3% and 84% for rolling once, twice and three times, respectively. Similarly, at 14 DAR, the control had a significantly lower termination rate (15.3%) compared to higher rates for rolling once (90.4%), rolling twice (94.0%) and rolling three times (97.7%), although there was no significant difference among rolling treatments. At 21 DAR, a significantly lower termination rate was found for the control (18.7%) compared to rates for rolling once (93.4%), rolling twice (95.0%) and rolling three times (98.2%), without significant differences among rolling treatments. Overall, in 2015, the results indicate that rolling once was as effective as rolling twice or three times due to complete engagement of crimping bars with pearl millet stalks resulting in a successful mechanical termination.

2.

The 2016 Growing Season

In 2016, no significant difference was found in termination rates between the control (17.5%) and rolling treatments, averaging 19.4% at week 0. At 7 DAR, a lower termination rate for the control (18.8%) was observed compared to the higher rate for rolling once (60.0%). Compared to rolling once for this interval, significantly higher termination rates were reported for rolling twice (72.8%) and rolling three times (75.6%) without difference between rolling twice and three times. At 14 DAR, the lowest termination rate of 20.5% was reported for the control compared to a significantly higher rate of 63.9% for rolling once and 73.3% for rolling twice, and the highest rate of 84.4% was found for rolling three times. At 21 DAR, the control had the lowest rate of 20.9% compared to the significantly higher rate of 79.4% for rolling once, and the highest rates of 86.7% and 90.9% were reported for rolling twice and three times, respectively. The 2016 season shows the advantages of recurrent rolling, with termination rates of 9.2–14.5% greater for rolling two and three times compared to only rolling one time.

3.

The 2017 Growing Season

In 2017, at day 0, no difference in termination rate was found among the control and rolling treatments with an overall average termination of 8.4%. At 7 DAR, significant differences were measured among all rolling treatments with the control having termination of 12.3% and rolling once with 70.9%, followed by higher termination of 82.7% for rolling twice, and the highest termination of 89.9% was recorded for rolling three times. These results indicate that increasing the number of rolling passes over the same pearl millet cover crop area generates higher termination rates compared to rolling once or twice. At 14 DAR, the control had a termination rate of 13.9%, compared to higher rates of 76.3% for rolling once followed by a numerically higher rate of 83.0% for rolling twice and a higher termination of 89.7% for rolling three times, without significant difference between rolling twice and three times. A similar relationship among rolling treatments was recorded at 21 DAR with the lowest termination of 20.7% for the control; a significantly higher rate of 74.4% was found for the rolling once, 80.8% for rolling twice, and rolling three times generated termination rates of 86.7%, without a significant difference in termination rates for rolling twice. This slight numerical decrease in the termination rate between 14 and 21 DAR was associated with a rainfall amount of 121.7 mm that fell on the experimental area during that period and inhibited cover crop desiccation.

Overall, rolling/crimping three times over the same cover crop area resulted in higher termination rates for both cover crops compared to rolling once or twice. These results can be explained with the working principle of the two-stage roller/crimper. Since mechanical termination of the cover crop with the roller/crimper is based on crushing stems at equal intervals against firm soil, it restricts nutrients and water flow throughout the plant, causing accelerated death. With rolling/crimping three times, this process repeats itself three times and triples the number of injuries sustained by the cover crop without severing the whole plant. For effective crimping of the cover crop, the firmness of soil surface, which is solely dependent on the soil moisture, must be much greater compared to the softer green cover crop tissue. Based on previous research, the optimum volumetric soil moisture content to provide soil firmness for effective crimping is approximately between 8 and 9% [7]. Since the crimping bar’s downward force crimps flattened cover crops, this force dissipates on the flattened cover crop layer and does not damage the firm soil structure. In contrast, if the soil is wet, the soil firmness is substantially diminished, the soil structure becomes damaged (disrupted) [28], and crimping is ineffective (cover crop stems are imprinted into the soft soil by steel crimping bars). Results obtained in this experiment agree with previous research findings that assessed benefits from multiple rolling/crimping of cover crops over the same area of different cover crops. Field experiments with different rollers conducted by [29,30] at different locations with diverse cover crops showed that cover crop termination rates significantly increased with rolling twice or three times compared with a single pass over the same cover crop area. These findings are important for organic no-till systems with cover crops where commercial herbicides are not permitted to be used and only effective mechanical termination through rolling/crimping can be utilized.

3.3. Volumetric Soil Moisture Content

ANOVA results for each cover crop for soil volumetric moisture content with respect to growing seasons (year), rolling treatments (TRT), and evaluation period (week) and their corresponding probabilities (p-values) are presented in

Table 7. ANOVA results for volumetric soil moisture content (VMC).

Source	DF	ICP Volumetric Soil VMC		Pearl Millet Soil VMC
		F-Value	Pr > F	F-Value	Pr > F
YEAR	2	1410.18	<0.0001	435.05	<0.0001
TRT	3	5.20	0.0020	38.84	<0.0001
WEEK	3	27.22	<0.0001	37.17	<0.0001
REP	3	8.56	<0.0001	4.17	0.0074
REP*TRT	9	3.24	0.0014	1.52	0.1454
WEEK*TRT	9	2.39	0.0152	3.78	0.0003
YEAR*TRT	6	4.79	0.0002	4.09	0.0009

. Initial statistical analysis revealed that there were significant differences in soil VMC with respect to the variables YEAR, TRT, and WEEK. In addition, significant interactions occurred between YEAR*TRT and WEEK*TRT; therefore, ANOVA analysis was performed separately for variables YEAR by each WEEK.

3.3.1. Iron Clay Peas

VMC results for iron clay peas during the 2015–2017 growing seasons for each rolling treatment separately according to evaluation period (0, 7, 14, and 21 DAR) are presented in

Table 8. Volumetric soil moisture content (%) for rolling treatment means for iron clay pea during 2015, 2016, and 2017 growing seasons.

Treatment	Growing Season
Number of Rolling Passes	2015				2016				2017
	DAR				DAR				DAR
	0	7	14	21	0	7	14	21	0	7	14	21
No-rolled	7.6	10.3 b *	18.6 a	8.9	2.6	2.7	1.7 a	0.3	8.3	7.2 b	4.4 b	14.2
One	7.5	13.9 a	15.8 c	7.9	3.4	1.8	1.1 ab	0.3	8.6	10.6 a	6.4 a	15.2
Two	8.3	14.0 a	18.0 ab	9.0	3.3	1.3	1.3 ab	0.1	8.5	11.4 a	7.6 a	14.8
Three	7.5	13.5 a	17.0 bc	8.6	3.1	1.5	0.7 b	0.1	8.9	11.1 a	7.0 a	15.0
p-value	0.924	0.0387	0.0214	0.7203	0.8233	0.229	0.0768	0.7204	0.8354	0.0048	0.0324	0.5948
LSD	N/S	2.2007	1.3839	N/S	N/S	N/S	0.6255	N/S	N/S	1.6795	1.7011	N/S

* Same lower-case letters in each column (each week) indicate no significant differences among means for each week and year separately (N/S = no significant differences in column).

.

• The 2015 Growing Season

For iron clay peas in 2015, no significant differences in soil VMC were found on day 0 with an average VMC of 7.7%. In contrast, at 7 DAR, VMC was not different among rolling treatments and was significantly higher with an average of 13.8% compared to a lower VMC of 10.3% for non-rolled control, indicating water conservation due to rolling. At 14 DAR, significant differences in VMC were found among all treatments including control with VMC values ranging from 15.8% (rolling once) to 18.6% (control); however, these differences are not associated with the treatment effect, and are most likely due to the rainfall event and differences in rainfall intensity on the experimental area. At 21 DAR, there were no significant differences among all treatments and control with an average soil VMC of 8.6%.

2.

The 2016 Growing Season

In 2016, on days 0 and 7 after rolling, no significant differences in VMC were reported for all treatments with corresponding VMC averages of 3.1% and 1.8%, respectively. These unusually low VMC values were related to prolonged periods of no rainfall that extended to the 21-day period of VMC reading. At 14 DAR, VMC values continued to decrease with averages ranging from 0.7% to 1.7%. At 21 DAR, VMC values were the lowest with average VMC values of only 0.2%, due to severe continuous drought conditions.

3.

The 2017 Growing Season

In 2017, on the day of rolling, the average VMC was 8.6% without differences among treatments. At 7 and 14 DAR, VMC for all rolling treatments was significantly higher compared to the control, but without differences between rolling once, twice or three times. The average VMC at 7 DAR was 11.5%, whereas the control had 7.2%. Likewise, at 14 DAR, an average VMC for rolling treatments was 7.0%, compared to a lower 4.4% for the control. Based on these results, when the cover crop is rolled down and crimped against the soil surface, water is conserved by preventing evaporation from the soil. In contrast, for non-rolled residue, VMC is lower because of plant evapotranspiration and increased soil evaporation as there is no adequate soil coverage by the cover crop.

Overall, during the 2015–2017 growing seasons, results from this field experiment for both cover crops agreed with other research findings confirming that rolling (flattening) the cover crop against the soil surface is beneficial for conserving soil water in no-till conservation systems with cover crops [7,31].

3.3.2. Pearl Millet

The results of pearl millet VMC according to year and week for each rolling treatment during three growing seasons are presented in

Table 9. Volumetric soil moisture content (%) for rolling treatment means for pearl millet during 2015, 2016, and 2017 growing seasons.

Treatment	Growing Season
Number of Rolling Passes	2015				2016				2017
	DAR				DAR				DAR
	0	7	14	21	0	7	14	21	0	7	14	21
No-rolled	11.9	15.7 b *	9.7 b	12.4 b	7.9 b	7.5 b	7.3	6.4 c	11.9	9.2 c	15.0 c	9.6 b
One	12.1	18.2 ab	14.3 a	15.2 a	8.3 b	9.2 a	8.2	7.7 a	12.0	12.6 b	16.7 bc	13.0 a
Two	10.9	19.9 a	14.4 a	14.7 a	9.1 a	9.0 a	7.8	7.0 b	12.3	13.5 ab	18.7 ab	14.8 a
Three	11.9	20.0 a	15.3 a	15.2 a	9.5 a	9.7 a	7.2	7.3 ab	11.3	14.8 a	20.8 a	14.6 a
p-value	0.1832	0.0522	0.0023	0.0602	0.0094	0.0354	0.1872	0.0165	0.6900	0.0057	0.0110	0.0077
LSD	N/S	2.6923	1.9651	1.8112	0.7194	1.1781	N/S	0.5807	N/S	2.153	2.519	2.265

* Same lower-case letters in each column (each week) indicate no significant differences among means for each week and year separately (N/S = no significant differences in column).

.

• The 2015 Growing Season

In 2015, at rolling, the VMC for pearl millet was not different across rolling treatments, including the control, with an average VMC value of 11.7%. At 7 DAR, no differences in VMC were detected for the control (15.7%) and rolling once treatment (18.2%). However, the two and three rolling treatments each had a higher VMC of 20% without differences among all rolling treatments. At 14 and 21 DAR, the VMC for all rolling treatments was significantly higher compared to the control. At 14 DAR, the average VMC for rolling once, twice and three times was 14.7%, compared to 9.7% for the control. However, 21 DAR VMC for rolling treatments was 15% compared to lower (12.4%) for the control.

2.

The 2016 Growing Season

In 2016, upon the application of rolling treatments, lower VMC was measured for the control (7.9%) and for rolling once (8.3%) compared to higher VMC for rolling twice (9.1%) and three times (9.5%). Despite these significant differences (p-value = 0.0094), differences in VMC values were most likely related to variability among replications (p-value = 0.0666), indicating some differences in soil properties or rainfall intensity in the experimental area among replications. At 7 DAR, the VMC for all rolling treatments was higher, averaging 9.3% compared to a lower VMC of 7.9% for the control, indicating soil water conservation. At 14 DAR, there was no difference in VMC among rolling treatments and the control averaging 7.6%. At 21 DAR, higher VMC was measured for all rolling treatments ranging from 7.0% to 7.7%, compared to a lower VMC of 6.4% for the control.

3.

The 2017 Growing Season

In the 2017 growing season, at rolling, there was no difference in VMC among rolling treatments and between the control averaging 11.9%. At 7 DAR, a higher VMC was measured for rolling treatments ranging from 12.6% to 14.8% compared with a lower VMC of 9.2% for the control. Similarly, at 14 DAR, VMC for rolling treatments was higher, ranging from 16.7% to 20.8%, compared to a lower VMC of 15.0% for the control. Again, at 21 DAR, a higher average VMC was reported for all rolling treatments, without differences among rolling once, twice and three times, averaging 14.1%, compared to a lower VMC of 9.6% for the control. Overall, changes in soil VMC were associated with rainfall events in the 3-week evaluation period. However, a higher VMC was always observed with rolling treatments compared to the control.

These results across three growing seasons agreed with other research findings confirming that rolling (flattening) a cover crop against the soil surface is beneficial with respect to conserving soil water through a mulching barrier, which reduces water evaporation from soil [31].

3.4. Soil Strength

ANOVA results (

Table 10. ANOVA results for soil strength during 2 years of iron clay pea and 3 years of pearl millet growth.

Source	ICP Soil Strength			Pearl Millet Soil Strength
	DF	F-Value	Pr > F	DF	F-Value	Pr > F
YEAR	1	214.94	<0.0001	2	357.89	<0.0001
TRT	2	1.28	0.2835	2	1.66	0.1944
POS	1	0.03	0.8575	1	5.17	0.0245
DEPTH	1	19.53	<0.0001	1	7.19	0.0082
REP	3	1.83	0.1470	3	1.56	0.2029
REP*TRT	6	3.19	0.0072	6	0.58	0.7470
DEPTH*YEAR	1	7.92	0.0061	2	13.45	<0.0001
YEAR*TRT	2	0.85	0.4319	4	1.84	0.1242

) indicate that significant differences in soil strength for iron clay peas occurred between variable YEAR and DEPTH with p-values < 0.0001. In addition, significant interactions between DEPTH*YEAR occurred with p-value < 0.0001.

3.4.1. Iron Clay Peas

Results from two growing seasons for iron clay peas indicate that there were significant differences in soil strength between years (p-value < 0.0001) and depths (15 and 30 cm; p-value < 0.0001), with significant interactions between depth and year (p-value = 0.0061). In addition, across two growing seasons, no difference was reported among rolling treatments (p-value = 0.2835) with soil strength ranging from 1.29 MPa to 1.45 MPa. Soil strengths in reference to position (p-value = 0.8575) showed CI values of 1.38 MPa, 1.36 MPa, and 1.29 MPa for wheel traffic, center of plot and control, respectively. These soil strength values were below the restrictive root penetration limit of 2.0 MPa established for cotton roots by [14]. Because of significant differences in soil strength detected between years, analysis of variance was performed again separately by year and depth with results presented in

Table 11. Soil strength (MPa) of respective rolling treatments on iron clay pea cover crop during 2015 and 2017 growing seasons.

Treatment	Growing Season
Position	2015		2016		2017
	Depth		Depth		Depth
	0–15 cm	15–30 cm	0–15 cm	15–30 cm	0–15 cm	15–30 cm
Control (No-rolled)	1.47 b	2.06	No data available for iron clay pea plots, due to an excessive soil hardness due to drought conditions.		0.78 a	0.85
Wheel traffic	1.57 ab	2.16			0.86 a	0.91
Center of plot	1.82 a	2.16			0.64 b	0.82
p-value	0.0467	0.9965			0.0009	0.1432
LSD	0.264	N/S			0.108	N/S

Same lower-case letters in each column (each year and depth) indicate no significant differences among means for each position, depth, year separately (N/S = no significant differences in column).

. Due to a prolonged drought, no data were available for the 2016 growing season as sandy loam soil was too hard to penetrate the soil with a hand-held penetrometer.

In 2015, at the 0–15 cm depth, soil strength for the control was significantly lower (1.47 MPa) compared to the center of plot (1.82 MPa), but no different than wheel traffic (1.57 MPa), but these soil strength levels were below a root restriction level of 2.00 MPa [14]. Although statistically significant, soil strength differences might be related to differences in soil properties and soil moisture. At a 15–30 cm depth, there was no significant difference among wheel traffic, center of plot, and the control, with an average soil strength of 2.13 MPa. In 2017, the 0–15 cm depth was significantly higher for the control (0.78 MPa) and the wheel traffic (0.91 MP) and lower (0.64 MPa) for the center of the plot. These results are directly related to higher soil moisture content caused by rainfall events and supported by previous research findings [28,30]. No significant differences were reported among any positions and the control at 15–30 cm, with the soil strength ranging from 0.82 MPa to 0.91 MPa.

3.4.2. Pearl Millet

Preliminary ANOVA results for pearl millet showed significant differences in soil strength among three growing seasons with a p-value < 0.0001. Similarly, differences in soil strength existed between positions measured at the center of the plot and in the wheel track, as well as between depths, with respective p-values of 0.0245 and 0.0082. Moreover, there was a significant interaction between YEAR*DEPTH with p-value < 0.0001 (Table 10), although no significant interactions between treatments and year existed (p-value = 0.1242). Because significant difference in soil strength occurred among years and depths, statistical analysis was performed again separately by year and depth, the results of which are shown in

Table 12. Soil strength (MPa) with respect to positions (wheel traffic, center of plot, untreated control) and the respective rolling treatments (rolling once, twice and three-times) on Pearl Millet cover crop during 2015, 2016 and 2017 growing seasons.

Treatment	Growing Season
Position and Rolling Treatment	2015		2016		2017
	Depth		Depth		Depth
	0–15 cm	15–30 cm	0–15 cm	15–30 cm	0–15 cm	15–30 cm
Control	0.80 c *	1.09	1.83	2.00	0.72 b	0.69
Wheel traffic	1.19 a	1.25	2.16	1.84	0.84 a	0.63
Center of plot	0.99 b	1.19	2.16	1.74	0.70 b	0.58
p-value	0.0086	0.5716	0.9725	0.2979	0.0002	0.1497
LSD	142.47	N/S	N/S	N/S	0.056	N/S
Control	0.80 c	1.09	1.83	2.00	0.72 b	0.69 a
Rolled once	0.98 b	1.16	2.11	1.87	0.69 b	0.55 c
Rolled twice	1.04 b	1.20	2.18	1.85	0.81 a	0.60 bc
Rolled three times	1.25 a	1.31	2.19	1.64	0.83 a	0.66 ab
p-value	0.0127	0.4621	0.4553	0.1070	0.0029	0.0444
LSD	0.151	N/S	N/S	N/S	0.059	0.074

* Same lower-case letters in each column (each year and depth) indicate no significant differences among means for each position and rolling treatment according todepth and year separately (N/S = no significant differences in column).

.

In 2015, at 0–15 cm, significant differences in soil strength were detected among positions, with 0.80 MPa for the control, which was higher for the center of plot (0.99 MPa) and highest (1.19 MPa) for wheel traffic. Moreover, no difference in soil strength was detected between rolling once (0.98 MPa and twice (1.04 MPa) compared to higher soil strength (1.25 MPa) for rolling three times. At a 15–30 cm depth, no significant difference in soil strength occurred among positions, averaging 1.18 MPa, and among rolling treatments, also averaging 1.19 MPa. In 2016, at a depth of 0–15 cm, no significant difference in soil strength was observed among positions, and among rolling treatments. Similarly at a depth of 15–30 cm, no differences in soil strength were indicated among positions and rolling treatments. In 2017, at a depth of 0–15 cm, lower soil strength values were obtained for the control and center of plot (0.72 MPa) compared to higher values for wheel traffic (0.84 MPa). For rolling treatments, lower soil strength was reported for the control and rolling once (0.69 MPa) compared to rolling twice (0.81 MPa and three times (0.83 MPa). Despite these differences, the soil strength was below a restrictive root penetration resistance of 2.00 MPa. Similarly, these lower soil strength values were observed at a 15–30 cm depth, without differences in soil strength among positions, averaging 0.6 MPa. In contrast, statistical differences in soil strength were found among rolling treatments; however, the control (0.69 MPa) and rolling three times (0.66 MPa) had higher soil strength compared to lower soil strength of 0.55 MPa and 0.60 MPa for rolling once and twice, respectively. Overall, the soil strength was not affected by rolling treatments but depended on changes in soil moisture [28,29,30]. It also should be pointed out that statistical (significant) differences in soil strength were very small (0.10 MPa) and were most likely related to differences in soil moisture content rather than being influenced by rolling treatments.

4. Conclusions

Summer cover crops are an important component to effectively cover the soil surface to prevent soil erosion and runoff associated with high-intensity rainfall events. Results have shown that rolling twice or three times over the same cover crop area increased cover crop termination rates compared to rolling once. For iron clay peas, the termination rate averaged over the years and weeks was 51% for rolling once, 61% for rolling twice, and 72% for rolling three times. Similarly, the overall termination rates for pearl millet were 60%, 65% and 70% for rolling once, twice and three times, respectively. Compared with the control (non-rolled cover crops), rolling down and crimping resulted in increased soil volumetric soil moisture (VMC), indicating improved soil-water conservation for both cover crops due to rolling. During the 21-day evaluation period in the summer of 2016, a lack of rainfall and high average max temperatures (32.6 °C) created drought conditions that significantly reduced the volumetric soil moisture content on sandy loam soil planted with iron clay peas. This unusually dry period in 2016 also affected the biomass production of cover crops, especially pearl millet, generating significantly less biomass compared to 2017 (3733 kg ha⁻¹ (31%)) and 2015 (5540 kg ha⁻¹ (39%)). The soil strength for sandy loam planted with iron clay peas was assessed in 2015 and in 2017. No measurements were performed in 2016 due to severe drought preventing the hand compaction meter from moving down through the soil profile. The soil strength was not different for rolling treatments and among wheel traffic, center of plot and control at a 15–30 cm depth. At a 0–15 cm depth, the soil strength was below the 2.0 MPa threshold for root penetration resistance. Similarly, the soil strength for clay soil planted with pearl millet in 2015 and 2017 was lower than the 2.0 MPa threshold. In contrast, in 2016, the soil strength between positions and between rolling treatments was not different; however, the soil strength slightly exceeded 2.0 MPa for these parameters at a 0–15 cm depth due to unusually low soil volumetric moisture content related to drought conditions. Based on results from this experiment, summer cover crop termination rates vastly increased with rolling/crimping twice or three times over the same cover crop area compared with a single pass. Such findings are important for organic systems with cover crops, where only mechanical termination such as rolling/crimping can be utilized, as terminating cover crops with commercial herbicides is prohibited. Another important benefit from flattening and crimping cover crop against soil surface is conserving soil water through a mulching barrier, which resulted in higher volumetric soil moisture due to decreased soil water evaporation. Notably, rolling/crimping twice or three times over the same cover crop area did not cause soil compaction. Changes in soil strength were not affected by rolling treatments but depended on changes in soil moisture.