Evaluation of Different Weight Configurations and Pass Numbers of a Roller Crimper for Terminating a Cover Crop Mixture in the Vineyard

Abstract

Viticulture, a key economic activity in the Mediterranean area, is facing several challenges including soil degradation. Among the sustainable practices available, the management of cover crops in vineyard inter-rows using a roller crimper to create dead mulch is gaining pace as an effective strategy for soil conservation. Nevertheless, the effectiveness of roller crimpers in terminating groundcovers in vineyards may be reduced by pedoclimatic conditions, type of vegetation and roller crimper configuration and operational parameters. This study aimed to evaluate the effectiveness of a roller crimper with two different weight configurations, light (LR) and ballasted (HR), each tested with one (P1) or two passes (P2), in terminating a cover crop mixture in a vineyard. To evaluate the termination performance, plant green cover data were modeled using a one phase exponential decay nonlinear regression. The four systems were also assessed for their ability to conserve soil moisture and their impact on soil compaction. Although the HR + P2 showed the highest termination performance, the system using the HR + P1 obtained comparable results, with k values of 0.07 and 0.11 days⁻¹ and half-life values of 9.50 and 6.09 days in 2023 and 2024, respectively. Given the need to coordinate multiple vineyard operations within short and weather-dependent timeframes, a one-pass approach such as HR + P1 offers operational advantages, providing a practical compromise between efficacy and efficiency.

1. Introduction

Viticulture represents a key economic activity in the Mediterranean area where vineyards are increasingly recognized as integral components of the agricultural landscape, capable of providing multiple environmental, economic and social benefits [1]. However, several challenges still need to be addressed, including soil degradation [2]. Intensive tillage for inter-row and under-row management is among the main drivers of this issue, as it adversely affects soil structure and stability, accelerates organic matter loss, and increases the susceptibility to erosion [3]. In this context, the adoption of conservation agriculture practices represents a valuable opportunity [4]. Among these, soil cover through spontaneous vegetation or cover crop sowing, combined with reduced soil disturbance, as is the case with no-till or minimum tillage, can enhance soil health and fertility in the vineyard [5]. Indeed, these practices can increase organic matter content, improve soil structure through greater aggregation, and mitigate erosion [6,7]. They can also promote microbial activity, increase nutrient availability, conserve soil moisture, and foster biodiversity [8,9]. Reducing tillage intensity also provides benefits in terms of lower energy consumption and reduced carbon dioxide emissions [10]. For the inter-row area, a commonly adopted practice consists in alternating row sowing of cover crops, thus some rows are sown, while others are left with spontaneous vegetation managed by mowing to allow for mechanical operations during the growing season (e.g., crop protection product application). Frequently, the sown cover crops are incorporated in the soil as green manures at the beginning of the vegetative season of the vines [11]. An interesting alternative is managing sown cover crops with a roller crimper, which creates a surface layer of dead mulch [12]. This occurs without tilling the soil, thus making it a more conservation-oriented practice than green manure. Dead mulch, in addition to supplying organic matter and conserving soil moisture, it is particularly effective in weed control as the thick layer of plant residues hinders weed germination and development [13], representing an useful tool in vineyard conducted based on organic agriculture principles.

Roller crimper is an implement consisting of a cylindrical roller with blades mounted on its outer surface in various shapes and arrangements. As it passes through the field, the blades crush the cover crop stems without cutting them [14,15]. This crimping causes injuries that promote plant desiccation [16]. Several models of roller crimpers are available on the market for vineyard applications. These are generally rear- or front-mounted implements with different frame types, such as fixed, hydraulically side-shifting, or telescopic frames with multiple modules to adjust working width [16,17,18]. Cutter arrangements can be tangential and staggered, as in the widely used arrangement, or helical [16,19]. Cutter profiles are usually chisel-shaped or flat. Roller crimpers produce larger residue fragments even distributed on the field, resulting in more persistent mulch compared to mulching mowers, which, instead, produce finer residues [20,21]. Optimal termination of the cover crop is crucial to avoid water and nutrient competition with vines and to ensure efficient agronomic management [13]. Studies have shown that roller crimper intervention is more effective when the cover crops reach certain phenological stages [22]. For example, according to Miville and Leroux [23] the optimal stage to use roller crimper is from the 70% flowering stage in broadleaf species and from anthesis in grasses. Specifically, regarding Vicia villosa Roth, Kenee et al. [24] found that terminating the crop at full flowering resulted in 98% biomass suppression, compared to 80% at an earlier stage.

To date, several efforts have been made to improve the effectiveness and speed up the termination of cover crops using roller-crimpers. In the context of organic farming the use of chemical herbicide is forbidden, prompting exploration of alternative methods. In this regard, Frasconi et al. [25] tested the combined use of roller crimper and flame weeding, observing up to 90% of cover crops devitalization. However, the high cost of LPG limits the large-scale applicability of flame weeding for this purpose. Prototypes equipped with under-cutting blades have also been developed in order to damage root system thereby accelerating plant death and preventing regrowth [26]. However, this tool is primarily suited for use in fields where raised beds are created as occurs in some horticultural systems.

Another solution to improve the effectiveness of the termination with roller crimper is to increase the roller crimper weight. According to Cerruti [27], adding weight to the roller crimper, for example by placing water in the rollers, can increase the pressure exerted on the plants, improving the termination effectiveness. Indeed, a heavier roller crimper can exert greater downward pressure on the stems, improving their crushing and accelerating sap loss. Similarly, other studies have shown that adding weight through a hopper can reduce the risk of resprouting and improve weed control [26]. However, the potential increase in effectiveness of a roller crimper with added weight has not been extensively investigated, representing a notable gap in the current body of research that warrants further investigation. In addition to the increase in roller crimper weight, the number of passages can also influence the effectiveness of cover crops termination. For example, Kenee et al. [24] observed a further improvement in roller crimper termination efficacy on V. villosa with a second pass. Similarly, performing two or three passes with a roller crimper on a cover crop mixture of cereal rye and crimson clover led to greater and more rapid termination compared to a single pass. In particular, three passes consistently achieved higher termination values than one pass, reaching 90% more frequently and more quickly [28]. However, increasing the weight may negatively affect soil health by causing compaction [29]. Repeated passes over the field raise similar concerns, and this remains an active area of research. In this regard Kornecki et al. [30] demonstrated that repeated passes of the roller crimper (up to three times) did not lead to increased soil compaction over three growing seasons. Additionally, the rolled mulch helped to reduce soil strength and preserve volumetric water content compared to untreated standing crops. According to another study [31], which evaluated the effect of different numbers of roller crimper passes (one, two, or three), no increase in soil compaction was observed with a higher number of passes.

The aim of this study is to evaluate the effectiveness of a roller crimper used with two different weight configurations, light and ballasted, each tested with a different number of passes (one or two passes), in terminating a cover crop mixture in a vineyard. In addition to termination effectiveness, the four systems under comparison were also evaluated for their ability to conserve soil moisture as a result of the created dead mulch, and for their impact on soil compaction.

2. Materials and Methods

The two-year field trial (2023–2024) was carried out at Tenuta Montefoscoli in Palaia (Pisa, Italy) (43°33′54″ N, 10°44′24″ E) on a vineyard (cv: Vermentino) managed according to organic farming regulations (Reg. CE 2018/848). The vineyard, established in 2019, is trained to a unilateral Guyot system, with a spacing of 2.30 m between rows and 0.8 m within rows. The rows are oriented along a northeast–southwest axis, with an average slope of approximately 4.2%. The soil texture is sandy loam, consisting of 61.81% sand, 21.89% silt, and 16.30% clay, according to the USDA [32], and the area is characterized by a Mediterranean climate, with seasonal peaks in rainfall during spring and autumn. Total monthly rainfall (mm) and multi-year data (mm) from 2002 to 2024, along with average minimum and maximum air temperatures of the experimental area during the trial period are shown in Figure 1.

2.1. Experimental Layout

The trial was conducted during the 2023 and 2024 vine growing seasons. In both years, the effectiveness of the roller crimper in terminating the cover crop mixture was tested using two different weight configurations, each applied with one or two passes. In both years, a mixture of cover crops was sown in alternate rows at a seeding rate of 90 kg ha⁻¹. The mixture consisted of Hungarian vetch (Vicia pannonica Crantz) (40%), triticale (×Triticosecale × neoblaringhemii A. Camus nothosubsp. neoblaringhemii) (30%), and phacelia (Phacelia tanacetifolia Benth.) (30%).

In both trial years, before sowing, 21 kg·ha⁻¹ of N and 21 kg·ha⁻¹ of P₂O₅ were applied in the form of organic dried and pelleted manure (HUMIPROMOTER, Eurovix, Entratico, Italy, 3-3-0). Subsequently, sub-soiling was carried out, followed by sowing with a combined pneumatic seeder Sharp (Moreni, Montichiari, Italy).

In the first year of the trial, the cover crop mixture was sown in October 2022, and the first pass of the roller crimper with both the weight configuration was carried out on 8 June 2023, when triticale was at the milk growth stage, while vetch and phacelia were in bloom. The second pass with both the roller crimper weight configurations was performed on 20 June 2023. In the second year sowing was carried out late, in February 2024, due to weather conditions characterized by frequent rainfall. In 2024, the first pass with roller crimper was performed on June 7. At the time of termination, phacelia and vetch were in bloom, while triticale was at the milk growth stage. The second pass was carried out two weeks later, on 21 June 2024.

Table 1. Dates of roller crimper interventions for each of the four systems compared.

	LR + P1	LR + P2	HR + P1	HR + P2
8 June 2023	✔	✔	✔	✔
20 June 2023	-	✔	-	✔
7 June 2024	✔	✔	✔	✔
21 June 2024	-	✔	-	✔

✔, The pass with the roller crimper has been carried out; -, The pass with the roller crimper has not been carried out. LR + P1—light roller crimper configuration with one pass; LR + P2—light roller crimper configuration with two passes; HR + P1—heavy roller crimper configuration with one pass; HR + P2—heavy roller crimper configuration with two passes.

shows the dates on which the roller crimper interventions were carried out in each of the four systems compared over the two years of the trial.

The different roller crimper systems were evaluated using a 2-w randomized complete block design with three replications. During both years, the same roller crimper settings, in terms of weight configurations and number of passes, were evaluated.

2.2. Description of the Roller Crimper Employed

During the trial the roller crimper Rotovitis 180 (Dondi, Bastia Umbra, Italy) [16] was employed. The implement consists of a three-point-hitch-provided two-beam chassis 185 cm wide, supporting roller crimper with a width of 180 cm. Cutters are arranged tangentially and in a staggered pattern; they are made of treated boron steel with chisel-shaped profiles (Figure 2). This type of roller crimper can be mounted on the front or rear part of the tractor and is suitable for vineyards with a maximum row spacing of 245 cm.

The roller crimper has a net mass of 460 kg. It can be filled with water, resulting in a mass increase of 132 kg. In this model, it is also possible to increase the roller’s weight by adding ballast to the frame, with an additional weight gain of up to 391 kg. During the test, the light configuration had no added weight systems, while the heavy configuration was achieved by applying ballast to the frame (Figure 3). Therefore, the light configuration weighed 460 kg, while the heavy one with ballast weighed 851 kg. During the trial, the roller crimper was mounted at the front of the New Holland T4.90 F powered by a 63 kW diesel engine in the first year, while in the second year it was mounted on a Fendt 209-f Vario, powered by a 71 kW diesel engine. The New Holland T4.90 F tractor had a weight of 2910 kg, while the Fendt 209 F Vario weighed 2940 kg. Both the tractors employed were equipped with front tires sized 280/70 R18 and rear tires sized 340/85 R28. Roller crimper has always been used with a working speed of 6 km h⁻¹.

2.3. Data Collection

During both years, before the first pass with the roller crimper, the vegetation to be managed in the inter-row area was evaluated to obtain the average plant biomass in the vineyard involved in the trial. A 0.25 square meter frame (0.5 × 0.5 m) was used, within which all the plant biomass present was collected. Subsequently, in the laboratory, the different species were separated (i.e., the three cover crop species and the weeds), dried in an oven at 100 °C for 3–4 days until a constant weight was reached, and then weighed to determine the average dry matter weight. As a result, a separate average dry weight was obtained for each of the three cover crop species and a combined weight for all the weed species before the first rolling treatments. During the first year of the trial, both parameters were measured on 7 June 2023, while in the second year, parameters were measured on 6 June 2024.

The plant termination rate over time after the roller crimper’ interventions was estimated as percentage of plant material greenness. To measure this parameter, digital images of the lodged plants were acquired and subsequently analyzed using the Canopeo app (MathWorks, Inc., Natick, MA, USA) [25,33]. The app calculates for each processed image the percentage of green pixels relative to the total number of pixels, thus allowing to obtain the percentage of green cover of the plant material. This method was chosen because the color of the tissues in plants cut down by the roller crimper, in which photosynthetic activity is likely compromised, tends to shift away from green hues over time. During both years of the trial, photographic surveys were carried out immediately after the first pass with the roller crimper, daily for one week, and subsequently at a frequency of approximately one survey every two days. Specifically, in 2023, photographic surveys were carried out on June 8, 9, 10, 11, 12, 13, 14, 15, 17, 20, 22, 24, 26, 28, and 30, as well as on July 3 and 5. In 2024, photos were taken on June 7, 8, 9, 10, 11, 12, 14, 15, 18, 20, 21, 24, 26, and 28, and on July 1, 3, and 5. During each photographic survey, three digital images were collected for each replicate within a square meter frame (1 m × 1 m) that was fixed over dates (Figure 4).

To assess the impact of the four systems under comparison on soil compaction, soil penetration resistance (kPa) was assessed in each plot before the first pass with the roller crimper and after the second pass, including plots where only one pass had been carried out. In 2023, measurements were taken on June 8 and 20, while in 2024 they were taken on June 7 and 21. For this parameter, the FieldScout SC 900 Soil Compaction Meter (Spectrum Technologies Inc., Aurora, IL, USA) was used. Soil penetration resistance was measured from 0 to 30 cm depth, at 2.5 cm intervals. During both trial years, one measurement per replicate was conducted for this parameter for each survey. Alongside each measurement, a soil sample was collected to determine the soil water content, in order to evaluate its impact on penetration resistance. To account for within-plot spatial variability, for each measurement, three sampling points were selected within each plot, resulting in three subsamples of soil penetration resistance and soil water content per replicate. Soil samples were collected in cylinders with a volume of 144.3 cm³. The samplings conducted throughout the trial were consistently performed at the same locations, centered within the inter-row, thereby avoiding areas trampled by the tractor and including only the zones affected by the roller crimper.

During both trial years, soil moisture content was measured in each plot to evaluate the effect of the dead mulch created by the roller crimper under different weight configurations and number of passes. Therefore, soil samples were collected at three different depths—between 0–15 cm, 15–30 cm, and 30–45 cm—and subsequently oven-dried at 105 °C until a constant weight was reached to determine the moisture content. During both trial years, measurements were taken before the first intervention with the roller crimper, before the eventual second pass, and finally at the end of the trial. Therefore, in 2023, the measurements were conducted on 7 and 20 June, and on 27 July. In 2024, the measurements were conducted on 6 and 20 June, and 22 July. During each survey, one measurement per replicate was taken for this parameter. For each measurement, three sampling points were selected, resulting in three subsamples of soil water content per depth range per replicate, using cylinders with a volume of 144.3 cm³. The samplings conducted throughout the trial were consistently performed at the same locations.

2.4. Statistical Analysis

The plant termination rate over time parameter was analyzed using GraphPad Prism software version 10.0.0 (GraphPad Software, Boston, MA, USA) through nonlinear regression, applying a one-phase exponential decay model.

$$Y=\left(Y_0-Plateau\right)\times e^{(-kX)}+Plateau$$

(1)

where, Y is the dependent variable, in our case the percentage of green cover; X is time expressed in days after treatment; Y₀ is the percentage of green cover when X equals 0, which in our case corresponds to the time immediately after the first intervention with roller crimper; Plateau is the asymptotic value of Y, representing the percentage of green cover approached as time tends to infinity; k is the exponential decay rate, expressed in days⁻¹, for which higher values indicate a faster decline in green cover. Based on these parameters estimated through nonlinear regression, the span was calculated as the difference between the initial value of the dependent variable (Y₀) and the asymptotic value (Plateau). The half-life, calculated as ${\displaystyle \frac{\mathit{ln}2}{k}}$, shares the same unit of measurement as X and indicates the time required for Y (the dependent variable) to decrease by 50% of the span.

Regarding soil moisture content, the Shapiro–Wilk test and Bartlett test were performed to assess data normality and homoscedasticity, respectively. Data underwent transformations when necessary to meet the normality assumption. Soil moisture content was modelled with a linear mixed effect model. Data collected before the first roller crimper intervention and before the second one were processed separately with a three-way ANOVA. Roller crimper weight configuration (heavy and light configurations), soil depth at which moisture was measured (i.e., 0–15 cm, 15–30 cm, and 30–45 cm), and the trial year (2023 or 2024) were considered as fixed factors, while blocks were treated as random effect factor. Data collected at the end of the trial were instead analyzed using a four-way ANOVA, where roller crimper weight configuration, the number of passes with the roller crimper (one or two), soil depth at which moisture was measured, and the trial year were considered fixed factors, while blocks as random effect factor. For this data analysis the R software (version 4.4.3, R Core Team, Vienna, Austria) extension package “lme4” was used [34].

Soil penetration resistance data were analyzed using multivariate analysis of variance (MANOVA), conducted separately for the phase before the first roller crimper pass and the phase after the second pass. Before the first pass, MANOVA was used to evaluate the effect of roller crimper weight on soil penetration resistance. After the second pass, the effects of both roller crimper weight and the number of passes, as well as their interaction, were assessed. The dependent variables consisted of mean soil penetration resistance values grouped by depth (2.5 cm, 5.0 cm, 7.5 cm, 10.0 cm, 12.5 cm, 15.0 cm, 17.5 cm, 20.0 cm, 22.5 cm, 25.0 cm, 27.5 cm, and 30.0 cm). Pillai’s trace was used to test the multivariate effect of each factor on the combined dependent variables, i.e., soil penetration resistance values measured at different depths. These analyses were performed using SPSS statistical software (IBM Corp., Released 2019, Version 26.0, Armonk, NY, USA).

3. Results

3.1. Effect of the Roller Crimper Different Weight Configurations and Number of Passes on Plant Termination

The biomass of triticale, vetch, phacelia, and weeds at the beginning of the trial before the first roller crimper pass was evenly distributed across the different plots both in 2023 and 2024. Average values of triticale, vetch, phacelia, total weed, total cover crops and total plant community (i.e., cover crops together with weeds) dry biomass are displayed on

Table 2. Average values of triticale, vetch, phacelia, total weed, total cover crops and total plant community dry biomass (g d.m. m⁻²) before the first roller crimper intervention in 2023 and 2024, respectively. SE is the standard error of the means.

	Plant Biomass (g d.m. m−2)
	2023	2024
	Value (SE)	Value (SE)
Triticale	129.52 (14.57)	361.20 (42.33)
Hungarian vetch	235.62 (16.98)	84.34 (15.40)
Phacelia	24.16 (5.84)	30.92 (8.99)
Weeds	440.54 (36.60)	177.73 (21.28)
Total cover crop biomass	389.30	476.46
Total plant community biomass	829.84	654.19

.

In 2023, the major weeds present in the vineyard belonged to the genera Rapistrum, Lolium, Medicago, Orobanche, and Papaver, whereas in 2024, the dominant genera were Cirsium, Rapistrum, Trifolium, Papaver, and Lolium.

Nonlinear regression performed on the plants percentage values of green cover over time after termination showed that the data from the four systems under comparison are better described by four distinct curves for each trial year (F-test, p < 0.0001).

Table 3. Degrees of freedom, coefficient of determination (R²), and adjusted R² for the regressions analyzing the temporal trends in plant termination rates of the four compared systems during the two years of the trial.

	LR + P1	LR + P2	HR + P1	HR + P2	LR + P1	LR + P2	HR + P1	HR + P2
	2023	2023	2023	2023	2024	2024	2024	2024
Degrees of Freedom	150	150	150	150	150	150	150	150
R2	0.986	0.987	0.988	0.967	0.990	0.992	0.989	0.994
Adjusted R2	0.985	0.987	0.988	0.967	0.990	0.992	0.989	0.994

LR + P1—light roller crimper configuration with one passage; LR + P2—light roller crimper configuration with two passages; HR + P1—heavy roller crimper configuration with one passage; HR + P2—heavy roller crimper configuration with two passages.

presents the parameters indicating the goodness of fit of the green cover data over time to various curves, for the four systems compared across the two experimental years, based on the adopted single-phase exponential decay model.

The four distinct monophasic exponential nonlinear regression curves for the 2023 and 2024 trial years are shown in Figure 5.

Concerning the main parameters of the nonlinear regression curves, no significant differences emerged among the four systems in terms of Y₀ within the same trial year (Figure 6). This implicates that the four systems operated under similar initial conditions of plant green cover. The four systems Y₀ values ranged from 88.02 to 89.02% in 2023, and from 79.45 and 80.79% in 2024. Instead, a statistically significant difference emerged in the comparison of the systems parameter values of the two trial years, with higher Y₀ values in 2023 compared to 2024.

Regarding the plateau parameter, no difference emerged among the four systems in 2023, while in 2024 the system HR + P2 obtained a lower plateau value (3.66%), compared to HR + P1 (6.85%), LR + P2 (9.49%) and LR + P1 (8.72%) (Figure 7). HR + P2 allowed percentage reductions of 58.03%, 61.43% and 46.57% compared to LR + P1, LR + P2, and HR + P1 respectively. HR + P1 obtained a lower plateau value compared to LR + P2 and similar to LR + P1.

In both the trial years, the k parameter values of the HR + P2 system (0.14 and 0.16 in 2023 and 2024, respectively) were higher compared to HR + P1 (0.07 and 0.11 days⁻¹ in 2023 and 2024, respectively), LR + P2 (0.08 and 0.11 in 2023 and 2024, respectively) and LR + P1 (0.05 and 0.08 in 2023 and 2024, respectively) with an average percentage increase of 72.7, 60.2 and 140.0%, respectively (Figure 8). No difference emerged among HR + P1 and LR + P2, while LR + P1 obtained the lowest values.

Concerning the half-life parameter, the LR + P1 system (14.45 days) obtained a higher value compared to LR + P2 (9.16 days), HR + P1 (9.50 days), and HR + P2 (4.97 days) in 2023 (Figure 9). The same trend was observed in 2024, with values of 8.16, 6.37, 6.09, 4.40 days for LR + P1, LR + P2, HR + P1 and HR + P2, respectively.

Therefore, on average HR + P2 allowed a percentage reduction equal to 54.70, 37.72 and 38.33% of the hall-life compared to LR + P1, HR + P1, and LR + P2, respectively. Instead, LR + P1 obtained the lowest values during both years of the trial.

3.2. Effect on Soil Moisture Content and Soil Penetration Resistance

The three-way ANOVA revealed that neither the roller crimper weight configuration, the depth at which soil moisture was measured, the trial year, nor their interactions affected soil moisture content measured before the first roller crimper intervention. Therefore, it can be stated that the initial conditions were uniform across plots. The same factors and interactions also had no significant effect on the soil moisture parameter before the second roller crimper passages. The four-way ANOVA highlighted that none of the factors and interactions affected the parameter at the end of the trial. Three-way and four-way ANOVA results were showed in

Table 4. Results of the three-way ANOVA on soil moisture content before the first and second roller crimper passes, and of the four-way ANOVA at the trial end.

	Before the First Roller Crimper Pass	Before the Second Roller Crimper Pass	the End of the Trial
	p-Value	p-Value	p-Value
Factors
Weight	0.641	0.891	0.922
Soil moisture depth	0.063	0.638	0.861
Year	0.923	0.053	0.145
Weight × Soil moisture depth	0.990	0.989	0.826
Weight × Year	0.903	0.650	0.730
Soil moisture depth × Year	0.270	0.151	0.846
Weight × Soil moisture Depth × Year	0.333	0.437	0.984
Passes	-	-	0.624
Weight × Passes	-	-	0.963
Soil moisture depth × Passes	-	-	0.989
Year × Passes	-	-	0.650
Weight × Soil moisture depth × Passes	-	-	0.236
Weight × Year × Passes	-	-	0.670
Soil moisture depth × Year × Passes	-	-	0.962
Weight × Soil moisture depth × Year × Passes	-	-	0.469

.

Table 5. Average values of soil moisture content in weight terms (% W/W) for each system in three different phases in 2023 and 2024. SE is the standard error of the means.

		Before the First Roller Crimper Pass		Before the Second Roller Crimper Pass		End of the Trial
		Value (%)	SE	Value (%)	SE	Value (%)	SE
2023	LR + P1	11.97	0.15	13.47	0.07	10.83	0.19
	LR + P2	9.54	0.45	15.09	0.20	12.22	0.10
	HR + P1	8.52	0.29	13.33	0.06	9.93	0.34
	HR + P2	10.50	0.21	14.23	0.02	15.88	0.08
2024	LR + P1	12.60	0.18	11.43	0.10	7.90	0.28
	LR + P2	12.28	0.15	10.11	0.06	5.84	0.22
	HR + P1	11.08	0.19	11.54	0.03	4.88	0.03
	HR + P2	12.05	0.23	12.38	0.21	7.24	0.27

displays the average soil moisture content values recorded in the three different phases in 2023 and 2024. Each system value is the average among the three different soil moisture depths (i.e., between 0–15 cm, 15–30 cm, and 30–45 cm).

Regarding soil penetration resistance, the MANOVA was initially conducted on the pooled data from both years of the trial, referring to the phase before the first roller crimper pass. However, the analysis showed that the factor year had a significant effect on the dependent variables (Pillai’s Trace, p = 6.89 × 10⁻⁵). Therefore, separate analyses were performed for each year of the trial, 2023 and 2024. The analysis of the 2023 soil penetration resistance data, referring to the phase before the first roller crimper pass, showed that the factor roller crimper weight had no significant effect on the set of dependent variables, i.e., soil penetration resistance at different depths. The same result was observed in 2024. Therefore, it can be stated that soil penetration resistance conditions were uniform across the different plots where the four systems were tested at the beginning of the trial. The MANOVA analysis performed on the data from the phase after the second roller crimper pass showed that roller crimper weight, number of passes, and their interaction did not significantly affect the set of dependent variables in either 2023 or 2024. MANOVA results are shown on

Table 6. Pillai’s trace p values from MANOVA analyses showing the effects of roller crimper weight on soil penetration resistance at different depths before the first intervention, and the effects of weight, number of passes, and their interaction on multiple dependent variables after the second pass.

	Before the First Roller Crimper Pass		After the Second Roller Crimper Pass
	2023	2024	2023	2024
	p-Value	p-Value	p-Value	p-Value
Factors
Weight	0.776	0.418	0.305	0.304
Passes	-	-	0.758	0.666
Weight × Passes	-	-	0.211	0.128

.

Average values of soil penetration resistance of the systems under comparison after the second roller crimper pass in 2023 and 2024 are showed in

Table 7. Average soil penetration resistance (kPa) at different soil depths (cm) after the second roller crimper pass in 2023 and 2024 for the four systems under comparison. SE is the standard error of the means.

		HR + P1		HR + P2		LR + P1		LR + P2
2023	Depth (cm)	Value (kPa)	SE	Value (kPa)	SE	Value (kPa)	SE	Value (kPa)	SE
	0.0	310.67	111.06	218.33	75.31	299.00	94.19	230.00	50.01
	2.5	448.67	206.67	345.33	34.67	494.33	167.14	402.67	70.09
	5.0	437.00	162.19	494.67	8997	460.33	98.35	460.00	46.00
	7.5	494.33	102.25	494.67	89.97	390.67	80.67	471.67	109.63
	10.0	540.33	102.18	471.33	98.16	356.67	69.81	437.00	128.06
	12.5	471.67	57.33	379.33	86.89	333.33	60.60	345.33	103.33
	15.0	471.33	114.67	345.33	86.89	264.33	30.30	287.33	80.33
	17.5	471.67	22.67	437.00	82.93	264.67	99.95	322.00	63.76
	20.0	402.67	11.33	425.33	80.48	379.67	99.59	322.00	82.93
	22.5	402.33	30.30	483.00	155.83	564.00	121.70	333.33	100.56
	25.0	460.00	30.62	667.00	201.75	66700	92.00	379.67	121.26
	27.5	885.33	11.67	736.33	174.15	656.00	159.35	678.67	149.33
	30.0	1495.00	237.56	1346.00	421.60	1000.67	527.98	1115.33	196.71
2024	Depth (cm)
	0.0	149.33	41.42	471.33	100.56	288.00	121.70	322.00	109.44
	2.5	460.33	41.42	655.33	52.81	437.33	267.32	747.67	212.89
	5.0	575.33	57.33	862.67	224.39	598.00	179.64	954.33	180.72
	7.5	724.67	110.75	885.33	316.49	598.33	113.14	966.33	189.86
	10.0	644.00	23.00	805.33	242.13	782.00	11.67	908.33	19.92
	12.5	609.33	23.33	804.67	154.88	782.00	30.62	759.00	71.74
	15.0	793.33	242.13	736.33	154.56	517.67	19.92	632.67	69.81
	17.5	747.33	265.10	655.33	381.75	483.00	39.84	505.67	41.42
	20.0	828.00	363.24	632.33	109.63	494.33	30.55	471.67	94.29
	22.5	896.67	381.75	678.33	247.53	51767	19.92	552.33	110.75
	25.0	874.00	224.16	586.67	206.67	402.67	11.33	712.67	80.48
	27.5	977.67	213.07	678.67	128.18	586.67	111.06	781.67	132.78
	30.0	1161.33	383.06	736.33	80.33	667.00	116.66	908.33	129.45

LR + P1—light roller crimper configuration with one passage; LR + P2—light roller crimper configuration with two passages; HR + P1—heavy roller crimper configuration with one passage; HR + P2—heavy roller crimper configuration with two passages. SE is the standard error of the means.

.

4. Discussion

The initial green cover conditions of the total lodged plant biomass by the four systems were uniform across the plots. However, in 2023, the initial percentage of green cover was slightly higher compared to 2024. Although the plants were at the same phenological stage in both years, rainfall recorded in May 2023 was higher than in 2024. An increased availability of water can positively influence the development of cover crops, leading to more vigorous and sustained growth, which in turn can result in enhanced soil coverage [35,36,37].

In both 2023 and 2024, all systems reached plateau values below 10% green cover, indicating that over 90% of the total ground cover was devitalized. In agreement with Kornecki [38] and Ashford [14], achieving 90% devitalization reduces the risk of competition between the cover crop and the cash crop. If cover crops were to persist, they could deplete a substantial amount of the soil water reservoir [12]. In general, grasses, having more fibrous tissues and higher lignin content in their stems, may devitalize more slowly than legumes, whose leaf structure and water content make them more susceptible to crushing [39,40,41]. Phacelia behaves similarly to legumes [42]. Effective devitalization is achieved by intervening at the appropriate phenological stage of the plants and allows the establishment of a well-structured dead mulch [22]. Dead mulch can reduce soil evaporation, stabilize near-surface moisture, and buffer temperature swings, creating a more favorable rhizosphere for vines in the upper soil layer. [43]. Moreover, the roots of rolled plants remain in situ, and their decay creates biopores while releasing organic inputs that enhance soil aggregation and pore connectivity [44]. No significant differences for the plateau parameter were detected among the systems in 2023. Conversely, in 2024 the HR + P2 treatment reached a lower plateau value (3.66%), compared to HR + P1 (6.85%), LR + P2 (9.49%), and LR + P1 (8.72%), with a mean reduction of 55.34%. The absence of differences among the systems in 2023 may be associated with the slightly higher initial green cover of the flattened vegetation. Moreover, greater rainfall was recorded in June 2023 compared to 2024. Rainfall following rolling treatments can enhance the recovery of damaged plants, thereby attenuating the effects of roller-crimper systems [31,39]. HR + P2 also recorded higher k values in both trial years (0.14 and 0.16 days⁻¹ in 2023 and 2024, respectively) compared to the other systems. Higher values indicate a faster decline in plant green cover. The difference was particularly marked in comparison with LR + P1 (0.05 and 0.08 days⁻¹ in 2023 and 2024, respectively), for which HR + P2 showed a 140% increase.

The improved results obtained with the HR + P2 treatment can be attributed to the greater weight of the roller, which likely exerted higher pressure on the plant biomass and enhanced the crimping effect. This interpretation is consistent with Cerrutti [27], who observed that increasing the roller crimper’s weight improves termination effectiveness and reduces the risk of cover crops resuming upright growth. Similarly, other studies [30,38] reported that increasing the number of passes enhances the devitalization effect. Focusing on the roller crimper’s mechanism of action, it crushes plant stems at regular intervals against the soil surface, restricting the flow of water and nutrients within the plant and ultimately leading to plant death. When the roller passes multiple times over the same area, this process is repeated, effectively doubling the number of injuries to the cover crop (each crimping point being different from the ones caused by the former passage), as was the case in this study [31]. This repetition can accelerate the devitalization process.

The findings for the k parameter are consistent with the results observed for the half-life, i.e., the time required for the lodged plant green cover to decrease by 50% of the span (the difference between Y₀ and the plateau). In fact, the HR + P2 system showed the lowest half-life values, thus allowing the 50% threshold to be reached more rapidly (4.97 and 4.40 days in 2023 and 2024, respectively) compared to the other systems under comparison. Conversely, LR + P1 recorded the highest values (14.45 and 8.16 days in 2023 and 2024, respectively). However, the results obtained with the HR + P1 system are promising and not far from those of HR + P2, both in terms of the k parameter (0.07 and 0.11 days⁻¹ in 2023 and 2024) and half-life (9.50 and 6.09 days in 2023 and 2024, respectively). These positive results were also achieved because, in both 2023 and 2024, the first roller crimper pass was carried out, also by decision of the farm itself, on cover crops at an advanced phenological stage, the optimal timing for successful termination [22,23]. It would be interesting to investigate all four systems also at earlier phenological stages of the cover crops, to identify the most effective system. Tendentially, better devitalization performance, both for the k parameter and for half-life, were generally observed in 2024 compared to 2023. This could be attributable to the slightly higher initial green cover in 2023, as well as the greater rainfall recorded in June 2023 compared to 2024, for the reasons outlined above.

No differences emerged among the different systems compared in terms of soil water content and soil penetration resistance. This is in agreement with the observations of Kornecki and Kichler [31], who reported that increasing the number of roller crimper passes does not lead to an effective increased soil compaction. Therefore, the force exerted by the roller crimper may have been dissipated across the layer of flattened plant material, thereby avoiding damage to the underlying soil structure. Moreover, all systems recorded soil penetration resistance values below 2000 kPa. This threshold is considered critical, as impediments to grapevine root penetration can occur above penetrometer readings of 2000 kPa [45,46].

Overall, the HR + P2 system demonstrated the most effective performance in terms of devitalization of the lodged plant biomass, as indicated by the lowest plateau values in 2024 and the most favorable k and half-life values across both years. Notably, this system did not result in soil compaction. Nonetheless, the results obtained with the HR + P1 system were also positive, suggesting that this approach could represent a valuable compromise between efficacy and operational efficiency. Indeed, rolling with a single pass can offer significant advantages in vineyard management, especially when growers must coordinate multiple operations within short operational windows. These windows are frequently constrained not only by the complexity of vineyard operations but also by unstable weather conditions, such as recurrent rainfall events, which can limit field accessibility [47]. Therefore, reducing the number of passes could allow growers to optimize machinery use, freeing equipment for other critical tasks and facilitating timely interventions. From this perspective, the practicality of implementing dead mulch with a roller crimper in vineyard inter-rows may become a key factor for the broader adoption of this technique. Promoting such practices contributes to a more sustainable cover crop management strategy, aimed at enhancing soil conservation, a fundamental resource for the long-term resilience of viticultural systems increasingly threatened by degradation processes [48].

5. Conclusions

This study evaluated the effectiveness of four roller crimper-based systems for groundcover termination in vineyard inter-row area. No significant differences were observed among systems in terms of soil water content or soil penetration resistance, indicating that none of the configurations caused soil compaction or modified water evaporation rate under the tested conditions.

The system using the heavier roller with two passes (HR + P2) consistently achieved the best termination performance, with lower green cover plateau values in 2024 (3.66%), and faster biomass devitalization with higher k values (0.14 and 0.16 days⁻¹ in 2023 and 2024, respectively) and shorter half-life in both years (4.97 and 4.40 days in 2023 and 2024, respectively). However, the system using the same heavy roller with only one pass (HR + P1) also yielded favorable results, with termination metrics comparable to those of HR + P2, with k values equals to 0.07 and 0.11 days⁻¹ and half-life values of 9.50 and 6.09 days in 2023 and 2024, respectively. Given the need to coordinate multiple vineyard operations within short and weather-dependent timeframes, a one-pass approach such as HR + P1 offers clear operational advantages. This configuration represents a practical compromise between efficacy and efficiency, and its adoption could facilitate the wider use of roller crimping techniques in vineyards, thereby promoting sustainable cover crop management practices focused on soil conservation, an increasingly vital resource threatened by degradation. Further studies could focus on evaluating the effectiveness of the roller crimper in different pedoclimatic contexts to assess its adaptability and performance under varying environmental conditions. Different types of groundcovers and earlier termination times should be tested to allow validating the results of the performance of the specific roller crimper and its configuration/operational settings tested in this study. It would also be useful to investigate the devitalization effectiveness of roller crimpers with different designs, such as a two-stage model, i.e., with two rolling drums, and alternative blade profiles, to explore potential improvements in the speed and completeness of plant biomass termination.