Effects of Italian Ryegrass (Lolium multiflorum) Cultivation for Green Manure and Forage on Subsequent Above- and Below-Ground Growth and Yield of Soybean (Glycine max)

Abstract

To diversify upland cropping systems, Italian ryegrass (Lolium multiflorum; IRG) can be incorporated as forage or green manure to soybean (Glycine max L.). The current study was conducted to analyze the effect of IRG cultivar and usage methods on the subsequent soybean above- and below-ground growth, as well as the yield, under different conditions. Three cycles of crop rotation were implemented with the combination of IRG cultivar (early maturing “Kowinearly”; late maturing “Winterhawk”), IRG usage (green manure for which both above- and below-ground biomass was incorporated, +CC; and forage for which only below-ground biomass was incorporated, −CC), and fallow soil as control. The soybean yield of +CC was consistently high regardless of the IRG cultivar, and it demonstrated an increase even under unfavorable weather conditions, while −CC treated did not differ from control. Incorporated IRG could serve as a starter fertilizer (79 to 156 kg ha⁻¹ of N input). Soybean root characteristic differences showed substantial variability depending on the year and treatments. However, when compared to the control, no adverse effects of IRG were evident. Thus, using IRG as a green manure has the potential to enhance soybean yield, while using IRG as a forage could bring additional harvested matter to the rotational system.

1. Introduction

Soybean (Glycine max L.) is the most important leguminous grain crop in the world. In an association with rhizobium bacteria, soybean fixes N₂ through biological nitrogen fixation (BNF). It is an important part of the human diet and contributes to livestock feed because of its high protein concentration [1]. In addition, after grain harvest, crop residues supply nitrogen to subsequent crops and improve soil properties [2].

Due to climate change and global warming, many areas where intensive soybean production occurs have experienced challenges associated with changing weather patterns and frequent occurrences of extreme weather events [3,4]. For example, soybean yield reduction has followed the occurrence of flooding during germination and early seedling growth [5]. Waterlogging-restricted seed germination and initial root development damage the water and nutrient uptake capability of the crop [5]. The scarcity of water negatively impacts the symbiotic association at its early stages by damaging the signal exchange between host plants and rhizobia. Thus, BNF by soybean was highly susceptible to drought [6].

Crop rotation refers to the practice of cultivating various crops in sequence on the same field. Utilizing cover crops in a rotation may have several benefits that could ameliorate challenges associated with climate change, including increased soil fertility, reduced soil erosion, and enhanced soil moisture retention and weed suppression [7,8,9,10].

Italian ryegrass (Lolium multiflorum, IRG) is a winter-annual forage and cover crop that can be rotated with summer crops. IRG is rotated with cotton and peanut in the southern United States, and it is currently widely grown as a winter forage crop in rice paddies during the winter season in Korea [11]. If IRG is used for forage, growers can have an additional source of income. As a forage crop, above-ground biomass of IRG can be utilized for animal feed, while its roots and stubble residue can protect the soil. Previous studies have reported that IRG had a greater root biomass, root length, and root surface area than other cover crops, and that it continues to produce roots until late growth stages [10,12]. As a result, it can provide more below-ground biomass than other cover crops, and longer root persistence is associated with the provision of a greater amount of nutrients to the subsequent primary crop [13]. As a cover crop, IRG increased soil organic carbon more than leguminous crops [14]. Additionally, nine years of IRG use resulted in greater soil organic N accumulation than other green manure crops such as hairy vetch [15]. When 40 Mg ha⁻¹ of IRG were incorporated as a green manure, it improved soil total N and available K, and increased seedling biomass of rice in an IRG–rice rotation [16]. Although there are significant potential benefits of using IRG as a green manure, research is limited, especially regarding its effects on subsequent crops in the rotation. As a result, it is necessary to experiment with various crop rotations involving IRG, particularly to determine their impact on subsequent crops such as soybean. For instance, physiological and morphological responses of above- and below-ground parts of following soybean crop have not been tested in a multi-year field scale by different regimes of IRG usage.

Thus, the objectives of the current study were to determine: (1) The effects of incorporating IRG to existing soybean production as a novel rotational cropping system for the region; (2) The comparison of different usage regime and ecotypes of IRG on subsequent soybean shoot and root development; and (3) The evaluation of IRG biomass and nitrogen contents as a potential N supply to following soybean production.

2. Materials and Methods

2.1. Research Site and Field Conditions

The field experiment was conducted at Gyeongsang National University (GNU) Research Farm in Jinju, Korea (35°14′ N 128°09′ E). Soils are sandy loams. They are, from the 0-to-30-cm depth, 68% sand-, 30% silt-, and 2% clay-sized particles. Soil samples were collected at a 30-cm depth using a soil auger (diameter 2.54 cm) before and after IRG planting. Twenty soil samples were collected in a “W” pattern and then composited. Soil chemical properties are shown in

Table 1. Soil chemical properties of experimental site.

			Exchangeable Cations
pH	OM	Av. P2O5	K	Ca	Mg	EC
(1:5)	(g kg−1)	(mg kg−1)	(cmol+ kg−1)			(dS m−1)
6.5	11	94	0.3	4.1	0.5	0.2

. The weather data during three growing seasons were obtained from the Automatic Weather Station, Korea Meteorological Administration (available at https://data.kma.go.kr/, accessed on 24 September 2023, in Korean).

Table 2. Monthly mean temperature, growing degree days (GDD), and precipitation during the research period.

Month	2019–2020	2020–2021	2021–2022	2019–2020	2020–2021	2021–2022	2019–2020	2020–2021	2021–2022
	Temperature (°C)			GDD (°C) *			Precipitation (mm)
November	9.4	9.1	8.4	61	22	49	22	26	44
December	2.9	1.3	2.5	0	0	0	42	9	2
January	3.0	0.6	0.4	3	7	0	106	16	0
February	4.5	4.8	1.4	7	18	0	46	25	0
March	8.7	9.8	8.7	44	52	51	41	159	116
April	11.4	13.6	14.3	61	174	218	68	82	70
May	18.2	17.6	18.8	415	358	401	110	117	1
June	23.0	22.2	22.7	405	376	393	234	102	121
July	22.7	26.2	26.5	413	516	526	588	359	218
August	27.3	26.0	26.3	554	517	515	357	320	157
September	20.7	21.9	21.8	332	369	365	200	122	167
October	14.5	15.9	15.1	102	173	96	29	46	22

* Base temperatures for GDD were 5 and 10 °C for IRG (November–May) and soybean (June–October), respectively.

shows the monthly mean temperature, growing degree days (GDD), and precipitation for the study period.

2.2. Experimental Design Layout and Field Management

The plowed and rotovated field received 45, 36, and 36 ha⁻¹ of N, P₂O₅, and K₂O, respectively, using mixed fertilizer (21–17–17) each season for three years of the study. Seeds of two IRG ecotypes (Kowinearly and Winterhawk) were broadcasted (40 kg ha⁻¹). Additional N fertilizer was broadcasted with a handheld fertilizer sprayer (GE –US 18 Li, Einhell, Landau an derIsar, Germany) at N level of 60 kg ha⁻¹ using urea (60–0–0) in February. The fallowed control (F) did not receive any fertilizer during the winter season.

The experimental design was a randomized complete block with four replications of five treatments. Four of the treatments were combinations of two IRG ecotypes and two usage options for IRG prior to planting soybean. The two IRG cultivars (early maturing ecotype “Kowinearly”, Ko; late maturing ecotype “Winterhawk”, Win) were selected to assess their adaptability to different usage regimes and whether IRG ecotype affected subsequent soybean performance. Two different IRG usage regimes included either use as a forage in which shoots were removed but the roots were soil-incorporated (−CC), or usage as a green manure cover crop with both shoots and roots incorporated into the soil (+CC). The other treatment was fallow control (F).

When the winter season ended, IRG herbage was harvested from −CC plots using a gasoline-powered brush cutter (FS55, STIHL, Waiblingen, Germany) leaving 3–5 cm of stubble height from the soil surface. After the IRG was harvested, fields were shallowly rotovated (10–15 cm) using a tractor-mounted rotary tiller to mix existing IRG stubble and below-ground biomass with soil. The above- and below-ground IRG herbage from +CC plots (used as a green manure) was incorporated in the same manner as the −CC residue.

The “Daewon” soybean cultivar (one of the most commonly grown cultivars in Korea with relatively large grain size and 120–140 days of growing period) was planted at 4 weeks after IRG harvest to allow sufficient time for incorporated IRG decomposition (

Table 3. Schedule of field management operations and agronomic trait measurements during IRG and soybean cultivations.

Crop	Operation	2019–2020	2020–2021	2021–2022
IRG	(before) Soil sampling	6 November	30 October	4 November
	Planting and preplant fertilizer input	14 November	7 November	5 November
	Additional N input	8 February	7 March	9 March
	Harvest	1 May	7 May	11 May
	Plowing	1 May	18 May	13 May
	(after) Soil sampling	1 June	24 May	18 May
Soybean	Planting	3 June	8 June	8 June
	Samplings for root characteristics	9 July (V4)	22 July (V7)	20 July (R1)
		5 September (R3)	18 August (R3)	16 August (R3)
		10 October (R5)	8 October (R6)	5 October (R6)
	Harvest	24 October	21 October	24 October

). Before soybean planting, the field was rotovated and a ridge was made to increase infiltration during the monsoon season. Soybean seeds were planted at two seeds per hill, on the ridge with 90 cm inter-row spacing and 15 cm intra-row spacing using a hand-pushed disk planter (TP–10RA, Korea Agritechno Search Corp., Cheongju, Republic of Korea). Soybean was planted each year at 150,000 seeds ha⁻¹. The herbicide alachlor (2.0 kg a.i. ha⁻¹) was applied immediately following planting for weed management. Other management practices followed the guidelines of agricultural standard cultural practices of the Korea Rural Development Administration (RDA, 2003) [17].

2.3. IRG and Soybean Plants Responses Measured

The IRG shoot and root samples were collected at the IRG harvest (Table 3). For IRG above-ground biomass yield, samples were taken from an area of 2-m-×-2-m area per plot. Then, approximately 1 kg of fresh sub-sample was dried at 70 °C until constant dry weight was reached to determine the ratio of fresh to dry matter. In order to collect the IRG below-ground parts, soil was removed in a cylinder shape with 30 cm of diameter and 30 cm of depth in two random locations within each plot. Collected root samples were washed to remove attached soil, then dried at 70 °C for 72 h to determine dry matter. Harvested IRG above-ground biomass was dried, and then the N concentration of above-ground IRG was determined using a Leco TruMac CNS Analyser (LECO Corp., St. Joseph, MI, USA). Above-ground biomass nitrogen content was calculated.

Soybean leaf area index (LAI) and leaf chlorophyll content were measured approximately every two weeks from V7 to R6 stages. LAI was estimated using LAI2200–C with a 90°-angle cap (LI–COR BioSciences, Lincoln, NE, USA). Measurements of LAI were obtained with one above-canopy reading as a reference and six below-canopy readings at 0, 25, and 50% between row centers. The soybean showed leaf color change differences by visual observation in the first year, so in the second and third year, chlorophyll content was measured on fully developed uppermost leaves of 10 plants per plot using the CCM–300 (Opti–Sciences Inc., Hudson, NH, USA), which records chlorophyll content on a surface basis (mg m⁻²).

The soybean root characteristics were assessed three times to analyze growth development in each stage (Table 3). Cylindrical sampling plots were established with a radius of 15 cm and a depth of 30 cm. A total of seven samples per plot were collected, and after removing the two that were most different, five individuals with similar growth conditions were selected. The selected plants were divided into shoot and root fractions. Washed roots were placed in plastic trays filled with water and then scanned using a flatbed scanner (Expresson 12000XL, Seiko Epson Corp., Nagano, Japan). Scanned root images were analyzed using WinRhizo Pro system (WinRhizo Pro 2017a, Regent Instruments Inc., Sainte Foy, QC, Canada) to determine the total root length, total root surface area, and total root volume.

For soybean yield components, 10 plants from each plot were assessed to determine the number of pods and number of seeds. One hundred-seed weights were obtained for each plot. Grain yields were conducted by assessing a plot size of 3.6 m² (4 m × 0.9 m) and adjusted for 14% moisture concentration.

2.4. Statistical Method

PROC UNIVARIATE function of SAS 9.4 software (SAS Institute Inc., Cary, NC, USA) was used to assess the normality of data distribution. All measurements were analyzed with ANOVA using PROC MIXED model. IRG cultivar (Kowinearly, Winterhawk) was treated as a fixed effect. In soybeans, treatment (F, Ko−CC, Ko+CC, Win−CC, Win+CC) was treated as a fixed effect. For seasonal responses (LAI, Chlorophyll content, total root length, surface area, volume), different seasons were considered as repeated measures. The block was treated as a random effect. Least significant difference (LSD) tests (significance threshold of p = 0.05) were performed for mean separation. To have an in-depth understanding of treatment effect between IRG-treated group and F, different IRG usages (F versus IRG-treated, F versus +CC, F vs. −CC) were compared using preplanned contrasts for root morphological responses [18].

3. Results

3.1. Soybean Yield Components and Grain Yield

Soybean yield components were compared for three years (

Table 4. Soybean yield components and grain yield after various Italian ryegrass cultivation treatments for three years.

Year	Treatment	Pods No.	Grain No.	100-Grain Weight	Grain Yield
		(plant−1)		(g)	(Mg ha−1)
2020	F *	52 ns †	40 ns	25.0 ns	1.83 ns
	Ko−CC	61	50	25.5	2.03
	Ko+CC	55	43	27.0	2.30
	Win−CC	59	47	25.6	1.94
	Win+CC	47	35	26.7	2.14
	mean	55	43	26.0	2.05 A ‡
2021	F	76 ns	99 ns	26.5 ns	2.00 b
	Ko−CC	89	128	28.0	2.02 b
	Ko+CC	73	107	27.9	2.40 a
	Win−CC	73	105	28.0	1.83 b
	Win+CC	74	104	28.5	2.46 a
	mean	77	111	28.1	2.14 A
2022	F	66 ns	76 ns	28.5 bc	1.57 ab
	Ko−CC	64	82	28.6 bc	1.53 ab
	Ko+CC	71	83	29.8 ab	1.87 a
	Win−CC	72	81	27.8 c	1.30 b
	Win+CC	68	88	31.4 a	1.89 a
	Mean	68	82	29.2	1.63 B

* Fallow field as control, F; early-maturing cultivar “Kowinearly”, Ko; late-maturing “Winterhawk”, Win; a forage, −CC; a green manure cover crop, +CC. ^† Within each year, values followed by different lowercase letters are significantly different between treatments at p ≤ 0.05; ns, not significant. ^‡ Values followed by different uppercase letters are significantly different between years at p ≤ 0.05.

). The number of pods and grains did not differ between the treatments for three years of study. Across the treatments during 2020, the average number of grains tended to have lower values compared to those of 2021 and 2022. There were no significant differences between the treatments in 100-grain weight both in 2020 and 2021, while in 2022, a significant difference was observed (p = 0.0163). In 2022, the Win+CC had the highest 100-grain weight value (31.4 g), while the Win−CC had the lowest value (27.8 g). The average 100-grain weight across the treatment was 26.0, 28.1, and 29.2 g in 2020, 2021, and 2022, respectively.

For soybean grain yield, there were no differences in 2020, but both 2021 and 2022 showed significant differences among treatments (Table 4). In 2020, the average grain yield for +CC was 2.22 Mg ha⁻¹, while F yields were 1.83 Mg ha⁻¹. In 2021, the grain yield was greatest for Ko+CC (2.40 Mg ha⁻¹) and Win+CC (2.46 Mg ha⁻¹), while it was least for Win−CC (1.83 Mg ha⁻¹). When soybean yield was compared based on IRG usage (forage versus green manure), Ko+CC showed a 19% greater yield compared with Ko−CC, and Win+CC showed a 34% higher yield compared with Win−CC (Table 4). Furthermore, the average grain yield of the IRG-treated sample was 2.18 Mg ha⁻¹ compared with 2.00 Mg ha⁻¹ for F. The average of +CC (2.43 Mg ha⁻¹) was 22 and 26% greater compared with F (2.00 Mg ha⁻¹), and the average of −CC (1.93 Mg ha⁻¹), respectively. This pattern of response was similar to 2021 LAI values at the R4 stage (See Section 3.4).

In 2022, Ko+CC (1.87 Mg ha⁻¹) and Win+CC (1.89 Mg ha⁻¹) had the highest grain-yield values, while Win−CC (1.30 Mg ha⁻¹) had the lowest value. Grain yield for Win+CC was 45% greater than the Win−CC. The average of +CC (1.88 Mg ha⁻¹) was 20 and 32% higher compared to F (1.57 Mg ha⁻¹) and −CC (1.42 Mg ha⁻¹), respectively. In the comparison between years, it was observed that 2020 received more rainfall in July compared to 2021. However, there was no significant difference in soybean grain yield, while the yield in 2022 was 24% lower than in 2021 (p = 0.0013).

3.2. IRG Above- and Below-Ground Biomass

The above- and below-ground biomass of two different ecotypes of IRG were compared for 3 years (

Table 5. Above- and below-ground biomass of two Italian ryegrass cultivars for 3 years.

Cultivar	Above-Ground Dry Matter				Below-Ground Dry Matter
	(Mg ha−1)
	2020	2021	2022	Mean	2020	2021	2022	Mean
Kowinearly	6.8	6.7	6.9	6.8	2.8	1.9	1.2	2.0
Winterhawk	6.4	6.0	7.2	6.5	2.1	2.5	1.4	2.0
Mean	6.6	6.4	7.0		2.4 A *	2.2 A	1.3 B

* Values followed by different uppercase letters are significantly different between years at p ≤ 0.05.

). Only a year main effect occurred for IRG below-ground biomass (p = 0.0274). On average, IRG above-ground biomass yield was 6.6, 6.4, and 7 Mg ha⁻¹ of dry matter for 2020, 2021, and 2022, respectively. This was slightly less than in a previous study (8.7 Mg ha⁻¹ of dry matter) in Korea with the Kowinearly cultivar [19]. In 2022, IRG below-ground biomass was 46 and 41% lower than the average of 2020 and 2021 (average of 1.3 versus 2.4 and 2.2 Mg ha⁻¹ of dry matter). However, if IRG was used as a green manure (+CC), the sum of biomass incorporated into the soil was 9.0, 8.6, and 8.3 Mg ha⁻¹ of dry matter in 2020, 2021, and 2022, respectively. Thus, with different usage scenarios such as green manure (+CC) versus forage (−CC), dry-matter input to soil can differ by up to 7-fold.

3.3. IRG Above-Ground Biomass Nitrogen Content

To estimate the nitrogen supply to soybeans from IRG above-ground sources, IRG above-ground nitrogen content was measured (

Table 6. Above-ground biomass nitrogen content of two Italian ryegrass cultivars for 3 years.

Cultivar	Above-Ground Nitrogen Content (kg ha−1)
	2020	2021	2022	MEAN
Kowinearly	156	83	107	115
Winterhawk	147	86	79	104
Mean	151 A *	84 B	93 B

* Values followed by different uppercase letters are significantly different between years at p ≤ 0.05.

). Significant differences were observed in the main effect of Year (p < 0.0001). The IRG nitrogen content in 2020 (151 kg ha⁻¹ of N) was 80 and 62% higher than in 2021 and 2022, respectively. No significant variations in IRG nitrogen content were found across different IRG cultivars. This may indicate that IRG above-ground incorporation did not affect soil organic matter (Table S1). Additional input of N from IRG above-ground may have contributed to the increased soybean grain yield of +CC treatments (Table 4) as compared to F or −CC.

3.4. Soybean LAI and Chlorophyll Content

To understand the photosynthetic capacity of soybean during the growing season, LAI and chlorophyll contents were measured during 2021 and 2022 (

Table 7. Soybean LAI and chlorophyll content in different growth stages after various Italian ryegrass cultivation treatments for three years.

Year	Treatment	LAI				Chlorophyll Content
						(mg m−2)
		V7	R2	R4	Mean	R1	R3	R6	Mean
2021	F *	1.3 C † b ‡	1.8 Ba	3.2 Aab	2.1 ns	316	399	382	365 ab
	Ko−CC	1.8 Ba	1.6 Ba	3.1 Ab	2.1	303	391	368	354 c
	Ko+CC	1.4 Bab	1.7 Ba	3.6 Aa	2.2	322	405	381	369 a
	Win−CC	1.3 Bb	1.6 Ba	3.1 Ab	2.0	297	397	378	357 bc
	Win+CC	1.5 Bab	1.7 Ba	3.5 Aab	2.2	320	398	371	363 ab
	Mean	1.4 C	1.7 B	3.3 A		312 C §	398 A	376 B
		R1	R2	R5		R2	R3	R6	Mean
2022	F	2.2 ns	3.7 ns	5.3 ns	3.7 ns	331	378	296	335 ns
	Ko−CC	2.4	3.5	5.3	3.7	343	383	294	340
	Ko+CC	2.3	3.4	5.3	3.7	324	371	282	326
	Win−CC	2.5	3.7	5.1	3.7	340	359	281	326
	Win+CC	2.8	3.9	5.3	4.0	330	379	288	332
	Mean	2.4 C	3.6 B	5.3 A		334 B	374 A	288 C

* Fallow field as control, F; early maturing cultivar ‘Kowinearly’, Ko; late maturing ‘Winterhawk’, Win; a forage, −CC; a green manure cover crop, +CC. ^† Within a treatment, values followed by different uppercase letters are significantly different between growth stages at p ≤ 0.05. ^‡ Within a growth stage, values followed by different lowercase letters are significantly different between treatments at p ≤ 0.05; ns, not significant. ^§ Values followed by different uppercase letters are significantly different between growth stages at p ≤ 0.05.

). For LAI, there was a treatment × growth-stage interaction in 2021 (p = 0.0388), while there was a growth-stage main effect only in 2022 (p < 0.0001). Hence, chlorophyll content showed growth-stage main effects for both years (p < 0.0001), and there was a treatment main effect in 2021 (p = 0.0061).

Specifically, Ko−CC treatment had the highest LAI value of 1.8 at V7 stage in 2021. The average LAI of IRG-treated plots was slightly higher than F at V7 (1.5 versus 1.3). There were no differences among the treatments at R2. The mean LAI values were 3.2, 3.1, and 3.6 for F with an average −CC, and +CC, respectively, at R4. In 2022, the average LAI values at R5 were the greatest and followed by R2 and R1 (5.3, 3.6, and 2.4, respectively). At R1, F tended to have the lowest LAI value (2.2), while the IRG-treated plots had an average LAI of 2.5. However, there were no statistical differences between the treatments.

In 2021, the chlorophyll content of Ko+CC was greatest (369 mg m⁻²), while Ko−CC had the lowest value (354 mg m⁻²) averaged across growth stages. The average value for each growth stage was greatest during R3, followed by R6 and R1 (398, 376, and 312 mg m⁻², respectively). In 2022, the chlorophyll content was highest during R3, followed by R2 and R6 (374, 334, 288 mg m⁻²) across the treatments. Interestingly, the chlorophyll content at R6 in 2021 (376 mg m⁻²) was 31% higher than that of the R6 stage in 2022 (288 mg m⁻²), suggesting that the soybean stay-green period was maintained through the later reproductive stage in 2021. There was no difference in chlorophyll content between different usages of IRG in either 2021 or 2022.

3.5. Soybean Total Root Length, Surface Area, and Volume

To investigate the differences in root development in early (vigorous growth), middle (maximum development), and late (before major decline) plant developmental periods, measurements of total root length, surface area, and volume were conducted for 3 years. The total root length, surface area, and volume showed significant treatment × growth stage interactions in 2020 (p < 0.05). In 2021, treatment- and growth-stage main effects were observed (p < 0.05). In 2022, treatment- and growth-stage interactions were observed only for total root length and volume, while significant differences were observed in total root length, surface area, and volume under the main effect of growth stage (p < 0.0001).

In 2020, the total root length, surface area, and volume in the plants showed significant differences among the treatments at R3, with the highest values observed at this stage (avg. of 2030 cm, 286 cm², 8.3 cm³,

Table 8. Soybean total root length, surface area, and volume in different growth stages after various Italian ryegrass cultivation treatments for three years.

Treatment	Total Root Length				Total Root Surface Area				Total Root Volume
	(cm plant−1)				(cm2 plant−1)				(cm3 plant−1)
	2020
	V4	R3	R5	Mean	V4	R3	R5	Mean	V4	R3	R5	Mean
F *	664 B † a ‡	1590 Ab	876 Ba	1040 b	77 Ca	219 Ab	152 Bb	149 c	1.6 Ba	6.3 Ac	4.5 Ab	4.1 b
Ko−CC	732 Ba	2680 Aa	992 Ba	1470 a	87 Ca	361 Aa	182 Bab	210 ab	1.9 Ca	10.0 Aab	5.9 Bab	5.9 ab
Ko+CC	707 Ba	1890 Ab	1020 Ba	1206 ab	81 Ca	257 Ab	187 Bab	175 bc	1.7 Ba	7.3 Abc	5.8 Aab	4.9 b
Win−CC	794 Ca	2480 Aa	1290 Ba	1521 a	91 Ca	385 Aa	239 Ba	238 a	2.0 Ca	12.3 Aa	8.0 Ba	7.4 a
Win+CC	574 Ba	1540 Ab	909 Ba	1007 b	69 Ba	207 Ab	162 Aab	146 c	1.6 Ba	5.6 Ac	4.9 Aab	4.0 b
Mean	694 C	2030 A	1020 B		81 C	286 A	184 B		1.8 C	8.3 A	5.8 B
	2021
	V7	R3	R6	Mean	V7	R3	R6	Mean	V7	R3	R6	Mean
F	593 ns	2290	1460	1450 bc	121 ns	409	337	289 b	2.0 ns	6.2	6.6	4.9 abc
Ko−CC	567	2380	1430	1460 bc	123	429	366	306 b	2.2	6.7	8.0	5.6 ab
Ko+CC	688	2100	1270	1350 c	131	332	271	245 b	2.0	4.3	4.9	3.7 c
Win−CC	709	3490	1740	1980 a	135	559	452	382 a	2.1	7.5	9.6	6.4 a
Win+CC	856	2850	1630	1780 ab	142	418	330	296 b	1.9	5.1	5.6	4.2 bc
Mean	683 C	2620 A	1510 B		130 C	429 A	351 B		2.0 B	6.0 A	6.9 A
	2022
	R1	R3	R6	Mean	R1	R3	R6	Mean	R1	R3	R6	Mean
F	2100 Aa	2550 Abc	2100 Aa	2250	292 ns	452	500	415	3.3 Ca	6.5 Bab	9.7 Aab	6.5
Ko−CC	1970 Ba	3100 Aab	1960 Ba	2350	275	536	489	434	3.1 Ca	7.5 Ba	9.9 Aab	6.8
Ko+CC	1820 Aa	2220 Ac	2410 Aa	2150	264	350	438	351	3.1 Ba	4.4 ABb	6.5 Ac	4.7
Win−CC	1980 Aa	2670 Abc	2260 Aa	2300	275	461	575	437	3.1 Ca	6.4 Bab	11.9 Aa	7.1
Win+CC	1670 Ca	3720 Aa	2700 Ba	2690	247	549	498	431	2.9 Ba	6.5 Aab	7.5 Abc	5.6
Mean	1910 C	2850 A	2290 B		271 B	470 A	500 A		3.1 C	6.3 B	9.1 A

* Fallow field as control, F; Early-maturing cultivar “Kowinearly”, Ko; Late-maturing “Winterhawk”, Win; a forage, −CC; A green manure cover crop, +CC. ^† Within a treatment, values followed by different uppercase letters are significantly different between growth stages at p ≤ 0.05. ^‡ Within a growth stage, values followed by different lowercase letters are significantly different between treatments at p ≤ 0.05; ns, not significant.

). Among the treatments at R3, the average total root length, surface area, and volume in the plants were highest for the −CC avg., measuring 2580 cm, 373 cm², and 11.2 cm³, respectively. When different IRG usage was compared with the F using preplanned contrast at R3, the total root length was 62% greater for avg. of −CC (2580 cm) and 35% overall IRG-treated (2148 cm) than that of the F (1590 cm) (

Table 9. Preplanned contrast p-value of soybean total root length, surface area, and volume in different growth stages after various Italian ryegrass cultivation treatments for 3 years.

Year	Preplanned Contrast	Total Root Length			Total Root Surface Area			Total Root Volume
		V4	R3	R5	V4	R3	R5	V4	R3	R5
2020	IRG vs. F	0.4371 *	0.0023 †	0.4804	0.6498	0.0035	0.3922	0.8511	0.0102	0.3802
	−CC vs. F	0.3721	<0.0001	0.2728	0.5794	<0.0001	0.1759	0.7763	0.0002	0.1465
	+CC vs. F	0.5987	0.3416	0.8512	0.7836	0.3821	0.8410	0.9534	0.4004	0.8895
		V7	R3	R6	V7	R3	R6	V7	R3	R6
2021	IRG vs. F	0.4248	0.2952	0.5539	0.7671	0.7612	0.5811	0.8755	0.3866	0.1861
	−CC vs. F	0.4631	0.7779	0.8663	0.8068	0.1543	0.1416	0.8597	0.1413	0.0215
	+CC vs. F	0.4688	0.0319	0.3628	0.7671	0.3774	0.6318	0.9130	0.9220	0.9485
		R1	R3	R6	R1	R3	R6	R1	R3	R6
2022	IRG vs. F	0.3208	0.0008	0.0963	0.5963	0.0850	0.9688	0.8498	0.7716	0.0465
	−CC vs. F	0.3528	0.0152	0.0847	0.6426	0.4176	0.5844	0.8900	0.6701	0.0028
	+CC vs. F	0.3760	0.0002	0.1876	0.6147	0.0208	0.6341	0.8357	0.3414	0.5730

* Significance evaluated at 0.05 probability level. ^† Values in boldface are statistically significant.

). Similarly, the total surface area was 70% greater for avg. of −CC (373 cm²) and 38% overall IRG-treated (303 cm²) than that of the F (219 cm²) (Table 9). Similarly, the total volume was 78% greater for avg. of −CC (11.2 cm³) and 40% overall IRG-treated (8.8 cm³) than that of the F (6.3 cm³) (Table 9).

In 2021, across the growth stages, total root length, surface area, and volume in the plants were the highest trend in Win−CC with values of 1980 cm, 382 cm², and 6.4 cm³, respectively. Among the growth stages, R3 showed the largest average values for total root length (2620 cm) and surface area (429 cm²), followed by R6 and V7 (Table 8). When comparing F using preplanned contrasts at the R3 stage, total root length significantly increased by 8% (p = 0.0319) under an average of +CC (2475 cm) compared to F (2290 cm). At the R6 stage, total root volume showed a significant 33% increase (p = 0.0215) with an average of −CC treatment (8.8 cm³) compared to F (6.6 cm³).

In 2022, the total root length showed significant differences between the treatments at R3. The Win+CC had the greatest length (3720 cm), while Ko+CC had the smallest (2220 cm). The total root length of Ko−CC was 40% greater for Ko+CC (3100 versus 2220 cm), and Win+CC was 39% greater for the Win−CC (3720 versus 2670 cm). When comparing between growth stages within a treatment, Ko−CC and Win+CC had a greater total root length at R3. When different IRG usages were compared with F using preplanned contrast at the R3 stage, the IRG-treated (2928 cm) and the average of −CC (2885 cm) and +CC (2970 cm) resulted in total root length increases of 15% (p = 0.0008), 13% (p = 0.0152), and 16% (p = 0.0002) compared to that of the F (2550 cm). In the case of total root surface area, the average values across treatments were relatively higher at the R3 and R6 (Table 8). The total root volume at the R3 stage was greatest for Ko−CC, while Ko+CC had the smallest (7.5 versus 4.4 cm³). At the R6 stage, Win−CC had the highest volume, while Ko+CC had the lowest (11.9 versus 6.5 cm³). Unlike total root length, root volume had the largest value at R6 across the treatment. When different IRG usage was compared with F using preplanned contrast at the R6 stage, the average of the −CC group was of 12% greater root volume compared to F (10.9 versus 9.7 cm³) (p = 0.0028). Overall, the total root length, surface area, and volume in the plants for both R3 and R6 tended to have greater values in 2022 compared to 2021 (Table 8).

4. Discussion

To understand the effects on above- and below-ground soybean development, various IRG-soybean rotation regimes were compared for 3 years under field conditions. Soybean yield component analysis indicated IRG usage as a green manure (+CC) had generally greater grain yield than other treatments, while IRG for forage (−CC) was not different from F across IRG cultivars. This is in line with previous studies that found higher soybean grain yields when both the above-ground and below-ground parts were placed as cover crops in comparison to F [20]. In addition, the introduction of a cover crop increased soybean yield by 14% compared with no cover crop [9]. In contrast, it differed from previous research where the soybean grain yield following IRG cultivation was lower than that after winter fallow [21]. A previous cover crop treatment showed that there was no significant difference in soybean yield compared to no cover crop control [8]. It is likely that the increased input of organic matter and subsequent mineralization in +CC compared with −CC positively influenced soybean grain yield (Table 4). There was also a difference among years in soybean grain yield. The overall mean of the treatments in 2022 was significantly lower than in 2020 or 2021 (p = 0.0013). This significant decrease in grain yield can be attributed to the reduced rainfall in 2022, particularly during the reproductive stage (Table 2). This is in line with previous studies where yield reduction occurred due to water stress during the reproductive growth stage of soybean [22,23]. Particularly, the precipitation in August was reduced by 200 mm and 163 mm in comparison to 2020 and 2021, respectively (Table 2). This period corresponds to the developmental stages of pod development (R3–R4), during which drought stress has an impact on grain yield [24]. In addition, GDD in 2022 was lower by 6 and 77 °C compared to 2020 and 2021 in October, respectively (Table 2). Lee et al. (2022) suggested that conditions with higher GDD levels are favorable to an increase in soybean grain yield [25]. Therefore, it appears that the combination of drought conditions and lower GDD resulted in reduced grain yield. Thus, for 3 years of study indicated that even under unfavorable weather condition, +CC treatments showed greater yield compared to other treatments.

The above- and below-ground biomass of two different ecotypes of IRG showed that IRG below-ground biomass in 2022 was 43% lower than the average of 2020 and 2021. This was also due to a different weather pattern for each growing season (Table 2). Even though the above-ground biomass yield was not affected by the year, lower precipitation during the winter-growing season negatively affected root development. Specifically, the third-year winter growing season had lower precipitation than the previous two years (435, 434 versus 233 mm; November to May). Previous studies addressed that under moderate drought-stress conditions, wheat exhibited reduced root mass in both topsoil and subsoil compared to well-watered conditions [26]. Previous research indicated that as soil penetration resistance increased, root length density decreased in the compacted soil layer [27,28]. Under moderate water stress, the decrease in root weight was more pronounced than that of above-ground biomass [29]. Despite large differences in dry matter input among IRG-treated plots (Table 5), there was no difference in soil organic matter content among the treatments (Table S1). According to previous studies, only 10–20% of dry matter input becomes part of the soil organic matter [30]. Although the amount of biomass input to soil was significantly different, actual differences in the organic matter pool may have been unaffected among treatments.

Starter fertilizer to soybean is a common practice as prior research indicated that supplying 50 kg ha⁻¹ of starter nitrogen fertilizer resulted in the highest soybean grain yield [31]. Based on the IRG above-ground biomass nitrogen content analysis, the amount of N supply from +CC exceeded suggested initial N fertilization of the previous study. Thus, to promote the initial growth of soybeans, implementing crop rotation with IRG as a cover crop may be an alternative method of using inorganic nitrogen starter fertilizer. This is in line with the recommendation of the previous study that use of a gramineous cover crop stored 25% more N on the soil surface than no cover crop control and may replace starter N fertilizer [32]. Above-ground IRG had 79–156 kg ha⁻¹ of N, which subsequently decomposed; incorporated IRG could be served as a starter fertilizer. Previous studies also reported that N accumulation of 26 kg ha⁻¹ from buried oat straw and 39 kg ha⁻¹ from cereal rye residue could be supplied to following crops [33,34]. Sievers and Cook (2018) observed that cereal rye residue exhibited a slower and more gradual release of nitrogen compared to hairy vetch, emphasizing the need for optimal synchronization of cover crop nitrogen release with subsequent crop nitrogen uptake to reduce nitrogen fertilizer [35].

The soybean LAI of IRG incorporated (+CC) had relatively higher values than other treatments for 2021 at R4. This is similar to a previous study that LAI of soybean at the R4 stage was 41% higher when the cover crop was incorporated compared with the no-cover crop control [20]. It is reported that under abiotic stress conditions, rice exhibited higher LAI and yield when intercropped with green manure, compared to the control conditions [36]. Previous studies also indicated that when the LAI of soybean during the R1–R4 stage ranges from 3.5 to 4, the photosynthetic potential is maximized [37,38]. The chlorophyll content is a crucial indicator reflecting the photosynthetic activity of plants [39]. In this study, the chlorophyll content during the R3 and R6 stages decreased sooner in 2022 compared to that of 2021, likely due to reduced rainfall in July and August (p < 0.0001, Table 2). Elsalahy and Reckling (2022) reported a 7% reduction in chlorophyll content of drought-stressed soybeans during the vegetative and flowering stages compared with under rainfed conditions [40]. Gholamin et al. (2011) also observed that drought stress diminished chlorophyll content and yield while also revealing a strong correlation between chlorophyll content and harvest yield [41]. Therefore, the decrease in chlorophyll content during the R3 and R6 stages occurred due to drought conditions in 2022 regardless of IRG input regimes.

In 2020, the total soybean root length, surface area, and volume was greater for IRG-treated crops than that of the F across the IRG cultivars. It was reported that as the root length density of the cover crops (triticale and pearl millet) increases, there is a corresponding increase in soybean root length density in 5–10 cm of soil [42]. Additionally, it was noted that the vigorous root growth of triticale and pearl millet led to a reduction in soil-penetration resistance [42]. During favorable conditions in 2020 and 2021, −CC showed a tendency for greater soybean total root length and surface area than +CC. When plants had moderate nutrient depletion, they tend to develop excessive roots to scavenge the nutrients and water. It was reported that low phosphorus concentrations lead to a high density of long roots [43]. Lecompte et al. (2008) observed a 180% increase in root length when nitrogen was limited compared to when it was sufficient [44]. Additionally, they observed an enhanced root density extending into deeper soil layers under conditions of low-soil nitrogen levels, and they characterized root expansion as continuous across the entire soil profile [44]. In the current study, this is probably why there were root development differences in later growing seasons between −CC and +CC. Hence, due to the predominant contribution of fine roots at root length, the total root length was significantly influenced by fine root development [45]. Fine roots are directly associated with water uptake compared to main roots, and they enable stable crop production through root distribution and development [46]. In the current study, overall root length development was highest at R3 in 2022, and Win+CC had the highest total root length. This indicates fine root development reached maximum at R3, suggesting Win+CC probably had greater adaptability to unfavorable weather conditions, including drought. The previous research suggested that the substantial proportion of fine roots in the root domain allows for the optimization of the capacity to explore the soil in preparation for drought conditions [47]. Thus, the current study suggested that incorporating IRG in soybean production may enhance the root system of soybeans and soybean yield.

In later seasons (R3 and 6), the total root length, surface area, and volume tended to have greater values in 2022 compared to 2021 (Table 8). This can be attributed to an overall lower precipitation, particularly during the months of July and October in 2022, leading to the development of larger and elongated roots for enhanced water acquisition (Table 2). Fenta et al. (2014) reported that most cultivars had increased root length, surface area, and volume under drought conditions, excluding drought-sensitive varieties [48]. Hence, it was reported that ruzigrass was found to enhance the growth of soybean roots in compacted soil conditions compared with other preceding crops (maize, oat, and wheat) [49]. Additionally, it has been suggested that the continuous presence of biopores created by ruzigrass in the preceding crop can promote soybean root growth even under water-stressed conditions, potentially leading to increased yields [49]. According to the previous study, cover crops resulted in an approximately 80% increase in water infiltration compared with no cover crop [9]. Basche et al. (2016) reported that a previous cover crop treatment increased soil water content at a depth of 30 cm relative to the no-cover crop [8]. Therefore, it can be suggested that the increased water-holding capacity due to the cover crop may have contributed to enhanced soybean root growth and maintenance.

When overall yield was compared between years, the yield in 2022 was lower compared to the years 2020 and 2021 (Table 4). This is attributed to the extended development of root volume and LAI during later seasons (Table 7 and Table 8). This is most likely due to the drought in 2022, which disrupted nutrient flow as plants progressed from the vegetative stage to the reproductive stage. Islam et al. (2019) suggested that the decline in LAI during the reproductive stage could be attributed to a reduction in leaf nitrogen concentration required for grain filling in rice [50].

5. Conclusions

The soybean yield of +CC appears to be consistently high regardless of the IRG cultivar, and it demonstrates an increase in yield even under unfavorable weather conditions. Consequently, growers can select IRG varieties based on their specific objectives. For instance, to enhance soybean yield through early sowing, the early maturity IRG ecotype, such as “Kowinearly”, could be cultivated and incorporated as a green manure. On the other hand, for enhancing IRG biomass and soil fertility, the late maturity ecotype, such as “Winterhawk”, could be grown and harvested. Soybean root characteristic differences showed substantial variability depending on the year and treatments. However, when compared to the F treatment, no adverse effects were evident. As cover-crop decomposition rates vary, future research should focus on quantifying the decomposition rate and nitrogen release of IRG, estimating nutrient supply to subsequent crop growth stages.