Hay Yield, Chemical Composition, and In Vitro Digestibility of Five Varieties of Common Vetch

Abstract

Diversifying feed with annual legumes could help to reduce reliance on imported sources of alfalfa hay, ensuring a consistent supply of forage throughout year, and potentially decreasing the cost of raising ruminants. This study evaluated the varietal variability in hay yield, chemical composition, carbohydrate and protein fractions, and in vitro digestibility of five common vetch varieties used in this experiment during the full-flowering period in Jianghuai region during the 2022 crop season. Results showed that improved varieties (Lanjian No. 1 and Lanjian No. 2, 6.30 and 6.11 t DM/ha) had significantly higher hay yields than the local variety (333A, Longjian No. 1), with Lanjian No. 3 showing intermediate yields. Variety Lanjian No. 1 had the highest non-protein nitrogen concentration and in vitro digestibility, while Lanjian No. 2 had the highest starch concentration, fraction of intermediately degradable pectin and starch, and fraction of neutral detergent-soluble protein and acid detergent-soluble protein. Pearson correlation showed that hay yield was not significantly correlated with quality parameters. Principal component analysis showed that Lanjian No. 1 and Lanjian No. 2 received higher nutritive value. In summary, Lanjian No. 1 and Lanjian No. 2 had better potential use as high-protein feedstuffs for dairy cattle and sheep in the Jianghuai region based on yield, protein fractions, and in vitro digestibility.

1. Introduction

China is the largest lamb meat producer worldwide and was home to 194.0 million sheep and 132.2 million goats in 2022, which yielded 5.2 million tons of mutton and 14,648.5 tons of cashmere wool [1,2]. The rapid development of sheep production necessitates an increase in forage production and nutritive value, as it is a crucial material base for animal husbandry and constitutes the primary source of livestock nutrition. Although forage cultivation and processing are continuously improving in China, one of the primary challenges facing the forage industry is the insufficient availability of high-quality feedstuff [3,4], particularly with the steady rise in high-protein forage prices such as alfalfa hay [5]. Therefore, annual pasture legumes have been introduced to replace imported alfalfa hay to minimize dependency on traditional sources or imported hay, to maintain the feed supply throughout the year, and to potentially reduce the cost of raising animals while simultaneously increasing soil fertility through nitrogen (N) fixation [6,7].

Common vetch (Vicia sativa L.), a cool-season annual legume widely cultivated in southern Europe and western Asia, is a green manure crop and can be used as high-protein forage for ruminants [7,8,9]. Its yield is high, nutritious, mineral-rich, and easily digestible, as evidenced by an average dry matter (DM) yield of 6.06 t DM/ha, a crude protein (CP) concentration of 18.9%DM, a calcium concentration of 1.39%DM, in vitro organic matter (OM) digestibility (IVOMD) of 72.78% DM and metabolizable energy of 9.59 MJ/kg [7,10,11]. It has thus attracted huge interest as a high-protein resource for ruminants. Previous studies showed that hay yield and nutritive value are influenced by multiple factors in the farming system, i.e., variety, growing condition (e.g., climate and soil type), harvesting growth stage and postharvest handling methods [12,13], of which variety seems to be the most important production factor. Although there is research demonstrating the value of improving forage production and the quality of common vetch at various phenological stages [14], less attention has been given to the comparative analysis of old and new common vetch varieties concerning their contribution to the feedbase. To this end, Wang et al. [15] suggested that the challenge of identifying an optimal variety for a specific region is a common concern across various ecological and geographical landscapes.

To formulate ruminant diets, the Cornell Net Carbohydrate (CHO) and Protein System (CNCPS) is an integrated chemical analysis of plant cell components with the digestibility of the ruminant, based on the neutral and acid detergent systems for analyzing CHO and CP fractions [16]. The CHO and CP fractions are reliable indicators for accurately predicting the biological value and performance of ruminant feeds [13]. In addition, Huang et al. [13] and Marcos et al. [17] indicated that chemical composition combined with in vitro digestibility serve as the crucial parameters to assess the nutritive value of forage. Although claims of nutritional benefits of a variety of common vetch hay have been made, there is currently limited information available regarding the evolution of varietal differences in chemical composition, CHO and CP fractions, and in vitro digestibility of the whole hay at the flowering stage.

Considering the lack of information regarding a suitable variety of common vetch hay that comprehensively addresses these nutritional and digestibility aspects, this study was conducted to (1) evaluate the hay nutritive value from five common vetch varieties in terms of production, chemical composition, CHO and CP fractions, and in vitro digestibility; (2) identify the most suitable common vetch feed varieties by principal component analysis (PCA); (3) examine the relationship between hay yield and these nutritive values.

2. Materials and Methods

2.1. Bioethics Committee Approval

The Institutional Animal Care and Use Committee of Anhui Agricultural University has approved the procedures used in this study for collecting rumen liquors from sheep (permit number SYXK 2016-007).

2.2. Location, Experimental Design, and Sampling

This experiment was conducted at the National Agricultural Green Development Long-Term Fixed Observation Yingshang Test Station located in Yingshang City, Anhui Province, China (32°42′18″ N, 116°0′13″ E; altitude 20.7 m), which belongs to the Jianghuai region of China. The location features a transitional climate situated between the north temperate and subtropical areas. It has an average annual temperature of 15.0 °C, a frost-free period of 221 days, and an average annual rainfall of 904.6 mm [18]. The test soil was classified as shajiang black soil, with a pH, OM, available potassium and available phosphorus concentrations of 6.0, 89.3 g/kg, 1.66 mg/kg, and 17.2 mg/kg, respectively. The previous crop was corn (Zea mays).

Two local varieties of vetch, 333A and Longjian No. 1 (LOJ1), and three improved varieties, Lanjian No. 1 (LAJ1), Lanjian No. 2 (LAJ2), and Lanjian No. 3 (LAJ3), were sown in late October 2021 (

Table 1. Agronomic characteristics of the common vetch varieties utilized in this experiment [7].

Agronomic Characteristic	333A	Longjian No. 1	Lanjian No. 1	Lanjian No. 2	Lanjian No. 3
Days to mature (d)	134	115	145	132	124
1000 grain weight (g)	54	55	79	71	76
Plant height (cm)	92	100	106	80	69
Altitude (m.a.s.l) 1	–	–	<3000	<3500	<4000
Year of release	1987	1985	2014	2015	2011

¹—no reported.

). The improved varieties, LAJ1, LAJ2 and LAJ3, were bred by combining single plant selection with mixed selection. A completely randomized design was organized with four replicates per variety under the same agronomic conditions. Each plot was 5 × 7 m with a spacing of 20 cm between them and a total of 21 rows. Seeds were planted by hand at a rate of 150 viable seeds per square meter. No irrigation nor fertilization were applied during the growing season, and plots were diligently weeded by hand.

Plants were manually harvested using a sickle during the full-flowering stage in mid-April 2022. Plants from an inner plot of each variety with two representative plots (1.0 × 1.0 m²) were cut 5 cm above ground level and weighed for hay fresh yield. To determine the hay %DM, representative samples were collected and weighed before and after being air-dried for 4 days. Hay DM yield was calculated as hay fresh yield × hay %DM. Then, 500 g of forage samples were collected, oven-dried (65 °C for 48 h), and ground through a 1.0 mm sieve to determine the chemical composition and in vitro digestibility.

2.3. In Vitro Digestibility

The in vitro DM digestibility (IVDMD) and in vitro neutral detergent fiber digestibility (IVNDFD) of the hay samples were determined using an artificial rumen incubator, according to Huang et al. [14]. Ground samples (250 mg per bag, duplicated) were carefully placed into pre-weighed filter bags and subsequently heat-sealed. Rumen liquor was collected from two adult Huang-huai sheep rams fed a total mixed ration, 10 min after slaughter. Then, 400 mL of rumen liquor was passed through four layers of sterilized cheesecloth and intermittently agitated under CO₂ flushing in a water bath at 39 °C. Incubation jars were filled with 1596 mL pre-warmed (39 °C) buffer solution (266 mL solution A of CaCl₂·2H₂O 0.1 g/L, MgSO₄·7H₂O 0.5 g/L, NaCl 0.5 g/L, urea 0.5 g/L and KH₂PO₄ 10 g/L; and 1330 mL of solution B consisting of Na₂S·9H₂O 1.0 g/L and Na₂CO₃ 15.0 g/L), and a sample of rumen liquor was added to each. Then, 24 fetch bags along with three blank controls were incubated in each incubation jar. After 48 h of incubation at 39 °C, the bags were removed and gently washed in cold tap water, and the neutral detergent fiber (NDF) concentration was determined as detailed below. The replicated analyses were averaged by sample. The IVDMD (g/kg DM) and the IVNDFD (g/kg NDF) were calculated following the equations recommended by Huang et al. [14].

2.4. Laboratory Analysis

Determination of DM, N, ether extract (EE) ash, acid detergent fiber (ADF) and acid detergent lignin (ADL) was carried out according to the methods outlined by the Association of Official Agricultural Chemists [19]. The CP concentration was calculated by multiplying the N value by 6.25, and the NDF concentration was determined following the method described by Van Soest et al. [20] using heat-stable α-amylase and sodium sulfite. The concentrations of NDF and ADF were expressed inclusive of residual ash, while acid detergent-insoluble protein (ADIP), and neutral detergent-insoluble protein (NDIP) were measured using Kjeldahl analysis of the ADF and NDF bag residues, respectively, as described by Licitra et al. [21]. The condensed tannins (CT) in the hay were measured using the butanol-HCl method as described by Zhang et al. [22], with purified CT serving as the standard.

The CHO fractions were analyzed as proposed by the CNCPS [23]. The system divides CHO in terms of degradation rate into four fractions: namely, C_A, rapidly degradable sugars, CB₁, intermediately degradable pectin and starch, CB₂, slowly degradable cell wall and C_C, and undegradable/lignin-bound cell wall. The total CHO (TCHO, %DM) was calculated as TCHO = 100 − (CP + EE + Ash). Structural carbohydrates (SC) and non-structural carbohydrates (NSC) were calculated according to the equations of Caballero et al. [24]: SC = NDF − NDIP and NSC = TCHO − SC. Starch concentration was determined through the enzymatic hydrolysis of α-linked glucose polymers [13].

The CP fractions of samples were divided into five fractions according to the CNCPS [21]. These fractions included non-protein CP (fraction P_A) (NPCP) and were measured with sulfuric acid and sodium tungstate and calculated as the difference between total CP and true protein CP; buffer soluble protein (fraction P_B1) was measured by subtracting the buffer-insoluble protein precipitated with freshly prepared (1 g/10 mL) sodium azide and borate-phosphate buffer (pH 6.7–6.8) solution from true protein; neutral detergent-soluble protein (fraction P_B2) was calculated by subtracting NDIP from the buffer-insoluble protein; acid detergent soluble protein (fraction P_B3) was calculated by subtracting ADIP from NDIP; and ADIP as indigestible protein (fraction P_C). All measurements were performed in duplicate, and appropriate chemical standards were included in each analytical run [16].

2.5. Statistical Analyses

Analysis of variance was conducted using the general linear model procedure in SPSS software (Version 21.0 IBM Corporation, Armonk, NY, USA). The effect of variety on all variables was analyzed with the statistical linear model: Yi = M + Vi + R, where Yi is the dependent variable, M is the mean value, Vi is the fixed effect of variety, and R is the random effect. The mean values were compared using the Duncan significant difference test at a significance level of p ≤ 0.05. A tendency was defined as 0.05 < p ≤ 0.10.

Pearson’s correlation was utilized to assess the relationships between common vetch yields and nutritive value (i.e., CP, NDF, ADL, IVDMD and IVNDFD). The PCA of the nutritive value of common vetch hay was performed to quantify the contribution of each constituent to the variation in nutritive value. The CP, NDF, ADL and in vitro digestibility were the primary factors influencing the hay’s nutritive value, and they were used in the PCA. The sign and magnitude of the eigenvectors were determined to assess their relevance in explaining nutritive value. All eigenvectors were normalized to have unit variance [25]. The varieties were ranked based on principal component scores derived from the relevant eigenvectors.

3. Results

3.1. Hay Yield

Significant differences (p = 0.038) were observed in hay yield between the five common vetch varieties (Figure 1). The hay yield was highest in the LAJ1 and LAJ2 varieties and lowest in the 333A and LOJ1 varieties, with LAJ3 showing intermediate yields.

3.2. Chemical Composition

The CP concentration was not affected by vetch variety (p = 0.084;

Table 2. Mean values of chemical composition and standard error of the mean (% dry matter (DM unless otherwise specified)) for hay from five common vetch varieties (n = 20).

Nutrients	333A	Longjian No. 1	Lanjian No. 1	Lanjian No. 2	Lanjian No. 3	p-Value
Ash	7.83 ± 0.06 b	8.08 ± 0.04 ab	7.32 ± 0.11 b	9.00 ± 0.13 a	8.01 ± 0.19 ab	0.049
Crude protein	20.74 ± 0.11	23.32 ± 0.12	22.57 ± 0.17	24.20 ±0.20	22.50 ± 0.54	0.084
Ether extract	1.19 ± 0.04 bc	1.45 ± 0.04 ab	1.29 ± 0.02 bc	1.16 ± 0.03 c	1.60 ± 0.02 a	0.014
Neutral detergent fiber	32.18 ± 0.57	35.19 ± 0.58	30.19 ± 0.44	31.89 ± 0.23	32.16 ± 0.25	0.058
Acid detergent fiber	20.56 ± 0.35	21.88 ± 0.37	21.25 ± 0.28	22.72 ± 0.19	22.59 ± 0.36	0.366
Acid detergent lignin	7.25 ± 0.48	6.36 ± 0.45	5.60 ± 0.47	4.37 ± 0.04	4.89 ± 0.08	0.188
Hemicellulose	11.72 ± 0.33 ab	13.35 ± 0.38 a	9.19 ± 0.31 bc	10.48 ± 0.11 bc	8.52 ± 0.40 c	0.010
Cellulose	13.32 ± 0.29 b	15.25 ± 0.29 ab	15.64 ± 0.41 ab	18.35 ± 0.21 a	18.47 ± 0.64 a	0.013
Phosphorus (mg/g DM)	0.42 ± 0.005 b	0.47 ± 0.006 a	0.40 ± 0.001 b	0.42 ± 0.005 b	0.47 ± 0.005 a	<0.001
Calcium	0.37 ± 0.01	0.48 ± 0.01	0.48 ± 0.04	0.51 ± 0.06	0.59 ± 0.08	0.828
Condensed tannins (mg/g DM)	2.97 ± 0.25 b	2.53 ± 0.14 b	4.87 ± 0.19 a	5.81 ± 0.02 a	5.50 ± 0.18 a	<0.001

^a–c Means with different letters in the same row are statistically different.

), with an average value of 22.6%DM. The EE concentration was highest (p = 0.014) in LAJ3 and lowest in LAJ2. The ash concentration of 333A was similar to those of LAJ1 and LAJ2, and significantly lower (p = 0.049) than that of LAJ2, but LOJ1 had an ash concentration similar to those of the improved varieties.

The cell wall (NDF, ADF and ADL) concentrations were not significantly different in all varieties. The hemicellulose concentration was significantly different (p = 0.010) among varieties, being the highest in LOJ1 and the lowest in LAJ3. The cellulose concentration was clearly lower (p = 0.013) in 333A than in LAJ2 and LAJ3, but similar to those of the other varieties. The phosphorus concentration was lower (p < 0.001) in the LOJ1 and LAJ3 varieties compared with the other varieties. The calcium concentration was not affected by variety (p > 0.05), with an average value of 0.49%DM. The concentration of CT was significantly lower (p < 0.001) in 333A and LOJ1 compared with the improved varieties.

3.3. CHO and CP Fractions

Except for P_B2, P_C, SCP and ADIP, protein fractions differed significantly (p < 0.05;

Table 3. Carbohydrate and protein fractions and standard error of the mean for hay from five common vetch varieties (n = 20).

Items	333A	Longjian No. 1	Lanjian No. 1	Lanjian No. 2	Lanjian No. 3	p-Value
Carbohydrates (% dry matter)
TCHO	70.24 ± 0.11 a	67.14 ± 0.08 ab	68.82 ± 0.17 a	65.64 ± 0.28 b	67.89 ± 0.66 ab	0.043
SC	27.35 ± 0.55 ab	29.93 ± 0.50 a	24.06 ± 0.35 b	25.01 ± 0.17 b	27.15 ± 0.32 ab	0.012
NSC	42.89 ± 0.48 a	37.21 ± 0.46 b	44.75 ± 0.34 a	40.64 ± 0.37 ab	40.74 ± 0.70 ab	0.011
Carbohydrate fractions (% CHO)
CA 1	56.06 ± 1.10 b	54.07 ± 1.17 b	62.26 ± 0.43 a	55.83 ± 0.24 b	57.81 ± 0.58 ab	0.044
CB1	5.02 ± 0.42 a	1.35 ± 0.48 b	2.78 ± 0.07 b	6.05 ± 0.19 a	2.14 ± 0.04 b	<0.001
CB2	14.21 ± 0.89	21.79 ± 0.85	15.38 ± 1.69	22.13 ± 0.16	22.72 ± 0.71	0.095
CC	24.72 ± 1.62	22.79 ± 1.52	19.58 ± 1.67	15.99 ± 0.20	17.33 ± 0.31	0.265
Protein fractions (% CP)
PA	9.41 ± 0.66 b	14.00 ± 0.65 ab	19.18 ± 1.33 a	17.95 ± 0.25 a	14.72 ± 0.62 ab	0.040
PB1	34.24 ± 0.19 a	25.15 ± 0.10 b	19.77 ± 0.80 b	19.93 ± 0.89 b	23.41 ± 0.21 b	<0.001
PB2	33.08 ± 0.99	38.34 ± 0.99	33.79 ± 0.95	33.55 ± 0.61	39.45 ± 0.81	0.134
PB3	13.16 ± 1.16 b	13.91 ± 0.64 b	18.47 ± 0.53 ab	22.56 ± 1.11 a	14.81 ± 0.37 b	0.030
PC	10.11 ± 0.70	8.60 ± 0.19	8.80 ± 0.24	6.02 ± 0.11	7.61 ± 0.40	0.141
PB2 + PB3	46.24 ± 0.61 b	52.25 ± 0.39 ab	52.25 ± 0.64 ab	56.11 ± 0.90 a	54.26 ± 0.63 a	0.010
NPCP (%DM)	1.94 ± 0.13 b	3.21 ± 0.13 ab	4.35 ± 0.31 a	4.35 ± 0.08 a	3.37 ± 0.20 ab	0.020
SCP (%DM)	9.04 ± 0.09	9.15 ± 0.08	8.80 ± 0.17	9.18 ± 0.27	8.62 ± 0.29	0.945
NDIP (%DM)	4.82 ± 0.11 c	5.25 ± 0.08 bc	6.13 ± 0.12 ab	6.88 ± 0.22 a	5.02 ± 0.14 bc	0.007
ADIP (%DM)	2.09 ± 0.14	2.03 ± 0.05	1.98 ± 0.04	1.46 ± 0.03	1.71 ± 0.10	0.361
Starch (%DM)	3.53 ± 0.30 a	0.91 ± 0.34 b	1.91 ± 0.05 b	3.97 ± 0.13 a	1.45 ± 0.03 b	<0.001

^a–c Means with different letters in the same row are statistically different. ¹ C_A, rapidly degradable sugars; C_B1, intermediately degradable pectin and starch; C_B2, slowly degradable cell wall; C_C, unavailable/lignin-bound cell wall; CP, crude protein; TCHO, total carbohydrates; NPCP, non-protein crude protein; NSC, non-structural carbohydrates; P_A, non-protein nitrogen; P_B1, buffer-soluble protein; P_B2, neutral detergent-soluble protein; P_B3, acid detergent-soluble protein; P_C, indigestible protein; SC, structural carbohydrates; SEM, standard error of the mean.

) among the common vetch varieties. The P_A and NPCP concentrations of 333A were similar to those of LOJ1 and LAJ3, but significantly lower (p < 0.05) in the other two varieties. The PB1 concentration was significantly higher (p < 0.001) in 333A than in the other varieties, with no difference between the other varieties. The P_B3 concentration of 333A and LOJ1 did not differ from those of LAJ1 and LAJ3 but was lower (p = 0.030) than that of LAJ2. The P_B2 + P_B3 concentration was significantly lower (p = 0.010) in 333A than in LAJ2 and LAJ3, with the P_B2 + P_B3 concentration in LOJ1 and LAJ1 being intermediate. The variation in NDIP concentration was the highest in LAJ2 and the lowest in 333A (p = 0.007).

Significant differences (p < 0.05) among the common vetch varieties were found for carbohydrate fractions except C_B2 and C_C. The TCHO concentration of 333A was similar to those of LOJ1, LAJ1 and LAJ3, and significantly higher (p =0.043) than that of LAJ2. The LOJ1 variety had the same SC concentration as LAJ3 but a higher concentration (p < 0.05) compared to LAJ1 and LAJ2. The NSC concentration of LOJ1 was not significantly different from those of LAJ2 and LAJ3, but lower (p = 0.011) compared to that of LAJ1. The TCHO concentration of LOJ1 was similar (p > 0.05) to those of the improved varieties, and the SC and NSC concentrations of 333A were similar (p > 0.05) to those of the improved varieties. The C_A fraction of 333A and LOJ1 was significantly higher (p = 0.044) compared to that of LAJ1, but similar to those of LAJ2 and LAJ3. The C_B1 fraction of 333A was similar to that of LAJ2 but higher (p < 0.001) than those of LAJ1 and LAJ3, while the C_B1 fraction of LOJ1 was significantly higher (p < 0.001) than that of LAJ2, but similar to those of LAJ1 and LAJ3.

3.4. In Vitro Digestibility

The IVDMD concentration of 333A and LOJ1 was clearly lower (p = 0.004; Figure 2) than those of LAJ1 and LAJ2, but similar to that of LAJ3. In contrast, the IVNDFD concentration was the highest in LAJ1 and the lowest in 333A (p =0.039).

3.5. Pearson Correlations and Principal Component Analysis

Pearson correlations between the hay yield and nutritive value (i.e., CP, NDF, ADL, IVDMD and IVNDFD) are shown in Figure 3. The hay yield was not significantly correlated with CP (r = 0.210; p = 0.374), NDF (r = −0.261; p = 0.266), and ADL (r = −0.112; p = 0.638), but tended to have a positive correlation with IVDMD (r = 0.423; p = 0.063) or IVNDFD (r = 0.393; p = 0.087).

The results of the PCA generated five principal components: PC1, PC2, PC3, PC4, and PC5. The PC1, PC2, and PC3 collectively explained 87.76% of the variation (

Table 4. Principal component analysis of the hay nutritive value derived from various common vetch varieties.

Item	PC1 1	PC2	PC3
Eigenvalues	2.50	0.984	0.904
Variance (%)	50.00	19.68	18.08
Eigenvectors
Crude protein (%DM)	0.311	0.877	0.277
Neutral detergent fiber (%DM)	−0.639	−0.015	0.701
Acid detergent fiber (%DM)	−0.681	−0.264	0.203
IVDMD (%DM)	0.954	−0.238	0.081
IVNDFD (%NDF)	0.789	−0.298	0.536

¹ PC: principal component; IVDMD, in vitro dry matter digestibility; IVNDFD, in vitro neutral detergent fiber digestibility.

). An examination of the eigenvectors showed that PC1 (50% of the variance) provided the best explanation for the nutritional value of hay. The eigenvectors of the soluble constituents of CP, IVDMD, and IVNDFD were all positive, indicating that these variables contributed positively to the hay’s nutritive value. The eigenvectors of the insoluble constituents of NDF and ADL were all negative, indicating that these variables had a negative effect on the hay’s nutritive value. The eigenvectors from PC1 showed that the highest magnitude was IVDMD with 0.954, followed by IVNDFD, NDF and ADL with 0.789, 0.681 and 0.639, and followed by CP with 0.3114. Accordingly, the eigenvectors of PC1 were used to rank the varieties by nutritive value. The ranking of the varieties in terms of nutritive value from best to worst was as follows: LAJ1 (105.60) > LAJ2 (102.06) > LAJ3 (94.13) > LOJ1 (88.70) > 333A (88.29).

4. Discussion

4.1. Hay Yield

The present results showed significant varietal differences in hay yields, which could be partly due to differences in days to the flowering stage, seedling robustness, cold resistance, winter and spring growth, plant height, DM accumulation from germination to flowering, and leaf–stem ratio [13,26,27]. These findings were consistent with results previously reported for common vetch, grasspea (Lathyrus sativus L.) and narbon vetch (Vicia narbonensis L.) in northwestern Syria [26,28]).

In the present study, the hay yield obtained (4.82–6.30 t DM/ha) was in agreement with values reported by Rahmati et al. [11] (6.06 t DM/ha), but higher than the ranges reported for common vetch (2.25 –3.38 t DM/ha) by Larbi et al. [26]. Hay yield values vary among the studies due to the differences in the varieties, agronomic practices and growing conditions (e.g., climate and soil type) [13,29]). Accordingly, the LAJ1 and LAJ2 varieties were identified as having higher hay yield, making them attractive as an alfalfa alternative for ruminants.

4.2. Chemical Composition

Greater CP and lesser cell wall (NDF, ADF, ADL and cellulose) concentrations in forage are generally considered indicators of better forage quality [13,23]. In the present study, the CP concentrations of stover from common vetch varieties exceeded 20.0%DM, where the average CP concentration requirements reported for dairy cattle were 19.8% and 17.4% for sheep [30]. For this standard, the sole provision of hay from these varieties will meet the CP demands of dairy cattle and sheep. Accordingly, all the varieties tested were identified as potential sources of CP supplementation, primarily comprising low-quality cereal straw for ruminants in smallholder crop-livestock systems. In this study, the NDF, ADF, cellulose and ADL concentrations of the hay were in the ranges of 30.19–35.19, 20.56–22.72, 13.32–18.47 and 4.37–7.25%DM, respectively. Larbi et al. [28] reported mean NDF and ADF concentrations of 33.7 and 25.9%DM for hay of different common vetch accessions at the 50% bloom stage, which is comparable to the current results. However, the CP concentrations (18.7%DM) found in their study were lower than in the current results. Huang et al. [7] recorded lower CP (18.93%DM) and higher NDF, ADF, and cellulose (39.03, 26.73, and 20.55%DM) and similar EE and ADL (1.51 and 6.17%DM) concentrations in the hay of three improved and local common vetch varieties at the full-bloom stage compared to this study. These inconsistent findings between studies may be related to varying hay NDF concentrations depending on the crop cycle, night temperatures and CHO levels, as well as growing conditions (e.g., soil type and climate) [13,23]. The present results showed that the CT of the three improved varieties were higher than those of the two local varieties, indicating that these improved varieties can reduce methane production in the rumen, as CT influences enteric methane emissions and protein digestion positively [31,32].

4.3. Carbohydrate and Protein Fractions

Protein is one of the limiting nutrients in most legume forage [23], of which NPN could be enhanced to maximize the efficiency of microbial protein synthesis [33]. The LAJ1 and LAJ2 varieties had higher NPN concentrations in hay, suggesting that these varieties can improve microbial protein synthesis efficiency. Furthermore, P_B2 and P_B3 fractions are indicative of the bypass protein within forage, while the P_c fraction is indicative of the non-degraded fraction [34]. In the present study, the PB2 fraction was the largest and similar to findings from sorghum stover [23] and crop straw [16]. The LAJ2 variety hay had higher NPN and P_B2 + P_B3 fractions, although the P_c was numerically minimal, indicating that the variety could serve as a superior CP source for ruminants. Data on the concentrations of CP fractions in common vetch hay are limited; however, the pattern of CP fractions found in the present study is similar to reports on legume crops, such as lucerne (Medicago Sativa L.) [35] and black gram (Vignamungo (L.) Hepper) [34].

Higher NSC fractions are indicative of a higher digestible fraction of total CHOs [36]. Accordingly, the 333A and LAJ1 varieties had higher potential herbage and a higher digestible fraction of carbohydrates. The variation noted in TCHO, SC, NSC, C_A and C_B1 fractions of the hay from common vetch varieties may be due to differences in days to flowering or leaf–stem ratio. In addition, the accumulation and lignification of cell wall concentrations from germination to flowering could also contribute to these variations [26]. Among the CHO fractions, the C_A fraction was the highest in the straw of the common vetch varieties analyzed, and variety LAJ1 had a higher concentration of rapidly degradable sugars. Studies on the pattern of CHO fractions in common vetch hays are partial, however, other studies reported similar results for CHO fraction patterns in other legume forages, such as cowpea and berseem [36].

4.4. In Vitro Digestibility

Higher NDF digestibility of forages is preferred since it improves animal performance [37]. The substantial variability differences in hay IVDMD and IVNDFD observed in this study are consistent with previous findings in common vetch [8]. Variable responses of in vitro digestibility among common vetch varieties could be explained by variable levels of hay cell wall concentrations and anti-nutritive activity (CT), or by variations in the proportion of plant fractions across common vetch varieties [37]. Consistent with results from the present study, Larbi et al. [28] reported that variety had a significant effect on the in vitro digestibility of common vetch, narbon vetch and grasspea. The LAJ1 variety showed IVDMD and IVNDFD values as high as 84.40%DM and 52.19%NDF, suggesting that feeding ruminants LAJ1 hay may result in increased growth performance. Common vetch hay IVDMD (81.4~83.8% DM) concentrations reported by Huang et al. [8] are comparable with the current results; however, the reported IVNDFD concentrations were lower than in this study.

4.5. Pearson Correlations and Principal Component Analysis

Pearson correlation analysis showed a lack of significant correlation between hay yields and the CP, NDF, ADL, IVDMD and IVNDFD of hay. These observations indicate a superior opportunity to breed improved common vetch, without impacting nutrient quality character, while potentially enhancing nutritive value potentially at the same time. In line with the current findings, a lack of or weak correlation between stover yields and the nutritive value of the stover was found in faba bean [38] and maize varieties [39].

Principal component analysis was applied for an unbiased ranking of varieties from best to worst, although the analytical procedure was not mathematically precise [38,40]. The fact that PC1 explained more than 50% of the variation points to the nutritional differences between varieties, and it allowed for an objective ranking of varieties based on hay nutritive value from their component scores. The top two varieties in the ranking were LAJ1 and LAJ2 based on hay nutritional quality. Similarly, PC scores were utilized to rank the nutritive value of forage of faba bean varieties [38] and oat varieties [15]. Furthermore, the IVDMD had the highest positive eigenvector in PC1. Similarly, Wamatu et al. [25] demonstrated that IVOMD is indicative of the nutritive value of lentil stover. Thus, common vetch breeders may utilize IVDMD as a criterion for optimization in breeding hay varieties with superior nutritive value.

5. Conclusions

Results of this study showed significant variations in hay yield and nutritive value between different common vetch varieties and that these differences are of importance to ruminant productivity. A lack of significant correlation was shown between the hay yields and the nutritive value of hay. Principal component analysis showed that the IVDMD was the highest positive eigenvector and the top two varieties in the ranking were Lanjian No. 1 and Lanjian No. 2 based on principal component scores. Further studies are necessary to increase the number of common vetch varieties and samples in the study to improve the breeding progress of common vetch varieties. Meanwhile, feeding hay of Lanjian No. 1 and Lanjian No. 2 and investigating their contribution to the differences in the growth performance and meat quality of ruminants should be considered. From all the varieties tested, the Lanjian No. 1 and Lanjian No. 2 varieties showed the best potential as ruminant protein fodder for dairy cattle and sheep, in terms of hay, protein fractions, condensed tannins and in vitro digestibility.