Comprehensive Evaluation and Screening of Autumn-Sown Oat (Avena sativa L.) Germplasm in Different Agropastoral Regions

Abstract

In light of current global challenges of climate change, the over-exploitation of natural resources, and increasing demand for livestock products, the exploration of excellent forage crop resources holds great potential for development. Therefore, selecting forage crops that are high-yield, high-quality, and have excellent resistance to pests and diseases can greatly promote the development of the livestock industry. Oat (Avena sativa L.), a dual-purpose crop for grain and forage, plays a vital role in the development of animal husbandry. Autumn-sown oats have a significantly longer growth cycle than spring-sown oats, ensuring a year-round forage supply and achieving higher yields. The agropastoral transitional zone in southwest China is a key region for autumn-sown oats, but the systematic evaluation of oat germplasm there is still limited. Therefore, we conducted a two-growing-season (2022–2024) field experiment across four locations to evaluate nine oat genotypes for growth phenotypes, forage productivity, and nutritional quality through 11 agronomic traits and nutritional parameters during the filling stage (Zadok’s 75). The results revealed the following: (1) agronomic performance: dry matter yield (DMY) ranged from 10.72 to 14.58 t/ha, with line ‘WC109’ achieving the highest DMY (14.58 t/ha) and crude protein (CP, 9.66%); (2) nutritional quality: CP exhibited a significant negative correlation with fiber content (NDF: r = −0.72, p < 0.01; ADF: r = −0.68, p < 0.01), highlighting a yield–quality trade-off; ‘WC109’ demonstrated superior forage value, with the highest relative forage quality (RFQ: 115.45) and grading index (GI: 19.30); (3) environmental adaptation: location-specific climatic conditions significantly influenced productivity, with Wenjiang (WJ) showing optimal performance due to favorable temperature and precipitation. These results position ‘WC109’ as a promising candidate for autumn-sown cultivation in southwest China, addressing winter forage shortages while enhancing livestock nutrition. Our findings further elucidate the mechanisms linking yield and feeding value to growth performance indicators, providing references for trait-based measures to enhance forage oat productivity and quality.

1. Introduction

With the growth in China’s population and the continuous upgrading of residents’ dietary structures, the demand for protein-rich foods is steadily increasing [1]. Livestock products, as a primary source of protein, play a crucial role in ensuring dietary health. Animal husbandry can effectively convert non-edible crops, such as fodder, into human food [2,3]. However, to meet the demand for livestock products, overgrazing and unreasonable development have led to the deterioration of the ecological environment of natural grasslands [4,5,6]. In recent years, the Chinese government has attached great importance to grassland restoration and environmental protection, recognizing that only by developing forage cultivation and animal husbandry can the contradiction between human and livestock needs for grain be effectively alleviated, the safety of the ecological environment be maintained, and the transformation and upgrading of the livestock industry be promoted [7,8,9].

With the structural adjustment of agriculture and animal husbandry, the promotion and application of grain to feed, the rapid reform of the agricultural supply chain, and the implementation of the rural revitalization strategy, China’s livestock enterprises have undergone radical changes, and animal husbandry occupies a large proportion of economic production [10]. In China’s southwestern region, the conventional agricultural practice involves sowing in late spring and harvesting in early autumn during the frost-free period (90–125 d), with subsequent fallow periods in late autumn and winter [11]. However, research indicates that cultivating fodder crops during the fallow season can significantly enhance soil nutrients [12,13]. In this region, the seasonal imbalance of forage supply—especially the feed shortage during the winter and spring seasons (from November to April of the following year)—has become a key bottleneck restricting the sustainable development of the livestock industry (https://www.gov.cn/gzdt/2010-02/21/content_1538234.htm, accessed on 21 February 2010). The scarcity of winter forage has consistently hindered the development of the livestock sector; thus, by optimizing the agricultural planting structure and increasing the area of winter forage cultivation, the issue of supply shortage can be effectively alleviated [14,15]. Moreover, selecting the appropriate forage varieties is of paramount importance. Oat (Avena sativa L.), as an annual crop that serves both as grain and fodder, is a principal source of feed for livestock globally and is also a significant type of cold-season grassland forage [16,17,18]. Oats are known for their wide adaptability, robust resistance to adverse conditions, and short growth cycle [19,20]. Appreciated for their excellent palatability and high nutritional value, oats are cultivated and utilized extensively both within China and abroad. The market demand for oats continues to rise due to their high nutritional value, good palatability, and easy digestion and absorption (https://www.renrendoc.com/paper/400944282.html, accessed on 23 March 2025). Furthermore, oat planting techniques are being continuously improved, resulting in significant increases in yield and quality. Meanwhile, oat processing technologies are also being refined, leading to a growing variety of products and enhanced market competitiveness (https://max.book118.com/html/2025/0219/5010132140012102.shtm, accessed on 17 April 2025).

China, a major player in the global animal husbandry sector and a traditional region for oat cultivation, confronts unique challenges regarding oat variety adaptation, yield, and quality due to its diverse climate and intricate topography. Although preliminary screening of oat varieties has been conducted [21,22], the existing varieties are often unable to meet the urgent demand for autumn-sown high-quality forage in agricultural and pastoral areas. In this study, we conducted a two-year trial of high-performing lines based on the preliminary screening. We measured growth characteristics, forage yield, and feeding value. The goals of the trial were as follows: (1) to comprehensively assess the adaptability, yield, and nutritional value of different oat varieties/lines; (2) to identify high- and stable-yield forage oats and evaluated their forage potential; and (3) to provide superior germplasm for autumn sowing in agricultural and pastoral areas of China to promote the sustainable development of the local livestock industry.

2. Materials and Methods

2.1. Plant Materials

Nine oat genotypes were tested, including ‘Longyan 3’ and ‘Intimidator’, which are widely promoted forage oat varieties in China and were used as the control group. The remaining seven lines (‘WC109’, ‘WC130’, ‘WC179’, ‘WC283’, ‘WC286’, ‘WC291’, ‘WC299’) were selected from Sichuan Agricultural University based on previous screening efforts in autumn-sown areas of China [23]. The growth period of each genotype in different environments is shown in

Table 1. Growth periods of nine oat genotypes in four locations (WJ, CZ, GY, and XC) from 2022 to 2024.

Genotypes	Growth Stage (d)
	2023WJ	2023CZ	2023GY	2023XC	2024WJ	2024CZ	2024GY	2024XC
Longyan 3	178	172	180	166	180	176	185	162
Intimidator	172	170	183	167	176	172	183	170
WC109	160	156	171	154	166	167	176	161
WC130	176	170	182	162	175	171	179	167
WC179	170	169	180	161	174	166	181	169
WC283	164	161	179	159	169	161	173	158
WC286	176	173	188	170	177	170	182	167
WC291	169	164	174	161	171	164	177	166
WC299	166	159	169	154	170	162	172	157

. Based on our initial screening, all the selected genotypes showed high disease resistance potential and good lodging resistance.

2.2. Experimental Site and Design

The trial was conducted during the 2022–2023 and 2023–2024 growing seasons at four different locations. Wenjiang (WJ) (103°51′ E, 30°43′ N), Chongzhou (CZ) (103°38′ E, 30°32′ N), and Guangyuan (GY) (105°86′ E, 32°43′ N) have typical subtropical humid monsoon climates, with elevations of 507, 514, and 1030 m; the total amounts of annual precipitation are 966, 1162, and 1136 mm; and the annual mean temperatures are 16.3, 16.7, and 16.1 °C, respectively. Xichang (XC) (102°26′ E, 27°89′ N) has a typical subtropical plateau monsoon climate, with an altitude of 1500 m, a total annual precipitation of 873 mm, and an annual mean temperature of 17.6 °C. The minimum temperature, maximum temperature, altitude, and monthly precipitation from seeding to harvest at each location are depicted in Figure 1. Before sowing, soil samples from 0 to 20 cm were collected using a five-point sampling method (

Table 2. Determination of soil nutrient content in 0–20 cm before sowing in four locations (WJ, CZ, GY, and XC) during the 2022–2024 growing season.

Soil Nutrient	2023WJ	2023CZ	2023GY	2023XC	2024WJ	2024CZ	2024GY	2024XC
Origin matter (g/kg)	32.31	28.54	29.20	28.40	32.49	28.62	29.31	28.44
Total nitrogen (g/kg)	1.56	1.48	1.49	1.45	1.62	1.48	1.54	1.46
Available nitrogen (mg/kg)	105.88	100.29	98.28	97.86	106.03	100.30	98.33	97.93
Available phosphorus (mg/kg)	15.53	12.24	12.01	11.46	15.57	12.31	12.03	11.55
Available potassium (mg/kg)	148.32	150.26	119.37	115.53	148.33	150.34	119.42	115.55
pH	6.48	7.12	6.84	7.43	6.46	7.14	6.83	7.45

). The determination of soil nutrients was conducted in accordance with the methods of soil agrochemical analysis (https://www.docin.com/p-2219784431.html, accessed on 12 June 2019).

The experiment was conducted using a completely randomized block design with four replications. Seeds were sown by hand on 17–24 October, during the 2022–2024 cropping seasons. The individual plots were 3 m × 5 m with a 0.7 m pass way. The seeds were planted by the strip method with a seeding rate of 2.4 million plants per hectare, plots were sown with 10 rows at a spacing of 0.3 m, and the sowing depth was 4 cm. Before sowing, the soil was prepared with a chisel plow and a disk harrow. A week before sowing, based on local fertilization recommendations, all plots were treated with a compound fertilizer (N:P₂O₅:K₂O = 15:15:15), at a rate of 80 kg/ha. During the experimental period, other field management practices adhered to local protocols. No additional fertilization or irrigation was applied. All test plots were kept weed-free by manual hoeing. Chemicals were used to control pests and diseases when necessary.

2.3. Measurement Index and Methods

2.3.1. Agronomic Characteristics

All agronomic characteristics were assessed at the filling stage (Zadok’s 75) [24]. The plant height (PH), tiller number (TN), and stem–leaf ratio (SLR) were measured using 10 plants randomly selected from the center of each plot. Ten plants were randomly selected from the central part of each plot to measure the PH, TN, and SLR. The PH was measured using a tape measure from the base to the top of the plant. The TN was counted directly. After measuring the PH and TN, the plants were cut at ground level, and their stems and leaves were separated. The dry weights of the stems and leaves were calculated after drying to obtain the SLR. The yield was determined by manually harvesting a 9.60 m² area of forage (Zadok’s 75) from the central region of each 15 m² plot, cutting at a height of 5 cm above the soil surface, and promptly weighing the biomass. Thereafter, approximately 2 kg of fresh tissue samples were collected from each plot, and the samples were placed in paper bags and oven-dried at 105 °C for 30 min to terminate physiological processes, then re-oven-dried at 65 °C to constant weight, which was used to determine the dry matter yield (DMY). Subsequently, the desiccated samples were individually pulverized, sifted, and sealed in paper bags for subsequent analysis.

2.3.2. Nutritional Indices

In order to assess the nutritional value of different genotypes of forage oats, the neutral detergent fiber (NDF) and acid detergent fiber (ADF) were estimated by the sequential method [25], the crude protein (CP) was determined using the Kjeldahl method [26], the crude fat (EE) was determined using the Soxhlet extraction method [27], and the crude ash (ASH) was measured using the Mafu furnace burning method [28]. The values of each trait were standardized as a percentage of the DMY, with the mean values being employed for further analysis.

Indicators for evaluating the nutritional value of forages include relative forage quality (RFQ) and grading index (GI), which were calculated using the following formulas [29,30]:

RFQ = TDN × DMI

(1)

TDN = 82.38 − 0.7515 × ADF

(2)

DMI = 120 ÷ NDF

(3)

GI = NEL × VDMI × CP ÷ NDF

(4)

NEL = (1.004 − 0.0124 × ADF) × 9.29

(5)

VDMI = 1.2 × BW × NDF

(6)

where TDN was total digestible nutrients (%), DMI was dry matter intake (%), NEL was net energy for lactation (MJ/kg), VDMI was voluntary dry matter intake (kg/d), and BW was the body weight of an average cow, calculated as 600 kg.

2.4. Statistical Analysis

Agronomic characteristics and nutritional indices were initially collated using Microsoft Excel 2021 and subjected to statistical analysis using IBM SPSS Statistics 28 (USA) software. All data were subjected to one-way and three-way (genotype × year × location) analysis of variance (ANOVA) and presented as mean ± standard deviation. Significant differences among means were determined using the least significant difference (LSD) test at a 5% significance threshold. Correlation analyses between traits were performed in Origin 2021 and plotted using GraphPad Prism 8. In order to more clearly demonstrate the direct and indirect effects among multiple variables, we conducted a structural equation model (SEM) analysis using the data collected from eight experimental sites, nine materials, and four replicates for each material. This analysis aimed to reflect how growth performance and nutritional composition influence forage yield and feeding value. The SEM was constructed by using the ‘Lavaan’ function in the R package (v4.4.2, R Core Team, 2024). Subsequently, we simplified the model by eliminating nonsignificant paths and selecting the model with a chi-square-to-degrees of freedom ratio of less than 2.0, root mean square error of approximation of less than 0.05, comparative fit index of greater than 0.9, and p-value of greater than 0.05. The AMMI model analysis was conducted using the ‘agricolae’ package in R.

3. Results

3.1. Agronomic Performance

In this study, we evaluated nine oat genotypes sown in the autumns of 2022 and 2023 for key agronomic traits including DMY, PH, TN, and SLR. Our findings indicated that all assessed agronomic traits were significantly influenced by the factors of G, Y, L, G × Y, G × L, Y × L, and G × Y × L (except for TN) (

Table 3. Dry matter yield mean values of different forage oat genotypes tested at four locations (WJ, CZ, GY, and XC) from 2022 to 2024.

Genotypes	DMY(t/ha)								Combined
	2023WJ	2023CZ	2023GY	2023XC	2024WJ	2024CZ	2024GY	2024XC
Longyan 3	14.77 ± 1.15 bc	11.69 ± 1.38 bc	12.37 ± 0.93 bc	10.98 ± 0.84 ab	17.56 ± 0.76 abc	12.62 ± 0.50 bc	13.01 ± 0.61 bc	12.97 ± 0.65 abc	13.25
Intimidator	15.82 ± 0.82 ab	11.88 ± 1.02 bc	13.87 ± 1.03 a	10.73 ± 0.69 abc	18.06 ± 0.41 a	12.51 ± 0.53 bc	13.17 ± 0.67 b	13.16 ± 0.66 ab	13.65
WC109	16.85 ± 0.66 a	14.20 ± 0.60 a	13.96 ± 0.63 a	11.21 ± 0.69 a	17.90 ± 0.32 ab	14.94 ± 0.50 a	14.47 ± 0.81 a	13.84 ± 1.02 a	14.58
WC130	13.17 ± 0.82 d	10.77 ± 1.07 cd	11.87 ± 0.96 bc	10.04 ± 1.38 abcd	15.27 ± 0.62 d	10.85 ± 0.47 de	12.96 ± 0.64 bc	10.88 ± 0.13 ef	11.98
WC179	13.01 ± 1.05 d	10.87 ± 0.67 bcd	12.46 ± 0.89 bc	10.84 ± 0.78 abc	16.41 ± 0.89 c	11.85 ± 0.23 cd	11.69 ± 0.30 d	12.28 ± 0.35 bcd	12.43
WC283	14.46 ± 0.68 c	12.42 ± 0.99 b	13.79 ± 0.96 a	10.32 ± 0.82 abcd	17.83 ± 0.54 ab	13.22 ± 0.61 b	13.35 ± 0.59 b	12.88 ± 0.79 abc	13.53
WC286	12.92 ± 0.67 d	11.85 ± 0.97 bc	12.85 ± 0.51 ab	9.70 ± 0.94 bcd	16.85 ± 0.99 bc	11.77 ± 0.50 cd	11.97 ± 0.68 cd	11.95 ± 0.43 cd	12.48
WC291	11.91 ± 1.00 d	12.28 ± 0.90 bc	11.99 ± 0.59 bc	9.51 ± 0.57 cd	14.89 ± 0.46 d	11.14 ± 0.23 de	11.85 ± 0.63 d	11.73 ± 0.40 de	11.91
WC299	10.47 ± 0.42 e	9.99 ± 0.73 d	11.19 ± 0.75 c	9.05 ± 0.41 d	13.35 ± 0.43 e	10.58 ± 1.02 e	11.18 ± 0.43 d	9.93 ± 0.36 f	10.72
Mean	13.64	11.77	12.71	10.27	16.46	12.16	12.63	12.18	12.73
CV(%)	13.96	11.93	9.19	9.75	9.90	11.14	8.64	10.21	10.58
G	74.481 ***
Y	118.475 ***
L	345.238 ***
G × Y	4.786 ***
G × L	4.145 ***
Y × L	57.318 ***
G × Y × L	2.127 **

Note: numbers in the table were means ± standard errors; means with different letters within a column are significantly different (p < 0.05); G, genotype; Y, year; L, location. ** significantly different at (p < 0.01); *** significantly different at (p < 0.001).

,

Table 4. Plant height mean values of different forage oat genotypes tested at four locations (WJ, CZ, GY, and XC) from 2022 to 2024.

Genotypes	PH(cm)								Combined
	2023WJ	2023CZ	2023GY	2023XC	2024WJ	2024CZ	2024GY	2024XC
Longyan 3	163.6 ± 4.2 d	162.4 ± 6.6 ab	133.4 ± 4.3 b	135.4 ± 3.9 bc	174.1 ± 5.1 b	158.6 ± 5.0 cd	147.0 ± 3.7 b	148.5 ± 5.5 abc	152.9
Intimidator	187.1 ± 2.6 a	165.0 ± 4.6 a	136.5 ± 3.9 b	134.6 ± 3.2 bc	184.4 ± 4.4 a	166.6 ± 3.8 b	147.4 ± 1.7 b	154.1 ± 3.3 ab	159.5
WC109	179.8 ± 3.0 b	160.6 ± 5.3 ab	149.4 ± 5.7 a	138.5 ± 0.6 b	175.4 ± 2.7 b	175.6 ± 3.6 a	157.4 ± 5.3 a	150.0 ± 4.3 a	160.8
WC130	179.0 ± 4.8 b	151.5 ± 4.1 cd	125.3 ± 4.9 c	144.8 ± 0.7 a	173.1 ± 1.6 b	157.4 ± 3.8 cd	157.3 ± 4.6 a	145.8 ± 3.0 ef	154.3
WC179	175.7 ± 4.8 bc	159.4 ± 5.9 abc	123.5 ± 5.5 c	135.5 ± 2.4 bc	176.0 ± 1.7 b	166.4 ± 4.4 b	123.9 ± 6.4 d	139.2 ± 1.1 bcd	149.9
WC283	177.6 ± 3.7 bc	154.7 ± 4.4 bc	133.0 ± 5.7 b	146.7 ± 1.4 a	178.7 ± 3.1 ab	163.2 ± 3.3 bc	136.3 ± 1.5 c	159.4 ± 5.9 abc	156.4
WC286	172.5 ± 2.5 c	152.5 ± 5.9 cd	131.0 ± 5.6 bc	133.1 ± 1.2 c	179.1 ± 5.7 ab	159.6 ± 4.0 cd	148.0 ± 2.2 b	145.7 ± 4.2 cd	152.7
WC291	132.6 ± 5.2 e	146.5 ± 2.8 d	110.4 ± 3.6 d	134.8 ± 4.1 bc	163.7 ± 6.2 c	153.1 ± 1.7 d	131.3 ± 1.5 c	103.1 ± 3.5 de	134.5
WC299	119.9 ± 3.6 f	117.6 ± 5.6 e	95.9 ± 5.2 e	119.1 ± 2.0 d	129.3 ± 3.1 d	120.2 ± 1.9 e	97.3 ± 3.8 e	99.1 ± 5.3 f	112.3
Mean	165.3	152.2	126.5	135.8	170.5	157.9	138.4	138.5	148.1
CV(%)	13.42	9.29	12.10	5.69	9.28	9.48	13.21	15.20	9.85
G	356.302 ***
Y	140.298 ***
L	910.628 ***
G × Y	5.050 ***
G × L	14.288 ***
Y × L	14.295 ***
G × Y × L	15.765 ***

Note: numbers in the table were means ± standard errors; means with different letters within a column are significantly different (p < 0.05); G, genotype; Y, year; L, location. *** significantly different at (p < 0.001).

,

Table 5. Tiller number mean values of different forage oat genotypes tested at four locations (WJ, CZ, GY, and XC) from 2022 to 2024.

Genotypes	TN								Combined
	2023WJ	2023CZ	2023GY	2023XC	2024WJ	2024CZ	2024GY	2024XC
Longyan 3	6.1 ± 0.6 abc	5.2 ± 0.6 bc	6.2 ± 0.7 a	6.2 ± 0.7 a	5.7 ± 0.8 b	5.2 ± 0.6 b	6.4 ± 0.6 a	6.2 ± 0.7 a	5.9
Intimidator	6.3 ± 0.6 ab	5.7 ± 0.6 ab	5.3 ± 0.4 bc	6.1 ± 0.5 a	7.1 ± 0.5 a	5.3 ± 0.6 b	5.9 ± 0.3 a	6.3 ± 0.6 a	6.0
WC109	7.0 ± 0.7 a	6.5 ± 0.6 a	5.7 ± 0.5 ab	5.7 ± 0.6 ab	7.8 ± 0.5 a	6.6 ± 0.4 a	6.0 ± 0.2 a	6.7 ± 0.7 a	6.5
WC130	4.8 ± 0.6 d	4.6 ± 0.6 c	4.7 ± 0.4 c	4.7 ± 0.3 cd	5.9 ± 0.5 b	4.3 ± 0.7 b	4.8 ± 0.4 c	5.2 ± 0.4 bc	4.9
WC179	5.0 ± 0.9 d	4.3 ± 0.4 c	4.5 ± 0.5 c	4.9 ± 0.5 bcd	6.0 ± 0.7 b	4.4 ± 0.3 b	4.8 ± 0.6 c	4.5 ± 0.2 c	4.8
WC283	5.4 ± 0.7 bcd	5.2 ± 0.4 bc	4.8 ± 0.6 bc	5.4 ± 0.7 abc	6.2 ± 0.5 b	5.2 ± 0.4 b	5.8 ± 0.7 ab	5.9 ± 0.4 ab	5.5
WC286	5.3 ± 0.5 cd	4.7 ± 0.6 c	4.6 ± 0.5 c	4.5 ± 0.4 d	5.7 ± 0.4 ab	4.5 ± 0.7 b	4.7 ± 0.3 c	4.9 ± 0.4 c	4.9
WC291	4.9 ± 0.5 d	4.6 ± 0.7 c	4.4 ± 0.6 c	4.4 ± 0.5 d	5.6 ± 0.2 b	5.0 ± 0.7 b	4.2 ± 0.6 c	4.7 ± 0.6 c	4.7
WC299	5.2 ± 0.7 cd	4.6 ± 0.5 c	4.7 ± 0.7 c	5.4 ± 0.6 abc	5.3 ± 0.4 b	5.2 ± 0.5 b	4.9 ± 0.7 bc	5.3 ± 0.5 bc	5.1
Mean	5.5	5.0	4.9	5.2	6.1	5.1	5.3	5.5	5.4
CV(%)	16.13	16.03	15.23	15.09	13.92	15.83	15.40	15.33	11.01
G	31.824 ***
Y	15.887 ***
L	25.754 ***
G × Y	0.971 NS
G × L	1.799 *
Y × L	1.620 *
G × Y × L	0.995 NS

Note: numbers in the table were means ± standard errors; means with different letters within a column are significantly different (p < 0.05); G, genotype; Y, year; L, location. * significantly different at (p < 0.05); *** significantly different at (p < 0.001). ^NS not significant (p > 0.05).

and

Table 6. Stem–leaf ratio mean values of different forage oat genotypes tested at four locations (WJ, CZ, GY, and XC) from 2022 to 2024.

Genotypes	SLR(%)								Combined
	2023WJ	2023CZ	2023GY	2023XC	2024WJ	2024CZ	2024GY	2024XC
Longyan 3	75.64 ± 5.80 ab	73.90 ± 7.19 c	66.14 ± 5.22 ab	61.68 ± 6.81 bc	85.67 ± 8.02 ab	57.00 ± 5.29 e	59.00 ± 3.24 b	84.29 ± 2.85 a	70.42
Intimidator	77.79 ± 6.49 ab	64.23 ± 6.11 de	58.51 ± 2.62 c	54.77 ± 4.66 c	88.00 ± 3.61 a	68.33 ± 5.11 cd	59.71 ± 1.30 b	81.17 ± 4.54 a	69.06
WC109	73.57 ± 6.73 abc	52.84 ± 5.46 f	60.35 ± 5.45 bc	61.22 ± 8.60 bc	86.33 ± 3.22 ab	62.00 ± 5.20 de	62.15 ± 5.02 ab	84.33 ± 3.93 a	67.85
WC130	75.29 ± 1.77 ab	68.21 ± 4.99 cd	61.56 ± 2.66 abc	63.42 ± 5.00 bc	83.33 ± 4.16 ab	82.00 ± 4.58 ab	62.49 ± 2.31 ab	78.78 ± 8.07 ab	71.88
WC179	72.16 ± 6.79 bc	57.72 ± 4.76 ef	62.32 ± 2.77 abc	63.56 ± 5.14 bc	78.67 ± 4.51 b	63.67 ± 6.43 de	63.60 ± 1.31 ab	71.13 ± 7.28 bc	66.60
WC283	75.22 ± 5.79 ab	84.94 ± 7.74 ab	59.66 ± 4.94 bc	59.42 ± 7.06 bc	89.67 ± 5.03 a	65.67 ± 7.10 de	60.60 ± 2.06 ab	63.92 ± 4.64 c	70.32
WC286	76.96 ± 6.08 ab	77.36 ± 4.76 bc	63.36 ± 3.09 abc	74.09 ± 6.85 a	88.67 ± 6.03 a	76.00 ± 3.61 bc	63.87 ± 1.30 ab	78.15 ± 3.14 ab	75.75
WC291	65.62 ± 4.24 c	53.06 ± 4.91 f	61.68 ± 3.46 abc	69.50 ± 6.96 ab	85.00 ± 2.00 ab	63.67 ± 2.31 de	62.70 ± 3.41 ab	80.01 ± 2.82 ab	67.65
WC299	81.87 ± 6.86 a	88.40 ± 7.18 a	67.63 ± 5.95 a	69.41 ± 7.64 ab	88.33 ± 1.53 a	86.33 ± 6.51 a	65.78 ± 3.43 a	86.47 ± 2.61 a	77.90
Mean	74.90	68.96	62.36	64.12	85.96	69.41	62.21	78.69	71.22
CV(%)	8.75	19.54	7.37	12.51	5.76	14.81	4.97	9.90	5.07
G	15.184 ***
Y	87.610 ***
L	108.015 ***
G × Y	5.588 ***
G × L	5.769 ***
Y × L	27.904 ***
G × Y × L	3.574 ***

Note: numbers in the table were means ± standard errors; means with different letters within a column are significantly different (p < 0.05); G, genotype; Y, year; L, location. *** significantly different at (p < 0.001).

). The selection of genotypes with superior yield in a specific environment does not necessarily facilitate the identification of their capability of exhibiting comparable performance across diverse environmental conditions. The mean DMYs of oats across locations over the years are presented in Table 3. The mean DMY varied between 10.72 and 14.58 t/ha. Notably, ‘WC109’ exhibited the highest DMY, followed by ‘Intimidator’, ‘WC283’, and ‘Longyan 3’, with each of these genotypes achieving a DMY exceeding 13 t/ha. This performance indicates a pronounced high-yielding tendency and is significantly superior to the remaining five lines. Furthermore, the data revealed that the maximum DMY was observed at location WJ, as compared to other sites, suggesting that the climatic conditions prevalent in WJ are particularly conducive to the cultivation of forage oats.

Mean values of PH at the harvest stage for forage oats across all locations are presented in Table 4. The overall mean PH ranged from 112.3 to 160.8 cm, with ‘WC109’ exhibiting the highest PH value, followed by ‘Intimidator’, while ‘WC299’ displayed the lowest, showing highly significant differences (p < 0.05). Additionally, PH was significantly higher in the WJ location than in the other three test sites (p < 0.05), and PH was higher in the 2023–2024 growing season than in the 2022–2023 growing season.

The mean TNs of forage oats across locations over the years are presented in Table 5. The TN range for these oat genotypes was 4.7–6.5, with three genotypes showing elevated TNs: ‘WC109’ (6.5), ‘Intimidator’ (6.0), and ‘Longyan 3’ (5.9), which exhibited an increase of 9.26% to 20.37% above the mean. ‘WC109’, ‘Intimidator’, and ‘Longyan 3’ show no difference among themselves (p > 0.05), but their results are significantly higher than the other six lines (p < 0.05).

Analysis of the SLR for the nine forage oat genotypes revealed that ‘WC179’ had the lowest content, of 66.60% (average data of four locations and two years), while ‘WC299’ had the highest content, of 77.90%, marking a difference of 16.97% between the two lines (Table 6). There was a variation in SLR between the years, with 2023–2024 showing higher values than 2022–2023. Additionally, the SLR values across different locations were ordered as WJ > CZ > XC > GY.

3.2. Nutrient Composition

The NDF, ADF, CP, EE, and ASH are five important indicators for measuring the nutritional value of forage oats. Our results indicate that these indices are influenced by genotype, year, location, and their interactions (p < 0.05) (

Table 7. Analysis of genotype, year, location, and their interactions regarding forage oat quality.

Treatment	NDF	ADF	CP	EE	ASH
G	30.543 ***	19.814 ***	47.620 ***	30.476 ***	14.843 ***
Y	580.737 ***	88.225 ***	16.727 ***	16.564 ***	160.224 ***
L	69.907 ***	77.559 ***	16.302 ***	5.559 ***	97.711 ***
G × Y	5.124 ***	2.840 **	3.864 ***	3.278 **	1.922 *
G × L	2.587 ***	3.334 ***	1.702 *	4.982 ***	3.558 ***
Y × L	77.148 ***	128.155 ***	6.463 ***	27.228 ***	88.681 ***
G × Y × L	3.251 ***	1.957 **	1.738 *	6.129 ***	3.157 ***

G, genotype; Y, year; L, location. * significantly different at (p < 0.05); ** significantly different at (p < 0.01); *** significantly different at (p < 0.001).

). The average NDF content range of these oat genotypes was 50.17–53.69% (Figure 2a). Compared to other genotypes, ‘WC109’ has the lowest NDF content at 50.17%, which is 2.07% to 7.02% lower than that of others, and it showed the lowest NDF content at all four locations. The difference in the data of ‘WC109’ (30.76%), which has the lowest ADF content, and ‘WC299’ (33.74%), with the highest ADF content, is 9.69%, and the difference is extremely significant. In the other tested genotypes, the ADF content ranges from 31.64% to 32.30%, with no significant differences among them (Figure 2b).

The average CP, EE, and ASH of forage oats across the four locations and over the two years are presented in Figure 3. Compared to other genotypes, ‘WC109’ (9.63%) had the highest CP content, significantly higher than that of other genotypes, which ranged from 4.12% to 30.84%. This was followed by ‘Intimidator’ at 8.99%, indicating that these two genotypes have superior amino acid profiles (Figure 3a). The results showed that ‘Intimidator’ (3.74%) and ‘WC283’ (3.48%) have a significantly higher EE content than the other genotypes, with increases of 16.15% to 28.08% and 8.07% to 19.18%, respectively. ‘WC299’ (2.92%) had the lowest EE content, followed by ‘WC286’ (2.94%), with no significant difference between them (Figure 3b). In terms of ASH content, ‘Longyan 3’ had the highest content of 9.37%, while ‘WC109’ had the lowest at 7.73%, with a difference of 21.22%. ‘Longyan 3’ showed no significant difference compared to ‘WC299’ and ‘WC179’, but was significantly higher than other genotypes (p < 0.05) (Figure 3c).

In the realm of practical application, the RFQ and GI emerge as critical metrics for evaluating the nutrient content and overall quality of forage. The mean RFQs and GIs of the oats across the locations over the years are presented in Figure 4. The RFQ value variation (p < 0.05) among genotypes ranged from 104.72 to 115.45. ‘WC109’ obtained a higher RFQ than its counterparts in the test (Figure 4a). The highest GI was for ‘WC109’ (19.30), followed by ‘Intimidator’ (16.23) and ‘Longyan 3’ (15.48), whereas the lowest GI was for ‘WC299’ (12.50). The recorded variation (p < 0.05) among oat genotypes for GI ranged from 12.50 to 19.30, with a mean value of 14.82 (Figure 4b).

3.3. Correlation of Agronomic Traits and Nutrients

The forage oat DMY displayed a significant positive correlation with PH, TN, SLR, EE, and CP, but it exhibited a significant negative correlation with NDF (p < 0.05). CP was highly significantly positively correlated with EE, PH, and TN (p < 0.05), and it had a positive but not significant correlation with ASH (p > 0.05), and a highly significant negative correlation with NDF and ADF (p < 0.05). PH displayed an extremely significant negative correlation with NDF, and TN displayed an extremely significant negative correlation with ADF (p < 0.05) (Figure 5).

3.4. Key Factors for Changes in Yield and Feeding Value

Most traits of forage oats were strongly correlated with each other. Structural equation model (SEM) analysis was further employed to evaluate the direct and indirect effects of growth performance and nutritional components on yield and feeding value (Figure 6). PH, TN, and SLR all enhanced the yield of forage oats, but the SLR was significantly lower than TN, which was significantly lower than PH. The main factors contributing to the increased forage oat yield and feeding value for PH included CP and NDF. The main causes of the increased forage oat yield and feeding value for TN included CP and NDF. The SLR primarily reduced the forage oat feeding value by increasing the contents of NDF and ADF. In addition, there was a trade-off between the direct and indirect pathways through which the SLR affected yield.

3.5. Principal Component Analysis

Principal component analysis (PCA) facilitates the study of differences in agronomic characteristics, nutritional indices, and the evaluation of feeding value among various forage oat genotypes, thereby enabling the selection of the most optimal materials. A PCA of nine oat genotypes was conducted across eleven parameters (Figure 7), and no significant variability was found among the eight genotypes except for the ‘WC109’. For PC1 and PC2, the total variabilities were 35.3% and 20.9%, respectively. The PCA scores for PC1 and PC2 indicate that ‘WC109’ is located in the positive quadrant of PC1 and PC2. This suggests that ‘WC109’ has high DMY, CP, PH, TN, RFQ, and GI, with low fiber content, demonstrating its characteristics as a high-quality and high-yield line.

3.6. Cluster Analysis

To elucidate the relationships among the oat genotypes, a systematic cluster analysis using the Euclidean distance square method was conducted on eleven phenotype traits. At Euclidean distance 3, the genotypes were categorized into four principal clusters (Figure 8). Cluster I (‘Longyan 3’, ‘Intimidator’, ‘WC283’, and ‘WC109’) is characterized by high DMY, CP content, TFQ, and GI, showing excellent performance. Within a distance of about 1, they were further subdivided into three sub-clusters. The classification results of the sub-clusters indicate that ‘WC109’ has a superior DMY and quality compared to the control varieties, making it suitable for promotion and cultivation in the autumn-sown region.

3.7. The AMMI Model Was Employed to Evaluate the Yield Stability and High-Yield Potential of the Tested Genotypes

In this study, the AMMI biplot was constructed with average yield on the x-axis and IPCA1 (interaction principal component axis 1) on the y-axis (Figure 9). A horizontal line was drawn at IPCA1 = 0, and a vertical line was drawn at the mean yield of all genotypes. Horizontally, genotypes with higher x-axis values have higher yields. Vertically, genotypes closer to the horizontal line have more stable yields. Line ‘WC109’ had a significantly higher yield than the control, while the yield difference between ‘WC283’ and the control was not significant. The other five genotypes had yields lower than the control. The genotypes with better yield stability than the control were ‘WC286’, ‘WC179’, ‘WC130’, ‘WC283’, and ‘WC109’. Therefore, ‘WC109’ was identified as the most high-yielding and stable genotype.

In the horizontal direction, the locations are more dispersed than the genotypes, indicating that the variation among locations is much greater than that among genotypes. Genotypes located above or below the horizontal line have positive interactions with the locations on the same side, meaning that the four high-yielding materials (‘WC283’, ‘WC109’, ’Longyan 3’, and ‘Intimidator’) have the best adaptability in WJ.

4. Discussion

4.1. Oat Genotypes’ Agronomic Characteristics Comparison

PH and TN are critical determinants of forage yield, constituting key breeding targets for plant breeders [31]. The present study observed a range of 112.3 to 160.8 cm for PH and 4.7 to 6.5 for TN across nine genotypes. Additionally, our research found significant differences in PH among the same genotypes under different environments and among different genotypes under the same environment, indicating that PH is influenced by both. Our research findings are consistent with those of Dinkale [32] and Tulu [33], who both reported that the agronomic and nutritional values of the same oat variety vary across different locations. Notably, compared with the 2022–2023 growing season, all nine oat lines showed higher PH and TN in 2023–2024, which in turn increased yields. This may be due to an increase in soil nutrient content. Forage yield is a critical metric for assessing the productive capacity and economic value of diverse oat genotypes; it is the foremost factor to consider when determining whether a genotype can be promoted for cultivation in a particular region [6,34]. The characteristics of a variety are decisive for the latent productivity potential of forage. In this study, the mean DMY of the genotypes across various environments ranged from 10.72 to 14.58 t/ha, which is higher than the previous findings of Tulu, who documented a range of 7.36 to 9.03 t/ha [33]. Meanwhile, the large fluctuations in precipitation and temperature at the four locations during the reproductive period resulted in large differences in the DMY of oats. This indicates that precipitation and climatic factors have an important effect on forage yield. The superior performance of ‘WC109’ across diverse environments underscores its broad adaptability, likely due to genetic traits enhancing stress tolerance and nutrient partitioning. The significant location effect (p < 0.05) on DMY aligns with findings by Tulu [33], emphasizing the role of temperature and precipitation in oat productivity. The SLR is a commonly utilized indicator of forage palatability, with a lower ratio indicating a more abundant leaf component and enhanced palatability [35]. In this study, the SLR of the nine oat genotypes ranged from 66.60% to 77.90%, with ‘WC179’ having the lowest ratio and ‘WC299’ the highest. Line ‘WC299’ exhibited the lowest plant height and fewer tillers but had a higher stem diameter, a plant architecture that confers better resistance to lodging, thereby enabling it to maintain a favorable growth status across various environments. The variation in the SLR under different environments is quite large, indicating that it is affected by soil fertility and temperature conditions [36].

4.2. Oat Genotypes Nutritional Indices Comparison

Forage quality affects animal growth and livestock products by influencing digestibility and energy intake [37]. Previous studies have employed various chemical parameters, including NDF, ADF, and CP, to assess forage quality [38]. Oat plants contain a variety of nutrients, including CP, EE, and ASH, which are significantly superior to the straw of crops like wheat and maize, and have important forage value [39,40]. The concentration of CP is a paramount nutritional metric for assessing the dietary value of forage oats [6]. Our research has determined that lines ‘WC109’ (9.63%) and ‘Intimidator’ (8.98%) exhibit elevated CP levels, while ‘WC299’ has the lowest CP content (7.36%). The reasons for these differences may be related to differences in genotype and environment [41,42]. Given the indispensable role of CP in the growth and development of livestock, the strategic selection and cultivation of oat varieties with elevated CP concentrations are imperative for enhancing the nutritional efficacy of forage. EE plays an important nutritional role in animal feed, providing essential energy and fatty acids for livestock and poultry. In this experiment, the EE content of all genotypes was below 5%, ranging from 2.92% to 3.74%, which is consistent with previous studies [43]. NDF and ADF serve as two key indicators for assessing forage grass quality, with higher contents indicating lower quality. In this experiment, the NDF content ranged from 50.17% to 53.69%, which is lower than that reported by Mut [44] and Tulu [33], and consistent with the evaluation results of 21 oat varieties in the northwestern Shanxi region by Wang [43]. In this study, the ADF content was 30.76% to 33.74%, significantly lower than that reported in previous studies [45,46]. This discrepancy may be attributed to differences in genotype and agricultural ecological variations between experimental environments. The negative correlation between CP and fiber content (r = −0.68) mirrors trends observed in temperate forages [24]. ‘WC109’ uniquely balances high CP (9.63%) with low fiber, suggesting its potential for genetic improvement without compromising yield.

4.3. Comprehensive Evaluation of Forage Oat Genotype Screening

Previous studies have demonstrated that RFQ and GI serve as important composite indicators for evaluating forage quality [24,47], providing a basis for screening superior materials. This study showed that line ‘WC109’ had an RFQ value of 115.45 and a GI value of 19.30, which were higher than those of other genotypes, indicating its greater potential as a forage and its suitability for popularization of cultivation in autumn-sown areas. ‘WC109’ addresses critical gaps in winter forage supply, offering a 21.2% increase in DMY over regional averages. Its high RFQ and GI values support its adoption in dairy and beef production systems to enhance feed efficiency. Furthermore, climate conditions exhibit regional diversity, substantially influencing the selection of oat genotypes. Temperature, precipitation, and photoperiod are key climatic factors that directly affect the growth cycle, biomass accumulation, and yield of oats [48,49,50]. In this study, the nine oat genotypes did not show any susceptibility to pests and diseases across different environments, indicating their high resistance. Moreover, the yield variation of the same genotype was relatively small across different years in the same environment, demonstrating their stability in production. The analysis of high yield and stable yield showed that ‘WC109’ and ‘WC283’ had certain advantages.

The intrinsic genetic characteristics of oat genotypes, together with fluctuations in environmental conditions, had a key impact on yield and quality, leading to significant differences between genotypes. This observation highlights the importance of prioritizing the balance between yield and quality in practical agricultural production to optimize economic returns. In addition, SEM analysis is capable of simultaneously evaluating the direct and indirect effects among multiple variables, thereby providing a more comprehensive understanding of the complex relationships between growth performance, nutritional components, yield, and feeding value. It can help identify key factors that significantly influence yield and quality, thus providing a direction for crop variety improvement.

5. Conclusions

This study offers prospects for livestock development by screening superior forage oat genotypes for autumn-sown oats in agricultural and pastoral regions of China. Over a two-year period, nine forage oat genotypes were evaluated across four locations for agronomic traits (DMY, PH, TN, and SLR), nutritional indices (NDF, ADF, CP, EE, and ASH), and feeding value (RFQ and GI). Line ‘WC109’ consistently exhibited the highest DMY, CP, RFQ, and GI. Additionally, it showed low levels of NDF and ADF, indicating excellent palatability and nutritional quality. Therefore, it can be effectively utilized in agricultural systems to address feed gaps in autumn-sown areas. This study identifies ‘WC109’ as an elite autumn-sown oat genotype for southwest China’s agropastoral zones, combining high biomass yield (14.58 t/ha), superior nutritional quality (CP: 9.63%), and environmental adaptability. Future research should explore the molecular mechanisms underlying its stress tolerance and nutrient efficiency to guide targeted breeding programs.