Selecting Fore and Succeeding Crops to Construct a Double-Cropping System That Increases the Forage Yields of Alpine Pastoral Areas on the Qinghai–Tibetan Plateau

Abstract

Double-cropping systems in which two kinds of crops are harvested per year can elevate forage yields significantly. This is the first report on a double-cropping system in the northeastern margin of alpine pastoral areas on the Qinghai–Tibetan Plateau with an elevation of 3500 m. In this experiment, eight triticale genotypes, including five varieties (‘Gannong No. 2’, ‘Gannong No. 3’, ‘Gannong No. 4’, ‘Gannong No. 7’, and ‘Zangsi No. 1’), and three lines (C16, C23, and C25) were used as the fore crops, with the four succeeding crops being 50% of triticale mixed with 50% of forage pea (B1), 50% of triticale mixed with 50% of common vetch, 50% of oat mixed with 50% of forage pea, and 50% of oat mixed with 50% of common vetch. Over 2 years (2020–2021), among the fore crops, ‘Gannong No. 4’ had the highest average hay yield (9.00 t·ha⁻¹), crude protein content (114.97 mg·g⁻¹), and relative feeding value (91.77), as well as the lowest average neutral detergent fiber content (598.17 mg·g⁻¹). Among the succeeding crops, B1 had the highest average hay yield (11.45 t·ha⁻¹) and nutritional quality. Among the interactions between the fore and succeeding crops, the highest hay yield (21.72 t·ha⁻¹), crude protein content (262.22 mg·g⁻¹), and relative feeding value (219.34) were obtained when ‘Gannong No. 4’ was doubled with B1. The results provide a theoretical basis for carrying out a double-cropping system in the alpine pastoral areas on the Qinghai–Tibetan Plateau, and this has very important implications for crop production in this area.

1. Introduction

The Qinghai–Tibetan Plateau has a wealth of pastural resources, thus constituting a pivotal base for animal husbandry in this area [1]. However, in recent years, due to overgrazing as well as serious pest and rodent infestations, there has been a significant deterioration in the pastural ecological environment [2]. The lack of balance between forage and livestock has become increasingly severe, with serious shortages of forage in winter and spring. Therefore, it is necessary to make full use of the limited artificial grassland and produce more forage for animals.

Double cropping, which makes full use of land, is an excellent method for increasing forage yield per unit area and enriching crop species [3]. It provides a buffer against the effects of environmental disasters and promotes livestock conservation [4]. Relevant research indicates that in the rye (Secale cereale L.)–oat (Avena sativa L.) double-cropping system within the agricultural area of the Lhasa Valley, the annual unit hay yield reaches 27.69 t·ha⁻¹, which is 84.97% and 117.69% higher than that of mono-cultivated rye and mono-cultivated oats, respectively [5]. In heat-sufficient agricultural areas, farmers increase their incomes using double-cropping techniques, primarily flat cropping [3,6].

Triticale (× Triticosecale Wittmack) is a human-made polyploid cereal produced by the crossing of wheat (Triticum aestivum L.) and rye. It is an alternative crop to grow in unfavorable biotic and abiotic conditions, including cold and saline environments [7]. In comparison with wheat, triticale varieties often produce greater biomasses and adjust better to various soil and environmental conditions [8]. Triticale is increasingly accepted by consumers, and its planting area is expanding [9]. Usually, triticale is good feed for livestock due to its high forage yield and nutritional quality, and high grain yield that is rich in protein, fat, essential amino acids, and carbohydrates [10]. It has a high tolerance to drought and pests and requires low inputs, such as fertilizers [11]. Triticale has good adaptability to stressful environmental conditions, especially coldness. Studies have shown that winter triticale can overwinter at low temperatures of −30 °C to −25 °C when there is snow during the winter [5]. Thus, it can be cultivated in the autumn in areas having very low temperatures [9]. In the alpine pastoral areas of the Qinghai–Tibetan Plateau, triticale variety ‘Gannong No. 2’ safely survives the winter when planted at the end of September at elevations below 3200 m [12] and produces hay yields of 13–15 t·ha⁻¹, which is 1.4 times than that of the oat varieties popularly cultivated in the spring in this area [13]. In the following year, another crop can be double cropped while autumn-sown triticale grows to the heading stage (early May) [14]. Mono-cultivated oats, triticale, rye, or any of these three in combination with common vetch, can be used as a succeeding crop [15]. Adding a certain proportion of legumes to the succeeding crops can increase the nitrogen source in the soil, thereby reducing the amount of nitrogen fertilizer used for the fore crops. Additionally, it enhances contents of soil phosphorus, potassium, and organic matter while promoting forage yields [16,17]. Legumes can improve the nutritive value of mixtures and increase in vitro true digestibility and crude protein (CP) levels, while reducing neutral detergent fiber (NDF) concentrations, compared with cereal monocultures [18]. Ren et al. [14] showed that in Hezuo City (with an elevation of 2954 m) in the alpine pastoral areas of the Qinghai–Tibetan Plateau, the highest hay yield was obtained by a 50% triticale variety mixed with 50% common vetch (Vicia sativa L.), followed by a mixture of oat and common vetch. In our 2-year study, we aimed to determine (1) whether the double-cropping system was practical in the alpine pastoral areas on the Qinghai–Tibetan Plateau at an elevation of 3505 m; (2) the best fore and succeeding crops; and (3) whether the double-cropping system could increase the forage yield and improve the nutritional value. The results serve as a theoretical basis for establishing a double-cropping cultivation system in the alpine pastoral areas on the Qinghai–Tibetan Plateau at an elevation of 3505 m.

2. Materials and Methods

2.1. Field Conditions

The experiments were carried out at the Forage Base of the Grassland Research Institute of Sichuan Province (31°50′ N,101°51′ E), which is located in Hongyuan County, Aba Tibetan and Qiang Autonomous Prefecture, China, with a continental plateau cold temperate monsoon climate. The site has an average elevation of 3505 m above sea level, an annual temperature of 2.9 °C, and an annual precipitation rate of 750 mm. The dominant soil is subalpine meadow, which contained 32.54 g·kg⁻¹ of soil organic matter, 144.20 mg·kg⁻¹ of available nitrogen, 93.16 mg·kg⁻¹ of available phosphorus, and 214.27 mg·kg⁻¹ of available potassium at the beginning of the study. The pH was 7.60. Monthly precipitation and average temperature during the experiment are shown in Figure 1.

2.2. Experimental Genotypes

The fore crops, including the triticale varieties ‘Gannong No. 2’, ‘Gannong No. 3’, ‘Gannong No. 4’, ‘Gannong No. 7’, and ‘Zangsi No. 1’, and the triticale lines C16, C23, and C25, were provided by Gansu Agricultural University, China.

Succeeding crops, including triticale variety ‘Zangsi No. 1’, Canadian forage oat, Canadian forage pea, and Qinhai common vetch, were also provided by Gansu Agricultural University.

2.3. Experimental Design

A split-plot design with three replicates was used. The main plot was the fore crop, with eight treatments: A1, ‘Gannong No. 2’; A2, ‘Gannong No. 3’; A3, ‘Gannong No. 4’; A4, ‘Gannong No. 7’; A5, ‘Zangsi No. 1’; A6, C16; A7, C23; and A8, C25. The subplot was the succeeding crop, with four treatments: B1, 50% triticale variety ‘Zangsi No. 1’ and 50% Canadian forage pea; B2, 50% triticale variety ‘Zangsi No. 1’ and 50% common vetch; B3, 50% Canadian forage oat and 50% Canadian forage pea; and B4, 50% Canadian forage oat and 50% common vetch. The area of each plot was 10 m² (2 m × 5 m). The fore crops were cultivated on 10 September 2019 and 6 September 2020, using a seeding density of 750 × 10⁴·ha⁻¹. They were planted at a row spacing of 20 cm in each plot. After planting three rows of triticale (refer to the 3 green solid lines in Figure 2B), a 60 cm wide blank row (refer to the area between the two sets of green solid lines in Figure 2B) was reserved for planting succeeding crops (refer to the 2 green dashed lines in Figure 2B) the following year at the late growth stage of the fore crop (Figure 2A–C). The sowing time and planting method for the fore crops and succeeding crops are shown in

Table 1. Sowing time and planting method for the fore crops and succeeding crops.

Treatment	Sowing Time	Planting Method
Fore crops (A1–A8)	10 September 2019 6 September 2020	Unicast
Succeeding crops (B1–B4)	8 June 2020 10 June 2021	Mixture sowing

.

Before sowing the fore crops, fine soil preparation was carried out and 30 t·ha⁻¹ of sheep manure was applied. The chemical composition was as follows: pure N = 0.72%, P₂O₅ = 0.45%, K₂O = 0.31%, and organic matter = 25%. At the green-up stage of triticale the following year (15 March 2020 and 20 March 2021), 300 kg·ha⁻¹ of organic fertilizer (N + P₂O₅ + K₂O ≥ 5%; organic matter ≥ 45%; humic acid ≥ 10%) was dressed evenly in each plot. Different succeeding crops were sown in these reserved rows according to the experimental design on 8 June 2020 and 10 June 2021. The seeding densities and treatment amounts are shown in

Table 2. Seeding densities and amounts for the succeeding crops.

	Seeding Amount (kg·ha−1)				Seeding Density (×104·ha−1)
Treatment	Triticale	Oat	Common Vetch	Forage Pea	Triticale	Oat	Common Vetch	Forage Pea
B1	313	—	—	276	600	—	—	160
B2	313	—	106	—	600	—	160	—
B3	—	189	—	276	—	600	—	160
B4	—	189	106	—	—	600	160	—

. After harvesting the fore crops, 300 kg·ha⁻¹ of organic fertilizer was dressed evenly in each plot. Weeds were controlled artificially. No irrigation was provided in this experiment.

2.4. Parameter Determination

2.4.1. Fore Crops

The plant height of the fore crop was determined at the flowering stage before harvesting. Ten plants were randomly selected from each plot, and the distance from the ground to the highest point of each plant was measured. The average value of 10 plants was taken as the plant height of the plot [19].

The number of tillers was measured at the flowering stage. A 100 cm long row (excluding side rows) was randomly selected in each plot, and all the branches longer than 20 cm were counted [12].

At the flowering stage of gramineous forage, fresh weight was determined by cutting the aboveground parts of fore crops in each plot (excluding plants closer than 50 cm to either end and plants in the two side rows) and weighing them. Then, 500 g samples from each plot were weighed and dried at 105 °C for 30 min and at 65–70 °C for 7–8 h to a constant weight to determine the dry yield. The fresh/dry weight ratio of each sample was then used to calculate the hay yield of the entire plot. The nutritional quality of the dried samples was determined as described later.

2.4.2. Succeeding Crops

In each plot, the community height of the succeeding crop was determined at the flowering stage of the gramineous crop (i.e., triticale or oats). The average of the 10 measurements was regarded as the community height of the plot.

The number of branches in the succeeding crop was measured at the flowering stage of the gramineous crop [14]. The method used for determination was the same as described in Section 2.4.1 for fore crops.

Hay yield of succeeding crops was determined using the same methods as described in Section 2.4.1 for fore crops. After drying and weight determinations, the samples from each plot were used for a nutritional quality assessment.

2.4.3. Nutritional Quality

To determine nutritional quality, samples were first crushed and passed through a 1 mm sieve. The CP content was determined using the Kjeldahl method [20], and NDF and acid detergent fiber (ADF) contents were determined using the Van Soest method [20]. Relative feeding value (RFV) was calculated using the following formulae [21]:

$$\mathrm{Dry} \mathrm{Matter} \mathrm{Intake} \left(\mathrm{DMI}\right) (\%)=120/\mathrm{NDF} (\%)$$

$$\mathrm{Digestible} \mathrm{Dry} \mathrm{Matter} \left(\mathrm{DDM}\right) (\%)=88.9-0.779 \times \mathrm{ADF} (\%)$$

$$\mathrm{RFV}=\mathrm{DMI} \times \mathrm{DDM}/1.29$$

2.4.4. Interactions Between Fore and Succeeding Crops

For forage, the plant height, branch number, and hay yield were generally positively correlated with nutritional quality [22,23]. Therefore, in this experiment, the values for each parameter of the fore and succeeding crops were added and regarded as the interaction values between the crops.

2.5. Statistical Analysis

All the calculations were performed in Microsoft Excel 2019. Analyses of variance of plant height, tiller number (branches), hay yield, and contents of CP, NDF, ADF, and RFV were carried out in SPSS 26.0 using split-plot designs for fore crops and succeeding crops. Duncan’s multiple range test was used to test for significant differences. Histograms and Pearson’s correlation analyses [24] were constructed using Origin 2024 software (Origin Lab Corporation, Northampton, MA, USA). The membership function method was used to comprehensively evaluate the results and identify the optimal treatment [12]. Since no significant differences were found between 2 years, 2019 and 2020, data for the same crop parameters over the 2 years were averaged and analyzed.

3. Results

As shown in

Table 3. Analysis of variances in production performance and nutritional qualities of fore crops, succeeding crops, and the interactions between the fore and succeeding crops.

Treatment	F-Value
	Plant Height (cm)	Branch Number (×104·ha−1)	Hay Yield (t·ha−1)	CP (mg·g−1)	NDF (mg·g−1)	ADF (mg·g−1)	RFV
Fore crops	1.004	8.959 **	13.711 **	3.974 **	4.049 **	1.102	2.363 *
Succeeding crops	60.487 **	62.956 **	191.654 **	18.597 **	34.675 **	57.629 **	55.430 **
Fore crops × Succeeding crops	15.848 **	108.757 **	173.972 **	28.832 **	23.555 **	30.862 **	25.218 **

Note: * indicates significant differences at the 0.05 level; ** indicates significant differences at the 0.01 level. CP: crude protein; NDF: neutral detergent fiber; ADF: acid detergent fiber; RFV: relative feeding value.

, significant differences existed in the branch number, hay yield, CP, NDF content, and RFV for the fore crops. For the succeeding crops, as well as the fore and succeeding crop interactions, all of the parameters showed significant differences.

3.1. Production Performance and Nutritional Qualities of Fore Crops

As shown in Figure 3A, among different fore crops, A3 had the highest average tiller number (732.25 × 10⁴·ha⁻¹), followed by A1 (685.58 × 10⁴·ha⁻¹), and both were significantly higher than those of A4 (498.33 × 10⁴·ha⁻¹) and A8 (610.83 × 10⁴·ha⁻¹). On the contrary, A4 had a significantly lower value (p < 0.05).

The highest average hay yield was obtained in A3 (9.00 t·ha⁻¹), which was significantly higher (p < 0.05) than the other crops, except for A1 (8.44 t·ha⁻¹), and it was 2.18 times that of A4 (4.14 t·ha⁻¹) (Figure 3A).

As shown in Figure 3B, among different fore crops, the highest average CP content was in A3 (114.97 mg·g⁻¹), which was significantly higher than those of the other treatments (p < 0.05), followed by A1 (101.17 mg·g⁻¹), which was only significantly higher than A8 (87.47 mg·g⁻¹) (p < 0.05).

The lowest average NDF content was found in A3 (598.17 mg·g⁻¹), followed by A1 (619.87 mg·g⁻¹) and A2 (614.93 mg·g⁻¹), and these three values were significantly lower than those of A5 (669.63 mg·g⁻¹) and A8 (666.12 mg·g⁻¹) (p < 0.05) (Figure 3B).

The highest average RFV was that of A3 (91.77), followed by A1 (87.94), which both were significantly higher than that of A5 (72.80) (p < 0.05) (Figure 3B).

3.2. Production Performance and Nutritional Qualities of Succeeding Crops

Due to different crop combinations and the different genetic backgrounds of succeeding crops, the average plant heights of the various succeeding crops were significantly different (

Table 4. Differences in forage production for succeeding crops.

Treatment	Plant Height (cm)	Branch Number (×104·ha−1)	Hay Yield (t·ha−1)
B1	119.70 ± 1.71 a	478.23 ± 4.09 a	11.45 ± 0.11 a
B2	120.98 ± 1.53 a	490.57 ± 5.38 a	9.89 ± 0.17 b
B3	100.49 ± 1.03 b	395.00 ± 5.35 c	9.62 ± 0.15 b
B4	99.47 ± 1.69 b	432.24 ± 6.93 b	6.46 ± 0.17 c

Note: The data are presented as the means ± standard errors. Different letters within the same parameters indicate significant differences at the 0.05 level. B1: 50% triticale variety ‘Zangsi No. 1’ and 50% Canadian forage pea; B2: 50% triticale variety ‘Zangsi No. 1’ and 50% common vetch; B3: 50% Canadian forage oat and 50% Canadian forage pea; B4: 50% Canadian forage oat and 50% common vetch.

). B1 and B2 had similar average plant heights that were significantly higher than those of B3 and B4 (p < 0.05). Thus, the average plant height of triticale mixed with legumes was significantly higher than that of oat mixed with legumes.

The highest average branch number was in B2, followed by B1, and both values were significantly higher than those of B3 and B4 (p < 0.05). Thus, the average number of branches on triticale mixed with legumes was significantly greater than that on oat mixed with legumes.

The highest average hay yield was that of B1, which was significantly higher than those of all the other treatments (p < 0.05), followed by B2, which was significantly higher than that of B4. Thus, there was a better mixture effect for triticale mixed with forage pea.

There were significant differences in the average CP contents among the succeeding crops (Figure 4A). The highest average CP content was that of B1 (142.52 mg·g⁻¹), which was significantly higher than those of B2 (134.19 mg·g⁻¹), B3 (137.21 mg·g⁻¹), and B4 (129.48 mg·g⁻¹) (p < 0.05). Additionally, B3 had a higher average CP content than B2. Thus, triticale and oats separately mixed with forage pea had higher average CP contents than when mixed with common vetch.

As shown in Figure 4B,C, the average contents of NDF and ADF showed the same trend. The lowest average NDF and ADF contents were those of B1 (504.30 and 335.38 mg·g⁻¹, respectively), followed by B3 (520.36 and 350.89 mg·g⁻¹, respectively), and these values were significantly lower than those of B2 (550.89 and 385.67 mg·g⁻¹, respectively) and B4 (548.36 and 393.01 mg·g⁻¹, respectively) (p < 0.05). Thus, triticale and oats separately mixed with forage pea had lower average NDF and ADF contents than when mixed with common vetch.

The highest average RFV was that of B1 (116.20), which was significantly higher than those of the other treatments (p < 0.05), followed by B3 (110.40), and it was also significantly higher than B2 (99.65) and B4 (99.25) (p < 0.05). Thus, triticale and oats separately mixed with forage pea produced higher average RFVs than when mixed with common vetch (Figure 4D).

3.3. Differences in the Fore and Succeeding Crop Interactions

As shown in Figure 5A, with the A2 treatment, the plant height of B2 (224.14 cm) was higher than after other treatments, but there were no significant differences. After A1, A3, A4, A5, A6, A7, and A8 treatments, B1 and B2 produced similar heights that were significantly greater than those of B3 and B4 (p < 0.05).

After the A1, A6, A7, and A8 treatments, B2 had more branches (1186.00, 1183.50, 1131.25, and 1084.58 × 10⁴·ha⁻¹, respectively) compared with B1, B3, and B4. After the A2, A3, and A4 treatments, B1 had more branches, at 1117.08, 1251.42, and 998.33 × 10⁴·ha⁻¹, respectively. After the A5 treatment, B4 had more branches (1189.00 × 10⁴·ha⁻¹), followed by B2 (1183.17 × 10⁴·ha⁻¹). The branch number in B1 (1251.42 × 10⁴·ha⁻¹) after the A3 fore crop was significantly higher than those of the other treatments (p < 0.05). Thus, the number of branches of triticale mixed with legumes was significantly higher than that of oat mixed with legumes (Figure 5B).

The hay yield of B1 was significantly higher than that of any other treatment (p < 0.05) having the same fore crop, except for A5. The hay yield of B1 after the A3 fore crop (21.72 t·ha⁻¹) was significantly higher than yields of all the other treatment combinations (p < 0.05), followed by A3 + B2 (20.72 t·ha⁻¹). This demonstrated that the hay yield of triticale mixed with legumes was significantly greater than that of oat mixed with legumes (Figure 5C).

As shown in Figure 6A, after the A1, A2, A3, A4, and A5 treatments, succeeding crop B1 had the highest CP contents, which were 246.25, 241.22, 262.22, 236.05, and 241.47 mg·g⁻¹, respectively. However, for the A6 and A8 fore crops, the highest CP contents were obtained in combination with B3, at 241.68 and 229.50 mg·g⁻¹, respectively, and for A7, the highest CP content was obtained in combination with B4 (233.02 mg·g⁻¹). Overall, the highest CP content (262.22 mg·g⁻¹) was obtained in A3 + B1, followed by A3 + B2 (261.98 mg·g⁻¹).

As shown in Figure 6B, after the A1, A2, A3, A6, A7, and A8 fore crops, the NDF contents of the B1 (1131.50, 1106.13, 1073.90, 1147.90, 1136.50, and 1170.03 mg·g⁻¹, respectively) and B3 (1127.63, 1118.55, 1113.72, 1157.15, 1161.38, and 1187.05 mg·g⁻¹, respectively) succeeding crops were lower than those of B2 (1155.62, 1130.40, 1140.07, 1185.97, 1194.22, and 1243.20 mg·g⁻¹, respectively) and B4 (1163.87, 1195.68, 1147.50, 1174.10, 1194.93, and 1210.20 mg·g⁻¹, respectively). Overall, the lowest NDF content (1073.90 mg·g⁻¹) was obtained in A3 + B1, and it was significantly lower than those of all the other treatments (p < 0.05), followed by A2 + B1 (1106.13 mg·g⁻¹). It may be that mixing with forage pea reduces the NDF content more than mixing with common vetch.

After the A1, A2, A3, A4, A5, A7, and A8 treatments, B1 had the lowest ADF contents, which were 704.83, 754.73, 690.43, 770.47, 807.37, 761.00, and 803.20 mg·g⁻¹, respectively, followed by B3 (730.33, 775.18, 731.43, 778.40, 842.90, 771.65, and 818.20 mg·g⁻¹, respectively). However, for the A6 fore crop, the lowest ADF content was obtained in B3 (730.02 mg·g⁻¹), followed by B1 (761.97 mg·g⁻¹). Thus, when cereals were mixed with forage pea rather than common vetch, the ADF contents were reduced (Figure 6C).

As shown in Figure 6D, after the A1, A2, A3, A5, A7, and A8 treatments, succeeding crop B1 had the highest RFVs, which were 205.24, 204.56, 219.34, 187.87, 198.78, and 189.12, respectively, followed by B3 (202.36, 198.40, 203.64, 171.90, 191.80, and 183.22, respectively). After the A4 treatment, B3 had the highest RFV (194.68), followed by B4 (190.98). For the A6 fore crop, the highest RFV was obtained in B3 (198.10), followed by B1 (195.50). Overall, the highest RFV (219.34) was obtained in A3 + B1, and it was significantly higher than that of any other treatment (p < 0.05), followed by A1 + B1(205.24).

3.4. Correlation Analysis

For the fore crops (Figure 7A), plant height was positively independently correlated with hay yield and RFV at the 0.01 level, and with the tiller number at the 0.05 level. It was negatively correlated with the ADF content at the 0.01 level. The tiller number was positively correlated with hay yield at the 0.01 level. Hay yield was positively correlated with RFV at the 0.01 level and with CP content at the 0.05 level, but it was negatively separately correlated with ADF and NDF at the 0.01 level. The CP content was negatively correlated with NDF at the 0.05 level. NDF and ADF were positively correlated, and they both negatively correlated with RFV at the 0.01 level.

For succeeding crops (Figure 7B), plant height was positively correlated with branch number, hay yield, and CP content at the 0.01 level. The branch number was positively correlated with hay yield at the 0.01 level and with CP content at the 0.05 level. Hay yield was positively separately correlated with CP content and RFV at the 0.01 level, but it was negatively separately correlated with NDF and ADF contents at the 0.01 level. The CP content was negatively correlated with ADF at the 0.05 level. NDF and ADF were positively correlated, and they both negatively correlated with RFV at the 0.01 level.

The correlation relationships of fore and succeeding crop combinations (Figure 7C) were similar to those of the fore crops alone. Plant height was positively correlated with branch number, hay yield, and CP content at the 0.01 level. Branch number was positively correlated with hay yield and CP content at the 0.01 level. Hay yield was positively separately correlated with CP content and RFV at the 0.01 level, but it was negatively separately correlated with ADF and NDF at the 0.01 level. The CP content was positively correlated with RFV and negatively separately correlated with NDF and ADF at the 0.01 level. NDF and ADF were positively correlated, and they both negatively correlated with RFV at the 0.01 level.

4. Discussion

Double cropping can increase forage yield [25], but in double-cropping systems, the selection of fore crop is particularly important due to the cold resistance of crops [26]. Crops with high overwintering rates and tiller numbers have high surface coverage and soil–water conservation capabilities [27]. In the high-altitude pastoral areas, due to the extremely cold climate and short growth season, forage crops are usually sown in spring, and there is no forage species cultivated in autumn [28]. Some triticale varieties released by Gansu Agricultural University, China, have very strong cold resistance and can be autumn sown and overwintered in alpine pastures at elevations below 3200 m and 3850 m on the Qinghai–Tibetan Plateau in high- and low-latitude areas, and the forage yields have been higher than those of crops sown in spring [29]. Pei [15] showed that when triticale variety ‘Gannong No. 2’ (A1 in the current study) is cultivated in autumn as a winter crop in the Qinghai–Tibetan Plateau’s alpine pastoral area, its plant height ranges from 102.80 to 147.67 cm, the tiller number ranges from 609.33 to 728.89 × 10⁴·ha⁻¹, and the hay yield ranges from 14.56 to 15.28 t·ha⁻¹. Additionally, the CP, NDF, and ADF contents range from 10.48% to 11.80%, 51.77% to 56.92%, and 30.65% to 32.78%, respectively. In this study, the plant heights of eight triticale genotypes ranged from 114.65 to 123.45 cm, the tiller number ranged from 498.33 to 732.25 × 10⁴·ha⁻¹, and the hay yield ranged from 4.14 to 9.00 t·ha⁻¹. The CP, NDF, and ADF contents ranged from 87.47 to 114.97 mg·g⁻¹, 598.17 to 669.63 mg·g⁻¹, and 384.58 to 470.15 mg·g⁻¹, respectively. Low yields in this study resulted from a lack of nitrogen nutrition, because no nitrogen fertilizer was applied to investigate the numbers and types of protozoa in the soil.

The eight fore crops showed significant differences in tiller numbers and hay yields. This mainly depended on their genetics and overwintering rates. Varieties A3 and A1 had strong cold resistance levels, with high winter survival rates of 95% and 90%, respectively. Consequently, their tiller numbers, especially for A3, were higher than those of the other genotypes. This resulted in high hay yields because the tiller number was positively correlated with hay yield at the 0.01 level (Figure 7A), which was in accordance with Ren et al. [14]. On the contrary, the tiller number of A4 was the lowest, significantly lower than those of the other fore crops, and resulted in the lowest hay yield (Figure 3A). Compared with previous studies [12,30], the CP content in this study was low, but the ADF and NDF contents were high, mainly due to the delayed harvesting. The best time to harvest triticale for high CP content and yield is the heading stage, and if it is harvested too late, the CP content decreases and both the ADF and NDF contents increase [31]. Among the eight fore crops, A3 had a high CP content and RFV and a low NDF due to its greater plant height, tiller number, and hay yield, which were positively independently correlated with the CP content and RFV and negatively correlated with NDF (Figure 7A). Thus, it was concluded that triticale variety A3 was the most suitable for cultivation as a fore crop in this area.

In the alpine pastoral areas on the Qinghai–Tibetan Plateau, mixing cereals and legumes is an effective way to increase forage yield and quality [32]. Previously, the yield and nutritional quality obtained when 50% of triticale variety ‘Gannong No. 2’ was mixed with 50% of common vetch variety ‘Lvjian 431’ were high, and the hay yield was significantly greater than that of the common vetch monoculture [32]. Yang et al. [33] observed similar results in which the highest comprehensive evaluation value was obtained when 50% Canadian forage oat was mixed with 50% Canadian forage pea (common vetch), whereas the Canadian forage pea monoculture had the lowest comprehensive evaluation value. Therefore, for cereal and legume rotations, 50:50 mixtures of a cereal and legume were used as succeeding crops in this study. Cornelissen et al. [34] pointed out that plant heights of crops reflect their competitiveness and fecundity in mixed communities. Genetics is the main determinant of plant height, branch number, and nutritional quality [35]. Branch number has an important influence on the forage yield, which reflects the development condition of plants. Studies have shown that the stronger the branching ability, the higher the forage yield [36,37,38]. In this study, the greatest plant height and branch number among the four succeeding crops were obtained in B2 (Table 4), and both were positively correlated with hay yield at the 0.01 level (Figure 7B). However, the highest hay yield was obtained in B1. The main reasons were as follows: First, there were no significant differences in plant height and branch number between treatments B1 and B2; and second, the growth and development rate of common vetch was slower than that of forage pea, and its branches were thinner and its leaves were smaller than those of forage pea [15,39]. Nutritionally, B1 obtained the highest CP content and RFV, followed by B3, and B2 and B4 had similar results due to the higher CP content and RFV in triticale and the high contribution of forage pea to the mixed community [32].

In agricultural production, forage is mainly evaluated based on yield, followed by quality [37,40]. In this study, for interactions between fore and succeeding crops, the highest hay yield, CP content, and RFV were obtained when triticale variety A3 was double cropped with B1. In this combination, the plant heights, tiller (branch) numbers, CP contents, and RFVs of the fore and succeeding crops were the highest (Figure 3) or high (Table 4), and the NDF and ADF contents were the lowest (Figure 3 and Figure 4). Plant height and branch number were positively separately correlated with hay yield at the 0.01 level, whereas the NDF and ADF contents were negatively separately correlated with RFV at the 0.01 level (Figure 7C). Overall, this double-cropping system, in which the best fore crop was triticale A3 and the best succeeding crop was B1, could be utilized in alpine pastoral areas on the Qinghai–Tibetan Plateau. The hay yield was increased by 117% compared with that of the local oat monoculture, which normally obtains a hay yield of approximately 8 t·ha⁻¹. In addition, autumn-sown triticale is of vital ecological importance in the local environment.

5. Conclusions

Among eight fore crops, triticale variety A3 (‘Gannong No. 4’) had the highest hay yield and RFV. Among the four succeeding crops, B1 (50% triticale variety ‘Zangsi No. 1’ and 50% Canadian forage pea) was the best. Among fore and succeeding crop combinations, the best was A3 (‘Gannong No. 4’) double cropped with B1 (50% triticale variety ‘Zangsi No. 1’ and 50% Canadian forage pea). This double-cropping system resulted in a hay yield of 21.72 t·ha⁻¹, CP content of 262.22 mg·g⁻¹, and an RFV of 219.34. This system could significantly increase forage yields and improve the nutritional levels of forage crops produced in Qinghai–Tibetan alpine pastoral areas at elevations greater than 3505 m.