A Comparative Study of Cutting Height and Fermentation Method on Cenchrus fungigraminus Silage: Effects of Natural Fermentation Versus Microbial Inoculant on Silage Quality and Fiber Degradation

Simple Summary

Cechrus fungigraminus has strong potential for feed production, and silage technology can effectively reduce the hardness of forage fiber and increase the nutrient content. By studying the effects of different cutting heights and fermentation methods on the quality of C. fungigraminus silage, this study provides feasible methodological guidance for the research on the feed production of C. fungigraminus.

Abstract

Cenchrus fungigraminus (Juncao) is a high-yielding, fast-growing forage crop with considerable potential for livestock feed; however, optimizing its processing is essential for cost reduction and quality enhancement. This study comprised three components: (1) a comprehensive analysis of 25 on-farm silage samples from five locations in Southwest China using Grey Relational Analysis (GRA); (2) an assessment of the effects of three cutting heights (low: 100–150 cm; mid: 150–200 cm; high: 200–250 cm) on silage quality; and (3) a comparison of silage quality between natural fermentation and microbial inoculant treatments using mature Juncao (250–300 cm). The results showed that: (1) in the on-farm silage samples, carbon supplementation was significantly positively correlated with total digestible nutrients (TDN), relative feed value (RFV), ether extract (EE), and sensory evaluation (p < 0.05), and the GRA identified the top-ranked treatments, including J2, J3, J6, X6, and J5; (2) in the cutting height trials, fiber content increased significantly with cutting height (p < 0.05), while crude protein (CP) and TDN decreased significantly (p < 0.05). The 200–250 cm group exhibited optimal fermentation quality, characterized by the highest total volatile fatty acids (total VFA) and lactic acid concentrations, alongside the lowest pH and ammonia nitrogen/total nitrogen ratios (NH₃-N/TN); (3) in the inoculant comparison, the natural fermentation group demonstrated significantly higher degradation rates of acid detergent fiber (ADF), neutral detergent fiber (NDF), and acid detergent lignin (ADL) compared to the microbial inoculant group, while also maintaining a lower pH, higher total VFA and lactic acid. Consequently, for on-farm production, carbon supplementation is recommended to improve silage quality. Although cutting Juncao below 200 cm provides higher nutritional value, a height of 200–250 cm is advised to ensure optimal fermentation characteristics. Furthermore, natural fermentation proves superior to microbial inoculant treatment for mature Juncao. Together, these measures offer an effective strategy for producing high-quality Juncao silage.

1. Introduction

Feed consumption for ruminants currently accounts for approximately 60% of total production costs. At present, primary feed ingredients such as corn and soybeans are edible for both humans and livestock, which exacerbates the ″food-feed competition″ issue [1]. Consequently, identifying high-yield, low-cost forage resources has become a priority in ruminant nutrition research. Cenchrus fungigraminus (Juncao) is a perennial grass widely cultivated in the tropical regions of Asia and Africa; it possesses a favorable nutritional profile and demonstrates significant potential as a forage crop [2].

Ensiling technology can effectively optimize the nutritional compositions of Juncao by increasing crude protein (CP) content while reducing fiber levels [3], thereby enhancing its palatability for ruminants. Due to its high-yield characteristics, Juncao can effectively alleviate the competition for grain between humans and livestock and reduce feeding costs, presenting immense prospects for large-scale feed utilization. Current research indicates that appropriate ensiling methods can mitigate feed losses caused by adverse weather and storage conditions, subsequently improving the voluntary intake rate of cattle and sheep [4]. However, the cutting height, moisture content, and the application of silage additives all exert significant influences on fermentation quality and feed characteristics [5].

Regarding cutting height, corn cut at a greater height exhibits significantly higher dry matter (DM), CP, and starch contents compared to lower cutting height [6]. Similarly, Pennisetum glaucifolium cut at 90–110 cm shows a significantly improved fermentation effect and higher intake rates in cattle compared to that cut at 70 cm [7]. In terms of additives, the application of lactic acid bacteria (LAB) before ensiling, compared to natural fermentation, optimizes the microbial community structure. This promotes the utilization of water-soluble carbohydrates (WSC) to produce lactic acids and acetic acids [8], which enhances fermentation stability, inhibits the proliferation of pathogenic microorganisms [9], and reduces both the spoilage rate and butyric acid production [10].

Specific research on Juncao silage indicates that adding compound probiotics (Lactobacillus plantarum, Bifidobacterium animalis, and Bacillus coagulans) effectively lowers pH, increases CP and WSC content, and inhibits butyric acid formation [11]. Furthermore, feeding Juncao silage to sheep can significantly increase average daily gain and reduce muscle water loss [12]. Despite these advancements, as Juncao is a relatively new forage species, research on its systematic feed application remains incomplete. Specifically, definitive reports regarding the optimal cutting height or standardized additive protocols are lacking, leading to fragmented ensiling techniques without unified standards. Therefore, this study was conducted to systematically evaluate and optimize Juncao silage production. First, on-farm silage samples were collected from Southwest China and comprehensively evaluated using Grey Relational Analysis (GRA) to understand current practices. Subsequently, this study investigated the effects of three cutting heights on silage quality, and compared the efficacy of natural fermentation and microbial inoculants. The aim was to identify optimal processing parameters and provide a scientific basis for high-quality Juncao silage production.

2. Materials and Methods

2.1. Collection of Juncao Silage Samples in Southwest China

In 2023, a total of 25 on-farm silage samples were collected from five locations in Southwest China: Jianshui County (23°46′57″ N, 102°43’5″ E; hereinafter referred to as J), Yanshan County (23°36’47″ N, 104°20’28″ E; Y), Simao District (22°46’39″ N, 100°58’35″ E; S), Changning County (24°49’38″ N, 99°36’36″ E; C), and Xundian County (25°33’48″ N, 103°15’21″ E; X). All Juncao plants were cultivated in June 2022 and cut for ensiling when they reached a plant height of approximately 3 m. The characteristics of the collected samples are listed in

Table 1. Characteristics of the 25 on-farm Juncao silage samples collected from five locations in Southwest China.

Number	Region	Ensiling Duration (Month/Day)	Cut	Ensiling Method	Treatment
J1	Jianshui County	5/25	1	Bag storage
J2		5/25	1	Bag storage	Added 4% cornmeal
J3		6/27	1	Barrel storage
J4		7/25	1	Barrel storage
J5		7/25	1	Barrel storage	Added 10% rice bran
J6		7/25	1	Barrel storage	Added 10% rice bran and 0.1% brown sugar
Y1	Yanshan County	6/21	1	Bag storage
Y2		11/1	2	Bag storage	After harvesting, the Juncao was allowed to wilt for 48 h.
S1	Simao District	7/30	1	Bag storage
S2		7/30	1	Lower layer of cellar
S3		7/30	1	Upper layer of cellar
C1	Changning County	7/10	1	Bag storage
C2		7/24	1	Bag storage	After harvesting, the Juncao was allowed to wilt for 24 h.
C3		9/10	2	Bag storage
C4		10/19	2	Bag storage	After harvesting, the Juncao was allowed to wilt for 24 h.
X1	Xundian County	7/15	1	Barrel storage
X2		7/15	1	Barrel storage	After chopping, the Juncao was allowed to wilt for 24 h.
X3		7/15	1	Barrel storage	After chopping, the Juncao was allowed to wilt for 48 h.
X4		7/15	1	Barrel storage	Added hay to adjust the moisture content to 70%
X5		7/15	1	Barrel storage	Added hay to adjust the moisture content to 78%
X6		11/12	2	Barrel storage
X7		11/12	1	Barrel storage
X8		11/12	2	Barrel storage	Added local hay to adjust moisture content to 75%
X9		11/12	2	Barrel storage	Added hay to adjust the moisture content to 75%
X10		11/12	2	Barrel storage	Added hay to adjust the moisture content to 80%

, and the sampling locations are shown in Figure 1.

After a natural fermentation period of 60 days, silage samples were obtained using a silage sampler (Dairy One, Ithaca, NY, USA; diameter: 4.8 cm; sampling depth: 25 cm) following a five-point sampling method. Three replicates were collected for each sample type. The replicates were mixed, and a 2 kg subsample was obtained using the quartering method. These subsamples were subsequently sealed, frozen, and transported to the laboratory for silage quality analysis. Vacuum packaging was performed using a single-chamber vacuum machine (Model: HR320, Wuhan Kelibo Animal Husbandry Technology Co., Ltd., Wuhan, China). The vacuum sealing bags (capacity: 2 kg; dimensions: 34 cm × 30 cm; thickness: 0.12 mm) were also supplied by Wuhan Kelibo.

2.2. Experiments on Cutting Height and Fermentation Method Comparison

2.2.1. Experimental Site and Materials

The experiment was conducted in Zaikai Village, Xundian County, Yunnan Province (103°10′59″ E, 25°31′47″ N; elevation: 2001 m). The region is characterized by a plateau monsoon climate, featuring dry winters and springs, and humid, rainy summers and autumns. The site has an average annual precipitation of 1018.8 mm, an average annual temperature of 15.3 °C, and 2046.6 h of annual sunshine. The soil is classified as red soil with moderate fertility.

The Juncao was provided by the National Engineering Research Center for Juncao Technology at Fujian Agriculture and Forestry University. The crops were planted on 1 June 2022, with a plant-to-row spacing of 60 cm × 80 cm. For the silage additive, a microbial inoculant (Chr. Hansen, supplied by Wuhan Kelibo Animal Husbandry Technology Co., Ltd., Wuhan, China) was used. The active ingredients consisted of Lactococcus lactis (≥6.2 × 10¹⁰ CFU/g) and Lactobacillus buchneri (≥6.2 × 10¹⁰ CFU/g). A stock solution with a concentration of 0.02% (w/v) was prepared by dissolving 0.2 g of the product in 1 L of sterile water (resulting in a Lactococcus lactis concentration of ≥2.6 × 10⁸ CFU/g) for subsequent use. Vacuum packaging was performed using a single-chamber vacuum machine (Model: DZ-600/700/800 2E, Shanghai Jiahe Packaging Machinery Co., Ltd., Shanghai, China). The silage was packed in 2 kg fermentation bags equipped with one-way degassing valves (dimensions: 23 cm × 40 cm; supplied by Wenzhou Huaguan Packaging Co., Ltd., Wenzhou, China).

2.2.2. Experimental Design

A randomized block design was utilized, classifying the Juncao into three cutting heights: 100–150 cm (Low), 150–200 cm (Mid), and 200–250 cm (High). The experimental plots measured 30 m² (5 m × 6 m), with three replicates per treatment. The stubble height was maintained at 10 cm. For each treatment, 6 kg of fresh Juncao was collected, chopped into 2–3 cm lengths, and packed into silage bags. The samples were vacuum-sealed and allowed to undergo natural fermentation at room temperature for 60 days. Subsequently, samples were taken for the determination of nutritional compositions and fermentation characteristics. The nutritional compositions of fresh Juncao at different cutting heights used in this study are summarized in

Table 2. Nutritional compositions of Juncao at different cutting heights.

Item	Cutting Height
	Low (100–150 cm)	Mid (150–200 cm)	High (200–250 cm)
DM, % FM	11.85	11.65	14.30
CP, % DM	14.20	10.50	9.70
ADF, % DM	40.35	43.40	45.15
NDF, % DM	66.25	70.90	70.60
CF, % DM	34.75	39.30	39.10
ADL, % DM	4.23	5.10	5.61
Starch, % DM	0.20	0.55	0.45
EE, % DM	2.60	2.10	1.60
TDN, % DM	57.05	53.55	51.60
RFV	80.50	72.50	71.00
WSC, % DM	5.35	4.20	4.41

Abbreviations: DM, dry matter; FM, fresh matter; CP, crude protein; ADF, acid detergent fiber; NDF, neutral detergent fiber; CF, crude fiber; ADL, acid detergent lignin; EE, ether extract; TDN, total digestible nutrients; RFV, relative feed value; WSC, water-soluble carbohydrate.

.

Mature Juncao with a cutting height of 250 cm to 300 cm (nutritional compositions shown in Table 3) was selected, as this stage exhibited the highest fiber content. Fermentation tests were conducted after cutting, with a stubble height of 10 cm. Three treatments were established:

Non-ensiled control (T): Fresh Juncao without ensiling.

Natural fermentation (T1): Juncao ensiled without any inoculant.

Microbial inoculant (T2): Juncao ensiled with 20 mL of microbial inoculant stock solution (0.02% w/v; Lactococcus lactis ≥ 2.6 × 10⁸ CFU/g).

Each treatment consisted of three replicates. For the ensiled groups, the respective liquids were sprayed uniformly onto 2 kg of chopped forage (2–3 cm length) per replicate. The samples were then vacuum-sealed and subjected to natural fermentation at room temperature for 60 days. This study evaluated the effects of these fermentation methods on the silage quality and fiber degradation of Juncao.

Table 3. Nutritional Compositions of Mature Juncao (250–300 cm).

Height	DM, % FM	CP, % DM	ADF, % DM	NDF, % DM	CF, % DM	ADL, % DM	Starch, % DM	EE, % DM	TDN, % DM	RFV	WSC, % DM
250–300 cm	15.27	11.30	47.70	73.13	40.80	5.62	0.33	1.48	52.80	66.00	6.10

Abbreviations: DM, dry matter; FM, fresh matter; CP, crude protein; ADF, acid detergent fiber; NDF, neutral detergent fiber; CF, crude fiber; ADL, acid detergent lignin; EE, ether extract; TDN, total digestible nutrients; RFV, relative feed value; WSC, water-soluble carbohydrate.

Table 3. Nutritional Compositions of Mature Juncao (250–300 cm).

Height	DM, % FM	CP, % DM	ADF, % DM	NDF, % DM	CF, % DM	ADL, % DM	Starch, % DM	EE, % DM	TDN, % DM	RFV	WSC, % DM
250–300 cm	15.27	11.30	47.70	73.13	40.80	5.62	0.33	1.48	52.80	66.00	6.10

Abbreviations: DM, dry matter; FM, fresh matter; CP, crude protein; ADF, acid detergent fiber; NDF, neutral detergent fiber; CF, crude fiber; ADL, acid detergent lignin; EE, ether extract; TDN, total digestible nutrients; RFV, relative feed value; WSC, water-soluble carbohydrate.

2.3. Determination of Indices and Methods

2.3.1. Sensory Quality Evaluation

The sensory quality of the silage was evaluated based on the sensory evaluation methods and grading standards established by the German Agricultural Society (Deutsche Landwirtschafts-Gesellschaft, DLG) [13]. The evaluation focused on three primary parameters: odor (smell), color, and structure (texture) to determine the overall quality grade of the silage mixtures (

Table 4. Sensory Evaluation Standards for Silage Quality.

Sensory Indicators	Scoring Criteria				Score
Odor	It has a strong butyric or ammonia odor, or almost no sour taste.				2
	It has a strong butyric acid taste, or a pungent, burnt, or musty smell.				4
	It has a weak butyric acid taste, or a strong sour taste and a weak aromatic taste.				10
	It has a strong or distinct aroma of bread, without any butyric acid odor.				14
Structure	Stem and leaf rot or severe pollution				0
	The stem and leaf structure is visibly damaged, or there is mild contamination.				1
	Slight damage to stem and leaf structure				2
	The stem and leaf structure is intact and clearly visible.				4
Color	Severe discoloration, turning dark green or brown.				0
	Slight discoloration, turning light yellow or yellowish-green.				1
	It closely resembles the color of the raw material, turning light brown after drying.				2
Total score	16–20	10–15	5–9	0–4
Grade	Level 1 Excellent	Level 2 is acceptable	Level 3 Intermediate	Level 4 corruption

).

2.3.2. Determination of Nutritional Compositions and Fermentation Characteristics

The nutritional compositions were determined according to the following standards: DM was determined according to GB/T 6435-2014 [14]; CP according to GB/T 6432-2018 [15]; neutral detergent fiber (NDF) according to GB/T 20806-2022 [16]; acid detergent fiber (ADF) according to NY/T 1459-2022 [17]; crude fiber (CF) according to GB/T 6434-2022 [18]; and starch according to GB/T 20194-2018 [19].

The pH of the silage extract was measured using a portable pH meter (Model: PH-30, Shanghai Yueping Scientific Instrument Co., Ltd., Shanghai, China). The concentrations of acid detergent lignin (ADL), ether extract (EE), total digestible nutrients (TDN), WSC, ammonia nitrogen (NH₃-N), and total nitrogen (TN), total volatile fatty acids (total VFA), lactic acid, acetic acid, propionic acid were analyzed using Near-Infrared Spectroscopy (NIRS) based on the CVAS (Cumberland Valley Analytical Services, Inc., Waynesboro, PA, USA) detection and data analysis model. The relative feed value (RFV) was calculated based on the NDF and ADF contents using the following formula:

$$\mathrm{RFV}=(88.9-0.779\times \mathrm{ADF})\times (120/\mathrm{NDF})/1.29$$

2.3.3. Grey Relational Analysis Method for Evaluating Silage Quality

Consider the 25 silage samples produced under actual field conditions by local farmers and herdsmen as a grey system. Select the optimal values of 12 indicators—DM, CP, NDF, ADF, EE, WSC, TDN, RFV, pH, total VFA, lactic acid, and Sensory evaluation—to form an ideal reference sample, and perform GRA. The original data were normalized using initialization to make them dimensionless. According to Formulas (1)–(5), the grey relational coefficient, absolute difference, equal weight relational degree, weight coefficient, and weighted relational degree were calculated, respectively. Here, ρ is the distinguishing coefficient, with ρ ∈ [0, 1] (typically set as 0.5), and a value of 0.5 was used in this study [20]. A higher value of either the equal weight relational degree or the weighted relational degree indicates a greater similarity between the sample and the optimal indicator set, reflecting better overall performance [21].

$${\epsilon }_i(k)={\displaystyle \frac{\underset{i}{\min } \underset{k}{\min } | X_0(k)-X_i(k)|+\rho \underset{i}{\max } \underset{k}{\max } | X_0(k)-X_i(k)|}{ | X_0(k)-X_i(k)|+\underset{i}{\max } \underset{k}{\max } | X_0(k)-X_i(k)|}}$$

(1)

$${\mathsf{\Delta }}_i\left(k)=|X_0\right(k)-X_i(k\left)\right|$$

(2)

$${\gamma }_i={\displaystyle \frac{1}{n}}{\sum }_{k=1}^n{\epsilon }_i(k)$$

(3)

$${\omega }_i={\displaystyle \frac{{\gamma }_i}{\sum {\gamma }_i}}i$$

(4)

$${{\gamma }^{‘}}_i={\sum }_{k=1}^n{\omega }_i(k)\epsilon (k)$$

(5)

where $\left|X_0\right(k)-X_i(k\left)\right|$ was the absolute difference between $X_0$ series and $X_i$ series at $k$, $\underset{i}{\min } \underset{k}{\min } | X_0(k)-X_i(k)|$ was the secondary minimum, $\underset{i}{\max } \underset{k}{\max } | X_0(k)-X_i(k)|$ was the secondary maximum. ${\epsilon }_i\left(k\right)$ is the grey relational coefficient, ${\Delta }_i\left(k\right)$ is absolute difference, ${\gamma }_i$ is equal weight relational degree, ${\omega }_i$ is weight coefficient, and ${{\gamma }^{‘}}_i$ is weighted relational degree.

2.3.4. Statistical Analysis

Statistical analyses were performed using SPSS 22.0 (IBM Corp., Armonk, NY, USA). Data from the three cutting heights were subjected to one-way analysis of variance (ANOVA). Where significant differences were found, post hoc multiple comparisons were conducted using the Least Significant Difference (LSD) test. Comparisons between the two fermentation treatments were analyzed using t-test. Correlation heatmaps depicting relationships among the 25 on-farm silage samples, as well as between cutting height and silage quality indicators, were generated using R software (Version 4.3.2, R Foundation for Statistical Computing, Vienna, Austria), with Spearman’s rank correlation coefficients calculated. Data visualization was conducted using GraphPad Prism (Version 9.0, GraphPad Software Inc., San Diego, CA, USA) and R software. Statistical significance was defined as follows: p < 0.05 (*) was considered significant, and p < 0.01 (**) or p < 0.001 (***) was considered extremely significant.

3. Results

3.1. Comprehensive Evaluation of On-Farm Juncao Silage Quality

As shown in

Table 5. Silage quality of on-farm Juncao silage from five locations in Southwest China.

Number	DM (% FM)	CP (% DM)	NDF (% DM)	ADF (% DM)	EE (% DM)	WSC (% DM)	TDN (% DM)	RFV	pH	Total VFA (% DM)	Lactic Acid (% DM)	Sensory Evaluation
J1	20.10	9.70	69.10	43.00	2.88	0.97	56.10	75	4.22	7.40	5.69	17.85
J2	25.20	8.70	50.30	31.40	3.18	1.71	61.80	119	3.70	7.14	6.21	18.10
J3	18.80	10.40	64.90	44.00	3.33	2.83	58.40	78	3.50	9.61	8.88	15.80
J4	19.70	4.50	70.60	49.60	2.04	1.90	52.80	66	3.45	10.75	8.81	17.00
J5	27.10	6.30	64.40	48.10	3.43	1.65	56.90	74	3.35	9.18	6.68	17.56
J6	27.40	6.10	64.00	47.10	3.39	1.85	57.80	76	3.34	9.03	6.39	17.56
Y1	19.60	7.10	71.10	45.40	2.62	2.00	53.10	70	4.47	5.68	4.21	17.57
Y2	32.60	3.20	76.30	56.20	0.63	1.10	50.30	55	3.63	8.93	4.79	17.75
S1	23.90	7.60	73.20	54.10	1.52	1.03	51.00	59	3.42	7.53	7.15	17.46
S2	19.90	5.90	72.40	51.50	1.68	1.80	51.70	63	4.14	8.43	3.74	17.00
S3	18.10	5.20	75.60	53.50	2.32	1.40	48.50	58	4.80	10.06	0.79	13.55
C1	16.80	8.30	71.20	49.00	2.24	0.08	47.80	66	5.08	8.11	2.37	9.78
C2	27.60	6.00	69.40	46.50	1.66	0.38	49.40	71	4.16	6.80	5.45	15.17
C3	16.80	8.10	73.70	52.60	2.40	0.60	50.80	61	4.47	5.93	2.18	17.00
C4	22.70	5.50	71.50	49.70	1.87	1.30	53.90	65	3.58	9.78	6.90	16.81
X1	15.50	10.90	69.90	48.50	2.99	1.60	56.20	68	4.70	5.52	2.60	14.10
X2	15.30	8.20	71.90	49.60	2.96	1.30	51.40	65	5.24	8.61	1.79	8.90
X3	17.10	10.20	71.70	47.70	3.37	1.10	48.70	67	5.23	10.25	2.30	9.00
X4	30.10	5.80	73.20	48.40	1.61	1.00	45.80	65	4.36	4.92	1.15	14.80
X5	22.00	7.40	72.90	48.70	2.08	1.10	46.90	65	4.72	7.25	1.66	16.10
X6	18.40	10.50	66.30	44.40	2.79	0.90	58.90	76	3.48	9.79	7.71	17.79
X7	29.00	7.90	68.60	47.40	1.73	1.00	55.30	71	3.63	9.38	6.32	16.29
X8	25.30	9.20	68.00	44.30	2.10	1.00	52.20	74	3.95	7.42	4.52	14.58
X9	26.00	7.80	70.70	46.50	1.47	1.10	51.60	69	3.96	7.40	4.81	14.25
X10	22.60	8.20	69.10	46.60	1.80	0.90	51.70	71	4.32	5.24	3.30	12.21

Abbreviations: DM, dry matter; FM, fresh matter; CP, crude protein; NDF, neutral detergent fiber; ADF, acid detergent fiber; EE, ether extract; WSC, water-soluble carbohydrate; TDN, total digestible nutrients; RFV, relative feed value; VFA, volatile fatty acids.

and Figure 2, correlation analysis revealed that carbon supplementation was extremely significantly positively correlated with EE (p < 0.01) and significantly positively correlated with TDN, RFV, and sensory evaluation (p < 0.05). Hay inclusion was significantly negatively correlated with total VFA (p < 0.05). Furthermore, natural fermentation showed a significantly negative correlation with DM content (p < 0.05).

In this study, GRA was used to comprehensively evaluate 25 silage treatments based on 12 indicators. The resulting weight order of the evaluation indicators was as follows (

Table 6. Equal weight relation degree and weighting coefficient for each indicator.

Item	Equal Weight Relational Degree	Weight Coefficient	Rank
DM	0.6297	0.0809	8
CP	0.6466	0.0830	7
NDF	0.6470	0.0831	6
ADF	0.6003	0.0771	9
EE	0.6478	0.0832	5
WSC	0.4901	0.0629	12
TDN	0.7771	0.0998	2
RFV	0.5508	0.0707	11
pH	0.7626	0.0979	3
Total VFA	0.6848	0.0879	4
Lactic acid	0.5517	0.0709	10
Sensory Evaluation	0.7984	0.1025	1

Abbreviations: DM, dry matter; CP, crude protein; NDF, neutral detergent fiber; ADF, acid detergent fiber; EE, ether extract; WSC, water-soluble carbohydrate; TDN, total digestible nutrients; RFV, relative feed value; VFA, volatile fatty acids.

): sensory evaluation > TDN > pH > total VFA > EE > NDF > CP > DM > ADF > lactic acid > RFV > WSC.

Based on equal weight relational degree and weighted relational degree (

Table 7. Comprehensive evaluation and ranking of on-farm Juncao silage quality using GRA.

Number	Equal Weight Relational Degree	Rank	Weighted Relational Degree	Rank
J1	0.6750	7	0.6916	7
J2	0.8209	1	0.8314	1
J3	0.8096	2	0.8107	2
J4	0.7038	6	0.7156	6
J5	0.7397	5	0.7578	5
J6	0.7424	3	0.7600	4
Y1	0.6307	15	0.6424	14
Y2	0.6340	12	0.6529	11
S1	0.6459	10	0.6634	10
S2	0.6089	17	0.6230	16
S3	0.5759	23	0.5875	22
C1	0.5544	25	0.5636	25
C2	0.6003	18	0.6122	18
C3	0.5829	19	0.6000	19
C4	0.6566	9	0.6729	9
X1	0.6340	11	0.6447	13
X2	0.5825	20	0.5897	21
X3	0.6333	14	0.6407	15
X4	0.5726	24	0.5835	24
X5	0.5818	21	0.5946	20
X6	0.7412	4	0.7609	3
X7	0.6741	8	0.6902	8
X8	0.6337	13	0.6457	12
X9	0.6103	16	0.6217	17
X10	0.5778	22	0.5874	23

), the top ten treatments were identified as J2, J3, J6, X6, J5, J4, J1, X7, C4, and S1. The weighted relational degree analysis revealed several key findings: (1) Harvest timing: The silage quality of the second ratoon (second crop) was superior to the first crop (e.g., Y2 > Y1, C4 > C2, and X6 > X1). (2) Additives: Treatments with carbon supplementation (e.g., cornmeal or rice bran) outperformed those without additives (e.g., J2 > J1; J5 > J4). (3) Storage method: The quality of bag storage was superior to cellar storage (e.g., S1 > S2). (4) Moisture regulation: Reducing moisture through wilting after chopping was detrimental to fermentation (e.g., X2 (wilted 24 h) and X3 (wilted 48 h) ranked low). (5) Hay inclusion: While adding dry hay successfully reduced the moisture content of Juncao, it negatively impacted silage quality due to the poor nutritional value of the hay itself (e.g., X4, X5, X9, and X10 ranked between 17th and 24th, whereas the corresponding groups without hay, X1 and X6, ranked 13th and 3rd, respectively).

3.2. Effects of Cutting Height on Nutritional Compositions and Fermentation Characteristics

As shown in

Table 8. Effects of cutting height on the nutritional compositions and pH of Juncao silage.

Item	Group			SEM	p-Value
	Low	Mid	High
DM, % FM	12.90 Bb	12.65 Bb	14.98 Aa	0.34	<0.01
CP, % DM	13.93 Aa	10.18 Bb	7.85 Cc	0.76	<0.01
ADF, % DM	40.76 Cc	43.40 Bb	45.40 Aa	0.64	<0.01
NDF, % DM	62.98	64.33	65.25	0.42	0.67
CF, % DM	36.75 Cc	39.40 Bb	40.83 Aa	0.56	<0.01
ADL, % DM	3.50 Cc	3.95 Bb	5.01 Aa	0.20	<0.01
Starch, % DM	0.40 b	0.90 a	0.40 b	0.09	0.02
EE, % DM	3.89 Aa	2.83 Bb	2.22 Cc	0.21	<0.01
TDN, % DM	59.58 Aa	57.58 Bb	55.88 Cc	0.50	<0.01
RFV	84.50 Aa	79.75 Bb	76.00 Bb	1.21	<0.01
WSC, % DM	1.64	1.50	1.23	0.10	0.27
pH	4.41 Aa	4.00 Bb	3.32 Cc	0.14	<0.01

Note: Low, cutting height at 100–150 cm, Mid, cutting height at 150–200 cm, High, cutting height at 200–250 cm. Abbreviations: DM, dry matter; FM, fresh matter; CP, crude protein; ADF, acid detergent fiber; NDF, neutral detergent fiber; CF, crude fiber; ADL, acid detergent lignin; EE, ether extract; TDN, total digestible nutrients; RFV, relative feed value; WSC, water-soluble carbohydrate. Note: Within a row, means with different capital superscript letters differ extremely significantly (p < 0.01), and those with different lowercase letters differ significantly (p < 0.05). Means without superscript letters are not significantly different (p > 0.05).

, the contents of ADF, CF, and ADL in Juncao silage increased extremely significantly (p < 0.01) with increasing cutting height. Conversely, the contents of CP, EE, and TDN, as well as pH, decreased extremely significantly (p < 0.01). Regarding the RFV, the Low group was extremely significantly higher than both the Mid and High groups (p < 0.01). The starch content in the Mid group was significantly higher than that in the Low and High groups (p < 0.05), while the DM in the High group was extremely significantly higher than the other two groups (p < 0.01). No significant differences were observed among the groups for the remaining indicators (p > 0.05).

As illustrated in Figure 3, the NH₃-N/TN in the Low group was extremely significantly higher than that in the High group (p < 0.01), while no significant difference was observed between the Low and Mid groups (p > 0.05). The lactic acid in the High group was extremely significantly higher than that in the Low group (p < 0.01), with no significant difference found between the Low and Mid groups (p > 0.05). Similarly, the total VFA in the High group was significantly higher than in the Low group (p < 0.05), whereas the difference between the Low and Mid groups was not significant (p > 0.05). No significant differences were detected among the three groups regarding acetic acid and propionic acid contents (p > 0.05).

As shown in Figure 4, the cutting height of Juncao exhibited a significant positive correlation (p < 0.05) with DM, ADF, CF, ADL, total VFA, and lactic acid contents. Conversely, cutting height was significantly negatively correlated (p < 0.05) with CP, EE, TDN, RFV, pH, NH₃-N/TN, and the contents of acetic acid and propionic acid.

3.3. Comparison Between Natural Fermentation and Microbial Inoculant on Silage Quality

As shown in

Table 9. Comparative effects of natural fermentation and microbial inoculant on nutritional compositions and pH in Juncao silage.

Item	Group		SEM	p-Value
	T1	T2
DM, % FM	14.33 ± 0.51	14.07 ± 0.49	0.41	0.55
CP, % DM	13.47 ± 0.57	12.20 ± 0.44	0.41	0.04
ADF, % DM	45.63 ± 1.07	48.00 ± 0.26	0.63	0.05
NDF, % DM	67.70 ± 1.73	69.93 ± 0.68	1.07	0.11
CF, % DM	40.13 ± 1.10	43.23 ± 0.57	0.72	0.02
ADL, % DM	4.50 ± 0.33	5.64 ± 0.05	0.19	0.02
Starch, % DM	0.43 ± 0.06	0.47 ± 0.21	0.12	0.80
EE, % DM	2.52 ± 0.13	2.11 ± 0.04	0.08	0.02
TDN, % DM	58.10 ± 0.79	54.57 ± 0.51	0.55	0.005
RFV	73.67 ± 3.21	68.67 ± 0.58	1.88	0.11
WSC, % DM	1.57 ± 0.06	1.80 ± 0.10	0.07	0.04
pH	4.06 ± 0.03	4.57 ± 0.78	0.04	0.001

Note: T1, natural fermentation without inoculant, T2, fermentation with microbial inoculant (0.02% w/v). Abbreviations: DM, dry matter; FM, fresh matter; CP, crude protein; ADF, acid detergent fiber; NDF, neutral detergent fiber; CF, crude fiber; ADL, acid detergent lignin; EE, ether extract; TDN, total digestible nutrients; RFV, relative feed value; WSC, water-soluble carbohydrate.

, the CP and EE contents in the T1 group were significantly higher than those in the T2 group (p < 0.05), and the TDN in the T1 group was extremely significantly higher than that in the T2 group (p < 0.01). Conversely, the CF, ADL contents, and WSC in the T1 group were significantly lower than those in the T2 group (p < 0.05), while the pH value in the T1 group was extremely significantly lower than that in the T2 group (p < 0.01).

As illustrated in Figure 5, by comparing the fiber degradation effects of the T1 and T2 groups against the fresh forage, it was observed that the ADF content in the T1 group was significantly reduced (p < 0.05). Furthermore, the NDF and ADL contents in the T1 group decreased extremely significantly (p < 0.01). These results indicate that natural fermentation exhibits a superior fiber degradation effect compared to the addition of silage inoculants.

As illustrated in Figure 6, the concentrations of total VFA and lactic acid in the T1 group were significantly higher than those in the T2 group (p < 0.05).

4. Discussion

4.1. Implications from the Comprehensive Evaluation of On-Farm Silage Quality

In this study, GRA was applied to comprehensively evaluate 25 on-farm Juncao silage samples. This method allows for a more holistic and accurate objective assessment [22]. The comprehensive analysis revealed that the top-ranked treatments were J2, J3, J6, X6, J5, J4, J1, X7, C4, and S1. These treatments all had sensory evaluation scores above 15, TDN contents greater than 52.8%, pH values below 4.2, total VFA contents above 7.14%, and DM contents ranging from 18.4% to 29.0%. Furthermore, carbon supplementation during ensiling, through materials such as cornmeal or rice bran, increased the DM content and enhanced the nutritional quality of the silage. For example, compared with the untreated group (J1), the addition of 4% cornmeal (J2) increased DM content by 25.37%, TDN content by 10.16%, reduced NDF content by 27.20% and ADF content by 26.98%, lowered pH by 12.32%, and raised lactic acid content by 9.14%. These findings are consistent with those reported by Wu et al. [23], who also observed that adding 3–9% cornmeal significantly improved fermentation characteristics and nutritional compositions in silage.

Additionally, correlation analysis was conducted on these 25 samples, which further identified that natural fermentation was significantly negatively correlated with DM content, while hay inclusion showed a significant negative correlation with total VFA. This could be attributed to the fact that lower DM content often led to contamination in silage, which subsequently reduced the concentration of total VFA [24]; or to the fact that hay inclusion increased the DM content during ensiling, but the resulting poor compaction might have resulted in inferior fermentation characteristics [25]. Also, undesirable microorganisms attached to the hay surface might have also contributed, the underlying mechanisms of which require further investigation. Furthermore, carbon supplementation showed a significant positive correlation with sensory quality and nutritional compositions. Current research generally maintains that incorporating carbon supplementation during the ensiling process can accelerate fermentation and enhance silage quality [23,26]. These additives not only provide sufficient substrates for the rapid accumulation of lactic acid and a sharp decline in pH but also promote the synthesis of microbial proteins, thereby improving the overall nutritional profile [27]. These findings are consistent with our results and demonstrate that carbon supplementation effectively improves Juncao silage quality.

4.2. Balancing Nutritional Compositions and Fermentation Characteristics at Optimal Cutting Height

Ensiling is one of the effective methods for improving nutritional quality and extending storage life. Generally, ensiling requires controlling moisture content between 60% and 70% and having a certain buffering capacity. However, for Juncao, its relatively high contents of WSC and CP make it a suitable substrate for anaerobic fermentation [28]. In this study, the nutritional compositions of Juncao silage showed that DM, ADF, ADL, and CF increased with cutting height, while CP, EE, TDN, and RFV decreased. Regarding fermentation characteristics, total VFA and lactic acid contents increased with cutting height, whereas the NH₃-N/TN, propionic acid content, and pH decreased. These trends were corroborated by our correlation analysis, showing significant positive correlations of cutting height with DM, ADF, ADL, CF, total VFA, and lactic acid, and significant negative correlations with CP, EE, TDN, RFV, pH, NH₃-N/TN, acetic acid, and propionic acid. These results were consistent with the findings of Zhao et al. [29]. Furthermore, the directional trends observed in our study for CP, ADF, RFV, and fermentation characteristics were in agreement with those reported by Li et al. [30] in a separate study.

The pH serves as a primary visual and biochemical indicator for evaluating silage fermentation quality. In this study, the pH of Juncao silage decreased significantly as the cutting height increased, reaching 4.01 and 3.33 in the 150–200 cm and 200–250 cm groups, respectively. These results indicate that cutting Juncao at a height between 150 cm and 250 cm resulted in a pH decrease to a favorable range (approximately 4.0 or below). This change was primarily driven by lactic acid. Current research generally identifies lactic acid as the predominant factor influencing silage pH [31], which was consistent with the lactic acid results observed in this study. This indicated that higher levels of lactic acid were produced during the ensiling of Juncao when cut above 150 cm. The resulting low-pH environment helped stabilize fermentation by inhibiting acid-intolerant microorganisms [32]. Regarding the effect of cut height on pH, previous studies on buffel grass silage reported that higher cut heights resulted in significantly higher pH values compared to lower heights; however, similar to our findings, pH showed a linear decrease as height increased [33]. This phenomenon could be attributed to the linear increase in DM content with cut height. Although the DM content in this study was relatively low, the reduction in CP and WSC after ensiling provided energy for Juncao fermentation, which aligned with the research of Zhao et al. [28,29]. Furthermore, Juncao contains a certain amount of starch. A higher carbohydrate content typically increases microbial abundance in silage, allowing LAB to utilize WSC to produce substantial amounts of lactic acid [34], thereby promoting a rapid decline in pH [35].

NH₃-N/TN is considered the most reliable variable for detecting the fermentation quality of silage. Our results demonstrated that the NH₃-N/TN concentration in Juncao silage from the 200–250 cm group was extremely significantly lower than in the other two groups. A lower proportion of NH₃-N indicated reduced protein degradation, which enhanced the apparent digestibility and overall quality of the silage [36]. Some researchers have pointed out that the NH₃-N/TN in Juncao silage produced at a later growth stage was significantly reduced by 40.12%, and therefore recommended harvesting Juncao at a later growth stage for anaerobic fermentation to produce silage, in order to optimize fermentation quality [29].

Silage fermentation is generally divided into six distinct stages: aerobic phase, acetic acid fermentation, initiation of lactic acid fermentation, completion of lactic acid fermentation, storage, and feeding [37]. This general finding diverged from the pattern of higher acetic acid concentrations in low-DM silage [38], where acetic acid was reported as a crucial precursor for milk fat synthesis in ruminants [39]. However, our findings diverged from this pattern. In this study, although DM content increased with cutting height, correlation analysis indicated a significant negative correlation between acetic acid content and cutting height. This discrepancy may be attributed to the specific types of LAB present. Mainstream theory classifies LAB into homofermentative and heterofermentative strains [40]. Certain heterofermentative LAB were reported to degrade lactic acid into acetic acid and propionic acid [41]. In our results, the trends for propionic and acetic acids were similar, while the lactic acid trend was the opposite. This suggests that during Juncao ensiling, the present LAB did not degrade lactic acid into other organic acids but instead facilitated lactic acid accumulation, further depressing the pH.

In conclusion, while elevated cutting heights led to a decline in nutritional components such as CP, TDN, and RFV along with an increase in fiber fractions including ADF, ADL, and CF, they consistently promoted enhanced fermentation, as evidenced by desirable pH, high lactic acid, and low NH₃-N/TN. Balancing these effects, a cutting height of 200–250 cm is proposed to maximize the overall fermentation quality of Juncao.

4.3. Efficacy and Mechanisms of Natural Fermentation vs. Microbial Inoculant

For Juncao with high moisture content, strategies such as adding materials to increase DM [25], wilting to reduce moisture [42], applying microbial inoculant [43], and mixing microorganisms with fermentation substrates [23] can promote lactic acid production during ensiling, inhibit harmful microorganisms, and enhance the nutritional value and fermentation quality of the silage.

While adding microbial inoculants is generally expected to increase protein content and minimize nutrient loss [44], the CP content in the microbial inoculant group of this study was significantly lower than that in the natural fermentation group, contradicting this expectation. This finding aligns with the conclusion of Chen et al. [45], who reported that adding Lactiplantibacillus plantarum and Bacillus licheniformis to Pennisetum sinese silage significantly reduced NH₃-N, NDF, and ADF contents and pH compared to natural fermentation, but also resulted in decreased CP content. The reduced CP content observed in both studies might be attributed to the abundance of epiphytic (endogenous) bacteria naturally present on the forage, which may exhibit an antagonistic effect when exogenous LAB strains are supplemented.

Regarding fiber degradation, natural fermentation significantly reduced the contents of NDF, ADF, and ADL, whereas the addition of microbial inoculants showed no significant impact. This finding contrasts with several mainstream studies in which additives are typically employed to degrade cellulose. For instance, Dong et al. [46] found that adding Lactobacillus plantarum and cellulase significantly reduced fiber content in Pennisetum Purpureum silage. Similarly, Silva et al. [47] reported a linear decrease in NDF with increasing cellulase levels, achieving the lowest ADF at a 4.5% inclusion rate. The inverse results observed in this study may be explained by the higher lactic acid levels in the natural fermentation group; the sustained low pH likely facilitated the acid hydrolysis of plant cell walls, leading to the reduction in NDF and ADF values. However, other studies have also indicated that additives such as Lacticaseibacillus rhamnosus and Lentilactobacillus buchneri had no significant effect on NDF and ADF contents [23].

Regarding fermentation quality, in this study, the pH of naturally fermented Juncao silage was significantly lower than that of the groups with microbial inoculants added, while the contents of total VFA and lactic acid were significantly higher. In contrast, some studies suggest that adding microbial inoculants can improve fermentation quality, increase the abundance of dominant bacterial communities, elevate lactic acid content, and lower pH [48]. However, Tian et al. [49] reported that among additives such as malic acid, glucose, cellulase, or Bacillus subtilis applied to hybrid Pennisetum silage, only malic acid significantly reduced pH, while the other methods showed no significant effect. Zhu et al. [50] found that adding Lactobacillus plantarum or Bacillus licheniformis during ensiling did not significantly lower pH, and an increase in lactic acid content was only observed with L. plantarum, and only when added at 1 × 10⁷ CFU/g. Wu et al. [23] concluded that adding microbial inoculants along with an appropriate amount of cornmeal could improve fermentation quality. However, when the cornmeal addition reached 6%, the effect of the microbial inoculants diminished, showing no clear impact on pH, whereas the role of cornmeal increased, and the lactic acid content peaked when no inoculant was added but with 6% cornmeal.

Therefore, we speculate that in this study, adding microbial inoculants such as Lactococcus lactis and Lactobacillus buchneri to Juncao silage did not significantly improve its post-ensiling nutritional compositions or fermentation characteristics. This may be related to the types and dosage of microorganisms added, as well as the type and amount of fermentation substrates. Additionally, the DM content of Juncao in this study was only 15.27%, which is below 20%. The lack of significant improvement in Juncao silage quality upon adding microbial inoculant may be due to the fact that the key factors determining successful ensiling are WSC and DM contents, rather than the addition of microbial inoculant [51].

Although this study comprehensively evaluated on-farm silage quality, determined the optimal cutting height, and compared natural fermentation with microbial inoculant, certain limitations remain. The selection of microbial inoculant was not exhaustive; instead, a mainstream microbial inoculant was utilized. Given that Juncao is a tropical plant rich in endogenous microbiota, future research should prioritize comprehensive microbial profiling and strain identification to develop specialized inoculant tailored to this species. Nevertheless, this study demonstrates that compared to conventional microbial inoculant additives, natural fermentation of Juncao can achieve excellent results. This finding not only suggests a potential reduction in input and labor costs but also provides a theoretical foundation and new perspectives for subsequent research on high-quality Juncao silage production.

5. Conclusions

This study comprehensively compared the fermentation of Juncao silage and demonstrated that carbon supplementation (e.g., cornmeal, rice bran) significantly improved sensory evaluation and nutritional quality. With increasing cutting height, fiber content rose while CP and TDN declined; however, fermentation characteristics improved markedly, reflected in higher lactic acid, lower pH, and reduced NH₃-N/TN. Balancing nutritional and fermentation outcomes, a cutting height of 200–250 cm is recommended as optimal. Furthermore, for mature Juncao, natural fermentation outperformed microbial inoculant treatment, particularly in fiber degradation, pH control, and lactic acid accumulation.