Is Grass Planting Suitable for Orchard Sustainability in Xizang? Insights from the Ecosystem Services Valuation of a 4-Year Apple Orchard Grass Planting Practice

Abstract

Orchard grass intercropping offers a promising strategy to mitigate forage scarcity and boost fruit yield. However, its applicability in high-altitude regions such as the Xizang Plateau remains poorly understood. During the four-year experiment, the ecosystem service values (ESV) of an apple orchard intercropped with different grass species in Linzhi, southeast Xizang, were investigated in order to assess the applicability of orchard grass planting at high altitudes and identify optimal grass species combinations. Seven treatments were established, including six artificial grass systems (annual legume, annual gramineous, annual legume + gramineous, perennial legume, perennial gramineous, perennial legume + gramineous) and a natural grass control group. Results showed that artificial grass systems, particularly perennial ones, yielded higher total ESV than the natural grass control. Perennial grasses outperformed annual grasses and natural grass in provisioning services (with higher fruit and forage yields) and regulating services (with enhanced carbon sequestration and oxygen production). However, perennial grasses also led to reduced supporting services, primarily due to decreased soil nutrient availability (especially available phosphorus and potassium) and lower plant diversity. The optimal treatments were Dactylis glomerata monoculture and Medicago sativa + D. glomerata mixed culture, which achieved the highest total ESV. Notably, the nutrient depletion observed in perennial grass treatments highlighted the need for supplemental fertilization to ensure long-term sustainability of the system. In conclusion, artificial orchard grass systems significantly enhanced total net ESV in high-altitude regions, whereas individual ecosystem services demonstrated divergent responses to different grass species. Specifically, D. glomerata monoculture and M. sativa + D. glomerata intercropping emerged as the optimal orchard grass patterns in Linzhi. To maintain long-term orchard productivity, adaptive fertilizer management strategies are recommended to counteract potential soil nutrient depletion associated with these grass cultivation systems.

1. Introduction

The Xizang Plateau, renowned as the “Roof of the World” with an average altitude exceeding 4000 m, constitutes a unique and highly ecologically sensitive region [1]. As a crucial provider of multiple ecosystem services, it faces escalating threats from climate change and intensified human activities, particularly in the context of grassland-livestock husbandry, which has a three-millennia-long history on the plateau [2]. In recent decades, the surging demand for livestock products has led to a sharp increase in grazing intensity [3]. Coupled with the adverse impacts of global climate change on forage production [4], this overgrazing has severely degraded the grassland ecosystem [5]. To address this challenge and ensure the sustainable development of the Xizang Plateau, it is imperative to adopt a multi-strategy approach to augment available forage resources and balance livestock needs with ecological conservation.

The establishment of orchard-grass intercropping ecosystems presents a promising strategy to alleviate feed shortages. Firstly, the orchard area in Xizang nearly doubled from 4.20 to 7.44 thousand hectares between 2020 and 2023 (National Bureau of Statistics of China) [6]. These orchards are primarily established on barren land to avoid competing with cropland, and “green cultivation”—typically orchard grass planting—has been vigorously promoted by the local government [6,7]. Given the expanding orchard areas, utilization of barren land, and policy-backed green cultivation practices, the orchard grass planting ecosystem has established a solid foundation for producing additional forage, offering a practical pathway to alleviate feed shortages on the Xizang Plateau. Secondly, in lower-altitude regions of China (e.g., Fujian, Zhejiang, Guangdong, Guangxi, and Hubei provinces), orchard grass planting has been shown to increase soil water infiltration, maintain soil fertility, and promote soil microbial abundance, diversity, and extracellular enzyme activities, thereby benefiting fruit tree growth [8,9,10,11,12]. A meta-analysis showed that orchard grass planting increased fruit yield by an average of 20.7% compared with clear tillage [13], with leguminous species outperforming gramineous or mixed systems. Thus, by integrating forage production into orchard systems, this approach holds the potential to simultaneously address feed shortages and enhance fruit production—a critical need for fragile ecosystems in Xizang.

However, three key knowledge gaps impede the application of orchard grass planting on the Xizang Plateau. Firstly, most prior studies were conducted in lower-altitude regions. Given the unique environmental conditions of the Xizang Plateau, characterized by extreme cold, strong ultraviolet radiation, and fragile soil, the applicability and performance of orchard grass planting in higher-altitude regions remain largely unexplored. Secondly, the utilization of grass in lower-altitude orchards mainly includes turnover and mulching, both of which leave fresh grass in the field [13]. Mineral nutrients and large amounts of organic matter contained in grass will return to the soil during the later decomposition process, thereby contributing to the fruit production by improving soil fertility [14]. However, in Xizang, fresh grass is harvested as forage and removed from the orchard, which may alter soil nutrient cycling and requires specific evaluation. Thirdly, previous studies focused solely on snapshot assessments of fruit yield, fruit quality, or soil fertility improvement, neglecting forage yield and quality—critical metrics for livestock systems in Xizang. In addition, the fragile ecological environment in Xizang necessitates sustained monitoring of the orchard grassland ecosystem dynamics to ensure the long-term sustainability of grass-orchard integration.

Ecosystems provide a wide range of direct and indirect services, including provisioning, regulating, and supporting services [15]. An economic valuation of these services related to orchard grass planting will provide a more comprehensive consideration of the range of impacts on the environment, income benefits, and production sustainability [1]. Thus, this study investigates an apple orchard intercropped with different grass species in Linzhi, southeastern Xizang. Using the market price method, alternative cost method, and other ecological economic approaches, we quantified the ecosystem service value (ESV). We hypothesized that (1) artificial orchard grass planting enhances the total net ESV in an apple orchard compared to natural grass cover; (2) individual ecosystem services exhibit differential responses across grass species. The objectives of this study were to: (1) characterize the dynamic performance of orchard grass planting at high altitudes; (2) identify optimal grass species combinations and planting systems, and propose sustainable orchard grass management strategies tailored to plateau-specific environmental constraints.

2. Materials and Methods

2.1. Study Area

The Xizang Plateau has a long history of separated, specialized crop and forage production due to the different climate conditions. Rangelands are mainly distributed in the northern Xizang Plateau with cold and dry environment [16]. In contrast, Linzhi, located in southeast Xizang, has relatively moderate weather and has become the most important fruit production area on the plateau. The investigated apple orchard is situated in Bujiu, Bayi District of Linzhi City (29°34′14.73″ N, 94°25′56.10″ E). The area has an average altitude of 3005 m and features a plateau temperate humid monsoon climate. The annual sunshine duration in this area exceeds 2022 h, with an average annual temperature of 8.7 °C. The annual precipitation is 650 mm, primarily concentrated in summer (June to August). The soil is classified as yellow brown soil with a sandy loam texture. The apple cultivar is Black Diamond (Malus pumila), with a tree age of 15 years, a crown width of 2.5 m × 2 m, and a tree height of 3 m; all trees were in the full fruit stage. Prior to the experiment, initial soil chemical properties were analyzed: pH was 6.67, bulk density was 1.24 g/cm³, total carbon (TC) content was 11.25 g/kg, and available nitrogen (AN), available potassium (AK), and available phosphorus (AP) contents were 32.67, 176.05, and 21.12 mg/kg, respectively.

2.2. Experiment Design

Four different grass species, including annual legume (Vicia sativa), annual gramineous (Avena sativa), perennial legume (Medicago sativa), and perennial gramineous (Dactylis glomerata), were introduced into the orchard since 2021. The soil between apple tree rows was ploughed using a rotary tiller to remove the original grass. Then six artificial orchard grass systems and one natural grass control were established as follows: (1) annual legume single sowing (A_L); (2) annual gramineous single sowing (A_G); (3) annual legume and gramineous mixed sowing (A_L+G); (4) perennial legume single sowing (P_L); (5) perennial gramineous single sowing (P_G); (6) perennial legume and gramineous mixed sowing (P_L+G); and (7) natural grass (no sowing) (N_G). Each treatment included three replicates, with a plot area of 300 m².

For annual grasses, seeds were sown every May at a rate of 15 g/m² of legume for A_L, 30 g/m² of gramineous for A_G, and 10.5 g/m² of legume + 21 g/m² of gramineous for A_L+G. For perennial grasses, seeds were sown only at the beginning of the experiment at a rate of 1.5 g/m² of legume for P_L, 3 g/m² of gramineous for P_G, and 1.05 g/m² of legume + 2.1 g/m² of gramineous for P_L+G. These seeding rates were determined based on the growth characteristics of each grass species and previous results [17]. Throughout the experiment, irrigation, fertilization, and orchard maintenance were kept consistent across all treatments, adhering to traditional management practices. Chemical fertilizers, including urea (200 kg N/ha), calcium superphosphate (130 kg P₂O₅/ha), and potash (50 kg K₂O/ha), were furrow-applied in October.

2.3. Sampling and Determination

Apple fruit sampling was conducted annually in October, coinciding with the full fruit ripening period (typically late October, when fruit skin coloration exceeded 70% red pigmentation based on visual assessment). The apple yield of each treatment plot was determined by complete harvesting of all fruits from mature trees. Then yield was calculated as kg/ha by normalizing to the tree density and plot area. Fifteen individual fruits were randomly selected from each plot, with sampling stratified across tree canopy positions (5 fruits from the upper east, upper west, middle, lower east, and lower west zones to minimize spatial bias). Each fruit was weighed using an electronic balance (accuracy ±0.1 g), and the average single fruit weight (g) was calculated per plot. Total soluble solid (TSS) was measured using a handheld refractometer (PAL-1, ATAGO Co., Ltd., Kyoto, Japan) with the following protocol: Peel a 1 cm² section of fruit skin, squeeze 2–3 drops of juice onto the refractometer prism. Read TSS values at 20 °C (automatic temperature compensation function enabled), with 3 replicates per fruit and an average recorded per plot.

Grass diversity was investigated annually in August, during the full bloom stage of the forage. The percent coverage of grass in each plot was determined using a photogrammetric method. In each plot, the grass community was investigated within three randomly selected quadrats (1 m × 1 m) placed on grass-covered patches between tree rows. The grass species composition, coverage, and height were surveyed, and species were identified to the species level in the laboratory. Aboveground grass biomass was then harvested to determine forage yield, and crude protein content of the forage was determined using the Kjeldahl method.

Soil samples were collected annually in August. As root depth investigations in grasslands indicate that 51–100% of total root biomass is contained in the upper 30 cm of soil, and root biomass at 30–60 cm depth increases with grass species richness [18], soil from a depth of 0–50 cm was collected. Using a soil auger, samples were taken from ten randomly selected points in each plot and mixed into one composite sample. Soil AN, AP, and AK were determined by the diffusion method, Olsen method, and flame photometric method, respectively. Soil TC content was determined by the dichromate oxidation method. Soil bulk density was determined by the ring-knife method. Soil water content was determined using the time-domain reflectometry (TDR) method: the TDR probe (TDR300, Spectrum Technologies Ltd., Haltom, TX, USA) was inserted into the soil, and data were automatically collected at noon each day. The average soil water content for each year was then calculated.

2.4. Construction of Ecosystem Services Value Evaluation Index System

Based on the classification of ecosystem services of Costanza et al. [19], Millennium-Ecosystem-Assessment [20], and Xie et al. [21], and considering the characteristics of apple orchard ecosystems, the ecosystem services in the present study were divided into three categories: provisioning, regulating, and supporting services, with eight evaluation indices. The methods for calculating the economic values of specific components of ecosystem services are shown in

Table 1. Connotation definition of the ESV evaluation system.

Categories		Value Evaluation Based on Ecological Economics
Provisioning services (V1, CNY/ha)	Fruit yield (V11, CNY/ha)	V11 = M11 × P11, M11 (t/ha) is the yield of apple. P11 (CNY/kg) is the price of apple.
	Forage yield (V12, CNY/ha)	V12 = M12 × P12, M12 (t/ha) is the yield of forage. P12 (CNY/kg) is the price of forage.
	Cost of production (V13, CNY/ha)	V13 = ∑ Im × Pm, Im includes material costs, such as fertilizer, pesticides, seeds, and labor force. Pm represents the market price of m (orchard management costs, such as fertilization, seeds, and labor, Table S1).
Regulating services (V2, CNY/ha)	Water conservation (V21, CNY/ha)	V21 = V21s + V21f, V21s (CNY/ha) and V21f (CNY/ha) are the value of water conservation by soil and forage, respectively. V21s = Ws× Pw, V21f = Wf × Pw, Ws (t/ha) is the amount of water held in 0–50 cm soil, which is calculated by the soil water content and soil bulk density. Wf (t/ha) is the amount of water hold in forage tissue, which is calculated by the difference between fresh and dry biomass of forage. Pw is agricultural irrigation water price at 0.4 CNY/t.
	Carbon sequestration (V22, CNY/ha)	V22 = V22s + V22f, V22s (CNY/ha) and V22f (CNY/ha) is the value of carbon sequestration in soil and forage, respectively. V22s = Cs × Pc, V22f = Cf × Pc Cs (t/ha) is the change in soil sequestrated carbon in the 0–50 cm soil layer between two adjacent years, which is calculated based on the soil total carbon content and soil bulk density determined in each of the two adjacent years, respectively. Cf (t/ha) is the total carbon sequestrated in forage tissue. Pc is the afforestation cost for carbon fixation, at price of 1200 CNY/t.
	Oxygen production (V23, CNY/ha)	V23 = Cf × 2.67 × Po, According to the equation of photosynthesis, 2.67 g of oxygen (O2) are released for every 1 g of carbon (C) fixed. Po is the price of oxygen production at price of 1000 CNY/t.
Supporting services (V3, CNY/ha)	Soil nutrient maintenance (V31, CNY/ha)	V31 = ANs × PN + APs × PP + AKs × PK, ANs (kg/ha), APs (kg/ha), and AKs (kg/ha) are the change in available N, P, and K storage in 0–50 cm soil layer between two adjacent years, which is calculated based on the soil available nutrients content and soil bulk density determined in each of the two adjacent years, respectively. PN, PP, and PK are chemical fertilizer price for N, P, and K, respectively (Table S1).
	Biodiversity conservation (V32, CNY/ha)	The Shannon–Wiener index of plant diversity was calculated at the forage planting area. The Shannon–Wiener index was divided into seven level according to Liu. The value of biodiversity conservation was calculated based on the unit value of each level of the Shannon–Wiener index (Table S2) multiplied by the area of forage planting.

.

(1)

Provisioning services (V₁)

Ecosystem provisioning services refer to the benefits that ecosystems provide in the form of tangible goods. Fruit and forage sales constituted the primary source of economic revenue. However, part of these profits is used to cover daily orchard management costs, such as fertilization, seeds, and labor (Table S1). Thus, provisioning services were determined by the net economic value of fruit (V₁₁) and forage (V₁₂) production, calculated as total economic yield minus management costs (V₁₃) [22]:

V₁ = V₁₁ + V₁₂ − V₁₃

(1)

(2)

Regulating services (V₂)

Ecosystem regulating services refer to the benefits derived from the regulation of ecosystem processes that maintain environmental balance and stability. Soil and vegetation can retain water, reduce runoff, and improve water infiltration. In orchard grass systems, grass tissue can retain water, and grass coverage may reduce soil evapotranspiration, thereby benefiting soil water storage [11]. Thus, the economic value of water retained (V₂₁) in soil and grass was determined based on the agricultural irrigation water price of 0.4 CNY/t [23].

Plants function as biological “pumps”, dynamically transferring atmospheric carbon to plant biomass through the process of photosynthesis. Specifically, grasses fix carbon dioxide, with a fraction of this fixed carbon being converted into soil organic carbon compounds through dual mechanisms: the decomposition of plant litter (including fallen leaves, stems, and roots) and the release of root exudates. Concurrently, another portion of carbon may be physically removed from the orchard ecosystem during forage harvest, representing a carbon loss pathway. To quantify these carbon fluxes, the present study calculated two key parameters: (1) Changes in soil total carbon content: Tracking the net accumulation or depletion of carbon in the soil matrix, influenced by litter decomposition and root exudation. (2) Carbon removal via forage harvest: Measuring the mass of carbon exported from the system when forage is harvested. Subsequently, the economic value of carbon sequestration (V₂₂) was derived by referencing afforestation costs for carbon fixation, using a benchmark price of 1200 CNY/t [15]. Plants convert carbon dioxide and water into glucose and release oxygen (O₂) using light energy in chloroplasts, which is the primary source of atmospheric oxygen. The equation of photosynthesis (6CO₂ + 6H₂O → C₆H₁₂O₆ + 6O₂) reveals a fixed molar ratio of carbon (C) to oxygen (O₂) released during photosynthesis: for every 6 moles of CO₂ fixed (containing 6 moles of C), 6 moles of O₂ are produced. This simplifies to a 1:1 molar ratio of C to O₂, translating to a mass ratio of 12:32 (or 3:8) based on atomic weights (C = 12 g/mol, O₂ = 32 g/mol). Thus, for every gram of carbon sequestered by grasses, 8/3 g of oxygen are generated. Based on this equation, the economic value of oxygen production by grasses (V₂₃) was determined using grass carbon content and the industrial oxygen production price of 1000 CNY/t [15]. Overall, the economic value of regulating services (V₂) was estimated as

V₂ = V₂₁ + V₂₂ + V₂₃

(2)

(3)

Supporting services (V₃)

Ecosystem supporting services are the invisible vital processes that sustain ecosystems’ functional and sustainable. Soil nutrient content denotes the soil’s capacity to sustain plant growth, which is fundamentally linked to the availability of macronutrients—specifically nitrogen (N), phosphorus (P), and potassium (K) [24]. Thus, changes in soil available N, P, and K were calculated to determine the soil nutrient maintenance value (V₃₁).

Compared with traditional orchards, orchard grass planting is highly valued for its contribution to the spatial heterogeneity of agricultural ecosystems [25]. Different grass species are expected to provide the habitat and diverse food resources required for arthropod predators and parasitoids, insectivorous birds and bats, and microbial pathogens that act as natural enemies to agricultural pests and provide biological control services in agroecosystems [26]. As pest abundance is highly dynamic and difficult to quantify, only plant species were identified in this study. The Shannon–Wiener index of plant diversity was calculated and classified into seven levels according to Liuet al. [15] (Table S2). Then, the value of biodiversity conservation (V₃₂) was calculated based on the unit value for each level of the Shannon–Wiener index. Overall, the economic value of supporting services was estimated as

V₃ = V₃₁ + V₃₂

(3)

(4)

Total Ecosystem Service Value Calculation (V)

The total ecosystem service value was calculated as the sum of values from the three abovementioned functional categories:

V = V₁ + V₂ + V₃

(4)

2.5. Statistical Analyses

Statistical analyses were performed using SPSS 19.0 (SPSS Inc., Chicago, IL, USA). The effects of different treatments on response variables were assessed by one-way analysis of variance (ANOVA) and least significant difference (LSD) at the 5% significance level.

3. Results

3.1. Economic Values of Provisioning Services

Provisioning services encompassed fruit and forage sales, net of daily orchard management costs. Apple yield fluctuated between 16.64 and 19.04 t/ha during the investigation period (

Table 2. Fruit yield, quality, and market price.

Year	Index	NG	AL	AG	AL+G	PL	PG	PL+G
2021	Yield (t/ha)	17.46 b	18.68 a	18.09 ab	18.57 a	18.42 a	18.23 ab	18.21 ab
	Fruit weight (g/individual)	211.45 a	210.47 a	213.42 a	215.33 a	214.17 a	218.22 a	220.17 a
	TSS (%)	12.37 a	12.44 a	12.15 a	12.35 a	11.98 a	12.08 a	12.44 a
	Market price (CNY/kg)	18.00	18.01	17.92	18.15	17.83	18.08	18.42
2022	Yield (t/ha)	17.18 b	18.05 ab	17.85 b	18.28 a	18.77 a	18.14 ab	18.20 ab
	Fruit weight (g/individual)	208.04 b	229.16 ab	210.80 b	234.22 a	248.39 a	243.68 a	237.70 a
	TSS (%)	10.97 b	10.47 b	11.40 b	11.20 b	9.27 b	11.40 ab	12.93 a
	Market price (CNY/kg)	20.00	20.56	20.52	21.47	20.39	22.11	23.22
2023	Yield (t/ha)	16.64 b	17.82 ab	16.92 b	17.48 ab	18.21 ab	19.04 a	19.02 a
	Fruit weight (g/individual)	209.88 b	219.54 ab	222.86 ab	218.06 ab	226.47 a	231.72 a	232.40 a
	TSS (%)	11.70 ab	12.13 ab	12.50 a	12.26 ab	11.47 b	12.39 a	12.81 a
	Market price (CNY/kg)	21.00	21.87	22.37	21.91	21.62	22.71	23.12
2024	Yield (t/ha)	16.69 b	17.50 b	17.64 ab	18.07 ab	18.46 a	18.70 a	19.04 a
	Fruit weight (g/individual)	210.47 a	213.14 a	210.22 a	214.87 a	209.41 a	213.47 a	212.84 a
	TSS (%)	11.87 a	12.66 a	12.08 a	11.86 a	11.88 a	12.73 a	12.44 a
	Market price (CNY/kg)	21.00	21.83	21.17	21.21	20.96	21.91	21.62

Note: According to the average market price of Black Diamond apples in Linzhi City, the fruit prices of the CK group were 18, 20, 21, and 21 CNY/kg for 2021, 2022, 2023, and 2024, respectively. For other treatment groups, the fruit prices were calculated based on the quality comparison with the CK group, using the following formula: Price _treatment = Price _CK × [(Fruit weight _treatment − Fruit weight _CK)/Fruit weight _CK + (TSS _treatment − TSS _CK)/TSS _CK]/2 + Price _CK. Different lowercase letters in the same line indicated significantly different (p < 0.05).

). Generally, artificial grass treatments (A_L, A_G, A_L+G, P_L, P_G, P_L+G) resulted in higher apple yields than natural grass (N_G). Among artificial grass planting treatments, perennial grass (P_L, P_G, P_L+G) performed better than annual grass (A_L, A_G, A_L+G). For most of the time, there were no significant differences in fruit quality indicators (total soluble sugar and individual fruit weight) among treatments (Table 2). However, artificial grass treatments (A_L, A_G, A_L+G, P_L, P_G, P_L+G) showed a slight improvement in fruit quality over time, leading to higher market prices compared with natural grass (N_G). Specifically, the market price for P_L+G increased from 18.42 CNY/kg in 2021 to 23.12 CNY/kg in 2023.

Natural grass (N_G) had the lowest forage yield (3.16–4.25 t/ha) (

Table 3. Forage yield, quality, market price, and species diversity.

Year	Index	NG	AL	AG	AL+G	PL	PG	PL+G
2021	Yield (t/ha)	3.16 b	4.33 a	4.11 ab	4.44 a	4.40 a	4.79 a	4.49 a
	Crude protein (%)	11.48 b	15.14 a	12.91 ab	15.27 a	14.32 a	12.27 ab	14.12 a
	Market price (CNY/kg)	1.70	2.24	1.91	2.26	2.12	1.82	2.09
	Water content (%)	71.60 a	68.00 a	72.20 a	69.20 a	74.70 a	73.70 a	75.00 a
	Shannon–Wiener index	1.14 ab	0.98 ab	1.14 ab	1.32 a	0.87 b	0.68 b	0.74 b
	Total C content (g/kg)	40.66 a	39.14 a	39.16 a	38.57 a	42.20 a	39.75 a	41.60 a
2022	Yield (t/ha)	3.48 c	4.53 b	4.50 b	4.68 b	6.41 bc	11.55 a	9.10 a
	Crude protein (%)	11.73 bc	17.04 a	13.31 b	14.47 b	17.32 a	10.47 c	14.32 b
	Market price (CNY/kg)	1.70	2.47	1.93	2.10	2.51	1.52	2.08
	Water content (%)	72.20 a	67.30 a	71.80 a	69.50 a	74.10 a	74.70 a	75.80 a
	Shannon–Wiener index	1.39 ab	0.82 b	1.55 a	1.52 a	0.91 b	0.46 c	0.60 bc
	Total C content (g/kg)	38.03 a	37.97 a	33.87 b	38.27 a	38.92 a	37.84 a	39.38 a
2023	Yield (t/ha)	3.73 d	7.35 c	4.88 d	5.58 cd	10.43 b	15.35 a	14.48 a
	Crude protein (%)	10.98 c	16.64 ab	13.51 b	13.47 bc	19.22 a	11.27 c	13.52 b
	Market price (CNY/kg)	1.70	2.58	2.09	2.09	2.98	1.74	2.09
	Water content (%)	70.60 a	67.20 a	72.90 a	69.80 a	75.20 a	73.20 a	74.90 a
	Shannon–Wiener index	1.28 b	1.62 a	1.05 b	1.73a	1.14 b	0.11 c	0.22 c
	Total C content (g/kg)	40.35 a	40.12 a	41.99 a	40.73 a	42.12 a	40.69 a	41.79 a
2024	Yield (t/ha)	4.25 c	6.67 bc	4.62 c	4.81 c	8.99 b	12.38 a	12.44 a
	Crude protein (%)	11.24 b	17.24 a	12.51 b	14.47 ab	18.22 a	11.47 b	14.32 ab
	Market price (CNY/kg)	1.70	2.61	1.89	2.19	2.76	1.73	2.17
	Water content (%)	70.60 a	67.20 a	71.30 a	68.20 a	75.50 a	73.60 a	75.90 a
	Shannon–Wiener index	1.30 ab	1.58 a	1.34 ab	1.65 a	0.92 b	0.18 c	0.19 c
	Total C content (g/kg)	33.79 a	37.06 a	37.63 a	40.07 a	36.54 a	34.95 a	36.59 a

Note: The forage price of CK is 1.7 CNY/kg according to the average market price of forage in Linzhi City, while the forage price of other treatments is calculated according to the quality of forage compared with CK, using the formula: Price _treatment = Price _CK × [(Crude protein _treatment − Crude protein _CK)/Crude protein _CK] + Price _CK. Different lowercase letters in the same line indicated significantly different (p < 0.05).

). Annual grass treatments (A_L, A_G, A_L+G) maintained relatively stable forage yields (4.11–7.35 t/ha) throughout the experiment. Perennial grass treatments (P_L, P_G, P_L+G) showed a ~3-fold increase in forage yield from 2021 (4.4–4.79 t/ha) to 2023 (10.43–14.48 t/ha), with a slight decrease in 2024 (8.99–12.44 t/ha). Legume grasses (A_L, P_L) had the highest crude protein content (14.32–19.22%), followed by legume–gramineous mixed grasses (A_L+G, P_L+G) (13.47–15.27%), then gramineous grasses (A_G, P_G) (10.47–13.51%), and finally natural grasses (N_G) (10.98–11.73%). Due to their high crude protein content, legume grasses (A_L, P_L) had higher market prices than other treatments (Table 3).

Fruit yield accounted for >80% of the net value of provisioning services (3.1 × 10⁵–4.4 × 10⁵ CNY/ha), with forage sales contributing 5.4 × 10³–3.1 × 10⁴ CNY/ha and management costs totaling 5.0 × 10⁴–6.2 × 10⁴ CNY/ha (Figure 1). The net value of provisioning services for natural grass (N_G) showed a slight increasing trend with no significant annual differences. In contrast, artificial grass treatments (A_L, A_G, A_L+G, P_L, P_G, P_L+G) showed a rapid increase from 2021 to 2023, followed by a slight decrease in 2024, with perennial grass treatments (P_L, P_G, P_L+G) performing best and significantly exceeding natural grass (N_G) from 2022 onward.

3.2. Economic Values of Regulating Services

Regulating services included water conservation, carbon sequestration, and oxygen production. No significant differences in soil water content were observed among treatments (

Table 4. Soil nutrient content and water content.

Year	Index	NG	AL	AG	AL+G	PL	PG	PL+G
2021	Bulk density (g/cm3)	1.24 a	1.24 a	1.25 a	1.25 a	1.26 a	1.25 a	1.26 a
	AN (mg/kg)	56.88 a	53.10 a	50.33 ab	43.90 b	49.78 ab	47.75 ab	48.75 ab
	AP (mg/kg)	18.63 a	18.33 a	16.57 ab	16.39 ab	14.91 b	16.84 ab	19.60 a
	AK (mg/kg)	131.50 a	129.50 a	125.50 a	129.50 a	132.00 a	129.50 a	128.67 a
	TC (g/kg)	11.31 a	11.40 a	11.28 a	11.34 a	11.26 a	11.28 a	11.27 a
	Water content (%)	12.70 a	8.61 b	11.53 ab	11.98 a	14.70 a	9.41 ab	10.55 ab
2022	Bulk density (g/cm3)	1.24 b	1.26 ab	1.27 ab	1.28 a	1.28 a	1.28 a	1.27 a
	AN (mg/kg)	52.15 a	54.93 a	42.00 b	47.58 ab	49.32 ab	51.08 a	57.48 a
	AP (mg/kg)	21.07 a	16.32 ab	12.78 b	15.88 ab	12.99 b	12.73 b	13.26 b
	AK (mg/kg)	128.45 a	112.64 b	121.22 a	117.56 ab	122.87 a	111.97 b	112.17 b
	TC (g/kg)	11.29 ab	11.24 ab	11.19 b	11.17 b	11.37 a	11.34 a	11.48 a
	Water content (%)	19.57 a	19.29 a	21.14 a	20.08 a	21.33 a	16.81 a	12.92 b
2023	Bulk density (g/cm3)	1.25 b	1.26 b	1.26 b	1.27 ab	1.30 a	1.31 a	1.31 a
	AN (mg/kg)	51.23 ab	53.55 a	47.55 b	54.33 a	56.75 a	44.60 b	52.38 ab
	AP (mg/kg)	23.20 a	21.22 ab	19.96 b	21.36 b	22.42 a	22.07 ab	21.21 b
	AK (mg/kg)	122.51 a	100.13 ab	104.13 ab	99.28 b	92.13 b	94.89 b	93.07 b
	TC (g/kg)	11.35 b	11.35 b	11.41 ab	11.64 a	11.44 a	11.48 a	11.52 a
	Water content (%)	21.66 a	20.73 a	19.07 a	24.23 a	21.08 a	20.17 a	20.03 a
2024	Bulk density (g/cm3)	1.26 b	1.25 b	1.27 ab	1.32 a	1.32 a	1.33 a	1.33 a
	AN (mg/kg)	48.18 a	51.53 a	48.17 a	47.19 a	46.38 a	41.90 b	45.68 ab
	AP (mg/kg)	24.25 a	19.82 a	19.02 ab	17.62 b	17.15 b	16.74 b	17.35 b
	AK (mg/kg)	115.42 a	92.93 b	95.69 b	94.15 b	93.09 b	87.30 b	86.88 b
	TC (g/kg)	11.51 b	11.78 ab	11.73 ab	11.81 a	11.81 a	12.07 a	11.96 a
	Water content (%)	20.58 a	20.09 a	18.13 a	22.54 a	20.58 a	18.62 a	19.60 a

Note: Different lowercase letters in the same line indicate significantly different (p < 0.05).

). Soil total carbon increased in all treatments from 2021 (11.26–11.40 g/kg) to 2024 (11.51–12.07 g/kg), with a more pronounced trend in perennial grass treatments (P_L, P_G, P_L+G). Perennial grass treatments (P_L, P_G, P_L+G) had significantly higher soil TC content than natural grass (N_G) from 2023 onward. Forage water content and total carbon content showed no significant differences among treatments (Table 3). Consequently, the amount of carbon and water fixed in grass tissues, as well as the oxygen produced by grasses, exhibited similar dynamics to the variation in grass biomass.

The net value of regulating services ranged from 3.6 × 10³ to 2.1 × 10⁴ CNY/ha, with carbon sequestration and oxygen production accounting for >90% (Figure 1). Natural grass (N_G) and annual grass treatments (A_L, A_G, A_L+G) showed slight increases in regulating service values over time. Perennial grass treatments (P_L, P_G, P_L+G) exhibited a 3-fold increase from 2021 to 2023, followed by a slight decrease in 2024, and were significantly higher than natural grass (N_G) from 2022 onward.

3.3. Economic Values of Supporting Services

Supporting services encompassed soil nutrient maintenance and biodiversity. Soil available phosphorus (AP) and available potassium (AK) decreased over time across all treatments (Table 4). Legume grass treatments (A_L, P_L) resulted in higher soil available nitrogen (AN) content (Table 4). The economic value associated with soil nutrient content decreased by 23.06–27.68% in perennial grass treatments (P_L, P_G, P_L+G), 11.83% in natural grass (N_G), relative to the experiment’s start (Figure 1). Perennial grass treatments (P_L, P_G, P_L+G) had the lowest indices on grass species diversity (Shannon–Wiener index), with P_G and P_L+G decreasing to 0.18 and 0.19 in 2024 (Table 3). Annual grass treatments (A_L, A_G, A_L+G) and natural grass (N_G) exhibited fluctuating indices (0.82–1.73) over time (Table 3).

The net value of supporting services ranged from −177 to 321 CNY/ha (Figure 1). Natural grass (N_G) had the highest value, followed by annual grass treatments (A_L, A_G, A_L+G), while perennial grass treatments (P_L, P_G, P_L+G) were significantly lower than N_G throughout the experiment.

3.4. Total Economic Values of Ecosystem Services

The net value of ecosystem services ranged from 2.7 × 10⁵ to 4.4 × 10⁵ CNY/ha (Figure 2), with provisioning services accounting for >95% and supporting services contributing the least. Artificial grass treatments (A_L, A_G, A_L+G, P_L, P_G, P_L+G) showed rapid increases in net value from the experiment’s start, with a slight decline in 2024, while natural grass (N_G) showed a slight, non-significant increase. Artificial grass treatments (A_L, A_G, A_L+G, P_L, P_G, P_L+G) had higher net values than natural grass (N_G), with P_G and P_L+G achieving the highest values during 2022–2024. Perennial grass treatments (P_L, P_G, P_L+G) had the highest provisioning and regulating services values but the lowest supporting services values, whereas N_G showed the opposite pattern.

4. Discussion

4.1. Perennial Grass Treatments Sharply Increased the Value of Provisioning Services

Ecosystem provisioning services refer to the tangible goods from ecosystems. Fruit and forage sales were the main sources of economic revenue, partially offset by daily orchard management costs. Overall, the economic benefit of fruit yield was the primary factor influencing the ecosystem provisioning value. Generally, artificial grass planting led to a higher apple yield than natural grass. This is likely because artificial grass (e.g., A. sativa, V. sativa, M. sativa, and D. glomerata) is selected for its ability to outcompete aggressive weeds while minimizing competition with fruit trees [27]. In contrast, natural grass communities may include species that compete strongly for water and nutrients, particularly in young orchards. Results of the present study further indicate that perennial grass performed better than annual grass, which may be because perennial grasses develop deeper and more extensive root systems than annual grass over time. These characteristics improve soil structure, promote organic matter accumulation, and create a persistent effect on soil temperature moderation and evaporation reduction. A meta-analysis of 62 studies in Chinese orchards found that annual grasses increased soil organic matter by 19.3% and reduced soil temperature by 4.5%, while perennial grasses led to a 20.7% increase in soil organic matter and an 11.7% reduction in soil temperature [28].

Previous studies have demonstrated that orchard grass application could significantly increase fruit hardness, total soluble sugar and vitamin C content [13]. However, for most of the time, there were no significant differences in the determined fruit quality indicators among different treatments. This phenomenon may be attributed to Linzhi’s relatively low annual temperature, whereas studies have shown that orchard grass significantly improves fruit taste and nutritional quality with rising temperatures by increasing soil moisture, enhancing soil buffering against temperature fluctuations, and enabling orchard trees to better withstand heat stress while maintaining stable high yields [13]. Nevertheless, the slightly improved fruit quality in artificial grass treatments led to a higher market price compared with N_G.

The optimal growth temperature for A. sativa, V. sativa, M. sativa, and D. glomerata ranges from 15 to 25 °C, making them widely planted in orchards in China’s lower-altitude regions with higher annual temperature (~20 °C) [8,9,10,11,12,29,30]. The present study showed that these species can also be used as orchard grasses in Linzhi (annual temperature 8.7 °C), although their fresh biomass is lower compared with previous studies in lower-altitude regions. In the present study, the forage yield from natural grasses remained the lowest value among all treatments, annual grasses stabilized throughout the experimental period, while perennial grasses increased by approximately 3-fold. Variations in forage yield may be attributed to the different survival strategies of grass species [31]. Natural grass communities typically comprise unmanaged, diverse grass species that have evolved to survive under low-resource conditions. These species often exhibit r-selected traits (e.g., rapid germination, short lifespan), prioritizing seed production over vegetative growth, which directly results in low aboveground biomass. In contrast, perennial grasses generally display K-selected traits, including deep root systems and persistent growth. This study found that perennial grasses allocated biomass preferentially to root development during the establishment phase (1–2 years), a growth strategy enabling efficient resource capture from both topsoil and subsoil layers. Consequently, from 2023 onward, these grasses achieved a significant increase in forage yield.

In the present study, legume grasses showed the highest crude protein content, followed by legume-gramineous mixed grasses, then gramineous grasses, and finally natural grasses. This difference stems primarily from the ability of legumes to form symbiotic relationships with nitrogen-fixing bacteria in their root nodules, which convert atmospheric nitrogen into organic nitrogen compounds [32]. This biological nitrogen fixation enhances their ability to synthesize proteins, resulting in a higher crude protein level than gramineous grasses. Consequently, legume grasses commanded higher market prices due to their superior crude protein content.

All orchard management practices across treatments were standardized and followed regional traditional methods, including identical fertilization, irrigation, and pruning regimes. Consequently, annual cost differences between treatments primarily stemmed from two factors: grass seed procurement and sowing labor. Annual grass species imposed the highest costs, as they required annual seed purchases and hired labor for sowing. In sharp contrast, natural grass communities entailed the lowest costs, as they eliminated both seed expenses and additional labor investments for establishment or maintenance.

4.2. Perennial Grass Treatments Sharply Increased the Value of Regulating Services

Ecosystem regulating services refer to the benefits derived from ecosystem processes that maintain environmental balance and stability through regulatory processes and functions, including water conservation, carbon sequestration, and oxygen production. Previous studies indicated that grass coverage can reduce soil evaporation by altering net radiation, wind speed, and surface temperatures, thereby benefiting soil water storage [33,34]. However, there were no significant differences in soil water content among treatments in the present study. Instead, P_G and P_L+G, which should have a stronger effect on altering net radiation and wind speed due to their high aboveground biomass, showed relatively lower soil water content throughout the experiment. This finding somewhat contradicts previous studies from orchards in southern China [35] to some extent. A potential explanation is that, although grass cover can reduce soil evaporation, the increased grass transpiration may partially offset this reduction [36]. Given the relatively low rainfall in Linzhi (annual precipitation: 650 mm), grass cultivation likely increased overall soil water consumption. In contrast, soil water consumption by grasses is less critical for orchards in southern China, where annual precipitation is higher (800–2000 mm).

Soil total carbon content increased in all treatments, with a more pronounced trend in perennial grass treatments. Zhuo et al. [37] indicated that grass cultivation measures promote soil carbon accumulation through two mechanisms: (1) increasing organic matter input and (2) reducing soil organic carbon loss. In the present study, fresh grass biomass was harvested for forage, so carbon input to the soil via grass litter decomposition was limited. Instead, the higher fresh biomass of perennial grass treatments may enhance microbial activity, accelerate the excretion of cementing substances, and promote the formation and stability of soil aggregates, thereby increasing the physical protection of organic carbon [11].

In addition to altering soil properties, grasses directly contribute to carbon sequestration and water retention through photosynthesis, transferring atmospheric carbon and water into plant biomass. The carbon and water fixed in grass tissues were calculated separately, as this portion of carbon and water is removed from the orchard system when forage grasses are harvested. In addition, grasses convert CO₂ into glucose and release oxygen (O₂) using light energy in chloroplasts, which is the primary source of atmospheric oxygen [38]. There were no significant differences in forage water content or total carbon content among treatments. Thus, the amount of carbon and water fixed in grass tissues, as well as the oxygen produced by grasses, exhibited similar dynamics to the variation in grass biomass.

4.3. Artificial Grass Treatments Lower the Value of Supporting Services

Ecosystem supporting services are the fundamental processes that maintain ecosystems’ functional and sustainable. Soil nutrient maintenance was included in the present study as it directly supports plant growth by providing essential nutrients (e.g., nitrogen, phosphorus, potassium). Nutrient uptake for fruit and forage production resulted in nutrient outputs, so fertilization is necessary for agriculture to maintain soil nutrient balance. Achieving the correct balance of nutrient application is particularly important because, unless this is achieved, food production will be constrained or nutrient losses will occur [39]. In the present study, fertilization followed local practices used for 15 years prior to the experiment. However, results showed that soil AP and AK contents decreased over time, indicating nutrient deficiency across all treatments. This contrasts with the findings of Wei, Xiang [40], who found that grass cultivation increased soil TN, available N, available P and total K content in the orchard soils by meta-analysis across China. This difference may be attributed to the fact that grass, when harvested as forage, removes the nutrients it has absorbed from the orchard soil. In contrast, in most previous studies [13], grass was left in the field, meaning the nutrients it took up would eventually return to the soil. The impact of nutrient depletion was more pronounced in perennial grass treatments, where the economic benefits associated with soil nutrient content decreased by 23.06–27.68% compared to the initial stage of the experiment. In comparison, natural grass showed an 11.83% reduction in the economic benefits of nutrient content relative to the experiment’s start. This discrepancy arises because the higher biomass of perennial grasses results in greater nutrient loss during forage harvesting. Compared with AP and AK, sowing legume grasses resulted in higher soil AN content, possibly due to their specific role in nitrogen fixation [41]. Leguminous green manure can assimilate atmospheric nitrogen at rates of 110–227 kg/ha [42]. Nevertheless, our findings indicate that implementing grass cultivation in Xizang orchards requires increased fertilizer inputs to compensate for nutrient removal.

Perennial grasses planting resulted in the lowest grass species Shannon–Wiener index among all treatments. In comparison, the Shannon–Wiener indices of annual grasses planting and natural grass planting fluctuated over time. This is because established perennial grass communities achieve a high stability structure. Their long life cycles allow them to dominate ecological niches, impeding the settlement and germination of seeds from external grass species and making it difficult for new species to integrate into the community [43]. Short-lived grasses, however, fail to establish long-term stable community structures. After each growing season, their vegetation withers and dies, creating open niches for other species to survive and reproduce. Moreover, annual grasses typically have less competitive morphological traits, such as a shallow root system, resulting in less competition for water and nutrients compared to perennial grasses, which reduces the resource competition pressure faced by other species. Grass diversification is key to ecosystem supporting services through the regulation of nutrient cycling, pest populations, etc., as different grass species are expected to provide diverse habitats and food resources for birds, insects, and microorganisms [44,45]. The present study indicates that perennial grasses reduced the economic benefit of biodiversity compared with the cultivation of natural grass and annual grasses.

4.4. Artificial Grass Treatments Result in a Higher Net Value of Ecosystem Services

Appraising the economic costs and benefits of ecosystem services related to orchard grass provides a comprehensive consideration of impacts on the environment, income benefits, and production sustainability. Provisioning services are the main component (accounting for more than 95%), while supporting services contribute the least to the net value of ecosystem services. The net value of ecosystem services increased rapidly with the planting of artificial grasses from the beginning of the experiment, then began to decline slightly in 2024. In contrast, the net value of ecosystem services under natural grass showed a slight, non-significant increasing trend over the years. These results support Hypothesis 1, which states that artificial grasses planting in orchards result in a higher net value of ecosystem services than natural grass, even though artificial grass treatments require additional economic input for seed purchase and labor costs. Regarding individual ecosystem services, perennial grasses planting showed the highest provisioning and regulating services values, and the lowest supporting services values. Natural grass planting, however, showed the highest supporting services value and the lowest provisioning and regulating services values. These findings support Hypothesis 2, indicating that different grass species plantings elicit varied responses from individual ecosystem services.

Overall, P_G and P_L+G resulted in the highest total economic value of ecosystem services, attributed to increased fruit and forage yields, soil carbon sequestration, and oxygen production. This result indicates that D. glomerata monoculture and M. sativa + D. glomerata mixed culture are the optimal orchard grass patterns in Linzhi. However, it should be noted that there was a slight decrease in the total economic value of ecosystem services in 2024, which may be attributed to nutrient depletion, indicating that increasing fertilizer input is necessary to ensure the sustainable production of orchards. To address this challenge, future research should focus on developing site-specific fertilization strategies. For example, soil nutrient monitoring could be conducted quarterly to determine the optimal NPK ratios for different growth stages of orchard trees and grasses. Additionally, exploring organic fertilization alternatives, such as compost or manure application, could mitigate the negative environmental impacts associated with synthetic fertilizers while maintaining soil fertility.

This research endeavors to comprehensively assess the value of ecosystem services from multiple dimensions. However, it is important to note the considerable uncertainties inherent in such assessments. Some ecosystem services (such as genetic resources and cultural services) and ecosystem disservices (such as wildlife habitat loss and waterway sedimentation) were not assessed due to the lack of effective evaluation methods. Furthermore, the economic evaluation method for each specific component of ecosystem services may also introduce some uncertainties. For example, the average market price was used to evaluate the economic value of fruit production. However, market prices fluctuate and are influenced by various factors, which may lead to under- or overestimation of provisioning services. In addition, the slight decrease in net ecosystem service value in artificial grass treatments indicates that a longer-term investigation is needed to evaluate the sustainability of orchard grass planting on the Xizang Plateau.

5. Conclusions

Orchard grass has become a mainstream model for ecological orchard construction and organic fruit production in China’s lower-altitude regions. However, its sustainability on the Xizang Plateau remains unclear. Our results demonstrate that, compared to natural grass, perennial grasses lead to a rapid increase in net ecosystem service value, primarily driven by substantially higher fruit and forage yields. In addition, perennial grass cultivation enhances soil carbon sequestration and oxygen production, thereby increasing the value of ecosystem regulating services. However, it is crucial to recognize that cultivating perennial grass can exacerbate soil nutrient depletion, potentially reducing the value of ecosystem supporting services and posing long-term risks to orchard sustainability. Overall, D. glomerata monoculture and M. sativa + D. glomerata mixed culture are identified as the optimal orchard grass patterns in Linzhi. To ensure sustainable orchard productivity, enhanced fertilizer input is essential. Concurrent research on the economic and environmental implications of increased fertilization is also warranted.