Introduction

In the global ecosystem, the research on the response of aboveground net primary productivity (ANPP) of grassland ecosystems to climate change and human activities is of great significance. Many studies have widely and deeply explored the relationships between ANPP and climate change and human activities in different ecosystems and have achieved fruitful results [1,2]. These studies provide a crucial basis for understanding the functions and dynamic changes of ecosystems and promote the development of the field of ecology. Three key scientific questions remain insufficiently addressed. First, while the dynamic responses of ANPP are a core topic in global change studies, existing studies have mostly focused on the responses of the dry ANPP to climate change and human activities, with insufficient attention paid to fresh ANPP [3,4]. As a key indicator directly reflecting plant water status, immediate metabolic activity, and forage availability, the responses of fresh ANPP to climate change and human activities have not yet formed a systematic understanding. For example, the spatiotemporal dynamic patterns of fresh ANPP still lack clear answers. Second, the impact of climate change and human activities on the ANPP of plant community has been mostly quantified [5,6]. However, on a large spatial scale, the research on the impact of ANPP of different plant functional groups (e.g., sedges and graminoids) is lacking [4,7]. Third, the main focus has been on the direction of ANPP change (increase, decrease or no change). However, relatively little attention has been paid to whether this increasing or decreasing trend is gradually intensifying or weakening. In general, although there are many research results related to ANPP of global grassland ecosystems, the above-mentioned problems still need to be further explored in order to accurately understand the functions and dynamic change mechanisms of grassland ecosystems.

The Qinghai-Xizang Plateau’s alpine grasslands are globally unique and highly sensitive to climate change, distinct from other grassland systems. As the “Roof of the World” and “Third Pole,” its average elevation exceeds 4,000 m, creating extreme topography, shaping a climate with low temperatures, strong radiation, and large diurnal variations. It acts as a critical “climate regulator,” influencing East Asian monsoons and global climate, making its grasslands key in biosphere-atmosphere interactions. Ecologically fragile yet biodiverse, these grasslands support endemic species adapted to harsh conditions, serve as major carbon sinks, and conserve water for rivers sustaining billions. However, low productivity and slow nutrient cycling make them vulnerable to disturbances [8]. Notably, the Qinghai-Xizang Plateau warms twice as fast as the global average, with altered precipitation and radiation [9]. These changes disproportionately affect its grasslands, causing amplified, nonlinear responses. Studying their fresh ANPP dynamics is vital for understanding regional resilience and high-altitude ecosystem responses to global change [10]. In the alpine grasslands of the Qinghai-Xizang Plateau, the ecological significance of the fresh ANPP is particularly prominent. First, fresh ANPP directly determines the palatability and nutritional value of forage, which is crucial for the sustainable development of grazing animal husbandry. Its dynamic fluctuations directly affect the feeding efficiency and health status of livestock. Second, in alpine regions with drastic day-night temperature differences and frequent precipitation, the response sensitivity of plant fresh ANPP to short-term changes in water and heat is much higher than that of dry ANPP, and it can more quickly and accurately reflect the immediate functional state of the ecosystem. Therefore, focusing on the response patterns of fresh ANPP is a key entry point for analyzing the correlation between the ecological functions of alpine grasslands and the sustainability of animal husbandry.

The unique sensitivity of the Qinghai-Xizang Plateau’s alpine grasslands makes understanding their ANPP responses critical. While existing studies using in-situ experiments and models have advanced knowledge of dry ANPP dynamics in alpine grasslands [3,11], which is beneficial for understanding the evolution of the alpine ecosystems. Two further uncertainties remain. First, the dominant functional group (sedges, graminoids, or forbs) driving plant community ANPP changes are still unresolved [4,11]. Second, the influence of changes in total shortwave radiation (a crucial component of surface energy balance affecting evapotranspiration and photosynthetic potential) on ANPP has received comparatively less attention than warming or precipitation changes. Addressing these uncertainties is essential for a comprehensive understanding of alpine grassland ecosystem functioning under global change. Therefore, further exploration is needed to comprehensively and accurately understand the operation mechanism and development trend of the alpine grassland ecosystems in this area.

In view of this background, this study was based on the modelled data of the fresh ANPP of plant community and different functional groups (sedges, graminoids and forbs), soil carbon, nitrogen and phosphorus of alpine grasslands on the Qinghai-Xizang Plateau from 2000 to 2022. It statistically analyzed the change characteristics of the data under the combined and individual impacts of climate change and human activities, used structural equation models to quantify the impact of various factors on the change of fresh ANPP, and deeply explored the dominant factors and their geographical variability of the changes of fresh ANPP of different plant functional groups and plant community. The main aim was to comprehensively reveal the change and driving mechanisms of fresh ANPP of this alpine grassland ecosystems, provide theoretical support for formulating ecological protection and management strategies, respond to the challenges of climate change and human activities, and maintain its sustainable development.

Materials and methods

Statistical analyses

Referring to previous studies [22,23], we calculated the change rate of AT (ΔAT), AP (ΔAP), ARad (ΔARad) and NDVI_max (ΔNDVI_max) for all pixels over the alpine grasslands on the Qinghai-Xizang Plateau from 2000 to 2022. Similarly, we also calculated the change rate of the fresh ANPP of plants and soil variables for all pixels over same region and time period. Following previous studies [22,23], this study calculated the raster dataset of the relative change of fresh ANPP of plant community, sedges, graminoids and forbs, and soil data, over the period 2000–2022, respectively. For example, first, we obtained the change rate of the fresh ANPP of plant community under the combined influence of climate change and human activities (ΔANPP_{C_C+H}) for the period 2000–2022 using the sens.slope function of ‘trend’ package. Then, we multiplied ΔANPP_{C_C+H} by 22 and divided the result by the fresh ANPP of plant community in 2000 to get the relative change of ANPP_{C_C+H} (RC_ANPP_{C_C+H}). The abbreviations of all other relative changes were similar to RC_ANPP_{C_C+H}. In addition to quantifying the overall change rate of the fresh ANPP from 2000 to 2022, this study also quantified the change rate of the change rate (i.e., to reflect whether the change rate of a variable was increasing or decreasing). The relevant steps were detailed below taking the fresh ANPP of plant community under the combined effects of climate change and human activities (ANPP_{C_C+H}) as an example. First, the change rates of the fresh ANPP of plant community from 2000 to 2022 were calculated using a sliding window (window size = 3). Thus, 19 change rates of this variable were obtained. Then, based on the above 19 change rates, their change rate, that is, the change rate of the change rate (ΔΔANPP_{C_C+H}), was calculated. If the overall change rate of this variable from 2000 to 2022 was > 0, and the change rate of the change rate was also > 0, it indicated that the positive response of this variable was strengthening year by year (Table 1). If the overall change rate of this variable from 2000 to 2022 was < 0, but the change rate of the change rate was > 0, it indicated that the negative response of this variable was weakened year by year (Table 1). This study quantified the dominance of the fresh ANPP of sedges, graminoids and forbs over that of plant community (Table 2). Referring to previous studies [19,22], we quantified the relative impacts of climate change and human activities to the fresh ANPP of plant community, sedges, graminoids, and forbs. Finally, the structural equation models were utilized to quantify the influence of independent variables on the changes in the fresh ANPP of plants. All statistical analyses were completed using R software (version 4.2.2).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Classification of the change patterns of fresh aboveground net primary productivity of the plant community under combined climate and human influences (ANPP_{C_C+H}).

https://doi.org/10.1371/journal.pone.0349705.t001

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Relative contributions of three plant functional groups (sedges, graminoids, and forbs) to the change in fresh aboveground net primary productivity of the plant community under combined climate and human influences.

https://doi.org/10.1371/journal.pone.0349705.t002

Results

The spatial averages RC_ANPP_{C_C+H}, RC_ANPP_{S_C+H}, RC_ANPP_{G_C+H} and RC_ANPP_{F_C+H} were –1.15%, 5.67%, –1.55%, and –2.26%, respectively. The spatially averaged negative change of ANPP_{C_C+H} (–0.13 per decade) and ANPP_{F_C+H} (–0.55 per decade) was increasing year by year, while the spatially averaged positive change of ANPP_{S_C+H} (0.19 per decade) was also increasing. By contrast, the spatially averaged negative change of ANPP_{G_C+H} (0.01 per decade) was decreasing year by year. The relative changes of ANPP were spatially heterogeneous (Fig 1 and S1 Fig). For example, the positive responses of the ANPP_{C_C+H} in 13.25%, 7.42% and 2.56%, the ANPP_{S_C+H} in 23.65%, 9.55% and 3.33%, the ANPP_{G_C+H} in 10.37%, 8.79% and 0.98%, and the ANPP_{F_C+H} in 9.29%, 9.25% and 3.32% of the grassland areas were strengthening, weakening and remaining unchanged year by year, respectively (Fig 1 and S1 Fig). In contrast, the negative responses of the ANPP_{C_C+H} in 9.16%, 15.03% and 2.28%, the ANPP_{S_C+H} in 7.10%, 7.35% and 0.42%, the ANPP_{G_C+H} in 10.49%, 15.48% and 3.69%, and the ANPP_{F_C+H} in 7.50%, 19.82% and 1.05% of the grassland areas were weakening, strengthening and remaining unchanged year by year, respectively (Fig 1 and S1 Fig). In addition, the ANPP_{C_C+H}, ANPP_{S_C+H}, ANPP_{G_C+H}, and ANPP_{F_C+H} did not change in 50.30%, 48.61%, 50.20% and 49.77% of the grassland areas, respectively (Fig 1 and S1 Fig).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Spatial distribution for the changes in fresh aboveground net primary production (ANPP) of plant community (ANPP_C), sedges (ANPP_S), graminoids (ANPP_G) and forbs (ANPP_F) in 2000–2022.

Note: The subscripts _C + H, _C, and _H denote the combined effects of climate change and human activities, the single effect of climate change, and the single effect of human activities, respectively.

https://doi.org/10.1371/journal.pone.0349705.g001

The dominant influences of the changes in fresh ANPP of sedges, graminoids and forbs on the changes in fresh ANPP of plant community were spatially heterogeneous (Fig 2). For example, sedge fresh ANPP changes dominated the ANPP_{C_C+H}, ANPP_{C_C} and ANPP_{C_H} changes by 18.27%, 8.28% and 12.47%, respectively; graminoid ANPP changes by 7.18%, 13.66% and 9.55%, respectively; forb ANPP changes by 35.43%, 54.40% and 61.85%, respectively; and the combined ANPP changes of all three by 39.12%, 23.65% and 16.13%, respectively (Fig 2). In addition, the increases in fresh ANPP_{C_C+H} of 9.14%, 1.40% and 12.40% of the grasslands were attributed to the increases in fresh ANPP_{S_C+H}, ANPP_{G_C+H}, and ANPP_{F_C+H} (Fig 2). Conversely, the decreases in fresh ANPP_{C_C+H} of 5.02%, 2.66% and 18.42% of the grasslands were attributed to the decreases in fresh ANPP_{S_C+H}, ANPP_{G_C+H}, and ANPP_{F_C+H}, respectively (Fig 2).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Spatial distribution for the relative contributions of changes in fresh aboveground net primary production (ANPP) of sedges, graminoids, and forbs to the change in ANPP of the plant community under different scenarios.

Note: The subscripts _C + H, _C, and _H denote the combined effects of climate change and human activities, the single effect of climate change, and the single effect of human activities, respectively. ΔANPP_C represents the change rate of plant community ANPP.

https://doi.org/10.1371/journal.pone.0349705.g002

The areas of grasslands where the fresh ANPP of plants increased, decreased and remained unchanged due to human activities were each larger than those caused by climate change (Fig 3). The proportions of grassland areas where the ΔANPP_{C_C+H}, ΔANPP_{S_C+H}, ΔANPP_{G_C+H} and ΔANPP_{F_C+H} were attributed to climate change were 20.79%, 17.02%, 24.08% and 18.01%, respectively, while those attributed to human activities were 54.26%, 60.73%, 52.36% and 56.87%, respectively (Fig 3). The remaining areas were jointly dominated by climate change and human activities (Fig 3).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Spatial distribution of the relative contribution of climate change and human activities to the change in fresh aboveground net primary production (ANPP) under different scenarios for (a) plant community, (b) sedges, (c) graminoids, and (d) forbs.

Note: The subscripts _C + H denote the combined effects of climate change and human activities. ICC: increases due to climate change; DCC: decreases due to climate change; NCC: no changes due to climate change; IHA: increases due to human activities; DHA: decreases due to human activities; NHA: no changes due to human activities; ICCHA: increases due to the combined effects of climate change and human activities; DCCHA: decreases due to the combined effects of climate change and human activities; NCCHA: no changes due to the combined effects of climate change and human activities.

https://doi.org/10.1371/journal.pone.0349705.g003

Geographical location, climate change, changes in NDVI_max, and changes in soil variables directly or indirectly affected the changes in fresh ANPP of plant community (Fig 4 and S2 Fig). The direct impacts of climate change on the RC_ANPP_{S_C+H} and RC_ANPP_{G_C+H} were negative, whereas the direct impact of climate change on the RC_ANPP_{F_C+H} was positive (Fig 4). The ΔNDVI_max had a positive effect on the RC_ANPP_{S_C+H}, but negative effects on the RC_ANPP_{G_C+H} and the RC_ANPP_{F_C+H} (Fig 4). Among these, the impact magnitude of ΔNDVI_max on the RC_ANPP_{F_C+H} was the greatest (Fig 4). The RC_SAP_C+H affected RC_ANPP_{C_C+H} more strongly than RC_NH₄⁺-N_C+H and RC_NO₃^--N_C+H (Fig 4). The impact direction of RC_NH₄⁺-N_C+H on the RC_ANPP_{C_C+H} was completely opposite to that of RC_NO₃^--N_C+H on the RC_ANPP_{C_C+H} (Fig 4).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Structural equation model of fresh aboveground net primary production (ANPP) of plant community under the combined effects of climate change and human activities (ANPP_{C_C + H}).

Note: ΔAT, ΔAP, ΔARad and ΔNDVI_max indicate the change rate of annual temperature, annual precipitation, annual radiation and maximum normalized difference vegetation index, respectively. RC_SOC_C+H, RC_TN_C+H, RC_TP_C+H, RC_C:N_C+H, RC_C:P_C+H, RC_N:P_C+H, RC_NH₄⁺-N_C+H, RC_NO₃^--N_C+H, RC_SAP_C+H, RC_pH_C+H, RC_ANPP_S__C+H, RC_ANPP_G__C+H, RC_ANPP_F__C+H and RC_ANPP_C__C+H indicated the relative change of soil organic carbon, total nitrogen, total phosphorus, ratio of SOC to TN, ratio of SOC to TP, ratio of TN to TP, ammonium nitrogen, nitrate nitrogen, soil available phosphorus, pH, ANPP of sedges, ANPP of graminoids, ANPP of forbs, and ANPP of plant community under the combined effects of climate change and human activities, respectively.

https://doi.org/10.1371/journal.pone.0349705.g004

Discussion

The RC_NH₄⁺-N_C+H and RC_NO₃^--N_C+H were negatively and positively correlated with RC_ANPP_{C_C+H}, respectively (Fig 4), reflecting the multiple action mechanisms of NH₄⁺-N and NO₃^--N on plants. NO₃ ⁻ -N is negatively charged, resulting in weak adsorption by soil and strong mobility [24]. It is easily absorbed efficiently by plant roots along with water movement, promoting the accumulation of aboveground biomass. In contrast, NH₄ ⁺ -N is positively charged and tends to be adsorbed by soil colloids, leading to uneven spatial distribution and low absorption efficiency [25]. In acidic soils, it combines with H⁺ to exacerbate acidification [26], further inhibiting plant growth. On the other hand, after being absorbed by the roots, NO₃ ⁻ -N is transported to the aboveground parts through the xylem. In mesophyll cells, it is catalytically converted to NH₄⁺ by nitrate reductase (NR) and nitrite reductase (NiR), and finally synthesized into amino acids [27]. This process works in synergy with photosynthesis, enhancing carbon assimilation efficiency by promoting chlorophyll synthesis and enzyme activity [28]. In contrast, the assimilation of NH₄ ⁺ -N relies on the GS-GOGAT cycle, which consumes large amounts of carbon skeletons and ATP [29]. Excessive accumulation triggers ammonium toxicity, inhibits photosynthetic assimilation capacity, and depletes energy reserves [30].

The area of grasslands where the change in fresh ANPP dominated by human activities was larger than that where it was dominated by climate change (Fig 3). Moreover, the proportion of the area where the change in grassland productivity was dominated by human activities was also greater than that reported in previous studies [19]. This finding may be related to the following mechanisms. Human activities directly alter surface cover through fence installation, ecological migration, and transportation construction [31,32]. Although such disturbances damage soil structure and organic matter in the short term (Fig 4, S2 Fig), they promote vegetation restoration through enclosure management in the long run [33]. In contrast, the indirect impacts of climate change on land use (such as climate warming expanding the upper limit of grazing elevation) are characterized by lag and uncertainty [34–38]. Livestock foraging exerts direct and high-intensity pressure on plant communities, and their selective grazing (preferring high protein and low fiber content) rapidly alters the composition of functional groups [39]. In contrast, the impacts of climate change are gradual, and plants can buffer short-term fluctuations through phenological adjustment [3]. Current intensive animal husbandry (such as breed improvement and high-density breeding) has expanded the depth of impact of human activities, resulting in its regulatory range on ANPP exceeding historical levels [19,40,41].

This functional group-specific response (Fig 3b-d and S1d-l Fig), as well as the geographical variability in their contributions to community productivity (Fig 2), is similar to some previous studies [4,11,42]. These phenomena are mainly driven by the following mechanisms. Similar to some previous studies [3,43], the degrees of closeness between the changes in the fresh ANPP of sedges, graminoids and forbs and the changes in climate, soil nutrients and soil pH differed, and even the directions also differed (Fig 4). Seeds of sedges and graminoids are primarily wind- or animal-dispersed, whereas forbs often rely on specific insects or small mammals for dispersal [44,45]. Global warming indirectly affects forb reproduction by altering pollinator insect distributions [46,47]. Overgrazing damages soil seed banks, inhibiting regeneration across functional groups [48]. Sedges and graminoids exhibit compensatory growth (increased tillering, enhanced photosynthetic capacity) under low-to-moderate grazing, whereas forbs are more vulnerable to soil degradation from intensive grazing and trampling [49,50]. Climate and anthropogenic activities indirectly regulate plant growth by altering soil/phyllosphere microbial community structures [51, 52]. Significant differences in symbiotic relationships between functional groups and microorganisms [50,53], drive divergent fresh ANPP responses [54–56]. Forbs expand rapidly in resource-rich areas, while in resource-limited regions, graminoids dominate fresh ANPP changes through efficient resource utilization and deep-rooted competition. Sedges maintain fresh ANPP dominance in low-temperature areas (such as alpine meadows), and warming promotes the growth of graminoids and forbs. Forbs have strong water-capturing capabilities in humid areas, while graminoids in arid regions enhance their ANPP contributions through deep root systems. Sedges prefer moist soils with high organic matter content, graminoids are adapted to loose and well-aerated substrates, and forbs dominate under specific nutrient conditions (e.g., high phosphorus levels). The foraging preferences of different herbivores for functional groups (e.g., livestock prefer graminoids with low fiber content) and their spatiotemporal activity patterns further amplify the spatial differentiation in ANPP contributions.

The acceleration of changes in fresh ANPP across different regions (i.e., the rate of change in ANPP itself) exhibits diversity (Fig 1). This diversity essentially arises from the spatial heterogeneity of five key dimensions: the rate of climate-driven changes, variations in the intensity of human activities, resource limitation thresholds, the direction of biological feedbacks, and the intensity of multi-factor interactions. Climate serves as the fundamental driver of fresh ANPP (Fig 4), but the intensity and trends of changes in climatic factors vary significantly across regions [9], directly leading to differences in the acceleration of fresh ANPP changes. Changes in the intensity of human activities (such as grazing), whether increasing, decreasing, or shifting direction, directly alter the rate of fresh ANPP change (S2 Fig), and such “intensity variations” show substantial regional differences. Fresh ANPP growth depends on resources such as water and nutrients (nitrogen, phosphorus, etc.) (Fig 4). When the “matching degree” between resource supply and vegetation demand changes [57], it triggers a turning point in acceleration, with significant regional variations in the thresholds of this process. Changes in vegetation itself (e.g., coverage, species composition) form feedbacks through “resource capture and cycling”. The “positive/negative direction” and “intensity changes” of these feedbacks alter fresh ANPP acceleration, with marked regional differences.

While our analysis revealed spatially heterogeneous patterns in the intensification or weakening of fresh ANPP trends (Fig 1), we acknowledge inherent difficulty in precisely quantifying the rate of change of these trends (ΔΔANPP). The relatively short temporal span of our dataset (23 years) and the natural interannual variability in grassland productivity data introduce significant uncertainty into estimates of second-order derivatives. Short time series are particularly sensitive to the influence of extreme years or short-term fluctuations, which can amplify noise and potentially obscure underlying long-term accelerations or decelerations. Therefore, the ΔΔANPP values presented in Table 1 and Fig 1 should be interpreted primarily as qualitative indicators of directional changes in trend magnitude (i.e., whether a positive or negative trend was generally strengthening or weakening over the study period) rather than precise quantitative measures of acceleration. The spatial patterns indicating where intensification or weakening occurred (Fig 1) are likely more robust and ecologically informative than the exact numerical values of ΔΔANPP itself. These patterns, when viewed together with the dominant drivers identified (e.g., human activities dominating larger areas, Fig 3; the differential roles of functional groups, Fig 2; and the influence of soil N forms, Fig 4), provide valuable insights into the dynamics and potential vulnerability of the Tibetan grassland ecosystems under changing environmental pressures. Future research based on longer-term monitoring will be crucial to confirm the persistence of these observed trends dynamics and refine estimates of their rates of change.

To directly address the core issue of spatiotemporal decoupling, this study employed a key methodological strategy: model construction relied exclusively on annually varying dynamic drivers (temperature, precipitation, radiation, NDVI_max), completely excluding static covariates (e.g., soil properties or elevation) as inputs. This design forces the model to resolve spatiotemporal signals through the dynamic drivers themselves. Model validation was conducted through strict randomization partitioning of all site-year observations, ensuring simultaneous evaluation of predictive performance on unseen locations and unseen years. The dataset encompasses three main grassland types (reflecting spatial heterogeneity) with a temporal depth of ≥6 years, establishing a foundation for assessing the model’s generalization capability across novel spatiotemporal combinations. Although the aforementioned methods effectively address the core concern, the following residual uncertainties must be clarified. The model utilizes annually aggregated driver data, yet key ecological processes governing fresh ANPP (e.g., timing of snowmelt, onset of the growing season, occurrence of extreme events such as late frosts/short-term droughts, intra-seasonal grazing patterns) operate at finer temporal scales. Annual aggregation inevitably smooths critical intra-annual dynamics, potentially masking their nuanced effects on productivity, particularly for different plant functional groups with varying phenological sensitivities. Although excluding static variables forces the model to learn dynamics, the sensitivity of fresh ANPP to interannual fluctuations in climatic and human activity proxies still varies spatially. This stems from unmodeled factors (e.g., inherent soil fertility, microtopography, historical legacy effects). While model validation assessed overall spatiotemporal predictive performance, it cannot guarantee uniform accuracy in capturing the magnitude of temporal change across all heterogeneous grassland types. Compared to the model’s generalized relationships, sites with inherent low nutrient availability or poor water retention capacity may exhibit dampened or amplified responses to climatic drivers. The identified dominant role of forb ANPP changes (Fig 2) and the core function of nitrate/ammonium dynamics (Fig 4) emerge from landscape-scale patterns. While statistically robust at this scale, they may not be directly extrapolated to mechanistic understanding at the patch or individual plant level. Manipulative experiments targeting specific factors (e.g., N form addition, selective grazing exclusion) remain crucial for validating the physio-ecological mechanisms underpinning the observed large-scale correlations. Consequently, estimates of the absolute rates of change and the precise areal proportions dominated by climate vs. human activities (Fig 3) should be regarded as the best possible spatially explicit reconstruction given the model structure and input data, with their quantitative accuracy constrained by the aforementioned uncertainties. Nevertheless, the core contribution of this study lies in revealing the following key new insights. The spatial differentiation pattern of change trends (intensification/weakening) (Fig 1), the differential responses of plant functional groups, the pan-plateau dominance of human activities relative to climate change, the contrasting ecological roles of soil nitrogen forms (nitrate-N vs. ammonium-N). These findings provide important evidence for understanding the ecosystem functioning of Tibetan alpine grasslands under global change pressures. Future research needs to integrate data with higher temporal resolution, manipulative experiments across environmental gradients, and process-based models to refine the quantitative assessments presented herein and elucidate their underlying mechanisms.

Conclusions

This study aimed at the deficiencies in previous studies on the fresh aboveground net primary productivity (ANPP), the ANPP of functional groups, and the intensity of change trends. Taking the alpine grasslands on the Qinghai-Xizang Plateau from 2000 to 2022 as the research objects, it thoroughly analyzed the change patterns of the fresh ANPP of the plant community and three functional groups (i.e., sedges, graminoids, and forbs), as well as their correlations with independent variables such as soil carbon, nitrogen, and phosphorus, striving to reveal the change rules and internal mechanisms of fresh ANPP under the dual influences of climate change and human activities. The relevant research conclusions are of great significance both scientifically and practically. Scientifically, they provide detailed and reliable bases for further exploring the functions of alpine grassland ecosystems and their dynamic change mechanisms. Practically, they can build a solid foundation for formulating precise and effective ecological protection and management strategies.

Regarding the change trends of fresh ANPP, the plant community and each functional group showed different performances. Under the combined effects of climate change and human activities, the fresh ANPP of the community generally showed a negative growth, with the spatially averaged relative change rate being –1.15%. Among them, the fresh ANPP of sedges increased by 5.67%, while the fresh ANPP of graminoids and forbs decreased by 1.55% and 2.26%, respectively. This phenomenon reveals a broader ecological principle: alpine ecosystems are undergoing functional restructuring under global change, rather than simply experiencing an overall decline. Such restructuring triggers significant shifts in key ecological processes. The expansion of sedges, with their dense root systems, might enhance soil carbon sequestration capacity, while the reduction of forbs could disrupt trophic networks. In high-altitude regions sensitive to global change, the differential dynamics of functional groups serve as a core indicator for predicting ecosystem resilience. As an iconic site for studying climate–ecosystem interactions, this pattern on the Qinghai-Xizang Plateau not only demonstrates the mechanisms of species filtering and community restructuring under extreme environmental stress but also provides a typical example for understanding the non-linear responses of high-altitude biomes worldwide.

The positive responses in some regions showed increasing or decreasing trends over time, and the negative responses showed similar changes, while in other regions the changes were not significant. This scientific discovery warns that regional differences should be considered when formulating ecological management strategies. For regions with an increasing positive response, especially those where the fresh ANPP of sedges and graminoids increased simultaneously, measures can be taken to further promote this positive change. For regions with a decreasing positive response, it is necessary to analyze the reasons, which might be overgrazing or the emergence of new disturbing factors, so as to adjust management strategies in a timely manner (e.g., reducing grazing intensity and controlling alien species invasion). For regions with an increasing negative response, large-scale ecological restoration projects may be required, such as vegetation reconstruction and soil improvement. However, for regions with a decreasing negative response, some relatively mild measures (e.g., fencing for natural restoration) can be adopted to help the ecosystem recover.

From the perspective of the dominance of influencing factors, the area where human activities dominated the changes in grassland fresh ANPP was larger than that of climate change. In this regard, managers could formulate targeted grazing management measures to effectively reduce the negative impacts of human activities on grassland ecosystems and enhance the positive impacts, thereby helping grassland ecosystems achieve healthy and stable development.

The contributions of fresh ANPP of different functional groups to the changes in fresh ANPP of the plant community showed geographical differences. Under the combined effects of climate change and human activities, the area of grassland jointly regulated by the three functional groups accounted for the largest proportion, and the areas dominated by fresh ANPP of forbs, sedges, and graminoids decreased in that order. This difference is closely related to plant preferences for temperature and humidity, soil conditions, and the feeding preferences of herbivores. Based on a clear understanding of this situation, the management of grassland resources can be made more scientific and reasonable. For example, in specific regions, the utilization methods of grasslands could be carefully planned according to the characteristics of the dominant functional groups to achieve optimal resource allocation, thereby enhancing grassland productivity and fully exerting the service functions of grassland ecosystems.

In addition, the impacts of nitrate nitrogen and ammonium nitrogen in soils on the fresh ANPP of the community are also different. When carrying out ecological restoration of alpine grasslands, considering the differences between nitrate nitrogen and ammonium nitrogen is helpful to optimize restoration strategies. For example, in some grasslands degraded due to nitrogen loss, if the loss of nitrate nitrogen seriously affects plant growth, measures can be taken to reduce nitrate nitrogen leaching, such as improving soil structure or increasing soil organic matter to enhance the soil’s ability to retain nitrate nitrogen. Meanwhile, according to plant requirements for different forms of nitrogen, suitable plant species can be selected for reseeding to improve the success rate of ecological restoration.

Supporting information