Response of Aboveground Net Primary Production, Species and Phylogenetic Diversity to Warming and Increased Precipitation in an Alpine Meadow

Abstract

The uncertain responses of aboveground net primary productivity (ANPP) and plant diversity to climate warming and increased precipitation will limit our ability to predict changes in vegetation productivity and plant diversity under future climate change and further constrain our ability to protect biodiversity and ecosystems. A long-term experiment was conducted to explore the responses of ANPP, plant species, phylogenetic α–diversity, and community composition to warming and increased precipitation in an alpine meadow of the Northern Tibet from 2014 to 2019. Coverage, height, and species name were obtained by conventional community investigation methods, and ANPP was obtained using observed height and coverage. Open–top chambers with two different heights were used to simulate low- and high-level climate warming. The low- and high-level increased precipitation treatments were achieved by using two kinds of surface area funnel devices. The high-level warming reduced sedge ANPP (ANPP_sedge) by 62.81%, species richness (SR) by 21.05%, Shannon by 13.06%, and phylogenetic diversity (PD) by 14.48%, but increased forb ANPP (ANPP_forb) by 56.65% and mean nearest taxon distance (MNTD) by 33.88%. Species richness, Shannon, and PD of the high-level warming were 19.64%, 9.67%, and 14.66% lower than those of the low-level warming, respectively. The high-level warming-induced dissimilarity magnitudes of species and phylogenetic composition were greater than those caused by low-level warming. The low- rather than high-level increased precipitation altered species and phylogenetic composition. There were significant inter-annual variations of ANPP, plant species, phylogenetic α–diversity and community composition. Therefore, climate warming and increased precipitation had non-linear effects on ANPP and plant diversity, which were due to non-linear changes in temperature, water availability, and/or soil nutrition caused by warming and increased precipitation. The inter-annual variations of ANPP and plant diversity were stronger than the effects of warming and especially increased precipitation on ANPP and plant diversity. In terms of plant diversity conservation and related policy formulation, we should pay more attention to regions with greater warming, at least for the northern Tibet grasslands. Besides paying attention to the responses of ANPP and plant diversity to climate change, the large inter-annual changes of ANPP and plant diversity should be given great attention because the large inter-annual variation indicates the low temporal stability of ANPP and plant diversity and thus produces great uncertainty for the development of animal husbandry.

1. Introduction

Aboveground net primary productivity (ANPP) of plants is an important material basis for human survival, and biodiversity is an important guarantee to maintain a high and stable yield of ANPP [1,2,3]. Temperature and precipitation jointly regulate ANPP and plant diversity [4,5]. In the context of global warming and increased precipitation, a growing number of studies have explored the effects of climate warming [6,7], increased precipitation [8,9], or climate warming plus increased precipitation [10] on ANPP and plant diversity. However, two non-completely mutually nonexclusive debates remain. First, does experimental warming or increased precipitation have consistent effects on ANPP and plant diversity? Experimental warming and/or increased precipitation have negligible [10,11], positive [12,13], or negative effects on ANPP and/or plant diversity [14,15]. These diverse effects of warming and increased precipitation on ANPP and plant diversity are related to their distinct vegetation types, soil nutrition conditions, climatic conditions, and the magnitudes of warming or increased precipitation [5,13,16,17,18,19,20,21]. Second, do ANPP and plant diversity respond nonlinearly to experimental warming and/or increased precipitation? ANPP and plant diversity show linear relationships with air temperature and precipitation [12,22,23], indicating ANPP and plant diversity may respond linearly to experiment warming and increased precipitation. In contrast, the magnitude of warming and increased precipitation can not be significantly correlated with the response of ANPP and/or plant diversity to experimental warming and increased precipitation, respectively [24,25], implying that experimental warming and increased precipitation may have nonlinear effects on ANPP and plant diversity. Therefore, it is necessary to further explore the effects of climate warming and increased precipitation on ANPP and plant diversity.

As a sensitive region to climate change, the responses of ANPP and plant diversity to climate change on the Tibetan Plateau are important indicators of the global responses of ANPP and plant diversity to climate change [26,27,28]. Under the background of the Qinghai–Tibet Plateau as a whole tending to be warmer and wetter [1,29,30], a great deal of studies have investigated the effects of climate warming and increased precipitation on ANPP and plant diversity [4,5], which can provide important basic data and scientific theories for the conservation of global plant diversity, the high-quality development of animal husbandry, etc. However, besides the above debates, two other problems remain unresolved. First, several studies have shown a single-level experimental warming [15,31,32] or a multi-level experimental warming effect on ANPP and plant diversity [24,25]. No studies have reported whether there is an optimum warming magnitude for ANPP and plant diversity. Second, compared with α–diversity of plant species, there are few studies on the responses of plant phylogenetic α–diversity to climate change, and plant species and phylogenetic α–diversity have significant differences in reflecting plant α–diversity [1,33,34]. At the same time, compared with plant α–diversity, few studies have examined the responses of plant community composition to climate change, and changes in plant community composition can reflect plant β–diversity, which is essentially different from plant α–diversity [15,19]. Thus, it is not yet clear how ANPP and plant diversity will respond to future climatic change in alpine grasslands on the Qinghai–Tibetan Plateau.

Here, we reported a multi-level warming and increased precipitation experiment in an alpine meadow. The main objectives of this study were to examine (1) whether there was an optimum warming magnitude for the responses of ANPP and plant diversity to warming and whether the low- and high-level warming had different influences on ANPP and plant diversity; (2) whether the responses of ANPP and plant diversity to increased precipitation were related to the magnitude of increased precipitation; and (3) whether the inter-annual variations of ANPP and plant diversity were stronger than the effects of warming and increased precipitation on ANPP and plant diversity in the alpine meadow of Northern Tibet.

2. Results

2.1. Effects of Experimental Warming and Increased Precipitation on ANPP_community, ANPP_sedge, ANPP_graminoid, ANPP_forb, Species and Phylogenetic Diversity, and Enviromental Variables

There were significant main effects of experimental warming on ANPP_sedge, ANPP_forb, species richness, Shannon, PD, MNTD, species composition, and phylogenetic composition; significant or marginally significant main effects of increased precipitation on ANPP_community, species composition, and phylogenetic composition; and significant interactive effects of experimental warming and increased precipitation on species composition and phylogenetic composition, respectively (

Table 1. Repeated-measures analysis of variance was used to estimate the main and interactive effects of experimental warming (W), increased precipitation (IP), and measuring year (Y) on aboveground net primary production at community level (ANPP_community), sedge ANPP (ANPP_sedge), graminoid ANPP (ANPP_graminoid), forb ANPP (ANPP_forb), species richness (SR), Shannon, Simpson, Pielou, and Faith’s phylogenetic diversity (PD), and mean nearest taxon distance (MNTD).

Model	ANPPcommunity	ANPPsedge	ANPPgraminoid	ANPPforb	SR	Shannon	Simpson	Pielou	PD	MNTD
Warming (W)	1.00	4.00 *	1.00	4.00 *	15.00 ***	5.00 *	2.00	1.00	12.00 ***	4.00 *
Precipitation (IP)	3.00 +	2.00	1.00	1.00	0.00	0.00	0.00	0.00	0.00	1.00
Year (Y)	43.00 ***	7.00 ***	18.00 ***	11.00 ***	38.00 ***	25.00 ***	16.00 ***	4.00 **	19.00 ***	23.00 ***
W × IP	0.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00	0.00	1.00
W × Y	2.00	1.00	1.00	6.00 ***	4.00 ***	3.00 **	2.00 +	1.00	4.00 ***	2.00 *
IP × Y	3.00 **	3.00 **	1.00	0.00	1.00	1.00	1.00	1.00	1.00	1.00
W × IP × Y	1.00	1.00	1.00	0.00	1.00	2.00 +	1.00	1.00	1.00	2.00 *

⁺, *, **, and *** indicate p < 0.10, p < 0.05, p < 0.01, and p < 0.001, respectively.

and

Table 2. Permutational multivariate analysis of variance was used to estimate the main and interactive effects of experimental warming (W), increased precipitation (IP), and measuring year (Y) on species composition and phylogenetic composition.

Model	Species Composition	Phylogenetic Structure
Warming (W)	5.82 **	10.78 **
Precipitation (IP)	1.96 *	6.21 **
Year (Y)	29.67 **	52.89 **
W × IP	2.89 **	7.19 **
W × Y	1.91 **	2.08 *
IP × Y	0.79	−0.16
W × IP × Y	0.61	0.30

* and ** indicate p < 0.05 and p < 0.01, respectively.

). There were significant inter-annual variations of ANPP_community, ANPP_sedge, ANPP_graminoid, ANPP_forb, species richness, Shannon, Simpson, Pielou, PD, MNTD, species composition, and phylogenetic composition (Table 1 and Table 2).

The differences in species richness, Shannon, Simpson, Pielou, PD, MNTD, ANPP_community, ANPP_sedge, ANPP_graminoid, and ANPP_forb among the nine treatments varied with years (Figures S1–S3). The low- and high-level experimental warming-induced and the low- and high-level increased precipitation-induced dissimilarity of species composition and phylogenetic composition varied with years (Figures S4 and S5).

The high-level experimental warming reduced ANPP_sedge by 62.81% (−2.64 g m⁻²), species richness by 21.05% (−1.56), Shannon by 13.06% (−0.22), PD by 14.48% (−115.29), but increased ANPP_forb by 56.65% (5.41 g m⁻²) and MNTD by 33.88% (40.93) across the six growing seasons regardless of increased precipitation. Species richness, Shannon, and PD of the high-level experimental warming was 19.64% (−1.43), 9.67% (−0.15), and 14.66% (−117.02) lower than those of the low-level experimental warming across the six growing seasons regardless of increased precipitation, respectively. The low-level (species composition: F = 2.90, p = 0.013; phylogenetic composition: F = 4.16, p = 0.041) and high-level (species composition: F = 9.00, p = 0.001; phylogenetic composition: F = 16.16, p = 0.001) experimental warming altered species composition and phylogenetic composition across the six growing seasons regardless of increased precipitation. The high-level experimental warming-induced dissimilarity magnitude of species composition and phylogenetic composition was greater than that caused by low-level experimental warming across the six growing seasons regardless of increased precipitation, respectively (species composition: F = 11.10, p = 0.002; phylogenetic composition: F = 12.13, p = 0.001).

The low-level (species composition: F = 3.61, p = 0.006; phylogenetic composition: F = 14.49, p = 0.001) rather than the high-level (species composition: F = 0.80, p = 0.573; phylogenetic composition: F = 0.97, p = 0.433) increased precipitation altered species composition and phylogenetic composition across the six growing seasons, regardless of experimental warming. However, there were few differences between the low- and high-level increased precipitation-induced dissimilarity magnitudes of species composition (F = 1.50, p = 0.226) and phylogenetic composition (F = 3.15, p = 0.082) across the six growing seasons, regardless of experimental warming.

Significant or marginally significant main effects of experimental warming on SM, VPD, GSP/AccT, and T_s were observed (Table S1, Figure S6). Significant main effects of increased precipitation on T_s, SM, T_a, VPD, AccT, GSP/AccT, NH₄⁺–N, and AP were detected (Table S1, Figure S6). There were significant or marginally significant interactive effects of experimental warming and increased precipitation on T_s, SM, T_a, VPD, AccT, GSP/AccT, and AP (Table S1, Figure S6). Inter-annual variations of T_s, SM, T_a, VPD, AccT, GSP/AccT, NH₄⁺–N, NO₃⁻–N, AP, and pH were observed (Table S1, Figure S6).

The change magnitude of SM caused by experimental warming (Δ_W_SM) was negatively correlated with that of T_s (Δ_W_T_s) and T_a (Δ_W_T_a), while that of VPD (Δ_W_VPD) was positively correlated with Δ_W_T_s and Δ_W_T_a (Figure S7). The magnitude of the change in GSP/AccT caused by experimental warming (Δ_W_GSP/AccT) decreased with increasing Δ_W_T_a (Figure S7). Regardless of increased precipitation, the low- and high-level experimental warming increased T_s by 1.13 °C and 2.59 °C, T_a by 2.39 °C and 3.86 °C, AccT by 296.26 °C and 476.16 °C, and VPD by 0.20 kPa and 0.35 kPa, but decreased SM by 0.02 m³ m⁻³ and 0.05 m³ m⁻³, and GSP/AccT by 0.06 °C mm⁻¹ and 0.09 °C mm⁻¹ across the six growing seasons, respectively. The T_s, T_a, AccT, and VPD of the high-level experimental warming were greater than those of the low-level experimental warming, whereas the SM and GSP/AccT of the high-level experimental warming were lower than those of the low-level experimental warming, respectively. Increased precipitation-induced change magnitudes of T_s (Δ_IP_T_s) and VPD (Δ_IP_VPD) decreased with increasing that of growing season precipitation (Δ_IP_GSP), but that of SM (Δ_IP_SM) and GSP/AccT (Δ_IP_GSP/AccT) increased with increasing Δ_IP_GSP (Figure 1). Regardless of experimental warming, the low- and high-level increased precipitation increased SM by 0.02 m³ m⁻³ and 0.04 m³ m⁻³, and GSP/AccT by 0.03 °C mm⁻¹ and 0.10 °C mm⁻¹, but decreased T_s by 0.28 °C and 0.36 °C, and VPD by 0.05 kPa and 0.10 kPa across the six growing seasons, respectively. The SM and GSP/AccT of the high-level increased precipitation were greater than those of the low-level increased precipitation, respectively.

The comparison of soil NH₄⁺–N, NO₃⁻–N, AP, and pH among the nine treatments were illustrated in Figure S8. The change magnitude of NO₃⁻–N caused by increased precipitation (Δ_IP_NO₃⁻–N) increased with increasing Δ_IP_GSP (Figure 1). Regardless of increased precipitation, the low- and high-level experimental warming increased soil NH₄⁺–N by 30.93% (3.02 mg kg⁻¹) and 32.74% (3.20 mg kg⁻¹), and soil AP by 28.60% (2.49 mg kg⁻¹) and 28.81% (2.51 mg kg⁻¹) across the four growing seasons, respectively. The high-level experimental warming increased soil NO₃⁻–N by 74.47% (9.51 mg kg⁻¹), and the soil NO₃⁻–N of the high-level experimental warming was 43.92% greater than that of the low-level experimental warming across the four growing seasons.

2.2. Relationships between Experimental Warming–Induced Change Magnitude and Response Ratio of Biotic Variables, βBray_W and βMNTD_W, and Experimental Warming–Induced Change Magnitude of Abiotic Variables and Warming Duration

The effects of warming on species richness, PD, ANPP_community, and ANPP_graminoid decreased with warming magnitude (Figure 2 and Figure S9–S11). The effects of warming on Pielou, MNTD, and ANPP_forb increased with Δ_W_T_a (Figure 2 and Figure S9). The effects of warming on Pielou and ANPP_forb showed quadratic relationships with Δ_W_T_s (Figures S10 and S11). The response ratio of ANPP_sedge to experimental warming (R_W_ANPP_sedge) and the change magnitude of ANPP_sedge caused by experimental warming (Δ_W_ANPP_sedge) both showed quadratic relationships with Δ_W_T_a, and R_W_ANPP_sedge and Δ_W_ANPP_sedge reached their minimum values when Δ_W_T_a was about 2.82 °C and 3.57 °C, respectively (Figure 2 and Figure S9). The higher was Δ_W_T_a, the greater were the warming-induced differences in species composition (βBray_W) and phylogenetic composition (βMNTD_W) (Figure 2). The βBray_W showed a quadratic relationship with Δ_W_T_s, and the βBray_W reached its maximum value when Δ_W_T_s was about 2.08 °C (Figure S10).

The greater the experimental warming-induced soil and/or air drying, the greater the experimental warming-induced reductions in species and phylogenetic α–diversity, ANPP_community, ANPP_sedge, and ANPP_graminoid (Figures S12 and S13). The effect of warming on ANPP_forb showed opposite relationships with experimental warming-induced soil and air drying (Figures S12 and S13). The reduction in GSP/AccT caused by experimental warming can cause species loss, ANPP_graminoid reduction, and ANPP_forb increase (Figures S12 and S13). The greater the experimental warming-induced reduction in GSP/AccT, the greater the warming-induced differences in species composition (βBray_W) and phylogenetic composition (βMNTD_W) (Figure S12).

The effects of experimental warming on species richness Shannon, Simpson, and PD showed quadratic relationships with the magnitude of the change in soil pH caused by experimental warming (Δ_W_pH), and βMNTD_W decreased with Δ_W_pH (Figures S14 and S15). The effects of experimental warming on species α–diversity, ANPP_community, PD, ANPP_sedge, and ANPP_graminoid decreased with increasing Δ_W_NO₃⁻–N (the change magnitude of NO₃⁻–N caused by experimental warming), but the effect of experimental warming on MNTD increased with increasing Δ_W_NO₃⁻–N (Figures S14 and S15). Both βBray_W and βMNTD_W increased with increasing Δ_W_NO₃⁻–N (Figure S14). The effect of experimental warming on ANPP_forb showed quadratic relationships with the change magnitude of NH₄⁺–N caused by experimental warming (Δ_W_NH₄⁺–N) (Figures S14 and S15).

The effects of warming on Shannon, Simpson, and ANPP_graminoid decreased with warming duration, and the effects of warming on MNTD and ANPP_forb increased with warming duration (Figure 3 and Figure S16). Moreover, the effects of warming on species richness, PD, ANPP_community, and ANPP_sedge showed quadratic relationships with warming duration, and there was a time point when the lowest effect of warming on species richness, PD, ANPP_community, and ANPP_sedge occurred, respectively (Figure 3 and Figure S16). The longer the warming duration, the greater the warming-induced differences in species composition (βBray_W) and phylogenetic composition (βMNTD_W) (Figure 3).

The varpart results showed that warming duration, experimental warming-induced changes in environment temperature and moisture, soil nitrogen and phosphorus, and/or soil pH together controlled the variation of species, phylogenetic diversity, and aboveground plant production under controlled warming conditions (Figures S17–S21 and Figure 4).

2.3. Relationships between Experimental Warming–Induced Change Magnitude and Response Ratio of ANPP_community and Experimental Warming–Induced Change Magnitude and Response Ratio of α–Diversity, βBray_W and βMNTD_W

R_W_ANPP_community increased with increasing R_W_SR and R_W_PD, but decreased with increasing R_W_Pielou (Figure S22). Δ_W_ANPP_community decreased with increasing Δ_W_Pielou and Δ_W_Simpson but increased with increasing βMNTD_W (Figure S22). The effect of experimental warming on the ANPP_community was simultaneously regulated by species and phylogenetic α– and β–diversity (Figure S23).

2.4. Relationships between Increased Precipitation–Induced Change Magnitude and Response Ratio of Biotic Variables, βBray_IP and ΒMNTD_IP, and Abiotic Variables and Increased Precipitation Duration

Increased precipitation-induced change magnitude of ANPP_sedge (Δ_IP_ANPP_sedge) increased with increasing Δ_IP_GSP, and there was a turning point from a negative effect of increased precipitation to a positive effect of increased precipitation on ANPP_sedge (Figure 1). In contrast, increased precipitation-induced change magnitude of MNTD (Δ_IP_MNTD) and increased precipitation-induced response ratio of MNTD (R_IP_MNTD) decreased with increasing Δ_IP_GSP, and there was a turning point from a positive effect of increased precipitation to a negative effect of increased precipitation on MNTD (Figure 1).

Increased precipitation-induced change in T_s, SM, T_a, VPD, NH₄⁺–N, NO₃–N, and/or AP may have some relationships with the effects of increased precipitation on species, phylogenetic diversity, and aboveground plant production (Figures S24–S27). For example, the effect of increased precipitation on species richness increased with Δ_IP_AP (the change in magnitude of AP caused by increased precipitation) (Figure S26).

The effects of increased precipitation on SR, ANPP_community, ANPP_sedge, and ANPP_graminoid increased with increased precipitation duration, and the effect of increased precipitation on Pielou decreased with increased precipitation duration (Figure 5 and Figure S28). There was a turning point from a negative effect of increased precipitation to a positive effect of increased precipitation on SR, ANPP_community, ANPP_sedge, and ANPP_graminoid, respectively (Figure 5 and Figure S28). There was a turning point from a positive effect of increased precipitation to a negative effect of increased precipitation on Pielou (Figure 5 and Figure S28). The longer the increased-precipitation duration, the greater the increased precipitation-induced difference in species composition (βBray_IP) and phylogenetic composition (βMNTD_IP) (Figure 5).

The shared and excluded effects of increased precipitation duration, Δ_IP_T (Δ_IP_T_s and/or Δ_IP_T_a), Δ_IP_W (Δ_IP_GSP, Δ_IP_SM, and/or Δ_IP_VPD), and Δ_IP_GSP/AccT on the response ratio of biotic variables to increased precipitation and the change magnitude of biotic variables caused by increased precipitation were illustrated in Figures S29 and S30, respectively. The shared and excluded effects of increased precipitation duration, Δ_IP_N (Δ_IP_NH₄⁺–N and/or Δ_IP_NO₃⁻–N), Δ_IP_AP and Δ_IP_pH on the response ratio of biotic variables to increased precipitation and the change magnitude of biotic variables caused by increased precipitation were illustrated in Figures S31 and S32, respectively. The shared and exclusive effects of increased precipitation duration, Δ_IP_T&W (Δ_IP_T_s, Δ_IP_T_a, Δ_IP_GSP, Δ_IP_SM, Δ_IP_VPD and/or Δ_IP_GSP/AccT), Δ_IP_N&P (Δ_IP_NH₄⁺–N, Δ_IP_NO₃⁻–N and/or Δ_IP_AP), and Δ_IP_pH on the response ratio of biotic variables to increased precipitation and the change magnitude of biotic variables caused by increased precipitation were illustrated in Figure 6 and Figure S33, respectively. These varpart results showed that increased precipitation duration, increased precipitation-induced changes in environment temperature and moisture, soil nitrogen and phosphorus, and/or soil pH together controlled the variation of species, phylogenetic diversity, and aboveground plant production under controlled increased precipitation conditions.

2.5. Relationships between Increased–Precipitation–Induced Change Magnitude and Response Ratio of ANPP_community and Increased–Precipitation–Induced Change Magnitude and Response Ratio of α–Diversity, βBray_IP and βMNTD_IP

R_IP_ANPP_community decreased with increasing R_IP_Shannon, R_IP_Simpson and R_IP_Pielou, and Δ_IP_ANPP_community decreased with increasing Δ_IP_Shannon, Δ_IP_Simpson and Δ_IP_Pielou (Figure S34). Both R_IP_ANPP_community and Δ_IP_ANPP_community increased with increasing βBray_IP and βMNTD_IP (Figure S34). The effect of increased precipitation on ANPP_community was simultaneously regulated by species and phylogenetic α– and β–diversity under increased precipitation conditions (Figure S35).

2.6. Shared and Excluded Effects of Experimental Duration, Environmental Variables, Species and Phylogenetic Diversity on ANPP_community

All the variables in the varpart analysis explained about 82%, 84%, 79%, and 79% variations of R_W_ANPP_community, Δ_W_ANPP_community, R_IP_ANPP_community, and Δ_IP_ANPP_community, respectively (Figure 7). The excluded effects on R_W_ANPP_community was in an order of Δ_W_Env, R_W_α–diversity, βBray_W&βMNTD_W and Duration_w (Figure 7a). The excluded effects on Δ_W_ANPP_community was in an order of Δ_W_α–diversity, Δ_W_Env, Duration_w, and βBray_W&βMNTD_W (Figure 7b). The excluded effects on R_IP_ANPP_community was in an order of R_IP_α–diversity, Δ_IP_Env, Duration_IP, and βBray_IP&βMNTD_IP (Figure 7c). The excluded effects on Δ_IP_ANPP_community were in the order of Duration_IP, Δ_IP_Env and Δ_IP_α–diversity, and the βBray_IP&βMNTD_IP had no excluded effect on Δ_IP_ANPP_community (Figure 7d).

3. Discussion

3.1. Warming Effects

The high-level warming caused greater reductions in species richness and PD than did the low-level warming. First, low temperature is a key limited factor for alpine growth, but excessive temperature may lead to the death of temperature-sensitive species and, in turn, species loss [4,14] and the decline in PD. A greater warming can often cause greater increases in T_s and T_a [10,35]. Second, water availability is another key limited factor for plant growth, and drying may lead to stomatal closure and photorespiration [17,36]. Warming-induced drying may cause the death of drying-sensitive species and, in turn, the loss of species and related phylogenetic information [37,38]. A greater warming can often cause greater soil and air drying [10,35]. Third, GSP/AccT is often positively correlated with plant growth [4,10,39]. The greater decline in GSP/AccT caused by high-level warming caused a greater reduction in species richness and related phylogenetic information. Fourth, soil NO₃⁻–N is an important nitrogen resource for alpine plant growth [2,3,20,40,41,42], and there can be species-specific preferences for soil NO₃⁻–N [43,44]. However, only high-level warming increased soil NO₃⁻–N.

The high-level warming caused a greater reduction in Shannon than did the low-level warming, which may be due to the high-level warming-induced greater reduction in SM and a greater increase in NO₃⁻–N (Figures S12–S15). The high- rather than low-level warming increased MNTD, which may be due to the greater increase in T_a and NO₃⁻–N, and the greater reduction in GSP/AccT caused by the high-level warming (Figure 2, Figures S9 and S12–S15).

Low- and high-level warming did not alter Simpson and Pielou. First, the effect of warming on Simpson may be mainly related to warming-induced increases in NO₃⁻–N, whereas the effect of warming on Pielou may be mainly related to warming–induced increases in T_s and T_a (Figure 2, Figures S9–S11, S14 and S15). Second, the low- and high-level warming-induced increases in T_s, T_a, and NO₃⁻–N were close to the turning points of T_s, T_a, and NO₃⁻–N where the R_W_Simpson and R_W_Pielou were equal to 1, or Δ_W_Simpson and Δ_W_Pielou were equal to zero (Figure 2, Figures S9–S11, S14 and S15). Third, the decreased magnitude of Simpson increased with warming duration (Figure 3 and Figure S16), which indicated that short-term warming caused a negligible effect on Simpson.

The high-level warming resulted in greater dissimilarities in species and phylogenetic composition than did the low-level warming. First, there was an optimum of Δ_W_T_s (about 2.08 °C) when the βBray_W was the greatest. The high-level warming-induced by the increase in the Δ_W_T_s (about 2.59 °C) was closer to the optimum of Δ_W_T_s than the low-level warming–induced in the increase in the Δ_W_T_s (about 1.13 °C) (Figure S10). Second, compared to the low-level warming-induced lower increases in Δ_W_T_a and Δ_W_NO₃⁻–N, the high-level warming-induced greater increases in Δ_W_T_a and Δ_W_NO₃⁻–N can tend to cause greater βBray_W and βMNTD_W (Figure 2 and Figure S14). Third, compared to the low-level warming-induced lower reduction in GSP/AccT, the high-level warming-induced greater reduction in GSP/AccT can tend to cause greater βBray_W and βMNTD_W (Figure S12).

The high- rather than low-level warming reduced ANPP_sedge, which was similar to a previous study demonstrating an increase in T_a can reduce sedge coverage in a Northern Tibet alpine meadow [25], and may be due to the high-level warming-induced greater increase in T_a, soil drying, AP, and NO₃⁻–N (Figures S9 and S13–S15).

The low- and high-level warming did not alter ANPP_graminoid, and there was no difference in ANPP_graminoid between the low- and high-level warming. Similarly, no significant main effect of warming on graminoid coverage was observed in a four-level warming (control, +1.00, +2.70, and +4.00 °C, respectively) experiment [45]. First, the low-level warming–induced the increases in T_s, T_a and VPD tended to increase R_W_ANPP_graminoid and Δ_W_ANPP_graminoid, but the high-level warming-induced increases in T_s, T_a, and/or VPD tended to reduce R_W_ANPP_graminoid and Δ_W_ANPP_graminoid (Figure 2 and Figure S10). The low-level warming-induced reduction in SM tended to increase Δ_W_ANPP_graminoid, but the high-level warming-induced reduction in SM tended to decrease Δ_W_ANPP_graminoid (Figure S13). The absolute values of the changes of R_W_ANPP_graminoid caused by the increases in T_s, T_a, and VPD under the low-level warming were greater than those under the high-level warming, whereas the absolute values of the changes of Δ_W_ANPP_graminoid caused by the changes in T_s, T_a and SM under the low-level warming were lower than those under the high-level warming. Second, the low- and high-level warming-induced the reduction in GSP/AccT tended to increase the R_W_ANPP_graminoid (Figure S12), and with a greater increase caused by the low-level warming-induced change in GSP/AccT. Third, the high- rather than the low-level warming-induced the increase in NO₃⁻–N tended to increase R_W_ANPP_graminoid (Figure S14). Fourth, the low- and high-level warming-induced the increase in NH₄⁺–N tended to increase the Δ_W_ANPP_graminoid (Figure S15), and with a greater increase caused by the low-level warming-induced change in GSP/AccT. Fifth, a greater reduction in ANPP_graminoid can occur along with warming duration (Figure 3), which indicated that short–term warming (<7 years) caused negligible effect of warming on ANPP_graminoid.

The high- rather than low-level warming increases ANPP_forb. First, the high-level warming-induced greater increase in T_s and T_a tended to cause a greater increase in R_W_ANPP_forb and Δ_W_ANPP_forb than did the low-level warming (Figure 2 and Figures S9–S11). Second, the high-level warming-induced greater soil drying and reduction in GSP/AccT tended to cause a greater increase in R_W_ANPP_forb and Δ_W_ANPP_forb than did the low-level warming (Figures S12 and S13).

The low- and high-level warming did not alter ANPP_community, and there was no difference in ANPP_community between the low- and high-level warming. Likewise, no main effect of warming on ANPP_community was detected in a four-level (control, +1.00, +2.70, and +4.00 °C, respectively) [45] and five-level warming experiment (control, +1.13, +1.66, +2.10, and +2.72 °C, respectively) [25]. There was a negligible difference in ANPP between low- and high-level warming (1.05 and 1.69 °C, respectively) in an alpine meadow [24]. First, the quite contrary effects of warming on ANPP_sedge and ANPP_forb may weaken the effect of warming on ANPP_community. Meanwhile, the negligible effect of warming on ANPP_graminoid can further weaken the warming effect on ANPP_community. Second, considering the obvious relationships of ANPP_community with Simpson and Pielou (Figure S22), the negligible effect of the low- and high-level warming on ANPP_community should be related to that on Simpson and Pielou. Third, warming-induced changes in T_s, T_a, VPD, SM, NO₃–N, and AP may have exclusive effects on ANPP_community. For example, drying climate conditions can dampen and even mask the effect of increased T_s and/or T_a on ANPP_community in grasslands [25,45] by reducing leaf area and inducing stomatal closure [46], and the average GSP (405.5 mm) in 2014–2019 was only equal to that (400.5 mm) in 1963–2019.

3.2. Increased Precipitation Effects

Increased precipitation did not change species and phylogenetic α–diversity, ANPP_community, ANPP_sedge, ANPP_graminoid and ANPP_forb. Likewise, several previous studies demonstrated that increased precipitation did not affect species richness in alpine meadows [47], annual grasslands [48], infertile grasslands [38], and tallgrass prairies [49]. A low- and high-level increase in precipitation (20% and 40%, respectively) did not impact sedge aboveground production in an alpine meadow on the Tibetan Plateau [50]. No differences in graminoid coverage were detected among control, 20% and 40% increased precipitation in an alpine meadow [50], and control, 15% and 30% increased precipitation in an annual forb-dominated desert steppe [51]. A previous study also demonstrated that increased precipitation did not increase forb coverage [52]. First, except for MNTD and ANPP_sedge, increased precipitation may only have indirect effects on species, phylogenetic α–diversity, and plant production, considering that only the effects of increased precipitation on MNTD and ANPP_sedge had significant correlations with Δ_IP_GSP. Second, although the increases in SM and GSP/AccT and the reduction in VPD under increased precipitation may be favorable for alpine plants, these probable positive effects may be dampened and even masked by the probable negative effect of the decrease in T_s under increased precipitation. Third, no effects of increased precipitation on soil pH, nitrogen, or phosphorus availability can explain the negligible effects of increased precipitation on species and phylogenetic α–diversity, ANPP_community, ANPP_sedge, ANPP_graminoid, and ANPP_forb. Fourth, the negligible effect of increased precipitation on ANPP_community can be related to that on species and phylogenetic α–diversity, ANPP_sedge, ANPP_graminoid, and ANPP_forb. Fifth, the negligible effects of increased precipitation may also be related to the short-term duration (<7 years), because a greater increase or reduction in species and phylogenetic α–diversity, ANPP_community, ANPP_sedge, and ANPP_graminoid may occur along with an increasing duration of increased precipitation (Figure 5 and Figure S28).

3.3. Interactive Effects of Experimental Warming and Increased Precipitation

Several previous studies indicated that the main effect of warming and increased precipitation on species diversity and plant production may overestimate/underestimate the interactive effect of warming and increased precipitation on species diversity and plant production in grasslands [10,38], which was supported by this study. The negligible interactive effects of warming and increased precipitation on species richness and ANPP_community were also in line with some previous studies [53,54]. Although warming can elevate T_s and T_a, it can also cause drying. Although increased precipitation can increase water availability, increased precipitation can also decrease T_s. Moreover, both plant phyllosphere and soil microbial communities can also affect plant growth, α–diversity and community composition [55,56,57,58]. Climate change may cause plant phyllosphere and soil microbial communities to develop in a direction that is unfavorable to plant growth (e.g., the increase of plant pathogens) [55].

3.4. Stronger Inter-Annual Variations than Effects of Warming and Increased Precipitation

Inter-annual variations of species, phylogenetic diversity, and aboveground plant production were stronger than the effects of experimental warming or increased precipitation on these plant variables. This phenomenon was similar to some previous studies [21,35]. First, the maximum difference in growing season air temperature among years in 1963–2019 (2.7 °C) under natural conditions was greater than that (2.4 °C) under low-level warming. The maximum GSP difference among years in 2014–2019 (288.9 mm) and 1963–2019 (364.9 mm) under natural precipitation conditions was greater than the Δ_IP_GSP (45.0–176.7 mm) under increased precipitation. Second, aboveground plant production, plant species richness, and PD may be restored to their original levels with increasing warming duration (Figure 3 and Figure S16). Therefore, significant interannual variation can be related to the high inter-annual variations of environmental temperature and moisture and strong adaptation and resilience [59].

4. Materials and Methods

4.1. Study Area, Experimental Design and Microclimate Measurements

This experiment was conducted at an alpine grassland site (30°30′ N, 91°04′ E, 4313 m) from June 2014 to 2019. The mean annual temperature and mean annual precipitation were 1.96 °C and 476.36 mm in 1963–2019, respectively [17,21]. The dominant species are Carex atrofusca, Stipa capillacea, and Kobresia pygmaea [2,17]. Open-top chambers with 40 cm and 80 cm openings were used to simulate low- and high-magnitude warming, respectively. The opening sizes of these open-top chambers are hexagons with a side length of 60 cm. All the materials in the open top chambers are polythene. We also simulated two levels of increased precipitation (15% and 30%). Each treatment had three replicates. The nine treatments are control (CK), low- and high-level warming (LW and HW), low- and high-level increased precipitation (LP and HP), and their interactive effects (i.e., LW + LP, HW + LP, LW + HP, and HW + HP). We measured soil temperature (T_s, 5 cm), soil moisture (SM, 10 cm), air temperature (T_a, 15 cm), and relative humidity (RH, 15 cm) by HOBO weather stations (Onset Computer, Bourne, USA) during the growing season (June–September). We then calculated vapor pressure deficit (VPD) using measured T_a and RH and calculated the ratio of growing season precipitation to accumulated ≥5 °C daily air temperature (GSP/AccT). The effects of experimental warming and increased precipitation on T_s, SM, T_a, VPD, and/or GSP/AccT in 2014–2017 were reported in previous studies [10,35].

4.2. Community Investigation, ANPP Estimation, Soil Sampling, and Analyses

Plant community investigations for each plot in August 2014–2019 were conducted. The quadrat size of the community investigation was 50 cm × 50 cm. When we did the community investigation, the 50 cm × 50 cm quadrat was placed right in the center of each treatment plot. We recorded species coverage and height for each species within each of the twenty-seven plots. All the species names were artificially identified, and species coverage was artificially estimated. The 50 cm × 50 cm quadrat was divided into twenty-five 10 cm × 10 cm subquadrats to better estimate species coverage. The species height was measured using a steel tape with millimeter accuracy. There were three plant functional groups: sedge, graminoids, and forbs. Aboveground net primary production of sedge, graminoids, and forbs (i.e., ANPP_sedge, ANPP_graminoid and ANPP_forb) was estimated using observed coverage and height of plant function groups [5], respectively.

$$ANP{P}_{sedge}=-0.77+0.93Coverage-0.24Height,R^2=0.95,p<0.001,n=120$$

(1)

$$ANP{P}_{graminoid}=-2.25+1.03Coverage+0.06Height,R^2=0.95,p<0.001,n=120$$

(2)

$$ANP{P}_{forb}=-1.00+0.50Coverage+1.71Height,R^2=0.91,p<0.001,n=120$$

(3)

The aboveground net primary production of plant community (ANPP_community) was the sum of ANPP_sedge, ANPP_graminoid, and ANPP_forb. Topsoil (0–10 cm) samples within all plots (areas outside the community investigation but inside each plot) were collected, sieved, and used to measure ammonium nitrogen (NH₄⁺–N), nitrate nitrogen (NO₃⁻–N), available phosphorus (AP), and pH in August 2014 and 2016–2018. Both NH₄⁺–N and NO₃⁻–N were measured on a LACHAT Quikchem Automated Ion Analyzer [60]. We measured soil pH using a soil pH meter [33,61,62]. We measured AP using the ammonium bicarbonate extraction molybdenum antimony resistance colorimetric method [18].

4.3. Statistical Analyses

We used the “specnumber” and “diversity” (vegan package) of R.4.1.2 software to calculate species α–diversity (SR—species richness; Shannon, Simpson, and Pielou).

$$Shannon=-{\sum }_{i=1}^n{p}_i\times log{p}_i$$

(4)

$$Simpson=1-{\sum }_{i=1}^n{p}_i^2$$

(5)

$$Pielou=\frac{Shannon}{logSR}$$

(6)

where p_i is the important value of each species within each sample. The important value of each plant species was the mean value of its relative coverage and relative height. We used the “vegdist” (Vegan package) of R.4.1.2 software to calculate species β–diversity (i.e., βBray, the dissimilarity indices of Bray–Curtis) between two treatments. We used “TPL” and “taxa.table” (Plantlist package) of R.4.1.2 software to obtain the taxonomy information (i.e., family, genus, and species) and then used Phylomatic software to generate a phylogenetic tree. We used the “pd”, “mntd”, and “comdistnt” (Picante package) of R.4.1.2 software to get phylogenetic α–diversity (PD: Faith’s phylogenetic diversity, i.e., the sum of the total phylogenetic branch length in one sample, and MNTD: mean nearest taxon distance for taxa in one sample) and phylogenetic β–diversity (i.e., βMNTD: beta mean nearest taxon distance) between two treatments. We used repeated-measures analysis of variance to estimate the main and interactive effects of experimental warming, increased precipitation, and measuring year on T_s, SM, T_a, VPD, AccT, GSP/AccT, NH₄⁺–N, NO₃⁻–N, AP, pH, SR, Shannon, Simpson, Pielou, PD, MNTD, ANPP_community, ANPP_sedge, ANPP_graminoid, and ANPP_forb. We then used Duncan multiple comparisons to examine the differences among the three experimental warming or increased precipitation levels. We used permutational multivariate analysis of variance (i.e., the “adonis2” of Vegan package) for the effects of experimental warming, increased precipitation, and measuring year on the βBray and βMNTD based on the R.4.1.2 software. The change magnitude of the abiotic and biotic variables caused by experimental warming (Δ_W) or increased precipitation (Δ_IP), and the response ratio of biotic variables to experimental warming (R_W) or increased precipitation (R_IP) were used as the effect size of experimental warming or increased precipitation for each year, respectively [63].

$${\mathsf{\Delta }}_{\mathrm{W}} \mathrm{or} {\mathsf{\Delta }}_{\mathrm{IP}}=\overline{X_t}-\overline{X_c}$$

(7)

$$R_{\mathrm{W}} \mathrm{or}{R}_{\mathrm{IP}}=\overline{X_t}/\overline{X_c}$$

(8)

For the Δ_W and R_w, $\overline{X_t}$ and $\overline{X_c}$ were concerned variables of the “LW” and “CK”, “HW” and “CK”, “LW + LP” and “LP”, “HW + LP” and “LP”, “LW + HP” and “HP”, or “HW + HP” and “HP”, respectively. For Δ_IP and R_IP), $\overline{X_t}$ and $\overline{X_c}$ were concerned variables of the “LP” and “CK”, “HP” and “CK”, “LW + LP” and “LW”, “LW + HP” and “LW”, “HW + LP” and “HW”, or “HW + HP” and “HW”, respectively. We used “varpart” (Vegan package) of R.4.1.2 software to partition the variation of the plant α– and β–diversity and aboveground net plant production by four explanatory matrices of biotic and/or abiotic variables. All the statistical analyses were examined at p < 0.05.

5. Conclusions

In summary, warming and increased precipitation were not always favorable for aboveground net primary production, species and phylogenetic α–diversity and community composition, which were related to the duration and magnitude of warming and increased precipitation, respectively. Species and phylogenetic α–diversity and composition did not have entirely uniform responses to warming and increased precipitation, and they had different correlations with aboveground net primary production. The combination of species and phylogenetic α–diversity and composition can better reflect the effects of climate warming and increased precipitation on plant α–diversity and community composition, and also better explain the variation of aboveground net primary production under controlled warming and increased precipitation conditions. Aboveground net primary production, species and phylogenetic α–diversity, and community composition had obvious inter-annual variations, and their variations were greater than their responses to warming and increased precipitation. Sedge, graminoid, and forb aboveground net primary production had different responses to warming.

The scientific findings of this study can provide some guidance for the conservation of plant diversity and the development of animal husbandry. First, compared to the regions with lower warming, more attention should be paid to the conservation of plant diversity in regions with greater warming. Second, in the context of climate change, biodiversity conservation policies should take into consideration both species and phylogenetic diversity. Third, we may pay more attention to the large inter-annual changes of ANPP and plant diversity than the effects of climate change on ANPP and plant diversity in terms of the stability of livestock and herders. Fourth, we need to be aware that the actual effects of warmer-wetter climate change trends on ANPP and plant diversity may be lower than their expected effects.