1. Introduction
Arid areas, including hyper-arid, arid, and semiarid regions, are of particular importance to humans since they support more than one-fifth of the global population [
1]. The ecosystems in the arid area are fragile and sensitive to climate change and anthropogenic activities [
2,
3,
4,
5] and play a dominant role in the interannual variability of global net primary productivity (NPP) [
6]. The Tianshan North Slope in Xinjiang province of China, located in the inland of Eurasia and far away from the oceans, is one of the driest regions in the world. This region has experienced a faster warming trend with a rate of 0.40 °C warming per decade [
7], which is much higher than the global mean rate of 0.27 °C per decade [
8]. A wetter trend is also detected by observations during the last several decades [
9,
10]. These climate changes play a very important role in the carbon and water flux dynamics of the arid ecosystems [
11,
12,
13,
14], and have drawn great concern about the sustainable development of the arid ecosystems. Due to the development of the industry and the intensive human activities, the observations from Mauna Loa show that the atmospheric CO
2 concentration has increased by 29.3% since 1959 and exceeded 400 ppm in 2015 [
15]. Moreover, previous experimental and modeling studies have shown that arid ecosystems benefit from the increase of atmospheric CO
2 concentration, which is called the CO
2 fertilization effect [
2,
16,
17]. However, different vegetation types have different responses to these changes resulting from the temporal and spatial heterogeneity of the climate change and the unique characteristics of each vegetation type [
12,
18]. Therefore, the dominant factor differs for each vegetation type. For example, by scenario simulations, Zhang et al. [
17] quantitatively explored the responses of the arid ecosystems in Central Asia to the changes in climate and CO
2 concentrations by scenario simulations, and they found that the NPP of the natural ecosystems was sensitive to precipitation change, while the NPP of cropland was sensitive to the CO
2 fertilization effect.
The carbon and water fluxes are key functions of the ecosystem, and are widely used to measure the status of the ecosystem [
19,
20,
21,
22,
23,
24]. The carbon flux, NPP, which is the net carbon retained by the vegetation in the ecosystem, is a balance between the carbon uptake from the photosynthetic function and the carbon losses caused by the plant respiration. The water flux, evapotranspiration (ET), is the sum of the evaporation of the soil, evaporation of the canopy interception, and the transpiration of vegetation. The NPP is the key component in the global carbon cycle while the ET is the major component of the global hydrological cycle [
21,
23]. On one hand, both NPP and ET are vulnerable to the changes in the climate and the atmospheric CO
2 concentration and the anthropogenic activities. Based on model simulations, Han et al. [
12] found that the annual NPP of grassland in arid Central Asia followed a decreasing trend (−0.21 g C m
−2 year
−1) from 1979 to 2011, resulting from the increase of annual mean temperature and the decline of precipitation. During the same period, the ET of grassland experienced significant declines of 1.47–2.72 mm per decade due to the climate change [
25]. On the other hand, the two indicators coexist in the photosynthesis and respiration processes and are controlled by the status of the stomas [
26,
27,
28]. It is therefore necessary to explore the responses of the ecosystem to the climate change and CO
2 fertilization effect simultaneously, especially for the future projections. Additionally, it is of significant importance for the sustainable development of the arid ecosystems and the policy making for climate adaption and mitigation to study the NPP and ET dynamics of different arid ecosystems under multiple representative concentration pathway (RCP) scenarios.
For an ecosystem, the water use efficiency (WUE) is defined by the ratio between carbon sequestration and water consumption, and it can be used to describe the tight link between carbon and water cycles within the ecosystem. Compared to NPP and ET, WUE is more sensitive to the ambient changes such as changes in precipitation, temperature, and CO
2 concentration. So it has been widely used to detect the responses of the ecosystems to historical changes in climate and CO
2 concentration. Under the global climate change, the annual WUE shows an increasing trend over middle and high northern latitudes [
6,
29]. Previous studies showed that the increasing CO
2 contributes to the increase of WUE by influencing the stomatal closure [
30]. In addition, the climate change can enhance the effects of CO
2 on WUE [
29]. Moreover, the WUEs of different ecosystems respond differently in climate change [
31,
32]. However, previous studies mainly focused on the WUE responses to historical changes in climate and CO
2. So it is very urgent for us to know about the WUE dynamics of the ecosystems in the arid area where the ecosystems are sensitive to the ambient changes.
The ecosystem process-based modelling is one of the best choices for predicting the carbon and water dynamics in global carbon and water cycles [
33,
34]. The arid ecosystem model (AEM) is selected to explore the dynamics of the NPP and ET in the Tianshan North Slope under multiple RCP scenarios at a horizontal resolution of 25 km. Compared to the popular ecosystem models, the AEM model can represent the unique root and canopy structure of the xeric vegetation and ecophysiological processes [
35]. Additionally, it has been well validated and successfully applied to study the impacts of the climate change and CO
2 fertilization effect on the carbon and water cycles of the arid ecosystems in Central Asia, including the Tianshan North Slope area [
3,
4,
14,
17,
36,
37]. The AEM model is forced by dynamically downscaled data from the RCP2.6, RCP4.5, and RCP8.5 at 25 km resolution from 2006 to 2055 during which significant climate change has been detected [
38]. These three RCP scenarios (RCP2.6, RCP4.5, and RCP8.5), which were predicted to lead to radiative forcing levels of 2.6, 4.5, and 8.5 W/m
2 by the end of the 21st century, present the low, medium, and high greenhouse gas emissions, respectively [
39]. The CO
2 concentration reaches the maximum around 2052 and keeps at a high level from 2053 to 2055 under RCP2.6, while it increases continuously from 2006 to 2055 under RCP4.5 and RCP8.5. These three RCP scenarios are widely used to study future climate change and its ecological effects [
40,
41,
42].
Previous studies mainly focused on the historical effects of changes in climate and/or CO
2 concentration on one or two elements (NPP, ET, and WUE) at a large scale [
12,
17,
18,
25,
36]. How the fragile ecosystems in the arid area will respond to the future changes in climate and CO
2 concentration remains unclear. In this study, we take the arid area in Tianshan North Slope as our study area to explore the dynamics of NPP, ET, and WUE for different arid ecosystems under multiple RCP scenarios (RCP2.6, RCP4.5, and RCP8.5) in the near future. The Tianshan North Slope, one of the economic centers in the northwest China, is an important region in the Silk Road. It is an ideal region to explore the climate change and CO
2 fertilization effects for it contains various vegetation types, changing from the desert to the mountains, and is sensitive to the climate change [
43,
44]. Therefore, it is of particular importance to study the response of the arid ecosystem in Tianshan North Slope to the near future changes in climate and CO
2 concentration. In this study, we first analyzed the temporal and spatial patterns of the climate change in the Tianshan North Slope region under different RCP scenarios. Then, the spatial and temporal dynamics in NPP and ET of the region are detected. Finally, we focus on the NPP, ET, and WUE temporal dynamics of different vegetation types, and compare the different responses among different vegetation types. Thus, the forthcoming results should enhance our understanding of the climate change and CO
2 fertilization effects on the arid ecosystems and provide scientific guidance for the decision makers to take proper actions to maintain the sustainable development of the arid ecosystem in the future.
5. Conclusions
In this study, we explored the NPP and ET dynamics of the arid ecosystem in Tianshan North Slope under RCP2.6, RCP4.5, and RCP8.5 by using AEM. The climate in Tianshan North Slope will experience a wetter and warming trend from 2006 to 2055 under each RCP scenario. In response to the changes in climate and CO2 concentration, the regional mean annual NPP increases by a rate of 1.16, 1.62, and 2.15 g C m−2 year−1 under RCP2.6, RCP4.5, and RCP8.5, respectively. Similar to the regional mean annual NPP, the regional annual ET presents an increasing trend with the rate of 0.38 mm year−1 under RCP2.6, 0.43 mm year−1 under RCP4.5, and 0.52 mm year−1 under RCP8.5. Hence, for the entire region, the Tianshan North Slope will benefit from the changes during 2006–2055.
The spatial pattern of the climate change indicates that the temperature in the basin increases while it decreases in the mountains from 2006 to 2055, although the decreased area in the mountains region decreases from RCP2.6 to RCP8.5. The annual precipitation in the northeastern part of the study area presents a declining trend under RCP2.6 and RCP4.5, and the precipitation in the mountains increases much more than in the other regions. However, the spatial patterns of regional annual NPP difference under different RCP scenarios show that the increase of NPP mainly occurs in the central and western part where the CRP is abundant. The decline in regional annual NPP mainly occurs in the northeastern part under RCP2.6. The spatial patterns of regional ET difference indicate that the increase of ET mainly occurs in the central and western part, which is similar to those of the NPP difference. However, the ET decreases in the high mountains, which is mainly caused by the decline in temperature. The decline of ET in the northeast part is mainly induced by the decline in precipitation under the RCP2.6 scenario.
There is temporal heterogeneity in the climate patterns, in which the climate changes more rapidly in the first 30 years (2006–2035) than in the next 20 years (2036–2055) under each RCP scenario. Except the declining trend of precipitation in the next 20 years under RCP4.5, the increasing trends are found for precipitation and temperature in each study period under all three RCP scenarios. In response to the heterogeneity of the climate change, the increase rates of NPP and ET in the first 30 years are higher than those in the next 20 years under each RCP scenario. Moreover, the ET decreases in the next 20 year under RCP4.5 and RCP8.5, and the NPP decreases by a rate of 0.75 g C m−2 year−1 in the next 20 years under RCP4.5.
The sensitivity of different vegetation types is well analyzed in this study. We find that different vegetation types respond differently to the changes in climate and CO2 concentration under different RCP scenarios. All vegetation types in Tianshan North Slope experience increased NPP and ET with various rates under each RCP scenario. Under each RCP, the NPS, PS, and GRS are more sensitive to the changes in climate and CO2 concentration compared to the other vegetation types, while those of the TBF and CRP are less sensitive to the changes. The sensitivity of the different vegetation types varies during different study periods and under different RCP scenarios. However, the WUE of each vegetation type shows an increasing trend under all RCP scenarios. For each vegetation type, the increasing rate of WUE increases from RCP2.6 to RCP8.5, resulting from the warm and wet climate change and increasing CO2 concentration.