The Qinghai-Tibetan Plateau is the highest biogeographic unit on earth and widely regarded as its ‘third pole’. The high-altitude, frigid and arid alpine ecosystems that form the Plateau are extremely sensitive to climate change and human disturbance. Unsurprisingly, the Plateau is therefore a global epicenter of ecological and global change research and provides the ideal conditions and context to study the impacts of global change. Ecological research conducted on the Plateau can be partitioned into four developmental and chronological phases, beginning with the gathering of primitive knowledge and progressing towards a description of mechanistic processes. Throughout the course of Plateau research paradigm shifts from standalone surveys of biogeographic patterns to fixed monitoring and mechanism research; from isolated population, community and ecosystem approaches to more integrated, multidisciplinary research; and from pure theoretical research to an emphasis on effective resource utilization and sustainable development. Future ecological research will likely pay increasing attention to quantifying the impacts of climate warming and human activity on ecosystem change, and climate and ecosystem feedback processes. Multidisciplinary and comprehensive research should be strengthened amongst fields such as ecosystem ecology, physical geography, environmental science and remote sensing in order to support climate change adaptation and sustainable development in this fragile and unique region.

Elevation dependency amongst climate change signals has been found in major mountain ranges around the world, but the main factors causing this dependency have not been clarified. In this study, four different datasets of observation and reanalysis for China were used to examine the elevation dependency of climate change. A lack of consistency was found in dependency between warming magnitude and elevation across the Tibetan Plateau and China. However, a dependency of climate change on water vapor was detected whereby the temperature trend initially increased at low specific humidity, and then decreased as specific humidity increased. At ground level the maximum trend in temperature appeared in the specific humidity range 2.0–3.0 g kg^-1. This suggests that water vapor is a mediator of climate change and may be responsible for elevation-dependent climate change.

The Koshi River Basin is in the middle of the Himalayas, a tributary of the Ganges River and a very important cross-border watershed. Across the basin there are large changes in altitude, habitat complexity, ecosystem integrity, land cover diversity and regional difference and this area is sensitive to global climate change. Based on Landsat TM images, vegetation mapping, field investigations and 3S technology, we compiled high-precision land cover data for the Koshi River Basin and analyzed current land cover characteristics. We found that from source to downstream, land cover in the Koshi River Basin in 2010 was composed of water body (glacier), bare land, sparse vegetation, grassland, wetland, shrubland, forest, cropland, water body (river or lake) and built-up areas. Among them, grassland, forest, bare land and cropland are the main types, accounting for 25.83%, 21.19%, 19.31% and 15.09% of the basin's area respectively. The composition and structure of the Koshi River Basin land cover types are different between southern and northern slopes. The north slope is dominated by grassland, bare land and glacier; forest, bare land and glacier are mainly found on northern slopes. Northern slopes contain nearly seven times more grassland than southern slopes; while 97.13% of forest is located on southern slopes. Grassland area on northern slope is 6.67 times than on southern slope. The vertical distribution of major land cover types has obvious zonal characteristics. Land cover types from low to high altitudes are cropland, forest, Shrubland and mixed cropland, grassland, sparse vegetation, bare land and water bodies. These results provide a scientific basis for the study of land use and cover change in a critical region and will inform ecosystem protection, sustainability and management in this and other alpine transboundary basins.

Kobresia pygmaea Willd. dominates the alpine meadow ecosystem on the Qinghai-Tibet Plateau. Knowledge of this species' distribution and ecological environment could provide valuable insights into the alpine ecosystem and key species living there, support species and ecosystem conservation in alpine regions, and build on species origin and evolutionary research. To avoid modelling uncertainty encountered in a single approach, four species distribution model algorithms (Surface Range Envelope (SRE), Generalized Linear Model (GLM), Generalized Boosted Regression (GBM) and Maximum Entropy (MAXENT)), were used to simulate the distribution of K. pygmaea based on occurrence samples that were verified using DNA sequencing techniques. Species distribution modelling revealed a vast distribution region of K. pygmaea in the northern Tibetan Highlands and alpine meadows in the southern and eastern declivity of the plateau. A high evaluation performance was found for the GLM, GBM and MAXENT models. Different potential range size patterns for the four models were found between 374340–482605 km² (average = 421591 km²). Precipitation during growing seasons was found to be the dominant factor accounting for the distribution, consistent with patterns of heat and water patterns conditions of alpine ecosystems on the plateau. Species distribution models provide a simple and reliable approach to simulating the spatial patterns of species inhabiting the Qinghai-Tibet Plateau.

Inter-annual variability in total precipitation can lead to significant changes in carbon flux. In this study, we used the eddy covariance (EC) technique to measure the net CO₂ ecosystem exchange (NEE) of an alpine meadow in the northern Tibetan Plateau. In 2005 the meadow had precipitation of 489.9 mm and in 2006 precipitation of 241.1 mm, which, respectively, represent normal and dry years as compared to the mean annual precipitation of 476 mm. The EC measured NEE was 87.70 g C m^-2 yr^-1 in 2006 and –2.35 g C m^-2 yr^-1 in 2005. Therefore, the grassland was carbon neutral to the atmosphere in the normal year, while it was a carbon source in the dry year, indicating this ecosystem will become a CO₂ source if climate warming results in more drought conditions. The drought conditions in the dry year limited gross ecosystem CO₂ exchange (GEE), leaf area index (LAI) and the duration of ecosystem carbon uptake. During the peak of growing season the maximum daily rate of NEE and P_max and α were approximately 30%–50% of those of the normal year. GEE and NEE were strongly related to photosynthetically active radiation (PAR) on half-hourly scale, but this relationship was confounded by air temperature (T_a), soil water content (SWC) and vapor pressure deficit (VPD). The absolute values of NEE declined with higher T_a, higher VPD and lower SWC conditions. Beyond the appropriate range of PAR, high solar radiation exacerbated soil water conditions and thus reduced daytime NEE. Optimal T_a and VPD for maximum daytime NEE were 12.7°C and 0.42 KPa respectively, and the absolute values of NEE increased with SWC. Variation in LAI explained around 77% of the change in GEE and NEE. Variations in R_e were mainly controlled by soil temperature (T_s), whereas soil water content regulated the responses of R_e to T_s.

Accurate quantification of aboveground biomass of grasslands in alpine regions plays an important role in accurate quantification of global carbon cycling. The monthly normalized difference vegetation index (NDVI), enhanced vegetation index (EVI), mean air temperature (Ta), ≥5°C accumulated air temperature (AccT), total precipitation (TP), and the ratio of TP to AccT (TP/AccT) were used to model aboveground biomass (AGB) in grasslands on the Tibetan Plateau. Three stepwise multiple regression methods, including stepwise multiple regression of AGB with NDVI and EVI, stepwise multiple regression of AGB with Ta, AccT, TP and TP/AccT, and stepwise multiple regression of AGB with NDVI, EVI, Ta, AccT, TP and TP/AccT were compared. The mean absolute error (MAE) and root mean squared error (RMSE) values between estimated AGB by the NDVI and measured AGB were 31.05 g m^-2 and 44.12 g m^-2, and 95.43 g m^-2 and 131.58 g m^-2 in the meadow and steppe, respectively. The MAE and RMSE values between estimated AGB by the AccT and measured AGB were 33.61g m^-2 and 48.04 g m^-2 in the steppe, respectively. The MAE and RMSE values between estimated AGB by the vegetation index and climatic data and measured AGB were 28.09 g m^-2 and 42.71 g m^-2, and 35.86 g m^-2 and 47.94 g m^-2, in the meadow and steppe, respectively. The study finds that a combination of vegetation index and climatic data can improve the accuracy of estimates of AGB that are arrived at using the vegetation index or climatic data. The accuracy of estimates varied depending on the type of grassland.

Vegetation phenology is a sensitive indicator of global warming, especially on the Tibetan Plateau. However, whether climate warming has enhanced the advance of grassland phenology since 2000 remains debated and little is known about the warming effect on semiarid grassland phenology and interactions with early growing season precipitation. In this study, we extracted phenological changes from average NDVI in the growing season (GNDVI) to analyze the relationship between changes in NDVI, phenology and climate in the Northern Tibetan Damxung grassland from 2000 to 2014. The GNDVI of the grassland declined. Interannual variation of GNDVI was mainly affected by mean temperature from late May to July and precipitation from April to August. The length of the growing season was significantly shortened due to a delay in the beginning of the growing season and no advancement of the end of the growing season, largely caused by climate warming and enhanced by decreasing precipitation in spring. Water availability was the major determinant of grass growth in the study area. Warming increased demand for water when the growth limitation of temperature to grass was exceeded in the growing season. Decreased precipitation likely further exacerbated the effect of warming on vegetation phenology in recent decades due to increasing evapotranspiration and water limitations. The comprehensive effects of global warming and decreasing precipitation may delay the phenological responses of semiarid alpine grasslands.

Alpine swamp meadows on the Tibetan Plateau, with the highest soil organic carbon content across the globe, are extremely vulnerable to climate change. To accurately and continually quantify the gross primary production (GPP) is critical for understanding the dynamics of carbon cycles from site-scale to global scale. Eddy covariance technique (EC) provides the best approach to measure the site-specific carbon flux, while satellite-based models can estimate GPP from local, small scale sites to regional and global scales. However, the suitability of most satellite-based models for alpine swamp meadow is unknown. Here we tested the performance of four widely-used models, the MOD17 algorithm (MOD), the vegetation photosynthesis model (VPM), the photosynthetic capacity model (PCM), and the alpine vegetation model (AVM), in providing GPP estimations for a typical alpine swamp meadow as compared to the GPP estimations provided by EC-derived GPP. Our results indicated that all these models provided good descriptions of the intra-annual GPP patterns (R²>0.89, P<0.0001), but hardly agreed with the inter-annual GPP trends. VPM strongly underestimated the GPP of alpine swamp meadow, only accounting for 54.0% of GPP_EC. However, the other three satellite-based GPP models could serve as alternative tools for tower-based GPP observation. GPP estimated from AVM captured 94.5% of daily GPP_EC with the lowest average RMSE of 1.47 g C m^-2. PCM slightly overestimated GPP by 12.0% while MODR slightly underestimated by 8.1% GPP compared to the daily GPP_EC. Our results suggested that GPP estimations for this alpine swamp meadow using AVM were superior to GPP estimations using the other relatively complex models.

Climate warming and economic developments have created pressures on the ecological systems that human populations rely on, and this process has contributed to the degradation of ecosystems and the loss of ecosystem services. In this study, Landsat satellite data were chosen as the data source and the Koshi River Basin (KB) in the central high Himalayas as the study area. Changes in land cover and changes in the value of ecosystem services between 1990 and 2010 were analyzed and the land cover pattern of the KB in 2030 and 2050 was modeled using the CA-Markov model. Changes in land cover and in the value of ecosystem services in the KB for the period 2010–2050 were then analyzed. The value of ecosystem services in the KB was found to decrease by 2.05×10⁸ USD y^-1 between 1990 and 2010. Among these results, the services value of forest, snow/glacier and barren area decreased, while that of cropland increased. From 1990 to 2050, forest showed the largest reduction in ecosystem services value, as much as 11.87×10⁸ USD y^-1, while cropland showed the greatest increase, by 3.05×10⁸ USD y^-1. Deforestation and reclamation in Nepal contributed to a reduction in the value of ecosystem services in the KB. Barren areas that were transformed into water bodies brought about an increase in ecosystem services value in the lower reaches of the Koshi River. In general, this process is likely to be related to increasing human activity in the KB.

Because of its landscape heterogeneity, Koshi Basin (KB) is home to one of the world's most abundant, diverse group of species. Habitat change evaluations for key protected species are very important for biodiversity protection in this region. Based on current and future world climate and land cover data, MaxEnt model was used to simulate potential habitat changes for key protected species. The results shows that the overall accuracy of the model is high (AUC > 0.9), suggesting that the MaxEnt-derived distributions are a close approximation of real-world distribution probabilities. The valley around Chentang Town and Dram Town in China, and Lamabagar and the northern part of Landtang National Park in Nepal are the most important regions for the protection of the habitat in KB. The habitat area of Grus nigricollis, Panax pseudoginseng, and Presbytis entellus is expected to decrease in future climate and land cover scenarios. More focus should be placed on protecting forests and wetlands since these are the main habitats for these species.

Climate change and forage-intake are important components of livestock population systems, but our knowledge about the effects of changes in these properties on livestock is limited, particularly on the Northern Tibetan Plateau. Based on corresponding independent models (CASA and TEM), a human-induced NPP (NPP_H) value and forage-intake threshold were obtained to determine their influences on livestock population fluctuation and regrowth on the plateau. The intake threshold value provided compatible results with livestock population performance. If the forage-intake was greater than the critical value of 1.9 (kg DM d^-1 sheep^-1), the livestock population increased; otherwise, the livestock population decreased. It takes four years to transfer a disturbance in primary productivity to the next trophic level. The relationships between livestock population and NPPH value determined population dynamics via the forage-intake value threshold. Improved knowledge on lag effects will advance our understanding of drivers of climatic changes on livestock population dynamics.

Xainza County on the Qiangtang Plateau (QP) spans a transitionally climatic and eco-environmental zone and is therefore a very sensitive region. Glaciers in this area are one of the most important kinds of land cover as well as key water resources, yet our understanding of their current status and spatio-temporal change remains limited. Using Landsat images, this study investigated the current distribution of glaciers (2015) in the Xainza Xiegang Mountains as well as the spatio-temporal changes that took place over six time periods between 1976 and 2015. Results show that, in 2015, 131 glaciers covered a total area of 74.59 ± 5.25 km², mainly located between 5,600 and 6,000 m above sea level (a.s.l). Between 1976 and 2015, the total number of glaciers increased by 12, while their areas decreased by 24.98% (24.83 km²). Glacial retreat has induced a loss of water resources of 11.77 × 108 m3 over the last 39 years, while spatial heterogeneities in glacial changes across various sub-basins, aspects, and altitudinal zones are also clearly observed. Climate warming is the key factor driving this continuous glacial retreat; the high-quality dataset presented in this paper for the Xainza Xiegang study area is crucial for the ongoing assessment of climatic and eco-environmental changes.

In order to understand whether or not the response of vegetation indices and biomass production to warming varies with warming magnitude, an experiment of field warming at two magnitudes was conducted in an alpine meadow on the northern Tibetan Plateau beginning in late June, 2013. The normalized difference vegetation index (NDVI), green normalized difference vegetation index (GNDVI) and soil adjusted vegetation index (SAVI) data were obtained using a Tetracam Agricultural Digital Camera in 2013–2014. The gross primary production (GPP) and aboveground plant biomass (AGB) were modeled using the surface measured NDVI and climatic data during the growing seasons (i.e. June–September) in 2013–2014. Both low and high warming significantly increased air temperature by 1.54 and 4.00°C, respectively, and significantly increased vapor pressure deficit by 0.13 and 0.31 kPa, respectively, in 2013–2014. There were no significant differences of GNDVI, AGB and ANPP among the three warming treatments. The high warming significantly reduced average NDVI by 23.3% (–0.06), while the low warming did not affect average NDVI. The low and high warming significantly decreased average SAVI by 19.0% (–0.04) and 27.4% (–0.05), respectively, and average GPP by 24.2% (i.e. 0.21 g C m^-2 d^-1) and 44.0% (i.e. 0.39 g C m^-2 d^-1), respectively. However, the differences of the average NDVI, SAVI, and GPP between low and high warming were negligible. Our findings suggest that a greater drying may dampen the effect of a higher warming on vegetation indices and biomass production in alpine meadow on the northern Tibetan Plateau.