This is the accepted version of the journal article: Shen, Xiangjin; Shen, Miaogen; Wu, Chaoyang; [et al.]. «Critical role of water conditions in the responses of autumn phenology of marsh wetlands to climate change on the Tibetan Plateau». Global change biology, Vol. 30, issue 1 (Jan. 2023), art. e17097. DOI 10.1111/gcb.17097 This version is available at https://ddd.uab.cat/record/287436 under the terms of the license 1 Critical role of water conditions in the responses of autumn 2 phenology of marsh wetlands to climate change on the Tibetan 3 Plateau 4 Xiangjin Shen1, Miaogen Shen2, Chaoyang Wu3, Josep Peñuelas4,5, Philippe Ciais6, Jiaqi 5 Zhang1, Chris Freeman7, Paul I. Palmer8,9, Binhui Liu10, Mark Henderson11, Zhaoliang 6 Song12, Shaobo Sun12, Xianguo Lu1, Ming Jiang1 1 7 Northeast Institute of Geography and Agroecology, Chinese Academy of Sciences, 8 Changchun, China. 2 9 State Key Laboratory of Earth Surface Processes and Resource Ecology, Faculty of 10 Geographical Science, Beijing Normal University, Beijing, China. 3 11 Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of 12 Sciences, Beijing, China. 4 13 CREAF, Cerdanyola del Vallès, Barcelona, Spain. 5 14 CSIC, Global Ecology Unit CREAF-CSIC- UAB, Barcelona, Spain. 6 15 Laboratoire des Sciences du Climat et de l'Environnement, LSCE/IPSL, CEA-CNRS- 16 UVSQ, Université Paris-Saclay, Gif-sur-Yvette, France. 7 17 School of Natural Sciences, Bangor University, Bangor LL57 2UW, UK. 8 18 School of GeoSciences, University of Edinburgh, Edinburgh, UK. 9 19 National Centre for Earth Observation, University of Edinburgh, Edinburgh, UK. 10 20 College of Forestry, Northeast Forestry University, Harbin, China. 11 21 Public Policy Program and Environmental Studies Program, Mills College, Oakland, 22 California, USA. 12 23 Institute of Surface-Earth System Science, School of Earth System Science, Tianjin 24 University, Tianjin, China. 1 25 26 27 Abstract: 28 The Tibetan Plateau, housing 20% of China's wetlands, plays a vital role in the regional 29 carbon cycle. Examining the phenological dynamics of wetland vegetation in response to 30 climate change is crucial for understanding its impact on the ecosystem. Despite this 31 importance, the specific effects of climate change on wetland vegetation phenology in 32 this region remain uncertain. In this study, we investigated the influence of climate change 33 on the end of the growing season (EOS) of marsh wetland vegetation across the Tibetan 34 Plateau. Utilizing satellite-derived Normalized Difference Vegetation Index (NDVI) data 35 and observational climate data, we observed a significant delay in the regionally averaged 36 EOS of marsh vegetation by 4.10 days/decade from 2001 to 2020. Warming preseason 37 temperatures were found to be the primary driver behind the delay in the EOS of marsh 38 vegetation, whereas preseason cumulative precipitation showed no significant impact. 39 Interestingly, the responses of EOS to climate change varied spatially across the plateau, 40 indicating a regulatory role for hydrological conditions in marsh phenology. In the humid 41 and cold central regions, preseason daytime warming significantly delayed the EOS. 42 However, areas with lower soil moisture exhibited a weaker or reversed delay effect, 43 suggesting complex interplays between temperature, soil moisture, and EOS. Notably, in 44 the arid southwestern regions of the plateau, increased preseason rainfall directly delayed 45 the EOS, while higher daytime temperatures advanced it. Our results emphasize the 2 46 critical role of hydrological conditions, specifically soil moisture, in shaping marsh EOS 47 responses in different regions. Our findings underscore the need to incorporate 48 hydrological factors into terrestrial ecosystem models, particularly in cold and dry regions, 49 for accurate predictions of marsh vegetation phenological responses to climate change. 50 This understanding is vital for informed conservation and management strategies in the 51 face of current and future climate challenges. 52 53 KEYWORDS Marsh vegetation, autumn phenology, climate change, water condition, 54 Tibetan Plateau 55 56 1 Introduction 57 Vegetation phenology refers to the seasonal timing of life cycle events in plants and 58 reflects the dynamic responses of terrestrial ecosystems to global climate change (Chen 59 et al., 2017; Peñuelas et al., 2001; Piao et al., 2007, 2008, 2019; Richardson et al., 2013; 60 Wu et al., 2022). Several studies have demonstrated that autumn phenology, signaling the 61 end of the growing season (EOS), reflects the vegetation’s growing period better than 62 spring phenology and has a significant effect on carbon sequestration in terrestrial 63 ecosystems (Bao et al., 2020; Fu et al., 2018; Garonna et al., 2014; Zhu et al., 2012). 64 Global climate change has significantly changed the autumn phenology worldwide, 65 affecting the regional and global energy balance, water flux, and carbon budget (Che et 66 al., 2014; Estiarte et al., 2015; Kelsey et al., 2021; Richardson et al., 2013; Yang et al., 3 67 2021). Although the variations in the autumn phenology of vegetation and its response to 68 regional and global climate changes have been extensively analyzed, most studies have 69 focused on grassland or forest ecosystems, with few investigations of wetland ecosystems 70 (Coleman et al., 2022; Ge et al., 2015; Ma et al., 2022; Rice et al., 2018; Yang et al., 2015). 71 Due to the unique environmental conditions in wetlands, climate change may have 72 different effects on the autumn phenology of vegetation compared to other ecosystems 73 (Keppeler et al., 2021; Ma et al., 2022; Molino et al., 2022; Shen et al., 2022). Such 74 differences must be considered if we are to better understand the responses of the global 75 carbon cycle and ecosystem vegetation to climatic variation in the context of global 76 climate change. 77 As the highest plateau in the world, the Tibetan Plateau is highly sensitive to climate 78 change. And yet the roles of temperature or precipitation in determining the vegetation 79 phenology of the Tibetan Plateau appear contradictory; Some studies assert that 80 temperature plays a dominant role in determining vegetation phenology (Yu et al., 2010), 81 while others argue that precipitation is critical to the vegetation phenology of the region 82 (Shen et al., 2014, 2015a). Known as the “water tower of Asia”, the Tibetan Plateau 83 features a large area of marsh wetlands with relatively high water content (Che et al., 84 2014; Shen et al., 2011; Shen et al., 2021a) that provide an ideal opportunity for clarifying 85 the dominant effects of temperature and precipitation on the phenology of the vegetation 86 of the Tibetan Plateau. 87 Recent studies have analyzed the effects of climate change on autumn phenology in 4 88 various ecosystems of the region (Shen et al., 2022b), reported the phenology of grassland 89 vegetation as positively correlated with precipitation and negatively correlated with 90 daytime maximum temperature. Increased precipitation can enhance the water use 91 efficiency of grassland vegetation, potentially delaying the EOS (Shen et al., 2015b; Wu 92 et al., 2018). However, an increase in daytime maximum temperature also promotes 93 evaporation, reducing water use efficiency and consequently advancing the EOS. Clearly, 94 an improved understanding of the influence of climate variation on the autumn phenology 95 of marsh vegetation in this region can greatly improve our knowledge of the relationships 96 between the vegetation of ecosystems and climate change. 97 This study utilizes data from 2001 to 2020, incorporating the Normalized Difference 98 Vegetation Index (NDVI) and observational climate data. The aim is to explore 99 spatiotemporal variations in the end of the growing season (EOS) and vegetation 100 responses to climatic variations in the marshes of the Tibetan Plateau. The objective is to 101 enhance our understanding and predictive capabilities regarding phenological changes in 102 marsh vegetation. By illuminating the intricate relationship between vegetation and 103 climate change in this ecologically significant area, our study's findings can offer valuable 104 new insights for ecological management and conservation efforts. 105 106 2 Material and method 107 2.1 Study region 108 The Tibetan Plateau is located in southwestern China at an altitude of 3,000–5,000 m 5 109 (average altitude > 4,000 m) and is characterized by a semi-arid and cold climate (Figure 110 1) (Shen et al., 2022). The annual precipitation exceeds 1,000mm in the southeastern 111 region and is < 100mm in the northwestern region (Cheng et al., 2021; Gao et al., 2013; 112 Piao et al., 2006a). The average temperature in the northwestern and southeastern areas 113 is approximately −6°C and 20°C, respectively (Qin et al., 2022). 114 6 115 FIGURE 1 Spatiotemporal change of the end of the growing season (EOS) in the marshes 116 of the Tibetan Plateau from 2001 to 2020. (a) Distribution of marshes at different altitudes 117 on the Tibetan Plateau. (b) Spatial patterns of regionally averaged EOS. (c) Temporal 118 trends in EOS. (d) Temporal variations of regionally averaged EOS. The inset histograms 119 at the bottom of (b) and (c) describe the frequency distributions of the average EOS and 120 EOS trend. - and + in C show the negative (advancing) and positive (delaying) trend, 121 respectively; * and ** indicate the trend is significant (P < 0.05) and extremely significant 122 (P < 0.01), respectively. The error bar and bold black line in D show standard error and 123 linear trend of the regionally averaged EOS, respectively. 124 As the highest plateau in the world, the Tibetan Plateau is extremely sensitive to 125 climate change (Dong et al., 2012; Zhang et al., 2013). Changes in vegetation phenology 126 in this region serve as crucial indicators of global climate change (Chen et al., 2015). 127 Large wetland areas characterized by marshes are distributed on the Tibetan Plateau and 128 are important for the ecological security of the region and the major river systems that 129 originate there, including the Huanghe (Yellow), Changjiang (Yangzi), and Brahmaputra 130 rivers. The main species of marsh plants distributed on the Tibetan Plateau are Phragmites 131 australis, Blysmus sinocompressus, Carex pseudosupina, and Kobresia littledalei (Shen 132 et al., 2021a). 133 134 2.2 Data 135 Satellite-derived NDVI data covering the 2001−2020 period were obtained from the 7 136 MOD13Q1 NDVI dataset, with temporal and spatial resolutions of 16 d and 250 m, 137 respectively (Shen et al., 2023). This dataset was provided by the Earth Science Data 138 Systems of the National Aeronautics and Space Administration. 139 The distribution of marshes in the study area was obtained from the wetland 140 distribution datasets for China for years 2000 and 2015 at a resolution of 30 m × 30 m 141 (Mao et al., 2020). These digital maps were available from the National Earth System 142 Science Data Center. The accuracy of datasets had been verified through field 143 observations, and the producer's and user’s accuracies were over 95% and 98%, 144 respectively (Mao et al., 2020). We used the marsh distribution data for two specific years 145 (2000 and 2015) to extract the unchanged marsh distribution (Shen et al., 2023). 146 The soil moisture data used in this study were extracted from a 1-km daily soil 147 moisture dataset of in situ measurements conducted in China from 2000 to 2020 (Li et al., 148 2022). These dataset was produced using spatially dense in situ observations and machine 149 learning, and was obtained from the National Tibetan Plateau Scientific Data Center. 150 The Climate Change Research Center of the Chinese Academy of Sciences provided 151 the daily gridded climate data from more than 2400 meteorological stations distributed 152 across China. These included daily precipitation as well as the minimum, maximum, and 153 mean temperatures with a spatial resolution of 1 km from 2001 to 2020. Marsh 154 distribution, soil moisture, and climatic data were resampled at a resolution of 250 m × 155 250 m to maintain consistency with the spatial resolution of the NDVI data (Shen et al., 156 2021b). 8 157 158 2.3 Methods 159 Considering that snow cover decreases the NDVI value, consequently affecting the 160 accuracy of satellite-derived phenology data, we replaced the snow-contaminated NDVI 161 values with the median value of the uncontaminated winter NDVI values between 162 November and the following March for each pixel (Shen et al., 2014, 2015). This 163 preprocessing of data has been validated and included in numerous previous studies 164 (Ganguly et al., 2010; Shen et al., 2015a; Wang et al., 2021). Based on previous research 165 (Cong et al., 2013; Liu et al., 2016; Ma et al., 2021; Piao et al., 2006b; Shen et al., 2016; 166 Su et al., 2022), we used the poly fit-maximum approach to extract phenological 167 information (Piao et al., 2006b). In this method, the growth curve of the growing season 168 is fitted with a daily temporal resolution in order to remove abnormal NDVI values (Piao 169 et al., 2006b, 2011), and the EOS date is set to correspond to the time of the largest 170 decrease in NDVI at the end of the growth period. The polyfit-maximum method has been 171 widely used to extract vegetation phenology owing to its excellent performance (Fu et al., 172 2014; Ma et al., 2022; Piao et al., 2006b, 2011; Shen et al., 2018, 2019, 2023; Su et al., 173 2022) and consists in a number of steps. 174 First, we calculated the annual and multiyear average rates of NDVI variation to 175 obtain the corresponding day of the year (DOY) for the vegetation’s EOS (Piao et al., 176 2006b). The following equation was used to calculate the rate of NDVI variation: NDVI(t+1)−NDVI(t) 177 NDVIratio (t) = (1) NDVI(t) 9 178 where t is time (temporal resolution of 16 days), NDVIratio(t) is the rate of NDVI 179 change corresponding to period t, NDVI(t) is the NDVI value for period t, and NDVI(t+1) 180 is the NDVI value for period t+1.We detected the time t with the minimum NDVIratio 181 and used the corresponding NDVI(t+1) at time (t+1) as the NDVI threshold for the EOS. 182 Then, we used the maximum value method of multivariate fitting to construct a 183 unary sixth-degree polynomial function (Piao et al., 2006b) and fitted the annual and 184 multiyear average daily NDVI fitting curve by pixel. The formula used was: 185 NDVI = a + a1 x1 + a2 x 2 + a3 x 3 + a4 x 4 + a5 x 5 + a6 x 6 (2) 186 where x is the day of each year (DOY); and a1, a2, a3, ... a6 are the regression 187 coefficients determined by least-squares regression. 188 Finally, we substituted the DOY into the fitting curve of the multiyear average daily 189 NDVI to obtain the NDVI threshold corresponding to the EOS (Piao et al., 2006b). We 190 applied the NDVI threshold values to the daily NDVI fitting curve of each year to obtain 191 the corresponding EOS values for marsh vegetation in that year. 192 To analyze the EOS trend and climatic variables on the Tibetan Plateau from 2001 193 to 2020, we performed a linear regression analysis using the following equation (Piao et 194 al., 2011): (𝑛×∑𝑛 𝑛 𝑛 𝑖=1 𝑖×𝑥𝑖 )−(∑𝑖=1 𝑖 ∑𝑖=1 𝑥𝑖 ) 195 𝜃𝑠𝑙𝑜𝑝𝑒 = 2 (3) 𝑛×∑𝑛 2 𝑛 𝑖=1 𝑖 −(∑𝑖=1 𝑖) 196 where n is the number of years analyzed (i.e., 20 years for this study); θslope indicates 197 the trend (of the EOS or climatic variables) for each pixel; and xi is the climate variable 198 (or EOS) during the i year. A negative θslope implies that the temporal variation shows an 10 199 advancing (i.e., decreasing) trend, whereas a positive θslope implies a delaying (i.e., 200 increasing) trend. 201 To investigate the seasonal changes in the EOS in response to climate variations, we 202 analyzed the simple correlation coefficients between monthly and seasonal climate 203 factors and the EOS in previous winter (November to February of the following year), 204 spring (March to May), summer (June to August), and autumn (September to October). 205 In addition, we carried out a partial correlation analysis to further examine correlations 206 between climate variables and the EOS. Through this analysis, it was possible to 207 determine the relationship between two parameters after removing the influence of other 208 factors (Peng et al., 2013; Shen et al., 2016). 209 The partial correlation coefficient between the time series of the EOS and daytime 210 maximum temperature (or nighttime minimum temperature) was calculated to assess the 211 effect of maximum temperature (or minimum temperature) on the EOS, with precipitation 212 and minimum temperature (or maximum temperature) as the controlling variables. In line 213 with previous studies, the duration of the preseason was calculated for the maximum 214 temperature (or minimum temperature) based on the period preceding the long-term 215 average date of the EOS. When the maximum temperature (or minimum temperature) had 216 the highest absolute value for the partial correlation coefficient with the EOS, this period 217 is referred to as the preseason for the maximum temperature (Shen et al., 2016; Wu et al., 218 2018) (or the minimum temperature). In this study, an interval of 10 days was adopted to 219 determine the duration of the preseason period and smooth out potential extreme values 11 220 (Shen et al., 2015a). 221 The effect of preseason cumulative precipitation on the EOS was similarly analyzed, 222 and preseason precipitation was determined by setting the maximum and minimum 223 temperatures as the controlling variables (Shen et al., 2016). We did not constrain the 224 duration of preseason precipitation to be equal to that of preseason temperature. In 225 addition, to further analyze the influences of climatic variations on the EOS, we compared 226 the partial correlation coefficients between the EOS and preseason precipitation and 227 between the EOS and maximum and minimum temperatures at different levels of soil 228 moisture. 229 230 3 Results 231 3.1 Spatial and temporal variations in the EOS 232 The multiyear mean EOS on the Tibetan Plateau occurred primarily between 260th and 233 300th day of year (DOY), with a regional average of 277th DOY (October 4, or October 234 3 in leap years) (Figure 1b). The EOS was later in the low altitude (below 4,000 m) humid 235 areas of the eastern Tibetan Plateau and earlier in the high-altitude (above 4,000 m) 236 central areas (Figure 1a,b). 237 The regionally averaged EOS from 2001 to 2020 across the Tibetan Plateau 238 exhibited a significant delay of 4.10 days per decade (P < 0.05) (Figure 1d). The 239 percentage of pixels showing a trend of delayed EOS (68.5%, with a significant 240 proportion of 36.2%, P < 0.05) was higher than that showing a trend of advanced EOS 12 241 (31.5%, with a significant proportion of 6.4%, P < 0.05). The former trend was more 242 evident in the northern and western (high altitude permafrost areas) of the Tibetan Plateau, 243 while the latter was concentrated in the central regions (Figure 1a,c). 244 245 3.2 Relationships between the EOS and climate variables 246 We first analyzed partial correlations between seasonal climate variables and the EOS 247 without using pre-seasons. In autumn, the EOS showed a significant positive correlation 248 with temperatures (P < 0.05) but a weak negative correlation with precipitation (P > 0.05) 249 across the Tibetan Plateau (Figure S1). The partial correlations between EOS and climatic 250 variables in other seasons were not statistically significant (P > 0.05). 251 We further analyzed the partial correlations between EOS and preseason climatic 252 variables. Across the Tibetan Plateau, the regionally averaged EOS showed a weak 253 negative correlation with preseason cumulative precipitation but a significant positive 254 correlation with preseason maximum and minimum temperatures (Figure 2). Spatially, 255 the proportion of pixels displaying a negative correlation between EOS and preseason 256 cumulative precipitation was approximately 52.1%, and that with a significant negative 257 relationship was approximately 7.6% (Figure 3a). The proportions of pixels showing a 258 positive relationship between EOS and preseason maximum and minimum temperatures 259 were approximately 65.2% and 89.4%, respectively, and the proportions of pixels with a 260 significant (P < 0.05) positive relationship were 19.5% and 34.0%, respectively (Figure 261 3c,e). 13 262 263 FIGURE 2 Partial correlation coefficients between preseason climatic factors and EOS 264 of marsh vegetation on the Tibetan Plateau. *P < 0.05; **P < 0.01. Correlations lacking 265 an asterisk are non-significant (P > 0.05). 14 266 267 FIGURE 3 Relationship between EOS and preseason climate variables for the marshes 268 on the Tibetan Plateau from 2001 to 2020. Spatial patterns of partial correlation 269 coefficients between EOS and preseason cumulative precipitation (a), maximum 15 270 temperature (c), and minimum temperature (e). The changes in partial correlation 271 coefficients between EOS and preseason cumulative precipitation (b), maximum 272 temperature (d), and minimum temperature (f) along the spatial gradient of long-term 273 preseason soil moisture on the Tibetan Plateau from 2001 to 2020. The inset histograms 274 at the bottom of left figures (a, c, e) display the frequency distributions of partial 275 correlation coefficients. - and + show the negative and positive correlation, respectively; 276 * and ** indicate the correlation is significant (P < 0.05) and extremely significant (P < 277 0.01), respectively. The body lines in the right figures (b, d, f) indicate the linear fit for 278 the partial correlation coefficients, and the shading represents the 95% confidence band 279 of the fits. 280 The partial correlations between EOS and preseason climatic variables indicated 281 spatial heterogeneity across the Tibetan Plateau. In the southwestern region, the EOS 282 showed a positive and negative partial correlation with preseason cumulative 283 precipitation and preseason maximum temperature, respectively; however, in the central 284 area, it exhibited a significant negative and positive partial correlation with these two 285 parameters, respectively (Figure 3a,c). The partial relationships between EOS and 286 preseason minimum temperature were positive in most of the study area, and the positive 287 relationships were extremely significant (P < 0.01) in the western and eastern regions 288 (Figure 3e). 289 Subsequently, we examined the partial correlations between EOS and preseason 290 climatic variables under different soil moisture levels in the study region (Figure 3). The 16 291 results showed that as the preseason soil moisture increased, the positive relationships 292 between EOS and preseason cumulative precipitation gradually weakened, whereas those 293 between EOS and preseason maximum and minimum temperatures became gradually 294 stronger (Figures 3 and 4). 295 296 FIGURE 4 Impact of climate change on marsh EOS in different regions of the marshes 297 on the Tibetan Plateau. (a) Conceptual diagrams showing the effects of climate change on 298 the EOS. (b) Spatial distribution of long-term average preseason soil moisture (m3/m3). 299 (c) The long-term average preseason temperature (°C) for marsh vegetation on the Tibetan 300 Plateau. (d) long-term average preseason soil moisture (m3/m3) and preseason 301 temperature (°C) in different regions of the Tibetan Plateau. “+” and “−” indicate that the 17 302 climatic variable had a significant positive and negative effect on the EOS, respectively 303 (P < 0.05). “++” indicates an extremely significant positive effect (P < 0.01). In the 304 southwestern Tibetan Plateau (circled in blue), increased preseason precipitation can 305 significantly delay the EOS, while a higher preseason maximum temperature will advance 306 it. In contrast, in the central Tibetan Plateau (circled in red), a higher preseason maximum 307 temperature significantly delayed the EOS, while an increased preseason precipitation 308 advanced it. In the eastern Tibetan Plateau (circled in green), a higher preseason minimum 309 temperature significantly delayed the EOS. 310 311 4 Discussions 312 4.1 Autumn phenology of marsh vegetation on the Tibetan Plateau 313 The long-term average EOS for marsh vegetation occurred later in the eastern Tibetan 314 Plateau than in the central and southwestern regions (Figure 1b). This aligns with 315 observations that the eastern area has a lower altitude and a warmer climate (Shen et al., 316 2021b), thus allowing the growing season to continue for longer. This result is consistent 317 with the findings of (Liu et al., 2021), which showed that the combined EOS for all 318 vegetation types occurred earlier in the central Tibetan Plateau and later in the eastern 319 regions. From 2001 to 2020, the average EOS was delayed by 4.10 days per decade across 320 the plateau (Figure 1c), a trend consistent with the results of Shen et al. (Shen et al., 2022), 321 which showed a delay of 8.2 days in the EOS for vegetation on the plateau over the same 322 period. 18 323 4.2 Climatic effects on the regionally averaged EOS for marsh vegetation 324 The regionally averaged EOS exhibited significant positive partial correlations with 325 preseason daytime maximum and nighttime minimum temperatures but a weak negative 326 correlation with preseason cumulative precipitation across the Tibetan Plateau. The 327 phenology of marsh vegetation on the plateau observed in this study differed from that of 328 grasslands reported in previous studies (Dorji et al., 2013; Shen et al., 2022; Yang et al., 329 2021). 330 Previous studies have shown that in the grassland vegetation of the Tibetan Plateau, 331 increased precipitation significantly enhances water use efficiency, delaying the EOS. In 332 these relatively arid systems, the higher maximum temperatures increase evaporation and 333 reduce water use efficiency, thus advancing the EOS (Dorji et al., 2013; Yang et al., 2021). 334 In contrast, the marshes examined in the present study contained far more water 335 (Ganjurjav et al., 2022; Shen et al., 2022), making it less likely that preseason 336 precipitation would affect the regionally averaged EOS for their vegetation, and explain 337 how the abundance of water allowed the EOS to be delayed by the increased preseason 338 temperatures on the Tibetan Plateau. 339 Our results indicate that preseason temperature is a key factor affecting the EOS, and 340 an increase in this parameter significantly delayed the EOS on the Tibetan Plateau (Figure 341 3). It is known that increased preseason maximum temperature can promote 342 photosynthesis by enhancing the daytime photosynthetic activity of enzymes (Piao et al., 343 2007; Turnbull et al., 2022). This increased photosynthesis, together with raised nighttime 19 344 minimum temperatures that reduce frost and low-temperature constraints (Shen et al., 345 2016), would further delay the EOS on the plateau. 346 347 4.3 Elucidating spatial variations in the effects of climatic change on the EOS 348 We identified 3 characteristic areas within the Tibetan Plateau, which exhibited differing 349 primary drivers of the changes in EOS; south western, central and eastern. 350 In the relatively arid areas of the southwestern Tibetan Plateau, the EOS exhibited a 351 significant positive partial correlation with preseason cumulative precipitation but a 352 significant negative correlation with preseason maximum temperature (Figure 3a, c). This 353 indicated that, in this region, higher preseason precipitation significantly delayed the EOS 354 although this was constrained by preseason maximum temperatures tending to 355 significantly advance the EOS. As the climate in this southwestern region is dry and the 356 preseason soil moisture is low (Figure 4), an increase in preseason precipitation can 357 alleviate water stress, enhancing water use efficiency (Liu et al., 2016; Munné-Bosch et 358 al., 2004) and delaying the EOS. At the same time, an increase in maximum temperature 359 would increase hydrological losses through evaporation and reduce the amount of 360 available water (Kelsey et al., 2021; Shen et al., 2016), inhibiting the growth of marsh 361 vegetation. These dynamics fit well with our results, indicating that the EOS was 362 primarily affected by precipitation in the southwestern arid regions of the study area. For 363 the first time, our results indicate that, even in marsh ecosystems with their relatively high 364 water contents, available water may be insufficient for vegetation growth in the drier 20 365 regions of the Tibetan Plateau. 366 In the central Tibetan Plateau, a higher preseason temperature delayed the EOS, 367 while increased preseason precipitation advanced it (Figure 3). In the humid and cold 368 areas of the central Tibetan Plateau, soil moisture and temperature are higher and lower, 369 respectively, than those in the southwestern region (Cong et al., 2017) (Figure 4). In light 370 of these conditions, our results indicate that water was not the key factor affecting the 371 growth of marsh vegetation in the central Plateau, although temperature remained a 372 limiting factor. 373 An increase in maximum temperature can decelerate the process of chlorophyll 374 degradation (Shi et al., 2017) and retard the progression of leaf senescence (Estiarte et al., 375 2017). In addition, high preseason nighttime temperatures reduce the occurrence of frost 376 damage (Shen et al., 2022). By calculating the number of the frost days, we confirmed 377 that warming preseason nighttime temperatures had the most notable negative effect on 378 the frost days in the central Tibetan Plateau, the coldest region of the study area (Figure 379 S2). This could account for the role of increasing minimum temperature in delaying the 380 EOS in the central Tibetan Plateau. In contrast, increased preseason precipitation could 381 retard the growth of marsh vegetation due to the accompanying cooling effect, which 382 would advance the EOS in the already cold and humid areas of the central Tibetan Plateau. 383 In the low altitude humid regions in the east and high-altitude cold permafrost regions 384 in the west, the EOS showed a significant positive correlation with preseason minimum 385 temperature (Figure 3), indicating that an increase in this parameter delayed the EOS in 21 386 the marshes distributed in these regions of the Tibetan Plateau. However, the mechanisms 387 through which minimum temperature affects the EOS may differ between these two areas. 388 On one hand, increased nighttime temperatures cause a greater loss of organic matter due 389 to enhanced respiration, but on the other hand it can also stimulate accumulation of more 390 organic matter via the overcompensation effect (Shen et al., 2021b, 2022), a phenomenon 391 through which the vegetation recovers and exceeds its original state by promoting 392 photosynthesis the day after the enhanced respiration (Peng et al., 2013). It has been 393 reported that the occurrence of this effect is favored by the presence of sufficient water 394 and nutrients (Peng et al., 2013; Shen et al., 2022). 395 In cold high-altitude regions, the physiological processes causing vegetation 396 senescence are typically determined by low temperatures during cold nights (Tang et al., 397 2016). In the high-altitude cold permafrost regions of the western Tibetan Plateau, 398 nighttime minimum temperature is generally low (Nan et al., 2005), and an increase in 399 preseason minimum temperature would delay the EOS by alleviating frost damage and 400 retarding vegetation senescence (Cong et al., 2017). In contrast, in the low altitude areas 401 of the eastern Tibetan Plateau, the climate is relatively warm and humid (Figure 4) and 402 marsh vegetation has access to sufficient water (Shen et al., 2022). Therefore, although 403 organic matter may be depleted through the respiration of marsh vegetation due to 404 nighttime warming, increased temperatures can also promote photosynthesis, leading to 405 accumulation of more organic matter the following day via the over compensation effect 406 (Belsky et al., 1986; Shen et al., 2021b, 2022). This would contribute to delaying the EOS 22 407 in the eastern region as a consequence of the increased preseason minimum temperatures. 408 By building a multi-variable regression for each pixel, we showed spatially which 409 preseason variable is the most important for affecting the EOS. The results confirmed that 410 preseason cumulative precipitation and minimum temperature played a crucial role in the 411 relatively arid southwestern and humid eastern Tibetan Plateau, respectively (Figure S3). 412 By comparing the partial correlation coefficients between EOS and climatic factors in 413 different regions, we confirmed that as annual mean temperature increased, the delaying 414 effects of higher preseason maximum and minimum temperatures on the EOS gradually 415 weakened, while the delaying effect of increased preseason precipitation became 416 gradually stronger (Figure 5). Based on our results, we propose that the effects of climate 417 variations on the EOS differ depending on the hydrological constraints on soil moisture 418 in the marshes within the Tibetan Plateau. As such, we further compared the partial 419 correlations between EOS and climatic factors for preseason soil moisture gradients of 420 0.05 m3/m3. The results showed that as soil moisture increased, the delaying effect of 421 increased preseason precipitation on the EOS gradually weakened (Figure 3), while the 422 delaying effects of higher preseason maximum and minimum temperatures became 423 gradually stronger (Figure 3). This finding supports the proposal that increased 424 precipitation and warming temperatures significantly delayed the EOS for marsh 425 vegetation in the arid southwestern Tibetan Plateau and the humid central and eastern 426 areas, respectively. In contrast, increasing daytime temperatures and precipitation 427 advanced the EOS in the arid southwestern and humid central areas due to the reduced 23 428 soil moisture and cooling effect, respectively (Figure 3). 429 430 FIGURE 5 The changes in partial correlation coefficients between EOS and preseason 431 cumulative precipitation, maximum temperature, and minimum temperature along the 432 spatial gradient of long-term annual mean temperature on the Tibetan Plateau from 2001 433 to 2020. The body lines in the figures indicate the linear fit for the partial correlation 24 434 coefficients, and the shading represents the 95% confidence band of the fits. 435 436 4.4 Attribution of temporal changes in the EOS 437 To further explain the temporal and spatial variations in the EOS, we calculated the rates 438 at which precipitation and maximum and minimum temperatures varied on the plateau 439 from 2001 to 2020 (Figure 6 and Table 1). The preseason cumulative precipitation and 440 maximum and minimum temperatures exhibited increasing trends (0.52 mm/a, 0.02 °C/a, 441 and 0.05 °C/a, respectively), but only the increase in the minimum temperature was 442 significant (P < 0.05). Because the EOS showed a significant positive relationship with 443 preseason minimum temperature, the increase in this parameter may partly account for 444 the delayed EOS on the Tibetan Plateau (Figure 1 and Table 1). 25 445 446 FIGURE 6 Preseason climate change in the marshes of the Tibetan Plateau. Spatial 447 patterns of temporal trends in preseason cumulative precipitation (a), maximum 26 448 temperature (b), and minimum temperature (c) in the marshes of the Tibetan Plateau from 449 2001 to 2020. The inset histograms at the bottom of figures display the frequency 450 distributions of the trends. - and + show the negative and positive trend, respectively; * 451 and ** indicate the variation trend is significant (P < 0.05) and extremely significant (P 452 < 0.01), respectively. 453 Table 1 Temporal trends of preseason and seasonal precipitation (mm/a), maximum 454 temperature (°C/a), and minimum temperature (°C/a) in marshes of the Tibetan Plateau 455 from 2001 to 2020. maximum minimum Precipitation temperature temperature Preseason 0.52* -0.02 0.05* Spring 0.22 0.01 0.05** Summer 0.31 0.05* 0.06* Autumn 0.14 0.05 0.08** Winter 0.79 -0.03 0.03 ** 456 P < 0.01; *P < 0.05. 457 In the high-altitude arid area of the southwestern Tibetan Plateau, the EOS was 458 positively correlated with preseason cumulative precipitation (Figure 3). As preseason 459 cumulative precipitation showed significant increasing trends in the high-altitude arid 460 area of the southwestern Tibetan Plateau (Figure 6), the increase in preseason 461 precipitation may partly account for the delayed EOS in this region (Figure 1). In the low 27 462 altitude humid areas in the east and high-altitude cold permafrost areas in the west, the 463 EOS was positively correlated with preseason minimum temperature (Figure 3). 464 Therefore, we deduce that the extremely significant increases of preseason minimum 465 temperature may contribute to the delayed EOS in these regions (Figures 1 and 6). In the 466 northeastern area of the plateau, the EOS appeared to occur earlier throughout the study 467 period (Figure 1), possibly due to the rise in preseason minimum temperature (Figures 3 468 and 6). 469 470 5 Conclusions 471 Our study uncovers several crucial findings regarding the end of the growing season (EOS) 472 in marsh vegetation across the Tibetan Plateau. Firstly, we observed a significant delay in 473 the EOS by 4.10 days/decade during the study period. Secondly, while average preseason 474 cumulative precipitation did not notably impact the regionally averaged EOS, warmer 475 preseason temperatures led to a significant delay in the average EOS of marsh vegetation. 476 Notably, the delaying effect of higher nighttime temperatures on the regionally averaged 477 EOS was more pronounced than that of daytime temperatures. This asymmetric response 478 to diurnal temperature variations can be attributed to the widespread delaying effect of 479 nighttime warming and the spatially diverse relationship between EOS and daytime 480 temperature, influenced by water conditions. Furthermore, our evidence indicates that 481 hydrological factors influencing soil water content play a regulatory role in the impact of 482 climate change on the EOS. As soil moisture decreased, the delaying effect of increasing 28 483 preseason maximum temperatures gradually weakened and even reversed, while the 484 delaying effect of increased preseason precipitation was strengthened. In the humid, cold 485 regions of the central Tibetan Plateau, higher preseason maximum temperatures 486 significantly delayed the EOS, whereas increased precipitation advanced it, potentially 487 due to a cooling effect. In the low altitude, humid regions in the east, higher minimum 488 temperatures delayed the EOS, possibly due to an overcompensation effect. In the arid 489 southwestern area, increased precipitation directly and significantly delayed the EOS, 490 whereas higher daytime temperatures advanced it, likely due to limited water availability. 491 These findings suggest that the EOS in these regions is constrained by water conditions, 492 even within marsh ecosystems. Overall, our study highlights the asymmetric influences 493 of daytime and nighttime temperatures on the EOS of marsh vegetation, particularly in 494 the context of global diurnal asymmetric warming (stronger warming during nighttime 495 than during daytime). It underscores the importance of considering water conditions in 496 EOS simulations conducted by terrestrial ecosystem models in cold and dry regions 497 worldwide. 498 499 Acknowledgements 500 We gratefully acknowledge the National Natural Science Foundation of China 501 (42230516), Key Research Program of Frontier Sciences, Chinese Academy of Sciences 502 (ZDBS-LY-7019), and the Spanish Government grant TED2021-132627 B–I00, funded 503 by MCIN, AEI/10.13039/ 501100011033 European Union Next Generation EU/PRTR. 504 29 505 Conflict of interest statement 506 The authors declare no conflict of interest. 507 508 Data availability statement 509 The MOD13Q1 NDVI dataset was provided by the Earth Science Data Systems of the 510 National Aeronautics and Space Administration 511 (https://ladsweb.modaps.eosdis.nasa.gov). The distribution of marshes was obtained from 512 the wetland distribution datasets for China (http://www.geodata.cn). The climate data was 513 available from the National Meteorological Information Center of China 514 (http://data.cma.cn/data). The soil moisture data were obtained from the National Tibetan 515 Plateau Scientific Data Center (https://data.tpdc.ac.cn). 516 517 References 518 Bao, G., Tuya, A., Bayarsaikhan, S., Dorjsuren, A., Mandakh, U., Bao, Y., Li, C., & 519 Vanchindorj, B. (2020). Variations and climate constraints of terrestrial net primary 520 productivity over Mongolia. Quaternary International, 537, 112-125. 521 https://doi.org/10.1016/j.quaint.2019.06.017 522 Belsky, A. J. (1986). Does herbivory benefit plants? A review of the evidence. The 523 American Naturalist, 127, 870-892. https://doi.org/10.1086/284531 524 Che, M., Chen, B., Innes, J. L., Wang, G., Dou, X., Zhou, T., Zhang, H., Yan, J., Xu, G., 525 & Zhao, H. (2014). Spatial and temporal variations in the end date of the vegetation 30 526 growing season throughout the Qinghai–Tibetan Plateau from 1982 to 2011. 527 Agricultural and Forest Meteorology, 189, 81-90. 528 https://doi.org/10.1016/j.agrformet.2014.01.004 529 Chen, X. (2017). Plant phenology of natural landscape dynamics. In Spatiotemporal 530 processes of plant phenology. Springer, Berlin, Heidelberg. 531 Chen, X., An, S., Inouye, D. W., & Schwartz, M. D. (2015). Temperature and snowfall 532 trigger alpine vegetation green-up on the world's roof. Global Change Biology, 21, 533 3635-3646. https://doi.org/10.1111/gcb.12954 534 Cheng, M., Jin, J., & Jiang, H. (2021). Strong impacts of autumn phenology on grassland 535 ecosystem water use efficiency on the Tibetan Plateau. Ecological Indicators, 126, 536 107682. https://doi.org/10.1016/j.ecolind.2021.107682 537 Coleman, D. J., Schuerch, M., Temmerman, S., Guntenspergen, G., Smith, C. G., & 538 Kirwan, M. L. (2022). Reconciling models and measurements of marsh vulnerability 539 to sea level rise. Limnology and Oceanography Letters, 7, 140-149. 540 https://doi.org/10.1002/lol2.10230 541 Cong, N., Shen, M., & Piao, S. (2017). Spatial variations in responses of vegetation 542 autumn phenology to climate change on the Tibetan Plateau. Journal of Plant 543 Ecology, 10, 744-752. https://doi.org/10.1093/jpe/rtw084 544 Cong, N., Wang, T., Nan, H., Ma, Y., Wang, X., Myneni, R. B., & Piao, S. (2013). Changes 545 in satellite-derived spring vegetation green-up date and its linkage to climate in 546 China from 1982 to 2010: a multimethod analysis. Global Change Biology, 19, 881- 31 547 891. https://doi.org/10.1111/gcb.12077 548 Dong, M., Jiang, Y., Zheng, C., & Zhang, D. (2012). Trends in the thermal growing season 549 throughout the Tibetan Plateau during 1960-2009. Agricultural and Forest 550 Meteorology, 166, 201-206. https://doi.org/10.1016/j.agrformet.2012.07.013 551 Dorji, T., Totland, Ø., Moe, S. R., Hopping, K. A., Pan, J., & Klein, J. A. (2013). Plant 552 functional traits mediate reproductive phenology and success in response to 553 experimental warming and snow addition in Tibet. Global Change Biology, 19, 459- 554 472. https://doi.org/10.1111/gcb.12059 555 Estiarte, M., & Peñuelas, J. (2015). Alteration of the phenology of leaf senescence and 556 fall in winter deciduous species by climate change: effects on nutrient proficiency. 557 Global change biology, 21, 1005-1017. https://doi.org/10.1111/gcb.12804 558 Fu, Y. H., Piao, S., Delpierre, N., Hao, F., Hänninen, H., Liu, Y., Sun, W., Janssens, I. A., 559 & Campioli, M. (2018). Larger temperature response of autumn leaf senescence than 560 spring leaf ‐ out phenology. Global Change Biology, 24(5), 2159-2168. 561 https://doi.org/10.1111/gcb.14021 562 Fu, Y. H., Piao, S., Op de Beeck, M., Cong, N., Zhao, H., Zhang, Y., Menzel, A., & 563 Janssens, I. A. (2014). Recent spring phenology shifts in western Central Europe 564 based on multiscale observations. Global Ecology and Biogeography, 23, 1255-1263. 565 https://doi.org/10.1111/geb.12210 566 Ganguly, S., Friedl, M. A., Tan, B., Zhang, X., & Verma, M. (2010). Land surface 567 phenology from MODIS: Characterization of the Collection 5 global land cover 32 568 dynamics product. Remote Sensing of Environment, 114, 1805-1816. 569 https://doi.org/10.1016/j.rse.2010.04.005 570 Ganjurjav, H., Gornish, E. S., Hu, G., Schwartz, M. W., Wan, Y., Li, Y., & Gao, Q. (2020). 571 Warming and precipitation addition interact to affect plant spring phenology in 572 alpine meadows on the central Qinghai-Tibetan Plateau. Agricultural and Forest 573 Meteorology, 287, 107943. https://doi.org/10.1016/j.agrformet.2020.107943 574 Gao, Y. C., & Liu, M. (2013). Evaluation of high-resolution satellite precipitation 575 products using rain gauge observations over the Tibetan Plateau. Hydrology and 576 Earth System Sciences, 17, 837-849. https://doi.org/10.5194/hess-17-837-2013 577 Garonna, I., De Jong, R., De Wit, A. J., Mücher, C. A., Schmid, B., & Schaepman, M. E. 578 (2014). Strong contribution of autumn phenology to changes in satellite‐derived 579 growing season length estimates across Europe (1982-2011). Global Change 580 Biology, 20, 3457-3470. https://doi.org/10.1111/gcb.12625 581 Ge, Q., Wang, H., Rutishauser, T., & Dai, J. (2015). Phenological response to climate 582 change in China: a meta ‐ analysis. Global change biology, 21(1), 265-274. 583 https://doi.org/10.1111/gcb.12648 584 Kelsey, K. C., Pedersen, S. H., Leffler, A. J., Sexton, J. O., Feng, M., & Welker, J. M. 585 (2021). Winter snow and spring temperature have differential effects on vegetation 586 phenology and productivity across Arctic plant communities. Global Change 587 Biology, 27, 1572-1586. https://doi.org/10.1111/gcb.15505 588 Keppeler, F. W., Olin, J. A., López‐Duarte, P. C., Polito, M. J., Hooper‐Bùi, L. M., 33 589 Taylor, S. S., Rabalais, N. N., Fodrie F. J., Roberts, B. J., Turner, R. R., Matin, C. 590 W., & Jensen, O. P. (2021). Body size, trophic position, and the coupling of different 591 energy pathways across a saltmarsh landscape. Limnology and Oceanography 592 Letters, 6, 360-368. https://doi.org/10.1002/lol2.10212 593 Li, Q., Shi, G., Shangguan, W., Nourani, V., Li, J., Li, L., Huang, F., Zhang, Y., Wang, C., 594 Wang, D., Qiu, J., Lu, X., & Dai, Y. (2022). A 1 km daily soil moisture dataset over 595 China using in situ measurement and machine learning, Earth Syst. Sci. Data, 14, 596 5267–5286. https://doi.org/10.5194/essd-14-5267-2022 597 Liu, Q., Fu, Y. H., Zeng, Z., Huang, M., Li, X., & Piao, S. (2016). Temperature, 598 precipitation, and insolation effects on autumn vegetation phenology in temperate 599 China. Global Change Biology, 22, 644-655. https://doi.org/10.1111/gcb.13081 600 Liu, X., Chen, Y., Li, Z., Li, Y., Zhang, Q., & Zan, M. (2021). Driving Forces of the 601 Changes in Vegetation Phenology in the Qinghai–Tibet Plateau. Remote Sensing, 13, 602 4952. https://doi.org/10.3390/rs13234952 603 Ma, R., Shen, X., Zhang, J., Xia, C., Liu, Y., Wu, L., Wang, Y., Jiang, M., & Lu, X. (2022). 604 Variation of vegetation autumn phenology and its climatic drivers in temperate 605 grasslands of China. International Journal of Applied Earth Observation and 606 Geoinformation, 114, 103064. https://doi.org/10.1016/j.jag.2022.103064 607 Mao, D., Wang, Z., Du, B., Li, L., Tian, Y., Jia, M., Zeng, Y., Song, K., Jiang, M., & Wang, 608 Y. (2020). National wetland mapping in China: A new product resulting from object- 609 based and hierarchical classification of Landsat 8 OLI images. ISPRS Journal of 34 610 Photogrammetry and Remote Sensing, 164, 11-25. 611 https://doi.org/10.1016/j.isprsjprs.2020.03.020 612 Molino, G. D., Carr, J. A., Ganju, N. K., & Kirwan, M. L. (2022). Variability in marsh 613 migration potential determined by topographic rather than anthropogenic constraints 614 in the Chesapeake Bay region. Limnology and Oceanography Letters, 7, 321-331. 615 https://doi.org/10.1002/lol2.10262 616 Munné-Bosch, S., & Alegre, L. (2004). Die and let live: leaf senescence contributes to 617 plant survival under drought stress. Functional Plant Biology, 31, 203-216. 618 https://doi.org/10.1071/FP03236 619 Nan, Z., Li, S., & Cheng, G. (2005). Prediction of permafrost distribution on the Qinghai- 620 Tibet Plateau in the next 50 and 100 years. Science in China Series D: Earth Sciences, 621 48, 797-804. https://doi.org/10.1360/03yd0258 622 Peng, S., Piao, S., Ciais, P., Myneni, R. B., Chen, A., Chevallier, F., Dolman, A. J., 623 Janssens, I. A., Penuelas, J., Zhang, G., Vicca, S., Wan, S., Wang, S., & Zeng, H. 624 (2013). Asymmetric effects of daytime and night-time warming on Northern 625 Hemisphere vegetation. Nature, 501, 88-92. https://doi.org/10.1038/nature12434 626 Peñuelas, J., Filella, I. (2001). Responses to a warming world. Science, 294(5543): 793- 627 795. https:// doi.org/10.1126/science.10668. 628 Piao, S., Ciais, P., Friedlingstein, P., Peylin, P., Reichstein, M., Luyssaert, S., Margolis, 629 H., Fang, J., Barr, A., Chen, A., Grelle, A., Hollinger, D. Y., Laurila, T., Lindroth, A., 630 Richardson, A. D., & Vesala, T. (2008). Net carbon dioxide losses of northern 35 631 ecosystems in response to autumn warming. Nature, 451, 49-52. 632 https://doi.org/10.1038/nature06444 633 Piao, S., Cui, M., Chen, A., Wang, X., Ciais, P., Liu, J., & Tang, Y. (2011). Altitude and 634 temperature dependence of change in the spring vegetation green-up date from 1982 635 to 2006 in the Qinghai-Xizang Plateau. Agricultural and Forest Meteorology, 151, 636 1599-1608. https://doi.org/10.1016/j.agrformet.2011.06.016 637 Piao, S., Fang, J., & He, J. (2006a). Variations in vegetation net primary production in the 638 Qinghai-Xizang Plateau, China, from 1982 to 1999. Climatic Change, 74, 253-267. 639 https://doi.org/10.1007/s10584-005-6339-8 640 Piao, S., Fang, J., Zhou, L., Ciais, P., & Zhu, B. (2006b). Variations in satellite-derived 641 phenology in China's temperate vegetation. Global Change Biology, 12, 672-685. 642 https://doi.org/10.1111/j.1365-2486.2006.01123.x 643 Piao, S., Friedlingstein, P., Ciais, P., Viovy, N., & Demarty, J. (2007). Growing season 644 extension and its impact on terrestrial carbon cycle in the Northern Hemisphere over 645 the past 2 decades. Global Biogeochemical Cycles, 21. 646 https://doi.org/10.1029/2006GB002888 647 Piao, S., Liu, Q., Chen, A., Janssens, I. A., Fu, Y., Dai, J., Liu, L., Lian, X., Shen, M., & 648 Zhu, X. (2019). Plant phenology and global climate change: Current progresses and 649 challenges. Global Change Biology, 25, 1922-1940. 650 https://doi.org/10.1111/gcb.14619 651 Qin, G., Adu, B., Li, C., & Wu, J. (2022). Diverse Responses of Phenology in Multi- 36 652 Grassland to Environmental Factors on Qinghai–Tibetan Plateau in China. 653 Theoretical and Applied Climatology, 148, 931-942. 654 https://doi.org/10.1007/s00704-022-03963-3 655 Rice, K. E., Montgomery, R. A., Stefanski, A., Rich, R. L., & Reich, P. B. (2018). 656 Experimental warming advances phenology of groundlayer plants at the boreal‐ 657 temperate forest ecotone. American journal of botany, 105, 851-861. 658 https://doi.org/10.1002/ajb2.1091 659 Richardson, A. D., Keenan, T. F., Migliavacca, M., Ryu, Y., Sonnentag, O., & Toomey, 660 M. (2013). Climate change, phenology, and phenological control of vegetation 661 feedbacks to the climate system. Agricultural and Forest Meteorology, 169, 156-173. 662 https://doi.org/10.1016/j.agrformet.2012.09.012 663 Shen, M., Piao, S., Chen, X., An, S., Fu, Y. H., Wang, S., Cong, N., & Janssens, I. A. 664 (2016). Strong impacts of daily minimum temperature on the green-up date and 665 summer greenness of the Tibetan Plateau. Global Change Biology, 22, 3057-3066. 666 https://doi.org/10.1111/gcb.13301 667 Shen, M., Piao, S., Cong, N., Zhang, G., & Jassens, I. A. (2015a). Precipitation impacts 668 on vegetation spring phenology on the Tibetan Plateau. Global Change Biology, 21, 669 3647-3656. https://doi.org/10.1111/gcb.12961 670 Shen, M., Piao, S., Dorji, T., Liu, Q., Cong, N., Chen, X., An, S., Wang, S., Wang, T., & 671 Zhang, G. (2015b). Plant phenological responses to climate change on the Tibetan 672 Plateau: research status and challenges. National Science Review, 2, 454-467. 37 673 https://doi.org/10.1093/nsr/nwv058 674 Shen, M., Tang, Y., Chen, J., Zhu, X., & Zheng, Y. (2011). Influences of temperature and 675 precipitation before the growing season on spring phenology in grasslands of the 676 central and eastern Qinghai-Tibetan Plateau. Agricultural and Forest Meteorology, 677 151, 1711-1722. https://doi.org/10.1016/j.agrformet.2011.07.003 678 Shen, M., Wang, S., Jiang, N., Sun, J., Cao, R., Ling, X., Fang, B., Zhang, L., Zhang, L., 679 Xu, X., Lv, W., Li, B., Sun, Q., Meng, F., Jiang, Y., Dorji, T., Fu, Y., Iier, A., Vitasse, 680 Y., Steltzer, H., Ji, Z., Zhao, W., Piao, S., & Fu, B. (2022). Plant phenology changes 681 and drivers on the Qinghai–Tibetan Plateau. Nature Reviews Earth & Environment, 682 1-19. https://doi.org/10.1038/s43017-022-00317-5 683 Shen, M., Zhang, G., Cong, N., Wang, S., Kong, W., & Piao, S. (2014). Increasing 684 altitudinal gradient of spring vegetation phenology during the last decade on the 685 Qinghai–Tibetan Plateau. Agricultural and Forest Meteorology, 189, 71-80. 686 https://doi.org/10.1016/j.agrformet.2014.01.003 687 Shen, X., Jiang, M., & Lu, X. (2023). Diverse impacts of day and night temperature on 688 spring phenology in freshwater marshes of the Tibetan Plateau. Limnology and 689 Oceanography Letters, 1-7. https://doi.org/10.1002/lol2.10285 690 Shen, X., Jiang, M., Lu, X., Liu, X., Liu, B., Zhang, J., Wang, X., Tong, S., Lei, G., Wang, 691 S., Tong, C., Fan, H., Tian, K., Wang, X., Hu, Y., Xie, Y., Ma, M., Zhang, S., Cao, 692 C., & Wang, Z. (2021a). Aboveground biomass and its spatial distribution pattern of 693 herbaceous marsh vegetation in China. Science China Earth Sciences, 64, 1115-1125. 38 694 https://doi.org/10.1007/s11430-020-9778-7 695 Shen, X., Liu, B., Henderson, M., Wang, L., Wu, Z., Wu, H., Jiang, M., & Lu, X. (2018). 696 Asymmetric effects of daytime and nighttime warming on spring phenology in the 697 temperate grasslands of China. Agricultural and Forest Meteorology, 259, 240-249. 698 https://doi.org/10.1016/j.agrformet.2018.05.006 699 Shen, X., Liu, B., Jiang, M., Wang, Y., Wang, L., Zhang, J., & Lu, X. (2021b). 700 Spatiotemporal change of marsh vegetation and its response to climate change in 701 China from 2000 to 2019. Journal of Geophysical Research: Biogeosciences, 126, 702 e2020JG006154. https://doi.org/10.1029/2020JG006154 703 Shen, X., Liu, B., Xue, Z., Jiang, M., Lu, X., & Zhang, Q. (2019). Spatiotemporal 704 variation in vegetation spring phenology and its response to climate change in 705 freshwater marshes of Northeast China. Science of the Total Environment, 666, 1169- 706 1177. https://doi.org/10.1016/j.scitotenv.2019.02.265 707 Shen, X., Liu, Y., Zhang, J., Wang, Y., Ma, R., Liu, B., Lu, X., & Jiang, M. (2022). 708 Asymmetric impacts of diurnal warming on vegetation carbon sequestration of 709 marshes in the Qinghai Tibet Plateau. Global Biogeochemical Cycles, 36, 710 e2022GB007396. https://doi.org/10.1029/2022GB007396 711 Shi, C., Sun, G., Zhang, H., Xiao, B., Ze, B., Zhang, N., & Wu, N. (2014). Effects of 712 warming on chlorophyll degradation and carbohydrate accumulation of alpine 713 herbaceous species during plant senescence on the Tibetan Plateau. PLoS One, 9, 714 e107874. https://doi.org/10.1371/journal.pone.0107874 39 715 Su, M., Huang, X., Xu, Z., Zhu, W., & Lin, Z. (2022). A Decrease in the Daily Maximum 716 Temperature during Global Warming Hiatus Causes a Delay in Spring Phenology in 717 the China–DPRK–Russia Cross-Border Area. Remote Sensing, 14, 1462. 718 https://doi.org/10.3390/rs14061462 719 Tang, J., Körner, C., Muraoka, H., Piao, S., Shen, M., Thackeray, S. J., & Yang, X. (2016). 720 Emerging opportunities and challenges in phenology: a review. Ecosphere, 7, 721 e01436. https://doi.org/10.1002/ecs2.1436 722 Turnbull, M. H., Murthy, R., & Griffin, K. L. (2002). The relative impacts of daytime and 723 night‐time warming on photosynthetic capacity in Populus deltoides. Plant, Cell 724 & Environment, 25, 1729-1737. https://doi.org/10.1046/j.1365-3040.2002.00947.x 725 Wang, Y., Shen, X., Jiang, M., Tong, S., & Lu, X. (2021). Spatiotemporal change of 726 aboveground biomass and its response to climate change in marshes of the Tibetan 727 Plateau. International Journal of Applied Earth Observation and Geoinformation, 728 102, 102385. https://doi.org/10.1016/j.jag.2021.102385 729 Wu, C., Peng, J., Ciais, P., Peñuelas, J., Wang, H., Beguería, S., Andrew Black, T.,Jassal, 730 R. S. , Zhang, X. , Yuan, W., Liang, E. , Wang, X., Hua, H., Liu, R. , Ju, W., Fu,Y., 731 & Ge, Q.(2022). Increased drought effects on the phenology of autumn leaf 732 senescence. Nature Climate Change, 12(10), 943-949. 733 https://doi.org/10.1038/s41558-022-01464-9 734 Wu, C., Wang, X., Wang, H., Ciais, P., Peñuelas, J., Myneni, R. B., Desai, A. R., Gough, 735 C. M., Gonsamo, A., Black, A. T., Jassal, R. S., Ju, W., Yuan, W., Fu, Y., Shen, M., 40 736 li, S., Liu, R., Chen, J., & Ge, Q. (2018). Contrasting responses of autumn-leaf 737 senescence to daytime and night-time warming. Nature Climate Change, 8, 1092- 738 1096. https://doi.org/10.1038/s41558-018-0346-z 739 Yang, Y., Guan, H., Shen, M., Liang, W., & Jiang, L. (2015). Changes in autumn 740 vegetation dormancy onset date and the climate controls across temperate 741 ecosystems in China from 1982 to 2010. Global Change Biology, 21, 652-665. 742 https://doi.org/10.1111/gcb.12778 743 Yang, Y., Qi, N., Zhao, J., Meng, N., Lu, Z., Wang, X., Kang, L., Li, R., Ma, J., & Zheng, 744 H. (2021). Detecting the Turning Points of Grassland Autumn Phenology on the 745 Qinghai-Tibetan Plateau: Spatial Heterogeneity and Controls. Remote Sensing, 13, 746 4797. https://doi.org/10.3390/rs13234797 747 Yu, H., Luedeling, E., & Xu, J. (2010). Winter and spring warming result in delayed 748 spring phenology on the Tibetan Plateau. Proceedings of the National Academy of 749 Sciences, 107(51), 22151-22156. https://doi.org/10.1073/pnas.1012490107 750 Zhang, G., Zhang, Y., Dong, J., & Xiao, X. (2013). Green-up dates in the Tibetan Plateau 751 have continuously advanced from 1982 to 2011. Proceedings of the National 752 Academy of Sciences, 110, 4309-4314. https://doi.org/10.1073/pnas.1210423110 753 Zhu, W., Tian, H., Xu, X., Pan, Y., Chen, G., & Lin, W. (2012). Extension of the growing 754 season due to delayed autumn over mid and high latitudes in North America during 755 1982–2006. Global Ecology and Biogeography, 21, 260-271. 756 https://doi.org/10.1111/j.1466-8238.2011.00675.x 41