For many plants, nectar is hidden in the depth of nectar spurs. The diverse morphology and length of nectar spurs can attract pollinators with various proboscis lengths and promote the differentiation of pollination niche, which plays an important role in the formation of species diversity. Therefore, nectar spur is considered as a key innovation trait in promoting speciation and an ideal trait for studying floral evolution and plant-pollinator coevolution in angiosperm. This review systematically summarizes the worldwide plant taxonomic databases and published articles, classifies the nectar spur plants, and reviews the diversity and evolutionary mechanisms of nectar spur plants. We investigated 3 427 nectar spur species belonging to 13 orders, 23 families and 271 genera of angiosperm. Among them, Orchidaceae has the largest number of species, with 1 536 species, following by Papaveraceae with 487 species, Ranunculaceae with 351 species, Balsaminaceae with 284 species and Violaceae with 197 species, accounting for 83.3% of all the species. According to the sources of the development of nectar spurs, the nectar spurs can be divided into six types: petal spur, corolla spur, sepal spur, calyx spur, tepal spur and hypanthium spur. The petal spurs are the predominant type, detected in 212 genera. The average length of nectar spurs in Balsaminaceae is the longest (23.8 mm) among the 2 546 species whose nectar spurs have been measured. The shortest nectar spurs are reported in Lycelandaceae, which are only 1.6 mm on average. Most nectar spurs have different degrees of curvature, likely helping to improve the pollination efficiency. The number of nectar spur(s) in a flower ranges from 1 to 6, and most species have only one nectar spur in a flower. Multiple nectar spurs may increase the visiting time of pollinators. Some plants in Orchidaceae and Violaceae have appendages inside the nectar spurs, which may have the function of nectar secretion. The pollinators of plants with nectar spur mainly comprise insects from Hymenoptera, Lepidoptera, and some long-proboscis species from Diptera and birds. Nectar robbers are mainly bees and birds. There is a general association of geographic mosaic theory of coevolution between the length of nectar spur and the length of pollinator proboscis. Evolutionary ecologists have proposed the hypothesis of “Darwin’s Mechanistic Model” and “The Pollinator Shift Model” to explain this phenomenon. In future studies of the mechanisms of nectar spur evolution, we should not only focus on the morphological characteristics of spur length and proboscis length and the pollination ecology, but also explore the driving forces of nectar spurs evolution from cell division/elongation, functional genes, abiotic factors, with the help of novel technological approaches including developmental anatomy and genomics.

Aims The southwestern China is a region with complex topography and diverse ecosystem and vegetation types. However, its role as an ecological barrier is being weakened by the effects of climate change and increasing pressure of human activities. This study examines the temporal dynamics of vegetation carbon use efficiency (CUE) in this region using the CMIP6 models, aiming to effectively reducing the uncertainties in prognostic results of future predictions.
Methods We used MODIS remote sensing data for the period 2001-2014 and simulations from 15 models in the Phase 6 of the Coupled Model Intercomparison Project (CMIP6), to determine the capability of the new generation models in simulating the seasonal and annual vegetation CUE in the southwestern China. The performance of the models was ranked based on the composite rating index (MR).
Important findings Most of the models used in this study underestimated the annual vegetation CUE, and their ability to simulate the spatial patterns in the trends of vegetation CUE is generally poor. However, some models performed relatively well in simulating the spatial distribution of multi-year average vegetation CUE; the top 1/3 tier included BCC-CSM2-MR, CMCC-ESM2, TaiESM, EC-Earth3-Veg and CAS-ESM2-0 in the order of performance. Among the seasons, the models best simulated the spatial distribution of vegetation CUE in summer, with better results from BCC-CSM2-MR, EC-Earth3-Veg, TaiESM, CMCC-ESM2 and CAS-ESM2-0. The simulation capability of the models for winter is second only to that for summer, and relatively poor for spring and autumn. Compared to the simulations by individual models, the multi-model ensemble mean (MME-S) reduced the uncertainties and exhibited a strong simulation capability, especially in the spatial distribution of vegetation CUE in local areas such as the Sichuan Basin. There was a lack of good simulation capability for the spatial distribution of vegetation CUE in Qingzang Plateau, Hengduan Mountains and other topographically complex areas. In general, before applying the CMIP6 models for regional vegetation CUE simulation, it is necessary to comprehensively evaluate the specific models from multiple perspectives to select the models with better simulation performance.

Aims Meteorological data show faster warming of the global land surface during the night than during the day in the past 50 years. However, most of the previous studies were focused on the effects of whole-day equivalent warming, and the understanding of effects of diurnal asymmetric warming remains elusive.
Methods This study evaluated the effects of diurnal asymmetric warming on carbon sink capacity using a optimization model considering photosynthetic gain and hydraulic cost in a broadleaf Korean pine (Pinus koraiensis) forest in Changbai Mountains.
Important findings Results show that the model simulations matched well with observations of net primary production based on the data measured from eddy covariance flux towers. Warming promoted carbon sequestration (11.2%-13.8%) in our study area but did not significantly affect the water use efficiency, and the positive effects on annual carbon sequestration had no statistical difference among different warming scenarios. In addition, warming increased the water stress for forest plants, subsequently increasing the loss percentage of conductivity (PLC, hydraulic vulnerability; 1.1%). In conclusion, all warming scenarios significantly enhanced the current carbon sink capacity of forests compared with ambient condition, but warming may increase the risk of forest death through hydraulic failure, which would significantly affect the future forest carbon sink.

Aims Due to complex root-soil interactions, the responses of carbon (C) dynamics in the rhizosphere soil to nitrogen (N) deposition may be different from those in bulk soil. However, the potentially different responses of C dynamics between the rhizosphere and bulk soil and their contributions to soil C sequestration under N deposition are still not elucidated.
Methods In this study, a typical subalpine coniferous plantation (Picea asperata) with chronic N addition treatments in southwestern China was selected as the research object. Based on the experimental plots of simulated N deposition (control: 0 kg·hm^-2·a^-1; N addition: 25 kg·hm^-2·a^-1), we measured the contents of soil organic carbon and its different physical and chemical fractions. Afterwards, by combining the rhizosphere spatial numerical model, we explored the differences in the C pool size of SOC and its fractions and their relative contribution to SOC pools between the rhizosphere and bulk soil, and further quantified the effects of N addition on soil C sequestration in rhizosphere soil.
Important findings The results showed that: 1) Although the addition of N increased the content of SOC and its physical and chemical components in the rhizosphere and non-rhizosphere at the same time, it only reached a significant level in the rhizosphere. Specifically, the rhizosphere SOC content increased by 23.64% under N addition, in which particulate organic carbon (POC), mineral-associated organic carbon (MAOC), labile carbon (LP-C) and recalcitrant carbon (RP-C) content increased by 19.63%, 18.01%, 30.48% and 15.01%, respectively. 2) The total SOC pool increment of spruce forest (0.88 kg·m^-2) was verified with the results of the rhizosphere space numerical model, and the effective rhizosphere extent of the southwest mountain coniferous forest was estimated to be 1.6 mm. Within this extent, N addition increased the SOC stocks of the rhizosphere and bulk soil by 33.37% and 7.38%, contributing to 45.45% and 54.55% of the total SOC pool increment, respectively. Among them, labile C components (POC and LP-C) are the major contributors to rhizosphere SOC accumulation under N addition. These results suggested that the rhizosphere and bulk soil of coniferous forest in southwestern mountainous area had great C sequestration potential under N deposition, and the C sink was more obvious in the rhizosphere soil. Our results highlight the importance of integrating rhizosphere processes into land surface models to accurately predict ecosystem functions in the context of increasing N deposition.

Aims Grassland is an important component of terrestrial ecosystems around the world and plays an important role in terrestrial carbon cycling. However, large uncertainties still exist in predictions of soil organic carbon (SOC) dynamics in grassland ecosystems using earth system models, partly due to an inadequate understanding of the spatial patterns and drivers of soil carbon components and the rate of decomposition. In this study, we explored the determinants of the contents of total SOC and its components as well as the rate of soil carbon decomposition in the topsoil of temperate grasslands of Nei Mongol.
Methods Soil samples at depths of 0-10 cm were collected during July to August 2015 from field sites on the Nei Mongol Plateau. We measured the contents of total SOC and its partitioning in three soil aggregate size-classes, and the decomposition rate based on laboratory incubation. In addition, we acquired a suite of explanatory factors including climatic, edaphic, vegetation, and mineral variables. Variance partitioning analyses were then used to investigate the relative importance of the four factors in affecting the contents of total SOC, aggregate-classified carbon fractions and soil carbon decomposition rate.
Important findings The contents of total SOC and three carbon fractions displayed an increasing trend from southwest to northeast of the study area, while soil carbon decomposition rate (standardized by SOC) showed a reverse trend. The carbon contents in bulk soil and different aggregate fractions are highest in the meadow steppe, followed by the typical steppe and the desert steppe; whereas soil carbon decomposition rate (standardized by SOC) was highest in the desert steppe, followed by the typical steppe and the meadow steppe. The spatial variations of carbon contents in the three soil aggregate fractions were mainly driven by climatic and mineral factors, with finer soil particles attaching greater relative importance in the effect by the mineral factor. The soil carbon decomposition rate (standardized by SOC) was affected by mineral, edaphic, and climatic factors. These findings highlight the importance of considering the differential influences by minerals in different soil aggregate carbon fractions, particularly the silt- and clay-associated carbon in the Earth system models, so as to improve the accuracy in the prediction of SOC dynamics in grassland ecosystems under changing environment.

Aims Grazing, one of the primary ways of grassland utilization in Nei Mongol, has essential influences on plant community properties of grasslands. However, the comprehensive response patterns of Nei Mongol grassland plant community properties to grazing remain unclear.
Methods Based on a dataset derived from 76 studies, the plant community characteristics and soil physicochemical properties of Nei Mongol grasslands under different grazing intensities, different grassland types and different grazing years were integrated and analyzed in order to comprehensively evaluate the response patterns of Nei Mongol grasslands to grazing.
Important findings Our results showed that grazing significantly reduced plant above/below ground biomass, cover, height, density, species richness, Shannon-Wiener diversity index, Pielou evenness index, Simpson diversity index, and soil water content. The negative effects of grazing were strengthened with increasing of grazing intensity and duration. Moreover, grazing had a greater negative effect on the grasslands with sparse vegetation and low environmental carrying capacity (e.g., desert grasslands, sandy areas, etc.). This study shows that the responses of plant community characteristics to grazing in Nei Mongol grasslands are regulated by multiple factors, and appropriate grazing intensity and grazing time should be set according to different types of grasslands to achieve sustainable utilization of grasslands. The standards of grazing intensity in current grazing studies are not uniform, making it difficult to compare different studies, and the results from some studies do not have statistical significance due to a lack of replications in the experiment. The exploration of uniform quantitative standards for grazing intensity will be an important and challenging issue in future grazing studies, and the rationality of experimental design should also be emphasized.

Aims Comparing functional traits of coexisting tree species in natural forests can reveal the adaptation strategies of different tree species to different habitats, which is essential for understanding the coexistence mechanism of the local forest community and guiding forest restoration.
Methods In this study, we compared the functional traits of five coexisting tree species in a Quercus aliena var. acuteserrata forest, the zonal vegetation in the Qinling Mountains, and further examined the difference in the traits space defined by the first two principal component analysis (PCA) axes to reveal the differences of these tree species in ecological strategies.
Important findings The results showed that there were obvious differences in the above- and belowground functional traits of different tree species, especially in the strategies of leaf resource acquisition and defense. Specifically, Acer davidii tended to have more acquisitive leaf traits and fine root traits, while Carpinus cordata was relatively more conservative in leaf and fine root traits. Lindera obtusiloba was inclined to have leaves with high resistance and defense ability, while leaves of Cornus kousa subsp. chinensis showed low resistance and defense ability. The dominant species, Q. aliena var. acuteserrata, occupied the middle position along the strategy gradient. The differences in the functional traits and their combinations of coexisting tree species in Q. aliena var. acuteserrata forest reflect their differences in ecological strategies and niches, which may contribute to their stable coexistence.

Aims Syzygium hancei is the dominant species in the seasonal rainforests of the coastal zone of Guangxi. It plays an important role in restoration of the coastal ecological environment. Hence, investigating the population dynamics of natural S. hancei populations in Guangxi is of significant importance.
Methods In this study we monitored seedlings of S. hancei populations in three distinct communities (A, B and C) in Fangchenggang City over a period of two-years. Data on seedling survival rates were collected, and dynamic life tables were established. Survival curves, mortality curves, and survival function curves were constructed and analyzed.
Important findings Our results showed that: (1) The survival curves of populations A and C exhibited Deevey-II type characteristics, whereas the curve for population B displayed Deevey-III type characteristics. The mortality curves experienced substantial fluctuations during the early stage, but eventually stabilized, oscillating between 0.2 and 0.4 in the later stage. (2) Survival analysis demonstrated that the survival rates of populations A, B, and C followed an initial increase followed by a decrease. The cumulative mortality rate in the later period consistently exceeded the survival rate. Survival function fluctuations of population B were more stable compared to populations A and C. (3) The comprehensive analysis showed that all the natural regeneration capacities of those three S. hancei populations were low, with population B exhibiting the most stable capacity. The low natural regeneration capacity can be attributed to a limited germination base, high mortality rate of seedlings, and insufficient light, resulting in fewer seedlings progressing to the subsequent growth stage.

Aims The morphology and photosynthetic functions of compound leaf are one of the most intuitive manifestations of plant response to habitats. The changes of photosynthetic capacity and number of leaflets of compound leaf directly reflect the adaptability of plants to habitats. There are some morphological differences between male and female Fraxinus mandschurica, and their compound leaf may have different adaptabilities in various habitats.
Methods This experiment took compound leaves male and female mature F. mandschurica growing in the drought saline-alkali habitat and suitable habitat as the research materials. The gender differences between female and male plants and the differences of morphology and photosynthetic function between compound leaf with different numbers of leaflets under two habitats were compared and analyzed.
Important findings The results showed that (1) compared with the suitable habitat, the morphology and photosynthetic capacity of compound leaf with the same number leaflets of female and male F. mandschurica under the drought saline-alkali habitat did not show gender difference, but the increment of percentages of compound leaf with 9 leaflets of female plants was 29.13% higher than that of the male plants, the decrement of percentages of compound leaf with 11 leaflets was 33.74% higher than that of the male plants, and the decreament of photosynthetic rate of compound leaf with 9 leaflets of female plants was 10.98% higher than that of the male plants. (2) Under the suitable habitat, the proportion of compound leaves of male plants was mainly concentrated within 9 and 11 leaflets, while the proportion of compound leaf of female plants was mainly concentrated within 11 leaflets. Meanwhile, the specific leaf area and leaf dry mass of compound leaves with 11 leaflets of female and male plants were greater than that of compound leaf with 7 and 9 leaflets. Under the drought saline-alkali habitat, the proportion of compound leaf of female and male plants was mainly concentrated in 9 leaflets. The specific leaf area of compound leaf with 9 leaflets of male and female plants was less than that of 11 leaflet compound leaf, whereas the leaf dry mass showed the opposite trend. There was also no significant difference in the net photosynthetic rate of compound leaf with 9 and 11 leaflets between female and male plants in the two habitats. (3) Among the indexes of compound leaf morphology and stomatal gas exchange capacity, the percentages of compound leaf with different number of leaflets, leaf dry mass, specific leaf area and net photosynthetic rate of compound leaf have high plasticity. Therefore, under the suitable habitat, there is no gender difference in the photosynthetic function between male and female plants, but there are gender differences in the morphology of compound leaf of male and female F. mandschurica, which are shown as follows: the female plants mainly developed the compound leaf with 11 leaflets, and the male plants predominantly developed the compound leaf with 9 leaflets and 11 leaflets. However, under the drought saline-alkali habitat, the morphology and photosynthetic function of compound leaf of male and female F. mandschurica did not show gender differences, and both female and male plants mainly developed the compound leaf with 9 leaflets. This study provides a theoretical basis for the gender difference in the growth and development of compound leaf in different habitats, and also provides data support for the ecological adaptability of compound leaf of dioecious trees.

Aims This study aimed to illustrate community structure, molecular ecological network and driving factors of ectomycorrhizal fungi associated with Pinus sylvestris var. mongolica at different growth periods, and provide a theoretical basis for sustainable management of P. sylvestris var. mongolica plantation.
Methods Pinus sylvestris var. mongolica with different tree ages (23, 33 and 44 a) in Mau Us Sandy Land were targeted. The species compositions, interaction and main drivers of ectomycorrhizal fungi at the early growth period (Apr.), vigorous growth period (July) and the late growth period (Sept.) were identified by field investigation and sampling, illumina high-throughput sequencing and molecular ecological network analysis, respectively.
Important findings 1) The growth period had significant effect on the richness and Chao1 index of ectomycorrhizal fungi, with significantly higher Chao1 and Simpson diversity index in vigorous growth period than in early and late growth periods. The tree age had no significant effect on the ectomycorrhizal fungal diversity index. 2) In Mau Us Sandy Land, the ectomycorrhizal fungi associated with P. sylvestris var. mongolica were identified into 2 phyla, 4 classes, 7 orders, 18 families, and 28 genera. The dominant genera were Tomentella, Inocybe and Geopora at the early, vigorous and late growth periods, respectively. The Tomentella and Inocybe were the common indicator genera in both vigorous and late growth periods. 3) At the end of the growing season, the nodes and edges of the ectomycorrhizal fungal networks were the largest, indicating that the ectomycorrhizal fungal community was more complex and there was strong interaction between fungal species. 4) Soil pH and annual relative humidity were the key environmental factors that significantly affected the community composition of ectomycorrhizal fungi. Our results demonstrated that the diversity and community composition of ectomycorrhizal fungi were affected by the growth stage and tree age, with a stronger effect in growth stage than in tree age. The seasonal dynamic distribution of ectomycorrhizal fungi mainly depended on soil properties and climate conditions.

Aims Soil respiration is one of the most critical components of carbon cycle in terrestrial ecosystems. The study on temporal dynamics of soil respiration and its linkage with environmental factors in desert steppes under changing precipitation can provide data supports for a deep understanding of the regulatory mechanisms of key carbon cycling processes in fragile ecosystems.
Methods A field experiment involving five precipitation treatments (50% reduction, 30% reduction, natural, 30% increase, 50% increase) was set up in 2014 in a desert steppe in Ningxia. The temporal dynamics of soil respiration rate were explored during the growing season (from June to October) in 2019, and the relationships between soil respiration rate and soil properties and plant characteristics were analyzed.
Important findings Soil respiration rate showed a seasonal variation of an increasing and a decreasing trend across the growing season, with the maximum values (2.79-5.35 μmol·m^-2·s^-1) occurring in late July or early August. Compared with the natural condition, 30% reduction in precipitation did not result in a significant effect on soil respiration rate, reflecting the adaptability of soil respiration to moderate drought. Overall, 50% reduction in precipitation reduced soil respiration rate, whereas increased precipitation (especially the 30% increase) enhanced soil respiration rate, and this positive effect was especially obvious in the early growing season (June to July). Soil respiration rate had a significantly exponential relationship with soil temperature and a significantly linear relationship with soil water content. Soil physicochemical property had a highly independent explanatory power for soil respiration rate (R² = 0.36), and its effect was highly correlated with soil biological property and plant diversity (R² = 0.31). Precipitation could affect soil respiration rate either directly or indirectly through the influences on soil biological property and plant biomass. The results indicated that a moderate increase in precipitation could accelerate soil respiration by alleviating soil water limitation, stimulating soil enzyme activity, promoting microbial activity and plant growth in the desert steppe, and that an extreme increase in precipitation would lead to a decrease in soil permeability and a hindrance to microbial metabolic activity, thus inhibiting soil respiration.

Arid and semi-arid ecosystem areas, which constitute an important component of the global land surface, act to regulate the long-term trend and interannual variations in global carbon and water cycles. Previous studies on the mechanisms underlying ecosystem carbon and water cycling and the development of relevant data products focus primarily on forest, grassland, and cropland ecosystems, with few research attention given to semi-arid shrublands. This research gap hinders the evaluation and projection of ecosystem functions at the regional scale. Since 2011, we used the eddy covariance technique to make continuous in situ measurements of carbon, water and heat fluxes in a shrubland ecosystem at Yanchi Research Station, the Mau Us Sandy Land. Data processing steps mainly included data collection, post-processing of raw data, quality control, gap-filling and carbon flux partitioning. We produced flux and micro-meteorological datasets at half-hourly, daily, monthly, and annual temporal resolutions for the years 2012-2016, and analyzed the overall quality of the datasets in terms of the proportion of valid data and the energy balance closure of flux measurements. Results showed: (1) After quality control, the proportion of valid data for half-hour net ecosystem CO₂ exchange (NEE), latent heat flux (LE), and sensible heat flux (Hs) was 56.23%-62.19%, 79.40%-94.12%, and 77.56%-91.27%, respectively. (2) Annual and monthly energy balance ratio ranged 0.78-0.83 and 0.59-1.19, respectively. (3) The energy balance closure estimated using the “ordinary least squares” regression method showed that the interannual and seasonal variations in the slope of regression curves varied with a range of 0.73-0.79 at interannual scale and 0.73-0.92 at seasonal scale, respectively. These results indicate that our datasets have a high proportion of valid data and a reasonable energy balance closure, and thus can be used in studies related to ecosystem processes and functions at varing spatio-temporal scales.