Previous studies showed that moss stoichiometric characteristics were influenced by moss patch size, shrubs and the environment in the desert. The study of moss stoichiometry in different spatial distribution areas is crucial for an understanding of growth and adaptation strategy of the mosses. In this study, the dominant moss (Syntrichia caninervis) of biological soil crusts and soil under the moss patches in the Gurbantunggut Desert were selected to determine their stoichiometry in different dunes and sites. Moss stoichiometry and soil available nutrients were significantly influenced by different distribution areas except for moss C. The N_aboveground vs. N_belowground' P_aboveground vs. P_belowground and K_aboveground vs. K_belowground scaling exponents of moss were 0.251, 0.389 and 0.442, respectively. The N vs. P scaling exponents were 0.71, 0.84 in above- and below-ground parts of moss. Moss stoichiometry was disproportionately distributed in the above-ground and below-ground parts. Moreover, moss N, P and K elements were influenced by mean annual precipitation (MAP), longitude and soil nutrients. The nutrients of moss were affected by spatial distribution, mean annual temperature (MAT), MAP and soil nutrients. The growth of moss was limited by N element in the temperate desert. This study provides the stoichiometric characteristics of C, N, P and K of moss at different spatial scales and explores their relationships with environmental variables, which can help understand nutrient patterns and utilization strategy of N, P and K, and their potential responses to global climate changes in desert.

Plant litter decomposition is critical for the carbon (C) balance and nutrient turnover in terrestrial ecosystems, and is sensitive to the ongoing anthropogenic biologically nitrogen (N) input. Previous studies evaluating the N effect on litter decomposition relied mostly on short-term experiments (<2 years), which may mask the real N effect on litter decomposition. Therefore, long-lasting experiments are imperative for the overall evaluation of the litter decomposition dynamics under N enrichment. We conducted a relative long-term (4-year) N-addition experiment with N levels ranging from 0 to 50 g N m^-2 yr^-1 to identify the potential abiotic and biotic factors in regulating the decomposition process of litterfall from the dominant species Leymus chinensis. The results showed a consistent decrease of decomposition rate with increasing N-addition rates, providing strong evidence in support of the inhibitory effect of N addition on decomposition. The N-induced alterations in soil environment (acidification and nutrient stoichiometry), microbial activity (microbial biomass and enzyme activity), changes of litter quality (residual lignin and nutrient content) and plant community (aboveground productivity and species richness) jointly contributed to the lowered decomposition. During the whole decomposition process, the changes of litter quality, including accumulation of lignin and the concentrations of nutrient, were mainly driven by the soil and microbial activity in this N-enriched environment. The findings help clarify how increasing N input rates affect long-term litter decomposition, and advance the mechanistic understanding of the linkages between ecosystem N enrichment and terrestrial C cycling.

Rodents attack oak (Quercus wutaishanica) seeds based on their sizes and frequencies before germination. However, the predation of oak seeds post-germination (seedling cotyledons) is not well studied. Here, we not only tested the preference of rodents for Q. wutaishanica seedling cotyledons based on the frequency of large- versus small-seeded (FLS), but also evaluated the effects of predation on seedlings growth and survival in different habitats. We transplanted seedlings with the FLS set as 9:1, 7:3, 5:5, 3:7 and 1:9, respectively, in the forest gap and under the canopy in the Liupan Mountains National Nature Reserve in Ningxia Hui Autonomous Region, Northwest China. The results showed that: (i) in 1-7 days after transplanting seedlings, rodents prefer the cotyledon of large-seeded seedings while small-seeded seedlings were preferred in 8-60 days, and the positive frequency-dependent predation was observed. (ii) The cotyledons were preyed on, the apical buds were bitted off, and the whole seedlings were uprooted, which mostly occurred under the forest canopy. At the end of a growing season, the survival rate of seedlings in the forest gaps was more than twice that under forest canopies. (iii) If cotyledons were preyed on, the growth of Q. wutaishanica seedlings would not be affected, but the seedlings growth was severely inhibited when the apical bud was bitten off. These results not only provide new insights into the coexistence between rodents and seedlings of different phenotypes, but also reveal the ecological characteristics of deciduous Quercus regeneration.

In legume-based grasslands, legumes supply the sward with nitrogen (N) from biological N₂-fixation. Previous studies in silvopastoral systems have emphasized that legume proportions decline near trees which would cause spatial variation in the N supply and the concentration of N in the harvested herbage of grass swards between trees. In natural ecosystems, it was shown that the N resorption efficiency (NRE) was increased by nutrient limitation, hence, near trees where legumes are scarce. Therefore, we tested the hypothesis that the NRE is increased near trees and compensates for the loss of legumes with respect to N concentration in live herbage. Two vegetation compositions consisting of legume-based grass swards were analysed across positions between tree lines in an alley cropping system established 5 years prior to this study. Legume proportion declined up to 45% towards the tree line, and also N concentration in live and dead herbage but on average only declined by 15%. As a consequence of a reduced N concentration and also herbage mass close to trees, the N yield decreased by up to 50%. Despite the loss of legumes near trees, the NRE was unaffected by the tree line irrespective of vegetation composition. Further, the results indicate that internal N cycling of managed grass swards in silvopastoral grassland with short defoliation intervals between harvests is of lesser relevance than in unmanaged ecosystems. Legume proportions control the N concentrations of live and dead herbage irrespective of tree shading in silvopastoral systems.

Soil microorganisms in tropical forests can adapt to phosphorus (P)-poor conditions by changing the activity ratios of different types of phosphatases. We tested whether microorganisms in P-poor tropical forest soils increased the phosphomonoesterase (PME) to phosphodiesterase (PDE) activity ratio, because a one-step enzymatic reaction of monoester P degradation might be more adaptive for microbial P acquisition than a two-step reaction of diester P degradation. A continuous 10-year P addition experiment was performed in three tropical forests. The activities of PME and PDE, and their ratio in soil, were determined under the hypothesis that the P-fertilized plots where P shortage is relieved would have lower PME:PDE ratios than the unfertilized controls. We demonstrated that long-term P addition in tropical forest soil did not alter the PME:PDE ratio in primary and secondary forests, whereas P fertilization elevated the PME:PDE ratio in planted forest. These results were in contrast to previous results. The long-term, large-scale P fertilization in our study may have reduced litter- and/or throughfall-derived PDE, which negated the lowered PME:PDE ratio via exogenous P inputs.

To clarify the role of tree characteristics and slope positions in the time lag between tree stem sap flux density (J_s) and solar radiation (R_s). Plants of different diameter classes in a Larix olgensis near-mature forest (31 years old) in the hilly area of the Sanjiang Plain were used. The relationships between the time lag J_s-R_s and tree characteristics, adjacent tree characteristics and slope positions were evaluated. Though both J_s and R_s exhibited diurnal variation, they were not synchronized, thus leading to a time lag between J_s and R_s. During the growing season, the change in J_s lagged behind the change in R_s by 21.1 ± 6.9 min. Compared with tree height and crown width, the time lag J_s-R_s was more dependent on diameter at breast height (DBH). The time lag between J_s and R_s showed a linear increase with DBH. Compared with the characteristics of neighboring trees, the time lags J_s-R_s were more dependent on their own tree characteristics. A significant relationship was not observed between the time lag J_s-R_s and soil volumetric water content. The effects of tree characteristics, adjacent tree characteristics and slope positions on the formation of the time lag J_s-R_s were compared. The time lag of J_s on R_s was mainly controlled by the tree characteristics (DBH). DBH is an important factor that affects the time lag between J_s and R_s under sunny conditions during the growing season of L. olgensis.

Previous studies have shown that plant-pollinator mutualistic interactions experience highly interannual variation. Given that pollinators often move across multiple plant species, the plant-plant interactions that take place via heterospecific pollen (HP) transfer may also vary temporally, which could have important implications for floral evolution and community assembly. Here, we evaluated the interannual variation in plant-pollinator networks and plant-plant heterospecific pollen transfer (HPT) networks of a subalpine meadow community in Southwest China for three consecutive years. The interactions largely varied among years for both network types. The composition of donor-species HP deposited on the plants varied less than did the visit composition of the pollinators, suggesting that HP could be transferred from identical donor species to recipient species through different shared pollinators among years. The plant species were at more similar positions in the HPT network than they were in the plant-pollinator network across years. Moreover, the more generalized plant species in the plant-pollinator network tended to export their pollen grains and more strongly influence HPT. We evaluated the relatively stable structure of the HPT network compared with the plant-pollinator network, which represents an important step in the integration of plant-pollinator and plant-plant interactions.

The assumption that climatic growing requirements of invasive species are conserved between their native and non-native environment is a key ecological issue in the evaluation of invasion risk. We conducted a growth chamber experiment to compare the effect of water regime and temperature on the growth and mortality of native and invasive populations of common gorse seedlings (Ulex europaeus L.). Seeds were sampled from 20 populations of five areas from both native (continental France and Spain) and non-native areas (New Zealand, Canary and Reunion islands). The seedlings were grown over 36 days in two temperature treatments (ambient and elevated) combined with two water treatments (irrigated or droughted). The elevated temperature (ET) was defined as the highest temperature observed at the niche margin in the different countries. While ET increased seedlings growth, the drought treatment increased mortality rate and limited seedlings growth. Under ET and drought, native populations showed a greater mortality rate (53%) than invasive populations (16%). Invasive seedlings also showed higher above- and belowground development than native ones under these constrained climatic conditions. While phenotypic plasticity did not differ between native and invasive populations, the difference between populations in terms of total dry mass could be related to differences in the climate of origin (precipitation in particular). Assessing the importance of phenotypic changes between populations within invasive species is crucial to identify the margins of their climatic distribution range and to highlight areas where management efforts should be concentrated in order to limit its spread.

Given the current scenario of climate change and anthropogenic impacts, spatial predictions of biodiversity are fundamental to support conservation and restoration actions. Here, we compared different stacked species distribution models (S-SDMs) to forest inventories to assess if S-SDMs capture emerging properties and geographic patterns of species richness and composition of local communities in a biodiversity hotspot. We generated SDMs for 1499 tree species sampled in 151 sites across the Atlantic Forest. We applied four model stacking approaches to reconstruct the plant communities: binary SDMs (bS-SDMs), binary SDMs cropped by minimum convex polygons (bS-SDMs-CROP), stacked SDMs constrained by the observed species richness (cS-SDMs) and minimum convex polygons of species occurrences (MCPs). We compared the stacking methods with local communities in terms of species richness, composition, community prediction metrics and components of beta diversity—nestedness and turnover. S-SDMs captured general patterns, with bS-SDMs-CROP being the most parsimonious approach. Species composition differed between local communities and all stacking methods, in which bS-SDMs, bS-SDMs-CROP and MCPs followed a nested pattern, whereas species turnover was most important in cS-SDMs. S-SDMs varied in terms of performance, omission and commission errors, leading to a misprediction of some vulnerable, endangered and critically endangered species. Despite differing from forest inventory data, S-SDMs captured part of the variation from local communities, representing the potential species pool. Our results support the use of S-SDMs to endorse biodiversity synthesis and conservation planning at coarse scales and warn of potential misprediction at local scales in megadiverse regions.

In Mongolia, overgrazing and the resulting degradation of rangelands are recognized as serious issues. To address rangeland degradation, we sought to develop a broad-scale vegetation classification of Mongolian rangeland communities focusing on regional characteristics. Moreover, we sought to clarify the spatial distributions of communities and the environmental drivers of the distributions. Between 2012 and 2016, we surveyed vegetation in 278 plots (each 10 m × 10 m) in different regions of Mongolia (43-50° N, 87-119° E) in plots where grazing pressure is low relative to adjacent areas. The data were grouped into vegetation units using a modified two-way indicator species analysis (TWINSPAN). We then explored the regional characteristics of species compositions and community distributions, as well as relationships between distributions and climatic variables. The modified TWINSPAN classified the vegetation data into three cluster groups, each of which corresponds to a particular type of zonal vegetation (i.e. forest steppe, steppe and desert steppe). The aridity index was identified as an important driver of the distributions of all cluster groups, whereas longitude and elevation were important determinants of the distribution of clusters within cluster groups. Western regions, which are characterized by higher elevation and continentality compared with eastern regions, have lower mean temperature and precipitation during the wettest quarter, leading to differences in species composition within cluster groups. Regional differences in species composition reflect differences in phytogeographic origin. Thus, the framework of species composition and distributional patterns in Mongolian rangeland communities was demonstrated in relation to climatic and geographical factors.

Plant density and nitrogen (N) availability influence plant survival and nutrient use strategies, but the interaction between these two factors for plant growth and the balance of elements remains poorly addressed. Here, we conducted experimental manipulations using Arabidopsis thaliana, with the combination of four levels of plant density and four levels of N addition, and then examined the corresponding changes in plant biomass production (indicated by total plant biomass and biomass partitioning) and nutrient use strategies (indicated by leaf N and phosphorus (P) stoichiometry). The biomass-density relationship was regulated by N availability, with a negative pattern in low N availability but an asymptotic constant final yield pattern at high N availability. Excessive N addition reduced plant growth at low plant density, but this effect was alleviated by increasing plant density. The root to shoot biomass ratio increased with plant density at low N availability, but decreased at high N availability. N availability was more important than plant density in regulating leaf N and P stoichiometry, with the increasing leaf N concentration and decreasing leaf P concentration under increasing N addition, resulting in a negative scaling relationship between these two elemental concentrations. Our results show that N availability and plant density interactively regulate plant biomass production and leaf stoichiometry of A. thaliana, and highlight that the interactive effects of these two factors should be considered when predicting plant growth behaviour under intraspecific competitive environments in the context of nutrient changes.

Tropical forests are characterized by vast biomass, complex structures and mega-biodiversity. However, the adaptation processes of these forests to seasonal water availability are less understood, especially those located in the monsoonal and mountainous regions of tropical Southeast Asia. This study used four representative tropical forests spanning from 2° N to 22° N in continental Southeast Asia to address dry-condition photosynthesis at the seasonal scale. We first provided novel and reliable estimations of ecosystem photosynthesis (gross primary production; GPP) seasonality at all four sites. As expected, both evergreen and deciduous seasonal forests exhibited higher GPPs during the rainy season than during the dry season. A bimodal pattern corresponding to solar radiation occurred in the GPP of the perhumid forest. The surface conductance (G_s) was consistently lower both in the dry season and during dry spells (DSPs) than during the wet season and non-dry spells. However, this did not prevent GPP from increasing alongside increasing irradiance in the perhumid forest, suggesting that other ecosystem physiological properties, for example, the light-saturated photosynthetic rate, must have increased, thus surpassing the effect of G_s reduction. Thus, perhumid forests could be defined as light-demanding ecosystems with regard to their seasonal dynamics. Seasonal forests are water-stressed ecosystems in the dry season, as shown by the reductions in GPP, G_s and related ecosystem physiological properties. At all four forest sites, we observed a lack of consistent adaptive strategy to fit the water seasonality due to the diversity in leaf phenology, soil nutrient availability, root depth and other potential factors.

The reproductive strategies of alpine plants are often altered by environmental changes caused by changes in the spatial distribution of the gradient. However, few studies have investigated whether reproductive patterns of the same species vary with elevation. Three natural populations of Primula atrodentata, which are distributed in the eastern Himalayas and have a long flowering period, were selected along the elevation gradients in Shergyla Mountain, Tibet, China. Morph ratio investigation, floral trait measurement, pollinator observation and manipulated pollination experiments were conducted to explore the changes in self-compatibility and floral traits associated with the selfing syndrome along elevation gradients. We found that the breeding system of the S-morph is facultative outcrossing, and that of the L-morph is obligatory outcrossing. We further found that with increasing elevation, the number of pollen and ovules, anther-stigma distance, and inbreeding depression index first increased and then decreased, whereas the seeds per fruit and seed-setting rate under hand self-pollination, pollen limitation and self-incompatibility index tended to decrease first, but then increased. In addition, pollinator diversity and visiting frequency were the highest at the middle elevation (4050 population), which can better explain the nonlinear change in self-fertility with elevation. Our findings provide insights into the evolutionary pattern of self-compatibility in alpine plants along elevational gradients.

Dominant species may strongly influence biotic conditions and interact with other species, and thus are important drivers of community dynamics and ecosystem functioning, particularly in the stressed environment of alpine grasslands. However, the effects of dominant species on the community stability of different ecosystems remain poorly understood. We examined the mechanisms underlying temporal stability (2014-2020 year) of aboveground productivity and community stability in four alpine grasslands (alpine meadow, alpine meadow steppe, alpine steppe and alpine desert steppe) of the northern Tibetan with different species composition and dominance. Our results showed that community stability was significantly higher in the alpine meadow than in the other three types of grasslands. This difference was mainly attributed to the higher compensatory effect and selection effect in the alpine meadows. Furthermore, dominant species strongly affected community stability by increasing dominant species stability and species asynchrony. However, species richness had little effect on community stability. Our findings demonstrate that dominant species, as foundation species, may play leading roles in shaping community stability in the alpine grasslands, highlighting the importance of conserving dominant species for stable ecosystem functioning in these fragile ecosystems under increasing environmental fluctuations.

Alpine forests in the eastern Tibetan Plateau are important ecological barriers in the upper reaches of the Yangtze River. However, due to continuous high-intensity harvesting, a large number of plantings, and the complete harvesting ban measures in recent decades, the forest tree species and age cohorts have become relatively homogenous, and the biodiversity and ecological functions have been reduced. To design effective forest management options to optimize forest structure and increase carbon sequestration capacity, Mao County in Sichuan Province was selected as the study site and six forest management options (harvesting, planting) of different intensities were tested using the LANDIS-II model to simulate and compare the differences in forest aboveground carbon sequestration rate (ACSR) between these options and the current management option over the next 100 years. Our results showed that (i) the different harvesting and planting intensities significantly changed the ACSR compared with the current management options; (ii) different communities responded differently to the management options, with the ACSR differing significantly in cold temperate conifers and temperate conifers but not in broad-leaved trees (P < 0.05); and (iii) a comprehensive consideration of forest management options at the species, community and landscape levels was necessary. Our results suggest that implementing a longer harvesting and planting interval (20 years) at the study site can maximize forest ACSR. This study provides an important reference for evaluating the ability of forest management options to restore forest ecological functions and increase carbon sequestration capacity and for selecting effective forest management programs in the eastern Tibetan Plateau.

Architectural plasticity in traits should be useful for understanding morphological and allocation plasticity at the whole-plant level and associated growth strategies of plants in dealing with variation in emergence time. To investigate how plants respond to emergence timing via dynamic architectural plasticity, we conducted a field experiment by sowing plants of Abutilon theophrasti on different dates to make them emerge in spring, late spring, summer and late summer, as four treatments of emergence time (ET1-ET4), and measured a series of mass and morphological traits in different vertical layers (with an inter-distance of 10 or 20 cm per layer) of plants at three stages. The ‘vertical arch' and ‘inverted-cone' shapes displayed at reproductive stage by plants emerged between spring and summer (ET1-ET3) and in late summer (ET4), respectively, revealed that they had contrasting timing schemes for vegetative and reproductive growth. Plants emerged earlier had relatively isolated vegetative and reproductive stages, while late-summer emerged plants had overlapped vegetative and reproductive stages due to accelerated leaf and reproductive growth simultaneously, as a result of shortened lifetime. In spite of this, plants emerged earlier were still able to adopt different optimal strategies in dealing with their corresponding circumstances. Architectural analysis should be a useful approach to better understanding diverse plant strategies in the context of varying environmental conditions.