Aims Australian alpine ecosystems currently experience high precipitation in the snow-free season, but they are predicted to experience drier conditions under climate change. We observed high mortality of the dominant alpine grasses following drought in 2007. Our aims were as follows: to test the involvement of plant-available water (PAW) and other environmental variables in grass mortality in the field; to detect possible species differences in drought response and to link soil moisture to precipitation using soil properties and climate data.
Methods The dominant tussock grasses of the Australian alpine zone, Poa hothamensis var. hothamensis N.G. Walsh, P oa hiemata Vickery and P oa phillipsiana Vickery (Poaceae), all exhibited mortality following drought in the Bogong High Plains, Victoria, Australia in 2007. PAW was calculated using soil water potential measurements, and past drought occurrence was modelled using climate data. We then tested the effects of PAW and soil depth on grass survival both at a large spatial scale spanning the elevational range of the alpine zone and at a smaller scale. Poa hothamensis and P. phillipsiana were compared in a common-garden experiment to test drought tolerance.
Important findings Poa hothamensis survival was predicted by dry-season PAW at the small spatial scale; at the large scale, soil depth and elevation were more important predictors of P. hothamensis survival, but dry-season PAW predicted P. hiemata survival. Common-garden experiments supported field observations that P. hothamensis is more drought-sensitive than is P. phillipsiana. We also present a simple polynomial relationship between rainfall and field soil moisture, which predicts that the alpine soils dry below wilting point several times a year. We suggest the timing of long rain-free periods may be more important than their duration.

Aims To determine if an experimentally applied anomalous weather year could have effects on species composition and community structure that would carry over into the following year.
Methods We conducted a field experiment applying two levels of temperature (ambient and +4°C) and two levels of precipitation (ambient and doubled) and followed cover of plant species during the treatment year and one post-treatment year. Data analysis included ordination analysis, examination of species frequency distributions and comparison of cover of functional groups and individual species.
Important findings A drought during the summer and fall of the treatment year resulted in significant differences in community structure between the 2 years. C₃ and winter annual species were depressed in the spring of the second year following the dry autumn. Species richness and legume cover increased in the second, wetter, year. Treatments caused no overall differences in community structure but did alter the dominance hierarchy of species among treatments as well as years. Warming decreased relative cover of winter annuals and early spring-flowering species but increased other annuals. Warming and double precipitation together increased cover of C₄ perennial graminoids. In particular, the warming and precipitation treatments both increased the abundance of Andropogon gerardii, not individually altering the dominance hierarchy but together nearly doubling the relative cover of A.gerardii, making it the most abundant species in the combined treatment, while the cover of Bromus arvensis, the former dominant, decreased by 25%. The following year, Andropogon relative cover increased further in the former warmed plots, becoming dominant in both the formerly warmed and warmed plus double precipitation treatments. The year following treatments also saw an increase in relative cover of summer-blooming species in the formerly warmed plots and differences among the former treatments in species richness of functional groups. If the effects of one anomalous year on plant abundance can carry over into the following year, several warm years could have a significant impact on plant community structure.

Aims Forest growth and recruitment and their relationships to climate are complex. The aims of our study are (i) to examine the patterns in tree radial growth and recruitment along an altitudinal gradient in Mt. Everest region and (ii) to identify the climatic factors that are responsible to the observed patterns in tree growth and recruitment.
Methods Four plots, each 30 × 60 m in size, were established from the lower to upper limits of the eastern Himalayan fir forest in Dingjie County of the Mt. Everest Nature Reserve, China. Dendrochronological techniques were applied to obtain information about the radial growth and age of the trees in the plots. Correlation analysis was used to identify the relationships between radial growth and recruitment of trees and climatic variables, i.e. monthly mean temperature and Palmer Drought Severity Index (PDSI). The population age structure was analyzed to investigate the recruitment history.
Important findings The timberline plot was characterized by significant tree recruitments in the recent three decades and sporadic recruitments in earlier periods. The other three plots showed recruitment pulses during 1880–1910 in Plot 3?700 m, during 1870–80 and 1920–30 in Plot 3?520 m and during 1900–40 in Plot 3?410 m. The recruitment of fir trees in the timberline was sensitive to summer (June–September) temperatures, but it was mainly controlled by episodic disturbances in lower altitudes. Fir radial growth in the upper two plots was positively correlated with previous winter and current August temperature. Fir radial growth at the two lower plots was positively correlated with PDSI from previous September to current September.

Aims In recent years, coastal mangroves have been frequently affected by large disturbances (cyclones, hurricanes, flooding and tsunamis) and post-disturbance vegetation is often dominated by small stature mangrove, mangrove-associate and non-mangrove species potentially affecting ecosystem functioning. Knowledge on the processes of mangrove vegetation development and recovery (succession) following normal and large disturbances will benefit practitioners in designing robust ecosystem management/restoration plans. Here we propose a conceptual model of disturbance-mediated succession in mangroves.
Methods Based on field observations and species' life history traits, we develop conceptual models of mangrove succession under normal disturbance regime and recently experienced increased frequency of large disturbances. We evaluate our conceptual models by conducting a scenario testing experiment.
Important findings We suggest two predominant processes affecting mangrove succession after disturbance: propagule limitation due to damage of seed producing mature trees and dispersal barrier resulting from biological invasion associated with large disturbance. We argue that large disturbances affect mature trees more than the small-stature non-tree (shrubs, herbs and climbers) species creating a larger propagule shortage for mangrove tree species than non-tree species. Secondly, large disturbances facilitate invasion of free-floating aquatics, which may interfere with the flow-facilitated propagule dispersal and seedling establishment of mangrove species. In a scenario testing experiment, we have shown that similar levels of disturbance impact vegetation development and recovery differently depending on the presence or absence of invasive species. We conclude that since biological invasion is one of the major drivers of post-disturbance mangrove succession, the dimension of biological invasion should be included in prediction, management and restoration of mangrove forests.

Aims In view of the growing interest in modelling the potential spread of invasive species, prediction of plant invasiveness on the basis of native range size holds considerable promise. Our objective was to use a simple model to evaluate whether a wider native range predisposes plant species to become invasive in non-native regions and to easily identify potential invaders on this basis. The Kashmir Himalayan alien flora, of which a large proportion is native to Europe, was used to test this model.
Methods The Kashmir Himalayan alien flora comprises 436 species of vascular plants at different stages of invasion. We focussed on plant species at two critical invasion stages (sensu Colautti and MacIsaac 2004), i.e. Stage II (species that are just at the earliest phase of introduction) and Stage V (species that are widespread and dominant in the invaded region and are thus considered invasive). We used the territorial distribution in Europe (number of countries) as a surrogate for the native range size of plants of European origin.
Important findings Using a subset of 88 species, for which information on the native European range was available, we showed that a large proportion (68%) of Stage II species growing in the Kashmir Valley had a relatively restricted European range (present in ≤20 countries); on the other hand, 77% of Stage V species had an extensive native range (present in>20 countries). We consequently hypothesized that 14 Kashmir Himalayan Stage II species of European origin that are distributed in>20 European countries are at risk of becoming future invaders in Kashmir. On the other hand, those Kashmir Himalayan Stage II species of European origin distributed in ≤20 European countries are less likely to become invasive. Although this analysis is quite simple, the data suggest that a wider native range is a good predictor of plant invasiveness and could be used as a simple and low-cost early warning tool in predicting potential invasive species.

Aims Possible shifts in the phenotypic performance along invasive plants' spreading route are rarely examined due to the discontinuous and incomplete records of exotic species. As the invasion history of common ragweed (Ambrosia artemisiifolia L.) is well documented in Hungary, its residence time is known for each location. By sampling a sequence of older to more recently established populations, we aimed to determine (i) whether there are phenotypic divergences along the historical spreading route of A.artemisiifolia; (ii) which traits are under selection during the invasion process and (iii) the extent of maternal effects on the individual's performance.
Methods We used a hierarchical sampling design to collect seeds from 64 individuals belonging to eight sites in four residence time categories (seven populations along the historical spreading route of ragweed in Hungary and one recently invaded site in Romania). We selected four large and four small individual plants in each population to control for maternal effects. The offspring were reared in a common garden located in Romania. Five vegetative phenotypic traits were measured at the end of the experiments and used in the subsequent analysis (plant height, basal diameter, number of secondary axes, length of the longest secondary axis and biomass). To summarize the variation of these highly correlated traits, principal component analysis was performed first and then the important components were used in linear mixed effect models.
Important findings The residence time categories were significantly distinguished by the first component, which compresses the variation of all five measured traits. The measures gradually decrease from populations with the longest residence time (introduced more than 65 years ago) towards the most recently established populations (established less than 30 years ago). These differences might reflect the invasion history of the populations: the longer the residence time the higher the chance to develop relevant traits beneficial in invasion process. The size of the mother plant significantly influenced not only the seed mass (inversely) but also the adult performance of its offspring (directly).

Aims Numerous studies have showed that the balance between negative and positive plant–plant interactions shifted along environmental gradients. But little is known how the positive or negative plant–plant interactions varied with temporal fluctuating habitat conditions and plant ontogenetic phases.
Methods In a 2-year experiment, the four perennial grasses (Kobresia humilis, Stipa aliena, Elymus nutans and Saussurea superba) were grown under four interaction treatments (no root or shoot interaction, only shoot interaction, only root interaction, root and shoot interaction). Intensity of above- and belowground interactions is proposed to vary with the fluctuation of seasonal climatic conditions and soil available nutrients. Here we report measurements of above- and belowground interactions during entire growing season. Correlation between plant interaction intensity and seasonal soil available N as well as habitat climate conditions was also performed.
Important findings Our experiment found that root interactions had negative effect on plant growth for the four species during growing season. However, both negative and positive shoot interactions occurred among the four species. Despite there being shoot facilitative effect for E. nutans and S. superba, the full interaction was negative, suggested that root interaction take more important role on plant growth than that of shoot interaction. The interaction between root and shoot effect varied as a function of species identity and growth phases. The weak correlation of plant interaction intensity to habitat environmental factors suggested that plant ontogenetic characteristics may be primary factors causing temporal variation in plant interaction.

Aims Combining field data analysis and modeling, this study investigates factors influencing the diurnal boundary layer (BL) development in boreal forest.
Methods Field data analysis used both air sounding and surface flux measurements collected during the Boreal Ecosystem–Atmosphere Study field campaigns in central Canada. Model study applied a non-local transilient turbulence theory (TTT) to simulate the impact of the heterogeneous boundary conditions together with initial conditions on the BL development at the Candle Lake and Thompson release sites over boreal forests. Boundary conditions were characterized by the integrated surface flux measurements from different forest stands. The lake effect was included in constructing the surface fluxes at Candle Lake release site.
Important findings Analyses of serial upper air sounding data and tower flux data indicate strong linear impacts of surface sensible heat forcing on the diurnal BL development above boreal forests. The regression slopes on the relationship between the BL development and the surface fluxes reflect the influences of initial boundary conditions to the BL developments. Both the modeled and the measured diurnal BLs show that lakes reduce sensible heat flux, leading to a shallower boundary in Candle Lake than in Thompson. Comparison of the model results and field measurements on the BL profiles indicates that the TTT model has the capability to simulate the BL development above boreal forests for sunny, rainfall or cloudy days. This study demonstrates the importance of lake on surface fluxes and BL development. The modeling effort shows the potential to couple the transilient theory with a land surface process model to study land surface and atmosphere interaction in boreal forest.

Aims Road effects from maintenance and traffic have the potential to alter plant communities, but the exact relationships between these effects and changes in plant community composition have not often been studied in diverse environments. To determine the direction and level of community composition changes in saline environment due to road effects, we conducted a study along roads of different ages and in nearby non-road (i.e. natural) areas in the Yellow River Delta, China. Additionally, to potentially elucidate the mechanisms underlying the changes in the richness and composition of plant communities along roads, we evaluated physiochemical changes in soil of roadside and non-road areas.
Methods Floristic and environmental data were collected along roadside of different ages and nearby non-road areas. To evaluate plant communities at each site, six 2 m × 2 m quadrats were placed at 3-m intervals along roads and six quadrats were arranged randomly in non-road areas. To determine the difference in plant community composition between roadside and non-road areas, we measured species richness and the abundance of each species, examined species turnover and floristic dissimilarity between the two areas and positioned plant species and sites in an abstract multivariate space. Plant community (species richness, percentage of halophytes) and soil physicochemical properties (pH, salinity, moisture content, bulk density, nitrate and ammonium nitrogen concentration) were compared between roadside and non-road areas (young roadside vs. corresponding non-road areas, old roadside vs. corresponding non-road areas) by using t -tests. Classification and ordination techniques were used to examine the relationship between vegetation and related environmental variables in both roadside and non-road areas.
Important findings For both the young and old roadside areas, species richness in roadside areas was significantly higher than in non-road areas and high floristic dissimilarity values indicated that roadside and non-road areas differed greatly in community composition. In both the young and old roadside areas, the plant communities in roadside areas had lower percentages of halophytes than non-road communities. Correspondence analysis and two-way indicator species analysis showed that halophytes dominated in the non-road areas, while a number of typical non-salt-tolerant species dominated in the roadside areas. Compared to non-road areas, activities associated with roads significantly decreased soil moisture, bulk density and salinity and increased soil pH and nitrate content. Forward selection for the environmental variables in canonical correspondence analysis showed that soil salinity was the most important factor related to the variation of species composition between roadside and non-road areas. Our study demonstrates that road effects have a significant impact on the associated vegetation and soil, and these changes are consistent across roads of different ages in our system.

Aims Vegetation type is important in determining variations in soil carbon (C) efflux under grassland managements. This study was conducted to examine the effects of mowing and nitrogen (N) addition on soil respiration and their dependences upon vegetation types in an oldfield grassland of northern China.
Methods Soil respiration, temperature, moisture and aboveground net primary productivity (ANPP) and belowground net primary productivity (BNPP) were examined in response to mowing and N addition among the three patches dominated by different species (named as grass, forb and mixed patches, respectively) in the growing seasons (May–October) from 2006 to 2008.
Important findings Across the 3 years, soil respiration in the grass patch was greater than those in the forb and mixed patches, which could have been ascribed to the higher soil moisture (SM) in the grass patch. Mowing had no impact on soil respiration due to unaltered SM and plant growth. Soil respiration was stimulated by 6.53% under N addition, and the enhancement was statistically significant in 2006 but not in 2007 or 2008 because of the limited water availability in the later 2 years. There were no interactive effects between mowing and N addition on soil respiration. Soil respiration showed positive dependence upon SM, ANPP and BNPP across plots. The results suggest that soil water availability and plant growth could be the primary factors in controlling the temporal and spatial variations in soil respiration and its response to different treatments. Our observations indicate that grassland managements (i.e. mowing for hay once a year) may have little influence on soil respiration of the oldfield grassland in northern China.

Aims Although stem CO₂ efflux is critical to ecosystem carbon and energy balance and its feedback to future climate change, little information is available on stem CO₂ efflux and its responses to temperature, especially in subtropical China. This study aims to (i) evaluate the temporal and spatial variations of stem CO₂ efflux of three species, including oak (Quercus acutissima Carr.), masson pine (Pinus massoniana Lamb.) and loblolly pine (Pinus taeda Linn.) in subtropical China and (ii) analyze the temperature sensitivity of stem CO₂ efflux in the three species based on 2-year field measurements.
Methods We measured stem CO₂ efflux and stem temperature (at 3 cm depth) of the three species using the horizontally oriented soil chamber technique from September 2008 to August 2010. We also conducted a 24-h measurement to examine the diurnal variation of stem CO₂ efflux in three consecutive days in April 2009.
Important findings The temporal dynamics of stem CO₂ efflux followed the change of the stem temperature in a 3-cm depth with a bell-shaped curve in the three species. Stem temperature explained 77–85% of the seasonal variations of stem CO₂ efflux over the entire study period in the three species. The temperature sensitivity (Q 10) of stem CO₂ efflux was obviously different among the three species with higher Q 10 value found in oak (2.24) and lower values in the coniferous species (1.76 and 1.63). Our results also showed that the Q 10 values of stem CO₂ efflux in all the three species were lower in the growing season than that in the non-growing season, indicating that the growth and maintenance respiration had different temperature responses. Moreover, we found that the temperature-normalized stem CO₂ efflux (R 10) changed greatly between the growing and non-growing seasons in oak and masson pine, but not in loblolly pine. Additionally, we also found that in the non-growing season, the principal factor responsible for the spatial variation of stem CO₂ efflux among the 15 sampling trees was sapwood volume, whereas in the growing season, stem CO₂ efflux was closely related to annual dry-matter production in the three subtropical species.

Aims Ecosystem carbon models often require accurate net ecosystem exchange of CO₂ (NEE) light-response parameters, which can be derived from the Michaelis–Menten equation. These parameters include maximum net ecosystem exchange (NEE max), apparent quantum use efficiency (α) and daytime ecosystem respiration rate (R e). However, little is known about the effects of land conversion between steppe and cropland on these parameters, especially in semi-arid regions. To understand how these parameters vary in responses to biotic and abiotic factors under land conversions, seasonal variation of light-response parameters were evaluated for a steppe and a cropland of Inner Mongolia, China, during three consecutive years (2006–08) with different precipitation amounts.
Methods NEE was measured over a steppe and a cropland in Duolun, Inner Mongolia, China, using the eddy covariance technique, and NEE light-response parameters (NEE max, α and R e) were derived using the Michaelis–Menten model. Biophysical regulations of these parameters were evaluated using a stepwise regression analysis.
Important findings The maximum absolute values of NEE max occurred in the meteorological regimes of 15°C ≤ T a < 25°C, vapor pressure deficit (VPD) < 1 KPa and 0.21 m 3 m ? 3 ≤ volumetric soil water content at 10 cm (SWC) < 0.28 m 3 m ? 3 for both the steppe and the cropland ecosystems. The variations of α and R e showed no regular variation pattern in different T air, VPD and SWC regimes. Under the same regime of T air, VPD and SWC, the cropland had higher absolute values of NEE max than the steppe. Canopy conductance and leaf area index (LAI) were dominant drivers for variations in NEE light-response parameters of the steppe and the cropland. The seasonal variation of NEE light-response parameters followed the variation of LAI for two ecosystems. The peak values of all light-response parameters for the steppe and the cropland occurred from July to August. The values of NEE light-response parameters (NEE max, α and R e) were lower in the driest year (2007). Seasonally averaged NEE light-response parameters for the cropland surpassed those for the steppe. Land conversion from steppe to cropland enhanced NEE light-response parameters during the plant growing period. These results will have significant implications for improving the models on regional NEE variation under climate change and land-use change scenarios.