Aims Process-based models are basic tools for predicting the response of forest carbon to future climate change. The models have commonly been tested for their predictions of spatial variation in forest productivity, but much less for their ability to predict temporal variation. Here, we explored methods to test the models with tree rings, using BIOME-BGC as an example.
Methods We used net primary productivity (NPP) data and tree rings collected from five major forest types along the altitudinal gradient of Mt. Changbai, northeast China, to test local-parameterized BIOME-BGC model. We first test the model's predictions of both spatial (Test 1) and temporal changes (Test 2) in productivity. Then we test if the model can detect the climatic factors limiting forest productivity during historical climate change, as revealed by dendroclimatic analyses (Test 3).
Important findings Our results showed that BIOME-BGC could well simulate NPP of five forest types on Mt. Changbai, with an r 2 of 0.69 between modeled and observed NPP for 17 plots along the altitudinal gradient (Test 1). Meanwhile, modeled NPP and ring-width indices were correlated and showed similar temporal trends for each forest type (Test 2). While these tests suggest that the model's predictions on spatial and temporal variation of NPP were acceptable, a further test that relate the correlations of modeled NPP with climate variables to the correlations of ring widths with climate (Test 3) showed that the model did not well identify the climatic factors limiting historical productivity dynamics for some forest types, and thus cannot reliably predict their future. Both dendrochronology and BIOME-BGC showed that forest types differed markedly in the climate factors limiting productivity because of differences in tree species and climate condition, and thus differed in responses to climate change. Our results showed that a successful prediction of spatial NPP patterns cannot assure that BIOME-BGC can well simulate historical NPP dynamics. Further, a correlation between modeled NPP and tree-ring series cannot assure that the limiting climatic factors for productivity have been correctly identified by the model. Our results suggest the necessity to test the temporal predictions of process-based models in a more robust way, and further integration of dendrochronology and biogeochemistry modeling may be helpful for this purpose.

Aims Biotic homogenization results in novel communities containing multiple exotic species. Many attributes that contribute to the dominance of invaders and increase invasibility of communities have been identified, but rarely have factors governing the dominance of co-occurring invaders been considered. Here we assess the presence and performance of five exotic and one native C₄ grass species across local (microsite) and more broad-scale variation in fire history, soil age, altitude and precipitation. We ask whether species show preferences for the same conditions, whether all are constrained by poor soil development, and how their performance changes when alone versus co-occurring. In addition to providing information useful to managers who want to control some of these species, such information can help to predict how distributions might change with climate change
Methods We measured percent cover and plant height of each grass species in 64 plots in each of 15 sites, located in young volcanic substrates across environmental gradients in Hawai'i Volcanoes National Park, Hawai'i, USA. Sites varied in fire history and soil age, across altitudinal and precipitation gradients. We estimated microsite heterogeneity by measuring rock cover and soil depth, where each species was present and in plots as a whole. We analyzed effects of possible controlling variables using chi-square tests and Generalized Additive Mixed Models.
Important findings Two species (Schizachyrium condensatum and Hyparrhenia rufa) were absent in shallow soil, restricted to more mesic sites and showed increasing performance with soil depth. By contrast, two other species (Andropogon virginicus and Melinis repens) tended to occur in shallow soil with high rock cover across a wide range of sites. One additional species (Melinis minutiflora) was weakly affected by soil depth and was the dominant at higher elevation (more mesic) and on burned sites, especially in old soil. This species was largely absent from lower elevation sites, where the more widespread and opportunistic congener, M. repens, and a drought tolerant native grass, Heteropogon contortus, were more abundant. The latter was confined to this low hot zone. Introduced C₄ grasses in this region of Hawai'i are distributed according to an interaction between soil requirements, climate (correlated with elevation), competitive ability and fire response. No one strategy can explain the distributions, which will likely persist unless climate changes toward warmer and drier conditions or new fires occur. The latter would promote dominance of M. minutiflora in more mesic sites, while the former would promote either native (H. contortus) or other exotic grasses (M. repens, A. virginicus).

Aims Topography has long been recognized as an important factor in shaping species distributions. Many studies revealed that species may show species–habitat associations. However, few studies investigate how species assemblages are associated with local habitats, and it still remains unclear how the community–habitat associations vary with species abundance class and life stage. In this study, we analyzed the community–habitat associations in a subtropical montane forest.
Methods The fully mapped 25-ha (500×500 m) forest plot is located in Badagongshan Nature Reserve in Hunan Province, Central China. It was divided into 625 (20×20 m) quadrats. Habitat types were classified by multivariate regression tree analyses that cluster areas with similar species composition according to the topographic characteristics. Indicator species analysis was used to identify the most important species for structuring species assemblages. We also compared the community–habitat associations for two levels of species abundances (i.e. abundant and rare) and three different life stages (i.e. saplings, juveniles and adults), while accounting for sample size effects.
Important findings The Badagongshan plot was divided into five distinct habitat types, which explained 34.7% of the variance in tree species composition. Even with sample size taken into account, community–habitat associations for rare species were much weaker than those for abundant species. Also when accounting for sample size, very small differences were found in the variance explained by topography for the three life stages. Indicator species of habitat types were mainly abundant species, and nearly all adult stage indicator species were also indicators in juvenile and sapling stages. Our study manifested that topographical habitat filtering was important in shaping overall local species compositions. However, habitat filtering was not important in shaping rare species' distributions in this forest. The community–habitat association patterns in this forest were mainly shaped by abundant species. In addition, during the transitions from saplings to juveniles, and from juveniles to adults, the relative importance of habitat filtering was very weak.

Aims Tropical and subtropical karst forests of south China are under increasing pressure from over-exploitation causing widespread habitat degradation and biodiversity loss. Previous research has demonstrated that topography, as a proxy for resource availability, plays an important role in shaping tree species distributions in tropical forests. However, the association between growth stages and habitats types has not been considered in this analysis. Our aim was to examine the differences among different habitat types to determine whether tree species show similar species–habitat associations at young and mature life stages.
Methods We used multivariate regression tree analysis to examined species–habitat associations among eight topographically defined habitats. The results were tested with a torus-translation test and canonical correspondence analysis (CCA) for 74 species in a 15 ha karst tropical seasonal rain forest at the Nonggang National Natural Reserve in south China. We considered two life stages (young and mature) of trees species that showed a positive association with topography.
Important findings We found marked differences in community characteristics and number of associations among the eight habitats. Of the 74 species subjected to torus-translation test, 63 had significant positive and 70 had significant negative associations with one or more of the eight habitats. Positive associations were more frequent in higher elevation habitats and negative associations were more frequent in lower elevation habitats. This suggests that edaphic and hydrological variation related to topography play important roles in habitat partitioning in heterogeneous karst forests. For the 63 tree species with significant positive associations to at least one habitat, 40 of them had the same positive association at young and mature life stages. The CCA revealed that the six topographic variables considered had consistent relationships with species distribution among all individuals and their two life stages. This indicates that most of the karst forest tree species show consistent associations with a single habitat throughout their life. We conclude that niche differentiation plays an important role in maintaining the diversity of this heterogeneous species-rich karst forest.

Aims We studied diversity, patterns of endemism and turnover of vegetation composition in dry inter-Andean valleys (DIAVs) where little is known about the influence of the abiotic drivers controlling plant species composition and occurrences, and the life forms that contribute most to α - and β -diversity, respectively. This study was focused on DIAVs located in the highlands (800–2800 m) around the Equator (1°N–5°S). The following questions were addressed: (i) what differences exist between endemic and non-endemic species in terms of species number, frequency and abundance? (ii) are patterns of α - and β -diversity correlated with latitude? (iii) what are the major environmental drivers controlling spatial patterns in species composition and occurrence?
Methods We established 63 transects of 5×100 m in areas with DIAV vegetation, impacted as little as possible by human activities. In each transect, all mature trees and shrubs were identified and counted to determine their density. The coverage of terrestrial herbs was estimated. Generalized additive models were used to quantify the relationship between α - and β -diversity with latitude. To record α -diversity, we used the exponential Shannon index. The S?rensen index was used to measure β -diversity or species turnover. We used canonical correspondence analysis to determine species composition and generalized linear mixed models to quantify simultaneously the determinants of species occurrence across species and sites. The models were evaluated using the Akaike information criterion. All analyses were run separately for trees, herbs and shrubs.
Important findings Overall, we found that shrubs was the most diverse life form, constituting 43% (12% endemic) of the 313 species, followed by herbs (36%; 11.5% endemic) and trees (20%; 5% endemic). Trees comprise mostly species with low abundance (e.g. <3 individuals ha^-1), which was also true for herbs, whereas shrubs showed a larger number of abundant species (>30 individuals ha^-1), including four endemic species. Most of the endemic species were locally rare, and most of them were restricted to southern valleys, suggesting low migration rates due to local barriers. We identified seven species distribution patterns that gave rise to high β -diversity in the dry inter-Andean valleys. The β -diversity of trees was the lowest. Herbs had the highest β -diversity, which increased steadily with geographic distance. Our results further highlight the influence of disturbance, water availability and low temperature on plant species composition and occurrence. We also found significant, contrasting patterns in responses to environmental drivers, when analyzing our data separately by life form. Our results show that analyzing diversity patterns separately by life form can shed new light on the mechanisms that have generated present-day patterns of plant diversity in DIAVs.

Aims Ecosystem respiration represents an important component of the carbon cycle. The response of respiration to climate change can have a significant effect on carbon sequestration in terrestrial ecosystems in the future when, according to climate scenarios, spring drought and consequent summer heavy rains are expected. Aims of our study were to determine the effect of repeated spring drought on biomass production and CO₂ efflux from a mountain grassland in Beskydy Mountains in the Czech Republic and to evaluate factors driving the differences among the study years.
Methods CO₂ efflux was measured on plots with ambient precipitation conditions (AMB) and on plots where drought conditions (DRY) in the first half of the growing seasons and consequent heavy rain were simulated in 2011–14.
Important findings The spring drought significantly decreased the amount of above-ground biomass sampled just after the simulated drought in all years except for 2014. On the contrary, the spring drought stimulated root production. The drought also resulted in a rapid decrease in CO₂ efflux. It was lower by up to 46% for the DRY treatment compared to AMB treatment. After the simulated drought period, differences in CO₂ efflux between the treatments gradually decreased. Simulated heavy rains in DRY resulted in fast but temporary increase in CO₂ efflux. We can assume that the future spring drought will have a significant effect on carbon balance of grassland ecosystems. The impact will depend on the length of the dry period and the time between the beginning of the growing season and the dry period.

Aims We aimed to improve the understanding of the carbon and nutrient physiological responses and adaptation of semi-arid grassland plants to environmental changes.
Methods We investigated plant leaf non-structural carbohydrate (NSC = soluble sugars + starch), nitrogen (N) and phosphorus (P) levels in an Inner Mongolian semi-arid grassland community treated with water, N and P additions for 8 years. Two dominant grasses (Agropyron cristatum (L.) Gaertn., Stipa krylovii Roshev.) and two forbs (Artemisia frigida Willd., Potentilla bifurca L.) were analyzed.
Important findings Water addition decreased plant leaf N and P concentrations, whereas N and P addition increased them, indicating that the semi-arid grassland studied suffers from a shortage of N and P supply. Both N and P addition decreased the levels of soluble sugars, starch and thus also NSC in plant leaves, which may be attributed to (i) increased carbohydrate consumption associated with a higher growth rate, and (ii) a dilution effect of greater plant size under N and P addition. Water addition tended to increase the leaf NSC levels both in the grasses (+9.2%) and forbs (+0.6% only), which may be a result of increased photosynthesis of plants with increased water availability. Under conditions of ambient and increased water supply in the present study, N addition resulted in an N/P ratio of>16 in the grasses but a significantly lower N/P ratio of <11 in the forb species. This finding implies that growth of the two grass species will be limited mainly by P availability but the forbs will still be mainly limited by N supply if N deposition, alone or in combination with summer precipitation, continues to increase as predicted in Inner Mongolia.

Aims Understanding how environmental factors and human activity drive plant community assembly remains a major challenge in community ecology. Two opposing processes, namely deterministic environmental filtering and nondeterministic elements such as interspecies competition, can be quantified by analyzing trait distributions in the community-assembly process.
Methods We examined the trait-mediated effects of environmental filtering and stochastic process and the distribution over time of nine traits related to vegetative growth, regenerative phase, dispersal capability, decomposition and interspecific competition in plant communities along a degradation gradient in the Xilin River Basin, Inner Mongolia, China. We analyzed the turnover of environmental trait filtering and the divergence/convergence of different traits along the degradation gradient.
Important findings Our results showed the following. (i) The patterns of trait distribution and filtering were strongly dependent upon the degradation gradient and trait types. Most traits were filtered intensely in degraded grasslands. (ii) Plants with two different strategies showed contrasting trait-distribution patterns. The traits that were related to biological matter cycling showed divergent patterns in highly degraded grasslands, while convergent patterns along the overall gradient were demonstrated in traits associated with other plant strategies. This suggests that the coexistence of multiple 'biological matter cycling-related niches' might be a basic structuring pattern of plant communities in our study area. (iii) The simultaneous occurrence of strong filtering and divergence revealed that environmental filtering does not necessarily prevent competition, and that different traits show different signatures.

Aims Community characteristics, such as disturbances and interspecific competition that affect the availability of microsites and resources, contribute to the success or failure of the establishment of exotic plant species. In particular, these two community characteristics may have adverse effects on plant emergence and survival, which are particularly important for population establishment and therefore it may be necessary to consider both these vital rates simultaneously when assessing demographic mechanisms. Here, we investigated the impacts of disturbance and interspecific competition on the establishment of a perennial invasive herb, Lupinus polyphyllus Lindl.
Methods Over the course of 2 years, we conducted an experiment in 10 populations of this species in Finland in which we manipulated the levels of soil disturbance. We recorded community characteristics (i.e. the number of vascular plant species, vegetation height, and the proportions of bare ground, litter and moss), and observed the emergence and survival of L. polyphyllus individuals in study plots.
Important findings A mild disturbance (breaking the soil surface mechanically) slightly increased seedling emergence but did not affect plant survival. Instead, an intense disturbance (vegetation and litter removal) had no effect on seedling emergence, although it significantly increased the proportion of bare ground and, consequently, seedling survival. Survival was not affected by the height of the surrounding vegetation, but both seedling emergence and plant survival increased with an increasing number of plant species in the study plots. These findings demonstrate that single disturbance events may considerably promote the establishment of invasive herbs, although the overall effect and demographic mechanisms behind the increased establishment are likely to vary depending on disturbance type. Moreover, our results suggest that species diversity per se may not be a crucial mechanism for locally preventing the establishment of exotic plants.

Aims Tiller recruitment of perennial grasses in mixed-grass prairie primarily occurs from belowground buds. Environmental conditions, such as temperature, soil moisture and grazing can affect bud outgrowth of both invasive and native perennial grasses. Differential bud outgrowth responses of native and invasive species to climate change and grazing could alter competitive interactions that have implications for future land management. The aims of this work were to (i) compare how spring temperature altered bud outgrowth of native Pascopyrum smithii (Rydb.) á. L?ve (western wheatgrass) and introduced Bromus inermis Leyss.(smooth brome), (ii) compare how watering frequency altered bud outgrowth of these two species and (iii) evaluate how clipping interacts with spring temperature or watering frequency to affect P. smithii bud outgrowth.
Methods Individual plants of B. inermis and P. smithii were harvested from North American mixed-grass prairie. Bud outgrowth from tillers of both species were evaluated under three spring temperature regimes (Average: 12, 18 and 24°C) and two watering frequencies (frequent and intermittent) in a growth chamber experiment. The response of P. smithii bud outgrowth to clipping was also examined.
Important findings Bromus inermis had more buds per tiller and initiated a greater proportion of these buds than P. smithii under all temperature and moisture conditions. Pascopyrum smithii bud development was reduced at 24°C. Intermittent watering did not significantly impact bud outgrowth of either species. Clipping increased P. smithii bud mortality and reduced its bud development for the 2-week period of the study. The robust vegetative reproductive capacity of B. inermis under a range of environmental conditions is a key mechanism enabling the expansion of B. inermis into P. smithii -dominated mixed-grass prairie in North America. Mixed-grass prairie dominated by P. smithii experiencing repeated defoliation may require longer recovery times and be more susceptible to B. inermis invasion due to the negative impact of grazing on P. smithii bud outgrowth. Successful tiller recruitment and establishment of native perennial grasses via the bud bank will be necessary for mixed-grass prairie to be resilient to climate change, plant invasions and grazing.

Aims Clonal plant species have the potential for high relative performance in heterogeneous environments, and this might increase the competitive ability and invasiveness of introduced clonal plant species. It was hypothesized that clonal species whose performance responds more to heterogeneity of a resource have higher competitive ability in habitats where this resource is more heterogeneous and that this relationship is stronger when other resources are less limiting.
Methods To test these hypotheses, the perennial clonal herb Alternanthera philoxeroides, which is invasive in China, was grown alone or with each of four clonal perennial, co-occurring herbs native to China, i.e. Alternanthera sessilis, Cynodon dactylon, Hemarthria altissima and Wedelia chinensis. Plants were given homogeneous or heterogeneous soil substrate crossed with low and high levels of soil moisture.
Important findings Effects of heterogeneity on the accumulation of mass and ramets and on competitive effect and response of A. philoxeroides differed between native species and interacted with effects of soil moisture. A. philoxeroides reduced the final total mass or ramet number of the native species except A. sessilis, and the negative competitive effects on H. altissima and C. dactylon were more pronounced in heterogeneous than in homogeneous soil. Competitive response of A. philoxeroides was more negative to A. sessilis than to the other native species. Across native species, the competitive response of A. philoxeroides was more negative in heterogeneous than in homogeneous soils at low moisture level, but the reverse was true at high moisture level. Results do not consistently support either hypothesis, but do suggest that competitive ability can be partly explained by individual species traits such as size, and that some competitive effects and responses are emergent properties of interspecific interactions.

Aims Foraging behavior was concerned less in plants especially for clonal aquatic species. The goal of this study was to test the foraging behavior and the function of phenotypic plasticity of a clonal submerged plant, Potamogeton maackianus A. Benn in variable manipulative environments.
Methods To explore the foraging behavior of a clonal submerged plant, clones of P. maackianus were cultivated in variable environments of different substrate types and light conditions. We measured biomass, specific root area (SRA), specific root length (SRL), branching angle of the ramet, chlorophyll content of apical leaves (CCAL), Fv/Fm (chlorophyll fluorescence parameter that indicates maximal photosystem II (PS II) efficiency) and biomass distribution in soil and light contrast treatments after harvest.
Important findings We found that the atypical non-clonal-like foraging behavior homogenized the performance of the stoloniferous clonal species, P. maackianus, in the measurement of biomass in our study. Substrate type had significant effects on most traits of P. maackianus, except on total biomass, shoot biomass distribution and CCAL. Patchy substrate mediated more root distribution into sand patch rather than into clay patch. Increase of SRA, SRL and Fv / Fm might be adaptive to substrate heterogeneity. Moreover, substrate heterogeneity induced 'sidewalk' stolon lineage on the edge of sand patch along the patch border. Light condition only had significant effects on branching angle, shoot biomass distribution and Fv / Fm. Shade led to more shoot assignment in light-rich zone than in light-poor zone. Potamogeton maackianus adjusted the branching angle of ramet rather than clonal exploitation in heterogeneous substrates with patchy shade. Moreover, shade led to higher Fv / Fm. A compromising foraging strategy promoted P. maackianus to escape unfavorable environments. Phenotypic plasticity, which maximized the performance of P. maackianus in the measurement of biomass in variable environments, may explain its domination in some shallow lakes of the middle and lower reaches of Yangtze River.

Aims Water level is one of the most important determinants of the distribution and composition of submersed aquatic vegetation in shallow lakes. Without changes in nutrient loading, large-scale declines in submerged macrophytes in shallow lakes are largely attributed to strong external or internal forces, including changes in water level. As a winter-active submerged macrophyte, Potamogeton crispus has important functions in aquatic ecosystem. The objectives of our study were to reveal the effects of water depths on the germination, growth, reproduction and morphological changes of P. crispus under natural environmental conditions; identify the optimum water depth range for colonization of P. crispus; and predict the succession course for P. crispus in Lake Taihu.
Methods A long-term in situ experiment was performed to study the effects of water depth on the growth and morphology of P. crispus across the entire life cycle. Plants were grown in the improved cross type of rhizotrons for 30 weeks at three different water depths (60, 120 and 200cm) in Dongtaihu Bay of Lake Taihu. We measured the plant height, root length, the length and width of leaves and counted the numbers of leaves, ramets, turion spikes and roots of each plant on each observation day during the experiment.
Important findings Water depths ranging from 60 to 120cm were favourable for above-ground vegetation growth, root tissue growth and the reproductive ability of P. crispus. At water depth of 200cm or beyond, the distribution of P. crispus will be limited in the following year because of the lack of turion formation and the severe inhibition of ramet production. The relationship between above-ground biomass and growth time at three different water depths fits a logistic growth curve well (P < 0.001). The rapid growth and distribution with high density of P. crispus at water depth near 60cm during the reproductive stage is not favourable for growth of other submerged species. So a continuous water depth of ~60–120cm is the optimum growth depth range for the conservation and restoration of P. crispus in Lake Taihu. In addition, the morphological parameters values acquired can replace or provide the ranges for the value of G maxSB (the maximum growth rate of submerged plants, in the unit of d^-1) obtained via calibration in the control equations of submerged plant growth in the Eco-Taihu model, which is a three-dimensional ecological model of Lake Taihu.

Aims Reproductive fitness of different floral phenotypes varies within and/or among populations. These variations are important to understand the process of natural selection and the evolution of floral traits. In this study, we focused on a distylous, self-incompatible species, Primula poissonii, to investigate fitness-related selection on floral traits. Our aim was to determine how traits vary as targets of natural selection and whether morph-specific selection occurs.
Methods This study was conducted at two sites (Yushuizhai at 2700 m and Haligu at 3200 m) in the Lijiang Alpine Botanical Garden, northwestern Yunnan, southwestern China. Insects visiting flowers of P. poissonii were observed, captured and identified. Randomly selected plants of long- and short-styled morphs were labeled. Five floral/inflorescence traits were measured including floral display, corolla width (CW), floral tube length (FTL), tube opening width (TOW) and floral scape height. Fruit and seed set were recorded. The total number of seeds per individual plant (plant fitness) and seed production per capsule (flower fitness) were calculated. Multiple regression analyses were used to quantify selection gradients.
Important findings The frequencies of the two morphs did not deviate from the expected 1:1 ratio at both sites. Except for FTL, the four other traits did not differ significantly between the long- and short-styled morphs. Floral scape height, floral display and FTL differed between two sites. The selection regimes differed between two morphs and between two sites. At the Yushuizhai site, linear selection for shorter floral tubes was stronger in the short-styled morph. However, nonlinear selection on the floral display was stronger in the long-styled morph than selection on the short-styled morph. At the Haligu site, linear selection for a smaller corolla was stronger in the long-styled morph. A morph-specific nonlinear selection on CW and floral display was also detected. Morph-specific selections were detected through the estimation of flower fitness only in Haligu population. In this site, morph-specific linear selection was also detected for CW and floral display. Morph-specific nonlinear selection on traits was detected only in CW. We found that butterflies and sphingid moths dominated at Yushuizhai, while long-tongued bees dominated at Haligu. The difference in pollinator fauna suggested that selection on floral tubes may be due to differences in pollinator assemblages. Overall, variation of floral and/or inflorescence traits in P. poissonii was probably driven by pollinator selection. Selection regime differences between two morphs, in part, due the inter-morph divergences of sexual functions in distylous plant.

Aims Many polyploids are geographically separated from their lower-ploidy progenitors at a variety of spatial scales. Diverse adaptive and non-adaptive mechanisms have been suggested to contribute to these spatial patterns. A long-standing hypothesis holds that polyploids are established and persist due to specialization to different ecological conditions compared to their progenitors. However, experimental approaches testing this hypothesis remain rare. Empirical evidence for niche differences among ploidy levels has been found previously for the European polyploid geophyte Allium oleraceum. Here, we conducted a 5-year reciprocal transplant experiment with tetraploid, pentaploid and hexaploid A. oleraceum to test for adaptation at the ploidy and population levels.
Methods In 2006, we sowed 2160 aerial bulbils among six sites (two per cytotype, central and western Moravia, the Czech Republic) and monitored their emergence in 2007, their survival until 2011 and their performance in 2011.
Important findings The emergence of bulbils was high at all target sites, and no evidence for niche differentiation among cytotypes was recorded in the first year. Survival in the following years was relatively high, but significantly better survival for plants of home ploidy than foreign ploidy was recorded. No plant reached the reproductive stage during our experiment. Adopting a strict definition of local adaptation, local plants only rarely performed better than foreign plants at both sites in pairwise comparisons within a cytotype. Together, our results suggest weak niche differentiation among cytotypes and provide limited support for the existence of local adaptation within cytotypes. Thus, all cytotypes are able to establish in habitats typically populated by other cytotypes via aerial bulbil spread, but will likely die out gradually in non-native habitats. This ability may explain the occurrence of some multiple-cytotype populations. The complex spatial pattern of A. oleraceum cytotypes is therefore the result of several interacting factors, including niche differentiation among cytotypes, the different evolutionary ages of cytotypes, the existence of gene flow between and within cytotypes, the dominance of clonality and stochastic events linked to anthropogenic dispersal and disturbance.