Journal of Ecology

Climate change drives shifts in species composition, but turnover in many communities lags behind the current pace of change. Anticipating the impact of the resulting community-climate disequilibria on ecosystem functioning is critical. Present-day communities may already be out of equilibrium with climate, providing an opportunity to estimate the effects of disequilibrium before they become more widespread. We analyzed plant community composition and function data from [~]60,000 rangeland monitoring sites across the western US to measure how community-climate disequilibrium contributes to spatial and temporal variation in net primary productivity (NPP) - a key ecosystem function. We found that communities were already substantially out of equilibrium with climate and accounting for this disequilibrium helped explain patterns of NPP. Communities farthest from equilibrium were less productive than those that were closely matched with climate. Our findings suggest that future increases in community-climate disequilibrium may further impair ecosystem functioning.

How plants allocate carbon determines their productivity, responses to stress, and interactions with other organisms. A substantial amount of plant carbon is stored as non-structural carbohydrates (NSC), which sustain turgor via osmoregulation and fuel metabolism when carbon is limited. NSC also support root-colonizing mycorrhizal fungi, thus we hypothesized that under carbon-limiting conditions such as drought, a trade-off between feeding mycorrhizal fungi and maintaining turgor may arise. We reduced carbon allocation to ectomycorrhizal (EcM) networks by girdling Pinus ponderosa trees exposed to drought or ambient conditions and manipulated putative fungal connections between trees by trenching. We show that, in droughted plots, trees putatively connected to girdled trees by EcM networks had 33 % less needle NSC and >10% less turgor than those connected to ungirdled trees. Trees disconnected from the mycorrhizal network by trenching had increased NSC likely from the increased water availability with girdling, but these gains were offset in the presence of networks. Our results demonstrate that the increased carbon demand by EcM fungi in response to reduced carbon inputs from some trees can deplete NSC in neighboring trees via shared mycorrhizal networks. At least in the short term, allocation trade-offs under carbon-limiting conditions may expose networked trees to carbon deficits. This may increase vulnerability to drought, which may be particularly acute given shifts in climate.

Grasslands are vulnerable to increasing drought with global warming, but process-based models lack the mechanistic knowledge required to predict the magnitude of drought impacts. While a plant hydraulics framework has been successful in advancing process understanding of drought responses in trees, and how drought responses vary across rainfall gradients, similar approaches have rarely been applied to grasses. Here, we quantified the progression of key drought response processes in sixteen dominant perennial grasses (seven C3 and nine C4) with differing climatic origins across eastern Australia. We found that stomatal closure, hydraulic impairment and leaf browning occurred concurrently, in contrast to the progressive sequence typically observed in trees. We also found that drought response traits were not correlated with species climate of origin. The early impairment of leaf hydraulic conductance and leaf browning along with the lack of correlation with climate of origin suggest that grasses may employ fundamentally different strategies to adapt to low water availability than trees. These results highlight the need for grass-specific parameterization of drought responses in process-based models.

Repeated grazing and trampling can reduce forb populations in South African mesic grasslands by limiting photosynthesis and depleting carbohydrate reserves stored in underground organs. We examined the effects of repeated defoliation on growth, bulb starch reserves and water status in the geophyte Eucomis autumnalis over 468 days. Four intense summer and autumn defoliations reduced above-ground production, bulb mass, starch reserves and bulb water pools, with the strongest impacts occurring in late summer and autumn when plants normally replenish reserves. Repeated defoliation disrupted seasonal source sink dynamics, forcing resources into regrowth and driving bulbs towards energetic bankruptcy. These results highlight the importance of avoiding heavy autumn grazing and providing periodic full-year rests to conserve geophytic forbs in mesic grasslands.

Understanding how habitat connectivity shapes biodiversity remains a major ecological challenge. In particular, the roles of connectivity and ecological heterogeneity on co-variation in plant species diversity and intraspecific genetic diversity is not understood. We combined species distribution modelling, resistance-to-movement mapping, landscape connectivity analysis and population genomics to investigate diversity patterns in three wet meadow herbs, Scorzonera humilis, Oenanthe peucedanifolia and Lychnis flos-cuculi, and their surrounding plant communities. Genetic diversity patterns differed strongly among co-occurring species. Connectivity metrics explained genetic diversity only in O. peucedanifolia, and environmental drivers of genetic diversity were highly species specific. Importantly, genetic diversity changed with the presence of some species in the community, but it was consistently unrelated to indicators of local plant community diversity. Overall, the processes shaping within-species biodiversity may differ fundamentally from those structuring habitat connectivity and plant species communities, with important implications for conservation.

Most plants require animal pollination to reproduce, prompting concern that pollinator declines immediately threaten plant populations. This concern is warranted if pollinator-mediated seed losses cause declines in plant population growth rates ({lambda}). However, demographic trade-offs might reduce the risk of population decline if seed loss improves performance elsewhere in the life cycle. We conducted a multi-year pollination manipulation on four species and measured how demographic vital rates and {lambda} responded. Seed responses did not predict net changes in {lambda}. Reduced pollination decreased seed production, but only caused a net decrease in {lambda} in one species; in the others, improved survival buffered {lambda}. Increased pollination boosted seed production, but at a cost to survival that caused a net reduction in {lambda} in three species. Our results highlight the importance of demographic trade-offs for understanding the impacts of pollinator declines on plant biodiversity and, more broadly, the population-level impacts of changing mutualisms.

Climate change is expected to alter plant populations not only through direct environmental shifts but also via changes in biotic interactions, such as with herbivores and pathogens. As plant species are also expected to differ in their responses to both climate and antagonists, plant responses to both factors are expected to be variable and species-specific. To assess whether interactive effects of climate and antagonists on plant population dynamics are common and whether the strength and direction of plant responses vary across species, we conducted a multi-year field experiment that manipulated realistic climate change and experimentally reduced insect herbivores and fungal pathogens. We measured responses of plant vital rates, such as survivorship, growth, and reproduction across six grassland species. Using Integral Projection Models (IPMs) and Life Table Response Experiments (LTREs), we quantified changes in population growth rate across experimental treatments and the contribution of each vital rate to that observed change. Two of the study species declined so drastically over the course of the experiment that demographic quantification of population growth rates was not possible. From the remaining species, Bromus erectus and Plantago lanceolata show significant interactive responses of climate and antagonist reduction on population growth rates. In contrast, Dianthus carthusianorum and Tragopogon orientalis showed limited responses to experimental treatments. Notably, our results indicate that in some species biotic interactions may amplify the effects of climate change: the presence of plant antagonists exacerbates the negative effects of the future climate treatment on plant population dynamics. Our findings highlight the complexity in predicting plant population responses to climate change and provide insights for grassland management under future environmental conditions.

Tropical rainforests generate and maintain their own microclimate regimes and the resultant cooler, more humid and stable environments foster the hyperdiversity typical of these ecosystems. Although the temporal and spatial (horizontal) distributions of microclimates have been relatively well studied, the vertical dimension has received less attention, and little is known about how forest disturbance affects the vertical stratification of microclimates in tropical forests. In this study, we examine how the vertical distribution of temperatures varies between undisturbed and burned Amazonian forests. We installed five vertical transects with temperature dataloggers distributed at 7 different heights to collect data over multiple days during the end of the dry season. We investigated how anthropogenic disturbance (fire) mediates the vertical stratification of microclimate and whether microclimate buffering (i.e, the difference between understorey and canopy temperatures) varies according to the forest structure. We showed that anthropogenic disturbance can cause an inversion in the vertical stratification of microclimates, with burned forests having hotter temperatures (up to 2 {degrees}C) in the understorey than in the canopy during the day - the opposite of what is found in undisturbed forests (typically 3 {degrees}C cooler). During the night, while understorey and canopy temperatures are similar in undisturbed forests, we found that, in burned forests, understorey temperatures were up to 2 {degrees}C cooler than in the canopy. Microclimate buffering by day was best explained by aboveground carbon stocks, with higher temperature buffering in more carbon rich forests. Our study shows that anthropogenic disturbance alters the vertical stratification of temperatures in Amazonian forests, leading to significant temporal changes along the diel cycle. Future research should focus on understanding these changes across a wider range of disturbance regimes, and explore the consequences for biodiversity and ecosystem functions from the canopy to the forest floor.

Forest regeneration depends not only on how many seeds trees produce, but on the physiological quality of those seeds. Yet while climate-driven shifts in seed quantity and masting have received sustained attention, the parallel question of whether climate change degrades seed quality remains poorly resolved. Using a nationwide dataset of seed mass and viability in European beech (Fagus sylvatica L.) collected between 1996 and 2024 (13,349 seed lots from 381 forest districts across Poland), with climate-quality analyses focused on 5,374 freshly harvested seed lots from 353 districts (2004-2023), we asked whether the two components of seed quality respond to different seasonal climatic windows, and whether harvest-year climate also shapes seed performance during long-term cold storage. Seed mass and seed viability were only weakly correlated (Spearmans {rho} = 0.15), acting as two independent dimensions of seed quality. Both revealed substantial temporal variation over the study period, but along distinct trajectories. Seed mass declined markedly between segmented-regression breakpoints in 2009 and 2019, more steeply at higher latitudes, coinciding spatially and temporally with the masting breakdown reported at the species northeastern range margin. Climatic associations were correspondingly divergent. Viability was positively associated with previous summer temperature, consistent with temperature-cued flower initiation, and negatively with spring temperature in the harvest year, plausibly reflecting thermal disruption of early embryogenesis. Seed mass showed no significant association with any seasonal climatic predictor, indicating control by slower or unmeasured processes. Storage duration progressively reduced viability, and this decline was further modulated by climate during seed development, with seeds developing under climatically suboptimal conditions losing viability faster. These results expose a hidden decoupling between seed quantity and seed quality under contemporary climate change, with direct consequences for forest regeneration and for ex situ conservation strategies that assume mast-year seeds will remain viable for decades.

O_LIIntraspecific variability in hydraulic and morphological traits may alter projections of forest vulnerability to climate change. Recognizing multiscale nature is essential, as trait relationships within populations may not reflect those across the species range. C_LIO_LIWe assessed variability in 11 traits in 10 natural populations of Abies alba across environmental and phylogeographic gradients to evaluate potential impacts on drought vulnerability. We quantified trait variation within and among populations, tested effects of climate, phylogeny and local factors, and evaluated trait coordination across scales. C_LIO_LIHydraulic safety traits and wood density showed low variability, indicating strong constraints. In contrast, water-use and efficiency traits were highly variable. Aridity influenced several traits, but reduced variance was detected only for leaf residual conductance and succulence consisted with stabilising selection. Trait coordination was weak within populations than among populations. C_LIO_LIOverall, A. alba combines constrained hydraulic safety traits with variable water use traits. This may buffer drought impacts but limits shifts in hydraulic safety margins, potentially increasing risks of hydraulic failure, especially in humid environments where populations lack the capacity to tolerate prolonged dry periods. Thus, our findings highlight the need to account for scale-dependent trait variability, coordination, and their underlying drivers when predicting species adaptive capacity to drought. C_LI

O_LIPlants take up nutrients from the soil while investing in absorptive root surface or mycorrhizal partners. Root hairs - a major structure for nutrient uptake and cheap to build - increase the absorptive root surface. As such they are an important component of plant resource economics but largely neglected in root economic concepts so far. C_LIO_LIThis is mainly due to data scarcity, which we set out to overcome by measuring root-hair traits on 82 European grassland species in a greenhouse experiment. Using fluorescence and light microscopy, root-hair length and incidence was measured along with mycorrhizal colonization. C_LIO_LIWe found a phylogenetically conserved trade-off between plant investment into root hairs and mycorrhiza. A similar trade-off between root-hair incidence and mycorrhiza occurred at the intraspecific level, while patterns were heterogeneous among species. Plant species with high colonization rates showed the highest variability in root-hair incidence. C_LIO_LIWe conclude that plants vary along a gradient ranging from investment into root hairs as part of a "do-it-yourself" strategy to collaboration with mycorrhizal fungi while showing intraspecific variation in root-hair incidence. These findings demonstrate that root hairs play a fundamental role in fine-root trait variation and need to be considered when studying belowground plant economic strategies. C_LI

Devils gardens are among the most striking ant-plant mutualisms in Amazonian forests. In this system, the tree Duroia hirsuta is associated with the ant Myrmelachista schumanni, which actively removes neighboring vegetation and maintains nearly monodominant patches of the host plant. Despite the apparent efficiency of this system, repeated field observations at the Tiputini Biodiversity Station, Yasuni Biosphere Reserve, revealed that individuals of Palicourea alba recurrently occur within active devils gardens. Palicourea alba closely resembles dead plant material, exhibiting leaf morphology and coloration that strongly mimic the surrounding litter layer, and appears to be uncommon outside these gardens. To our knowledge, this "dead-leaf" masquerade has not been previously documented in this intensively studied system, making it a particularly striking and unexpected observation. To evaluate whether this masquerade facilitates persistence within devils gardens, we surveyed 35 gardens and recorded P. alba in 19 (52.8%). When present, P. alba covered on average 27% of plot area, while mean herbivory across sampled leaves remained low (8.6%). Generalized linear mixed models showed that P. alba cover decreased significantly with increasing herbivory (F = 8.09, p = 0.0159), whereas herbivory increased with leaf-litter cover (F = 8.73, p = 0.0120). Field observations further revealed that many individuals are nearly indistinguishable from dry leaf litter, suggesting a role for visual crypsis or masquerade. Together, these results indicate initially, that the persistence of P. alba within devils gardens is mediated by a multi-layered ecological filtering process. First, masquerade likely reduces detection by M. schumanni, allowing seedlings to escape ant-mediated removal. Second, low herbivory suggests either enemy avoidance or reduced apparency to herbivores within the simplified understory. Third, spatial heterogeneity in leaf-litter cover may create microhabitats where both ant activity and herbivore pressure are modulated, reinforcing establishment success. This system thus represents a previously undocumented mechanism in which plant-litter resemblance enables persistence within a highly structured, biotically filtered habitat, highlighting how subtle trait-mediated interactions may modulate outcomes in otherwise strongly deterministic mutualisms.

Disturbance is widely recognized as a catalyst of invasion, but growing evidence suggests even protected communities are susceptible to severe infestation. We used k[i]pukas--naturally isolated patches of minimally-disturbed vegetation surrounded by lava flows--as a large-scale natural experiment to test the long-term biotic resistance of protected sagebrush ecosystems threatened by Bromus tectorum and other invasive annual grasses. Employing a robust causal inference approach combining matching with regression adjustment, we compared protected communities within k[i]pukas to otherwise similar communities exposed to contemporary disturbance regimes. Despite their near-total protection from fire and livestock grazing, k[i]pukas were extensively invaded by annual grasses (18.9 {+/-} 0.28% cover), with abundance comparable to or slightly exceeding disturbed sites (16.7 {+/-} 0.29% cover). These findings challenge the notion that protection from disturbance confers effective long-term resistance to invasion, instead demonstrating that invaders can establish, proliferate, and drive ecosystem transformation where favorable abiotic conditions prevail. Our findings reveal the limits of passive protection as a conservation strategy, suggesting active management may be necessary to prevent ecosystem degradation by aggressive invaders.

Atmospheric inorganic nitrogen (N) deposition has been linked to increased tree phosphorus (P) deficiency and shifts in ectomycorrhizal (ECM) fungal community composition across Europe, but the underlying mechanisms remain poorly understood due to the scarcity of species-level studies of fungal physiology at large spatial scales. Here, we characterized ECM communities in nine Scots pine (Pinus sylvestris L.) stands across Europes largest N deposition gradient to gain mechanistic insight into N-driven ECM community shifts, by assessing morpho-physiological traits (i.e. soil exploration types and ECM root-tip level exoenzyme activities involved in organic N and P acquisition) on individual ectomycorrhizas. Our data revealed high functional variation in foraging strategies across species and sites, including within dominant ECM genera (Cortinarius, Elaphomyces, Lactarius, Piloderma, Russula). Shifts in community-level exoenzyme activities along the N deposition gradient were consistent with increasing P limitation, with a buffering effect of phosphomonoesterase activity on host nutritional status (i.e. reduced foliar N:P). These trends were mainly driven by interspecific differences in enzymatic profiles and species turnover along the gradient, rather than intraspecific variation within widespread species. Dominant low-biomass species in high N sites (e.g. E. citrinopapillatus, L. subdulcis, R. ochroleuca) were efficient P-foragers, with some displaying high oxidative activity, potentially hampering soil carbon storage under elevated N loads. These findings highlight the role of ECM species-specific traits in mediating ecosystem processes and can help understand the future of pine forests under chronic N pollution, with potential implications for applied forestry.

O_LIBiodiversity loss can alter interactions not only through changes in tree species richness, but also through the loss of particular functional strategies from ecological communities. Working in a subtropical forest diversity experiment we asked whether tree species richness effects on arthropod herbivory and leaf pathogen infestation depend on community functional diversity, and whether trait dissimilarity-based, non-random species loss alters these relationships compared to random loss. To address this, we combined already established planted scenarios with newly constructed extinction pathways. C_LIO_LIWe tested the responses of herbivory and leaf pathogen infestation (i) to tree species richness, functional diversity (Raos Q), community structure and resource strategies (i.e. ever-greenness) and community-weighted trait means as well as predation, and (ii) trait dissimilarity-based extinction pathway analyses that contrasted directed loss of functionally similar versus functionally distinct tree species. C_LIO_LIHerbivory increased with tree species richness and this increase was significantly stronger in communities with higher tree functional diversity. Under directed species loss scenarios, herbivory differed most strongly from random-loss expectations when similar tree species were lost first. By contrast, losing functionally distinct species first produced richness effects that were much closer to the random-loss scenarios. Trait-based species loss will therefore modify trophic interactions more strongly than random loss. For pathogen infestation tree richness effects depended on evergreenness and among planted extinction scenarios (three-way interaction), with only minor deviations of trait-based extinction path-ways from random-loss expectations. Pathogen infestation also tended to increase with community-weighted mean leaf nitrogen. Predation showed no clear relationship with tree species richness or functional diversity but was positively associated with herbivory. The strength of this association differed among extinction scenarios, providing no evidence for consistent top-down regulation. C_LI SynthesisThe ecological consequences of biodiversity loss for leaf damage depend on which functional strategies are lost, not only on how many tree species remain. By integrating ob-served tree diversity gradients with trait-based extinction pathways, this study shows that functional diversity and host redundancy help explain why herbivores and pathogens are shaped by the same changes in tree diversity through different functional constraints and im-prove predictions of interaction strength under non-random species loss.

Forests play a crucial role in mitigating climate change as primary terrestrial carbon sinks. While some studies suggest that global warming enhances forest productivity, a growing body of evidence highlights detrimental impact primarily driven by increased water stress. Yet the extent to which positive effects of climate change offset its negative impacts on tree species productivity remains unclear at large spatial extents. We assessed forest growth and mortality for the 21 most abundant tree species in Europe using National Forest Inventory data from more than 50,000 plots and 700,000 trees to disentangle the relative importance of climate and forest structure. Specifically, we examined how vapor pressure deficit (VPD) anomalies across species climatic edges and stand developmental stages affect forest growth and mortality occurrence and intensity (i.e. whether mortality occurred and the amount of basal area lost). Then, we aggregated the responses across species and separately for broad-leaved and needle-leaved species to assess whether forest growth and mortality differed between major functional groups. Although the importance of forest growth and mortality drivers varied markedly among species, climate had a stronger influence on mortality than on growth, particularly in needle-leaved species. Forest growth declined and mortality increased along VPD anomaly in most species and forests studied. Responses were most pronounced at arid species edges in early-stage broad-leaved forests and at wet edges in late-stage needle-leaved forests, where differences between functional groups were also highest. We evidence the need to parametrise species-specific models of forest growth and mortality across large spatial extents to better understand and predict effects of climate change on forest productivity. In addition, our results emphasize the importance of improving the understanding of forest mortality processes given the strong influence of climate on mortality, while also further studying vulnerable populations to climate change in arid edges of species distributions.

Regeneration failure is a bottleneck in Mediterranean oak woodlands. Cattle can hinder or promote recruitment, depending on grazing location, timing and intensity. Herbivory theory predicts that repeated defoliation and trampling deplete seedling reserves, whereas resprouting can extend survival; yet field studies rarely separate intensity from recency or combine long-run grazing records with individual fates and microhabitat/climate context. We test how management-driven heterogeneity shapes cork oak seedling survival and resprouting by combining 12 years of paddock-level grazing records with individual tracking of 8431 seedlings across 24 paddocks. Bayesian mixed-effects survival models related seedling lifespan to grazing history x pressure (moderate [≤]150; high >150 LSU ha-1 days yr-1) and to key covariates, including seedling height, resprouting status, shrub distance, cattle dung counts (as a proxy of very recent grazing), and 1-month SPEI (as recent water balance). Bayesianlogistic mixed models were then used to relate resprouting probability to grazing treatments. Survival was lower in grazed than ungrazed paddocks and declined along management gradients: median lifespan fell from 460 (moderate grazing) to 256 days (high), and from 460 (old grazing; two-year absence) to 199 days (recent). A two-year cattle absence increased survival under moderate pressure but was insufficient where pressure was high, indicating legacy effects and that recovery windows must scale with pressure. Resprouting dominated persistence: resprouters lived >5x longer than non-resprouters (2351 vs 460 days). Taller seedlings lived longer, and shrub proximity conferred a modest benefit. Climate modulated outcomes: wetter recent periods (higher SPEI) markedly boosted survival. Cattle reduced the odds of resprouting, with the strongest penalty under recent use. By disentangling grazing intensity from recency and linking both to seedling survival and resprouting, we show why recruitment falters under continuous, heavy grazing and when it can recover. Because drought intensifies cattle impacts, managers should combine moderate stocking rates with multi-year rest periods to rebuild oak bud banks and below-ground reserves; a two-year hiatus can help under moderate pressure but appears insufficient where pressure is high. Aligning rotational plans with drought outlooks and tracking simple field cues (seedling height, recent resprouting) offers a practical path to reconcile production with regeneration in Mediterranean wood-pastures. HighlightsO_LITwelve years of grazing records linked to 8431 cork oak seedling fates C_LIO_LIRecent grazing reduced survival and resprouting versus a two-year cattle absence C_LIO_LIHigh grazing shortened lifespan; two-year rest helped only under moderate pressure C_LIO_LIResprouting was the strongest survival correlate; resprouters lived over 5x longer C_LIO_LIWetter short-term water balance increased cork oak seedling longevity C_LI

Forests are essential to the global carbon cycle with light use efficiency (LUE) as a key parameter for assessing carbon sequestration capacity. However, the variations and drivers of LUE remain inadequately understood. Using remote sensing data, we analyzed global LUE patterns across five forest types and identified the main drivers. The global average annual LUE of forests is 0.93 {+/-} 0.36 g C MJ-1 during the period 2001-2022, with an increasing trend of 0.0034 g C MJ-1 yr-1. Among forest types, evergreen broadleaf forests exhibited the highest LUE, followed by evergreen needleleaf forests. Deciduous broadleaf forests and mixed forests showed similar levels, while deciduous needleleaf forests exhibiting the lowest LUE. Variations in LUE were jointly driven by plant traits and climatic conditions, with generalized linear models explaining 86% and 98% of spatial and temporal LUE variations, respectively. These findings highlight the critical role of plant traits and climate in shaping forest LUE, providing insights for enhancing carbon cycle models and informing forest management strategies in the context of global change.

SummaryO_LIIn woody perennials, reproductive allocation considered as a fraction of NPP (RA) has been rarely quantified, especially tracked across plant lifetimes, measured in biologically meaningful currencies, and separated from standing biomass. C_LIO_LIWe addressed this gap by measuring dry mass, nitrogen, and phosphorus for every aboveground tissue type for 14 iteroparous perennial shrubs tracked across their full lifetimes, calculating RA using four accounting schemes and three currencies. C_LIO_LIRA was substantially higher by mid-life than typical estimates from ecosystem-scale studies. Distinguishing standing biomass from yearly production further revised RA upwards. Using a nutrient currency (especially P) increased the perceived costs of reproduction. Propagules were highly nutrient-enriched and reproductive accessory costs consumed more nutrients than did the propagules themselves. Considering N and P resorption from senescing leaves and wood shrank the effective vegetative nutrient budget, further concentrating net annual nutrient demand in reproductive tissues. C_LIO_LIOur results highlight high investment in RA for woody perennials, especially using nutrient currencies. Broadly similar allocation patterns were observed across species with different functional traits and lifespans, suggesting generality that may apply across biomes. Widespread underestimation of RA in forest growth models likely overestimates the proportion of NPP available for vegetative growth, leading to substantial errors in predictions. C_LI

Background and AimsIn the plant kingdom, gene flow occurs through pollen and seed dispersal, shaping both within-population spatial genetic structure and among-population genetic differentiation. Anthropogenic land-use change can affect levels of gene flow by reducing pollinator abundance and altering pollen and seed dispersal pathways. Yet, how environmental features shape gene flow events in common herbaceous plants - a fundamental building block of many ecosystems - remains poorly understood. We address this gap by investigating how environmental context influences population genetic structure at regional and local scales in the red campion (Silene dioica). MethodsBy sampling 1,005 individuals from 29 populations across habitats ranging from semi-natural to strongly human-altered, we assessed whether population size and landscape composition influenced within-population genetic diversity and population genetic differentiation. At the local scale, we examined whether landscape composition affected fine-scale spatial genetic structure and pollen dispersal distances in a subset of six populations representing the two extremes of an anthropogenic gradient. Key ResultsNo effect of either demographic or landscape factors was found on levels of genetic diversity. We detected moderate levels of genetic differentiation among populations that matched an isolation-by-distance pattern, with high levels of population admixture, while landscape composition did not explain variation in population genetic differentiation. At the local scale, five of the six studied populations exhibited significant spatial genetic structuring, indicating distinct neighbourhoods at spatial scales less than 10 m. Paternity analyses based on 4,800 offspring further revealed predominantly short-distance pollen dispersal together with substantial immigration from external sources. Neither fine-scale genetic structure nor pollen dispersal distances differed between habitat types. ConclusionsAltogether, our results demonstrate a multimodal pattern of gene flow and a remarkable resilience of this common herbaceous species to anthropogenic habitat change, as substantial genetic connectivity is maintained despite insect-mediated pollen dispersal and gravity-driven seed dispersal.