Nature Ecology & Evolution

Multiple genetic variants associated with diet-related traits show strong signatures of natural selection1-17. To test whether these signals were indeed diet-driven, we conducted an empirical investigation. We compiled an isotopic dataset comprising 6,064 ancient human samples and 5,635 food resource samples from Britain. We developed a Bayesian mixing model to estimate individual dietary proportions based on isotopic data and subsequently constructed a temporal dietary model. A dairy-use time series was also constructed18. Using 1,038 ancient DNA samples, we reconstructed derived allele frequency trajectories for 14 strongly selected single nucleotide polymorphisms (SNPs) via bootstrap resampling19. A generalized additive model (GAM) was then applied to estimate mean and time-varying selection coefficients, while accounting for evolutionary forces beyond selection. Finally, we applied the convergent cross mapping (CCM)20 algorithm for causal discovery between the time-varying selection coefficients and their corresponding dietary variables. Our findings indicate that C3 plant consumption drove selection on rs12401678 and rs653178, and dairy consumption on rs4988235. Selection signals at rs174570 and rs174594 are likely linked to marine fish and terrestrial meat intake, whereas the remaining SNPs show more complex selection dynamics in which any diet-related signal cannot be clearly identified. Our results underscore the complexity of natural selection at the genetic level and highlight the need for more careful evaluation when identifying its potential drivers.

Interspecific comparisons of cell-type-specific gene expression levels can provide information about the evolutionary processes that drove divergence between species. From these comparisons, it is now evident that the predominant mode of gene expression evolution has been stabilizing selection, both on the steady-state (mean) protein levels as well as on the mRNA levels with additional lineage-specific shifts resulting from directional selection. However, as all previous work has used bulk RNA measurements, it has been impossible to determine which of the many cellular processes that contribute to mean abundances are highly constrained and which are more evolutionary labile. Assessing this is further complicated by the expectation that components of complex systems will evolve over time independent of changes in the selective regime so long as the net output of a system (i.e., mean expression) remains near the evolutionary optima. This process, known as evolutionary systems drift (ESD), has been frequently invoked as a non-adaptive explanation for changes in cellular phenotypes but has never been quantitatively tested or accounted for in any statistical test for selective constraints. Here, we develop a new paradigm that addresses both of these open problems simultaneously. Using single-cell expression data and biophysical models, we estimate mRNA transcriptional bursting rates, splicing rates, and decay rates across multiple vertebrate species. We then derive new mathematical results that describe how these various biophysical parameters are expected to co-evolve under ESD and then test whether we need additional evolutionary constraints to explain the divergences in these parameters. We find evidence that the biophysical parameters are indeed evolving in a coordinated manner as predicted by ESD and that there are additional strong constraints on transcriptional bursting, likely as a consequence of selection to reduce noise in expression. More broadly, this work opens up a whole new approach for studying the evolutionary dynamics of complex cellular systems.

Emerging infectious diseases threaten public health and biodiversity across the globe1,2. Disease outcomes are frequently dependent on local environmental conditions3-5, but how these factors shape host adaptation and long-term recovery are often unknown6. Here we combine two decades of population, disease, and environmental data with a common garden experiment to investigate the drivers of variable declines and recovery for remnant bat populations following the emergence of the fungal disease, white-nose syndrome. We find that initial declines were greater and faster in warmer sites (88.3% vs. 74.2% in cold sites), but these populations recovered more quickly and hosts developed higher resistance (1.5x reduction of fungal loads) than populations from colder sites that were buffered from initial impacts. Our experimental data suggest that warm sites served as hotspots of host adaptation where selective pressures were stronger because thermal conditions approached optimal growth for the pathogen, which eventually favored the development of high pathogen resistance. Populations in colder sites experienced weaker selective pressure and thus remain more susceptible, although bats from larger colonies were more likely to survive, suggesting that adaptive traits exist in these populations, but at much lower frequency. These findings show that the environmental conditions that initially buffer populations from collapse can simultaneously constrain their evolutionary response to emerging threats, and ultimately determine differential recovery following disease-induced declines.

Tropical rainforests, and Amazonia in particular, contain more tree species than anywhere else, most of which arose through rapid evolutionary radiations 1-3. Rapid radiations are often catalysed by ecological opportunity 4-6, which in rainforest trees is presented by intense insect herbivore pressure, spurring the evolution of novel plant defence chemistry to escape it 7. However, we do not understand how long-lived trees can adapt quickly enough to keep pace with rapidly-evolving insect herbivores. Here we show that hybridisation in rainforest trees, which was considered rare, allows exchange of gene clusters used in chemical defence against herbivore attack, facilitating rapid adaptation and diversification. Using genome sequencing for 461 individuals from the genus Inga, a characteristic Amazonian tree radiation, we find that regional tree communities form syngameons - networks of closely related, co-occurring species connected by gene flow. Integrating these genomes with herbivore abundance data from the same communities across the tropical Americas, we show that herbivore compositional turnover coincides with local, recurrent interspecific transfer of defence gene clusters that are retained by balancing selection, consistent with fluctuating selective pressure imposed by shifting herbivore communities. Together, our results demonstrate that hybridisation allows long-lived tropical trees to rapidly evolve chemical defences, fuelling adaptation to the relentless insect herbivory that structures the worlds most species-rich forests.

Tumour prevalence varies dramatically throughout the animal kingdom despite broadly conserved cellular and developmental processes, raising the question of how evolution has shaped susceptibility 1,2. Here, we link macroevolutionary variation in tumour prevalence to gene-level selection by integrating comparative genomics data from 109 species of birds and mammals using a Bayesian phylogenetic framework to estimate pangenome-wide rates of genetic evolution across >150 million years of evolutionary change. We identify 3,206 genes in which natural selection is associated with shifts in tumour prevalence, with more than 80% of which are linked to reduced prevalence, suggesting pervasive selection for cancer suppression. Using causal phylogenetic inference, we show that genes associated with reduced tumour prevalence act predominantly through indirect effects on body size, revealing growth as a key mediator of cancer risk across species. In contrast, genes associated with increased tumour prevalence exert direct effects independent of body size. Finally, at the species-level, we demonstrate that exceptionally low rates of benign tumours do not necessarily coincide with reduced malignancy, revealing that benign and malignant tumour processes are evolutionarily decoupled. Together, these results reveal how natural selection has fine-tuned the link between genotype, phenotype, and cancer risk across species.

Climate warming is altering the timing of seasonal events across ecosystems, impacting the temporal synchrony of interactions among species1,2. For trophic interactions, the match-mismatch hypothesis predicts that when consumers become phenologically asynchronous with key ephemeral resources their fitness will decline3-5. Most studies of mismatch focus on single resource-consumer species pairs, and implicitly assume trophic specialisation. However, many consumers exploit more than one resource species, giving rise to several mechanisms whereby the negative impacts of mismatch on individuals and populations could be buffered6. Here we experimentally manipulate phenological asynchrony across 48 plant-caterpillar interactions in a spring woodland food-web system and assay caterpillar performance. As asynchrony increases, we find strong evidence for a decline in survival that generalises across host-caterpillar interactions, whereas caterpillar growth and development are largely unaffected. We also show that focus in the literature on a single model interaction (Oak-Winter Moth)7,8 has likely overestimated the general impact asynchrony in this system. The strength of the effect of mismatch varies markedly among host-plants, caterpillars, and their interactions--with a small number of interactions showing little or no decline in consumer performance despite substantial asynchrony. Our results demonstrate that the fitness consequences of phenological mismatch are widespread but interaction-specific, revealing substantial heterogeneity in how trophic interactions are expected to respond to climate-driven shifts in seasonal timing. This variation in response could allow resource diversity and resource switching to buffer consumer guilds against the phenological impacts of ongoing climate change, stabilising the abundance of caterpillars for higher trophic levels.

Eupercarian spiny-rayed fishes are one of the largest vertebrate radiations, rivaling mammals and occupying nearly every aquatic habitat. We present a densely sampled, time-calibrated phylogenomic framework for Eupercaria, supporting a revised classification, combined with the largest cranial phenomics dataset for fishes. Habitat and trophic ecology make independent, complementary contributions to skull shape. Most species cluster around a conserved generalized architecture, the Percomorph Pile, from which one clade of pufferfishes, anglerfishes, butterflyfishes, and surgeonfishes repeatedly invaded novel morphospace; exceptionally high rates on its deep branches indicate that rapid skull evolution arose early in this clade. Freshwater lineages converge on the ancestral condition, reflecting late arrival into systems occupied by older otophysans, whereas durophages show the greatest disparity and converge on derived forms. Cranial diversity was partitioned among subclades during the Cretaceous and later within them across the Cenozoic, showing that clade-level differences in evolutionary rates and ecological opportunity jointly shaped skull diversification.

Algal blooms are frequently dominated by motile species1,2 whose vertical migration enhances resource acquisition and bloom development3,4. Yet bloom conditions present a paradox: high cell densities intensify nutrient depletion5 and self-shading6, making individual swimming increasingly costly under severe resource limitation. How motile blooms persist and remain resilient under such stress remains unresolved7, particularly as climate-driven warming strengthens stratification and resource scarcity8,9. Here we show that the red-tide-forming phytoplankton Heterosigma akashiwo overcomes bloom-induced constraints through bioconvection, a self-generated active flow that emerges above a critical cell density (>1.5x105 cells/ml). Using a custom ocean-on-chip platform that recapitulates bloom-relevant constraints, we identify an optimal synergy of cell concentration, swimming speed and gravitactic stability that promotes the formation of persistent bioconvective plumes. At constant cell density, plume onset is governed by two phenotypic traits-- vertical swimming velocity and reorientation time--demonstrating that collective transport is governed by the biophysical traits of single cells. We show that bioconvection drives ecologically relevant multiscale transport, enhancing exchange of molecules and micro-cargo across stratified interfaces, mimicking transport of nutrients, extracellular vesicles10 and co-existing species in a bloom environment11. By enabling cells to hitch a hike on self-generated flows when active propulsion becomes energetically prohibitive, bioconvection-mediated transport improves nutrient delivery, restores photosynthetic performance, reverses lipid accumulation associated with nutrient-stress, and facilitates recovery of cellular motility to ultimately mitigate resource limitations. Our findings identify bioconvection as a population-level adaptive mechanism that sustains algal blooms, and reveal a previously unrecognised role of collective microbial motion in bloom persistence under ecological stresses. One sentence summarySelf-organised bioconvection drives multiscale transport and resilience in algal blooms.

The impact of projected extinctions on global animal colour diversity remains unknown. Combining citizen science with self-supervised deep learning, we built novel representations of bird plumage colour patterning based on >125,000 museum specimen images covering 9,143 species. We demonstrate that losing currently threatened bird species will drive a disproportionate reduction in avian plumage diversity, with the most severe losses occurring in tropical and subtropical regions. Furthermore, while humans generally find non-typical plumage phenotypes more aesthetically attractive, threatened species are unexpectedly deemed less visually appealing despite their comparatively unusual plumages. Overall, our results highlight severe, imminent threats to the existing avian colourscape and raise critical questions about the future of animal colour diversity in a changing world.

Ancient admixture is a powerful catalyst for evolutionary innovation, yet its long-term genomic consequences for newly formed lineages remain poorly understood. Here, using 53 whole genomes across 19 species of the Asian langur genus Trachypithecus, we explore ancient admixtures role in shaping a reticulated radiation. Anchoring our phylogenetic analysis on the X-linked recombination desert (XLRD)--which shows an 84.5% reduction in introgression compared to autosomes--mitigates genome-wide historical noise to resolve the true ancestral species tree. Genome-wide analyses of phylogenomic triplet topologies demonstrate that phylogenetic discordance across this radiation is primarily driven by widespread introgression rather than incomplete lineage sorting, establishing reticulation as the dominant force shaping genomic diversity. Within this network, we identify Delacours langur as a clear case of homoploid hybrid speciation, arising from ancient admixture with [~]30:70 genomic contributions from northern and southern ancestral limestone langur lineages. This hybrid lineage appears to have fixed key loci related to reproductive isolation, including a chimeric RNF175 allele potentially determining its diagnostic pelage phenotype, illustrating how introgressed variation can underpin rapid phenotypic divergence. However, subsequent spatial isolation within fragmented karst landscapes forced a major conversion of genetic burden into realized load. Genus-wide, homozygous loss-of-function variants exceed 82%, consistent with the expression of lethal recessive mutations within long runs of homozygosity. Together, our findings demonstrate that ancient admixture can simultaneously trigger homoploid hybrid speciation and lock derived lineages into severe, long-term genomic erosion, revealing a fundamental trade-off at the heart of reticulate evolutionary radiations.

Climate is a strong predictor of global species richness, but the effects of climatic conditions are difficult to separate from the geography of the climates themselves. Recent work in climate space has shown that the area and isolation of discrete climatic conditions explain broad-scale richness gradients, yet the internal spatial cohesion of those climates remains poorly characterized. Here, we introduce climate percolation as a complementary descriptor of climate geography, measuring the degree to which the total area of a climate bin is concentrated within effectively connected fragments. Using global range maps for amphibians, birds, mammals and reptiles, we quantified species richness across a two-dimensional climate space defined from 12 climatic variables and evaluated the independent and joint effects of climate area, climate isolation and climate percolation across multiple climate-space resolutions. Climate isolation and percolation were strongly coupled: their first joint axis explained, on average, more than 95% of their shared variation, revealing a dominant gradient of climate fragmentation along which geographically isolated climates are also internally subdivided. Despite this collinearity, percolation consistently outperformed isolation in cross-validation across all four vertebrate groups, with particularly strong predictive gains for birds and mammals. The largest improvements, however, came from the shared isolation-percolation axis, indicating that vertebrate richness in climate space is more strongly associated with the integrated geographical structure of climates than with either inter-fragment distance or internal cohesion alone. These results suggest that climate fragmentation is a multidimensional property of environmental space, combining both the distance among climate fragments and the dominance structure of connected areas. By extending climate-space approaches from area and isolation to percolation, our framework provides a more complete description of how the geography of climate may shape global richness gradients and offers a structural basis for anticipating how future changes in climate connectivity could alter biodiversity patterns.

Complex environments combining multiple stressors are the new norm worldwide. Adaptive evolution will be critical to population persistence under these combined challenges, but how environmental complexity affects the pace of evolution remains poorly understood. Using a meta-experimental evolution approach, we exposed 14 species, from bacteria to unicellular eukaryotes and plants, to single stressors and their pairwise combinations for multiple generations, while keeping the overall stress level comparable. Populations evolving under combined stressors had lower fitness increase in the selective environments, higher fitness reductions in the control environment, and shallower relation between initial maladaptation and fitness gain, than under single stressors. However, these responses varied with species and stressor type. Accounting for such constraints on evolutionary dynamics should prove crucial for the management of biodiversity.

Why independent origins of the same complex physiological trait repeatedly produce similar body forms and physiological changes remains central in evolutionary biology. Endothermy, the internal production and retention of metabolic heat, evolved at least four times in ray-finned fishes, providing natural replicates for convergent genomic signatures. We present a chromosome-scale analysis of three transitions (opah, tunas, and swordfish), including a new chromosome-level genome of the rare, charismatic Pacific oarfish, analyzed with 31 other teleost genomes. The strongest signal is regulatory: of 253,680 conserved noncoding elements, 577 are rate-accelerated in endothermic lineages, with 67 accelerated in all three, exceeding matched ectothermic controls and enriched near developmental transcription factors and Wnt-signaling genes (e.g., irx1a, irx5a, her9, and lmo1). These elements overlap zebrafish developmental enhancers more than expected by chance but are not tied to genes emphasized by expression or coding-selection studies of endothermic lineages, marking a regulatory layer distinct from that metabolic layer. This convergence is part of a broader mosaic: endothermic lineages also share transition-biased substitution and convergent duplication signatures, including excess tandem duplications and lineage-specific gene-family expansion, whereas chromosome organization and protein-coding sequence change little. Endothermic convergence therefore leaves its clearest signal in regulatory remodeling, alongside shifts in substitution bias and gene-family evolution.

Human disturbance drives rapid biodiversity loss, yet how prey diversity shapes individual niche differentiation and scales up to structure social organisation remains poorly understood, particularly in marine systems where prey are largely invisible to observation. Here we mapped prey community structure ( prey fields) using environmental DNA metabarcoding, integrating these maps with stable isotope data and >20 years of behavioural observations on bottlenose dolphins. We show that dolphins specialised on specific prey fields to varying degrees, and that none of 471 tracked individuals changed their predominant prey field -- even through climate-driven habitat reorganisation. Isotope profiles confirmed dietary differentiation scaling with specialisation strength. Across four independent sex-subpopulation networks, prey field overlap predicted social affinity in non-foraging contexts, revealing that prey community structure organises social ties beyond foraging. These results expose a socio-ecological link between prey diversity, trophic ecology, and social organisation, providing a transferable eDNA framework for predators with elusive prey.

During tumor progression, it has been assumed that individual cells that have acquired advantageous mutations overtake the population. Cancers driven by extrachromosomal DNA (ecDNA) do not follow this paradigm. Instead, these tumors have a spectrum of oncogene copy numbers across cells, and graded ecDNA variation may function as a form of bet-hedging that equips tumors with a broad range of phenotypes. Using imaging, single-cell multiomics, and multiplexed proteomics, we systematically characterized ecDNA levels across thousands of single cells. Higher ecDNA dosage produces proportional changes in transcript abundance, chromatin accessibility, protein levels, cell-cycle progression, and proliferation. Genes amplified on ecDNA exhibit distinct transcriptional scaling regimes that shift when the same genes are reintegrated into chromosomal homogeneous staining regions. When we experimentally disrupted the continuum of ecDNA dosage by sorting cells into low- and high-copy number states, the population rapidly recovered its original, continuous distribution. Our time-course data, live-cell imaging, and stochastic models collectively show that restoring this spectrum is an active, deterministic process rather than the passive outcome of random segregation. Together, these findings position ecDNA-mediated expression as a distinct evolutionary mechanism that endows tumors with rapid, population-level adaptability. These findings offer insight into why ecDNA-driven cancers are among the most aggressive and treatment-resistant.

Humans possess human-specific traits1,2 such as spoken language and lineage-specific traits such as ape-specific taillessness3,4. Previous efforts to identify the DNA sequences responsible for such human traits were limited by necessary accommodations for poor genome assembly quality and lack of population genomic sampling5-16. Here, we implement new algorithms that combine the near-complete human reference pangenome alignment with a new near-complete simian cross-species alignment to define human- and lineage-specific DNA sequences fixed across human haplotypes. Previously reported FOXP2/NOVA117,18 amino acid substitutions linked to human spoken language and TBXT transposable element insertion contributing to ape taillessness3,4 were unique to their respective clades and fixed in sampled humans. In contrast, widely used sets of candidate human-mutated loci showed limited enrichment for either human specificity or fixation. Integration with candidate cis-regulatory elements19-22 identified putative regulatory sequences specific to humans and linked to human-specific traits like hair reduction23,24 and brain transcriptome patterning25. Although brain-associated fixed regulatory changes were present in all lineages, enrichment for spoken language was human-specific and enrichment for receptive language was ape-specific. This study provides a new pangenome-aware comparative framework and catalogs of candidate genomic loci to trace the evolutionary origins of common human traits and disease risks.

Tropical rivers in Australia and New Guinea (Sahul) provide a rare natural experiment in vertebrate evolution: unlike other continental systems, their freshwater ichthyofaunas are composed almost entirely of marine-derived lineages rather than primary freshwater fishes. This unique biogeographic setting enables replicated tests of why some marine-to-freshwater transitions give rise to extensive adaptive radiations whereas others remain species-poor, and whether these outcomes reflect ecological opportunity or temporally structured paleoenvironmental constraints. Using a densely sampled, time-calibrated phylogenomic framework spanning 2,303 teleost species, we identified a likely range of 27-34 marine-to-freshwater transitions during the Cenozoic, including a pronounced Middle Miocene peak (16-11 Ma). Although ecological opportunity in Sahul rivers enabled repeated colonization in the absence of dominant primary freshwater incumbents, younger freshwater lineages nevertheless diversify faster than older ones, contradicting the expectation that early arrivers should undergo elevated diversification when accessing vacant niche space. Although some colonizations coincide with bursts of speciation consistent with adaptive radiation, many yielded few species despite long residence times. Functional trait analyses likewise revealed no consistent relationship between colonization timing, ecological breadth, or diversification rate, although expanded functional space characterizes previously proposed Sahul adaptive radiations. Comparisons with paleoenvironmental curves indicate that colonization success correlates with sea-level minima and low-oxygen conditions, suggesting that Earth history dynamics modulated when ecological opportunity was accessible. Our results show that although ecological opportunity enabled repeated freshwater invasions into the Sahul region, diversification outcomes are governed by the interaction of paleoenvironmental dynamics and possibly lineage-specific traits, generating stark asymmetries in freshwater radiations. Significance statementTropical rivers in Australia and New Guinea host one of the most unusual continental freshwater fish assemblages on Earth, composed almost entirely of marine-derived lineages. This system allows asking why some colonizing lineages diversify dramatically while others remain species-poor on a continental scale. Using large-scale phylogenomic and functional trait data, we show that early arrival alone does not predict diversification success. Instead, the lineages that radiate most successfully are those whose arrival coincides with windows of paleoenvironmental opportunity created by sea-level and oxygen fluctuations. These results reveal that the fates of colonizing lineages are shaped not only by ecological opportunity, but also by Earth-history dynamics that govern when, where, and how species can invade and diversify.

The evolution of antibiotic resistance is traditionally understood as a selective sweep to fixation, yielding easily detectable, population-wide resistance. Many clinical isolates, however, exhibit a subtle phenotype in which resistance remains hidden within a susceptible majority despite a clonal genetic background: a phenomenon clinically recognized as heteroresistance (HR). Treatment failure driven by HR has been widely reported across bacterial and fungal infections and in cancer therapy. To understand when and how HR evolves, and why it is selected over classical population-wide resistance, we conducted de novo evolution experiments starting from susceptible Escherichia coli and analyzed the genetic changes and fitness effects in the evolved strains. Prolonged gaps in antibiotic exposure are required for HR to evolve, implicating treatment interruptions as a key driver. HR emerges rapidly and reproducibly with minimal antibiotic use, yet its emergence is not readily detected by routine susceptibility testing. Unlike classical resistance, an evolved HR population partitions at the single-cell level into multiple phenotypes with distinct growth-resistance trade-offs. Their relative abundance shifts dynamically with antibiotic exposure, enabling robust population survival while avoiding the constitutive fitness burden associated with classical resistance. Despite this phenotypic flexibility, stable single mutations including a missense substitution and a short in-frame deletion are sufficient to generate HR, indicating a low evolutionary barrier. Additionally, we found that clinical isolates exhibit genetic and fitness signatures resembling those of our lab-evolved strains, suggesting that clinical HR emerges through the selective mechanism uncovered in our experiments. Together, our results establish HR as a readily evolvable adaptive strategy under treatment interruptions that leverages phenotypic flexibility to maintain resistance at minimal fitness cost, providing mechanistic insight into its emergence and prevalence.

Genetic interventions to modify wild population densities are typically framed around pest suppression, with parallel approaches for boosting beneficial or endangered populations remaining largely undeveloped. Imposing a sustained but non-eliminative genetic load could in principle address both objectives, but existing designs rely on genes with intermediate fitness effects whose loads are difficult to predict under field conditions. Here we describe engineered balanced lethal systems, in which CRISPR-based gene drive establishes two complementing recessive-lethal alleles at a single locus, producing a sustained 50% load through Mendelian segregation. Modelling shows these systems spread from small releases, and that the resulting population-level consequences depend on density regulation and on the timing of lethality: the same 50% load can suppress pests, boost populations of beneficial or endangered species, dampen boom-bust cycles, or raise effective population size. Additional systems at independent loci scale the effect in stepwise increments, and a split-drive variant localises it geographically. These results demonstrate that gene drives imposing genetic load can be expanded beyond elimination, to support and preserve beneficial and endangered populations.

Plant pests threaten 10-40% of global food production, resulting in $55-220 billion in annual economic losses. Despite these escalating risks, biosecurity remains largely reactive, lacking anticipatory frameworks that integrate pest-specific drivers governing transboundary spread. We present GIRAF 1.0 (Global Invasion Risk Assessment Framework), the first quantitative, data-driven system that unifies pest-specific multi-host landscapes, abiotic suitability, and global trade networks with international phytosanitary policies. We applied GIRAF to four globally devastating pests - ranging from viral to insect taxa - to reconstruct a century of transcontinental spread and generate the first multiscale atlases of future invasion potential. GIRAF reveals that 22-37% of Earths land surface can contain host communities that largely overlap with environmentally suitable hotspots. Over 115 countries are highly vulnerable to trade-mediated pest introductions despite adopted phytosanitary policies. GIRAF provides a foundation for proactive surveillance and pandemic preparedness, offering a scalable path for transnational biosecurity agencies and global food industries.