A multi-city examination of neutral and adaptive evolution in the native wildflower Impatiens capensis

A multi-city examination of neutral and adaptive evolution in the native wildflower Impatiens capensis | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL This is a preprint and has not been peer reviewed. Data may be preliminary. 6 February 2025 V1 Latest version Share on A multi-city examination of neutral and adaptive evolution in the native wildflower Impatiens capensis Authors : Ruth Rivkin 0000-0003-2632-3388 [email protected] , Colin Garroway 0000-0002-0955-0688 , and Marc Johnson Authors Info & Affiliations https://doi.org/10.22541/au.173884755.59553944/v1 389 views 236 downloads Contents Abstract Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract Urbanization is a major driver of environmental change that shapes the evolution of populations. However, the effects of environmental differences among cities on neutral and adaptive evolution remains poorly understood. We investigated evolutionary patterns in Impatiens capensis, a native wildflower that can be found in many cities in eastern North America. We used genotype-by-sequencing to evaluate genetic variation, contemporary demographic history, and local adaptation across 10 cities in Ontario, Canada. Urbanization and city size shaped the amount of genetic diversity present at sites and contributed to fine-scale spatial genetic structure. We identified a signal of repeated population bottlenecks occurring across all cities that corresponded to the timing of rapid urban expansion in the region. City size was an important environmental predictor of local adaptation, highlighting the role of cities in driving the adaptive evolution of populations. Our findings provide one of the first examples of parallel demographic shifts in response to urbanization in plants and offer insights into why I. capensis may be particularly resilient to urbanization. Taken together, our results emphasize the role that urban parks can play in maintaining genetic diversity and facilitating adaptation, suggesting that prioritizing greenspace conservation is critical for maintain urban biodiversity. A multi-city examination of neutral and adaptive evolution in the native wildflower Impatiens capensis Running Head Urban evolution across multiple cities L. Ruth Rivkin 1,2,3,4,* , Colin J Garroway 1 , Marc T.J. Johnson 5 1 Department of Biological Sciences, University of Manitoba, Winnipeg, Manitoba, Canada; 2 Polar Bears International, Bozeman, Montana, USA; 3 San Diego Zoo Wildlife Alliance, Escondido, California, USA; 4 University of Toronto, Toronto, Ontario, Canada; 5 University of Toronto Mississauga, Mississauga, Ontario, Canada *Author for correspondence: [email protected] Author details LRR: [email protected] , ORCID: 0000-0003-2632-3388 CCG: [email protected] , ORCID: 0000-0002-0955-0688 MTJJ: [email protected] , ORCID: 0000-0001-9719-0522 Data Availability Statement Raw sequence reads are deposited in the NCBI SRA (BioProject PRJNA1216546). R code and scripts c an be found at https://github.com/ruthrivkin/Impatiens-multicity-evolution. Funding Statement LRR was funded by an NSERC Postdoctoral Research Fellowship and Queen Elizabeth II Scholarship. MTJJ was funded by a Canada Research Chair, Steacie Fellowship and NSERC Discovery Grant. Conflict of Interest Disclosure The authors have no conflict of interest to disclose. Abstract Urbanization is a major driver of environmental change that shapes the evolution of populations. However, the effects of environmental differences among cities on neutral and adaptive evolution remains poorly understood. We investigated evolutionary patterns in Impatiens capensis, a native wildflower that can be found in many cities in eastern North America. We used genotype-by-sequencing to evaluate genetic variation, contemporary demographic history, and local adaptation across 10 cities in Ontario, Canada. Urbanization and city size shaped the amount of genetic diversity present at sites and contributed to fine-scale spatial genetic structure. We identified a signal of repeated population bottlenecks occurring across all cities that corresponded to the timing of rapid urban expansion in the region. City size was an important environmental predictor of local adaptation, highlighting the role of cities in driving the adaptive evolution of populations. Our findings provide one of the first examples of parallel demographic shifts in response to urbanization in plants and offer insights into why I. capensis may be particularly resilient to urbanization. Taken together, our results emphasize the role that urban parks can play in maintaining genetic diversity and facilitating adaptation, suggesting that prioritizing greenspace conservation is critical for maintain urban biodiversity. Effective population size; genetic diversity; Jewelweed; local adaptation; urbanization Introduction Understanding population responses to human-driven environmental change is essential for mitigating biodiversity losses. Changes in land use, such as urbanization, deforestation, and expansion of agricultural lands, are key drivers of recent extirpation and extinction events across the globe (IPBES 2019). Populations that are resilient to environmental change are important contributors to local and regional biodiversity. The amount of genetic variation in a population is a critical component of resiliency to stress and perturbations (Hughes et al. 2008). Populations with more genetic variation are more robust to environmental change because the rate of adaptation directly scales with the amount of genetic variation in fitness (Barrett and Schluter 2008; Kardos et al. 2021). Population size, habitat connectivity, and selection pressure shape the amount of genetic variation within populations. Thus, habitat loss or fragmentation may affect the persistence of populations and can lead to adaptive or maladaptive responses (Huxel and Hastings 1999; Frankham et al. 2019). Species that respond positively or negatively to environmental change can enhance our understanding of how population resiliency can be maintained or lost in the face of habitat fragmentation. Urbanization has caused massive changes to the environment. Cities are one of the fastest growing environments in the world, with more than half of the human population currently living in urban areas (Ritchie et al. 2024). Cities can serve as replicated natural experiments, allowing for repeated tests of environmental change on ecological and evolutionary processes (Johnson and Munshi-South 2017; Santangelo et al. 2020). Urban habitats share many characteristics including impervious surfaces, high levels of disturbance, managed greenspaces, and fragmented remnant natural habitats (McDonald et al. 2020). Additionally, cities that occur in close geographic proximity share regional climatic conditions. Despite these similarities, even nearby cities can differ dramatically due to differences in size, developmental history, and socioeconomic factors (Alberti et al. 2020; Des Roches et al. 2020). Large cities and quickly growing cities can alter genetic variation and exert strong selection pressure on species due to the pace and magnitude of environmental change (Charmantier et al. 2024). Given these differences, studies that compare population genetic responses between nearby cities of different sizes and histories have the power to detect fine-scale differences in evolutionary patterns that may go undetected when examining only a single city (Combs et al. 2018; Miles et al. 2019; Schmidt et al. 2020; Fidino et al. 2021; Santangelo et al. 2022). While many species are extirpated from cities due to habitat loss or stressful conditions, others persist and thrive. These species often exploit urban spaces for habitat, food resources, or dispersal mechanisms. Several well-studied plant species (e.g., white clover, Trifolium repens) have evolved to grow in urban habitats, including lawns, sidewalk cracks, and roadside verges (Cheptou et al. 2008; Larson et al. 2014; Santangelo et al. 2022). Other species occupy urban greenspaces or remnant natural habitats, such as parks or waterfronts. These species may have previously occupied larger areas prior to the growth of the city, but are now restricted to the remaining suitable habitat within a city (Callaghan et al. 2019; Rivkin et al. 2019). Understanding the mechanisms that contribute to the persistence or extirpation of species in greenspaces and remnant natural habitats can provide useful information on how and when populations might be expected to survive in the face of rapid environmental change. We explored the role of environmental heterogeneity within and among cities on patterns of genetic variation and local adaptation in Impatiens capensis (Meerb.) populations. Impatiens capensis is a wildflower native to eastern North America. Populations thrive in wet, shady habitats, such as in forest depressions, ditches, and along the margins of marshes, ponds, lakes, and rivers. In cities, these habitats are found almost exclusively within parks and remnant natural habitats, where I. capensis can be found growing at high densities. Urbanization was associated with declines in genetic variation in I. capensis populations in one large city, Toronto, Canada (Rivkin and Johnson 2022). Despite reductions in genetic diversity, pollination and outcrossing rates remained high for I. capensis , perhaps contributing to its persistence in urban habitats (Barker and Sargent 2020; Rivkin and Johnson 2022). Toronto is the second largest city within the native range of I. capensis , and city size could have driven the observed losses in genetic diversity (Alberti et al. 2020). A general understanding of how urbanization shapes genetic diversity requires examining the genetics of species sampled across cities that differ in size and growth rate. To better understand how environmental variation across cities shapes evolutionary processes, we explored the emergence of population structure, demographic history, and patterns of local adaptation in I. capensis . We expanded on our previous study in I. capensis by exploring patterns of genetic variation and local adaptation across 10 cities located within 100 km of Toronto. These cities varied in potentially important characteristics that could shape the ecology and evolution of I. capensis including human population growth rate, city size, and developmental history. We asked three questions: 1) Do city characteristics and urbanization patterns shape genetic diversity and genetic structure across sites? 2) What is the demographic history of I. capensis across cities? And 3) Do environmental differences within and between cities drive patterns of local adaptation in I. capensis ? Our results shed light on the contributions of city variation to patterns of neutral and adaptive evolution and identify possible avenues for population resilience to large-scale environmental change. Materials and Methods Sampling protocol Impatiens capensis exhibits a mixed mating system, where each plant produces both self-compatible, predominantly outcrossing (chasmogamous) flowers, and closed, obligately self-fertilizing (cleistogamous) flowers. Chasmogamous flowers are pollinated by Bombus spp. , Apis mellifera and Archilochus colubris (Ruby-Throated Hummingbird; (Schemske 1978)). In September 2019, we sampled 53 I. capensis populations growing in urban and rural sites in 10 cities in southern Ontario (Figure S1). Impatiens capensis is native to the region, but due to agricultural and urban development, the species primarily persists in small pockets of natural habitat. Urban sites were sampled within the city borders and were located in parks and along the edges of watersheds. Rural sites were sampled outside of the city borders and were also located near watersheds or along the borders of farmland. Because seed dispersal occurs primarily along waterways, we ensured that sample sites were not part of the same watershed system to avoid sampling populations that were likely to be genetically related. We haphazardly collected 10-15 seeds from five to eight individuals per site that were spaced at least 2 m apart. Immediately after collection, we placed the seeds in a cooler with ice packs. At the end of each day, we transferred the seeds to a -20 °C freezer until DNA extraction. Environment Quantification To investigate the impact of city size on I. capensis genetics, we calculated the area (km 2 ), human population density (number of people per km 2 ), and population growth rate between 2016 and 2021 for each city (Table S1). We obtained all estimates of city size from Government of Canada Census of Population dataset (Statistics Canada 2023). We quantified the environmental conditions at each site to investigate how environmental heterogeneity influences genetic variation in I. capensis. We extracted the percent impervious surface area (ISA) surrounding each site from the 2010 Global Man-made Impervious Surface dataset v1at a spatial resolution of 30 m (Brown de Colstoun et al. 2017). This dataset bins pixels into urban and non-urban categories and uses surface reflectance to estimate the extent of area contained within an urban pixel that is covered by impervious surface (Brown de Colstoun et al. 2017). We measured the vegetation surrounding each site using the Landsat Normalized Difference Vegetation Index (NDVI) from the United States Geological Survey AppEARS database (Didan 2021). We averaged monthly NDVI raster maps from May-September 2019, at a spatial scale of 1 km to obtain a mean growing season NDVI value for our sampling year. Lastly, we estimated climatic variation between populations using mean annual temperature and mean annual precipitation at each site from the Worldclim2 database at a spatial scale of 1 km (Fick and Hijmans 2017). We compiled the separate environmental raster files into a single raster with four layers. We resampled the NDVI and climate layers to match the resolution and extent of the ISA layer with the projectRaster function from the terra v1.7-71 package (Hijmans 2024) in R v4.4.1 (R Development Core Team 2008). We evaluated differences in environment across the study area using a Principal Component Analysis (PCA). Lastly, we extracted values from each layer at every site using a 20 m 2 buffer radius. DNA extraction and sequencing Sequencing was performed in two batches because of delays caused by the COVID-19 pandemic. The first batch was sequenced in 2020 and included samples from Toronto. The results from this batch can be found in Rivkin and Johnson (2022). We sequenced the remaining samples in 2021, following the same DNA extraction and sequencing protocol (Rivkin and Johnson 2022). Prior to extraction, we removed the seed coat from the seeds to prevent contamination from maternal DNA. We extracted DNA from one seed from each plant using the DNeasy Plant Mini Kit in tube format from QIAGEN® (Germantown, MD, USA) and assessed the quantity of DNA from each extraction using a Qubit™ Fluorometer dsDNA High Sensitivity Assay (Thermo Fisher Scientific, Waltham, MA, USA). Library preparation and sequencing were performed by the Elshire Group (Palmerston North, New Zealand) using genotype-by-sequencing (Elshire et al. 2011). Samples were digested with the restriction enzyme, ApeKI, tagged with combinatorial barcodes, and multiplexed into a single library. The samples were sequenced on an Illumina HiSeq XTen machine (Illumina, Inc., San Diego, CA, USA) with 150 bp paired end reads. Variant discovery and filtering To avoid batch effects, we processed reads from all samples, including Toronto, in a single pipeline. We demultiplexed the raw sequences using Axe v0.3.3 (Murray and Borevitz 2018), trimmed adaptor and reverse barcode sequences with the custom batch_trim.pl script (https://github.com/relshire/GBS-PreProcess), removing reads with a Phred score less than 20. We inspected the quality of the reads with fastQC v0.11.4 (Andrews 2015). We used BWA v0.7.12 (Li and Durbin 2009) to map reads to the scaffold level using the I. capensis reference genome (Schoen and Speed 2024), then aligned and sorted reads with samtools v1.17 (Li et al. 2009). We removed duplicate PCR and technical reads with Picard Tools v2.27.4 (Broad Institute 2019). We jointly called SNPs using bcftools v1.17 (Li et al. 2009) with a maximum per sample depth of 50x. We then used bcftools filter to remove reads with low quality scores, strand bias, and read mismatch, multi-allelic calls, and retain reads with a depth >10x and genotype quality >30. As a final step, we used VCFtools v0.1.17 (Danecek et al. 2011) to remove indels and variants with call rates < 80% in all samples and minor allele frequency < 0.01. After filtering, we retained 9,021 high quality SNPs called across nine scaffolds, with an average variant depth of 6.5x. We used a PCA of genetic variation to confirm the absence of batch effects between sequencing runs. The code for variant discovery and filtering and the exact parameters used can be found at https://github.com/ruthrivkin/Impatiens-multicity-evolution. Genetic diversity We first assessed patterns of genetic diversity among sample sites. We used pixy v1.2.10.beta2 (Korunes and Samuk 2021) to calculate nucleotide diversity (π) for each site using both variant and invariant sites. We calculated observed heterozygosity (H O ) for each site with the gl.report.heterozygosity function from the dartR R package v2.9.7 (Gruber et al. 2018). Lastly, we calculated the number of private alleles (PA – i.e., alleles that only occurred in a single population) per site using the gl.report.pa function from the dartR package. For both H O and PA we included only variant sites and used PLINK v1.90 (Chang et al. 2015) to filter SNPs for HWE (p < 0.001) and in linkage disequilibrium ( –indep-pairwise 10 1 0.1 ), retaining 3,261 SNPs. We fit three linear models in R to test for differences in genetic variation across habitats and cities. We modeled changes in π, H O , and PA associated with ISA, NDVI, mean annual temperature, mean annual precipitation, log-transformed city area, and city growth rate. We confirmed that the residuals were not spatially autocorrelated using a Durbin-Watson test (Cliff and Ord 1972) and assessed the assumptions of the models by plotting residuals and examining collinearity between predictors. Population structure We quantified population genetic structure to identify the potential role of environment on relatedness among individuals. We first filtered the dataset for linkage disequilibrium and HWE. We assessed pairwise genetic differentiation (F ST ) among cities using the stammpFst function in the StaMMP v1.6.3 R package (Pembleton et al. 2013). We bootstrapped across loci 1000 times to generate 95% confidence intervals for each pairwise F ST value to determine if cities were significantly differentiated from one another. We also investigated the inbreeding coefficient (F­ IS ) for each site with the gl.report.heterozygosity function from the dartR package. We evaluated genetic structure among individuals using a PCA generated with the –pca flag from PLINK. In addition, we estimated individual ancestry coefficients using sparse nonnegative matrix factorization (sNMF; Frichot et al. 2014), implemented with the snmf function from the R package LEA v3.12.2 (Frichot and François 2015). sNMF takes a non-model based approach to estimate individual admixture coefficients from multilocus genotype data comparable to those from STRUCTURE or ADMIXTURE (Frichot et al. 2014). To test the hypothesis that each city would form a unique genetic cluster, we compared the fit of models with K = 1-15 clusters. We ran each model 10 times and selected the value of K with the lowest cross-entropy score as the best fitting model (Frichot et al. 2014). We next evaluated fine-scale spatial genetic structure using a Moran’s Eigenvector mapping (MEM) based analysis. We generated a genetic distance matrix for all samples using the –distance-matrix flag from PLINK, which calculates pairwise Identity-by-State distances for all pairs of individuals (Chang et al. 2015). We then used the R package MEMGENE v1.0.2 (Galpern et al. 2014) to evaluate genetic relatedness associated with spatial distance between sites. We used the mgQuick function with default parameters to calculate the Euclidean distance among shared alleles between individuals and extracted the first three MEMGENE variables (i.e., the eigenvectors from a PCA of the fitted values from a redundancy analysis of genetic distances). We visualized the variables by mapping them on to the geographic coordinates for each sample. Demographic history We investigated how urbanization has shaped the demographic history of I. capensis across cities. Very recent changes in N e can be estimated by considering the contribution of each generation to linkage disequilibrium among pairs of loci, allowing for detection of N e within 200 generations (Novo et al. 2023). We used GONE (Santiago et al. 2020) to obtain historical estimates of effective population size over 200 generations, with the greatest resolution within 100 generations. Because I. capensis is an annual, this timeframe represents the last 200 years, capturing the development of each city in our dataset (Table S1). We ran GONE to include pairs of loci within 1 cM (hc = 0.01) to account for admixture within cities, with all other settings set to the default. Given the high level of gene flow between cities, we also ran GONE for all samples included as a single population, with 100 replications, to assess if the patterns we observed within each city were driven primarily by admixture. Lastly, we estimated the N e of the parental generation for each city using NeEstimator (Do et al. 2014) implemented through a wrapper function (gl.LDNe) in the dartR.popgen v1.0.0 R package (Gruber et al. 2018). We performed a Waples correction based on the number of chromosomes in the I. capensis genome to improve the accuracy of the estimates (Gruber et al. 2018). Genotype-by-Environment association tests We assessed evidence for local adaptation using three complementary approaches. For all approaches, we filtered for linkage disequilibrium, retaining 3,965 SNPs. We first conducted a Redundancy Analysis (RDA) to identify adaptive loci that may be associated with environmental variation among populations (Rao 1964). RDA is a constrained ordination method that implements multivariate regressions to identify loci that can be linearly explained by different components of the environment (Capblancq et al. 2018). Importantly, RDA can identify both the loci most strongly associated with environmental variables and the selective gradients from combined environmental variables (Capblancq et al. 2018). This approach allows for adaptive variation to be identified in complex environments rather than across a single gradient. We implemented the RDA using the rda function from the vegan v2.6-4 R package (Oksanen et al. 2008) including the same environmental predictors as in the genetic diversity analysis. Prior to running the RDA, we imputed missing SNPs with the impute function from the LEA package with the most likely genotype value computed from the genotype matrix. We identified adaptive loci following Forester et al. (2018). We ran a partial RDA with all environmental variables and conditioned on the first Principal Component (PC1) from the PCA of genetic distances to control for population structure. We identified significant constrained axes with 999 permutations (p < 0.05). We selected candidate adaptive loci from SNPs on the significant axes with loadings greater than ±3 standard deviations away from the mean, which is equivalent to a 1% false discovery rate (FDR). Finally, we identified the environmental variable most strongly correlated with each candidate loci. We also conducted genome scans for outlier loci using sNMF and pcadapt. The sNMF method detects outlier loci that may be under selection using F ST values that account for underlying population structure (Martins et al. 2016). We computed p-values for each loci with K=6 clusters (the best fitting model identified by sNMF) using the snmf.pvalues function from the LEA package. We applied an FDR of 1% when selecting significant outliers to minimize the likelihood of false positives (Benjamini and Hochberg 1995). We also took a PCA-based approach to detect outlier loci using the R package pcadapt v4.3.5 (Luu et al. 2017). This method handles the presence of admixed individuals well and assumes that loci that are correlated with population structure above a specified level are the indicators of local adaptation (Luu et al. 2017). We specified an FDR of 1% and K = 3 clusters based on the PCA plotting results from pcadapt. We considered outlier loci that were identified by all three detection scans to be robust candidates for natural selection. To evaluate the functional role of the shared loci, we investigated whether outlier loci overlapped genes within 1 kb upstream and downstream of the reference genome. We identified regions using BEDTools v2.31.2 (Quinlan and Hall 2010) then queried sequences with BLASTx searches (Altschul et al. 1997), identifying up to 100 matches per region with a cutoff e-value = 0.0001. Results Environmental variation The environmental PCA revealed patterns of environmental variation within and among cities. PC1 explained 57% of the variation in environments across the region and corresponded to increased ISA and temperature and decreased NDVI and precipitation, explaining broad differences between urban and rural environments (Figure S1a). Variation in precipitation was less than 200 mm across the entire region, while temperature varied by 4 °C. PC2 accounted for 21% of the variation across the region, and corresponded to increased ISA, NDVI, and precipitation, and decreased temperature. This variation is best explained by fine-scale heterogeneity likely due to the presence of urban parks and greenspaces (Figure S1b). Genetic diversity We identified clear effects of urban habitat and city size on genetic diversity (Figure 1, Tables S and S2). Population mean π = 0.014 (range = 0.011-0.017), mean H O = 0.108 (range = 0.083-0.131), and mean PA = 2,327 (range = 1,860-2,170). Both π and H O increased with ISA (π: R 2 = 0.08, p = 0.008; H O : R 2 = 0.02, p = 0.021; Table S2), and marginally with NDVI (π: R 2 = 5.67 x 10 -5 , p = 0.071; H O : R 2 = 0.01, p = 0.085; Table S2), suggesting that genetic variation increased in more urban areas and in habitats with more vegetation. In contrast, PA decreased with city area (R 2 = -0.07, p = 0.024), suggesting that sites in large cities contain fewer unique alleles than sites in small cities. Lastly, PA increased with temperature (R 2 = 0.02, p = 0.008) and precipitation (R 2 = 0.05, p = 0.048), suggesting that sites in warmer and wetter habitats were more genetically unique than sites in other environments. Figure 1. Genetic variation associated with urbanization and city size for 53 Impatiens capensis sample sites. A) Nucleotide diversity and B) observed heterozygosity increased with the percentage of impervious surface area (ISA) surrounding a site. C) The number of private alleles per site declined with city size. City size has been log transformed. Population Structure Pairwise F ST between cities were low, varying between 0.01 and 0.04, although all comparisons were statistically significant (Table S3). All estimates of F IS were negative (mean = -0.136, range= -0.272 to -0.018; Table S1), suggesting that there was an excess of heterozygotes within populations. Cities did not form discrete genetic clusters and instead showed evidence of admixture. Plotting the PCA results identified three genetic clusters associated with the first two PC axes (Figure 2A). PC1 explained 7.65% of genetic variation among samples, and while most samples clustered together, one urban site from Toronto separated out as a distinct cluster along the axis. PC2 explained 4.34% of genetic variation, with one rural site from Toronto forming a distinct cluster. The sNMF analysis showed a similar degree of admixture (Figure 2B). The ancestry model fit plateaued at K = 6 clusters (Figure S2), with no single city standing out as genetically unique, although certain ancestral haplotypes were more common in specific cities (e.g., purple in Guelph, red, pink, and orange in Toronto). We identified broad- and fine-scale signals of spatial genetic structure among individuals. The overall MEMGENE model explained 2.9% of genetic variation across all individuals. Within the model, the first three MEMGENE variables explained 38.1%, 22.8%, and 21.5% of shared genetic variation between individuals. Visualization of the MEMGENE eigenvectors identified different components of spatial genetic structure (Figure 2C). The first variable (MEMGENE-1) described variation between western and eastern cities, while MEMGENE-2 identified spatial variation between the southern and northern cities. MEMGENE-3 identified within-city genetic variation, with some urban and rural habitats differing in genetic similarity within cities. Figure 2. A) Principal Component Analysis (PCA) of genetic structure of I. capensis sampled from southern Ontario. Samples are color coded by city, with circles denoting rural sites and triangles denoting urban populations. The percent of variation explained by each PC axis is included in brackets. B) Ancestry proportions for K = 6 clusters generated by the best fitting sNMF model. Individuals are grouped by city ordered from east to west, with urban and rural sites denoted with a U or R, respectively. The colors represent shared genetic ancestry. C) Visualization of MEMGENE axes 1-3, explaining 38.1%, 22.8%, and 21.5% of shared genetic variation between individuals. Shapes denote urban (triangle) and rural (circle) sites, colors denote positive (blue) and negative (pink) eigenvectors, and the size of the shape is proportional to the magnitude of eigenvectors. Demographic history Contemporary N e estimated by NeEstimator was low for all cities (Table S1). The average N e among all cities was 92 plants and was lowest in Toronto ( N e = 33) and highest in Cambridge ( N e = 182). The reconstruction of recent demographic history by GONE identified a steep decline in N e within the last 50 years, consistent with a population bottleneck in each city (Figure 3). Prior to the bottlenecks, N e was between 2 and 5 million plants for all cities. Following the bottleneck events, contemporary N e dropped below 100 plants in all 10 cities. Four cities showed some level of recovery N e following the bottleneck, including Toronto, Milton, Brantford, and Orangeville, with an N e between 300,000 and 1 million plants (Figure 3). When we ran all samples as a single population, we identified a similar bottleneck event occurring being driven primarily by admixture between cities. Figure 3. Recent demographic history of I. capensis within 10 cities in the last 200 generations (1 generation = 1 year) estimated with GONE. Genotype-by-Environment association tests The RDA identified 175 putatively adaptive loci associated with genetic structure (PC1), city size (i.e., city area and growth rate) and habitat heterogeneity (i.e., NDVI, ISA, temperature, and precipitation). The RDA explained 3% of genetic variation across loci, suggesting that the majority of SNPs in our dataset are neutral alleles. The first three RDA axes were significantly associated with genetic variation: explaining 42%, 22%, and 8% of the variation respectively. Toronto samples formed a distinct group from the rest of the cities along all three RDA axes (Figure S4). These differences were most strongly associated with population structure and city size (Figure 4A). RDA3 was also associated with temperature, precipitation, and NDVI (Figure 4B). Population structure was the strongest contributor to multilocus adaptation, with 112 outlier loci most strongly correlated with PC1. City area was most strongly correlated with 50 outlier loci, while the remaining outliers were associated with temperature (4), NDVI (3), and precipitation (2). The sNMF F­ ST -based method identified 459 outlier SNPs, and the pcadapt method identified 126 outliers. Together, these methods identified 117 shared outlier loci. The RDA identified 130 SNPs shared with sNMF, and 13 shared SNPs with pcadapt. In total, 13 outlier loci were identified across all three methods. Of these, two loci were associated with city area and the rest with PC1. The BLASTx query of the 13 consensus outlier loci flagged 290 hits, 75 (26%) of which were associated with city area. Most hits were uncharacterized proteins, although several of the outliers associated with city area were related to defensive functions (extensin and mucin-like proteins). Figure 4. Ordination plots of A) RDA1 and RDA2 and B) RDA1 and RDA3 showing the association between candidate SNPs and environment. SNPs are color coded by their most highly correlated environmental predictor. Neutral SNPs are shown in light grey, and the vectors denote the environmental predictors. Both plots are scaled symmetrically by the square root of the eigenvalues. PC1 is the first principal component axis included from the genetic structure PCA to account for genetic structure in the model. Growth rate refers to the growth in human population size of each city over a 5-year period. Discussion We identified clear effects of urbanization on patterns of genomic evolution in I. capensis . City area and urbanization shaped the amount of genetic diversity present in sites and contributed to fine-scale spatial genetic structure. We identified a signal of repeated demographic shifts across all cities that corresponded to the timing of rapid urban expansion in the region. City size was an important environmental driver of local adaptation among sites, highlighting the role of cities in the adaptive evolution of populations. However, despite the evidence of recent genetic bottlenecks, contemporary genetic diversity and gene flow were high across the area. Together, our results demonstrate the robustness of I. capensis to rapid environmental changes and provide one of the first examples of parallel demographic change in response to urbanization in plants. Genetic Diversity City area and urbanization shaped the amount of genetic diversity present in populations and contributed to fine-scale spatial genetic structure. Surprisingly, increased ISA was associated with increases in nucleotide diversity and heterozygosity. Urban sites were often found in higher quality habitats than rural sites, which may have contributed to the increased amount of genetic diversity with ISA. Within cities, I. capensis grows is often found in large populations in parks or remnant natural habitats, while in rural habitat, populations often grow in roadside ditches with minimal canopy cover and inconsistent water availability. This type of habitat is prone to drying out and is poorly suited for I. capensis growth and survival relative to the more stable habitats found in cities (Simpson et al. 1985). Inconsistent water availability can lead to seasonal population crashes, which can trigger genetic bottlenecks as plants die off during suboptimal growing conditions, reducing the amount of genetic variation found in the population (Bouzat 2010). However, given that genetic variation was on average high across all sites, it is possible that the elevated rates gene flow we observed between urban and rural sites allowed for rapid recovery of genetic variation in populations that experience a crash (Bell et al. 2019). This result suggests that urban sites may act as genetic reservoirs for the more ephemeral rural sites. Interestingly, we found different patterns of genetic variation when we compared the effects of urbanization across 10 cities versus a single city. We had previously identified a negative association between urbanization and genetic diversity across six sites in Toronto (Rivkin and Johnson 2022). However, when we expanded our results to 10 cities, we identified a positive association between urbanization and genetic diversity. By including more populations and cities, we increased the power of our analyses to detect relatively small differences in genetic diversity and expanded the range of habitat variation included in our analysis. We also aligned our results to a reference genome instead of calling variants de novo and likely called a different set of SNPs in the current study, which may have contributed to these differences. When we look to other systems, the effects of urbanization on genetic diversity are mixed. A review of 194 urban genetic diversity studies identified a positive effect of urbanization on genetic diversity in 32% of cases (Miles et al. 2019). A separate review of mammal and bird genetic diversity identified a negative effect of urbanization on genetic diversity in mammals, but no effect in birds (Schmidt et al. 2020). In plants, genetic diversity has been negatively and positively associated with urbanization (Bartlewicz et al. 2015; Johnson et al. 2018; Rivkin and Johnson 2022; Caizergues et al. 2024), and in some cases urban populations act as reservoirs that harbor essential genetic diversity that maintains variation across the broader metapopulation (Roberts et al. 2007). Our results add to the growing body of evidence that demonstrate the complex responses of species to urbanization. Although we found that genetic diversity was negatively associated with ISA, larger cities harbored less unique genetic variation than smaller cities. City area was negatively associated with the number of private alleles in a population, a trend that was primarily driven by Toronto. Increased gene flow and admixture in larger cities could have contributed to the decline in PA with city size. Pollinators are the primary contributors of gene flow in I. capensis , and pollinator services were maintained across an urbanization gradient in the nearby city of Ottawa, CA (Barker and Sargent 2020). Large cities can harbor larger pollinator communities than small cities, potentially contributing to increased gene flow (Udy et al. 2020; Fauviau et al. 2024). However, despite high levels of admixture across all cities, we also observed the most genetic structure in Toronto, which is inconsistent with the hypothesis that increased gene flow in large cities reduced the genetic uniqueness of those cities. An alternative explanation is that our results are capturing the after-effects of a population bottleneck event. The number of unique alleles in a population may be lower following a bottleneck event or founder effect if the event reduces population genetic diversity to a subset of original levels or results in the loss of rare alleles (Slatkin and Takahata 1985; Kalinowski 2004). This effect may be particularly noticeable in populations with historically high levels admixture that experienced a relatively recent bottleneck and have not yet fully recovered. The current N e of cities estimated by NeEstimator was lowest in Toronto, suggesting that this city may not have recovered from the most recent bottleneck event, perhaps explaining the lack of unique alleles in the city. Demographic History We identified repeated shifts in demographic history associated with a period of rapid urban expansion across cities. Results from the GONE analysis suggest that all cities experienced a severe population bottleneck that occurred between 10-50 years before sampling. We identified the same pattern when we grouped all cities together in a separate analysis, suggesting that the region may have experienced a broader population crash. Although approximately half of the cities exhibited some level of recovery, current N e values (generations 1-5) from GONE can be unreliable when dealing with reduced representation genotype data and high levels of admixture (Novo et al. 2023). Gene flow within cities may also upwardly skew GONE N e estimates (Gargiulo et al. 2024), and we observed much lower contemporary N e with NeEstimator, suggesting that most cities have not recovered from the bottleneck. While the exact N e from GONE should be interpreted with caution, to our knowledge this is the first evidence in plants of parallel population crashes documented across cities. In contrast to our results, a global study of 160 cities that tested for parallel shifts in demographic history in white clover found that the species maintained higher N e in urban relative to rural habitats (Caizergues et al. 2024). However, demographic declines were observed in fire salamander ( Salamandra salamandra ), corresponding to the historical development of neighborhoods in Oviedo, Spain (Lourenço et al. 2017). Given that urbanization can lead to massive changes to the environment, it is perhaps unsurprising that we observed such parallel responses to urban development. The mixed mating system of I. capensis may have contributed to the demographic shifts we observed. Impatiens capensis produces two flower types, open flowers that outcross via pollinators, and closed flowers that self-fertilize. We demonstrated that populations primarily relied on outcrossing across an urbanization gradient in Toronto (Rivkin and Johnson 2022). However, it is possible that during times of rapid urban development that affect pollinator services, I. capensis populations switch to a reliance on selfing. Selfing increases inbreeding rates and can drive down N e , leading to severe bottlenecks such as those observed here (Wright et al. 2013; Wang et al. 2016). Our results suggest that mating system and life history strategies contribute to demographic responses to urbanization. Gene Flow Despite the evidence of recent bottlenecks in all cities, gene flow was maintained across the region. At the broadest scale, most samples clustered together. Finer scaled reconstruction of population structure identified six clusters that were unrelated to city or habitat. While Toronto largely formed two distinct clusters, the other cities were admixed. This result suggests that gene flow between cities remains elevated, possibly because pollination services to I. capensis are robust to the effects of urbanization (Barker and Sargent 2020). However, we detected a signal of population structure even at the finest scales (i.e. within a city). Latitude, longitude, and within-city environmental variation all contributed to population differentiation. Population differentiation was likely driven by fine-scale environmental variation, including ISA, precipitation and temperature, which were also associated variation in genetic diversity. Our results are consistent with findings from other studies which have identified weak or non-existent signatures of urbanization on plant genetic structure (Mollashahi et al. 2023; Ruas et al. 2024; Taichi et al. 2024). Gene flow is likely a core contributor to the resilience of I. capensis to rapid urbanization. Local Adaptation City size was the main environmental driver of local adaptation among populations, highlighting the role that city variation can play in adaptive evolution. Our genome scans identified population structure as the biggest predictor of outlier loci, which is unsurprising given that we used a reduced-representation sequencing approach (Hoban et al. 2016). After population structure, city area was the predictor most strongly associated with the majority of the outlier loci, a trend that was again driven primarily by Toronto. Toronto is the second largest city in eastern North America and may impose stronger selection pressure on species than smaller cities. Although most putatively adaptive loci were unidentified, several were associated with defensive functions. It is possible that increased herbivory pressure in larger cities may have led to selection on defensive genes. City parks can support greater abundances of herbivores than surrounding rural habitats, particularly compared to sites near agricultural fields that are regularly sprayed with pesticides (Rivkin and De Andrade 2023). Increased herbivore pressure could select for novel genes or traits that promote survival in an urban environment. Although I. capensis is a classic system used to study rapid adaptation to environmental variation (Donohue and Schmitt 1999; Heschel and Riginos 2005; Stinchcombe et al. 2010; Zhao and Schoen 2022), much less is known about the genomic architecture of adaptive traits in the species. With the development of a robust reference genome (Schoen and Speed 2024) , it is now possible to use a common garden approach to identify the underlying genetic contributions of adaptation to urbanization in I. capensis . Conclusions Our study examined patterns of neutral and adaptive evolution in a species that can grow at high abundance in cities. We identified the first evidence of repeated demographic shifts across cities in plants, despite increased genetic diversity in more urban sites. Our findings demonstrate the complexity of evolutionary responses to urbanization and highlight the need for studies to consider multiple cities when assessing the influence of urbanization on patterns of genetic variation. Parks and remnant natural habitats are important resources for many urban species. These habitats can play a large role in supporting evolutionary processes in cities. For I. capensis , and many other native species, parks are the only habitat which allows for survival in cities. These essential greenspaces facilitate eco-evolutionary processes in many species and provide long-term benefits for humans (Rivkin et al. 2019; Des Roches et al. 2020). Prioritizing the development and conservation of parks during city planning will contribute to the preservation and resilience of urban biodiversity. Acknowledgments We thank Inder Sheoran for valuable help with DNA extractions and Evelien de Greef for helpful R Code. The Elshire Group prepared the libraries and sequenced the DNA Literature Cited Alberti, M., E. P. Palkovacs, S. Des Roches, L. De Meester, K. I. Brans, L. Govaert, N. B. Grimm, N. C. Harris, A. P. Hendry, C. J. Schell, M. Szulkin, J. Munshi-South, M. C. Urban, and B. C. Verrelli. 2020. The complexity of urban eco-evolutionary dynamics. BioScience 70:772–793. Altschul, S. F., T. L. Madden, A. A. Schäffer, J. Zhang, Z. Zhang, W. Miller, and D. J. Lipman. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Research 25:3389–3402. Andrews, S. A. 2015. Quality control tool for high throughput sequence data. Babraham Institute. Barker, C. A., and R. D. Sargent. 2020. Pollination services to Impatiens capensis (Balsaminaceae) are maintained across an urbanization gradient. International Journal of Plant Sciences 181:937–944. Barrett, R. D. H., and D. Schluter. 2008. Adaptation from standing genetic variation. Trends in Ecology and Evolution 23:38–44. Bartlewicz, J., K. Vandepitte, and H. Jacquemyn. 2015. Population genetic diversity of the clonal self-incompatible herbaceous plant Linaria vulgaris along an urbanization gradient. Biological Journal of the Linnean Society 116:603–613. Bell, D. A., Z. L. Robinson, W. C. Funk, S. W. Fitzpatrick, F. W. Allendorf, D. A. Tallmon, and A. R. Whiteley. 2019. The exciting potential and remaining uncertainties of genetic rescue. Trends in Ecology & Evolution 34:1070–1079. Benjamini, Y., and Y. Hochberg. 1995. Controlling the false discovery rate: A practical and powerful approach to multiple testing. Journal of the Royal Statistical Society. Series B (Methodological) 57:289–300. Bouzat, J. L. 2010. Conservation genetics of population bottlenecks: the role of chance, selection, and history. Conservation Genetics 11:463–478. Broad Institute. 2019. Picard Toolkit. Broad Institue . Brown de Colstoun, E. C., C. Huang, P. Wang, J. C. Tilton, B. Tan, J. Phillips, S. Niemczura, P.-Y. Ling, and R. E. Wolfe. 2017. Global Man-made Impervious Surface (GMIS) dataset from Landsat. Palisades, NY: NASA Socioeconomic Data and Applications Center (SEDAC). Caizergues, A. E., J. S. Santangelo, R. W. Ness, F. Angeoletto, D. N. Anstett, J. Anstett, F. Baena-Diaz, E. J. Carlen, J. A. Chaves, M. S. Comerford, K. Dyson, M. Falahati-Anbaran, M. D. E. Fellowes, K. A. Hodgins, G. R. Hood, C. Iñiguez-Armijos, N. J. Kooyers, A. Lázaro-Lobo, A. T. Moles, J. Munshi-South, J. Paule, I. M. Porth, L. Y. Santiago-Rosario, K. S. Whitney, A. J. M. Tack, and M. T. J. Johnson. 2024. Does urbanisation lead to parallel demographic shifts across the world in a cosmopolitan plant? Molecular Ecology 33:e17311. Callaghan, C. T., G. Bino, R. E. Major, J. M. Martin, M. B. Lyons, and R. T. Kingsford. 2019. Heterogeneous urban green areas are bird diversity hotspots: insights using continental-scale citizen science data. Landscape Ecology 34:1231–1246. Capblancq, T., K. Luu, M. G. B. Blum, and E. Bazin. 2018. Evaluation of redundancy analysis to identify signatures of local adaptation. Molecular Ecology Resources 18:1223–1233. Chang, C. C., C. C. Chow, L. C. A. M. Tellier, S. Vattikuti, S. M. Purcell, and J. J. Lee. 2015. Second-generation PLINK: Rising to the challenge of larger and richer datasets. GigaScience 4:1–16. Charmantier, A., T. Burkhard, L. Gervais, C. Perrier, A. I. Schulte-Hostedde, and M. J. Thompson. 2024. How does urbanization affect natural selection? Functional Ecology 38:2522–2536. Cheptou, P.-O., O. Carrue, S. Rouifed, and A. Cantarel. 2008. Rapid evolution of seed dispersal in an urban environment in the weed Crepis sancta . Proceedings of the National Academy of Sciences 105:3796–3799. Cliff, A., and K. Ord. 1972. Testing for spatial autocorrelation among regression residuals. Geographical Analysis 4:267–284. Combs, M., K. A. Byers, B. M. Ghersi, M. J. Blum, A. Caccone, F. Costa, C. G. Himsworth, J. L. Richardson, and J. Munshi-South. 2018. Urban rat races: spatial population genomics of brown rats ( Rattus norvegicus ) compared across multiple cities. Proceedings of the Royal Society B: Biological Sciences 285:20180245. Danecek, P., A. Auton, G. Abecasis, C. A. Albers, E. Banks, M. A. DePristo, R. E. Handsaker, G. Lunter, G. T. Marth, S. T. Sherry, G. McVean, and R. Durbin. 2011. The variant call format and VCFtools. Bioinformatics 27:2156–2158. Des Roches, S., K. I. Brans, M. R. Lambert, L. R. Rivkin, A. M. Savage, C. J. Schell, C. Correa, L. De Meester, S. E. Diamond, N. B. Grimm, N. C. Harris, L. Govaert, A. P. Hendry, M. T. J. Johnson, J. Munshi-South, E. P. Palkovacs, M. Szulkin, M. C. Urban, B. C. Verrelli, and M. Alberti. 2020. Socio-eco-evolutionary dynamics in cities. Evolutionary Applications 14: 248-267. Didan, K. (2021). MODIS/Terra Vegetation Indices Monthly L3 Global 1km SIN Grid V061 . NASA EOSDIS Land Processes Distributed Active Archive Center. https://doi.org/10.5067/MODIS/MOD13A3.061. Accessed March 22, 2024 Do, C., R. S. Waples, D. Peel, G. M. Macbeth, B. J. Tillett, and J. R. Ovenden. 2014. NeEstimator v2: re-implementation of software for the estimation of contemporary effective population size ( N e ) from genetic data. Molecular Ecology Resources 14:209–214. Donohue, K., and J. Schmitt. 1999. The genetic architecture of plasticity to density in Impatiens capensis . Evolution 53:1377–1386. Elshire, R. J., J. C. Glaubitz, Q. Sun, J. A. Poland, K. Kawamoto, E. S. Buckler, and S. E. Mitchell. 2011. A robust, simple genotyping-by-sequencing (GBS) approach for high diversity species. PLoS ONE 6:1–10. Fauviau, A., W. Fiordaliso, A. Fisogni, L. Fortel, F. Francis, B. Geslin, N. Hautekèete, C. Heiniger, O. Lambert, V. L. Feon, F. Massol, A. Michelot-Antalik, D. Michez, H. Mouret, G. Noël, Y. Piquot, L. Ropars, L. Schurr, C. V. Reeth, V. Zaninotto, I. Dajoz, and M. Henry. 2024. Larger cities host richer bee faunas, but are no refuge for species with concerning conservation status: Empirical evidence from Western Europe. Basic and Applied Ecology 79:131–140. Fick, S. E., and R. J. Hijmans. 2017. WorldClim 2: new 1-km spatial resolution climate surfaces for global land areas. International Journal of Climatology 37:4302–4315. Fidino, M., T. Gallo, E. W. Lehrer, M. H. Murray, C. A. M. Kay, H. A. Sander, B. MacDougall, C. M. Salsbury, T. J. Ryan, J. L. Angstmann, J. Amy Belaire, B. Dugelby, C. J. Schell, T. Stankowich, M. Amaya, D. Drake, S. H. Hursh, A. A. Ahlers, J. Williamson, L. M. Hartley, A. J. Zellmer, K. Simon, and S. B. Magle. 2021. Landscape-scale differences among cities alter common species’ responses to urbanization. Ecological Applications 31:e02253. Forester, B. R., J. R. Lasky, H. H. Wagner, and D. L. Urban. 2018. Comparing methods for detecting multilocus adaptation with multivariate genotype–environment associations. Molecular Ecology 27:2215–2233. Frankham, R., J. D. Ballou, K. Ralls, M. Eldridge, M. R. Dudash, C. B. Fenster, R. C. Lacy, and P. Sunnucks. 2019. A Practical Guide for Genetic Management of Fragmented Animal and Plant Populations. Oxford University Press. Frichot, E., and O. François. 2015. LEA: An R package for landscape and ecological association studies. Methods in Ecology and Evolution 6:925–929. Frichot, E., F. Mathieu, T. Trouillon, G. Bouchard, and O. François. 2014. Fast and efficient estimation of individual ancestry coefficients. Genetics 196:973–983. Galpern, P., P. R. Peres-Neto, J. Polfus, and M. Manseau. 2014. MEMGENE: Spatial pattern detection in genetic distance data. Methods in Ecology and Evolution 5:1116–1120. Gargiulo, R., V. Decroocq, S. C. González-Martínez, I. Paz-Vinas, J.-M. Aury, I. Lesur Kupin, C. Plomion, S. Schmitt, I. Scotti, and M. Heuertz. 2024. Estimation of contemporary effective population size in plant populations: Limitations of genomic datasets. Evolutionary Applications 17:e13691. Gruber, B., P. J. Unmack, O. F. Berry, and A. Georges. 2018. dartr: An R package to facilitate analysis of SNP data generated from reduced representation genome sequencing. Molecular Ecology Resources 18:691–699. Heschel, M. S., and C. Riginos. 2005. Mechanisms of selection for drought stress tolerance and avoidance in Impatiens capensis (Balsaminaceae). American Journal of Botany 92:37–44. Hijmans, R. 2024. terra: Spatial Data Analysis. CRAN: Contributed Packages . Hoban, S., J. L. Kelley, K. E. Lotterhos, M. F. Antolin, G. Bradburd, D. B. Lowry, M. L. Poss, L. K. Reed, A. Storfer, and M. C. Whitlock. 2016. Finding the genomic basis of local adaptation: pitfalls, practical solutions, and future directions. The American Naturalist 188:379–397. Hughes, A. R., B. D. Inouye, M. T. J. Johnson, N. Underwood, and M. Vellend. 2008. Ecological consequences of genetic diversity. Ecology Letters 11:609–623. Huxel, G. R., and A. Hastings. 1999. Habitat loss, fragmentation, and restoration. Restoration Ecology 7:309–315. IPBES. 2019. Global assessment report on biodiversity and ecosystem services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. IPBES Secretariat, Bonn, Germany. Johnson, M. T. J., and J. Munshi-South. 2017. Evolution of life in urban environments. Science 358:eaam8327. Johnson, M. T. J., C. M. Prashad, M. Lavoignat, and H. S. Saini. 2018. Contrasting the effects of natural selection, genetic drift and gene flow on urban evolution in white clover ( Trifolium repens ). Proceedings of the Royal Society B: Biological Sciences 285:20181019. Kalinowski, S. T. 2004. Counting alleles with rarefaction: private alleles and hierarchical sampling designs. Conservation Genetics 5:539–543. Kardos, M., E. E. Armstrong, S. W. Fitzpatrick, S. Hauser, P. W. Hedrick, J. M. Miller, D. A. Tallmon, and W. C. Funk. 2021. The crucial role of genome-wide genetic variation in conservation. Proceedings of the National Academy of Sciences 118:e2104642118. Korunes, K. L., and K. Samuk. 2021. pixy: Unbiased estimation of nucleotide diversity and divergence in the presence of missing data. Molecular Ecology Resources 21:1359–1368. Larson, J. L., A. J. Kesheimer, and D. A. Potter. 2014. Pollinator assemblages on dandelions and white clover in urban and suburban lawns. Journal of Insect Conservation 18:863–873. Li, H., and R. Durbin. 2009. Fast and accurate short read alignment with Burrows–Wheeler transform. Bioinformatics 25:1754–1760. Li, H., B. Handsaker, A. Wysoker, T. Fennell, J. Ruan, N. Homer, G. Marth, G. Abecasis, R. Durbin, and 1000 Genome Project Data Processing Subgroup. 2009. The sequence alignment/map format and SAMtools. Bioinformatics 25:2078–2079. Lourenço, A., D. Álvarez, I. J. Wang, and G. Velo-Antón. 2017. Trapped within the city: integrating demography, time since isolation and population-specific traits to assess the genetic effects of urbanization. Molecular Ecology 26:1498–1514. Luu, K., E. Bazin, and M. G. B. Blum. 2017. pcadapt: an R package to perform genome scans for selection based on principal component analysis. Molecular Ecology Resources 17:67–77. Martins, H., K. Caye, K. Luu, M. G. B. Blum, and O. François. 2016. Identifying outlier loci in admixed and in continuous populations using ancestral population differentiation statistics. Molecular Ecology 25:5029–5042. McDonald, R. I., A. V. Mansur, F. Ascensão, M. Colbert, K. Crossman, T. Elmqvist, A. Gonzalez, B. Güneralp, D. Haase, M. Hamann, O. Hillel, K. Huang, B. Kahnt, D. Maddox, A. Pacheco, H. M. Pereira, K. C. Seto, R. Simkin, B. Walsh, A. S. Werner, and C. Ziter. 2020. Research gaps in knowledge of the impact of urban growth on biodiversity. Nature Sustainability 3:16–24. Springer US. Miles, L. S., L. R. Rivkin, M. T. J. Johnson, J. Munshi‐South, and B. C. Verrelli. 2019. Gene flow and genetic drift in urban environments. Molecular Ecology 28:4138–4151. Mollashahi, H., J. Urbaniak, T. H. Szymura, and M. Szymura. 2023. Genetic structure of Trifolium pratense populations in a cityscape. PeerJ 11:e15927. Murray, K. D., and J. O. Borevitz. 2018. Axe: Rapid, competitive sequence read demultiplexing using a trie. Bioinformatics 34:3924–3925. Novo, I., P. Ordás, N. Moraga, E. Santiago, H. Quesada, and A. Caballero. 2023. Impact of population structure in the estimation of recent historical effective population size by the software GONE. Genetics Selection Evolution 55:86. Oksanen, J., R. Kindt, P. Legendre, B. O’Hara, G. L. Simpson, P. Solymos, H. M. H. Stevens, and H. Wagner. 2008. The vegan Package: Community Ecology Package. R package version 2.6-4. Pembleton, L. W., N. O. I. Cogan, and J. W. Forster. 2013. StAMPP: an R package for calculation of genetic differentiation and structure of mixed-ploidy level populations. Molecular Ecology Resources 13:946–952. Quinlan, A. R., and I. M. Hall. 2010. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics 26:841–842. R Development Core Team. 2008. R: A language and environment for statistical computing. R. Foundation for Statistical Computing, Vienna, Austria. Rao, C. R. 1964. The use and interpretation of Principal Component Analysis in applied research. Sankhyā: The Indian Journal of Statistics, Series A 26:329–358. Ritchie, H., V. Samborska, and M. Roser. 2024. Urbanization. Our World in Data. Rivkin, L. R., and A. C. De Andrade. 2023. Increased herbivory but not cyanogenesis is associated with urbanization in a tropical wildflower. Austral Ecology 48:388–398. Rivkin, L. R., and M. T. J. Johnson. 2022. The impact of urbanization on outcrossing rate and population genetic variation in the native wildflower, Impatiens capensis . Journal of Urban Ecology 8:juac009. Rivkin, L. R., J. S. Santangelo, M. Alberti, M. F. J. Aronson, C. W. de Keyzer, S. E. Diamond, M. J. Fortin, L. J. Frazee, A. J. Gorton, A. P. Hendry, Y. Liu, J. B. Losos, J. S. MacIvor, R. A. Martin, M. J. McDonnell, L. S. Miles, J. Munshi-South, R. W. Ness, A. E. M. Newman, M. R. Stothart, P. Theodorou, K. A. Thompson, B. C. Verrelli, A. Whitehead, K. M. Winchell, and M. T. J. Johnson. 2019. A roadmap for urban evolutionary ecology. Evolutionary Applications 12:384–398. Roberts, D. G., D. J. Ayre, and R. J. Whelan. 2007. Urban plants as genetic reservoirs or threats to the integrity of bushland plant populations. Conservation Biology 21:842–852. Ruas, R. de B., S. M. de Godoy, D. C. Feliciano, C. de F. Ruas, and F. Bered. 2024. A bromeliad living in the city: a case of a native species resilient to urbanization in South Brazil. Botanical Journal of the Linnean Society 205:161–176. Santangelo, J. S., R. W. Ness, B. Cohan, C. R. Fitzpatrick, S. G. Innes, S. Koch, L. S. Miles, S. Munim, P. R. Peres-Neto, C. Prashad, et al , and M. T. J. Johnson. 2022. Global urban environmental change drives adaptation in white clover. Science 375:1275–1281. Santangelo, J. S., L. S. Miles, S. T. Breitbart, D. Murray-Stokeſ, L. R. Rivkin, M. T. J. Johnson, and R. W. Ness. 2020. Urban environments as a framework to study parallel evolution. in M. Szulkin, J. Munshi-South, and A. Charmantier, eds. Urban Evolutionary Biology. Oxford University Press, USA. Santiago, E., I. Novo, A. F. Pardiñas, M. Saura, J. Wang, and A. Caballero. 2020. Recent demographic history inferred by high-resolution analysis of linkage disequilibrium. Molecular Biology and Evolution 37:3642–3653. Schemske, D. W. 1978. Evolution of reproductive characteristics in Impatiens (Balsaminaceae): the significance of cleistogamy and chasmogamy. Ecology 59:596–613. Schmidt, C., M. Domaratzki, R. P. Kinnunen, J. Bowman, and C. J. Garroway. 2020. Continent-wide effects of urbanization on bird and mammal genetic diversity. Proceedings of the Royal Society B: Biological Sciences 287. Schoen, D. J., and D. Speed. 2024. The heritability of fitness in a wild annual plant population with hierarchical size structure. Evolution 78:1739–1745. Simpson, R. L., M. A. Leck, and V. T. Parker. 1985. The comparative ecology of Impatiens capensis Meerb. (Balsaminaceae) in central New Jersey. Bulletin of the Torrey Botanical Club 112:295–311. Torrey Botanical Society. Slatkin, M., and N. Takahata. 1985. The average frequency of private alleles in a partially isolated population. Theoretical Population Biology 28:314–331. Statistics Canada. 2023. Census Profile. 2021 Census of Population. Statistics Canada Catalogue number 98-316-X2021001. Ottawa. Released November 15, 2023. https://www12.statcan.gc.ca/census-recensement/2021/dp-pd/prof/index.cfm?Lang=E. Accessed March 14, 2024. Stinchcombe, J. R., R. Izem, M. S. Heschel, B. V. McGoey, and J. Schmitt. 2010. Across-environment genetic correlations and the frequency of selective environments shape the evolutionary dynamics of growth rate in Impatiens capensis . Evolution 64:2887–2903. Taichi, N., N. Nakahama, N. Ohmido, and A. Ushimaru. 2024. Habitat diversification associated with urban development has a little effect on genetic structure in the annual native plant Commelina communis in an East Asian megacity. Ecology and Evolution 14:e10975. Udy, K. L., H. Reininghaus, C. Scherber, and T. Tscharntke. 2020. Plant–pollinator interactions along an urbanization gradient from cities and villages to farmland landscapes. Ecosphere 11:e03020. Wang, J., E. Santiago, and A. Caballero. 2016. Prediction and estimation of effective population size. Heredity 117:193–206. Wright, S. I., S. Kalisz, T. Slotte, and S. I. Wright. 2013. Evolutionary consequences of self- fertilization in plants. Proceedings of the Royal Society B: Biological Sciences 280:20130133. Zhao, Y., and D. J. Schoen. 2022. Relaxed selection and the evolution of the chasmogamous flower of Impatiens capensis (Balsaminaceae). Evolutionary Ecology 36:233–250. Data Accessibility and Benefit-Sharing Section Raw sequence reads are deposited in the NCBI SRA (BioProject PRJNA1216546). R code and scripts c an be found at https://github.com/ruthrivkin/Impatiens-multicity-evolution. Benefits Generated: Benefits from this research accrue from the sharing of our data and results on urban biodiversity as described above. Information & Authors Information Version history V1 Version 1 06 February 2025 Copyright This work is licensed under a Non Exclusive No Reuse License. Keywords effective population size genetic diversity jewelweed local adaptation urbanization Authors Affiliations Ruth Rivkin 0000-0003-2632-3388 [email protected] University of Manitoba View all articles by this author Colin Garroway 0000-0002-0955-0688 University of Manitoba View all articles by this author Marc Johnson University of Toronto Mississauga View all articles by this author Metrics & Citations Metrics Article Usage 389 views 236 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Ruth Rivkin, Colin Garroway, Marc Johnson. A multi-city examination of neutral and adaptive evolution in the native wildflower Impatiens capensis. Authorea . 06 February 2025. DOI: https://doi.org/10.22541/au.173884755.59553944/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); Cited by Aiden M. Stanley, Tia‐Lynn Ashman, Urbanization Alters Phenology, Mating System Allocation, and Life History of Impatiens capensis (Balsaminaceae) via Trait‐Specific Plasticity and Genetic Differentiation , Ecology and Evolution, 15 , 6, (2025). https://doi.org/10.1002/ece3.71583 Crossref Loading... View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.173884755.59553944/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9ff75eb228688650',t:'MTc3OTQwNjIzNg=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();