Comparative genomics revealed new insights into the plastome evolution of Ludwigia (Onagraceae, Myrtales).

The primrose-willow (Ludwigia L.), a well-defined genus of the Onagraceae family, comprises 87 species widely distributed worldwide. In this study, we sequenced and characterized the complete chloroplast (cp) genomes of three species in the genus, including Ludwigia adscendens, Ludwigia hyssopifolia, and Ludwigia prostrata. Three Ludwigia cp genomes ranged from 158,354 to 159,592 bp in size, and each contained 113 genes, including 79 unique protein-coding genes (PCGs), four rRNA genes, and 30 tRNA genes. A comparison of the Ludwigia cp genomes revealed that they were highly conserved in gene composition, gene orientation, and GC content. Moreover, we compared the structure of cp genomes and reconstructed phylogenetic relationships with related species in the Onagraceae family. Regarding contraction/expansion of inverted repeat (IR) region, two kinds of expansion IR region structures were found in Oenothera, Chamaenerion, and Epilobium genera, with primitive IR structures in Ludwigia and Circeae genera. The regions clpP, ycf2, and ycf1 genes possessed highly divergent nucleotides among all available cp genomes of the Onagraceae family. The phylogenetic reconstruction using 79 PCGs from 39 Onagraceae cp genomes inferred that Ludwigia (including L. adscendens, L. hyssopifolia, L. prostrata, and Ludwigia octovalvis) clade was monophyletic and well-supported by the bootstrap and posterior probability values. This study provides the reference cp genomes of three Ludwigia species, which can be used for species identification and phylogenetic reconstruction of Ludwigia and Onagraceae taxa.

Immediate vs. evolutionary consequences of polyploidy on clonal reproduction in an autopolyploid plant.

Background and Aims: Clonal reproduction in polyploids is expected to exceed that in diploids, due to either the immediate direct effects of whole-genome duplication (WGD) or selection during establishment. The timing of polyploidy effects on clonality are largely unknown despite its hypothesized influence on polyploid success. This study tests the direction and timing of divergence in clonal traits in diploid and polyploid Chamerion angustifolium. Methods: Root bud production and biomass allocation patterns were compared between diploids and synthesized tetraploids (neotetraploids), and between neotetraploids and naturally occurring tetraploids grown in a common environment. Key Results: Neotetraploids produced more root buds and fewer sexual structures than diploids and natural tetraploids; diploids and natural tetraploids had similar root bud numbers and sexual investment. The root bud:inflorescence biomass ratio was 71 % higher in neotetraploids than in natural tetraploids. Root bud location suggests that ramet density in neotetraploid genets could be higher than in diploid genets. Conclusions: WGD immediately increases investment in asexual vs. sexual reproduction in C. angustifolium, potentially promoting within-cytotype mating and establishment for neopolyploids. However, evolutionary change after the polyploidization event negates the direct effects of WGD. Natural polyploids and diploids have similar root bud production and biomass allocation patterns, probably resulting from habitat- and ploidy-mediated selection on polyploids to become more like diploids. These results highlight the value of studying the effects of polyploidization in young vs. established polyploids.

Genome skimming provides new insight into the relationships in Ludwigia section Macrocarpon, a polyploid complex.

PREMISE OF THE STUDY: Interpreting relationships within groups containing polyploids, which are frequent in angiosperms, can be greatly assisted by genomic techniques. In this study, we used a genome-skimming approach to investigate the evolutionary relationships and origins of polyploids in the monophyletic group, Ludwigia section Macrocarpon (Onagraceae), which includes diploid, tetraploid, and hexaploid taxa. METHODS: We sampled all known taxa and ploidy levels in the section and conducted shotgun sequencing. We assembled plastomes, mitochondrial sequences, and completed nuclear ribosomal regions, reconstructed phylogenies, and conducted comparative genomic analyses for plastomes to gain insights into the relationships among studied taxa. KEY RESULTS: Within the section, results showed that the South American diploid taxa L. bonariensis and L. lagunae were closely related. We reported the first chromosome count (2n = 4× = 32) for L. neograndiflora, which is closely related to the two South American diploid taxa, although its exact origin remains unclear. The samples of the widespread, polyploid taxon L. octovalvis do not form a monophyletic group. Both tetraploid and hexaploid L. octovalvis lineages have originated more than once. At least one tetraploid in the L. octovalvis lineage may have been involved in the origins of hexaploids. One or more extinct/unsampled intermediate tetraploids in the L. octovalvis lineages had also likely been involved in the origins of hexaploids. CONCLUSIONS: Genome skimming provided important insights into the complex evolutionary relationships within sect. Macrocarpon, but additional sampling and data from single-copy nuclear regions are necessary to further elucidate the origins of the polyploids in this section.

Consequences of multiple mating-system shifts for population and range-wide genetic structure in a coastal dune plant.

Evolutionary transitions from outcrossing to selfing can strongly affect the genetic diversity and structure of species at multiple spatial scales. We investigated the genetic consequences of mating-system shifts in the North American, Pacific coast dune endemic plant Camissoniopsis cheiranthifolia (Onagraceae) by assaying variation at 13 nuclear (n) and six chloroplast (cp) microsatellite (SSR) loci for 38 populations across the species range. As predicted from the expected reduction in effective population size (Ne ) caused by selfing, small-flowered, predominantly selfing (SF) populations had much lower nSSR diversity (but not cpSSR) than large-flowered, predominantly outcrossing (LF) populations. The reduction in nSSR diversity was greater than expected from the effects of selfing on Ne alone, but could not be accounted for by indirect effects of selfing on population density. Although selfing should reduce gene flow, SF populations were not more genetically differentiated than LF populations. We detected five clusters of nSSR genotypes and three groups of cpSSR haplotypes across the species range consisting of parapatric groups of populations that usually (but not always) differed in mating system, suggesting that selfing may often initiate ecogeographic isolation. However, lineage-wide genetic variation was not lower for selfing clusters, failing to support the hypothesis that selection for reproductive assurance spurred the evolution of selfing in this species. Within three populations where LF and SF plants coexist, we detected genetic differentiation among diverged floral phenotypes suggesting that reproductive isolation (probably postzygotic) may help maintain the striking mating-system differentiation observed across the range of this species.

Pollinator identity and spatial isolation influence multiple paternity in an annual plant.

The occurrence and extent of multiple paternity is an important component of variation in plant mating dynamics. However, links between pollinator activity and multiple paternity are generally lacking, especially for plant species that attract functionally diverse floral visitors. In this study, we separated the influence of two functionally distinct floral visitors (hawkmoths and solitary bees) and characterized their impacts on multiple paternity in a self-incompatible, annual forb, Oenothera harringtonii (Onagraceae). We also situated pollinator-mediated effects in a spatial context by linking variation in multiple paternity to variation in plant spatial isolation. We documented pronounced differences in the number of paternal sires as function of pollinator identity: on average, the primary pollinator (hawkmoths) facilitated mating with nearly twice as many pollen donors relative to the secondary pollinator (solitary bees). This effect was consistent for both isolated and nonisolated individuals, but spatial isolation imposed pronounced reductions on multiple paternity regardless of pollinator identity. Considering that pollinator abundance and pollen dispersal distance did not vary significantly with pollinator identity, we attribute variation in realized mating dynamics primarily to differences in pollinator morphology and behaviour as opposed to pollinator abundance or mating incompatibility arising from underlying spatial genetic structure. Our findings demonstrate that functionally distinct pollinators can have strongly divergent effects on polyandry in plants and further suggest that both pollinator identity and spatial heterogeneity have important roles in plant mating dynamics.

Strong genetic differentiation but not local adaptation toward the range limit of a coastal dune plant.

All species have limited geographic distributions; but the ecological and evolutionary mechanisms causing range limits are largely unknown. That many species' geographic range limits are coincident with niche limits suggests limited evolutionary potential of marginal populations to adapt to conditions experienced beyond the range. We provide a test of range limit theory by combining population genetic analysis of microsatellite polymorphisms with a transplant experiment within, at the edge of, and 60 km beyond the northern range of a coastal dune plant. Contrary to expectations, lifetime fitness increased toward the range limit with highest fitness achieved by most populations at and beyond the range edge. Genetic differentiation among populations was strong, with very low, nondirectional gene flow suggesting range limitation via constraints to dispersal. In contrast, however, local adaptation was negligible, and a distance-dependent decline in fitness only occurred for those populations furthest from home when planted beyond the range limit. These results challenge a commonly held assumption that stable range limits match niche limits, but also raise questions about the unique value of peripheral populations in expanding species' geographical ranges.

Direct vs. indirect effects of whole-genome duplication on prezygotic isolation in Chamerion angustifolium: Implications for rapid speciation.

PREMISE OF THE STUDY: The depiction of polyploid speciation as instantaneous implies that strong prezygotic and postzygotic isolation form as a direct result of whole-genome duplication. However, the direct vs. indirect contributions of genome duplication to phenotypic divergence and prezygotic isolation are rarely quantified across multiple reproductive barriers. METHODS: We compared the phenotypic differences between diploid and both naturally occurring and synthesized tetraploids (neotetraploids) of the plant Chamerion angustifolium. Using this information and additional published values for this species, we compared the magnitude of isolation (ecological, flowering, pollinator, and gametic) between diploids and natural-occurring tetraploids to that between diploids and neotetraploids. KEY RESULTS: Differences among ploidy cytotypes were observed for eight of 12 vegetative and reproductive traits measured. Neotetraploids resembled diploids but differed from natural tetraploids with respect to four traits, including flowering time and plant height. Diploid-neotetraploid (2x-4xneo) experimental arrays exhibited lower pollinator fidelity to cytotype and seed set compared with 2x-4xnat arrays. Based on these results and published evidence, reproductive isolation between diploids and neotetraploids across all four life stages averaged 0.48 and deviated significantly from that between diploids and natural tetraploids (RI = 0.96). CONCLUSIONS: Genome duplication causes phenotypic shifts and contributes directly to prezygotic isolation for some barriers (gametic isolation) but cannot account for the cumulative isolation from diploids observed in natural tetraploids. Therefore, conditions for species formation through genome duplication are not necessarily instantaneous and selection to strengthen prezygotic barriers in young polyploids is critical for the establishment of polyploid species in sympatry.

Effect of inbreeding on pollen tube growth in diploid and tetraploid Chamerion angustifolium: Do polyploids mask mutational load in pollen?

PREMISE OF THE STUDY: Deleterious recessive mutations are an important determinant of fitness (mutational load) in the sporophytic phase of plants and a major cause of inbreeding depression; however, their role in gametophyte function is less well documented but may account for variation in pollen tube growth and siring ability, especially between diploid and polyploid plants, which can mask the load. METHODS: We investigated the role of mutational load in pollen performance using the perennial polyploid Chamerion angustifolium by comparing tube growth of pollen, in styles and in growth medium, from inbred (selfed) and outbred diploids to that of inbred and outbred tetraploids. Pollen from tetraploids is expected to mask deleterious mutations more effectively in the outbred condition but reveal them after inbreeding. In contrast, gametophytes from diploids should express the same genetic load in inbred or outbred plants. KEY RESULTS: Pollen tube growth measured in growth medium was highest in outbred tetraploids and generally lower in inbred than outbred plants. The effect of selfing was significant in pollen from tetraploids but not diploids. The differential effect of selfing was also evident in the proportion of pollen reaching the base of styles, but the ploidy × pollination interaction was not significant. Selfing also had a negative effect on sporophyte fitness but was greater in diploids than tetraploids. CONCLUSIONS: Pollen performance is influenced by the expression of mutational load, which is masked in polyploids. This effect may partly explain strong siring success of tetraploids in this species.

Divergence times, historical biogeography, and shifts in speciation rates of Myrtales.

We examine the eudicot order Myrtales, a clade with strong Gondwanan representation for most of its families. Although previous phylogenetic studies greatly improved our understanding of intergeneric and interspecific relationships within the order, our understanding of inter-familial relationships still remains unresolved; hence, we also lack a robust time-calibrated chronogram to address hypotheses (e.g., biogeography and diversification rates) that have implicit time assumptions. Six loci (rbcL, ndhF, matK, matR, 18S, and 26S) were amplified and sequenced for 102 taxa across Myrtales for phylogenetic reconstruction and ten fossil priors were utilized to produce a chronogram in BEAST. Combretaceae is identified as the sister clade to all remaining families with moderate support, and within the latter clade, two strongly supported groups are seen: (1) Onagraceae+Lythraceae, and (2) Melastomataceae+the Crypteroniaceae, Alzateaceae, Penaeaceae clade along with Myrtaceae+Vochysiaceae. Divergence time estimates suggest Myrtales diverged from Geraniales ∼124Mya during the Aptian of the Early Cretaceous. The crown date for Myrtales is estimated at ∼116Mya (Albian-Aptian). BioGeoBEARS showed significant improvement in the likelihood score when the "jump dispersal" parameter was added. South America and/or Africa are implicated as important ancestral areas in all deeper nodes. BAMM analyses indicate that the best configuration included three significant shifts in diversification rates within Myrtales: near the crown of Melastomataceae (∼67-64Mya), along the stem of subfamily Myrtoideae (Myrtaceae; ∼75Mya), and along the stem of tribe Combreteae (Combretaceae; ∼50-45Mya). Issues with conducting diversification analyses more generally are examined in the context of scale, taxon sampling, and larger sets of phylogenetic trees.

Climatic niche differences between diploid and tetraploid cytotypes of Chamerion angustifolium (Onagraceae).

PREMISE OF THE STUDY: Polyploidy-the possession of more than two copies of each chromosome in the nucleus-is common in flowering plants. Polyploid plants can occupy different geographic ranges than their diploid progenitors, but the factors responsible for maintaining these range differences are poorly understood. Polyploidy can have significant physiological consequences, and the present study aims to determine whether previously described physiological differences between cytotypes are correlated with climatic niches and geographic distributions. METHODS: Prior research indicates that tetraploid plants of Chamerion angustifolium (Onagraceae) are more tolerant of drought and less tolerant of freezing than diploids, which suggests that they should occupy a niche that is warmer and drier than that of diploids. We extracted climate data for 134 populations of C. angustifolium classified as pure diploid, pure tetraploid, or mixed-ploidy. We compared climatic conditions between these population categories and generated ecological niche models to compare their geographic distribution with prior qualitative estimates. KEY RESULTS: Pure tetraploid populations occupy habitats that are warmer and drier than those of pure diploid populations. Mixed-ploidy populations occur in habitats that are not strictly intermediate between pure diploid and pure tetraploid populations, but are as cold as pure diploid populations and have intermediate soil moisture deficits. Our niche models were similar to previous qualitative estimates of cytotype geographic distribution. CONCLUSIONS: The correspondence between the physiological tolerances of cytotypes, their climatic niches, and their geographic distributions suggests that physiological traits are at least partially responsible for differences in the realized climatic niches of diploid and tetraploid C. angustifolium.

Adaptation of diploid and tetraploid chamerion angustifolium to elevation but not local environment.

Polyploid organisms often have different geographic ranges than their diploid relatives. However, it is unclear whether this divergence is maintained by adaptation or results from historical differences in colonization. Here, we conducted a reciprocal transplant experiment with diploid and autotetraploid Chamerion angustifolium to test for adaptation at the ploidy and population level. In the Rocky Mountains, pure diploid populations occur at high elevations and pure autotetraploid populations occur at low elevations with mixed ploidy populations between. We planted 3134 seedlings in 2004 and 3890 juveniles (bolting) in 2005 among nine plots, three in each of the diploid, mixed ploidy, and tetraploid zones, and monitored survival until 2008. For both seedlings and juvenile plants, elevation significantly influenced survival. The juvenile plants also showed a significant ploidy by elevation interaction, indicating that diploids and tetraploids survived best at their native elevations. In contrast, we found no evidence of local adaptation to plot within elevation. This suggests that the current distribution of diploids and tetraploids across elevations is the result of adaptation and that genome duplication may have facilitated the invasion of lower elevation habitats by limiting the movement of maladapted alleles from diploid populations at higher elevations.

Cytotype coexistence leads to triploid hybrid production in a diploid-tetraploid contact zone of Chamerion angustifolium (Onagraceae).

PREMISE OF THE STUDY: Polyploids are often geographically segregated from their diploid progenitors, but the extent of sympatry and the consequences for reproductive isolation and coexistence are rarely quantified. • METHODS: In this study, we document the distribution and co-occurrence of diploid and tetraploid Chamerion angustifolium among 57 populations within the diploid-tetraploid contact zone in the Canadian Rocky Mountains. Rates of hybrid mating in mixed-ploidy populations were inferred from the frequency of triploid offspring in open-pollinated seed families. • KEY RESULTS: Twenty-three of 57 populations sampled contained a single cytotype; 20 (87%) were tetraploid and three (13%) were diploid. Thirty-four populations (60%) contained multiple ploidies. Diploid and tetraploid plants occurred in all mixed-ploidy populations; triploids occurred in 13 populations and averaged 1.4% of plants per population. The proportion of tetraploids in a population was negatively related to elevation (partial regression: F = 27.2, P <0.0001) and latitude (partial regression: F = 17.4, P < 0.0001). Triploids were detected in seed from all eight mixed-ploidy populations sampled ( = 3.7% of seed per population), comprising 7% of that expected with random mating (G = 2589.2, df = 1, P <0.0001, n = 2628), and were more often produced by diploid maternal parents than tetraploid parents. • CONCLUSIONS: Our results indicate that tetraploids regularly coexist with diploids in the contact zone and that this coexistence is likely promoted by both strong reproductive isolation and asymmetric intercytotype mating between diploid and tetraploid C. angustifolium.

The association between polyploidy and clonal reproduction in diploid and tetraploid Chamerion angustifolium.

Clonal reproduction is associated with the incidence of polyploidy in flowering plants. This pattern may arise through selection for increased clonality in polyploids compared to diploids to reduce mixed-ploidy mating. Here, we test whether clonal reproduction is greater in tetraploid than diploid populations of the mixed-ploidy plant, Chamerion angustifolium, through an analysis of the size and spatial distribution of clones in natural populations using AFLP genotyping and a comparison of root bud production in a greenhouse study. Natural tetraploid populations (N = 5) had significantly more AFLP genotypes (x¯ = 10.8) than diploid populations (x¯ = 6.0). Tetraploid populations tended to have fewer ramets per genotype and fewer genotypes with >1 ramet. In a spatial autocorrelation analysis, ramets within genotypes were more spatially aggregated in diploid populations than in tetraploid populations. In the greenhouse, tetraploids allocated 90.4% more dry mass to root buds than diploids, but tetraploids produced no more root buds and 44% fewer root buds per unit root mass than diploids. Our results indicate that clonal reproduction is significant in most populations, but tetraploid populations are not more clonal than diploids, nor are their clones more spatially aggregated. As a result, tetraploids may be less sheltered from mixed-ploidy mating and diploids more exposed to inbreeding, the balance of which could influence the establishment of tetraploids in diploid populations.

Experimental and genetic analyses reveal that inbreeding depression declines with increased self-fertilization among populations of a coastal dune plant.

Theory predicts that inbreeding depression (ID) should decline via purging in self-fertilizing populations. Yet, intraspecific comparisons between selfing and outcrossing populations are few and provide only mixed support for this key evolutionary process. We estimated ID for large-flowered (LF), predominantly outcrossing vs. small-flowered (SF), predominantly selfing populations of the dune endemic Camissoniopsis cheiranthifolia by comparing selfed and crossed progeny in glasshouse environments differing in soil moisture, and by comparing allozyme-based estimates of the proportion of seeds selfed and inbreeding coefficient of mature plants. Based on lifetime measures of dry mass and flower production, ID was stronger in nine LF populations [mean δ = 1-(fitness of selfed seed/fitness of outcrossed seed) = 0.39] than 16 SF populations (mean δ = 0.03). However, predispersal ID during seed maturation was not stronger for LF populations, and ID was not more pronounced under simulated drought, a pervasive stress in sand dune habitat. Genetic estimates of δ were also higher for four LF (δ = 1.23) than five SF (δ = 0.66) populations; however, broad confidence intervals around these estimates overlapped. These results are consistent with purging, but selective interference among loci may be required to maintain strong ID in partially selfing LF populations, and trade-offs between selfed and outcrossed fitness are likely required to maintain outcrossing in SF populations.

Whole genome duplication affects evolvability of flowering time in an autotetraploid plant.

Whole genome duplications have occurred recurrently throughout the evolutionary history of eukaryotes. The resulting genetic and phenotypic changes can influence physiological and ecological responses to the environment; however, the impact of genome copy number on evolvability has rarely been examined experimentally. Here, we evaluate the effect of genome duplication on the ability to respond to selection for early flowering time in lines drawn from naturally occurring diploid and autotetraploid populations of the plant Chamerion angustifolium (fireweed). We contrast this with the result of four generations of selection on synthesized neoautotetraploids, whose genic variability is similar to diploids but genome copy number is similar to autotetraploids. In addition, we examine correlated responses to selection in all three groups. Diploid and both extant tetraploid and neoautotetraploid lines responded to selection with significant reductions in time to flowering. Evolvability, measured as realized heritability, was significantly lower in extant tetraploids (^b(T) =  0.31) than diploids (^b(T) =  0.40). Neotetraploids exhibited the highest evolutionary response (^b(T)  =  0.55). The rapid shift in flowering time in neotetraploids was associated with an increase in phenotypic variability across generations, but not with change in genome size or phenotypic correlations among traits. Our results suggest that whole genome duplications, without hybridization, may initially alter evolutionary rate, and that the dynamic nature of neoautopolyploids may contribute to the prevalence of polyploidy throughout eukaryotes.

Broad geographic covariation between floral traits and the mating system in Camissoniopsis cheiranthifolia (Onagraceae): multiple stable mixed mating systems across the species' range?

BACKGROUND AND AIMS: Plants vary widely in the extent to which seeds are produced via self-fertilization vs. outcrossing, and evolutionary change in the mating system is thought to be accompanied by genetic differentiation in a syndrome of floral traits. We quantified the pattern of variation and covariation in floral traits and the proportion of seeds outcrossed (t) to better understand the evolutionary processes involved in mating system differentiation among and within populations of the short-lived Pacific coastal dune endemic Camissoniopsis cheiranthifolia across its geographic range in western North America. METHODS: We quantified corolla width and herkogamy, two traits expected to influence the mating system, for 48 populations sampled in the field and for a sub-sample of 29 populations grown from seed in a glasshouse. We also measured several other floral traits for 9-19 populations, estimated t for 16 populations using seven allozyme polymorphisms, and measured the strength of self-incompatibility for nine populations. KEY RESULTS: Floral morphology and self-incompatibility varied widely but non-randomly, such that populations could be assigned to three phenotypically and geographically divergent groups. Populations spanned the full range of outcrossing (t = 0·001-0·992), which covaried with corolla width, herkogamy and floral life span. Outcrossing also correlated with floral morphology within two populations that exhibited exceptional floral variation. CONCLUSIONS: Populations of C. cheiranthifolia seem to have differentiated into three modal mating systems: (1) predominant outcrossing associated with self-incompatibility and large flowers; (2) moderate selfing associated with large but self-compatible flowers; and (3) higher but not complete selfing associated with small, autogamous, self-compatible flowers. The transition to complete selfing has not occurred even though the species appears to possess the required genetic capacity. We hypothesize that outcrossing populations in this species have evolved to different stable states of mixed mating.

A unified model of autopolyploid establishment and evolution.

The prevalence of polyploidy among flowering plants is surprising given the hurdles impeding the establishment and persistence of novel polyploid lineages. In the absence of strong assortative mating, reproductive assurance, or large intrinsic fitness advantages, new polyploid lineages face almost certain extinction through minority cytotype exclusion. Consequently, much work has focused on a search for adaptive advantages associated with polyploidy such as increased competitive ability, enhanced ecological tolerances, and increased resistance to pathogens. Yet, no consistent adaptive advantages of polyploidy have been identified. Here, to investigate the potential for autopolyploid establishment and persistence in the absence of any intrinsic fitness advantages, we develop a simulation model of a diploid population that sporadically gives rise to novel autopolyploids. The autopolyploids have only very small levels of initial assortative mating or niche differentiation, generated entirely by dosage effects of genome duplication, and they have realistic levels of reproductive assurance. Our results show that by allowing assortative mating and competitive interactions to evolve, establishment of novel autopolyploid lineages becomes common. Additional scenarios where adaptive optima change over time reveal that rapid environmental change promotes the replacement of diploid lineages by their autopolyploid descendants. These results help to explain recent empirical findings that suggest that many contemporary polyploid lineages arose during the Cretaceous-Tertiary extinction, without invoking adaptive advantages of polyploidy.

Effect of recurrent selfing on inbreeding depression and mating system evolution in an autopolyploid plant.

Genome duplication resulting in polyploidy can have significant consequences for the evolution of mating systems. Most theory predicts that self-fertilization will be selectively favored in polyploids; however, many autopolyploids are outcrossing or mixed-mating. Here, we examine the hypothesis that the evolution of selfing is restricted in autopolyploids because the genetic cost of selfing (i.e., inbreeding depression) increases monotonically with successive generations of inbreeding. Using the herbaceous, autotetraploid plant Chamerion angustifolium, we generated populations with different inbreeding coefficients (F= 0, 0.17 and 0.36) through three consecutive generations of selfing and compared their magnitudes of inbreeding depression in a common environment. Mating system estimates for four natural populations confirmed that tetraploid selfing rates (s(m) = 0.25, SE = 0.02) are similar to those of diploids (s(m) = 0.12, SE = 0.12; F1,2 = 1.34, P= 0.37) indicating that both cytotypes are predominantly outcrossing. Compared to an outbred control line, mean inbreeding depression for seed production, survival, and height (vegetative and total) in the inbred line differed among generations (inbreeding coefficients). Across all stages, inbreeding depression (relative to control) was positively related to generation (inbreeding coefficient). Although the initial costs of inbreeding in extant and newly synthesized polyploids may be low compared to diploids, the monotonic increase in inbreeding depression with repeated inbreeding may limit the extent to which selfing variants are favored.