Disentangling Crocus Series Verni and Its Polyploids

SIMPLE SUMMARY: In plants, the occurrence of polyploid lineages, which are plants with multiple instead of two sets of chromosomes, is quite common. Polyploids can originate as autopolyploids within a species or by combining the genomes of different species resulting in allopolyploids. Within the group of spring crocuses, a polyploid complex exists where it is unclear how it evolved and which species eventually contributed to polyploid formation. Among Crocus species, evolutionary analyses are further complicated by widely varying chromosome numbers that do not clearly correlate with di- or polyploidy. To reconstruct the evolution of these polyploids, we combine chromosome counts, genome size estimations, phylogenetic analyses based on maternally and bi-parentally inherited genomes, co-ancestry analysis, and morphometric data for all species potentially involved in polyploid formation. Through this approach, we show that polyploids in the Crocus heuffelianus group are allopolyploids that originated multiple times involving different parental genotypes and reciprocal crosses. Chromosome numbers partly changed after polyploidization. Numbers found in polyploids are therefore no longer in all cases additive values of their parents’ chromosomes. We conclude that in crocuses, only an approach combining evidence from different analysis methods can uncover the evolutionary history of species if polyploidization is involved. ABSTRACT: Spring crocuses, the eleven species within Crocus series Verni (Iridaceae), consist of di- and tetraploid cytotypes. Among them is a group of polyploids from southeastern Europe with yet-unclear taxonomic affiliation. Crocuses are generally characterized by complex dysploid chromosome number changes, preventing a clear correlation between these numbers and ploidy levels. To reconstruct the evolutionary history of series Verni and particularly its polyploid lineages associated with C. heuffelianus, we used an approach combining phylogenetic analyses of two chloroplast regions, 14 nuclear single-copy genes plus rDNA spacers, genome-wide genotyping-by-sequencing (GBS) data, and morphometry with ploidy estimations through genome size measurements, analysis of genomic heterozygosity frequencies and co-ancestry, and chromosome number counts. Chromosome numbers varied widely in diploids with 2n = 8, 10, 12, 14, 16, and 28 and tetraploid species or cytotypes with 2n = 16, 18, 20, and 22 chromosomes. Crocus longiflorus, the diploid with the highest chromosome number, possesses the smallest genome (2C = 3.21 pg), while the largest diploid genomes are in a range of 2C = 7–8 pg. Tetraploid genomes have 2C values between 10.88 pg and 12.84 pg. Heterozygosity distribution correlates strongly with genome size classes and allows discernment of di- and tetraploid cytotypes. Our phylogenetic analyses showed that polyploids in the C. heuffelianus group are allotetraploids derived from multiple and partly reciprocal crosses involving different genotypes of diploid C. heuffelianus (2n = 10) and C. vernus (2n = 8). Dysploid karyotype changes after polyploidization resulted in the tetraploid cytotypes with 20 and 22 chromosomes. The multi-data approach we used here for series Verni, combining evidence from nuclear and chloroplast phylogenies, genome sizes, chromosome numbers, and genomic heterozygosity for ploidy estimations, provides a way to disentangle the evolution of plant taxa with complex karyotype changes that can be used for the analysis of other groups within Crocus and beyond. Comparing these results with morphometric analysis results in characters that can discern the different taxa currently subsumed under C. heuffelianus.