Cytotype coexistence in the field cannot be explained by inter-cytotype hybridization alone: linking experiments and computer simulations in the sexual species Pilosella echioides (Asteraceae).

BACKGROUND: Processes driving ploidal diversity at the population level are virtually unknown. Their identification should use a combination of large-scale screening of ploidy levels in the field, pairwise crossing experiments and mathematical modelling linking these two types of data. We applied this approach to determine the drivers of frequencies of coexisting cytotypes in mixed-ploidy field populations of the fully sexual plant species Pilosella echioides. We examined fecundity and ploidal diversity in seeds from all possible pairwise crosses among 2x, 3x and 4x plants. Using these data, we simulated the dynamics of theoretical panmictic populations of individuals whose progeny structure is identical to that determined by the hybridization experiment. RESULTS: The seed set differed significantly between the crossing treatments, being highest in crosses between diploids and tetraploids and lowest in triploid-triploid crosses. The number of progeny classes (with respect to embryo and endosperm ploidy) ranged from three in the 2x-2x cross to eleven in the 3x-3x cross. Our simulations demonstrate that, provided there is no difference in clonal growth and/or survival between cytotypes, it is a clear case of minority cytotype exclusion depending on the initial conditions with two stable states, neither of which corresponds to the ploidal structure in the field: (i) with prevalent diploids and lower proportions of other ploidies, and (ii) with prevalent tetraploids and 9% of hexaploids. By contrast, if clonal growth differs between cytotypes, minority cytotype exclusion occurs only if the role of sexual reproduction is high; otherwise differences in clonal growth are sufficient to maintain triploid prevalence (as observed in the field) independently of initial conditions. CONCLUSIONS: The projections of our model suggest that the ploidal structure observed in the field can only be reached via a relatively high capacity for clonal growth (and proportionally lower sexual reproduction) in all cytotypes combined with higher clonal growth in the prevailing cytotype (3x).

The intriguing complexity of parthenogenesis inheritance in Pilosella rubra (Asteraceae, Lactuceae).

Neither the genetic basis nor the inheritance of apomixis is fully understood in plants. The present study is focused on the inheritance of parthenogenesis, one of the basic elements of apomixis, in Pilosella (Asteraceae). A complex pattern of inheritance was recorded in the segregating F(1) progeny recovered from reciprocal crosses between the facultatively apomictic hexaploid P. rubra and the sexual tetraploid P. officinarum. Although both female and male reduced gametes of P. rubra transmitted parthenogenesis at the same rate in the reciprocal crosses, the resulting segregating F(1) progeny inherited parthenogenesis at different rates. The actual transmission rates of parthenogenesis were significantly correlated with the mode of origin of the respective F(1) progeny class. The inheritance of parthenogenesis was significantly reduced in F(1) n + n hybrid progeny from the cross where parthenogenesis was transmitted by female gametes. In F(1) n + 0 polyhaploid progeny from the same cross, however, the transmission rate of parthenogenesis was high; all fertile polyhaploids were parthenogenetic. It appeared that reduced female gametes transmitting parthenogenesis preferentially developed parthenogenetically and only rarely were fertilized in P. rubra. The fact that the determinant for parthenogenesis acts gametophytically in Pilosella and the precocious embryogenesis in parthenogenesis-transmitting megagametophytes was suggested as the most probable explanations for this observation. Furthermore, we observed the different expression of complete apomixis in the non-segregating F(1) 2n + n hybrids as compared to their apomictic maternal parent P. rubra. We suggest that this difference is a result of unspecified interactions between the parental genomes.

Bridging global and microregional scales: ploidy distribution in Pilosella echioides (Asteraceae) in central Europe.

BACKGROUND AND AIMS: A detailed knowledge of cytotype distribution can provide important insights into the evolutionary history of polyploid systems. This study aims to explore the spatial distribution of different cytotypes in Pilosella echioides at various spatial scales (from the whole distributional range to the population level) and to outline possible evolutionary scenarios for the observed geographic pattern. METHODS: DNA-ploidy levels were estimated using DAPI flow cytometry in 4410 individuals of P. echioides from 46 populations spread over the entire distribution range in central Europe. Special attention was paid to the cytotype structure in the most ploidy-diverse population in south-west Moravia. KEY RESULTS: Five different cytotypes (2x, 3x, 4x, 5x and 6x) were found, the last being recorded for the first time. Although ploidy-uniform (di- or tetraploid) sites clearly prevailed, nearly one-quarter of the populations investigated harboured more (up to all five) cytotypes. Whereas penta- and hexaploids constituted only a minority of the samples, a striking predominance of the triploid cytotype was observed in several populations. CONCLUSIONS: The representative sampling confirmed previous data on cytotype distribution, i.e. the spatial aggregation of mixed-ploidy populations in south-west Moravia and Lower Austria and the predominance of ploidy-uniform populations in other parts of the area investigated. Recurrent origin of polyploids from diploid progenitors via unreduced gametes and their successful establishment are considered the key factors promoting intrapopulational ploidy mixture ('primary hybrid zones'). As an alternative to the generally accepted theory of cytotype co-existence based on the development of different means of inter-ploidy reproductive isolation, it is suggested that a long-term ploidy mixture can also be maintained in free-mating populations provided that the polyploids originate with a sufficient frequency. In addition, the prevalence (or subdominance) of the triploid cytotype in several mixed-ploidy populations represents the first evidence of such a phenomenon in plant systems with exclusively sexual reproduction.

Expressivity of apomixis in 2n + n hybrids from an apomictic and a sexual parent: insights into variation detected in Pilosella (Asteraceae: Lactuceae).

Reproductive variation was studied in the tetraploid Pilosella aurantiaca, hexaploid P. rubra (both species with facultative autonomous apospory) and in their 2n + n hybrids, which were obtained by crossing with a sexual pollen parent (tetraploid P. officinarum). The different DNA content in P. aurantiaca and P. officinarum demonstrated the actual 2n + n origin, both spontaneous from the field and through experimental crosses, of their hexaploid hybrids. The octoploid 2n + n progeny were recovered from an experimental cross of P. rubra and P. officinarum. The reproductive pathways operating in two maternal facultatively apomictic species and in the hybrids were quantified using a flow cytometric analysis of seeds obtained from either open-pollinated or emasculated plants. Whereas both maternal species displayed a high penetrance of apomixis, the level of apomixis among the majority of 2n + n hybrids was much lower and variable. Some of the hexaploid hybrids had a reduced seed set. Compared to the respective maternal parents, the decrease in apomixis due to haploid parthenogenesis and/or n + n mating was evident in almost all unreduced hybrids, irrespective of their field/experimental origin and ploidy. Hence, the reproductive behaviour in the apomictic maternal parent was profoundly different from that of the 2n + n hybrids with a sexual parent in spite of the preservation of the complete maternal genome in the hybrids. The regulatory interactions in hybrid genomes, such as effects of modifiers, heterochrony, and epigenetic control, may be consistent with the different expressivity of apomixis observed under different genetic backgrounds.

Genome size variation and species relationships in Hieracium sub-genus Pilosella (Asteraceae) as inferred by flow cytometry.

BACKGROUND AND AIMS: Hieracium sub-genus Pilosella (hawkweeds) is a taxonomically complicated group of vascular plants, the structure of which is substantially influenced by frequent interspecific hybridization and polyploidization. Two kinds of species, 'basic' and 'intermediate' (i.e. hybridogenous), are usually recognized. In this study, genome size variation was investigated in a representative set of Central European hawkweeds in order to assess the value of such a data set for species delineation and inference of evolutionary relationships. METHODS: Holoploid and monoploid genome sizes (C- and Cx-values) were determined using propidium iodide flow cytometry for 376 homogeneously cultivated individuals of Hieracium sub-genus Pilosella, including 24 species (271 individuals), five recent natural hybrids (seven individuals) and experimental F(1) hybrids from four parental combinations (98 individuals). Chromosome counts were available for more than half of the plant accessions. Base composition (proportion of AT/GC bases) was cytometrically estimated in 73 individuals. KEY RESULTS: Seven different ploidy levels (2x-8x) were detected, with intraspecific ploidy polymorphism (up to four different cytotypes) occurring in 11 wild species. Mean 2C-values varied approx. 4.3-fold from 3.53 pg in diploid H. hoppeanum to 15.30 pg in octoploid H. brachiatum. 1Cx-values ranged from 1.72 pg in H. pilosella to 2.16 pg in H. echioides (1.26-fold). The DNA content of (high) polyploids was usually proportional to the DNA values of their diploid/low polyploid counterparts, indicating lack of processes altering genome size (i.e. genome down-sizing). Most species showed constant nuclear DNA amounts, exceptions being three hybridogenous taxa, in which introgressive hybridization was suggested as a presumable trigger for genome size variation. Monoploid genome sizes of hybridogenous species were always between the corresponding values of their putative parents. In addition, there was a good congruency between actual DNA estimates and theoretical values inferred from putative parental combinations and between DNA values of experimental F(1) hybrids and corresponding established hybridogenous taxa. CONCLUSIONS: Significant differences in genome size between hawkweed species from hybridogenous lineages involving the small-genome H. pilosella document the usefulness of nuclear DNA content as a supportive marker for reliable delineation of several of the most problematic taxa in Hieracium sub-genus Pilosella (including classification of borderline morphotypes). In addition, genome size data were shown to have a good predictive value for inferring evolutionary relationships and genome constitution (i.e. putative parental combinations) in hybridogenous species.