Joel S. Shore

Phylogenetic relationships, chromosome and breeding system evolution in Turnera (Turneraceae): inferences from its sequence data

Turnera provides a useful system for exploring two significant evolutionary phenomena-shifts in breeding system (distyly vs. homostyly) and the evolution of polyploids. To explore these, the first molecular phylogeny of Turnera was constructed using sequences of the internal transcribed spacer region (ITS) of nuclear ribosomal DNA for 37 taxa. We attempted to resolve the origins of allopolyploid species using single-strand conformation polymorphism and sequencing of homeologous copies of ITS. Two allohexaploid species possessed putative ITS homeologues (T. velutina and T. orientalis). A phylogenetic analysis to identify progenitors contributing to the origins of these polyploids was unsuccessful, possibly as a result of concerted evolution of ITS. Breeding system evolution was mapped onto the phylogeny assuming distyly to be ancestral in Turnera. Self-compatible homostyly appears to have arisen independently at least three times in Turnera; however, we were not able to determine whether there have been independent origins of homostyly among hexaploid species in series Turnera. Our phylogenetic analyses suggest that series Turnera is monophyletic. Neither series Microphyllae nor Anomalae, however, appear to be monophyletic. Future taxonomic revisions may require new circumscriptions of these latter series.

Tetrasomic inheritance and isozyme variation in Turnera ulmifolia vars. elegans Urb. and intermedia Urb. (Turneraceae)

Tetrasomic inheritance of three isozyme loci is demonstrated in Turnera ulmifolia vars. elegans and intermedia. The data support the occurrence of an autopolyploid origin for tetraploids of both taxonomic varieties. The extent of isozyme variation at 14 loci was determined for diploid and tetraploid populations. Tetraploid Turnera ulmifolia var. intermedia showed the lowest levels of isozyme variation, perhaps a result of founder effect, upon island colonization. In contrast, Turnera ulmifolia var. elegans showed the greatest levels of isozyme variation, suggesting that hybridization among locally differentiated diploid or tetraploid populations might have occurred. The number of independent origins of autotetraploidy in Turnera ulmifolia is unknown.

Characterization and localization of short-specific polygalacturonase in distylous Turnera subulata (Turneraceae)

We describe for distylous Turnera subulata a polygalacturonase specific to short-styled plants that is localized to the style transmitting tissue (the tissue through which pollen tubes grow). The polygalacturonase gene is linked to and may be upregulated by the S allele of the distyly locus. Because of its tissue-specific location, the polygalacturonase may be involved in the self-incompatibility response, acting in a complementary or antagonistic manner, or possibly in signalling downstream events. A pollen-specific polygalacturonase was also identified and may be a member of a small multigene family of pollen polygalacturonases. The role, if any, played by the pollen polygalacturonase in distyly, is presently unknown.

Habitat selection by Drosophila melanogaster larvae

Drosophila melanogaster larvae are used to examine habitat choice behavior and its effect on a component of preadult fitness (pupal survivorship). We established strains of flies by collecting pupae from two microhabitats from an orchard. Strain differences in pupation site choice (on versus off fruit) persisted in a field‐like laboratory assay without artificial selection. To produce heterogeneous environments, air temperature and soil water content were varied in these assays. A habitat suitability difference measure was used to determine for each environment, which microhabitat (on or off fruit) resulted in greater pupal survivorship. We found 1) that habitat choice behavior had both plastic and heritable components, 2) that strain‐by‐environment interactions influenced habitat choice behavior and pupal survivorship and, 3) a significant positive correlation between habitat suitability and larval habitat choice behavior.

Cyanogenesis, herbivory and plant defense in Turnera ulmifolia on Jamaica

Field surveys of eight populations of Turnera ulmifolia L., a Jamaican weed exhibiting quantitative genetic variation for cyanogenesis, were undertaken to assess the effectiveness of cyanogenesis a...

Pollination genetics of the common milkweed, Asclepias syriaca L.

Electrophoretic methods are used to determine the genotype of individual pollinia simultaneously at several isozyme loci for the common milkweed, Asclepias syriaca. Isozymes of pollinia are the result of gene expression in pollen. Genotype frequencies of maternal plants, pollinia carried by honey bees and pollinia captured by flowers were determined and found to be comparable. Over 50% of bees carried genotypically mixed pollen loads. Methods for estimating the rate of self-pollination, assuming the mixed-pollination model, are developed and the rate of self-pollination in a population is found to be 66%. The use of the methods developed here for testing various hypotheses, including the operation of sexual selection in plant populations, is discussed.

Morphological differentiation and crossability among populations of the Turnera ulmifolia L. complex (Turneraceae)

The Turnera ulmifolia complex is a polymorphic assemblage of perennial weeds native to the Neotropics. Thirteen population samples from a wide geographical range, representing six taxonomic varieties, were selected for a comparison of morphological differentiation and crossa- bility. The complex is polyploid with diploid and tetraploid populations exhibiting distyly and hexaploid populations homostyly. Thirty-eight morphological characters were measured on plants grown under uniform glasshouse conditions and seed set data collected from 2195 controlled crosses made in all combinations among the thirteen populations. Discriminant analysis and clus- tering techniques were employed to investigate phenetic relationships among populations and quantitative methods were used to compare the structure of the data from the morphological and crossability studies. Analysis of the phenetic and crossability data sets, using the Kc statistic, in- dicated that these data sets were significantly associated although discrepancies were apparent in the dendrograms produced. A dendrogram generated from the crossability data was composed solely of clusters of populations of the same ploidal levels, while the dendrogram based on phenetic data included one cluster containing populations at all ploidal levels. It is suggested that autopo- lyploidy or segmental allopolyploidy is responsible for the discordance between data sets. Mor- phological, cytological, and crossability data indicate that the complex might be better treated taxonomically as a number of separate species. The use of controlled crosses among differ- entiated populations of plants is a basic exper- imental technique in systematic studies. Cross- ability data are often used to infer genetic or patristic relationships among groups and are useful in demonstrating hybridization poten- tial or the existence of isolating mechanisms. In association with morphological, cytological, and chemical characters, data from artificial hy- bridization studies provide the most common- ly used sources of information in the construc- tion of classifications of flowering plants. Where both morphological and crossability data are available a comparison of the relationships im- plied by the alternative data sets may be help- ful in clarifying taxonomic decision-making as well as providing insights into the processes of evolutionary differentiation among taxa. The Turnera ulmifolia L. complex is variable in morphology, chromosome number, and breeding system. The plants are native throughout the Neotropics and occur as weeds

High Level of α‐Dioxygenase in Short Styles of Distylous Turnera Species

To identify proteins responsible for self‐incompatibility and/or other dimorphisms in distylous species of Turnera, we used SDS‐polyacrylamide gel electrophoresis to compare protein profiles of long and short styles. We detected a prominent 68‐kD protein in styles of short‐ but not long‐styled plants. This result was consistent in all five of the distylous species examined, except in a more divergent species, Piriqueta caroliniana, in which short styles did not express the protein but long styles may exhibit a low level of expression. Sequencing of tryptic peptides, genomic DNA, and phylogenetic analyses revealed the 68‐kD protein to be an α‐dioxygenase. Immunoblotting using antiserum against a pathogen‐induced α‐dioxygenase from tobacco further confirmed that the 68‐kD protein is an α‐dioxygenase. Using immunoblotting, we could not detect the α‐dioxygenase in other floral or vegetative organs, in styles of homostylous species, or in three mutants. The α‐dioxygenase reaches detectable levels in short styles 1 d before anthesis and high levels at flowering. Immunocytochemical analysis localized the α‐dioxygenase to the transmitting tissue of short styles in all three species examined. An α‐dioxygenase assay using crude style extracts revealed twice the activity for short compared with long styles. The tissue‐specific and temporal patterns of expression, as well as its consistent appearance in short‐styled plants of all Turnera species examined, indicates that the α‐dioxygenase plays some as yet unspecified role in distyly.

Characterization of X-ray-generated floral mutants carrying deletions at the S-locus of distylous Turnera subulata

To investigate the genetic architecture of distyly in Turnera subulata and test the hypothesis that a supergene determines distyly, we used X-ray mutagenesis to generate floral mutants. Based upon the crossing design, all progeny were expected to be short-styled. Of 3982 progeny screened, 10 long-styled mutants, one long homostyle and one short homostyle were recovered. Assays for molecular markers tightly linked to the S-locus showed that the mutants were missing 1–3 markers indicating they are deletion mutants. We investigated the incompatibility phenotype of the mutants and found that both their styles and pollen behaved like those of the long-styled morph. There was a variation in the absolute length of styles, stamens and pollen size of the long-styled mutants. Furthermore, long-styled mutants possessing larger deletions tended to have their anthers and stigmas in closer proximity. We explored the inheritance of the S-locus mutations and found that only one of the deletion mutations was transmitted to progeny where we recovered seven such progeny. Remarkably, our data are consistent with the supergene model (GPA/gpa) of Primula. The long homostyle mutant appears to have deletions involving both the G and P loci. The other mutants appear to have deletions of the entire S-locus. The mutants generated will serve as a valuable resource for the molecular dissection of the S-locus region, and in the identification of genes determining distyly.

Meiotic recombination in Turnera (Turneraceae): extreme sexual difference in rates, but no evidence for recombination suppression associated with the distyly ( S ) locus

To explore the rate of recombination resulting from male vs female meiosis, crosses were performed using distylous Turnera subulata as well as a cross involving the introgression of genes from T. krapovickasii into T. subulata. We assayed four loci on the chromosome bearing the S-locus as well as two loci on each of two other linkage groups. Substantial and consistent dimorphism in recombination rates was found with female meiosis resulting in as much as a ∼6-fold increase relative to male. Aberrant single locus segregation ratios occurred for some loci, particularly when the male (pollen) parent was heterozygous and the cross involved introgressed genes. The extreme trend of greater recombination resulting from female meiosis was, however, maintained in crosses where no aberrant ratios occurred, indicating that the sex dimorphism in recombination is not the result of aberrant segregation. We also exploited this distylous species and tested whether there is recombination suppression around the S-locus because of an inversion or other chromosome rearrangement(s). We found no significant evidence for recombination suppression.