Browsing by Subject "Systematics"

Background

Results

Conclusions

Not all ferns grow in moist and shaded habitats. One notable example is the ecologically unusual clade of notholaenids. With approximately 40 species, the notholaenids have adapted to and diversified within the deserts of Mexico and the southwestern United States. In my dissertation, I studied the evolution and diversification of notholaenid ferns, using an approach that integrates data from multiple sources: biochemistry, biogeography, cytology, ecological niche modeling, molecular phylogeny, morphology, and physiology.

In Chapter 1, I infer a species phylogeny for notholaenid ferns using both nuclear and plastid DNA sequences, and reconstruct the evolutionary history of “farina” (powdery exudates of lipophilic flavonoid aglycones), a characteristic drought-adapted trait, that occurs on both the gametophytic and sporophytic phases of members of the the clade. Forty-nine notholaenid and twelve outgroup samples were selected for these analyses. Long (ca. 1 kb) low-copy nuclear sequences for four loci were retrieved using a recently developed amplicon sequencing protocol on the PacBio Sequel platform and a bioinformatics pipeline PURC; plastid sequences from three loci were retrieved using Sanger sequencing. Each nuclear/plastid dataset was first analyzed individually using maximum likelihood and Bayesian inference, and the species phylogeny was inferred using *BEAST. Ancestral states were reconstructed using likelihood (re-rooting method) and MCMC (stochastic mapping method) approaches. Ploidy levels were inferred using chromosome counts corroborated by spore diameter measurements. My phylogenetic analyses results are roughly congruent with previous phylogenies inferred using only plastid data; however, several incongruences were observed between them. Hybridization events among recognized species of the notholaenid clade appear to be relatively rare, compared to what is observed in other well-studied fern genera. All characters associated with farina production in the group appear to be homoplastic and have complex evolutionary histories.

In Chapter 2, I focus on the infraspecific diversification of Notholaena standleyi, a species that thrives in the deserts of the southwestern United States and Mexico and has several “chemotypes” that express differences in farina color and chemistry. Forty-eight samples were selected from across the geographic distribution of N. standleyi. Phylogenetic relationships were inferred using four plastid makers and five single/low-copy nuclear markers. Sequences were retrieved using PacBio and the PURC pipeline. Ploidy levels were inferred from relative spore size measurements calibrated with chromosome counts, and farina chemistry was compared using thin-layer chromatography (TLC). My studies of Notholaena standleyi reveal a complex history of infraspecific diversification traceable to a variety of evolutionary drivers including classic allopatry, parapatry with or without changes in geologic substrate, and sympatric divergence through polyploidization. Four divergent clades were recognized within the species. Three roughly correspond to previously recognized chemotypes: gold (G), yellow (Y), and pallid/yellow-green (P/YG). The fourth clade, cryptic (C), is newly reported here. The diploid clades G and Y are found in the Sonoran and Chihuahuan Deserts, respectively; they co-occur (and hybridize) in the Pinaleño Mts. of eastern Arizona. Clades G and Y are estimated to have diverged in the Pleistocene, congruent with the postulated timing of climatological events that divide these two deserts. Clade P/YG is tetraploid and partially overlaps the distribution of clade Y in the eastern parts of the Chihuahuan Desert. However, PY/G is apparently confined to limestone, a geologic substrate rarely occupied by members of the other clades. The newly discovered diploid clade, cryptic (C), is distributed in the southern Mexican states of Oaxaca and Puebla and is highly disjunct from the other three clades.

In Chapter 3, I study the ecological niche differentiation among the three major chemotypes––G, Y, and P/YG. Using both ordination and species distribution modelling techniques, the ecological niches for each chemotype were characterized and compared. The main environmental drivers for their distributions were identified, their suitable habitats in both geographic and environmental spaces were predicted, and their niche equivalencies and similarities were tested. My ecological niche analyses results suggest that all three chemotypes are ecologically diverged. The ecological niches of the two parapatric, sister diploid chemotypes, G and Y, are significantly different from one another. Chemotype G occupies a very extreme niche with higher solar radiation, and lower rainfall and higher temperatures in the wettest quarter. The niche space of tetraploid chemotype P/YG is similar but not equivalent to the other two chemotypes. Its distribution model is highly influenced by the high percentage of Calcids and warmer temperatures in the wet season, reflecting the fact that it is confined to limestone in areas of lower elevation/latitude.

In Chapter 4, I gather together all my other studies related to Notholaena standleyi, including: 1) morphological and anatomical observations of its desiccation-tolerant leaf, with special focus on the farina; 2) two cases of hybridization between the chemotypes, one between the diploid chemotype Y and tetraploid chemotype YG, and another between diploid chemotypes Y and G; and 3) morphological and physiological comparisons between the two diploid chemotypes Y and G. My plan is to finalize these studies and submit them for publication in the near future.

In Chapter 5, I summarize collaborative contributions that I made to other fern studies during my Ph.D.

How geographical distance and historical events affect patterns of population divergence and gene flow is an important question in evolution and biogeography. Sphagnum subgenus Acutifolia is one of the four major subgenera of Sphagnum peatmoss comprising numerous species with broad geographic ranges and diverse ecological niches within wetland habitats. One group of particular interest within the subgenus is the S. capillifolium complex which contains at least seven closely related species. Five species within the complex are circumboreal and all have overlapping geographic ranges, one species is endemic to subtropical region of eastern North America, and one species is in the tropical regions of Central and South America. The presence of both species with overlapping and disjunct distributions makes the Sphagnum capillifolium complex a natural experiment to investigate gene flow and population divergence in multiple phylogenetic scales (within and between species). Chapter 1 describes patterns of phylogenetic discordance across the genome in the Sphagnum capillifolium complex using whole genome resequencing data. The species tree phylogeny was generally well supported but phylogenetic discordance among genomic regions was prevalent, especially at nodes in the backbone. Alternative topologies for each of the backbone nodes were not random, suggesting the presence of introgression, in addition to incomplete lineage sorting (ILS). Analyses of introgression signals using ABBA/BABA tests and branch length distributions (QuIBL) showed that there were several possible introgression events within the S. capillifolium complex involving both extant and ancestral species. Most of the introgression events occurred between species that currently have overlapping geographic ranges. Further investigation of one introgression event using comparisons of terminal branch lengths showed that the biased pattern of shared derived alleles likely derives from introgression, not ancient polymorphism. These findings show that introgression played a significant role in generating phylogenetic incongruence within the S. capillifolium complex. We also show that the use of multiple phylogenomic methods and investigating localized genomic regions are essential to infer complex introgression scenarios. Chapter 2 describes phylogenetic structure of Sphagnum subgenus Acutifolia and population structure of circumboreal species within the S. capillifolium complex. Genome scale data (RAD-seq) was generated for the subgenus, with an intensive population sampling of circumboreal species within the S. capillifolium complex. Most of the species are resolved as monophyletic, although relationships among species are weakly supported in some parts of the phylogeny. Some currently recognized species are phylogenetically discernable while others are not distinguishable from the well-supported species. Within the S. capillifolium complex, five circumboreal species show similar patterns of population structure. One population system comprises plants in eastern North America and Europe, and sometimes includes plants from eastern Eurasia and the Pacific Northwest of North America. Another group comprises plants in the Pacific Northwest, or around the Beringian and Arctic regions. Our results suggest that populations of circumboreal species survived in multiple refugia during the last glacial maximum (LGM). Long-distance dispersal out of refugia, population bottlenecks, and possible adaptations to conditions unique to each refugium contribute to current geographic patterns. There are patterns of genetic admixture among distinct genotype groups within species in some restricted areas. Alaska is a hotspot for both intraspecific genetic diversity and admixture. These genetic results indicate the important role of historical events, especially Pleistocene glaciation, in shaping the complex population structure of plants with broad distribution ranges. Chapter 3 assesses the pattern of gene flow between a pair of sister Sphagnum species within the S. capillifolium complex: S. warnstorfii and S. talbotianum. The two species have different distribution ranges: S. warnstorfii is circumboreal while S. talbotianum is circumarctic, but they overlap in Alaska. Genetic data from chapter 2 were used in this chapter. Analyses of interspecific gene flow and population sizes were accomplished using coalescent simulations of site frequency spectra (SFSs), and the signature of gene flow was further corroborated by ABBA/BABA statistics. Our results indicate that S. warnstorfii and S. talbotianum were isolated after divergence. S. warnstorfii was relatively recently established in Alaska and Alaska is the only region that shows evidence of gene flow between S. talbotianum and S. warnstorfii. Gene flow occurred in only one direction from S. talbotianum into S. warnstorfii, which can possibly help S. warnstorfii survive in subarctic conditions. Molecular evidence further suggests gene flow from Alaska S. warnstorfii to other regional populations of that species. S. warnstorfii suffered a stronger population bottleneck than S. talbotianum, suggesting that Beringia could have harbored larger populations during the last glacial maximum than other, likely more southern, refugia. Although the two species are very closely related, S. talbotianum has larger pores on the convex surfaces of branch leaf apices than S. warnstorfii. Our results represent a case study of a recent gene flow between geographically sympatric peatmoss species using genomic data. Our results also support S. talbotianum as a distinct species from S. warnstorfii.