Phylogenetic approaches for investigating the taxonomic and functional diversity of taxonomically complex taxa within the genus Dianthus
Loading...
Date
2026-07-09
Authors
Φάσσου, Γεωργία
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The Caryophyllaceae Juss. is one of the largest families of flowering plants, comprising approximately 100 genera and 3,000 species distributed predominantly throughout the Northern Hemisphere, with its greatest diversity in the Mediterranean and Irano-Turanian regions. Recent molecular phylogenetic studies have substantially improved our understanding of relationships within the family, leading to a revised tribal classification and providing a robust framework for investigating the evolution of its major lineages.
Within Caryophyllaceae, the tribe Caryophylleae Lam. & DC. is the second largest after Sileneae DC. and includes approximately 630 species distributed among 14 genera. The largest genus of the tribe, Dianthus L., comprises approximately 300 species and represents one of the most rapidly diversifying lineages of angiosperms. Although recent molecular phylogenetic studies have considerably improved our understanding of evolutionary relationships within Dianthus, they have also challenged traditional classifications and demonstrated that several infrageneric groups are not monophyletic. Furthermore, the generic delimitation of Dianthus remains under discussion following the proposed inclusion of Velezia and several species formerly assigned to Petrorhagia.
Among the most taxonomically challenging groups within the genus are the Greek endemic Dianthus fruticosus and the Cretan endemic D. juniperinus. Dianthus fruticosus is distributed across the Cyclades, the southeastern Aegean Islands, the Ionian Islands, southern Peloponnese, and Crete, and has undergone numerous taxonomic revisions since the early eighteenth century, until Runemark's (1980) treatment of the complex as a single species comprising eight allopatric subspecies. Dianthus juniperinus is an obligate chasmophyte confined to the calcareous cliffs of Crete. Following a similarly complex taxonomic history, Turland in 1992 recognized seven geographically structured subspecies. The restricted distributions, noticeable geographical isolation, and morphological variability of both complexes, together with the dynamic palaeogeographical history of the Aegean region, make them an excellent model for investigating patterns of diversification, taxonomic evolution, and conservation in Mediterranean plants.
The present PhD thesis aims to clarify the taxonomy, phylogenetic relationships, evolutionary history, ecological differentiation, and conservation status of the Dianthus fruticosus and D. juniperinus complexes using an integrative approach that combines molecular, morphological, ecological, and conservation analyses. Throughout this integrative investigation, the study also contributes to a broader understanding of species diversification, taxonomic evolution, and phylogenetic relationships within Dianthus.
To place the Dianthus fruticosus and D. juniperinus complexes within a broader evolutionary framework, extensive phylogenetic analyses were conducted using representatives from across the geographical and taxonomic diversity of Dianthus, together with species of Petrorhagia and Velezia.
Phylogenetic reconstructions based on four plastid regions (matK-trnK-psbA, rpl32-trnL, trnQ-rps16) and the nuclear ITS, inferred using both Bayesian and Maximum Likelihood approaches. The resulting phylogenetic framework confirms that the current circumscription of Dianthus should include Velezia and several species previously assigned to Petrorhagia. These findings largely support previous molecular studies while providing a substantially improved taxon sampling and a more robust assessment of generic boundaries. Within the core of Dianthus, three principal evolutionary lineages were consistently recovered. The first comprises predominantly Eurasian taxa and includes the type species, D. caryophyllus. The second contains mainly Irano-Turanian and Caucasian species together with the African representatives, whereas the third corresponds to a distinct eastern Mediterranean lineage that includes all representatives of the D. juniperinus complex together with part of the D. fruticosus complex, specifically the subsp. carpathus, subsp. occidentalis (S Peloponnese), subsp. karavius, subsp. sitiacus and subsp. creticus, (SE Crete). Representatives of D. fruticosus complex was found also in the first clade with the Eurasian taxa, and specifically the subsp. amorginus, subsp. fruticosus, subsp. occidentalis (NW Crete), subsp. creticus (NW Crete), and subsp. rhodius. Furthermore, Comparison of the recovered phylogeny with the traditional infrageneric classifications of Williams (1893) and Pax and Hoffmann (1934) revealed little overall congruence.
In contrast to the plastid phylogeny, the ITS dataset recovered the D. fruticosus–D. juniperinus complex as one of the few moderately supported lineages emerging from the large polytomy that characterizes core Dianthus. Most representatives of D. fruticosus were recovered within a single lineage, except for subsp. creticus from northwestern Crete and subsp. occidentalis from the Peloponnese. Dianthus juniperinus was divided into two main lineages: one comprising subsp. juniperinus, subsp. heldreichii, subsp. idaeus, subsp. pulviniformis, and representatives of subsp. aciphyllus from the central and northern Dikti mountain range, and a second including subsp. kavusicus and representatives of subsp. aciphyllus from eastern Dikti and the Asterousia mountain range.
The monophyly of the D. fruticosus and D. juniperinus complexes was already questioned by the analyses based on plastid and nuclear markers. Moreover, the incongruence between the plastid and nuclear datasets made the interpretation of their phylogenetic relationships even more vague. Therefore, to obtain a more highly resolved phylogenetic framework and further evaluate their taxonomic relationships, complete chloroplast genomes and nuclear ribosomal DNA sequences were reconstructed and analyzed.
The complete chloroplast genome analyses largely corroborated the patterns inferred from the plastid marker dataset and confirmed that D. fruticosus is not monophyletic, with the subsp. amorginus, subsp. fruticosus, and subsp. rhodius recovered within one clade, whereas subsp. occidentalis and subsp. creticus were distributed between clades two clades, a pattern showing a strong geographical component. Dianthus fruticosus subsp. rhodius was recovered as a distinct lineage from the eastern Aegean. Overall, the chloroplast data suggest that plastid variation within the D. fruticosus complex is structured more strongly by geography than by current taxonomic boundaries, probably reflecting the influence of historical Aegean fragmentation, retention of ancestral chloroplast variation, and subsequent lineage sorting. The complete chloroplast dataset also provided a clearer picture of relationships within the D. juniperinus assemblage. Dianthus juniperinus subsp. juniperinus and subsp. heldreichii formed a strongly supported lineage, while subsp. pulviniformis was recovered together with representatives of subsp. aciphyllus from the Asterousia Mountains. A further well-supported lineage comprised subsp. idaeus, whereas representatives of subsp. aciphyllus from the Dikti massif, together with subsp. bauhinorum and subsp. kavusicus, were associated with D. fruticosus subsp. creticus from the Asterousia Mountains and subsp. sitiacus. These results indicate that the chloroplast variation within D. juniperinus is also geographically structured and that some lineages are more closely associated with Cretan representatives of D. fruticosus than with the core D. juniperinus taxa.
The nuclear ribosomal phylogeny revealed a pattern broadly consistent with the ITS dataset, while providing substantially improved resolution and support. In contrast to the chloroplast phylogeny, the D. fruticosus complex was recovered as a well-supported clade, supporting its recognition as a single evolutionary lineage. Within this clade, however, relationships among subspecies were not always fully resolved or strongly supported. Dianthus fruticosus subsp. rhodius formed the earliest-diverging lineage, followed by subsp. amorginus. The remaining subspecies were divided into three main lineages. The Asterousia representative of subsp. creticus formed an independent lineage, whereas the northwestern Cretan representative of subsp. creticus was recovered as sister to the northwestern Cretan representative of subsp. occidentalis, thus reproducing the geographical pattern observed in the chloroplast phylogeny. Together with the Peloponnesian representative of subsp. occidentalis, these taxa formed a second lineage within the clade. The third lineage included the remaining representatives of the complex, namely subsp. sitiacus, subsp. carpathus, subsp. fruticosus, and subsp. karavius. Regarding Dianthus juniperinus, the ribosomal data cannot confirm the monophyly of the complex. Dianthus juniperinus subsp. kavusicus and subsp. aciphyllus from Asterousia, formed a lineage sister to the remaining representatives of D. juniperinus together with the entire D. fruticosus complex. The remaining representatives of D. juniperinus formed a strongly supported clade comprising two principal lineages, one consisting of subsp. aciphyllus (Dikti) together with subsp. bauhinorum, while the other includes subsp. juniperinus, subsp. idaeus, subsp. pulviniformis, and subsp. heldreichii. Notably, the separation between the Dikti and Asterousia representatives of subsp. aciphyllus was consistently recovered in both chloroplast and nuclear datasets, indicating a stable phylogenetic pattern within this taxon.
Overall, the molecular analyses consistently recovered D. fruticosus and D. juniperinus as closely related lineages, supporting their close evolutionary affinity despite differences in topology among datasets. Within each complex, a relatively stable core assemblage can be recognized. In D. fruticosus, this core comprises subsp. amorginus, subsp. fruticosus, and subsp. rhodius, whereas in D. juniperinus it includes subsp. juniperinus, subsp. heldreichii, subsp. idaeus, and subsp. pulviniformis. Several taxa show more complex or unstable phylogenetic placements, particularly D. juniperinus subsp. bauhinorum and subsp. aciphyllus, and D. fruticosus subsp. occidentalis and subsp. creticus.
Morphological variation within the D. fruticosus and D. juniperinus complexes was analyzed to evaluate patterns of phenotypic diversity and taxonomic differentiation among the examined taxa. Thirteen morphological characters were selected, including two qualitative and eleven quantitative characters, covering leaf, floral, and flowering stem features. Among the analyzed characters, the most informative were floral traits, particularly petal limb length and width, calyx length, calyx teeth length, and the number of bracts in the first and later flowers, together with leaf width among the vegetative traits. Across these characters, a consistent pattern was observed with the D. fruticosus subspecies generally exhibited higher values and broader ranges of variation, whereas most D. juniperinus subspecies were characterized by lower values and more compact morphological distributions. Notable exceptions within D. juniperinus were subsp. aciphyllus, subsp. bauhinorum, and, to a lesser extent, subsp. kavusicus, which displayed intermediate or elevated values approaching those of the D. fruticosus group, indicating that these taxa occupy an intermediate morphological position between the compact D. juniperinus core and the broader D. fruticosus assemblage. The overall morphological pattern revealed a clear contrast between the two complexes. Most D. juniperinus subspecies formed a compact and cohesive assemblage, whereas the D. fruticosus subspecies occupied a broader and more heterogeneous morphological space. The morphological core of D. juniperinus, comprising subsp. juniperinus, subsp. heldreichii, subsp. idaeus, and subsp. pulviniformis, was consistently recovered as a distinct group across the morphological analyses, a pattern observed in the molecular results. A clear geographical pattern was observed within subsp. aciphyllus, with specimens from Asterousia clustered with the core of D. juniperinus, while those from the E Dikti region showed intermediate positions toward the D. fruticosus group, also supported by molecular data, indicating geographical differentiation within the taxon. the morphological analyses revealed a broader and more heterogeneous morphospace for the D. fruticosus complex, which was partially approached by D. juniperinus subsp. bauhinorum and some representatives of subsp. aciphyllus.
Ecological niche analyses revealed additional patterns of differentiation within the D. fruticosus and D. juniperinus complexes. Overall, the D. fruticosus subspecies showed greater climatic niche overlap, although several taxa, particularly subsp. occidentalis, subsp. rhodius, subsp. sitiacus, and subsp. creticus, displayed more distinct or complex ecological affinities. These patterns partly correspond to the morphological and molecular evidence, suggesting that ecological differentiation reflects, at least in part, the evolutionary and geographical structure of the complex. In contrast, the D. juniperinus complex exhibited lower climatic niche overlap and stronger ecological differentiation among subspecies. Dianthus juniperinus subsp. juniperinus occupied a relatively isolated climatic niche, whereas subsp. idaeus and subsp. pulviniformis exhibited extensive niche overlap, representing one of the strongest ecological associations within the complex. One of the broadest climatic niche spaces, can be found in subsp. aciphyllus, a pattern consistent with the pronounced morphological and phylogenetic heterogeneity observed among its populations. Its ecological affinity with subsp. bauhinorum and subsp. kavusicus further supports its intermediate and heterogeneous position within the complex. A notable ecological association was also observed between subsp. heldreichii and subsp. kavusicus, which showed substantial climatic similarity despite their different positions in the molecular phylogenies.
The extinction risk assessments revealed substantial differences in conservation status among the examined taxa. Although both D. fruticosus and D. juniperinus are currently assessed as Least Concern at the species level, largely due to their relatively broad overall distributions and large total population sizes, several subspecies qualify for threatened categories. Within the D. fruticosus complex, four subspecies were assigned to threatened categories, whereas within D. juniperinus, five subspecies were classified as threatened.
Species distribution models consistently predicted substantial future habitat loss across both the D. fruticosus and D. juniperinus complexes, irrespective of the General Circulation Model, Representative Concentration Pathway, or modelling approach applied. Among the different modelling approaches, climate-change-only projections produced the most severe habitat losses, indicating that future climatic change is likely to represent the main driver of range contraction. Overall, the D. juniperinus complex appears considerably more vulnerable, with projected habitat losses approaching 59% by 2080, compared with approximately 37% for the D. fruticosus complex.
Overall, the integrative analyses reveal that the D. fruticosus and D. juniperinus complexes represent closely related but highly complex evolutionary lineages, whose present diversity cannot be fully explained by current taxonomic boundaries alone. The incongruence among molecular datasets, together with the morphological and ecological differentiation observed among taxa, suggests a history shaped by geographical isolation, lineage sorting, and possibly past gene flow. The strong correspondence between several molecular patterns and regional distribution further emphasizes the role of Aegean paleogeography in structuring diversification within the complex. Although the present study provides a comprehensive framework for interpreting these relationships, several taxa remain taxonomically and phylogenetically ambiguous. Further studies, including broader population-level sampling and additional genomic data, will be necessary to clarify their evolutionary position and to support a more robust taxonomic revision. Such clarification is essential not only for understanding the evolutionary history of the group, but also for establishing reliable conservation priorities for its narrowly distributed and threatened taxa.
Description
Keywords
Integrative taxonomy, Phylogenetics, Endemic species, Evolution history