To lump or to split? Revision of Cuscuta section Indecorae using a combined morphometric, phylogenetic and host range approach
Data files
Sep 12, 2024 version files 294.56 KB
-
Cuscuta-sect-indecorae-host-matrix.xlsx
89.59 KB
-
README.md
3.96 KB
-
Section_sect_Indecorae_MORPHOMETRIC-matrix.xlsx
201.01 KB
Abstract
Cuscuta section Indecorae is an infrageneric clade of subgenus Grammica that originated in N America which includes three species: C. indecora, C. coryli, and C. warneri. Some forms of C. indecora are agricultural weeds, while C. warneri and C. coryli face conservation problems regionally or globally. Cuscuta indecora is a morphologically diverse species with three recognized varieties: indecora, longisepala, and attenuata. Two recent broad-scale molecular phylogenetic studies raised questions about the circumscription of C. indecora and the validity of its infraspecific taxa. Therefore, we conducted an extensive morphometric study and used DNA sequences from plastid and nuclear ribosomal non-coding regions to reconstruct evolutionary relationships among taxa. In addition, we studied the host range of all taxa using herbarium specimens, bipartite networks, and Venn diagrams. Results from all types of evidence confirmed the circumscription of the three species but differed on the infraspecific structure of C. indecora. The molecular phylogeny inferred C. indecora as paraphyletic, consisting of two distinct clades, with C. coryli nested within one of them. The morphometric and host range studies uncovered extensive overlaps both among the traditional varieties of C. indecora as well as between the two molecular lineages of C. indecora. Previously, Iva annua had been considered the sole host of C. indecora var. attenuata (C. attenuata) and this presumed host specificity was the main criterion for accepting this taxon. We found that the type of var. longisepala was also growing on Iva annua, which together with the morphological and genetical similarity, indicates that var. attenuata is identical to var. longisepala. Overall, considering the lack of morphological, host range, and geographical patterns distinguishing either the current three varieties of C. indecora or the two cryptic lineages, the most appropriate solution is to tentatively recognize C. indecora as variable species without recognizing any infraspecific taxa.
README: To lump or to split? Revision of Cuscuta section Indecorae using a combined morphometric, phylogenetic and host range approach
https://doi.org/10.5061/dryad.p8cz8wb02
Description of the data and file structure
Two Excel files are included:
1) Section_sect_Indecorae_MORPHOMETRIC-matrix (Morphology and Morphometric Analyses)
This is the full dataset consisting of an Excel spreadsheet that includes all measurements and scores for 47 characters (columns) obtained from 188 herbarium specimens (lines) representing the taxa of sect. Indecorae. See more details about the rehydration of specimens and microscopy in the article. Empty cells in the Excel spreadsheet represent missing data ("gaps") for those herbarium specimens and characters specific values could not be obtained.
The characters measured/scored were:
Continuous Characters.
Pedicel length (mm)
Bract length (mm).
Bract width (measured at the base, mm).
Angle of bract apex (°).
Flower length (measured from the tips of corolla lobes to the receptacle base; mm).
Total calyx length (mm)
Calyx lobe length (mm). 6. Maximum calyx lobe width (mm).
Middle calyx lobe width (mm).
Maximum calyx lobe length divided by maximum calyx lobe width.
Calyx tube length divided by total calyx length.
Angle of calyx lobe apex (°).
Calyx tube length (mm).
Circumference of calyx tube (mm).
Total corolla length (mm).
Corolla lobe length (mm).
Maximum corolla lobe width (mm).
Middle corolla lobe width (mm).
Corolla lobe apex width (mm).
Length of inflexed corolla lobe apex (mm).
Angle of corolla lobe apex (°).
Ratio of corolla lobe apex length divided by corolla lobe apex width.
Corolla tube length (mm).
Circumference of corolla tube at base of staminal filaments (mm).
Maximum infrastaminal scale length (mm).
Infrastaminal scale width at base (mm).
maximum infrastaminal scale width (mm).
Length of free portion of infrastaminal scale (mm).
Infrastaminal scale bridge length (mm).
Length of longest infrastaminal scale fimbria (mm).
Number of infrastaminal scale fimbriae.
Stamen filament length (mm).
Anther length (mm).
Anther width (mm).
Length of longest style (mm).
Length of shortest style (mm);
Stigma width (mm);
Length of thickened apical portion of capsule (mm);
Capsule length (mm);
Capsule width (mm);
Seed length (mm);
Seed width (mm);
Hilum scar length (mm);
Vascular scar length (mm).
Binary Characters.
Flowers mostly 4-merous, rarely 3-merous: present (1), absent (0).
Flowers mostly 5-merous, rarely 4-merous present (1), absent (0).
Horn-like projections on calyx lobe apices: present (1), absent (0).
Corolla lobes connivent, not spreading or reflexed at full anthesis: present (1), absent (0).
Infrastaminal scale bifid: present (1), absent (0).
Capsule with distinct collar-like apex: present (1), absent (0).
The extent of morphological variation was visualized with both principal coordinates analysis (PCoA or multidimensional scaling) and unweighted pair-group average (UPGMA), using the Gower’s coefficient of similarity. Please see the article for more details and analyses.
2) Cuscuta-sect-indecorae-host-matrix (Host Range Assessment)
This is the dataset that includes 1) the raw host data obtained from 965 herbarium specimens at three taxonomic levels: family, genus, and species; and 2) the Host Frequency Index (HFI) calculated for the host family, genus, and species taxonomic levels. HFI data were mapped into a directed bipartite network using Cytoscape to visualize host range and frequency of occurrence across sect. Indecorae. In C. indecora, only the hosts of specimens that could be unambiguously assigned to a variety were included. To visualize the numbers of shared versus unique hosts, Venn diagrams we also generated for the entire dataset using InteractiVenn. Please see the article for more details and analyses.
Methods
1) Section_sect_Indecorae_MORPHOMETRIC-matrix. A total of 47 morphological characters were measured/scored using rehydrated plant material from 188 herbarium specimens. A total of 47 morphological characters were measured/scored (see article and Readme file). Empty cells in the Excel spreadsheet represent missing data for that particular trait and specimen. The extent of morphological variation was visualized with both principal coordinates analysis (PCoA or multidimensional scaling) and unweighted pair-group average (UPGMA), using the Gower’s coefficient of similarity.
2). Cuscuta-sect-indecorae-host-matrix. 964 herbarium specimens for which the host plants could be identified at species, genus, or family level were used to calculate a HFI index at these three taxonomic levels. The HFI shows the frequency (percentage) of a Cuscuta taxon to be encountered on a particular host (at family, genus, and species level), in relation to the total number of host occurrences. HFI data were then mapped into a directed bipartite network using Cytoscape to visualize host range and frequency of occurrence across section.