Pollen morphology of Ellisiophyllum and Sibthorpia (Plantaginaceae, tribe Sibthorpieae) and phylogenetics of the tribe

Pollen morphology of six species belonging to genera Ellisiophyllum and Sibthorpia (Plantaginaceae tribe Sibthorpieae) was studied using light and scanning electron microscopy. The data were analyzed in the light of the first phylogenetic analysis including all but one species of the tribe using DNA sequence data from nuclear ribosomal (ITS) and plastid trnL-F region. Pollen grains in representatives of this tribe are 3-colpate, occasionally 3-porate, suboblate to prolate; mainly medium-sized, rarely small. One major pollen type (3-colpate) is recognized in the tribe. Within this pollen type, six subtypes are distinguished based on their exine sculpture, pollen grain size, length of the apertures, and exine thickness. The obtained results confirm that pollen characters are useful for species identification. Palynomorphological data are consistent with the results of the molecular phylogenetic analyses. All studies support a sister relationship of the widespread European Sibthorpia europaea with the widespread South American Sibthorpia repens and a sister relationship of two insular species, the Balearic Sibthorpia africana and the Madeiran Sibthorpia peregrina. Pollen grains in the tribe Sibthorpieae have both reticulate exine sculpture characteristic for representatives of the Russelieae–Cheloneae–Antirrhineae clades of Plantaginaceae, and also nanoechinate sculpture, which is typical for the Veroniceae and Plantagineae clades of that family. Also, in Sibthorpia repens, we observe a possible transition from the colpate type to the porate type typical for taxa of Plantago and Littorella.


Introduction
The circumscription of the family Scrophulariaceae has greatly changed since the first report of its polyphyly (Olmstead and Reeves 1995), and members of the traditional Scrophulariaceae are now split among at least eight families representing monophyletic lineages. Polyphyly extends also to traditional subfamilies and tribes of the family, and thus, reevaluation of the importance of characters in genera of traditional Scrophulariaceae is necessary. The tribe Sibthorpieae Benth. was established by Bentham (1846) with eleven genera, two now belonging to Phrymaceae, three to Scrophulariaceae, and seven to Plantaginaceae. However, later systems combined these genera with Digitalis L., Veronica L., and related genera, placing them in Digitalideae (Wettstein 1891-1893), or subsumed Sibthorpia (with Hemiphragma Wall., Scoparia L. and Capraria L., the latter now in Scrophulariaceae sensu stricto) under Hemiphragmeae (Rouy 1909). Wettstein's system was followed by most authors, for example by Takhtajan (1987Takhtajan ( , 1997, who included them in the tribe Veroniceae. Fischer (2004) restricted Sibthorpieae to only two genera, Ellisiophyllum Maxim. and Sibthorpia L. and placed the tribe in subfamily Digitalidoideae. Molecular phylogenetic studies of Ellisiophyllum and Sibthorpia were first conducted by Albach et al. (2005) who confirmed that they are phylogenetically closely related to each other and unrelated to genera previously considered close to them. Sibthorpieae, as outlined now, thus includes only the genera Ellisiophyllum and Sibthorpia (Albach et al. 2005;Tank et al. 2006;Reveal 2012;Olmstead 2016).
The genus Sibthorpia includes five currently recognized species that occur in tropical America, the Azores, Madeira, Europe (two species), and African mountains (Hedberg 1955(Hedberg , 1975Diaz-Miranda 1988;Mabberley 1997Mabberley , 2017Fischer 2004;Albach et al. 2005;Tank et al. 2006;Olmstead 2016). A comprehensive taxonomic treatment of Sibthorpia was published by Hedberg (1955). The morphological features of flowers, fruits, seeds, and chromosome numbers of the genus in general (Hedberg 1975) and in Sibthorpia europaea L. in particular (Juan et al. 1999) were investigated. Based on his investigations, Hedberg (1955) suggested that the Balearic Sibthorpia africana L. and the Madeiran Sibthorpia peregrina L. are sister species, which was supported by the same chromosome number (Hedberg 1975). In turn, he hypothesized that the Neotropical Sibthorpia repens (L.) Kuntze and the closely related S. conspicua Diels are tetraploid derivatives of the diploid European-African S. europaea (Hedberg 1955(Hedberg , 1975. To date, this phylogenetic hypothesis has not been tested in a phylogenetic analysis. The genus Ellisiophyllum is represented by the only species, E. pinnatum (Benth.) Makino, which is distributed from India to Japan and Taiwan, and to eastern New Guinea (Hedberg 1975;Mabberley 1997Mabberley , 2017Fischer 2004;Olmstead 2016). The species was originally described by Bentham (1846) based on the specimen(s) collected by Wallich in Nepal or adjacent regions of India and listed in his handwritten catalog under No. 3915.
Earlier opinions on the proper phylogenetic position and relationships of Ellisiophyllum varied greatly. Wallich provisionally listed the species under the name Mazus pinnatus Wall. (nom. inval., nom. nudum), in a genus now placed in Phrymaceae, but Bentham validly published it as Ourisia pinnata Benth. (Bentham 1835;see also Hayata 1911;Meudt 2006, etc.). Later, Bentham (1846) described the genus Hornemannia Benth. for it, an illegitimate later homonym of Hornemannia Willd., and put the species in his order close to Sibthorpia. Maximowicz (1871) established the new genus Ellisiophyllum with one species, E. reptans Maxim. The names of the genus and its only species were simultaneously validated by one description (descriptio generico-specifica, Art. 38.5 of the ICN; Turland et al. 2018). Most probably Maximowicz was unaware of the identity (or at least similarity) of his newly described species with the species earlier described by Bentham as Ourisia pinnata, which is understandable, partly because these taxa were described from distant territories: Japan and Nepal (or India), respectively. Maximowicz (1871: 223) characterized his genus as being intermediate "inter Hydrophyllaceas et Polemoniaceas." It was consequently included in the family Hydrophyllaceae by Peter (1897). Hooker (1885), however, considered Ellisiophyllum to be a synonym of Sibthorpia. Hemsley (1899) disagreed with that generic placement and, being aware of the illegitimacy of Bentham's generic name Hornemannia but evidently not knowing about the availability of the name Ellisiophyllum, coined the replacement name Mosleya Hemsl. (to replace Hornemannia Benth.) and validated the combination M. pinnata (Benth.) Hemsl. Evidently, Ellisiophyllum has priority over Mosleya at the genus rank. Brand (1913: 185-186) definitely excluded Ellisiophyllum from Hydrophyllaceae and confirmed instead its placement in Scrophulariaceae ("Genus Scrophulariaceis attribuendum"). Recent molecular and other findings (see an overview above) firmly placed Ellisiophyllum and Sibthorpia in the extended and re-circumscribed Plantaginaceae.
With the gained certainty in the familial relationships and phylogenetic hypotheses available, it is timely to reinterpret trends in character evolution and investigate poorly known pollen characters in a phylogenetic framework. For example, very little information is available on pollen grains of representatives of Sibthorpieae. The morphological features of pollen grains of S. europaea (Juan et al. 1999) have been described. However, as far as we know, pollen grains of the monotypic (monospecific) genus Ellisiophyllum and the other species of Sibthorpia have not been investigated before.
The purpose of the present research was to study and analyze the phylogenetic relationships among members of the tribe Sibthorpieae using DNA sequence data and to compare them with data on morphological features of pollen grains of these taxa.

DNA-based phylogenetic analysis
For the DNA-based part of the study, we have sampled four of the five species of Sibthorpia and the only species of Ellisiophyllum, with two or three samples of three of the species (Table 1). Only samples of S. conspicua were not available for DNA sequencing. Outgroups were chosen based on the analysis of Plantaginaceae by Albach et al. (2005) to ensure a wide variety of taxa and sufficient representation of the family (Table 1). DNA was isolated from about 20 mg of tissue from either living material, silica gel-dried or herbarium material with the NucleoSpin Plant II (Macherey and Nagel, Düren, Germany) or the DNeasy plant Mini Kit (Qiagen, Hilden Germany) following the provided protocol. The quality of the extracted DNA was checked on a 0.8% TBE-agarose-gel and the concentration measured spectrophotometrically with a GeneQuant RNA/DNA calculator (Pharmacia, Cambridge, UK). The nuclear ribosomal ITS region (hereafter ITS) and the plastid trnL intron, trnL 3´ exon and trnL-F spacer (hereafter trnL-F region) were amplified using primers ITS A (Blattner 1999) and ITS4 (White et al. 1990) for ITS, and the trnL-F region with primers c and f and sometimes including internal primers d and e (Taberlet et al. 1991). PCR reactions included 2-2.5 mM MgCl 2 , 8 mM bovine serum albumin, 0.4 µm primer, 0.2 mM dNTP, 1U/µl Taq polymerase (New England Biolabs, Ipswich, MA, USA), 1 × polymerase buffer and 1-5 µl DNA for a final volume of 25 µl. ITS sequences were amplified with a program consisting of 2 min at 95 °C followed by 36 cycles of 1 min at 95 °C, 1 min at 50-55 °C, and 1.5-2 min at 72 °C with a final extension of 5 min at 72 °C on either a Mastercycler gradient (Eppendorf) or TProfessional Standard thermocycler (Biometra). The trnL-F region was amplified after 1 min denaturation at 95 °C followed by 35 cycles with 30 s at 95 °C, 30 s at 52 °C and 1 min at 72° with a final extension of 8 min at 72 °C. PCR products were cleaned using QIAquick PCR purification kits (Qiagen, Hilden, Germany) following the provided protocol. Sequencing reactions of 10 µl were carried out using 1 µl of the Taq DyeDeoxy Terminator Cycle Sequencing mix (Applied Biosystems, Foster City, CA, USA) and the same primers as for PCR. Sequences were generated by Sanger sequencing at commercial sequencing companies. All sequences are available from GenBank (Table 1). The data matrices are available at http:// purl. org/ phylo/ treeb ase/ phylo ws/ study/ TB2: S25825.
Sequences were manually aligned in Phyde v.0.9971 (Müller et al. 2010) and evaluated for the best model of evolution in jModeltest2 (Darriba et al. 2012). No indel coding was conducted due to the high variability of the ITS region across Plantaginaceae. Phylogenetic analyses were conducted in IQ-TREE (Trifinopoulos et al. 2016) using the GTR + Γ + I for ITS and GTR + Γ for trnL-F with 8 different rates and 1000 ultrafast bootstrap replicates.

Pollen analysis
Pollen grains of two species belonging to two genera of Sibthorpieae (Ellisiophyllum and Sibthorpia) were sampled in the herbarium of the Missouri Botanical Garden (MO; St. Louis, Missouri, U.S.A.). Pollen grains of four species of Sibthorpia were sampled in the herbarium of the Conservatoire et Jardin botaniques de la Ville de Genève (G, Genève, Switzerland). Pollen grains of two species of Sibthorpia were sampled in the National Herbarium of Ukraine (KW-herbarium of the M.G. Kholodny Institute of Botany, National Academy of Sciences of Ukraine, Kyiv, Ukraine). The specimens examined are listed in "Appendix" section. Herbarium acronyms are given following Index Herbariorum (Thiers 2008-onward).
The methods used in the present study are essentially the same as we used earlier Tsymbalyuk 2015a, b, 2017). Pollen morphology was studied using light microscopy and scanning electron microscopy. For light microscopy (LM) studies (Biolar, × 700), the pollen was acetolyzed following Erdtman (1952), mounted on slides with glycerinated gelatin and analyzed and photomicrographed using light microscopy. Pollen morphometric features of 20 properly developed pollen grains from each specimen were measured on the acetolyzed pollen grains, and the measurements included the following parameters: polar axis (P), equatorial diameter (E), mesocolpium diameter, exine thickness, and 10 measurements of the apocolpium diameter, the width and length of apertures were performed. The P/E ratio was calculated in order to determine pollen shape. For all the quantitative characters, descriptive statistics was applied and the range (minimum and maximum values), arithmetic mean and standard deviation were calculated (Tables 2 and 3). The slides were deposited in the Palynotheca (reference pollen collection) at the National Herbarium of Ukraine (KW) (Bezusko and Tsymbalyuk 2011).
For scanning electron microscopy (SEM) studies (JEOL JSM-6060LA), dry pollen grains were treated with 96%-ethanol; then, these samples were sputter-coated with gold and investigated at the Center of Electron Microscopy of the M.G. Kholodny Institute of Botany. Terminology used in descriptions of pollen grains mainly follows the glossaries by Punt et al. (2007) and Halbritter et al. (2018).
Evolution of pollen characters was analyzed with the ancestral character state model using the package phytools (Revell 2012) in RStudio v. 1.4 (RStudio Team 2021) and R version 4.0.3 (R Development Core Team 2020) using the ITS species tree restricted to Sibthorpieae.

DNA-based phylogenetic analysis
The ITS dataset included 38 sequences with a final alignment of 832 characters with 352 potentially parsimony informative, whereas the trnL-F region included 34 sequences with 1137 characters with 254 potentially parsimony informative. The optimal tree from the maximum likelihood analyses of each dataset separate are shown in Figs. 1 and 2. Analyses of ITS and trnL-F region were congruent for relationships within the Sibthorpieae. Relationships among the outgroups are inconclusive because of incongruence among markers. Noteworthy is the difference among both datasets regarding the closest relatives of Sibthorpieae. However, in both cases Sibthorpieae branch deeply within Plantaginaceae. In turn, the Sibthorpieae clade itself is strongly supported to be monophyletic by analyses of both ITS and trnL-F region (Figs. 1, 2; 100% and 99% bootstrap support (BS), respectively) with Ellisiophyllum pinnatum sister to Sibthorpia in both analyses (100% BS). Within Sibthorpia, all species sampled by multiple individuals are monophyletic. Amplification of S. africana was unsuccessful for ITS but is sister to S. peregrina in the analysis of the trnL-F region (99% BS). Sibthorpia europaea and S. repens are sisters (100% BS).

General description of pollen grains of Ellisiophyllum
Pollen grains are monads, radially symmetrical, isopolar, tricolpate. Ellisiophyllum pollen is medium-sized (P = 30.59-42.56 µm, E = 25.27-34.58 µm). According to P/E ratio, pollen grains are oblate-spheroidal to prolate (P/E = 0.96-1.63) in shape. Outline of pollen grains in equatorial view is elliptic. Outline of pollen   (Tables 2 and 3). Colpus membranes are rugulate-nanoechinate (Fig. 3c). Exine is 1.59-2.66 µm thick (Table 2). Sexine is thicker than nexine. Tectum is nearly equal to infratectum, columellae distinct. Exine sculpture is rugulate-nanoechinate, nanoechinate (Fig. 3b, c).  The data obtained demonstrated that the pollen grains of Sibthorpieae differ in their shape, outline, and size, length and width of the colpi, exine thickness, exine sculpture, and aperture membranes between species. This confirms that pollen grain characteristics are useful for species identification. Pollen grains of the studied species can be included in one type (3-colpate). This type in Sibthorpieae contains six subtypes segregated according to the exine sculpture, grain size, length of apertures, and thickness of the exine (Table 4).

Discussion
The phylogenetic analyses based on both ITS (Fig. 1) and plastid trnL-F region (Fig. 2) are congruent with the hypothesis of Hedberg (1955) that S. europaea is sister to S. repens while S. africana is sister to S. peregrina. Hedberg (1955) hypothesized these relationships based on marked difference in seed and pollen size between the two species pairs, and later (Hedberg 1975) also added base chromosome numbers and crossability between the species as the characters supporting that phylogenetic scheme, which agrees with our analyses (Fig. 6). Species of S. africana and S. peregrina have the basic chromosome number x = 10 and larger pollen grains (Table 2; Fig. 6), while in S. europaea, S. repens and S. conspicua the basic chromosome number is x = 9. The pollen grains of these three species have smaller sizes as compared to pollen of S. africana and S. peregrina (Hedberg 1955;Juan et al. 1999; Table 2). Also, pollen grains of S. europaea, S. repens and S. conspicua all have perforate to reticulate exine ornamentation (Fig. 4) and also agree in their general shape and outline despite that S. repens is tetra-to octoploid compared to S. europaea based on known chromosome numbers (Hedberg 1975). These results suggest that a long-distance dispersal event occurred across the Atlantic Ocean relatively recently, and that migration was unidirectional, from Europe to America. Thus, Sibthorpia adds to the known examples of Mediterranean-American disjunctions (Raven 1973). Similar to most other examples, in that case, the phylogenetic relationships suggest a Mediterranean origin of the group. However, the Sibthorpia case has notable differences as compared to other examples of similar disjunctions. A number of studies have demonstrated a Miocene origin of the Madrean-Tethyan type of disjunctions between California and the Mediterranean region (e.g., Wen and Ickert-Bond 2009;Vargas et al. 2014) contributing to the evolution of the typical Mediterranean floras in both regions. Others have shown even more recent origins (within the last 500.000 years) of disjunctions between both regions in plants living in deserts (e.g., Coleman et al. 2003;Meyers and Liston 2008;Martín-Bravo et al. 2009). Sibthorpia europaea and S. repens, however, do not occur in typical Mediterranean, at least seasonally arid environments but instead are mostly confined to moist and shady places of montane forests (Hedberg 1955). Additionally, they differ from other examples in their more widespread occurrence in the New World, from Mexico southward to Argentina. The timing of the disjunctions is uncertain since molecular dating in Sibthorpieae is problematic due to the scarcity of fossils in the predominantly herbaceous family, the nucleotide substitution rate heterogeneity among species, and the incongruence among the outgroup taxa (Albach et al. 2005).
The sister-group relationship previously found between Sibthorpia and Ellisiophyllum (Albach et al. 2005) has been supported here with increased taxon sampling in Sibthorpia and is also supported by such pollen characters as the type of apertures, exine sculpture, shape, outline, size, and exine thickness (Tables 2,3;Figs. 3,4,and 5). Whereas comparison with Ellisiophyllum may help in explaining evolutionary trends in phenotypic characters, it adds even more complexity to the biogeographic scenario in the tribe. Ellisiophyllum shares with S. europaea/S. repens the base chromosome number of x = 9 (Borgmann 1964) and with the former the white color of the flower. It shares, however, with S. africana / S. peregrina the larger pollen (Table 2) and also the larger seeds (Hong et al. 1998). Also, pollen grains of Ellisiophyllum are similar to those in S. africana and S. peregrina by the type of apertures, shape, and outline. The exine sculpture is rugulate-nanoechinate, nanoechinate in Ellisiophyllum (Fig. 3b, c), nanoechinate-perforate, nanoechinate in S. peregrina (Fig. 3f), and rugulate-perforate in S. africana (Fig. 3h, i). Biogeographically, the Himalayan-to-East Asian distribution area suggests either another case of longdistance dispersal or, in this case more likely, a Himalayan-Mediterranean vicariance event similar to the one seen in the related Veroniceae (Surina et al. 2014). Based on ancestral character estimation, the larger pollen and seeds seem to be the ancestral condition (Figs. 6 and 7) and suggest an ancient Tethyan distribution of early evolved (ancestral) Sibthorpieae. However, this character evolution needs to be considered in the light of character evolution in the family.
Pollen grains in taxa of Sibthorpieae are characterized by a perforate to reticulate exine sculpture that is common in most of species of the Russelieae-Cheloneae-Antirrhineae clades of Plantaginaceae (Tsymbalyuk 2013(Tsymbalyuk , 2016Mosyakin 2013, 2014). Also, in Ellisiophyllum pinnatum and Sibthorpia peregrina, the types of exine sculpture were observed (such as rugulate-nanoechinate, nanoechinate, nanoechinate-perforate), which are more typical for the Veroniceae-Plantagineae clade of the family (Hong 1984;Fernández et al. 1997;Martínez-Ortega et al. 2000;Saeidi-Mehrvarz and Zarrei 2006;Tsymbalyuk 2008;Mosyakin and Tsymbalyuk 2008;Sánchez-Agudo et al. 2009;Tsymbalyuk et al. 2011;Tsymbalyuk and Mosyakin 2013;Tsymbalyuk 2016;Halbritter 2015Halbritter , 2016Halbritter and Svojtka 2016a, b). In species of Sibthorpia, we observed a transition from the colpate type to the porate type; the latter is also typical for representatives of some taxa of Veronica L., and especially for Littorella Asch. and Plantago L., but this seems to be a parallel trend. Furthermore, pollen with a perforate and reticulate exine sculpture is hypothesized to be a plesiomorphic condition within Plantaginaceae. The porate pollen probably represents an apomorphy in this tribe. However, this requires a more robust phylogenetic hypothesis for relationships within the family.
We noted that there is not just a topological difference between DNA regions analyzed but also between our analysis and that of Albach et al. (2005) as well as between different types of analyses (preliminary parsimony and neighbor-joining analyses and maximum likelihood analyses). Based on our experience with the dataset, we especially assume that different alignments of highly variable regions of the ITS region are prone to cause different relationships. Plantaginaceae are congruently divided into two clades, Plantaginoideae and Gratioloideae, with Sibthorpieae being one of ten tribes in the former. The five tribes, Plantagineae, Veroniceae, Digitaleae, Globularieae and Hemiphragmeae, consistently form clades in phylogenetic analyses (Albach et al. 2005;Figs. 1 and 2) but the relationship between this PVDGH-clade and the other tribes, Cheloneae, Antirrhineae, Callitricheae, Russelieae, and Sibthorpieae, differs considerably between analyses. In the analyses of ITS and the plastid rps16 intron of Albach et al. (2005), Sibthorpieae are even sister to Gratioloideae but this was not confirmed here, although in ITS it is sister to all Plantaginoideae. Based on the uncertainty in relationships between tribes of Plantaginaceae and the large variation of pollen and seed characters in the family, we will await a more robust phylogenetic hypothesis for relationships in the family to conduct a family wide analysis of pollen and seed characters.

Conclusions
The present study provides the first characterization of pollen grains of Ellisiophyllum. Images using scanning electron microscopy (SEM) were obtained for the first time for Ellisiophyllum and S. peregrina, S. africana, S. conspicua, and S. repens, which allowed more detailed descriptions of pollen characters in this group. We found variation in pollen grains morphology in Sibthorpieae, confirming its eurypalynous nature. Palynomorphological data support the placement of Ellisiophyllum and Sibthorpia in the welldefined tribe Sibthorpieae based on shared peculiarities such as shape, outline, size, exine thickness, exine sculpture, and the tricolpate type of pollen grains. The results of the current study expand the palynomorphological data for Sibthorpieae in particular and Plantaginaceae in general and will also contribute to future phylogenetic and taxonomic studies in this group.
Acknowledgements The authors express their gratitude to James C. Availability of data and material All DNA sequence data are freely available from GenBank after publication. All other data are included in the manuscript.

Declarations
Conflict of interest The authors declare that they have no conflict of interest. Fig. 7 ContMaps of quantitative characters generated using the package phytools (Revell 2012) in RStudio v. 1.4 (RStudio Team 2021) and R version 4.0.3 (R Development Core Team 2020) using the ITS species tree restricted to Sibthorpieae: a colpus length, b seed length, c mesocolpium length, d pollen size, e P/E ratio ◂ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.