Plantago campestris (Plantaginaceae), a rare new species from southern Brazil, supported by phylogenomic and morphological evidence

High-throughput sequencing, when combined with taxonomic expertise, is a powerful tool to refine and advance taxonomic classification, including at the species level. In the present work, a new species, Plantago campestris, is described out of the P. commersoniana species complex, based on phylogenomic and morphological evidence. The main morphological characters that distinguish the new species from P. commersoniana are the glabrous posterior sepals and the slightly broader leaves. The new species is known from only three localities, all in natural high-elevation grasslands in Paraná and Santa Catarina states, southern Brazil. According to the IUCN criteria new species should be assessed as Endangered (EN). We present field photographs of P. campestris and related species, and we provide an identification key to the species previously included within the circumscription of P. commersoniana.

The Plantago commersoniana alliance constitute a group of plants which is hypothesised to have thrived during colder and drier periods in South America, when grasslands covered most of the land (Behling, 1998;Behling, 2002;Behling et al., 2007a). With the notable exceptions of P. guilleminiana, which occurs in high-elevation grasslands in southern Brazil, and P. berroi, which occurs in pampas in Uruguay and eastern Argentina, all members of this species complex are rare and threatened with extinction (Rahn, 1974;Hassemer, 2019), so that a good understanding of this group is therefore critical for biodiversity conservation (Ely et al., 2017;Thomson et al., 2018). It is illustrative that even the type-population of P. commersoniana was for a long time thought to be lost, having only been rediscovered two centuries after the collection of the type (Hassemer & Marchesi, 2016).
A number of new species were recently discovered and described out of the Plantago commersoniana complex (Villarroel & Wood, 2011;Hassemer, Baumann & Trevisan, 2014;Hassemer, 2016;Hassemer, 2019), and one species (P. bradei, see Pilger, 1949) was reestablished (Hassemer, 2017a). Phylogenetic reconstructions based on high-throughput sequencing, when combined with morphological data and taxonomic expertise, is a powerful tool to refine and advance taxonomic classification at the species level (e.g., Gardner et al., 2016;Hou et al., 2016;Simpson et al., 2017;Uribe-Convers et al., 2017;Hassemer et al., 2019). In the present work, a new species is described out of the P. commersoniana species complex, based on a whole-chloroplast phylogenomic analysis and morphology. We present field photographs of the new species and related species, and we provide an identification key to the species previously (Rahn, 1974) included within the circumscription of P. commersoniana, which is still the most commonly used/seen name in herbaria worldwide for specimens of the P. commersoniana alliance. Specimens representative of the recorded geographic distributions of the species included in the P. commersoniana alliance were studied, comprising several hundred specimens in total. The classification of trichome types follows Rahn (1992). The diagnosis was prepared according to the recommendations in Hassemer, Prado & Baldini (2020a). The species concept adopted follows Hołyński (2005) and De Queiroz (2007). The taxonomic work followed the hypothesis-driven framework as described in Henderson (2005), Hołyński (2005 and Sluys (2013). The conservation status assessment follows the IUCN (2012), IUCN (2019) criteria. The distribution of the type specimens of the new species is pending the return of normalcy of herbaria activities, which is currently interrupted due to the COVID-19 pandemic.

Plantago
To place the new species in a phylogenetic context, we generated phylogenetic trees based on whole chloroplast and nuclear ribosomal DNA (nrDNA) sequences. The chloroplast dataset consisted of 47 samples from the alignment used by For the sample of the new species, DNA was extracted from silica-dried leaf tissue of the type specimen (which serves as voucher) using the Qiagen DNeasy Plant Mini kit (Qiagen, Germany) following the manufacturer's protocol and the modifications described in Hassemer et al. (2019). DNA was quantified using high sensitivity reagents on Qubit 2.0 fluorometer (Life Technologies, USA) and fragmented to ca. 300 bp with a Bioruptor (Diagenode, Belgium) for four cycles of 15 s ON / 90 s OFF. Illumina TruSeq-style libraries were prepared using the NEBNext DNA Ultra II kit following the manufacturer's protocol. Libraries were amplified using AmpliTaq Gold (Life Technologies, USA) and quality checked on a TapeStation 2200 (Agilent Technologies, USA). Sequencing took place alongside other samples on a 2 × 125 bp run on an Illumina HiSeq 2000.
To generate the chloroplast assembly, sequences were assembled de novo using NOVOPlasty 4.3. 1 (Dierckxsens, Mardulyn & Smits, 2016). Because the sequences were not sufficient to assemble the entire plastome into a single contig in one run, we carried out two assemblies and combined them. The first assembly was seeded with a P. major L. trnL-trnF sequence extracted from the Hassemer et al. (2019) alignment, and the second was seeded with a psbA sequence from the same source. The three longest contigs from the two assemblies were aligned to P. major using MAFFT (Katoh & Standley, 2013) and manually combined using AliVew (Larsson, 2014) into a single scaffold with a gap of approximately 10kb. Sequences were aligned using MAFFT, columns with more than 75% gaps were removed with trimAl (Capella-Gutiérrez, Silla-Martínez & Gabaldón, 2009), and a maximum-likelihood tree was estimated under the best-fit model using IQtree (Minh, Nguyen & Haeseler, 2013). The figure was generated using ape 5.3 (Paradis, Claude & Strimmer, 2004) in R 3.5.1 (R Core Development Team, 2018). The plastome sequence of the sample of the type of the new species was deposited in GenBank (accession no. MW727694).
The electronic version of this article in Portable Document Format (PDF) will represent a published work according to the International Code of Nomenclature for algae, fungi, and plants (ICN), and hence the new names contained in the electronic version are effectively published under that Code from the electronic edition alone. In addition, new names contained in this work which have been issued with identifiers by IPNI will eventually be made available to the Global Names Index. The IPNI LSIDs can be resolved and the associated information viewed through any standard web browser by appending the LSID contained in this publication to the prefix ''http://ipni.org''. The online version of this work is archived and available from the following digital repositories: PeerJ, PubMed Central, and CLOCKSS.  (Fig. 1); several isotypes to be distributed, to C, FT, FURB, HBR, MBM, OLD, among other herbaria). Diagnosis: Similar to Plantago commersoniana, from which it differs by its darkening more noticeably on drying, the slightly broader narrow-lanceolate leaves, and the glabrous sepals. Description: Single-rosette herbs, up to 26 cm tall, perennial. Taproot developed, up to 5 cm long (probably much longer; it was broken in all specimens examined) and up to 0.9 cm wide; numerous unthickened (up to 1.5 mm wide) cord-like secondary roots also present. Caudex with negligible longitudinal growth, up to 1.7 cm wide, without a conspicuous crown of trichomes at its apex. Trichomes on leaves and scapes filiform, terete, with inconspicuous cellular articulations, very slender throughout their entire length, not perceptibly gradually tapering towards the apex (type K), whitish to light orange-coloured. Leaves 5.0-11.5 × 0.4-0.5 cm, papiraceous, 3-veined, basally attenuated, petiole indistinct from the narrow-lanceolate blade; apex acuminate; margin slightly to strongly involute, very   Fig. 2. Photographs: Fig. 1 (herbarium specimen); Fig. 3 (living specimen). Etymology: The epithet is a reference to the habitat of the species, i.e., the high-elevation grasslands of southern Brazil, which is perhaps the most neglected type of terrestrial environment in Brazil (Behling et al., 2007b;Overbeck et al., 2007;Overbeck et al., 2015), despite harbouring an impressive plant biodiversity and endemism (Iganci et al., 2011;Hassemer, Ferreira & Trevisan, 2015;Pla et al., 2020), frequently even greater than speciesrich tropical forest areas. With over 35,683 plant species recorded in its territory, Brazil harbours the greatest plant biodiversity in the world (The Brazil Flora Group, 2015;The Brazil Flora Group, 2018;The Brazil Flora Group, 2021). However, conservation attention and efforts in the country have almost always focused on forests, which has often led to poorly-informed conservation decisions, favouring forest advance (which involves common, pioneer tree species) over the conservation of grasslands rich in endemic species. We hope that the description of narrowly-endemic new species from the Brazilian grasslands will help to draw attention to the importance of conserving these environments, promoting better-informed conservation decisions that take these unique ecosystems into account. Distribution: Plantago campestris is recorded in three municipalities in southern Brazil: Candói and Cantagalo, in the central-southern part of the state of Paraná, and Campo Erê, in the western part of the state of Santa Catarina (Fig. 4). It is important to mention that the municipalities of Candói and Cantagalo were both until recently part of the municipality of Guarapuava, hence the indication of the latter in labels of herbarium specimens. Habitat: High-elevation grasslands ( Fig. 5; see also comments and references above, under Etymology), at elevations of 750-1,000 m asl. Conservation status: Endangered: EN-B2 (a, b[ii, iii, iv]). Only four populations are known, encompassing a total area of occupancy of <2 km 2 . As mentioned above, the high-elevation grasslands are a considerably neglected and highly threatened type of environment, with a clear ongoing tendency of area and quality reduction (Behling et al., 2007b;Overbeck et al., 2007;Overbeck et al., 2015;Hassemer, Ferreira & Trevisan, 2015).

Plantago campestris
The main threats to the high-elevation grasslands in the region are the agricultural advance, and to a lesser extent the urban advance and the forest advance (due to the removal of cattle, see e.g., Boldrini & Eggers, 1996;Sühs, Giehl & Peroni, 2020). Furthermore, none of the four recorded populations are within an environment protection area. Finally, the date of the most recent gathering is considerably old (>40 years) for all but one population, despite the fact that the states of Paraná and Santa Catarina can be considered well-sampled and are target of appreciable ongoing botanic sampling. This means that it is unfortunately possible that some of the recorded populations might be no longer extant. The first author, together with J.M. da Silva searched for the species in 2015 in suitable environments in adjacent areas in the state of Paraná but could not locate any further populations. Chromosome number: unknown. Discussion: In our plastome phylogenetic analysis (GTR+F+R5 model, log-lik =-334825.9250, AICc =669878.0542) (Fig. 6), the species in the Plantago commersoniana alliance included in this study (i.e., Plantago campestris, P. commersoniana, P. floccosa, P. guilleminiana, P. hatschbachiana and P. rahniana) formed a clade (BP = 100%) that does not include P. bradei and P. weddelliana. Within this P. commersoniana alliance clade, P. guilleminiana is sister to P. hatschbachiana (BS = 98%), followed by a grade of P. rahniana (BP = 97%), P. commersoniana (BP = 100%), P. campestris (BP = 100%) and P. floccosa (BP = 99%). Of the species in the P. commersoniana alliance, only P. berroi, P. pyrophila and P. veadeirensis were not sampled in this study. Based on morphology, we expect that P. hatschbachiana (sampled in this study), P. pyrophila and P. veadeirensis would form a clade. We cannot predict the position of P. berroi except that it most likely belongs to the clade of the P. commersoniana alliance. The remainder of the topology was identical to that found in Hassemer et al. (2019). The nrDNA tree (TVM+F+R2 model, log-lik =-12851.7151, AICc =25873.8488) was not as well resolved as the plastome tree, with only 13 nodes receiving at least 80% bootstrap support (Fig. 7). Thus, although the sectional clades agree with those in the plastome tree, the positions of individual species within sect. Virginica are often at odds with it. In the nrDNA tree, P. campestris is not part of the P. commersoniana clade, instead appearing in a clade with P. catharinea, P. napiformis, P. trinitatis and P. tomentosa.
The position of the new species as sister to a clade including Plantago commersoniana, P. guilleminiana, P. hatschbachiana and P. rahniana, with very high support values throughout, indicates that the circumscription of P. commersoniana prior to the description of P. campestris was paraphyletic. In the nrDNA tree, the latter two species do not even form a grade. Furthermore, the striking morphological (Fig. 8), ecological and geographic (Hassemer, 2017b;Hassemer, 2019) differences between the four species that constitute the clade sister to P. campestris provide a compelling argument for the recognition of P. campestris at the species rank. This molecular phylogenetic result confirmed long-time taxonomic suspicions by the first author, despite the subtle morphological differences between P. campestris and P. commersoniana (see the diagnosis above). See Fig. 3 in Hassemer (2019) for the recorded distributions of P. bradei, P. hatschbachiana, P. pyrophila, P. rahniana and P. veadeirensis, Fig. 7 in Hassemer (2017b) for the distribution of P. commersoniana, and Fig. 55 in Rahn (1974) and Fig. 2A in Hassemer, De Giovanni & Trevisan (2016)

Identification key to the species encompassed within Rahn's (1974) circumscription of Plantago commersoniana
Species distributions are presented within brackets.

ADDITIONAL INFORMATION AND DECLARATIONS Funding
This work was supported by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) through a doctoral scholarship granted to Gustavo Hassemer from 2014 to 2017 (process BEX 13765/13-2). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Grant Disclosures
The following grant information was disclosed by the authors: Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES).

Competing Interests
The authors declare there are no competing interests.

Author Contributions
• Gustavo Hassemer conceived and designed the experiments, performed the experiments, analyzed the data, prepared figures and/or tables, authored or reviewed drafts of the paper, and approved the final draft.
• Elliot M. Gardner analyzed the data, prepared figures and/or tables, authored or reviewed drafts of the paper, and approved the final draft.
• Nina Rønsted conceived and designed the experiments, analyzed the data, authored or reviewed drafts of the paper, and approved the final draft.

Field Study Permissions
The following information was supplied relating to field study approvals (i.e., approving body and any reference numbers): The type specimens of the new species were collected by G. Hassemer and J.M. da Silva in 2015. The specimens were collected outside any conservation area. G. Hassemer was then a PhD student financed by the Brazilian government, and J.M. da Silva was then a public servant working at a Brazilian herbarium, i.e., MBM herbarium (he has since retired). In accordance with the Brazilian legislation, Brazilian citizens do not need a permit to collect plants if the species is not in the official Red List and if the plants are not in an environmental conservation area. Thus, no permit was required for the collection of the type gathering of the new species.

DNA Deposition
The following information was supplied regarding the deposition of DNA sequences: The plastome sequence of the sample of the type of the new species is available in GenBank: MW727694.

Data Availability
The following information was supplied regarding data availability: The alignments used for our phylogenetic analyses are available in the Supplemental Information.
Voucher information for all samples used (except that of the new species) is available in Table 2  The voucher for the sample of the new species is the holotype of the species, which is deposited in the HTL herbarium.