Description of Longidorus bordonensis sp. nov. from Portugal, with systematics and molecular phylogeny of the genus (Nematoda, Longidoridae)

The genus Longidorus currently comprises 176 species of polyphagous plant ectoparasites, including eight species that vector nepoviruses. Longidorus is one of the most difficult genera to accurately identify species because of the similar morphology and overlapping measurements and ratios among species. Sequences of ribosomal RNA (rRNA)-genes are a powerful level-species diagnostic tool for the genus Longidorus. From 2015 to 2019, a nematode survey was conducted in vineyards and agro-forest environments in Portugal. The populations of Longidorus spp. were characterized through an integrative approach based on morphological data and molecular phylogenetic analysis from rRNA genes (D2-D3 expansion segments of the 28S, ITS1, and partial 18S), including the topotype of L. vinearum. Longidorus bordonensis sp. nov., a didelphic species recovered from the rhizosphere of grasses, is described and illustrated. Longidorus vineacola, with cork oak and wild olive as hosts, is also characterized. This is the first time that L. wicuolea, from cork oak, is reported for Portugal. Bayesian inference (BI) phylogenetic trees for these three molecular markers established phylogenetic relationships among the new species with other Longidorus spp. Phylogenetic trees indicated that i) L. bordonensis sp. nov. is clustered together with other Longidorus spp. and forms a sister clade with L. pini and L. carpetanensis, sharing a short body and odontostyle length, and elongate to conical female tail, and ii) all the other species described and illustrated are phylogenetically associated, including the topotype isolate of L. vinearum.

According to the morphological features and morphometric measurements of adult (mainly females) and that of juveniles [mainly first-stage juveniles (J1)], each Longidorus species is defined by 11 matrix codes (A-K) in the polytomous key published by Chen et al. (1997) and two additional supplements published by Loof and Chen (1999) and Peneva et al. (2013). However, the high intraspecific variability of some diagnostic features and the great diversity in phenotypic plasticity makes species identification based on metric and morphological data of external morphology and internal anatomy difficult and sometimes unreliable. Sequencing of RNA-based markers is a powerful molecular diagnostic approach within this group (Archidona-Yuste et al. 2016Palomares-Rius et al. 2017;Peraza-Padilla et al. 2017;Tzortzakakis et al. 2017;Xu et al. 2018). Recently, several studies (Archidona-Yuste et al. 2016Roshan-Bakhsh et al. 2016;Esmaeili et al. 2017;Tzortzakakis et al. 2017;Xu et al. 2017Xu et al. , 2018Barsalote et al. 2018;Cai et al. 2019;Lazarova et al. 2019) have shown the usefulness of a pair of molecular markers based on ribosomal RNA (rRNA) (D2-D3 domains of 28S gene and the ITS regions, particularly ITS1) for a fast and precise diagnosis of Longidorus species, even in extreme situations such as sibling and cryptic species. Subbotin et al. (2015) and Barsi and De Luca (2008) designed a PCR-RFLP of the D2-D3 segments of the 28S rRNA and ITS. Both assays were based on five restriction enzymes for genotyping of species-specific variations: the first from L. orientalis Loof, 1982, and the other from L. pius Barsi &Lamberti, 2001 (Barsi andDe Luca 2008;Subbotin et al. 2015). Additionally, studies have revealed that the mitochondrial marker gene, COI, is useful for the delineation of closely related species within Longidoridae Archidona-Yuste et al. 2019;Cai et al. 2019). Thus far, more than half of the valid Longidorus species have molecular markers deposited in the GenBank database; however, only a small number belong to topotypes. Genomic data of topotypes are very useful for confirming identifications and clarifying the composition of species complexes within the Longidoridae (Gutiérrez-Gutiérrez et al. 2010Kornobis et al. 2017;Archidona-Yuste et al. 2019;Fouladvand et al. 2019).
Members of the genus Longidorus have not been studied in detail during the past 18 years in Portugal, and updated information on the present occurrence and distribution is lacking, as well as molecular data (Gutiér-rez-Gutiérrez et al. 2016). This prompted us to carry out surveys in vineyards and agro-forestry systems in Portugal, from 2015 to 2019. The objectives of the present work are: 1) to characterise 11 populations of Longidorus species through an integrative approach based on morphological, morphometric, and molecular data, including topotypes of L. vinearum and of L. bordonensis sp. nov.; 2) to establish phylogenetic relationships of the identified Longidorus species from the surveys with available sequences of the known species.

Methods
Nematode population sampling, extraction, and morphological characterization Nematode surveys were conducted in spring and autumn from 2015 to 2019 in vineyards (Vitis sp.) and agro-forestry soils and included several host plants (Table 1). A total of 65 and 85 sampling sites of vineyards and agro-forestry areas, respectively, were arbitrarily chosen in Portugal. Field samples were taken in a zigzag pattern according to EPPO diagnostic protocols (OEPP/EPPO 2009). Each sample was collected using a drill from the upper 60 cm of the rhizosphere of 10-20 plants (sub-samples) from each field. Nematodes were extracted from 250 cm 3 of soil by a modification of Cobb's decanting and sieving method (Flegg 1967). Additional soil was collected to guarantee enough specimens for morphological and/or molecular analyses.
Nematodes were placed in a drop of water, killed in a hot fixative solution (4% formaldehyde + 1% glycerol + 85% distilled water), maintained for 48-72 h at room temperature (25 °C), and processed into pure glycerine by a modification of Seinhorst's method (Seinhorst 1966). Specimens were examined using an Olympus BX50 light microscope with differential interference contrast (DIC) up to 1,000× magnification. Photographs were taken with an Olympus DP70 camera. Cell software (Olympus Software Imaging for Life Sciences) was used for image analysis and measurements. All measurements were expressed in micrometers (µm). For line drawings of the new species, light micrographs were imported to CorelDraw v. X6 and the main features were outlined. All abbreviations are as defined in Jairajpuri and Ahmad (1992). According to metric (such as lip region length and width, body length, odontostyle length, maximum body width, guiding ring position, vulva position, pharyngeal length, and tail length and width) and non-metric (e.g., tail shape, size, and position of amphidial fovea, vulva size and shape, and lip region shape) morphological data of adult specimens and juveniles (J1-J4), each species was defined by the matrix code for the polytomous key (Chen et al. 1997;Loof and Chen 1999;Peneva et al. 2013). In addition, specimens of L. vinearum from its type locality, Dois Portos, Torres Vedras, Portugal, were collected. After verifying that their morphology was consistent with that of the original description, they were processed for their genotypic and phenotypic characterizations, and included as one of the 11 populations (Table 1). DNA extraction, PCR, and sequencing After nematodes were extracted from the soil, specimens were examined by light microscopy (LM) on temporary glass slide mounts and digital images were recorded. These photomicrographs were used to match each phenotype with its associated genotype. Temporary slides were dismantled and individual nematodes were placed in a 2 µl drop of sterile water on the cover of a PCR tube, and the specimen was cut into six small pieces with a surface-sterilized scalpel. Subsequently, they were centrifuged in 18 µl of solution containing 10 µl ddH 2 O, 6 µl 10× PCR buffer, and 2 µl of proteinase K (20 mg/ml) (Nalgene), and frozen at −80 °C (15 min). Samples were mixed for 15 sec and PCR assays were conducted as described by Gutiérrez-Gutiérrez et al. (2018). The tubes were incubated at 57 °C (1 h), 65 °C (1 h), and 95 °C (15 min). Half of one µl of extracted DNA was transferred to an Eppendorf tube containing reaction mixtures of 25 ul NZYTaq 2× Green Master Mix (2.5 mM MgCl 2 , 200 mM dNTPs, 0.2 U/µl DNA Polymerase) (NYZTech, Portugal), 0.4 µl of each primer (25 mM), and ddH 2 O was added to make a final volume of 50 µl. The D2-D3 expansion segments of 28S rRNA gene, the ITS1 region of rRNA, and a partial potion of 18S rRNA gene were amplified using several primer pairs (Suppl. material 3: Table S1).
PCR cycle conditions for markers of ribosomal DNA included one cycle of 95 °C for 3 min; followed by 30 cycles of 94 °C for 30 s; an annealing temperature of 54 °C (D2A/D3B or 28LeX/ D3B), 53 °C (rDNA1/18S), and 50 °C (988F/1912R, 1813F/2646R) for 30 s, 72 °C for 15-45 s; and one cycle of 72 °C for 7 min. PCR products were purified after amplification using NZYGelpure (NZYTech Genes & Enzymes, Portugal) following the manufacturer's instructions and used as template for direct sequencing at Eurofins Genomic (Germany) using the primers listed (Suppl. material 3: Table S1). The sequences were deposited in the GenBank database under accession numbers (Table 1) and used for constructing phylogenetic trees (Figs 3-5) were inferred using the methods descripted in the following section.
Type locality. Holotype and paratype specimens were extracted from a soil sample collected from the rhizosphere of an unidentified grass species at Bordonhos, São Pedro do Sul, Viseu district, Beira Alta province, northern Portugal (40°45'53"N, 8°5'12"W) (Table 1) Etymology. The specific epithet of this species refers to the region of the type locality (Bordonhos) where the new species was found.

Description of male.
Males are as common as females.

Longidorus vinearum Bravo & Roca, 1995
Suppl. material 1: Fig. S1(1-9), Suppl. material 5: Table S3 Remarks. Longidorus vinearum was originally described from around roots of grapevine (Vitis L.) in Dois Portos, Torres Vedras, Portugal (Bravo and Roca 1995). Subsequently, Bravo and Roca (1998) Table S3). These populations prompted us to characterize them genotypically and phenotypically, including the topotype specimens, in order to confirm their identification. Unfortunately, only one specimen was found at Picanceira, Mafra (Table 1) and used to complete the molecular analysis. These findings represent the third and fourth records of this species for Portugal and the Iberian Peninsula, respectively. We confirm a wider geographical distribution of this species in this geographical region.
Longidorus vinearum populations are characterized by a lip region, which is broadly rounded frontally, and more so laterally, and almost totally continuous with the outline of the body; a vulva near mid-body; the amphidial fovea large and clearly asymmetrically bilobed; the odontostyle long and robust; short tail characterized by having a bluntly rounded to hemispherical shape, dorsal side quite more convex than ventral side with rounded terminus; males characterized by large-sized spicules (average = 112.0 µm) and a large number of supplements, one pair of adanal and 18 or 19 mid-ventral supplements (Suppl. material 1: Fig. S1(1-9); Suppl. material 5: Table S3). Morphological and morphometrical traits of the topotype population from Dois Portos, Torres Vedras (Suppl. material 5: Table S3) agree very well with the original description (Bravo and Roca 1995). Morphometric measurements of adult specimens of the topotype population are coincident with those provided in the original description (Bravo and Roca 1995) Table S3), which may be due to intraspecific variability, as reported by Archidona-Yuste et al. (2016). Also, the topotype population shows similarity to four populations from Córdoba province, southern Spain (Archidona-Yuste et al. 2016); however, minor differences were detected in females such as L, a and c' ratios, lip region diameter length, odontostyle and odontophore lengths, distance from oral aperture to guiding ring, and, in males, spicule length. In addition, the topotype population agrees closely with the morphological features and morphometric measurements of all Portuguese populations examined (Suppl. material 5:  Table S3). Nevertheless, these differences further expand the intraspecific variation of the species and should be regarded as geographical intraspecific variation. According to the polytomous key by Chen et al. (1997) and its supplements (Loof and Chen 1999;Peneva et al. 2013), the topotypes and other studied Portuguese populations of this species have the following codes: A45, B45, C34, D2, E3, F45, G1, H1, I12, J?, K?. Unfortunately, we did not detect the first juvenile-stage. However, this stage was characterized in the original description (Bravo and Roca 1995) and later by Archidona-Yuste et al. (2016), who also characterized this species molecularly.
Molecular results and phylogenetic relationships of Longidorus bordonensis sp. nov. and other Longidorus spp.
Polymerase chain reaction (PCR) was used to amplify the D2-D3 expansion segments of 28S rRNA, ITS1 rRNA, and partial 18S rRNA from L. bordonensis sp. nov. and four other Longidorus spp. For each of the species studied, these three genes had an approximate size of 700-800, 900-1000, and 1600 bp, respectively, based on visualization of the band on the electrophoresis gel and the subsequent direct sequencing. , respectively, and differed in 27, 28-30, and 75 nucleotides, respectively. ITS1 sequence of L. bordonensis sp. nov. (MN150062) appropriately matched with other Longidorus spp. deposited in GenBank. This ITS1 sequence was 83-82, and 83% similar to L. carpetanenesis (MH429991-MH429993, Spain) and L. pini (MH430001, Spain), respectively. The variations among the ITS1 sequences of these species were from 143 to 157 nucleotides. The partial 18S rRNA gene sequences of L. bordonensis sp. nov. (MN129757) showed a high homology (more than 99% similarity) with two sequences deposited in GenBank belonging to L. carpetanensis (MH430006, Spain) and L. pini (MH430011, Spain). The variations among the 18S sequences of these species were from 8 to 15 nucleotides.
Using Bayesian inference (BI), we compared the phylogenetic position of L. bordonensis sp. nov. and other Longidorus spp. by using the D2-D3 expansion segments of 28S rRNA, the ITS1 region, and the partial 18S rRNA gene sequences (Figs 3-5). The BI tree (50% majority rule consensus tree) of the D2-D3 domains of 28S rRNA gene (Fig. 3) was based on a multiple-edited alignment (135 total sequences) of 722 total characters and revealed a major clade containing the majority of these species, including L. bordonensis sp. nov. and the remaining Iberian populations of Longidorus spp. (Fig. 3). The generated phylogenetic tree, using sequences of these D2-D3 fragments (Fig. 3)  Similarly, the BI tree (50% majority rule consensus tree) of a multiple-edited alignment, including 116 18S rRNA sequences and 1690 total characters (Fig. 4) and 116 ITS1 sequences and 570 total characters (Fig. 5), showed a topology similar to that of the D2-D3 fragments of the 28S gene. Both the partial 18S and ITS1 trees using BI (Figs 4, 5) Sturhan & Weischer, 1964 within the genera Longidorus and Paralongidorus. Bayesian 50% majority rule consensus trees as inferred from 18S rRNA sequences alignments under the SYM model. Posterior probabilities more than 70% are given for appropriate clades. Newly obtained sequences in this study are coloured in light blue. Scale bar: expected changes per site. and L. carpetanensis (18S, MH430006 Spain; ITS1, MH429991-MH429993, Spain). Both 18S and ITS1 trees showed a congruent position for all known species found in this study. For 18S and ITS1 trees, our L. vineacola sequences (18S, MN129758; ITS1, MN150064) were grouped in a well-supported clade also contain-  Sturhan & Weischer, 1964 within the genus Longidorus. Bayesian 50% majority rule consensus trees as inferred from ITS1 rRNA sequences alignments under the SYM model. Posterior probabilities more than 70% are given for appropriate clades. Newly obtained sequences in this study are coloured in green. Scale bar: expected changes per site.

Discussion
The main goal of our study was to identify and describe, morphologically and molecularly, Longidorus spp. parasitizing herbaceous and woody plants in vineyards and agro-forestry systems in Portugal. This was conducted in a nematological survey that included 150 sampling sites, with 43% of them in vineyards and the rest in agro-forestry areas. Eleven soil samples, each infested with only one needle nematode population, were selected for this study (Table 1). Our results confirmed the usefulness of developing an integrative approach based on the combination of morphometric and morphological characteristics and genotyping rRNA markers to correctly discriminate among Longidorus species. We described one new species (L. bordonensis sp. nov.) and identified several populations by integrating morphological analyses, morphometric measurements, and molecular data based on rRNA sequences to elucidate their phylogenetic relationships within Longidorus. New molecular markers were described for the new species, and the molecular diversity of three species (L. vinearum, L. vineacola, and L. wicuolea) was evaluated.
The comparative morphological taxonomic study of the 11 Portuguese populations of Longidorus spp. confirmed that the identification of species from phenotypic features including morphometric and morphometrical data is not easy due to inter-and intra-variability, overlapping of measurements and de Man ratios between species, and ambiguity caused by the presence of hundreds of species. As for previous biogeographic studies (Navas et al. 1993;Taylor and Brown 1997;, our study has revealed a new species (L. bordonensis sp. nov.), a first report of Longidorus spp. for Portugal (L. wicuolea), and new geographic records for other species, such as L. vinearum and L. vineacola. Our findings confirm that Longidorus is widespread in Southern Europe. This is in agreement with previous studies in Europe and around the Mediterranean Basin (Navas et al. 1993;Taylor and Brown 1997;, in which a dispersalist model was the main hypothesis to explain the large number of Longidorus spp. in Iberian Peninsula (Navas et al. 1993;Taylor and Brown 1997;Archidona-Yuste et al. 2019). However, Navas et al. (1993) proposed vicariant speciation as an alternative explanation for the current distribution of some Longidorus species; Navas et al. (1993) proposed that species survived the Pleistocene glaciation in refugia in the southern European peninsulas of Iberia, Italy, and the Balkans. We suggest that further molecular and phylogenetic studies are needed before assigning the correct model of speciation to explain the current biodiversity and distribution patterns of longidorids in Europe and the Mediterranean Basin. The biodiversity of Longidorus spp. in Portugal is considerable, with approximately 17 species reported (Bravo and Lemos 1997;Gutiérrez-Gutiérrez et al. 2016), including L. bordonensis sp. nov. and newly reported L. wicuolea. A major goal of our study was to generate DNA barcode markers that are useful tools for identifying new species and distinguishing among species within this genus. We update the geographical distribution and occurrence of Longidorus spp. in Portugal, which is important to the understanding of their current dispersion, the risks of spreading, the recognition of endemics and invasive species, and the diagnosis of quarantine pests.

Conclusion
Our work contributes greater understanding of the biodiversity within the genus Longidorus, describes Longidorus spp. by utilizing both morphological and molecular data, and establishes these species' phylogenetic relationships within the genus. Our study also establishes the value of using rRNA molecular markers, especially from topotype specimens, for the identification of Longidorus spp., when other methods are difficult and inconclusive. In addition, we establish molecular markers for precise and unequivocal diagnosis of a new species, L. bordonensis sp. nov., and show that molecular markers are useful to differentiate this species from other species that are virus vectors. Additionally, these markers were used to characterize the topotype of L. vinearum. To our knowledge, this is also the first time that L. wicuolea is reported for Portugal. ers to freely share, modify, and use this Dataset while maintaining this same freedom for others, provided that the original source and author (s)