Morphostatic Speciation within the Dagger Nematode Xiphinema hispanum-Complex Species (Nematoda: Longidoridae)

Dagger nematodes of the genus Xiphinema include a remarkable group of invertebrates of the phylum Nematoda comprising ectoparasitic animals of many wild and cultivated plants. Damage is caused by direct feeding on root cells and by vectoring nepoviruses that cause diseases on several crops. Precise identification of Xiphinema species is critical for launching appropriate control measures. We deciphered the cryptic diversity of the Xiphinema hispanum-species complex applying integrative taxonomical approaches that allowed us to verify a paradigmatic example of the morphostatic speciation and the description of a new species, Xiphinema malaka sp. nov. Detailed morphological, morphometrical, multivariate and genetic studies were carried out, and mitochondrial and nuclear haploweb analyses were used for species delimitation of this group. The new species belongs to morphospecies Group 5 from the Xiphinema nonamericanum-group species. D2-D3, ITS1, partial 18S, and partial coxI regions were used for inferring the phylogenetic relationships of X. malaka sp. nov. with other species within the genus Xiphinema. Molecular analyses showed a clear species differentiation not paralleled in morphology and morphometry, reflecting a clear morphostatic speciation. These results support the hypothesis that the biodiversity of dagger nematodes in southern Europe is greater than previously assumed.


Introduction
Plant-parasitic nematodes (PPN) are characterized by the presence of a stylet used for root tissue penetration, comprise about 15% of the total number of nematode species currently known, of which over 4100 species have been identified as PPN [1,2]. Annual crop losses caused by PPN are estimated Plants 2020, 9, 1649 2 of 27 to be about 8-15% of total crop production worldwide [3,4]. Accurate identification of PPN is essential for the selection of appropriate control measures against plant pathogenic species, as well as for a reliable method allowing distinction between species under quarantine or regulatory strategies and a better understanding of their implications in pest control and soil ecology [5,6]. PPN species have been defined historically based on morphological characteristics [7,8]. However, the adoption of molecular techniques in nematode taxonomy has revealed unexpected genetic diversity within species throughout the phylum Nematoda [9]. This has been especially accurate for the family Longidoridae, a large group of ectoparasitic nematodes feeding from the root tip zone to the hairy root region, and characterized by a substantial intra and interspecific homogeneity of the morphometric characters used for species discrimination [1,6,10,11]. Use of molecular data in species identification of dagger and needle nematodes over the last three decades has indicated that many widespread species actually comprise multiple genetically divergent and morphologically similar cryptic species [6,[11][12][13]. Complexes of cryptic species often result from nonecological speciation in which diversification is not accompanied by apparent ecological or morphological separation in traditional quantitative traits [14].
The genus Xiphinema is one of the most diversified group species of longidorid nematodes with more than 280 valid species [5,6,11,15]. The ecological and phytopathological importance of this group of nematodes lies in its wide range of host plants and cosmopolitan distribution [5,11], but some species of this genus are vectors of several important plant viruses (genus Nepovirus, family Comoviridae) that cause significant damage to a wide range of crops [10,16]. Considering the great diversity of this group, the genus Xiphinema was divided into two different species groups [5,17,18]: (i) the Xiphinema americanum-group comprising a complex of about 60 species [15,17]; and (ii) the Xiphinema nonamericanum-group which comprises a complex of more than 220 species [5,6,19]. Later, this group was divided into eight morphospecies groups for helping identification [18]. However, some cryptic species and species complexes within Xiphinema have been recently revealed based on integrative taxonomical approaches, including morphometric multivariate methods, genetic analyses based on ribosomal and mitochondrial DNA (rDNA and mtDNA, respectively) and species delimitation (haplonet tools) [6,11,20,21]. A paradigmatic example of these species complexes comprises the Xiphinema hispanum-complex, viz. didelphic Xiphinema species from the Iberian Peninsula characterized by a rounded tail in females with or without an inconspicuous bulge projecting slightly ventrally and a uterus showing spiniform structures [22]. The cryptic diversity of this species complex has been deciphered by our team over the last ten years applying integrative taxonomical approaches that allowed us to verify these species as valid, and the recent description of a new species, X. subbaetense [11,20]. Recent studies on this species complex clearly separated three species (X. adenohystherum, X. hispanum and X. subbaetense) revealing high levels of genetic diversity within them that showed little morphological differentiation [11]. In new nematode surveys carried out in natural areas in the provinces of Málaga and Almería, Andalusia, southern Spain, we have detected nine unidentified Xiphinema isolates resembling X. hispanum-complex morphology. Detailed morphological and morphometrical observations using light microscopy indicated that these isolates appeared undistinguishable from X. hispanum complex species, a fact which prompted us to undertake comprehensive multivariate and genetic analyses, compared with previous reported data, to decipher this taxonomic conundrum.
Morphostatic evolution can be defined as genetic modifications, and even complete speciation events, which are not reflected in morphology, often being a result of nonadaptive radiation marked by the rapid proliferation of species without ecological differentiation [23,24]. Although no data have yet been specifically mentioned in Nematoda, morphostatic evolution seems not to be a rare phenomenon in longidorids based on the numerous complexes and cryptic species documented [6,[11][12][13]15,20,25]. In Longidoridae, it is very common that molecular divergences among species are not reflected in morphological or morphometric traits, which conforms a morphostatic model of evolution with numerous cryptic species within this group [6,11,13,15,20,21,25,26].
In this context, we investigated (1) the existence of a new cryptic species within the X. hispanum-complex confirming a morphostatic speciation in this group using an integrative species delineation approach based on multivariate morphometric analysis and haplonet mitochondrial and nuclear haploweb tools; (2) a new species of the genus Xiphinema (Xiphinema malaka sp. nov.) described through integrative methods based on the combination of morphological, morphometric and molecular data; and (3) phylogenetic analyses based on D2-D3 expansion domains of the 28S rRNA gene, ITS1, the partial 18S rRNA gene, and the partial mitochondrial coxI gene sequences to clarify the relationships of the new Xiphinema species.

Results
Species boundaries within the Xiphinema complex included in this research ( Figure 1) were based on the integrative application of morphological, morphometric and molecular methods to unravel potential cryptic species diversity (Table 1). Species delimitation was carried out using two independent approaches based on morphometric (multivariate analysis) and molecular data using ribosomal and mitochondrial sequences (haplonet). Multivariate morphometric and haplonet methods were performed on the nine studied isolates including previous isolates from the X. hispanum-complex to verify species identifications. The integration of this procedure with the analysis of nematode morphology allowed us to verify Xiphinema malaka sp. nov. as a valid new species within the X. hispanum cryptic complex. Additionally, we maintained a consensus approach for the different species delimitation methods, including concordant results in phylogenetic trees inferred from nuclear and mitochondrial markers and/or different morphological or morphometric characteristics.

Results
Species boundaries within the Xiphinema complex included in this research ( Figure 1) were based on the integrative application of morphological, morphometric and molecular methods to unravel potential cryptic species diversity (Table 1). Species delimitation was carried out using two independent approaches based on morphometric (multivariate analysis) and molecular data using ribosomal and mitochondrial sequences (haplonet). Multivariate morphometric and haplonet methods were performed on the nine studied isolates including previous isolates from the X. hispanum-complex to verify species identifications. The integration of this procedure with the analysis of nematode morphology allowed us to verify Xiphinema malaka sp. nov. as a valid new species within the X. hispanum cryptic complex. Additionally, we maintained a consensus approach for the different species delimitation methods, including concordant results in phylogenetic trees inferred from nuclear and mitochondrial markers and/or different morphological or morphometric characteristics.

Multivariate Morphometric Analysis
In principal component analysis (PCA), the first three components (sum of squares (SS) loadings > 1) accounted for 65.1% of the total variance in the morphometric characteristics of the X. hispanum-complex ( Table 2). The eigenvalues for each character were used to interpret the biological meaning of the factors. First, the principal component 1 (PC1) was mainly dominated by a stylet with a high positive correlation (eigenvalue = 0.523). PC2 was mainly dominated by high negative correlation for the vulva position (eigenvalue = −0.547) as well as a high positive correlation for the a ratio (eigenvalue = 0.482) ( Table 2). This component was, therefore, related with the overall nematode size and shape. Finally, PC3 was mainly dominated by the highest positive correlation found for the c' ratio and lower, but also high, positive correlation for the hyaline region length (eigenvalues = 0.774 and 0.458, respectively). This component was then related with tail shape. Overall, these results suggest that all of the extracted PCs were related to the overall size and shape of nematode isolates. The results of the PCA were represented graphically in Cartesian plots in which isolates of the X. hispanum-complex were projected on the plane of the xand y-axes, respectively, as pairwise combinations of components 1 to 3 ( Figure 2). In the graphic representation of the X. hispanum-complex, and with the exception of X. adenohystherum, we observed that the specimens of all species were projected showing an expanded distribution along the plane for all the projected combinations of the components. One reason might be the wide morphometric variation detected in these species (Tables 3 and 4) [6,11]. As a consequence, we did not detect a clear separation amongst species within the X. hispanum-complex, all the specimens being projected at random for all the projected combinations. These patterns suggest a clear example of morphostatic speciation within the X. hispanum-complex. However, it should be noted that when projected on the plane of the combinations of PC1-2 and PC2-3, almost all specimens of X. malaka sp. nov. and X. subbaetense were separated among them ( Figure 2). This graphical separation was shown by the projection of PC2 (dominated by the V and a ratios). This graphical separation is due to the variation found in the ratio a among these species, as pointed out below. A minimum spanning tree (MST) superimposed on the plot of the first three principal components showed the same patterns observed with PCA, that is, not clear separation amongst species within the X. hispanum-complex ( Figure 2). All isolates were molecularly identified and located at southern Spain. The c' ratio was excluded by the multicollinearity test and then, it was not included in the multivariate analysis for the Xiphinema hispanum-complex; b Morphological and diagnostic characters according to Jairajpuri and Ahmad [7] with some inclusions. a = body length/maximum body width; c' = tail length/body width at anus; d = anterior to guiding ring/body diam. at lip region; d' = body diameter at guiding ring/body diameter at lip region; Oa-gr = Oral aperture-guiding ring distance; V = (distance from anterior end to vulva/body length) × 100. with some inclusions. a = body length/maximum body width; c' = tail length/body width at anus; d = anterior to guiding ring/body diam. at lip region; d' = body diameter at guiding ring/body diameter at lip region; Oa-gr = Oral aperture-guiding ring distance; V = (distance from anterior end to vulva/body length) × 100.

Molecular Characterization
The amplification of D2-D3 expansion domains of 28S rRNA, ITS1 rRNA, the partial 18S rRNA, and partial coxI genes, yielded single fragments of ~900 bp, 1100 bp, 1800 bp, and 500 bp, respectively, based on gel electrophoresis. D2-D3 for X. malaka sp. nov. (MT584052-MT584085) showed a low However, in coxI haplonet ( Figure 3B), six different haplotypes of X. malaka sp. nov. were detected, three in SN and three in SA. One from SA shared the same haplotype with the Tabernas isolate, and this haplotype kept a far molecular distance with the other two haplotypes from SA. It was worth noting that the number of D2-D3 haplotypes of X. malaka sp. nov. was higher than coxI haplotypes (13 vs. 6), but there were more mutations between these coxI haplotypes than D2-D3 haplotypes ( Figure 3A,B). Besides, X. subbaetense also comprised more haplotypes in the D2-D3 haplonet than the coxI haplonet (11 vs. 2); the situation of X. hispanum, X. adenohystherum were the same as previously described by Cai et al. [11].

Phylogenetic Relationships
Phylogenetic relationships among Xiphinema species inferred from analyses of D2-D3 expansion domains of 28S rRNA, ITS1, the partial 18S rRNA and the partial coxI mtDNA gene sequences using BI are shown in Figures 3C, 4, 5 and 6, respectively. The phylogenetic trees generated with the nuclear and mitochondrial markers included 136, 49, 65 and 95 sequences with 747, 1106, 1547 and 372 positions in length, respectively ( Figures 3C, 4, 5 and 6). The D2-D3 tree of Xiphinema spp. showed a well-supported clade (PP = 1.00), including 10 species from morphospecies Groups 5 and 6, seven of them belonging to morphospecies Group 5 and three to Group 6, all of them reported from the Iberian peninsula, and included X. malaka sp. nov. (MT584052-MT584085). All other clades followed the same pattern as previous studies. Xiphinema malaka sp. nov. was phylogenetically related with X. hispanum, X. celtiense and X. cohni in a moderately supported clade (PP = 0.88), but clearly separate from all of them ( Figure 3C). Finally, the phylogenetic relationships of Xiphinema species inferred from analysis of partial coxI gene sequences showed several clades that were not well defined ( Figure 6). Xiphinema malaka sp. nov. (MT580263-MT580274) was phylogenetically related to X. hispanum-complex species in a low supported clade (PP = 0.65), but clearly separate from all of them ( Figure 6). Bayesian 50% majority-rule consensus trees as inferred from ITS1 sequence alignments under transition model with a proportion of invariable sites and a rate of variation across sites (TIM2 + I + G). Posterior probabilities more than 70% are given for appropriate clades. Newly obtained sequences in this study are in bold letters, and each colour is associated with each species of the complex. Bayesian 50% majority-rule consensus trees as inferred from ITS1 sequence alignments under transition model with a proportion of invariable sites and a rate of variation across sites (TIM2 + I + G). Posterior probabilities more than 70% are given for appropriate clades. Newly obtained sequences in this study are in bold letters, and each colour is associated with each species of the complex. Bayesian 50% majority-rule consensus trees as inferred from 18S sequence alignments under the GTR + I + G model. Posterior probabilities more than 70% are given for appropriate clades. Newly obtained sequences in this study are in bold letters. Figure 5. Phylogenetic relationships of Xiphinema malaka sp. nov. within the genus Xiphinema. Bayesian 50% majority-rule consensus trees as inferred from 18S sequence alignments under the GTR + I + G model. Posterior probabilities more than 70% are given for appropriate clades. Newly obtained sequences in this study are in bold letters.
Difficulties were experienced with alignment of the ITS1 sequences due to scarce similarity. Thus, only related sequences were used for phylogeny. The 50% majority rule consensus ITS1 BI tree showed several clades low to moderately supported ( Figure 4). Xiphinema malaka sp. nov. was phylogenetically related to X. adenohystherum and X. iznajarense in a moderately supported clade (PP = 0.92), but clearly separate from all of them ( Figure 4).
Finally, the phylogenetic relationships of Xiphinema species inferred from analysis of partial coxI gene sequences showed several clades that were not well defined ( Figure 6). Xiphinema malaka sp. nov. (MT580263-MT580274) was phylogenetically related to X. hispanum-complex species in a low supported clade (PP = 0.65), but clearly separate from all of them ( Figure 6).  Bayesian 50% majority-rule consensus trees as inferred from cytochrome c oxidase subunit I (coxI) mtDNA gene sequence alignments under the GTR + I + G model. Posterior probabilities more than 70% are given for appropriate clades. Newly obtained sequences in this study are in bold letters, and each colour is associated with each species of the complex.      Paratypes. Female and juvenile paratypes were collected from the same soil sample as the holotype (Table 3); mounted in pure glycerin and deposited in the Institute for Sustainable Agriculture (IAS) of the Spanish National Research Council (CSIC), Córdoba, Spain (Slide numbers X-SA3-03-X-SA3-08); one female at Istituto per la Protezione delle Piante (IPP) of Consiglio Nazionale delle Ricerche (C.N.R.), Sezione di Bari, Bari, Italy (X-SA3-011); one female at the USDA Nematode Collection (T-7474p).

Discussion
The primary objective of this study was to decipher the cryptic diversity of the X. hispanumcomplex by applying an integrative taxonomical approaches on several new unidentified Xiphinema isolates from Málaga and Almería provinces (southern Spain), appearing morphologically and morphometrically indistinguishable from this species complex. Multivariate morphometric analyses proved to be useful tools for species delimitation within the genera Longidorus and Xiphinema [11,15,19,28]. These data support that X. hispanum-complex species comprise a model example of morphostatic speciation (genetic modifications not reflected in morphology and morphometry) [23,24], since independent approaches based on molecular analyses using ribosomal and mitochondrial sequences (haploweb and haplonet) revealed high levels of genetic diversity within these species complexes which clearly separated X. malaka sp. nov. from all other X. hispanumcomplex species. These results, as well as those from previous studies, may suggest that X. hispanumcomplex species comprises a Xiphinema endemic lineage, with members morphologically and morphometrically very similar, that have diversified in the Iberian peninsula, since no other records on these species have been reported outside this area [20,22,29].
Phylogenetic analyses based on three rDNA molecular markers (D2-D3 expansion domains of 28S rRNA gene, ITS1 region and the partial 18S rRNA) resulted in a general consensus of species phylogenetic positions for the majority of them, and were generally congruent with those given by previous phylogenetic analysis [6,11,19,[30][31][32][33].
The results of this research support our hypothesis that biodiversity of Longidoridae in southern Spain is still not fully clarified and needs additional sampling efforts given the significant gaps in soil nematode biodiversity regarding the large number of undescribed species [34,35] and the hypothesis suggesting the Iberian Peninsula as a possible center of speciation for some groups of the family Longidoridae [6,15,36]. The recognition of this extraordinary cryptic diversity has a direct bearing on estimates of global nematode biodiversity and concepts of nematode biogeography. Regional endemicity in plant-parasitic nematodes has seldom been recognized and cosmopolitan distributions in nematodes, like other microscopic organisms, are reportedly common [37,38]. Additional material examined. Additional nematode isolates were studied and characterized from the rhizosphere of maritime pine, black pine, cork oak and yellow broom at several localities at Málaga and Almería provinces (Table 4). Morphometric measurements were taken for 62 individuals, 40 females, one male and 21 juveniles from J1 to J4 from several localities in Málaga province, Tables 3  and 4. Unfortunately, the scarce nematode isolate detected in the isolte of Tabernas (Almería) did not allow us to take measurements of adult females.
Etymology. The species epithet refers to the Phoenician word Malaka, the name of the province of Málaga where the species was found in several localities.

Diagnosis. Xiphinema malaka sp. nov.
Belongs to morphospecies Group 5 from the Xiphinema nonamericanum-group species [18]. It is characterized by a moderate long body (3.5-4.9 mm), assuming a J-shaped when heat-relaxed; lip region hemispherical, separate from the body contour by a depression, 14.0-15.0 µm wide; a relatively long odontostyle 131.0-148.5 µm; vulva located at 47.1-53.8% of body length; female reproductive system didelphic-amphidelphic having both branches about equally developed, pseudo Z-differentiation containing numerous small granular bodies, uterus tripartite with small crystalloid bodies and spines in low number and presence of prominent wrinkles in the uterine wall that may be confused with spiniform structures; female tail short convex-conoid on both sides, and bearing 3 caudal pores, ending in a rounded and broad terminus with a very small bulge at the end in some specimens; c' ratio (0.9-1.0); male rare one individual out of 75 females. Four developmental juvenile stages were identified, the 1st-stage juvenile with tail elongate-conoid with characteristic subdigitate rounded terminus (c' ratio 3.2-3.8). According to the polytomous key of Loof & Luc [18] and matrix codes sorted by Archidona-Yuste et al. [19], codes for the new species are (codes in parentheses are exceptions): A4-B23-C6-D6-E65-F4(5)-G3-H2-I3-J6-K2-L1. The DNA sequences of D2-D3 expansion domains of 28S, ITS1 rRNA, 18S rRNA, and partial coxI were deposited in GenBank under the accession numbers MT584052-MT584085, MT584088-MT584099, MT584086-MT584087 and MT580263-MT580274, respectively.
Male. Extremely rare, only one male individual out of 75 female specimens was found in one sample near the type locality. Morphologically similar to female except for genital system and secondary sexual features. Male genital tract diorchic with testes containing multiple rows of different stages of spermatogonia. Tail short, convex-conoid with a broadly rounded terminus and thickened outer cuticular layer. Adanal supplements paired, preceded anteriorly by a row of five irregularly spaced ventromedians supplements. Spicules paired, dorylaimoid, moderately long and slightly curved ventrally, approximately 2.5 times longer than tail length; lateral guiding pieces more or less straight or with curved proximal end.
Juveniles. Four developmental juvenile stages were detected and distinguished by relative body length, odontostyle and replacement odontostyle length. The 1st-stage juveniles were characterized by the replacement odontostyle inserted into odontophore base (Figure 8). In all other stages, the replacement odontostyle was posterior to the flanges of odontophore in its resting position. The correlation between body length, replacement and functional odontostyle of the type population is given in Figure 10. Lip region in all juvenile stages looks similar to that in females. Other morphological characters similar to female, except for their size and immature sexual characteristics (developing genital primordium 16.0-87.0 µm long). The first-stage juvenile was characterized by a tail elongate-conoid with characteristic subdigitate rounded terminus (c' ratio 3.2-3.8). Tail of other developmental stages becoming progressively shorter and wider after each moult (Figure 8).
Xiphinema malaka sp. nov. is morphometrically almost undistinguishable from X. subbaetense and X. hispanum, from the former can only be differentiated in females by a higher a ratio (65.6-99.

Discussion
The primary objective of this study was to decipher the cryptic diversity of the X. hispanum-complex by applying an integrative taxonomical approaches on several new unidentified Xiphinema isolates from Málaga and Almería provinces (southern Spain), appearing morphologically and morphometrically indistinguishable from this species complex. Multivariate morphometric analyses proved to be useful tools for species delimitation within the genera Longidorus and Xiphinema [11,15,19,28]. These data support that X. hispanum-complex species comprise a model example of morphostatic speciation (genetic modifications not reflected in morphology and morphometry) [23,24], since independent approaches based on molecular analyses using ribosomal and mitochondrial sequences (haploweb and haplonet) revealed high levels of genetic diversity within these species complexes which clearly separated X. malaka sp. nov. from all other X. hispanum-complex species. These results, as well as those from previous studies, may suggest that X. hispanum-complex species comprises a Xiphinema endemic lineage, with members morphologically and morphometrically very similar, that have diversified in the Iberian peninsula, since no other records on these species have been reported outside this area [20,22,29].
Phylogenetic analyses based on three rDNA molecular markers (D2-D3 expansion domains of 28S rRNA gene, ITS1 region and the partial 18S rRNA) resulted in a general consensus of species phylogenetic positions for the majority of them, and were generally congruent with those given by previous phylogenetic analysis [6,11,19,[30][31][32][33].
The results of this research support our hypothesis that biodiversity of Longidoridae in southern Spain is still not fully clarified and needs additional sampling efforts given the significant gaps in soil nematode biodiversity regarding the large number of undescribed species [34,35] and the hypothesis suggesting the Iberian Peninsula as a possible center of speciation for some groups of the family Longidoridae [6,15,36]. The recognition of this extraordinary cryptic diversity has a direct bearing on estimates of global nematode biodiversity and concepts of nematode biogeography. Regional endemicity in plant-parasitic nematodes has seldom been recognized and cosmopolitan distributions in nematodes, like other microscopic organisms, are reportedly common [37,38].
In summary, the present study confirmed the extraordinary cryptic diversity of X. hispanum-complex species in Andalusia and comprises a paradigmatic example of morphostatic speciation of dagger nematodes in southern Spain, which can be a potential explanation of the unusual high biodiversity within Longidoridae, considering Andalusia as a hot spot of biodiversity. However, additional similar intensive taxonomic studies are needed in other areas which can confirm this statement.

Nematode Isolates and Morphological Studies
No specific permits were required for the indicated fieldwork studies. The soil samples were obtained in public areas, forests and other natural areas and did not involve any endangered species or those protected in Spain, nor were the sites protected in any way.
A total of 62 individuals including 41 adults and 21 juvenile specimens from several localities in Málaga and Almería provinces (southern Spain) were used for morphological analyses (Table 1, Figure 1). Nematodes were surveyed during spring season in 2019 in natural ecosystems in Andalusia, southern Spain (Table 1). Soil samples were collected for nematode analysis with a shovel randomly selecting four to five cores at each point, and considering the upper 5-50 cm depth of soil that was close to the active plant root at each sampling spot. Nematodes were extracted from a 500-cm 3 sub-sample of soil by centrifugal flotation [39] and a modification of Cobb's decanting and sieving [40] methods. For morphometric studies, Xiphinema specimens were killed and fixed by a hot solution of 4% formalin + 1% glycerol, then processed in pure glycerin [41] as modified by De Grisse [42].
Specimens for light microscopy were killed by hot fixative using a solution of 4% formaldehyde +1% propionic acid and embedded in pure glycerine using Seinhorst's [41] method. The morphometric study of each nematode isolate included morphology-based diagnostic features in Xiphinema (i.e., de Man body ratios), lip region width, amphid shape, oral aperture-guiding-ring, odontostyle and odontophore length and female tail shape [7]. For line drawings of the new species, light micrographs were imported to CorelDraw ver. X7 and redrawn. The light micrographs and measurements of each nematode isolate, including important diagnostic characteristics (i.e., de Man indices, body length, odontostyle length, lip region, tail shape, amphid shape and oral aperture-guiding ring; [7]) were performed using a Leica DM6 (Wetzlar, Germany) compound microscope with a Leica DFC7000 T digital camera. For the line drawings of the new species, CorelDraw software version X7 (Corel Corporation, London, UK) was used to redraw according to the selected light micrographs.

DNA Extraction, Polymerase Chain Reaction (PCR) and Sequencing
For molecular analyses, in order to ensure the selected nematodes for extracting DNA were from the same species, two live nematodes from each sample were temporary mounted in a drop of 1M NaCl containing glass beads (to avoid nematode crushing/damaging specimens) to ensure specimens conformed to the unidentified isolates of Xiphinema. Thus, 34 individuals collected from several sampling points in Andalusia were molecularly analyzed ( Table 1). All necessary morphological and morphometric data, by taking pictures and measurements using the above camera-equipped microscope, were recorded. This was followed by DNA extraction from a single specimen and polymerase chain reaction (PCR) cycle conditions as previously described [6,15]. Several sets of primers were used for PCR. A partial region of the 28S rRNA gene including the expansion domains D2 and D3 (D2-D3) was amplified by using the primers D2A (5 -ACAAGTACCGTGAGGGAAAGTTG-3 ) and D3B (5 -TCGGAAGGAACCAGCTACTA-3 ) [43]. The Internal Transcribed Spacer region 1 (ITS1) separating the 18S rRNA gene from the 5.8S rRNA gene was amplified using forward primer 18S (5 -TTGATTACGTCCCTGCCCTTT-3 ) [44] and reverse primer rDNA1 5.8S (5 -ACGAGCCGAGTGATCCACCG-3 ) [45]. A partial sequence of the 18S rRNA gene (18S) was amplified as previously described [46] using primers 988F (5 -CTCAAAGATTAAGCCATGC-3 ), 1912R (5 -TTTACGGTCAGAACTAGGG-3 ), 1813F (5 -CTGCGTGAGAGGTGAAAT-3 ), and 2426R (5 -GCTACCTTGTTACGACTTTT -3 . Finally, the portion of the cytochrome c oxidase subunit I gene (coxI) was amplified using the primers COIF (5 -GATTTTTTGGKCATCCWGARG-3 ) and COIR (5 -CWACATAATAAGTATCATG-3 ) [47]. The newly obtained sequences were deposited in the GenBank database under accession numbers indicated in Table 1 and on the phylogenetic trees. PCR cycle conditions were one cycle of 94 • C for two min, followed by 35 cycles of 94 • C for 30 s, annealing temperature of 55 • C for 45 s, 72 • C for three min, and finally one cycle of 72 • C for 10 min. PCR products were purified after amplification using ExoSAP-IT (Affimetrix, USB products, High Wycombe, UK), quantified using a Nanodrop spectrophotometer (Nanodrop Technologies, Wilmington, DE, USA) and used for direct sequencing in both directions using the primers noted above. The resulting products were purified and run on a DNA multicapillary sequencer (Model 3130XL genetic analyser; Applied Biosystems, Foster City, CA, USA), using the BigDye Terminator Sequencing Kit v.3.1 (Applied Biosystems, Foster City, CA, USA), at the Stab Vida sequencing facilities (Caparica, Portugal). The newly obtained sequences were submitted to the GenBank database under accession numbers indicated in Table 1 and on the phylogenetic trees.

Species Delimitation via Multivariate Morphometric Analysis and Haplotype Networks Construction
The nine new Xiphinema isolates detected in this study were included in the X. hispanum-complex species group given the close relationships morphologically with X. hispanum as outlined above. An iterative analysis of morphometric and molecular data using two independent strategies of species delimitation was utilized to asses described and undescribed specimens and to determine species boundaries within this species complex.
Species delineation using morphometry was conducted with principal component analysis (PCA) in order to estimate the degree of association among species within the X. hispanum-complex [48]. PCA was based upon the following morphological characters: L (body length), the ratios a, c, c', d, d', V, odontostyle and odontophore length, lip region width and hyaline region length (Table 2) [6,7,13]. Prior to the statistical analysis, diagnostic characters were tested for collinearity [49]. We used the collinearity test based on the values of the variance inflation factor (VIF) method that iteratively excludes numeric covariates showing VIF values > 10 as suggested by Montgomery and Peck [50]. PCA was performed by a decomposition of the data matrix amongst isolates using the principal function implemented in the package psych [51]. Orthogonal varimax raw rotation was used to estimate the factor loadings. Only factors with sum of squares (SS) loadings > 1 were extracted. Finally, a minimum spanning tree (MST) based on the Euclidean distance was superimposed on the scatter plot of the X. malaka sp. nov.-specimens complex against the PCA axes. MST was performed using the ComputeMST function implemented in the package emstreeR [52]. All statistical analyses were performed using the R v. 3.5.1 freeware [53].
In order to detect distinct phylogenetic groups possibly representing separate species, haplotype networks (briefly, haplonet) were constructed to each of the two separate datasets, i.e., the D2-D3 and coxI. Alignments were converted to the NEXUS format using DnaSP V.6 [54]; TCS networks [55] were applied in the program PopART V.1.7 [56]. Illustrations of networks were prepared using the program Adobe illustrator to add connecting curves between the haplotypes found co-occurring in heterozygous individuals [57].

Phylogenetic Analysis
Sequenced genetic markers in the present study (after discarding primer sequences and ambiguously aligned regions), and several Xiphinema spp. sequences obtained of GenBank, were used for phylogenetic reconstruction (Table 1). Outgroup taxa for each dataset were selected based on previous published studies [6,11,30,45,58]. Multiple sequence alignments of the newly obtained and published sequences were made using the FFT-NS-2 algorithm of MAFFT v. 7.450 [59]. Sequence alignments were visualized using BioEdit [60] and edited by Gblocks ver. 0.91b [61] in the Castresana Laboratory server (http://molevol.cmima.csic.es/castresana/Gblocks_server.html) using options for a less stringent selection (minimum number of sequences for a conserved or a flanking position: 50% of the number of sequences + 1; maximum number of contiguous no conserved positions: 8; minimum length of a block: 5; allowed gap positions: with half).
Phylogenetic analyses of the sequence data sets were based on Bayesian inference (BI) using MRBAYES 3.2.7a [62]. The best-fit model of DNA evolution was calculated with the Akaike information criterion (AIC) of JMODELTEST v. 2.1.7 [63]. The best-fit model, the base frequency, the proportion of invariable sites and the gamma distribution shape parameters and substitution rates in the AIC were then used in phylogenetic analyses. BI analyses were performed under a general time reversible, with a proportion of invariable sites and a rate of variation across sites (GTR + I + G) model for D2-D3, the partial 18S rRNA, and the partial coxI gene, and under a transition model with a proportion of invariable sites and a rate of variation across sites (TIM2 +I + G). These BI analyses were run separately per dataset with four chains for 2 × 10 6 generations. The Markov chains were sampled at intervals of 100 generations. Two runs were conducted for each analysis. After discarding burn-in samples of 30% and evaluating convergence, the remaining samples were retained for more in-depth analyses. The topologies were used to generate a 50% majority-rule consensus tree. Posterior probabilities (PP) were given on appropriate clades. Trees from all analyses were visualized using FigTree software version v.1.42 [64].