Characterization of Hoplolaimus seinhorsti and Hoplolaimus pararobustus (Tylenchina: Hoplolaimidae) from banana, with phylogeny and species delineation in the genus Hoplolaimus

Abstract The morphological and molecular characterisations of two lance nematode species isolated from the rhizosphere of banana, Hoplolaimus seinhorsti and H. pararobustus, are provided based on an integrative study that includes light and scanning electron microscopy, phylogenetic analysis and two tree-based molecular species delimitation methods (GMYC and bPTP). Nineteen new sequences were obtained, including 5 partial 18S rRNA, 6 D2-D3 of 28S rRNA, 1 ITS rRNA and 7 COI mtDNA (the first COI sequences of H. seinhorsti and H. pararobustus), and an updated morphological character comparison of 37 Hoplolaimus species is presented. The tree-based molecular species-delimitation approaches employed gave markedly differing results, and also showed remarkable discrepancies among the investigated genes, although the bPTP output was found to agree well with established morphological species delimitations. Both species-delimitation approaches did, however, provide the same output for the COI mtDNA sequences, and the COI mtDNA gene sequence was also found to correspond better to established morphological species. It is therefore recommended by this paper as representing the most suitable barcode marker for Hoplolaimus species identification. This integrative study also resulted in the corrective reassignment of 17 gene sequences that were previously unidentified or incorrectly classified, as well as concluding that H. pararobustus consists of two cryptic species.

The genus Hoplolaimus Daday, 1905, subfamily Hoplolaiminae Filipjev, 1934and family Hoplolaimidae Filipjev, 1934 was established based on a single female of Hoplolaimus tylenchiformis Daday, 1905 from Paraguay (Sher, 1963). To date, 37 valid species within this genus have been described (Handoo and Golden, 1992;Tiwari et al., 2001;Ali et al., 2009;Nguyen et al., 2015;Ma et al., 2019;Ghaderi et al., 2020). The descriptions of these species are primarily based on morphology and morphometrics alone, as molecular data are unavailable for all but for ten of these Hoplolaimus species. Commonly referred to as lance nematodes, species of the genus Hoplolaimus are known for their well-developed and robust stylet with 'tulip-shaped' knobs. These plantparasitic nematodes represent an important group of semi-endo to endo-parasitic nematodes that cause considerable damage to the host plant's cortex and vascular tissue as well as inducing root necrosis. They are widely-distributed worldwide, and target a broad range of host plants including cotton, citrus, sugarcane, mango, tamarind, cowpea, baobab tree, banana, corn and soybean (Sher, 1963;Bridge, 1973;Robbins et al., 1987Robbins et al., , 1989Henn and Dunn, 1989;Koenning et al., 2004;Ahmadi et al., 2016;Holguin et al., 2015). The infective stages of Hoplolaimus species are from J2 to adult (Bridge and Starr, 2007). Some species reproduce sexually (amphimictic), while others reproduce asexually with females producing offspring without fertilization (parthenogenetic) (Bae et al., 2008).
The identification of Hoplolaimus species is facilitated by the key proposed by Handoo and Golden (1992), which was developed using information resulting from a comparative study of 29 valid Hoplolaimus species. The species-informative morphological characters include, among others, the number and the pattern of lateral field incisures, the number of labial annuli, the position of scutella along the body, the number of esophageal gland nuclei, and the hemizonid position with respect to the secretory-excretory (SE) pore. Ghaderi et al. (2020) recently published a comprehensive book on the systematics of the genus Hoplolaimus, in which they presented an updated key for Hoplolaimus species identification. This new key incorporates a broader range of informative morphological characteristics compared to the earlier key proposed by Handoo and Golden (1992). However, Bae et al. (2008Bae et al. ( , 2009 have previously indicated that a relatively large intraspecific variation vs minor interspecific variation of the diagnostic characters renders the identification of Hoplolaimus species difficult when based solely on morphology and morphometrics, leading them to develop molecular identification tools for some Hoplolaimus species from the USA based on duplex and multiplex PCR and PCR-RFLP. Prior to Bae's work on Hoplolaimus, several authors had already deposited sequences of genes from this genus in GenBank, including the partial 18S rRNA from H. galeatus (AY912054, AY912053) and H. columbus (AY912052, AY912051) by Powers, et al. (2008) and H. galeatus (AY146452) by Mullin et al. (2005). In addition, Subbotin et al. (2008) sequenced the D2-D3 of 28S rRNA from H. seinhorsti (DQ328752) and Chen et al. (2006) deposited partial 18S rRNA gene and complete ITS1-5.8S-ITS2 sequence with partial 28S rRNA from H. columbus (DQ309584) in GenBank. Additionally, Powers et al. (1997) reported that some congeneric Hoplolaimus species exhibit ITS size length variation and this variation can affect the accuracy of identifying Hoplolaimus species based solely on molecular methods that rely on this region.
The aims of this current study were to (1) characterise Hoplolaimus pararobustus and H. seinhorsti from the rhizosphere of banana (Musa spp.) in Nigeria and Indonesia, respectively, based on morphology (light microscopy and scanning electron microscopy) and molecular data (partial 18S rRNA, D2-D3 expansion segment of 28S rRNA, ITS rRNA and COI mtDNA sequences); (2) investigate the phylogenetic relationships of the Hoplolaimus species in combination with tree-based species-delimitation; and (3) update the existing morphological comparison of the 37 known Hoplolaimus species.
Morphological analysis: Morphological and morphometric characterization of the two nematode species was conducted based on fresh and fixed specimens. For the preparation of permanent slides, a small suspension of nematodes in an embryo dish were killed and fixed by adding a few drops of Trump's fixative (2% paraformaldehyde, 2.5% glutaraldehyde in 0.1M Sorenson buffer (Sodium phosphate buffer at pH = 7.5)). Subsequently, the embryo dish was heated in a microwave (700 Watts) for about 5 sec, left to rest for 1 h at room temperature followed by 24 h at 4°C to ensure maximum penetration of the fixative as described in Singh et al. (2018). Afterwards, the nematodes were gradually transferred to anhydrous glycerin for permanent slides following the protocol of Seinhorst (1959) and mounted on glass slides, for further morphological study. Nematodes were examined, photographed, and measured using an Olympus BX51 DIC Microscope (Olympus Optical, Tokyo, Japan) equipped with an Olympus C5060Wz camera. Scanning electron microscopy (SEM) was performed for H. seinhorsti specimens (n = 3) fixed in Trump's fixative, washed in 0.1M phosphate buffer (pH = 7.5), dehydrated in a graded series of ethanol solutions and critical-point-dried with liquid CO 2 . The specimens were mounted on stubs with carbon tabs (double conductive tapes), coated with gold of 25 nm and photographed with a JSM-840 EM (JEOL) at 12 kV (Singh et al., 2018). The H. pararobustus population was compared with lectotype and paralectotype material of the Ghent University Museum, Zoology Collections, Belgium (UGMD 100061-63).
Molecular analysis: Nematode morphological vouchers were prepared prior to DNA extraction. These vouchers were made of LM pictures of individual nematodes in temporary slides with distilled water. Each nematode was subsequently removed from the temporary mount and cut into pieces in distilled water using a blade and the pieces were transferred to a PCR tube with 20 µl of worm lysis buffer (50 mM KCl, 10 mM Tris at pH = 8.3, 2.5 mM MgCl 2 , 0.45% NP 40 (Tergitol Sigma), 0.45% Tween 20). The PCR tube was then incubated at -20°C (10 min) followed by adding 1µl proteinase K (1.2 mg/ ml), incubation at 65°C (1 h) and 95°C (10 min) and ending by centrifuging the mixture at 14000 rpm for 1 min (Singh et al., 2018). PCR amplification of partial ITS and 18S regions of rDNA was conducted using the primer pairs Vrain2F: 5´-CTT TGT ACA CAC CGC CCG TCG CT-3´ / Vrain2R: 5´-TTT CAC TCG CCG TTA CTA AGG GAA TC-3´ (Vrain et al., 1992) and SSU18A: 5´-AAA GAT TAA GCC ATG CAT G-3´ / SSU26R: 5´-CAT TCT TGG CAA ATG CTT TCG-3´ (Mayer et al., 2007) with thermal profile described in Singh et al. (2018Singh et al. ( , 2019. For amplification of the D2-D3 expansion segment of the 28S rDNA sequence, the primer pair 391: 5´-AGC GGA GGA AAA GAA ACT AA-3´ / 501: 5´-TCG GAA GGA ACC AGC TAC TA-3´ was used as described in Nadler et al. (2006) and for the amplification of the COI region of mtDNA, the primer pair JB3: 5´-TTT TTT GGG CAT CCT GAG GTT TAT-3´ / JB4.5: 5´-TAA AGA AAG AAC ATA ATG AAA ATG-3´ was used as described in Derycke et al. (2010). The PCR products were enzymatically cleaned with alkaline phosphatase (1 U/ml) and exonuclease I (20 U/ml) for 15 min at 37°C, followed by 15 min at 8°C and sent for sequencing at Macrogen (https:// dna.macrogen.com).
Phylogenetic and species delimitation analysis: The construction of a supermatrix for the phylogenetic analysis of Hoplolaimus species was not possible due to the limited availability of relevant sequence data on GenBank. At the time of writing, only a few species were associated with both nuclear and mitochondrial sequences. Therefore, each genetic marker was analysed separately.
The phylogenetic relationship of H. seinhorsti with other related species was analyzed based on the D2-D3 of 28S and ITS of rDNA and partial COI sequences of mtDNA, while that of H. pararobustus was analyzed based on the partial sequences of 28S and 18S of rDNA and partial COI sequences of mtDNA.
All sequences were analysed using a suite of programs implemented in Geneious 10.0.9 (https://www.geneious.com). The newly generated sequences were first subjected to a Basic Local Alignment Search Tool (BLAST) search against a closely related set of species on GenBank to identify and collect homologous sequences for multiple sequence alignment and phylogenetic analysis. Multiple sequence alignments were constructed using MUSCLE with default parameters. The poorly aligned regions were manually trimmed to obtain high-quality alignments for subsequent analysis. Bayesian inference was performed using MrBayes 3.2.6, with the general time reversible substitution model and estimation of invariant sites, assuming a gamma distribution with four categories gene (GTR + I + G) model. The analyses were run under 1 × 10 6 generations with four independent chains to ensure convergence and to obtain the posterior probabilities for the phylogenetic tree. Convergence of the runs was also checked using Tracer v1.7.2 (Rambaut and Drummond, 2010), and the effective sample size (ESS) values were well above 200 (>3000) for each run, indicating that the chains had converged and that the results were reliable. The Markov chains were sampled at every 100 generations, and 20% of the converged runs regarded as burn-in (Huelsenbeck and Ronquist, 2001).
Molecular species-delimitation of Hoplolaimus spp. in this study was performed using two tree-based methods, a Bayesian implementation of the Poisson tree processes (bPTP; Zhang et al., 2013) and the generalized mixed-yule coalescent (GMYC; Pons et al., 2006). For the bPTP approach, the phylogenetic trees created by MrBayes were uploaded to the online server of bPTP (http://species.h-its.org/ptp/) excluding outgroups, with default parameters. For the GMYC analysis, ultrametric trees were constructed using BEAST v1.10.4 (Drummond et al., 2012). Strict clock model with a lognormal distribution for the clock rate prior, a Speciation: Yule process for the tree prior, and a Hasegawa-Kishino-Yano (HKY) Substitution Model rate prior were used, and analyses were run for 1 × 10 7 generations, saving trees every 1 × 10 3 generations. The final trees were produced after removing 2000 samples (20%) as burn-ins, and the maximum clade credibility tree was calculated using TreeAnnotator 1.10.4. Finally, GMYC species delimitation was performed using a python re-implementation of the single threshold GMYC model in the GitHub repository (https://github.com/ iTaxoTools/GMYC-pyqt5), using TreeAnnotator trees as input.

Description
Females: Vermiform cylindrical body slightly tapering at both the ends. Body slightly curve to open C-shape after fixation. Head with prominent cephalic framework, hemispherical, four labial annuli and distinctly set-off from the body by a deep constriction. In SEM, head region divided into six equal sectors. Irregular longitudinal indentations or striae can be seen on the basal lip annule. Slightly raised ovoid oral disc with a central oral opening. Lateral sectors smaller than the sub-ventral and sub-dorsal sectors and visible amphidial apertures. Lateral field, around the mid-body, four to eight not well-delineated irregular incisures with breaks, and towards the anterior and the posterior regions reduced to one incisures. Stylet strong and large with prominent tulip-shaped knobs. Metacorpus rounded with sclerotized valve. Esophageal glands overlapping the intestine dorsally with five to six gland nuclei. SE-pore at isthmus level anterior to hemizonid and hemizonid about three cuticular annuli long. Two scutella, one anterior to the vulva (about 520 µm from the anterior end) and the other posterior to the vulva (about 1170 µm from the anterior end). Oval vulval opening around mid-body surrounded by unsculptured lips and vulva sometimes appears swollen in live specimens. Posterior epiptygma more conspicuous than the anterior epiptygma. Reproductive system didelphic amphidelphic with two equally developed outreached ovaries, spermathecae round to oval. Tail hemispherical to conoid-hemispherical, 13-16 annuli long.

Male: Not found
Hoplolaimus pararobustus (Schuurmans Stekhoven & Teunissen, 1938) Sher, 1963 Figure 3, Table 1 Description Females: Vermiform cylindrical body, 1161-1552 µm long, and near C-shape when heat relaxed. Head with prominent cephalic framework, hemispherical, four-five labial annuli and distinctly set-off from the body by a deep constriction. The lateral field is relatively inconspicuous under the light microscope with irregular incisures or broken lines, not well delineated around the mid-body and at the level of vulva and reduced to merely a single incisure towards the posterior part of the body. Stylet strong and large with prominent tulip-shaped knobs. Esophageal glands overlapping intestine dorsally with three gland nuclei. SE-pore above hemizonid and relatively opposite the median bulb. Hemizonid about two to three cuticular annuli long. Two scutella, one anterior to (about 410 µm from the anterior end) and the other posterior to the vulva (about 880 µm from the anterior end). Vulva at 54-70%, reproductive system didelphic amphidelphic with two equally developed outreached ovaries. Spermathecae round to oval with sperm. Tail short (15-30 µm) hemispherical, 13-16 annuli long.
Male: Similar to female except for reproductive structures with broad enveloping bursa, and body length generally shorter. Long and prominent spicule and gubernaculum with large and conspicuous bursa extending to the tail tip.

Hoplolaimus pararobustus
(Namibian specimens according to Marais et al., 2020) Hoplolaimus pararobustus (Syntypes; according to Sher, 1963) Anterior   (1992) Kolombia et al. (2017), which is only 0-2 nucleotides different and Hoplolaimus sp. (KY639326) is therefore likely also H. pararobustus. However, our sequences are 24-25 nucleotides different to the Namibian population of H. pararobustus (MT302643) previously characterized by Marais et al. (2020) (Fig. 4). The GMYC approach indicated that the four H. pararobustus sequences generated in this study belong to two putative species, Figure 4: BI phylogenetic tree inferred from analysis of the D2-D3 expansion segment of 28S rDNA sequences from nine known and three unknown Hoplolaimus species using the GTR + I + G model. Bayesian posterior probabilities are given next to each node and H. seinhorsti and H. pararobustus is presented in bold. Intraspecific variation of a clade indicated by a bar is given to the right of the gray bars, nucleotide differences between sister clades is provided left to the gray bars. Red and blue bars represent species boundaries estimated by GMYC and bPTP methods, respectively. A star indicates data with too many non-nucleotide characters in the sequences.
despite the fact that all four sequences were from the same population. The bPTP approach indicated the four H. pararobustus generated in this study together with Hoplolaimus sp. (KY639326) as one species, and the Namibian population of H. pararobustus (MT302643) as a different species. For clade II, H. stephanus (HQ678722, HQ678724, HQ678717, KX347887) forms a maximally supported clade with Hoplolaimus sp. (EU586798, EU586797, EU626794, EU626795) and it is also confirmed by bPTP that these sequences belong to a single species, i.e. H. stephanus. H. galeatus (KY849910) forms a maximally supported clade with H. magnistylus, also supported by bPTP as being two putative species, and thus one of these species is likely to be mislabeled (Table 4). It is clear that both species delimitation results (GMYC and bPTP) provide highly different outcomes with 27 vs 9 putative Hoplolaimus species respectively (Table 3).

COI mtDNA
The COI gene of mtDNA gene alignment ( The H. pararobustus sequences generated in this study form a maximally supported clade with Hoplolaimus sp. (KY639326) from Kolombia et al . (2017), which is 17-23 nucleotides different from the five H. pararobustus sequences obtained (Fig. 5). Remarkably, both speciesdelimitation results indicated the five COI sequences of H. pararobustus of Nigeria and Hoplolaimus sp. (KY639374) as three separated species.

ITS rDNA
The ITS rRNA gene alignment (1094 long) included 42 Hoplolaimus sequences and two outgroup species. One ITS rDNA sequence was obtained for H. seinhorsti from Indonesia (MK521872; 1017 bp long) while ITS sequences for H. pararobustus from Nigeria were not obtained. The phylogenetic tree resolved two major clades, and the H. seinhorsti (MK521872) sequence of the current study is within a maximally supported clade with H. columbus and other H. seinhorsti sequences from GenBank without internal resolution (Fig. 6) Figure 5: BI phylogenetic tree inferred from analysis of COI mtDNA sequences from seven known and one unknown Hoplolaimus species using GTR + I + G model. Bayesian posterior probabilities are given next to each node, H. seinhorsti and H. pararobustus are given in bold. Intraspecific variation of a clade indicated by a bar is given to the right of the gray bars, nucleotide differences between sister clades is provided left to the gray bars. Red and blue bars represent species boundaries estimated by GMYC and bPTP methods, respectively.
in GenBank (see Table 4

18S rDNA
The 18S rRNA gene alignment (923 bp long) included 13 Hoplolaimus sequences and two outgroup species. Five new 18S rDNA sequences Figure 6: BI phylogenetic tree inferred from analysis of ITS rDNA sequences from seven known and four unknown Hoplolaimus species using the GTR + I + G model. Bayesian posterior probabilities are given next to each node and H. seinhorsti is provided in bold. Intraspecific variation of a clade indicated by a bar is given to the right of the gray bars, nucleotide differences between sister clades is provided left to the gray bars. Red and blue bars represent species boundaries estimated by GMYC and bPTP methods, respectively.
Figure 7: BI phylogenetic tree inferred from analysis of 18S rDNA sequences from three known and two unknown Hoplolaimus species using GTR + I + G model. Bayesian posterior probabilities are given next to each node and H. pararobustus is given in bold. Intraspecific variation of a clade indicated by a bar is given to the right of the gray bars, nucleotide differences between sister clades is provided left to the gray bars. Red and blue bars represent species boundaries estimated by GMYC and bPTP methods, respectively.   Mukherjee et al., 1983;Larizza et al., 1998;Quénéhervé et al., 2006;Sikora et al., 2018) and H. pararobustus from Musa plant in several Asian and African countries, including Nigeria (Saeed et al., 1979;Coomans 1983;Fargette and Quénéhervé, 1988;Gowen and Quénéhervé, 1990;Liu and Feng 1995;Vovlas and Lamberti, 1985;Larizza et al., 1998;Speijer et al., 2001;Van den Berg et al., 2003;Loubama et al., 2007;Gaidashova et al., 2009). Other Hoplolaimus species have also been found associated with banana (Musa sp.), including H. bachlongviensis (Nguyen, Bui and Trinh, 2015) from Vietnam (Nguyen et al., 2015), H. columbus from Pakistan (Maqbool and Ghazala, 1988;Pathan et al., 2004), H. indicus from India and Iran (Maafi and Kheiri 1993;Sundaram 1997;Tilwari et al., 2000;Khan & Hasan, 2010), and other undescribed Hoplolaimus species (Khan 1999;Sawadogo et al., 2001;Cannayane et al., 2007;Adriano-Anaya et al., 2008). It is to be remembered that the results of using morphometrics in species-level identification of nematodes must be examined carefully, as morphometrics of nematodes in general can be influenced by several factors such as environment, host type, geographical origin etc. (Lax et al., 2004;Loubama et al., 2007). In the case of this study, H. seinhorsti and H. pararobustus could not be unequivocally differentiated by morphometric measurements, but instead with identification being based rather on the number and pattern of lateral incisures, number of labial annuli, number of esophageal gland nuclei, position of SE pore, the absence or presence of an intestinal post-rectal sac and the absence or presence of males. According to Fortuner (1991), the genus Hoplolaimus may be divided into two groups based on several phenotypic traits that are either ancestral or derived, including the number of esophageal gland nuclei (3 vs 6), number of lateral incisures (4 vs <4), position of SE pore (below the hemizonid vs above the hemizonid), and the presence of either regular or irregular striae on the basal lip annulus. However, this supposed division between ancestral and derived traits was not reflected in the phylogenetic results obtained in this study.
The D2-D3 of 28S rDNA distinguish between closely-related species in clade Ia (H. seinhorsti, H. columbus, H. indicus, H. dubius), a finding that agrees with previous observations (Bae et al., 2008;Ahmadi et al., 2016). For 18S rDNA, the data obtained are too limited to draw clear conclusions. While results based on ITS rDNA agreed fairly well with morphologically-based species delimitation, H. seinhorsti and H. columbus were not able to be resolved. The current study revealed that only the use of COI mtDNA supplied a means of resolving these species, and even then it was only for COI that both species-delimitation approaches provided the same output. It is therefore based on this evidence that we propose that the COI mtDNA as representing the most suitable barcode region for Hoplolaimus, in line with previous observations (Holguin et al., 2015;Ma et al., 2019;Shokoohi et al., 2022).
Remarkably, all phylogenetic and molecular species delineation results indicate that the H. pararobustus population from Nigeria and the H. pararobustus population from Namibia are not appointed as one single species. However, both populations/species do only differ in morphometrics from each other, the Namibian population has a shorter body length, stylet length and mid-body diameter compared to the Nigerian population. The morphometrics of the Nigerian population are more close to the syntypes according to Sher (1963) and no differences were observed between the H. pararobustus population from Nigeria and the type material. This might indicate that the Nigerian population is more likely represents the genuine H. pararobustus. However, given the large morphological variation (e.g. the lateral field of the lectotype material ranges from a clear singular line in combination with irregular lines to only unclear irregular lines), it is not obvious to separate the type population and the Nigerian population from the Namibian population on morphological grounds. Therefore, the H. pararobustus population from Nigeria and the H. pararobustus population from Namibia must at present be considered as cryptic species.
Cryptic species represent a significant component of biodiversity, and are an important factor in quarantine decisions and management strategies (Palomares-Rius et al., 2014). In such cases, if morphological data cannot give a conclusive answer, molecular data of the type specimens are needed, i.e. non-fixed topotype material of H. pararobustus (Kanyabayongo, Parc National Albert, Congo;Sher, 1963). This is the only way to conclude the determination of which Hoplolaimus population (Namibia vs Nigeria) represents the genuine H. pararobustus.
In spite of the increasing use of the coalescence models to study closely-related species that are difficult to differentiate using phenotypic characteristics, these models have only rarely been applied to plant-parasitic nematode investigations (Palomares-Rius et al., 2014;Singh et al., 2021;Nguyen et al., 2022). The present study has investigated putative species boundaries using coalescent-based approaches based on two different models (GMYC and bPTP) and four gene fragments (D2-D3 of 28S, ITS, 18S rRNA and COI). Results of these observations show remarkable discrepancies among the genes as well as compared to morphologically-established species (Table 3). Only the COI-based results provided identical species delimitation results for both approaches, which is in agreement with the findings of Singh et al. (2021). The GMYC approach revealed many more putative species, while the bPTP is more conservative and agrees better with established species delimitations ( Table 3). The GMYC algorithm is based on the time interval to the most recent common ancestor of species and an inherent assumption of monophyly, which is not always the case (Fujisawa et al., 2013), whereas the bPTP algorithm delimits species based on the number of nucleotide substitutions (Prevot et al., 2013). Furthermore, species delimitation methods that are based on single gene trees, for example for the bPTP and GMYC algorithms referred to herein, suffer from serious limitations due to gene tree/species tree incongruence (Zhang et al., 2013).
When gene tree topologies are incongruent with one another, it is difficult to determine whether this incongruence is due to incomplete lineage sorting, trans-species polymorphism, hybridisation, or introgression (Leliaert et al., 2014). Therefore, the simultaneous acquisition of several gene sequences will allow for a more precise and substantiated coalescence-based, multilocus species delimitation for plant-parasitic nematodes (Singh et al., 2021). A multilocus approach was not possible in the current study as, at the time of writing, data for only very few multi-loci species are available that have been obtained from the same population, and furthermore, very few such species are associated with both nuclear and mitochondrial sequences.
The findings of this work reinforce the proposals made by Singh et al. (2021) and Nguyen et al. (2022) concerning the need to unambiguously link comprehensive morphological data with both nuclear D2-D3 of 28S rRNA and mitochondrial COI gene sequences at the very least. This is clearly particularly necessary for certain species, a case in point being the genus Hoplolaimus, an important group of highly damaging plant-parasitic nematodes that were found to display remarkable molecular variations that render their identification especially challenging.