First Report and Molecular Variability of Belonolaimus longicaudatus Associated with Turfgrass in Maryland

During a turfgrass nematode survey, a composite soil sample was collected from an asymptomatic athletic field with “North Bridge” bermudagrass in Baltimore County, MD in July 2023. The sample was processed at the U.S. Department of Agriculture, Agricultural Research Service, Mycology and Nematology Genetic Diversity and Biology Laboratory in Beltsville, MD. Nematodes were extracted from soil using sugar flotation and centrifugation (Jenkins, 1964). Adult and juvenile Belonolaimus were initially observed with an inverted microscope at a density of four individuals/100 cm³ of soil. Morphometric measurements were performed on 10 females and 10 males.

Molecular confirmation was performed by sequencing two ribosomal markers that have robust representation in GenBank. The ribosomal ITS regions ITS1 and ITS2 and the D2-D3 expansion regions of 28S large subunit rRNA were obtained from five females. ITS-CL-F2 (5′-ATTACGTCCCTGCCCTTTGTA-3′)/D2AR (5′-ACTTTCCCTCACGGTACTTGT-3′) and 28S-CL-F1 (5′-ACTTTCCCTCACGGTACTTGT-3′)/D3B (5′-TCGGAAGGAACCAGCTACTA-3′) primer sets were used to obtain comprehensive coverage of target loci with cycling conditions as described previously (Carta and Li, 2018; 2019). Forward and reverse sequences for each marker were aligned and assembled into a consensus sequence and subsequently queried using the BLASTn tool from the National Center of Biotechnology Institute (ncbi.nlm.nih.gov) to determine species identity. Newly obtained sequences were submitted to GenBank under the accession numbers OR520202 and OR520203 for ITS and OR520268 and OR520269 for 28S.

Separate ITS and 28S rRNA sequence alignments were constructed with B. longicaudatus populations from GenBank in MAFFT within Geneious and trimmed with GBlocks to eliminate poorly aligned regions. Outgroups included Tylenchus annulatus (EF030983) for ITS and T. leviterminalis (KJ461550) for 28S. In order to align data from GenBank that include different parts of the ITS rRNA gene, 18S-ITS1-5.8S sequences and outgroup taxa were added to the longer 18S-ITS1-5.8S-ITS2-28S alignment as described in Mundo-Ocampo et al. (2017) by using the MAFFT-add option on the MAFFT v7.0 server (http://mafft.cbrc.jp/alignment/server) (Katoh and Standley, 2013). For both ITS and 28S alignments, the model of base substitution was selected using jModelTest 2.1.2 (Darriba et al., 2012). The Akaike-supported model, a general time-reversible model including among-site rate heterogeneity and estimates of invariant sites (GTR + G + I), was used in further analysis. Bayesian analyses were run using the MrBayes CIPRES plugin within Geneious. Bayesian inference (BI) analysis for each gene included a random starting tree and was run with four chains for 2 x 10⁶ generations. Two runs were performed for each analysis. Trees were sampled every 1,000th generation, with 25% of the results discarded as burnin. The remaining samples were used to generate a 50% majority rule consensus tree. Posterior probabilities (PP) are shown on appropriate clades. The alignments minus outgroup taxa and outlier B. longicaudatus sequences were also used to construct phylogenetic networks via statistical parsimony (SP) using the program PopART (http://popart.otago.ac.nz) (Leigh and Bryant, 2015).

Morphometrics and characteristics of adult specimens were consistent with B. longicaudatus (Tables 1,2; Fig. 1). Bodies of females and males were slender and cylindrical, lip region was rounded and offset, stylet was slim and long with rounded stylet knobs, and the esophagus overlapped the intestine. Females were didelphic with hemispherical tails (Fig. 1A). Males possessed curved spicules and a bursa that enveloped the tail end (Fig. 1B).

Figure 1.

Photomicrographs of Belonolaimus longicaudatus female (A) and male (B) specimens.

The MD population morphometric means were compared to morphometric means of subclades IIIB and IIIA from Habteweld et al. (2021) due to interpopulation variation that exists in morphology and morphometrics in B. longicaudatus. MD morphometric means were closest in value to the IIIB subclade means in stylet knob width, excretory pore distance from anterior, esophagus length, and stylet/tail length ratio, relative to the IIIA subclade (Table 1). Excretory pore distance from the anterior of the MD population was the only measurement that did not have overlapping ranges with the IIIA subclade selected characteristics. The morphometrics of subclades IIIB and IIIA presented in Table 2 were found to be statistically different between the two subclades by Habteweld et al. (2021). The subclades were defined by geographical region with subclade IIIB composed of populations from north and west Florida, Texas, and South Carolina and subclade IIIA made up of south Florida populations.

For 28S rRNA, sequences of 721 bp were obtained from five females, of which four were identical and one differed at 2 bp from the others. BLASTn of the single representative was 100% identical to B. longicaudatus sequences obtained from Florida (peanut; KF963100), Texas (soil; MZ045460), and several others; the other sequence representing four individuals was 99.7% identical to the same accessions (0.3%, 2-bp differences). For ITS rRNA, sequences 867 bp in length were obtained from two females, with 3-bp differences between them. BLASTn showed 99.4% identity (0.6%, 3- to 5-bp differences) with B. longicaudatus sequences obtained from Delaware (soybean; GQ896549) and Florida (citrus; DQ672364) and high similarity to several others.

Newly obtained 28S rRNA sequences were aligned with all known Belonolaimus spp. sequences from GenBank, with Tylenchorhynchus leviterminalis (KJ461550) used as the outgroup. The 28S phylogenetic tree inferred by Bayesian analysis (Fig. 2) showed that B. longicaudatus 28S sequences formed a single clade, with the Maryland population included in the largest subclade (Group A2) along with others from Florida (various hosts), Alabama and Georgia (bermudagrass), Delaware and Indiana (soybean), South Carolina (bermudagrass and centipedegrass), Mississippi (bermudagrass), and Texas (bermudagrass). Group A1 contained populations from Florida, California, South Carolina, and Texas. Group B was placed as a sister clade to Group A and was composed of south Florida populations from bermudagrass, bentgrass, seashore paspalum, citrus, corn, turf plugs, or soil. Outlier sequences DQ672357 (from Oklahoma turfgrass) appeared in Group C with B. maluceroi from Mexico (banyan tree), and DQ672358 (from Florida pine) appeared in a polyphyletic clade Group D with B. euthychilus and B. gracilis populations from Florida pine. Support for the separation of subclades A1 and A2 was low; however, the separation between Group A1/A2 and Group B was maximally supported (PP = 1).

Figure 2.

Molecular phylogeny of Belonolaimus species as inferred from Bayesian inference of the 28S D2-D3 expansion region of ribosomal RNA under the GTR + I + G model of nucleotide substitution. The 50% majority rule consensus tree from Bayesian inference is presented. Posterior probabilities are given for appropriate clades. Newly obtained sequences are in bold. Shaded regions A to D indicate corresponding groupings in Figure 4 of Mundo-Ocampo et al. (2017). Roman numerals indicate groupings shown in Figure 1 of Habteweld et al. (2021).

Newly obtained ITS rRNA sequences were aligned with Belonolaimus spp. sequences from GenBank, with T. annulatus (EF030983) as the outgroup. Some of the B. longicaudatus data from GenBank included only ITS1; thus, following the MAFFT-add alignment strategy employed by Mundo-Ocampo et al. (2017) was effective in preserving the main alignment while allowing distant sequences and outgroups to be included without introducing unnecessary gaps. To simplify the comparison with previous phylogenetic trees, we not only followed the clade designations of Mundo-Ocampo et al. (2017) with some modifications, but also referred to the Roman numeral clade assignments (IIIA and IIIB) in Figures 3 and 4 of Habteweld et al. (2021). The phylogenetic tree (Fig. 3) inferred by BI produced three major groupings and placed the Maryland population of B. longicaudatus within the largest subclade (Group A2; PP = 1) that included populations from Delaware and Indiana (found on soybean); Florida (from a variety of hosts); North Carolina (corn), Louisiana, Texas, and South Carolina (bermudagrass); Georgia (cotton or bentgrass); and Barbados (grasses). Group A1 populations came mainly from Florida and Texas, but also included others from California, Mississippi, and South Carolina. Together, A1 and A2 correspond to Group A as previously described (Mundo-Ocampo et al., 2017). Group B was dominated by Florida populations and one population from Texas. There was strong support for the separation of subclades A1 and A2 (PP = 0.91) and for the separation of Group A from Group B (PP = 0.96). There was also strong support for Group C that included B. maluceroi (isolated from Banyan tree in Mexico), an unnamed Belonolaimus sp. (KY272122 and KY272123), and apparent outlier populations B. longicaudatus from Texas (grass; DQ494800), Nebraska (corn; DQ494801), and South Carolina (U89696). The most divergent B. longicaudatus population was from Florida pine (DQ672384), appearing basal to clade D, which included B. gracilis and B. euthychilus.

Figure 3.

Molecular phylogeny of Belonolaimus species as inferred from Bayesian inference of the ITS1 and ITS2 region of ribosomal RNA under the GTR + I + G model of nucleotide substitution. The 50% majority rule consensus tree from Bayesian inference is presented. Posterior probabilities are given for appropriate clades. Newly obtained sequences are in bold. Shaded regions A to D indicate corresponding groupings in Figure 5 of Mundo-Ocampo et al. (2017). Roman numerals indicate groupings shown in Figure 2 of Habteweld et al. (2021). Asterisks indicate topotype populations.

Figure 4.

28S rRNA statistical parsimony network. The sequences of each species are marked by different color circles as indicated in the key. Circles divided into sections represent sequences of each species with the same haplotype, and their size is proportional to the number of these sequences in the samples. Numbers of nucleotide differences between the sequences are indicated on lines connecting the pies. Small black dots represent missing haplotypes. New sequences are indicated in bold font. Color shaded regions A1, A2, and B correspond to groupings in the 28S phylogenetic tree.

Statistical parsimony was applied to the sequence alignments to study the haplotype networks among populations of B. longicaudatus. The network for 28S sequences of B. longicaudatus is shown in Figure 4. The size of the circles at node points is proportional to the number of sequences that share a haplotype, and geographic locations are indicated by color. For 28S, three main haplotype groupings consistent with the 28S MB tree were observed. Group A1 contained 10 haplotypes, Group A2 contained 8 haplotypes, and Group B contained 9 haplotypes. The Maryland isolate appeared in Group A2 and shared the most dominant haplotype with isolates from Florida, Alabama, Delaware, Georgia, Indiana, South Carolina, and Texas. Group A2 had haplotypes from Florida, California, Mississippi, South Carolina, and Texas. Group B was dominated by south Florida populations. Two outliers, DQ672357 from Oklahoma (grass) and DQ672358 from Florida (pine), were distantly separated from Group B.

Sequences from the ITS parsimony network are shown in Figure 5. GenBank datasets that included only ITS1 were omitted from this analysis. Three main population groupings consistent with the ITS MB tree emerged from the analysis. For ITS, there were only a few haplotypes shared by multiple populations; for the most part, each sequence formed its own branch in the network. Group A1 (composed of 15 haplotypes) was further divided into two sub-branches, with one consisting of sequences from Florida, California, and South Carolina, and the other composed of sequences from Florida, Mississippi, and Texas. The second group, A2 (18 haplotypes), included the two Maryland B. longicaudatus sequences and others from diverse locations including Delaware, Florida, Georgia, Indiana, Louisiana, South Carolina, Texas, and the island of Barbados. Group A2 contained the most shared haplotypes. One Maryland sequence (OR520202) shared a haplotype with Texas and Florida isolates. Group B (10 haplotypes) was composed entirely of sequences from Florida isolates, similar to the phylogenetic tree.

Figure 5.

ITS statistical parsimony network. The sequences of each species are marked by different color circles as indicated in the key. Pies (circles) represent sequences of each species with the same haplotype, and their size is proportional to the number of these sequences in the samples. Numbers of nucleotide differences between the sequences are indicated on lines connecting the circles. Small black dots represent missing haplotypes. New sequences are indicated in bold font. Color shaded regions A1, A2, and B correspond to groupings in the ITS phylogenetic tree.