TARANTOBELUS JEFFDANIELSI N. SP. (PANAGROLAIMOMORPHA; PANAGROLAIMIDAE), A NEMATODE PARASITE OF TARANTULAS

Members of the family Panagrolaimidae consist mostly of bacteria-feeding nematodes that occupy a diversity of niches ranging from Antarctic and temperate soils to terrestrial mosses (Shannon et al., 2005). Panagrolaimids can be found across the globe in some of the world's most extreme conditions. Some species of the genus Panagrolaimus are capable of survival in Antarctica, where they can live for long periods in a state of anhydrobiosis, avoiding desiccation (Wharton, 2003; Shannon et al., 2005). However, not all members of the family are soil-dwelling and entirely free living.

Some members of Panagrolaimidae are parasites and infect living organisms such as mammals, insects, and arachnids (Webster and Lam, 1971; Nadler et al., 2003; Pizzi, 2009; Camerota et al., 2016; Abolafia and Peña-Santiago, 2018). However, none of the family members are considered to be obligate parasites. They have only been described as facultative or opportunistic parasites. A nematode parasite of tarantulas, Tarantobelus arachnicidaAbolafia and Peña-Santiago, 2018, was recently described. This is the only reported parasitic nematode of tarantulas, and it is the only species of its genus (Abolafia and Peña-Santiago, 2018; Wyrobisz-Papiewska et al., 2019). There have been other reports of Panagrolaimids infecting tarantulas; however, aside from T. arachnicida, there are few peer-reviewed studies on these reported cases (Pizzi, 2009; Pizzi et al., 2003). When T. arachnicida was described, it was placed into the family Brevibuccidae based on 28S rDNA sequencing. Members of the family Brevibuccidae are primarily found in the tunnels of bark beetles, or compost. However, T. arachnicida was found near the oral cavity of captive-bred tarantulas, an environment in which members of Brevibuccidae had not previously been reported. Symptoms of infection with T. arachnicida were reported as lethargy, anorexia, and tip-toe behavior, where the tarantulas stand abnormally high on their feet, keeping their abdomens off of the ground. All cases of infection resulted in death (Abolafia and Peña-Santiago, 2018; Wyrobisz-Papiewska et al., 2019).

In 2019, a wholesale tarantula breeder in the United States observed unusual behavior and symptoms with some Grammostola pulchra and Monocentropus balfouri spiders. Some individual spiders displayed a white area around the mouth, as well as an unusual lack of appetite, tip-toe behavior, and an apparent loss of pedipalp use. Using specimens provided by the breeder, we described a new species of nematode, based on morphological and morphometrical data; genetic and molecular phylogenetic analysis, including aspects of biology; and pathogenicity on insects and arachnids.

An infected G. pulchra was placed on a 1% agar plate and pressed lightly to dislodge nematodes adhering to the body. From this seed plate, individual nematodes were transferred to pure culture on 5 nematode growth media (NGM) plates (3 g NaCl, 2.5 g Peptone, 20 g agar, 10 ml Uracil [2 g/L], 975 ml DI water, autoclave, 25 ml 1 M KPO₄ [pH 6.0], 1 ml 1 M MgSO₄, 1 ml 1 M CaCl₂, and 1 ml cholesterol [5 mg/ml in ethanol]) and incubated at 17 C. After at least 2 generations, 1 nematode was picked from each plate for molecular identification. Nematodes for morphological and morphometrical studies were maintained on NGM seeded with Escherichia coli OP50; nematodes were transferred to fresh plates as needed.

Infective juveniles (IJs) for virulence assays were reared by placing 15-μl droplets of tap water on the lids of NGM plates seeded with E. coli OP50 and Tarantobelus jeffdanielsi n. sp. Two weeks after incubation at 17 C, IJs were collected from the underside of the petri dish lids.

Nematodes were preserved in hot formalin–glycerin (4:1) for light microscopy. Formalin-preserved specimens were processed to anhydrous glycerol (De Grisse and Lagasse, 1969) and mounted on glass slides. Videos and images of specimens were recorded on a Nikon-Eclipse E600 (Nikon, Melville, New York). Measurements were taken using Volocity image analysis software (Quorum Technologies, Puslinch, Ontario, Canada). Line drawings were created using Adobe Photoshop by tracing over images captured with a Nikon Eclipse E600 microscope (Nikon).

Specimens preserved in double-strength formalin–glycerin were selected for observation under SEM as described in Abolafia (2015). The nematodes were hydrated in distilled water, dehydrated in a graded ethanol–acetone series, critical-point dried, coated with gold, and observed with a 5-kV Zeiss Merlin microscope (Zeiss, White Plains, New York).

Individual nematodes were picked out and cut in a drop of worm lysis buffer (WLB: 50 mM KCl, 10 mM Tris-Cl pH 8.3, 2.5 mM MgCl₂, 0.45% NP40, and 0.45% Tween 20, as described in Williams et al., 1992), pipetted into a microcentrifuge tube with Proteinase K (60 μg ml⁻¹) and frozen at −80 C until ready for use. Genomic DNA was extracted from the frozen lysate by incubating the tubes at 64 C for 1 hr followed by enzyme deactivation at 90 C for 10 min and centrifugation at 13,226 g for 3 min. PCR was performed in a 25-μl reaction consisting of 4 μl genomic DNA template, 2.5 μl of 10X reaction buffer with 2 μl of 50 mM MgCl₂, 0.75 μl of 10 mM dNTP-mix, 0.4 μl of 25 μM primers D2Ab and D3B (De Ley et al., 1999) for D2–D3 expansion segments of the 28S rDNA gene; G18S4 and 18P; 988F and 1912R; and 1813F and 2646R for the 18S rDNA gene (Blaxter et al., 1998; Holterman et al., 2006), and 1 unit of EconoTaq® DNA Polymerase (Lucigen, Middleton, Wisconsin). The PCR conditions for amplification of both loci were initial denaturation at 94 C for 5 min, followed by 35 cycles of 94 C for 1 min, 55 C for 90 sec and 72 C for 2 min, and a final extension at 72 C for 10 min. PCR products were sized with a 1-Kbp DNA ladder (Promega) on a 1% agarose gel stained with 0.0003% ethidium bromide. PCR products were cleaned with QIAquick® PCR Purification Kit (Qiagen, Germantown, Maryland) following the manufacturer's protocol. Purified products were sequenced with primers G18S4, 18p, 22F, 13R, 4F, 4R, 988F, 1912R, 1813F, and 2646R (Table I; Blaxter et al., 1998; Meldal et al., 2007) for SSU, and D2Ab and D3B primers for 28S LSU rDNA. Sequencing was performed at the UCR Core Instrumentation Facility using a 96-capillary ABI 3730xl (ThermoFisher, Waltham, Massachusetts) following the manufacturer's protocol.

Sequences were assembled and compared (CodonCode Aligner 7.0.1) with published sequences in GenBank using BLAST search (Altschul et al., 1997). Sequences generated as part of this study were deposited in GenBank with the following accession numbers: T. arachnicida (MW559214) and T. jeffdanielsi (MW560268) for near full-length 18S, and T. jeffdanielsi (MW560627) for D2–D3 rDNA sequences.

The sequences of 18S and 28S DNA were aligned in ClustalX (Thompson et al., 1997) using default parameters.

The program IQ-Tree (Trifinopoulos et al., 2016) selected GTR + F + I + G4 as the best-fitting model for the 18S data set using the “find best model” function, and TIM3 + F + R2 was the best fit for 28S. The phylogenetic trees were constructed with a maximum-likelihood (ML) algorithm. The 18S phylogenetic tree included 61 aligned sequences, utilizing ultrafast bootstrap branch support with 1,000 replicates, a perturbation strength of 0.5, and 100 unsuccessful iterations to stop. No single branch tests were performed, and Plectus aquatilis (AF036602.1) was used as an outgroup. The 28S tree utilized the same specifications, with 23 sequences, and Plectus murrayi (LC457699.1) as an outgroup.

Fifteen individual Acheta domesticus were placed inside sterile 60 × 15 mm petri dishes previously lined with filter paper moistened with 200 μl of tap water. Crickets were provided 0.75 g of rabbit food pellets per petri dish. There were 2 inoculum levels (40 and 200 T. jeffdanielsi infective juveniles or IJs) per cricket/petri dish adjusted to a total volume of 1 ml and introduced using a pipette. As a control, crickets were exposed to 1 ml of tap water. Each petri dish lid was secured with a rubber band and using a hot needle, holes were burned into the lids. Plates were stored in the dark at about 21 C. Tap water (200 μl) was added to each petri dish 6 days after exposure to rehydrate the filter paper. Mortality and symptomatology of nematode infection were observed for 10 days. Each experiment was repeated 3 times.

Fifteen Galleria mellonella wax worms were collected and split into 3 equal groups. The assay was performed on G. mellonella wax worms secured from www.crittergrub.com. The same arenas were prepared as described previously except that there were 5 wax worms per petri dish; lids did not have holes, and the inoculum levels were higher: 200 and 1,000 T. jeffdanielsi IJs or an equivalent of 40 and 400 IJs/waxworm, respectively). For the untreated control, wax worms were moistened with 1 ml of tap water. Each petri dish was then secured with a rubber band, and stored in the dark at about 21 C. After 6 days, 200 μl of tap water was added to each petri dish to rehydrate the filter paper. Mortality and symptomatology were observed for 10 days. Each experiment had 3 replicates and was repeated 3 times. To fulfill Koch's postulate, nematodes were collected from the cadavers of wax worms via a modified white trap as described in Kaya and Stock (1997) for identification via Sanger sequencing of the D2–D3 regions of the 28S rDNA.

A total of 5 methods of inoculation with T. jeffdanielsi were tested. After 3 unsuccessful attempts (see Suppl. Methods and Figs. S1, S2), the following methods were used.

Twenty-five A. domesticus were exposed to T. jeffdanielsi via a modified version of the method described in Lu et al. (2017). Fifty A. domesticus were anesthetized via incubation at 4 C for 15 min. Twenty-five wells of 2 microtube storage boxes were lined with paper towels and the anesthetized crickets were split evenly and placed inside the lined wells of each microtube storage box. One thousand T. jeffdanielsi IJs were pipetted evenly into the lined wells of 1 microtube storage box in a total volume of 500 μl. An equal volume of 500 μl of tap water was added to each lined well of the other microtube storage box as a control. A transparent film was placed over all the lined wells of the microtube storage boxes, and a petri dish filled with water was placed on top of the film to weigh it down and prevent cricket escape (Suppl. Fig. S3). The boxes were then stored in the dark at about 21 C. After 24 hr, crickets were collected and observed for signs of infection. Out of the 25 crickets inoculated with T. jeffdanielsi, 5 died and had nematodes present on the body; 11 survived and had nematodes visible on the body, and 9 survived with no visible nematodes on the body. These successfully preinoculated crickets with nematodes visible on the body were further used in the tarantula virulence assay.

Ten G. pulchra spiderlings with an average weight of 1.642 g were split into 2 groups. In the first group, tarantulas were fed a single live uninfected cricket, and in the second group, tarantulas were fed a single live infected cricket. The tarantulas were kept in 11 × 6 × 4.5 cm containers lined with 7 g of sterile coconut husk substrate in an incubator with a 12-hr night/day cycle at 23 C and 26 C, respectively. The containers were misted with tap water daily, the tarantulas were fed a single live uninfected A. domesticus weekly, and mortality and signs of infection were recorded daily. After the experiment, nematodes from the cadavers of infected tarantulas were observed by microscopy to fulfill Koch's postulate. Signs of infection included visible nematodes on the body of the tarantula, anorexia, lethargy, tip-toe behavior, and itching. One week postinfection, it was decided that the remaining tarantulas should be used for a virulence assay utilizing dead infected crickets. To do this, the cricket infection assay previously described was repeated, and a single dead infected cricket was fed to each of the 3 remaining tarantulas. The tarantulas were then kept in the same incubator and were treated under the same conditions.

Three weeks before the life-cycle experiments, mixed stages of T. jeffdanielsi were transferred to nematode growth gelite (NGG) culture plates seeded with E. coli OP50 to remove the influence of maternal effects (Muschiol and Traunspurger, 2007). The preparation and ingredients of NGG were analogous to those of standard NGM, the only modification being the replacement of Bacto-agar by 1.5 g/L gellan gum, a bacterial exopolysaccharide, Gelrite (Merck & Co. Kelco Division, Kenilworth, New Jersey; Eyre and Caswell, 1991). Tarantobelus jeffdanielsi were transferred to seeded NGG weekly and stored at 17 C, and life-cycle experiments and manipulations were carried out at room temperature (20 ± 1 C).

Tarantobelus jeffdanielsi egg-laying was observed to be delayed to the extent that hatching occurs within the uterus, also known as “bagging” or matricidal hatching (Luc et al., 1979). The time of hatching, rather than the time of egg deposition, was used as an indicator of ontogenetic age (Dolgin et al., 2007). Thus, we began our experiment with cohorts of juveniles that had hatched within a narrow period (less than 3 hr). Adult T. jeffdanielsi were transferred to 10 μl NGG droplets and were observed hourly until eggs laid began hatching. All first-stage juveniles (J1s) that hatched within 3 hr after initial observation were collected (n = 36). The starting point of the experiment (age x = 0) was based on the average hatching time of the 3 hr-interval of observation. Each J1 was placed into a 10-μl NGG hanging drop on the lid of a 12-well multiwell plate. The drop consisted of washed E. coli OP50 resuspended in 10 μl NGG. The wells were filled with moistened Greiner cellulose paper (Greiner Bio-One, Monroe, North Carolina). Bacterial density was set to 5 × 10⁹ cells ml⁻¹ which is suggestive of optimal population growth (Muschiol and Traunspurger, 2007). Juveniles were transferred to a new drop every 24 hr until they died, while the previous drop was checked for any progeny. Progeny of previous drops were observed using a light microscope and counted twice. Counting occurred after the initial transfer, and 24 hr after transfer to ensure that all fertile eggs hatched. Nematodes were considered dead when they did not respond to physical touch with an eyelash pick and showed a lack of turgor.

All experiments were repeated 3 times except for the tarantula virulence assay, which was only performed once due to time constraints and tarantula availability. All graphs and statistics were completed using Graphpad Prism 9. For analysis of all virulence assays, 3 Mantel-Cox log-rank analyses were performed comparing each treatment to each other.

Family Panagrolaimidae Thorne, 1937 Subfamily Tarantobelinae Abolafia & Vecchi, 2021 Genus TarantobelusAbolafia & Peña-Santiago, 2018 Tarantobelus jeffdanielsin. sp. (Figs. 1–8)

Body straight to slightly curved ventrad when relaxed (Fig. 1B, E); 833–1,288 μm long. Anterior end and lip region low, rounded, continuous with body contour. Oral aperture opening triangular (Fig. 2B), surrounded by 6 lips each with small, slightly raised papilla and each ending in small flaps (Figs. 2A–C, 3B–D). Four cephalic papillae are also similar in appearance with labial papillae. Amphidial aperture oval, about 3 μm from cephalic papillae. Cuticle with fine (width) annules, lateral field bearing 3 transverse incisures (Fig. 2E).

The females have a typical panagrolaimid stoma (Figs. 1A–C, 3B–D), with a length of 18.1 ± 0.6 (16.5–19) μm [presented as mean ± standard deviation (range)] with stegostom comprising half of its length. Cheilostom and gymnostom comprise the other half, of equal length, the former lacking refringent rhabdia; gymnostom well developed, with distinct and refringent rhabdia; stegostom short, funnel-shaped, with poorly refringent rhabdia (Fig. 3B–D). Typical panagrolaimid pharynx with cylindrical muscular corpus, slightly widened posteriorly but with no distinct procorpus and metacorpus; 2.3–3.4 times longer than isthmus (Figs. 1A, B, E; 3E). Isthmus about as long as the spheroid basal bulb. Pharyngeo-intestinal valve distinct, conoid. Nerve ring and excretory pore located 60–80% and 70–90%, respectively, of neck length from the anterior end (Figs. 1A, 3G). Deirid inconspicuous, and where observed, located just posterior to the excretory pore (Fig. 2D), at level with the posterior part of the basal bulb, 90% of neck length. Vulva located 46–68% of body length; vagina transverse, oblique where the body curvature is prominent, occupying more than a third of vulva body width, and often seen with secretion/copulation plug (Figs. 1B, 2F, 3A). Reproductive system monodelphic–prodelphic, with outstretched ovaries, oftentimes extending past anal aperture position in mature females and in few specimens with tip reflexed anteriorly (Figs. 1B, D; 3A). Uterus long, tubular, with eggs almost as wide as long (28–46 × 27–50 μm) in various developmental stages and oftentimes with hatched juveniles typical of endotokia matricida; spermatheca distinct, elongate, dextral relative to the intestine (Figs. 1D, 3H). Postvulval sac short, 0.4–0.7 times the vulva body width (Figs. 1D, 3H). Rectum length 1–1.6 times anal body with, with 3 prominent rectal glands. Tail conical with acute tip; phasmid inconspicuous, located less than half (mean = 0.25; range 0.22–0.35 μm) of tail length (Fig. 3H, I). Morphometrics of the holotype (female), female and male paratypes are presented in Table II.

Similar body habitus to female when relaxed, but posterior prominently curved ventrad, appearing as J-shaped; 842–1,288 μm long. All other morphological features of the anterior region (stoma, pharynx) typical panagrolaimid. Reproductive system monorchic, its length occupying 60–70% of total body length; testis ventrad, anteriorly reflexed. Cloacal aperture with opposing papilla-like processes (Figs. 2J, 3J, K). Genital papillae 2/1 + 1 + 3 + p, arranged as follows (Fig. 4H, I, K): 2 latero-ventral, precloacal pairs (GP1, GP2); 1 midventral papilla just anterior to the cloaca (MP); 1 pair, lateral, ad-cloacal (GP3); about halfway to the tail, 2 ventro-lateral pairs (G4 and G7); and 2 dorso-lateral pairs (G5, G6), with the phasmid positioned in between them, very posterior to the tail. Spicules typical panagrolaimid, curved ventrad with rounded manubrium, short calamus, ventrally curved lamina with very slight dorsal hump. Gubernaculum thick, curved, with small rounded manubrium and fine acute tip. Tail 61–78 μm long, conical, wider at more than halfway its length, and strongly curved ventrally in its tip; with very posterior phasmids, between GP6 and GP7. Entire body of the male can be found in Figure S4.

Grammostola pulchra.

Oral cavity, sternum, labium, various regions of the exoskeleton.

Captive-bred tarantulas in Virginia Beach, Virginia.

Eight females (holotype and paratypes) and 8 males (paratypes) were deposited to the nematode collection at the University of California–Davis, Davis, California.

urn:lsid:zoobank.org:act:8DE3E8EB-9465-4436-B372-8E7BD63FBF7A.

Tarantobelus jeffdanielsi n. sp. is named after American actor Jeff Daniels, whose character in the 1990 film Arachnophobia kills the queen spider and saves the fictional town of Canaima from a deadly infestation of spiders.

The new species conforms with taxonomic characters typical of the Panagrolaimomorpha, Panagrolaimoidea, family Panagrolaimidae (Thorne, 1937), subfamily Tarantobelinae (Abolafia and Vecchi, 2021). The genus Tarantobelus was proposed with a single species, T. arachnicida, a nematode isolated from mouthparts of a young adult, male greenbottle blue tarantula, Chromatopelma cyaneopubescens (Strand, 1907; Theraphosidae) that came from Venezuela and was bred in captivity in Poland. This new species is the second described in this genus and is proposed based on combined morphospecies (morphological and morphometrical data) and phylogenetic species concepts. Additionally, some biological features including its virulence to A. domesticus and G. pulchra, and species fecundity and mortality were characterized.

Tarantobelus jeffdanielsi is characterized by a simple lip region, each ending in small flaps; papilliform sensilla; stoma stegostom comprising half of stoma length; cuticle without loose sheath; annules ≤1 μm wide, panagrolaimid pharynx with cylindrical muscular corpus, slightly widened posteriorly with no distinct metacorpus; monodelphic, prodelphic female reproductive system with axial spermatheca; postequatorial vulva, vulva–anus distance longer than its tail length; conical tail with acute tip; male spicules relatively broad with slightly rounded manubrium; papillae formula 2/1 + 1 + 3 + p; life cycle with dauer stage.

Tarantobelus jeffdanielsi closely resembles T. arachnicida morphologically and morphometrically, and a number of their other measurements are reflective of their main difference in size. Tarantobelus jeffdanielsi female and male are longer, with mean ± standard deviation (range) of 1,090.4 ± 127.3 (833–1,289) vs. 874.7 ± 50.4 (769–954) μm. In addition, the following are also longer: female ovary and uterus, rectum, tail, and testis. Interestingly, despite its longer body, vulva position is more anterior compared to T. arachnicida (46–68% vs. 72–77%, respectively). Other parameters that are shorter are: length of pharyngeal corpus (91.7 ± 4 [83–98.8 μm]) vs. 125.9 ± 4.9 (116–133 μm) and postvulval sac (7.7 ± 0.9 [6.1–9.8 μm] vs. 32.7 ± 5.8 [23–45 μm]).

The ranges of the lengths of stoma (female and male), spicules, and gubernaculum were similar for both species, suggesting that these taxonomic characters are relatively stable within the genus.

The near full-length sequences of T. arachnicida (MW559214) and T. jeffdanielsi (MW560268) 18S; and D2–D3 expansion segments of T. jeffdanielsi (MW560627) rDNA were obtained via Sanger sequencing. The nearly complete 18S sequences of both Tarantobelus species consisted of 1,736 base pairs (bp); and differed by 4 point mutations: 2 transversions and 2 insertions/deletions. However, between the 2 species, of 735 nucleotides, 28 point mutations consisting of 7 transitions, 11 transversions, and 10 insertions/deletions in the D2–D3 domains of the 28S were observed.

Figure 4 shows the evolutionary relationships of 61 different nematodes as inferred from 18S rDNA sequences. Tarantobelus jeffdanielsi and T. arachnicida belong to the family Panagrolaimidae, Infraorder Panagrolaimomorpha (De Ley and Blaxter, 2002, 2004) with the closest relative being the vinegar eel, Turbatrix aceti AF202165.2. Furthermore, its position is corroborated by its placement in the 28S tree as a sister clade to (Medibullinae + Tricephalobinae + Panagrolaiminae + Turbatricinae; Fig. 5) and confirming that it does not belong in the subfamily Brevibuccinae. Family Brevibuccidae Paramonov 1956 was considered a taxon of uncertain position (incertae sedis, De Ley and Blaxter, 2002). The 18S and 28S rDNA phylogenetic trees are not completely congruent, as the 28S rDNA tree shows Halicephalobus gingivalis and Panagrellus sp. as the closest relatives, whereas the 18S rDNA tree indicates that Turbatrix aceti is the closest relative.

As with Aryal et al. (2019), house crickets were in poor health for the virulence assays for unknown reasons. All crickets died within 10 days of the experiment regardless of treatment type. No significant differences in mortality were observed between control crickets A. domesticus and those exposed to T. jeffdanielsi IJs (Fig. 6A).

Virulence assays with waxworms showed a dose-dependent effect to exposure to T. jeffdanielsi IJs (Fig. 6B). Ten days after inoculation, wax worms showed a 47% average survival rate when exposed to 200 IJs, 49% when exposed to 40 IJs, and 78% when exposed to tap water (Fig. 6B), with significant differences among all treatments (P = 0.0008 and 0.0044, respectively). Infected waxworms died and showed visible melanization (Fig. S5). Nematodes recovered from waxworm cadavers were confirmed as T. jeffdanielsi by D2–D3 28S rDNA sequencing.

The tarantula virulence assay resulted in 75% mortality of G. pulchra after 94 days when fed dead crickets inoculated with 1,000 IJs, 20% mortality when fed live crickets inoculated with 1,000 IJs, and 0% mortality when fed live uninoculated crickets (Fig. 6C). There was a significant difference in survival between G. pulchra fed with dead, infected crickets and control (P = 0.0221). No significant differences were observed between G. pulchra fed with live, infected crickets and the control (P = 0.3173); and G. pulchra fed with live, infected crickets and G. pulchra fed with dead, infected crickets (P = 0.1877).

Infected tarantulas showed large masses of mixed-stage nematodes in the mouth and around the sternum (Fig. 7). The most affected areas of the tarantulas included the upper section of the sternum, the labium, and the mouth (Fig. 7). Individual nematodes were also observed on the leg hairs and head of the tarantula. Infected tarantulas showed all signs of infection including lethargy, anorexia, itching, and tiptoe behavior throughout the infection period.

Tarantobelus jeffdanielsi cultured at 20 C in 10 μl NGG droplet had an average life span of 11.2 days, where the longest-lived individual died on day 18 (Fig. 8A). Tarantobelus jeffdanielsi hermaphrodites began their reproductive period on day 3. A total of 474 offspring were produced among all 36 nematodes. The total fertility rate (TFR, i.e., the total number of offspring a hermaphrodite would have, on average, if individuals were to live to the maximum age) of a T. jeffdanielsi hermaphrodite was 158 progeny (Fig. 8B).

Based on a combination of morphological and morphometrical data, T. jeffdanielsi is distinguishable from the only other species in the genus, T. arachnicida. Tarantobelus jeffdanielsi is longer, and consequently, measurements of the female ovary, uterus, tail and male rectum, testis, and tail are longer. However, the vulva position is more anterior, and the distance between vulva and anus (VAD) is longer (range 218.8–397.3; mean = 304.9 ± 45 μm) than T. arachnicida (range 143–213; 168 ± 17.5 μm). Some structures are shorter in T. jeffdanielsi, namely, the pharyngeal corpus and postvulval sac. Additionally, T. jeffdanielsi has 3 longitudinal lines in the lateral field, whereas no lines were visible on T. arachnicida even with the SEM (Fig. 2D,E). The comparison of the morphological traits between T. arachnicida and T. jeffdanielsi alone is enough to distinguish T. jeffdanielsi as a new species.

The molecular work performed on T. jeffdanielsi and T. arachnicida reinforces the separation of the 2 species. Tarantobelus arachnicida was originally placed in the family Brevibuccidae (Paramonov 1956) based on morphology and D2–D3 28S rDNA gene phylogeny (Abolafia and Peña-Santiago, 2018). Traditionally, this family includes Brevibucca species whose vulva position are highly posterior and near the anus, with VAD shorter than the tail length (Goodey, 1935); and categorized under superfamily Panagrolaimoidea. This family was later considered “of an uncertain position” (incertae sedis) based on morphology (De Ley and Blaxter, 2002) and combined morphology and 18S phylogenies (De Ley and Blaxter, 2004). The addition of a new taxon and the use of 2 genetic loci (28S and 18S) for estimating phylogenies (maximum likelihood and Mr. Bayes on XSEDE version 3.2.7a) resulted in a different grouping of the clade; that is, T. arachnicida previously grouped with members of Brevibuccidae, whereas our analyses place Tarantobelus in its own clade within the family Panagrolaimidae, superfamily Panagrolaimoidea, and infraorder Panagrolaimomorpha, to the exclusion of the family Brevibuccidae (Figs. 4, 5). Tree topology for both ML and MB algorithms was congruent.

Measuring the fecundity and mortality of T. jeffdanielsi using solid medium cultures proved difficult for fecundity and mortality assays. As mentioned in Muschiol et al. (2009), solid medium offered poor visibility of eggs and juveniles, and variable food concentrations, which can affect the number of offspring produced, counted, and affect the life span of the nematode. Therefore, we used the hanging drop method of Muschiol et al. (2009) to address these concerns. However, multiple bacterial densities were not tested to establish what density creates an ideal environment for T. jeffdanielsi, and which other factors would affect fertility and life span.

Evaluations of Tarantobelus species as bona fide parasites had not been explored previously. Observations of the nematode presence in the oral cavity of tarantulas have been reported following the death of the spiders. However, experiments had not been performed while assessing host biology and fulfilling Koch's postulates (Berman, 2019). The origin of the infections in the tarantula breeding facility is unknown, but exposure to the nematodes by insects used to feed the tarantulas is a likely possibility. We have demonstrated that healthy, uninfected G. pulchra can be infected when fed live or dead insects that had been exposed to the nematode. Tarantobelus jeffdanielsi was also recovered after the virulence assays after causing significant mortality to the tarantulas, fulfilling Koch's postulates. We have also demonstrated that T. jeffdanielsi IJs are capable of infecting, killing, and reproducing in G. mellonella larvae. The potential of these nematodes to infect and reproduce using arachnids and insects as hosts, and the lack of obligate parasites among nematodes in Panagrolaimidae, suggests T. jeffdanielsi is a facultative, opportunistic parasite. The fact that the nematodes survive and reproduce on NGM culture plates seeded with E. coli is convenient for research, and their self-fertile hermaphroditic reproduction increases their potential as a model system. It should be noted that tarantulas were difficult to infect by the methods used in this study. Three different methods of inoculation were attempted before successful methods were discovered (Supplemental Methods). More host preference assays should be performed to determine the host range of T. jeffdanielsi and its parasitic relationship with the hosts. Host biology, along with nematode biology, should be measured within these assays to determine the nature of parasitism accurately.

Tarantulas exhibited multiple behaviors during infection with T. jeffdanielsi. Noted behaviors included itching, tip-toeing, anorexia, and lethargy. Tarantulas that were used as controls did not exhibit any of these pathologies. Symptoms of anorexia and lethargy proved difficult to measure with conclusive results because of typical behavior of tarantulas while they are preparing to molt. Symptoms of infection occurred as early as 33 days postexposure and continued until death.

Tarantobelus jeffdanielsi appeared in small numbers on the leg hairs of the tarantulas during the early stages of infection (20 days at the earliest), and in large numbers around the mouth and outline of the sternum and labium during late infection (44 days at the earliest). However, the vast majority of nematodes appeared in the substrate of the tarantula containers. Tarantula infection may have resulted from exposure to nematodes on the soil, rather than nematodes having been orally introduced to the tarantula via an infected cricket. Soil inoculation may have occurred during the initial placement of infected crickets into the container with the tarantulas. Exposing the infected crickets to the soil may have inoculated the soil with T. jeffdanielsi. After the initial introduction of a nematode-exposed cricket into the containers, tarantulas were fed 1 live, uninfected cricket each subsequent week. We observed that these crickets died and decomposed in a matter of 1–2 days in the chambers, whether the tarantulas interacted with them or not. Nematodes were observed in and around the cadavers of the crickets, indicating the possibility that the nematodes were responsible for the death and/or decomposition of the crickets. A single infected tarantula was dissected to determine which host tissues T. jeffdanielsi had infected. No nematodes were discovered within the oral cavity of the tarantula during dissection. All nematodes were only discovered externally, with a majority located near the mouth and lining of the sternum. This observation is similar to a previous report of nematode-infected spider dissections (Pizzi, 2009). Further experimentation is required to evaluate the mechanism by which T. jeffdanielsi infects G. pulchra and other hosts.

We are grateful to Tanya Stewart and Michelle Black from Fear Not Tarantulas (https://fearnottarantulas.com) for providing samples, tarantulas, and for introducing us to this issue and providing feedback. We thank Frank Marshall, Don Jakoby, Wesley Strick, Al Williams, Jeff Daniels, and everyone involved in the making of the film Arachnophobia, for inspiring our interest in arachnids. This research was supported by a U.S. Department of Agriculture National Institute of Agriculture Hatch project (accession 1011296), the California Association of Pest Control Advisors (CAPCA), Plant California Alliance, and the Research Support Plans PAIUJA 2019/2020: EI_RNM02_2019 and PAIUJA 2021/2022: EI_RNM02_2021, University of Jaén, Spain. SEM pictures were obtained with the assistance of technical staff (Amparo Martínez-Morales and Alba N. Ruiz Cuenca) and equipment of the Centro de Instrumentación Científico-Técnica (CICT) at the University of Jaén.