To determine if P. pacificus and other species of the genus Pristionchus are associated with scarab beetles in the United States, we analyzed beetles from Massachusetts, New York, Ohio, Iowa, Kansas, Texas, and Nebraska. The scarab beetle fauna differs between North America and Europe. For example, the cockchafer Melolontha melolontha and the dung beetle Geotrupes stercorosus that host the predominant European Pristionchus species are not known from North America. However, the subfamily Melolonthinae, which includes the june beetles and chafers, contains more than 500 species in North America, some of which are serious pests 12. We used the same sampling strategies in Europe and the United States and collected scarab beetles from the US that are comparable with those obtained in our European study. However, we have not been able to sample identical species on both continents. In total, we investigated 1241 beetles of more than 15 genera yielding 285 Pristionchus isolates (Fig. 1, Table 1). SSU sequence analysis and mating experiments of the 285 isolates revealed that they fall into seven species, two of which were known to us (Fig. 2). We obtained two strains of P. entomophagus (both from beetles from Ohio), a hermaphroditic species that is one of the most common Pristionchus species in Europe 8. We also found P. pacificus, but in no more than three beetle samples, one from Ohio, one from Nebraska and one from Massachusetts. Surprisingly, the remaining 280 (98%) isolates represent five Pristionchus species, which had up to then neither been found by beetle sampling in Western Europe 8 nor by previous soil sampling in the United States 13.

Nematode species of the genus Pristionchus have little diagnostic morphological characters and usually show overlapping ranges of morphometric values 148. Prior to our own studies on P. pacificus, only one Pristionchus species, P. aerivorus, was described for North America by Cobb in 1916 915. All previously obtained strains of P. pacificus from North America result from soil samples taken by various researchers 813. It is important to note that all of these soil-derived P. pacificus strains represent random findings within a sampled locality. P. pacificus was a rare nematode in these soil samples and multiple isolates from the same soil sample are not available.

Morphometric analysis revealed that the most common species in our analysis is very similar to P. aerivorus (Table 2). Thus, we have classified this species as P. aerivorus. The four other species are described as novel species and we provide morphological measurements, molecular SSU sequence data, the results of mating experiments and species designations for all of them (Fig. 2, Table 2, see species diagnosis below). The reference strains of all species are available as living cultures and frozen stocks in our laboratory and can be provided to other researchers upon request. Also, type material has been delivered to museums. Together, these five species account for the majority of isolates from scarab beetles (Table 1). Specifically, we obtained P. aerivorus (85 isolates), P. pseudaerivorus n. sp. (71 isolates), P. marianneae n. sp. (45 isolates), P. pauli n. sp. (35 isolates) and P. americanus n. sp. (1 isolate) in the US scarab beetle survey. In the following, we provide the diagnoses for the four novel Pristionchus species:

Diagnostic character: The species is characterized by a 471 bp sequence of a 1 kb fragment of the small subunit ribosomal RNA gene (SSU) that was amplified by PCR with the primers shown in Materials and Methods. This SSU sequence is distinct from the SSU sequence of all other Pristionchus species, but is identical between all isolates of P. marianneae. The sequence is shown in Fig. 2.

GenBank accession code: DQ419901

A second diagnostic character is provided by mating experiments. P. marianneae males mate with females of other isolates but not with females/hermaphrodites of other species.

Morphological measurements: see Table 2

Type host and locality: On Popilia japonica (Coleoptera: Scarabaeidae) near Geneva, New York, United States of America.

Etymology: R. J. S and M. H. want to dedicate this species to their mothers.

Frozen strain number: RS5108

Holotype: One male permanent slide No: SMNK-NEMA-T0022 (Natural History Museum Karlsruhe, Germany)

Paratypes: One male and one female permanent slide No: SMNK-NEMA-T0023; One male and one female permanent slide No: SMNK-NEMA-T0024

Diagnostic character: The species is characterized by a 471 bp sequence of a 1 kb fragment of the small subunit ribosomal RNA gene (SSU) that was amplified by PCR with the primers shown in Materials and Methods. This SSU sequence is distinct from the SSU sequence of all other Pristionchus species, but is identical between all isolates of P. pauli. The sequence is shown in Fig. 2.

GenBank accession code: DQ419900

A second diagnostic character is provided by mating experiments. P. pauli males mate with females of other isolates but not with females/hermaphrodites of other species.

Morphological measurements: see Table 2

Type host and locality: On Lichnanthe vulpina (Coleoptera: Scarabaeidae) in Carver, Massachusetts, United States of America

Etymology: M. H. wants to dedicate this species to Dr. Paul S. Robbins, an outstanding beetle expert and true friend.

Frozen strain number: RS5130

Holotype: One male permanent slide No: SMNK-NEMA-T0025 (Natural History Museum Karlsruhe, Germany)

Paratypes: One male and one female permanent slide No: SMNK-NEMA-T0026; One male and one female permanent slide No: SMNK-NEMA-T0027

Diagnostic character: The species is characterized by a 471 bp sequence of a 1 kb fragment of the small subunit ribosomal RNA gene (SSU) that was amplified by PCR with the primers shown in Materials and Methods. This SSU sequence is distinct from the SSU sequence of all other Pristionchus species. The sequence is shown in Fig. 2.

GenBank accession code: DQ419904

P. americanus males do not mate successfully with females/hermaphrodites of other species.

Morphological measurements: see Table 2

Type host and locality: On Polyphylla fullo (Coleoptera: Scarabaeidae) in Centerville, Massachusetts, United States of America

Etymology: This species has been found only in the United States of America so far. Frozen strain number: RS5140

Holotype: One male permanent slide No: SMNK-NEMA-T0019 (Natural History Museum Karlsruhe, Germany)

Paratypes: One male and one female permanent slide No: SMNK-NEMA-T0020; One male and one female permanent slide No: SMNK-NEMA-T0021

Diagnostic character: The species is characterized by a 472 bp sequence of a 1 kb fragment of the small subunit ribosomal RNA gene (SSU) that was amplified by PCR with the primers shown in Materials and Methods. This SSU sequence is distinct from the SSU sequence of all other Pristionchus species, but is identical between all isolates of P. pseudaerivorus. The sequence is shown in Fig. 2.

GenBank accession code: DQ419902

Morphological measurements: see Table 2

Type host and locality: On Phyllophaga sp. (Coleoptera: Scarabaeidae) near Lincoln, Nebraska, United States of America.

Etymology: The species is very similar in respect to morphological measurements and molecular sequence to P. aerivorus. Mating experiments revealed that crosses between P. pseudaerivorus and P. aerivorus can result in sterile females.

Frozen strain number: RS5139

Holotype: One male permanent slide No: SMNK-NEMA-T0028 (Natural History Museum Karlsruhe, Germany)

Paratypes: One male and one female permanent slide No: SMNK-NEMA-T0029; One male and one female permanent slide No: SMNK-NEMA-T0030

The result of the US survey indicates a striking difference in the mode of reproduction of North American and European beetle associated Pristionchus species. P. aerivorus, P. pseudaerivorus n. sp., P. pauli n. sp., P. marianneae n. sp. and P. americanus n. sp. are gonochoristic (male/female) species and account for 98% of the isolates from scarab beetles (Fig. 3). In contrast, more than 60% of the isolates from scarab beetles in Western Europe belong to two hermaphroditic species, P. maupasi and P. entomophagus, respectively (Fig. 3) 8. Thus, North America and Europe differ in the Pristionchus species associated with scarab beetles and in the mode of reproduction of these nematodes.

To determine if the Pristionchus biogeography would match species phylogeny we incorporated the North American species into the phylogenetic framework of the genus Pristionchus with a representative of the closely related genus Koerneria as outgroup (Fig. 4). Interestingly, P. aerivorus, P. pseudaerivorus n. sp., P. pauli n. sp., P. marianneae n. sp. and P. americanus n. sp. form a single clade and are much more closely related to one another than the species found in Europe. P. aerivorus, P. pseudaerivorus n. sp. and P. americanus n. sp. are the most closely related species and differ in only one out of 471 base pairs in their SSU sequence (Fig. 2). Surprisingly, the cockchafer-associated hermaphroditic species P. maupasi from Europe belongs to the same clade and carries an SSU sequence identical to the P. aerivorus sequence (Figs. 2, 4). A more detailed analysis including several ribosomal protein encoding nuclear genes distinguishes the two species and indicates that P. aerivorus and P. pseudaerivorus n. sp. are the two most closely related species and that P. maupasi is the sister taxon to the P. aerivorus/P. pseudaerivorus n. sp. species pair (Fig. 5). Specifically, P. aerivorus and P. pseudaerivorus n. sp. differ in 19 nucleotides of the concatenated sequences of the rpl-26, rpl-28 and rps-14 genes. In contrast, P. maupasi and P. aerivorus differ in 31 nucleotides and P. maupasi and P. pseudaerivorus n. sp. differ in 33 nucleotides of the corresponding sequences (Fig. 5). P. americanus n. sp. and P. marianneae n. sp. show even more nucleotide differences to the other species (Fig. 5). Thus, the investigated genes allow us to clearly distinguish these four species.

The unusually high DNA sequence similarity between P. aerivorus and P. pseudaerivorus n. sp. suggests that they result from a recent speciation event. Therefore, these two species might still possess the potential to mate and form hybrid offspring. To obtain support for this hypothesis, we carried out cross-species mating experiments and found a strong case of Haldane's rule (Fig. 6). Haldane's rule states that in hybrids between diverging species the sterile absent or underrepresented offspring is of the heterogametic sex 11. We found F1 hybrids in 12 of 24 crosses between P. aerivorus and P. pseudaerivorus n. sp. and all of the F1 hybrid animals were phenotypically female (Fig. 6A). Specifically, F1 hybrids were obtained after reciprocal crosses of P. pseudaerivorus n. sp. from Texas with four strains of P. aerivorus. In contrast, P. pseudaerivorus n. sp. from Nebraska produced F1 hybrids only when males were used for mating with P. aerivorus females. Another P. pseudaerivorus n. sp. strain from Lincoln, Nebraska (LNE) did not produce F1 hybrids with any of the P. aerivorus strains. Thus, the occurrence of species hybrids is highly strain-specific and does not correlate with the geographic origin of the strains used in the analysis.

The primary signal in P. pacificus and C. elegans sex determination is the ratio of the set of X chromosomes to autosomes 1. Animals with two X chromosomes develop as females (XX), whereas animals with a single X chromosome develop as males (XO). Theoretically, the absence of males among the F1 hybrid animals could result from the death of the putative XO animals or the sexual transformation of these animals into females. The latter scenario was observed in crosses of Caenorhabditis remanei with Caenorhabditis briggsae 16.

Additional support for the results of the laboratory crosses came from genotyped Pristionchus nematodes from wild scarab beetles. We found two cases of hybrids between P. aerivorus and P. pseudaerivorus n. sp. emerging from individual beetles (Fig. 6B). Genotyping of other Pristionchus individuals from the same two beetles revealed that they carried both P. aerivorus and P. pseudaerivorus n. sp., suggesting that the observed hybrids are the result of cross-species fertilization (Fig. 6B). However, both hybrid animals were isolated more than a week after the beetle was placed on the petri dish. Therefore, it remains unknown if the observed hybrids came from the generation that lived on the beetle or whether it originated from fertilization in the laboratory. Taken together, P. aerivorus and P. pseudaerivorus n. sp. co-exist on the same beetles in nature, they can form hybrids and laboratory-derived F1 hybrids follow Haldane's rule. In addition, both species have a similar geographic distribution in our sampling.

Finally, we wanted to know if Pristionchus nematodes occur in association with the CPB Leptinotarsa decemlineata in the United States. Our analysis of CPB in Europe had indicated that one species, P. uniformis, infests these beetles with high specificity 8. The CPB is a pest species that originated in Mexico, where it extended its range of host plants from the wild solanaceae buffalo bur (Solanum rostratum) to include the potato (Solanum tuberosum). It migrated further north to the United States and was first described in Colorado in 1811. Spreading towards the East it reached the Atlantic coast in 1874 and was first reported in Europe in 1877 1718.

We collected CPB from two different locations in Ohio and New York State and found Pristionchus nematodes with an infestation rate of up to 60% (Table 1). Sequence analysis and mating experiments revealed that all of these isolates belong to P. uniformis, the same species that is found on European CPB. Thus, in contrast to scarab beetles, the CPB hosts the same (and only one) Pristionchus species on both continents, P. uniformis. Given the recent invasion of the CPB to Europe we speculate that P. uniformis was introduced to Europe together with its beetle vector, the CPB, at the end of the 19th century.

We collected different beetles at adult stage using sweeping nets, blacklight traps and pitfall traps baited with dung. The beetles were transferred to the lab alive, sacrificed by cutting them in half transversely, and put on NGM agar plates (6 cm diameter) seeded with 300 μl of the slowly growing E. coli strain OP50. The plates were checked daily using a Zeiss Stemi 2000 dissecting scope over a period of one to three weeks for emerging and reproducing nematodes. From the emerging nematodes we produced isogenic lines by transferring single gravid females or hermaphrodites to new plates. To see whether the isolated nematodes belonged to gonochoristic or hermaphroditic species, virgin larvae were singled out onto plates. The presence of offspring indicated that they represent a hermaphroditic species.

Emerging nematodes were determined to family level with a Zeiss Stemi 2000 dissecting scope and to genus level with a Zeiss Axioplan 2 microscope using the key by Sudhaus and Fürst von Lieven 9. For determination several worms were transferred onto microscopic slides covered with a 0.5 mm thick layer of 5% agar and either immobilized by heating the slide over an open flame to about 60°C for a few seconds or anesthesized with sodium azide. Within the genus Pristionchus many species can not be identified by morphological methods. We therefore chose to use molecular tools and mating experiments with reference strains to distinguish the different species.

We have originally isolated up to three isogenic female lines per beetle and have processed them for sequence analysis. In nearly all cases, these isogenic female lines had identical SSU sequences. The exceptions were the observed hybrids between P. aerivorus and P. pseudaerivorus n. sp. described above, and two cases where isogenic female lines of P. aerivorus and P. pseudaerivorus or P. aerivorus and P. marianneae were derived from the same beetle. When the SSU sequences of different isogenic female lines from the same beetle were identical, only one isolate was further considered. Throughout the text (and in the Tables and Figures), we consider only one of these identical isolates per beetle. In particular, the numbers and frequencies provided in Table 1 and Figure 3 consider only one of these isolates. In Table 1, beetle individuals from the different sampling sites are pooled. There was little variation in incidence of Pristionchus species on beetles among sampling sites.

Molecular species identification was done using the small subunit rRNA gene (SSU) as described before 8. In short, genomic DNA from single nematodes was prepared using the NaOH digestion procedure described by Floyd et al. 23. A single worm was transferred to 20 μl of 0.25 M NaOH, incubated overnight at 25°C and heated to 99°C for 3 min before 4 μl of 1M HC1, 10 μl of 0.5 M Tris-HCl (pH 8.0) and 5 μl of 2% Triton X-100 were added. The mixture was heated to 99°C for 3 min, frozen at -20°C and reheated at 99°C for further 3 min. Two microliters of this extract were used for subsequent polymerase chain reaction (PCR).

A 1 kb fragment of the SSU was amplified by PCR using the primers SSU18A (5'-AAAGATTAAGCCATGCATG-3') and SSU26R (5'- CATTCTTGGCAAATGCTTTCG-3') 2324. The reactions were performed in 25 μl of 1× PCR buffer (Amersham Biosciences, Freiburg, Germany) containing 2.5 mM of MgCl₂, 0.16 mM of each deoxynucleoside triphosphate, 0.5 μM of each primer, 2 μl of the lysate, and 2 units of Taq DNA polymerase (Amersham). The reactions were started by initial denaturation at 95°C for 2 min in a PTC-200 (MJ Research, Biozym, Hess. Oldendorf, Germany) or T gradient (Biometra, Göttingen, Germany) thermocycler, followed by 35 to 40 cycles of denaturation at 95°C for 15 sec, primer annealing at 50°C for 15 sec, and extension at 72°C for 2 min. A final incubation step at 72°C for 7 min concluded the reaction. PCR products were purified by the Qiagen PCR product gel extraction kit (Qiagen, Hilden, Germany). Approximately 500 bp of the 5'-terminal end were sequenced using the primer SSU9R (5'-AGCTGGAATTACCGCGGCTG-3') and the Big Dye terminator protocol (Applied Biosystems, Darmstadt, Germany).

Sequences were aligned manually using the Seqpup 0.6f software for Macintosh 25. The substitution model for the reconstruction of phylogenetic relationships was selected by the hierarchical likelihood ratio test as implemented in the Modeltest 3.7 software 26. The selected substitution model corresponds to the Kimura 2-parameter model 27 with Ts/Tv = 1.4727, equal base frequencies, and a γ-correction with shape parameter α = 0.2381.

Phylogenetic trees were determined using the heuristic search algorithm under the maximum likehood (ML) criterion using PAUP*4.0bl0 program 28. Trees were rooted by the Koerneria sp. sequence as outgroup. Neighbour joining (NJ, 29) and maximum parsimony (MP) trees were drawn by the same program. Alignment gaps were eliminated from the analysis. The topological stability of the trees was assessed by 1000 bootstrap replications 30.

To confirm the species identification by the molecular sequence of a novel isolate we performed mating experiments with the reference strain of the respective species (see below for definitions). Five virgin females were put on a plate with a small spot of OP50 together with five males of the reference strain of a certain species. On a second plate we picked the opposite sexes of the two strains to test for reciprocity. If there was no offspring after one week, the experiments were repeated two more times. If fertile offspring occurred we considered the two strains to belong to the same species.

We use the following definitions to distinguish "isolates" and "strains". An isolate is an isogenic female line, which is derived from a beetle sample and subjected to molecular and experimental analysis. After species identification we established one isolate per species and location as a strain. The strains are permanently cultured in the lab, have a strain number and are also kept as frozen stocks. For each new species designated by molecular sequence analysis and mating experiments, one strain was defined as a reference strain (see below).

According to the designation of reference strains, the 285 isolates obtained from American scarab beetle material fell into seven Pristionchus species, two of which were already cultured in our lab (P. pacificus, reference strain PS312 and P. entomophagus ref. strain RS0144). The other five could not be identified. We found the morphometric data of the most common of the unidentified species to coincide with the description of the only North American Pristionchus species listed in the catalog provided by Sudhaus and Fürst von Lieven 9, namely P. aerivorus. The other four species are novel and are described based on morphometric analysis, SSU sequence analysis and mating experiments (Fig. 2, Table 2 and species diagnoses). In total, the following eight Pristionchus species were obtained from scarab beetles (seven species) and the Colorado potato beetle (one species) in North America:

Pristionchus entomophagus (Steiner, 1929) (ref. str. RS0144)

Pristionchus pacificus Sommer, Carta, Kim & Sternberg, 1996 (ref. str. PS312)

Pristionchus uniformis Fedorko & Stanuszek, 1971 (ref. str. RS0141)

Pristionchus aerivorus (Cobb in Merrill & Ford, 1916) (ref. str. RS5106)

Pristionchus pseudaerivorus n. sp. Herrmann, Mayer & Sommer 2006 (ref. str. RS5139)

Pristionchus marianneae n. sp. Herrmann, Mayer & Sommer 2006 (ref. str. RS5108)

Pristionchus pauli n. sp. Herrmann, Mayer & Sommer 2006 (ref. str. RS5130)

Pristionchus americanus n. sp. Herrmann, Mayer & Sommer 2006 (ref. str. RS5140)

In order to obtain nuclear protein coding genes to distinguish P. americanus n. sp. from its sister species P. aerivorus, P. pseudaerivorus n. sp. and P. maupasi we isolated total RNA from 50 to 100 μg of nematodes using the TRIZOL reagent (Invitrogen). The RNA was reverse transcribed into cDNA with the help of the Omniscript reverse transcriptase kit (Qiagen, Hilden, Germany) and the primer RH5620 (5'-GAAGATCTAGAGCGGCCGCCCTTTTTTTTTTTTTTT-3'). Based on EST data from SL1-transspliced genes from European Pristionchus species (unpublished data) we designed generic RT-PCR primers for Pristionchus ribosomal protein genes (WM8220, 5'-TCGACAACGACAGAAAGAAGA-3', rpl-26, sense; WM8221, 5'-ACGGAGTCRTCGCTGTRCTTGC-3', rpl-26, antisense; WM8263, 5'-CCGTCAGCGYGGMATCCAGAAG-3', rpl-28, sense; WM8264, 5'-GCTGGASGGAGCGGAGRAGCTG-3', rpl-28, antisense; WM8114, 5'-GCYCAYATYTTCGCYTCTTTCAA-3', rps-14, sense; WM8113, 5'-GGRGTCTTNGTTCTRGTTCCTC-3', rps-14, antisense) and used them to synthesize the complete transcript in two overlapping fragments by PCR. The SL1-specific primer BJ234 (5'-GGTTTAATTACCCAAGTTTGAG-3') was used with the antisense primers to obtain the 5' part of the transcripts and the combination of RH5620 and the sense primers to obtain the 3' parts. The PCR was performed with the help of the HotStar Taq DNA polymerase kit (Qiagen) including the Q solution in the reaction mix. Conditions were initial activation of the enzyme at 95°C for 15 min, followed by 40 cycles of denaturation at 94°C for 30 sec, primer annealing at 50°C for 30 sec, and primer extention at 72°C for 3 min. The reaction was completed by an incubation at 72°C for 10 min. PCR fragments were gel purified using the Wizard SV gel purification kit (Promega) and sequenced directly using the PCR primers.