Fundam. appl. NemalOl., 1995,18 (5),419-425

Selection of a bacterium for the mass production of Phasmarhabditis hermaphrodita (Nematoda : ) as a biocontrol agent for

Michael J. WILSON *, David M. GLEN *, Susan K GEORGE * and Jeremy D. PEARCE ** * Department ofAgricullural Sciences, University ofBristol, !nstitute ofArable Crops Research, Long Ashton Research Station, Bristol BSIB 9AF, u.K. and ** MiCToBio Ltd., Church Street, Thriplow, Royston, Herts. SGB 7RE, u.K. Accepted for publication 16 September 1994.

Summary - Dauer larvae of the -parasitic , Phasmarhabditis hermaphrodir.a, were found to carry a mean of ca . 70 bacterial colony forming units with up to five different bacterial types present in individual . Nine bacterial isolates found associated with P. hermaphrodir.a, were injected into the body cavity of the slug, DerOGeras reliculalum. Only two isolates (Aeromonas hydrophila and Pseudomonasfluorescens, isolate 140), were found to be pathogenic. Phasmarhabditis hermaphrodila grown in monoxenic culture with five bacterial isolates ( osloensis, Providencia rellgeri, Serralia proteamaculans, P. fluorescens, isolate 140 and P. fluorescens, isolate 141) which supported strong growth were bioassayed against D. reliculalum. Nematodes grown with M. osloensis or P. fluorescens (isolate 141) were pathogenic, those grown with P. reugeri gave inconsistent results and those grown with the other two were non-pathogenic. These results demonstrate that differences in pathogenicity between nematode/ bacterium combinations did not result from differences in bacteriaJ pathogenicity a1one. M. osloensis was selected as the bacterium for rearing P. hermaphrodita as a biocontrol agent for slugs.

Résumé - Sélection d'une bactérie pour la production en masse ck Phasmarhabditis hennaphrodita (Nematoda : Rhabditidae), agent de contrôle biologique cks limaces -Il a été observé que le nématode parasite des limaces Phasmarhab­ ditis hermaphrodita était porteur d'environ 70 « unités formant une colonie » (CFU) appartenant jusqu'à cinq types différents de bactéries par individu. Neuf isolats bactériens associés à P. hermaphrodir.a sont injectés dans la cavité du corps de la limace Deroceras reliculalUm. Seuls deux isolats - Aeromonas hydrophila et Pseudomonas fluorescens isolat 140 - se sont révélés pathogènes. Phasma­ rhabditis hermaphrodir.a en élevage monoxénique avec cinq isolats bactériens (Moraxella osloensis, Providencia reugeri, Serralia proteomaculans, P. fluoresens isolat 140 et P. fluorescens isolat 141) permettant un taux élevé de croissance a été testé envers D. reliculalUm. Les nématodes élevés avec M. osloensis et P. fluorescens isolat 141 se sont montrés pathogènes, ceux élevés avec P. rellgeli donnent des résultats inconstants et ceux élevés avec les autres bactéries ne sont pas pathogènes. Ces résultats prouvent que les différences dans la nocuité des combinaisons nématode/bactérie ne provielU1ent pas seulement de la nocuité de la bactérie. M. osloensis a été sélectiolU1é en vue de l'élevage de P. hermaphrodir.a comme agent de contrôle biologique envers les limaces.

Key-words : Phasmarhabditis hermaphrodir.a, rhabditid nematodes, bacteria, parasite, slug, Deroceras reliculalUm, monoxeny, pa­ thogenicity, biologicaJ control.

The rhabditid nematade, Phasmarhabdùis herma­ & Poinar, 1979), respectively, which are carried monox­ phrodita (Schneider), is a parasite capable of killing sev­ enically in the anterior portion of the nematode gut. eral species of pest slugs and has the potential to be used These murualistic bacteria are thought to play an impor­ as a biocontrol agent (Wilson el al., 1993 a). It is a tant raIe in the pathogeniciry of the nematode/bacterium bacterial feeder and has been mass-produced in xenic complex and when these nematades are mass-produced culrure on nutrient media supporting a mixed bacterial for commercial use, they are grown in monoxenic cul­ flora (Wilson el al., 1993 b) using similar techniques to rure with their preferred species of Xenorhabdus or Pho­ those developed to mass-praduce entomopathogenic nematodes of the families Heterarhabditidae and Stei­ lorhabdus. This generally resuJts in consistent and pre­ nernematidae, which have been used successfully ta dictable yields of reliably virulent nematodes. These control several insect pest species. bacteria, however, can spontaneously alternate between Heterorhabditid and Steinernematid nematodes are two phases, and phase two provides a Jess suitable envi­ symbiotically associated with bacteria of the genera Pho­ ronment for nematode growth than phase one (Akhurst lorhabdus (Boemare, 1993) and Xenorhabdus (Thomas & Boemare, 1990).

ISSN 1164-5571/95/05 S 4.00/ © Gauthier-Villars - ORSTOM 419 M. J. Wilson et al.

There are many ways in which nematodes can inter­ To obtain monoxenic cultures of P. hennaphrodita, act with bacteria (Poinar & Hansen, 1986)", but Iittle is xenically-cultured individuais were first freed ofbacteria known about such interactions for P. hennaphrodita. then grown with the selected bacterium on a kidney­ Wilson et al. (1995) found that many species of bacteria based medium on 3 cm diameter Petri dishes, as de­ supported growth ofP. hennaphrodita in monoxenic cul­ scribed by Wilson et al. (1995 a). Gravid adult P. her­ tures, but that different bacterial isolates caused differ­ maphrodita were transferred, through n'lO changes of ences in the total numbers of nematodes produced in sterile water, to a sterile watch glass containing water culture and influenced the proportion of the population with 0.02 % sodium ethyl mercurithiosalicylate (" Thi­ which formed dauer larvae after three weeks. As dauer merosal ", Sigma) where they were left for 16 h at 10°C. larvae are the stage of P. hemzaphrodita which infect Nter this period, the larvae (L11L2) produced by the slugs, this has important implications for production of aduJts were transferred to centrifuge tubes containing P. hennaphrodita as a biocontrol agent. Wilson et al. la rrù quarter strength Ringer's solution (Poinar & Tho­ (1995) did not investigate differences in the quality ofP. mas, 1984) with 500 units ml -1 ofboth benzyl penicillin hennaphrodita grown in monoxenic culture with differ­ and streptomycin sulphate (Sigma). The larvae were ent bacteria, in particular, any effect that the bacterial kept in this solution at la oC for 24 h then concentrated food-source might have on the ability of dauer larvae to by centrifugation at 90 g for la min, re-suspended in infect and kill slugs. sterile quarter strength Ringer's solution and centri­ The investigations described herein into the relation­ fuged again. Re-suspension and centrifugation were re­ ships between P. hennaphrodita and bacteria had two peated once more to remove any traces of , related aims. First to determine the direct effects of bac­ then sterile (axenic) larvae were transferred to a sterile teria associated with P. hermaphrodita on the field slug, watch glass, where they could be handled individually Deroeeras retieulatum (Müller). And, second, to deter­ using sterile micro-pipettes. For larger scale production mine whether the pathogenicity of P. hennaphrodita to of monoxenic nematodes for use in bioassays, mono­ D. retieulatum is influenced by growth with different xenic P. hennaphrodita from Petri dish-cultures were bacteria. Bacteria were isolated from dauer larvae of P. introduced into 250 ml flasks by pippeting 1 ml of sterile hennaphrodita, from xenic cultures of P. hennaphrodita broth over the surface of the agar, then re-pippeting the which were producing rughly pathogenic nematodes liquid containing a suspension of bacteria and 50 000­ and from living or dead individuals of D. reticulatum 100 000 nematodes into each flask, containing either a infested by P. hennaphrodita. The effects of nine isolates solid or liquid medium. For solid phase cultures, flasks on slugs were tested by injecting cultures of bacteria were filled with foam chips impregnated with kidney directly into the haemocoel of D. retieulatum. Five of medium (Wilson et al., 1993 b) and incubated at 15 oC these bacteria which had been found to support good for 3 weeks; dauer larvae were then harvested by placing monoxenic growth of P. hennaphrodita (Wilson et al., the foam on sieves in water (Wilson et al., 1993 b). For 1995) were used to grow monoxenic cultures of P. her­ liquid cultures, nematodes were introduced into 250 rrù maphrodita wruch were then bioassayed for their effects flasks containing 50 rrù of liquid kidney medium (KYO) on D. retieulatum. (Wilson et al., 1995) which were incubated at 15 oC in a gyrotatory flask shakerlincubator at 200 rpm for 3 weeks. Dauer larvae were harvested by aUowing the Materials and methods nematodes to settle out in tap water.

COLLECTION AND MAINTENAt"lCE OF SLUGS ISOLATION AND MAINTENANCE OF BACTERlAL CUL­ Slugs (Deroceras retieulatum) were collected from TURES traps consisting of inverted plastic flowerpot saucers Bacteria were isolated from within dauer larvae of P. baited with bran in rough grassland at Long Ashton hennaphrodita, from living and dead D. retieulatum in­ Research Station (LARS). They were kept at 10°C with fected with P. hermaphrodita, and from xenic foam chip a 12 h night and day cycle in non-airtight plastic boxes cultures of P. hennaphrodita, as described below. lined with moistened absorbent cotton wool. Dauer larvae were first surface-sterilised by immer­ sion in 0.1 % (w/v) sodium ethyl mercurithiosalicylate COLLECTION AND MAINTENANCE OF NEMATODES (" Thimerosal ", Sigma Ltd) for 1 h, transferred to fresh In the initial phase ofthe investigation, dauer larvae of Thirnerosal for a further 3 h, then transferred into sterile P. hennaphrodita were either produced in xenic cultures quarter strength sterile Ringer's solution (Poinar & Tho­ using foam culture and harvested as described by Wil­ mas, 1984). Bacteria were liberated from these surface­ son et al. (1993 b), or they were isolated from individuals sterilised nematodes in one of two ways: i) 48 individual of D. retieulatum that were found to be infected with P. dauer larvae were transferred into a drop of sterile saline hennaphrodita when collected from the field or from D. on a flame-sterilised microscope slide, and then eut at retieulatum infected in the laboratory with xenically cul­ several sites along their bodies. The drop of saline com­ tured P. hermaphrodita. plete with the dead nematode was transferred to a

420 Fundam. appl. NemaLOI. Produaion ofPhasmarabditis hermaphrodita

nutrient agar plate where it was spread over the agar removed from these boxes after five days and kept indi­ surface with a glass spreader; il) an estimated 20 000 vidually in Petri dishes each containing a 3 cm diameter dauer larvae were suspended in 1 mJ of sterile Ringer's leaf disc of Brassica ehinensis (L) which was renewed solution in a 5 mJ Teflon tissue homogeniser, ground for every 3 days. Slug mortality was recorded after 9 days in c. 1 min then added to 9 mJ of sterile nutrient broth. the Petri dishes (fourteen days after initial exposure to This suspension was seriaily (ten-fold) diluted and nematodes). 0.1 mJ a1iquots of the resulting bacterial suspensions Three experiments were done. In the first, nematodes were plated out onto nutrient agar, using standard mi­ grown in monoxenic foam-chip cultures with five bacte­ cro-biological techniques. ria (Moraxella osloensis, Providencia rellgeri, Serratia pro­ Bacteria were isolated from xenic foam-chip cultures teamaculans and two isolates of Pseudomonas fluorescens by transferring foam chips, using sterile forceps, to nu­ (numbers 140 and 141) were bioassayed using 50, 70, trient broth tubes which were shaken, serially diluted 50,30 and 70 slugs per nematode dose, respectively, for and plated out onto nutrient agar. each nematode/bacterium combination, as perrnitted by Bacteria were isolated from within the mantle ofliving yields ofdauer larvae from flasks ofdifferent monoxenic nematode-infected D. retieulatum, and from cadavers of cultures. The bacterial flora in ail flasks was examined D. reticulatum which had died foilowing nematode in­ immediately before harvest to ensure that cultures were fection and were supporting growth oflarge numbers of still monoxenic. Foam chips were removed from each nematodes. The surface of the mantle ofliving D. retieu­ flask and suspended in nutrient broth, which was diluted latum was fust swabbed with 70 % (v/v) ethanol to re­ serially and plated out onto nutrient agar so that move surface bacteria. The mantle was then pierced individual colonies could be observed. Any cultures with with a sterile needle and drops of the exuding fluid obvious contamination were discarded. spread over the surface of nutrient agar plates. Loopfulls Two further experiments used nematodes grown in of bacterial mix were taken from the slug cadaver and monoxenic liquid culture with either M. osloensis or P. transferred ta nutrient broth which was seriaily diluted rellgen·. In both experiments, 30 slugs were used for ail and plated out on nutrient agar plates. six nematode doses of each nematode/bacterial combi­ In ail cases, nutrient agar plates were incubated at nation. As before, bacteria present in the liquid cultures 25 oC for 2 days. Numbers of colony-forming units were tested prior to harvesting to ensure monoxenicity. were then estimated and different bacterial isolates, as For aIl bioassays, slug mortality data were analyzed by distinguished by colonial morphology, were subcultured probit analysis using the GENSTAT 5 (Payne et aL, and maintained on nutrient agar slopes. 1987) Probitan procedure (Payne, 1991).

INJECTION OF SLUGS wrrH BACTERlA Results Nine bacterial isolates were grown in static cultures of nutrient broth for 24 h at 25 oC before their effects on D. ISOLATION OF BACTERlA retieulatum were tested. D. reticulatum were anaesthe­ Over 150 bacterial isolates were obtained, 54 from the tized by exposure ta carbon dioxide for 15 min (Hen­ intestine ofdauer larvae, 7 from infected slugs and ca 90 derson, 1969) prior to injection with la 1-11 of sterile from xenic nematode cultures. Results of tests used in nutrient broth or bacterial culture, using a micro-appli­ identifying bacteria have been deposited with the editor cator fitted with an ail-glass syringe. After injection, of Fundamental and Applied NemalOlogy. Moraxella os­ twenty slugs per treatrnent were kept in plastic boxes loensis was described as Moraxella phenylpymvica in an (ten slugs per box) lined with moist absorbent cotton earlier paper (Wilson et aL, 1995) but further tests have wool and mortality was recorded after 8 days. Three identified it as the former species (A. von Graevenitz, different experiments were done in which at least three pers. comm.). different bacteria were screened. In the first two experi­ Nine individual bacterial isolates were selected and ments, slugs injected with sterile nutrient broth and un­ identified for use in this study (Table 1). The mean treated slugs were used as controls, whereas in the third number of bacterial colony-forrning units, estimated by experiment only slugs injected with sterile nutrient broth cutting up individual dauer larvae, was 26 (s.e. =22.2) were used as controls. and the number estirnated by the tissue homogeniser method was 71 (s.e. = 8.5). The number of different BIOASSAy OF EFFECTS OF P. HERMAPHRODITA ON D. bacterial types from individual nematodes, as distin­ RETICULATUM guished by colonial morphology, ranged from one to Bioassays were done at la oC and 12 h daylength us­ five. ing a soil-based bioassay system as described by Wilson et al. (1993 b). Ten individual D. retieulatum were INJECTION OF D. RETICULATUM WfTH BACTERlA placed in plastic boxes ofmoistened soil aggregates con­ Two of the nine species of bacteria tested for patho­ taining either 0, 15 000, 23 000, 35 000, 55 000, or genicity to D. retieulatum, Aeromonas hydrophila and 75 000 dauer larvae of P. he1maphrodita. The slugs were Pseudomonas fluorescens (isolate number 140), were

Vol. 18, n° 5 - 1995 421 M. J Wilson et al.

Table 1. Species and sources of bacœria leSledfor lheir effeClS on todes grown monoxenicaUy on four other bacteria which . Ali bacleria were isolaled al Long AshLOn supported good nematode growth (Wilson el al., 1995) Research Stalion. but wruch were not themselves pathogenic when in­ jected into D. relicuLaLUm. NemaLOdes grown in foam chip cultures Bacterium Isolation source There was no significant increase in slug mortality Aeromonas hydrophila Intestine of P. hermaphrodiLa with increasing nematode dose for nematodes grown dauer larvae with P. rellgen, P. fluorescens (isolate 140) or S. pTOlea­ Aeromonas sp. * Cadaver of field collected D. maeuLans (Fig. 2). However, probit analysis showed a reliculalum which had lied fol­ significant (P < 0.05) positive dose-response relation­ lowing infection with P. her­ ship for nematodes grown in monoxenic culture with M. maphrodita osloensis and P. fluorescens (isolate 141) (Fig. 2). The Flavobaclerium breve Xenic foam-chip culture of P. EDso values (effective dose required to kiU 50 % ofslugs, hermaphrodita 14 days after treatment) for these two nematodelbacteri­ FlavobaClerium odoralum Xenic foam-chip culture of P. um combinations were 28 900 (lOglO s.e. = 0.0605) and hermaphrodita 49700 (log\O s.e. =0.0947), respectively. There was no Moraxella osloensis** Xenic foam-chip culture of P. evidence of lack of fit for either probit mode!. hermaphrodita Providencia reugen' Xenic foam-chip culture of P. Nemalodes grown in liquid cultures hermaphrodiLa In the first experiment (Fig. 3A), nematodes grown Pseudomonas fluorescens Cadaver of field-collected slug with M. osloensis were found to be pathogenic with an (isolate no. 140) (D. reliculalum) which had estimated ED50 value of 35900 (log\O s.e. = 0.0583), died following infection with similar to that for nematodes grown with this bacterium P. hermaphrodita on foam chips. However, nematodes grown with P. Pseudomonas fluorescens Manùe cavity of living field­ rellgen were not pathogenic, because slug mortality did (isolate no. 141) collected slug (D. reliculatum) not increase with nematode dose for this nematodelbac­ infected with P. hermaphrodiLa terium combination. Again, this is similar to the results Serralia proleamaculans Xenic foam-chip culture of P. obtained using foam chip cultures. hermaphrodita In the second experiment (Fig. 3B), nematodes grown with M. osloensis were again highly pathogenic * SimiJar [0 Aeromonas salmonicida National Collection of Indusrrial with an ED50 value of 20000 (log1o s.e. = 0.0355) but, and Marine Bacteria NC1MB, pers. comm. in contrast tO earlier experiments, nematodes grown "* This was identified by NCLMB as an arypica! isolare of Momxella with P. rellgeri were also pathogenic with an estimated phenylpyruvica, bur has since been identified by A. von Graevenirz ED50 of 32 300 (loglo S.e. =0.0511). (pers. comm.) as M. osloemis. Discussion found to be highly pathogenic when injected into D. relicuLalum (Fig. 1), causing significantly (P < 0.01) Although the studies on the gut flora of P. herrnaph­ greater mortality than that for untreated slugs or slugs rodita presented in this paper are limited in their scope, injected with sterile nutrient broth. Injection with other they have established that P. hermaphrodita can retain bacteria, inciuding another isolate of P. fluorescens many different types of bacteria within its gut and that (No. 141), or sterile nutrient broth, did not cause mor­ no one isolate seemed to predominate. Since P. her­ tality significantly greater than that in untreated slugs. maphrodita can grow and reproduce on a wide range of bacteria (Wilson el al., 1995), and it is known that in­ BIOASSAYS OF THE EFFECTS OF NEMATODES, GROWN gested bacteria may sometimes survive within the gut of IN MONOXENIC CULTURE WITH DIFFERENT BACTERlA, bacteriophagous nematodes (Poinar & Hansen, 1986), ON D. RETICULATUM this finding is not surprising. However, it contrasts with Aeromonas hydrophiLa, which was found to be patho­ entomopathogenic nematodes, which are generally con­ genic when injected into D. reticulalum, did not support sidered to retain only their symbiotic Xenorhabdus or good growth of P. hermaphrodita in monoxenic cultures PhOlOrhabdus bacteria. Nevertheless, it should be noted (Wilson el al., 1995) and it was not therefore possible to that Boemare el al. (1983) found eight species of non­ test nematodes grown on this bacterium for their effects symbiotic bacteria within Steinemema carpocapsae strain on D. reliculalum. However, P. fluorescens, isolate DD-136 and that Lunau el al. (1993) found evidence of No. 140, which was pathogenic when injected into D. non-symbiotic bacteria associated with various Sleiner­ relicuLaLUm, supported good growth of P. hennaphTOdita nema spp. and Helerorhabdùis spp. in monoxenic cultures and nematodes grown in cultures It is difficult to quantify accurately retention of baete­ with this bacterium were tested, together with nema- ria within the gut of dauer larvae of P. hennaphrodila,

422 Fundam. appl. NemaLOI. Production of Phasmarabditis hermaphroclita

Experiment 1 Experiment 2 Experiment 3 100

80 >. :t: -ai t 0 E 60 Cl .2 (J') -c 40 Cl) ....ü Cl) Cl... 20

o

Fig. 1. Percen/age slug mortaliLy after 8 days at 10°C in three experimerlls in which D. reticularum was injected wiLh (lO /û of sterile Tlutrieru broth or ovemight cullUres ofdifferent bacteria. I.s.d. =least significant difference, P < 0.05).

because the double cuticle is difficult to disrupt com­ African snail, Achatinafulica (Mead, 1961), and in snails pletely. However, numbers of bacteria per P. hermaph­ reared in snail farms in the O.K. (Runham, 1989). rodita larva are similar to estimates made for entomo­ Our study clearly shows that growing P. hermaph­ pathogenic nematodes, e.g., Poinar (1975), states that rodita in monoxenic culture with different bacteria can 20-250 bacterial cells are associated with each dauer lar­ dramatically affect its pathogenicity to D. reticulatum va of Steinemema carpocapsae. and that healthy dauer larvae of P. hermaphrodita grown Of nine selected bacterial strains associated with P. monoxenically with certain bacteria are unable to kill hermaphrodita, only A. hydrophila and one strain of P. slugs. The bacterium which produced the most patho­ fluorescens (isolate number 140) were found to be patho­ genic nematodelbacterium combination (M. osloenis) genic to D. reticulatum when injected into the haemo­ was not itself pathogenic when in;ected into D. reticu­ coel. It is, perhaps, surprising that other bacterial strains latum, and, furthermore, nematodes grown in monox­ were not pathogenic considering the large numbers of enic culture with P. fluorescens (isolate 140) were not bacteria which were injected. Little is known about the pathogenic to slugs, even though this bacterium was immune systems of gastropod molluscs, but phagocytic itself pathogenic when injected into slugs. It is possible cells and agglutinins are thought to be important (South, that the way in which the bacteria interact with both P. 1992) and the present study indicates that antibacterial hermaphrodita and the immune system of the host slug is mechanisms are very effective. It should be noted that important in determining the pathogenicity of different while A. hydrophila and P. fluorescens (isolate 140) were combinations of nematodes and bacteria. Dunphy et al. pathogenic when injected, they are only opportunistic (1985), showed that Steinemema glaseri was more viru­ pathogens. Neither bacterium can cause disease in D. lent when grown on certain subspecies of Xenorhabdus reticulatum as a result of skin contact or ingestion (Wil­ nematophilus than others and found that virulence corre­ son, 1992). Hm,vever, interestingly, A. hydrophila has lated weil with the proportion of infective juveniles been reported to be associated with disease of the giant which retained bacteria. Also, Akhurst (1986) observed

Vol. 18, n° 5 - [995 423 M. J. Wilson et al.

100 100 A

80 80 • >. :'=ro t (tJ o 60 t E o 60 Cl E :::J Cl en :::::l c: li) -Q) 40 C 40 e --Q) Q) o ~ ~ • ID • • Q.. 20 0• • • • U • 20 • 0 0

O'------l_----l._--'-_--'-_---l-_--J...._-'-~ L-..-.J-----I.-----'-----I..----'-.~~a...J__•• o •...:..l__...... 4.2 4.4 4.6 4.8 4 4.2 4.4 4.6 4.8 5 Log nematode dose Log nematode dose Fig. 2. Percenlage m01'lalùy ofD. reticulatum measured 14 days after inùial exposure 10 five differenl doses of P. hermaphrodita dauer larvae grown in monoxenic culture on foam chips wùh M. osloensis (.), P. rettgeri (.), S. proteamacLÙans (*), P. f1uo­ rescens (isolale 140) (0), or P. f1uorescens (isolale 141) (0). 100 that neither axenic S. glaseri nor its symbiont, X. poinarii were pathogenic to Gallen'a mellonella on their own, but were in combination. Growth on different bacteria may influence the ability 80 of the nematode to find and penetrate slugs. CUlTan ~ (1993), referring tO entomopathogenic nematodes, (tJ states that under controlled conditions, apparently fa­ ~ -­o vourable for infection, only a proportion of individual E 60 nematodes are capable of establishing in a host insect. Cl This proportion ranges from 13 to 71 % for differenr :::::l species of Steinemema and Heterorhabditis. Curran li) (1993) also states that the proportion of the population c 40 which is capable ofinfecting and establishing within host --Q) o insects can be significantly altered by the nematode cul­ ~ ture conditions. It is possible that different conditions in ID monoxenic cultures with differenr bacteria iniluenced ll.. the proportion of P. hermaphrodiLa capable of infecting 20

~ Fig. 3. Percenlage morlalùy ofD. reticulatum measured 14 days after inùial exposure ta five differenl doses of P. hermaphrodita OL--'---l..-~_l....--'---l..-~_J.....--'----L dauer larvae grown in liquid monoxenic culture wùh M. osloensis 4 4.2 4.4 4.6 4.8 5 (.) 01' P. rettgeri (.) in (A) che jirsc and (B) che second IYioassay experimenl using nematad.es produced in liquid culcures. Log nematode dose

424 Fundam. appl. Nema10 1. Production ofPhasmarabditis hermaphrodita

D. reliculalum in our experimems, Providencia rellgen~ LUNAu, S., STOESSEL, S., SCH,WDT-PE1SKER, A. J & EH­ the bacterium which produced the highest yields of LERS, R-U, (1993). The establishment of monoxenic in­ dauer Jarvae of P. hermaphrodita in an earlier study (Wil­ ocula for scaling up in vitro cultures of the entomopathogen­ son el al., 1995) was found to produce nematodes which ic nematodes Steinemema spp, and Helerorhabditis spp. were inconsistem in their effects on D, reliculalUm, NemalOlogica, 39 : 385-399, showing no detectable effects in two experiments, but a MEAD, A. R (1961). The gianl African snail: A problem in strong dose response in a third experiment. The two economic malacology. Chicago, Universiry of Chicago Press, experiment were done using identical methods, each 257 p. with five doses of nematodes and 30 slugs per dose, The PAYNE, R. W, (1991), Procedure probitanalysis. In: Payne, R, results of both experimems \Vere statistically valid, but W., Arnold, G, M. & Morgan, G, W, (Eds) , Genslal proce­ contradictory, for reasons which are unknown. Moraxel­ dure library manual 2.2[2]. Oxford, Numerical Algorithms la osloensis produced nematodes which were consistently Group: 201-204, pathogenic to D. reticulmum in a11 three experimems PAYNE, R. W., LANE, P, W" fuNSLEY, A, E, BrCKNELL, K. reported here and since. It produced yields of dauer E" DIGBY, P. G, N., HARDING, S, A" LEECH, P, K., SYMP­ larvae only slightly less than those achieved with P. rellg­ SON, H, R, TODD, A. D., VERRIER, P. ]., WHlTE, R P., eri (Wilson el al., 1995). Therefore, it was selected as the GOWER, J c., TUNCLIFFE-WILSON, G. & PATERSON, L J preferred bacterial species to produce P, hermaphrodita (1987),Genslal 5 Reference Manual. Oxford, Oxford Uni­ for further studies and for commercial production as a versiry Press, xviü + 749 p, biocontrol agent for slugs, POINAR, G. 0" Jr. (1975). EnlOmogenous nemalOdes, A manual Acknowledgements and hoSI liSI of inseCI-nemalOde associalions. Leiden, Brill, We thank the Agricultural Genetics Company Ltd, ofwhich 749 p, MicroBio Ltd is a wholly-owned subsidiary, for funding, Dr POINAR, G. O.,]r. (1989), Non-inseet hosts for the entomoge­ P, Brain for statistical advice, NCIMB Aberdeen, for identify­ nous rhabditoid nematodes Neoaplectana (Steinernemati­ ing Aeromonas spp" Prof. A. von Graevenitz, Universiry of dae) and Heterorhabdùis (Heterorhabditidae), Revue Néma­ Zurich, for identifying M, osloensis and Dr N. Boemare, Uni­ lOI., 12 : 423-428. versité de Montpellier, for his comments on the manuscript. POINAR, G. 0" Jr. & HANSEN, E L. (1986), Associations References between nematodes and bacteria, Helminth, Abslr, Sel', B, AKHuRST, R J (1986), Xenorhabdus poinarii .' its interaction 33: 61-81. with insecr pathogenic nematodes, Sysl. appl, Microb" 8 : POINAR, G. O.,Jr. & THOMAS, G, M, (1984). LaboralOry guide 142-147. 10 inseCl palhogens and parasites. New York & London, Plen­ AKHuRST, R. J & BOEMARE, N, E. (1990), Biology and raxo­ um Press, xvi + 392 p, nomy of Xenorhabdus. In : Gaugler, R. & Kaya, H. K (Eds). RUNHAlvl, N, W, (1989), Snail farrning in the U.K In: Hen­ EnlOmopalhogenic nemalOdes for biologual contro!. Boca Ra­ derson 1. F. (Ed,), Slugs and snails in world agricullure. ton, Florida, CRC Press: 75-90, Thomton Heath, British Crop Protection Council : 49-56. BOEMARE, N, E., AKHuRsT, RJ, & MOURANT, R. G. (1993). SOUTH, A. (1992), Terreslrial slugs : Biology, ecology and con­ DNA relatedness between Xenorhabdus spp, (Enterobacte­ lro!. London, New York, Tokyo, Melbourne & Madras, riaceae), symbiotic bacteria of entomopathogenic nema­ Chapman & Hall, vii + 428 p, todes, and a proposal to transfer Xenorhabdus luminescens to THOMAS, G, M. & POINAR, G. O., Jr. (1979), Xenorhabdus a new genus, PholOrhabdus gen, nov. Int. J. SYSI. Baclerio!., 43 : 249-255, gen, nov, a genus of entomopathogenic bacteria of the fam­ ily Enterobacteriaceae. lm. J, SYSl. Bacleno!., 29 : 352-360. BOEMARE, N" BONrFASSI, E, LAUMOND, C. & LUCIANI, J (1983), Érude expérimentale de l'action pathogène du né­ WILSON, M. J (1992), A nemalOde parasite for biological control matode Neoapleclana carpocapsae Weiser; recherches gnoto­ ofslugs. PhD Thesis, Universiry of Bristol, 178 p. biologiques chez l'insecte Galleria mellonella L. Agronomie, WILSON, M, J" GLEN, D. M., & GEORGE, S, K (1993 a). The 3:407-415, rhabditid nematode Phasmarhabditis hermaphrodita as a po­ CURRAN, J. (1993). Post-application biology ofentomopatho­ tential biological control agent for slugs, Biocontr. Sci. & genic nematodes in soil. In: Kaya, H. K, Bedding, R. & TechlWl., 3 : 503-511. Akhurst, R. (Eds). NemalOdes and lhe biological control of WILSON, M, J, GLEN, D. M., GEORGE, S, K. & BUTLER, R. peSIS. East Melbourne, CSIRO : 67-77. C. (1993 b). Mass cultivation and storage of the rabditid DUNPHY, G, B" RUTHERFORD, T. A. & WEBSTER, J M, nematode, Phasrnarhabditis hermaphrodita, a biocontrol (1985). Growth and virulence ofSteinernema glaseri influen­ agent for slugs, Biocontr, Sci, & Techno!., 3 : 513-521, ced by different subspecies of Xenorhabdus nemalOphilus, J, WILSON, M.]., GLEN, D, M., PEARCE, l D, & RODGERS, P. Nemalol., 17: 476-484, B, (1995), M.onoxenic culture of the slug parasite Phasmar­ HENDERSON, 1. p, (1969). A laboratory method for assessing habditis herrnaphrodita (Nematoda : Rhabditidae) with dif­ the toxiciry ofstomach poisons to slugs. Ann. appl. Bio!., 13 : ferent bacteria in liquid and solid phase. Fundam, appl. Ne­ 167-171. malOl., 18 : 159-166,

Vol. 18, n° 5 - 1995 425