Romanomermis culicivorax: Penetration of larval mosquitoes

JOURNAL

OF

INVERTEBRATE

PATHOLOGY

Romanomermis

54, 191-199 (1989)

cu/icivorax:

Penetration

M. M. SHAMSELDEAN' AND E.G. Department

of Nematology,

of California,

University

of Larval

Mosquitoes

PLATZER Riverside,

California

92521

Received September 20, 1988; accepted January 3, 1989 In the presence of second larval instars of three mosquito species the preparasites of Romanoculicivorax swam near the water surface in an orthokinetic manner. When the preparasites were ca. 1 mm from the host, they stopped and swam klinotactically toward the host. During this phase, the preparasites secreted a small amount of a putative adhesive material from the anterior region and host contact was completed. The adhesive appeared to aid in attachment of the preparasites to the host and initiation of the search-boring phase. The preparasites glided over the host until a suitable penetration site was found. The penetration phase was initiated by probing with the odontostyle. This was followed by partial paralysis, decreased intestinal peristaltic movement, and temporary cardiac arrest in all host mosquitoes which was probably related to injection of esophageal secretions. SEM observations showed that the abdominal walls were the most frequent site for penetration. As the preparasites entered through the penetration hole, microorganisms adhering to the cuticle of the preparasites were retained by the adhesive which accumulated around the penetration site. Thus, microbial contamination of the host was avoided by a mechanical cleansing mechanism. Penetration was usually completed in less than 10 min. o 1989Academic press, I~C. mermis

KEY

WORDS:

Romanomermis

culicivorax;

Aedes

aegypti;

Anopheles

quadrimaculatus;

Culex

nematode adhesive; nematode behavior; cuticle penetration; mosquito paralysis; insect cardiac arrest.

pipiens;

INTRODUCTION

Cuticular penetration of insects by entomophilic nematodes often involves use of a stylet which is aided by glandular secretions (Christie, 1936; Poinar, 1968; Obiamiwe and MacDonald, 1973). Whereas the tylenchids Tripius sciarea (Poinar and Doncaster, 1965) and Dkladenus siricidicola (Bedding, 1972) use the stylet for penetration, the rhabditid nematode Heterorhabditis heliothidis penetrates the host mainly through the integument using its terminal tooth (Bedding and Molyneux, 1982). Although some mermithids enter the host via the oral cavity and subsequent penetration of the gut wall (Strickland, 1911; Phelps and Defoliart, 1964), penetration of the cuticle is the primary route of entry for many mermithids (Comas, 1927; Iyengar, 1927; Svabenik, 1928; Wiilker, 1965; Petersen and Chapman, 1970; Molloy and Jamnback, ’ Current address: Department of Zoology and Nematology, Faculty of Agriculture, Cairo University, Giza, Egypt.

1975). The general process of infection of mosquito larvae by mermithid nematodes has been observed previously by many authors but the process has not %een described fully. In this study of Romanomermis culicivorax the penetration process was followed with light microscopy aided by video recording and with scanning electron microscopy. MATERIALS

AND METHODS

Experimental Infections R. culicivorax was cultured after the techniques of Platzer and Stirling (1978). An autogenous strain of Culex pipiens served as the host and the larvae were reared in tap water at 27°C. Second-instar larvae of both Anopheles quadrimaculatus and Aedes aegypti were obtained from Dr. Brian Federici (University of California, Riverside). Experimental infections for scanning electron microscopy studies were performed by placing 10 second-instar mosquito larvae in 10 ml of deionized water in a 10 x 35mm Petri dish to which ca. 1000 R. 191 0022-201 l/89 $1.50 Copyright Q 1989 by Academic Press, Inc. A!J rights of reproduction in any form reserved..

192

SHAMSELDEAN

culicivorux were added in a small volume of water. After ca. 10 min at room temperature (24°C) host larvae were collected individually and fixed for scanning electron microscopy. The penetration process was followed by light photomicrography and video recordings of mosquito larvae exposed individually to preparasitic nematodes at a ratio of 1: 10 (host/parasite) under similar conditions. Video Equipment

and Techniques

Mosquito and nematode behavior during the penetration process were observed with a Wild dissecting microscope or a Zeiss inverted compound microscope. Video records were made with a Panasonic TV camera (WV 1850) and a Sony U-matic videocassette recorder (VO-5600). Real time display was provided by a 9-in. Panasonic video monitor (WV-5360) and a video timer (FOR.A, VTG-33) supplied a timing record. With the aid of a dissecting microscope nematode penetration was observed in a 10 x 35mm Petri dish, whereas a standard glass microslide with 1 ml of water was the arena on the inverted compound microscope. SEM Techniques After mosquito penetration was initiated by infective nematodes,. mosquito larvae were recovered individually in ca. 1 ml of water with a plastic pipette at various time intervals and added to modified specimen holders placed in the wells of a tissue culture plate (Costar). The specimen holders were fashioned from BEEM capsules with the bottoms removed and closed at both ends with a 3-pm nylon mesh held in place with perforated (6-mm holes) BEEM caps. Ca. 1 ml of 7% glutaraldehyde was rapidly added to the specimens to yield a final glutaraldehyde concentration of 3.5%. The nematodes were killed within an hour. After 2 hr, specimens were washed in deionized water and gradually dehydrated in a graded ethanol series. Specimens were left

AND

PLATZER

for ca. 25 min in each of the following ethanol concentrations: 10, 20, 40, 60, 70, 80, 95, 100%. They were subsequently treated three times with 100% ethanol for 25 min each and critical point dried using carbon dioxide (Eisenback, 1985). Dried specimens were mounted on stubs, sputtercoated with 20 nm of gold/palladium alloy, and examined with a JEOL-JSM-35C SEM at 5 and 15 kV. Infected mosquito larvae were opened for internal SEM examination by a simple fracturing procedure. After the specimens were critical point dried as above, they were fractured by pressing firmly on the specimens with the tip of a sharp insect pin. Specimens were then sputter-coated and examined by SEM. Measurement

of Mosquito

Heart Rate

Heartbeats were easily observed through the transparent dorsal body of second instar mosquito larvae. Individual mosquitoes were infected as described above. Mosquito larvae were placed individually on a glass slide in one drop of distilled water and excess water was removed with tissue paper to minimize mosquito movement. The main heart chamber at the posterior end of each mosquito larva was watched carefully as the host was penetrated by a single preparasite. A stop watch and a manual counter were used to record heart beats per minute. Heart beats were counted before, during, and after penetration. Time needed for each mosquito larva to regain mobility was recorded. Statistical

Analysis

Data from the experiments were analyzed by Student’s t test, analysis of variance, and Duncan’s multiple range test (Snedecor and Cochran, 1980). RESULTS

General-Description of the Penetration Process After infective juveniles

of R. culicivorax

R. culicivorax

AND

LARVAL

were added to water containing host larvae, the preparasites swam around randomly. During this phase, the nematodes moved their heads in an orthokinetic manner but as they came within ca. 1 mm of a host larva they stopped their body movement. In a few seconds the nematodes resumed moving klinotactically toward the host. At contact, the parasites appeared to attach to the host by means of an adhesive material secreted from the anterior portion of the body. The attachment phase when successful was completed within 1 min (Fig. 1). This was followed by a search-boring phase, which took 2 to 3 min (Fig. 2). When the nematodes found a suitable site near the edge of a body segment, for penetration, they began stylet thrusting. Thrusting usually lasted for 2 min, after which the nematodes opened a small hole to fit only the diameter of their odontostylets (Fig. 3) through which they appeared to inject material into the body of their host. This action resulted in temporary paralysis of mosquito larvae which lasted for almost 2 min (Table 1). The infective juveniles started the entrance attempt during paralysis of the host. The nematodes contorted their bodies and forced their way inside, boring through the mosquito integument (Fig. 4). A peripheral secretion covered the cuticle of the nematode and adhered to microorganisms surrounding the nematode. This material appeared to prevent transport of adherent organisms (e.g., bacteria) into the host along with the nematode. When almost half way in, the nematode then quickly drew in the rest of its body, sliding into the hemocoel of the host. At the end of this phase the peripheral secretion remained in and around the penetration hole (Fig. 5). After entrance into the mosquito larva, infective juveniles had an outer coat on the cuticle which is presumably a thin glycocalyx (Fig. 6). The nematodes in the hemocoel thrusted their spears against some of the body tissues for several minutes and moved about actively for ca. 48 hr inside

MOSQUITOES

193

the mosquito

larva. As shown in Table 1, often penetrated mosquito larvae in less than 9 min, starting with the attachment phase and ending with total entrance into the host. Romanomermis

Quantitative Data for the Penetration Process

The time required for half and complete penetration into both A. quadrimaculatus and C. pipiens mosquito larvae were recorded. No significant differences (t test, P > 0.05) were found in time required for nematode penetration into A. quadrimaculatus and C. pipiens mosquito larvae (Table 1). Multiple penetration could occur and the time for host penetration was not significantly different for subsequent penetrators. In two cases, quadruple infections were observed (one anopheline and one culicine larva) and time for complete penetration by the fourth nematode was 11.2 and 9.1 min, respectively, i.e., ca. 4 min above the mean time required for the third nematode to penetrate in Anopheles and ca. 2 min above the mean in Culex. SEM observations showed that juveniles of R. culicivorax preferred to penetrate through the main body cuticle near the edge of a body segment rather than through intersegmental membranes (Fig. 4). The percentage of penetration via intersegmental membranes was 10.9% in A. quadrimaculatus and 9.1% in C. pipiens (Table 2). In a total of 91 Anopheles and 44 Culex larvae, there was a statistically significant difference between penetration sites on the mosquito bodies for both species: the nematodes preferred to penetrate through the abdomen more than the other parts of the body (F test, P < 0.01). Many nematodes (ca. 60%) preferred to penetrate in the abdominal region of A. quadrimaculatus larvae near the base of a sensory hair (Fig. 4). Paralysis and cardiac arrest phenomena were recorded by video techniques and visual observations for larvae of A. quadrimaculatus, A. aegypti, and C. pipiens. Host body movement ceased when the

194

SHAMSELDEAN

AND

PLATZER

R. cuficivorax AND LARVAL MOSQUITOES

195

FIG. 1. Romanomermis culicivorax infective juvenile attached by adhesive material (AM) to the quadrimaculatus second instar mosquito larva. edge of the fourth abdominal segment of Anopheles (AD, amphid; C, mosquito cuticle; bar = 1 km.) FIG. 2. Two Romanomermis culicivorax infective juveniles starting the exploration phase attached to the mseo- and metathorax of Anopheles quadrimaculatus second instar larva. (C, mosquito cuticle; bar = 100 pm.) FIG. 3. SEM photomicrograph of a fracture through the body cuticle of second instar mosquito larva quadrimaculatus showing the everted odontostylet (0) of an infective juvenile of Roof Anopheles manomermis culicivorax during paralysis phase. (C, mosquito cuticle; bar = 1 km.) FIG. 4. An infective juvenile of Romanomermis culicivorux forcing its way through the lateral side of the fifth abdominal segment of a second instar Anopheles quadrimaculatus larva. (S, sensory hair; bar = 10 pm.) FIG. 5. SEM photomicrograph showing the penetration hole in the third abdominal segment of a quadrimaculatus surrounded by the remnants of the adhesive mass (AM). second instar Anopheles Bacterial cells (B) are stuck to the mosquito body surface. (Bar = 10 km.) FIG. 6. The infective juvenile of Romanomermis culicivorux inside a fractured second instar Anopheles quadrimaculatus larva. The nematode has an outer coat which is probably glycocalyx (G). (I, mosquito intestine; bar = 10 pm.)

nematode stylet penetrated the cuticle. Subsequently, the heart slowed to half of the normal rate, then it stopped beating completely while the nematode was entering. Heart beats recorded before and during penetration were significantly different (P < 0.01) (Table 3). The peristaltic movements of the mosquito intestine were decreased by the nematode penetration, but the intestinal movements were inconsistent and could not be reliably quantitated.

Within 1 to 2 min of completion of nematode penetration mosquito larvae resumed normal activity and the heart rate returned to normal. DISCUSSION

Early observations

of the infection proshowed that the preparasites are positively thigmotactic and negatively geotactic, behavioral characteristics that greatly increase their chance of

cess in R. culicivorax

196

SHAMSELDEAN

AND PLATZER

TABLE TIME

FOR PENETRATION

Mosquito species Anopheles

quadrimaculatus culicivorax

AND Culex

pipiens

BY

Penetration sequenceb

Mean time for 50% penetration (mitt 2 SD)

Mean time for complete penetration (min f SD)

7 7

1 2 3

6.64 ” 2.8 7.28 f 4.1 7.61 2 1.9

7.70 f 3.1 8.20 + 4.1 8.76 + 1.8

9 7 4

1

2 3

7.86 f 2.7 7.87 f 2.7 7.13 * 1.5

8.31 f 2.8 8.41 2 2.7 7.67 f 1.5

N”

quadrimaculatus I

Culex

1

OF SECOND INSTAR LARVAE OF Anopheles INFECTIVE JUVENILES OF Romanomermis

pipiens

a Total number of mosquito larvae infected. b Sequence of nematodes infecting mosquito larvae. C Time for nematodes to penetrate halfway into the host.

contact with a suitable mosquito host (Petersen, 1973). It was also reported that the preparasites of most aquatic mermithids attach to the host by means of a stylet and enter the hemocoel through a hole made in the host’s cuticle (Petersen, 1985). Our recent observations on R. culicivorax have established that six stages can be identified in the infection process: host detection, orientation to the host, attachment, searchTABLE

2

PENETRATION SITES FOR Romanomermis culicivorax IN SECOND INSTAR MOSQUITO LARVAE OF Anopheles quadrimaculatus AND Culex pipiens AS OBSERVED WITH THE SCANNING ELECTRON MICROSCOPY Anopheles quadrimaculatus

Penetration sites

C&x

pipiens

NO. nematodes

%

NO. nematodes

%

Head” Cervix” Thorax= Abdomen” Intersegmental membranesb

13 A 6A 18 A 54 B 10

14.3 6.6 19.8 59.3 11.0

7A 2A 4A 31 B 4

15.9 4.5 9.1 70.5 9.1

Total

91’

44’

Note. Duncan’s multiple range test was used. Within each species data with different letters are significantly different. a Includes attached and penetrating nematodes. b Sum of nematodes penetrating through cervix, between meso- and metathorax, and intersegmental membranes of abdomen. ’ Fifteen A. quadrimaculatus and 12 C. pipiens larvae were infected by 91 and 44 nematodes, respectively.

boring, host immobilization, and cuticle penetration. In this study we found that the preparasites could not detect the host from relatively remote distances. Even in the presence of a mosquito larva the nematodes swam randomly around in an orthokinetic manner. A specific behavioral pattern was developed only when the nematodes were ca. 1 mm from the host. The first phase has been designated as host detection. After this point, the nematodes stopped their body movement but resumed within a few seconds in a klinotactic orientation to approach the host. This phase has been designated as orientation to the host. In other insect parasitic nematodes the direction of movement of infective stages is often influenced by chemical stimuli; e.g., the migration of Steinernema feltiae through sand is greatest in the presence of a host (Georgis and Poinar, 1983). The same nematode was attracted along a chemical and bacterial gradient (Pye and Burman, 1981). The pronounced response of infective juveniles of S. feltiae to CO, suggests that this compound aids host finding (Gaugler et al., 1980). However, there is no concrete evidence for host specificity based on chemical attractants (Webster and Dunphy, 1987). The second step toward successful pene-

R. culicivorax

AND

LARVAL

TABLE EFFECT

OF Romanomermis

culicivorax quadrimaculatus,

197

MOSQUITOES

3

ON THE HEART RATE IN SECOND INSTAR Culex pipiens, AND Aedes aegyptib

LARVAE

Heart rate (beats/mm k SD) Mosquito species

N”

Before penetration

Anopheles quadrimaculatus Culex pipiens Aedes aegypti

30 30 30

93.9 k 7.4A 90.1 2 15.3A 93.1 + 7.6A

OF Anopheles

During penetration

Duration of cardiac arrest (set -t SD)’

48.6 f 4.3B 44.8 2 11.8B 48.1 ” 4.9B

93.4 + 18.4A 61.7 + 21.4B 99.1 * 12.OA

n Number of mosquito larvae observed for each species. b Duncan’s multiple range test was used. Data in columns with different letters are significantly different. ’ Different set of data not comparable to the heart rate data. Duncan’s multiple range test was used.

tration is the orientation to the host. In the present study, there was no clear differentiation between the end of the host detection phase and the start of the orientation phase. Nematode chemotaxes are generally regarded to be klinotactic responses, as indicated in studies by Klinger (1965) and Ward (1973). The behavioral response of R. culicivorax juveniles to the host from relatively close distances is also attributed to klinotactic orientation, since attracted nematodes were regularly observed to make side-to-side head motions as if comparing attractant intensity. The same behavior was observed in S. feltiae juveniles responding to CO, (Gaugler et al., 1980), and this agrees with Croll’s (1970) definition of klinotaxis. The third step of penetration by R. culicivorax begins with the secretion of an adhesive material from the anterior region of the nematode body, during which the attachment and exploration start. A similar behavior was described earlier in T. sciarea, a sphaerulariid nematode, infecting a mycetophilid fly. T. sciarea lives in a terrestrial habitat and the published observations showed that penetration behavior started with the exploration phase followed by the secretion of an adhesive mass which could be considered an attachment phase (Poinar and Doncaster, 1965). In Hydromermis, the preparasites produced an adhesive surface coat that enabled them to

stick to the surface of Chironomus larva and prepare an invasion hole (G&z, 1976). Observations on Gastromermis sp., a mermithid nematode infecting blackflies, and on Hydromermis contorta, a mermithid infecting chironomid midges, are similar to our findings. In Gastromermis the nematodes were observed affixing their mouths to the substrate to resist water currents while waiting to find and penetrate larvae of Simulium damnosum (Mondet and Poinar, 1976). According to the observations of Gotz (1976) and on the basis of our results, it is presumed that those mermithids are also capable of secreting an adhesive material which enables them to adhere to the surface of the aquatic host larvae. Another function could be attributed to the adhesive material secreted by R. culicivorax. When the infected mosquito larvae were fractured immediately after penetration, SEM examinations showed no bacterial cells inside the body cavity. This suggests a second function of the nematode adhesive material as a cleansing or surface sterilizing factor for the nematode cuticle during the penetration of the host. This would prevent sepsis and, hence, selectively enhance the successful completion of the parasitic phase. In R. culicivorax the attachment phase was followed by a search-boring phase. In contrast the attachment in T. sciarae was followed by spear thrusting and salivary se-

198

SHAMSELDEAN

cretions which ended with the boring phase (Poinar and Doncaster, 1965). Presumably by the end of the boring phase, esophageal injections in Romanomermis was the cause of temporary immobilization and cardiac arrest of the host. The infective stages of most aquatic mermithids are relatively large in comparison to the host, therefore, the injection of anesthetics would seem vital for successful penetration. The suggested function for such injections is to facilitate cuticle penetration by reducing extraneous host activity that might dislodge the penetrating nematode. Although anesthesia was not observed during penetration by T. sciarea, successful penetration in Paramermis contorta, a mermithid nematode infecting chironomid larvae, required a certain immobility as well as a softening of the cuticle of the host (Comas, 1927). In Gastromermis rosea, Wulker (1965) described the injection of a toxin by the infective stages during the penetration process into the body of their host which caused a temporary paralysis and facilitated the penetration by the nematodes. A similar behavior was observed in blackflies penetrated by the mermithid nematode Neomesomermis jlumenalis. In the latter case, the infected host larvae appeared partially paralyzed with the body often in an atypical contorted position (Molloy and Jamnback, 1975). Whereas mosquito larvae penetrated by R. culicivorax recovered in a matter of minutes, blackfly larvae infected with N. jlumenalis slowly resumed normal movement and feeding within an hour. The final step toward a successful host entrance is cuticle penetration. Comas (1927) observed juveniles of P. contorta penetrating the cuticle of Chironomus rhummi larvae. The site of penetration in the latter mermithid was the anal foot of the young host larvae (Svabenik, 1928). Both Romanomermis iyengari and Agamermis decaudata were observed penetrating the cuticle of their hosts. Whereas the whole process from the time of attachment to the time of complete entry took from 1 to 2 min

AND PLATZER

in R. iyengari (Iyengar, 1927), it was variable for A. decaudata since it completed penetration sometimes in 1 to 2 min and at other times in 5 to 10 min (Christie, 1936). In comparison, our data show that the preparasites of R. culicivorax accomplish the same task in less than 9 min. The observations presented here have established a framework for further analysis of the initial parasite-host interface in mermithids and aquatic arthropods. At this interface, R. culicivorax exhibits a hexapartite response: host detection only at short range, subsequent orientation to the host, attachment, search-boring, host immobilization, and host cuticle penetration. Subsequent investigations are in progress to define further the mechanisms developed and adapted by mermithids to successfully achieve the transition from the external to the internal host milieu. ACKNOWLEDGMENTS We thank Dr. Abubaker A. Othman and Arnold Bell for their assistance in development of the SEM procedures. M. M. Shamseldean expresses his appreciation for support by the Amideast Peace Fellowship Program for Egypt. This project was supported in part by Research Grant AI-15717 from NIAID, U.S. Public Health Service, National Institute of Health.

REFERENCES BEDDING, R. A. 1972. Biology of Deladenus siricidicola (Neotylenchidae) an entomophagousmycetophagous nematode parasitic in siricid woodwasps. Nematologica, 18, 482-493. BEDDING, R. A., AND MOLYNEUX, A. S. 1982. Penetration of insect cuticle by infective juveniles of Heterorhabditis spp. (Heterorhabditidae: Nematoda). Nematologica, 28, 354-359. CHRISTIE, J. R. 1936. Life history of Agamermis decaudata, a nematode parasite of grasshoppers and other insects. .I. Agric. Res., 52, 161-198. COMAS, M. 1927. Sur le mode de penetration de Paramermis contorta. V. Linst. dans la larvae de Chironomus rhummi Kief. C. R. Sot. Biol. Paris, 9, 673-675. CROLL, N. A. 1970. “The Behavior of Nematodes: Their Activity, Senses and Responses.” Arnold, London. EISENBACK, J. D. 1985. Techniques for preparing nematodes for scanning electron microscopy. In “An Advanced Treatise on Meloidogyne,” Vol. II,

R. culicivorax AND LARVAL “Methodology” (K. R. Barker et al., Eds.), pp. 79105. Department of Pathology, North Carolina State University, Raleigh. GAUGLER, R., LEBECK, L., NAKAGAKI, B., AND BOUSH, G. M. 1980. Orientation of the entomogenous nematode Neoaplectana carpocapsae to carbon dioxide. Environ. Entomol., 9, 649-652. GEORGIS, R., AND POINAR, G. O., JR. 1983. Effect of soil texture on the distribution and infectivity of Neoaplectana glaseri (Nematoda: Steinernematidae). .I. Nematol., 15, 32s332. G~Tz, P. 1976. Parasit-Wirt-Beziehungen zwischen dem Nematoden Hydromermis contorta und der Zuckmucke Chironomus thummi. Publ. Wiss. Filmen, 9, 67-87. IYENGAR, M. 0. T. 1927. Parasitic nematodes of Anopheles in Bengal. Far East. Assoc. Trap. Med., Trans. 7th Congr. Brit. India, 3, 128-135. KLINGER, J. 1965. On the orientation of plant nematodes and of some other soil animals. Nematologica, 11, 4-18. MOLLOY, D., AND JAMNBACK, H. 1975. Laboratory transmission of mermithids parasitic in blackflies. Mosq. News, 35, 339-342. MONDET, B., AND POINAR, G. O., JR. 1976. Recent developments on mermithid parasitism of Simulium damnosum in West Africa. In “Proceedings of the 1st International Colloquium on Invertebrate Pathology and IXth Annual Meeting, Society for Invertebrate Pathology, Queen’s University at Kingston, Canada,” pp. 232-235. OBIAMIWE, B. A., AND MACDONALD, W. W. 1973. A new parasite of mosquitoes, Reesimermis muspratti sp. nov. (Nematoda: Mermithidae), with notes on its life-cycle. Ann. Trop. Med. Parasitol., 67, 439+$4. PETERSEN, J. J. 1973. Relationship of density, location of hosts, and water volume to parasitism of larvae of the southern house mosquito by a mermithid nematode. Mosq. News, 33, 516-520. PETERSEN, J. J. 1985. Nematodes as biological control agents: Part I. Mermithidae. Adv. Parasitol., 24, 307-344.

MOSQUITOES

199

PETERSEN, J. J., AND CHAPMAN, H. C. 1970. Parasitism of Anopheles mosquitoes by a Gastromermis sp. (Nematoda: Mermithidae) in southwestern Louisiana. Mosq. News, 30,420-424. PHELPS, R. J., AND DEFOLIART, G. R. 1964. Nematode parasitism of Simuliidae. Univ. Wisconsin Madison. Res. Bull., 245. PLATZER, E. G., AND STIRLING, A. M. 1978. Improved rearing procedures for Romanomermis culicivorax. Proc. Calif. Mosq. Vect. Cont. Assoc., 46, 87-88. POINAR, G. O., JR. 1968. Hydromermis conopophaga n. sp., parasitizing midges (Chironomidae) in California. Ann. Entomol. Sot. Amer., 61, 593-598. POINAR, G. 0.. JR. AND DONCASTER, C. C. 1%5. The penetration of Tripius sciarae (Bovein) (Sphaerulariidae: Aphelenchoidae) into its insect host, Bradysia paupera Tuom. (Mycetophilidae: Diptera). Nematologica, 11, 73-78. PYE, A. E., AND BURMAN, M. 1981. Neoaplectana carpocapsae: Nematode accumulations on chemical and bacterial gradients, Exp. Parasitol., 51, 13-20. SNEDECOR, G. W., AND COCHRAN, W. G. 1980. “Statistical Methods,” 7th ed. Iowa State Univ. Press, Ames. STRICKLAND, E. H. 1911. Some parasites of Simulium larvae and their effects on the development of the host. Biol. Bull., 21, 302-338. SVABENIK, J. 1928. Aus dem Leben der Paramermis contorta (V. Linstow). Zool. Anzeiger, 77, 259-267. WARD, S. 1973. Chemotaxis by the nematode Caenorhabditus elegans: Identification of attractants and analysis of the response by use of mutants. Proc. Natl. Acad. Sci. USA 70, 817-821. WEBSTER, J. M., AND DUNPHY, G. B. 1987. Host compatibility of insects to nematodes. In “Vistas on Nematology” (J. A. Veech and D. W. Dickson, Eds.), pp. 237-245. Society of Nematologists, Inc., Hyattsville, Maryland. WALKER, W. 1965. Der Mechanismus des Eindringens parasitarer Mermithiden (Nematoda) in Chironomus-larven (Dipt., Chironomidae). Z. Parasite& 26, 2-9.