With few exceptions, knowledge of environmental limitations is restricted to that developed for R. culicivorax, first, because this parasite is easily manipulated and can be field tested, and second, because no other mermithid species have been investigated to the point that this type of information is important for their use as biological control agents.

1. Physicalfactors

The effects of temperature on development and parasitism are fairly well documented for a number of mermithid species. However, specific temperature limitations become critical when a nematode is to be used to control host populations. Romanomermis culicivorax, naturally found only in the southern United States, is active only when temperatures are above 15'C (Petersen, 1973b). Brown and Platzer (1977) reported parasitism in the laboratory at 12'C; and Galloway and Brust (1977) reported only limited parasitism in field releases at IO'C and concluded that the use of R. culicivorax for mosquito control in colder temperature zones is precluded. Mermithids vary greatly in their tolerance to low temperatures as certain species parasitize snowmelt mosquitoes and have been isolated from mosquitoes as far north as the Arctic Circle (Frohne, 1955).

The effect of photoperiod on parasitism by R. culicivorax remains controversial. Brown and Platzer (1974) reported that mosquito larvae were more susceptible to preparasites when kept in continuous darkness. However, Galloway and Brust (1977) reported that photoperiod had no significant effect on infectivity. They also observed that there was no interaction between the temperature and the photoperiod. In a more recent study, Sharma and Gupta (1982a) determined that there was a definite interaction between the photoperiod and temperature, and in contrast to Brown and Platzer (1974), reported that the highest rate of larval infection occurred in continuous light at ambient temperature.

For aquatic mermithids, water movement can be an important factor. Romanomermis culicivorax was ineffective in releases against pasture mosquitoes in California because irrigation water was continually flowing through the field and the preparasites were apparently washed away before they could make contact with the host (Hoy and Petersen, 1973). Also, attempts to parasitize simuliid larvae with the mosquito parasite R. culicivorax in running water environments were ineffective. Limitations associated with water movement were further substantiated in field releases of R. culicivorax at a lake in El Salvador. Releases made prior to heavy wave action on the lake resulted in levels of parasitism of about half those when releases were made following heavy wave activity (Petersen et al., 1978b).

Desiccation is not a serious limiting factor for aquatic mermithids. Although the nematodes and eggs cannot tolerate desiccation even for a few minutes, there is sufficient moisture even when the habitat dries out to maintain the parasite populations until the habitat again becomes flooded. In contrast, moisture is a major factor for terrestrial mermithids. The highest levels of parasitism by A. decaudata and M. nigrescens are always associated with persistently moist environments (Mongkolkiti and Hosford, 1971; Christie, 1936).

2. Chemicalfactors

Petersen and Willis (1970) found that R. culicivorax was inhibited by mild salinity (0.04 M NACI). This was confirmed by Brown and Platzer (1978a) who also ranked ion toxicity for R. culicivorax for the following ions: cations, sodium < potassium < calcium; and anions, chloride < carbonate = sulfate < nitrate < nitrite < phosphate. Therefore, R. culicivorax is ineffective for mosquito control in habitats under conditions of increased water salinity. A similar tolerance was shown for R. iyengari (Bheema Rao et al., 1979). In contrast, the mosquito mermithid 0. muspratti has been shown to be tolerant of diluted seawater (3000-4000 Amhos cm-') and water from tree holes (10,000 timhos cm-') (Petersen, 1981). The mermithid Perutilimermis culicis develops in Aedes sollicitans, a salt marsh mosquito, and therefore has adapted to tolerate very high salinity.

Chen (1976) reported that the optimum pH for infection by R. culicivorax was 6.7-7.2. In contrast, recent studies showed that aquatic mermithids tolerate a broad pH range. Brown and Platzer (1978a) reported that parasitism occurred over a pH range of 3.6-8.6. Similarly, Petersen (1979a) showed that host mosquitoes were readily infected at all pH concentrations tested (5.4-7.9) and that infection increased at the lower pH ranges. More recently, Sharma and Gupta (1982b) reported a limited tolerance for higher pH levels and concluded that R. culicivorax did not appear suitable for biological control of Culex quinquefasciatus because of the high pH characteristic of habitats of this mosquito.

Mermithids have a high tolerance for many agricultural chemicals. Mitchell et al. (1974) reported that levels of Abate, Dieldrin, and GammaHCH, normally used then for mosquito control, did not adversely affect host parasitism by R. culicivorax. Levy and Miller (1977a) reported similar tolerance by R. culicivorax to four pesticides and a growth regulator. Finney et al. (1977) found that Altosid 5E, an insect growth regulator, did not interfere with parasite development, and that host mortality increased when a combination of Altosid and R. culicivorax was used in laboratory experiments. Also, copper-based organic algacides and copper sulfate did not compromise the infectivity of R. culicivorax at concentrations used for algae and weed control (Platzer and Brown, 1976).

Brown and Platzer (1978b) showed that transient exposure to low oxygen tension increased the survival and infectivity of preparasites of R. culicivorax. However, Platzer (1981) found that preparasites of this mermithid stopped moving within 8 hours in water rich in organic content and low in oxygen content. This explained, at least in part, why R. culicivorax has proved ineffective against mosquitoes in polluted environments. Mermithid species also vary in their tolerance to pollution. Octomyomermis muspratti has been shown to tolerate much higher levels of organically rich tree- hole water than R. culicivorax (Petersen, 198 1), again demonstrating the diversity in tolerance of aquatic mermithid species.

3. Biological limitations

Mermithids, with few exceptions, parasitize the immature stages of their hosts. The age of the immature host influences the susceptibility of that host to attack by some mermithid species. Petersen and Willis (1970) showed that second-instar larvae of Cx. quinquefasciatus were most readily invaded by R. culicivorax, third instars were slightly less susceptible than first instars and limited parasitism occurred in fourth-instar larvae. Host age can greatly influence the outcome of attempts to control fast growing floodwater mosquitoes with mermithids because some floodwater mosquito species can develop to a stage where control with mermithids is'ineffective within 48 hours during midsummer temperatures.

Research suggests that predators and parasites of mermithids can reduce their effectiveness especially in aquatic environments. Mitchell et al. (1974) were the first to report the predation of preparasites of R. culicivorax by ostracods. Platzer and MacKenzie-Graham (1978) have since shown that the preparasitic stage is also preyed upon by copepods, young gammarids, diving beetles, dragonfly and damselfly naiads, and small crayfish. In the laboratory, mermithid parasite populations were reduced significantly by copepod densities of 20-100 liter -1 (Platzer and MacKenzie-Graham, 1980). The authors concluded that copepods may play a significant role in the success or failure of field applications of mermithids.

Pathogens may also limit populations of both aquatic and terrestrial mermithids. Nolan (1977) reported the presence of a phycomycete and hyphomycete in M. flumenalis and suggested that the occurrence of fungi might affect the success of this mermithid for the control of simuliids. A highly virulent chytridiomycetous fungus, Cantenaria anguillulae, reached epizootic proportions in laboratory cultures of R. culicivorax, but was eliminated by rearing the mosquito hosts in acidified water (Sterling and Platzer, 1978). Recently a soil hyphomycete was reported attacking eggs of 0. muspratti in the laboratory (Platzer, 1982b). The impact of pathogens and parasites of mermithid nematodes in nature remains unknown.