« Back

Evaluation of Lagenidium Giganteum for Biocontrol of Florida Mansonia Mosquitoes


Executive Summary

Many unreclaimed pits and settling ponds created by the phosphate mining industry have become infested with waterhyacinth, Eichhornia crassipes (Mart.) Solms., and waterlettuce, Pistia stratiotes L. (Haeger 1979, Lounibos and O’Meara 1982, Morris et al. 1986). The prolific growth of waterhyacinth and waterlettuce associated with these excavations not only displaces more desirable native vegetation but the extensive mats formed by these floating plants produce enormous populations of mosquitoes which threaten the public health and comfort (Haeger 1979, Morris et al. 1986). Specifically, the immature stages of important disease-vectoring Mansonia mosquitoes are dependent upon these aquatic plants for their survival. Unlike other immature mosquitoes which breathe at the surface of the water and are susceptible to chemical controls, the larvae and pupae of Mansonia mosquitoes are unique because they remain attached to the roots of waterhyacinth and waterlettuce to presumably obtain oxygen (Wesenberg-Lund 1918) or escape predation (Van Den Assem 1958, Lounibos et al. 1992). Except for chemical or mechanical plant removal, no environmentally safe yet effective methods have been developed for controlling the cryptic aquatic stages of these pestiferous mosquitoes (Lounibos and O’Meara 1982).

Through a cooperative effort involving the U. S. Dept. of Agriculture, U. S. Army Corps of Engineers, Florida Dept. of Environmental Protection and the University of Florida, several host specific insects have been introduced into Florida from South America which provide substantial control of waterhyacinth and waterlettuce under certain conditions (Cofrancesco 1991). Although biological control of these plants may offer the most promising long-term solution to both the weed and Mansonia problems by reducing the density of the host plants upon which the mosquitoes breed, these weed biocontrol agents do not provide rapid control of the plants in all situations (Cofrancesco 1991). Consequently, control of waterhyacinth and waterlettuce in the mining area of central Florida is still accomplished mainly with herbicides such as 2,4-D, diquat and glyphosate. Unfortunately, there is increasing evidence to suggest that these herbicide applications provide only temporary control and may also be detrimental to weed biocontrol agent populations (Grodowitz and Cofrancesco 1990, Haag and Buckingham 1991). Furthermore, because the herbicides are applied directly to plants growing in the water the perception that these compounds are a health threat has been created by the fact that many surface water systems in central Florida are closely interconnected with the underlying ground water system through springs and sinkholes (Southwest Florida Water Management District, unpublished report). For example, the surficial aquifer is highly susceptible to ground water contamination due to the shallow depth to the water table and high recharge rate.

Faced with society’s growing concern over pesticide contamination in surface and ground water supplies, it is imperative that innovative management strategies for Mansonia mosquitoes be developed which contribute minimally to the water quality problem, are cost effective and are also compatible with the biological agents purposefully introduced into Florida to control waterhyacinth and waterlettuce.

Lagenidium giganteum Couch is a naturally occurring entomopathogenic fungus that attacks a broad spectrum of mosquito species and appears to be restricted to this group (McCray 1985, Hornby et al. 1992). The non-discriminating host range within the insect family Culicidae, or mosquitoes, results from the ability of the fungus to differentiate between the cuticle of mosquito larvae and other aquatic animals (Kerwin et al. 1991). Upon application to a breeding site, the fungus is activated in water where motile biflagellate zoopores are produced. Unlike bacterial larvicides which must penetrate the foliage and be ingested by the larvae, the free-swimming zoospore actively seeks out its larval mosquito host and infects resident larvae. One to three days after infection by L. giganteum, the mosquito larvae die, new spores are produced and larva-to-larva transmission continues.

We experimentally tested the effect of inoculative releases of the mosquito-specific fungus Lagenidium giganteum Couch on a population of Mansonia dyari during a typical breeding season at one site in eastern Hillsborough County.Mansonia dyari was selected as the target species for this investigation because it is the most abundant species on waterlettuce (Slaff and Haefner 1985), which formed almost a pure stand on the surface of the unreclaimed phosphate pit at the project site.

The temporal distribution of Ma. dyari in the central Florida phosphate mining region has been described in detail elsewhere (Lounibos and Escher 1985, Slaff and Haefner 1985) but is briefly summarized here. There is a peak emergence of adults in the spring and another in the late summer and early fall. Adult emergence ceases during the winter months due to the cooler water temperatures and plant mortality. In general, the fall emergence is greater than in the spring. Unlike other mosquitoes which have a relatively short larval/pupal period, the aquatic stages of Ma. dyari have a protracted life cycle. Data on the closely related Ma. titillans (Haeger 1960) indicates development from egg to adult occurs in approximately six weeks.

We introduced L. giganteum (California isolate), cultured in the laboratory on yeast extract, dextrose and egg yolk, into outdoor caged replicated test pools containing water lettuce and larvae of Ma. dyari collected from an inactive phosphate pit. We added susceptible larvae (first and second instars) twice a week between August and December 1993 to simulate natural oviposition by Ma. dyari. We also pumped fresh water into the test pools from the phosphate pit twice a week (August 1993 to March 1994 ) to prevent stagnation and replenish nutrients extracted by the water lettuce plants. We collected data on weather at the project site, water temperature in one of the test pools, and water quality in the test pools and inactive phosphate pit.

The data from the emergence trap samples and sentinel larvae indicated that Ma. dyari was highly susceptible to L. giganteum at the field application rates of 400 and 800 ppm. Adult emergence was reduced by more than 77 % in comparison to untreated pools. Sentinel larvae provided evidence of recycling of the fungus approximately 14 days post-treatment. We also observed continuous mortality of larval Ma. dyari for a period of 46 days after the second pool inoculation in November 1993 when the water temperature in the test pools did not exceed 38°C, which is lethal to the zoospores. Except for the high water temperature we observed in the test pools prior to the first inoculation (September), our data showed water quality in the test pools and phosphate pit was suitable for fungal growth and zoosporogenesis for the remainder of the study period.