The removal of populations of plants can have significant consequences that depend on the extent and speed of the removal (Thayer et al. 1990). The killing of large amounts of aquatic vegetation can significantly affect the ecosystem, especially if the killed plants are left to decay in the water. This is especially true if the percentage of the biomass to total volume of the lake is great (Thayer et al. 1990). If so, the decomposing plant material can use up all dissolved oxygen in the water and kill fishes. The ecological changes in a water body as the vegetation is slowly removed reveal the changes that must have occurred in the ecosystem as the vegetation originally increased. The major concern is that improper control methods can remove all or most vegetation and reduce the ecosystem's ability to support native fish species. Excessive removal of plants can also affect birds and terrestrial animals that use aquatic habitats.
The food preferences of grass carp frequently do not match the needs for aquatic-plant control by the managers for a particular water body. Target plant species may rank lower in preference by grass carp than the more valuable plants, and management of the undesirable plants takes place only after the more valuable plants are consumed. Sometimes, the removal of rooted aquatic plants by the grass carp triggers the flow of carbon and energy through the plankton populations and creates an algae bloom (Hichling 1966; Michewichewicz et al. 1972; Canfield et al. 1983). Stanley (1972) and others found that the proper level of stocking water bodies with grass carp can increase the yield of native fishes. Leslie et al. (1987) determined how much vegetation should be removed for a specific level of control.
Taylor et al. (1984) discussed the results of evaluations of the use of grass carp to control aquatic--usually nonindigenous--nuisance plants. Canfield et al. (1983) discussed the possible effects on ponds in Florida from the removal of hydrilla with grass carp. We were unable to find any evidence that the grass carp are established in Florida. Additional information on the ability of this species to control aquatic vegetation is discribed later under the section entitled "Biological Agents other than Insects". Bain (1993) developed a set of guidelines for managers for the gathering of basic data to assess the significant effects of grass carp on large systems. He concluded that the effects would increase in relation to the percentage of the affected vegetation and that maximum changes would be related to complete removal of the vegetation. Fine tuning of the level of removal by grass carp has been difficult (Shireman and Haller 1982). Either the stocking is inadequate to remove much of the vegetation in a given timeframe or the stocking rate exceeds the desirable level of control and all vegetation is removed (Shireman et al. 1980, 1991). The U.S. Army Corps of Engineers (Theriot and Decell 1978) attempted to develop a computer model to predict the correct size of grass carp stock for Florida's Lake Conway. A population of female grass carp was released into Lake Conway in 1977 (Theriot and Decell 1978; Lazor 1983). A comprehensive study to measure the effect of the hydrilla reduction on the flora and fauna in the lake was initiated.
Introductions of grass carp as weed control agents can be advantageous. They are less expensive than mechanical removals, chemical applications, water level manipulations, or other biological agents (Shireman and Smith 1983), and the results usually last longer than results from other methods (Decell 1975; Rottmann 1977). Chemical treatment may be faster, but slower removal of the vegetation by grass carp avoids depletion of oxygen from the decomposition of dead plants, which may cause fish kills (Thayer et al. 1990). Native biological control agents have also been evaluated (Leslie et al. 1987). Native nematodes attack hydrilla in Florida (Esser et al. 1985; Gerber and Smart 1987), but no exhaustive surveys for nematodes have been made in the native range of hydrilla (Buckingham and Habeck 1990).
The present method for appropriate levels of weed control in Florida is the treatment of infested areas with an herbicide to return the weed to an acceptable level (Shireman et al. 1980, 1991). Applications of herbicide are followed with lighter stockings with grass carp to control regrowth of the nuisance plants. Shireman et al. (1980, 1991) discussed the problems from stocking with too many grass carp.
