76 J. Aquat. Plant Manage. 41: 2003. J. Aquat. Plant Manage. 41: 76-81 Combining Plant Pathogenic Fungi and the Leaf-Mining Fly, Hydrellia pakistanae, Increases Damage to Hydrilla YASSER M. SHABANA 1 , J. P. CUDA 2 , AND R. CHARUDATTAN 3 ABSTRACT Four fungal species, F71PJ Acremonium sp., F531 Cylindro- carpon sp., F542, Botrytis sp., and F964 Fusarium culmorum [Wm. G. Sm.] Sacc. were recovered from hydrilla [Hydrilla verticillata (L. f.) Royle] shoots or from soil and water sur- rounding hydrilla growing in ponds and lakes in Florida and shown to be capable of killing hydrilla in a bioassay. The iso- lates were tested singly and in combination with the leaf-min- ing fly, Hydrellia pakistanae (Diptera: Ephydridae), for their capability to kill or severely damage hydrilla in a bioassay. For the fungus plus insect treatments, two ready-to-hatch H. paki- stanae (HP) eggs were placed on hydrilla shoot in each assay tube. The shoots were rated for insect damage 21 days later and grouped into two damage levels, 15% (HP1) and 25% (HP2). The insect-damaged hydrilla shoots were then ex- posed to each fungus added to the water in the assay tube at a final concentration of 6 × 10 4 to 1 × 10 6 propagules per ml. In two of the fungus-insect combinations (HP2 plus Acremoni- um sp. and HP2 plus F. culmorum), the level of damage on hy- drilla was increased in comparison with the damage caused by either agent alone. Among the isolates tested alone, the F. culmorum isolate was the most effective. The maximum damage to the shoots by this isolate was achieved at 20 to 30 C as compared to 15 or 35 C. The synergistic effect of com- bining F. culmorum and H. pakistanae was 1.22 to 1.56 times greater, respectively, for the two herbivory levels of HP1 and HP2, than the effect of the fungus alone. The damage levels were 3.21 to 2.78 times greater for the F. culmorum plus H. pa- kistanae treatments compared to the insect alone at the re- spective levels of herbivory. Thus, the combined use of this pathogenic fungus and the fly appears a promising approach for integrated control of hydrilla. Key words: Hydrilla verticillata, Hydrellia pakistanae, Fusarium culmorum, Acremonium sp., Botrytis sp., fungi, biocontrol, in- sect-pathogen synergism, integrated control. INTRODUCTION Hydrilla is a submersed freshwater macrophyte that be- longs to the family Hydrocharitaceae. It is widely distributed in the African-Asian region (Pieterse 1981) and has become one of the most invasive weeds of waterways in tropical and subtropical regions of the world (Langeland 1996). Hydrilla was first reported on the west coast of Florida in 1958 (Black- burn et al. 1969). Since then, it has spread throughout Flori- da and to many other states (Langeland and Burks 1998). Two factors enable hydrilla to out-compete other submersed aquatic plants: it thrives under low light conditions and it pro- duces an abundance of viable vegetative propagules (Bowes et al. 1977). Hydrilla forms dense surface mats that can se- verely reduce water flow, interfere with boating, water sports, fishing, and navigation. It can also significantly reduce the wa- ter holding capacity of storage ponds. From 1980 to 1993, ap- proximately $39 million was spent in Florida to manage hydrilla in the state’s public waters (Schardt 1997), mainly for chemical herbicides containing copper, diquat, endothall, or fluridone as an active ingredient (Langeland 1996). Biological control of hydrilla has been a high priority in Florida since the 1970s. Among the biocontrol agents released to control hydrilla is Hydrellia pakistanae Deonier (Diptera; Ephydridae [HP]), a small fly native to tropical and temperate regions of Asia (Deonier 1993). The leaf-mining larvae of this fly cause extensive damage to hydrilla (Baloch and Sana-Ullah 1974, Deonier 1978, Baloch et al. 1980, Buckingham et al. 1989). Host-range tests conducted in Pakistan (Baloch and Sa- na-Ullah 1974) and in Florida (Buckingham et al. 1989) dem- onstrated that H. pakistanae is highly specific to hydrilla. The fly was released in Florida in 1987, but to date its population density has never exceeded more than 15 adults per m 2 of hyd- rilla stands and the level of damage has not been more than one-fifth the level estimated to be necessary to produce a sig- nificant impact on the plant (Wheeler and Center 2001). Several plant pathogens have been discovered and shown to be capable of killing hydrilla under controlled conditions (Charudattan 1990, Verma and Charudattan 1993, Shearer 1998, Nachtigal and Pitelli 1999, Shabana et al. 2003). A mul- ticomponent, integrated control approach rather than a sin- gle control tactic offers the best prospect for long-term management of aquatic weeds (Pieterse 1977; Charudattan 2001). In this respect, the potential to exploit synergistic in- teractions between insects and plant pathogens or a patho- gen and a herbicide has been suggested as an option (Charudattan et al. 1978, Charudattan 1986, Netherland and 1 Corresponding author: Plant Pathology Department, Faculty of Agricul- ture, Mansoura University, El-Mansoura, Egypt, ymsh@mans.edu.eg & yass- ershabana@hotmail.com. 2 Entomology and Nematology Department, University of Florida, Gainesville, FL, USA. 3 Plant Pathology Department, University of Florida, Gainesville, FL, USA. Received for publication October 16, 2002 and in revised form June 11, 2003.