Non-indigenous invasive species are causing ever increasing economic and environmental damage to the Caribbean Region, including Florida. Two species of non-native pest mole crickets, Scapteriscus vicinus Scudder in Florida and S. didactylus (Latreille) in Puerto Rico, are among the most damaging. Their tunneling and feeding cause significant economic losses to pastures, golf courses, sod farms, sports fields, residential lawns, and vegetable farms. In Puerto Rico, S. didactylus was recognized long ago as a general pest of agriculture, and the loss of forage for livestock grazing on pastures severely infested with S. vicinus has become a serious problem for beef and dairy cattle in Florida. For decades, the worst pest insects on the 20 plus golf courses in Puerto Rico and 1400 or more in Florida have been mole crickets. Puerto Rico and the southern states suffer in excess of $200 million annually in damage and control costs due to invasive mole crickets (Frank & Parkman 1999). However, this situation is changing as populations of biological control agents increase, spread, and become effective in suppressing these pests (Frank & Walker 2006).

Chemical controls are available for mole crickets in lawns, golf courses, and sod farms (Buss et al. 2002) but can be expensive and are not always effective, in addition to incurring environmental risks and providing only short-term relief. Public concern has been increasing because of the possible overuse and misuse of chemicals on turf and the potential for ground water contamination (FDEP 2002). Some golf course and sports complex managers have attempted to minimize the use of insecticides due to local restrictions and concern about the health of their members. This has led to unacceptable levels of damage to turf as a result of mole cricket tunneling and feeding. Chemical control in pastures is prohibitively expensive (Frank & Parkman 1999). Consequently, cattlemen, turf producers, golf course managers, and others have been seeking a cost-effective, environmentally safe, and permanent means of minimizing mole cricket damage.

This project was conducted to advance biological control of invasive mole crickets, Scapteriscus spp., in Florida and Puerto Rico by fully deploying and evaluating an entomopathogenic nematode, Steinernema scapterisci (Nguyen and Smart), and a parasitic wasp, Larra bicolor F. (Fig. 1). The nematode was introduced into Florida from Uruguay in the 1980s, commercialized in 2001, and applied on bahiagrass in at least half of the 67 Florida counties (Adjei et al. 2006). Nematode applications were also made at several sites in Puerto Rico from 2001 to 2004. The nematode is effective, multiplies within infected mole crickets, and spreads rapidly as the insects disperse (Parkman et al. 1993).

The wasp was imported from Bolivia and released at Gainesville, Florida in 1988 (Frank & Sourakov 2002). It became established and is spreading (Frank et al. 1995) but has not yet colonized all of Florida (Frank & Walker 2005). The adult wasps can fly and distribute themselves but they require a nectar source, especially the wildflower, Spermacoce verticillata (L.). This wildflower is widespread in southern Florida (Wunderlin & Hansen 2003) and considered an unimportant weed (Murphy et al. 1998). Plants growing in road margins and turf are typically small and produce few flowers because of mowing and the use of herbicides. Nevertheless, they probably are capable of sustaining sparse L. bicolor populations because they can flower all year. In northern Florida, the plant is reported only from Alachua and St. Johns counties (Wunderlin & Hansen 2003); and it freezes to the ground in winter (Arévalo & Frank 2005).

Since L. bicolor was introduced into Puerto Rico from Brazil in the late 1930s (Wolcott 1938), it could have colonized the entire country (Castner & Fowler 1984) but it is not conspicuous. Its primary nectar source, S. verticillata, is native to Puerto Rico (Liogier 1982) and most abundant in sandy places near the coasts. Urbanization, especially along the coasts, has diminished S. verticillata habitats and the plant is typically viewed as a weed and destroyed. Wolcott (1941a, 1941b) speculated that, if S. verticillata were permitted to grow as a beneficial plant, L. bicolor would be more abundant and widespread. Regardless, even unmanaged populations of the wasp have achieved partial biological control of pest mole crickets (Cruz & Segarra 1992), a vast improvement over the situation in the early 1900s when mole crickets were considered to be the most serious general pests of agriculture in Puerto Rico (Zwaluwenburg 1918).

While significant research has been conducted on S. scapterisci and L. bicolor, and their effectiveness in controlling mole crickets has been established (Frank et al. 2005), additional work was needed to more efficiently produce, distribute, and evaluate the impact of these natural enemies in new locations (Leppla et al. 2004). The objectives of the current project conducted in Florida and Puerto Rico are to (1) continue trapping mole crickets and determining the infection levels and spread of S. scapterisci, (2) plant plots of S. verticillata in as many locations as possible to provide nectar for adult L. bicolor, (3) accelerate the natural spread of the wasps by capturing them from established populations and moving them to new sites, and (4) educate collaborators, extension agents, master gardeners, cattlemen, golf course superintendents, and the general public about the use and benefits of S. scapterisci and L. bicolor for biological control of invasive mole crickets.

Materials and Methods

The S. scapterisci nematode product, Nematac®S, was produced by MicroBio, UK, under an exclusive license for its use from the University of Florida, the patent holder, and provided by Becker Underwood, Ames, Iowa ( http://www.beckerunderwood.com). The nematodes were shipped directly to the University of Florida or the University of Puerto Rico (UPR) at Mayagüez and stored for no more than 1 month at about 10°C. As recommended by Becker Underwood, nematodes were mixed in water and applied as an amount equivalent to about 2.5 billion per ha. Applications were made without irrigation, early in the morning or late in the afternoon to avoid direct sunlight.

Nematode applications were made throughout much of Florida in 2000 and 2001 by injecting them 1.5 cm below the soil surface with a modified slit-seeder (Adjei et al. 2005). To determine the degree of spread, nematodes were applied in 6.08 m wide by 41.3 m long strips covering 1/2, 1/4, or 1/8 of 0.4-ha plots distributed across a 10-ha bahiagrass pasture during Sep 2000 in Polk County. Pitfall traps used to capture mole crickets for estimating levels of infection and impact of the nematode (Hudson 1989) were installed in each plot: 3 in nematode-treated strips, 3 in untreated strips, and 6 in the untreated control plots. More than 600 mole crickets were trapped in the plots for 3 months prior to applying the nematodes and none were found to be infected. The percentage of bahiagrass in the pasture was recorded as green, yellow, or dead (brown, bare ground, or weeds). This was accomplished by randomly tossing a m² quadrat with 100 subdivisions 5 times in each plot.

In Puerto Rico, nematodes were applied for the first time ever in Nov 2001 and again in Aug 2003 on an irrigated golf course with clay loam soil near Aguadilla and at an organic vegetable farm in Aguada. Commercial spray equipment was used at the golf course to apply nematodes on 0.5-acre plots. A pitfall trap was installed near the center of each of 5 treated and 5 untreated plots. At the vegetable farm, a hand-held watering container was used to sprinkle nematodes on five 2 × 7-m plots. Applications were made early in the morning at both locations followed by enough irrigation to soak the surface of the soil.

Mole crickets from all sites in Florida and Puerto Rico were pitfall-trapped at 7-d intervals. Additional 24-h collections were made biweekly in Florida to assure that the mole crickets were alive when taken to the laboratory. Regardless of source, captured mole crickets were provided with food and water, held until dead, and assayed for presence of nematodes (Woodring & Kaya 1988). Effectiveness and spread of the nematodes was evaluated by determining the percentage of trapped mole crickets that were infected and by monitoring an associated decline in the number of mole crickets.

To detect and redistribute the wasp, S. verticillata plants were grown at sites in Florida and Puerto Rico. Wasps destined for release in Florida were captured at S. verticillata plots in the Gainesville area, chilled, and transported to the new sites. Releases were made no earlier than dusk on the day of capture, and each site received 24-30 presumably mated females. Wasp populations were highest in Sep through Nov, so these were the most appropriate months for the adults to be captured. In Puerto Rico, small S. verticillata plants collected from an overgrown pasture at Playa Anasco were transplanted into several plots at the golf course and vegetable farm where the nematodes had been applied. These plots provided nectar sources to attract and retain the wasps. Plants were also planted within 1-2 meters of each of 5 pitfall traps at 5 other locations in Nov-Dec 2004. The plants had been dug from sandy soil as small plants in Jul and grown for 3-4 months in pots containing wet peat to recover from transplantation. They were placed in the field to determine the presence of wasps.

Results and Discussion

During spring 2001, after the nematodes had been applied in the Florida pasture during the previous fall, 80% of the 328 trapped mole crickets were infected, including those from both the treated and untreated plots. Infected mole crickets thus spread the nematodes across the entire 10-ha pasture in less than 1 year. This indicated that nematodes can be introduced in strips receiving about 0.3 billion infective juveniles per ha to avoid much more expensive broadcast applications of 2.5 billion per ha. Three years later, the proportion of infected mole crickets stabilized at 20-35%, the number of mole crickets trapped declined by 85% (Fig. 2), and the bahiagrass cover increased by 40-95% (Adjei et al. 2006).

In Puerto Rico, S. didactylus, was trapped at both sites where nematodes were applied and relatively small numbers of S. abbreviatus were discovered at the organic vegetable farm. Mole crickets captured at the golf course in Aguadilla contained steinernematid nematodes but those from the vegetable farm were not infected. Persistence of infected mole crickets at the golf course for more than 7 months indicated that the nematode became established. According to research conducted in Florida, the nematode is not likely to persist for more than 10 weeks in soil without passing through host mole crickets (Nguyen & Smart 1989). At Aguadilla, more than 225 mole crickets were trapped between Nov 2001 and Oct 2002, and some of these insects were infected every month until the end of Jun 2002 when assays ceased. A total of 28 mole crickets were infected. Thus, we determined that S. scapterisci infected and killed S. didactylus mole crickets in Puerto Rico

Parasitism of S. vicinus by L. bicolor averaged about 24% at 2 locations near Gainesville, Florida throughout the fall seasons of 1999 and 2000. Larra bicolor has several annual generations, whereas S. vicinus has but one, so the average generational mortality of the mole cricket may approach 70% (Frank & Walker 2005). Due to this initial success, wasps were released in fall 2002 in northern Florida at Quincy in Gadsden County and Palmetto in Manatee County. They were recovered in both counties during May and Nov 2004. Wasps have not yet been captured following releases in fall 2004 in Bay and St. Lucie counties. The wasps spread naturally to Bradford, Pinellas, and Sumter counties during the summer of 2004 and to Hernando County by Oct 2004. In all, the wasps were reported from 22 counties by Jun 2005. Wasps observed at S. verticillata plots in Puerto Rico have not yet been confirmed as L. bicolor.

Larra bicolor feeds at nectaries of plants in addition to S. verticillata and could have used them to disperse widely in northern Florida. However, the duration of visits to S. verticillata compared with 4 other plants, Conoclinium coelestinum (L.) DC, Elephantopus elatus Bertol., Passiflora coccinea Aubl., and Solidago fistulosa Mill., indicated that S. verticillata is significantly more attractive (Arévalo & Frank 2005). Consequently, plots of S. verticillata are being maintained in Duval, Hillsborough, and Pasco counties where the wasp is assumed to be present but has not been detected. Extension agents and master gardeners in 30 counties have each been provided with at least one potted S. verticillata plant to attract the wasps.

Cattle, landscape, turf and allied industries, and the general public in Florida and Puerto Rico will benefit greatly by having long term, cost-effective biological control that minimizes the damage caused by invasive mole crickets. Establishing natural enemies that recycle, leaving only periodic pest resurgence in isolated areas, can eliminate most of the damage caused by these pests. Nematac®S nematodes persist in irrigated clay loam soils, spread rapidly, induce high levels of infection, and kill pest mole crickets. The nematodes can be applied in strips from which they will move to fill the gaps, thus reducing the cost to 1/8 as much as broadcast applications. Hopefully, the wasp will function additively with the nematode to reduce mole cricket populations. The wasps will forage for mole crickets more effectively if S. verticillata plants are available as a source of nectar. The long-term financial and environmental costs of this biological approach to mole cricket management will be much less than continued reliance on chemical insecticides.