The redbay ambrosia beetle (RAB), Xyleborus glabratus Eichhoff (Coleoptera: Curculionidae: Scolytinae), is an invasive pest of Asian origin and vectors Raffaelea lauricola T.C. Harr., Fraedrich & Aghayeva (Ophiostomatales: Ophiostomataceae), a fungus that causes laurel wilt in plants of the family Lauraceae (Crane et al. 2008; Fraedrich et al. 2008; Hanula et al. 2008). In the USA in natural hammocks of Georgia, Florida, and North and South Carolina (Hanula et al. 2008; Mayfield et al. 2008), laurel wilt has decimated redbay, Persea borbonia (L.) Spreng. (Laurales: Lauraceae), and swamp bay, P. palustris Sarg. The disease is now present in commercial avocado (P. americana Mill.) orchards of south Florida (FDOACS 2012). Enhancing the disease effect, other species of ambrosia beetles (i.e., Xylosandrus crassiusculus [Wood], Xyleborinus saxeseni [Reitter]), appear to be responsible for the horizontal spread of the pathogen (Carrillo et al. 2014).

Research to control the spread of the beetle has focused on the beetle's biology (Brar et al. 2012), seasonality (Brar et al. 2013), host preferences (Kendra et al. 2014a), development of attractants (Kendra et al. 2012b, 2014b; Kuhns et al. 2014a,b), repellents (Peña et al. 2011), and chemical control of the vector (Peña et al. 2011; Carrillo et al. 2013), but little effort has been directed toward determining the presence or absence of natural enemies of X. glabratus. Studies on natural enemies of Scolytinae have provided mixed results. For instance, Wood (1982) reported that natural enemies efficiently caused reduction of more than 90% of a bark beetle brood in a given generation whereas other reports focus on the potential of parasitoids and predators against important pests such as the coffee berry borer, Hypothenemus hampei Wood & Bright (Coleoptera: Curculionidae: Scolytinae) (Vega et al. 1999; Perez-Lachaud et al. 2002), without an evaluation of their effectiveness. Although bark and ambrosia beetles fill a slightly different niche and differ somewhat in their biology, they do have other attributes in common and may share natural enemies. For the purposes of this work, we considered all cases of scolytine predator/ prey relationships.

The diversity and richness of key natural enemies of X. glabratus across the Florida landscape or in its areas of origin (i.e., Japan, India, Myanmar, Taiwan, and Bangladesh) are unknown (Rabaglia et al. 2006). During a survey conducted during 2010 in Taiwan by the senior and junior authors (JEP and SSL) X. glabratus was only collected in Cinnamomum osmophloeum Kanehira (Laurales: Lauraceae) but not on other reported original host trees such as Lindera latifolia Hook. fil., Litsea elongata (Nees) Bentham & Hooker, Phoebe lanceolata (Nees) (Laurales: Lauraceae), Shorea robusta Gaertner (Malvales: Dipterocarpaceae), Lithocarpus edulis (Makino) Nakai (Fagales: Fagaceae), and Leucaena glauca L. (Fabales: Fabaceae), indicating that RAB densities are not prominent in that area (Peña & Lu, unpublished data). Natural enemies could be a biological factor regulating X. glabratus populations in Taiwan and thereby limiting the host range in the pest's native range.

Finding natural enemies of bark and ambrosia beetles can be a complex task due to the concealed nature of the insects. Direct observations of predators or parasitoids attacking the beetle are difficult. The ambrosia beetle may also be rare in an area, thereby compounding the difficulty in finding any associated natural enemies. In addition, multiple scolytines and other insects may also be infesting field-collected wood making it difficult to determine what insect or group of insects were the target of a particular natural enemy. Placing laboratory-reared insects in the field (called the sentinel system) is one technique to lure the natural enemies to specifically attack a target pest. This technique has been quite useful in locating natural enemies in a wide range of host/prey systems (Meurisse et al. 2008). One of the advantages of this portable technique includes the ability to place infested logs in an array of environments. However, there are disadvantages as well. For instance, the sentinel system might lack enough prey to provide the important cues that natural enemies use to locate their prey. It also is unknown whether an artificially cut small log may attract most natural enemies seeking prey.

A search for natural enemies should include a variety of techniques in order to help obtain the most diverse and comprehensive collection of potential natural enemies. The objectives of this study were to determine the natural enemy community associated with X. glabratus in Florida and Taiwan using 3 different methods: 1) field-collected wood naturally infested with X. glabratus and other scolytines (Florida); 2) bolts of avocado artificially infested with X. glabratus in the laboratory (Florida); and 3) direct collection of natural enemies attracted to a trap area baited with infested wood and known X. glabratus lures (Florida and Taiwan).

The purpose of this work was to explore the possibility of finding natural enemies to help control the scolytine vectors of the fungus that leads to laurel wilt disease. The surveys in Taiwan centered on classical biocontrol. However, most efforts were focused on finding native or previously introduced natural enemies. A classical biocontrol program takes long to implement, and the threat of this disease is an immediate concern. Finding an endemic or naturalized predator or parasitoid may make a big difference in saving the avocado industry of Florida, which is a US$ 13 million industry (FDOACS 2012).

Materials and Methods

NATURAL ENEMIES FROM LABORATORY-INFESTED AVOCADO LOGS

Approximately forty 25 to 30 cm bolts of P. americana ‘Booth 7' were cut in Miami-Dade County during fall 2010 and spring, summer, and fall 2011. Bolts were infested with 30 adult male and female X. glabratus taken from a colony reared at the UF Entomology and Nematology Department (Gainesville, Florida, USA), then held in the laboratory at 25 ± 2 °C under complete darkness for 4 wk (Brar et al. 2013). The infested X. glabratus bolts were hung approx. 1 m above the ground in shady bay stands (Persea spp.) located in Alachua (29°44.84′N, 82°12.60′W), Brevard (28°13.006′N, 82°10.2024′W), Highlands (27°25.35′N, 81°9.42′W), Osceola (29°24.445′N, 82°09.756′W), St. Luc ie (27°25.576′N, 80°24.354′W), Miami-Dade (25°43.37′N, 80°28.12′W), and Indian River (27°35.285′N, 80°24.410′W) Counties, all known to have established populations of X. glabratus. The bolts were left for 1 mo to allow additional colonization by X. glabratus, by other scolytines, and by natural enemies. Exposed bolts were then collected and held in 15 × 45 cm cardboard containers (Single Wall Air Sea Containers, Miami, Florida, USA) at the UF (Gainesville, Florida, USA) indoor insectary (25 ± 2 °C; 75-80% RH) and at the Indian River Research and Education Center in Fort Pierce, Florida, USA, at similar temperature and humidity conditions. Sixty days were given for insects to develop and emerge in the sealed containers. After this time, adult insects (scolytines and/or natural enemies) were collected from all interior surfaces and grooves of cardboard containers, and on the surface and under the bark of bolts and stored in 70% ethanol until identification. Second generation X. glabratus adults emerge from bolts approx. 60 d after the first infestation (Brar et al. 2013); therefore, exposed bolts were considered to host cohorts of different ages and possibly the parent adults. Insects collected from the bolts were sent to the taxonomic identification resources mentioned above for identification.

Table 1.

Hymenoptera and Coleoptera associated with Persea spp., trees infected with R. lauricola and Raffaelea spp.

Results

USING BAIT STATIONS TO ATTRACT NATURAL ENEMIES OF XYLEBORUS GLABRATUS IN FLORIDA

Four hymenopteran species were collected using attractants. Those included Leptacis sp. (n = 1) and Telenomus sp. (both Platygastridae) (n = 1) in Miami-Dade County and Prosacantha sp. (n = 4), Telenomus sp. (n = 1) (Platygastridae), and Anteon xanthothorax (Dryinidae) (n = 1) in Highlands County.

Table 2.

Arthropods associated with Persea americana logs infested with the red bay ambrosia beetle in different Florida counties in 2010–2012. The date indicates when the logs were deployed.

Continued

Continued

Discussion

Parasitoids. Among the collected parasitoids from this survey, we report Tetrastichus (Eulophidae), Heterospilus, and Chelonus (Braconidae), which were noted by Wood (1982) as genera that parasitize Scolytinae. We also report Trisacantha spp. (Platygastridae: Teleasinae). However, this genus has not previously been reported from Scolytinae. Because the other collected parasitoids were identified only to the family level, i.e., Bethylidae, Platygastridae, Pteromalidae, their identification to the genus or species level is necessary in order to determine whether they could be associated with the Scolytinae reported in this survey. Wood (1982) reported among the Hymenoptera containing parasitoids of scolytines the families Braconidae, Bethylidae, Chalcididae, Diapriidae, Encyrtidae, Eulophidae, Eupelmidae, Eurytomidae, Gasteruptiidae, Ichneumonidae, Mymaridae, Podagrionidae, Proctotrupidae, Pteromalidae, and Torymidae. Cephalonomia stephanoderis and Prorops nasuta (Bethylidae) were reported as parasitoids of Hypothenemus hampei (Perez-Lachaud et al. 2002), but their importance as mortality factors is open for discussion. However, Vega et al. (1999) reported that C. stephanoderis was the most abundant coffee berry borer parasitoid in Cote d'Ivoire, followed by Phymasticus coffeae (Eulophidae). Moreover, Boone et al. (2009) reported that from a survey of hymenopterans attracted to logs colonized by Ips pini Woods & Bright (Coleoptera: Curculionidae: Scolytinae) in western Montana, pteromalids were the most important group related to Scolytinae, but less than 1% of the trapped specimens were associated with Scolytinae. These lists and examples are of dubious worth because they are primarily parasites of the less-hidden bark beetles. There is not much information available concerning ambrosia beetle parasites, but the previously listed families and genera may be good places to start.

Predators. Wood (1982) also cited the coleopteran families Cleridae, Colydiidae (= Zopheridae), Cucujidae, Elateridae, Histeridae, Cybocephalidae, Rhizophagidae (Monotomidae), Staphylinidae, and Trogositidae as reported by Chamberlin (1939) as major predators of Scolytidae. Here we report undetermined species of the genera Microsicus (Zoopheridae) and Laemophloeus as possible predators associated with arthropods inhabiting the logs. For instance, Laemophloeus juniperici (Grouvelle) has been cited as the most important predator of Phloeotribus scarabeoides (Bern.) (Scolytinae) in olive groves (Gonzalez & Campos 1990), and Leptophloeus near punctatus was observed preying on H. hampeii larvae in Cote d'Ivoire (Vega et al. 1999). Among the Thysanoptera, we report A. ramachandrai, which has not been associated with Scolytinae but as a predator of insects causing galls (Boyd & Held 2006). The genus Europs (Monotomidae) is also reported. Whereas we could not find reports of this genus as a predator of Scolytinae, it is important to note that other Monotomidae (Rhizophagus grandis Gyllenhal) have been reported as specific predators of the great spruce bark beetle, Dendroctonus micans Erichson (Coleoptera: Curculionidae) (Meurisse et al. 2008).

Problems arose by using laboratory-infested avocado bolts. First, these bolts desiccated quickly, which might have increased beetle mortality. Second, avocado is not one of the preferred hosts of X. glabratus, nor is it a host in which the beetle reproduces in high numbers as it does in P. palustris (Brar et al. 2012). For example, the numbers of X. glabratus collected from the set of sentinel logs per site averaged 10 and ranged from 0 to 34. These 2 factors might have influenced the results. Therefore, we suggest that in other surveys, other Persea species, such as P. borbonia and P. palustris should be used as sentinel logs.

Use of attractants proved to be a fast and easy method of parasitoid collection. However, as with the other studies, this method does not offer the assurance that the parasitoids caught are indeed parasitoids of Scolytinae.

It is necessary to determine whether native parasitoids are key mortality factors of X. glabratus in its areas of invasion. Also, if similar surveys are conducted in those states that were invaded earlier by X. glabratus (i.e., Georgia, South Carolina), a richer natural enemy fauna than the one found in Florida, which was invaded during 2007, will likely be found. Surveys of natural enemies associated with X. glabratus in its different areas of “origin” in Asia should be conducted in order to determine the existence or lack of any key mortality factor.