IN RELATION TO PLANT DISEASE Potential of Contact to Control Xyleborus glabratus (Coleoptera: ), a Vector of Laurel Wilt Disease in Avocados

1 DANIEL CARRILLO, JONATHAN H. CRANE, AND JORGE E. PEN˜ A

Tropical Research and Education Center, University of Florida, 18905 SW 280 Street, Homestead, FL 33031

J. Econ. Entomol. 106(6): 2286Ð2295 (2013); DOI: http://dx.doi.org/10.1603/EC13205 ABSTRACT Xyleborus glabratus Eichhoff (Coleoptera: Curculionidae: Scolytinae) is an invasive ambrosia that vectors laurel wilt, a new disease that threatens avocado and other species in the

Lauraceae Family. The lethal concentrations (LC50&90) of nine commercial insecticides to X. glabratus were determined by using a bolt-dip bioassay. Different formulations of , per- methrin, , z- ϩ bifenthrin, l- ϩ thiamethoxam, , , , and were tested. Four concentrations of each were tested (0.5, 0.1, 0.03, and 0.01 of the label rate) and with water as a control. were exposed to treated bolts and mortality registered 48 h later. After 2 wk, bolts were destructively sampled to determine the number of beetles that constructed galleries and were alive inside the wood. Probit analysis was

used to determine the LC50&90. Six pesticides were applied directly to the trunk and limbs of avocado trees in a commercial grove. Limbs of treated trees were cut weekly after the application and exposed to X. glabratus to determine the number of beetles boring into the logs. The toxicity of pesticides to X. glabratus was greatly reduced 2 wk after application. Among the tested pesticides, malathion and z-cypermethrin ϩ bifenthrin provided the best suppression of X. glabratus. Among the insecticides registered for use in avocado, fenpropathrin and malathion were the most effective in protecting trees from attack by X. glabratus. Other pesticides that are currently not registered for use in avocados could be useful for managing this ambrosia beetle.

KEY WORDS redbay ambrosia beetle, exotic , insectÐdisease complex, Lauraceae, chemical control

The redbay ambrosia beetle, Xyleborus glabratus Eich- 35Ð100 cm above the ground, mostly between 1600 and hoff (Coleoptera: Curculionidae: Scolytinae), is an 1800 hours, and observed two peaks of trap catches invasive species that vectors a phytopathogenic fun- (MarchÐApril and October). Maner (2012) reported gus, Raffaelea lauricola T.C. Harr., that causes laurel that symptoms characteristic of laurel wilt develop in wilt, a lethal disease of several plant species within redbay trees with as few as two X. glabratus entry the Lauraceae, including avocado (Persea americana holes. Little is known about the life history of X. Miller), redbay (Persea borbonia (L.) Spreng), glabratus in avocado. MayÞeld et al. (2008) evaluated swampbay (Persea palustris (Raf.) Sarg.), and sassafras young potted avocado trees in no-choice tests and (Sassafras albidum (Nutt.) Nees; Fraedrich et al. 2008, found boring of X. glabratus and transmission of the Pen˜ a et al. 2012). The spread of X. glabratus has af- laurel wilt pathogen. Kendra et al. (2011) found that fected large areas of native Lauraceae trees in the avocado wood volatiles were attractive to dispersing southeastern United States (Fraedrich et al. 2008) and female X. glabratus in natural areas infested with this is now threatening the avocado industry in south Flor- beetle. In addition, Brar et al. (2013) studied the life ida (Crane et al. 2008; Carrillo et al. 2012, 2013). Most cycle of X. glabratus in logs of avocado, redbay, and studies about the life history of X. glabratus have been swampbay. They found comparable rates of develop- conducted in natural areas with large stands of Lau- ment in the three hosts, but fewer progeny produced raceae trees, primarily redbay and swampbay. In a 2-yr in avocado. study in South Carolina and Georgia, X. glabratus The detection of X. glabratus in commercial avo- adults were active throughout the year, with peak cado groves is quite recent, and it is unclear if it will activity in September (Hanula et al. 2008). In north follow the same life-history patterns observed in hosts Florida conditions, Brar et al. (2012) reported that in natural areas. MiamiÐDade County is the main largest number of beetles was trapped at heights of (Ͼ95%) commercial avocado production area in Flor- ida. During February 2010, one X. glabratus female 1 Corresponding author, e-mail: dancar@uß.edu. was collected by FDACS-CAPS (Florida Department

0022-0493/13/2286Ð2295$04.00/0 ᭧ 2013 Entomological Society of America December 2013 CARRILLO ET AL.: CONTACT INSECTICIDES AGAINST X. glabratus 2287 of Agriculture and Consumer Services, Cooperative cides against X. glabratus; their Þrst approach used a Agricultural Pest Survey) personnel from a trap in “hanging bolt” technique to test insecticides applied as MiamiÐDade County. The trap was placed in a natural a barrier treatment to prevent beetles from boring into area Ϸ11 miles north of the avocado production area. avocado bolts. In that study, the contact insecticides One year later, laurel wilt was detected in native z-cypermethrin ϩ bifenthrin and l-cyhalothrin ϩ thia- swampbay trees only a few miles from the initial beetle methoxam provided the most consistent suppression detection. In February 2012, the Þrst avocado tree in of Scolytinae in areas infested by X. glabratus. The a commercial grove located in the northeastern quad- hanging bolt technique is limited in that it does not rant of the avocado growing area was diagnosed with allow identiÞcation of beetles attacking the bolt. R. lauricola (Florida Department of Agriculture and Thus, a detailed study to determine the effectiveness Consumer Services [FDACS] 2012). As of July 2013, of different insecticides speciÞcally on X. glabratus is 90 trees have been have diagnosed R. lauricola posi- needed. tive, and Ͼ1,900 symptomatic trees have been re- In the current study, controlled laboratory bioas- moved as part of a suppression and sanitation strategy says and a Þeld trial were conducted to test the efÞ- (J. H. Crane, personal communication). Interestingly, cacy and persistence of several insecticides applied as only six X. glabratus individuals have been captured in barriers to prevent X. glabratus from boring into host commercial avocado groves (A. Derksen and D. Car- trees. A range of insecticides from different chemical rillo personal observation). Because of the lack of groups were tested against X. glabratus, including in- alternative pest management strategies (e.g., biologi- secticides that are registered for use on avocado or cal control, repellents, etc.), private landowners and recommended for management of other wood-boring avocado producers rely on applications of chemical . The speciÞc objectives were 1) to determine insecticides to complement sanitation practices and the lethal concentrations (LC50&90) of commercial protect trees in groves affected by this beetleÐdisease insecticides to X. glabratus, and 2) to determine the complex. persistence of selected insecticides to manage X. Most chemical control strategies against scolytines glabratus in avocados, under normal Þeld conditions have focused on those species that are pests of co- during the summer rainy season in south Florida. nifers or ornamental plants. Carbaryl, , cyßuthin, , , chlorpyrifos, Materials and Methods and/or bifenthrin have been used to manage pine bark beetles (Hall et al. 1982, Shea et al. 1984, Haverty et al. Insects. All X. glabratus used in the bioassays 1998, Hastings et al. 2001, DeGomez et al. 2006, Fettig emerged from logs cut from swampbay (P. palustris) et al. 2006, 2013). insecticides and chlor- trees affected by laurel wilt. Collection sites were pyrifos have been proposed to manage the native elm swampy natural areas in Flagler County (29Њ 36Ј47.48Љ , Hylurgopinus rufipes (Eichhoff; Co- NÐ81Њ 13Ј39.41Љ W) during November 2011, MiamiÐ leoptera: Curculionidae), and the smaller European Dade County (25Њ 43Ј37.96Љ NÐ80Њ 28Ј36.16Љ W) dur- elm bark beetle, multistriatus (Marsham), ing February 2012, and Highlands County (27Њ vectors of the Dutch elm disease (Gardiner and Web 12Ј53.94Љ NÐ81Њ 20Ј49.38Љ W) during August 2012. In 1980, Lanier et al. 1984, Phillipsen et al. 1986, Pajares these sites, trees showing signs of X. glabratus infes- and Lanier 1989, Jin et al. 1996, Jin and Webster 1997, tation (small strings of compacted sawdust protruding Oghiakhe and Holliday 2011). Much less is known from the bore holes along the trunk of the tree) were about insecticide efÞcacy for management of bark and cut and measured, and wood Ͼ10 cm in diameter was ambrosia beetles in fruit crops. Saeed et al. (2011) stored inside emergence chambers (165 liters, Brute evaluated the toxicity of several insecticides against container 2643Ð60, Rubbermaid, Peoria, IL) with a the mango bark beetle, Hypocryphalus mangiferae mason jar placed in a hole in one of its sides to collect Stebbing, and found that among the contact insecti- emerging beetles. Wood collected from Flagler cides tested, chlorpyrifos had greater efÞcacy than County was transported and held at the Department and bifenthrin. Lozano et al. (2001) sug- of Entomology and Nematology, University of Florida, gested that deltamethrin can provide effective pro- Gainesville, FL, whereas the wood collected from tection to olive trees from attack by the olive bark MiamiÐDade and Highland Counties was transported beetle, Phloeotribus scarabaeoides (Bernard). and held at the Containment Facility of the University Very few studies have investigated chemical control of Florida, Tropical Research and Education Center of ambrosia beetles. Mizell and Riddle (2004) re- (TREC), Homestead, FL. Wood was kept at photo- ported that high rates of bifenthrin provided better period of 14:10 (L:D) h, 80% relative humidity (RH), suppression of the granulate ambrosia beetle, Xylosan- and 25ЊC. X. glabratus adult females that emerged from drus crassiusculus (Motschulsky), than other tested the wood were collected daily and placed inside petri insecticides. Reding et al. (2013) reported that per- dishes (5 cm in diameter) provided with a moistened methrin and bifenthrin provided better protection to Þlter paper. All healthy X. glabratus female beetles nursery plants from attack by Xylosandrus germanus used in the bioassays had less than2dofemergence (Blandford) and X. crassiusculus. In addition, endo- from the wood. sulfan has been used against the coffee berry borer, Host Plant Material. Limbs (7Ð10 cm in diameter) Hypothenemus hampei (Ferrari; Damon 2000). Pen˜ aet of healthy and nonsprayed ÔLulaÕ avocado trees from al. (2011) initiated efforts to identify effective pesti- an avocado orchard at TREC were cut with a chainsaw 2288 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 106, no. 6 and divided into smaller bolts (7Ð9 cm in diameter by 10 cm in length) by using a table saw (DeWalt DW713). All plant material was cut 1 d before each bioassay. Lethal Concentration of Contact Insecticides on X. glabratus. Nine insecticides, including products reg- (ml/liter) istered for use in avocado, and others recommended Bioassay concn for use against other wood boring insects, were tested in the bioassays (Mizell and Riddle 2004; Reding et al. 2013). Each insecticide was tested in a serial dilution where the starting solution was half the recommended label rate for Þeld applications, assuming a standard volume of water of 935 liters/ha. For each pesticide,

Þve concentrations (treatments) were tested: 0.5, 0.1, (ml/liter) 0.03, 0.01, and 0 (control) of the label concentration. Label concn Each treatment was replicated Þve times. The full label rates for Þeld applications were not included in the lethal concentration determinations because these are usually higher than the concentrations used in 0.6 0.62 0.31, 0.06, 0.020, 0.001, 0

laboratory toxicological bioassays. The tested pesti- 2.3Ð11.7 2.50 1.25, 0.25, 0.075, 0.003, 0 (liter/ha) cides (including trade name, manufacturer, chemical Label rates class, active ingredient, application rate, label rate, and bioassay concentrations) are listed in Table 1. Avo- cado bolts were dipped for5sinthedifferent pesticide solutions or in water (control) and air-dried at ambi- ent temperature for 24 h. Treated bolts were placed individually inside plastic containers with a screen lid for ventilation (11 cm in diameter by 14.5 cm in height, Instawares APCTR32) and stored at the TREC Con- tainment Facility (26ЊC, 70 Ϯ 10% RH, 14:10 [L:D] h). Ten X. glabratus females were placed on the bark of each bolt and allowed to interact with the treated bolts. After 48 h, the numbers of dead and live beetles, and the number of beetles that bored into the bolts, were recorded. Two weeks later, bolts were opened to determine the number of beetles that constructed galleries and alive or dead inside the bolts. Persistence of Selected Insecticides Against X. Cary, NC NC Arysta LifeScience North America Corporation, Valent USA Corporation, Walnut Creek, CAMicro Flo Company, Memphis, TN 1.2Ð1.6DuPont, Wilmington, DE 1.25 1.7Ð5.2 0.62, 0.13, 0.04, 0.001, 0 1.88 1.7Ð3.5 0.94, 0.19, 0.06, 0.002, 0 1.88 0.94, 0.19, 0.06, 0.002, 0 glabratus Under Field Conditions. Six insecticides were selected to test their residual efÞcacy protecting avocado from X. glabratus infestations under Þeld con- * ditions. The trunk of ÔPetersonÕ avocado trees were (including class/a.i., trade name, manufacturer, label rate, label concentration, and bioassay concentrations) * sprayed (May 2012). The insecticide solutions were * prepared by using the highest label concentration of * each pesticide (Table 2) and a standard application

volume of 935 liters H2O/ha. The treatments were X. glabratus

applied with a handgun sprayer calibrated to deliver HeroEndigo ZC FMC Syngenta Corporation, Crop Philadelphia, Protection, PA Greensboro, NC 0.37Ð0.44 0.75 0.43 0.87 0.21, 0.04, 0.01, 0.0004, 0 0.44, 0.09, 0.026, 0.001, 0 3.8 liters of solution per tree, directing the application to the tree trunk and limbs. Twenty-one trees were

treated; three with each insecticide and three with ) are registered for use in avocado. *

water (control). Each treated tree was separated by at ϩ least three untreated trees. On days 4, 8, 15, and 22 after application, Þve ran- domly selected limbs per tree (Ϸ50 cm in length and thiamethoxam

7Ð10 cm in diameter) were cut with a chainsaw and ϩ subsequently cut into two smaller bolts (8Ð9 cm in diameter by 10 cm in length). The bolts were placed inside a plastic container with a screen lid for venti- Class/a.i. Trade name Manufacturer lation (11 cm in diameter by 14.5 cm in length) and kept at the TREC Containment Facility. Ten X. glabra-

tus females were placed on the bark of each bolt. After Table 1. Commercial pesticides tested against bifenthrin (11.25%) (12.6%) For each pesticide, ÞveTrade concentrations names (treatments) followed were by tested: asterisks 0.5, ( 0.1, 0.03, 0.01, and 0 (control) of the label concentration. Pyrethroid/bifenthrin (25.1%)Pyrethroid/permethrin (36.8%) Brigade 2EC Permethrin 3.2AG FMC Corporation, Philadelphia, PA 0.88Ð2.3 0.94 0.47, 0.1, 0.03, 0.001, 0 Pyrethroid/z-cypermethrin (3.75%) /chlorpyrifos (44.9%)/carbaryl (44.1%) Lorsban 4E Sevin XLR Micro Flo Company, Memphis, TN CropScience, Research Triangle Park, 3.5 1.88 0.94, 0.19, 0.06, 0.002, 0 48 h, the numbers of dead and live beetles were re- Pyrethroid/fenpropathrin (30.9%)l-Cyhalothrin (9.48%) Organophosphate/malathion (56%) Danitol 2.4 EC Methomyl (99.3%) Malathion 5EC Lannate LV December 2013 CARRILLO ET AL.: CONTACT INSECTICIDES AGAINST X. glabratus 2289

Table 2. Insecticides applied to avocado trees at the highest the highest acute toxicity on X. glabratus among the label rate to determine the persistence of X. glabratus control under tested insecticides (Table 3). The LC of chlorpyrifos field conditions 50 was similar (as indicated by 95% FLs overlap) to z-cypermethrin ϩ bifenthrin and fenpropathrin but a.i. Trade name liter/ha ml/liter H2O was signiÞcantly lower than malathion, permethrin, Bifenthrin Brigade 2EC 1.8 1.87 bifenthrin, and l-cyhalothrin ϩ thiamethoxam (Table Permethrin Permethrin 3.2AG* 0.6 0.62 3). Carbaryl and methomyl had signiÞcantly higher Fenpropathrin Danitol 2.4 EC* 1.6 1.66 z-cypermethrin ϩ bifenthrin Hero 0.8 0.87 LC50s; consequently, these two insecticides had the low- ϩ l-cyhalothrin thiamethoxam Endigo ZC 0.4 0.43 est acute toxicity on X. glabratus. Similarly, the LC90 of Malathion Malathion 5EC* 5.6 5.63 chlorpyrifos was signiÞcantly lower than the other in- secticides, followed by malathion and z-cyperme- Trade names followed by asterisks (*) are registered for use in ϩ avocado. thrin bifenthrin. The estimated LC90s of malathion and z-cypermethrin ϩ bifenthrin were similar to those corded. Fifteen days later, the bolts were destructively of fenpropathrin, permethrin, and bifenthrin, but sig- sampled to determine the number of beetles that niÞcantly lower than l-cyhalothrin ϩ thiamethoxam bored through the bark and were found alive. Each (Table 3). Carbaryl and methomyl had the highest treatment was replicated Þve times. Rainfall during LC90s. The LC50&90of chlorpyrifos, malathion, this experiment was recorded through the Florida z-cypermethrin ϩ bifenthrin, bifenthrin, permethrin, Automated Weather Network (FAWN), which has a fenpropathrin, and l-cyhalothrin ϩ thiamethoxam meteorological station located Ϸ200 m away from the were within the range of the tested concentrations experimental site. (Tables 1 and 3). By contrast, the estimated lethal Data Analysis. Lethal concentrations 50 and 90 concentrations of carbaryl and methomyl were out of (LC ) were calculated 48 h after beetles were the range of the tested concentrations and were 50 & 90 exposed to each pesticide by using the SAS-PROBIT higher than the label concentrations (Tables 1 and 3). procedure (SAS Institute 2012). AbbottÕs transforma- Contact with fresh residues of most insecticides tion was used to correct for control mortality (Abbott caused death on X. glabratus in a concentration-de- 1925), which was usually Ͻ10%. SigniÞcant differ- pendent fashion (Fig. 1). Contact with chlorpyrifos ences between lethal concentrations were indicated resulted in signiÞcantly higher mortality of beetles when the 95% Þducial limits (FLs) of one pesticide did exposed to the three highest concentrations (␹2 ϭ not overlap with the FLs of the other pesticides. Be- 20.81; P Ͻ 0.001) compared with all others tested. cause of variance heterogeneity and non-normality of Exposure to the two highest concentrations of mala- data, beetle mortality and beetle boring into bolts thion caused signiÞcantly higher mortality than - treated with the various concentrations of each pes- tles exposed to the two lowest concentrations, which ticide were analyzed with KruskalÐWallis tests (SAS had mortality rates similar to the untreated control (␹2 Institute 2012). ϭ 19.07; P Ͻ 0.001; Fig. 1). Contact with fresh residues Beetle mortality and boring values in the insecticide of bifenthrin, permethrin, fenpropathrin, z-cyperme- persistence Þeld trial were normally distributed (Kol- thrin ϩ bifenthrin, and l-cyhalothrin ϩ thiamethoxam mogorov P Ͼ 0.005) and analyzed through repeated- at the highest tested concentration (i.e., 0.5 Ϸ half of measures analysis of variance. Differences among the the label rate) resulted in 100% death of X. glabratus treatments were detected through TukeyÕs range tests within 48 h (Fig. 1). However, beetle mortality de- (SAS Institute 2012). creased signiÞcantly (␹2 ϭ 22.16, ␹2 ϭ 20.77, ␹2 ϭ 14.69, ␹2 ϭ 21.86, ␹2 ϭ 19.37; P Ͻ 0.001; respectively) when exposed to lower insecticide concentrations Results (Fig. 1). No effect of the insecticide concentration Lethal Concentration of Contact Insecticides on X. was observed in the mortality of beetles exposed to ␹2 ϭ ϭ glabratus. Chlorpyrifos showed the lowest LC50 com- fresh residues of carbaryl ( 4.21; P 0.37). Ex- pared with the other insecticides, and therefore had posure to the two highest concentrations of methomyl

Table 3. Lethal concentrations (LC50&90) at 48 h after X. glabratus adults were exposed to avocado bolts treated with nine pesticides at four different concentrations

␹2 Pesticide LC50 (ml/liter) LC50 FL (ml/liter) LC90 (ml/liter) LC90 FL (ml/liter) Slope Bifenthrin 0.08 0.03Ð0.42b 0.25 0.11Ð91.2bcd 9.15 2.59 Permethrin 0.07 0.03Ð0.26b 0.23 0.1Ð12.41bc 8.19 2.60 Fenpropathrin 0.06 0.02Ð0.35ab 0.19 0.08Ð44.6cb 29.54 2.27 z-cypermethrin ϩ bifenthrin 0.03 0.02Ð0.04ab 0.1 0.08Ð0.14b 3.25 2.60 l-cyhalothrin ϩ thiamethoxam 0.08 0.07Ð0.10b 0.22 0.18Ð0.30c 2.72 3.01 Malathion 0.03 0.03Ð0.04b 0.1 0.08Ð0.13b 1.15 2.84 Chlorpyrifos 0.02 0.01Ð0.02a 0.07 0.05Ð0.07a 0.97 2.28 Carbaryl 23.6 3.6Ð12360c 1607 56.6Ð1.4 by 108d 2.72 0.70 Methomyl 2.71 1.1Ð17.38c 126.3 19.05Ð6550d 1.94 0.77

ϭ Fiducial limits (FL) 95%. The LC50 and LC90 of each pesticide followed by the same letter are not signiÞcantly different because of FL overlap. 2290 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 106, no. 6

Fig. 1. Percent mortality of X. glabratus 48 h after exposure to Þve different concentrations of pesticides on avocado bolts. The insecticide concentration “0” represents the no-insecticide control treatment. Error bars represent the SEM. caused signiÞcantly higher mortality than beetles secticide concentration was observed in the per- exposed to the two lowest concentrations centage of boring and surviving beetles in the (␹2 ϭ 10.49; P Ͻ 0.001); however, mortality rates were carbaryl (␹2 ϭ 5.73; P ϭ 0.21) and methomyl (␹2 ϭ Ͻ40% even after exposure to the highest concentra- 8.85; P ϭ 0.06) treatments (Fig. 2). tion of methomyl (Fig. 1). Persistence of Selected Insecticides Against X. The percentage of beetles that bored into the bolts glabratus Under Field Conditions. Evaluation of the and were found alive inside galleries after 15 d was efÞcacy of the tested insecticides 4 d after the appli- affected by the pesticide concentration of all insecti- cation showed that all insecticides but permethrin cides but carbaryl and methomyl (Fig. 2). There was caused a signiÞcant reduction in the number of beetles signiÞcantly less boring and survival of beetles ex- boring in avocado wood (F ϭ 13.86; df ϭ 6, 34; P Ͻ posed to the three highest concentrations of chlor- 0.001; Fig. 3). Few beetles bored through the bark of pyrifos compared with the control (␹2 ϭ 17.28; P Ͻ bolts treated with bifenthrin, fenpropathrin, z-cyper- 0.002). Exposure to the two highest concentrations of methrin ϩ bifenthrin, l-cyhalothrin ϩ thiamethoxam, malathion, fenpropathrin, and permethrin resulted in and malathion (Fig. 3). Contact with permethrin res- little beetle boring and survival, but exposure to the idues caused signiÞcantly less beetle mortality than two lowest concentrations resulted in beetle boring the other insecticides (F ϭ 17.59; df ϭ 6, 34; P Ͻ 0.001) and survival rates similar to the untreated control and more beetles were found alive 15 d after inside (␹2 ϭ 19.59, ␹2 ϭ 17.10, ␹2 ϭ 19.28; P Ͻ 0.001, re- galleries constructed within the wood. Beetles ex- spectively; Fig. 2). Contact with fresh residues of posed to the water control treatment had a low mor- z-cypermethrin ϩ bifenthrin resulted in signiÞcantly tality and more boring through the bark. No rain was less boring and survival than the control treatment recorded during the Þrst 4 d after application. across all the tested concentrations; the two highest Eight days after application, only the z-cyperme- concentrations resulted in less beetle boring and sur- thrin ϩ bifenthrin and malathion treatments caused vival (␹2 ϭ 21.97; P Ͻ 0.001). The bifenthrin and signiÞcantly higher beetle mortality than the control l-cyhalothrin ϩ thiamethoxam treatments showed a treatment (F ϭ 3.50; df ϭ 6, 34; P ϭ 0.01; Fig. 3). concentration-dependent effect with signiÞcantly less Consequently, fewer beetles bored through bark boring and survival at the highest concentrations treated with these two insecticides compared with the (␹2 ϭ 21.48, ␹2 ϭ 18.40; P Ͻ 0.001, respectively; Fig. other treatments (F ϭ 5.72; df ϭ 6, 34; P Ͻ 0.001). The 2). No beetle boring was recorded in the bolts number of beetles that died 48 h after exposure to treated with the three pyrethroid insecticides at the l-cyhalothrin ϩ thiamethoxam was higher than the highest rates (z-cypermethrin ϩ bifenthrin, perme- number of beetles that bored and were found alive thrin, bifenthrin). By contrast, no effect of the in- 15 d later (Fig. 3). In the rest of the treatments, the December 2013 CARRILLO ET AL.: CONTACT INSECTICIDES AGAINST X. glabratus 2291

Fig. 2. Percent of X. glabratus that bored through the bark and were found alive inside galleries 15 d after exposure to Þve different concentrations of insecticides. The insecticide concentration “0” represents the no-insecticide control treat- ment. Error bars represent the SEM. number of beetles boring through the treated bark was pletely lost 2 wk after application under typical sum- higher than the number of beetles that died because mer rainy conditions in south Florida. These results of the treatments. In all, 60.9 mm of rain fell between are strikingly different from those reported when days 4 and 8 after application. bifenthrin is used against bark beetles. According to Little beetle mortality occurred 15 d after appli- DeGomez et al. (2006) and Fettig et al. (2006), bifen- cation in all treatments, and only malathion and thrin confers protection to conifers against bark bee- z-cypermethrin ϩ bifenthrin caused signiÞcantly tles for one or two seasons applied as trunk-directed higher beetle mortality than the control treatments sprays at similar rates as in our experiments. Similarly, (F ϭ 3.90; df ϭ 7, 34; P ϭ 0.006; Fig. 3). The number Oghiakhe and Holliday (2011) reported that bifen- of beetles boring through the bark and constructing thrin provides long-term suppression of the native elm galleries was not affected by the insecticide treat- bark beetle H. rufipes. It is unclear whether the sub- ment (F ϭ 0.78; df ϭ 6, 34; P ϭ 0.59; Fig. 3). In all, tropical conditions of south Florida or the avocado 86.4 mm of rain was recorded between days 8 and 15 bark characteristics affect the persistence of bifen- after application. thrin, or whether some ambrosia beetles are more None of the treatments had any effect on X. glabra- tolerant to bifenthrin than bark beetles. Reding et tus mortality or in the number of beetles boring al. (2013) reported that bifenthrin did not effec- through the treated bark at 22 d after treatment (F ϭ tively prevent attacks by ambrosia beetles in nurs- 0.73; df ϭ 6, 34; P ϭ 0.63 and F ϭ 0.78; df ϭ 7, 34; P ϭ ery trees. Our results suggest that contact with fresh 0.59, respectively; Fig. 3). In all, 58.4 mm of rain was residues of bifenthrin applied at the highest label registered between days 15 and 22 after application. rate can kill X. glabratus, but its suppression will last only few days. Permethrin is another pyrethroid insecticide re- Discussion garded as effective for elm bark beetle control (Phil- Four pyrethroid insecticides with different active lipsen et al. 1986, Pajares and Lanier 1989). Reding et ingredients were tested in this study. Bifenthrin al. (2013) reported that permethrin products were the showed a high acute toxicity on X. glabratus, but under most effective among a range of insecticides tested to Þeld conditions, effective suppression of X. glabratus protect nursery plants from ambrosia beetles (i.e., X. lasted only 4 d. The effect of bifenthrin was com- germanus). By contrast, other researchers reported a 2292 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 106, no. 6

other insecticides. Fenpropathrin is registered for mite and thrips control at a rate of 1.56 liters/ha. However, used at label recommended rates, its efÞ- cacy diminished a few days after application. These results coincide with those reported by Pen˜ aetal. (2011) using potted avocado plants infested with known numbers of X. glabratus in a controlled fashion. They found that fenpropathrin-treated plants had sig- niÞcantly lower numbers of entry holes when checked at 1 d after treatment. However, there were no dif- ferences in entry holes between the treated trees and the untreated controls 1 wk later. The pyrethroid mixture of z-cypermethrin ϩ bifenthrin had a high acute toxicity on X. glabratus and was one of the two treatments that provided longer beetle suppression in the Þeld experiment. These results coincide with those reported by Pen˜ a et al. (2011), who found that z-cypermethrin ϩ bifenthrin was one of the insecti- cides that provided the most consistent suppression of Scolytinae as a contact insecticide. This insecticide is not currently registered for use in avocado but it could be useful in managing X. glabratus infestations. Addi- tional studies toward registration of z-cypermethrin ϩ bifenthrin are underway. l-cyhalothrin ϩ thiamethoxam caused a high mor- tality of X. glabratus, but its persistence under Þeld conditions was inferior to z-cypermethrin ϩ bifen- thrin. The pyrethroid part of the formulation is rec- ommended for management of the long horned Fig. 3. Persistence of contact insecticides against X. beetle, Anoplophora glabripennis (Coleoptera: Ceram- glabratus. Avocado trees were treated with insecticides at the bycidae), an invasive species in the United States highest label rate and limbs cut periodically and exposed to (Smith et al. 2007). Pen˜ a et al. (2011) reported that groups of 10 X. glabratus under controlled conditions. The l-cyhalothrin ϩ thiamethoxam was one of the mixtures black bar represents the number of beetles that were found dead 48 h after exposure to the treated bolts. The white that provided the most consistent suppression of Sco- bars represent the number of beetles that bored through lytinae. This mixture is not registered for use on av- the bark and were found alive in galleries inside the bolts ocado. However, if the persistence of this pesticide 15 d after. DAT, days after treatment. Error bars represent could be improved by the addition of adjuvants, it the SEM. might be useful for managing or preventing X. glabra- tus infestations. Organophosphate insecticides were highly toxic to lack of persistence of permethrin; when tested under X. glabratus. Chlorpyrifos and malathion caused high Þeld conditions, its efÞcacy diminished within few X. glabratus mortality in the laboratory bioassays. Mal- days of application (Oghiakhe and Holliday 2011). athion was among the two insecticides that provided Our results agree with the latter authors; permethrin longer persistence under Þeld conditions. Malathion is showed high acute toxicity on X. glabratus in labora- registered for use in avocado, whereas chlorpyrifos is not, tory bioassays, but caused low beetle mortality only 4 d and there is no interest by the manufacturer to pur- after application in the Þeld. However, the rate of sue registration (IR4 http://ir4.rutgers.edu/FoodUse/ permethrin registered for use in avocado is signiÞ- Food_Use1.cfm). cantly lower than the one used against bark beetles. In a previous assessment, malathion caused a sig- Prelude (AMVAC) is a formulation of permethrin niÞcant reduction in the number of scolytines entry registered for control of termites and bark beetles (i.e., holes and beetle emergence from treated logs in sev- vectors of Dutch elm disease) as a bark spray at a eral experiments but also provided erratic suppression Ϸ concentrated rate of 5.1 ml a.i./liter H2O. Perme- in one experiment (Pen˜ a et al. 2011). Smith (1982) thrin 3.2 AG is registered for use in avocado at 0.22 ml reported that malathion was ineffective on western a.i./liter H2O. At this rate, suppression of X. glabratus pine beetle, Dendroctonus brevicomis LeConte, 5 mo lasted Ͻ8d. after application. Fenpropathrin is another pyrethroid insecticide Carbaryl and methomyl were ineffective at control- that is registered for use in avocado. Contact with ling X. glabratus. Carbaryl has been widely used to fresh residues of this pesticide under controlled con- manage several pine bark beetles (Hastings et al. ditions showed a high acute toxicity on X. glabratus, 2001). However, carbaryl also lacks toxicity toward but when applied under normal grove conditions, the the southern pine beetle, Dendroctonus frontalis Zim- effect of fenpropathrin was intermediate relative to all mermann, which has a high degree of tolerance for this December 2013 CARRILLO ET AL.: CONTACT INSECTICIDES AGAINST X. glabratus 2293 insecticide (Hastings et al. 2001). Our experiments contact insecticides might reduce the number of suc- showed that X. glabratus tolerated carbaryl in the cessful attacks, but is not sufÞcient to completely elim- laboratory bioassays. Similarly, X. glabratus bored inate the risk of disease transmission. The authors are through the bark of avocado bolts treated with metho- currently investigating systemic insecticides as an al- myl showing tolerance toward this insecticide, which ternative management tactic against X. glabratus. is one of the pesticides currently registered for avo- Moreover, the potential of fungicides used as preven- cado. Our results agree with those of Pen˜ aetal. tive and curative treatments against R. lauricola is (2011), who reported higher number of entrance currently under investigation (Ploetz et al. 2011). The holes and higher beetle emergence in methomyl- current strategy is based in early detection and re- treated bolts relative to the nontreated control. moval of diseased trees to eliminate beetle breeding X. glabratus is the main vector of a fungal pathogen sites and fungal inoculum sources. The diseased that can kill a tree with an inoculation of relatively few trees are uprooted, the stump and roots burned, the spores. This increases the challenge of achieving an trunk and limbs are chipped, and the chips and integrated approach and also the likelihood that con- adjacent trees are sprayed with insecticides. Our ventional insecticides will be used by avocado grow- results suggest that pesticide applications under ers. None of the tested insecticides could completely these conditions can cause a short-term suppression prevent X. glabratus attack, and their persistence was of beetle populations, but other practices are low during the rainy summer season in south Florida. needed to mitigate the adverse effects of this bee- Organophosphate and pyrethroid insecticides appear tleÐdisease complex. to be the most effective groups for suppressing X. glabratus. Among the pesticides registered for use in Acknowledgments avocado, malathion and fenpropathrin represent the best chemical tools currently available; however, the We thank J. Capinera (UF Entomo. Nematol. Department, persistence of all tested insecticides is low, requiring Gainesville, FL) for providing laboratory space for the lab- frequent repeated applications. z-cypermethrin ϩ oratory bioassays. Steve Sorrell (Hammock Dunes Golf bifenthrin and l-cyhalothrin ϩ thiamethoxam at pos- Club) and the Florida Water Management for facilitating sibly higher rates and other pesticides may be required access to X. glabratus collecting sites. We thank Eileen Buss (UF Entomo. Nematol. Department), Brian J. Ulmer (Syn- to manage X. glabratus outbreaks. Moreover, it is im- genta), and Andrew Derksen (FDACS) for suggestions to perative to identify effective adjuvants to prolong the improve the manuscript. We thank G. Brar, S. McLean, Rita efÞcacy of these contact insecticides to manage X. H. Duncan, Paul Ruppert, Jose Alegria, Katia Santos, Wanda glabratus. Mantas, Jacob Hall, Armando Garza, and Reed Olszack for Pest resurgence is a major concern of avocado their help. This research was partially funded by a Specialty growers regarding insecticide applications against X. Crop Grant and a Florida Avocado Committee grant to J. E. glabratus. The choice of an insecticide to be applied Pen˜ a and J.H. Crane. against X. glabratus should be deÞned not only by its effectiveness against the vector but also by its effect on References Cited direct and indirect pest resurgence. The complex of spider mites and insects that affect avocado in south Abbott, W. S. 1925. A method of computing the effective- Florida has been under a 20-yr integrated pest man- ness of an insecticide. J. Econ. Entomol. 18: 265Ð267. Brar, G. S., J. L. Capinera, S. McLean, P. E. Kendra, R. C. agement program (Pena et al. 2013). The most com- ˜ Ploetz, and J. E. Pen˜ a. 2012. Effect of trap size, trap mon pests are mirids, thrips infesting ßowers, mites, height and age of lure on sampling Xyleborus glabratus lace bugs, and loopers affecting leaves (Pen˜ aetal. (Coleoptera: Curculionidae: Scolytinae), and its ßight 2013), whereas soft, armored scales, mealybugs, and periodicity and seasonality. Fla. Entomol. 95: 1003Ð whiteßies are seldom observed. Spray interventions 1011. are kept to a minimum because of the effective Brar, G. S., J. L. Capinera, S. McLean, and J. E. Pen˜ a. 2013. natural enemy complex keeping these pests under Life cycle, development and culture of Xyleborus glabra- low density levels. Pest resurgence represents a tus (Coleoptera: Curculionidae: Scolytinae). Florida En- major challenge to design sound chemical manage- tomol. 96: 1158Ð1167. Carrillo, D., R. E. Duncan, and J. E. Pen˜ a. 2012. Ambrosia ment strategies against this invasive vector. The beetles (Coleoptera: Curculionidae: Scolytinae) that indirect effect on nontarget pests of several contact breed in avocado wood in Florida. Fla. Entomol. 95: insecticides recommended against X. glabratus is 573Ð579. currently being tested by the authors in south Flor- Carrillo, D., R. E. Duncan, J. N. Ploetz, A. F. Campbell, R. C. ida avocado groves. Ploetz, and J. E. Pen˜ a. 2013. Lateral transfer of a phy- More management tools including more registered topathogenic symbiont among native and exotic ambrosia pesticides are needed to protect avocados from this beetles. Plant Pathol. (doi: 10.1111/ppa.12073). beetleÐdisease complex. In this study, no data were Crane, J. H., J. E. Pen˜ a, and J. L. Osborne. 2008. Redbay collected about inoculation of R. lauricola by X. ambrosia beetle-laurel wilt pathogen: a potential major problem for the Florida avocado industry. University of glabratus in the limbs; however, previous experiments Florida, IFAS Extension, EDIS, HS1136, 8 pp. (http:// infesting small potted avocado trees suggest that a few edis.ifas.uß.edu/hs379). beetle attacks can be enough to transmit the laurel wilt Damon, A. 2000. A review of the biology and control of the pathogen. In our experiments we seldom observed coffee berry borer, Hypothenemus hampei (Coleoptera: 100% beetle mortality, which suggests that the use of Scolytidae). Bull. Entomol. Res. 90: 453Ð465. 2294 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 106, no. 6

DeGomez, T. E., C. J. Hayes, J. A. Anhold, J. D. McMillin, Lanier, G. N., J. F. Sherman, R. J. Rabaglia, and A. H. Jones. K. M. Clancy, and P. P. Bosu. 2006. Evaluation of insec- 1984. Insecticides for control of elm bark beetles that ticides for protecting southwestern ponderosa pines from spread Dutch elm disease. J. Arboric. 10: 265Ð272. attack by engraver beetles (Coleoptera: Curculionidae: Lozano, C., A. Pen˜ a, A. J. Sanchez–Raya, and M. Campos. Scolytinae). J. Econ. Entomol. 99: 393Ð400. 2001. Effectiveness of deltamethrin for protecting olives Fettig, C. J., K. K. Allen, R. R. Borys, J. Christopherson, C. P. from attack by the beetle Phloeotribus scarabaeoides (Sco- Dabney, T. J. Eager, K. E. Gibson, E. G. Hebertson, D. F. lytidae). Intl. J. Pest Manage. 47: 75Ð78. Long, and A. S. Munson. 2006. Effectiveness of bifen- Mizell, R. F., and T. C. Riddle. 2004. Evaluation of insecti- thrin (Onyx) and carbaryl (Sevin SL) for protecting cides to control Asian ambrosia beetle, Xylosandrus cras- individual, high-value conifers from bark beetle attack siusculus, pp. 152Ð155. In B. L. James (ed.), Proceedings (Coleoptera: Curculionidae: Scolytidae) in the western 49th Annual Southern Nursery Association Research United States. J. Econ. Entomol. 99: 1691Ð1698. Conference. Atlanta, GA. Southern Nursery Association, Fettig, C. J., D. M. Grosman, and A. S. Munson. 2013. Ad- Inc., Acworth, GA. vances in insecticide tools and tactics for protecting co- Maner, M. L. 2012. Life history, rearing, and habits of the nifers from bark beetle attack in the Western United redbay ambrosia beetle (Coleoptera: Curculionidae: Sco- States, pp. 471Ð492. In T. Stanislav (ed.), Insecticides- lytinae) in the Þeld and laboratory. M.S. thesis, University Development of Safer and More Effective Technolo- of Georgia, Athens, GA. gies. InTech, Rijeka, Croatia. Mayfield III, A. E., J. E. Pen˜ a, J. H. Crane, J. A. Smith, C. L. (FDACS) Florida Department of Agriculture and Con- Branch, E. D. Ottoson, and M. Hughes. 2008. Ability of sumer Services. 2012. FDACS-identifes laurel wilt dis- the redbay ambrosia beetle (Coleoptera: Curculionidae: ease in Avocado Production Area of MiamiÐDade Scolytinae) to bore into young avocado (Lauraceae) County. (http://www.freshfromßorida.com/newsroom/ plants and transmit the laurel wilt pathogen (Raffaelea press/2012/05012012.html). sp.). Fla. Entomol. 91: 485Ð487. Fraedrich, S. W., T. Harrington, R. J. Rabaglia, M. D. Oghiakhe, S., and N. J. Holliday. 2011. Evaluation of insecti- Ulyshen, A. E. Mayfield, J. L. Hanula, J. M. Eicwort, and cides for control of overwintering Hylurgopinus rufipes (Co- D. R. Miller. 2008. A fungal symbiont of the redbay leoptera: Curculionidae). J. Econ. Entomol. 104: 889Ð894. ambrosia beetle causes a lethal wilt in redbay and other Pajares, J. A., and G. N. Lanier. 1989. Pyrethroid insecticides Lauraceae in the southeastern United States. Plant Dis. for control of European elm bark beetle (Coleoptera: 92: 215Ð224. Scolytidae). J. Econ. Entomol. 82: 873Ð878. Pena, J. E., J. H. Crane, J. L. Capinera, R. E. Duncan, P. E. Gardiner, L. M., and D. P. Web. 1980. Tests of chlorpyrifos ˜ Kendra, R. C. Ploetz, S. Mclean, G. Brar, M. C. Thomas, and for control of the North American elm bark beetle, R. D. Cave. 2011. Chemical control of the redbay ambrosia Hylurgopinus rufipes Eichh. Can. For. Serv. Rep. 0-X- beetle, Xyleborus glabratus, and other Scolytinae (Co- 311. leoptera: Curculionidae). Fla. Entomol. 94: 882Ð896. Hall, R. W., P. J. Shea, and M. J. Haverty. 1982. Effectiveness Pen˜ a, J. E., D. Carrillo, R. E. Duncan, J. L. Capinera, G. Brar, of carbaryl and chlorpyrifos for protecting ponderosa S. McLean, M. L. Arpaia, E. Focht, J. A. Smith, M. Hughes, pine from attack by the western pine beetle (Coleoptera: et al. 2012. Susceptibility of Persea spp. and other Lau- Scolytidae). J. Econ. Entomol. 75: 504Ð508. raceae to attack by redbay ambrosia beetle, Xyleborus Hanula, J. L., A. E. Mayfield III, S. W. Fraedrich, and R. J. glabratus (Coleoptera: Curculionidae: Scolytinae). Fla. Rabaglia. 2008. Biology and host associations of red- Entomol. 95: 793Ð797. bay ambrosia beetle (Coleoptera: Curculionidae: Sco- Pen˜ a, J. E., M. S. Hoodle, M. Aluja, E. Palevsky, R. Ripa, and lytinae), exotic vector of laurel wilt killing trees in the M. Wysoki. 2013. Insect and mite pests, pp. 423Ð488. In Southeastern United States. J. Econ. Entomol. 101: B. Schaffer, N. Wolstenholme, and A. W. Whiley (eds.), 1276Ð1286. The Avocado: Botany, Production and Uses, 2nd ed. Hastings, F. L., E. H. Holsten, P. J. Shea, and R. A. Werner. CABI, Boston, MA. 2001. Carbaryl: a review of its use against bark beetles in Phillipsen, W. J., M. E. Ascerno, and V. R. Landwehr. 1986. coniferous forests of North America. Environ. Entomol. Colonizations, emergence, and survival of Hylurgopinus 30: 803Ð810. rufipes and Scolytus multistriatus (Coleoptera: Scolyt- Haverty, M. I., P. J. Shea, J. J. Hoffman, J. M. Wenz, and K. E. idae) in insecticide-treated elm wood. J. Econ. Entomol. Gipson. 1998. Effectiveness of esfenvalerate, cyßuthrin, 79: 1347Ð1350. and carbaryl in protecting individual lodgepole pines and Ploetz, R. C., J. M. Perez–Martinez, E. A. Evans, and S. A. ponderosa pines from attack by Dendroctonus spp. Pac. Inch. 2011. Toward fungicidal management of laurel Southwest Res. Stn. For. Ser. U.S. Dep. Agric. Res. Pap wilt of avocado. Plant Dis. 95: 977Ð982. PSW-RP-237. Reding, M. E., J. B. Oliver, P. B. Schultz, C. M. Ranger, and Jin, H., and N. J. Webster. 1997. Persistence, penetration, N. N. Youssef. 2013. Ethanol injection of ornamental and surface availability of chlorpyrifos, its , and 3,5,6- trees facilitates testing insecticide efÞcacy against am- trichloro-2-pyridinol in elm bark. J. Agric. Food Chem. 45: brosia beetles (Coleoptera: Curculionidae: Scolytinae). J. 4871Ð4876. Econ. Entomol. 106: 289Ð298. Jin, H., G.R.B. Webster, N. J. Holliday, P. A. Pines, and A. R. Saeed, S., A. Masood, A. H. Sayyed, and Y. J. Kwon. 2011. Westwood. 1996. An elm bark beetle bioassay for resid- Comparative efÞcacy of different pesticides against ual efÞcacy of chlorpyrifos and cypermethrin used for the mango bark beetle Hypocryphalus mangiferae Stebbing control of Dutch elm disease in Manitoba. J. Environ. Sci. (Coleoptera: Scolytidae). Entomol. Res. 41: 142Ð150. Health B 31: 751Ð761. SAS Institute. 2013. SAS version 9.3. SAS Institute, Cary, Kendra, P. E., W. S. Montgomery, J. Nigoret, J. E. Pen˜ a, J. L. NC. Capinera, G. Brar, N. D. Epsky, and R. R. Heath. 2011. Shea, P. J., M. I. Haverty, and R. C. Hall. 1984. Effectiveness Attraction of the redbay ambrosia beetle, Xyleborus of fenithrothion and permethrin for protecting pon- glabratus, to Avocado, Lychee, and essential oil lures. derosa pine trees from attack by western pine beetle. J. J. Chem. Ecol. 37: 932Ð942. Ga. Entomol. Soc. 19: 427Ð433. December 2013 CARRILLO ET AL.: CONTACT INSECTICIDES AGAINST X. glabratus 2295

Smith, R. H. 1982. Log bioassay of residual effectiveness of control of the Asian long horned beetle, pp. 73Ð74. U.S. insecticides against bark beetles. PaciÞc Southwest Forest Dep. Agric. Interagency Research Forum GTR-NRS- and Range Experiment Station, Forest Service, U. S. De- P-28. partment of Agriculture, Res. Paper PSW-168. Berkeley, CA. Smith, M. T., J. Wu, J. M. Tropp, W. He, H. Su, G. Zhang, X. Xu, and J. Li. 2007. EfÞcacy of lambda-cyhalothrin for Received 30 April 2013; accepted 29 August 2013.