Pesticide Compatibility With Natural Enemies for Pest Management in Greenhouse Gerbera Daisies

Transcription

1 BIOLOGICAL AND MICROBIAL CONTROL Pesticide Compatibility With Natural Enemies for Pest Management in Greenhouse Gerbera Daisies CHERI. M. ABRAHAM, 1,2 S. K. BRAMAN, 3 R. D. OETTING, 3 AND N. C. HINKLE 1 J. Econ. Entomol. 106(4): 1590Ð1601 (2013); DOI: ABSTRACT Pesticides commonly used in commercial greenhouse management were evaluated for compatibility with two biological control agents: a leafminer parasitoid (Diglyphus isaea [Walker]), and a predatory mite (Neoseiulus californicus [McGregor]). These natural enemies were exposed to miticides, fungicides, and insecticides targeting leafminers, thrips, and whiteßies, according to label directions in laboratory vial assays, after which mortality at 12, 24, and 48 h was recorded. Greater mortality of predatory mites than leafminer parasitoids was observed overall, illustrating that fewer pesticides were compatible with predatory mites compared with the parasitoid. However, some commonly used pesticides were found to cause high mortality to both the leafminer parasitoid and predatory mites. Twospotted spider mite (Tetranychus urticae Koch) infestations often disrupt leafminer (Liriomyza trifolii [Burgess]) biocontrol programs. Therefore, potentially compatible miticides (bifenazate, hexythiazox, spiromesifen, acequinocyl, etoxazole, and clofentezine) identiþed in laboratory trials were also evaluated in a greenhouse study and found to be compatible with leafminer biocontrol. KEY WORDS Diglyphus isaea, greenhouse pest management, greenhouse biocontrol, leafminer biocontrol, safe pesticide 1 Department of Entomology, University of Georgia, 413 Biological Sciences Bldg., Athens, GA Corresponding author, cherimabraham@gmail.com. 3 Department of Entomology, University of Georgia, 1109 Experiment St., GrifÞn, GA Serpentine leafminers, Liriomyza trifolii (Burgess) (Diptera: Agromyzidae), key pests of greenhouse gerberas, have a wide distribution and attack 400 species (Reitz and Trumble 2002) of plants, including vegetables and ornamentals. The larvae feed on the palisade mesophyll (Parrella et al. 1985) and decrease photosynthesis and yield, directly affecting the marketable produce. Intensive and extended use of pesticides has rendered leafminers with reduced susceptibility to many chemistries (Keil and Parrella 1982). Leafminers are also protected from topical pesticides by being concealed within the leaves in their larval stages. Successful and effective biological control has been implemented by augmentative releases of parasitoids, only where disruptive use of chemical controls has been avoided (Liu et al. 2009). The inßux of secondary pests (like mites, thrips, whiteßies, and aphids) and pathogens (e.g., causing powdery mildew) through the season necessitates pesticide sprays that could negatively impact the leafminer parasitoids and hence disrupt biological control. The unique situation in greenhouse gerbera production suggests the potential for integrated pest management (IPM) as a solution. Knowing the compatibility of pesticides with natural enemies is a prerequisite for implementation of any IPM program. Although Biobest 2011 (biobest.be) and Koppert biological systems 2011 (koppert.com) and several studies (Medina et al. 2003, Cloyd 2006, Cloyd and Dickinson 2006, Bethke and Cloyd 2009) have compiled information about compatibility of pesticides to parasitoids and predators in numerous production systems, gaps exist in the greenhouse gerbera system. Pesticides targeting leafminers (L. trifolii), mites (Tetranychus urticae Koch), thrips (Frankliniella occidentalis [Pergande]), whiteßies (Trialeurodes vaporariorum [Westwood] and Bemisia tabaci [Gennadius]), aphids (Myzus persicae [Sulzer]), and pathogens causing powdery mildew (from the genera Podosphaera, Erysiphe, Leveillula, Golovinomyces, and Oidium) were evaluated for compatibility with a leafminer parasitoid (Diglyphus isaea [Walker]) and a predatory mite (Neoseiulus californicus [McGregor]). Materials and Methods Laboratory Study. Experimental Protocol. An experiment evaluated a set of chemicals targeting each of the Þve pests, namely, leafminers, mites, thrips, whiteßies, and powdery mildew causing pathogen. Nine pesticides and a water control were evaluated in each experiment. As pesticides recommended against aphids are also used against other pests, but at a higher rate, they were not evaluated as a separate group. Vial assay methods (Bjorksten and Robinson 2005, Wu and Miyata 2005) were modiþed and used as leaf-dip assays for the parasitoid wasps and as pesticide swirl assays for predatory mites /13/1590Ð1601$04.00/ Entomological Society of America

2 August 2013 ABRAHAM ET AL.: PESTICIDE COMPATIBILITY WITH GREENHOUSE BIOCONTROL 1591 Table 1. Summary of compatibility of pesticides with natural enemies following IOBC guidelines (Stark et al. 2007) Leafminer materials Miticides Thripicides Whiteßy chemicals Fungicides Harmless ( 30% mortality within 48 h) Novaluron a Clofentezine a Pyriproxyfen a Butanone a Petroleum oil a Acequinocyl a Spiromesifen a Fosetyl-aluminum a Leafminer parasitoid (D. isaea): Gerbera plugs that had previously not been treated were obtained from Speedling Inc., Blairsville, GA. A single leaf was removed from the plug and covered with cotton around the petiole and inserted into one end of a 1.5-cm-long section of Tygon tubing and hydrated with a squirt bottle when necessary. The leaf was then completely dipped in the respective treatments (aqueous pesticide solutions at label rates or water control) for 10 s each and allowed to dry for at least 3 h. After the inside of the vial (23 by 85 mm, six drams, BioQuip Products, Inc., Rancho Dominguez, CA) was streaked with honey (as a food source for the parasitoids), 10 D. isaea parasitoids were introduced. The tubing with the leaf inside was then inserted at the neck region of the vial and sealed using ParaÞlm. Predatory Mites (N. californicus): A solution (10Ð15 ml) of the designated treatment was poured into each glass vial (26.5 by 56 mm, 4.5 drams, BioQuip Products, Inc.) and swirled for even coverage over the glass surface. After allowing at least 3 h for drying, a drop of honey was streaked inside each vial, and then 10 adult N. californicus mites were inserted and the vial capped. Design and Data Collection. The experimental unit was a vial and the experiment consisted of 10 replicates for each of the 10 treatments, all of which were placed on a laboratory counter with a photoperiod of 14:10 (L:D) h and held at 22Ð25 C. Three trials for each experiment were conducted. Live adult parasitoids and adult mites (viewed through a microscope) were Azoxystrobin a Potassium bicarbonate a Pyraclostrobin a Copper sulfate a Piperalin a Slightly harmful (30Ð79% mortality within 48 h) Azadirachtin a Bifenazate a,b Flonicamid a,b Flonicamid a,b Sulfur a Cyromazine a,b Hexythiazox a,b Cyßuthrin a Chlorpyrifos a Rosemary oil a Petroleum oil b Spiromesifen a,b Insecticidal soap a,b Spirotetramat a,b Butanone b Acetamiprid a Milbemectin a B. bassiana a Pyridaben a Copper sulfate b Novaluron b Etoxazole a,b Acetamiprid a Thiamethoxam b Clofentezine b Spiromesifen b Moderately harmful (80Ð98% mortality within 48 h) cyhalothrin a Abamectin a Abamectin a Kinoprene a Sulfur b Azadirachtin b Acequinocyl b Fluvalinate a Thiamethoxam a Chlorfenapyr a Imidacloprid a B. bassiana b Lambda Cyhalothrin a Acetamiprid b Pyriproxyfen b Chlorpyrifos b Harmful ( 99% mortality within 48 h) Dinotefuran a,b Spinosad a,b Spinosad a,b Kinoprene b Fosetyl-aluminum b Bifenthrin a,b Milbemectin b Abamectin b Imidacloprid b Rosemary oil b cyhalothrin b Abamectin b Cyßuthrin b Pyridaben b Azoxystrobin b Acetamiprid b Fluvalinate b Lambda cyhalothrin b Potassium bicarbonate b Chlorfenapyr b Pyraclostrobin b Piperalin b Safety to natural enemies denoted by following legends: D. isaea a, and N. californicus b. counted 12, 24, and 48 h after the treatment. Any movement by the natural enemies designated them as alive, whereas the lack of movement when disturbed resulted in counting them as dead. Greenhouse Miticide Study. Location and Experimental Design. The study was conducted at the UGA- GrifÞn campus. After selecting and housing 170 potted gerbera plants of the Gerbera ÔFestival Mini Yellow ShadeÕ cultivar in similar growth stages from among plants that we managed and not sprayed any insecticides in the past 30 d, an excess of 500 adult L. trifolii collected from grower and research greenhouses were released into the greenhouse. Flies from the grower greenhouses were under regular and possibly high pesticide pressure, while our research greenhouses were maintaining leafminers (hence no pesticide application apart from fungicides). Treatments included six miticides (bifenazate, hexythiazox, spiromesifen, acequinocyl, etoxazole, and clofentezine) and a (water) control and were applied a week after the ßies were introduced. Four potted plants caged together consisted an experimental unit (BugDorm rearing cage, #1452, BioQuip Products, Inc.). There were seven replicates. Twenty-four hours later, 10Ð12 parasitoids (D. isaea) purchased from Rincon Vitova Insectaries Inc., Ventura, CA, were released into each cage. During the test period, the greenhouse was maintained at 25Ð32 C and 85% relative humidity (RH). Data Collection and Evaluation. Seven days after the parasitoids were released, three leaves were sampled from each experimental unit and inspected under

3 1592 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 106, no. 4 Table 2. Means ( SE) of number of live natural enemies (D. isaea and N. californicus) at each observation time of 12, 24, and 48 h in each of three trials (trials 1, 2, and 3) after exposure to leafminer-targeted materials at median label rates out of 10 natural enemies in each experimental unit Treatment D. isaea N. californicus Trial 1 Trial 2 Trial 3 Trial 1 Trial 2 Trial 3 12 h Control a a a a a a Spiromesifen ab a a ab bc bc Cyromazine ab a a a a ab Novaluron bc 10 0a a abc bcd cd Petroleum oil cd a a c ab ab Azadirachtin ab a a bc ab cd Acetamiprid abc a b d cd cd Dinotefuran e b c d cd 0.0 0d Bifenthrin e c c 0.0 0d 0.0 0d 0.0 0d cyhalothrin de b c 0.0 0d d 0.0 0d F h Control ab a a ab a a Spiromesifen bcd ab a abc bc bc Cyromazine a b a a a ab Novaluron cd ab a bc bcd cd Petroleum oil de ab a c ab ab Azadirachtin abc ab a c ab cd Acetamiprid abc ab b d cd cd Dinotefuran f cd c d cd 0.0 0d Bifenthrin f d c 0.0 0d 0.0 0d 0.0 0d cyhalothrin ef c c 0.0 0d 0.0 0d 0.0 0d F h Control a a a a a a Spiromesifen ab abc ab a bc bc Cyromazine a c bc a a ab Novaluron b ab ab ab bcd cd Petroleum oil bc abc abc b ab ab Azadirachtin a c bc c cd 0.0 0d Acetamiprid ab bc c c d cd Dinotefuran c d d 0.0 0c 0.0 0d 0.0 0d Bifenthrin c d d 0.0 0c 0.0 0d 0.0 0d cyhalothrin c d d 0.0 0c 0.0 0d 0.0 0d F a microscope for parasitoid and leafminer activity. After the Þrst sampling date, cages were removed so that the leafminer pressure and the parasitoid availability would be equal for all the plants, while residual toxicity would determine the actual activity of leafminer and D. isaea. The greenhouse was ßooded with an excess of 600 adult leafminers and 72 h later, 250 parasitoids. Sampling was then repeated every seventh day thereafter for 3 wk spanning 14 June through 5 July Statistical Analyses. The experiments were analyzed as randomized complete block designs. Replications were considered as the block factor. Data were subjected to analysis of variance (ANOVA) using the general linear model procedure (PROC GLM, SAS Institute, Cary, NC) and means were separated using TukeyÕs honestly signiþcant difference test. Laboratory Study. Treatment means were analyzed separately for each study. When initial analysis determined that date was signiþcant (P 0.05), trials for each experiment were subsequently analyzed separately. The tiered method advocated by the International Organization of Biological Control (IOBC) considers pesticides from laboratory studies causing mortality rates of 30Ð79% to be slightly harmful and 30% mortality harmless (Stark et al. 2007), and chemicals falling in both these categories to qualify to be part of IPM programs. Pesticides in this study that caused mortalities within these values at least twice out of the three trials were considered at least less harmful. Greenhouse Study. Data were analyzed as previously described, Þrst to Þnd the difference in parasitism rate (average number of parasitoids/total number of leafminers in the experimental unit) between the treatments. Additional analyses investigated the differences based on average number of leafminers, average number of parasitoids, number of live leafminers, and total (sum of live and dead) leafminers.

4 August 2013 ABRAHAM ET AL.: PESTICIDE COMPATIBILITY WITH GREENHOUSE BIOCONTROL 1593 Table 3. Means ( SE) of number of live natural enemies (D. isaea and N. californicus) at each observation time of 12, 24, and 48 h in each of three trials (trials 1, 2, and 3) after exposure to miticides at median label rates out of 10 natural enemies in each experimental unit Treatment D. isaea N. californicus Trial 1 Trial 2 Trial 3 Trial 1 Trial 2 Trial 3 12 h Control b a a a a b Hexythiazox a a a ab a ab Milbemectin a a a c d d Clofentezine a a a a a a Spiromesifen a a a a a a Bifenazate a a a a a ab Etoxazole bc a a a a ab Acequinocyl b a a bc b c Abamectin b b b bc bc cd Spinosad c 0.0 0c c c cd d F h Control b a a abc a b Hexythiazox a a a cdef a ab Milbemectin a a a f d c Clofentezine a a a bcde a a Spiromesifen a a a a ab a Bifenazate a a a abcd a b Etoxazole bc a a ab a ab Acequinocyl bc a a def bc c Abamectin bc b b ef cd c Spinosad c 0.0 0b 0.0 0c ef d c F h Control b ab a abc a b Hexythiazox a ab ab cde a ab Milbemectin a b ab 0.0 0e 0.0 0c c Clofentezine a ab ab bcd a a Spiromesifen a ab b a ab a Bifenazate a ab ab abcd a b Etoxazole b a ab ab 6.2a 0.61a ab Acequinocyl b ab ab de bc c Abamectin b 0.0 0c c 0.0 0e bc c Spinosad 0.0 0b 0.0 0c 0.0 0c 0.0 0e c c F Results Laboratory Study. Following the criteria accepted by the IOBC (Stark et al. 2007), chemicals tested in laboratories are divided into four categories based on their toxicity. Those causing 30% mortality are considered harmless, 30Ð79% slightly harmful, 80Ð98% moderately harmful, and 99% considered harmful. The same criteria were used to elucidate our laboratory experiment results. Leafminer Chemicals (D. isaea at 48 h). Novaluron and petroleum oil were harmless (Table 1). Azadirachtin, cyromazine, and acetamiprid were slightly harmful, causing mortality in the range of 30Ð 79%. Lambda cyhalothrin was found to be moderately harmful, with a mortality of 80Ð98%. Dinotefuran and bifenthrin were harmful and caused mortality 99% within 48 h (F range 27.04, 47.96, 39.45; df 9, 99; P ) (Tables 2 and A1). Leafminer Chemicals (N. californicus at 48 h). At the 48 h mark, none of the pesticides were harmless to the predatory mites (Table 1). Cyromazine, novaluron, and petroleum oil were slightly harmful. Azadirachtin was moderately harmful, whereas dinotefuran, bifenthrin, lambda cyhalothrin, and acetamiprid were harmful (F range 46.24, 16.84, 18.32; df 9, 99; P ) (Table 2). The low mortality in cyromazine and novaluron at the 48 h mark does not ensure their harmlessness because of their being insect growth regulators. Miticides (D. isaea at 48 h). Clofentazine and acequinocyl were harmless (Table 1), whereas bifenazate, hexythiazox, spiromesifen, etoxazole, and milbemectin were slightly harmful. Abamectin and spinosad were moderately harmful and harmful, respectively, to D. isaea (F range 23.53, 84.97, 17.46; df 9, 99; P ) (Table 3). Miticides (N. californicus at 48 h). Etoxazole, bifenazate, hexythiazox, clofentazine, and spiromesifen were slightly harmful (F range 12.85, 16.68, 43.56; df 9, 99; P ) (Tables 1 and 3). The rest were

5 1594 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 106, no. 4 Table 4. Means ( SE) of number of live natural enemies (D. isaea and N. californicus) at each observation time of 12, 24, and 48 h in each of three trials (trials 1, 2, and 3) after exposure to whitefly-targeted materials at median label rates out of 10 natural enemies in each experimental unit Treatment D. isaea N. californicus Trial 1 Trial 2 Trial 3 Trial 1 Trial 2 Trial 3 12 h Control a a a bc a a Spirotetramat ab a a ab cde bc Pyriproxyfen a a a bcd cde ef Flonicamid abc ab a a ab ab Kinoprene ab a a bcd c,d,f de Chlorpyrifos cde bc a ab cd cde Pyridaben cd cd a cd f ef cyhalothrin e cd b d ef f Imidacloprid bc d b cd def ef Thiamethoxam de d b ab bc bcd F h Control a a a bcd a a Spirotetramat abc ab a ab b abc Pyriproxyfen ab ab a cde bcd de Flonicamid bcd b a a a ab Kinoprene abc b a de bcd de Chlorpyrifos de bc a bc bc cd Pyridaben cde cd a de 0.0 0d e cyhalothrin e cd b 0.0 0e cd 0.0 0e Imidacloprid bcd cd b cde cd de Thiamethoxam e d b bcd b bc F h Control a a a bcd a a Spirotetramat ab b a b bc ab Pyriproxyfen a b a de d d Flonicamid b bc a a a ab Kinoprene c cde b 0.0 0e 0.0 0d 0.0 0d Chlorpyrifos c cd 7.7a 0.41b cde bcd cd Pyridaben bc de a e 0.0 0d 0.0 0d cyhalothrin c de b 0.0 0e cd 0.0 0d Imidacloprid b de b 0.0 0e 0.0 0d d Thiamethoxam c e b bc b bc F at least moderately harmful, and none of them were known to have preferential toxicity only to pest mite species. Whitefly Chemicals (D. isaea at 48 h). Pyriproxyfen, and spiromesifen were harmless (F range 20.07, 24.71, 20.39; df 9, 99; P ) (Tables 1 and 4) whereas spirotetramat, ßonicamid, pyridaben, and chlorpyrifos were slightly harmful. Kinoprene, thiamethoxam, imidacloprid, and lambda cyhalothrin were moderately harmful and hence are probably best not used in a biological-based IPM program. Whitefly Chemicals (N. californicus at 48 h). Flonicamid, spirotetramat, thiamethoxam, and spiromesifen were slightly harmful (F range 21.7, 24.94, 24.88; df 9, 99; P ) (Tables 1 and 4). Pyriproxyfen and chlorpyrifos were moderately harmful, whereas kinoprene, imidacloprid, pyridaben, and lambda cyhalothrin were harmful. Thripicides (D. isaea at 48 h). Flonicamid, cyßuthrin, insecticidal soap, Beauveria bassiana, and acetamiprid were slightly harmful (F range 32.47, 31.2, 40.96; df 9, 99; P ) (Tables 1 and 5); abamectin, ßuvalinate, and chlorfenapyr were moderately harmful; and spinosad was harmful. Thripicides (N. californicus at 48 h). Flonicamid and insecticidal soap were slightly harmful, whereas B. bassiana and acetamiprid were moderately harmful (F range 15.04, 32.61, 27.01; df 9, 99; P ) (Tables 1 and 5). Abamectin, spinosad, cyßuthrin, ßuvalinate, and chlorfenapyr were all harmful. Fungicides (D. isaea at 48 h). All tested fungicides showed 79% mortality in D. isaea within 48 h and thus qualify to be used in IPM programs. Butanone, fosetyl-aluminum, azoxystrobin, potassium bicarbonate, pyraclostrobin, and copper sulfate were harmless (F range 4.92, 5.50, 1.53; df 9, 99; P , , 0.15) (Tables 1 and 6). Rosemary oil (EcoSMART Technologies, Alpharetta, GA) and sul-

6 August 2013 ABRAHAM ET AL.: PESTICIDE COMPATIBILITY WITH GREENHOUSE BIOCONTROL 1595 Table 5. Means ( SE) of number of live natural enemies (D. isaea and N. californicus) at each observation time of 12, 24, and 48 h in each of three trials (trials 1, 2, and 3) after exposure to thrips materials (thripicides) at median label rates out of 10 natural enemies in each experimental unit Treatment D. isaea N. californicus Trial 1 Trial 2 Trial 3 Trial 1 Trial 2 Trial 3 12 h Control a a a a a a Acetamiprid ab ab b de ab c Flonicamid a a b ab a bc Insecticidal soap a ab ab bcd ab ab B. bassiana a ab b abc ab bc Cyßuthrin c bc c e cd d Fluvalinate bc cd c e d d Abamectin c cd ab cde bc bc Carbonitrile bc d b e d c Spinosad c e c de ab ab F h Control a a a a a a Acetamiprid ab a b cde ab cde Flonicamid ab a b ab abc bc Insecticidal soap a a ab bcd cd ab B. bassiana ab a ab bc bc c Cyßuthrin c b cd e e ef Fluvalinate bc b d e 0.0 0e f Abamectin cd b bc cde de cd Carbonitrile cd bc d e e def Spinosad d c e de ab c F h Control ab a a a a a Acetamiprid cd a a cd b bcd Flonicamid bc a a ab c b Insecticidal soap a a a bc bc bc B. bassiana bc a a bc bc 0.0 0d Cyßuthrin e b bc d 0.0 0c 0.0 0d Fluvalinate de bc cd 0.0 0d 0.0 0c 0.0 0d Abamectin e bc b 0.0 0d 0.0 0c 0.0 0d Chlorfenapyr e bc de 0.0 0d 0.0 0c 0.0 0d Spinosad e c e cd 0.0 0c cd F fur were the only ones that caused higher mortality and were slightly harmful. Fungicides (N. californicus at 48 h). Butanone and copper sulfate were slightly harmful (F range 16.11, 70.13, 40.97; df 9, 99; P ) (Tables 1 and 6), whereas sulfur was moderately harmful. Fosetyl-aluminum, rosemary oil, azoxystrobin, potassium bicarbonate, pyraclostrobin, and piperalin were harmful. Greenhouse Study. Treatments did not differ from the control in parasitism rates over 4 wk, conþrming compatibility observed in laboratory studies (F range 0.22Ð1.38; df 6, 41; P range Ð0.9673) (Table A2). Trends were same for other parameters analyzed: Average number of leafminers (F range 0.95Ð1.27; df 6, 41; P range Ð0.4774) (Table A3), average number of parasitoids (F range 0.18Ð 1.54; df 6, 41; P range Ð0.9800) (Table A4), number of live leafminers (F range 0.95Ð1.27; df 6, 41; P range Ð0.4774) (Table A5), and total (sum of live and dead) leafminers (F range 0.31Ð 1.51; df 6, 41; P range Ð0.9276) (Table A6). Parasitism, which started high in the Þrst week, fell in the second week and returned to its highest level by the fourth week. Discussion Laboratory Study. A majority of the tested pesticides were toxic to the leafminer parasitoid D. isaea and the predatory mite N. californicus at the 48 h mark. Several studies have looked at effects of fewer pesticides on leafminer parasitoids in either Þeld (Poe et al. 1978, Johnson et al. 1980, Oetting 1985, Hara 1986, Weintraub and Horowitz 1998, Civelek and Weintraub 2003) or laboratory (Bjorksten and Robinson 2005, Wu and Miyata 2005) studies and demonstrated toxic effects or the lack thereof on natural enemies. This study however looked at a larger number of pesticides commonly applied against at least six major pests in the greenhouse gerbera system and investigated their compatibility with

7 1596 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 106, no. 4 Table 6. Means ( SE) of number of live natural enemies (D. isaea and N. californicus) at each observation time of 12, 24, and 48 h in each of three trials (trials 1, 2, and 3) after exposure to fungicides at median label rates out of 10 natural enemies in each experimental unit Treatment D. isaea N. californicus Trial 1 Trial 2 Trial 3 Trial 1 Trial 2 Trial 3 12 h Control a b a a a a Sulfur a a a abc abc bc Piperalin a ab a 0.0 0e 0.0 0e c Pyraclostrobin a a a cd cd ab Fosetyl-aluminum a a a de abcd bc Copper sulfate a ab a abc bcd bc Butanone a ab a ab ab a Potassium bicarbonate a a a bcd d ab Azoxystrobin a ab a bcd ab a Rosemary oil a ab a 0.0 0e e c F P value * h Control ab b a a a a Sulfur ab ab a b c bcd Piperalin ab ab a 0.0 0c 0.0 0d d Pyraclostrobin ab a a c d bc Fosetyl-aluminum ab a a c d cd Copper sulfate ab ab a b b cd Butanone ab ab a ab a a Potassium bicarbonate a ab a 0.0 0c d bc Azoxystrobin ab ab a c d ab Rosemary oil b b a 0.0 0c 0.0 0d d F P value ** * h Control a bc a a a a Sulfur ab a b c b Piperalin a ab a 0.0 0c 0.0 0c 0.0 0b Pyraclostrobin a a a 0.0 0c 0.0 0c b Fosetyl-aluminum a ab a 0.0 0c c b Copper sulfate a ab a b b b Butanone a abc a ab ab a Potassium bicarbonate a ab a 0.0 0c 0.0 0c 0.0 0b Azoxystrobin ab ab a 0.0 0c c b Rosemary oil b c a 0.0 0c 0.0 0c 0.0 0b F P value *, P 0.001; **, P natural enemies that have the potential of controlling the two most important pests. Effects on D. isaea. As L. trifolii are often chemically resistant, most of the chemicals labeled for use against them rarely control populations to a signiþcant level. Growers often rely on pesticides as the only solution to pest problems, as they (when effective) allow for tangible and observable effects immediately, as opposed to biological control methods, which take more time and do not eliminate a pest completely. The knowledge that novaluron, petroleum oil, azadirachtin, cyromazine, and acetamiprid are at most slightly harmful to the leafminer parasitoid could encourage the use of such chemicals for leafminer control when inevitable. Mites are the most commonly encountered among the secondary pests in this system, and chemicals are effective in controlling them. A majority of the tested miticides were potentially harmless to D. isaea. That abamectin is toxic to parasitoids has been shown previously (Hara 1986, Schuster 1994, Bjorksten and Robinson 2005, Kaspi and Parrella 2005). Our results on the effect of spinosad corroborate similar Þndings in protected cultivation (Jones et al. 2005) and Þeld situations, where high mortality was observed in hymenopterans despite it being accepted by many as a biorational pesticide (Williams et al. 2003). This also cautions and emphasizes the importance of individual components of an integrated management program in cut ßowers. Spirotetramat and spiromesifen demonstrated potential as whiteßy insecticides that could integrate with biological control of the leafminer. However, both are in the insecticide class 23, which inhibits acetyl CoA carboxylase (IRAC 2011). As a thrips control material, ßonicamid, cyßuthrin, acetamiprid, insecticidal soap, and B. bassiana were seemingly safe to the leafminer parasitoid, but from a growerõs perspec-

8 August 2013 ABRAHAM ET AL.: PESTICIDE COMPATIBILITY WITH GREENHOUSE BIOCONTROL 1597 tive, the natural products are not Þrst choice options because they do not immediately show effects. However, ßonicamid, a feeding blocker, class 9c chemical (IRAC 2011); cyßuthirin, a pyrethroid; and acetamiprid, a neonicotinoid, could all be effective options. Spinosad is effective for thrips control (Jones et al. 2005), but demonstrated negative effects on parasitoid populations. Fungicides, in general, were found to cause low mortality in D. isaea. EcoSmart, a readyto-use rosemary oil concoction, and sulfur were the only fungicides (Tables 1 and 6) that caused signiþcant mortality in D. isaea, but still usable in IPM programs. Our data suggest that fungicides do not cause immediate negative effects on leafminer parasitoids. Effects on N. californicus. Unless a miticide specifically toxic to pest mite species is available, integration of miticides and predatory mites would not be possible in an IPM program. Cyromazine is accepted as being safe for natural enemies in general (Biobest 2011, Koppert 2011), and our study noted the same. However, we observed heightened activity by the surviving mites in the vial closer to the lid. Whether the phenomenon is a synergistic or repellent effect needs closer investigation. Spiromesifen and spirotetramat add to the number of rotational options as whiteßy chemicals. Among commonly used thrips control materials, only ßonicamid and insecticidal soap showed potential to integrate with pest mite biocontrol. Whereas miticides, in general, were not completely toxic to the insect natural enemy (leafminer parasitoid), insecticides, in general, seemed to harm the noninsect natural enemy (predatory mite). Reevaluating our control options from the available compatible chemistries to effectively rotate, and convincing growers to adopt only those options in an IPM program would be the challenge going forward. Greenhouse Miticide Study. Our study showed that the residual effect of miticides was not detrimental to D. isaea in the long run. Even though the parasitism rate dropped to 30% in the second week, the fact that the ßuctuation occurred in all treatments, including the control, and that there were no differences in other parameters that were analyzed, indicates that the effect was because of life history traits. All the treatments followed a similar pattern and reached peak parasitism by the fourth week. This meant that the miticides did not detrimentally affect D. isaea development in the weeks before (second or third week) when the parasitoids were in younger and more vulnerable stages. Bifenazate, hexythiazox, spiromesifen, acequinocyl, etoxazole, and clofentazine are hence not injurious in the long run for the development and population buildup of D. isaea. This gives us valuable information for integrating biological and chemical control in this system. The primary pest can be controlled using its natural enemy, and the major secondary pest can be controlled by rotating safe chemicals that do not harm the leafminer parasitoid, D. isaea. Additionally, from these results (Table 1), we would be able to use less disruptive options from among the chemicals to control other secondary pests. The beneþts from such a strategy are multifold: 1) reduced pesticide footprint in the premises and environment, 2) enhanced safety to the workers and producers alike, 3) better management of the pest and diseases leading to a better crop, and 4) overall a sustainable production system. Acknowledgments We greatly appreciate the technical assistance of Tina Thomas, Monica Townsend, and Sherrie Stevens in the conduct of these experiments. Funding was provided by the Georgia Department of Agriculture as a Specialty Crop Initiative Grant. References Cited Bethke, J. A., and R. A. Cloyd Pesticide use in ornamental production: what are the beneþts? Pest Manage. Sci. 65: 345Ð350. Biobest Side effects manual. ( biobest.be/neveneffecten/3/search-itmq/). Bjorksten, T. A., and M. Robinson Juvenile and sublethal effects of selected pesticides on the leafminer parasitoids Hemiptarsenus varicornis and Diglyphus isaea (Hymenoptera: Eulophidae) from Australia. J. Econ. Entomol. 98: 1831Ð1838. Civelek, H. S., and P. G. Weintraub Effects of bensultap on larval serpentine leafminers, Liriomyza trifolii (Burgess) (Diptera: Agromyzidae), in tomatoes. Crop Prot. 22: 479Ð483. Cloyd, R. A Compatibility of insecticides with natural enemies to control pests of greenhouses and conservatories. J. Entomol. Sci. 41: 189. Cloyd, R. A., and A. Dickinson Effect of insecticides on mealybug destroyer (Coleoptera: Coccinellidae) and parasitoid Leptomastix dactylopii (Hymenoptera: Encyrtidae), natural enemies of citrus mealybug (Homoptera: Pseudococcidae). J. Econ. Entomol. 99: 1596Ð1604. Hara, A. H Effects of certain insecticides on Liriomyza trifolii (Burgess) (Diptera: Agromyzidae) and its parasitoids on chrysanthemums in Hawaii. Proc. Hawaii. Entomol. Soc. 26: 65Ð70. IRAC IRAC MoA classiþcation scheme, pp. 1Ð23. ( 09/MoA_ClassiÞcation.pdf). Johnson, M. W., E. R. Oatman, and I. A. Wyman Effects of insecticides on populations of the vegetable leafminer and associated parasites on summer pole tomatoes. J. Econ. Entomol. 73: 61Ð66. Jones, T., C. S. Dupree, R. Harris, L. Shipp, and B. Harris The efþcacy of spinosad against the western ßower thrips, Frankliniella occidentalis, and its impact on associated biological control agents on greenhouse cucumbers in southern Ontario. Pest Manage. Sci. 61: 179Ð185. Kaspi, R., and M. P. Parrella Abamectin compatibility with the leafminer parasitoid Diglyphus isaea. Biol. Control 35: 172Ð179. Keil, C. B., and M. P. Parrella Liriomyza trifolii on chrysanthemums and celery: managing an insecticide resistant population, pp. 162Ð167. In S. L. Poe (ed.), 3rd Annual Industry Conference on the Leafminer, San Diego, CA. Koppert, B. S Side effects. ( koppert.nl/). Liu, T. -X., L. Kang, K. M. Heinz, and J. T. Trumble Biological control of Liriomyza leafminers: progress and

9 1598 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 106, no. 4 perspective. Perspect. Agric. Vet. Sci. Nutr. Natural Resour. 4: 1Ð16. Medina, P., G. Smagghe, F. Budia, L. Tirry, and E. Vinuela Toxicity and absorption of azadirachtin, dißubenzuron, pyriproxyfen, and tebufenozide after topical application in predatory larvae of Chrysoperla carnea (Neuroptera: Chrysopidae). Environ. Entomol. 32: 196Ð203. Oetting, R. D Effects of insecticides applied to potting media on Oenonogastra microrhopalae (Ashmead) parasitization of Liriomyza trifolii (Burgess). J. Entomol. Sci. 20: 405Ð410. Parrella, M. P., V. P. Jones, R. R. Youngman, and L. M. LeBeck Effect of leaf mining and leaf stippling of Liriomyza spp. on photosynthetic rates of chrysanthemum. Ann. Entomol. Soc. Am. 78: 90Ð93. Poe, S. L., P. H. Everett, D. J. Schuster, and C. A. Musgrave Insecticidal effects on Liriomyza sativae larvae and their parasites on tomato. J. Ga. Entomol. Soc. 13: 322Ð327. Reitz, S. R., and J. T. Trumble InterspeciÞc and intraspeciþc differences in two Liriomyza leafminer species in California. Entomol. Exp. Appl. 102: 101Ð113. Table A1. Schuster, D. J Life-stage speciþc toxicity of insecticides to parasitoids of Liriomyza trifolii (Burgess) (Diptera: Agromyzidae). Int. J. Pest Manage. 40: 191Ð194. Stark, J. D., R. Vargas, and J. E. Banks Incorporating ecologically relevant measures of pesticide effect for estimating the compatibility of pesticides and biocontrol agents. J. Econ. Entomol. 100: 1027Ð1032. Weintraub, P. G., and A. R. Horowitz Effects of translaminar versus conventional insecticides on Liriomyza huidobrensis (Diptera: Agromyzidae) and Diglyphus isaea (Hymenoptera: Eulophidae) populations in celery. J. Econ. Entomol. 91: 1180Ð1185. Williams, T., J. Valle, and E. Viñuela Is the naturally derived insecticide Spinosad compatible with insect natural enemies? Biocontrol Sci. Technol. 13: 459Ð475. Wu, G., and T. Miyata Susceptibilities to methamidophos and enzymatic characteristics in 18 species of pest insects and their natural enemies in crucifer vegetable crops. Pest. Biochem. Physiol. 82: 79Ð93. Received 16 December 2012; accepted 3 May ANOVA of number of natural enemies (D. isaea, and N. californicus) alive 12, 24, and 48 h after chemical treatments Treatments Nontarget Run Time (h) F df P value Treatments Replicates Miticides D. isaea , , , , , , , , , N. californicus , , , , , , , , , Leafminer materials D. isaea , , , , , , , , , N. californicus , , , , , , , , , Continued on following page

10 August 2013 ABRAHAM ET AL.: PESTICIDE COMPATIBILITY WITH GREENHOUSE BIOCONTROL 1599 Table A1. Continued Treatments Nontarget Run Time (h) F df P value Treatments Replicates Thripicides D. isaea , , , , , , , , , N.californicus , , , , , , , , , Whiteßy materials D. isaea , , , , , , , , , N.californicus , , , , , , , , , Fungicides D. isaea , , , , , , , , N. californicus , , , , , , , , , Table A2. D. isaea ANOVA and means ( SE) for the residual toxicity of miticides, over a 4-wk period, on average percent of parasitism by Treatments (a.i.) Wk 1 Wk 2 Wk 3 Wk 4 Control a a a a Hexythiazox a a a 100 0a Bifenazate a a a a Etoxazole a a a a Spiromesifen a a a a Acequinocyl a a a a Clofentezine a a a a df 6, 41 6, 41 6, 41 6, 41 F value P value

11 1600 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 106, no. 4 Table A3. ANOVA and means ( SE) for the residual toxicity of miticides, over a 4-wk period, on average L. trifolii populations Treatments (ai) Wk 1 Wk 2 Wk 3 Wk 4 Control a a a a Hexythiazox a a a 0.0 0a Bifenazate a a a a Etoxazole a a a a Spiromesifen a a a a Acequinocyl a a a a Clofentezine a a a a df 6, 41 6, 41 6, 41 6, 41 F value P value Table A4. (D. isaea) ANOVA and means ( SE) for the residual toxicity of miticides, over a 4-wk period, on average population of parasitoid Treatments (ai) Wk 1 Wk 2 Wk 3 Wk 4 Control a a a a Hexythiazox a a a a Bifenazate a a a a Etoxazole a a a a Spiromesifen a a a a Acequinocyl a a a a Clofentezine a a a a df 6, 41 6, 41 6, 41 6, 41 F value P value

12 August 2013 ABRAHAM ET AL.: PESTICIDE COMPATIBILITY WITH GREENHOUSE BIOCONTROL 1601 Table A5. ANOVA and means ( SE) for the residual toxicity of miticides, over a 4-wk period, on sum of live leafminers in the three sampled leaves from each experimental unit Treatments (a.i.) Wk 1 Wk 2 Wk 3 Wk 4 Control a a a a Hexythiazox a a a 0.0 0a Bifenazate a a a a Etoxazole a a a a Spiromesifen a a a a Acequinocyl a a a a Clofentezine a a a a df 6, 41 6, 41 6, 41 6, 41 F value P value Table A6. ANOVA and means ( SE) for the residual toxicity of miticides, over a 4-wk period, on sum of leafminers live and dead ( number of parasitoids) Treatments (a.i.) Wk 1 Wk 2 Wk 3 Wk 4 Control a a a a Hexythiazox a a a a Bifenazate a a a a Etoxazole a a a a Spiromesifen a a a a Acequinocyl a a a a Clofentezine a a a a df 6, 41 6, 41 6, 41 6, 41 F value P value