Biological Control 60 (2012) 39–45

Contents lists available at SciVerse ScienceDirect

Biological Control

journal homepage: www.elsevier .com/locate /ybcon

Variability in response of four populations of Amblyseius largoensis(Acari: Phytoseiidae) to Raoiella indica (Acari: Tenuipalpidae) and Tetranychus gloveri(Acari: Tetranychidae) eggs and larvae

Daniel Carrillo a,⇑, Martha E. de Coss b, Marjorie A. Hoy c, Jorge E. Peña a

a University of Florida, Department of Entomology and Nematology, Tropical Research and Education Center, Homestead, FL 33031, USAb Universidad Autónoma de Chiapas, Terán Tuxtla Gutiérrez, Chiapas, CP 29050, Mexicoc University of Florida, Department of Entomology and Nematology, Gainesville, FL 32611, USA

h i g h l i g h t s

" Amblyseius largoensis with differentprevious exposure to Raoiella indicawere tested.

" Predators consumed high numbersof R. indica eggs regardless ofprevious exposure.

" Experienced predators consumedmore R. indica larvae than naïvepredators.

" Plasticity in response could beassociated to genetic selection,learning, or both.

" Implications for biological control ofthe invasive species R. indica arediscussed.

1049-9644/$ - see front matter � 2011 Elsevier Inc. Adoi:10.1016/j.biocontrol.2011.09.002

⇑ Corresponding author. Fax: +1 305 2467003.E-mail address: dancar@ufl.edu (D. Carrillo).

g r a p h i c a l a b s t r a c t

a r t i c l e i n f o

Article history:Received 24 May 2011Accepted 2 September 2011Available online 10 September 2011

Keywords:Invasive speciesRaoiella indicaAmblyseius largoensisLearningSelection

a b s t r a c t

Raoiella indica (Acari: Tenuipalpidae) is a phytophagous mite that recently invaded the Neotropicalregion. A predatory mite Amblyseius largoensis (Acari: Phytoseiidae) has been found associated with R.indica in Florida. This study evaluated A. largoensis by determining its likelihood of consuming eggsand larvae of R. indica and Tetranychus gloveri (Acari: Tetranychidae) under no-choice and choice condi-tions. To detect variations in the response of A. largoensis to R. indica, four populations of predators wereexamined: (1) predators reared exclusively on R. indica in the laboratory for 2 years, (2) predators rearedon T. gloveri in the laboratory for 2 months but reared on R. indica for 2 years previously, (3) predatorscollected from a field infested with R. indica, and (4) predators collected from a field that had never beeninfested with R. indica. Results of this study suggest that A. largoensis is likely to accept and consume highnumbers of R. indica eggs regardless of their previous feeding experience. In contrast, all populations con-sumed relatively fewer R. indica larvae than the other prey tested. Predators previously exposed to R.indica were more likely to consume R. indica larvae. By contrast, predators not previously exposed to R.indica showed the lowest likelihood of choosing to feed on this prey item. Plasticity in the response ofA. largoensis to R. indica larvae could be associated to selection, learning, or a combination of both. Thepossible implications of the observed differences in terms of biological control of R. indica are discussed.

� 2011 Elsevier Inc. All rights reserved.

ll rights reserved.

40 D. Carrillo et al. / Biological Control 60 (2012) 39–45

1. Introduction

Raoiella indica Hirst (Acari: Tenuipalpidae) is a phytophagouspest native to tropical and subtropical areas of Asia that recentlyinvaded the Neotropical region (Etienne and Flechtmann, 2006;Vásquez et al., 2008; Marsaro et al., 2009). This mite is a multivol-tine and gregarious species that can reach high population densi-ties and cause significant damage to a variety of plant species,especially palms (Arecaceae). Establishment of this invasive spe-cies in the Neotropical region has given rise to concerns about itspotential impact on a number of economically and ecologicallyimportant plants (Carrillo et al., in press).

Efforts were undertaken to identify and evaluate potentialbiological control agents of R. indica in the Neotropical region(Peña et al., 2009; Carrillo et al., 2010). The predatory miteAmblyseius largoensis Muma (Acari: Phytoseiidae) has been foundassociated with R. indica in several areas of recent invasionincluding Florida, Puerto Rico, Trinidad and Tobago, Colombia,Cuba, and Mexico. Moreover, this predator has been found asso-ciated with R. indica in India (Taylor et al., in press), Philippines(Gallego et al., 2003), Mauritius and Dominica (Hoy, personalcommunication), Benin and Tanzania (Zannou et al., 2010). Itwould be important to know if all the predators identified asA. largoensis are equally efficient natural enemies of R. indica.The geographically diverse distribution of this phytoseiidsuggests that strains, biotypes, or cryptic species could exist(Noronha and Moraes, 2004; Bowman, 2010), and thatdifferences in predatory behaviors among the various popula-tions of A. largoensis could affect their propensity to feed on R.indica, as has been found for other phytoseiid species.

Previous studies suggested that R. indica can be considered anappropriate prey for Florida populations of A. largoensis in termsof its reproductive success and consumption rates (Carrillo et al.,2010; Carrillo and Peña, in press). The predator also showed gooddevelopmental and reproductive parameters when feeding on Tetr-anychus gloveri Banks (Acari: Tetranychidae) (Carrillo et al., 2010).However, all previous experiments were conducted offering preyin a no-choice condition. In the present study, A. largoensis was of-fered the choice between eggs and larvae of R. indica and T. gloveri,which represents a more realistic test for this predator that has tochoose among prey found on coconut leaves in Florida.

Knowledge of the mechanisms that underlie prey acceptanceand choice in predatory mites has recently expanded. Studies onthe predatory mites Hypoaspis aculeifer (Canestrini) (Acari: Laelap-idae) and Phytoseiulus persimilis Athias-Henriot (Acari: Phytoseii-dae) demonstrated that foraging traits and prey preferencescould be genetically determined and that genetic polymorphismmay occur within local populations (Lesna and Sabelis, 1999,2002; Jai et al., 2002; Tixier et al., 2010). In addition, there isincreasing evidence suggesting that learning plays an importantrole in prey recognition and acceptance in phytoseiid mites(Rahmani et al., 2009; Schausberger et al., 2010). Until now allinformation on predation by A. largoensis upon R. indica is limitedto a single population of predators that was collected from a coco-nut plantation in Florida in 2007 and used to initiate a laboratorycolony. Because prolonged laboratory rearing could affect the preypreference of phytoseiid mites (Mwansat, 2001), it is important todetermine whether other populations of A. largoensis couldrespond differently to R. indica.

This study was designed to determine whether the feeding his-tory of Florida populations of A. largoensis could influence theirlikelihood of utilizing the invasive R. indica as prey. To detect pos-sible behavioral plasticity in the response of A. largoensis to R. indi-ca this study evaluated four populations of A. largoensis that haddifferent feeding histories (experience feeding on R. indica).

2. Materials and methods

2.1. General experimental procedures

Experiments were conducted at 26.5 ± 1 �C, 70 ± 5% RH under a12:12 L:D photoperiod. The experimental arenas consisted of rect-angles (4 � 2.5 cm) cut from mature unblemished coconut (Cocosnucifera L.) leaves. Leaf rectangles were placed with the abaxialsurface facing up on cotton squares (6 � 6 � 2 cm) saturated withwater in a plastic tray (hexagonal polystyrene weighing dishes12.7/8.9 cm, Fisherbrand� Cat. No. 02-202-103). Paper strips(Kimwipe�, Kimberly–Clark Corporation, Roswell, GA) were placedalong the edges of the leaf squares to minimize mite escape. Asmall plastic square (0.5 � 0.5 cm) was placed on top of each arenaas a shelter for A. largoensis.

2.2. Prey mites

The two prey species, R. indica and T. gloveri, were obtained fromstock colonies reared on potted coconut palms kept in separategreenhouses following the procedure previously described (Carrilloet al., 2010). Eggs of both prey were obtained by placing 100 ovipos-iting females on each arena and allowing them to deposit eggs for5 days. After this period the females were removed and the desirednumber of eggs adjusted by removing excess eggs with a fine brush.In the case of T. gloveri, extensive webbing resulted from thecrowded conditions of females on the ovipositing arenas. Suchextensive webbing is not observed under ‘natural’ field conditions,so a layer of webbing was carefully removed to expose T. gloveri eggsand allow more common conditions for A. largoensis to access itsprey. Larvae of both prey species were transferred from the stockcolonies onto the arenas according to the desired prey ratios. Arenaswith two prey species had an additional moistened paper strip onthe mid vein of the leaf to allow separate placement of prey, butthe strip was removed just before releasing the predators.

2.3. Predatory mites

Four populations of A. largoensis from Florida with differentfeeding histories (previous diet) were used. Two populations werelaboratory colonies that had been fed different diets (followingthe procedures previously described, Carrillo et al., 2010). The firstcolony had been fed exclusively with R. indica for more than 2 yearsand will be referred to as the R-colony. This colony was initiated inNovember 2008 with 150 predators obtained from a coconut palmplantation (Malayan dwarf variety, pesticide-free, 26�02.93 N,80�09.82 W) that had been infested with R. indica since 2007. Indi-viduals collected from the same site and confirmed as A. largoensiswere added to the colony (�10 females/rearing plate) every6 months. The second laboratory colony was initiated with 50 fe-males obtained from the R-colony which were subjected to a dietshift from R. indica to T. gloveri 2 months prior to testing, whichwould have allowed the predators to have undergone approxi-mately six generations on the new diet. This colony will be referredto as the T-colony. The other two populations were collected fromcoconut palm plantations and maintained in the laboratory for1 week prior to testing on arenas constructed with the same leavesand prey that were present at the time they were collected. The firstfield population was obtained from the same coconut plantationwhere 2 years before A. largoensis was obtained to initiate the R-colony and will be referred to as the experienced field populationbecause of its long-term previous exposure to R. indica, althoughwe cannot exclude the possibility that this population fed on pollenand T. gloveri, as well. The second field population was collectedfrom a pesticide-free coconut plantation (Malayan dwarf variety)

D. Carrillo et al. / Biological Control 60 (2012) 39–45 41

that had never been infested with R. indica (25�35.49 N,80�04.03 W, Tropical Research and Education Center, Universityof Florida, Homestead) and will be referred to as the naïve field pop-ulation for its lack of previous exposure to R. indica. The naïve fieldpopulation was approximately 80 km away from any A. largoensispopulation that could have fed on R. indica and most likely, it hadfed on pollen, T. gloveri, Oligonychus aff. modestus and Oligonychussp. (Acari: Tetranychidae); Nipaecoccus nipae (Maskell) (Hemiptera:Pseudococcidae); Aonidiella orientalis (Newstead) (Hemiptera:Diaspididae); and Aleurocanthus woglumi Ashby (Hemiptera: Aley-rodidae) (Peña et al., 2009; Carrillo et al., 2010).

2.4. No-choice tests

No-choice tests were included to ensure that each population ofpredators could feed on R. indica and T. gloveri and to determinewhether feeding experience and laboratory rearing could affectprey consumption. Predators from the four populations were of-fered separately 50 eggs or 50 larvae of either R. indica or T. gloveri.The number of prey offered was chosen based on a previous studyin which A. largoensis showed a maximum prey consumption of 45R. indica eggs/day (Carrillo and Peña, in press). Our experiments didnot involve a test with R. indica nymphs and adults as prey. How-ever, previous studies indicated that female A. largoensis (i.e., the R-colony) had difficulty handling nymphs or adults and consumedsignificantly more R. indica eggs or larvae than nymphs and adults(Carrillo and Peña, in press). No choice experiments were repli-cated 10 times. In addition, 10 absolute controls consisting of are-nas with prey but without predators were included to ensure thatmortality was caused by predation.

2.5. Choice tests

Prey preference of the four populations was addressed bychoice experiments offering 50 individuals of each prey item inthe following combinations: R. indica eggs and T. gloveri eggs,R. indica eggs and T. gloveri larvae, R. indica larvae and T. gloverieggs, and R. indica larvae and T. gloveri larvae. Prey type(s) wererandomly assigned to either side of the arena to discard any effectof prey location on the choice by A. largoensis. Fifteen replicates pertreatment (prey combination) were used, including the samenumber of absolute controls to ensure that prey mortality wascaused by predation.

Predators used in all experiments were A. largoensis mated fe-males in their oviposition peak (3–5 days after adult emergence).To ensure females were mated and of the same age deutonymphswere isolated and males added to the rearing arenas. In addition,only replicates in which the female remained within the test arenaand produced at least one egg within 48 h were considered foranalyses. Females were starved for 8 h before being transferredto the experimental arenas. The numbers of each R. indica orT. gloveri stage consumed during a period of 24 h were recordedby counting the number of shriveled corpses (larvae) and bysubtracting the number of eggs remaining from the number of eggsprovided. A period of 24 h was selected to ensure a behavioral basisin the response because longer term tests could be influenced bypost-ingestive processes that are not necessarily a reflection ofthe predator’s prey preference.

2.6. Data analysis

Data were separately analyzed for each experiment using SAS9.2 (SAS Institute Inc., 2010). Predation by the four populationsof A. largoensis in the no-choice experiments was analyzed byKruskal–Wallis tests due to variance heterogeneity and non-

normality of data. The likelihood of choosing one prey item overanother (choice experiments) was quantified using the preferenceindex b proposed by Manly et al. (1972):

b ¼ln N0

N0c

� �

ln NNc

� �þ 1

24

35�1

where N and N0 are the numbers of each prey-stage provided and Nc

and N0c are the numbers of each prey consumed. The index assignspreference values from 0 to 1, where 0.5 represents no preference.Mean b-values were considered significant when 95% confidenceintervals based on the t-distribution did not overlap with b = 0.5.Preference (b) values were normally distributed (Kolmogorov–Smirnov P > 0.05) and had homogeneous variances (Levene testP > 0.05) in the experiments involving choice between R. indica lar-vae and T. gloveri eggs or larvae. Differences in the prey preference(b) among the four populations of A. largoensis were analyzedthrough ANOVAs and means separated by Tukey’s test. Data ofthe experiments involving choice tests between R. indica eggs andT. gloveri eggs or larvae were not normally distributed, so Krus-kal–Wallis tests were used in this case.

3. Results

Prey survival in the absolute controls was always greater than95%, which ensures that mortality in the choice and no-choiceexperiments was caused by predation by A. largoensis females.

3.1. No-choice tests

The four populations of A. largoensis consumed equivalent num-bers of R. indica eggs, with consumption ranging from 44 to 47 eggsper female/day regardless of their previous feeding experience, [H(3,N = 40) = 3.90, P = 0.27] (Table 1). By contrast, the four popula-tions consumed relatively fewer R. indica larvae (ranging from 9to 23 larvae per female/day) compared with the other prey itemsoffered and they differed in their predation rates (P < 0.001, Table1). The naïve field population, not previously exposed to R. indica,consumed significantly fewer R. indica larvae than the other threepopulations [H (3,N = 40) = 19.0, P < 0.0001]. With respect to utili-zation of T. gloveri as prey, the four populations preyed on 39–44T. gloveri eggs per female/day, showing no statistical differencesamong the predator populations [H (3,N = 40) = 2.96, P = 0.4]. Con-sumption of T. gloveri larvae ranged from 36 to 46 larvae per fe-male/day. The R-colony, fed on R. indica for 2 years withoutexposure to T. gloveri prey, consumed significantly fewer T. gloverilarvae than the other three populations [H (3,N = 40) = 7.5, P = 0.05].

3.2. Choice tests

All four populations of A. largoensis consumed more R. indicaeggs than T. gloveri eggs or larvae (Fig. 1A and B). Consequently,preference indices (b) were significant (not overlapping withb = 0.5) and favoring R. indica eggs without significant differencesamong the four populations tested [H (3,N = 52) = 0.68, P = 0.871)and H (3,N = 58) = 6.54 P = 0.08 for the R. indica eggs vs. T. gloverieggs and R. indica eggs vs. T. gloveri larvae, respectively] (Fig. 2Aand B).

The likelihood that the four A. largoensis populations would con-sume R. indica larvae was lower and differed among them (Fig. 1Cand D). Preference analysis showed that the R-colony, fed on R.indica for 2 years, was more likely to choose to prey on R. indica lar-vae than the other three populations tested, which showed signifi-cant preferences for T. gloveri eggs (F = 5.29; df = 3, 52; P < 0.003,Fig. 2C). Finally, in the R. indica and T. gloveri larval-choice test, three

Table 1Predation of R. indica and T. gloveri eggs and larvae by four local populations of A. largoensis with different feeding history (previous diet) under no-choice conditions. Meannumber prey items consumed in 24 h Means ± S.D. Means within a column followed by the same letter are not significantly different (P < 0.05; Kruskal-Wallis test).

Amblyseius largoensis population Number of individuals consumed/day

R. indica eggs R. indica larvae T. gloveri eggs T. gloveri larvae

R-colony 47.2 ± 2.7 a 23.4 ± 6.7 a 39.4 ± 8.2 a 35.5 ± 11.1 bExperienced field 44.4 ± 4.2 a 24.3 ± 7.1 a 40.5 ± 3.9 a 41.7 ± 6.4 aT-colony 46.6 ± 3.5 a 21.7 ± 5.4 a 42.6 ± 5.8 a 46.3 ± 2.3 aNaïve field 45.5 ± 3.0 a 9.2 ± 4.8 b 43.7 ± 5.0 a 41.3 ± 9.6 aH (3,N = 40) 3.9 18.9 2.96 7.5P value 0.3 <0.001 0.4 0.05

Fig. 1. Percentage of individuals of R. indica (black bars) and T. gloveri (grey bars) consumed by four populations of A. largoensis with different feeding history (previous diet)under choice conditions. (A) R. indica eggs vs. T. gloveri eggs, (B) R. indica eggs vs. T. gloveri larvae, (C) R. indica larvae vs. T. gloveri eggs, and (D) R. indica larvae vs. T. gloverilarvae.

42 D. Carrillo et al. / Biological Control 60 (2012) 39–45

populations of A. largoensis showed no significant differences intheir consumption rates or preference among these prey items(Fig. 2D). However, the naïve field population, not previously ex-

posed to R. indica, showed a low consumption of R. indica larvae(Figs.1C and 2D) and a marked preference for T. gloveri larvae(F = 5.29; df = 3, 40; P = 0.004) (Fig. 2D).

Fig. 2. Prey preference of four local populations of A. largoensis with different feeding history (previous diet) when preying on R. indica and T. gloveri eggs and larvae underchoice conditions. (A) R. indica eggs vs. T. gloveri eggs, (B) R. indica eggs vs. T. gloveri larvae, (C) R. indica larvae vs. T. gloveri eggs, and (D) R. indica larvae vs. T. gloveri larvae. Thepreference index b assigns preference values from 0 to 1, where 0.5 represents no preference. Mean b-values we considered significant when 95% confidence intervals (errorbars) based on the t-distribution did not overlap with b = 0.5. Differences in b-values among the four populations of A. largoensis are represented by letters were analyzedthrough ANOVAs or Kruskal–Wallis tests depending on normality of data.

D. Carrillo et al. / Biological Control 60 (2012) 39–45 43

4. Discussion

Our experiments showed that populations of A. largoensis didnot differ in the extent to which they fed on R. indica eggs, but ratherin the extent to which they fed on R. indica larvae. R. indica eggswere consumed in high numbers under no-choice conditions, andchosen over T. gloveri eggs and larvae by the four populations ofA. largoensis in the choice tests, regardless of their previous feedingexperience. Naïve predators, not previously exposed to R. indica,responded to R. indica eggs similarly as the laboratory colony fedexclusively on R. indica for 2 years, suggesting that the four popula-tions of A. largoensis rapidly associated R. indica eggs with a rewardand started to consume and prefer this prey item. Thus, the cuesused by A. largoensis to recognize and associate R. indica eggs witha reward must be well defined and easily assimilated by the pred-ator. Previous studies showed that oviposition rates of A. largoensisfemales (i.e., the R-colony) fed on a diet of R. indica eggs(2.36 ± 0.1153 eggs/day, Carrillo and Peña, in press) were higherthan oviposition rates when the predator was provided a diet ofmixed R. indica stages (1.63 ± 0.27 eggs/day, Carrillo et al., 2010),suggesting that a diet of R. indica eggs is more rewarding in termsof the predator’s reproductive success than a mixed diet. A similarcondition was observed with Metaseiulus occidentalis (Nesbitt)(Acari: Phytoseiidae), which showed better reproductive parame-ters when fed on Tetranychus pacificus McGregor (Acari: Tetranychi-dae) eggs compared to a diet of mixed immature stages(Bruce-Oliver and Hoy, 1990). Blackwood et al. (2001) testedseveral species of phytoseiid mites for preferences between eggsand larvae of Tetranychus urticae Koch (Acari: Tetranychidae) andsuggested that specialized spider mite predators had a preferencefor eggs whereas generalist predators had no prey-stage preference.Our results suggest that A. largoensis, despite being a generalistpredator (McMurtry and Croft, 1997), showed a marked preferencefor eggs of R. indica over eggs and larvae of T. gloveri. In addition,three out of four populations of A. largoensis preferred eggs of

T. gloveri over larvae of R. indica. Apart from the phytoseiidae, otherpredatory mites have shown preference for the egg stage of prey.For example, Poecilochirus carabi G. & R. Canestrini (Mesostigmata:Parasitidae) females are oophagous on Nicrophorus vespilloidesHerbst (Coleoptera: Silphidae) (Beninger, 1993) and Macrochelesperegrinus (Krantz) (Acari: Macrochelidae) showed a strong prefer-ence for eggs of various dipteran species (Roth et al., 1988). In addi-tion, the predatory insect Delphastus catalinae (Horn) (Coleoptera:Coccinellidae) showed a strong preference for Bemisia argentifoliiBellows and Perring (Homoptera: Aleyrodidae) eggs (Legaspiet al., 2006).

A. largoensis displayed ability to rapidly recognize and prey on R.indica eggs. By contrast, results of our study suggest that predatorsneed to habituate to R. indica before starting to utilize R. indicalarvae as prey. The four A. largoensis populations differed in theirlikelihood of preying upon R. indica larvae. The laboratory R-colonyshowed the highest likelihood preying on R. indica larvae in thechoice experiments, which could have resulted from selectiondue to it having access only to R. indica for over 2 years (ca. 50 gen-erations). Interestingly, this selection was not eliminated by2 months of rearing exclusively on T. gloveri in the T-colony,suggesting that the changes in feeding behavior of A. largoensiscould have a genetic basis. In general, the three populations thathad a ‘long term’ exposure to R. indica (i.e., R-colony, T-colony,and the experienced field population) had a higher likelihood ofpreying on R. indica larvae than the naïve field population withno previous exposure to R. indica, suggesting that its lack ofexperience in preying on R. indica might have affected its abilityto consume this prey.

Overall, results of our studies suggest that multiple mechanismscan be involved in the response of A. largoensis to R. indica. It is pos-sible that these mechanisms do not act independently but are con-ditionally expressed depending on the conditions that the predatorfaces (Sabelis and Lesna, 2010). For instance, it is likely that popu-lations of A. largoensis are subject to relatively long periods of expo-

44 D. Carrillo et al. / Biological Control 60 (2012) 39–45

sure to R. indica which has become by far the dominant arthropodon coconuts in areas of recent invasion. These circumstances couldfavor selection processes and the likelihood of A. largoensis preyingon R. indica active stages may increase. Experienced predators couldlearn to optimize their foraging activities and also increase the like-lihood of preying on R. indica larvae. However, when food qualitydeclines and predators are forced to disperse, dispersal might bringA. largoensis to a location where another food type is dominant.Thus, predators must retain the ability to switch prey (Sabelis andLesna, 2010). Our results suggest that A. largoensis can switch to no-vel prey after a period of starvation despite possible conditioning ofprey preferences due to genetic selection or experience. In conclu-sion, A. largoensis showed plasticity in its utilization of R. indicalarvae which could be related to selection, learning or a combina-tion of both. Moreover, the environmental conditions thatpredators face can modulate the expression of these.

Another factor that could affect the prey preference ofphytoseiid mites is the presence of feeding attractants and deter-rents (Hoy and Smilanick, 1981; Vet and Dicke, 1992). It has beensuggested that R. indica could produce repellent compounds be-cause of the presence of droplets located at the tips of the dorsalbody setae and at the tip of the egg’s pedicel (Zanotto andRodrigues, 2010). The marked preference for R. indica eggsobserved in the present study and the high ‘profitability’ of thisprey for A. largoensis (Carrillo et al., 2010) do not support thehypothesis of toxic or repellent compounds being present in thepedicel droplets on R. indica eggs. However, it is possible that thedroplets found on other R. indica larvae (and other life stages) couldhave a different composition and affect the ability of A. largoensis tofeed on them. The possible existence of kairomones or allomonesproduced by R. indica should be investigated to better understandtheir role on the prey choice of this predator. In addition, mites canhave other defense mechanisms that could directly influence theirattractiveness to predators. The hardness of the cuticle could serveas a defense mechanism against predators (Alberti and Crooker,1985). Observations made during the experiments showed thatA. largoensis usually required a single attack to successfully pene-trate the chorion and imbibe all of the egg contents. In contrast,A. largoensis females probed several times in order to penetratethe larval cuticle, had a longer handling time, and did not consumethe larval prey entirely. Thus, preference of A. largoensis forR. indica eggs and the lack of preference for larvae could be relatedto the energy spent and the nutritional reward obtained whenfeeding on each of them.

Findings of this study suggest that different populations of A.largoensis may be responding to the invasion by R. indica eitherby learning or adapting to be better predators of this invasive pest.These findings may be important for future biological control pro-grams targeting R. indica. Since R. indica gained importance as aninvasive pest in the neotropics, surveys for natural enemies in sev-eral places of the world (Roda et al., 2008; Peña et al., 2009; Ramoset al., 2010; Zannou et al., 2010; Carrillo et al., 2011; Taylor et al., inpress; Hoy, personal communication) have identified A. largoensisas the most abundant predator, and often the only phytoseiid spe-cies associated with R. indica. Further studies are needed to deter-mine the efficacy of these predator populations in controlling R.indica, and whether there are additional differences in populationsof A. largoensis with regard to their utilization of R. indica. Results ofthis study suggest that a long time of association between A. largo-ensis and R. indica could result in more aggressive strains of thispredator. The possible existence of more aggressive strains, bio-types or cryptic species of A. largoensis should be investigated. Inaddition, methods of increasing the effects of local populations ofA. largoensis on R. indica merit investigation. Techniques such asaugmentative release and provision of alternative food sourcesfor increasing predator populations in R. indica infested areas

should be explored. Moreover, research towards identifying chem-ical control practices that can preserve these natural enemies fortheir use in integrated management plans targeting R. indicashould be conducted.

Acknowledgments

We are thankful with Dr. Cal Welbourn (FSCA-DPI) for the iden-tification of mites, Dr. Na-Lampang Sikavas (UF-TREC) for help withthe statistical analyses, and Dr. Josep Jacas (UJI Spain) for helpfulreviews of the original manuscript. We thank K. Santos, R. Duncan,D. Long, A. Vargas, and J. Alegria, for their help. This research waspartially funded by a TSTAR Grant to JEP.

References

Alberti, G., Crooker, A.R., 1985. Internal Anatomy. In: Helle, W., Sabelis, M.W. (Eds.),Spider Mites: Their Biology, Natural Enemies, and Control, vol. A. Elsevier,Amsterdam, Netherlands, pp. 29–311.

Beninger, C.W., 1993. Egg predation by Poecilochirus carabi (Mesostigmata:Parasitidae) and its effect on reproduction of Nicrophorus vespilloides(Coleoptera: Silphidae). Environmental Entomology 22 (4), 766–769.

Blackwood, J.S., Schausberger, P., Croft, B.A., 2001. Prey-stage preference ingeneralist and specialist phytoseiid mites (Acari: Phytoseiidae) when offeredTetranychus urticae (Acari: Tetranychidae) eggs and larvae. EnvironmentalEntomology 30 (6), 1103–1111.

Bowman, H.M., 2010. Molecular discrimination of phytoseiids associated with thered palm mite Raoiella indica (Acari: Tenuipalpidae) from Mauritius and SouthFlorida. MS Thesis, University of Florida, Gainesville.

Bruce-Oliver, S.J., Hoy, M.A., 1990. Effect of prey stage on life-table attributes of agenetically manipulated strain of Metaseiulus occidentalis (Acari: Phytoseiidae).Experimental and Applied Acarology 9, 201–217.

Carrillo, D., Peña, J.E., in press. Prey-stage preferences and functional numericalresponses of Amblyseius largoensis (Acari: Phytoseiidae) to Raoiella indica (Acari:Tenuipalpidae). Experimental and Applied Acarology. doi:10.1007/s10493-011-9488-7.

Carrillo, D., Peña, J.E., Hoy, M.A., Frank, J.H., 2010. Development and reproduction ofAmblyseius largoensis (Acari: Phytoseiidae) feeding on pollen, Raoiella indica(Acari: Tenuipalpidae), and other microarthropods inhabiting coconuts inFlorida, USA. Experimental and Applied Acarology 52, 119–129.

Carrillo, D., Amalin, D., Hosein, F., Roda, A., Duncan, R., Peña, J.E., in press. Host plantrange of Raoiella indica Hirst (Acari: Tenuipalpidae) in areas of invasion of thenew world. Experimental and Applied Acarology. doi:10.1007/s10493-011-9487-8.

Carrillo, D., Navia, D., Ferragut, F., Peña, J.E., 2011. First report of Raoiella indica Hirst(Acari: Tenuipalpidae) in Colombia. Florida Entomologist 94 (2), 370–371.

Etienne, J., Flechtmann, C.H.W., 2006. First record of Raoiella indica (Hirst, 1924)(Acari: Tenuipalpidae) in Guadeloupe and Saint Martin, West Indies.International Journal of Acarology 32, 331–332.

Gallego, C.E., Aterrado, E.D., Batomalaque, C.G., 2003. Biology of the false spidermite, Rarosiella cocosae Rimando, infesting coconut palms in Camiguin, northernMindanao (Philippines). Philippines Entomology 17 (2), 187.

Hoy, M.A., Smilanick, J.M., 1981. Non-random prey location by the phytoseiidpredator Metasieulus occidentalis: differential responses to several spidermitespecies. Entomologia Experimentalis et Applicata 29, 241–253.

Jai, F., Margolies, D.C., Boyer, J.E., Charlton, R.E., 2002. Genetic variation in foragingtraits among inbred lines of a predatory mite. Heredity 89, 371–379.

Legaspi, J.C., Simmons, A.M., Legaspi, B.C., 2006. Prey preference by Delphastuscatalinae (Coleoptera: Coccinellidae) on Bemisia argentifolii (Homoptera:Aleyrodidae): effects of plant species and prey stages. Florida Entomologist 89(2), 218–222.

Lesna, I., Sabelis, M.W., 1999. Diet dependent female choice for males with ‘goodgenes’. Nature 401, 581–584.

Lesna, I., Sabelis, M.W., 2002. Genetic polymorphism in prey preference at a smallspatial scale: a case study of soil predatory mites (Hypoaspis aculeifer) and twospecies of astigmatic mites as prey. In: Bernini, F., Nannelli, R., Nuzzaci, G., deLillo, E. (Eds.), Acarid Phylogeny and Evolution: Adaptation in Mites and Ticks –Proceedings of the IV Symposium of the European Association of Acarologists.Kluwer Academic Publishers, Dordrecht, The Netherlands, pp. 49–64.

Manly, B.F.J., Miller, P., Cook, L.M., 1972. Analysis of a selective predationexperiment. The American Naturalist 106, 719–736.

Marsaro Jr., A.L., Navia, D., Gondim Jr., M.G., Silva, F.R., Moraes, G.J., 2009. Chegou aoBrasil – o ácaro vermelho das palmeiras Raoiella indica. Cultivar, Hortaliças eFrutas 57, 31.

McMurtry, J.A., Croft, B.A., 1997. Life-styles of phytoseiid mites and their role inbiological control. Annual Review of Entomology 42, 291–321.

Mwansat, G.S., 2001. Prolonged laboratory rearing as it affects host preference andreproductive capacity of Neoseiulus idaeus Denmark and Muma (Acari:Phytoseiidae). Bioscience Research Communication 13 (3), 213–222.

D. Carrillo et al. / Biological Control 60 (2012) 39–45 45

Noronha, A.C.D.S., Moraes, G.J.D., 2004. Reproductive compatibility between mitepopulations previously identified as Euseius concordis (Acari: Phytoseiidae).Experimental and Applied Acarology 32, 271–279.

Peña, J.E., Rodrigues, J.C.V., Roda, A., Carrillo, D., Osborne, L.S., 2009. Predator-preydynamics and strategies for control of the red palm mite (Raoiella indica) (Acari:Tenuipalpidae) in areas of invasion in the Neotropics. In: Proceedings of thesecond Meeting of IOBC/WPRS, Work Group Integrated Control of Plant FeedingMites. Florence, Italy, 9–12 March. pp. 69–79.

Rahmani, H., Hoffman, D., Walzer, A., Shausberger, P., 2009. Adaptive learning in theforaging behaviour of the predatory mite Phytoseiulus persimilis. BehaviouralEcology 20, 946–950.

Ramos, M., Gonzales, A.I., Gonzales, M., 2010. Management strategy of Raoiellaindica Hirst in Cuba, based on biology, host plants, seasonal occurrence and useof acaricide. In: XIII International Congress of Acarology, Recife, Brasil, AbstractBook, pp. 218–219.

Roda, A., Dowling, A., Welbourn, C., Peña, J.E., Rodrigues, J.C.V., Hoy, M.A., Ochoa, R.,Duncan, R.A., De Chi, W., 2008. Red palm mite situation in the Caribbean andFlorida. Proceedings of Caribbean Food Crops Society 44 (1), 80–87.

Roth, J.P., Macqueen, A., Bay, D.E., 1988. Predation by the introduced phoretic mite,Macrocheles peregrinus (Acari: Macrochelidae) on the buffalo fly, Haematobiairritans exigua (Diptera: Muscidae) in Australia. Environmental Entomology 17(3), 603–607.

Sabelis, M.W., Lesna, I., 2010. Does artificial selection for fixed prey preference affectlearning in a predatory mite? Experiments to unravel mechanisms underlyingpolyphagy in Hypoaspis aculeifer. Acarologia 50 (2), 257–268.

SAS Institute Inc. 2010. SAS 9.2 for Windows. Cary, NC, USA.Schausberger, P., Walzer, A., Hoffman, D., Rahmani, H., 2010. Food imprinting

revisited: early learning in foraging predatory mites. Behaviour 147, 883–897.Taylor, B., Rahman, P.M., Murphy, S.T., Sudheendrakumar, V.V., in press. Within-

season dynamics of red palm mite (Raoiella indica) and phytoseiid predators ontwo host palm species in south-west India. Experimental and AppliedAcarology. doi:10.1007/s10493-011-9482-0.

Tixier, M.-S., Ferrero, M., Okassa, M., Guichou, S., Kreiter, S., 2010. On the specificidentity of specimens of Phytoseiulus longipes Evans(Mesostigmata:Phytoseiidae) showing different feeding behaviours: morphological andmolecular analyses. Bulletin of Entomological Research. doi:10.1017/S0007485309990617.

Vásquez, C., Quirós, G.M., Aponte, O., Sandoval, D.M.F., 2008. First report of Raoiellaindica Hirst (Acari: Tenuipalpidae) in South America. Neotropical Entomology37, 739–740.

Vet, L.E.M., Dicke, M., 1992. Ecology of infochemical use by natural enemies in atritrophic context. Annual Review of Entomology 37, 141–172.

Zannou, I.D., Negloh, K., Hanna, R., Houadakpode, Sabelis, M.W., 2010. Mite diversityin coconut habitat in west and east Africa. In: XIII Abstract book of the XIIIInternational Congress of Acarology, Recife, Brazil, 23–27 August, p. 295.

Zanotto, S.P., Rodrigues, J.C.V., 2010. Chemical composition of flat mite glandularexudates. In: Abstract book of the 58th Annual Meeting of the EntomologicalSociety of America, San Diego, California, 12–15 December, p. 56.