Biological Control 47 (2008) 167–171
Contents lists available at ScienceDirect
Biological Control
journal homepage: www.elsevier.com/locate/ybcon
Effect of host-plant genotypes on the performance of three candidate biological control agents of Schinus terebinthifolius in Florida
Veronica Manrique a,*, J.P. Cuda b, W.A. Overholt a, D.A. Williams c, G.S. Wheeler d a Biological Control Research and Containment Laboratory, University of Florida, 2199 South Rock Road, FL 34945, USA b Department of Entomology & Nematology, University of Florida, Building 970 Natural Area Drive, Gainesville, FL 32611, USA c Department of Biology, Texas Christian University, 2800 S. University Drive, Fort Worth, TX 76129, USA d USDA-ARS Invasive Plant Research Laboratory, 3225 College Avenue, Fort Lauderdale, FL 33314, USA article info abstract
Article history: Brazilian peppertree, Schinus terebinthifolius Raddi, native to South America, is an invasive weed in Received 23 January 2008 Florida, California, Texas, and Hawaii. Genetic studies have recognized two S. terebinthifolius haplotypes Accepted 15 July 2008 (A and B) in Florida, and extensive hybridization has occurred between these two populations. Three Available online 24 July 2008 candidate biological control agents were identified from the native range (Brazil); a leaflet rolling moth Episimus utilis Zimmerman, a thrips Pseudophilothrips ichini Hood from Ouro Preto, and an unindentified Keywords: thrips, referred to as Pseudophilothrips sp. near ichini, from Curitiba, Brazil. The objective of this study was Host-plant genotypes to compare the performance of these three candidate agents on different S. terebinthifolius genotypes Biological control found in Florida and Brazil. Survival (54%), adult longevity (9 days), fecundity (84 eggs laid), and fertility Local adaptation Schinus terebinthifolius (68% eggs hatched) of E. utilis were similar on all S. terebinthifolius genotypes tested from Florida. In Episimus utilis contrast, the two thrips species differed in their ability to utilize different genotypes of their host plant. Pseudophilothrips ichini Pseudophilothrips sp. near ichini exhibited low survival (0–4%) and short adult longevity (<10 days) when reared on Florida genotypes, whereas higher survival (50%) and longevity (30 days) were observed for P. ichini on these genotypes. These findings highlight the importance of examining insect performance on plant genotypes present in the introduced and native ranges when selecting biological control agents. The ecological significance of the results is discussed in the context of plant genotypes and possible local adaptation of their natural enemies. Ó 2008 Elsevier Inc. All rights reserved.
1. Introduction A classical biological control program was initiated in the 1980s against S. terebinthifolius in Florida. Several phytophagous insects Schinus terebinthifolius Raddi (Anacardiaceae), a woody peren- were identified as potential biological control agents from explor- nial plant native to South America (Barkley,1944, 1957), was intro- atory surveys conducted in southeastern Brazil, including a leaflet duced into Florida, USA, as an ornamental between 1898 and 1900 rolling moth Episimus utilis Zimmerman (Lepidoptera: Tortricidae), (Morton, 1978). Schinus terebinthifolius is recognized as one of the and a shoot and flower attacking thrips Pseudophilothrips ichini most widespread exotic plants in Florida (Cuda et al., 2006), and Hood (Thysanoptera: Phlaeothripidae) (Bennett et al., 1990; aerial surveys indicate that 2833 km2 in central and southern Flor- Bennett and Habeck, 1991; Habeck et al., 1994; Cuda et al., 1999, ida have been invaded by this noxious weed (Cuda et al., 2006). 2006). The larval stages of E. utilis feed on S. terebinthifolius leaflets Chloroplast DNA (cpDNA) analysis indicated that two different and can completely defoliate small plants (Martin et al., 2004). populations of S. terebinthifolius were introduced separately on Episimus utilis is an appropriate agent for S. terebinthifolius growing the east and west coasts of Florida (Williams et al., 2005, 2007). on sites exposed to seasonal flooding because it completes its en- Haplotype A is more common on the west coast whereas haplotype tire life cycle in the canopy of its host plant. In the case of P. ichini, B is more common on the east coast. Nuclear microsatellite DNA both larval and adult stages damage the host plant by feeding on analysis revealed that extensive hybridization has occurred the growing shoot tips and flowers causing flower abortion (Garcia, between these two populations since arriving in Florida (Williams 1977). Immature thrips undergo two larval stages on the plant, and et al., 2005, 2007). In addition, genetic studies in the native range three non-feeding pupal stages in the soil (Garcia, 1977; Cuda et have identified nine cpDNA haplotypes (A, C–J), with haplotype D al., 2008). Until recently, P. ichini was assumed to be a widespread being the most common and widespread (Williams et al., 2005). species in Brazil, but morphological (Mound, unpublished data) and mitochondrial genetic (Williams, unpublished data) studies now indicate that there are at least two close related species, * Corresponding author. Fax: +1 772 460 3673. E-mail address: vero72@ufl.edu (V. Manrique). P. ichini from the Ouro Preto region and an undescribed species,
1049-9644/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.biocontrol.2008.07.005 168 V. Manrique et al. / Biological Control 47 (2008) 167–171 referred to as Pseudophilothrips sp. near ichini, from Curitiba. These six nuclear microsatellite loci as described in Williams et al. two taxa are now treated as distinct entities in investigations of (2002, 2005, 2007). Individuals were classified as ‘‘pure bred” east- their value for biological control of S. terebinthifolius in Florida. ern or western if they had the ancestry coefficient value (q) >0.90 Studies of the biology and host specificity of E. utilis, and in a respective cluster and as hybrids if they contained a maximum Pseudophilothrips sp. near ichini (thought at the time to be P. ichini) of q = 0.40 for either western or eastern ancestry. Hybrid individu- were conducted in Brazil and in quarantine in Florida (Garcia, als were further categorized as having either the A cpDNA haplo- 1977; Harmuch et al., 2001; Martin et al., 2004; Cuda et al., type or B haplotype. 2006, 2008). The outcome of these studies was a recommendation The cut ends of different S. terebinthifolius plants (10 plants by the federal interagency Technical Advisory Group for Biological per genotype) from Florida were coated in rooting hormone pow- Control Agents of Weeds (TAG) that the thrips is suitable for field der (Schultz TakeRootÒ Rooting Hormone) and planted in pots release and a release petition for E. utilis is in preparation. (18 cm height, 17 cm diameter) using sterilized soil mix (FafardÒ Many ecologists and weed biological control practitioners germination mix). Pots were placed in the shade and misted every believe that highly specialized natural enemies which share an 10 min. After three months, cuttings with attached roots were evolutionary history with their hosts are likely to be the most transplanted to new pots (18 cm height, 17 cm diameter) contain- effective agents for controlling invasive plants (Zwolfer and Preiss, ing soil mixture (FafardÒ #3B mix), and all plants were placed in 1983; Strong and Pemberton, 2000; Myers and Bazely, 2003; the greenhouse under ambient conditions and watered as needed. Williams et al., 2005). However, hybridization between plants in Schinus terebinthifolius seeds from Brazil (5 plants per genotype) the introduced range from different source populations may con- were collected in Blumenau, Santa Catarina State (haplotype A) found ‘old associations’ between insect herbivores and their hosts. and in Curitiba (haplotype D), and grown inside a quarantine Moreover, the establishment success and eventual effectiveness of greenhouse at the BCRCL. Genetic analyses were conducted on all biocontrol agents may be determined by the host-plant genotypes the plants to confirm their genotypes (Williams et al., 2002, present in the introduced range. For example, the establishment of 2005). Schinus molle seedlings were purchased from a Nursery in two biocontrol agents on the invasive leafy spurge (Euphorbia esula California, and grown inside a greenhouse at the BCRCL. All plants L.) was influenced by different genotypes of the plant (Lym et al., were fertilized once with 15 g of OsmocoteÒ (a slow release fertil- 1996; Lym and Carlson, 2002). The objective of this study was to izer 15–9–12, N–K–P), and 400 ml per pot of liquid fertilizer (Mir- determine the suitability of different genotypes of acle GrowÒ 24–8–16) monthly. S. terebinthifolius found in Florida and Brazil as hosts for E. utilis, P. ichini, and P. sp. near ichini. 2.1. Development of E. utilis on different S. terebinthifolius genotypes
2. Materials and methods To determine the suitability of different S. terebinthifolius geno- types as hosts for E. utilis, groups were confined on the plants and This study was conducted at the Biological Control Research and the following life-history parameters were recorded: (1) pupal Containment Laboratory (BCRCL) located at the Indian River fresh weight (mg), (2) developmental time to adult (days), Research and Education Center of the University of Florida, Fort (3) percent survival to adult, (4) adult longevity (days), (5) fecundity Pierce, FL, USA. A colony of E. utilis was initiated in August 2006 (total eggs laid), and (6) fertility (percent eggs hatched). The at the BCRCL, and insects were reared on potted S. terebinthifolius experiments were conducted in an environmental growth chamber plants (Florida genotypes) (pots: 18 cm height, 17 cm diameter) in- (25 ± 2 °C, 60–70% RH, 14 L:10 D photoperiod). Four different S. side environmental growth chambers (25 ± 2 °C, 60–70% RH, 14 terebinthifolius genotypes were compared (6–8 replicates per treat- L:10 D photoperiod) (for specific details on rearing methods, see ment): (1) Florida haplotype B, (2) Florida haplotype A, (3) Florida Martin et al., 2004). Episimus utilis individuals were originally hybrid A, and (4) Florida hybrid B. Ten neonates of E. utilis were collected in 2003 from S. terebinthifolius haplotypes C and D found caged on each potted plant inside a clear acrylic cylinder (45 cm in the vicinity of Curitiba located in Parana State of southeastern height, 15 cm diameter) with six holes (6 cm diameter) and tops Brazil. Pseudophilothrips sp. near ichini was collected from covered by a fine mesh to allow air circulation. After 20 days, the S. terebinthifolius Brazil haplotypes C and D at Curitiba, while cylinders were removed and all plant foliage was removed. Last in- P. ichini was collected from S. terebinthifolius Brazil haplotype A stars or pupae were placed individually inside plastic vials (29.5 ml at Ouro Preto, located in Minas Gerais State 830 km northeast of Dixie PL1) containing moist filter paper and plant leaflets from the Curitiba. same genotype on which they were reared. Upon adult emergence, A colony of P. sp. near ichini was initiated in January 2007 at the individual pairs from each treatment were placed inside wax paper BCRCL. Preliminary trials revealed that survival of this thrips was oviposition cages made by stapling together six rectangles of wax very low on Florida types of S. terebinthifolius,soSchinus molle paper (19 Â 30 cm) to form a cube. Each cage containing a cotton Ò L., a more suitable host, was used for colony maintenance of this wick with Gatorade and one S. terebinthifolius leaflet (same plant Ò species. A colony of P. ichini was introduced into quarantine in genotype). The cages were placed inside Ziploc freezer bags and November 2007 at the BCRCL facility, and maintained on kept in the environmental growth chamber. After all adults had S. terebinthifolius potted plants (Brazilian haplotype A). Voucher died, the numbers of hatched and unhatched eggs were counted specimens of E. utilis, P. ichini, and P. sp. near ichini were deposited under a dissecting microscope (magnification 25Â). in the Florida State Collection of Arthropods, Florida Department of In a separate experiment, E. utilis larval development and sur- Agriculture and Consumer Services, Gainesville, Florida, USA. vival to the adult stage were compared on two S. terebinthifolius Schinus terebinthifolius plants representing four Florida geno- genotypes: Florida hybrid B, and Brazil haplotype D. The E. utilis types were grown from cuttings that were collected in the field colony was originally collected from haplotypes Brazil haplotype from west (haplotype A), east (haplotype B), and hybrid D around Curitiba, Brazil. Since E. utilis survived equally well on (haplotypes A or B) plants previously confirmed by genetic analysis all the Florida genotypes tested only one type was included in (Williams et al., 2002, 2005). A Bayesian clustering method imple- the experiment. Individual neonates were placed inside vials mented in the program STRUCTURE version 2.1 (Pritchard et al., containing moist filter paper and plant leaflets (10 vials for each 2000; Pritchard and Wen, 2003) was used to estimate the degree plant genotype), and a total of 5 replicates per treatment of hybridization between the two introductions into Florida using were conducted. Several insect parameters were recorded: V. Manrique et al. / Biological Control 47 (2008) 167–171 169
Table 1 Life-history parameters (means ± SE) of E. utilis on different genotypes of S. terebinthifolius from Florida
Host genotype Survival to adult (%) Developmental time to adult (days) Adult longevity (days) Fecundity (eggs per female) Fertility (% egg hatched) Haplotype A 55.0 ± 8.8 31.9 ± 0.6 8.4 ± 0.8 53.2 ± 3.6 74.1 ± 9.1 Haplotype B 55.0 ± 9.4 32.0 ± 0.5 8.2 ± 0.7 91.2 ± 30.8 81.1 ± 5.3 Hybrid A 56.2 ± 8.2 31.5 ± 0.3 9.6 ± 1.1 95.0 ± 21.4 81.2 ± 4.9 Hybrid B 48.3 ± 11.9 31.9 ± 0.6 9.5 ± 0.2 98.9 ± 17.0 78.8 ± 5.8
FF3,29 = 0.11 F3,29 = 0.12 F3,23 = 0.89 F3,17 = 0.87 F3,17 = 0.11 P 0.93 0.94 0.46 0.47 0.95
(1) developmental time to adult (days), (2) percent survival to formed prior to analysis (Zar, 1999). Means were separated adult, and (3) adult longevity (days). using the Student–Neuman–Keuls (SNK) test (SAS Institute, 1999). A significance level of a = 0.05 was used for all statistical 2.2. Development of two thrips species on different S. terebinthifolius analyses. genotypes and on S. molle 3. Results The experiments described below were conducted separately for the two thrips species inside an environmental growth cham- 3.1. Development of E. utilis on different S. terebinthifolius genotypes ber (28 ± 2 °C, 60–70% RH, 14 L:10 D photoperiod). Seven treat- ments were established (8 replicates per treatment): (1) Florida No differences were detected for survival, developmental time, haplotype B, (2) Florida haplotype A, (3) Florida hybrid B, (4) Flor- adult longevity, fecundity or fertility of E. utilis reared on different ida hybrid A, (5) Brazil haplotype A, (6) Brazil haplotype D, and (7) Florida S. terebinthifolius genotypes (Table 1). In addition, pupal S. molle. Ten neonate thrips were placed inside a plastic vial (11 cm weight did not differ between plant genotypes, but female pupae height, 5 cm diameter) contained a plant shoot and moist filter pa- (18.2 ± 0.8 mg) were larger than male pupae (16.9 ± 0.5 mg) (geno- per. Vials were checked every other day, and moisture and food types: F3,59 = 2.11, P = 0.11; sex: F1,59 = 5.48, P = 0.02; geno- were added as needed. The life-history parameters recorded were: type  sex: F3,59 = 0.3, P = 0.8). percent survival to adult and developmental time to adult (days). No differences were detected for survival to adult, developmen- Preliminary laboratory studies confirmed that the two thrips spe- tal time to adult, or adult longevity when E. utilis was reared in cies survived as well on excised shoots as on rooted plants. vials on either Florida hybrid B or Brazilian haplotype D (Table 2). In a separate experiment, adult longevity of each thrips species was measured on the same host plants mentioned above and in 3.2. Development of two thrips species on different S. terebinthifolius vials with no food (controls). Ten newly emerged adults (5 fe- genotypes and on S. molle males:5 males) of P. sp. near ichini, reared from S. molle or P. ichni, reared from S. terebinthifolius Brazil haplotype A, were placed in Survival of P. sp. near ichini to the adult stage was highest on each vial containing a plant shoot and moist filter paper. There S. molle, followed by Brazil haplotype D and Brazil haplotype A were 7–8 replicates per treatment. Vials were checked every other (Table 3). Survival was low on all S. terebinthifolius Florida genotypes day, and survival and pre-oviposition period was recorded. Exper- tested (<4%) (Table 3). No differences were detected among iments were terminated when all adults had died. host plants for developmental time to adult (Table 3). In contrast,
2.3. Data analysis Table 2 Life-history parameters (averaged values per replicate) for Life-history parameters (means ± SE) of E. utilis reared on different genotypes of E. utilis and for the two thrips were compared between host S. terebinthifolius from Florida and Brazil genotypes plants using one-way analysis of variance (ANOVA) (SAS Insti- Host genotype Survival to Developmental time Adult longevity tute, 1999). Two-way ANOVA with interaction was used to adult (%) to adult (days) (days) compare pupal weight of E. utilis between genders and plant Florida hybrid B 38.0 ± 8 34.7 ± 0.8 5.0 ± 0.4 treatments (SAS Institute, 1999). Data expressed as percentages Brazil haplotype D 0 28 ± 5.8 34.3 ± 0.9 5.3 ± 0.5 (e.g., survival and eggs hatched) were arcsine square root trans- F1,8 = 1.0, ns F1,31 = 0.07, ns F1,21 = 0.94, ns
Table 3 Life-history parameters (means ± SE) for two thrips species reared on different genotypes of S. terebinthifoliusi, from Florida and Brazil, and on S. molle
Host plant Survival to adult (%) Development to adult (days) Adult longevity (days) P. sp. near ichini P. ichini P. sp. near ichini P. ichini P. sp. near ichini P. ichini Florida haplotype A 3.8 ± 2.6 d 51.2 ± 8.9 a 18.7 ± 1.33 a 15.8 ± 0.5 b 9.09 ± 0.3 c 34.8 ± 1.6 a Florida haplotype B 0 d 57.6 ± 11.2 a – 16.2 ± 0.5 b – 27.8 ± 1.3 b Florida hybrid A 0 d 68.7 ± 10.4 a – 16.4 ± 0.6 b – 25.8 ± 1.9 b Florida hybrid B 0 d 52.5 ± 4.5 a – 17.4 ± 1 b – – Brazil haplotype A 24.4 ± 11.5 c 41.2 ± 5.8 a 18.8 ± 0.9 a 16.4 ± 0.5 b 18.4 ± 2.2 b 34.8 ± 1.4 a Brazil haplotype D 43.7 ± 6.6 b 6.6 ± 2.1 b 17.7 ± 0.3 a 20 ± 1 a 21.4 ± 1.1 b 5.7 ± 0.4 c S. molle 75.5 ± 7 a 42.5 ± 3.6 a 17.3 ± 0.5 a 17.1 ± 0.5 b 33.1 ± 1.4 a 25.8 ± 2.2 b Control – – – – 8.4 ± 0.4 c 5.4 ± 0.5 c
F6,51 = 29.95 F6,53 = 6.6 F3,22 = 1.39 F6,51 = 3.03 F4,39 = 60.75 F6,49 = 62.4 P < 0.0001 P < 0.0001 ns P < 0.01 P < 0.0001 P < 0.0001
Control, no food. Different letters in the same column indicate statistical differences between plant treatments (P < 0.05). –, no measurements, not included in analysis. 170 V. Manrique et al. / Biological Control 47 (2008) 167–171
P. ichini survived equally well on the host plants that were tested of S. terebinthifolius not present in Florida was not entirely unex- with the exception of Brazil haplotype D (Table 3). Developmental pected considering the close taxonomic association of this species time to adult did not differ significantly for P. sp. near ichini on the and the weed. Schinus molle is a popular ornamental in California, host plants that were tested, but was significantly longer on Brazil but the California Exotic Pest Plant Council recently listed this spe- haplotype D than on all the other plants for P. ichini (Table 3). cies as a Category B invasive species (Cal-IPC, 2006). Adult longevity was longest for P. sp. near ichini on S. molle, fol- Local host or ‘fine-tuned’ adaptation refers to a process whereby lowed by Brazil haplotype D and Brazil haplotype A (Table 3). Lon- locally occurring herbivores with shorter generation times than gevity of <10 days was recorded for adults exposed to either their hosts and poor dispersal capabilities adapt rapidly to become Florida haplotype A or those given no food (the control) (Table specialized on host genotypes to which they are exposed 3). Adults of P. sp. near ichini laid eggs on S. molle and all the Brazil (Edmunds and Alstad, 1978; Harley and Foreno, 1992; Ebert, genotypes, but no eggs were laid on Florida haplotype A. The pre- 1994; Gandon and Van Zandt, 1998; Goolsby et al., 2006). Several oviposition period was similar (approximately 12 days) for the studies have shown that the dispersal capacity of adult thrips is three host plants that received eggs (F2,23 = 2.45, P = 0.11). In con- typically limited, and movement occurs mostly between neighbor- trast, P. ichini survived longest on both Florida and Brazil haplo- ing host plants, which result in low gene flow and allows local types A, followed by Florida hybrid A and S. molle (Table 3). adaptation (Karban and Strauss, 1994; Rhainds and Shipp, 2003; Longevity of <7 days was recorded for adults exposed to either Bra- Rhainds et al., 2005). In an extreme case, thrips populations of zil haplotype D or the control treatment (Table 3). In addition, Apterothrips secticornis Trybom were specifically adapted to indi- P. ichini laid eggs on all host plants tested except for Brazil haplo- vidual clones of their host Erigeron glaucus Ker. (Asteraceae) type D. The pre-oviposition period (days) of this thrips varied (Karban, 1989). In contrast, lepidopteran species such as E. utilis among host plants (F4,33 = 3.2, P = 0.027), being shorter on Brazil are capable of flying over longer distances (Suckling et al., 1994; haplotype A (6.5 ± 0.5 days) and Florida haplotype A (7.7 ± 1.2 Showers et al., 2001), and this may explain the differences in diet days) compared to S. molle (8.5 ± 1.3 days), Florida haplotype B breadth between E. utilis and the two thrips species. Another (10.5 ± 1.7 days), and Florida Hybrid A (13 ± 1.6 days). important trait usually associated with locally adapted herbivores is parthenogenic reproduction or haplodiploidy (Rice, 1983; Boecklen and Mopper, 1998), which occurs in P. sp. near ichini 4. Discussion (published under P. ichini)(Garcia, 1977). Preliminary genetic stud- ies showed that P. sp. near ichini is only associated with Brazil The majority of S. terebinthifolius trees in Florida are intraspe- haplotypes C and D, while P. ichini was found on haplotypes A, K, cific hybrids between the two genotypes that were originally intro- N, and M in the native range (Williams, unpublished data). Thus, duced (Williams et al., 2005, 2007). In the native range, cpDNA it appears that Pseudophilothrips sp. near ichini is more specialized haplotypes are geographically structured with only one or two clo- than P. ichini. Haplotype K of S. terebinthifolius is closely related to sely related haplotypes occurring at a given location (Williams haplotype B (Williams et al., 2005, unpublished data), which may et al., 2005, unpublished data). The source location of Florida hap- explain the high performance of P. ichini on Florida haplotype B. lotype A is the coast of southeastern Brazil, whereas the origin of Biological control programs have been criticized for their lack of Florida haplotype B has recently been found to be the coast of Bahi- predictability in terms of agent establishment and success (Ehler, a in northeastern Brazil (Williams et al., 2005, unpublished data). 1990; Harris, 1998). In order to improve the predictability of bio- This allopatry prevents hybridization between these two geno- control, genetic studies may help to identify coadapted natural types in their native ranges (Williams et al., 2005, 2007). enemies in the native range that may be the most effective against Hybrid genotypes in the introduced range may have character- a particular genotype present in the introduced range (Williams et istics which make plants unsuitable hosts for the phytophagous in- al., 2005; Goolsby et al., 2006). This study showed that three can- sects that feed on the parent genotypes in the native ranges, didate biological control agents differed in their ability to utilize although this is difficult to predict a priori (Strauss, 1994; Fritz et different genotypes of their host plant. The leaflet roller E. utilis al., 1999). The potential mismatch requires that any insects se- performed well on all S. terebinthifolius genotypes tested (Florida lected as potential biological control agents should be tested and Brazil), whereas P. sp. near ichini and P. ichini differed in their against all plant genotypes present in the area of introduction. In- ability to utilize different host plant genotypes. Moreover, as each deed such possible mismatches may be common among other tar- thrips species responded to select plant haplotypes, factors such as geted weeds because recent molecular genetic work suggests that plant secondary chemistry (Koschier et al., 2000; Katerinopoulos et invasive plants are often introduced multiple times from different al., 2005), amino acid composition (Mollema and Cole, 1996; source regions (Bossdorf et al., 2005; Novak and Mack, 2005). Brodbeck et al., 2001), or the presence of leaf pubescence (Quisenberry According to this study, S. terebinthifolius genotypes did not affect and Rummel, 1979) may explain differences in their utilization. the performance of the leaflet roller E. utilis, but had a substantial These findings highlight the importance of testing biological con- effect on the performance of two thrips species. trol agents on different genotypes of the weed from both the native Host specificity tests on E. utilis in the laboratory showed that it and introduced range. feeds on several related plant species in the family Anacardiaceae (J.P. Cuda, unpublished data). This extended host range was not ob- Acknowledgments served in the native range of the moth or in Hawaii where it has been introduced (Cuda et al., 2006). In nature, E. utilis seems to The authors are grateful to the colleagues from the weed biocon- be only able to develop on S. terebinthifolius sensu lato, but its abil- trol laboratory (University of Florida, USA) for their constant support ity to develop on other species in the laboratory may preclude it and assistance during this study: J. Gillmore, J.C. Medal, R. Diaz, O. from consideration for release in Florida. Moeri, J. Markle, L. Markle, Y. Valenzuela, D. Gonzalez, and A. Sam- The two thrips species differed significantly in their ability to ayoa. Special thanks to the collaborators in Brazil for providing the utilize different genotypes of S. terebinthifolius. The thrips P. ichini insects: H.J. Pedrosa-Macedo (Federal University of Parana, Curitiba, which was originally collected from Brazil haplotype A is well Brazil), and M.D. Vitorino (Regional University of Blumenau, Brazil). adapted to Florida genotypes and should be considered as a poten- The authors also are thankful to Dr. L. Mound (CSIRO Entomology, tial biocontrol agent of S. terebinthifolius in Florida. The relatively Australia) for conducting the morphological studies on the two high performance of both thrips species on S. molle, a close relative thrips species. Earlier versions of this manuscript have been im- V. Manrique et al. / Biological Control 47 (2008) 167–171 171 proved by Dr. H. McAuslane and Dr. A. Fox (University of Florida, Karban, R., 1989. Fine-scale adaptation of herbivorous thrips to individual host USA). This project was supported by grants from the Florida Depart- plants. Nature 340, 60–61. Karban, R., Strauss, S.Y., 1994. Colonization of new host plant individuals by locally ment of Environmental Protection, South Florida Water Manage- adapted thrips. Ecography 17, 82–87. ment District, and Florida Exotic Pest Plant Council. Katerinopoulos, H.E., Pagona, G., Afratis, A., Stratigakis, N., Roditakis, N., 2005. Composition and insect attracting activity of the essential oil of Rosmarinus officinali. Journal of Chemical Ecology 31, 111–122. References Koschier, E.H., de Kogel, W.J., Visser, J.H., 2000. Assessing the attractiveness of volatile plant compounds to western flower thrips Frankliniella occidentalis. Barkley, F.A., 1944. Schinus L.. Brittonia 5, 160–198. Journal of Chemical Ecology 26, 2643–2655. Barkley, F.A., 1957. A study of Schinus L. Lilloa Revista de Botanica. Tomo 28. Lym, R.G., Carlson, R.B., 2002. Effect of leafy spurge (Euphorbia esula) genotype on Universidad Nacional del Tucuman, Argentina. feeding damage and reproduction of Aphthona spp.: implications for biological Bennett, F.D., Crestana, L., Habeck, D.H., Berti-Filho, E., 1990. Brazilian peppertree— weed control. Biological Control 23, 127–133. prospects for biological control. In: Delfosse, E.S. (Ed.), Proceedings Seventh Lym, R.G., Nissen, S.J., Rowe, M.L., Lee, D.J., Masters, R.A., 1996. Leafy spurge International Symposium on Biological Control of Weeds, 6–11 March 1988, (Euphorbia esula) genotype affects gall midge (Spurgia esulae) establishment. Rome, Italy. Ministero dell’Agriculture e delle Foreste, Rome/CSIRO, Melbourne, Weed Science 44, 629–633. Australia, pp. 293–297. Martin, C.G., Cuda, J.P., Awadzi, K.D., Medal, J.C., Habeck, D.H., Pedrosa-Macedo, J.H., Bennett, F.D., Habeck, D.H., 1991. Brazilian peppertree—prospects for biological 2004. Biology and laboratory rearing of Episimus utilis (Lepidoptera: control in Florida. In: Center, T., Doren, R.F., Hofstetter R.L., Myers, R.L., Tortricidae), a candidate for classical biological control of Brazilian Whiteaker, L.D. (Eds.), Proceedings of the Symposium of Exotic Pest Plants, 2–4 peppertree, Schinus terebinthifolius (Anacardiaceae), in Florida. Environmental November 1988. Miami, FL, pp. 23–33. Entomology 33, 1351–1361. Boecklen, W.J., Mopper, S., 1998. Local adaptation in specialist herbivores: theory Mollema, C., Cole, R.A., 1996. Low aromatic amino acid concentrations in leaf and evidence. In: Mopper, S., Strauss, S.Y. (Eds.), Genetic Structure and Local proteins determine resistance to Frankliniella occidentalis in four vegetable Adaptation in Natural Insect Populations: Effects of Ecology, Life History, and crops. Entomologia Experimentalis et Applicata 78, 325–333. Behavior. Chapman and Hall, New York, pp. 64–88. Morton, J.F., 1978. Brazilian peppertree—its impact on people, animals and the Bossdorf, O., Auge, H., Lafuma, L., Rogers, W.E., Siemann, E., Prati, D., 2005. environment. Economic Botany 32, 353–359. Phenotypic and genetic differentiation between native and introduced plant Myers, J.H., Bazely, D.R., 2003. Ecology and Control of Introduced Plants. Cambridge populations. Oecologia 144, 1–11. University Press, Cambridge, U.K. Brodbeck, B.V., Stavisky, J., Funderburk, J.E., Andersen, P.C., Olson, S.M., 2001. Flower Novak, S.J., Mack, R.N., 2005. Genetic bottlenecks in alien plant species. In: Sax, D.F., nitrogen status and populations of Frankliniella occidentalis feeding on Stachowicz, J.J., Gaines, S.D. (Eds.), Species Invasions: Insights into Ecology, Lycopersicon esculentum. Entomologia Experimentalis et Applicata 99, 165–172. Evolution, and Biogeography. Sinauer Associates, Inc. Publishers, Sunderland, California Invasive Plant Council (Cal-IPC), 2006. Invasive Plant Inventory. Available MA, pp. 201–228. from: