The leafmining Leurocephala schinusae (Lepidoptera: Gracillariidae): not suitable for the biological control of Schinus terebinthifolius (Sapindales: Anacardiaceae) in continental USA

The host range of Leurocephala schinusae Davis & Mc Kay (Lepidoptera: Gracillariidae) was studied to assess its suitability as a biological control agent of Schinus terebinthifolius Raddi (Anacardiaceae), a serious environmental weed in the USA and elsewhere in the world. The host range was determined in the laboratory with adult no-choice oviposition (Argentina and USA) and larval development tests (USA). Seventeen plant species in ten genera were selected based on taxonomic relatedness to S. terebinthifolius, economic importance, and availability. Additional information was obtained by sampling foliage of S. terebinthifolius and six other South American native Anacardiaceae species in north-eastern Argentina. In the laboratory, except for Lithrea molleoides and Spondias mombin, all of the tested species were accepted for oviposition with a marked preference for Rhus aromatica. Incipient mines successfully developed into complete mines, pupae and adults on R. aromatica, Rhus copallinum, Schinus molle, Schinus lentiscifolius and S. terebinthifolius. In the field, although L. schinusae showed a clear preference for S. terebinthifolius, the host range, as determined by samples of host use in the native range, included three other Schinus species (S. lentiscifolius, Schinus longifolius, Schinus weinmannifolius) and one Astronium species (Astronium balansae). In conclusion, L. schinusae will not be considered for the biological control of S. terebinthifolius in continental US. However, the utilisation of this species in other infested areas such as Hawaii and Australia should be further discussed.     

Risk assessment of Episimus unguiculus (Lepidoptera: Tortricidae), a biological control agent of Schinus terebinthifolia (Sapindales: Anacardiaceae) in Hawaii,USA

Schinus terebinthifolia Raddi (Anacardiaceae) is an introduced ornamental tree from South America that has become one of the most invasive weeds in Hawaii and Florida, USA. Exploratory surveys in the plant’s native range from 1950 to 2014 identified several potential biological control agents. One of these is the leaflet rolling moth Episimus unguiculus Clarke (Lepidoptera: Tortricidae), previously known as Episimus utilis Zimmerman. This biological control agent was released in Hawaii in the 1950s where high densities were occasionally observed, leading to partial control of S. terebinthifolia by the 1960s. Larvae are leaf tiers capable of completely defoliating small plants. In order to investigate the release of E. unguiculus in the continental USA, a series of laboratory no- choice, and multiple-choice tests were conducted in Florida, and a preliminary open field test with a native plant in Hawaii. Under the confined laboratory conditions imposed during the no-choice tests, E. unguiculus accepted the economically important Pistacia spp. and several other non-target plants for oviposition and development. However, in the multiple-choice tests E. unguiculus exhibited a clear preference for S. terebinthifolia relative to non-target plants accepted in the no-choice tests. Overall, the results of field observations during surveys in South America and Hawaii and host range studies completed in Hawaii and Florida showed that E. unguiculus is a narrow specialist on S. terebinthifolia, its natural host plant.     

Temperature-dependent development and potential distribution of Episimus utilis (Lepidoptera: Tortricidae), a candidate biological control agent of Brazilian peppertree (Sapindales: Anacardiaceae) in Florida

The invasive Brazilian peppertree (Schinus terebinthifolius Raddi), native to South America, is widely established throughout central and south Florida. The defoliating leaflet-roller Episimus utilis Zimmerman was selected as potential biocontrol agent of this invasive species. The objectives of this study were to determine development rate and survival of E. utilis at seven constant temperatures (10, 15, 20, 25, 30, 33, and 35 degrees C) and generate prediction maps of the number of generations per year this species may exhibit in the United States. The rate of development of E. utilis as a function of temperature was modeled using linear regression to estimate a lower developmental threshold of 9.6 degrees C and the degree-day requirement of 588. The Logan nonlinear regression model was used to estimate an upper developmental threshold of 33 degrees C. Cold tolerance of E. utilis was examined using all insect stages, and each stage was exposed to three constant temperatures (10, 5, 0 degrees C) for 0.5, 1, 2, 4, and 8 d (or until all insects died). The pupal stage was the most cold tolerant with 100% mortality after 12 d at 0 degrees C. The pupal lethal times at 5 (Ltime50 = 10 d, Ltime90 = 28 d) and 0 degrees C (Ltime50 = 5 d, Ltime90 = 9 d) were used to generate isothermal lines to predict favorable regions for E. utilis establishment. A GIS map was generated to predict the number of generations of E. utilis (range, 0.5-9.8) across all Brazilian peppertree range in the United States. The potential for establishment of E. utilis and its probable distribution in the continental United States was examined.     

Biology and host range of Omolabus piceus, a weevil rejected for biological control for Schinus terebinthifolius in the USA

Surveys for biological control agents of the invasive weed Schinus terebinthifolius (Anacardiaceae) discovered two Omolabus weevils (Coleoptera: Attelabidae) feeding on the plant in its native range. Molecular and morphological analysis indicated that one of these species consistently fed on the target weed and the other species fed more broadly. Aspects of the biology and host range of the more specific species, Omolabus piceus (Germar) were examined to determine its suitability as a biological control agent of S. terebinthifolius in the USA. Adults feed on newly formed leaves, and eggs, larvae and pupae develop in curled fragments of leaves, called nidi. Larvae consumed an average of 11.3 (±0.4) mg throughout their development which required 15.1 (±0.2) days. An average of 31.6 (±2.7) eggs were laid per female during their 23.8 (±2.2) day lifetime, after a 3.4 (±1.0) day preoviposition period. In no-choice tests, O. piceus adults fed and oviposited on all tested native North American, Caribbean and agricultural Anacardiaceae species except for M. indica. The field host-range of O. piceus, as determined by samples of host use in the native range, included three Schinus, two Lithrea and one Anacardium species. Therefore, we do not recommend O. piceus for biological control of S. terebinthifolius in the USA. However, the utilization of this species in other infested areas such as Hawaií and Australia should be considered.     

Biology, host specificity tests, and risk assessment of the sawfly Heteroperreyia hubrichi, a potential biological control agent of Schinus terebinthifolius in Hawaii

Heteroperreyia hubrichiMalaise (Hymenoptera: Pergidae), a foliagefeeding sawfly of Schinusterebinthifolius Raddi (Sapindales:Anacardiaceae), was studied to assess itssuitability as a classical biological controlagent of this invasive weed in Hawaii. No-choice host-specificity tests were conductedin Hawaiian quarantine on 20 plant species in10 families. Besides the target weed, adultfemales oviposited on four test species. Females accepted the Hawaiian native Rhussandwicensis A. Gray (Sapindales:Anacardiaceae) as an oviposition host equallyas well as the target species. The other threespecies received significantly fewer eggs. Neonate larvae transferred onto test plantssuccessfully developed to pupae on S.terebinthifolius (70% survival) and R.sandwicensis (1% survival). All other 18test plant species failed to support larvaldevelopment. A risk analysis was conducted toquantify the acceptability of non-targetspecies as host plants for H. hubrichi onthe basis of the insect's performance atvarious stages in its life cycle. Risk ofdamage to all plant species tested wasinsignificant except for R. sandwicensis. Risk to this native plant relative to S.terebinthifolius was estimated at 1%. Currently this level of risk is too high torequest introduction of this insect into theHawaiian environment. Detailed impact studiesin the native range of S. terebinthifoliusare needed to identify thepotential benefit that this insect offers. Also, field studies in South America withpotted R. sandwicensis would give a morereliable analysis of the risk this nativeHawaiian plant would face from naturalpopulations of H. hubrichi.     

Biology and host range of Platyphora semiviridis, a leaf beetle evaluated as a potential biological control agent for Solanum mauritianum in South Africa

The South American tree Solanum mauritianum is a major environmental weed in the high-rainfall regions of South Africa and has been targeted for biological control. Potential agents included five species of the genus Platyphora, which were imported from South America in 1994. Platyphora species associated with Solanaceae reputedly have very specific habitat requirements and host plant preferences in the field. Despite this, host-specificity tests on one species, Platyphora semiviridis, revealed a broad physiological host range. Although laboratory tests showed that P. semiviridis is confined to Solanum species and cannot survive on solanaceous crops outside that genus, it developed on potato and cultivated eggplant (aubergine) as well as on 10 native South African Solanum species. With few exceptions, there were no consistent differences in survival and duration of development on these compared with S. mauritianum. Furthermore, at least six of these non-target species, including potato and eggplant, supported breeding colonies of the beetles in cages. During choice tests in both small and larger cages, P. semiviridis avoided potato but did not consistently discriminate between S. mauritianum, eggplant and six native solanums for larviposition. Despite these findings, P. semiviridis has never been recorded on either potato or eggplant in South America, where it was only observed to feed on S. mauritianum. Although there are several reasons why P. semiviridis is unlikely to attack non-target Solanum species in the field, it will not be released in South Africa because there are other imported agents which have displayed narrower physiological host ranges and which may be more effective.     

Comparison of two populations of Pseudophilothrips ichini (Thysanoptera: Phlaeothripidae) as candidates for biological control of the invasive weed Schinus terebinthifolia (Sapindales: Anacardiaceae)

Brazilian peppertree, Schinus terebinthifolia Raddi (Sapindales: Anacardiaceae) (hereafter Schinus), is one of the worst invasive species in Florida and Hawaii. The thrips Pseudophilothrips ichini Hood (Thysanoptera: Phlaeothripidae) is being considered as a potential biological control agent of Schinus. Two populations of this thrips were collected in the weed's native range; one from central-east Brazil (Ouro Preto thrips) and a second from north-east Brazil (Salvador thrips). Temperature requirements, adult fecundity and impact on different plant haplotypes by P. ichini were examined in the laboratory. Complete development of thrips from both populations occurred at temperatures ranging from 20 to 30°C. Two approaches were used to model the predicted distributions of the thrips populations in the USA: the physiological model (NAPPFAST) based on cold tolerance and the ecological niche model based on climatic variables (MaxEnt). The physiological model predicted that both populations of P. ichini may establish in similar areas of the USA, overlapping with the distribution of Schinus. However, the niche model predicted that only the Ouro Preto thrips could establish in the USA. The difference in model predictions suggests an apparent preadaptation of the Salvador thrips to lower temperatures than those experienced at the locations they were collected in Brazil. The Ouro Preto thrips had similar fecundity on two Florida Schinus haplotypes, whereas lower fecundity on haplotype A was found for the Salvador thrips. Based on these results, the Ouro Preto population may be better adapted to the climatic conditions and plant haplotypes found in Florida. Moreover, greenhouse studies indicated that Schinus growth was greatly reduced by thrips feeding, which may result in lower weed reproduction and densities in the field.     

Laboratory host range of Austromusotima camptozonale (Lepidoptera: Crambidae), a potential biological control agent of Old World climbing fern,Lygodium microphyllum (Lygodiaceae)

Old World climbing fern, Lygodium microphyllum, is one of the most serious invasive weeds impacting south Florida and development of biological control is crucial for sustainable management. Larvae of a small moth, Austromusotima camptozonale, were discovered defoliating L. microphyllum in Australia. Preliminary testing suggested this moth was a Lygodium specialist. Laboratory host range testing was conducted on 65 species of test plants, from 31 families, comprising seven Lygodium species, four close relatives, 45 other species of ferns and fern allies, eight agricultural crops and one gymnosperm species plus the primary host L. microphyllum. Significant oviposition occurred only on other species of Lygodium. No eggs were laid on the agricultural crops, or about half the species of non-Lygodium ferns and fern allies tested. Oviposition on the other non-Lygodium ferns was very low, except on Anemia adiantifolia and Blechnum serrulatum, which received modest egg loads, but did not support development to adult. Larval feeding was low to non-existent on all the non-Lygodium species. Larvae developed to adult only on the native, American climbing fern L. palmatum, and to a lesser extent on L. japonicum. Lygodium japonicum is a naturalized invasive weed in the United States. Colonies of A. camptozonale were unable to persist on L. palmatum and died out in two to seven generations. Freezing winter temperatures in states where L. palmatum occurs would be lethal to A. camptozonale. It was concluded that A. camptozonale would pose no threat to native or cultivated plants in North America or the Caribbean and should be considered as a weed biological control agent against L. microphyllum.     

Biology and host range of Heterapoderopsis bicallosicollis: a potential biological control agent for Chinese tallow Triadica sebifera

Chinese tallow, Triadica sebifera, is an invasive weed that infests natural and agricultural areas of the south-eastern USA. A candidate insect for biological control of Chinese tallow has been studied under quarantine conditions. The biology and host range of a primitive leaf-feeding beetle, Heterapoderopsis bicallosicollis, were examined between July 2008 and February 2010. H. bicallosicollis adults fed on 16 of 29 non-target species and survived >10 days on all 29 species, more than the number of days that adults survived when deprived of food. Additionally, adult feeding, oviposition and partial larval development occurred on a Florida endangered species, Heterosavia bahamensis. These data indicated that H. bicallosicollis adult feeding and oviposition may not be limited to the target weed and included several valued North American natives. Therefore, as its release could pose unacceptable risks, testing of this species was discontinued and the quarantine colony destroyed.     

Biology and field host range of Ceutorhynchus cardariae,a potential biological control agent for Lepidium draba

Lepidium draba (Brassicaceae) is a clonal herb, originating from Eurasia, which is invasive in North America. A classical biological control project was initiated in 2001, and the gall‐forming weevil Ceutorhynchus cardariae was prioritized as a candidate agent. We studied its biology and field host range between 2003 and 2014 in the laboratory and a common garden in Switzerland and in the field in Romania. Ceutorhynchus cardariae is a univoltine to bivoltine species. In Switzerland, oviposition usually started at the beginning of March and can occur at temperatures as low as 2.5°C. Galls are formed on stems, leaf stalks and midribs of L. draba rosettes and bolting plants. Gall size increased with an increasing number of larvae per gall. The three larval instars feed inside the galls and leave the plant to pupate in the soil once mature. In Switzerland, development from egg to adult took about 12 weeks in spring. Adults emerged from May to July. After a brief feeding period, adults aestivate. From late summer, feeding recommenced and females may oviposit, forming a partial second generation. Eggs and all larval instars can be found in galls throughout winter. The rate of larval ectoparasitism reached 78%, while endoparasitism was low with a maximum of 2.3%. Lepidium draba populations differed in their suitability for development (number of C. cardariae produced), indicating that effectiveness of C. cardariae – in case released – may be variable. In the field, we observed that gall formation by C. cardariae can severely stunt or even kill shoots. Investigations on the field host range of C. cardariae indicated that only the closely related Lepidium campestre may act as an alternative host for the weevil in Europe. Host‐specificity tests are underway to determine its environmental safety before field release in North America is being considered.     

Life history and host range of the leaf blotcher Eucosmophora schinusivora: a candidate for biological control of Schinus terebinthifolius in the USA

The host range of Eucosmophora schinusivora Davis and Wheeler (Lepidoptera: Gracillariidae) was studied to assess its suitability as a biological control agent of Schinus terebinthifolius Raddi (Anacardiaceae), a serious environmental and agricultural weed in the USA and elsewhere in the world. The life history of this insect species and its host range were determined in the laboratory with adult no-choice oviposition and larval development tests. This species has five instars, the first three are sap-feeding miners and the last two are tissue feeding. Total development time was 31.7 days. To examine specificity of this species, 10 plant species in Anacardiaceae were selected based on taxonomic relatedness to S. terebinthifolius, economic importance, and availability. In the laboratory, except for Anacardium occidentale and Cotinus obovatus, all of the tested species were accepted for oviposition with a marked preference for the weed S. terebinthifolius, Schinus molle, Rhus copallinum, Rhus sandwicensis and Pistacia chinensis. Complete development, from egg to adult, was achieved only on S. terebinthifolius, S. molle, R. copallinum, P. chinensis and Metopium toxiferum. In conclusion, E. schinusivora will not be considered for the biological control of S. terebinthifolius in the continental USA. However, the utilisation of this species in other infested areas such as Hawai'i and Australia should be considered.     

Pyemotes tritici (Acari: Pyemotidae), a potential biological control agent of Anagasta kuehniella (Lepidoptera: Pyralidae)

The feasibility of utilizing Pyemotes tritici as a biological control agent against the Mediterranean flour moth, Anagasta kuehniella was investigated. The results from experiments designed to assess the effect of P. tritici on cohorts of A. kuehniella of different ages were not significant. They suggested that moth larvae were most severely affected if physogastric P. tritici females were introduced 10 days after moth eggs were added to the experimental arena. In part II, the impact of various densities of P. tritici on flour moths indicated that complete destruction of cohorts of 400 A. kuehniella larvae was achieved by adding about 80 female P. tritici 10 days after the eggs were laid or 6 days after the larvae hatched. Half the above density of pyemotid mites gave inconsistent, but at times complete control.     

Laboratory studies on the biology and host range of Dichrorampha odorata (Lepidoptera: Tortricidae), a biological control agent for Chromolaena odorata (Asteraceae)

Dichrorampha odorata (Lepidoptera: Tortricidae) is a moth from Jamaica whose larvae bore into, and kill, the shoot tips of the invasive alien plant, Chromolaena odorata (L.) King and Robinson (Asteraceae). This study reports aspects of the biology of D. odorata, and also determined the host specificity (larval and adult no-choice trials) of the moth. Adults were short lived (ranging from 2 to 7 days), with females laying a mean of 15.4 eggs. Eggs took 9 days to hatch, larvae 20–23 days to develop and the pupal stage lasted 11–12 days, giving an overall lifecycle period of 41–45 days. Larval no-choice tests using 34 asteraceous test species indicated that only C. odorata could sustain complete development of D. odorata to adulthood, although there was slight initial boring 14 test species (plus chromolaena). Results from the adult nochoice trials, in which seven test-plant species were exposed to D. odorata, were consistent with those from larval trials, with larval damage, pupae and adults of D. odorata recorded from only C. odorata. This confirmed that only C. odorata is a suitable host for D. odorata in South Africa. Permission has subsequently been granted for the release of D. odorata in South Africa, thus making it the first shoot-tip attacking agent to be released against C. odorata. It is hoped that in the field, high levels of damage by the moth will reduce the height and therefore competitiveness of C. odorata, thereby contributing to the success of biological control of this plant.     

Biology and host range of the leafminer,Pseudonapomyza sp. (Diptera: Agromyzidae), a potential biological control agent for Tecoma stans (Bignoniaceae) in South Africa

Tecoma stans (Bignoniaceae), is an evergreen shrub that has a wide natural distribution in the tropical and subtropical parts of the western hemisphere. This shrub is native to Mexico and the southern regions of the USA. This weed is widely distributed in South Africa and neighbouring countries. As part of the biological control initiative, a leafmining fly, Pseudonapomyza sp. (Diptera: Agromyzidae), was imported into South Africa, and was subsequently studied as a potential biological control agent for T. stans. During no-choice tests involving 46 plant species in 16 families, Pseudonapomyza sp. only oviposited and developed on T. stans. Neither oviposition nor larval development was recorded on the closely related and indigenous plant species. When six plant species in the Bignoniaceae family were exposed to Pseudonapomyza sp. during multi-choice tests, oviposition and larval development only occurred on T. stans. It was concluded that Pseudonapomyza sp. was sufficiently host-specific to be released against T. stans in South Africa. Pseudonapomyza sp. also displayed very promising biological attributes that could enhance its effectiveness to control T. stans.     

Biology ofPterolonche inspersa (Lep.: Pterolonchidae), a biological control agent forCentaurea diffusa andC. maculosa in the United States

The moth,Pterolonche inspersa (Staudinger) (Lepidoptera: Pterolonchidae), is widely distributed in southern Europe, north Africa, Turkey and the former Soviet Union. It occurs in both thick and scattered stands of knapweeds in disturbed sites, usually on sandy and/or stony soil. Larvae bore in the roots of diffuse and spotted knapweeds (Centaurea diffusa De Lamarck andC. maculosa De Lamarck). There is one generation per year in northern Greece, and larvae feed in the roots for about 11 months during the growing season (August–September, to the following July–August). In the laboratory garden, emergence took place between the second half of July and the end of August, with peak emergence during mid August. In the field, adults were observed from early to late July. Female moths oviposited on rosettes during the first ten days of July and continued through the end of July. Eggs were laid singly or in groups of five or six, firmly attached to the leaves of the host plant. In the laboratory, females mated within 24 hours of emergence and the preoviposition period lasted 2.6±0.8 days. The oviposition period lasted 7.4±2.2 days and the average number of eggs per female was 142.2±59.2. The incubation period was 12±4.7 days; the pupal stage lasted 14.7±2.4 days; and females lived 15.8±2.4 days, while males lived 10.7±1.4 days. First instar larvae failed to survive on economically important Compositae in the generaCynara L.,Helianthus L.,Zinnia L. andCalendula L. (Dunnet al., 1989).     

Biology and host range of Mada polluta,a potential biological control agent of Tecoma stans in South Africa

The host range of Mada polluta Mulsant (Coleoptera:Coccinnellidae) was studied to assess its suitability as a biological control agent of Tecoma stans (L.) Juss ex Kunth var stans (Bignoniaceae), an invasive weed in South Africa. Biology of M. polluta and its host range were determined in the laboratory using no-choice and multi-choice feeding, oviposition and larval survival tests. Out of 36 plant species from 12 plant families (Bignoniaceae, Acanthaceae, Asteraceae, Verbenaceae, Lamiaceae, Oleaceae, Cucurbitaceae, Fabaceae, Scrophulariaceae, Solanaceae, Apiaceae, Chenopodiaceae and Poaceae) within the order Lamiales that were tested during the host specificity testing, M. polluta showed a very strong preference for T. stans, depositing its eggs on T. stans and none on non-target plant species. In no-choice tests, an average of 246 eggs was laid on T. stans, and from these, 133 larvae developed to adulthood. The beetle also showed very promising biological attributes that will contribute to its success as a biological control agent of T. stans. These attributes include highly damaging larvae and adults, high fecundity (532 eggs/female) and a short life cycle (four weeks). The short life cycle will enable multiple generations per year and rapid population increase in the field. It is concluded that M. polluta is sufficiently host-specific to be released against T. stans in South Africa. It is, therefore, strongly recommended that permission be granted to release this beetle from quarantine for biological control of T. stans in South Africa.     

Biology of Mastrus ridibundus (Gravenhorst), a potential biological control agent for area‐wide management of Cydia pomonella (Linneaus) (Lepidoptera: Tortricidae)

The codling moth Cydia pomonella (Linnaeus) (Lepidoptera: Tortricidae) is a serious pest of pome fruit crops. A natural enemy of codling moth, the larval ectoparasitoid Mastrus ridibundus (Gravenhorst) (Hymenoptera: Ichneumonidae) has been imported into South America from the USA but little is known about the biology and ecology of the wasp, knowledge that is needed to design an efficient strategy of release and establishment. Experiments were carried out to assess important traits of the biology of the parasitoid in relation to its possible use as a biocontrol agent for codling moth. When M. ridibundus females were offered larvae ranging in weight from 37 to 78 mg, they oviposited more eggs on heavier hosts. In another study, the adult wasps were offered honey, diluted honey (10%) or pollen in paired choice tests and both males and females preferred honey over the other two foods. Females preferred 10% honey over pollen, while the males showed the opposite preference. Honey‐fed females lived longer than starved females. Adults died rapidly at 35°C, while they lived 20 days at 25°C and 12–17 days at 15°C. Female wasps had on average 25 ± 14 and 18 ± 11 progeny at 15 and 25°C, respectively, but they did not had progeny at 35°C. The development time (egg to adult emergence) was on average 44 ± 7 and 24 ± 2 days at 15 and 25°C respectively. Immature insects did not reach the adult stage at 35°C.     

Biology ofAllothrombium pulvinum (Acariformes: Trombidiidae), a potential biological control agent of aphids in China

Allothrombium pulvinum Ewing is a common natural enemy of aphids and other small arthropods in China. It is univoltine species. The eggs hatch in spring, nymphs emerge in early summer, and adults emerge in autumn. Larvae are ectoparasites of aphids, whereas deutonymphs and adults are free-living predators of aphids, spider mites and some lepidopterous eggs. Adults hibernate in soil during winter and females lay their eggs in spring. In the laboratory the duration of immature stages is 74 days at 20–30°C. Larvae kill their host within 1–3 days when the mite load of the host is 2 or more. They decrease the reproductive rate of adult aphids and arrest the development of early-instar aphid nymphs when the mite load is 1. Larvae are capable of controlling the host population growth when the level of parasitism is high. Owing to its wide occurrence,A. pulvinum could be a potential candidate in the biological control of aphids. It is suggested that its impact on the host pest be evaluated in IPM programs in the future.     

Predicting the host range of Nystalea ebalea: Secondary plant chemistry and host selection by a surrogate biological control agent of Schinus terebinthifolia

The safety of weed biological control depends upon the selection and utilization of the target weed by the agent while causing minimal harm to non-target species. Selection of weed species by biological control agents is determined by the presence of behavioral cues, generally host secondary plant compounds that elicit oviposition and feeding responses. Non-target species that possess the same behavioral cues as found in the target weed may be at risk of damage by classical biological control agents. Here we conducted host range tests and examined secondary plant compounds of several test plant species. We studied the specialist herbivore Nystalea ebalea (Lepidoptera: Notodontidae) a Neotropical species, present in Florida as a surrogate biological control agent of the weed, Brazilian peppertree Schinus terebinthifolia, invasive in Florida and Hawaii. We found that the larvae had the greatest survival when fed the target weed, the Neotropical species Spondias purpurea, the Florida native species Rhus copallinum, and the ornamental Pistacia chinensis. Reduced survival and general larval performance were found on the native species Metopium toxiferum and Toxicodendron radicans. Both the volatiles and the allergen urushiols were chemically characterized for all species but urushiol diversity and concentration best predicted host range of this herbivore species. These results provide insight into host selection and utilization by one oligophagous Schinus herbivore. Other potential biological control agents may also be sensitive to plants that contain urushiols and if so, they may pose minimal risk to these native species.