Journal of Science: Vol. 12 | Article 63 Morais et al.

Lepidopterans as potential agents for the biological control of the invasive plant, Miconia calvescens

Elisangela G.F. Morais1a*, Marcelo C. Picanço2b, Altair A. Semeão2c, Robert W. Barreto3d, Jander F. Rosado2e, and Julio C. Martins2f

1Embrapa Roraima - Brazilian Agricultural Research Corporation's, Boa Vista, Roraima, Brazil 2Department of Biology, Federal University of Viçosa, Viçosa, Minas Gerais, Brazil 3Department of Fitopathology, Federal University of Viçosa, Viçosa, Minas Gerais, Brazil

Abstract This work investigated eight species of associated with Miconia calvescens DC. (Myrtales: Melastomataceae) in Brazil, including six defoliators, Salbia lotanalis Druce (Lepidoptera: Pyralidae), Druentia inscita Schaus (Mimallonidae), Antiblemma leucocyma Hampson (Noctuidae), three Limacodidae species, a fruit borer Carposina cardinata Meyrick (Carposinidae), and a damager of flowers Pleuroprucha rudimentaria Guenée (Geometridae). Based on host specificity and the damage caused to plants, S. lotanalis and D. inscita are the most promising species for biological control of M. calvescens. Furthermore, if C. cardinata and P. rudimentaria have host specificity in future tests, these caterpillars could also be considered as appropriate biocontrol agents.

Keywords: caterpillar, host specificity, leaf roller, Salbia lotanalis, weed Correspondence: a [email protected], b [email protected], c [email protected], d [email protected], e [email protected], f [email protected], *Corresponding author Editor: J.P. Michaud was editor of this paper. Received: 14 June 2011, Accepted: 24 October 2011 Copyright : This is an open access paper. We use the Creative Commons Attribution 3.0 license that permits unrestricted use, provided that the paper is properly attributed. ISSN: 1536-2442 | Vol. 12, Number 63 Cite this paper as: Morais EGF, Picanço MC, Semeão AA, Barreto RW, Rosado JF, Martins JC. 2012. Lepidopterans as potential agents for the biological control of the invasive plant, Miconia calvescens. Journal of Insect Science 12:63 available online: insectscience.org/12.63

Journal of Insect Science | www.insectscience.org 1

Journal of Insect Science: Vol. 12 | Article 63 Morais et al.

Introduction damaging M. calvescens, including Salbia lotanalis Druce (Lepidoptera: Pyralidae) Exotic plant species can become great threats (mistakenly identified as Ategumia sp.), to the biodiversity of natural environments Druentia inscita Schaus (Mimallonidae) and and require expensive strategies for their Antiblemma leucocyma Hampson control. In most cas control with herbicides (Noctuidae). and removal is expensive and inefficient. In some cases, integration of biological control Several species of Lepidoptera have been and other strategies are also recommended. extensively used in the biological control of However, prior to the introduction, several invasive plants (Hight et al. 2002; Baars 2003; steps are required, such as a survey of Ostermeyer and Grace 2007; Boughton and feeding on the plant in its region of origin, and Pemberton 2009), and others have been an evaluation of damage and host specificity considered as potential agents (Dhileepan et of selected candidate agents (Morin et al. al. 2007). The major case of success was the 2009). introduction of Cactoblastis cactorum, a Miconia calvescens DC. (Myrtales: species native to Argentina that was Melastomataceae) is a severe weed found in introduced throughout the world for control of rainforest ecosystems on oceanic islands weedy Opuntia spp. (Pettey 1948). Among the (Medeiros et al. 1997), and is classified main groups used in the biological among the one hundred worst invasive species control programs of weeds (Acarina, around the world (Lowe et al. 2000; Murphy Coleoptera, Diptera, Hemiptera, et al. 2008). Miconia calvescens is native to Hymenoptera, Lepidoptera, and Central and South America and was Thysanoptera), Lepidoptera were the fastest introduced as an ornamental plant in French dispersers (Paynter and Bellgard 2011). The Polynesia, Hawaii, New Caledonia, and natural dispersal ability of a new agent is a Australia (Meyer 1996; Csurhes 1997; limiting factor for success of biological Medeiros et al. 1997). On the Hawaiian and control of weeds. Tahitian Islands, the problems caused by the invasion of this weed are more serious, and In fact, establishment and spread are two many areas have become extensive distinct issues in biological control programs monospecific stands of this plant (Meyer of weeds. The population dynamics of an 1996; Medeiros et al. 1997). In Australia, the agent post–introduction can be influenced by invasion of M. calvescens is more recent, but factors such as climate (Dennill and Gordon it has been classified as a major threat to the 1990; Zalucki and van Klinken 2006), natural native biodiversity in the country (Murphy et enemies (Hill and Hulley 1995; Paynter et al. al. 2008). 2010), resource availability (quantity and quality of host plant) (Foxcroft et al. 2007), To select agents for the biological control of and the appearance of more vigorous M. calvescens, surveys on the that genotypes of weeds (Sheppard 1992; Nissen attack this weed in Brazil have been carried et al 1995). Crawley (1986) found that 44% of out since 2001 (Picanço et al. 2005; Morais et the failures of weed control by biological al. 2010a, 2010b). In these previous studies, agents are to due the non–adaptation of these some species of Lepidoptera were found organisms to climate conditions in the country

Journal of Insect Science | www.insectscience.org 2

Journal of Insect Science: Vol. 12 | Article 63 Morais et al. of introduction. Surveys of abundance of altitude and was located at 19º 50' 34" S, 42º possible biocontrol agent can be used to 46' 36" W. During the experimental period, improve understanding of the climate effects the average temperature was 23.2 ºC, and on the agent and of its establishment in new rainfall reached 1003 mm/yr. areas (Zalucki and van Klinken 2006). According to Paynter and Bellgard (2011), the knowledge about parameters of a biocontrol The experimental area in Guaraciaba was agent, prior to its introduction (agent located in an abandoned pasture, under an voltinism, parasitoid diversity in the native initial natural recovery process, and was 1 km range, habitat, fecundity, taxon and life–style away from the Piranga River. This area was category) could be used to predict if it will 580 m in altitude and located at 20º 34' 36" S, establish and disperse in a new environment. 43º 02' 26" W. During the evaluation period, the average temperature was 19.6 ºC, and This work presents the damage, host rainfall reached 988.5 mm/yr. Miconia specificity, population dynamics, and the field calvescens plants in this area were discovered occurrence of natural enemies of Lepidoptera at the beginning of 2005; therefore, species associated with M. calvescens in evaluations at this site started only in that Brazil; the defoliators S. lotanalis, D. inscita, year. and A. leucocyma, three Limacodidae species, the fruit borer Carposina cardinata Meyrick (Carposinidae) and the flower damaging Evaluations were performed during three Pleuroprucha rudimentaria Guenée different periods: the first period spanned June (Geometridae). The mass–rearing techniques 2001 to June 2002; the second period was of these species are also described. from February 2004 to February 2005; and the third period was from March 2005 to February 2006. During the first period, the plants were Materials and Methods assessed every three weeks in Viçosa and Population dynamics Dionísio. During the second period, The population density of Lepidoptera evaluations were carried out at 10-day associated with M. calvescens was evaluated intervals in Viçosa and monthly in Dionísio. in the counties of Viçosa, Dionísio, and In the third period, evaluations were made Guaraciaba in Minas Gerais, Brazil. The every two weeks in Viçosa and Guaraciaba. experimental area in Viçosa was located in a secondary forest fragment under an advanced Each evaluation involved the randomised stage of natural recovery. The area was sampling of ten M. calvescens plants, each located at an altitude of 700 m (20º 46' 37" S, one with a height of 1-7 m and a basal trunk 42º 50' 36" W). During the experimental diameter of 6-15 mm. For each plant, the period, the average temperature was 19.4 ºC, number of Lepidoptera was counted in the and rainfall reached 1221 mm/yr. main stem, six secondary branches, six inflorescences, and six infructescences. Leaves and buds from the secondary branches In Dionísio, the experimental area was located were examined. Six infrutescences were also in a riparian forest at the Doce River margins, brought to the laboratory to determinate at the Parque Estadual do Rio Doce (Doce whether Lepidoptera larvae were present River State Park). The forest was at 344 m inside the fruits. During the evaluations, the

Journal of Insect Science | www.insectscience.org 3

Journal of Insect Science: Vol. 12 | Article 63 Morais et al. organ that each Lepidoptera attacked and the assessed during the site visits. Evaluations type and level of the damage caused by each were made until February 2006 because after species were recorded. this time, the seedlings did not establish in the Samples of adult Lepidoptera were forwarded field due to defoliation by ants (Atta spp.). to Dr. Vitor Becker for species identification. Samples of these species were delivered to the Effect of climatic parameters on caterpillar Regional Museum of Entomology of the population dynamics Federal University of Viçosa. Canonical correlations and Pearson correlations were carried out to determinate Climate data from Viçosa were provided by the relationship between the population the Climate Station of the Federal University density of Lepidoptera and the climatic of Viçosa; data from Guaraciaba were parameters (average, maximum, and provided by the Ponte Nova Climate Station minimum air temperature, relative humidity, (belonging to the Federal University of rainfall, photoperiod, and wind speed). Viçosa); and data from Dionísio came from Statistical analyses were performed using the the Ponte Alta Climate Station (belonging to SAS procedures PROC CANCOR and PROC the private company CAF Santa Bárbara COR. In these analyses, the related to Ltda). Lepidoptera density of each evaluation and the averages of climatic parameters 15 days preceding the evaluations were applied, Population dynamics: Hawaiian biotype except for rainfall, for which the total rain in In September 2005, 10 M. calvescens the 15 days before the evaluations was seedlings of Hawaiian biotype were applied. transplanted into the ground at the experimental area in Guaraciaba. These Parasitism seedlings were used because plants from During the population peaks of Lepidoptera, Hawaii have a distinct appearance, with leaves 10 caterpillars of each species were collected slightly softer and different in color (purple in in the field and brought to the laboratory to abbatial face; in Brazilian plants this part is verify the emergence of parasitoids. These green). Therefore, the objective of this study caterpillars were maintained in a rearing room was observed whether the seedlings of with a light regime of 12:12 L:D, 26 ± 2 ºC air Hawaiian biotype were naturally attacked by temperature, and 70 ± 10% relative humidity. caterpillars. The defoliator caterpillars were placed in plastic bags (40 × 50 cm) with leaves of M. Seeds were collected in Hawaii and calvescens. Fruit borer caterpillars were germinated in a box with sand. When the placed into transparent 2 L plastic pots (14 cm seedlings were at a height of 30-40 cm, they diameter, 14 cm height) with fruits inside. were transplanted into the ground at Flower feeding caterpillars were put inside Guaraciaba the experimental area next to M. wooden cage (0.5 × 1.0 × 1.2 m), enclosed in calvescens plants that grew naturally. organza containing inflorescences of M. Seedlings were arranged in a row with 2 m calvescens. The bags, pots, and cages that spacing between the plants. From September contained the caterpillars were maintained in a 2005 to February 2006, the natural attacks on rearing room until the caterpillars pupated. these seedlings by caterpillars were also The number of parasitized caterpillars and the

Journal of Insect Science | www.insectscience.org 4

Journal of Insect Science: Vol. 12 | Article 63 Morais et al. number of emerged adult parasitoids of each transplanted to 5 L pots with soil. When the caterpillar were counted every day. The adult seedlings were 70-100 cm tall, four seedlings parasitoids were collected and forwarded to of each Melastomataceae species were put in a Dr. Maria Angélica Penteado-Dias for species wooden cage (0.5 × 1.0 × 1.2 m), enclosed in identification. organza and infested with five S. lotanalis caterpillars of the third instar. The caterpillars Host specificity: field conditions were reared in rooms similar to that described During the evaluations, all vegetation in the 2 above. The instar of the caterpillars, their m radius of M. calvescens sampled during the survival and the damage they caused to the population dynamics experiment was also seedlings were recorded daily. observed to check whether the Lepidoptera species had attacked other plant species. Mass–rearing Among the species found close to M. Salbia lotanalis, D. inscita, A. leucocyma, C. calvescens, other Melastomataceae were cardinata, and P. rudimentaria were reared in found. The most common were as follows: rearing rooms similar to the room described Clidemia capitellata (Bonpl.) D. Don., above. Caterpillars of S. lotanalis, D. inscita, Clidemia hirta (L.) Don, Leandra lacunosa and A. leucocyma were collected and placed Cogn, Miconia albicans (Sw.) Triana, M. into transparent plastic bags (40 × 50 cm) with mendoncaei Cogn., Miconia ibaguensis leaves of M. calvescens. Five M. calvescens (Bonpl.), and Ossala confertiflora (DC.). leaves were placed in each bag along with five Organs of these species were carried to the caterpillars of similar size. These bags were laboratory to determinate whether Lepidoptera maintained in a rearing room until the larvae were present inside. caterpillars pupated. The pupae were separated by sex according to characters described by Butt and Cantu (1962) and Host specificity tests: Salbia lotanalis placed into plastic pots with wet vermiculite Host specificity tests of S. lotanalis were to avoid desiccation. carried out in glasshouses. These tests were Emerged adults were transferred to a wooden conducted with the following cage (0.5 × 1.0 × 1.2 m) and enclosed in Melastomataceae species: C. capitellata, C. organza with a young M. calvescens plant hirta, M. mendoncaei, M. albicans, M. (0.8-1.0 m high). 10 pairs were placed inside ibaguensis, L. lacunosa, O. confertiflora, each cage and fed with a 10% honey solution. Tibouchina granulose, and T. moricandiana. After the eggs hatched, the caterpillars were Hawaiian biotype of M. calvescens was also transferred with a fine brush to leaves of M. tested. calvescens, which were put inside plastic bags (40 × 50 cm) and maintained in a rearing The seeds of the Melastomataceae species room, following the same procedures were extracted from fresh fruits of trees mentioned above. located in the field where the evaluations were performed. The seeds of the Hawaiian biotype Miconia calvescens fruits attacked by C. of M. calvescens were collected in Hawaii. cardinata were collected in the field and put The seeds were germinated in a box with in into transparent 2 L plastic pots (14 cm sand. After about three months, when diameter, 14 cm height). Each pot received seedlings were 10-20 cm tall, they were about 80 g of fruit. The pots were capped with

Journal of Insect Science | www.insectscience.org 5

Journal of Insect Science: Vol. 12 | Article 63 Morais et al. organza and their bottoms were covered with where they feed and eventually pupate. One paper for absorbing the dampness of the fruits. caterpillar is able to destroy an entire leaf, When the fruits began to rot, the caterpillars causing its abscission. Adults are small brown were removed and transferred to pots with yellowish patterns on the wings containing healthy fruits. As caterpillars and caterpillars are green with black patterns attained the pupa stage outside the fruits, the on the body (Figure 1A-C). pupae could be seen and removed from the pots. The pupae were separated by sex The density of S. lotanalis was higher in according to characters described by Butt and Dionísio and Guaraciaba than Viçosa. In Cantu (1962) and placed into plastic pots with Dionísio, population peaks occurred from wet vermiculite. The emerged adults were January to April 2002 and from May to transferred to pots containing mature fruits of October 2004. In Guaraciaba, S. lotanalis M. calvescens. 10 pairs of C. cardinata were caterpillars were found throughout the whole put in each pot, and they were fed with a 10% evaluation period, with population peaks from honey solution. Emerged caterpillars from June to September 2005. In Viçosa, S. these pots were reared as mentioned above. lotanalis caterpillars were found occasionally and at a low density (Figure 2A). During Caterpillars of P. rudimentaria were reared in population peaks, the damage caused by these wooden cages (0.5 × 0.5 × 0.5 m) and caterpillars to M. calvescens plants was high, enclosed in organza with M. calvescens amounting to 50% defoliation. In these inflorescences. The inflorescences were periods, some seedlings of M. calvescens died collected in the field and maintained with their due the high levels of defoliation by this leaf peduncle immersed in glass bottles containing roller. 100 mL of water. 20 caterpillars of the same instar and four inflorescences were put in each Salbia lotanalis and one species of cage. The inflorescences were changed every Limacodidae were observed attacking M. five days. Pupae were separated by sex calvescens seedlings of the Hawaiian biotype according to characters described by Butt and transplanted into the experimental area in Cantu (1962) and placed into plastic pots with Guaraciaba (Table 1). Two weeks after the wet vermiculite. The emerged adults were transplantation of the seedlings, S. lotanalis transferred to wooden cages (same characters caterpillars were found on these plants. The described above) with four M. calvescens highest attack occurred from September to inflorescences. 10 pairs of P. rudimentaria December 2006 (Figure 2A). In this period, all were placed in each cage, and they were fed seedlings were attacked by at least one with a 10% honey solution. After the eggs Table 1. Host range of Salbia lotanalis as revealed by attack (+) hatched, caterpillars were transferred to new or non–attack (−) on Melastomataceae species in field M. calvescens inflorescences and reared as evaluations and in host specificity tests.

mentioned above.

Results

Salbia lotanalis Salbia lotanalis is a leaf roller. Its caterpillars

roll leaves longitudinally, forming tubes

n.e. = not evaluated

Journal of Insect Science | www.insectscience.org 6

Journal of Insect Science: Vol. 12 | Article 63 Morais et al. caterpillar. Furthermore, in December, one lotanalis, D. inscita caterpillars are found seedling died due to the damage caused by singly inside the leaf cylinders. This species this caterpillar species. Only four was also more abundant in Dionísio, with Limacodidae caterpillars were found population peaks from May to June 2002 and defoliating the seedlings in October. from February to June 2004. In Viçosa, this Nearly 10% of S. lotanalis caterpillars brought caterpillar species was found occasionally and from the field were parasitised. Three at low density. In Guaraciaba, it was found parasitoid species were found: two only in March 2005 (Figure 2B). Hymenoptera (Apanteles sp. (Braconidae) and Pediobius sp. (Eulophidae)) and a Diptera No parasitism was verified in D. inscita. No (Tachinidae). Apanteles sp. larvae emerged caterpillars of this lepidopteran were found in from the last instar of S. lotanalis and pupated other plant species during the field outside the caterpillars. The pupae of this evaluations, including the other braconid are white, and about 20-30 larvae Melastomataceae that were assessed (Table emerged from each caterpillar. Pupae of 1). Tachinidae were also found in the last instar of S. lotanalis. The pupae of this Diptera are Antiblemma leucocyma brown, and about 6-10 pupae emerged in each Antiblemma leucocyma caterpillars are caterpillar. Three S. lotanalis pupae were also defoliators that make irregular holes in leaves. parasitized by Conura sp. (Hymenoptera: They are green and have a distinctive looping Chalcididae) in October 2004. walk (Figure 1E-F), and were found in all three places (Figure 2C). Caterpillars of this In the field, S. lotanalis caterpillars were not species were found during the evaluation found on other plants species, including the period in Dionísio. In Viçosa, their population other Melastomataceae assessed (Table 1). peaks occurred from April 2005 to February However, a similar species, Ategumia 2006, while in Guaraciaba, they occurred only matutinalis Guenée, was observed attacking from May to July 2005 (Figure 2C). C. hirta. Ategumia matutinalis caterpillars are very similar to S. lotanalis, with similar wing About 30% of the A. leucocyma caterpillars color patterns as adults (Figure 1D). In host brought from the field were parasitized by specificity tests, S. lotanalis caterpillars were Hymenoptera (Apanteles sp. (Braconidae)). able to develop through to pupation on both C. Many A. leucocyma caterpillars found in the hirta and the Hawaiian biotype of M. field had dark coloration, indicating that they calvescens. In other Melastomataceae, no were parasitized. Caterpillars of this damage by caterpillars was verified, and they Lepidoptera were not found in the other plant died of starvation (Table 1). species assessed (Table 1).

Druentia inscita Limacodidae Druentia inscita caterpillars are also leaf Three species of Limacodidae caterpillars rollers. They feed mostly on apical foliage, were found occasionally in Viçosa, Dionísio, which they can completely destroy. The and Guaraciaba, causing defoliation of M. caterpillars are yellow with brown stripes and calvescens plants. These caterpillars were also the adults are brown with two darker stripes found on the other sampled plant species on the wings (Figure 1E-F). Similarly to S.

Journal of Insect Science | www.insectscience.org 7

Journal of Insect Science: Vol. 12 | Article 63 Morais et al. neighboring M. calvescens, including plants they were present from May to September belonging to the Melastomataceae family. 2005 (Figure 3B). During these population peaks, about 40% of inflorescences were Carposina cardinata completely stripped by this caterpillar species, Carposina cardinata is a fruit borer of M. and the fruit production of the M. calvescens calvescens and is found only in Guaraciaba. plants was extremely low. No P. rudimentaria Its caterpillars are beige and in the last instar caterpillars brought to the laboratory were they measure about 3 mm long. Adults are parasitized. This caterpillar was not found in micro moths, bright gray, and measure about 5 other plant species, including the other mm (Figure 1H-I). They live inside fruits and Melastomataceae assessed (Table 1). feed on the flesh and seeds, causing severe damage. One to five caterpillars were found in Effect of climatic parameters on caterpillar each M. calvescens fruit. Carposina cardinata population dynamics caterpillars were found in Guaraciaba in 2005, There was a significant canonical correlation from March to September, which between the density of Lepidoptera associated corresponded to the period of M. calvescens with M. calvescens and climatic parameters fruit production (Figure 3A). No parasitized (Wilks’ Lambda = 0.35, F = 2.72, df num/den= caterpillars were verified. During this period, 35/335, p < 0.01). Five canonical axes were the following Melastomataceae were also in generated, out of which only the first two their fruiting periods: C. capitellata, C. hirta, were significant at 5% probability and L. lacunosa, M. albicans, M. mendoncaii, M. explained 86% of the correlation. According ibaguenscens, and O. confertiflora. However, to the first canonical axis, among the no C. cardinata was found in the fruits of Lepidoptera, S. lotanalis was the most these species (Table 1). influenced by the climatic parameters (r = −0.89), and according to the second axis, D. Pleuroprucha rudimentaria inscita (r = 0.86) was the most influenced by Pleuroprucha rudimentaria caterpillars feed the climatic parameters. The climatic on flowers. These caterpillars are brown, with parameters that most affected the Lepidoptera white stripes and a distinctive looping walk density were the maximum (r = 0.75) and (Figure 1J-L). This species was found only in average (r = 0.79) air temperature according Viçosa and Guaraciaba. In Viçosa, caterpillars to the first axis, and the minimum air attacked M. calvescens from April to June temperature (0.57) according to the second 2005 and in December 2005. In Guaraciaba, axis (Table 2). Table 2. Canonical correlations and canonical pairs between the abundance of Lepidoptera associated with Miconia calvescens and climatic parameters in Viçosa, Dionísio and Guaraciaba, Minas Gerais, Brazil.

Journal of Insect Science | www.insectscience.org 8

Journal of Insect Science: Vol. 12 | Article 63 Morais et al.

There was a positive Pearson correlation Carposina cardinata and P. rudimentaria can among the minimum, maximum, and average be reared in the laboratory during the air temperatures and the population density of flowering and fruiting periods of M. D. inscita (r = 0.34, p < 0.01, r = 0.25, p < 0.01 calvescens. Only two generations of these and r = 0.28, p < 0.01, respectively). In fact, Lepidoptera were reared in the laboratory highest densities of D. inscita were observed because afterwards there were no fruits or Dionísio, where air temperatures were highest flowers in the field for feeding the caterpillars (28.6, 19.6, 23.2 ºC maximum, average, and or for egg–laying by the adults. Carposina minimum air temperatures, respectively) than cardinata caterpillars survived and fed on Viçosa (27.2, 15.1, 20.1 ºC maximum, average, fruits maintained in plastic pots. Their adults and minimum air temperatures, respectively) also laid eggs in M. calvescens fruits inside and Guaraciaba (22.9, 16.3, 16.3 ºC maximum, the pots. Pleuroprucha rudimentaria average, and minimum air temperatures, caterpillars fed on the inflorescences brought respectively) (Figure 4). There was a negative from the field, and adults laid their eggs in Pearson correlation among the maximum and these inflorescences. average air temperatures and the population density of S. lotanalis (r = −0.38, p < 0.01 and Discussion r = −0.39, p < 0. 01, respectively). On the other hand, S. lotanalis was more frequent in Salbia lotanalis has already been considered Guaraciaba, where the temperatures were as a potential agent due its high reproductive lowest. In addition the population peaks of capacity and impact on young plants (Morais this caterpillar occurred in colder months et al. 2010b). The constant presence of S. (Figure 4). lotanalis in the field indicates that this lepidopteran may establish well in places Mass–rearing where M. calvescens causes problems, since Salbia lotanalis, D. inscita, and A. leucocyma agents that are frequent in the native range can be reared throughout the year under may remain frequent in the introduced laboratory conditions. As M. calvescens trees environment (Sheppard 2002). The host produce a large number of leaves throughout specificity study demonstrated that S. the year, the caterpillars of these species had lotanalis could feed on the leaves of M. plenty of good quality food. Furthermore, M. calvescens and C. hirta. Additionally, its calvescens leaves stayed in good condition for caterpillars were not found in other species in a week because they were kept in the plastic the field. Although this species attacks C. bags. Salbia lotanalis was the easiest to rear in hirta in the field, it does not hinder its the laboratory because of its shorter life cycle introduction as a biocontrol agent for M. (about 70-100 days), high reproductive rate calvescens, because C. hirta is often also a (~90 eggs/female), and low mortality rate weed where M. calvescens causes problems (more than 70% of individuals reached the (Wester and Wood 1977; Medeiros et al. adult phase). The life cycles of D. inscita and of 1997). In this case, the S. lotanalis attack on A. leucocyma are longer, lasting 150-200 and C. hirta could be desirable. 100-120 days, respectively. The reproductive rate of both these species is about 30 A closely related leaf roller, Salbia eggs/female. haemorrhoidalis was used in successful

Journal of Insect Science | www.insectscience.org 9

Journal of Insect Science: Vol. 12 | Article 63 Morais et al. programs for the biological control of Lantana lepidopteran can reduce seed fruit production. camara in South Africa. This species showed Studies have verified that herbivores of good establishment and distribution in this reproductive structures can be very effective country, as well as high population densities weed biocontrol agents because they can (Baars 2003). Another great advantage of reduce recruitment to the seedling stage of the employing S. lotanalis as a biocontrol agent is populations (Bellows and Headrick 1999). A that it can easily be raised under laboratory similar Carposina, C. bullata, was used as a conditions, with high population growth rates biocontrol agent of C. hirta in Hawaii, where (Morais et al. 2010). it established and became very effective. This species feeds internally in the flower buds and Druentia inscita also causes severe damage to M. is specific to the Miconia and Clidemia genera calvescens plants in the field. During the and the damage to the flowers of C. hirta evaluations, dead branches were observed due prevents seed production (Nakahara et al. to the attack of this caterpillar on terminal 1992). The major difficulty in the introduction buds. Nakahara et al. (1992) verified that D. of C. bullata was achieving a large number of inscita is restricted to 13 species of individuals to be released because no mass– Melastomataceae out of the 39 species surveyed, rearing method was developed (Culliney and and no larval feeding damage was observed on Nagamine 2000). Therefore, we verified that non–Melastomataceae from the 15 families tested. C. cardinata could be reared in the laboratory Druentia inscita was introduced in Hawaii for the during the fruiting period of M. calvescens, control of C. hirta, occupying upper–elevation obtaining a great number of adults for sites and shaded environments (Nakahara et al. releasing in the field. Another species of 1992). However, its establishment was not very Carposina genera used in control of weed is successful probably because few individuals were the seed–, C. autologa, introduced in introduced (Culliney and Nagamine 2000). South Africa to control a serious invasive plant Hakea sericea, a small tree Australian Compared with S. lotanalis and D. inscita, A. origins (Gordon and Fourie 2011). leucocyma caused less damage to M. calvescens, and there were long periods where Pleuroprucha rudimentaria is also a potential these species was not recorded in the field. biocontrol agent for M. calvescens. Fruit The low density of A. leucocyma may be due production decreased about 40% in the to the high parasitism rates by Apanteles sp. flowering period of 2005 in Viçosa and However, high host specificity of this Dionísio. Furthermore, although it was not Lepidoptera is an advantage for the biological recorded, a severe attack was observed in control of M. calvescens (Badenes-Perez and Viçosa in 2006 (Morais, personal Johnson 2008). A similar caterpillar, observations). Pleuroprucha rudimentaria is Antiblemma acclinalis, from Trinidad was reported to occur in Costa Rica, Peru, introduced in Hawaii for the control of C. Jamaica, and Venezuela, but there is no hirta and had a confirmed establishment only information about its host specificity (BLDS in Oahu (Culliney and Nagamine 2000; 2010). During the field evaluations, no Conant 2002). caterpillars of P. rudimentaria and C. cardinata were found attacking others plant Carposina cardinata is an important fruit species. Therefore, it would be worth borer of M. calvescens. Attacks of this

Journal of Insect Science | www.insectscience.org 10

Journal of Insect Science: Vol. 12 | Article 63 Morais et al. conducting additional host–specificity tests lotanalis and D. inscita showed differences in for these species. their temporal distribution; S. lotanalis occurs in cooler periods, and D. inscita occurs in warmer Parasitism may be a limiting factor for periods. The same pattern was observed for C. establishment of lepidopteran species in new cardinata and P. rudimentaria, the first areas. Native parasitoids have reduced the species feeds on fruits and the second feeds on efficacy of some biocontrol agents, such as flowers. An effective result in the control of Neogalea sunia and S. haemorrhoidalis in the Sesbania punicea was obtained with the control of Lantana camara (Hill and Hulley combined action of a seed–feeder 1995; Baars and Neser 1999). Therefore, Rhyssomatus marginatus, two defoliators although S. lotanalis has been parasitized by Tyria jacobaeae, and a flea beetle Longitarsus several species, this caterpillar maintains its jacobaeae (McEvoy et al. 1990; Hoffmann high density in the field, causing severe and Moran 1992). damage to M. calvescens. Moreover, no parasitism was verified in D. inscita, C. Considering the difficulty and high cost of M. cardinata, or P. rudimentaria. calvescens control by removal (Burnett et al. 2007; Hester et al. 2010), the introduction of Climate is also an important factor for the effective biocontrol agents is necessary to establishment of biocontrol agents. Climatic reduce the population density of M. factors can limit the dispersal of biological calvescens in Tahiti and Hawaii. Therefore, agents to new areas within the geographic the traits of S. lotanalis and D. inscita, such as range of the weeds (Baars 2003). Air host specificity, high population growth rate, temperature is one of the most important high rate of damage to the plants, relatively factors interfering with the establishment, easy mass–rearing under laboratory dispersal, and impact of organisms in new conditions, and occurrence in the field areas of distribution (Baker 2002). Salbia throughout the year indicate that these lotanalis seems to develop better in periods of Lepidoptera can be good classical biocontrol cooler air temperatures and D. inscita in hotter agents against M. calvescens. Furthermore, if seasons (Figures 2A and 4). C. cardinata and P. rudimentaria show host specificity in future tests, these caterpillars The introduction of multiple herbivores as could also be considered biocontrol agents. biocontrol agents on a single weed species may be advantageous, particularly where they Acknowledgements attack different plant parts and suppress plant growth and reproduction. In this case, the This study was supported by the USGS BRD introduction of S. lotanalis and D. inscita Pacific Island Ecosystem Research Centre, the (defoliators) and C. cardinata and P. National Park Service, the Research rudimentaria (attack on reproductive organs) Corporation of Hawaii, and the National can be a good strategy for the control of M. Council for Scientific and Technological calvescens. Although S. lotanalis and D. Development (CNPq). We thank the company inscita are defoliators, S. lotanalis feed on CAF Santa Bárbara Ltda for providing the mature leaves, and D. inscita feed on young climatic data and permitting the experiment leaves in terminal buds, which prevents on their property. We also thank Dr. Vitor competition between them. Furthermore, S. Becker (Federal University of Brasília, Brazil)

Journal of Insect Science | www.insectscience.org 11

Journal of Insect Science: Vol. 12 | Article 63 Morais et al. for identification of Lepidoptera species and Boughton AJ, Pemberton RW. 2009. Dr. Maria Angélica Penteado-Dias (Federal Establishment of an imported natural enemy, University of São Carlos, Brazil) for Neomusotima conspurcatalis (Lepidoptera: identification of parasitoid species. ) against an invasive weed, Old World climbing fern, Lygodium References microphyllum, in Florida. Biocontrol Science and Technology 19(7): 769-772. Baars JR. 2003. Geographic range, impact, and parasitism of lepidopteran species Burnett K, Kaiser B, Roumasset J. 2007. associated with the invasive weed Lantana Economic lessons from control efforts for an camara in South Africa. Biological Control invasive species: Miconia calvescens in 28(3): 293-301. Hawaii. Journal of Forest Economics 13(2-3): 151-167. Baars JR, Neser S. 1999. Past and present initiatives on the biological control of Lantana Butt BA, Cantu E. 1962. Sex Determination of camara (Verbenaceae) in South Africa. Lepidopterous Pupae. pp. 33-75. USDA. African Entomology Memoir 1: 21-33. Conant P. 2002. Classical biological control Badenes-Perez FR, Johnson MT. 2008. of Clidemia hirta (Melastomataceae) in Biology, herbivory, and host specificity of Hawaii using multiple strategies. In: Smith Antiblemma leucocyma (Lepidoptera: CW, Denslow JS, Hight S, Editors. Workshop Noctuidae) on Miconia calvescens DC. on Biological Control of Invasive Plants in (Melastomataceae) in Brazil. Biocontrol Native Hawaiian Ecosystems. pp. 13-20. Science and Technology 18(2): 183-192. Pacific Cooperative Studies Unit, University of Hawaii at Manoa. Baker RHA. 2002. Predicting the limits to the potential distribution of alien crop pests. In: Crawley MJ. 1986. The population biology of Hallman GJ, Schwalbe CP, Editors. Invasive invaders. Philosophical Transactions of the Arthropods in Agriculture Problems and Royal Society of London B: Biological Solutions. pp. 207-241. Science Publishers. Sciences 314(1167): 711-731.

Bellows TS, Headrick DH. 1999. Arthropods Csurhes SM. 1997. Miconia calvescens, a and vertebrates in biological control of plants. potentially invasive plant in Australia’s In: Bellows TS, Fischer TW, Editors. tropical and sub–tropical rainforests. First Handbook of Biological Control. pp. 505-516. Regional Conference on Miconia Control, pp. Academic Press. 72-77.

Barcode of Life Data Systems. 2010. Culliney TW, Nagamine WT. 2000. Lepidoptera Barcode of Life: Geometridae. Introductions for Biological Control in Biodiversity Institute of Ontaro. Available Hawaii, 1987-1996. Proceedings of the online, Hawaiian Entomological Society 34(1): 101- http://www.lepbarcoding.org/geometridae/spe 113. cies.php?region=1&id=40667

Journal of Insect Science | www.insectscience.org 12

Journal of Insect Science: Vol. 12 | Article 63 Morais et al. Dennill GB, Gordon AJ. 1990. Climate– biocontrol agents introduced to South Africa. related differences in the efficacy of the Biological Control 5(2): 297-302. Australian gall wasp (Hymenoptera: Pteromalidae) released for control of Acacia Hoffmann JH, Moran VC. 1992. Oviposition longifolia in South Africa. Environmental patterns and the supplementary role of a seed– Entomology 19(1): 130-136. feeding weevil, Rhyssomatus marginatus (Coleoptera: Curculionidae), in the biological Dhileepan K, Snow EL, Rafter MA, Trevino control of a perennial leguminous weed, M, McCarthy J, Senaratne K. 2007. The leaf– Sesbania punicea. Bulletin of Entomological tying moth Hypocosmia pyrochroma (Lep., Research 82(3): 343-347. Pyralidae), a host–specific biological control agent for cat's claw creeper Macfadyena Lowe S, Browne M, Boudjelas S, De Poorter unguis-cati (Bignoniaceae) in Australia. M. 2000. 100 of the World’s Worst Invasive Journal of Applied Entomology 131(8): 564- Alien Species: A Selection from the Global 568. Invasive Species Database. The Invasive Species Specialist Group (ISSG) a specialist Foxcroft LC, Hoffmann JH, Viljoen JJ, Kotze group of the Species Survival Commission JJ. 2007. Factors influencing the distribution (SSC) of the World Conservation Union of Cactoblastis cactorum, a biological control (IUCN). Available online, agent of Opuntia stricta in Kruger National http://www.issg.org/ Park, South Africa. South African Journal of Botany 73(1): 113-117. McEvoy PB, Cox CS, James RR, Rudd NT. 1990. Ecological mechanisms underlying Gordon AJ, Fourie A. 2011. Biological successful biological weed control: Field Control of Hakea sericea Schrad. & experiments with ragweed Senecio jacobaea. J.C.Wendl. and Hakea gibbosa (Sm.) Cav. In: Delfosse E, Editor. Proceedings of the (Proteaceae) in South Africa. African Seventh International Symposium on Entomology 19(2): 303-314. Biological Control of Weeds, pp. 55-66.

Hester SM, Brooks SJ, Cacho OJ, Panetta FD. Medeiros AC, Loope LL, Conant P, 2010. Applying a simulation model to the Mcelvaney S. 1997. Status, ecology, and management of an infestation of Miconia management of the invasive plant, Miconia calvescens in the wet tropics of Australia. calvescens DC (Melastomataceae) in the Weed Research 50(3): 269-279. Hawaiian Islands. Bishop Museum Occasional Papers 48(1): 23-36. Hight SD, Carpenter JE, Bloem KA, Bloem S, Pemberton RW, Stiling P. 2002. Expanding Meyer JY. 1996. Status of Miconia calvescens geographical range of Cactoblastis cactorum (Melastomataceae), a dominant invasive tree (Lepidoptera: Pyralidae) in North America. in the Society Islands (French Polynesia). The Florida Entomologist 85(5): 527-529. Pacific Science 50(1): 66-76.

Hill MP, Hulley PE. 1995. Host–range Morais EGF, Picanco MC, Barreto RW, Silva extension by native parasitoids to weed NR, Campos MR. 2010a. Biological performance of Diclidophlebia smithi

Journal of Insect Science | www.insectscience.org 13

Journal of Insect Science: Vol. 12 | Article 63 Morais et al. (Hemiptera: Psyllidae), a potential biocontrol agent for the invasive weed Miconia Paynter Q, Fowler SV, Gourlay AH, calvescens. Biocontrol Science and Groenteman R, Peterson PG, Smith RL, Technology 20(1): 107-116. Winks CJ. 2010. Predicting parasitoid accumulation on biological control agents of Morais EGF, Picanço MC, Barreto RWS, GA weeds. Journal of Applied Ecology 47(3): Silva, Moreno SC, Queiroz RB. 2010b. 575-582. Biology of the leaf roller Salbia lotanalis and its impact on the invasive tree Miconia Paynter Q, Bellgard S. 2011. Understanding calvescens. Biocontrol 55(5): 685-694. dispersal rates of invading weed biocontrol agents. Journal of Applied Ecology 48(2): Morin L, Reid AM, Sims-Chilton NM, 407-414. Buckley YM, Dhileepan K, Hastwell GT, Nordblom TL, Raghu S. 2009. Review of Pettey FD. 1948. The Biological Control of approaches to evaluate the effectiveness of Prickly Pears in South Africa. Department of weed biological control agents. Biological Agriculture, Union of South Africa. Control 51(1): 1-15. Picanço MC, Barreto RW, Fidelis EG, Murphy HT, Hardesty BD, Fletcher CS, et al. Semeão AA, Rosado JF, Moreno SC, Barros 2008. Predicting dispersal and recruitment of EC, Silva GA, Johnson T. 2005. Biological Miconia calvescens (Melastomataceae) in Control of Miconia calvescens by Australian tropical rainforests. Biological Phytophagous Arthropods. University of Invasions 10(6): 925-936. Hawai'i.

Nakahara LM, Burkhart RM, Funasaki GY. Sheppard AW. 1992. Predicting biological 1992. Review and status of biological control weed control. Trends in Ecology and of Clidemia in Hawaii. In: Stone CP, Tunison Evolution 7(9): 290-291. JT, Smith CW, Editors. Alien Plant Invasions in Native Ecosystems of Hawaii: Management Sheppard AW. 2002. Improving the selection, and Research. pp. 452-465. University of testing and evaluation of weed biological Hawaii Cooperative National Park Resources control agents. In: Spafford JH, Briese DT, Studies Unit. Editors. Prioritising agents based on predicted efficacy: beyond the lottery Nissen SJ, Masters RA, Lee DJ, Rowe ML. approach. CRC for Australian Weed 1995. DNA-based marker systems to Management. pp. 11-21. Adelaide. determine genetic diversity of weedy species and their application to biocontrol. Weed Wester LL, Wood HB. 1977. Koster’s curse Science 43(3): 504-513. (Clidemia hirta), a weed pest in Hawaiian forests. Environmental Conservation 4(1): 35- Ostermeyer N, Grace BS. 2007. 41. Establishment, distribution and abundance of Mimosa pigra biological control agents in Zalucki MP, van Klinken RD. 2006. northern Australia: implications for biological Predicting population dynamics of weed control. Biocontrol 52(5): 703-720. biological control agents: science or gazing

Journal of Insect Science | www.insectscience.org 14

Journal of Insect Science: Vol. 12 | Article 63 Morais et al. into crystal balls? Australian Journal of Entomology 45(4): 331-344.

Figure 1. (A)Adult, (B) caterpillar, and (C) damage caused by Salbia lotanalis to Miconia calvescens seedlings (Hawaiian biotype); (D) adult of Ategumia matutinalis; (E) adult and (F) caterpillar of Druentia inscita; (G) caterpillar of Antiblemma leucocyma; (H) caterpillar and (I) adult of Carposina cardinata; (J) caterpillar and (K-L) damage caused by Pleuroprucha rudimentaria to inflorescences of Miconia calvescens. High quality figures are available online.

Journal of Insect Science | www.insectscience.org 15

Journal of Insect Science: Vol. 12 | Article 63 Morais et al.

Figure 2. Population dynamics of leaf feeders: (A) Salbia lotanalis in plants of Brazilian and Hawaiian biotypes, (B) Druentia inscita and (C) Antiblemma leucocyma in Viçosa, Dionísio and Guaraciaba, Minas Gerais, Brazil, between 2001-2006. Vertical bars = standard error. High quality figures are available online.

Figure 3. Population dynamics of reproductive organ feeders: (A) Carposina cardinata and and (B) Pleuroprucha rudimentaria in Viçosa, Dionísio and Guaraciaba, Minas Gerais, Brazil, between 2001-2006. Vertical bars = standard error. High quality figures are available online.

Journal of Insect Science | www.insectscience.org 16

Journal of Insect Science: Vol. 12 | Article 63 Morais et al.

Figure 4. Average air temperature (ºC), photoperiod (hours), wind speed (m/s), relative humidity (%), and total rainfall (mm) in Viçosa, Dionísio and Guaraciaba, Minas Gerais, Brazil, between 2001-2006. High quality figures are available online.

Journal of Insect Science | www.insectscience.org 17