Introduction

In Colombia, aromatic herbs have become an important option for agricultural production with potential for exportation and possibilities to compete on international markets such as in North America and the European Union (^{Fajardo and Sánchez 2007}). Eighty percent of the exported aromatic herbs are fresh (^{Forero 2005}). Among the species allowed to be imported into the USA, basil (Ocimum basilicum Linnaeus), mint (Mentha spicata Linnaeus), rosemary (Rosmarinus officinalis Linnaeus), and thyme (Thymus vulgaris Linnaeus) represent the major contingent (ibid.).

The insect pests that are present on these herbs belong to the four orders; Diptera, Hemiptera, Lepidoptera, and Thysanoptera (ibid.). Among these, the families Agromyzidae, Aleyrodidae, Aphididae, Noctuidae, and Thripidae are regularly intercepted by USDA-APHIS-PPQ (Lenis 2007 cited by ^{Patiño 2014}). Noctuid moths of the genus Copitarsia, subfamily Cuculliinae, are frequently intercepted at US border ports on produce and cut flowers from Mexico, Central America, and South America (^{Venette and Gould 2006}). The members of Copitarsia spp. are considered a quarantine pest as these cause significant losses. Especially, Copitarsia spp. spreads via commodities contained with eggs or larva. Majorities have been originated in Colombia followed by Mexico and Ecuador (ibid.).

The genus Copitarsia spp. comprises 21 species. Three of those species are grouped in the Copitarsia complex turbata (Herrich-Schaeffer); the other 18 species belong to the Copitarsia complex naenoides (Butler) (^{Angulo and Olivares 2003}). A very cryptic feature defines the difference between these two groups: the uncus vertex. The Copitarsia complex turbata is characterized by an uncus vertex present in a dorsally and in a plain manner, without overlapped or noticeable structure. The Copitarsia complex naenoides has two longitudinal plaques with recurved indentations (ibid.). ^{Simmons and Pogue (2004)} redescribed C. decolaradecolara Guenée giving C. turbata as synonym.

Copitarsia spp. occur along the western edge of South and Central America from the southern tip of Argentina through central Mexico attacking a wide range of crops, such as Allium, Alstroemeria, Brassica, Mentha, Rosa, Asparagus, strawberry, spinach, and many others (^{Castillo and Angulo 1991}; ^{Angulo and Olivares 2003}; ^{Gould et al. 2013}).

For pest insects, locating suitable host plants is a formidable task requiring a sophisticated detection mechanism. The host selection process can be seen as a continuum between two extremes, namely, insects ‘choosing’ their host from a distance using olfactory and visual cues, and insects ‘selecting’ their host only after contact, when gustatory cues are also employed (^{Bruce et al. 2005}). Host selection is crucial for the survival of herbivorous insects, and this decision is often based on the suitability of a plant for feeding and oviposition (ibid.). This selection goes together with host plant quality, which depends on several factors including the nutritional value (^{Scheirs et al. 2003}), defensive chemicals (^{Dicke and van Poecke 2002}), physical barriers, such as trichome distribution and density (Kennedy 2003), and associated natural enemies (^{Dicke and van Loon 2000}; ^{Scheirs and de Bruyn 2002}). Odors from host plants play a vital role in the attraction of insect herbivores (^{Guerin and Visser 1980}) allowing herbivores locating a host plant from a distance (^{Bernays and Chapman 1994}). Plant volatiles are detected by specifically tuned and highly sensitive olfactory receptor neurons (^{Bruce and Pickett 2011}).

These allelochemical interactions between Lepidoptera and plants have been extensively studied. ^{Pavela (2011)} pointed out the importance of allelochemicals in the defense system of plants against herbivores. These compounds can be toxic, repellent, attractive or nutritious to insects. They play an important role as stimulants or inhibitors of alimentation, as part of host plant resistance, or as communication vehicle between competitors, parasitoids and predators (^{Chapman 1994}). Plants liberate and produce many volatiles, specific for each plant and each insect (^{Torres 2009}). ^{Hilkerand Meiners (2002)} claimed that these compounds are basic for host recognition and acceptation by insects.

In order to understand better host preferences, various workers have conducted direct and indirect preference tests. ^{Bado et al. (2001)} offered three Solanaceae species to larvae of the chrysomelid Lema bilineata (Germar) in free choice and non-choice tests. The authors measured the consumed leaf area as direct preference tests and the time insects stayed on the different food types as indirect preference tests.

Olfactory experiments complete the above mentioned direct and indirect preference tests. They are mostly used for studies on the perception of plant emitted volatile compounds which insects use for orientation (^{Chapman 1994}). ^{McIndoo (1926)} developed one of the earliest olfactometers. He released beetles in Y-glass-tubes testing the attractiveness of odours from different host plants. These type “Y” olfactometers are still the most used approach where the insects have the option to choose between two different odours. Later, ^{Pettersson (1970)} designed a four-armed olfactometer. ^{Vet et al. (1983)} constructed an olfactometer where insects are introduced into a chamber connected to four arms with their respective odour source. The airflow in each arm can be regulated separately avoiding air turbulences as it happens at the junctions in Y-tubes when mixing the offered odours. The Y-tube olfactometer is used for measuring chemotactic or olfactive responses by insects within short distances or in situations where chemical signals are diffused into space without being transported by an airstream (^{López-Avila and Rincón 2006}). There is also an olfactometer designed as a tunnel for insects that have excellent flight capacities towards different odour sources. However, their disadvantage is that only one odour can be tested at the same time (ibid.).

This study thus evaluated the olfactory response and feeding preferences bioassays of C. uncilata on four aromatic herbs, basil, mint, rosemary, and thyme.

Materials and methods

The experiments were conducted in the Faculty of Agricultural Science of the National University in Bogotá (2.640 masl), Colombia, during 2012 and 2013. Laboratory maintained larvae of C. uncilata were used in this study. In order to rear C. uncilata in our laboratory, we recollected eggs, larvae, and pupae at the installations of the Alstroemeria aurea grower “Flores de Conchita Ltd.” in Bojacá (Cundinamarca, Colombia). The colonies were started with one couple of male and female adults. The female laid 350 viable eggs. These eggs hatched after three or four days; the larvae passed through five instars. All larvae of C. uncilata were taken from the colonies, which were maintained at 12:12-hours light/dark (natural photoperiod) at 21.1 ± 1,7 °C and 50.7 ± 7.5 % relative humidity according to the average of the environmental conditions in the laboratory, which were similar to those where the insects and plants were recollected. Larvae were individualized in plastic cups (diameter top/bottom: 3.52/2.22 inches, height: 4.3 inches16 ounces) in order to avoid cannibalism and fed with buds of Peruvian lily (Alstroemeria aurea Linnaeus). After 25-35 days the larvae pupated and were separated according to their sex following the descriptions by ^{Pogue and Simmons (2008)}: as in most insects the openings of the female genitals are located in the eighth and ninth segment, those of the male only in the ninth. The separated pupae were confined in new cups of the same size as mentioned before, filled up to a quarter with vermiculite or sawdust. After 6 to 10 days they developed into adults. These were paired and each pair was confined in a separate crystal bonbonniere. The adults were fed with honey, water, and sucrose solution (1:5:1 ratio) until they died within six to eight days. The complete life cycle oscillated between 45 and 52 days. The colonies were held at 21.1 ± 1.7 °C, 50.7 ± 7.5 % RH, and 12D:12L-hours (natural photoperiod). The age of the harvested herbs was as follows: basil (“Nufar”) at 90-120 days, mint (“Hierbabuena”) at 90 days, rosemary (“Romero común”) at 180 days, thyme (“Tomillo limonero”) at 90 days.

Olfactory response experiment

It was designed and elaborated an olfactometer with four arms based on ^{López-Ávila and Rincón (2006)} and ^{Pacheco et al. (2012)}. A pump equipped with an activated charcoal filter generated the airflow that entered the olfactometer after having passed a water trap in order to avoid the entrance of odours into the system. Then, the air reached the recipients filled with 0.5 ml essential oil (EO). From there the air diffused the aromatic odours through 6/8 inches’ tubes to the hermetical camera where six gravid females were confined (Fig. 1). The females could fly or walk towards the exit of the emanations. The olfactive perception of C. uncilata was assessed using the Randomized Complete Block Design (RCBD) with four treatments (the aromatic host plants) and nine repetitions (^{GLM procedure of SAS 2009}). The natural EO’s were acquired from Green Andina Colombia Ltda. They were processed without any organic solvents in order to obtain more reliable results.

The variable “number of females” was computed based on the studies of ^{Cerda et al. (1996)}, ^{Narayandas et al. (2006)} and ^{Saïd et al. (2006)}: The orientation of the gravid females was determined by counting the females that headed to one of the olfactometer arms with its specific odour. The experimental unit was revised during 10 minutes every 60 minutes during 12 hours (8 am to 8 pm). Based on these data we calculated the relative distribution of females by: (number of gravid females in one arm / all gravid females) x 100. Additionally, we determined the time that females remained in the olfactometer arms.

Figure 1 Overall illustration of design and process of olfactometer bioassay. 

Choice experiment

In the “preference” or “choice-test” the frequency of associations between larvae and the aromatic leaves was evaluated. Before starting the experiment larvae were starved as an adaptation for two hours. Then, each larva was placed in the centre of a filter paper disc (Whatman Grade 1, ø 10 cm) in a Petri dish and surrounded by one leaf (20 g) of each aromatic plant (Fig. 2). After this adaptation period a two hours observation was started and recorded the leaves the larva was associated with. The experiment was repeated for 30 times. We considered a larva present on a leaf as 1, absent as 0. These data of the indirect test were basis for the construction of the relative distribution as ^{Bado et al. (2001)} suggested: (Number of individuals associated with host 1, or 2, or 3, or 4 / total number of larvae on all host plants) x 100.

Figure 2 Arrangement of experimental units. 

In both experiments, 3^rd instar larvae were used because at this stage their body length and cephalic capsule width grow faster than in any other stage (Fig. 3) (^{Moreno and Serna 2006}). Hence, we assumed that this stage makes it easier to find a correlation between ingestion quantity and body growth. We considered a larva as third instar when the diameter of the cephalic capsule reached 0.84 mm (ibid.). was used the capsule width by means of the software ImageJ 1.46 (2012) and using a vernier scale as base.

Figure 3 Larval instars 2, 3, 4, and 5 of Copitarsia uncilata. This picture may explain why the third instar has the highest proportional weight increase. 

Results and discussion

The preferences are presented as percentage measuring the time that larvae remained on a leaf. There was a clear preference of the C. uncilata larvae to stay longer on mint or basil than on rosemary or thyme (Fig. 4). Time of permanency in association with thyme and rosemary was almost the same (F_2.7 = 87; P < 0.05, Scheffé-test).

These results are corroborated by our observation that leaves of mint and basil were more consumed by the larvae than thyme and rosemary (Fig. 5). ^{Moon et al. (2011}) reported that mint oils attract Lycorma delicatula (White) (Hemiptera: Fulgoridae). This stands in contrast to ^{Regnault-Roger et al. (2012)} who stated that the great majority of reviews clearly comment that most aromatic plant species have a repellent and deterrent effect on insects. However, the chemical composition of mint, particularly the presence of the terpenoid carvone, creates an attraction to insects that is superior to others compounds, which can be explained by its predominance (89.7 %) in its EO’s (^{Moon et al. 2011}). ^{Cohen (2005}) argued that the compound, which is a product of the secondary plant metabolism, permits the so-called token-stimuli. This basically stimulates feeding without being involved in the nutritional process. This might be considered as an explanation why C. uncilata larvae did prefer mint.

Figure 4 Permanence test of Copitarsia uncilata (3rd instar) versus four aromatic plants (basil, thyme, mint, and rosemary). Bars represent standard error (SE) or standard deviation (SD). Same letter indicates no significant differences between means (F_2.7 = 87; P ≤ 0.05; Scheffé test). 

Figure 5 Experimental unit of choice test. Larvae of Copitarsia uncilata didn’t cause any damage to rosemary and thyme, whereas basil and mint leaves showed clear feeding symptoms. 

^{Ferreira and Moore (2011)}, as well as ^{Moon et al. (2011)}, described that rosemary and thyme have a well-known repellent effect on Culicidae and Cicadidae, respectively. For this reason, we assumed that most C. uncilata larvae didn’t feed on those two aromatics either.

Regarding preferred daytime larvae showed a tendency to remain longer associated with mint after midday (Fig. 6). This may be explained by the increasing temperature. ^{Scriber and Slansky (1981)} mentioned that food quality (in terms of water and dry weight) could be affected by temperature. This was corroborated by our observation that basil loses more liquid than mint indicating that latter conserved water with increasing temperatures around midday, which did not occur with basil although the leaves laid on a moist filter paper in a sealed Petri dish.

The weight of larvae increased more when they fed on mint or basil in contrast to thyme and rosemary, where their weight even slightly decreased (F_2.7 = 87; P ≤ 0.05; Scheffé test) (Fig. 7). The treatments basil and mint were not significantly different, as well as the means of rosemary and thyme. The weight increase correlated with preference (r = 0.772; P < 0.0001).

Figure 6 Time of permanency of larvae (3rd instar) of Copitarsia uncilata on four host plants. Data are presented as percentage of the whole time of permanency during 12 hours. 

Figure 7 Weight increase of larvae feeding on four aromatic hosts, respectively, during 12 hours. Treatments with the same letter are not significantly different (F_2.7 = 87; P ≤ 0.05; Scheffé-test). 

Olfactory test

Figure 8 shows the preference of gravid females for the four offered EO’s in the olfactometer. We didn’t observe any significant differences between basil and rosemary; however, there was a clear difference between mint and thyme. Mint was significantly less attractive than basil and even less than rosemary, and but was more accepted than thyme. Although the difference between mint and thyme was not significant the result suggested the repellent character of thyme.

Figure 8 Response of essential oils by gravid females of Copitarsia uncilata. Treatments with the same letter are not significantly different (P < 0.005; Tukey-test). 

This behavior was not expected since the larvae showed a clear preference for basil and mint. The preference-performance hypothesis predicts that oviposition preference should correlate with host suitability for offspring development (=survival off- spring and later adult potential fecundity) because females are assumed to maximize their fitness by ovipositing on high quality hosts (^{Scheirsand de Bruyn 2002}; ^{Schowalter 2006}). However, some studies reported strong, positive correlations between host preference and offspring performance, many studies found poor correlations (^{Scheirs and de Bruyn 2002}). ^{Petit et al. (2015}) came to a similar conclusion, although they argued in a different way: Host selection may differ according to the host range given that poly-, oligo-, and monophagous species use different host selection strategies. For example, oligophagous and monophagous species are more efficient when making decision for host selection than polyphagous species (^{Bernays 2001}). This tendency might have affected the females of the polyphagous Copitarsia genus by their selection of volatiles. ^{Carrasco et al. (2015)} hypothesized that innate plant preferences in combination with the current environmental conditions would regulate the final oviposition choice. Nevertheless, the preference hierarchy is a dynamic one, since it can be modified, for instance, by previous experiences, especially in oligophagous and polyphagous species.

^{Regnault-Roger et al. (2012)} concluded that the efficacy of ethereal oils and their constituents varies according to the phytochemical profile of the plant extract and the entomological target. The bruchid A. obtectus is more sensitive to phenolic monoterpenes and the aphid Rhopalosiphum padi (Linnaeus) to methoxylated monoterpenes, whereas Ceratitis capitata (Wiedemann) flies respond to both types of compounds. Ethereal oils such those extracted from of thyme, rosemary, and Eucalyptus (L’Heritier) have antifeedant or repellent activity. The oil of citronella (Cymbopogon nardus, (Linnaeus) Rendle) repels mosquitoes and flies. Garlic (Allium sativum Linnaeus) oil is a deterrent to many insect herbivores. These are currently marketed to horticulturists, greenhouses, and home gardens in the United States and the United Kingdom (ibid.). ^{Scriber and Slansky (1981)} claimed that insect behavior is related to feeding habits and other characteristics, and influenced by a wide range of biochemical variables, morphological mechanisms, the host, and the pest. These authors also mentioned the repellent and antifeedant action of the EO’s of thyme and rosemary that exactly corroborated our results.

Conclusions

Future studies on the establishment of cultural practices focusing on repellent products based on thyme and rosemary as a strategy against Copitarsia spp. are required. It should the aim of these future investigations to include factors like plant material, insect’s developmental stage (larval instar, gravid or not gravid females), olfactometer design (number of arms, hermetical camera, air pump), and disposition of treatments (time and space) for the search for reliable answers in order to acquire more knowledge about ethology of this pest insect. Moreover, studies on larval development and longevity of adults on different herbs should be continued.