Advances in catching fruit butterflies for quick inventory

The family Nymphalidae comprises frugivorous butterflies that are frequently used as bioindicators in environmental assessment studies. The collection of these lepidopterans is easily made with traps filled with fruit baits. The most frequently used bait is a mixture of banana and sugarcane juice, though other mixtures are also attractive. Hence, the objective of the present study was to assess the attractiveness and aging effect of baits used to capture fruit-feeding butterflies. All tested baits contained fructose, glucose, and sucrose but it was not possible to determine an optimal capture period in terms of bait aging, as sugar concentrations varied throughout the fermentation process.


INTRODUCTION
Fruit-feeding butterflies are frequently used as bioindicators of the environmental quality, as there is a relation between total species richness and changes in physical factors of the habitat, and because they also present a close relationship with plants [1,2,3]. In addition, these insects are easy to visualize, manipulate and identify [4]. Although long diversity studies are essential for knowledge of the fauna, they sometimes become unfeasible. For example, monitoring in hard-toreach areas is almost impossible on a weekly or even monthly basis. Rapid biological evaluations have alternatively been used in entomofauna surveys.
Among the methods for sampling lepidopterans, such as active (insect nets) or passive collection (traps) [5], we highlight Van Someren-Rydon attractive traps, which are used to collect fruitfeeding butterflies [6,7,8] in several biomes [9,10,11]. Some advantages of collecting butterflies with attractive baits that justify their broad use in species inventories are their feasibility and low cost of capture, which are easily accomplished with fruit baits [12]. The most used bait is a mixture of banana and sugarcane juice [10,11]. Other blends, however, can also be attractive: other fruits, like pineapple, or even fish, shrimp and feces [13,14,15].
Aiming at optimizing practical and reliable methods for assessing changes in biodiversity and ecosystem function, the study evaluated the attractiveness of different fruit baits in the field to capture frugivorous butterflies. It was also evaluated the composition of the bait on the types of sugar in the laboratory.

MATERIAL AND METHODS
We carried out the present study in the Botanical Garden of the Federal University of Juiz de Fora (JB-UFJF), located in the municipality of Juiz de Fora, southeastern Minas Gerais State, at 750 m a.s.l. in an area of 84 ha (21°43'28"S; 43°16'47"W) [16]. The climate is warm subtropical with dry winter (Cwa), according to the Köppen-Geiger system [17]. The study was carried out from July 2013 to February 2014, in a total of six sampling events: three in the dry season (July, August and September) and three in the rainy season (October, January and February).
To sample the insects, we used nine modified Van Someren-Rydon traps. Based on the methods commonly used in inventories of fruit-feeding butterflies [10,13,18], we tested the following baits: pineapple with sugarcane juice (P&S), banana with sugarcane juice (B&S), and banana with water (B&W). All baits were prepared as juice, in a blender, at the proportion of three parts of mature fruit (300 g) to one part of liquid (100 ml) and were stored in plastic containers and aged for 48h at room temperature. We set up the traps at approximately 1.5 m above ground, and divided them in three sets. Each set comprised three traps, and each trap contained 200 ml of a single type of bait. The traps were located 25 m apart from each other. The trap sets were 225 m apart from each other and remained in the field for 48 h during each sampling event, which makes a total sampling effort of 432 h.
Small nymphalids were killed by thoracic compression and large ones were killed with an ethanol injection in the thorax, following Duarte et al. (2012) [5]. Specimens were identified using published keys [6,13,18] and through comparison with specimens of the entomological collection of the Natural History Museum of the Academy of Commerce and of the Center of Higher Education of Juiz de Fora. The collected material was deposited in the Laboratory of Behavioral Ecology and Bioacoustics (LABEC) at UFJF; some individuals were dry-mounted to make a butterfly display case.
To estimate the sugar content of the baits, we measured reducing sugars (fructose and glucose) using the 3,5-dinitrosalicylic acid method (DNS). We analyzed baits after 48, 72, and 96 h of fermentation. For building the standard curve of glucose and fructose we prepared solutions of both sugars at concentrations of 20, 40, 60, 80, and 100 mg/ml. Reading was made in a spectrophotometer with absorbance at 540 nm.
For paper chromatography, 4 µl of standard solution (glucose and fructose) and of each sample at a concentration of 1 mg/ml of carbohydrate (prepared by dilution based on the results of the DNS method) were applied on Whatman paper n.1 and submitted to descending chromatography in isobutyric acid: NH4OH 1.25 M (5:3 v/v). After 18-24 h the chromatogram was dried in an oven with heating and air circulation. Reducing products were revealed with silver nitrate in alkaline medium.
To test for variations in abundance and richness of nymphalids captured with different baits, we used a Kruskal-Wallis test using BioEstat 5.3, and for the dry and rainy periods we used Student's T-Test using Past 3.08.
Although they recorded exclusive species, there was no statistical difference between baits in relation to abundance (B&W There was also no significant difference between the dry and rainy periods in relation to abundance recorded by the different baits (B&W: t = 0.4756, p = 0.3181; B&S: t = 1.5133, p = 0.0687; P&S: t = 0.3656, p = 0.3581) and richness (B&W: t = 0.0000, p = 0.5000; B&S: t = -0.5317, p = 0.2985; P&S: t = -0.2628, p = 0.3969). All the baits contain fructose, glucose and sucrose according to the result obtained by the DNS method and confirmed by paper chromatography. Regarding to the concentration of reducing sugar was observed that P&S and B&S behaved as expected, that is, the concentration of reducing sugars increased with the progress of the fermentation process ( Table 2). The water present in the bait B&W may have reduced the fermentation process, so the concentration of reducing sugars remained low (Table 2) [15], suggest that lepidopterans are attracted, not by the sugar, but by substances present in baits or released by fermentation (volatile).

CONCLUSION
In this study, the use of attractive traps proved to offer a useful way to monitor changes in species abundance over time, compare species composition and abundance between sites, and track individual movement. For a quick inventory, where the goal is to record the greatest diversity in the shortest time, we recommend a set of all three baits, at any time of year, as each one captures a different set of species and, if possible, a consortium with an active collection method for sampling be maximized.

ACKNOWLEDGMENTS
This work was supported by the Universidade Federal de Juiz de Fora (UFJF), the University of Otago, the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), the Conselho de Desenvolvimento Cientifico e Tecnológico (CNPq), and the Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG).