The use of Polistes versicolor (Olivier, 1971) in the control of Ascia monuste orseis (Godart, 1819) in kale cultivation

Social wasps have been successfully utilized as means of biological control, particularly for the larvae of Lepidoptera. Ascia monuste orseis causes devastating effects on the kale, inducing production losses of up to 100%, therefore, requiring stringent control measures. This work aimed testing the species Polistes versicolor as potential means of biological control of A. monuste orseis, during autumn and winter, when the kale crop shows good development. These experiments were conducted at the IFMG Bambuí Campus, between May and June 2017, during which time four P. versicolor colonies were translocated to artificial shelters constructed in proximity to the kale crop, being registered daily the predation of the social wasp on like of A. monuste orseis caterpillars. The translocation of P. versicolor colonies onto the kale crop during the cold and drought seasons was shown to be ineffective in controlling the A. monuste orseis population. This was because this social wasp exhibited low foraging activity, and therefore a low degree of predation on the target pest; however, it became crucial to assess their activity during the hottest and most humid times of the year, as P. versicolor effectively preys upon the various species among the Lepidoptera.


Introduction
The social wasps (Hymenoptera: Vespidae) are usually included among the subfamilies Polistinae, Stenogastrinae and Vespinae, of which Polistinae alone comprises 343 species in Brazil (HER-MES et al., 2017).
Significant as natural predators of pest insects, these social wasps are effective even in small populations and make substantial contributions towards controlling and reducing the peak times of infestation (PICANÇO et al., 2010). Lepidopteran larvae are among the preferred prey of social wasps (PREZOTO; MACHADO, 1999;BICHARA-FILHO et al., 2009;FREITAS et al., 2015). Some species of polistine wasps, particularly those of the genus Polistes, have already been employed in biological control programs in Brazil (PREZOTO; MACHADO, 1999;MARQUES, 2005;ELISEI et al., 2010).
Ascia monuste orseis is one of the key pests of crops in the Neotropical region, mainly in Brazil (BITTENCOURT-RODRIGUES; ZUCOLOTO, 2009). As the caterpillars of this species devour the leaves, they cause great cultivation losses, which translate sometimes into as much as 100% production loss (VENDRAMIM; MARTINS, 1982).
Four or five days after the females of this pest oviposit the undersurface, most often, of the young leaves (BITTENCOURT-RODRIGUES; ZUCOLOTO, 2009), the caterpillars hatch out. During the first and second instars, they feed in the site where oviposition had occurred. However, the fourth and fifth instar caterpillars exhibit some degree of mobility and move to other leaves and sometimes even from one cultivar to another (BARROS-BELLANDA;ZUCOLOTO, 2003). The larval period extends for 20 to 25 days, after which the pupal stage lasts for a mean of 11 days (BITTENCOURT-RODRIGUES; ZUCOLOTO, 2009).
Ascia monuste orseis is mostly controlled by insecticide application, which can result in several problems, such as remaining as residues in food, mortality of natural enemies, poisoning to the applicators, and development of resistant pest populations. Thus, the use of agents of biological control is one method to minimizing the employment of these products. However, only one study is available, demonstrating the effectiveness of the social wasp Polybia ignobilis (Haliday, 1836) in the biological control of this pest (PICANÇO et al., 2010). Although there is no study of the specific relationship of the genus Polistes with this pest, it is recognized that wasps of this genus are excellent predators of agricultural pests, mainly Lepidoptera caterpillars (ELISEI et al., 2010).
The Polistes versicolor (OLIVIER, 1971) species in particular was selected as an agent of biological control for various reasons, as listed: a) availability of the predation records of this wasp on the lepidopteran caterpillars (MARQUES, 2005, ELISEI et al., 2010; b) as the dominant species in the IFMG -Bambuí Campus (JACQUES et al., 2015), which facilitates nest location; c) being a registered forage species on the kale culture (JACQUES et al., 2018); d) easy translocation of their nests (ELISEI et al., 2012), which enables them to accept the shelter; and e) wide distribution of this species across Brazil (BARBOSA et al., 2016), opening up opportunities to employ this methodology throughout the country.
The objective of this work, therefore, is to examine the effectiveness of the P. versicolor (Hymenoptera: Vespidae) found in the kale crop (B. oleracea var. acephala), on the biological control of A. monuste orseis (Lepidoptera: Pieridae), during the autumn and winter seasons, the time when this brassica crop develops well.
The experiment included two treatments: T1 -Kale planting with the P. versicolor wasp colonies and A. monuste orseis caterpillar infestation; T2 -Kale being planted without the wasp colonies and with A. monuste orseis caterpillar infestation. The second treatment was employed as a control, with which the mortality rate of the treatments with the wasp colonies and without them could be compared.
Each treatment was conducted in an area of 5.0 x 5.0 m, provided with 1.0 x 0,5 m spacing (FILGUEIRA, 2000), for a total of 5 rows containing 10 plants in each. The areas were 900 m apart, to ensure that the T1 treatment wasps did not consume the T2 treatment caterpillars. The longest return distance to the nest for the P. versicolor workers was established as 850 m (GOBBI, 1978).
In the T1 area, artificial wooden shelters, constructed using a 0.3 x 0.4 x 0.1 m sized board and attached to a deck of 1.8 m, were installed. In each shelter, a 0.13 x 0.17 x 0.11 m sized white plastic pot, open at the base alone, was installed, to offer protection against the rain and sun (ELISEI et al., 2012). The base of each batten, up to a height of around 0.50 m, was coated with burnt oil and grease, to protect against ant attacks (PREZOTO; MACHADO, 1999). Four such shelters were constructed at about one meter distance from the border of the crop.
Four P. versicolor colonies, with 32, 34, 21 and 14 wasps per nest, were identified and gathered from man-made constructions on campus, and, using the translocation method proposed by Elisei et al. (2012), the colonies were captured during the early hours of the night, when most of the wasps are in the nest (PREZOTO; MACHADO, 1999). After carefully bagging the colonies in a large plastic bag without destroying their peduncles, the wasps were separated from the nest. The wasps showed positive phototropism, moving immediately to the highest and brightest portion of the bag, facilitating their separation from the nest. Then, the colonies were immediately glued to the shelters by their peduncles, using cyanoacrylate ester glue or Super-Bonder® ( Figure 1). The plastic bag containing the individuals was wrapped in the overnight shelter of the translocation and removed the next morning after all the wasps had flown away from it. Once the nests were translocated, 30 numbers of the third and fourth instar larvae of A. monuste orseis were released at 7 o'clock in the morning, on five randomly selected kale plants, for both treatments T1 and T2, for fifteen days. In this experiment five plants corresponded to 10% of the total number of plants, demonstrating the degree of economic loss to the crop (PICANÇO et al., 2005). At 18 hours of the same day, the caterpillars were collected and the daily degree of predation was recorded for both treatments. As P. versicolor experiences a diurnal rhythm, the caterpillars were left undisturbed at night to minimize their death due to other factors. Daily observations were recorded between 11am and 1pm (and feeding was ad libitum), in order to confirm the cause of death of the A. monuste orseis caterpillars.
The A. monuste orseis were produced by first collecting the postures in the IFMG-Bambuí garden. They were packed in 500 ml pots and brought to a B.O.D. with the temperature controlled at 25 ± 2 ºC and a 12 h photoperiod. Fresh kale leaves were supplied daily to the caterpillars, until they reached the pupal stage. The pupae were transferred to a cage (2.5 x 2.5 x 2.5 m) in a greenhouse having 20 pots with kale seedlings for oviposition. The adults were fed with 10% honey and the new postures collected every day.
A comparison was made of the daily mortality rate between the two treatments and the means were compared using the T test at the 5% level of significance through the statistical program Past, v. 2.17c (HAMMER et al., 2005). The P. versicolor colonies were also closely tracked, and the number of individuals from the colonies was observed every day (PREZOTO; MACHADO, 1999). The colonies were then removed from the T1, 15 days after the experiment commenced and the predation each day was noted for another 15 days to record the effect of the treatment site on the final findings.
The findings recorded of the daily mortality rate were subjected to regression analyses, accounting for the variables of "daily temperature" and "relative air humidity", with p <0.05 using the BioEstat 5.3 program (AYRES et al., 2007). Relative temperatures and humidities were determined from the Bambuí weather station, situated on the IFMG-Bambuí Campus. Measurements were taken at nine and fifteen hours every day, and their means were utilized for the regression analyses.

Results and discussion
Treatment T1, involving the P. versicolor colonies, showed a value of 18.8% for the mean daily mortality rate of the A. monuste orseis. However, the T2 treatment, in the absence of the P. versicolor colonies, was statistically higher (p = 0.001), showing a mean value of 39.46% (FIGURE 2). Translocations of social wasp nests can cause stress in the colony, leading to abandonment of the nest. The success of the translocation technique is considered by the presence of the individuals in the colonies at least five days after the transfer (ELISEI et al., 2012), and this occurs in about 60 to 85% of the cases (BUTGNOL, 1992;MACHADO, 1999). In our study, the translocations of the colonies P. versicolor were completely successful and all four colonies remained unchanged throughout the experiment.
However, during the observation period, these colonies were only slightly active, with merely three predatory attacks by these wasps on A. monuste orseis being recorded. During the experiment, the mean temperature was 22°C and the mean relative humidity 62.2%. These values may have caused the decline in the foraging activity of P. versicolor, which is much greater under conditions of high light intensity and humidity and, particularly so at high air temperatures (ELISEI et al., 2010), decreasing between the months of May and September (ELISEI et al., 2013). The return of the P. versicolor workers to prey on the eucalyptus plantation during the cold and dry seasons is very minimal (ELISEI et al., 2010). Normally, the higher levels of temperatures, light intensity, humidity and air velocity favor the well being of the foraging of social wasps (DE CASTRO et al., 2011).
The foraging demand is likely stimulated by the climatic variables, as well as the biological needs of the colony (DETONI et al., 2015). When the experiment was concluded, the cells were empty, without eggs and larvae, indicating a probable abandonment of the colonies. Nests which support a decreasing number of immature ones, also experience a drop in the food collection. The prey supplies protein essential for offspring growth and development; therefore, the quantity of prey captured using fodder is an indirect estimation of the number of immature wasps and, thus, the protein demand of the colony (CANEVAZZI; NOLL, 2011).
The foraging behavior pattern evident in P. versicolor is not directed by a queen or any other individual exerting control, but arises only based on the colony requirements (DE SOUZA; PREZOTO, 2012a). Although they lack a queen, the workers carry on the foraging activity (DE SOUZA; PREZO-TO, 2012a). Aggressive behavior among the individuals induces exit foraging, reducing the number of workers as the needs of the colony decline PREZOTO, 2012b). Besides, the return of workers which exhibit foraging is the main stimulus which initiates the activities within the colony (DE SOUZA; PREZOTO, 2012a).
Carbohydrates and proteins are amenable to storage within the cells for consumption at a later date, and form a good store of reserve for the unfavorable seasons (MICHELUTTI et al., 2017); this results in a reduction in the foraging activities. Such storage abilities have been recorded earlier for the colonies of Polistes simillimus Zikán, Polybia paulista (Von Lhering), P. occidentalis, Mischocyttarus drewseni (Saussure) and Mischocyttarus cassununga (Von Ihering) (MICHELUTTI et al., 2017). However, foraging activities for water are also observed to decline during the cold and dry seasons, as water is principally utilized in the colony for keeping it cool, a finding earlier reported for P. versicolor (ELISEI et al., 2013).
The foraging behavior of P. versicolor for plant fiber is also at a minimum during the cold and dry seasons, as this resource is mainly required during the demographic explosion phase within the colony, to building and expand the nest cells (ELISEI et al., 2010), although such activities are not noticeable in the translocated nests. Foraging for vegetable fibers occurs when the worker uses the jaw to shave the vegetal substrate, including stems, trunks or other such materials. This process takes place more easily when the substrate is moist, and the plant fibers are soft (ELISEI et al., 2010).
During the experiment, we observed that 90 to 100% of the mortality of A. monuste orseis was found as a result of predation by another social wasp, Polybia ignobilis (Haliday, 1836), for both treatments. This wasp, recognized as the main predator of A. monuste orseis (Picanço et al., 2010), was found in large number in T2 treatment, which did not have the colonies of P. versicolor, especially during the initial stages of the experiment.
For the T1 treatment, the findings revealed a very low predation rate, particularly in the first five days (10.4%), which were statistically below (p = 0.031) the average recorded in the other ten days (25.0%). This was probably because of the necessity to separate the two areas (900 meters) in order to minimize the likelihood of the P. versicolor foraging in the T2 treatment. The sample area T2 was thus nearer to the horticulture section of the Campus, which supported a variety of cultures like cucurbitaceae, lettuce, tomato, sunflower and other brassicaceae, providing a more heterogeneous ecosystem, which positively affected the prey availability for P. ignobilis, with respect to the T1 sample area.
Comparing the two treatments in terms of the mortality rate, a similarity (p = 0.662) was noted between days 6 and 15 (FIGURE 3). This may have been due to decrease in the stress of translocated P. versicolor colonies. In addition, an increase in P. ignobilis activity and foraging evident in Treatment T1 was observed. Such an escalation in the P. ignobilis foraging activity might be linked to the more intense defoliation in the kale precipitated by A. monuste orseis, which most likely stimulated a higher degree of sinomonium secretion, which in turn attracted more numbers of these predatory wasps. When plants experience a herbivorous attack, they emit a wide variety of volatile substances in large quantities, induced by these herbivores (HIPVs). Their natural enemies use these secretions to find their hosts/prey. Such behavior has been recorded earlier for P. ignobilis which forages on Passiflora edulis Sims (RAW, 1998). Other species belonging to genus Polybia are also drawn to plants that have experienced herbivore attacks (SARAIVA et al., 2017). After being attracted to the T1 treatment area, characterized by abundant food supply, P. ignobilis began to forage frequently this area, exhibiting a behavior pattern reported earlier for this species (RAW, 1998). Social wasp workers search for food alone and opportunistically (MICHELUTTI et al., 2017), returning to hunt in places of success of the previous hunt and feeding several times of the same species of prey (BICHARA-FILHO et al., 2009).
For both treatments no regression analysis model was fitted to the daily data on the predation rate, taking into consideration the variables "daily temperature" and "relative humidity". This was most likely because of the low variations in the temperature and humidity during the 15 experimental days.
The two locations without the presence of the wasp colonies registered a similar daily predation rate (FIGURE 4), indicating that once the first few days of the experiment were over, the location of the treatment did not influence the results.

Conclusions
The translocation of the P. versicolor colonies onto the kale crop during the cold and drought periods was ineffective in controlling the A. monuste orseis population. This was because the forage activity of this social wasp was poor, therefore, predation on the target pest was also low.