Bioefficacy of Rhynocoris longifrons (Stål) (Heteroptera: Reduviidae) against multiple cotton pests under screen house and field conditions

Rhynocoris longifrons (Hemiptera: Reduviidae) is a generalist predator of many cotton insect pests. The hiding behaviour of this predator, which is one of the key factors of predation success, was investigated under screen house conditions. Moreover, we evaluated its biocontrol potential against Aphis gossypii (Hemiptera: Aphididae), Dysdercus cingulatus (Hemiptera: Pyrrhocoridae), Phenacoccus solenopsis (Hemiptera: Pseudococcidae), and Helicoverpa armigera Hübner (Lepidoptera: Noctuidae) under screen house and field conditions. Results showed that R. longifrons life stages preferred to hide under small pebbles in the screen house tests. All the R. longifrons life stages showed a biocontrol potential against the four insect pests under screen house conditions. However, their biocontrol potential had not varied in relation to day and night hours. Augmentative releases of R. longifrons were carried out for two seasons such as South-west monsoon, 2011 and post-monsoon, 2012. The augmentative release of R. longifrons reduced significantly insect pests on cotton. In fact, the release of this predator in cotton fields was capable to reduce the population of H. armigera (50%), P. solenopsis (28%), D. cingulatus (18.8%), and A. gossypii (11.8%) during the rain fed condition (south-west monsoon season). During irrigated condition (post-monsoon season), populations of D. cingulatus were reduced by 26%, than P. solenopsis (20.6%), and A. gossypii (16.8%). Except ants, no negative impact was reported on other natural enemies present in the cotton field. Significantly higher crop yield and cost benefit ratio was observed in the predator release plots indicating that R. longifrons can be used in an integrated pest management program for multiple cotton pests.

Many reduviid predators possess either morphological adaptive characters 25 or behavioural adaptive features 26 including hiding 27 to successfully capture and feed the preys' . Shelter provisioning with pieces of clay pots and stones in cotton field enhanced reduviids population and increased cotton production 21 . To date, specific information on the predatory behaviour and biological control potential of R. longifrons against any pests under field conditions has not been widely reported. The predator can exhibit a variety of necessary adaptations for predation as well as survival 28 . One such important adaptation is a hiding behavior, either to escape from natural enemies and/or to find shelter. This hiding behavior, referred to as anti-predator behaviour 29 has not been studied with reduviid predators. Moreover, Ambrose 30 reported that reduviid predators of the subfamily Ectrichodiinae are diurnal, whereas Peiratinae and Emesinae are nocturnal. To date, however, no systematic study has been made of the feeding potential of these predatory R. longifrons in relation to day and night hours. Therefore, we conducted a series of studies to assess their possible integration in an augmentative biological control program for cotton pests in Tamil Nadu, India. The specific objectives of this study were to: 1) understand R. longifrons hiding behaviour under open field conditions, 2) evaluate their biocontrol potential in relation to day and night hours, 3) test their efficacy in an augmentative release program under field conditions for two seasons, and 4) estimate the cost benefit ratio and percent avoidable loss.
Cotton production, Cost Benefit Ratio (CBR) and Percent avoidable loss. Cotton production was higher in the predator-release plots (837.0 and 753.4 kg /Hectare −1 for south-west monsoon and post-monsoon seasons respectively). Similarly, the cost benefit ratio was higher in the predator-release treatment (1:1.28) than in the control treatment (1:1.17) during the south-west monsoon season, as well as post-monsoon ( Table 5). Cost of cultivation was less in the control [= 4872.0 (73,08 US Dollar) and 26652.0 (399,78 US Dollar) for South-west monsoon and post-monsoon seasons respectively)] compared with the predator-released field [= 29096.0 (436,44 US Dollar) and 28134.0 (422,01 US Dollar), for South-west monsoon and post-monsoon seasons respectively)]. A higher PAL was recorded in the South-west monsoon season (14.5%) than in the post-monsoon season (4.9%) ( Table 5).

Discussion
Reduviids are distributed in cotton agro-ecosystems worldwide and considered as important biocontrol agents 10 . Hence, use of reduviid predators to manage cotton pests can be one of the effective components of IPM programme, thus basic information is needed about hiding behavior, predatory potential, and other factors. The present study clearly shows that R. longifrons hide under different objects and, the preference for an object is function of the time of the day. The reduviid predator Rhynocoris fuscipes (Fabricius) also showed similar hiding behavior 31 . This hiding behavior is an advantageous in the framework of an IPM because it allows this predator to better surprise the prey it captures. Moreover, it allows a better escaping from their natural enemies such spiders and wasps [31][32][33] . Therefore, we can conclude that R. longifrons nymphs and adults can survive and adapt in the cotton agro-ecosystem after its release.
Our results showed that, feeding potential of R. longifrons life stages on P. solenopsis and A. gossypii adults, D. cingulatus nymphs, and H. armigera larvae had not varied in relation to day and night hours. This result was surprising because some authors suggested that reduviid predators are either nocturnal or diurnal 34 . For example, Lira et al. 35 found immature Harpactorini assassin bug feeding on a scorpion during the night time only. In contrast, Sycanus sp. and Scipinia sp. life stages were more active in the morning (6.30 to 10.0 a.m.) 36 . Our results imply that R. longifrons life stages can be released at any time of the day and we recommend the release of the predator third-stadium for A. gossypii control in screen house.
The predominance of A. gossypii and P. solenopsis populations in cotton fields respectively during south-west and post-monsoon seasons are not surprising. In fact, it is known that the population dynamics of aphids and mealybugs can be affected by seasonal changes in weather conditions 37 . Moreover, Wang et al. 38 showed that high rainfall was unfavourable for serious infestation by P. solenopsis. Our study showed that, augmentative releases of R. longifrons significantly reduced the number of insect pests in the cotton fields. The first release of this predator in cotton fields was capable to reduce A. gossypii and P. solenopsis populations during south-west and post monsoon compared with the control. Similarly, the field release of some reduviid predators such as Pristhesancus plagipennis Walker 17,39 , Platymeris laevicollis Distant 40 and Rhynocoris marginatus (Fab.) 19,20 were successful in reducing various pests in coconut palm, cotton and groundnut fields. Our results showed that R. longifrons can be an effective predator of multiple preys, which is consistent with previous results for generalist predators 9,41,42 .
Further, no adverse interaction between R. longifrons and indigenous predators like coccinellids, spiders and, other reduviids were observed. This situation suggests reduviid predators do not interfere with other natural enemies. Consequently, reduviid predator may be an ecofriendly protection for cotton pest populations and are likely to be highly detrimental to their biological control. Reduviids can be utilized as part of a multiple species release program as suggested for Dephastus catalinae (Horn) (Coleoptera: Coccinellidae) 43 . However, Symondson et al. 44 reported that at natural ecosystems, change of biotic and abiotic variables is common and that under such circumstance, it was difficult to predict the interspecific and intraspecific interactions between released generalist predators, and the predators dwelled in the eco-system. The decrease of ants in R. longifrons treated plots could be explained among other factors, by the decrease of aphids and mealybugs populations, which bribe them with their honeydew excretion. In fact, it is known that P. solenopsis 45   www.nature.com/scientificreports www.nature.com/scientificreports/ Comparatively of the cotton production (1585 kg/ha) at Tamil Nadu of India, the yield recorded during our study in cotton fields during south-west monsoon (2068 kg/ha) and post-monsoon season (1862 kg/ha) reveal that release of reduviids reduced impact of insect pests thus enhance cotton production. Similar trends were also observed when P. plagipennis was release in cotton fields 17,39 . Although the cost benefit ratio was low (1:1.28 or 1:1.24) comparatively to entomopathogenic insecticides such as HaNPV (1:3.50) or Bt (1:1.23) used for cotton protection 47,48 , it had been higher than those of control highlighting the benefit of integrating this predator in cotton pest management. Furthermore, the small-scale mass production technology for R. longifrons developed by Sahayaraj and Ravi 49 is available for an augmentative biological control program. In view of these encouraging results, it would be interesting in the framework of a biocontrol of cotton pests program to integrate this predator in other agro-climatic zones in Tamil Nadu as well as in other states in India.

Conclusion
Based on the observed hiding behaviour of R. longifrons, we conclude that this predator can survive and adapt to the cotton agro-ecosystem after its release. In the framework of an IPM program, R. longifrons life stages can be releases at any time of the day. Augmentative releases of this predator reduced significantly the number of insect pests in the cotton fields. R. longifrons increased the cotton production and enhanced the cost benefit ratio increased in predator released cotton field. It is concluded that reduviid predators can be integrated into cotton integrated pest management program.
Adults and immatures of the A. gossypii, P. solenopsis, D. cingulatus, and eggs and larvae of H. armigera were collected from cotton and maintained on their natural host plant under the above mentioned laboratory conditions. Life stages of D. cingulatus were maintained on wet cotton seeds. Both A. gossypii and P. solenopsis life stages were maintained on cotton seedlings (5-10 days old) which were established in small plastic tubs (10 cm height × 8 cm upper diameter × 6 cm lower diameter). On each seedling 40-50 D. cingulatus nymphs or adults were accommodated. H. armigera larvae were maintained individually in plastic containers (8 × 6.5 cm) with a mixture of healthy cotton flowers, flower buds and young leaves.
Hiding behaviour of the predator inside the screen houses. To record the hiding places of the predator when it was released in cotton during an augmentative pest management programme, the methodology described by Tomson et al. 31    www.nature.com/scientificreports www.nature.com/scientificreports/ instar predators (previously starved for 24 hr) were released at 6.00 a.m., 8.30 a.m., 11.00 a.m. and 3.30 p.m in different pots. The reduviids were introduced individually by means of a camel's hair brush into the northeast corner of the base of the plant. Two hours after every release, the percentage of insects that settled at various places [cotton plant, fallen leaves, small pebbles in the plant and other objects (soil balls, weed plants)] were recorded. Similar procedure was followed for the fifth instars and adult predators.
Biocontrol potential of R. longifrons in relation to day and night hours. Modified methodology of Tomson et al. 31 was followed for bioefficacy experiments. For the experiments, cotton seeds of MCU-5 variety were sown in cement pots (36 × 30 × 22 cm) and maintained in screen house at St. Xavier's College, Palayamkottai. The bioefficacy of R. longifrons life stages (third, fourth and fifth instar nymphs and adults) was tested against preferred life stages of A. gossypii (4 adults/predator), D. cingulatus (second and third nymphal instars, 2 of each instar/predator), H. armigera (second and third instar larva, 2 of each instar/predator) and P. solenopsis (4 adults/predator). The predatory potential of R. longifrons was evaluated between 6 a.m. and 6 p.m. (day experiments) and between 6 p.m. and 6 a.m. (night experiments). pests were released on the 25-30-day old cotton plants covered by a nylon net at 6 a.m. for day or 6 p.m for night experiments respectively. One hour after pests release, 24-hour pre-starved R. longifrons (third, fourth and fifth instar nymphs, and adults -2/plant) were introduced separately on the infested cotton plants. Six replications were used for each pest and predator life stage for both experiments. Three replications without predator were established as control. After 12 h, the number of live and dead preys in each plant was counted and predation rate of the R. longifrons life stages was calculated 50 . Experimental design for the augmentative biocontrol potential evaluations. The augmentative biocontrol potential was evaluated during rain fed condition (South-west monsoon) 2011 (July to September) in the cotton field at Kothankulam (9.4692° N, 77.6046° E), in the Virudhunagar district, Tamil Nadu and also in irrigated condition (post-monsoon season) 2011-2012 (December to February) at K. Duraisamiyapuram (E 77° 35°, N 22.16°), in the Tuticorin district of Tamil Nadu. The methodology described by Tomson et al. 31 with slight modifications was used for these evaluations. For weed control and fertilization farmers' standard cultural practices were used to grow cotton (SVPR4 cultivar). Two treatments were evaluated, a cotton field into which R. longifrons was released and a control field, free of pesticides and predators. Unplanted buffer zone (2 m) was established between treatment and control plots. The treatment plots were arranged in randomized complete block design with five replications (10 plots total with a size of 10 m × 5 m each). To identify the predominant pests and predators in experimental cotton fields, we examined in each field 8-10 leaves of 10-randomly selected plants from two days before predators release. R. longifrons nymphal instars (50 individuals each) were released individually during 6-8 a.m. in each experimental cotton fields 40, 55 and 70 days after the seedling emergence (ASE). In addition, three cards (3 × 3 cm) containing each 50 eggs of R. longifrons were also placed onto the twigs of a plant in each cotton field. In total, a mixture of 900 R. longifrons (150 eggs, 250 nymphal instars × 3 releases = 900/plot) were released in each plot. Three days after the predators' release, 10 randomly selected cotton plants in each experimental cotton field were visually examined for the presence of A. gossypii, D. cingulatus, P. solenopsis and H. armigera. The number of each species of predators encountered was also recorded per plant.
Cost Benefit Ratio (CBR) and Percent Avoidable loss. At the completion of the growing seasons, the cotton in each plot was harvested. It was then cleaned, weighed and sold in the local market. The cost benefit ratio (CBR) was calculated based on the income per hectare thus generated. In addition, the percent of avoidable loss (PAL) was also calculated according to the method proposed by Krishnaiah 51 . Statistical analysis. Before variance analysis, the data normality was tested using Levene's test. Data on the percentage of predator hiding under various objects in relation to life stages was subjected to arcsine transformation, while the total number of prey consumed by a predator in 24 h, numbers of different insect pests and natural enemies present at each release (first, second and third), as well as their total mean populations were log-transformed to homogenize the variances before being subjected to variance analysis using IBM SPSS Statistics version 25 software package. Significant differences between the means were separated using the Student Newman Keuls test (p ≤ 0.05). Original means are presented in tables and figures.