Diligent production and economic strategies are important for cotton growing farmers due to of expanding cost of cultivation and stagnating productivity. Adoption of high-density planting system (HDPS) and newly released desi cotton varieties offer an alternative way to the sustainable production and decreased production cost (Kumar et al., 2017). In India, cotton is grown in 12.23 million ha with productivity of 524 kg lint/ ha (Anonymous, 2017), and productivity has not shown any remarkable improvement in the last 10 years. Bollworms (BW) especially, American bollworm Helicoverpa armigera (Hubner) and pink bollworm, Pectinophora gossypiella (Saunders) cause considerable damage (Deore et al., 2010). Theoretically, higher planting density ensures earlier crop canopy cover, higher sunlight interception leading to higher and earlier yields at reduced cost. The obvious advantage of this system is earliness (Rossi et al., 2004) since high density planting needs less bolls/plant to achieve the same yield as compared to conventional cotton and the crop does not have to maintain the late formed bolls to mature. In general, it was observed that lower plant densities produce high values of growth and yield attributes/ plant, but yield/ unit area was higher with higher plant densities (Namdeo et al., 1991; Dhoble et al., 1992; Sharma et al., 2001). Changes in plant density modify the microclimate and this may alter the incidence of pests and diseases (Venugopalan et al., 2013). Early cotton genotypes early in duration (<150 days), tolerant to sucking pests having compact plant architecture ideally suited for HDPS and mechanised harvesting. These genotypes can provide higher cotton yields under HDPS and they require very less of inputs, management and time thus, providing better economic returns to the cotton growers (Parihar et al., 2018).

Chemical control is still the most important method for managing pests (Korkor et al., 1995) and the synthetic insecticides are often a part of management programs to control lepidopterous pests (Aydin and Gurken, 2006). The success of bollworms complex control program relies on use of insecticides belonging to different chemical groups in certain sequences, application time and spray intervals (Watson et al., 1986; El-Feel et al., 1991; Abd El-Mageed et al., 2007). Conventional insecticides have not provided a longterm solution to the bollworms problem, moreover as a result of continued massive use of certain synthetic insecticides against the cotton pests has resulted in development of tolerant and resistant strains (Schmutterer, 1985). In addition to this, the toxicity of conventional insecticides to the natural enemies present in various agroecosystems has been demonstrated in laboratory tests and most of them were found harmful to the different parasites and predators (Balakrishnan et al., 2009; Sahito et al., 2011; Sabry and El-Sayed, 2011). Novel substances with different biochemical targets are effective at low doses and have less exposure in the environment. This study focuses on some new insecticides viz., chlorantraniliprole, flubendiamide, emamectin benzoate indoxacarb and spinosad (with new mode of actions, low dose and environmentally safe) by evaluating different spray regimes against cotton bollworms and its impact on natural enemies.

A field experiment were conducted to study the effect of different insecticide spray regimes on cotton (Gossypium hirsutum) for the management of bollworms and its impact on under HDPS. Sowing was done with Suraj (non Bt) variety with spacing of 45x 10 cm during 2013-2014 under rainfed condition in a deep black soil and followed all recommended agronomic practices to raise crop (Venugopalan et al., 2013). The five insecticides viz., chlorantraniliprole 18.5 SC, flubendiamide 480 SC, spinosad 45 SC, indoxacarb 14.5 SC and emamectin benzoate 5 SG at 27.75, 600, 56.25, 36.25 and 10 g.ai/ha, respectively were used at 60 DAS, 80 DAS, 100 DAS and 120 DAS in 8 treatments (spray regimes) including control in 3 replicates each were arranged in randomized complete block design. Details of different spray regimes are given in Table 1.

Treatment-wise application of insecticides was given as per the regimes by using high volume knapsack sprayer with required concentration. Control plots were kept without spray throughout the season for comparison. The incidence of H. armigera on squares and bolls damage were observed in six randomly selected plants/ plot. Number of spiders and coccinellids/ plant was also observed. In each plot, data was recorded from six pre-tagged plants. The observation was recorded at 60, 80, 100, 120 days after sowing by counting the number of insects prior to and after each spray applications i.e., 3 and 7 DAT. The data obtained for square damage and the data on natural enemies were subjected to arcsine and square root transformation respectively, and were subjected to ANOVA in randomized complete block design (RBD, p=0.05) (Gomez and Gomez, 1984).

The data on square damage as influenced with chlorantraniliprole in different spray regimes during 2013-14 and comparison of different spray regime are given in Table 2. The results showed that square damage at 80 DAS was nil in the regime 4 and which was at par with regime 2, 3, 6 and 7, recording 0.78 to 4.20% damage; maximum damage was noticed in the regime 1 (6.20%); similarly, at 100 DAS, results were non-significant among all treatments, but at 120 DAS regime 4 was the best (1.04%) followed by regime 5 and 6 (1.65 and 2.40%). Similarly, for emamectin benzoate, data at 80 DAS showed there was no damage was observed in regime 4 followed by regime 3 (0.37%) and all remaining regimes found at par with this and at 100 DAS lowest damage was noticed in the regime 3 (0.57%) and regime 4 (1.12%). Likewise, at 120 DAS non-significant results were observed among all regimes. However, no damage was observed in regime 4. The plots sprayed with flubendiamide showed that regime 3 (0.28%) was found superior at 80 DAS, followed by regime 4. Similarly, at 100 DAS non-significant results were observed, while at 120 DAS the damage was nil in regime 4. About indoxacarb, the results revealed less damage in spray regimes 2 and 3 (0.99 and 0.82%); at 100 DAS, the significantly superior ones were regime 3 and 5 (2.33 and 2.73%, respectively); maximum damage was in regime 7 (16.26%); and the damage was nil in regimes 2, 3, 4 and 5 at 120 DAS. Spinosad treated plots exhibited a minimum damage in regime 4 and 2 (1.74 and 2.17%) and these were at par with regime 1, 3 and 6; there was no damage was noticed in the regime 2 and it was followed by regime 4 (1.60%) at 100 DAS; at 120 DAS the results were found non-significant; spray regime 6 and 4 were found superior (0.69 and 1.47%, respectively (Table 2).

The boll damage was found negligible and below economic threshold level in all treatments except in untreated plots; this might be attributed to effective control of pest in square and flowering stage itself; and control plot was showing damage above ETL, indicating that insecticides were effective against bollworm on cotton under HDPS. The counts on spiders and coccinellids during 2013-14 in different spray regimes presented in (Fig. 1, 2) reveal the variations in their occurrence was non-significant among spray regimes including control; all the insecticidal treatments were observed equally safe; however, spray regime 1 (spray at 60 days after sowing) and 2 (spray at 60 and 80 days after sowing) recorded maximum counts. The data on seed cotton yield from insecticide spray regimes during 2013-14 given in Table 3 reveal that differences are statistically non-significant; however, regime 4 led to maximum (1973 kg/ ha); in case of emamectin benzoate it was 2220 kg/ ha.

The present study was carried out to evaluate spray regime windows to find the most effective spray window against bollworms and to find spray interval that pose minimal impact on natural enemies. Insecticide resistance rendered insecticides ineffective, thus increasing the need for repeated applications, wastage of resources and consequent environmental pollution (Kranthi, 2007). Resistance of key insect pests to insecticides continues to be a significant problem (Cook et al., 2005; Dhingra et al., 1988; McCaffery et al., 1989; Armes et al., 1992 & 1994; Kranthi et al., 2001). In India, the first case of control failure after spraying synthetic pyrethroids from suspected insecticide resistance in H. armigera (Hubner) was from Guntur in Andhra Pradesh (Reddy, 1990). The success of bollworm control programs relies mainly on the spraying those insecticides belonging to novel modes of action in a particular time interval. Chemicals such as spinosad, indoxacarb, emamectin benzoate, novaluron and lufenuron ensured effective control and less toxic to beneficial insects (Kranthi, 2007). The most probable reason for increased use of new chemistry insecticides is resistance (Razaq, 2006). There is a need to develop alternative insecticides/ techniques, allowing rational use of pesticides (Barrania et al., 2016). Sharah and Ali (2008) concluded that it is very important to determine which of the available insecticides are most effective and economical at particular stage of crop phenology, doses and intervals should the spray be done to achieve the best result. Spray regimes 2 and 3 was effective in taking care of bollworms in the initial stage itself by minimising square damage in early growth stages. Earlier reports of (Donnelly and Adeyemi, 1966; Akinlosotu, 1969; Fadare, 1984) clearly revealed that early chemical intervention reduced the damage levels of flowers.

In the present study, flubendiamide, emamectin benzoate and spinosad were found relatively superior. Flubendiamide 60 g a.i. ha^-1 showed marked reduction in the larvae and damage (Thilagam et al., 2010); Tomar et al., 2005) also obtained similar results with square damage caused by bollworms. Saleh et al. (2013) observed that emamectin benzoate achieved high efficacy against pink and spotted bollworms; Gupta et al. (2005) and Sontakke et al. (2007) found this as the most potent. Johnson et al. (1997) and Dandale et al. (2000) observed that spinosad was quite effective. Ghure et al. (2008) observed that indoxacarb and spinosad were effective. The present results agree with those of Johnson et al. (2000), Haidar et al. (2002); Omar et al. (2005), Ghure et al. (2008) and Gosalwad et al. (2009). Chlorantraniliprole also proved equally effective under high density planting system in the present study and this is in accordance with Dhengre et al. (2017); Ma et al. (2000) reported that chlorantraniliprole was found to be the most effective against H. armigera in cotton. These pesticides were effective against targeted arthropod pests but relatively non-harmful to natural enemies (De Clercq, 1995; Charleston, 2005).

Insecticides evaluated in present investigation not only effective for controlling bollworms but were relatively safe on spiders and coccinellids. Ruberson et al. (1998) and Lacey et al. (1997) assumed the compatibility of natural enemies with pesticides depends on a range of factors; Tohnishi et al. (2005) and Kubendran et al. (2006) found flubendiamide to be least toxic against beneficial arthropods. Chlorantraniliprole at doses ranging from 20- 50 g a.i./ ha was safe to natural enemies (Misra, 2011; Yang et al., 2012). These diamides are relatively safe to natural enemies (Brugger et al., 2010); spinosad was found to be harmless to Coccinella septempunctata grubs (Olszak and Sekrecka, 2008). Spray regime 1 (spray at 60 days after sowing) and 2 (spray at 60 and 80 days after sowing) recorded maximum counts of spiders and coccinellids. Interestingly, in all insecticidal spray regimes maximum natural enemy population was noticed 80 DAS. The insecticides, flubendiamide, emamectin benzoate, spinosad and chlorantraniliprole are thus superior as spray regimes against bollworms; in particular, spray regime 2 (spray at 60 and 80 DAS) was the best and can be recommended under HDPS system in cotton.

Figures in parentheses are arcsine transformed values; NS = Non-significant

References