Microbes, Dodonaea viscosa and Chlorantraniliprole as Components of Helicoverpa armigera IPM Program: A Three Region Open-Field Study

Abstract

Field trials were conducted on the efficacy of different control options against Helicoverpa armigera on tomato. We evaluated and compared pest control, safety to natural enemies, crop yield and economics of application for various treatments including: a mycoinsecticide based on Beauveria bassiana; a baculovirus, H. armigera nucleopolyhedrovirus (HaNPV); a plant extract from Dodonaea viscosa; and the insecticide chlorantraniliprole (Coragen^®). Trial sites were located in the Bahawalpur, Faisalabad and Rawalpindi regions in Punjab, Pakistan. A combined application of HaNPV + chlorantraniliprole was better than all other treatments in reducing pest larval populations and fruit damage, and in increasing crop yield. The least effective control was with D. viscosa, but plots treated with this plant extract also had the greatest number of natural enemies. Treatment with HaNPV + chlorantraniliprole resulted in the fewest natural enemies at all observation times and in all field sites, but the highest yields, varying from 42.8–46.6 tons ha⁻¹, and the highest cost:benefit ratios of 1:1.74–1:1.81. Our findings suggest that a combined application of a specific microbial agent and an insecticide effectively controls H. armigera, increases tomato yield, reduces costs and maximizes economic returns. This open-field study concludes that all of the microorganisms and other treatments have the potential to combat H. armigera populations and could be used in successful integrated pest management programs.

1. Introduction

Amongst vegetable crops in Pakistan, tomato (Lycopersicon esculentum Miller) is second only to potato in terms of area cultivated and production [1], with an estimated 61 thousand ha grown and average crop yields of about 9.5 tons per hectare [2]. Despite this, yields are quite low compared with other tomato-producing countries [3]. A complex of boring and sucking insect pests attack tomato plants, ultimately reducing the quality and quantity of fruit; a major pest worldwide is the highly polyphagous cotton bollworm, Helicoverpa armigera (Hübner) (Lepidoptera: Noctuidae).

In Pakistan, H. armigera is regarded as the key pest of tomato crops and is known locally as the tomato fruitworm (=Old World bollworm) [4]. A wide range of economically important crops such as cotton, maize, tobacco, sunflower, chickpea, soybean and fodder crops are damaged by this pest. Estimates of yield losses in different crops exceed 55% in tomato [5,6] 1015  kg ha^–1 in pigeon pea [7] and three-million-dollar losses annually in cotton [8]. Over 100 crops, and other non-crop plant species, have been recorded as hosts in Asia, Africa and Australia [9,10,11,12]. The pest is found in cultivated areas throughout the year, with weeds and shrubs acting as important reservoirs during non-cropping periods [13].

The most widespread and conventional practice for the control of H. armigera in Pakistan is the use of synthetic insecticides, but a number of hazards associated with human health concerns and environmental impacts are driving the search for alternative methods. Furthermore, the development of insecticide resistance to pyrethroids [14], organophosphates [15], carbamates [16] and the organochlorine compound endosulfan [17] has been noted in field populations. Alternative approaches to insect pest management include the possible use of microbial control agents.

Beauveria bassiana (Balsamo-Crivelli) Vuillemin (Cordycipitaceae: Hypocreales) is a ubiquitous entomopathogenic fungus readily isolated from soil and insects and grows endophytically within plants [18,19,20,21,22]. This mycopathogen has a broad host range and is acknowledged to be a vital biological control agent of many insect pests [23,24,25,26,27]. Isolates of B. bassiana have shown considerable promise in laboratory and field testing [28,29,30,31,32,33]. Amongst the hosts of B. bassiana, H. armigera is regarded as a quite susceptible insect and stories of success in field trials have been reported by various researchers [34,35,36,37].

Nucleopolyhedroviruses (NPVs) belong to a large group of viruses in the family Baculoviridae, and many NPVs are important pathogens of lepidopteran pests. As NPVs are host-specific and do not affect non-targets, including natural enemies, there has been considerable interest in their use as biological control agents [38,39,40,41], including the development of genetically modified NPVs with a quicker rate of kill in treated insects [42,43].

Chlorantraniliprole (Rynaxypyr^®, Coragen^®) belongs to a new class of selective insecticides (anthranilic diamides) exhibiting a novel mode of action based on the disruption of insect muscle cells leading to paralysis and death. It has low mammalian toxicity, high intrinsic activity against target pests, strong ovi-larvicidal and larvicidal properties, minimal impact on beneficial arthropods, long-lasting crop protection and no cross-resistance to any existing insecticide [44,45,46]. These features mean that chlorantraniliprole is a good option for use with other control methods in insect pest management (IPM) and insecticide resistance management (IRM) strategies [47,48].

Dodonaea viscosa Jacq. (hopseed bush or hopbush) is a widely distributed evergreen shrub found in tropical and subtropical regions. It contains low levels of cyanogenic glycosides and other compounds that have antifeedant activities [49], and this plant has the potential to be used as a botanical biopesticide for controlling lepidopteran pests [50].

The present study was conducted to determine whether combinations of biorational and chemical agents could provide better control and greater economic returns than if each control agent was used by itself. Combining several approaches would be expected to reduce the risk of insecticide resistance developing in H. armigera.

2. Materials and Methods

2.1. Experimental Sites and Crop Establishment

The study was conducted at three sites; in farmers’ fields at Faisalabad, Bahawalpur and Rawalpindi district (one field at each location). The well-known variety ‘Money Maker’ was raised as seedlings in tomato nurseries adjacent to the experimental sites for 30 days before being transplanted to raised beds at the field sites using the recommended distance of 45 cm between tomato plants; the transplanting took place in December. During the entire cropping period, standard agronomic practices were followed for irrigation, fertilizer use, fungicide application, hand weeding and the monitoring of insect pests.

2.2. Beauveria bassiana

The reference strain WG-06 of B. bassiana was isolated from infected cadavers (4th and 5th larval instar) of H. armigera collected from tomato fields in the district of Rawalpindi, Pakistan during 2009. The isolation of the fungus was firstly performed on the Sabouraud Dextrose Agar (SDA) (32.5 g SDA; 7.5 g of bacto agar; 5 g yeast in 1 liter of distilled water). It was sub-cultured on an SDA medium using the single spore technique for mass production of conidia at 25 °C and 75% rh with 16 h of illumination per day. After 14 days of incubation, the plates were allowed to dry under aluminum foil on the bench top for one week. The conidia were harvested using a sterilized (70% ethanol) scalpel by scraping the conidial masses formed on the surface of the plates. They were put in conical tubes (50 mL) with (30 mL 0.05% Tween-80) solution and eight glass beads were added inside each tube and vortexed for about 5 min. The desired concentration (1.3 × 10⁸ conidia mL⁻¹) was determined using a hemocytometer under the microscope. The conidia viability was determined by spreading 0.1 mL of solution @ 1 × 10⁶ conidia per mL on SDAY (SDA with 1% yeast) plates (60 mm), wrapping with parafilm and placing in an incubator at 25 °C with a 14:10 h (Light: Dark) photoperiod for 16 h [51]. A cover slip was placed on the SDAY plate and approximately 200 conidia were assessed for germination by counting four times, i.e., two counts from each plate [52]. A conidium was considered to have germinated if the germ tube was longer than the actual size of the conidia [51]. The desired concentration of conidia 1.3 × 10⁸ conidia mL⁻¹ was adjusted based on the germination percent (%) of the fungus.

2.3. Nucleopolyhedrovirus

The nucleopolyhedrovirus (NPV) commercial formulation was provided by AgriLife, Hyderabad, India. H. armigera second instar larvae were infected with a suspension of NPV by spraying and feeding them on an artificial diet. The virus was allowed to propagate in the midguts for seven days. The midguts of cadavers seven days after inoculation were homogenized in deionized water and strained through a fine mesh to remove large debris. The filtrate was centrifuged at 16,000 rpm for 45 min [48,53,54], and the purified virus was recovered and washed in distilled water three times and held in 0.1 mM NaOH at 5 °C. The polyhedral occlusion bodies mL⁻¹ suspension (POB mL⁻¹) was determined from ten counts under the microscope using an improved Neubauer hemocytometer [55].

2.4. Dodonaea viscosa

Leaves of D. viscosa from the mid-peripheral area of plants were collected from the Botanical Garden, University of Agriculture, Faisalabad, Pakistan. These plants were never sprayed with any chemicals. The leaves were washed twice in distilled water and surface sterilized with 1% sodium hypochlorite solution. After rinsing again the plant material was dried under shade for one day. The leaves were weighed and homogenized with distilled water 1:3 (w/v). The slurry of plant material obtained was passed through a muslin cloth and the filtrate was centrifuged at 5000 rpm for 10 min. The supernatant obtained after centrifugation was used as the standard aqueous solution. All plant extracts were heated at 40 °C for 15 min to avoid contamination [56]. The required concentration of 7 ppm was created by dilution for use in experiments.

2.5. Chlorotaniliprole

Coragen^® 20 SC (chlorantraniliprole) powered by Rynaxypyr (FMC United Private Limited, Pakistan) is a novel insecticide in the anthranilic diamide class. It is a semi-viscous liquid with an off-white color. It contains chlorantraniliprole (20% w v⁻¹) and other ingredients (80% w v⁻¹).

2.6. Field Trials

Experiments were conducted using a randomized complete block design (RCBD) in three blocks. Each block was divided into seven treatments which were allocated in 21 plots at each experimental site. Each plot was measured 5 by 5 m separated by 0.5 m buffers, and the blocks were separated by 1.5 m buffers. The treatments were as follows: B. bassiana @ 1.3 × 10⁸ conidia mL⁻¹; HaNPV @ 2.5 × 10⁷ POB mL⁻¹; D. viscosa plant extract @ 7 ppm; chlorantraniliprole @ 0.29 mL⁻¹; chlorantraniliprole plus B. bassiana; chlorantraniliprole plus HaNPV, and an untreated control.

The applications were conducted using a knapsack hydraulic sprayer (FUSITE Co. Ltd., Taizhou, China) with a flat fan nozzle. Each treatment was applied as two sequential sprays with an interval of 15 days between each spray starting in mid-March. The cross-contamination of test chemicals was prevented as a separate sprayer was used for each biorational or chemical agent.

2.7. Population Density Monitoring

Prior to experimentation, counts of H. armigera larvae were taken at least twice a week until a threshold level (one larva/plant) [57] was reached. To determine the effectiveness of the various treatments, pre-treatment counts were taken one day before each spray application by recording larval densities from ten plants selected at random from each plot, and then at 3, 5 and 10 days after each spray treatment (DAT). The larvae with no movement response to disturbance were considered dead and were not included. The population densities of several H. armigera natural enemies were also recorded from the same ten plants selected at random at 3, 5 and 10 DAT after the second spray. The natural enemies noted were green lacewing Chrysoperla carnea (Stephens), spiders, mites, ladybird beetles and predatory bugs. The data on the population of natural enemies were recorded from ten randomly selected plants in each treatment plot at 3, 5, and 7 days after the second application of treatments [34].

2.8. Fruit Damage Assessment

The degree of damage caused by H. armigera larvae was determined by the weight of all fruit (ripe and unripe) from the tomato plants (ten plants per occasion selected at random) that were sampled on three occasions: 3, 5 and 10 DAT in the months of February to April at each location. The percentage of damaged fruit was calculated using the following formula: Percent damaged fruit = (weight of H. armigera-infested fruits/total weight of fruits) × 100 [58].

2.9. Yield Evaluation and Cost-Benefit Analysis

All mature, undamaged, red-colored fruits (i.e., commercially saleable yield) were harvested from each plot on each of the three sampling occasions (3, 5 and 10 DAT) and weighed; the yields on each occasion were added together and converted to a per hectare figure to give a final cumulative yield for each plot at the end of the experiment. Tomato yield and costs of production were used to calculate the cost-benefit ratio (CBR) of production for each treatment. Briefly, it was calculated using the formula [cost:benefit ratio = Bt/(1 + i)n/Ct/(1 + i)n, where: Bt = benefit in year; Ct = cost in year; n = no. of years; i = interest rate] [59]. In addition, the percent increase in yield over control was calculated by the formula [T-C/C × 100, where: T = yield at treatment; C = yield at control] [58].

2.10. Statistical Analysis

The data for each insect population, larval density (3, 5, 10 d) and natural enemies (3, 5, 7 d) were submitted to ANOVA to determine differences among exposure intervals using the Minitab statistical package [60]. Analyses were performed separately for each location and each sample time. Then, all data were analyzed separately for each insect population according to the Repeated Measures Analysis. The repeated factor was exposure interval, while the response variable was larval density, and the main effects were treatments and spray application. Differences between treatment means were compared using Tukey’s HSD test at p = 0.05 [61].

3. Results

3.1. Larval Population Densities

Regarding H. armigera larvae, all main effects were significant while associated interaction spray application × interval and treatment × spray application × interval were non-significant (

Table 1. Repeated measures MANOVA for main effects and associated interactions for larval population of H. armigera (total df = 167).

	Region		Rawalpindi		Faisalabad		Bahawalpur
	Source	df	F	p	F	p	F	p
Between variables	Treatment	6	18.68	<0.01	4.10	0.0045	17.89	<0.01
	Spray application	1	18.33	<0.01	<0.01	1.00	31.10	<0.01
	Treatment × spray application	6	5.03	<0.01	<0.01	1.00	5.16	<0.01
Within variables	Interval	3	30.11	<0.01	39.22	0.00	39.48	<0.01
	Treatment × interval	18	3.26	<0.01	4.41	0.00	3.64	<0.01
	Spray application×interval	3	0.46	0.71	<0.01	1.00	0.02	0.99
	Treatment × spray application × interval	18	0.34	0.99	<0.01	1.00	0.22	0.99

). Prior to the first spray, pre-treatment data indicated there were significant differences in H. armigera larval population densities between treatments at Rawalpindi, Faisalabad and Bahawalpur (

Table 2. Mean larval density (larvae per ten plants ± SE) of H. armigera following field application of different biorational and chemical control options based on B. bassiana; HaNPV; D. viscosa and chlorantraniliprole (chloran.). Means for each location and sampling occasion followed by the same letters are not significantly different from each other; Tukey-Kramer HSD test at p = 0.05.

Region	Treatment	1st Spray				2nd Spray
		Pre-Treatment	3 DAT	5 DAT	10 DAT	Pre-Treatment	3 DAT	5 DAT	10 DAT
Rawalpindi	B. bassiana	6.47 ± 1.26 a	4.78 ± 1.05 b	3.76 ± 0.81 ab	3.50 ± 1.31 b	4.35 ± 0.79 b	2.70 ± 0.71 b	1.81 ± 0.40 b	0.84 ± 0.11 b
	HaNPV	7.35 ± 0.85 a	4.12 ± 1.11 bc	3.38 ± 0.60 ab	2.90 ± 0.94 b	3.55 ± 0.59 ab	2.35 ± 0.94 b	1.46 ± 1.04 b	0.74 ± 0.34 b
	D. viscosa	5.89 ± 0.41 ab	5.20 ± 1.21 ab	4.58 ± 0.85 ab	3.85 ± 0.96 ab	4.53 ± 1.15 ab	3.31 ± 0.77 b	2.26 ± 0.74 b	1.70 ± 0.50 b
	Chloran.	6.59 ± 0.50 a	3.97 ± 1.06 c	3.10 ± 0.77 b	2.18 ± 0.54 b	3.15 ± 0.90 b	2.20 ± 0.77 c	0.98 ± 0.80 c	0.58 ± 0.14 c
	B. bassiana + chloran.	5.32 ± 0.91 ab	3.70 ± 1.06 c	2.72 ± 0.84 b	1.92 ± 0.62 c	3.05 ± 0.94 c	1.41 ± 0.54 c	0.70 ± 0.26 c	0.48 ± 0.09 c
	HaNPV + chloran.	5.38 ± 1.16 ab	3.37 ± 1.23 c	2.16 ± 0.95 c	1.57 ± 0.88 c	2.90 ± 1.24 c	1.13 ± 0.73 c	0.54 ± 0.30 c	0.35 ± 0.06 c
	Control	4.92 ± 0.46 b	6.02 ± 1.13 a	6.55 ± 0.67 a	7.20 ± 0.73 a	8.15 ± 1.33 a	8.30 ± 1.01 a	9.79 ± 1.37 a	10.29 ± 1.07 a
	F6,20	0.94	2.55	4.79	7.23	4.48	12.45	25.25	89.02
	p	0.50	0.08	<0.01	<0.01	<0.01	<0.01	<0.01	<0.01
Faisalabad	B. bassiana	7.66 ± 1.27 ab	6.06 ± 1.06 ab	5.14 ± 0.83 ab	4.39 ± 1.33 b	5.04 ± 0.80 b	3.67 ± 0.70 b	2.68 ± 0.43 b	1.05 ± 0.08 b
	HaNPV	8.55 ± 0.87 a	5.44 ± 1.13 b	4.72 ± 0.62 ab	3.80 ± 0.96 b	4.24 ± 0.61 ab	3.21 ± 0.95 b	2.29 ± 1.01 b	0.96 ± 0.34 b
	D. viscosa	7.11 ± 0.44 ab	6.51 ± 1.22 b	5.93 ± 0.87 ab	4.74 ± 0.98 ab	5.22 ± 1.17 ab	4.18 ± 0.81 b	3.13 ± 0.75 b	1.91 ± 0.51 b
	Chloran.	7.78 ± 0.49 ab	5.27 ± 0747 b	4.44 ± 0.77 b	3.05 ± 0.96 b	3.84 ± 0.92 b	2.93 ± 1.00 c	1.82 ± 0.81 c	0.63 ± 0.16 c
	B. bassiana + chloran.	6.51 ± 0.92 b	5.03 ± 0.10 bc	4.08 ± 0.81 c	2.86 ± 0.67 c	3.73 ± 0.95 c	2.26 ± 0.54 c	1.09 ± 0.10 c	0.49 ± 0.14 c
	HaNPV + chloran.	6.52 ± 1.17 b	4.66 ± 1.24 c	3.51 ± 0.95 c	2.44 ± 0.88 c	3.60 ± 1.26 c	1.92 ± 0.77 c	0.98 ± 0.29 c	0.28 ± 0.16 c
	Control	5.78 ± 0.13 c	7.37 ± 1.16 a	7.92 ± 0.67 a	8.15 ± 0.75 a	9.08 ± 1.35 a	9.16 ± 1.08 a	10.61 ± 1.37 a	10.83 ± 1.13 a
	F6,20	1.14	2.56	4.90	7.03	4.68	13.08	25.56	68.90
	p	0.39	0.08	<0.01	<0.01	<0.01	<0.01	<0.01	<0.01
Bahawalpur	B. bassiana	9.61 ± 1.29 a	7.94 ± 1.09 ab	7.03 ± 0.87 ab	6.20 ± 1.33 b	6.95 ± 0.83 b	5.47 ± 0.73 b	4.46 ± 0.42 b	2.8 ± 0.10 b
	HaNPV	9.03 ± 0.94 a	7.32 ± 1.13 b	6.81 ± 0.54 b	5.66 ± 0.90 b	6.16 ± 0.60 ab	4.98 ± 0.92 bc	4.07 ± 1.03 bc	2.71 ± 0.35 b
	D. viscosa	8.95 ± 0.44 ab	8.37 ± 1.19 b	7.81 ± 0.85 ab	6.59 ± 0.95 b	7.14 ± 1.19 ab	5.95 ± 0.79 b	4.96 ± 0.72 b	3.72 ± 0.50 b
	Chloran.	9.58 ± 0.45 a	7.15 ± 1.07 b	6.36 ± 0.76 c	4.87 ± 0.98 c	5.78 ± 0.94 b	4.76 ± 0.92 bc	3.60 ± 0.80 c	2.44 ± 0.20 c
	B. bassiana + chloran.	8.40 ± 0.85 ab	6.95 ± 1.08 c	6.04 ± 0.76 c	4.68 ± 0.70 c	5.63 ± 0.97 b	4.04 ± 0.51 c	2.89 ± 0.13 d	2.24 ± 0.15 c
	HaNPV + chloran.	8.52 ± 1.13 ab	6.54 ± 1.26 c	5.39 ± 0.93 c	4.30 ± 0.82 c	5.52 ± 1.27 c	3.69 ± 0.80 c	2.75 ± 0.31 d	2.08 ± 0.17 c
	Control	7.70 ± 0.16 c	9.29 ± 1.14 a	9.78 ± 0.65 a	9.97 ± 0.72 a	10.97 ± 1.34 a	11.53 ± 0.74 a	12.47 ± 1.52 a	12.87 ± 1.33 a
	F6,20	4.48	2.54	4.97	7.21	4.48	0.93	21.96	54.53
	p	<0.01	0.08	<0.01	<0.01	<0.01	<0.01	<0.01	<0.01

). At all localities, there were significant decreases in larval densities in the biorational and chemical treatments compared with the control at 5 and 10 DAT. The final densities of H. armigera at 10 DAT were lowest in the combined treatments of HaNPV with chlorantraniliprole and B. bassiana with chlorantraniliprole. The mean H. armigera larval population densities in the control plots varied from 7.20 ± 0.73–9.97 ± 0.72 larvae per ten plants at all three localities while means were 2.1- to 4.6-fold lower and superior in the combination treatments of microbial and chemical control agents.

Pre-treatment means before the second spray were significantly higher in the control plots at all localities compared with the treatments, and the final densities at 10 DAT were the smallest for the combination treatments of HaNPV with chlorantraniliprole and B. bassiana with chlorantraniliprole; mean densities varied from 0.35 ± 0.06 to 2.24 ± 0.15 larvae per ten plants which were 5.8- to 29.4-times smaller than equivalent control densities at all three localities. The trend in the effectiveness of the various treatments in reducing larval densities was:

HaNPV + chlorantraniliprole > B. bassiana + chlorantraniliprole > chlorantraniliprole > HaNPV > B. bassiana > D. viscosa.

3.2. Effects on Non-Target Natural Enemies

Significant differences in population densities of natural enemies were observed between treatments at 3, 5 and 7 DAT after the second spray. The numbers of the five taxa were approximately double in the untreated control plots compared with the biorational and chemical treatments. Treatment with chlorantraniliprole alone or in combination with B. bassiana or HaNPV had the greatest negative effect on populations of the natural enemies of H. armigera while the D. viscosa plant extract appeared to have the least effect on natural enemy densities (Figure 1, Figure 2 and Figure 3).

3.3. Fruit Damage

Concerning fruit damage caused by H. armigera, all of the main effects were significant while associated interaction spray application × interval and treatment × spray application × interval were non-significant (

Table 3. Repeated measures MANOVA parameters for main effects and associated interactions for tomato fruit damage caused by H. armigera larvae (total df = 125).

	Region		Rawalpindi		Faisalabad		Bahawalpur
	Source	df	F	p	F	p	F	p
Between variables	Treatment	6	35.30	<0.01	30.81	<0.01	35.30	<0.01
	Spray application	1	26.50	<0.01	24.89	<0.01	26.50	<0.01
	Treatment × spray application	6	5.81	<0.01	4.91	<0.01	5.81	<0.01
Within variables	Interval	3	11.97	<0.01	12.57	<0.01	11.97	<0.01
	Treatment × interval	18	2.39	0.01	2.38	0.01	2.39	0.01
	Spray application × interval	3	2.35	0.10	3.18	0.04	2.35	0.10
	Treatment × spray application × interval	18	0.07	1.00	0.11	0.99	0.07	1.00

). There were significant differences in the degree of fruit damage between the control and the other treatments from 5 DAT after the first spray (the second sampling occasion) onwards at all localities (

Table 4. Fruit damage (% ± SE) caused by H. armigera larvae following field application of different biorational and chemical control options based on B. bassiana; HaNPV; D. viscosa and chlorantraniliprole (chloran.). Means within each observation followed by the same letters are not significantly different from each other (transformed values); Tukey-Kramer HSD test at p = 0.05.

Region	Treatment	Fruit Damage (%)
			1st Spray			2nd Spray
		3 DAT	5 DAT	10 DAT	3 DAT	5 DAT	10 DAT
Rawalpindi	B. bassiana	24.84 ± 4.90 b	20.59 ± 3.86 b	16.83 ± 4.12 b	13.63 ± 3.29 bc	9.09 ± 2.19 b	8.82 ± 1.61 bc
	NPV	21.96 ± 5.35 b	18.71 ± 2.88 ab	14.45 ± 3.17 b	12.53 ± 2.68 bc	7.52 ± 2.75 b	6.97 ± 2.66 bc
	D. viscosa	26.94 ± 5.75 b	24.39 ± 4.08 ab	18.61 ± 3.02 b	16.06 ± 2.56 b	11.17 ± 2.56 b	9.31 ± 1.85 c
	Chloran.	21.15 ± 5.05 bc	17.40 ± 3.61 bc	11.62 ± 1.89 c	10.37 ± 1.68 bc	6.91 ± 1.53 c	5.54 ± 1.75 bc
	B. bassiana + chloran.	20.02 ± 5.22 bc	15.25 ± 2.06 c	10.40 ± 2.58 c	7.69 ± 1.12 c	2.69 ± 0.58 d	1.73 ± 1.38 d
	NPV + Coragen	18.25 ± 4.44 c	13.14 ± 4.27 c	8.43 ± 3.27 c	5.39 ± 1.63 c	1.66 ± 1.08 d	0.96 ± 0.62 c
	Control	30.42 ± 4.85 a	34.26 ± 2.77 a	35.51 ± 3.34 a	40.10 ± 4.14 a	46.68 ± 5.30 a	51.20 ± 4.55 a
	F6,20	2.65	6.25	14.80	21.88	38.98	96.43
	p	0.07	<0.01	<0.01	<0.01	<0.01	<0.01
Faisalabad	B. bassiana	28.52 ± 4.99 b	24.11 ± 3.89 b	20.40 ± 4.12 b	17.04 ± 3.20 b	12.45 ± 2.16 b	12.21 ± 1.62 bc
	NPV	25.58 ± 5.38 b	22.19 ± 2.88 ab	18.04 ± 3.19 b	15.94 ± 2.69 bc	10.88 ± 2.73 bc	10.35 ± 2.66 bc
	D. viscosa	30.60 ± 5.76 ab	27.93 ± 4.10 ab	22.03 ± 3.97 bc	19.55 ± 2.58 b	14.52 ± 2.53 b	13.30 ± 2.36 b
	Chloran.	24.76 ± 5.07 b	20.89 ± 3.62 bc	14.41 ± 2.54 bc	13.80 ± 1.70 bc	9.053 ± 2.86 bc	8.913 ± 1.74 c
	B. bassiana + chloran.	23.61 ± 5.22 b	19.25 ± 3.79 c	13.63 ± 3.11 c	11.20 ± 1.12 c	5.29 ± 1.15 c	5.08 ± 1.40 c
	NPV + chloran.	21.87 ± 4.44 b	16.60 ± 4.28 c	11.67 ± 3.55 b	8.90 ± 1.63 b	4.64 ± 1.23 b	4.34 ± 1.59 c
	Control	34.60 ± 5.41 a	37.60 ± 2.69 a	38.84 ± 3.11 a	42.80 ± 4.10 a	49.57 ± 5.44 a	54.71 ± 4.56 a
	F6,20	2.73	5.64	14.55	22.30	37.39	103.13
	p	0.06	<0.01	<0.01	<0.01	<0.01	<0.01
Bahawalpur	B. bassiana	32.66 ± 5.22 bc	28.62 ± 3.75 b	24.87 ± 3.08 b	21.48 ± 3.19 b	16.87 ± 2.23 b	16.37 ± 1.57 bc
	NPV	30.05 ± 5.36 bc	26.74 ± 2.74 ab	22.54 ± 3.09 bc	20.44 ± 2.68 b	15.34 ± 2.82 b	14.56 ± 2.56 bc
	D. viscosa	35.58 ± 5.31 ab	32.33 ± 4.16 ab	26.46 ± 3.94 b	23.93 ± 2.44 bc	18.92 ± 2.50 bc	17.70 ± 2.39 b
	Chloran.	29.33 ± 5.11 bc	25.30 ± 3.58 bc	18.86 ± 2.60 bc	18.24 ± 1.68 bc	13.46 ± 2.86 bc	12.85 ± 1.39 bc
	B. bassiana + chloran.	24.65 ± 2.01 c	23.64 ± 3.81 c	18.28 ± 3.04 c	15.60 ± 1.12 c	9.64 ± 1.15 c	8.94 ± 0.96 c
	NPV + chloran.	26.28 ± 4.49 c	21.02 ± 4.25 c	16.22 ± 3.40 b	13.39 ± 1.63 c	9.08 ± 1.19 c	8.10 ± 1.07 c
	Control	38.83 ± 5.56 a	41.76 ± 2.51 a	43.32 ± 3.09 a	46.73 ± 3.70 a	53.10 ± 4.64 a	59.21 ± 4.57 a
	F6,20	3.12	5.78	16.02	23.69	44.89	106.59
	p	0.04	<0.01	<0.01	<0.01	<0.01	<0.01

). By 7 DAT after the second spray (the sixth and final sampling occasion), the mean percent damage in the control plots varied from 51–59%, while in the biorational and chemical treatments, it was significantly lower at between 1 and 18% damage, lowest in Rawalpindi and highest in Bahawalpur.

The trend for reductions in tomato damage at the three localities was as follows:

HaNPV + chlorantraniliprole > B. bassiana + chlorantraniliprole > chlorantraniliprole > HaNPV > B. bassiana > D. viscosa.

3.4. Fruit Yield and Cost-Benefit Ratio

Fruit yields were significantly lower at 987–1294 kg ha⁻¹ in the control plots at all localities. However, a significant increase in yield over the control was recorded. The highest yield among the treatments was observed in plots treated with HaNPV + chlorantraniliprole, with a maximum quantity of 4659 kg ha⁻¹ in the Rawalpindi region. The yield was increased by 3.6- to 4.3-fold higher in the treatments with chlorantraniliprole in combination with either HaNPV or B. bassiana (

Table 5. Cost-benefit ratio of tomato crop after treating following field application of different biorational and chemical control options based on B. bassiana; HaNPV; D. viscosa and chlorantraniliprole (chloran.). Yield means within each observation that are followed by the same letters are not significantly different from each other; Tukey-Kramer HSD test at p = 0.05.

Region	Treatment	Yield (kg/ha.)	Cost (ha.) USD	Increased Yield Over Control (kg/ha.)	Increased Yield Over Control (%)	Income Increased/ha. (USD)	Cost-Benefit Ratio
Rawalpindi	B. bassiana	3466 ± 2.20 ab	24.09	21.72	153.61	49.36	1.05
	HaNPV	3519 ± 2.64 ab	24.09	22.25	167.83	50.57	1.10
	D. viscosa	3282 ± 3.13 b	22.72	19.88	171.95	45.18	0.99
	Chloran.	3726 ± 2.21 ab	25.45	24.33	187.97	55.29	1.17
	B. bassiana + chloran.	4112 ± 2.86 ab	27.27	28.18	217.75	64.04	1.35
	HaNPV + chloran.	4659 ± 1.90 a	27.27	33.6	260.10	76.50	1.81
	Control	1294 ± 1.79 c	-	-	-	-	-
	F6,20	18.14	-	-	-	-	-
	p	<0.01	-	-	-	-	-
Faisalabad	B. bassiana	3350 ± 1.68 ab	19.99	54.98	156.30	25.89	1.07
	HaNPV	3419 ± 1.89 ab	20.61	56.70	191.05	27.45	1.14
	D. viscosa	2950 ± 3.62 b	16.35	44.98	197.05	18.16	0.80
	Chloran.	3626 ± 2.14 ab	25.45	24.75	215.03	56.25	1.21
	B. bassiana + chloran.	4037 ± 2.16 ab	27.27	28.87	250.80	65.60	1.41
	HaNPV + chloran.	4451 ± 2.97 a	27.27	33	286.74	75.00	1.75
	Control	1151 ± 1.47 c	-	-	-	-	-
	F6,20	17.42	-	-	-	-	-
	p	<0.01	-	-	-	-	-
Bahawalpur	B. bassiana	3062 ± 2.57 ab	18.86	51.88	162.24	23.07	0.96
	HaNPV	3280 ± 1.48 ab	20.84	57.33	210.27	28.02	1.16
	D. viscosa	2588 ± 3.25 b	14.55	40.03	232.35	13.66	0.60
	Chloran.	3469 ± 3.65 ab	25.45	24.82	251.50	56.40	1.22
	B. bassiana + chloran.	3889 ± 3.47 ab	27.27	29.03	294.12	65.97	1.42
	HaNPV + chloran.	4279 ± 3.68 a	27.27	32.92	333.60	74.82	1.74
	Control	987 ± 1.34 c	-	-	-	-	-
	F6,20	11.86	-	-	-	-	-
	p	<0.01	-	-	-	-	-

).

The cost:benefit ratios were highest for the combination of HaNPV with chlorantraniliprole and varied from 1:1.74 to 1:1.81 at the three localities. The lowest cost:benefit ratio was for the D. viscosa plant extract (Table 5).

4. Discussion

The application of chemical insecticides is the primary method for the management of lepidopterous pests around the globe [62,63]. However, this approach has not experienced sustained success due to the development of insecticide resistance [63]. Entomopathogenic microorganisms, plant extracts and new-chemistry insecticides are used as alternative methods for the sustainable management of H. armigera without damaging non-target organisms and the environment [63]. Our findings showed that maximum reduction in larval populations were recorded in the Rawalpindi district 10 DPT after a second spray of B. bassiana. Our findings are consistent with prior studies involving entomopathogenic fungi applied for the control of H. armigera [34]. For example, recently, Kalvnadi et al. [64] reported high levels of B. bassiana mortality in H. armigera. Entomopathogenic fungi viz., B. bassiana and Metarhizium anisopliae were effective in reducing fruit damage [65] when both were tested under field conditions against H. armigera on tomatoes. Jarrahi and Safavi [66] investigated the sub-lethal effects of M. anisopliae sensu lato (s.l.) (isolate M14) on the life table parameters of offspring from treated larvae of H. armigera. Fecundity decreased in females derived from H. armigera larvae treated with M. anisopliae s.l. Sub-lethal concentrations of the entomopathogen reduced the net reproduction rate (R0) of F1 insects for all treatments compared with the control [66]. A field study was conducted to evaluate the efficacy [67] of two virulent native isolates of entomopathogenic fungi (M. anisopliae and B. bassiana) against chickpea pod borer. The number of H. armigera larvae was significantly lower in plots treated with M. anisopliae and B. bassiana during the vegetative, flowering and pod setting stage of chickpea. Savitha et al. [68] found the entomopathogenic fungus N. rileyi caused the highest mortality in third instar larvae of H. armigera followed by M. anisopliae and B. bassiana; all three isolates of Verticillium lecanii used in the same study failed to cause infection. In another field experiment, a combination of B. bassiana + neem soap achieved the lowest population of H. armigera and pod damage [69].

Dodonaea viscosa extract alone showed good results but in comparison with other treatments was less effective in reducing larval populations. Our results are in agreement with other studies on botanical extracts. For example, Ali et al. [70] reported that plant extracts (neem seed, turmeric, garlic and marsh pepper) caused the higher mortality of H. armigera compared to the control, and neem seed resulted in the highest tomato yield. The hexane extract of hop, a bushy perennial shrub, Dodonaea angustifolia, was effective [71] against different life stages of H. armigera. In greenhouse experiments, the mortality induced by extracts of A. mucosa (as an emulsifiable concentrate formulation) and the botanical insecticide based on the extract of Annona squamosa L. (Anosom^® 1 EC) was comparable with a diamide-based commercial insecticide (flubendiamide 480 SC) against H. armigera larvae on tomato plants [72]. A synergistic interaction between NPV and azadirachtin (AZA) against H. armigera has also been reported [73]. Field investigations [74] were undertaken to determine the efficacy of N. rileyi, HaNPV and neem seed kernels for the management of the tomato fruit borer, H. armigera. Spraying (N. rileyi + HaNPV + neem) proved significantly superior to all other treatments in reducing the fruit damage.

Nucleopolyhedrovirus (NPVs), of the family Baculoviridae, is another microbial agent for the management of H. armigera on various host crops [8]. In our experiments, the sole application of HaNPV proved effective but least compared with chlorantraniliprole. Previously, NPV as an effective microbial insecticide has been reported in H. armigera [75]. Recently, Eroğlu et al. [76] reported a novel NPV strain from Turkey (HearNPV-TR) that showed promising results agent against H. armigera and other Helionthinae species.

We reported that the combined use of HaNPV and B. bassiana with chlorantraniliprole was more effective than individual treatments of chlorantraniliprole, HaNPV or B. bassiana in reducing the density of H. armigera larvae throughout the growing season and increasing tomato yields. Although the biorational and chemical control agents were effective in reducing fruit damage caused by H. armigera, in terms of the cost-benefit ratio, the application of HaNPV with chlorantraniliprole was superior to the other treatments, followed by B. bassiana with chlorantraniliprole.

A possible explanation for the superiority of the combined applications of either HaNPV with chlorantraniliprole or B. bassiana with chlorantraniliprole in reducing larval populations and fruit damage could be due to the quicker mode of action of chlorantraniliprole and a bit slower period of action of the microbial agents, providing protection earlier in the season against the feeding damage of flower buds and then developing fruits. The insecticide component may provide an initial level of mortality while the fungal pathogen or baculovirus have a lag period between infection/ingestion and insect death [77].

The efficacy of HaNPV, B. bassiana and an insecticide for the control of H. armigera in tomato has been reported by various researchers. Previous studies on the sustainable management of H. armigera have indicated the potential for using HaNPV, B. bassiana and plant extracts alone or in combination with a compatible insecticide. Controlling insect pests through the simultaneous use of both an insecticide and an insect pathogen is a most promising strategy [78]. The insecticidal efficacy of a binary co-occluded mixture (HearSP1B:LB6) of H. armigera single nucleopolyhedrovirus (HearNPV) was effective in greenhouse- and field-grown tomato crops in Spain and Portugal [79] and its efficacy and persistence were comparable with those of commercial insecticides in both field and greenhouse tomato crops. HearNPV has shown superior control of H. armigera compared with the standard chemical insecticide, endosulfan, or Bacillus thuringiensis, in two successive years [80]. Field experiments on the management of H. armigera in field tomatoes showed that the different sequential application of HaNPV, Bacillus thuringiensis (var. kurstaki) (Btk), azadirachtin and spinosad provided effective control against the larval population of this insect [77], which is in line with our findings. Gupta et al. [81] reported that the application of HearSNPV alone at 3.0 × 10¹² OBs/ha in tomato resulted in the significant suppression of the pest. Moreover, HearSNPV was compatible with the recommended insecticide endosulfan, Bacillus thuringiensis and egg parasitoid, Trichogramma pretiosum (Riley) in both combined and sequential applications. Similarly, the same HearSNPV was effective against H. armigera when the laboratory data were validated in a large tomato field [82]. Here, SNPV also proved economical and compatible with bio-agents such as the entomopathogenic fungus M. anisopliae, the parasitoid, Campoletus chloridae (Uchida) and trap cropping with marigold (Tagetes patula L.) when evaluated at three locations in a one ha field. Our results agree with Kumar et al. [83], who reported that treatment with biorational insecticides (B. thuringiensis, B. bassiana, azadirachtin and nucleopolyhedrovirus) significantly reduced pod damage by H. armigera. Similarly, IPM experiments conducted at two locations in Guam (Yigo and Inarajan), USA, significantly reduced fruit damage by H. armigera in plots treated with sprays of selective insecticides (petroleum spray oil, B. bassiana, azadirachtin and B. thuringiensis) compared with plots treated with carbaryl, malathion or the control treatments at both locations [84]. Raj et al. [85] tested different chemical and microbial insecticides and the results showed that among the chemical insecticides, the lowest larval population of H. armigera was recorded in the rynaxypyr treatment while the minimum larval population and pod damage was found in HaNPV treated plots. The results of field trials conducted by Manoj et al. [86] against H. armigera in chickpea are also comparable with our studies; the authors showed that the smallest larval population was in the plot treated with B. bassiana followed by HaNPV. Wakil et al. [48] reported that combinations of NPV with A. indica and chlorantraniliprole caused higher mortality, reduced pupation and had an additive effect against the different larval instars of H. armigera compared with their application singly under the laboratory conditions. The mortality of H. armigera larvae on diet into which Bt and NPV have been incorporated was higher than when each pathogen was applied alone [33].

The natural enemy complex recorded in tomato fields is comprised of a diverse complex of predators, parasitoids and spiders [87]. In our study areas, C. carnea, ladybirds and spiders were the most abundant predators. The numbers of all the natural enemies were highest in untreated control plots and lowest in the treatments with combinations of HaNPV with chlorantraniliprole, or B. bassiana with chlorantraniliprole. The reasons for the outcome observed in the natural enemy response are not clear as we did not directly observe the mortality of the natural enemies. Additionally, it may be important to note that the direct toxicity of some of the treatments to the natural enemies is unlikely, e.g., HaNPV is not pathogenic to the natural enemies tested. Thus, it might have been expected that treatments with the insecticide chlorantraniliprole would cause the greatest reduction in predatory insect populations [88,89]. Conceivably, observing fewer natural enemies in treatments with HaNPV or B. bassiana could be related to lower numbers of H. armigera larvae as potential prey/hosts for mobile predatory arthropods [77,90]. Significant associations have been found between members of predatory species and the infestation levels of insect pests in the Honduran highlands, Central America [87].

The estimation of cost-benefit ratios (CBR), or marginal rates of return (MRR), is considered to be a standard method for measuring the outcome of investment for a particular process or treatment. If CBR or MRR are greater than zero, then there would be a net gain on investment, and it would be economical to include the method in a control strategy [80]. In the present study, HaNPV or B. bassiana in combination with chlorantraniliprole were found to have the highest cost:benefit ratio (CBR) based on tomato yield. Cherry et al. [80] reported the use of insecticides or HearNPV gave the maximum return compared with other approaches for managing H. armigera in tomatoes. However, they only estimated the cost of production for applications of individual control agents rather than combinations of different approaches. Furthermore, improved production techniques and lower costs have recently meant that HaNPV is much more economically viable for field use [62].

5. Conclusions

Effective IPM requires a number of different control options so there is minimal reliance on just one method. Furthermore, the over-use of conventional insecticides increases selection pressure on target pests for resistance; combining insecticides with microbial agents can help to delay resistance and the resurgence of pests. Our findings suggest that a combination of HaNPV and chlorantraniliprole is economically viable for the control of H. armigera on tomato. The combination of B. bassiana and chlorantraniliprole also indicated substantial promise. Future research can be directed toward optimizing timing and the rates of applications including a diverse array of strategies within an IPM context.