The Black Pearl Pepper Banker Plant for Biological Control of Thrips in Commercial Greenhouses

Laboratory experiments:

Study Organisms
All O. insidiosus were purchased from Koppert Biologicals (Howell, MI), and used for experiments within one week of arrival. When not being used, O. insidiosus were kept in a refrigerator set on its least cool setting. Black Pearl pepper plants (128 plugs/flat; plants <7cm in height) were supplied by Van Wingerden International (Mills River, NC) and C. Raker and Sons Inc (Litchfield, MI). Plants were transplanted to 10.2 cm pots and allowed to grow for eight weeks until they were transplanted to 11.36 liter pots. Fafard 2 light weight soil mix was used for all plants with 473 ml of Osmocote/soil bag (14-14-14; 2.8cu ft or 0.85cu meters). Western flower thrips, Frankliniella occidentalis Pergande, were obtained from a colony started from field collected adults and maintained in the laboratory on green beans and cabbage.

Effect of flowers and thrips on female O. insidiosus longevity
To determine if Black Pearl pollen, thrips prey, or a mixed diet of both increase O. insidiosus adult longevity, we conducted a 2x2 factorial laboratory experiment that crossed two pollen treatments (absent or present) with two thrips treatments (absent or present). The experiment was conducted in arenas made from plastic 50mL Corning vials (Corning, Corning, NY). One hole (2 cm diameter) was made in the top and side of each vial for ventilation. Holes were covered with thrips screen and secured with hot glue to prevent escape of experimental organisms. A third smaller hole (0.7 cm diameter) was made through the bottom tip of the vial. A schematic of the vials is presented in Figure 1.
Black Pearl pepper stems were collected from greenhouse pepper plants. Each stem was 6-8 cm in length and had three leaves. All buds and flowers were picked from ‘flower absent’ treatment stems. Buds and 1-2 open flowers were left on stems in ‘flower present’ treatments. The bottom-half of each stem was wrapped in cotton then pulled through the small hole in the bottom of vials so that the stem fit snugly in the hole. The cut end was inserted into a #55 (12.7cm, 10 ml) floral-pick (Syndicate Sales Inc., Kokomo, Indiana) filled with tap water. Floral-picks were stuck into a Styrofoam board and a wire-grid was made from craft wire in order to support the experimental vials in an up-right position. An example of this set-up is presented in Figure 1. To establish ‘thrips present’ and ‘thrips absent’ treatments, 20-30 adult thrips were added to ‘thrips present’ treatment vials. Then one female O. insidiosus was added to each vial. 10 replicates of each treatment were conducted simultaneously. The entire procedure was repeated yielding 20 replicates of each treatment.
O. insidiosus survival was determined daily by gently tapping and turning each vial until the female was found alive or dead. Living females were transferred to a new vial, set up as described, twice a week so O. insidiosus would have fresh plant and insect material for refuge and food. Vials were placed in an incubator at 27-28? C and 55-60% RH for the entirety of the experiment.
Statistical Analysis. ANOVA was used to test the effect of pollen, thrips, and their interaction on total number of days females survived (SAS, 2008).

Effect of pollen and thrips on development of O. insidious nymphs
To determine how pollen and thrips prey affect development of O. insidiosus nymphs, we conducted a factorial laboratory experiment that crossed two pollen levels (absent or present) with two thrips levels (absent or present). Experiments were conducted using O. insidiosus nymphs within 24 hours of hatching. O. insidiosus nymphs were reared similarly to Kiman and Yeargan (1985) procedures. To obtain eggs, adult O. insidiosus were placed in a plastic container with two moist cotton balls for water, Heliothis subflexa Guenée (Lepidoptera: Noctuidae) eggs (obtained from NCSU colony) for food, and pole beans or green beans as an oviposition substrate. The adult colony container was placed in an incubator at 27-28? C and 55-60% RH so adult O. insidiosus could lay eggs. Every 2-3 days, beans were inspected under a stereo dissecting scope for eggs. Sections of bean that contained eggs were removed and placed on filter paper in 5 cm petri dishes and placed back in the incubator. Petri dishes were inspected for newly emerged nymphs daily.
New nymphs were placed into an experimental arena. Experimental arenas were 5 cm petri dishes that contained a piece of filter paper moistened with 1-2 drops of tap water. A disc of Black Pearl pepper leaf was cut out using a 2 cm cork borer, soaked in tap-water to regain turgidity, and placed on the filter paper. To establish ‘thrips present’ treatments, 10 second instar thrips were placed in each petri dish. To establish ‘pollen present’ treatments, eight anthers from Black Pearl pepper flowers were added to each petri dish. Anthers were obtained by picking open Black Pearl pepper flowers that had pollen producing anthers and cutting the anthers out with dissecting scissors. This was done once a week or as needed and anthers were stored in a sealed plastic container in a refrigerator. As a precaution, the plastic container was filled with CO2 and sealed for 20 minutes in order to kill any small arthropods, such as mites, in or on the anthers. Depending on the treatment, dishes were replenished with 4-5 anthers and 5-10 thrips (depending on how many were still alive from the previous day) so nymphs did not run out of food. Every two-three days nymphs were placed in new arenas with a new leaf disc and filter paper. Since nymphs were emerging daily, a few nymphs at a time were assigned to each treatment until a minimum total of 20 nymphs had been assigned to each treatment. All petri dishes were placed in an incubator at 27-28? C and 55-60% RH for the entirety of the experiment. Arenas were inspected daily with a dissecting scope to record nymphal survival. Observation of nymphal development lasted until all nymphs had died or matured into adults.
Statistical Analysis. A chi-square test was conducted to compare the total number of O. insidiosus completing development among the treatments. ANOVA was used to test for significant effects of pollen, thrips, and their interaction on the number of days nymphs took to complete development from egg hatch to adult eclosion (SAS, 2008).

Effect of pollen and thrips on O. insidiosus adult size
To determine the effect of pollen and thrips on adult O. insidiosus size we conducted a laboratory experiment with two treatments: ‘thrips’ and ‘thrips + pollen’. These were the only treatments that enabled O. insidiosus to complete development in the previous experiment. O. insidiosus was reared from first instar to adult according to the procedures described in Section 2.3. Once nymphs had matured into adults, they were sexed, and the length of their hind tibia measured.
Statistical Analysis. A t-test was used to compare the length of adult female and male hind tibias among treatments (SAS, 2008).

Effect of flowers and thrips on O. insidiosus abundance
To determine how flowers affect O. insidiosus abundance on Black Pearl pepper banker plants, we conducted a greenhouse experiment with individually bagged peppers as experimental units. Two treatments were created by allowing plants to flower versus continuously picking buds and flowers from other plants. There were a total of 40 plants, or 20 replicates per treatment.
Plants were dipped in insecticidal soap (Safer Brand, Lititz, PA) every two days for one week to remove thrips and other arthropods prior to establishing either ‘flower present’ or ‘flower absent’ treatments. One week before the experiment began, individual plants were placed in a bag made from organdy fabric (Jo-Anns Fabric, Raleigh, NC; 60cm wide x 121.9cm tall). Two 91.4 cm bamboo stakes were inserted on the outer edge of the pot to support the bag. The top of the bag was sealed by twisting the fabric and securing it with a large binder clip. Plants remained in bags for the entirety of the experiment.
Plants in the ‘flower present’ treatment were encouraged to flower by picking all peppers and dying flowers from each plant. To establish ‘flower absent’ treatment plants, open flowers and buds were removed from each plant. This procedure was repeated the day before the experiment began and twice per week for the duration of the experiment to maintain the ‘flower present’ and ‘flower absent’ treatments. After both flower treatments were established, 20 adult thrips were also added to each plant so that thrips could lay eggs on the plants and build populations in the absence of predators. Plant material removed before the experiment began was discarded. On week zero, one week after treatments were established and thrips were added to plants, 20 adult O. insidiosus (male:female ratio of 1:1) were added to each plant.
Plants were sampled for arthropods once a week. Each plant was beaten a total of four times for four seconds each time into a 33 x 40.5cm white tray. O. insidiosus adults and nymphs, and thrips were counted and collected using an aspirator. Plants were then visually inspected to count open flowers from ‘flowers present’ treatment plants and remove buds or open flowers from ‘flowers absent’ treatment plants. All trapping materials on the beat tray, picked peppers, flowers, buds, and aspirated insects were placed back on their respective plant. This was done so that any insects that were not aspirated, or eggs that had been laid on foliage, would be put back onto plants. This experiment ran for a total of six weeks from 6 June 2011 – 19 July 2011.
Statistical Analysis. A repeated measures ANOVA was used to test the effect of flowering and non-flowering plants on O. insidiosus adult and nymphal abundance and thrips abundance over a course of six weeks. Data for both nymphs and thrips was log (x+1) transformed to correct for non-normal distribution (SAS, 2008).

Greenhouse Experiments:

Effect of banker plants on pest abundance
To determine how Black Pearl pepper banker plants affect pest and natural enemy abundance compared to augmentative releases, we conducted an experiment at Hoffman Nursery in Rougemont, NC, which produces native and ornamental grasses in plastic covered hoop houses. Hoffman Nursery was selected for these experiments because they had been implementing the Black Pearl pepper banker plant system for two growing seasons and presently manages several of their hoop houses each season with biological control, using purchased natural enemies and compatible insecticides. The hoop houses used in these experiments were all 6.7m wide but had different lengths ranging from: smallest (25.6m) to largest (38.4m). Sides of the houses were rolled up in summer to maintain a cooler temperature. Each house had four rows of grasses in 30.48cm x 60.96cm flats with 32 grass plugs per flat. Each row was two flats wide (1.22 m) and as long as the houses.
We set up two treatments (Augmentation and Banker Plant) and one control replicated four times in 12 hoop houses. Augmentation and Banker Plant houses were treated the same as Control houses except that we released one bottle of 500 count O. insidiosus (Koppert Biologicals, Howell, MI) per house every three weeks (May 25th, June 15th, and July 7th, 2010). O. insidiosus were released by sprinkling them evenly on plant flats moving lengthwise from one end of the house to the other. Banker Plant houses had Black Pearl Pepper banker plants (2-3 feet tall; 9” pot) spaced six per row at alternating lengths within the house (Fig. 2) resulting in 24 banker plants per banker plant treatment house, with the exception of the smallest house which only had five banker plants per row for a total of 20 banker plants. To encourage continuous flowering, we picked peppers from each banker plant in banker plant houses for 1-2 minutes each week. Houses were not treated with synthetic insecticides but were spot-treated occasionally with insecticidal soap for aphids, spittle bugs, and leafhoppers which are not pests targeted by O. insidiosus. All houses were treated weekly with a rotation of fungicides (azoxystrobin, triazzole, iprodione, Trichoderma harzianum strain T-22, thiophanate-methyl, myclobutanil, polyoxin, phenylamide, strobilurine, carboxylic acid amide, mancozeb).
Arthropod sampling was conducted once a week for nine weeks (26 May – 20 July, 2010) by vacuuming plant flats using a modified Husky Blower Vac (Husqvarna, 125 BVX Series) fitted with organdy (Jo-Anns Fabric, Raleigh, NC) bags to catch insects. One randomly selected 1.22m x 1.83m section of grasses was vacuumed in each row (four per house) for a total of 8.93 m2 per house. The four samples from each house were vacuumed into a single organdy bag. Bags were placed in a sealed jar in the freezer for 24 h to kill arthropods. Then the contents of each bag were emptied into a jar of 70% ETOH for later observation. Samples were viewed under a dissecting scope to count thrips and spider mites. On weeks when natural enemies were released, sampling took place the day after release.
Statistical analysis. We used a Maximum Likelihood repeated measures ANOVA to determine how banker plants and augmentative release affected the seasonal abundance of thrips and spider mites. All data were log (x+1) transformed to correct for non-normal distribution (SAS, 2008).

Banker plant sampling
Individual banker plants were sampled to monitor for O. insidiosus, spiders, and thrips. Four plants from each banker plant treatment house were sampled every week from June 8th – July 23rd by randomly selecting one plant from each of the four rows in a house and beating the plants over a 33 x 40.5cm white tray. Organisms were counted then returned to the plant.

O. insidiosus predation in the presence and absence of Black Pearl pepper pollen
To determine if the presence of pollen affected O. insidiosus predation rate of Western Flower Thrips, we conducted a factorial experiment with two flower treatments (present or absent) crossed with two predator treatments (O. insidiosus present or absent). Experimental arenas were made from 50ml plastic Corning vials according to Wong, unpublished methods.
Each arena had a 6-8 cm Black Pearl pepper stem. All buds and flowers were removed from ‘flowers absent’ treatment stems; buds and 1-2 open flowers were left on stems in ‘flowers present’ treatments. We placed 30 adult thrips in each container and then added a single female O. insidiosus to ‘predator present’ treatment vials. After 48 hours all contents of the vials were emptied and thrips were counted as alive or dead.
Statistical Analysis. ANOVA was used to test the effect of pollen, predators, and their interaction on total number of thrips remaining alive (SAS, 2008).

How spiders affect O. insidiosus retention on banker plants
To determine how spiders affected O. insidiosus abundance and retention on Black Pearl pepper banker plants, we manipulated spider abundance on banker plants at Hoffman Nursery. In follow-up experiments we measured the consumptive effects of spiders on O. insidiosus abundance and retention by restricting O. insidiosus emigration with cages. Then we measured the non-consumptive (behavioral) effects of spiders on O. insidiosus by creating ‘risk only’ spiders that had glued mouth parts and could scare but not consume O. insidiosus.
In the first experiment, we tested the abundance and retention rate of O. insidiosus on banker plants at Hoffman Nursery by conducting factorial experiment with two spider treatments (present or absent) crossed with two sampling time treatments (1 hour or 3 hours). Eight banker plants from each Banker Plant treatment house were prepared by beating them to remove spiders and O. insidiosus. Spiders were then collected from the surrounding landscape by beating grasses and banker plants. Salticidae spiders were used because they were the most prevalent spider family on banker plants at Hoffman Nursery. Twenty O. insidiosus (Koppert Biologicals, Howell, MI) were added to banker plants, followed by 4 spiders. After 1 or 3 hours, plants were sampled by beating them over a white tray (33 x 40.5cm) to count O. insidiosus and spiders. This experiment was repeated 30 June and 14 July 2010 for a total of 16 replicates per treatment.
To test the consumptive effects of spiders on O. insidiosus abundance and retention rate, we conducted another factorial experiment in a research greenhouse at North Carolina State University on July 21st, 2010. Treatments were the same as in the first experiment at Hoffman Nursery. Plants were individually caged by placing them into 60 cm wide x 121.9 cm tall organdy bags (Jo-Anns Fabric, Raleigh, NC) so that O. insidiosus could not emigrate from plants. Approximately 120 spiders were captured from natural areas on campus, 75% of which were Oxyopidae, the third most prevalent spider family on banker plants at Hoffman Nursery, and 25% were Salticidae. After adding 20 O. insidiosus to each plant, 4-6 randomly selected spiders were added to ‘spiders present’ treatment plants. Plants were sampled at 1 or 3 hours as described above to count O. insidiosus and spiders. Banker plants that were similar in size, age, and maintenance to banker plants at Hoffman Nursery were used in this experiment for a total of 10 replicates for each treatment.
To test behavioral effects of spiders on O. insidiosus abundance and retention rate on banker plants, we conducted another factorial experiment at North Carolina State University on August 21st, 2010. The treatments and sampling methods were as described above in the previous two experiments. However, plants were not in cages so O. insidiosus could emigrate from plants. ‘Risk only’ spiders with non-functioning mouthparts were used to measure O. insidiosus that were emigrating from banker plants instead of being consumed by spiders. To do this, spider mouthparts were glued prior to the experiment with ‘Liquid Bandage’ (Rite Aid Corporation, Harrisburg, PA) surgical glue that is non-toxic and dries quickly. To apply glue, spiders were rendered unconscious with CO2 and then held upside down with soft forceps to expose their mouthparts so that a small drop of glue could be applied with a camel-hair paintbrush. Once the glue was dry, spiders were placed in individual petri dishes in a refrigerator and used within 24 hours for the experiment.
For each of the above experiments, we conducted control experiments in the lab to determine if the spiders in each experiment were able to capture and consume O. insidiosus. To do this, we placed individual spiders in each of five petri dishes with five O. insidiosus per dish and checked the petri dishes at 1 and 3 hours to count eaten or dead O. insidiosus.
Statistical Analysis. Two-way ANOVA was used to test the effect of spiders, time, and their interaction on number of O. insidiosus recovered on banker plants (SAS, 2008).

• . Experimental arenas made from 50ml Corning Vial held up by a wire grid with Black Pearl pepper stems in water picks (left), and schematic of vial with hole sizes (right).
• Pollen increases fitness and abundance of Orius insidiosus Say
• Influence of banker plants and spiders on biological control by Orius insidiosus
• Banker plants at Hoffman Nursery (left, picture by Steven Frank) and banker plant layout schematic displaying four rows of grasses with alternatively spaced banker plants (right).