This 1st annual research summary highlights many of the recent and on-going activities in biological control in Arizona. It is our hope that this document will broaden awareness of biological control activities to agricultural producers, extension personnel, and others throughout the state. We also hope that these brief summaries will encourage interested parties to seek further detail from individual scientific investigators or research teams. We strongly believe that biological control can be a significant component of overall pest management in Arizona crops.

Integrated pest management (IPM) is use of several diverse tactics (tools) to control arthropod pests in plants, shrubs, and trees important to man and animal. Some common methods used in IPM are familiar to most growers, e.g. selection of seed/rootstock resistant to particular pests, fire, crop rotation, and chemical control. The use of biological agents for pest control is generally less familiar e.g. lady bird beetles as predators, parasitic wasps or flies, and disease agents -- deliberately spread or encouraged to suppress pests. An important challenge to the practical application of IPM is to utilize both biological and chemical control in a compatible manner in the field to achieve true IPM, .e.g., bridge biological control with chemical control.. Bio-insecticides (biorationals) need to be incorporated in such regimes. Not only to help prevent insecticide resistance but to minimize detrimental insecticidal impact on beneficial arthropods and promote biological control of SLWF, (Akey 1992, 1993, Henneberry and Butler 1992)

Some fungi (Entomopathogenic) may be used against pests as disease agents. When such products are spread about to deliberately control pests, they are a form of biological control and also fall within the definition of being bio-insecticides and more broadly biorational insecticides.

We have been using entomopathogenic fungi against the silverleaf whitefly and have obtained control in upland and pima cotton in Arizona and in furrow and drip irrigation systems (Akey et al. 1994). In 1997 we conducted trials in upland cotton with two genera of entomopathogenic fungi from three different companies:

Beauveria bassiana Naturalis-l^TM Troy

Beauveria bassiana Mycotrol^TM Mycotech

Paecilomyces fumosoroseus PFR-97^TM Thermal-Trilogy

Our primary objectives are to identify efficacious products that can be used by growers within the practical restraints of application and cost, to demonstrate such products to the agricultural community, particularly extension personnel, and encourage the necessary technology transfer essential for the acceptance and use of these products.

All three entomopathogenic fungi effectively controlled silverleaf whitefly in furrow-irrigated cotton and were equal to or better than "best agricultural practices" that included insect growth regulators and conventional insecticides.

The portion of the work being conducted by our laboratory that concerns biological control centers around the characterization of flight behavior of whitefly parasitoids. Early indications are that in a vertical flight chamber female flight typically lasts three times as that of males. Both sexes fly on multiple days. Flight by 21 day-old females has been observed. Eretmocerus only minimally responds to visual cues during flight.

¹USDA-ARS-Western Cotton Research Laboratory

²USDA-APHIS-Phoenix Plant Protection Center

A temperature-based mathematical model that simulates Bemisia biotype B ( = B. argentifollii Bellows and Perring), parasite, and predator population dynamics (BIOCONTROL-WHITEFLY) has been constructed. The model, called BIOCONTROL-WHITEFLY, also simulates pesticide applications and augmentative releases of parasites and predators and their effects on Bemisia populations. Predictions from BIOCONTROL-WHITEFLY on Bemisia development and population growth at various temperatures on cotton are comparable with literature values (Table 1). The model predicts that temperature and migration are the most important factors influencing Bemisia population growth and the potential efficacy of biological control agents. Simulations can be used to compare the efficacy of various parasite and predator species both alone and in combinations for reducing Bemisia populations. Simulations where parasitism and predation rates were equivalent indicate that extraordinary numbers of predators per plant (>10) are required to achieve levels of whitefly population control comparable to that obtained with initial populations of Eretmocerus of 1.95 females per plant. However, the lowest whitefly populations occurred when combinations of predators and parasites were simulated even when predators could feed on parasitized immatures.

Table 1. Comparisons of actual Bemesia tabaci development times and predictions generated by the BIOCONTROL-WHITEFLY model.

* Actual and predicted development times are significantly different at p=0.05 as determined by a two-tailed t-test.

Investigation of digestive enzymes produced by nematodes and symbiotic bacteria carried within nematodes (Families Steinernematidae and Heterorhabditidae) have identified several key enzyme groups responsible for the breakdown of insect tissues. Namely: type I collagenase, chondroitinase, tripsinase and various other proteases. Further work is continuing to ascertain the exact sequence of events that leads to the breakdown of tissues.

Temporal association of entomopathogenic nematodes and bacteria

Galleria mellonella larvae were infected with ten species/strains of Steinernema spp. and Heterorhabditis spp. in Petri dishes. Every six hours over a 48-hour period, larvae were dorsally dissected. Hemolymph was collected with a sterile loop and streaked on tryptic soy agar plates. Each of the ten nematode strains was incubated at 22°C, 27°C and 32°C. Subsequently isolated bacterial colonies were grown for 48 hours at 27°C.

Bacterial contaminants of insect origin were identified as Salmonella subspecies 1 G and Xanthomonas campestris PV visicatoria B. Following nematode application, several other bacterial species were identified from nematode infected insect cadavers: Enterobacter gergoviae, Vibrio spp., Pseudomonas fluorescens type C, Serratia marcescens, Citrobacter freundii, and Serratia liquefaciens/grimesii.

Between 24-30 hours incubation, the primary symbiont occurred almost exclusively. Few agar plates during this period grew any other cultures with the exception of H. megidis (M-145). The primary symbiont from H. megidis (M-145) disappeared rapidly being replaced at 24 hours with Serratia liquefaciens/grimesii. After 30 hours of incubation the initial insect contaminants occurred once more. Suppression of these organisms occurred during 24-30 hours, however the bacteria remained viable. Secondary bacterial associates generally appeared outside the 24-30 hour window, often before the primary symbiont. Galleria mortality usually occurred within the first 18 hours. At higher temperatures the primary symbiont developed earlier, with the exception of symbionts from S. feltiae (St. 27) and H. megidis (M-145). Two strains of nematodes carried only the primary symbiont: S. riobravis (St. 355) and H. bacteriophora (Cruiser). Both nematodes came from monoxenic in vitro produced commercial products. We conclude that not all Galleria mortality in entomopathogenic nematode bioassays is caused by the primary bacterial symbiont.

Immunoflurescence and ELISA based studies confirm the production of a polyclonal anti-body for Xenorhabdus genera of bacteria. This will be used to study the persistence of bacteria in the soil.

Behavioral studies involving egg laying of Pectinophora gossypiella on Gossypium barbadense, and transgenic Bt. New cotton 33 G. hirsutum. The collection of bolls throughout the season and monitoring of entrance holes made by neonate larvae has indicated consistent preference for egg laying to take place on transgenic G. hirsutum cotton. Boll infestation levels in G. barbadense have remained low despite high numbers of moths caught in delta traps.

Cotton fields were treated with the entomopathogenic nematode, Steinernema riobravis, and Vydate® L for the control of plant parasitic nematodes. Short staple cotton grown near Coolidge, Arizona, was treated at a rate of 1 billion and 2 billion S. riobravis nematodes per acre, and 0.5 lb a.i. Vydate® L per acre. Untreated cotton received an application of water only. Treatments were applied through a subterranean drip system with 12 inch spaced outlets. Applications were made in the daily irrigation cycle of 0.33 inches of water, normal irrigation cycles followed.

Products were uniformly distributed over the treated fields. Entomopathogenic nematodes persisted throughout the 6 week experimental period at the 1 billion per acre rate. However, nematodes applied at 2 billion per acre rate disappeared rapidly. Populations of various plant parasitic nematode species were monitored subsequent to treatment application. Nematodes were extracted using a standard sugar flotation technique and counted in 1 ml slide samples. Both Meloidogyne incognita and Tylenchorhynchus spp. populations were reduced by S. riobravis applied at 1 billion per acre rate. Phytoparasitic nematodes were reduced following application of Vydate® L, but control was not sustained beyond one week.

At pre-sowing irrigation (mid-March), cotton fields were treated with two entomopathogenic nematode species; Steinernema riobravis and S. carpocapsae for control of diapausing Pectinophora gossypiella larvae. Pima S-6 cotton fields situated in Fort Hancock, Texas were treated at a rate of one billion nematodes per acre. Caged, diapausing larvae were buried in fields at a depth of one inch, in row tops and furrow bases. Nematodes were applied with a spray rig, fixed winged aircraft, or in furrow irrigation via a constant flow, battery box. Fields were irrigated after ground application, prior to aerial spraying and during furrow application. Caged larvae were recovered 48 hours after nematode application.

All application methods resulted in uniform distribution of nematodes over the treated fields. No significant differences in larval mortality between nematode species or application method could be determined. However, aerial and furrow application methods gave consistently better parasitism of larvae compared to ground rig delivery. Larval mortality in cassettes buried in furrow bases was significantly higher than in row tops. Larval mortality ranged from 53.26-79.14%. Both nematode species could be recovered 50 days post application.

At pin-head square Frustrate® PBW pheromone bands (biosys, Inc.) were applied at 100 bands per acre placement rate (16 g a.i./acre), giving a target release of 115 mg gossyplure/acre/day. Capillary gas chromatography was used to analyze bands throughout the growing season. A uniform release profile indicated sufficient release of pheromone for 144 days after placement.

Pink bollworm mating disruption was monitored in three ways: 1. Delta 2 traps were positioned throughout the farm, forming a continuous trap line. Significantly larger numbers of moths were recovered form untreated zones. 2. Virgin female moths were placed in mating stations at dusk. At sun rise moths were collected and later dissected for spermatophores. Significantly higher mating activity occurred in untreated fields (p=0.000). 3. Green bolls were collected at random and examined for larvae. Significantly higher infestation levels existed in untreated zones.

At harvest (November), seed cotton yields were weighed using trailer scales. Higher yields were recovered from pheromone (1,864 lb/acre), and pheromone + nematode fields (1,712 lb/acre), than control fields (1,450 lb/acre). However, due to large variations between fields, the differences were not statistically significant (p=0.436).

Release of Exotic Parasitoids for Establishment in AZ

The USDA silverleaf whitefly biocontrol project has identified several key crops that support the growth of whitefly populations but do not harbor a large fauna of parasitoids. Extensive laboratory and field cage trials have identified parasitoids that perform better than the native Eretmocerus on both cole crops and melons. These parasitoids (all Eretmocerus species) were the target of introduction efforts in Arizona for 1997.

Agricultural settings, while they are an ultimate target of the parasitoids we are attempting to establish, are harsh environments and inappropriate for inoculative releases. We felt that release of parasitoids in residential areas nestled among agricultural fields was a better approach. Home gardeners planted crops that support whitefly populations throughout the year. Parasites were released once every three weeks after whitefly nymphs became plentiful. Over 3 million Eretmocerus from Pakistan, Israel, and the United Arab Emirates were released in seven general areas of Arizona: 1) Coolidge/Casa Grande, 2) Gila Bend, 3) Colorado River, 4) Northwest Phoenix, 5) Southwest Phoenix, 6) Northeast Phoenix, 7) Southeast Phoenix.

All three species of Eretmocerus were recovered, although not all species were recovered at all the sites. Releases will continue this fall to selectively release parasites at sites where establishment has not been documented.

Parasitoids in the genus Eretmocerus from Israel, Spain, and Pakistan were released at 25 sites throughout Arizona in 1997. These exotic parasitoids exhibited within season reproduction at many of the sites. The parasitoids were released in home gardens in residential areas where Hibiscus, Lantana, and Cape Honeysuckle could provide overwintering sites for the parasitoids. It is hoped that parasitism of whiteflies on these ornamental plants not only provides a source of parasitoids for agriculture on a yearly basis, but that whitefly mortality will increase once exotic parasitoids have established. The result would be fewer whiteflies migrating to agricultural crops from residential areas in the spring.

To test this hypothesis, we initiated a life-table study to look at mortality factors affecting populations of whiteflies on Hibiscus. We will conduct the study at three release sites where exotic parasitoids are known to be reproducing. Three control sites will also be monitored. The control sites are ¼ mile from the release sites. We plan to repeat life-table analysis at these sites to capture the buildup of exotic parasitoids at the release sites and the eventual dispersal of exotic parasitoids to the control sites.

The silverleaf whitefly, Bemisia argentifolii, is a serious pest in many cropping systems in the desert southwest. Controlling this pest in any given crop or in a specific field is difficult, however, given the highly polyphagous nature of this insect and its high reproductive and migration capabilities. Utilizing the potential of natural enemies to control the whitefly is also problematic because insecticides used to control whiteflies and other pests usually also kill beneficial insects and disrupt natural control. Three of the major crops grown in the desert southwest (alfalfa, melons, and cotton) support whitefly populations during part of the year. Alfalfa is also a source of Lygus populations that can damage cotton. We propose that to effectively control whiteflies in cotton, we need to look at the entire agricultural system (both multiple crops and multiple pests) and utilize the most advanced technology that is the least disruptive to natural enemies. This includes 1) early season inoculation of whitefly parasites in melons, 2) Bt cotton or pheromones for controlling pink bollworm and other Lepidoptera in cotton, and 3) controlling Lygus in alfalfa with egg parasites and leaving small strips of alfalfa on border edges to reduce the numbers migrating to cotton.

Our ultimate goal is to control pests in cotton fields in an economical manner that utilizes the action of natural enemies to the maximum extent. Because several key cotton pests come from alfalfa and melons, we plan to control these pests in the alternate crops to reduce migration into cotton. It is therefore crucial that the alfalfa and melon fields be right next to the cotton field so that any control we do in these fields has an effect on insects migrating into the study cotton field. There were 5 general locations where we conducted this research. At each location there was one group of fields (alfalfa, melons, cotton) treated under an IPM regime and one group of fields treated under a conventional regime.

Poinsettia is a major greenhouse crop attacked by the silverleaf whitefly. Research has been conducted looking at the performance of the native Eretmocerus and some Encarsia species on poinsettias, and studies of the effectiveness of exotic Eretmocerus on field crops have been conducted. The goal of this project is to study the capacity of several exotic Eretmocerus species to locate and kill whiteflies on poinsettia. These parasites were compared to the native Eretmocerus.

Only 8 of 24 replicates have been completed at the time of printing. More parasite progeny were produced by the Eretmocerus species from Spain, Pakistan, and Unit Arab Emirates than by the Eretmocerus from the United States. There was no evidence from the number of dead whitefly nymphs that levels of host feeding were different among the species.

Aphilinid parasitoids are frequently released using several methods, however their fate after release is usually not known. Pupae can be subject to predation and adults can disperse or die prior to being able to parasitize hosts in the field. Ultimately, one wishes to use a release method that maximizes the amount of parasitism resulting from the release. Our goal in this project was to test four methods of releasing parasites and compare the resulting amount of parasitism. Four release methods were tested: 1) release of parasitoid pupae on leaves, 2) release of parasitoid pupae in vials, 3) release of parasitoid adults, and 4) release of parasitoid pupae on living plants (we used Hibiscus).

We have conducted this study twice, for a total of six replicates per treatment. We will count the parasitoids and whiteflies on the currently frozen leaves this fall and will run a final three replicates this spring.

The populations of whitefly predators and parasitoids were monitored in replicated field plots subjected to different whitefly management systems. Management systems included two whitefly growth regulators, a rotation of conventional insecticides, and an untreated control. All insecticides were applied according to recommended thresholds. Additionally, a study was conducted to evaluate the sub-lethal effects of whitefly growth regulators and conventional insecticides on predator foraging behavior. Gut contents of over 33,000 predators representing over 25 genera were evaluated for whitefly prey remains using a whitefly monoclonal antibody-based enzyme-linked immunosorbent assay (ELISA).

Laboratory and field studies were conducted to evaluate the efficacy of marking parasitoids and predators with a protein prior to ELISA. Data indicate that this marking technique is superior to conventional marking techniques. A field study was conducted to monitor the dispersal patterns of protein-marked Eretmocerus (a whitefly parasitoid) released into the center of a cotton field and recaptured at various time intervals.

A laboratory study was conducted that evaluated the feeding behavior of five insect predators when exposed to the various life stages of whitefly. The predators examined included: Collops vittatus, Geocoris punctipes, Orius tristicolor, Drapetis sp., and Hippodamia convergens. The behaviors monitored included; feeding (eggs, nymphs, and adults), resting, grooming, walking, and probing. The egg and immature stages of whitefly were not preyed on as frequently as the adult stage.

A laboratory study was conducted to examine the efficacy of a novel immunomarking technique on Anaphes iole Girault, a minute parasitoid of Lygus spp. eggs. Adult A. iole were marked with the readily available mammal protein, rabbit immunoglobulin G (IgG), by three different application methods. Adult parasitoids were marked internally by feeding them a honey solution "spiked" with rabbit IgG and externally by contact exposure or topical mist. Marked individuals were then assayed using a sandwich ELISA for the presence of the IgG marker using an antibody specific to rabbit IgG. Data indicate that the IgG marker was retained throughout the entire adult lifespan in almost (98.9%) every individual parasitoid assayed, regardless of the application method used.

Five different immunoassay formats were examined for their ability to detect a minute quantity of prey remains in predator guts. The convergent lady beetle, that had consumed either one or five pink bollworm eggs was evaluated by the following immunoassays: three variations of ELISA, a dot blot, and a western blot. Sandwich ELISA, dot blot, and western blot were the most sensitive immunoassays based on the proportion of individual predators scoring positive for prey remains. The direct ELISA and indirect ELISA were ineffective at detecting prey in the predators.

A field cage study was conducted that compared the retention time between a novel immunolabelling mark-release-recapture marking technique with the more conventional insect marker, Day-Glo dust. Commercial-purchased convergent lady beetles were marked with either a rabbit immunoglobulin G (IgG) or a chicken IgG solution. The beetles were then released into separate field cages, recaptured daily, and assayed by both sandwich ELISA and direct ELISA for the presence of IgG markers. A third group of lady beetles were marked with Day-Glo dust, released into a cage, recaptured daily, and examined under a dissecting microscope for the presence of the Day-Glo marker. Data indicate that the predators marked with the IgGs had a much longer retention time than those marked with Day-Glo. Additionally, the rabbit IgG had a greater retention time than chicken IgG.

The gut contents of three species of insect predators that were either fed variable amounts of pink bollworm eggs or a fixed amount of eggs but held at variable time and temperature regimes were assayed by an indirect ELISA. The sensitivity and efficacy of the ELISA was dependent on the predator species examined. Small predators were more immunoresponsive to the ELISA than large predators. Furthermore, the assay sensitivity was dependent on the number of prey consumed, on the elapsed time after feeding, and on the temperature at which the predators were held. The smaller predator species retained recognizable traces of prey remains for longer periods than larger predator species. The ELISA efficacy decreased with increasing ambient temperatures. A series of regression equations have been developed to estimate the median detection interval of prey in a predator's gut that takes into account the predator species examined, the amount of the prey consumed, and the ambient, after-meal temperature.

In qualitative predator gut content immunoassays, the sensitivity of the immunoassay is important for determining whether a predator contains a targeted prey antigen. If the immunoassay employed is insensitive, the probability of obtaining a false negative reaction is high. The sensitivity of an indirect ELISA developed to detect pink bollworm egg(s) in whole-body homogenized predators varied in efficacy between species. Specifically, the indirect ELISA was more effective at detecting egg antigen in small predators than in large predators. In this study, we examined the effect the predator:prey protein ratio has on the sensitivity of an indirect ELISA. Our results suggest that when assaying whole body homogenized predators, care must be taken not to overload an ELISA microplate with non-target (predator) proteins. Predator samples should be diluted to less than 125 µg of total protein per sample. Any protein concentration above 125 µg per ELISA microplate well will likely result in an ELISA false negative reaction. In another experiment, we compared the efficacy of an indirect ELISA with a dot blot immunoassay. Adult lady beetles that had eaten one pink bollworm egg were homogenized in variable dilutions of phosphate buffered saline (PBS) and each sample was analyzed using both immunoassays. The dot blot immunoassay was more reliable than the indirect ELISA for detecting the presence of minute traces of egg antigen in the samples. Generally, the volume of PBS that the beetles were homogenized in had little effect on the qualitative outcome of the dot blot. However, the indirect ELISA was more effective when the beetles were homogenized in larger volumes of PBS. This suggests that the efficacy of an indirect ELISA can be improved for large, protein-rich predators by grinding them in larger volumes of PBS, thus minimizing the total protein in a given sample.