Yellow Starthistle biocontrol insects: There have been several species of insects discovered by the USDA that attack yellow starthistle. All of these have been found in the plant's native range, that is, in the Mediterranean region. Three species of these insects have been released in the Arizona yellow starthistle biocontrol program. These insects were supplied to ADA by the California Department of Food and Agriculture (CDFA). The species involved in this program are: (1) yellow starthistle bud weevil - Bangasternus orientalis, (2) hairy weevil - Eustenopus villosus, and (3) yellow starthistle seedhead gall fly - Urophora siranaseva. All three are host specific. This means these insects will feed only on yellow starthistle and not native plant species.

Method of attack: The two species of weevils released by ADA attack the Starthistle by the female depositing eggs in or near the developing Starthistle seedhead. Yellow Starthistle bud weevils deposit a cluster of eggs below the seedhead. Upon hatching, the larvae burrow into the seedhead base and into the developing seeds. Larvae feed on seeds as they develop. The seeds are totally consumed by the time the adult emerges in late summer. The hairy weevil deposits a single egg into the developing seedhead. Seeds are consumed by the larva as it develops. The adult weevil emerges from the seedhead in June. Few of the yellow starthistle seeds remain. The adult weevil feeds on the seedhead as well. Yellow starthistle seedhead gall fly lays its eggs directly into the seedhead, and the developing fly larvae feed on the contents of seeds. Development into the adult (pupation) occurs withing the seedhead. Adults emerge in early summer to start the process over again.

Biological control project: This invasive, toxic range weed infests 1000 acres in Pleasant Valley (Community of Young) in northern Gila County. The infested area includes pastures, roadsides, yards, and waste areas. Cooperating with the ADA is the Young Town Council (Pleasant Valley Community Council) and the USDA/Agricultural Research Service (Biological Control of Weeds Laboratory, Albany, CA). Release sites were established within Young with emphasis on abandoned pastures. Releases of weevils began in July 1993 and continued during the summer months until July 1995. One release was made of the seedhead gall fly in June 1994.

Results of biocontrol project: Monitoring of release sites in 1995 indicated that the yellow starthistle bud weevil did not become established at any site at which it was released in Young. The hairy weevil became established at two sites in Young (2 five-acre pastures). The seedhead gall fly did not survive its first year at the release site near Jakes' Corners since monitoring in 1995 revealed no evidence of the insect. Severe drought during 1995 and 1996 caused a temporary decline in the yellow starthistle population in Pleasant Valley. Combined with the high elevation of Young (5,500 feet) this drought triggered a crash in the populations of hairy weevil at the two sites they were established at earlier. Monitoring of these sites during the late summer and fall of 1996 revealed extremely low populations of hairy weevil.

Future of biological control: Decisions have been made by ADA and the Young Town Council and by two additional participants, Cooperative Extension Service (University of Arizona) and Arizona Department of Transportation (ADOT), to approach management of yellow starthistle in Pleasant Valley through adoption of a program using the principles of Integrated Pest Management (IPM). In this developing program, chemical and mechanical methods of weed control will be used as well as biological control.

Arizona Department of Agriculture and the USDA/Animal and Plant Health Inspection Service began a biological control program against diffuse knapweed in Gila County (ranch in Pleasant Valley). This weed is an aggressive, spiny invader of rangeland like its relative the yellow starthistle. A 35-acre infestation of this weed was the target of the ADA - APHIS efforts in the summer of 1994. Biological control insects were supplied in 1994 and 1995 by the APHIS Biological Control Laboratory in Bozeman, Montana.

The release program: During the summers of 1994 and 1995, three species of diffuse knapweed biological control agents were released at various locations throughout the 35-acre infestation in Young. Insects used were (1) Agapeta moth - Agapeta zoegena, (2) bronze knapweed root-borer - Sphenoptera jugoslavica (a beetle), and (3) banded gall fly - Urophora affinis. The agapeta moth and the root-boring beetle larvae feed on the plant's roots, whereas the gall fly larvae consume the developing seeds.

Results of the project: Monitoring (by USDA, ADA, and Rancher-Cooperator) in 1995 and 1996 of the release sites in the diffuse knapweed population on the ranch in Young have revealed no established populations of the three species of biological control agents released.

Future of the project: Plans are being made by ADA, the Arizona Department of Transportation, the Young Town Council, and other cooperators to include diffuse knapweed in the IPM Management approach along with yellow starthistle.

Non-native populations and species of whitefly parasitoids and predators are evaluated for their ability to survive and reproduce in key crops in the desert environment. These crops include alfalfa, melons, cotton and cole crops. Evaluations are conducted with field-cage trials under prevailing desert environmental conditions. Biological characteristics of particular interest are further examined in laboratory studies. During the past four years, 8 different Eretmocerus and 7 Encarsia species or strains from 13 different countries have been evaluated. These evaluations have identified several non-native whitefly enemies that are promising candidates for improving whitefly management in desert regions: Eretmocerus mundus from Mediterranean countries and two new species of Eretmocerus from Pakistan and the United Arab Emirates. Each of these has been widely released in desert valleys of Arizona and California through inoculative and inundative releases.

Surveys of native enemies in crops, urban areas and surrounding desert vegetation have also been conducted. Significant differences have been found in natural enemy activity on different crops in southwestern deserts. For example, low parasitism rates of whitefly by native parasites on winter cole crops was identified as a major gap which the introduction of exotic species is intended to fill. The native species that usually has the greatest impact on whitefly in other crops including melons and cotton, Eretmocerus eremicus, does not become abundant in whitefly on cole crops. As a result, populations of this parasite tend to be low in the spring when they are most needed in spring crops. Because of the low level of activity of native Eretmocerus in winter cole crops, introductions of exotic species capable of development in these crops is very important. Other surveys examine non-target whitefly hosts, such as other whitefly species found on urban ornamentals and native desert vegetation. This information will be valuable in documenting the eventual impact, if any, on non-target species resulting from introductions of silverleaf whitefly enemies.

Knowledge gained of the relative performance of exotic species and the value of different potential reservoir host plants is used to design suitable approaches to implementing biological control. For instance, urban areas appear to be more useful as overwintering sites than desert plants. Diversity of vegetation and seasonal continuity for whitefly populations may also give urban reservoirs more stability for introducing exotic populations. Consequently, urban environments have been high-priority release sites for inoculative introductions.

Introductions have been made in commercial crops and suitable non-crop reservoirs of 30 or more different geographic populations or species of parasitoids and several predatory coccinellids in the genera Serangium and Delphastus. The size of these introductions has varied from several hundred to millions of individuals per species or geographic population. Many of these releases have resulted in local, within-season reproduction at the release site. Recovery surveys during 1996 and 1997 of several exotic Encarsia and Eretmocerus from a number of new locations indicate that successful overwintering and dispersal from the release sites has occurred. Several years of increasing recoveries will likely be required to demonstrate their permanent establishment.

The Brawley lab provides support for regional APHIS demonstration projects utilizing whitefly parasitoids as a management option in commercial agriculture. These projects aim to show that the use of parasitoids is compatible with the use of selective pesticides and other management methods. Introduced parasitoid species shown to perform well under desert conditions in field evaluations (see above) are mass-produced for use in these demonstrations. Augmentative releases of parasitoids in spring melons and cotton during 1994-97 have shown that high levels of parasitism can be achieved with early-season releases. The exotic species of Eretmocerus released reproduced extensively in the field. Besides demonstrating that biological control is possible in commercial crops, these projects also rapidly and inexpensively mass-produce the exotic species in the field, increasing their spread and the likelihood of permanent establishment.

The Brawley biocontrol program is also engaged in USDA efforts to secure the transfer of new and effective natural enemies to private industry. This will help to ensure their long-term availability for further research and demonstration projects and, eventually, buyers. Other related projects include cooperative studies to evaluate whitefly traps for their impact and specificity towards natural enemies and biological evaluations of Semidalis species, a promising native neuropteran predator of whitefly.

The main research project in our laboratory currently focuses on the interactions of two parasitic wasps attacking the sweetpotato whitefly, Bemisia tabaci, and their effect on biological control. This project has both particular and general applicability. One of the wasps is a native of Arizona, Eretmocerus eremicus, and the other is an exotic species from Spain introduced for biological control of the sweetpotato whitefly, Encarsia transvena. The exotic species has an unusual biology in that males are obligate hyperparasitoids and develop at the expense of females of their own species or other parasitic wasps. Some researchers have claimed that these wasps with hyperparasitic males pose a risk to effective biological control. We have initiated a study which includes laboratory and field investigations of this question.

In a series of three laboratory experiments, MH and SK investigated the preference and suitability of the immature stages of both wasps species for development of the hyperparasitic male E. transvena. The results suggest that male E. transvena may successfully develop on more stages of E. eremicus than of E. transvena immature females. While we observed no significant preference by the female E. transvena for either host type when suitable stages of both species were presented, the window of vulnerability of immature E. eremicus is much broader than that of E. transvena immatures, and may lead to a greater rate of attack on these wasp hosts.

This past summer, we performed the first of a two year series of field experiments to investigate the outcome of interactions between the two parasitoids species. We (MH, SK, TC, and two undergraduate student workers, Don Harp and Kim Hammond) caged 180 cotton plants, inoculated them with whiteflies, and then introduced either no parasitoids, one of the species alone, or both species together. We have just finished collecting data on this experiment and data has not been entered or analyzed.

TC is starting experiments to address the effects of direct competition between the two parasitoids for the whitefly host.

The effects of interactions between natural enemies on biological control has long been a subject of debate among biological control workers. If the introduction of a second natural enemy has the potential to disrupt the control provided by the first, then a different protocol for introduction biological control than multiple species releases is indicated.

MP, a postdoctoral associate with funding from the Danish Research Council, is studying the interactions of two species of pecan aphids and their lacewing predators in cooperation with FICO, Sahuarita, AZ. She is specifically addressing whether there is direct competition for resources between the aphids, or whether competition is mediated by predators. Aphids are the only major pests in pecans, and therefore anything we learn that might lead to enhanced natural enemy conservation could eventually lead to reduced pesticide use. In addition we expect that this study will contribute to current debate in the literature about the relative importance of direct and 'indirect' effects between species. Some researchers have taken the position that because all 'indirect effects' cannot be predicted prior to the introduction of a biological control agent we should not perform biological control programs at all. This project began in May with sampling of the pecan aphid complex, and field and laboratory experiments are starting this fall.

JN, a masters student, is looking at the occurrence of egg-killing in the whitefly parasitoid, Encarsia formosa. E. formosa females locate eggs laid by other females within the whitefly host, and may stab them with their ovipositor before laying their own egg. While observed and recorded in a one page paper ten years ago, this phenomenon has not been investigated in any depth and yet it appears to occur with great frequency. These findings may change the way we think about this extremely well-studied parasitic wasp, a successful biological control agent for greenhouse whitefly.

The Beneficial Insect Rearing Facility at the University of Arizona has had a record setting year in 1997. Improvements in rearing procedures have resulted in dramatic increases in parasitoid production. In 1997 three exotic strains of Eretmocerus were cultured for release in Arizona and the Imperial Valley to control silverleaf whitefly. The three strains and their country of origin are, M12 from Pakistan, M20 from Israel, and one from the United Arab Emirates. The Beneficial Insect Rearing Facility supplied these wasps to USDA-APHIS for field release.

Mass rearing of all three strains was done on eggplants. The host used for the parasitoid was the silverleaf whitefly. Successful rearing of Eretmocerus depends on a number of factors. Healthy plants, a vigorous Bemisia colony and correct timing of wasp introduction and harvest are all important to an efficient rearing operation. Careful cleaning at the final stage results in a pest and whitefly free product. The harvested and cleaned pupae are then ready for introduction. The following description gives the basics of our rearing procedure.

Starting with healthy plants is essential. These plants are exposed to large numbers of adult whiteflies. Once high densities of eggs have been deposited by the Bemisia, the adults are removed. The eggs are then allowed to develop into second instar nymphs. When the whitefly nymphs are between their second and third instars the Eretmocerus are introduced. If the timing and number of wasps introduced is correct, parasitism percentages of 80-90% are not uncommon. Depending on temperature and humidity, the wasps will develop and begin emerging in two to three weeks.

From January 1 until October 1, our facility has produced 33,614,750 Eretmocerus of all three exotic strains. A total of 23,358,750 wasps have been shipped to USDA-APHIS. Continuing improvements should increase production totals for 1998.

Large-scale studies were initiated in 1996 to examine the comparative effect of two new insect growth regulators (IGR) and conventional insecticides on the abundance and activity of native natural enemies of Bemisia. Eretmocerus eremicus and Encarsia meritoria were present throughout the season, although the former species was dominate, comprising about 75% of all parasitoids collected. Rates of parasitism reached 75% in some plots and was generally higher throughout the season in plots sprayed with either buprofezin (chitin inhibitor) or pyriproxyfen (JH analog) in comparison with those treated with organophosphate-pyrethroid mixes. Rates of parasitism generally increased over the season as Bemisia populations declined and averaged about 10% over the whole experimental area for the season. A severe storm in late July severely depressed populations of both host and parasitoids. Densities of coleopteran and heteropteran predators were depressed in plots receiving a rotation of conventional insecticides in comparison with those receiving IGRs. This effect was most apparent immediately following the first applications in early July. The same pattern was apparent for spiders except that populations were depressed in plots treated with pyriproxyfen and conventional insecticides in comparison with those treated with buprofezin. There was no clear pattern in relation to insecticide regime for Chrysoperla carnea and Drapetis spp. These studies were continued in 1997 and results are pending.

As part of the large-scale research project described above, comparative partial life table studies were conducted in 1996 and 1997. Cohorts of eggs and settled 1st instar nymphs were established from natural populations and the fate of each individual was tracked by visual observation. Mortality due to insecticides, predators, parasitoids, and weather were quantified. Predation was a major source of mortality for both eggs and nymphs prior to the use of any insecticides. Predation continued to be a major source of mortality, particularly of nymphs, after the first applications of insect growth regulators (IGR). Rates of natural enemy-induced mortality were comparable in IGR plots with that in untreated control plots. Natural enemy mortality was depressed in plots receiving mixtures of conventional compounds. Severe wind and rain also was a significant source of mortality for nymphs in both years. More detailed analyses are planned.

Studies were conducted in 1994-1995 to measure the effects of insecticide use frequency in cotton on populations of general predators of Bemisia tabaci in Imperial Valley, CA. Replicated plots were sprayed with a pyrethroid + organophosphate mix when densities of adult whiteflies exceeded 2.5, 5, 10, or 20 per leaf. Untreated plots served as controls. Geocoris punctipes was consistently the most abundant predator species, but increasing insecticide usage significantly reduced populations. Orius tristicolor and Nabis alternatus were moderately abundant and insecticide use frequency had a comparatively smaller negative effect on population density. Insecticide use frequency had little or no effect on several predators, including Collops vittatus, Spanogonicus albofaciatus, assassin bugs (Zelus and Sinea spp.), and Lygus hesperus. Spiders and Hippodamia convergens declined with increasing insecticide use. Populations of Chrysoperla carnea actually increased with greater insecticide use in 1994. Based on density and relative value (determined from previous analysis of predator gut contents), the insecticides used here would be expected to be most detrimental to natural control provided by G. punctipes and O. tristicolor.

The population dynamics and activity of native parasitoids (Eretmocerus and Encarsia) of Bemisia were studied in cotton growing regions of Israel and southern California over several years. The emphasis of the study was to examine the effects of insecticide use on parasitoid activity. Results from Israel showed that, in many cases, insecticide treatments for other pest, primarily bollworms, had only a minor affect on rates of parasitism. Percent parasitism often approached 80% over extended portions of the growing season even in fields subject to repeated use of synthetic insecticides. Studies in California involved controlled experiments with different frequencies of insecticide use. Percent parasitism neared 25% in 1994 and often exceeded 60% in 1995 and generally did not differ between treated and untreated control plots. Results from both Israel and California suggest that the insecticides used may be no more detrimental to parasitoids than to whitefly. Nonetheless, repeated insecticide use would be expected to reduce regional parasitoid populations and negatively impact the natural control provided by these organisms.

Large complexes of generalist predators are found in many agricultural systems, yet we have only a rudimentary knowledge of how they function in pest control. We used monoclonal antibodies developed to recognize egg antigens of pink bollworm, Pectinophora gossypiella, and sweetpotato whitefly, Bemisia tabaci, Strain B (= B. argentifolii), to study the native predators of these pests in cotton. Using a multiple-gut ELISA (enzyme-linked immunosorbent assay) we tested more than 22,000 individuals of seven species of predaceous Heteroptera over two field seasons in central Arizona. Based on the frequency of positive ELISA responses and population densities, Orius tristicolor and Lygus hesperus, a recognized pest species, were found to be the dominant predators of pink bollworm and sweetpotato whitefly eggs. Minor predation can be credited to Collops vittatus, Geocoris pallens, G. punctipes, Hippodamia convergens, and Nabis alternatus. Preliminary analysis of pink bollworm egg predation suggests that the cotton predator complex was responsible for removing approximately 20% of all pink bollworm eggs over the entire season.

We have developed a process-oriented simulation model of sweetpotato whitefly population dynamics and biological control in cotton. The overall structure of the model is highly flexible so that various components can be easily added as research progresses. The basic foundation of the model is temperature-dependent distributed development of eggs and nymphs, and temperature-dependent adult fecundity and longevity. Presently the model incorporates the effects of natural enemy activity, including mortality imposed by indigenous predators and augmentatively released parasitoids and predators. The model is user-friendly and utilizes menus and dialogue boxes to prompt the user for needed information such as weather data, initial densities of pest and natural enemies populations, and biological attributes of parasitoids and predators. The user can also enter field sample information, and make daily, weekly or monthly decisions concerning the release of natural enemies or the application of insecticide. Graphic and tabular output can be generated. The model is being tested against independent observations in the literature. Field validation of pest dynamics and augmentative releases of several species of Eretmocerus is currently underway.

The development and reproduction of Trichogrammatoidea bactrae was studied on pink bollworm under a variety of constant and fluctuating temperature regimes. Overall, the parasitoid appears well adapted to the relatively high temperatures associated with southwestern cotton. Females could survive for > 24 h and be potentially capable of parasitizing 20 or more eggs under the most severe temperatures typical of mid to late summer. Immature development and survival should be favorable and should not pose a significant barrier to establishment.

Studies of searching behavior revealed that T. bactrae females were capable of parasitizing eggs throughout the cotton plant canopy. However, rates of attack on eggs under the calyces of cotton bolls are low. These results suggest that this parasitoid may be best suited to early-season release when the majority of pink bollworm eggs are laid on foliage and stem surfaces.

Augmentative releases of T. bactrae were evaluated in small replicated field plots for control of pink bollworm. Weekly to biweekly releases at rates of «100,000/acre significantly depressed pink bollworm populations over the early part of the growing season in comparison with check plots. There was also a significant enhancement in yield and reduction in seed damage in parasitoid release plots. However, augmentative release are not economically feasible at present due to parasitoid production costs and variable quality.

Further studies were conducted to estimate rates of movement by this parasitoid. Preliminary results indicate that parasitoids readily moved over 85 m within 24 hours and persisted for at least 72 h after release. Movement was significantly influenced by wind direction during the first 24 h, but parasitoids were fairly evenly distributed within the recapture grid 48-72 h after release. Results also indicate that parasitoids moved more readily along the length of rows than across rows.

Silverleaf whitefly infesting cantaloupe and others melons grown in the spring, are an attractive candidate for biological control by augmentation with parasitoids of Eretmocerus species because most of the pest pressure is caused by within field reproduction, unlike crops grown later in the season which can be heavily impacted by immigration of whitefly from other crops. As such, early season inoculations of parasitoids can be relied upon to increase within the crop and cause significant reductions in whitefly levels. Furthermore, parasitoid release is compatible with Admire^tm (imidacloprid) which is widely used for early season whitefly control. Releasing parasitoids in Admire treated fields, can eliminate the need for applications of pyrethroids for late season whitefly control that are often needed after the effectiveness of the Admire application is reduced with time. Because of growing concern about the possibility of silverleaf whitefly evolving resistance to Admire, combining its use with a non-chemical control measure (such as parasitoid release) can help delay the development of resistance.

Similarly, whitefly infesting cotton is also an attractive candidate for augmentative biological control in areas (or years) where heavy immigration pressure is not a concern. The use of B.t cotton and several new whitefly selective insect growth regulators, reduces the pesticide pressure on cotton allowing beneficial insects to play a greater role in whitefly suppression. Parasitoid release combined with limited pesticide treatments may result in very effective and stable pest control.

Fours years of studies in small plots and entire fields has shown that levels of parasitism ranging from 30 to 90% and reductions of whitefly ranging from 50% to 90%, relative to no-release check fields, can be achieved with early season release of exotic Eretmocerus species. The next step is to determine the minimum number of parasitoids needed for efficacious and economical release rates. Replicated release rate studies are conducted within grower cooperator's fields in plots ranging from 5 to 10 acres in size. Release rates are investigated over a range that include those that would be considered economical ($80 to $180 an acre). Levels of parasitism achieved are analyzed with non-linear regression to determine if there are release rates that result in threshold levels parasitism above which no increase in parasitism can be achieved.