Summary

Introduction
Purpose
Background

Materials and Methods
General Method
Protocol Development
Insects Sampled
Sampling Method
Data Recording
Data Entry
Data Analysis

Results and Discussion
Apple samples
White apple leafhopper
Aphids and Natural Enemies
Western tentiform leafminer
Campylomma
Leafrollers
Mites
Fruit damage
Pear samples
Prebloom cluster counts
Leafroller counts
Beating tray counts
Leaf-brushing counts
Fruit damage

References Cited

Tables 1-30

Type Means Table #1-15
Site Means, Type I Table #16-30
Site Means, Type II Table #31-45

Figures 1-111

Bar Graphs
Aphid Population Graphs
Psylla Brush and Count Graphs
Psylla Tap Graphs
Moth Trap Catch Graphs

Appendices
A. Apple Sampling Protocols
B. Apple Data Forms
C. Pear Sampling Protocols
D. Pear Data Forms
E. Site Maps

The objective of this project was to document changes in secondary pest and natural enemy populations in blocks managed under large-scale mating disruption (MD) when compared with conventional (organophosphate-based) management regimes. Our hypothesis is that conditions will be more favorable in blocks under MD for integrated control of secondary pests, and that the reduced need for insecticide applications for secondary pests will offset the higher cost of MD technology.

Standardized sampling protocols were developed for the principal secondary pests of apple and pear and their associated natural enemies. Seven apple sites and three pear sites were sampled during the 1998 growing season, for a total of 9 sites (one site contained both apple and pear orchards). For each site, a subsample of the blocks in the boundary of the MD area was chosen to carry out the intensive sampling. These blocks were standardized, where possible, to 10 acres in size. Orchard blocks under conventional management that were representative of the regions were chosen as comparison blocks. The number of blocks, both MD and comparison, were proportional to the size of the MD acreage included in the site. Five of the sites were the primary MD projects (the CAMP project) established during the 1995 growing season (with the exception of Randall Island, begun in 1993). The remaining 4 sites (GRABs subproject) were apple acreage in central Washington, and differed from the CAMP sites in that they were managed by a single farmer or corporation as opposed to a group of cooperating independent fruit growers.

Apple: For the majority of the pest insects and natural enemies sampled, no differences occurred between the 2 management regimes. However, in most cases where differences occurred, the trend was for lower pest populations or fruit damage, or higher populations or impact by natural enemies. As in 1996 and 1997, the % parasitism of overwintering white apple leafhopper eggs was significantly higher in MD blocks. However, 1^st and 2^nd generation nymphs were no different in either MD or CONV blocks. Aphid populations were generally low, and no differences in population means were found. One of the exceptions to the general trend was that lacewing populations in particular and motile predators in general were slightly higher in conventional blocks. In the mite binomial counts the populations were higher in conventional blocks on all three sample dates. However, in all cases, the populations were subeconomic. In the leaf-brush mite samples, the tetranychid mites were higher in the conventional blocks in both evaluations. Predatory mites were higher in the MD blocks in the June sample. Fruit damage by codling moth and leafroller was lower in MD blocks, which continues the 1996-97 trend.

Pear: Psylla nymphs were higher in conventional blocks than in blocks under MD, although this difference could not be tied directly to increased levels of natural enemies. Pest mite populations did not differ between the 2 management regimes, although predatory mite populations were higher MD blocks. Fruit damage by psylla was higher in conventional blocks, corresponding to the higher in-season populations. Fruit damage by codling moth and leafroller was not significantly different between the 2 management regimes.

INTRODUCTION: The organophosphate-based IPM program that is currently in widespread use throughout the apple-growing regions of the Pacific Northwest has been relatively stable and effective for over thirty years. Until the mid- to late 1980s, key pests (codling moth, Cydia pomonella L. and leafrollers, Pandemis pyrusana Kearfott and Choristoneura rosaceana (Harris)) were controlled and many secondary pests were either controlled or suppressed with this program. Some pests developed resistance over time (e.g., mites, leafhoppers, leafminers, and aphids), causing other materials (primarily carbamates, and an organochlorine), to be incorporated into the control program. The dominance of OP and other conventional insecticides as the main pest control tactic has severely restricted the activities of natural enemies in the orchard ecosystem. A few natural enemies have developed a tolerance of certain insecticides, especially the predatory mite Typhlodromus occidentalis (Nesbitt) and the parasitic wasp Pnigalio flavipes (Ashmead), and have been important components of the OP-based IPM program. However, activities of even these natural enemies are disrupted by commonly used insecticides, e.g., encapsulated methyl parathion, carbaryl, formetanate hydrochloride (etc.). As a result of the almost universal use of OPs for key pests, relatively little effort has been directed at looking at integrated control of secondary pests other than mites and leafminers until mating disruption (MD) became available.

The removal of OPs from the IPM program provides opportunities both for a buildup of natural enemies, but also the release from suppression of many secondary pests. While OPs have been considered ineffective for aphids and leafhoppers, 50-75% mortality may have occurred with each cover spray (e.g., Beers & Elsner 1987, Beers et al. 1987, Beers & Browne 1991). The cumulative effect of this suppression has largely gone unnoticed. Thus, the potential exists for significant changes in secondary pest populations under MD-based programs. A documented example of this phenomenon is the outbreak of Eriosoma lanigerum Hausmann (woolly apple aphid) following the change from an OP-based program to a pyrethroid-based program in New Zealand (Penman & Chapman 1980).

Industry representatives have identified the control of secondary pests as being second only to efficacy of MD as a critical component of the areawide MD program. It is anticipated that at least in the initial phases of the areawide program, costs for codling moth control will be higher than in conventionally managed orchards. In orchards where leafroller is problematic, control using Bacillus thuringiensis-based products will add substantially to control costs. Thus, reductions in the use of control measures for secondary pests such as aphids, leafhoppers, leafminers, and mites are critical to balance the higher cost for control of key pests.

Growers committed to the MD technology will tolerate some cost differential especially when it is mitigated by the buy-down incorporated in pilot test projects. However, for implementation of MD to spread beyond the pilot areas, (where no buy down will be possible), the total program cost will have to approximate that of current conventional programs. The most likely scenario for this will be through 1) reduced cost of pheromone dispensers through more efficient production or economy of scale and 2) reduced cost of control of secondary pests. An alternative model for implementation of MD (where pheromones are used as a one-for-one substitution of azinphosmethyl and no other components of the control program are changed) has led to programs with substantially higher costs and a largely unrealized potential of reduced pesticide use (Williamson et al. 1994, Weddle 1994).

PURPOSE: The purpose of this project is to assess the impact on secondary pests and their natural enemies under large-scale pheromone mating disruption control tactics for codling moth. Our hypothesis is that the use of MD, and the concomitant reduction in OP use will foster integrated control of secondary pests. The potential benefits are 1) a more stable orchard ecosystem, which is less prone to outbreaks of secondary pests; 2) a lower control cost for secondary pests when compared to orchards under conventional (OP-based) management; 3) a lower overall insecticide load entering the environment, and 4) a lower detrimental impact of the pesticides used (e.g., more selective products). For purposes of this project, secondary pests are considered as all pests other than codling moth. The secondary pests of consequence vary by crop and region; however, sampling was standardized across regions for a given crop. The project is structured to achieve 3 objectives: 1) document changes that occur over time in a given block under mating disruption in terms of diversity and abundance of natural enemies and secondary pests; 2) compare MD versus conventional management strategies on a yearly basis across the 9 sites sampled for secondary pest and natural enemy populations and the resulting economic damage; and 3) assess the effect of MD versus conventional management on spray practices for secondary pests, and its economic significance.

Background: The first implementation of a large scale mating disruption project was the Randall Island Project, headed by Dr. Steve Welter and co-investigators in the Sacramento Delta pear growing region of California. This project was begun in 1993, focusing on MD as a resistance management tactic. The advantages (and perhaps the necessity) of this approach was recognized in the western region by the scientists involved in codling moth research, and a plan was developed to extend the approach to the major pome fruit production areas of Washington and Oregon. The conceptual framework for this extension was compiled in 1994 by Dr. Marcos Kogan in a document entitled "Areawide Management of the Codling Moth: Implementation of a Comprehensive IPM Program for Pome Fruit in the Western United States". The project was designed as an "areawide, multi-state, multi-institutional program in the region to assess, test, and implement a multi-tactical integrated pest management strategy for the codling moth that will reduce the negative impact of pesticides on natural enemies of secondary pests, the environment, and farm worker community" (Kogan et al. 1994). Preliminary work was accomplished in 1994 with the first full year of project implementation in 1995. Five project sites were implemented beginning in 1995, and were sampled in 1996 and 1997 for secondary pests and natural enemies:

The GRABS (Grower Resource Acquisition Baseline Study) program was started in 1995 by Dr. Alan Knight and Nick Stephens (Northwest IPM) and was charged with insect pest sampling in blocks under mating disruption and in conventional blocks located in the Brewster/Bridgeport areas of Washington as well as blocks near Vantage, Mattawa, and Chelan. These blocks differed from CAMP blocks in that they were farmed by a single grower/company whereas the Camp blocks encompassed numerous growers, packing houses and field representatives. For each site, 4 blocks inside the mating disrupted area were paired with 4 blocks under conventional management, attempting to select blocks of similar size, age, cultivar makeup, and management. The 32 GRABS blocks are divided into the following sites:

Sample blocks were about 10 acres in size. If only larger blocks were available, sampling was restricted to a specific 10-acre section of the larger block. If only smaller blocks were available in a specific class, then the entire block was sampled. Care was taken to match tree age, cultivar and training system as much as possible between the mating disruption blocks and the blocks under conventional management. As a minimum dataset, we utilized at least 1 intensively sampled block for every 100 acres in the project site (e.g., if the site is 1,200 acres, 12 intensively sampled blocks are needed). In CAMP sites where both apples and pears are grown, blocks were chosen in proportion to the percentage of the two crops at that site. Practically speaking, this was of concern only at the Parker project. Both the Medford and Randall Island CAMP sites contained predominately pears. One modification for 1997 was that the minimum number of blocks sampled of any one type was set at 4. The GRABs sites monitored in 1996 and 1997 included only apples.

MATERIALS AND METHODS:

General Method: Samples of secondary pests and their natural enemies were taken on a subset of blocks within the various CAMP/GRABS program project sites along with a comparable number of similar blocks outside the site and under conventional techniques of pest management (see tables above). Separate sets of protocols were developed for apple and pear to make them specific for the pest and natural enemy complex on each crop.

Protocol Development: The cooperators met in October of 1995 to discuss sampling protocols. There was agreement on the general principle that sampling protocols must be standardized in order to make cross-project comparisons valid. The types and frequency of the samples taken were a compromise between what was desirable and what was feasible given the resource constraints. A second principle that arose from these discussions is that the timing of the samples would target documenting populations that had occurred, rather than sampling in order to predict the necessity for a spray application. This principle applies particularly to insects with a limited number of discrete generations, such as white apple leafhopper and western tentiform leafminer. Insects with multiple overlapping generations were sampled more frequently, in hopes of capturing peak populations and important interactions with natural enemies. Some samples were based on a "trigger"; that is, the sample was taken only if some other sample or preliminary search revealed a sufficient population to perform a more detailed sample. This applied to 2 situations: Extra leaf-brushing samples for mites were done only if the binomial samples exceeded a certain percentage of leaves infested, and leafminer parasitism assessment was done only if the mine density evaluation exceeded a threshold. These principles were incorporated into a detailed set of instructions, and made concrete in a series of data forms for each insect sample type, crop, and location. In addition, for the apple sampling protocols, an approximate schedule of samples throughout the season was developed, keyed to crop and insect phenology. After two years of field use, the cooperators met again in January of 1998 and revised the protocols. The revised protocols and forms were distributed before the commencement of the 1998 field season.

1. White apple leafhopper overwintering egg parasitism

2. Leafhopper nymph density- Generation 1 & 2

3. Aphid population and aphid natural enemy counts (5 counts)

4. Leafminer population counts- Generation 2 & 3

5. Leafminer parasitism rates- Generation 2 & 3

6. Campylomma nymph counts (beating tray) (petal fall only)

7. Leafroller larval counts (overwintering and summer generations)

8. Mite binomial counts (multiple counts)

9. Mite leaf-brushing counts (3 in ca. June, July, and Sept)

10. Fruit damage samples, mid-season and preharvest.

1. Pre-bloom cluster sample for mites and psylla,

2. Post-bloom leaf brushings for mites and psylla

3. Limb taps for psylla and predators

4. Fruit damage counts at mid-season and pre-harvest

1. Leafroller parasitoid collection and preservation

2. Leafminer parasitoid collection and preservation

3. Leafhopper egg parasitoid collection and preservation

4. Preservation of unknown phytophagous or predatory mites

Sampling Method: Trees or units to be sampled were selected randomly throughout the test blocks. This was done by walking the block in an X pattern or, if the block was trellised, by subdividing the block into quadrants. Sampling units were selected at random from the tree, taking care to avoid bias. Care was taken to avoid subconscious sampling of infested fruit or leaves in preference to those units that did not have obvious damage. Sampling was done in situ, unless otherwise called for in the protocol. All sampling was done from the ground, without a ladder, unless top shoots were specified in the protocol. In blocks that contained multiple cultivars, sampling was restricted to the cultivar that comprised the majority of the block. As much as possible, the blocks sampled represented the predominant apple or pear cultivars in that region.

Data recording: Data was recorded in one of two ways. First, as in 1997, numbers were recorded in pencil, and header information was provided on each page. The date indicated on the datasheet represented the day the sample was collected, not when it was counted if samples were taken to the laboratory for further processing. If these dates were not the same, both dates were recorded with the date counted indicated in parentheses next to the date collected. All columns on the page were filled in, even if the entry was a zero. A zero meant the sample unit was looked at and that the insect of interest was absent, whereas a period (the missing data notation) entered in the column meant that the sample either was not taken or misplaced. Alternatively, the data was entered in the field into a HP 360LX palmtop unit. These units were set up with Pocket Excel versions of all the Secondary Pest Sampling datasheets. A separate file version of the datasheet was completed by the field scout for each block and the files downloaded into the Data Entry computer upon return to the office. These datasheets were then printed and a hard copy filed.

Data entry: The 1998 data were entered either directly from data forms or copied from field generated data files into Microsoft ExcelÒ ver. 5.0 workbook templates. In the workbook template the data were entered twice, in 2 worksheets in the same workbook; a 3^rd worksheet contained a formula that flagged data that were different in the 2 data worksheets. The use of spreadsheets greatly speeded the time required to teach new operators the data entry system, and to enter and verify the data. The data were then transferred using the Clipboard function to Kedit (Mansfield Software Group, Storrs, CT) ASCII files.

Data analysis: Data were tested for homogeneity of variance using Levene's (1960) test, and transformed (log[y+0.5]) where indicated prior to analysis. Data were analyzed using analysis of variance on each count date (PROC GLM; SAS Institute 1988). Means were separated with the Fisher's Least Significant Difference. Three types of analyses were conducted for each sample type:

Where multiple counts of a given type were made (e.g., aphids and natural enemies on apple, or psylla adult beating tray counts on pears) a seasonal average was calculated and analyzed. For samples of discrete generations or timings (e.g., 1st and 2nd generation white apple leafhopper), a separate analysis was done for each generation or timing.

RESULTS AND DISCUSSION: The analysis of large sets of data generated from different sites and compiled by different researchers has certain inherent limitations, as alluded to previously. Every attempt was made to standardize the data from the different sites and to compare them in a meaningful manner. However, there were some variations in the sample techniques and frequency used at the different sites. In addition, no account is made in the current analysis of the impact of insecticides on pest and natural enemy populations, thus these data should be interpreted with caution. The spray record information is being compiled at the time of this report and the analysis of this data may clarify many of the secondary pest population patterns noted during the 1998 growing season.

White apple leafhopper, Typhlocyba pomaria McAtee is an indirect pest on apple in the Pacific Northwest. Specific control measures were rarely needed until the mid- 1970's at which time populations were occurring at higher levels, probably due to resistance to organophosphate insecticides. Leafhopper overwinters as eggs inserted under the bark of apple and then have two generations during the summer. Damage is due to foliage feeding which can reduce photosynthesis. In addition, tar spots (droplets of excrement deposited on the fruit) can be a problem in heavily infested orchards, and the adults also represent a significant nuisance to field workers during harvest. The primary biological control agent of leafhopper is a mymarid wasp egg parasitoid, Anagrus epos Girault. This parasitoid attacks both overwintering and summer generation eggs and levels of parasitism can be quite high in unsprayed orchards.

Leafhopper samples were taken at the peak nymphal period of the 2 generations that occur in the Pacific Northwest, ca. late May and mid-August. Nymphs per leaf (10 leaves per tree on 20 trees per block) were counted in situ. Parasitism of the overwintering eggs was evaluated in early April by cutting 10-cm sections of the previous season's extension growth (2 shoots per tree on 20 shoots per block, or a total of 40 shoots), and dissecting all the leafhopper eggs found.

Overwintering leafhopper egg density was highest in the Oroville site, moderate in Howard Flat, P&G and Nickell (Tables 1, 16, 31; Figs. 1, 2). There was no correlation in the egg density and parasitism among the sites, as Vantage and Howard Flat both had >40% parasitism, but the latter site exhibited a 9-fold higher egg densitity. As in 1996-97, percentage parasitism of the eggs was higher in the MD treatment when compared to the conventional blocks. In addition, in 1998 the % live eggs was higher in the conventional sites, although this difference could not be detected in the total number of live (normal) eggs.

Nymph densities were low during the 1st generation, with the average of 0.03 nymphs/leaf in both the MD and conventional blocks (Tables 2, 17, 32; Figs. 3, 4). No statistical difference in 2nd generation nymph density was found due to management regime, although nymph densities were higher than the 1^st generation.

Aphids and their Natural Enemies: Aphids on apple are both a direct and indirect pest. Aphid feeding on heavily infested shoots can reduce growth potential, especially when another indirect pest infestations are present such as leafminer, leafhopper or mites. The most important damage is direct and caused by honeydew dripping on exposed fruit. This honeydew is then colonized by a black mold that can cause fruit to be downgraded. Aphids can have anywhere from 9-17 generations per summer and populations can grow quickly where conditions are favorable. Typically, there has been a high threshold on the part of growers and field advisors and chemical treatment has not been applied till 50-75% of the shoots are infested. At the present time, the recommendation of treatment has been for when there are 2-3 infested leaves per randomly chosen shoot. Aphids and the complex of generalist predators were sampled beginning in early June and then every three weeks until late summer. The primary target of this sample were the summer aphid complex, apple aphid (Aphis pomi De Geer) the spirea aphid (Aphis spireacola [Patch]). The natural enemy complex counted consisted of lady beetles (Coleoptera: Coccinellidae; adults and larvae); syrphids (Diptera: Syrphidae; larvae only); lacewings (Neuroptera: Chrysopidae and Hemerobiidae; adults, larvae, eggs and pupae); Deraeocoris brevis (Knight) (Hemiptera: Miridae, adults and nymphs grouped); Cecidomyiids (Diptera: Cecidomyiidae, larvae only); Campylomma verbasci (Meyer) (Hemiptera: Miridae; nymphs and adults); and parasitized aphids (prob. Aphidius spp., Hymenoptera: Aphidiidae).

Aphid populations were not different overall in the conventional blocks than in the mating disrupted blocks. As in 1997, lacewing populations were higher in the conventional blocks. In addition, the total number of motile predators were also higher in conventional blocks. Lacewings have been shown to tolerate field rates of organophosphate and carbamate insectecticides extremely well in bioassays (Beers, unpublished data). Coccinellids vary by species and material in terms of tolerance, but can also survive many spray regimes. A more detailed analysis of these data are warranted to determine the influence of prey populations on predator populations, as well as in-season pesticide applications.

Western Tentiform Leafminer (Phyllonorycter elmaella Doganlar & Mutuura). Leafminers have been present in Northwest orchards since the early 1900's, but only since ca. 1980 have they achieved pest status, probably due to the spread of an organophosphate-resistant strain. Leafminers have three complete generations in the Western U.S., with a partial fourth generation. Damage is caused by the larvae, which feed in the spongy mesophyll layers of the leaf, producing mines. Loss of chlorophyll-bearing tissue may reduce photosynthesis, and premature defoliation may result from a high mine density. Premature fruit drop on sensitive cultivars of apple has also been reported. Treatment thresholds of 1-2 mines/leaf for the 2nd generation, and 5 mines/leaf for the 3rd generation have been recommended. The primary biological control agent in PNW orchards is a small parasitic wasp, Pnigalio flavipes. If the level of parasitism is >35%, the potential for regulation below the treatment threshold is good. This parasitoid is quite sensitive to several frequently used insecticides and parasitism rates in conventional blocks are usually low.

There was a leafminer outbreak in northcentral Washington during the 1997 season, and leafminer populations were high again in 1998. The Oroville site in particular had high populations both 2^nd and 3^rd generations (Tables 5, 20, 35; Figs. 7, 8), and the Vantage and P&G sites also had substantial populations. In contrast to 1997 (where no difference in population occurred), leafminer densities were higher in conventional blocks in 1998. The relatively high population levels allowed parasitism to be evaluated on most blocks. In the 2^nd generation, both the total mines/leaf and the number of live mines/leaf was higher in conventional sites, while the percentage of parasitism was higher in the MD sites (Tables 6, 21, 36; Figs. 9-12). The pattern was the same during the 3^rd generation, but only the live mines were statistically higher. In 1997, no difference in parasitism was found due to management regime in either generation, although some of the OPs are quite toxic to P. flavipes, the principal parasitoid of leafminer. It appear that the reduced spray pressure may have a positive influence on biological control of leafminer.

Campylomma (Campylomma verbasci): C. verbasci is facultatively zoophytophagous, and depending on the crop species and time of year, can be regarded as either a pest or a predator. This pest was not considered to be significant until the early 1970's. Damage has increased over the past 2 decades and C. verbasci has become a pest that requires control in many areas. C. verbasci damages fruit only during a relatively short period of time during and after bloom. Insect feeding on developing fruitlets and flower parts produces a characteristic corky blemish on the fruit. For the remainder of its life-cycle, C. verbasci is found on herbaceous hosts and is also a predator on mites and aphids. Action thresholds for treatment of C. verbasci are tied to the petal fall beating tray count. Light skinned cultivars exhibit the damage more readily and a count of 1 C. verbasci nymph per beating tray on 'Golden Delicious' can cause economic damage. On darker skinned cultivars, the action threshold is 4 nymphs per tray. There are currently no known natural enemies of C. verbasci and biological control strategies are not at this time possible. Treatment is applied at bloom time and the choice of materials is severely limited by potential bee toxicity. C. verbasci was sampled using a beating tray at petal fall. A range of other pests and natural enemies were also recorded at this time.

C. verbasci populations were lower overall in 1998 than in 1997, with the Oroville site having the highest population (Tables 4, 19, 34; Fig. 13). Of the 85 blocks sampled, 11% exceeded the threshold of 1 nymph/tap, with a maximum of ca. 4 nymphs/tap. As in the two previous seasons, no differences were found due to management regime. Because this is an early season pest with very specific spray timing at bloom, the overall program may have little effect on this pest. In addition, the ovipositing adults migrate into orchards in late summer and fall, too late to be much influenced by the seasonal program. Of the other pest and natural enemy species recorded at this time, only stinging bugs were significantly more numerous in the mating disruption blocks. However, the number in either regime was too low to be biologically significant.

Leafrollers (Pandemis pyrusana and Choristoneura rosaceana): Leafrollers have become increasingly important direct pests of apple and pear in the Western states, and have emerged as the most significant problem in mating disrupted orchards. These species overwinter as early instar larvae and complete two generations each year. Fruit can be damaged by overwintering generation larval feeding during the early spring, as well as by summer generation larval feeding on larger fruits. In heavily infested orchards, damage by leafrollers can exceed that caused by codling moth. OP resistance has been documented, and is probably the cause of increased populations in orchards. There are several important groups of parasitoids which attack leafroller (e.g., Colpoclypeus florus (Walker), [Hymenoptera: Eulophidae], tachinids [Diptera: Tachinidae], ichneumonids [Hymenoptera: Ichneumonidae], braconids [Hymenoptera: Braconidae]), and biological control could play a significant role. Leafroller larvae were sampled near the end of the 2 generations, when the last 3 larval instars predominated (petal fall and late July in Washington) by examining 25 shoots per tree on 20 trees per block. Adult male moths were monitored with pheromone traps (1 per 10 acre block). Traps were checked weekly, and lures replaced every 4 weeks. Larval parasitism was assessed by collecting late-instar larvae from heavily infested blocks. These samples were sent to Eric LaGasa, Washington State Dept. of Agriculture, for rearing and identification of both moths and parasitoids. Fruit damage was assessed in the pre-harvest sample.

Overwintering leafroller larval populations were generally low at the petal fall sample in 1998, and no differences in population means due to management regime were found. However, in the mid-season sample, significantly more live larvae were found in the conventional blocks than in the mating disruption blocks (Table 3, 18, 33; Figs. 14-17). This same trend occurred in 1997. This is in contrast to previous experience in mating disruption blocks, where leafrollers became quite problematic with the reduction in OP usage (Knight 1995; Gut and Brunner 1994, 1996). It is also in contrast to 1996 results of this project, where higher levels of leafrollers (shoot damage) were found in mating disruption blocks. It is likely that an aggressive Bt program in the mating disruption blocks was pursued to counterbalance the reduction in OPs. Leafroller populations were highest in the Oroville project, whereas the Nickell and P&G sites had low populations. The within-project comparison of management regime showed no difference the number of leafroller larvae, although at the Oroville site the conventional blocks had a ca. 6-fold higher population (Table 3). There was a statistical difference, however, in the number of damaged shoots at this site. It is clear that the overall difference in leafroller population due to regime (Table 18) was attributable almost solely to the population trends at the Oroville site.

Mites: Mites are significant indirect pests on apple in Northwest orchards. The pest species belong primarily to the spider mite (Tetranychid) group and include McDaniel spider mite (Tetranychus mcdanieli McGregor), twospotted spider mite (Tetranychus urticae Koch) and the European red mite (Panonychus ulmi [Koch]). Eriophyid (rust) mites (Aculus schlechtendali [Nalepa]), although phytophagous, are significant primarily for their role as an alternate prey for predatory mites. The predominant predatory mites in the PNW are Typhlodromus occidentalis (Acari: Phytoseiidae) and Zetzellia mali Ewing (Acari: Stigmaeidae). Mites reproduce rapidly and can produce up to a dozen generations each growing season. Feeding damage reduces photosynthetic leaf area and causes premature defoliation; these may affect fruit size and quality of the current season's crop, and reduced return bloom or set of the subsequent season's crop. However, relatively high populations can be tolerated under Washington's irrigated growing conditions. Integrated control of tetranychid mites in apple is one of the classic examples of successful implementation of IPM. Regulation of pest mites in PNW apple orchards by predatory mites has been nearly complete on up to 80% of the acreage for the past 2 decades.

Phytophagous and predatory mite populations were sampled using 2 types of assessments: a standard leaf-brushing machine (a composite sample of 200 leaves per block, 2x per season in late-June and early August); and a binomial sample (presence/absence or % infested), done 3 times during the season (mid-June, mid-July, and early September). The binomial count was used to trigger a supplementary leaf-brushing sample when greater than 80% of the leaves were infested. This occurred only once during the 1998 season:

Of the 170 leaf-brush samples taken, only 16 (9%) had populations which exceeded 1 mite/leaf, with a maximum of 12.5 mites/leaf (Knapp/House1, August) (Tables 9, 24, 39; Figs. 18, 19). The average predatory mite population for the samples was 0.4/leaf, similar to 1997 levels. This is a strong indication that integrated mite control is still an important and widespread phenomenon in Washington apple orchards. OP usage, except where very extreme, is generally not detrimental to the principal mite predator, T. occidentalis, which is highly tolerant of OPs. Despite this, phytophagous mite populations were higher in conventional blocks during both the mid-June and mid-August evaluations, and predatory mites were higher in mating disruption blocks during the June evaluation. This indicates that there was some slight detrimental impact of conventional programs on integrated mite control, which resulted in higher mite populations in these blocks.

The mite binomial samples ranged from 0 to 93% infested leaves of the 186 samples taken (Tables 8, 23, 38). Only 3 samples (2%) exceeded the threshold of 80% (predicted 4.7 mites/leaf, Beers et al. 1993) necessitating a leaf-brush sample as a supplement. Oroville had the highest percentage infested leaves at all three evaluation periods. As with the mite brushed-leaf sample, the populations were higher in the conventional blocks on all three dates, although the majority of the populations did not exceed the threshold.

Fruit damage samples were taken at mid-season (timed to coincide with the end of the first generation of codling moth) and just before harvest. In 1998, only codling moth was evaluated at the mid-season sample, since the timing and intent are specific for codling moth. Forty half-fruit (20 from the upper canopy, 20 from the lower canopy) on 50 trees per block were examined for damage, for a total of 2,000 half-fruits/block. Ladders were used if necessary to examine fruit in the upper part of the canopy. For the mid-season sample, codling moth damage was recorded as either a sting or an entry. In the preharvest sample, whole fruit were examined, and the types of damage evaluated included that by codling moth (stings and entries recorded together), leafrollers, lygus, C. verbasci, thrips, aphids (honeydew and sooty mold), stink bugs and cutworms.

Both codling moth stings and entries were very low overall (<0.05%) during the mid-season sample (Tables 10, 25, 40; Figs. 20-22 ). However, the percentage codling moth entries was significantly higher in the conventional blocks than in the mating disruption blocks at this time. Similarly, the percentage of codling moth damage was 7-fold higher in conventional blocks at harvest in comparison to mating disruption blocks. However, of the 95 blocks sampled at harvest, 68% had no codling moth damage whatsoever, and only 11% had >0.5% damage, with a maximum of 11.1% damage (Oroville/Kelly). The overall level of codling moth damage in 1998 was higher than in 1997. While the mating disruption technique, as it is being implemented in Washington apples (with supplemental sprays of OPs) has produced superior codling moth control in comparison to conventional programs, there are still interseasonal fluctuations in damage levels.

Leafroller damage was also higher in 1998 in comparison to 1997, with 0.29% damage overall in conventional blocks and 0.10% damage in mating disruption blocks (Tables 10, 25, 40; Fig. 23). This difference was statistically different in 1998. The maximum level of damage was 1.6%, with 8% of the 95 blocks having „ 0.5% damage, and 35% of the blocks with no damage. As in the previous 2 years, the Oroville and Vantage sites had the highest levels of damage. The higher level of fruit damage found at harvest in Oroville corresponds with the higher level of leafroller larval populations found during the most critical damage period, midsummer; the same was not true, however, at the Vantage site, where no live larvae were found during midsummer.

Damage due to thrips was higher in mating disruption blocks than in conventional blocks, although this level was not significantly different. (Tables 10, 25, 40; Fig. 24). The level of damage in 1998 was somewhat higher than in previous years. As in 1997, Parker had the highest level of thrips damage found in the 7 apple sites; in fact, this was the only site that recorded thrips damage at harvest. Although not significant, there was a marked higher level of damage in mating disruption blocks in comparison to conventional blocks. The unequal number of blocks sampled from the two regimes (8 conventional vs. 25 mating disruption) may have masked the statistical differences.

Fruit damage by C. verbasci was generally low at harvest (0.05-0.06% average fruit damage) in mating disruption and conventional sites (Tables 10, 25, 40; Fig. 25); this level is comparable to that found in 1997. No differences in population between the 2 regimes were found. The highest level of damage found was 0.7%, and only 2 blocks of 95 sustained levels „ 0.5%.

Lygus damage did not vary due to regime; this is in contrast to 1997, when mating disruption blocks had higher levels in comparison to conventional blocks (Tables 10, 25, 40; Fig. 26). This is considered a sporadic pest in Washington, and the short window for damage and control make it difficult to manage. The trends for 2 direct sporadic pests (thrips and lygus) to be higher in mating disruption sites is disquieting, since neither is amenable to biological control, and would likely necessitate the use of one or more broadspectrum insecticides for control. However, intensive sampling should help identify blocks at risk, and reduce prophylactic sprays.

The incidence of honeydew-contaminated fruit was low throughout the 7 sites, and no differences occurred due to management regime (Tables 10, 25, 40;).

The incidence of stinkbug damage was highest at the Nickell (0.29%) Knapp (0.18%) and Vantage (0.13%) sites. The other sites had less that 1/10 of one perscent damage and no difference was seen between management regime. (Tables 10, 25, 40, Fig. 27 , 29.)

Incidence of cutworm damage was high throughout the state in 1998, repeating a trend of increasing damage in the past few years. While originally the outbreak was thought to be a sporadic occurrence, there may be a longer term trend toward increasing cutworm problems. No differences were found due to management regime (Fig. 28 , 29).

Pear Samples: The pear sampling was approached quite differently from the apple sampling because of the nature of the pest and natural enemy complex. Instead of focusing on individual insects and/or key events in life cycles, the insect samples were of 3 basic types.. The pre-bloom cluster counts were timed to key events, viz., delayed dormant and tight cluster. The leaf-brushing counts and beating tray samples were repeated at frequent intervals throughout the season, and targeted both pests and natural enemies. Because of an incipient problem with leafrollers at the Randall Island site, a leafroller sample was added in 1997. This protocol followed the one for apple. Fruit damage assessments were timed in a manner similar to those for apple, viz, after the 1^st generation of codling moth, and just before harvest.

Other than the key direct pest, codling moth, pear has 2 secondary pests of significance: pear psylla and tetranychid mites. There are regional differences in pressure from these pests. Psylla populations and resulting damage tend to be highest in Washington pear orchards, moderate in Medford, and low in California. Mite problems have historically been more acute in the Medford area than either California or Washington. The pest status varies with cultivar, however; 'd'Anjou', a widely grown cultivar in central Washington, is highly sensitive to mite damage, whereas 'Bartlett' , the predominant cultivar in California, is less sensitive. 'Comice' and the red strains of Bartlett are fairly tolerant of mite feeding, and these cultivars form an important part of the production in the Medford area, although 'Bartlett' and 'd'Anjou' are also a significant proportion of the production.

Pear psylla and mites have a long history of developing resistance to nonselective insecticides, and the history of pear pest management in the west has been driven largely be resistance considerations. Psylla is technically an indirect feeding pest, but fruit russeting caused by honeydew deposition has elevated its pest status. Mite feeding in sensitive cultivars causes a phenomenon known as transpiration burn, and defoliation can be severe during the hot part of the growing season where extensive mite damage has occurred. This defoliation has been shown to affect primarily the return bloom and set of affected pear trees (Westigard 1966).

Pre-bloom cluster count: This sample targeted pear psylla (Psylla pyricola Foerster) eggs, pear rust mite (Epitremerus pyri Nalepa), overwintering females of Tetranychus spp., and P. ulmi eggs. Forty clusters per block (2 clusters from 20 trees) were examined for presence of insects under a binocular microscope. All insects except pear rust mites were counted; the latter was rated on a scale of 0-3.

The Parker site had by far the highest level of psylla eggs in the delayed dormant cluster sample, but Medford had the highest number at the tight cluster evaluation Randall Island had no psylla eggs at either period. There were more eggs in the conventional sites than in the mating disruption sites (delayed dormant only), which is consistent with differences in pear psylla populations in the two management regimes over the past two years. This difference, however, did not persist to the tight cluster evaluation, in either the psylla egg or nymph variable (Tables 11, 26, 41; Figs. 29-32). European red mite eggs were also evaluation during these samples, and both Medford and Parker had moderate populations. No differences were found due to management regime.

Leafroller larval samples: The sampling timing and protocol for leafroller larvae on pear were the same as for apple, viz, visual examination of 20 shoots on 25 trees/block at petal fall and midsummer. In addition, sawfly damage was evaluated because this pest has been sporadically important in the Medford area, and could potentially become more problematic under mating disruption. However, this pest was not found in 1997 or 1998 in any of the sites.

Overall, the number of larvae and damaged shoots found was low in all sites at both sampling periods (Tables 12, 27, 42; Figs. 33, 34). No differences in populations were found due to management regime.

Beating tray counts: This sampling technique has been used historically as the primary means of assessing pear psylla adults, but is equally useful for tracking populations of beneficial insects. Samples were begun before delayed dormant, a typical timing for assessing the overwintering population of pear psylla adults, and continued at biweekly intervals throughout the rest of the season. Beneficial insects recorded included anthocorids, Deraeocoris, earwigs, C. verbasci, and spiders. Psylla nymphs mummified by the parasitoid Trechnites sp. were also recorded. Seasonal averages were used for analyses.

Psylla adults were higher (although not significantly) in the Parker site than the other 2 sites in 1998, similar to 1996-1997 (Tables 13, 28, 43; Figs. 35, 62-67). Although in past years virtually no psylla adults have been found in the Randall Island site, a moderate population was found in 1998. Unlike previous years, adult populations were not significantly higher in conventional sites than in mating disruption sites in 1998. The only difference due to management regime in terms of predators was a higher level of spiders in the mating disruption sites, although the biological importance of this difference is questionable, given the overall low numbers.

Leaf-brushing counts: Leaf-brushing counts targeted pear psylla nymph populations, as well as phytophagous and predatory mites. Sampling began at petal fall, and continued biweekly until about 2 weeks before harvest. Seasonal averages were calculated for the purpose of analysis.

Although Parker had high levels of psylla adults, Medford had the highest nymph numbers (differences not significant). (Table 14, 29, 39; Figs. 36, 56-61). Randall Island site had very low nymphal populations (seasonal average of zero). Like 1996-97 nymphal populations, the populations were significantly higher in conventional blocks than in mating disruption blocks in 1998. This trend for improved control of pear psylla has been one of the most consistent of the project.

Parker had the highest levels of phytophagous mites in 1998, in contrast to the 1997 (Table 14, 29, 44; Figs. 37). Four of the Parker blocks had „ 1.5 mites/leaf as a seasonal average, which is high for mite populations on pear. One of the four blocks (that with the highest seasonal average) was a mating disruption block. No differences due to mating disruption vs. conventional management regimes were found in phytophagous mite populations. The difference in predatory mite populations was significant (higher in mating disruption blocks); however, numbers were too low to draw conclusions.

Fruit Damage: Codling moth fruit damage was assessed at the end of the 1st generation by examining ª 2,000 half-fruits/block. Damage by codling moth, leafrollers, pear psylla, mealybug, lygus, stink bug, earwigs and cutworm was evaluated just prior to harvest by examining 2,000 whole fruits/block (20 high and 20 low from 50 trees/block).

Mid-season codling moth damage was low overall, and no differences occurred among the sites or between the management regime. For the harvest samples, damage was highest in Medford and lowest in Randall Island, a reversal of previous years (Tables 15, 30, 45; Fig. 38). Of the 39 blocks sampled at harvest, 21 (54%) had no detectable damage, and 2 (5%) had „ 0.5% damage. No differences between the 2 management regimes occurred in terms of codling moth damage.

Leafroller damage in 1998 was highest in Medford, whereas Randall Island had suffered the most leafroller damage in 1996-97. Although the mating disruption blocks had a higher mean percentage of leafroller damage, the difference was not significant when management regime was compared (Tables 15, 30, 45; Fig. 39).

Percentage psylla-damaged fruit was ca. 3.4-fold higher in the conventional blocks, which corresponds with the higher level of nymphs in conventional blocks in the seasonal counts (Tables 15, 30, 45; Fig. 40). The same trend occurred in 1997, although the differences were not significant.

Evaluation of true bug damage in 1997-98 distinguished between lygus and stink bugs, and lygus damage was by far the most prevalent. Parker had the highest level of lygus damage of the 3 sites (this occurred in Randall Island in 1997) (Tables 15, 30, 45; Fig. 41). No differences in mean damage due to management regime were found. Earwig, mealybug, and cutworm damage was low overall.

Kogan, M. Area wide management of the codling moth: Implementation of a comprehensive IPM program for pome fruit crops in the western U.S., IPPC/ Ent. OSU, July 1994