DEVELOPMENT AND IMPLEMENTATION OF REDUCED RISK PEST MANAGEMENT STRATEGIES FOR BLUEBERRIES

Transcription

1 DEVELOPMENT AND IMPLEMENTATION OF REDUCED RISK PEST MANAGEMENT STRATEGIES FOR BLUEBERRIES Multi-State Blueberry Project YEAR 3 MICHIGAN REPORT January, 26 Rufus Isaacs, Keith Mason, and Julianna Tuell Small Fruit Entomology Laboratory, Department of Entomology, Michigan State University, East Lansing, MI John Wise Trevor Nichols Research Complex, Fennville, Michigan Carlos Garcia-Salazar Ottawa County Extension, Grand Haven, Michigan Mark Longstroth Van Buren County Extension, Paw Paw, MI For more information regarding this report, contact Keith Mason at masonk@msu.edu or (517) Funded by the USDA Program and Michigan Agricultural Experiment Station

2 Acknowledgements A project of this scale takes a large effort to complete, and we are very grateful to many individuals who have helped with this project in 25. Thanks to the Michigan blueberry growers who provided their land and their ideas for this research; Karlis Galens, Larry Bodtke, Joe DeGrandchamp, Al Ochoa, Rick Kiel, and Bob Carini. Thanks also to Dave Trinka, Ed Wheeler and the staff of MBG Marketing for their assistance with this project. We thank various agrochemical companies for their generous donation of product for testing within this or related projects: Bayer CropScience, Dow Agrosciences, Gowan, DuPont, MGK, and Valent. Many undergraduate students worked to help gather and process the data, including Mike Playford, Kevin Lawracy, Jay Prescott, Jack Langdon and Mark Vander Werp. 2

3 CONTENTS PAGE INTRODUCTION. 4 EVALUATION OF REDUCED-RISK INSECT MANAGEMENT PROGRAM... 6 OVERALL METHODS.. 6 Cranberry Fruitworm 8 Cherry Fruitworm.. 1 Japanese Beetle. 12 JB Adult Monitoring 12 Product Performance for JB Adult Control 13 JB Larval Control Blueberry Maggot. 15 Leafroller Complex Blueberry Aphid 2 Sharp-Nosed Leafhopper.. 22 White-Marked Tussock Moth. 23 Oriental Beetle.. 24 Plum Curculio MEASURE CHANGE IN CONVENTIONAL INSECTICIDE RESIDUES AND THE EFFECTS ON ABUNDANCE OF NATURAL ENEMIES, POLLINATORS, AND SECONDARY PESTS 26 PESTICIDE RESIDUE ANALYSIS.. 26 OVERVIEW OF THE ECONOMICS OF THE REDUCED RISK AND GROWER STANDARD INSECT MANAGEMENT PROGRAMS 27 NATURAL ENEMIES 28 Pitfall Trapping. 28 Natural Enemy Survey.. 3 Traps... 3 Scouting for natural enemies 31 Abundance of Ground Foraging Ants Predator Exclusion Experiments.. 35 Aphid Natural Enemies 38 Aphid Parasitoid Identification POLLINATORS.. 4 SUMMARY. 41 3

4 INTRODUCTION The Blueberry Project is a large-scale, multi-year research and extension effort in the three largest blueberry-producing states in the US. Our overall goal is to measure the positive and negative aspects of reduced-risk insect management programs in blueberry, and to deliver information on this project to our industry clientele. This report contains the Year 3 results from Michigan in 25. The blueberry industry in Michigan and across the Eastern US faces a combination of challenges to achieving high quality insect-free product. To successfully manage multiple species of insect pests, growers must be able to economically monitor and scout their plantings for these pests, and must have effective insecticides available to them when control of pest populations is required. At the same time, the pressures from the food industry for zero contamination with blueberry maggot, Japanese beetle, and other insects creates significant pressure to consistently achieve excellent insect control. Passage of the Food Quality Protection Act in 1996 changed how pesticides are evaluated for use in food crops, and this is leading to reduced availability of conventional insecticides that growers have relied upon for the past 2-5 years. Organophosphate and carbamate insecticides are the first classes of pesticides to have been scrutinized, but others including some fungicides are expected to be re-evaluated in the future as FQPA implementation continues. At the same time, the new regulatory standards are forcing agrochemical companies to bring products to market that have better environmental, worker, and food safety than their predecessors. Products with potential for use in fruit crops, including blueberry, are evaluated by MSU Applied Entomologists, and the most promising are recommended for registration. In recent years, the combined efforts of land grant researchers, commodity group leaders, agrochemical companies and the IR-4 program have brought about the registration of new classes of reduced-risk insecticides for blueberry including naturalytes, neonicotinoids, insect growth regulators, and botanicals. These products have inherently different modes of action from conventional insecticides, and many (but not all) of these products have low impacts on non-target organisms. The availability of new insecticide classes presents opportunities and challenges. Some of the new products have shorter residual activity than conventional insecticides, and in general they are more expensive. At the same time, their lower environmental impact may allow greater survival of natural enemies and pollinators in fields managed with reducedrisk insecticides. There is a significant hurdle to more to adoption of these alternatives in agriculture, however, because growers have relatively little experience with reduced-risk insecticides, they are more confident in well-tested conventional insecticides, and there is a greater cost of using many of the newer types of insecticides. To bridge the gap to adoption, it is critical that new approaches are tested at the commercial production scale. This provides a venue for research and extension personnel to show growers the results and for leading growers to start gaining experience with the new alternatives. From this, further efforts toward adoption have a chance. Without it, change will be minimal. In addition to new insecticides, new tools for pest management are needed, including improved pest prediction models, sampling schemes, cultural controls and biological controls. This report focuses on the pest management, natural enemy, and pollinator measurements during the third year of this project. 4

5 Insecticides Used in Different Management Programs The products available for use within the two management programs tested in the on-farm components of the Michigan Blueberry project during 25 are listed in Table 1. Full insecticide application records are given below in Table 3. All listed products were registered under full Section 3 labels for use in Michigan blueberries during 25. Based on cooperating grower feedback at a project-planning meeting in February 25, we sought to make the insecticide program more economical. Toward this end we recommended that in 25 growers with pressure from CBFW and CFW substitute the pyrethroid Asana for one of the normally required applications of Confirm. Additionally Admire was not used for JB grub control, Evidence from the first two years suggested that while Admire was very effective at reducing white grub abundance, there was no lasting effect on the number of Japanese beetle adults found on bushes. Additionally, we avoided the use of SpinTor for fruitworm or fruit fly control to maintain costs within the level growers would be willing to pay. In some cases, as explained below, growers were unable to completely stay with the product choices in the or Grower Standard program column and applied a broad-spectrum insecticide to fields or reduced risk insecticides to Grower Standard fields to achieve insect control. This reflects the reality of current pest management challenges or confidence in new pesticides, respectively. Table 1. Insecticide options for insect management in Grower Standard and programs in blueberry during 25. Timing Target Pest* Grower Standard Pre bloom RBLR, OBLR Lannate, Asana Guthion, Diazinon Confirm, B.t., SpinTor, Esteem Malathion Bloom CBFW, CFW, RBLR, B.t. Confirm, B.t., SpinTor, Post bloom OBLR CBFW, CFW, OBLR, RBLR Diazinon, Guthion, Asana, Lannate, Imidan Esteem Confirm, B.t., SpinTor, Asana** BBA Diazinon, Lannate Asana, Provado, Mid-season JB adults JB adults Imidan, Malathion, Asana, Sevin Imidan, Sevin, Asana Malathion, Provado, Neem, Asana** Asana**, Provado Pre-harvest BBM BBA JB adults Diazinon, Imidan, Sevin, Asana Diazinon, Lannate Asana, Imidan, Malathion, Sevin Asana**, SpinTor, Provado Azadirect Provado, Evergreen BBM Imidan, Malathion, Sevin SpinTor, Provado *OBLR, obliquebanded leafroller; RBLR Red-banded leafroller, CBFW, cranberry fruitworm; CFW, cherry fruitworm; BBA, blueberry aphid; JB, Japanese beetle; BBM, blueberry maggot. ** Asana is an interim product for use in this initial reduced-risk program. Provado is not registered for use against BBM, however control may be attained when spraying for JB. 5

6 EVALUATION OF REDUCED-RISK INSECT MANAGEMENT PROGRAM OVERALL METHODS Project Design Six growers each provided access to two blueberry fields (4 to 2 acres) of either Bluecrop (5 farms) or Jersey (1 farm) cultivars. Three farms were located in each of the two main blueberry-producing counties in southwest Michigan, Van Buren and Ottawa counties. At each farm one field received a season-long insect management program using reduced risk insecticides () while the other received a program employing broad spectrum insecticides (the Grower-Standard - see Table 1). Plant developmental stages for each farm were monitored and dates of major developmental stages are given in Table 2. In each field key insect pests were monitored weekly with pheromone (or fruit attractant) traps (2 per field) and scouting procedures to determine the abundance of insect pests, damage and beneficial insects. Each field was divided into four blocks to ensure scouting covered as much of the field as possible. The first block, designated as the field border, was always adjacent to a woodlot. The remainder of the field was divided in three additional blocks and data from these blocks were pooled and represent the field interior. Particular methods for data collection are given below in the sections for each pest. The following key pests of blueberry were monitored: cranberry fruitworm, cherry fruitworm, Japanese beetle, and blueberry maggot. Pests of less importance were also monitored in each field, these included: blueberry aphid, oblique-banded leafroller, red-banded leafroller, plum curculio, sharp-nosed leafhopper, and tussock moth. Growers were informed weekly of all pest pressure in both and Grower Standard fields, and if pest populations reached a critical level, recommendations for insecticide applications were made for the plots using the pesticide options in Table 1. In some instances growers applied insecticides to fields even though scouting of those fields resulted in recommendations not to spray. Records of insecticide applications are given for each farm in Table 3. Fruit clusters were collected from each field directly before each harvest to determine levels of infestation caused by key pests (see below). Prior to harvest, leaf and fruit samples were collected from each field for pesticide residue analysis (described below, see p. 26). Data Analysis Season totals of insects captured in traps, insects detected on bushes or damage from pest insects were compared across the six farms using analysis of variance (ANOVA) on log(x+1) transformed data to determine significant differences between the and Grower Standard programs. Data were analyzed to test for differences between border and interior samples when scouting was split between these positions. The Mann Whitney U test was used when transformation did not stabilize variance. All data analysis was accomplished with Statview Ver. 4.57, unless noted otherwise. 6

7 Table Crop phenology. Major crop development stages observed at weekly visits for pest scouting. All fields were cv. Bluecrop except Farm 3 (cv. Jersey). Farms 1-3 are in Van Buren Co. and Farms 4-6 are in Ottawa Co. Plant Stage Farm 1 Farm 2 Farm 3 Farm 4 Farm 5 Farm 6 Bud Break Pink bud Bloom Petal Fall Fruit Coloring st Harvest nd Harvest rd Harvest* *Only two harvests were made at Farms 3 and 5. Table 3. Insecticide applications made to blueberry fields in the 25 Michigan Project. Farms 1-3 are located in Van Buren Co. and farms 4-6 are in Ottawa Co. All fields were cv. Bluecrop except Farm 3 where both fields were cv. Jersey. For key to abbreviations see Table 1. FARM GROWER STANDARD PROGRAM Compound Date Target Compound Date Target 1 Dipel 6 LR, FW Confirm 6 FW Guthion 6/8 FW Asana 6/8 FW Guthion 6/17 FW Confirm 6/17 FW Imidan 7/19 JB adult Provado 7/19 JB adult Imidan 8/2 JB adult Provado 8/2 JB adult 2 Confirm 6/8 FW Confirm 6/8 FW Imidan 6/2 FW Asana 6/17 FW Sevin JB adult Sevin 7/14 JB adult 3 Guthion 6/7 FW Confirm 6/7 FW Guthion 6/2 FW Asana 6/16 Provado 7/3 BBA Provado 7/9 BBA Imidan 7/14 JB adult Malathion 7/22 JB adult 4 Guthion 6/7 FW Confirm 6/2 FW Guthion 6/2 FW Confirm 6/25 FW Imidan 7/8 JB adult Provado 7/23 JB adult Imidan 7/2 JB adult 5 Guthion 6/8 FW Confirm 6/8 FW Lannate 6/8 BBA Confirm 7/1 FW Guthion 7/1 FW Lannate 7/1 BBA 6 Confirm 6/1 LR, FW Confirm 6/1 FW Asana 6/17 FW Confirm 6/15 FW Imidan 7/2 JB adult Sevin 7/1 JB adult, BBA Sevin 7/12 JB adult, BBA Sevin 7/28 JB adult Asana 8/23 JB adult Average # Sprays Standard Insecticides 4..7 Applied at grower s discretion after receiving scouting based recommendation to not spray. 7

8 Cranberry Fruitworm (CBFW), Acrobasis vaccinii (Riley) During the second week of April 25, two pheromone traps were placed in each field on border bushes at least 5m apart. The number of moths in traps was recorded each week and pheromone lures were replaced monthly. Beginning at the time of consistent male moth captures (sustained flight), twenty bushes along the border of each field were visually examined for one minute each and the number of CBFW eggs, cherry fruitworm (CFW) eggs, CBFW damaged clusters and CFW damaged clusters was recorded. Immediately prior to each harvest, 1 randomly selected clusters with ripe fruit (5 clusters from border bushes and 5 from interior bushes) were collected from each field. Clusters were returned to the laboratory where the number of fruitworm-infested clusters was recorded. Clusters were placed in 12 quart containers lined with sand to allow larvae within the collected fruit to complete development, drop from the fruit and form protective over-wintering sacks (hibernaculae). The sand was sifted at least one month after fruit collection and the number of hibernaculae and pupae were recorded. Sustained CBFW male flight began earlier in Van Buren Co. (3 May) compared to that in Ottawa Co. (6 June) and flight ended around 4 July in Van Buren and 11 July in Ottawa (Figure 1a, 1b). Eggs were detected 1 week after the start of male flight in both counties with peak egglaying occurring around the time of peak male flight, 13 June in both counties (Figure 1c). Very few eggs were detected at Ottawa County Farms (Figure 1d.). Feeding damage from CBFW larvae was seen throughout the period of egglaying and reached its maximum 2 weeks after peak moth capture in both counties (27 June). In general, pressure from CBFW was higher in Van Buren Co. than in Ottawa Co. These differences were not tested statistically due to the small sample size. The total number of moths captured per field was similar between programs in both counties. Across all farms there was no statistical difference between programs in the total number of CBFW trapped during the season (F 1,1 =.6, p =.99). In both counties, the total number of CBFW eggs and damage detected during scouting was higher in fields compared to that in Grower Standard fields, however these differences were not statistically different for eggs (F 1,1 =.27, p =.62) or damage (F 1,1 = 2.4, p =.15). Fruit samples taken from each field immediately prior to each harvest show a trend similar to that in the data from weekly scouting, i.e. there were more infested clusters in samples taken from fields compared to those from Grower Standard fields (Table 4). However the number of CBFW larvae that survived (hibernaculae) in the fruit was lower in samples from blocks than in those from Grower Standard fields (Table 4). The different modes of action of the insecticides used in and Grower Standard fields may help explain this difference. In general the program employed a growth regulator, Confirm 2F (16oz/acre) to control fruitworm, while a neurotoxic insecticide Guthion 5WP (1 lb/acre) was generally used in Grower Standard fields for these pests (Table 3). A growth regulator such as Confirm must be ingested to be effective and thus would take longer to kill an insect compared to a neurotoxin that kills on contact, such as 8

9 Guthion. Therefore we would expect to see higher levels of initial damage when Confirm is used. Nevertheless, the strong performance of Confirm seen in the low number of hibernaculae recovered from fruit samples in plots shows that this insecticide kills larvae later in development. A substantial reduction in pest population may result after Confirm is used over the four years of this study. After the first three years of this study we have not seen the decline of CBFW populations in fields. We did observe a more than two-fold increase in CBFW infested clusters in both and Grower Standard fields in 25 compared to 24. We suspect the heavy snow cover of the 24-5 winter and the warm spring of 25 contributed to the increase of CBFW infestation. In 25 there were no significant differences between programs in the number of infested clusters (F 1,28 =.3, p =.87) or the number of hibernaculae (F 1,28 =.58, p =.45). CBFW adults 5 4 a. 5 4 b. Moths per trap egg egg Cluster Damage Cluster Damage 5/3 1/3 5/3 Eggs or damaged clusters (per 2 bushes) 1/ egg egg Cluster Damage Cluster Damage 5/3 1/3 5/3 1/3 CBFW eggs and damage c. d. Van Buren Ottawa Figure 1. Cranberry fruitworm (CBFW) phenology during 25. Moth captures are given for (a.) 3 farms in Van Buren County and (b.) 3 farms in Ottawa County. Eggs and damage observed during weekly scouting at (c.) Van Buren and (d.) Ottawa County farms. 9

10 Table 4. Total fruitworm infestation and number of hibernaculae recovered from fruit collections in 25*. Data from field borders and interiors taken just prior to each of three harvests are pooled**. Infestation data includes cluster infestation from cranberry fruitworm and single berry infestation from cherry fruitworm. No cherry fruitworm pupae were recovered from these samples. GROWER STANDARD Farm FW Infestation Hibernaculae FW Infestation Hibernaculae CBFW CFW Total CBFW CFW Total Total (6.9%) (8.4%) 1 * 1 clusters (5-8 berries per cluster) per field at each of 3 pre-harvest timings (3 clusters total per field) ** Only two harvests were completed at Farms 3 and 5. percent infested clusters (16 clusters total per treatment program). Cherry Fruitworm (CFW), Grapholitha packardi (Zeller) Cherry fruitworm adult flight monitoring, scouting for eggs and damage and harvest fruit collection was performed as described above for cranberry fruitworm. During scouting and fruit collection, fruitworm damage was differentiated into multiple berry damage (CBFW) and single berry damage (CFW), but fruit were not opened to identify larvae. Cherry fruitworm flight began 9 May in Van Buren and Ottawa Counties, and the peak of CFW moth captures occurred on 3 May in Van Buren and Ottawa Co. (Figure 2). The number of moths caught in traps was generally higher at Ottawa Co. farms compared to that at Van Buren Co. farms. The total number of moths trapped during the season was not significantly different between and Grower Standard programs (F 1,1 =.1, p =.76). As seen in Figure 2, egglaying was first detected 23 May in Van Buren and 3 May in Ottawa Co. In both counties peak egglaying occurred after the date of peak moth captures. Paradoxically more CFW damage was observed in Van Buren Co. fields even though more moths were trapped at Ottawa Co. farms. This may have occurred because early cluster damage from CBFW is indistinguishable from early CFW damage. That is, a CBFW larva enters a single berry and later webs berries in the cluster together with silk and frass to form the characteristic multiple berry damage that is diagnostic of CBFW. So some of the damaged clusters originally scored as single berry damage may have been caused by cranberry fruitworm. In Van Buren County, the total number of CFW damage clusters detected during scouting was higher in fields compared to that in Grower Standard fields, but these numbers were similar at Ottawa County farms. Significant differences between programs were not detected for CFW eggs (F 1,1 =.15, p =.7) or damage (F 1,1 =.78, p =.4). Clusters with single berry damage were detected in fruit samples taken prior to harvest, and although the total number of infested clusters was considerably greater in fields, CFW infestation was not statistically different between programs (F 1,28 =.19, p =.67) No CFW pupae were recovered from preharvest fruit samples indicating both 1

11 programs controlled CFW effectively. Although in general more cranberry fruitworm and cherry fruitworm damage and infestation was observed in Fields compared to Grower Standard fields (Table 4), no statistical differences were detected when comparing these measurements across management programs. Moreover, fewer fruitworm larvae survived in program fields compared to that in Grower Standard fields, suggesting that even though a growth regulator such as Confirm may allow some damage to occur, it still prevents larval development and may be effective at suppressing populations of fruitworm pests over multiple seasons. This may not be acceptable to growers, however, for meeting fruit quality standards. CFW adults a. 8 b. Moths per trap /3 1/ egg egg single berry damage single berry damage egg egg single berry damage single berry damage 5/3 1/3 5/3 5/3 1/3 1/3 CFW eggs and damage Eggs or damaged clusters (per 2 bushes) c. d. Van Buren Ottawa Figure 2. Cherry fruitworm (CFW) phenology during 25. Moth captures are given for (a.) Van Buren County farms and (b.) Ottawa County farms. Eggs and damage observed during weekly scouting at (c.) Van Buren and (d.) Ottawa County farms. 11

12 Japanese Beetle (JB), Popillia japonica (Newman) JB Adult Monitoring The Japanese beetle is the most important insect pest for Michigan blueberry growers. Control of this pest is challenging because there is a zero tolerance for insects in harvested blueberries and adult beetles are present throughout the harvest period. Because applying insecticides with ground spray equipment destroys ripe fruit, growers need to utilize products that have both long residual activity and short pre-harvest intervals. This allows blueberry producers to protect fruit, and to reduce the risk of insect contamination while harvesting fields every 7 to 1 days. Beginning in the last week of June each field was scouted weekly for Japanese beetles and JB feeding damage. Fifteen bushes along the border and fifteen bushes in the field interior were examined visually (1 min per bush) and the number of beetles and the number of damaged berries on each bush was recorded. At all farms in 25 we found Japanese beetle pressure (beetles and damaged berries per bush) was generally similar to that in 24. Japanese beetle adults were first detected on 27 June in Van Buren County and 5 July in Ottawa County, and beetles were detected at farms in both counties until 29 August (Figure 3a and 3b). As seen in Figures 3c and 3d, JB feeding damage was first detected 27 June in both counties and feeding damage was observed until 15 August in Ottawa Co. and 29 August in Van Buren Co. Generally higher numbers of beetles were found in Van Buren County, but the number of damaged berries per bush were similar between counties (Figure 3). Statistical comparisons of these differences were not made due to small sample size. We compared the total number of beetles and damaged berries per bush (for the whole season) at field borders and in field interiors. No significant difference in the number Japanese beetle adults on bushes on the borders of fields compared to the field interiors was detected (F 1,22 =.12, p =.73), however significantly higher numbers of damaged berries were found on the interiors of fields compared to the field borders (F 1,22 = 19.8, p =.2). Because there was significantly more beetle feeding damage on interior bushes, we compared the season-long total number of beetles and damaged berries per bush at border and interior locations separately between Grower Standard and treatment programs. There were no significant differences in the number of beetles per bush (F 1,1 =.2, p =.99) or the number of damaged berries per bush (F 1,1 = 1.9, p =.19) between programs on border bushes. There were no significant differences between programs when comparing beetles per bush (F 1,1 =.2, p =.9) or beetle feeding damage in fields compared to Grower Standard fields (F 1,1 =.34, p =.57). These results demonstrate that, viewing the season as a whole, the program, which generally used Provado for JB adult control, and the Imidan-Malathion-Sevin based Grower Standard program controlled Japanese beetle equally well. 12

13 Japanese beetle adults Beetles per bush GSTD IN IN GSTD OUT OUT a. b GSTD IN IN GSTD OUT OUT 5/3 1/3 5/3 JB damaged berries per bush GSTD IN IN GSTD OUT OUT 5/3 1/3 1/3 Japanese beetle damaged berries c d. GSTD IN IN GSTD OUT OUT 5/3 1/3 Van Buren Ottawa Figure 3. Abundance of Japanese beetle in (JB) adults and damage at Van Buren and Ottawa County study sites in 25. Product Performance for JB Adult Control We were unable to evaluate the performance of specific insecticides against Japanese beetle in 25 because of low numbers of beetles in all fields during the harvest period when insecticides are normally applied for Japanese beetle control. Furthermore, growers used prophylactic insecticide applications before beetle pressure could increase, and this prevented the acquisition of the precount data that are needed to determine insecticide efficacy. 13

14 JB Larval Control In previous years of this project, we evaluated the efficacy of Admire 2F for control of JB larvae, and found this product to very extremely effective at reducing the abundance of JB grubs. Despite consistently showing that grub abundance is lower in Admire-treated fields, it appears this may be of limited benefit to growers when applied around individual fields. At the start of JB adult emergence in 24, lower numbers of adults were found in fields compared to that in Grower Standard fields but this difference dissipated midway through harvest (2 August), presumably due to beetles moving into blueberry fields from surrounding areas. The added expense of applying Admire ($75/acre for a full application) to reduce adult beetle numbers for 2-3 weeks may not be cost effective in a market where there is a zero tolerance for this pest in harvested fruit. Therefore in the interest of developing an economically feasible reduced risk spray program, Admire was eliminated from the spray program for fields. To determine the longevity of previous Admire treatments, the abundance of grubs was assessed in and around and Grower Standard fields during Spring (May and June) and again in Fall (September and October) 25 by taking 25 soil samples each at the border and in the interior of each field using a golf-course cup cutter. Because JB abundance can vary greatly between border and interior positions, these samples were analyzed separately. In Spring sampling, overall grub density was almost 19-fold lower in fields compared to Grower Standard fields, and this difference was significant at field borders (F 1,1 = 7.15, p =.2), but not in field interiors (F 1,1 = 2.2, p =.17; Figure 4a). The situation was reversed in Fall 25. There was significantly lower grub abundance in compared to Grower Standard field interiors (F 1,1 = 6.37, p =.3), but no significant difference at the field borders (F 1,1 = 4.1, p =.7; Figure 4b), despite a numerical difference in grub density a. GSTD * Spring Grubs per sq b. GSTD Fall * Interior Border Figure 4. Japanese beetle larval abundance in and GSTD fields in (a) Spring 25 and (b) Fall 25. * means are significantly different p <.5. 14

15 Blueberry Maggot (BBM), Rhagoletis mendax (Curran) On 6 June 25, two unbaited yellow sticky traps with ammonium acetate chargers were placed in each field on border bushes at least 5m apart. Traps were checked weekly and the number of adult BBM flies was recorded. Traps were replaced every other week and ammonium acetate chargers were replaced as needed. To estimate the level of fruit infestation by BBM, 1 randomly selected clusters with ripe fruit (5 clusters from border plants and 5 from interior bushes) were collected from each field immediately prior to each harvest. Clusters were returned to the laboratory and placed in 12-quart containers lined with sand to allow larvae within the collected fruit to complete development, drop from the fruit and form pupariae. The sand was sifted at least one month after fruit collection to determine if pupariae were present, as a measure of maggot survival. The first adult BBM were caught on 4 July in Van Buren Co. and on 18 July in Ottawa Co. (Figure 5). These emergence dates are earlier than those observed in 23 and 24. The number of adults trapped at Van Buren Co. farms was similar to that at Ottawa Co. farms. More adult BBM were trapped in fields as compared to Grower Standard fields in both counties (Figure 5), however there was no significant difference in the total number of adult BBM trapped when comparing programs (F = 1.8, p =.21). In 25 BBM emergence overlapped the period when insecticides were applied for Japanese beetle control, so growers in this project did not apply insecticides specifically for BBM (Table 3). BBM Adult.5.4 a..5.4 b /3 1/3 5/3 Adults per trap 1/3 Van Buren Ottawa Figure 5. Adult BBM abundance in 25 in and fields in (a) Van Buren and (b) Ottawa counties. No BBM pupariae were found in fruit infestation samples taken during 25, indicating that BBM was effectively controlled by both and Grower Standard programs (Table 5). This result is consistent with our findings from 23 and

16 Table 5. BBM pupariae recovered from preharvest fruit samples*. Data from field borders and interiors taken just prior to each of 3 harvests are pooled**. FARM GROWER STANDARD 1 2 3** 4 5** 6 Total * 1 clusters (8-12 berries per cluster) per field at each of 3 pre-harvest collections (3 clusters total per field). ** Only two pre-harvest samples were taken at Farms 3 and clusters total per treatment program. 16

17 Leafroller Complex At least two species of leafroller are sporadic pests of blueberry in Michigan. Historically there has been only occasional need to apply insecticides exclusively for leafrollers as conventional management of other insect pests has reduced the pressure from leafrollers in Michigan blueberries. The two most prevalent species, the Red Banded leafroller (RBLR), Argyrotaenia velutinana (Walker), and the Oblique banded leafroller (OBLR), Choristoneura rosaceana (Harris), are also pests of tree fruit crops throughout the fruit growing regions of Michigan. In blueberry, leafroller larvae feed on leaf and flower buds as well as developing leaves and fruit. Larvae construct shelters using plants parts webbed together with larval silk. RBLR overwinter as pupae and adults emerge in mid to late spring, whereas OBLR overwinter as late stage larvae, and adults emerge in late spring or early summer. Because larvae of these two species may be present during early spring, these species are managed as a complex and in the data below larval and damage data are pooled for the two pests. At all farms one RBLR pheromone trap was deployed in each field during the second week of April. Traps were placed on border bushes. OBLR traps were deployed in all fields on 23 May in similar fashion on a row at least 5m away from the RBLR trap. Beginning on 18 April each field was scouted weekly for the presence of leafroller feeding damage and larvae. During each visit 5 leaf clusters and 5 flower clusters on 1 shoot on each of 15 bushes (five border and 1 interior bushes) per field were visually examined for the presence of leafroller feeding damage and leafroller larvae. RBLR adults were first caught on 18 April at Ottawa County and Van Buren County farms. Moth catches indicate that flights may have already begun by the time of trap placement (Figures 6a and 6b.). Three similarly timed peaks in moth captures were detected in both Van Buren County and Ottawa County, and the average number of moths caught was similar between the two counties (Figure 6a and b). Across both counties there was no difference between and Grower Standard fields in the total number of RBLR caught during the season (F 1,1 =.11, P =.75). OBLR were first caught at Van Buren County on 14 June and on 2 June at Ottawa County farms (Figure 6c and d). In general, similar numbers of moths were caught at Ottawa and Van Buren County farms. In both counties two peaks of moth captures were detected, and these peaks occurred earlier in Van Buren than in Ottawa County (Figure 6c and d). There was no significant difference in the total number of moths caught in and Grower Standard fields (F 1,1 =.8, p =.78). 17

18 RBLR adult Moths per trap a b /3 1/3 5/3 1/3 OBLR adult Moths per trap c. d /3 1/3 5/3 1/3 Van Buren Ottawa Figure 6. Leafroller abundance in and Grower Standard fields. RBLR trapped in a) Van Buren County, and b) Ottawa County fields. OBLR caught in c) Van Buren County, and d) Ottawa County fields. Leafroller larvae were first observed on 25 April in Van Buren Co. and 2 May in Ottawa Co. The peak of larval abundance occurred around 3 May in both counties (Figure 7a and b). Larval abundance was low throughout the season, and there was no significant difference in the number of leafroller larvae recorded during the entire season in fields compared to that in Grower Standard fields (F 1,1 =.2, p =.89). Evidence of leafroller feeding was present throughout the season, and the incidence of feeding damage declined sharply at the end of May and remained low until the end of the season (Figure 7c and d). When comparing and Grower Standard programs, there was no significant difference in the season total of leafroller-damaged clusters (F 1,1 =.1, p =.98). 18

19 Leafroller larvae Damaged clusters (out of 5) Larvae per 5 clusters a. c. 5/3 1/ Leafroller Feeding b. d. 5/3 1/3 5/3 1/3 5/3 1/3 Van Buren Ottawa Figure 7. Leafroller larvae and feeding damage in and Grower Standard fields. Abundance of leafroller larvae in a) Van Buren County, and b) Ottawa County fields. Incidence of leafroller feeding damage in c) Van Buren County, and d) Ottawa County fields. Data for RBLR and OBLR larvae and damage are combined. In 25, two farms required insecticide applications specifically targeted for leafroller control, while at the remaining farms, control was attained with insecticides applied for fruitworm control (Table 3). Overall, the program (consisting of Confirm and Asana 3 farms) and the Grower Standard program (using Dipel, Guthion, Imidan or Asana) effectively controlled leafrollers. 19

20 Blueberry Aphid (BBA), Illinoia pepperi The blueberry aphid is a sporadic pest in Michigan blueberries. Aphid colonies are commonly found on the new shoot growth at the base of the bush, and in heavy infestations, colonies can be found in the fruiting zone of the plant. BBA is known to be a vector of shoestring virus and this is of great concern to growers with susceptible varieties such as Jersey. In general, Michigan blueberry growers do not target insecticide sprays specifically for BBA unless aphids are present in the fruiting zone or if aphids are detected in a cultivar susceptible to shoestring virus. Beginning on 9 May, two young shoots at the base of each of 15 plants (5 border bushes and 1 bushes in the field interior) in each field were examined for the presence of blueberry aphid and/or parasitized aphids (mummies), and the number of bushes with aphids or mummies present was recorded. Additional data on BBA and BBA parasitism were compiled for the biocontrol objective for this project and are presented below. BBA were first detected on 23 May in Ottawa Co.and 3 May in Van Buren Co. and aphids were present on new shoot growth until the end of August in both counties (Figure 8). Aphids were not seen in the fruiting zone during scouting for other pests. Throughout the season, the number of bushes with aphids was similar across counties (Figure 8a and b). Only the grower at Farm 3 applied an insecticide (Provado) for aphid control in fields while 2 growers (Farms 3, 5) used insecticides (Provado or Lannate) specifically for aphid control in Grower Standard fields (Table 3). A greater total number of bushes with aphids was recorded in Grower Standard compared to fields, but this difference was not statistically significant (F 1,1 = 3.5, p =.11), and this indicates both control programs worked equally well. Aphid mummies were first detected 2 weeks after aphids were detected in Van Buren County (6 June) and Ottawa County (13 June; Figure 8c and d). The number of bushes with aphid mummies was generally higher at Ottawa County farms compared to those in Van Buren County, but this difference was not tested statistically. There were significantly more total bushes with mummies recorded in Grower Standard fields compared to fields (F 1,1 = 1.83, p =.21). The greater number of bushes with mummies present recorded in Grower Standard fields may be the result of increased host availability, because aphids were more abundant in Grower Standard fields. 2

21 Blueberry aphid Bushes with aphids (out of 5) a b. 5/3 1/3 5/3 1/3 Parasitized aphid Bushes with parasitized aphids (out of 5) c. 5/3 1/ d. 5/3 1/3 Van Buren Ottawa Figure 8. Bushes with blueberry aphid present on new growth in and Grower Standard fields in 25 at (a) Van Buren County and (b) Ottawa County farms. Bushes with parasitized blueberry aphid present on new growth in and Grower Standard fields at (c) Van Buren County and (d) Ottawa County farms 21

22 Sharp-Nosed Leafhopper (SNLH), Scaphytopius magdalensis The sharp-nosed leafhopper is a vector of stunt, a virus-like disease of blueberry present in North America, yet this insect is currently considered a minor or secondary pest in Michigan blueberries. SNLH was monitored to determine whether outbreaks of this secondary pest might increase as broad spectrum insecticide use is decreased in fields managed with reduced-risk insecticides. On 11 April, two unbaited yellow sticky traps were placed in each field on border bushes at least 5m apart. Traps were checked weekly and replaced every 2 weeks through 1 October and the number of adult SNLH was recorded. On 6 June an ammonium acetate charger was added to each trap so that it could also be used to monitor blueberry maggot fly. Fewer SNLH were trapped in 25 (88) compared to 24 (187). In 25 the first catch of SNLH ocurred on 13 June in Van Buren and on 18 July in Ottawa County (Figure 9). As was seen in 24, the peak of SNLH captures occurred in September. There were more SNLH trapped at Van Buren County farms compared to that at Ottawa County farms, but this difference was likely due to the large population at Farm 3. There was no significant difference in the total number of SNLH caught during the season in fields as compared to that in Grower Standard fields (F 1,1 =.35; p =.56). No symptoms of stunt were observed during scouting. The fact that similar numbers of SNLH were caught in and Grower Standard fields indicates the abundance of this insect is not increasing in reponse to the reduction of broad-spectrum insecticide usage in fields. This pest will be monitored for the remainder of this project. Sharp-nosed leafhopper Insects per trap a. 3 b /3 1/3 5/3 1/3 Van Buren Ottawa Figure 9. Abundance of sharp-nosed leafhopper on yellow sticky traps at a) Van Buren County and b) Ottawa County farms. 22

23 White Marked Tussock Moth (TM) Orygia leucostigma (J.E. Smith) The white marked tussock moth is an occasional pest of blueberry. Tussock moth larvae feed on blueberry foliage, and these larvae are covered with hairs that can cause irritation to the skin of agricultural workers that encounter this pest. We began monitoring this insect in 24 to guard against secondary outbreaks in response to reduced broadspectrum insecticide input. One pheromone-baited trap was placed in each field on 25 April. Traps were checked weekly and pheromone lures were changed every 8 weeks. The incidence of larvae was recorded when bushes were scouted for other pests (see above). The first moth captures occurred on 27 June in Van Buren and Ottawa County, and based on Van Buren data it appears there may have been two flights of tussock moth in 25 (Figure 1a and b). The total number of moths trapped during the season was not significantly different between and Grower Standard fields (F 1,1 =.1.74, p=.21). Tussock moth larvae were present at farms in both counties for two periods during the season, 9 May to 27 June and 15 a. TM adults 15 b /3 1/3 5/3 Larvae per 5 bushes Moths per trap 5/3 1/3 1/3 TM larvae c. d. 5/3 1/3 Van Buren Ottawa Figure 1. Abundance of tussock moth adults in pheromone traps at a) Van Buren County and b) Ottawa County farms. Abundance of larvae in scouting samples at c) Van Buren and d) Ottawa County farms. 23

24 18 July to 29 August (Figure 1c and d). Therefore larvae were present for most of the harvest period. More larvae were observed at Van Buren County farms, especially during the later part of the season, but this difference was not compared statistically. The total number of larvae seen in fields compared to Grower Standard fields was not statistically different (F 1,1 =.41, p =.54) indicating the and Grower Standard programs are similarly effective at controlling tussock moth. Oriental Beetle (OB) Exomala orientalis (Waterhouse) The Oriental beetle is an invasive species that is known to be a pest of turfgrass, nursery stock and blueberry in the eastern United States. The larvae of this scarab beetle feed directly on the roots of blueberry bushes and can cause decreased plant vigor and reduced yield. We are monitoring for this pest to determine if this pest is present in Michigan blueberry. Prior to 24 this beetle had not been recorded in blueberries in Michigan. On 23 May, one pheromone-baited beetle trap was placed in each field. Traps were checked weekly and pheromone lures were changed every 8 weeks. Oriental beetle adults were found at one of the Ottawa County farms in July (Figure 11). This indicates this pest is invading the blueberry production areas of Michigan. We will continue to monitor this pest throughout the duration of this project to help determine the potential risks of oriental beetle to the Michigan blueberry industry. Oriental beetle 1 8 a. 1 8 b /3 5/3 Insects per trap 1/3 1/3 Van Buren Ottawa Figure 11. Abundance of Oriental beetle adults in pheromone traps at a) Van Buren County and b) Ottawa County farms. 24

25 Plum Curculio (PC), Conotrachelus nenuphar (Herbst) Plum curculio is generally considered a pest of tree fruits in Michigan, but it also is known to be a pest of blueberry in other production areas such as New Jersey. PC was monitored to guard against outbreaks of this secondary pest in response to reduced broad spectrum insecticide use in fields. During sampling for fruitworm eggs and damage (see above), the number of berries with plum curculio egglaying scars was recorded weekly from 6 June to 18 July on 2 bushes in each field. PC damage was generally more abundant at Van Buren County comapared to Ottawa County farms (Figure 12). There was no difference in PC damage in compared to Grower Standard fields. Our data suggests PC remains a secondary pest for blueberry that should be monitored for the remainder of this project, and we will continue to record incidence of egglaying scars on fruit during weekly scouting in 26. Plum Curculio Damaged berries per 1 bushes a b. 5/3 1/3 5/3 1/3 Van Buren Ottawa Figure 12. Abundance of Plum Curculio damaged fruit at a) Van Buren County and b) Ottawa County farms. 25

26 MEASURE CHANGE IN CONVENTIONAL INSECTICIDE RESIDUES AND THE EFFECTS ON ABUNDANCE OF NATURAL ENEMIES, POLLINATORS, AND SECONDARY PESTS. RESIDUE ANALYSIS In each field just prior to the first harvest, leaf and fruit samples from 2 randomly selected bushes spread throughout the sample area were collected for pesticide residue analysis. One leaf shoot was collected from each of the top, middle, and lower region of the fruiting zone and one ripe berry cluster was collected from the outer top, outer middle, outer lower, inner top, and inner middle part of the bush. Bushes on the border of a field were not sampled to minimize the effects of spray drift from adjacent fields. Leaf and berry samples were placed in resealable plastic bags and immediately stored in a cooler until they could be returned to the laboratory and transferred to a freezer (-2 O C). Generally, samples were frozen within 6h of collection. Additional berry collections as described above were made just prior to the 2 nd and 3 rd harvest. Results from the 23 and 24 growing season are complete and 25 samples are being analyzed at the time of writing this report. We show the data from collected fruit available for the broadspectrum insecticides residues that have been tested to date. Throughout this project, average levels of insecticide residues in fruit samples were all lower than the tolerances established by the Environmental Protection Agency. Additionally, transition to the program has resulted in lower broad-spectrum insecticide residues in harvested fruit (Table 6). In general in both programs, the amount of broad-spectrum insecticide residue on fruit is lower in 24 and 25 compared to that in the first year of the study (23). We attribute this to generally lower pest pressure in 24 and 25 as well as a change in grower behavior in response to scouting and decision making information given to them for managing the fields. Residue analysis will continue in the final year to track the relationship between compounds applied to the fields in response to weekly scouting, and the residues detected in the fruit. We will also conduct more detailed analysis of the residues detected in each harvest. Table 6. Total residue (ppm) of broad spectrum insecticides recovered from fruit samples taken before harvest. EPA tolerances (ppm) are given in parentheses Compound GSTD GSTD GSTD Azinphos-Methyl (5.) Carbaryl (1.) Esfenvalerate (3.) Malathion (8.) Methomyl (6.).... na* na* Phosmet (1.) Total * Methomyl data is not yet available for 25 samples. 26

27 OVERVIEW OF THE ECONOMICS OF THE REDUCED-RISK AND GROWER STANDARD INSECT MANAGEMENT PROGRAMS Over the course of this project we have collected grower spray records for and Grower Standard fields and calculated total insecticide cost (excluding cost of application) for each farm and program. These data are given for all three years in Table 7. In 23 and 24 the average cost of the program was 2 to 3 times more expensive than the average cost of the Grower Standard program. Through discussions at the annual meeting for this project after the 24 growing season, the researchers and growers involved in this project agreed to focus on making the program cost effective. With this in mind, the midseason application of Admire 2F ($75/acre) was dropped from the program and the pyrethroid Asana ($9/acre) was substituted for an application of Confirm 2F ($28/acre) for fruitworm control. Additionally we chose not to use SpinTor for fruitworm control. With these changes in effect, the average cost of the program was greatly reduced and made comparable to the average cost of the GSTD program (Table 7). Table 7. Insecticide costs (per acre) for and GSTD Management programs Farm GSTD GSTD GSTD 1 $48.38 $ $48.38 $ $48.38 $ $37.18 $ $46.62 $2.2 $59.4 $ $85.14 $24.2 $89.34 $18.62 $8.19 $ $44.23 $ $34.44 $11.49 $43.79 $ $77.79 $168.4 $29.76 $13.56 $75.2 $ $118.4 $ $83.33 $149.7 $72.33 $69.2 Average $68.52 $18.3 $55.31 $ $63.16 $

28 NATURAL ENEMIES Pitfall Trapping In Michigan, pitfall traps were deployed in each field in April. Three traps were deployed along the border of the field adjacent to a wood lot and a trap was placed 5m from the border along each row that contained a border trap. Rows for trap placement were evenly spaced across a blueberry field, 6-18 rows apart. Traps were set up between bushes, filled with approx. 2 ml antifreeze, and covered with a plywood rain shield. Traps were opened for one week each month from April to October and the entire contents from each pitfall trap was recorded, carabids identified to species and all other insects identified to family. We compared monthly totals of the entire carabid community and the 1 most common species between and Grower Standard programs using repeatedmeasures ANOVA with farms as replicates (SAS Institute, Cary, N.C.). Overall, we collected 4594 carabids representing 39 species. The total number of carabids trapped in fields was not different from that in Grower Standard fields for any sampling date. The 1 most common carabids in each program are listed below (Table 8). The most abundant species was Harpalus pensylvanicus, which accounted for 36% of the total carabids collected. In 25, we detected some significant differences in the activity/density (as represented by mean numbers of adults/field/month) between the two programs for some of the most common carabid species collected. Overall there was not a consistent trend amongst the carabids; some species were more abundant in the reduced risk program and others in the Grower Standard fields. The most important difference between the two programs was in the activity/density of a common spring emerging species, Amara aenea. This species accounted for 2% of the carabids caught in field, 5% of the carabids in Grower Standard fields, and 12% of all captures. There were significantly more A. aenea caught in traps in fields in June [Figure 12 (p <.9)], and this is consistent with the data from 24. Also in June 25, more Anisodactylus santaecrusis were caught in fields than in Grower Standard fields (p <.1). In July more Chalaenius tricolor were collected in than in Grower Standard fields (p <.8). Three common species (Table 8) were found in greater numbers in Grower Standard compared to fields. More Poecelius chalcites were trapped in Grower Standard fields than in fields in May (p <.8), higher activity/abundance of Scarites subterraneus was observed in Grower Standard compared to fields in June (p <.8), and more Harpalus indianus were caught in Grower Standard fields than in fields in September (p <.8). However the overall low abundances of Anisodactylus santaecrusis, Chalaenius tricolor, Poecelius chalcites, Scarites subterraneus and Harpalus indianus (Table 8) suggests the observed statistical differences will have little biological significance and no measurable effect on the abundance of blueberry pest insects. In contrast to our results from 23 and 24, we did not find any statistically significant differences in the number of Harpalus erraticus in compared to Grower Standard fields, in fact fewer H. erraticus were caught in fields than in Grower Standard fields. 28

29 Table 8. Percent composition of beetles trapped in pitfall traps placed in blueberry fields managed under two different programs. Grower Standard (total = 26) % (total = 1994) % Harpalus pensylvanicus (DeG.) 42 Harpalus pensylvanicus (DeG.) 28 Harpalus erraticus Say 22 Amara aenea (DeG.) 2 Pterostichus mutus (Say) 12 Harpalus erraticus Say 18 Patrobus longicornis (Say). 6 Pterostichus mutus (Say) 14 Amara aenea (DeG.) 5 Patrobus longicornis (Say). 5 Poecilus chalcites Say 3 Chlaenius tricolor Dejean 2 Harpalus herbivicus Say 1 Harpalus herbivicus Say 2 Pterostichus melanarius (Illiger) 1 Pterostichus melanarius (Illiger) 2 Harpalus indianus (Csiki) 1 Anisodactylus santaecrusis (F.) 1 Scarites subterraneus (F.) 1 Poecilus chalcites Say 1 All carabid species Amara aenea # Beetles per field a. GSTD b. GSTD 1 April May June July August Sept Oct April May June July August Sept Oct Figure 12. Carabid abundance in and Grower Standard fields. a.) All species combined. b.) The spring emerger, A.aenea. denotes highly significant difference between programs for June, p <.8. In summary, during the third year of this study, we have seen several carabid species respond with greater activity/density in the program than the grower standard program, and we have demonstrated significant differences between the two programs in the abundance of some key species. It will be crucial to determine how these species can function as biocontrol agents. For example A. aenea is present in blueberry fields when cranberry fruitworm larvae drop from fruit clusters to overwinter in the soil, but we do not know if A. aenea can be an effective predator against this key pest. Similarly, a fall emerging species could help reduce the abundance of key late season pests (blueberry maggot and Japanese beetle) by feeding on different life stages of these pests that are present on or in the soil during the fall. We hope to demonstrate this potential through feeding assays and additional exclusion/prey removal experiments (see below). 29

30 Natural Enemy Survey Traps At each farm, two Pherocon AM yellow sticky traps were deployed on bushes in both fields on border facing a woodlot. These traps were replaced every other week from April to October. On 6 June an ammonium acetate charger was added to each trap so that it could also be used to monitor blueberry maggot fly. We report the number of adult coccinellids, chrysopids, syrphids, formicids (ants) social/solitary wasps (vespids/sphecids), spiders, and parasitoid wasps trapped (Figure 13). We observed significantly higher numbers of social/solitary wasps on sticky traps in fields (F 1,1 = 6.16, p =.3). There does not appear to be a consistent trend amongst the other natural enemy groups on sticky traps. Some groups were more abundant in the program while others were more numerous in the Grower Standard program (Figure 13). Nevertheless this data suggests that some natural enemies are continuing to respond to the insecticide programs. Mean per farm ( 2 traps) GSTD RR Coccinellid Chrysopid Syrphid Formicid Wasp Spider Para wasp * Figure 13. Natural enemies observed on yellow sticky traps in and Grower Standard fields in 25. * denotes means are significantly different at p =.5. 3

31 Scouting for natural enemies The number of natural enemies encountered during visual inspection of bushes (1 minute per bush) during routine scouting was recorded each week on 15 bushes per field (5 border bushes and 1 interior bushes). We report the total number of adult Coccinellids, Chrysopid eggs, adult Syrphids, Formicids (ants) social/solitary wasps (Vespids/Sphecids), spiders, and parasitoid wasps found on blueberry bushes during scouting (Figure 14). Significantly more spiders (F 1,1 = 4.86, p =.5) and chrysopid eggs (F 1,1 = 5.6, p =.5) were found in fields. In general this is similar to the results from 24 where significantly greater numbers of spiders, coccinellids and formicids were observed in fields. Additionally if the total number of natural enemies encountered at each farm in 25 is compared between programs, there are significantly more natural enemies in fields (F 1,1 = 5.13, p =.5). This shows a consistent trend that remains in the third year; several natural enemy groups are continuing to increase in abundance in response to the reduced use of broad-spectrum compounds in fields. This remains true even though the reduced risk insecticide program at three farms included Asana, a pyrethroid. We will continue to monitor the response of these predators to the insecticide programs, and as with the carabids, we hope to further demonstrate how differences in natural enemy abundances can increase control of key pests in Michigan blueberries. 4 * 35 Mean per farm * GSTD RR * 1 5 Coccinellid Chrysopid egg Syrphid Formicid Wasp Spider Parasitic wasp All Groups Figure 14. Natural enemies observed during scouting in and Grower Standard fields in 25. * denotes means are significantly different at p =.5. 31

32 Abundance of Ground Foraging Ants The abundance of ants in and Grower Standard fields was measured by using tuna baits. In each field 15 test tubes containing ~5g of tuna fish were deployed every two weeks. Five baited tubes were placed in three separate rows, evenly spaced, within each field. In each row one tube was placed at the edge of an adjacent woodlot, a second tube was placed between bushes near (within 5m) the field border, the third tube was placed in the row middle adjacent to the second tube, the fourth tube was placed in the same row between bushes 3 to 5m inside the field and the last tube was placed in the row middle adjacent to the fourth tube. Tubes were left in a field for 1 hour and then collected. Tubes with ants present were corked, labeled and frozen. Tubes were emptied at a later date, ants were counted, identified to morphospecies and representative specimens were pinned for identification. The number of bait tubes with ants present, the number of ants per bait, and the species identification were recorded. Significantly more baits with ants were found in compared to Grower standard fields (F 1,1 = 5.8, p =.5; Figure 15). These data along with the results of the Natural Enemy Survey above indicates ants may be important natural enemies in the blueberry system. Samples are still being identified, but we have trapped 12 morphospecies. In accordance with data from 24, Lasius neoniger, Prenolepis imparis and Tapinoma sessile are the most common species (Table 9). Baits with ants per farm GSTD (N=115) (N=235) Average FARM * Figure 15. Ant abundance in and Grower Standard fields at the six farms. Total number of baits in each program is given in legend. * denotes mean is significantly different, p <.5. 32

33 Table 9. Percent composition of ants captured in tuna baited test tubes placed in blueberry fields managed under two different programs. Grower Standard (total = 1726) % (total = 3451) % Lasius neoniger 58.3 Lasius neoniger 47.9 Solenopsis sp 16.5 Prenolepis imparis 17.3 Tapinoma sessile 11. Tapinoma sessile 13.8 Prenolepis imparis 7.8 Leptothorax sp 9.5 Myrmica sp 4.6 Solenopsis sp 7.4 Aphaenogaster tennesseensis 1.6 Aphaenogaster tennesseensis 2.1 Formica argentea.2 Formica argentea 1.1 Leptothorax sp.1 Myrmica sp.8 Unidentified.1 We compared monthly totals of all ant species captured and monthly totals of all species between and Grower Standard programs using repeated-measures ANOVA with farms as replicates (SAS Institute, Cary, N.C.). Approximately twice as many ants were collected in compared to Grower Standard fields. During July the total number of ants that were captured in fields was significantly greater than that in Grower Standard fields [Figure 16a, (p =.5)], and it appears this result is influenced by the abundance of Lasius neoniger (Figure 16b). Prenolepis imparis was more abundant in fields in September (Figure 16c), and more Leptothorax sp. was recorded in fields in August and September (Figure 16d). Considerably more Leptothorax were recorded in 25 than in 24 and this species occurred almost exclusively in fields in both years. Total Number of Ants a. All Species * GSTD Total Number of Ants b. Lasius neoniger * GSTD 1 1 May June July August September May June July August September Total Number of Ants GSTD Prenolepis imparis c. 3 GSTD 25 * Leptothorax sp * d. * May June July August September May June July August September Figure 16. Abundance of some of the most common ant species in and Grower Standard fields. a) All species combined. b) Lasius neoniger. c) Prenolepis imparis. d) Leptothorax sp. * denotes monthly totals are significantly different, p <.5. 33

34 The greater abundance of several ant species in fields suggests these generalist predators are responding to the different management programs. It will be important to identify whether these species of ants have the potential to aid the control of blueberry pests. It is also possible that these ants function as indirect pests by tending aphids or interfering with other biocontrol agents. In the remaining year if this project, we intend to continue to use feeding and exclusion cage experiments to investigate how these species fit into the blueberry agroecosystem. 34

35 Predator Exclusion Experiments To determine relative rates of predation on blueberry pest insects by different groups of natural enemies between and Grower Standard fields, a series of exclusion cage experiments were employed. Exclusion cages were constructed out of ½ inch mesh hardware cloth, window screen and 17 mil clear plastic sheeting. Traps were made in the form of an 8 x 8 x 1.5in. box of mesh hardware cloth with a removable lid and hinges and clasps made from twist ties. Lids for all cages were covered with a layer of plastic sheeting to act as a rain shield. Cages were further modified to fit into one of three groups: 1) Open - cages not lined to allow passage of all natural enemies, 2) Mesh - cages lined with 16 mesh aluminum window screen cages to exclude large (> 1.5 mm) natural enemies, and 3) Closed - cages completely lined with 17 mil clear plastic sheeting and prey to exclude all natural enemies. Cages were placed in and Grower Standard fields at four farms. Three sets of the three cage types were placed in and Grower Standard fields. Cages were placed between bushes in three rows that were evenly spaced throughout the field. Cages were positioned at 1m intervals along the length of the row, and cage order was arranged in a Latin square so that one of each cage type was as in each position (1m, 2m, 3m) along the row. The same cages were used in three series of prey removal experiments described below. Removal of CBFW Hibernaculae - In the first series of experiments, five freeze-killed cranberry fruitworm hibernaculae were placed 1-2 mm below the surface of a moist sand filled 6cm Petri dish in each cage. A small mound of moist sand was placed around each dish to allow crawling insects access to the top of the dish. Cages were then sealed and left in the field for 7 days. At that time cages were opened and remaining hibernaculae were counted and collected in glassine envelopes. Any predatory insects in or near the cages were noted. Collected hibernaculae were returned to the laboratory and frozen until microscopic examination could be performed to determine if hibernaculae received injuries indicative of feeding damage. This experiment was repeated three times on consecutive weeks from 11 July to 2 August. We compared the effect of management program and predator exclusion (cage type) on the proportion of hibernaculae removed and the proportion of hibernaculae with signs of feeding damage with Factorial ANOVA using farms as replicates. Removal of JB eggs 1 - The second series of experiments used the same general protocol to place 1 Japanese beetle eggs in the cages as prey items. Eggs were collected as described above, but the number of prey items removed was determined by sifting the sand from the Petri dish with a # 18 mesh sieve. This was also repeated three times on consecutive weeks from 2 August to 23 August. We compared the effect of management program and predator exclusion (cage type) on the proportion of Japanese beetle eggs removed with Factorial ANOVA using farms as replicates. 35

36 Removal of JB eggs 2 - Upon examination of the eggs in the laboratory it became evident that many of these prey items were desiccated and difficult to find in the sifted sand, so we repeated this experiment using the same cages and five Japanese beetle eggs in plastic 4-dram vial lids that were filled with moist potting soil. Prey items were left in the fields overnight 8 to 9 September for 12 hours. Vial lids with eggs were collected and covered with parafilm and returned to lab and kept in a 4 o F fridge for 3 days until they could be examined under a microscope to determine the number of eggs removed. We compared the effect of management program and predator exclusion (cage type) on the proportion of Japanese beetle eggs removed with Factorial ANOVA using farms as replicates. Results Removal of CBFW Hibernaculae - Predation in fields was not significantly different than that in Grower Standard fields, however there was a significant effect of cage type on the rate of predation of CBFW hibernaculae (Figure 17a), and more hibernaculae were fed upon in open cages compared to the closed or screen cages (Figure 17b). This suggests that, in this experiment, larger predators that could pass through the mesh of the open cages were responsible for more predation on hibernaculae than smaller predators. That is carabids would seem to be more important than ants as predator of cranberry fruitworm hibernaculae. Proportion removed a. b a b b a b Closed Open Mesh a ns b a Proportion with feeding b. c a b b a ab Closed Open Mesh ns b a b Overall 18 Jul 25 Jul 2 Aug. Overall 18 Jul 25 Jul 2 Aug Figure 17. Effect of exclusion of natural enemy groups on predation of CBFW hibernaculae. a.) Proportion of hibernaculae removed b.) Proportion of hibernaculae with evidence of feeding. For a given date, columns with the same letter are not statistically different (p<.5). Removal of JB eggs 1 - Predation on Japanese beetle eggs was difficult to assess due to desiccation of eggs. When eggs were left in cages for a week, we did not find any significant differences between management programs in the proportion of Japanese beetle eggs removed, and we did not find any differences in the proportion of Japanese beetle eggs removed between different exclusion cage types. 36

37 Removal of JB eggs 2 When Japanese beetle eggs were only left in cages overnight, predation in fields was not significantly different than that in Grower Standard fields, however there was a significant effect of cage type on the rate of predation of CBFW hibernaculae (Figure 18). The proportion of prey removed was significantly greater in open cages compared to that in mesh or closed cages. Like the results from the experiments using hibernaculae, the data for Japanese beetle egg removal suggests carabids were more important predators than ants in this study..25 Closed a Proportion prey removed Open Mesh b b Figure 18. Effect of exclusion of natural enemy groups on predation of Japanese beetle eggs. Columns with the same letter are not statistically different (p<.5). 37

38 Aphid Natural Enemies Aphid densities were intensively sampled approximately every two weeks from 14 July to 8 September. Blueberry fields were divided into three blocks; the first block was the area 3m into the field along the entire edge adjacent to woods, and the remaining portion of the field was evenly divided into two blocks. On each of five sampling dates we located 5 bushes infested with aphids within each block, and flagged the base of infested branches of first-year growth. We returned to these branches and counted the total number of aphids (mature and immature), and the number of parasitized aphids (mummies). Although there were fewer aphids per branch in fields, there were no significant differences in the mean number of aphids per branch between programs for any date (Figure 19a). Aphid mummies were found throughout our sampling period (Figure 19b), and the mean number of mummies per branch was significantly higher in Grower Standard fields for the first, second and final collection dates and the overall average (Figure 19 b). This is likely due to the greater numbers of aphid hosts available in Grower Standard fields. Unlike 24, percent parasitism was similar in and Grower Standard fields at all sampling periods (Figure 19c). The high levels of parasitism in both programs suggests that as the aphid densities dropped in mid-late July, relatively high populations of parasitoids were able to parasitize all remaining aphids in both programs. Aphids per branch GSTD a. Mummies per branch * GSTD * * b. * Jul 27 Jul - 2 Aug 8-11 Aug Aug 7-8 Sep Overall average Collection date Jul 27 Jul - 2 Aug 8-11 Aug Aug 7-8 Sep Overall average Collection date Percent parasitism GSTD c Jul 27 Jul - 2 Aug 8-11 Aug Aug 7-8 Sep Overall average Collection date Figure 19. Abundance of aphids and mummies in and Grower Standard fields during 25. a.) Aphid density, b.) mummy density, and c.) percent parasitism. * denotes means are significantly different, p <.5. 38

39 Aphid Parasitoid Identification During the first two sampling periods (14-18 Jul and 22 Jul-2 Aug) we collected all the aphid mummies on five randomly selected bushes in each block in both fields at each farm until a total of 5 mummies per field was reached. All mummies were kept in individual plastic cups until the parasitoid adult emerged. Wasps were identified and grouped according to life history (primary parasitoid or hyperparasitoid). We did not observe a significant difference between the programs in the number of aphids parasitized by a primary parasitoid (Braconidae: Praon sp or Aphidius sp.) or hyperparasitoid species [Alloxysta or Pteromalidae (Figure 2)]. We will continue to monitor aphid parasitism throughout this project. Percent of Total Parasitoids Emerged GSTD (N=93) (N=73) a b. GSTD (N=93) (N=73) Praon Aphidius Alloxysta Pteromalidae Primary Parasitoid Hyperparasitoid Figure 2. Percentage of parasitoids reared from aphid mummies collected in 25 within each treatment. Total number of wasps recovered given in legend. a.) Parasitoids grouped by taxon. b.) Parasitoids grouped into primary parasitoids and hyperparasitoids. 1 and 2 are primary parasitoid species and 3 and 4 are hyperparasitoids (1. Praon sp., 2. Aphidius sp., 3. Alloxysta sp., 4 Pteromalidae corresponding with the wasps in the pictures above). 39