ARTHROPOD BIOLOGY Biology of Parasitoids (Hymenoptera) Attacking Dasineura oxycoccana and Prodiplosis vaccinii (Diptera: Cecidomyiidae) in Cultivated Blueberries BLAIR J. SAMPSON,1 TIMOTHY A. RINEHART,1 OSCAR E. LIBURD,2 STEPHEN J. STRINGER,1 1 AND JAMES M. SPIERS Ann. Entomol. Soc. Am. 99(1): 113Ð120 (2006) ABSTRACT The blueberry gall midge, Dasineura oxycoccana (Johnson), and blueberry tip midge, Prodiplosis vaccinii (Felt) (Diptera: Cecidomyiidae), are recurring cecidomyiid pests of cultivated blueberries in the southern United States and Mediterranean Europe. Insecticides can give short-term control, but overlap in parasitoid phenologies indicates the potential for natural control of midge populations. Using a combination of laboratory rearing and mitochondrial DNA analysis of Þeld samples, we identiÞed Þve species of solitary endoparasitoids that killed 30Ð40% of midges. These species include at least three undescribed platygastrids in the genera Synopeas, Platygaster, and Inostemma. An undescribed prepupal idiobiont, Aprostocetus sp. (Eulophidae: Tetrastichinae) was the only midge parasitoid that was consistently active when rabbiteye blueberries, Vaccinium ashei Reade, were in ßower. Six percent of midge prepupae, half of which already contained platygastrid larvae, were parasitized by Aprostocetus. KEY WORDS biological control, Platygastridae, Eulophidae, high-Þdelity polymerase chain reac- tion, parasitism BLUEBERRY GALL MIDGE, Dasineura oxycoccana (John- However, injury to blueberry leaf buds and meristems son), and blueberry tip midge, Prodiplosis vaccinii by P. vaccinii often induces excessive suckering, leaf (Felt), are two species of univoltine gall midges distortions, and leaf drop, which could result in lighter (Diptera: Cecidomyiidae) that feed exclusively on bud sets (Gagne´ 1986, Williamson and Miller 2000). Vaccinium buds. Midges are not easily detected be- Little is known about the interactions between cause their feeding damage only becomes visible 7Ð midges and their parasitoids on fruit crops. Proc- 14 d after larvae have left the blueberry bush. Fur- totrupoid wasps such as those belonging to the family thermore, bud injury is often mistaken for freeze dam- Ceraphronidae parasitize D. oxycoccana on Wisconsin age, plant disease, or nutrient deÞciency (Driggers cranberries, Vaccinium macrocarpon L. (Barnes 1948). 1926, Lyrene and Payne 1992, Bosio et al. 1998, Samp- In the southern United States (Florida, Georgia, Ala- son et al. 2002, Wei 2004). D. oxycoccana is the Þrst of bama, Mississippi, and Louisiana), endoparasitic these species to break winter dormancy in January or Platygastridae seem to replace ceraphronids as the key February in the southeastern United States. Females natural enemies of midges (Vlug 1976, Yoshida and principally oviposit between the developing scales of Hirashima 1979, Jeon et al. 1985, Sone´ 1986, Sampson ßower buds. Newly hatched larvae crawl deeper into et al. 2002, Sarzynski and Liburd 2003, Legner 2004). blueberry buds where they feed on the innermost Chalcids, like tetrastichine eulophids, are also midge meristematic tissue and can reduce up to 80% of ßower production and potential fruit yield (Lyrene and parasitoids found in both midwestern and southern Payne 1992, Sampson et al. 2002). As damaged ßoral habitats, where they attack midges on cranberries and buds disintegrate, D. oxycoccana move to leaf buds, blueberries, respectively (Sampson et al. 2002). but they are soon superseded by P. vaccinii (Driggers Here, we studied the ecology of parasitoids found in 1926; Gagne´ 1986, 1989; B.J.S., unpublished data). rabbiteye, Vaccinium ashei Reade, and southern high- Crop losses to P. vaccinii have yet to be assessed. bush blueberry (Vaccinium corymbosum ϫ Vaccinium darrowii Camp) ecosystems. Larval development, Mention of trade names or commercial products in this article is adult reproductive behavior and parasitoidÐhost phe- solely for the purpose of providing speciÞc information and does not nology was investigated during 2003Ð2004. Mitochon- imply recommendation or endorsement by the U.S. Department of drial DNA analysis was used to conÞrm the identity of Agriculture. adult parasitoids as well as to assist in interpreting 1 USDAÐARS Small Fruit Research Station, Poplarville, MS 39470. 2 Department of Entomology and Nematology, University of Flor- interactions between parasitoids and their midge ida, Gainesville, FL 32611Ð0620. hosts. 114 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 99, no. 1 Materials and Methods imens for 24Ð48 h in Histoclear (National Diagnostics, Atlanta, GA). Wings were removed from adults, la- Host and Parasitoid Sampling. Host Eggs and Larvae. beled, and separately slide mounted in water to pro- Juvenile midges were sampled from two blueberry tect Þne hairs during the clearing process. Adults were nurseries at the USDAÐARS Small Fruit Research Sta- cleared using 10% KOH, neutralized, and stained with tion, Poplarville MS (N 30Њ 50.21Ј,W89Њ 32.65Ј). Nurs- acid fushin in lactophenol, and then slide mounted in ery 1 contained 950 2- to 6-yr-old potted bushes rep- Ϸ euparal or silicon oil. The identity of parasitoids on resenting 50 clones of rabbiteye blueberry, southern sticky traps that belonged to species capable of par- highbush blueberry, and wild blueberries, Vaccinium Ϸ asitizing immature midges was conÞrmed by compar- elliottii Chapman. Nursery 2 housed 11,000 second- ing them with specimens captured among host eggs and third-year seedlings representing 300 clones of and neonates inside blueberry buds or reared from V. ashei and southern highbush blueberry. Additional larval or prepupal hosts inside the insect growth cham- midge hosts were collected from 10 commercial rab- bers. biteye blueberry farms in southern Louisiana, Missis- Host and Parasitoid Identification. The following sippi, and Alabama. Ϸ taxonomic keys and records were used for midge and A single sample consisted of 10 terminals bearing parasitoid identiÞcation: Cecidomyiidae (Diptera): 30 ßower or leaf buds enclosed in a resealable plastic Barnes (1948) and Gagne´ (1986, 1989). Platygastridae bag. Five or 10 samples were taken from random (Hymenoptera): Ashmead (1887, 1893), Marchal bushes in each nursery every 3 to 4 d over a 2-yr (1906), Fouts (1924), Kieffer (1926), Masner and period. Smaller samples were collected when bud Muesbeck (1968), Krombien et al. (1979), Velikan et availability waned in the winter. Bags were held for 2 al. (1984), Vlug (1984, 1995), Masner and Huggert to3dtoallow larvae to abandon buds. The bags were (1989), and Buhl (1998, 2001). Eulophidae (Hyme- then partly Þlled with water and shaken to ßush out noptera): Crawford (1907), Girault (1917), Myers remaining host eggs and larvae as well as larvae that (1930), Walter (1941), Burks (1943), Piore and Vig- were diseased, heavily parasitized, or dead. Eggs and giani (1965), Boucˇek (1986), Krombien et al. (1979), larvae were pipetted into clean petri dishes, counted, Velikan et al. (1984), LaSalle (1994), and Shauff et al. and preserved in 100% ethanol. The number of host (1997). Taxonomic experts afÞliated with the USDAÐ larval stages (instars) was established from a fre- ARS Systematic Entomology Laboratory Communi- quency histogram generated by PROC UNIVARIATE cations and Taxonomic Services Unit (Beltsville, MD) (SAS Institute 1990) by using the body width mea- conÞrmed our identiÞcations. Vouchers were depos- surements of 474 midge larvae. Mature midge larvae ited in the U.S. National Collection by Raymond J. (third instars or prepupae) were identiÞable by a dark Gagne´ (Cecidomyiidae), Michael E. Shauff (Eulophi- Y-shaped ventral sclerite called a spatula (Gagne´ dae), Michael W. Gates (Eulophidae and Platygastri- 1989). dae), and Terry Nuhn (Platygastridae). Juvenile Parasitoids. Preparing a parasitoid brood Ecological Data: Analysis of Host Parasitism and for slide mounting sometimes required varying a host Parasitoid Reproductive Behavior. We calculated the larvaÕs exposure to the clearing, staining, and Þxing average parasitism rate expected for each month of a agents. Normally, host fat tissues were adequately sol- single year (Fig. 1). Parasitism rates were expressed as ubilized with exposure to solutions of 5 and 10% KOH the percentage of host eggs and larvae with parasitoid for 24 h and 30 min, respectively. Two or three drops eggs, larvae, or pupae (Figs. 2 and 3). The chi-square of diluted double stain (BioQuip Products, Inc., Ran- goodness-of-Þt analysis (PROC FREQ) was used to cho Dominguez, CA) or acid fuschin in lactophenol test whether parasitoid eggs, neonates, and older Þrst was sufÞcient to clear, stain, and temporarily Þx spec- instars were randomly or uniformly distributed among imens. Smearing host larvae by applying mild pressure host larvae (SAS Institute 1990). to the coverslip entirely expulsed many parasitoids Foraging activity of adult parasitoids throughout the from the host or more clearly exposed them. Live day was evaluated using hourly sticky trap (16 by parasitized hosts, compressed by a coverslip, made it 23-cm) catches. The resulting data were Þtted to a easier to identify which host organs nourished the Gompertz function [y ϭ 163 eϪ49eˆ(Ϫ0.26x)], where x is parasitoid brood (e.g., midgut, brain, or ganglia). Us- the cumulative hours that elapsed starting at 0800 ing an ocular micrometer, we measured the cephalo- hours, and y is hourly cumulative catch expressed as thorax and caudal segments of juvenile parasitoids that a percentage of the total