Response of Direct-seeded Broccoli to Preemergence Oxyfluorfen and BAS 514 Preemergence treatments of oxyfluorfen at 0.14, 0.28, and 0.56 kg ai/ha caused a 40 to 99% reduction in stand of direct-seeded broccoli, while BAS 514 applied at the same rates caused no reduction in stand. All rates of oxyfluorfen completely controlled Venice mallow while only the 0.56 kg/ha rate of BAS 514 provided acceptable (>70%) control. Broccoli plants that survived preemergence applications of oxyfluorfen exhibited injury ranging from 32 to 97% 6 to 9 weeks after treatment. BAS 514 applied at 0.14, 0.28, 0.56 kg ai/ha provided excellent control of common ragweed with little or no visible crop injury. Common ragweed control with oxyfluorfen varied with soil moisture and organic matter. Applications of 0.28 and 0.56 kg/ha oxyfluorfen significantly reduced yield when compared to the cultivated check. Yields from plots treated with 0.14 kg/ha BAS 514 equalled or exceeded the cultivated check. Treatments of oxyfluorfen and BAS 514 to seeded broccoli caused approximately a one week delay in harvest. Nomenclature: oxyfluorfen, [2-chloro-1-(3-ethoxy-4- nitrophenoxy)-4-(trifluoromethyl)benzene]; BAS 514, quinclorac, (3,7-dicloro-8-quinoline-carboxylic acid); broccoli, [Brassica oleracea(L.) var botrytis]; common ragweed, (Ambrosia artemisifoliaL. #³ AMBEL); Venice mallow (Hibiscus trionum L. #3 HIBTR); Additional index words: Brassica oleracea (L.) botrytis, quinclorac. Weed Control and Crop Yield with Postemergence Applications of Oxyfluorfen, Pyridate, and BAS 514 to Direct-seeded Broccoli Postemergence applications of oxyfluorfen in field studies caused white chlorotic speckling on broccoli foliage. Plants treated with 0.07 kg/ha oxyfluorfen produced yields similar to the cultivated check. High rates of pyridate (4.03 kg/ha) caused yellow lesions on broccoli foliage. Total broccoli yields from plots treated with pyridate at 0.28 and 0.50 kg/ha were not significantly different from the cultivated check. BAS 514 caused moderate growth stunting to broccoli in field studies. At two out of four locations, injury increased with time. At these sites, total broccoli yield from plots treated with 0.28 and 0.56 kg/ ha BAS 514 was lower than the cultivated check. Total number of broccoli heads per hectare for plots treated with BAS 514 equaled or exceeded that of the cultivated check. Broccoli treated with BAS 514 plus the experimental adjuvant BAS 090 showed increased injury and decreased yield with increasing rates when compared to applications of BAS 514 alone. Broccoli developed foliar veinal chlorosis and whitish elongated heads following applications of BAS 514 plus BAS 090. High rates of BAS 514 (0.56 kg/ha) with or without BAS 090 reduced average head weight. Little to no stand reduction or delay in harvest was observed with any herbicide treatment. A two fold safety margin for herbicide rate was observed for oxyfluorfen at 0.07 kg/ha, pyridate at 0.50 kg/ha and BAS 514 at 0.14 kg/ha. In greenhouse studies, oxyfluorfen at 0.28 kg/ha provided >90% control of common ragweed and common lambsquarters, and 77% control of common pigweed. Pyridate at 2.02 kg/ha provided excellent control of horseweed, and greater than 70% control of prickly sida, common ragweed and common lambsquarters. All rates of pyridate controlled common pigweed. Applications of BAS 514 at 0.14 to 0.56 kg/ha provided 88 to 100% control of horseweed and 76 to 82% control of common ragweed. Greater than 90% control of large crabgrass was observed with BAS 514 plus BAS 090 at 0.28 kg/ha. Nomenclature: oxyfluorfen, 2-chloro-1-(3-ethoxy-4-nitrophenoxy)-4-(trifluoromethyl) benzene; pyridate 6-chloro-3-phenylpyridazine-4-yl-S-octyl carbonothionate; BAS 514, quinclorac, 3,7-dichloro-8-quinoline-carboxylic acid; broccoli Brassica oleracea L. var. botrytis; common ragweed Ambrosia artemisiifolia L. #3 AMBEL; redroot pigweed Amaranthus retroflesus L. #AMARE; common lambsquarters Chenopodium album L. #CHEAL; Prickly sida Sida spinosa L. #SIDSP; Horseweed Conyza canadensis L. #ERICA; Large crabgrass Digitaria sanguinalis L. #DIGSA; Additional index words. quinclorac, BAS 090, Brassica oleracea L. botrytis. / Master of Science

Three formulations of Bacillus thuringiensis Berliner subsp. kurstaki (Dipel 4L, Dipel 2X, and ABG-6167) were compared with the synthetic pyrethroid perrnethrin ( common name Pounce 3. 2 EC) for insecticidal activity and impact upon parasitism of the imported cabbageworrn, Pieris rapae (L.) (Lepidoptera: Pieridae) in field broccoli (CV Packrnan) from fall 1986 to spring 1988. Perrnethrin, Dipel 4L, and ABG-6167 were not significantly different in their efficacy towards imported cabbageworm larvae. Parasitization of P. rapae by the larval parasite Cotesia glomerata (L.) (Hymenoptera: Braconidae) and the pupal parasite Pteromalus puparum (L.) (Hyrnenoptera: Pteromalidae) continued after spraying in the B. thuringiensis plots, but was not significantly different from perrnethrin. However, 100% of the P. rapae chrysalids recovered were parasitized by P. puparum. The impact of seasonal hyperparasitism was compared between .Q. glomerata, and a Yugoslavian strain of Cotesia rubecula (Marshall), an exotic larval parasite of the imported cabbageworm. Four hyperparasites, two of them attacking both~. glomerata and~. rubecula, were found in field plots from 1986-1988. The level of hyperparasitization for the two primary parasites was significantly different. Hyperparasitization averaged only 8.1% for C. glomerata from 1986-1988, but was 37. 9% for C. rubecula from 1987-1988. During the early- to mid-season of 1988, hyperparasite activity was not detectable and C. rubecula outcompeted C. glomerata for hosts; but by mid-season, hyperparasite activity against C. rubecula increased to 100%, causing its populations to crash. C. glomerata then became the dominant parasite of P. rapae. C. rubecula was not recovered in 1989. Hyperparasites may be a limiting factor in establishing C. rubecula in southwestern Virginia. Mortality and successful pupation of P rapae fourth instars parasitized by C. rubecula to B. thuringiensis endotoxin at dosages of 850, 85, and 8.5 I.U./ml was examined. After day two, the LC₅₀'s of parasitized fourth instars were approximately thirty times higher than that of unparasitized larvae and by day four, the LC₅₀ response of parasitized fourth instars was 180 times higher than unparasitized larvae. Twenty-five percent of parasitized fourth instars exposed to a concentration of 850 I.U./ml successfully pupated, compared to 76% at 85 I.U./ml and 69% at 8.5 I.U./ml. Parasitized fourth instar P. rapae consume less food and are therefore less susceptible to B. thuringiensis than unparasitized larvae at the same dosages. / Ph. D.

Respondents of the 1986 Broccoli Growers Survey grew an average of 6.2 acres of broccoli in Fall 1986; popular varieties were Green Valiant (87.5% of respondents), Packman and Emperor (62.5%). Less than 5% of tobacco land was converted to broccoli production by 62.5% of the growers. Before heading, 2.4 insecticidal sprays were applied for worms at $20.51/acre. During heading, 2.4 insecticidal sprays were applied for worms at $19.40/acre. Under normal growing conditions, a nine-day mean of 8.4 and a 15-day mean of 12.1 cabbage looper larvae/plant did not significantly reduce yield in the Spring and Fall respectively. However, means of 7.4 and 11.1 cabbage looper larvae/plant reduced yield (head weight) under drought conditions. Harvest date was not affected by the above cabbage looper levels. In Fall 1986, a reduction in yield occurred when plants had 50% or more defoliation and harvest date was significantly delayed at 100% defoliation. In the drought stressed Spring 1987 planting, a stimulation in yield occurred at 25% defoliation and 75% defoliation was needed to significantly reduce yield, but there was no difference in harvest date. The imported cabbageworm was the most abundant pest of broccoli in Montgomery Co., VA. Based on counts of all immature stages of the insects for three seasons, Packman and Southern Comet varieties were less preferred compared to Green Defender. Laboratory studies indicated that imported cabbageworm larvae developed faster to the fifth instar on Packman broccoli than those reared on Green Defender, Southern Comet or Emperor. Difference in amount consumed between varieties was not detected. Pupae reared on Southern Comet weighed significantly more than those reared on Packman. / M.S.

Four key aspects of the relationship between predatory hoverflies and the aphid pest Brevicoryne brassicae L. on broccoli were investigated in the Willamette Valley, Oregon USA: 1) the relationship between aphid density and hoverfly oviposition, 2) the larval voracity of key hoverfly species, 3) the preferences of hoverflies and broccoli pests for candidate insectary plant species, and 4) the role of insectary plantings in enhancing hoverfly oviposition in aphid colonies in broccoli fields. Hoverfly oviposition at two commercial broccoli field sites increased at aphid densities greater than 50 aphids per plant, but did not peak at the highest aphid densities. A logistic regression model, selected with Akaike's Information Criterion, of the odds of oviposition in relation to aphid density included effects for sampling date and aphid species (B. brassicae and Myzus persicae (Suizer)). The magnitude of the response was similar for B. brassicae and M persicae. The oviposition response was greater in the upper part of the plant canopy, and different on individual leaves from that seen on whole plants. The voracity and development rate of Scaeva pyrastri (L.) were greater than those of Eupeodes fumipennis (Thomson), and both exceeded Syrphus opinator Osten Saken and Sphaerophoria suiphuripes (Thomson). Both E. fumipennis and S. opinator reduced aphid populations to low levels on plants over two weeks, but E. fumipennis spent much of the time foraging away form the caged plants. Alyssum (Lobularia maritima (L.) Desv.), coriander (Coriandrum sativa L.), buckwheat (Fagropyrum esculentum Moench), and phacelia (Phacelia tanacetifolia Benth.) were screened for arthropod visitation frequency. The three pest species Pieris rapae (L.), Diabrotica undecimpunctata undecimpunctata Mannerheim, and Lygus hesperus Knight visited phacelia the most frequently, while the hoverfly species visited mostly coriander, which was possibly influenced by competition from other foragers. Hoverfly oviposition was increased near within-field blocks of alyssum in broccoli fields at the end of the season however, hoverflies arriving at the field had pollen in their guts and were oviparous four to five weeks before their eggs were detected in aphid colonies. This late oviposition may have occurred because threshold densities for oviposition were not reached until this stage. / Graduation date: 2006