Evaluations of individually collected wheat heads and whole plants indicated that several factors may influence the efficacy of aerial sprays against the wheat midge, Sitodiplosis mosellana (Géhin). These factors related to methods of assessing midge damage, condition of wheat heads during spraying, and methods of spray application. Location- and distribution-related influences also were important.

Efficacy assessments were based on reductions in midge larvae, midge-damaged kernels, and percentage kernel damage. The latter data were confounded by differences in kernel numbers within wheat heads. In head and plant samples, sprays were more effective at reducing numbers of midge larvae than reducing the incidence of midge-damaged kernels. Evaluations of individual heads overestimated the efficacy of sprays in whole plants which contained one primary and two tiller heads. Sprays provided more effective midge control and kernel protection in the primary heads, which emerged before spraying, than tiller heads, which emerged mainly after spraying. Protection was usually better in apical than basal regions of each head type. In whole plants, efficacy declined as tillering increased.

Plant evaluations indicated that the high-volume (37.4 L water per hectare) chlorpyrifos spray provided the best midge control and kernel protection (95% and 87%, respectively), followed by the low-volume (18.7 L water per hectare) chlorpyrifos spray (87% and 76%, respectively), and low-volume dimethoate spray (66% and 53%, respectively). When improvements in both yield and grade were considered and calculations based on 1987/1988 wheat prices, net returns from the three aerial sprays ranged from $62 to $113 Canadian per hectare. Long-term benefits of the sprays probably were less favourable. Reduction in wheat midge after spraying would negate the benefits of an egg–larval parasite, Pirene penetrans (Kirby), which was present in low numbers.

Mandibles of four species and three subspecies of penicillate diplopods inhabiting Japan, and two species from outside Japan, were studied to clarify their taxonomic significance. The mandibular structure showed distinct differences both between genera and between species, indicating its usefulness in the taxonomy of penicillate diplopods.

We provide keys for identifying the 26 bumble bee species (Bombus and Psithyrus) found in Canada east of Manitoba, and information on their ecology and distribution. The keys are designed for field use and rely primarily on colour patterns rather than on microscopic features.

Aspergillus parasiticus Speare and Beauveria bassiana (Bals.) Vuillemin were isolated from melanopline grasshoppers in Saskatchewan in 1986 and 1987. The pathogenicity of these isolates for Melanoplus sanguinipes (Fabricius) nymphs was assessed by topical, oral, and injected applications of 1.0 × 103, 104, or 105 conidia per individual. The A. parasiticus isolate was pathogenic to M. sanguinipes by injection and topical application but not orally. The B. bassiana isolate was pathogenic when injected or topically applied resulting in high mortality (82–100%), at all doses tested, between days 3 and 5 post-inoculation. Beauveria bassiana was infectious, at the highest dose tested, when orally applied with the majority of mortality occurring between 7 and 14 days post-inoculation.

A no-choice feeding study was used to evaluate larch casebearer development and survival on eight species of conifer seedlings and two foliar stages of larch from molting to the fourth-instar larvae through adult eclosion. Percentage survival through adult eclosion varied on foliage of western larch (spring—flush stage, 44%; fall—bud set stage, 23%), Douglas-fir (40%), grand fir (26%), Englemann spruce (26%), western white pine (10%), and lodgepole pine (6%). No survival occurred on ponderosa pine or western hemlock. The rate of development differed among conifer seedling species, and between foliar stages of larch.

Glossina palpalis palpalis (Robineau-Desvoidy) and Glossina palpalis gambiensis Vanderplank hybridized readily in the laboratory but hybridized females produced fewer offspring than did females that mated with their own kind. Most hybrid females were fertile when backcrossed to either G. p. palpalis or G. p. gambiensis but almost all hybrid males were sterile. About half of the backcross males were able to fertilize G. p. palpalis and G. p. gambiensis. By using an X chromosome marker gene, tan, evidence was obtained that the X chromosome is involved in hybrid male sterility, either through interaction with the Y chromosome or the autosomes of the other subspecies. There was no evidence for maternally inherited sterility factors of a type that confer unidirectional sterility on hybrid or backcross males.

The efficiency and accuracy of sampling western flower thrips (Frankliniella occidentalis [Pergande]) from “Granny Smith” apple blossom clusters were analyzed during 1986–1987 to develop a sampling plan for research purposes. The accuracy of the “shake” method was compared with an “extraction” process of each of three blossom cluster types: pink, open, and petalless (petal fall). Thrip extractions from combined clusters revealed that a 9-s and 6-s “shake” removed 84 and 74%, of the thrips, respectively, but a 3-s “shake” removed 53%, and was more variable. Open blossom clusters always had higher thrips densities than either pink or petal fall clusters, regardless of the bloom state. The effects of cardinal position within trees were not consistent over time. Clusters from the top of the canopy had more thrips than lower canopy clusters, and apical clusters had more thrips than basal clusters during peak bloom. Variance component analyses indicated that thrips counts from clusters within tree were more variable than counts among trees, even when cluster types were analyzed separately. Two sets of indices (Iwao’s regression of mean crowding on mean density and Taylor’s regression of log variance on log mean density) for each cluster type indicated aggregated spatial patterns. Precision level sampling plans were developed using Iwao’s regression coefficients.

A list is given for 45 species of drepanosiphine aphids occurring in the Soviet Far East (South Primorye). Descriptions of Iziphya mordvilkoi sp. nov., Subsaltusaphis pulchra sp. nov., S. primoriensis sp. nov., Symydobius minutus sp. nov., and Sinochaitophorus mordvilkoi sp. nov. are presented. A new subgenus Primoriaphis subg. nov. for the species S. pulchra and S. primoriensis is proposed. The taxonomic relationships of the new species are discussed. The following new synonyms are established: Thripsaphis caricis (Mordvilko, 1921) = T. producta Gillette, 1917; Tiliaphis mordvilkoi Pashtshenko, 1984 = T. coreana Quednau, 1979; Tinocallis hemipteleae Zhang and Zhong, 1980 = T. takashihoensis Higuchi, 1972; and Sappocallis Matsumura, 1919 = Tinocallis Matsumura, 1919.

Larvae of Lacinipolia patalis (Grote) (Lepidoptera: Noctuidae) were discovered in Douglas-fir (Pseudotsuga menziesii [Mirb.] Franco) cones collected from the Louisiana-Pacific Corporation's Little River Seed Orchard near Trinidad Head in Humboldt County, CA (elevation 91 m) during the fall of 1985. Previous surveys have not reported this noctuid from Douglas-fir cones (Keen 1958; Tietz 1972; Furniss and Carolin 1977; Hedlin et al. 1980; Ruth 1980; Schowalter et al. 1985). Its usual hosts are blackberry and loganberry (Rubus spp.) and roses (Rosa spp.); it normally feeds on foliage and, to some extent, dead fruit bases and dead leaves (Crumb 1956). The shift from angiosperm leaves and fruit to gymnosperm cones is somewhat surprising. Blackberries, however, grow around the perimeter of the seed orchard and along the rows between trees, and so provide an abundance of the usual host for L. patalis nearby.

In an attempt to control the larch casebearer, Coleophora laricella (Hübner), in western North America with biological agents, insect parasites from Europe and Japan were introduced (Denton 1972, 1979; Ryan et al. 1987). The braconid Agathis pumila (Ratz.) has become widely established and apparently has been responsible for markedly reduced casebearer densities (Ryan 1983, 1986; Ryan et al. 1987). Parasitization by it in many research plots reached 80-95% (unpublished data), then declined following reductions in the casebearer populations in a typically delayed density-dependent manner. I present here a partial explanation for why casebearer densities declined so drastically, to levels even lower than could perhaps be explained by parasitization at such high levels.

In the eastern deciduous forests of North America, oaks (Quercus spp.) that have been weakened by stress are often attacked and killed by the twolined chestnut borer, Agrilus bilineatus (Weber) (Dunn et al. 1986a). This borer attacks stressed and dying trees but does not attack healthy or dead trees (Haack and Benjamin 1982; Dunn et al. 1986a, 1987).