The banana weevil, Cosmopolites sordidus (Germar), is the most important constraint to banana production in East African highlands. For most resource-poor farmers in the region, biological control is the only viable option because it requires little or no cash investment. I investigated patterns of predation by natural enemies of the banana weevil in its presumed native range in Indonesia to determine if there were natural enemies that could be imported to control the pest in Africa. Work was also done in Uganda to determine if predatory insects already present in local banana farming systems could be better conserved to produce banana weevil control. Through farm surveys in Indonesia, I confirmed existing reports that banana weevil damage is low in Sumatra and Java. However, I found no evidence of parasitism from 25,000 eggs and 3600 larvae collected from seven diverse geographical locations. The adult and larvae of the histerid, Plaesius javanus Erichson, were found to be important predators of C. sordidus. P. javanus larva entered tunnels of plants, presumably in search of banana weevil stages. This predator should be imported to Uganda for establishment as a classical biological control agent of the banana weevil. Experiments in Uganda showed that destruction of crop residues in bananas, as recommended for weevil control and practiced by some farmers, reduced predator numbers on farms, reduced predator: prey ratios and had no benefit to the plant with respect to damage prevention. Instead, I demonstrated that residue presence, through maintaining high predator: prey ratios, offsets damage to plants that would otherwise occur from increased numbers of banana weevils. Ants are the major predatory group now present in banana farming systems in Uganda. Surveys in banana farms found 55 species from pitfall traps, 17 from banana pseudostem residues and 34 from banana corms. Eleven species of ants came to banana weevil egg and larvae exposed in the field as baits and thirteen species were tested in the laboratory for their potential to attack the banana weevil. Two species—Pheidole sp. 2 (Myrmicinae) and Odontomachus troglodytes (Santschi) (Ponerinae)—caused significant banana weevil larvae mortality in crop residues and significant egg mortality in living plants in microcosm experiments and in the field.

Listronotus maculicollis (Dietz) is a major pest of golf course turf in the northeastern United States. Because the larval stage of this insect causes considerable damage to short-mowed turfgrass, such as tees, greens, fairways, golf course superintendents rely on chemical applications for control. I investigated physiological, ecological, and behavioral characteristics of L. maculicollis to enhance management strategies that will lead to reduced insecticide inputs in golf course turf. Among the cool season turfgrass species, L. maculicollis larvae are reportedly found primarily in Poa annua L., annual bluegrass, among cool-season turfgrass species. To confirm this observation, I conducted a quantitative investigation to determine how abundant L. maculicollis larvae were in P. annua compared with other grasses. I also investigated the influence of mowing height and fertilization on the abundance of larvae. L. maculicollis larvae were present in highly maintained grasses (P. annua and Agrostis palustris Huds., creeping bentgrass) in field studies; no differences in numbers of larvae were detected between P. annua and two types of creeping bentgrass. However, in choice and no-choice tests among five grass types, L. maculicollis were significantly abundant in P. annua. Additionally, one study showed L. maculicollis larvae collected from P. annua weighed more than larvae from other grass types. I found a significant effect of fertilizer application. More L. maculicollis were collected in non-fertilized turf compared to fertilized turf, and more larvae were collected in short-mowed plots than from long-mowed plots. I also investigated adult spring emergence and the distribution of adults and larvae on turfgrass hosts. Golf course superintendents observe primary damage on perimeters of short-mowed areas. In an attempt to corroborate their observations, I examined the distribution of adults and larvae across the width of a golf course fairway, but no differences were detected. Although no more larvae were detected at fairway edges, we determined by visual assessment that the turf on the edge of the fairway was poorer quality, and substandard turf quality is often a result of pest, mechanical, or environmental damage. Therefore, from our results, larval feeding alone does not account for the increased damage in edge areas. I also established that adult L. maculicollis emerge from overwintering sites and walk onto host plants in the spring. These results will be utilized to develop perimeter treatment. (Abstract shortened by UMI.)

This project focused on field evaluations of synthetic host and pheromonal odor combinations for attractiveness to plum curculios, Conotrachelus nenuphar (Coleoptera: Curculionidae), with the aim of developing a monitoring device that could track the onset and extent of immigration into apple orchards, thereby maximizing plum curculio control while minimizing exposure and cost by helping growers to accurately time insecticide sprays. In a 2000 field test I screened six individual host plant volatiles, each in association with grandisoic acid (synthetic plum curculio aggregation pheromone), using panel and pyramid traps. Benzaldehyde was the most attractive host plant odor when in combination with grandisoic acid. Evaluations conducted in 2001 confirmed the attractiveness of this binary combination to both males and females across the entire period of immigration, and also revealed that benzaldehyde interacts in a synergistic manner with grandisoic acid. Additional field studies conducted in 2002 and 2003 showed that benzaldehyde at 40 mg/day of release in association with grandisoic acid at 1 mg/day of release is the most cost-effective bait combination for use in traps to monitor plum curculio immigration. I also investigated temporal dynamics of plum curculio immigration into an unsprayed section of an apple orchard over a five-year period using traps. My main goal was to determine relationships among the timing of immigration, temperature and phenological tree stage. Based on my data, I propose a pre- and a post-petal-fall period of immigration, each of which is influenced to a different extent by spring temperature. Thermal constants (expressed in Degree Days [DD] base 6.1°C, computed starting January 1st) for the start of immigration, 50th and 80th percentile of cumulative captures were 113, 249 and 412 DD, respectively. Some spatial aspects of plum curculio immigration were also studied in commercial apple orchards over a two-year period. By petal fall, most adults were found on perimeter-row trees in sprayed blocks having large trees (M.7 rootstock), but more adults were found inside of blocks than on perimeter-row trees if blocks had small trees (M.9 rootstock). Plum curculios were able to overwinter inside blocks, regardless of the presence or absence of weeds beneath trees.