No description available.

The main object of this study was to investigate the effects of predation by starlings (Sturnus vulgaris) on populations of the grass grub (Costelytra zealandica) in Canterbury, with the secondary objective of investigating the feasibility of life table studies of grass grub populations under Canterbury conditions. Study plots were established in four areas, two on non-irrigated pastureland at Lincoln and Weka Pass and two on irrigated pastureland at Winchmore. A third study plot was established at Winchmore in the second year of field work.

Bumble bees (Bombus spp.: Apidae) are valuable pollinators of many crop and wildflower species. However, in some situations their potential is limited. Evaluation of, and management to improve bumble bee efficacy should include spatial information which is currently limited. Distance and direction determine the success of gene flow via pollen cross-over within and between plant populations at several scales. Studies of movement by bumble bees at large scales in semi-natural and intensively managed habitats are scarce. Few studies of bumble bee dispersal from the nest exist, particularly in relation to crops. At a small scale, directional rather than random movement between flowers has benefits for pollen flow. Results to date of directionality studies at small scales and their interpretation are inconsistent. The purpose of this thesis was to assess distances and directions moved by foraging bumble bees at a range of scales in two contrasting farm habitats in order to predict their pollination potential. A novel method was developed to mark automatically all the occupants of nests of bumble bees B. terrestris (L.) placed around a Lucerne seed crop Medicago sativa L. in New Zealand. Reobservation data from eight nests showed that of bumble bees which foraged within the crop, 81 % travelled ≤ 50 m and 56% ≤ 20 m from their nest. Results should be interpreted with extreme caution because fewer than 1 % of bumble bees marked at nests were reobserved in the crop. Because it was not established where the other 99% of the bumble bees went, foraging areas for nests could not be calculated as anticipated. Theories to explain the non-specificity of bumble bees to the crop include; resource depletion near nests, competition with honey bees in the crop, or an evolved strategy to disperse in order to minimise nest predation. Lucerne flowers contained a significantly lower concentration of sugar in nectar, and significantly fewer pollen grains than did those of purple loosestrife Lythrum salicaria L., a species on which bumble bees appeared to forage in preference Lucerne. The higher rewards offered by L. salicaria may have diverted bumble bees from the less-rewarding Lucerne crop. In a Norwegian meadow system, all foraging bouts by bumble bees B. lucorum (L.) within a patch of wood cranesbill flowers Geranium sylvaticum L. were random with respect to direction. This result is not consistent with predictions, based on optimal foraging theory, that movement should be directional to enable optimal pollen flow, and to avoid revisitation of just-emptied flowers by the pollinator. A medium-scale study of several bumble bee species moving between patches of northern wolfsbane Aconitum septentrionale Koelle in Norway revealed considerable loyalty by bumble bees to patches in which they were marked. In a different landscape-scale study (over 5 ha), several bumble bees exhibited a high degree of loyalty to areas in which they were marked (87% were reobserved ≤ 50 m from marking points). These restricted movement patterns are discussed in terms of potential pollen flow. Of 260 bumble bees marked, only five were recorded crossing between meadows, which could be a result of innate loyalty to small forage areas, an artefact of the sampling technique used, or forest boundaries acting as physical impediments to movement. In the future, spatial data of the type collected in this thesis will aid in the management of bumble bee populations to achieve both commercial and conservation goals. Spatial data can be applied to predict the optimal placement of artificially-reared nests, predict suitable isolation distances for pure seed crops, and aid in the positioning of supplementary forage sources and nest-site refuges.