The aim of this project was to provide a framework for future habitat management schemes aimed at conserving bumblebee populations in agro-ecosystems. I focused on the identification of landscape components that could be of value in the implementation of future conservation schemes. Specifically, two areas where individual bumblebees display ecological 'choice' were investigated; forage choice, and nest-site choice. The availability of suitable forage resources and nest sites are two important components in the continued survival of bumblebee populations on agricultural land, and must be considered in planning any conservation scheme. One way in which increased melliferous forage can be provided in agro-ecosystems having impoverished floral diversity is through allowing natural regeneration along field margins, or sowing these with a wildflower seed mix. I found that margins of both types attracted a higher density and diversity of foraging bumblebees than corresponding edges managed as conservation headlands. Floral diversity was found to be a much more important component of overall attractiveness than floral density, and perennials were a much more attractive resourse than annuals; the preference of different bee species for different forage plants indicates that conservation schemes could have a species-specific component. Growing certain crop species could greatly supplement the floral resources provided by areas of semi-natural vegetation, especially such nectar-rich 'novel' crops as Onobrychis viciifolia. Newly emerged queens displayed interspecific site preferences when nest searching along field boundaries. Subterraneous nesting species (B. terrestris, B. lapidarius, B. lucorum) showed a strong preference for searching along banks, whilst B. pascuorum, B. hortorum and B. ruderarius searched preferentially along edges with tussocky vegetation. B. lapidarius in particular was most frequently observed searching along exposed boundaries; this species has a higher temperature threshold, and nests along exposed edges may maximise the heat reservoir effect of the soil. It is clear that a range of different boundary types need to be provided if nesting requirements of a variety of bumblebee species are to be met. These findings are discussed in detail, with reference to the implications for conservation of bumblebees in agro-ecosystems. The possible effects on other species of flora and fauna are also considered.