The sampling design: across agricultural land, with some woods, hedgerows, and a nearby town in Hertfordshire. The sampling points (in red) are approximately 250 m apart.
They screened the workers using 8 to 9 very variable genetic markers (microsatellites, which are also used in human paternity testing) and identified sisters by assessing the number of markers sisters should be expected to share. Their results were then used to calculate the maximum foraging distances of nestmates of the four species. It was 449 m for B. pascuorum, 758 m for B. terrestris, 450 m for B. lapidarius and 674 m for B. pratorum. Quite a long way, I would say. B. terrestris is the species flying longer to forage, and also the species with the larger workers and more workers per nest. A much more striking result of this study, though, is the number of nests from which workers come to feed in a specific sampling point. Knight and coworkers estimated that, on average, B. lapidarius workers came from 75 different nests to a specific sampling point, these figures were 43 nests for B. pascuorum, 52 for B. terrestris and 37 for B. pratorum. Of course this study applies to farmland, but for B. terrestris, nest densities might be even higher in urban areas. It makes me feel good to know that even when, frustratingly, there are no bumblebee nests in my garden, it still supports a very high number of nests from a range of species.