Site constancy of bumble bees in an experimentally patchy habitat

Habitat fragmentation alters the spatial and temporal distribution of flora l resources in farmland. This will affect the foraging behaviour of bees ut ilising these resources and consequently pollen flow within and between pat ches of flowering plants. One element of bees' foraging behaviour, which is likely to be affected, is the degree to which individual bees remain const ant to a particular site or patch, both within and between foraging trips. Mark-re-observation was used to investigate whether foraging bumble bees sh owed site constancy over several days to regular patches of forage, even wh en those patches contained qualitatively and quantitatively similar resourc es. The authors also investigated whether site constancy was affected by th e arrangement of patches within the area. The experimental arena was a fiel d of barley containing patches of a grass/herb mixture, including Centaurea nigra L. (black knapweed) which provided nectar and pollen for bumble bees , particularly Bombus lapidarius L. Patches were either contiguous or non-c ontiguous in patch groups. Twenty to 28% of marked B. lapidarius were re-ob served in the experimental arena during the week following marking. The num ber of re-observations of bees decreased over time probably because floral density decreased, the bees sought alternative forage elsewhere or they die d from natural causes. The bees showed striking site constancy: 86-88% of r e-observations were constant to patch group (27 x 27 m(2) or 45 x 45 m(2)) and fewer re-observations were constant to small patches (9 x 9 m(2)) withi n a patch group. Most observed moves were to patches or patch groups adjace nt to those on which the bumble bees were marked. There was Limited evidenc e that bumble bees were more constant to 9 x 9 m(2) patches surrounded by b arley (70-73% of re-observations were patch constant) than to 9 x 9 m(2) pa tches that were contiguous (20-50% of re-observations were patch constant). The implications of the observed bumble bee site constancy for plant gene flow in the arable ecosystem are discussed. (C) 2001 Elsevier Science B.V. All rights reserved.