To account for differences in occurrence of social behavior in differe
nt lineages of bees, Michener (1985) hypothesized that ground nests ar
e move easily located by parasites than are twig nests, In the former
case parasites search in two dimensions, while in the fatter they sear
ch in three-dimensional space, One prediction derived from this hypoth
esis is that ground nests will have higher rates of parasitism than tw
ig nests. A survey of published reports on rates of cell parasitism fo
r 92 species of nesting bees and wasps (Apoidea) shows no significant
differences in mean parasitism razes between these two classes of nest
s. The analyses were repeated at the generic level (N = 44), yielding
the same pattern. These data may be biased due to phylogenetic effects
. Paired comparisons (n = 11 pairs) of related taxa that differ in nes
t site show that ground-nesting taxa more often have higher rates of p
arasitism than twig-nesters. The use of artificial ''trap-nests'' to s
tudy twig-nesters significantly enhances the success rare of parasites
. This bias, as well as several other limitations, suggests that exper
imental studies of the host-searching capabilities of parasites and pr
edators may be more efficacious than such comparative tests.