Pan traps were set out at all plots at 9 am and collected at 5 pm
once every 2 weeks from 2 February to 13 April, 2006 (six sampling
days). This period extends 6 weeks before and after the coffee
bloom, from the middle of the dry season to the beginning of the
wet season in southern Mexico. We processed the specimens on
each sampling day by pinning and identifying them to the genus
or subgenus level. All bees found in pan-traps in a single site were
summed to provide a site-level count per sampling date (for a total
of 42 samples). Bee pan traps were combined because they are not
independent from one another and because the pan-trap colors do
not differ significantly in the bees they attract (LeBuhn et al., 2005).
We identified all individual bees trapped. For species-level identifications,
we referred to local collections at El Colegio de la Frontera
Sur in Tapachula, Mexico and the expertise of T. Griswold, at the
USDA Bee Lab in Logan, Utah. For some individuals, we could not
arrive at species-level classifications, thus we left their classification
at the genus level (with a morphospecies label). Because many
bee groups do not have adequate taxonomic treatments (Michener,
2000), especially in the tropics, identification to the morphospecies
level was necessary. Previous work shows that morphospecies-level
identifications can serve as a good proxy in the estimation of
species richness (Oliver and Beattie, 1996).