Here's the Bradshaw and Schemske Monkeyflower speciation study. Here are two sister species of Monkeyflower, Mimulus (link). M. lewisii is bee pollinated, whereas M. cardinalis is hummingbird pollinated. Although the two species seldom hybridize naturally, due to their different pollinators, they can be cross pollinated in the greenhouse to yield F1 hybrids, and the F1's can be crossed with each other to yield F2's. In this figure (link, link), A is M. lewisii, C is M. cardinalis, B is the F1 hybrid, and D-L are different phenotypes of the F2's. Bradshaw and Schemske took these F2's (228 of them), along with 24 each of the parental species and the F1 hybrids, back out into the field and recorded 12,567 pollinator visits by bees and hummingbirds. From this, they were able to demonstrate natural selection for the parental species phenotypes. They also did QTL mapping (quantitative trait loci) of major floral traits, and are able to correlate selection on phenotype with selection on genotype.

Here's a summary of their results.

• Increased petal anthocyanins, petal carotenoids, and nectar loads actually reduced the proportion of pollinator visits by bees (by increasing the proportion of pollinator visits by hummingbirds); whereas, increased flower projection area (i.e. size) increased the proportion of visits by bees (link).
• QTL alleles for petal carotenoids were found to have a significant affect on pollinator visits. Whereas bees had a strong aversion to high petal carotenoids, which give orange to red flowers, hummingbirds apparently were indifferent to petal carotenoids, or at best had a weak and statistically non-significant preference for high petal carotenoids (link).
• QTL alleles for nectar load also had a signifcant affect on pollinator visits. Hummingbirds had a stronger preference for the cardinalis (i.e. high nectar volume) QTL allele; whereas bees did not (link). Somehow, hummingbirds were able to determine,or rapidly learn, which F2 hybrids (D-L in the figure, link) had the high nectar allele.

In subsequent studies, they used a breeding design of hybridization, backcrossing and inbreeding (i.e. recombinant inbred lines). With this design they were able to generate plants that only vary in a trait of interest; e.g. they could effectively recreate the original species except exchange only a QTL allele (e.g. for carotenoids) of interest. These plants have also been taken back to the field, and their fitnesses (pollinator visits) compared with each other and controls. All of this gives a very good picture of the genetics and evolution of reproductive isolation for this pair of species, and suggests that genes with large phenotypic effects can play a role in speciation. (Mimulus Class Slides)

References (available online):

Beardsley, P. M. Yen, A. and R. G. Olmstead. 2003. AFLP phylogeny of Mimulus section Erythranthe and the evolution of hummingbird pollination. Evolution 57: 1397-1410. (link)

Bradshaw, H.D., Jr. and Schemske, D.W. (2003) Allele substitution at a flower colour locus produces a pollinator shift in monkeyflowers. Nature 426: 176-178. (link)

Bradshaw, H.D., Jr., Otto, K.G., Frewen, B.E., McKay, J.K. & Schemske, D.W. (1998) Quantitative trait loci affecting differences in floral morphology between two species of monkeyflower (Mimulus). Genetics 149: 367-382. (link)

Ramsey, J. R., Bradshaw, H. D., Jr., & Schemske, D. W. (2003) Components of reproductive isolation in the monkeyflowers Mimulus lewisii and M. cardinalis (Phrymaceae). Evolution 57(7): 1520-1534. (link)

Schemske, D.W. & Bradshaw, H.D., Jr. (1999) Pollinator preference and the evolution of floral traits in monkeyflowers (Mimulus). Proceedings of the National Academy of Sciences USA 96(21): 11910-11915. (link)

Homepages:

Schemske: http://www.plantbiology.msu.edu/schemske.shtml

Bradshaw: http://faculty.washington.edu/toby/

Beardsley: http://www.isu.edu/~bearpau2/

Olmstead: http://depts.washington.edu/phylo/