Domestication, however, has eroded sub-species therefore distinctions through hybridization, particularly in regions such as North America where A. mellifera was not native. It is common practice among North American beekeepers to replace queens every one to two years to maximize productivity [5]. These queens originate from a restricted set of queen breeders situated in regions optimal for queen production and mating. In the United States these regions are located in Hawaii, central California and along a south-eastern band spanning from Florida through to Texas. While a small number of queens in Canada are produced domestically, the majority are imported from central California, Hawaii, New Zealand, Australia or Chile.

Since the genotypes of the individual workers in the colony are derived from the mated queen, this practice undermines the stock improvement goals of queen purchasers in two ways. First, purchasers frequently value traits differently than queen breeders [6]. Second, the agro-ecological conditions where queens are selected may not resemble those where the queens are used. Combined, these aspects results in a situation where many beekeepers operate without the full benefits of stock improvement. Like any livestock, the variation in phenotypes observed among honey bees are a product of artificial and natural selection. The common methodology for estimating variation among populations, however, provides only a limited picture of the adaptive significance of this variation. Such methods rely on quantifying neutral genetic variation among populations by correlating microsatellite markers with quantitative traits found in the populations.

Consequently, these techniques provide little insight into the biochemical mechanism(s) at work in adaptation [7]. Mutations that occur in protein coding regions are infrequent but can lead to mechanistic insight: in feral honey bees the identification of locally adapted population clines due to geographic diversity has been shown previously by the polymorphism of alloenzymes [8], [9]. Of the large-scale approaches available to study biological diversity, next-generation sequencing technology allows a deep and high-resolution probing of differences among groups or individuals in a species [10] but is too far removed from the level of proteins to provide much functional insight into the adaptations. Even mRNA expression profiling, either by RNA-Seq [11] or more classical microarrays Entinostat [12], [13], is not consistently correlated with protein expression [14], [15]. Proteomics [16], in contrast, directly measures biomolecules responsible for responding to a changing environment and so is ultimately the best approach for probing the underlying mechanisms at work in adaptation.