4. Annex: Evolutionary lineages of Apis mellifera
Besides serving the aims of identifying samples and discriminating subspecies or populations, the techniques discussed in this chapter have been used from the beginning (Ruttner et al., 1978) to shed light on the evolutionary history of Apis mellifera and its many subspecies and regional populations. Innumerable publications have been produced on this subject, with the results produced with morphological and molecular techniques partially supporting each other, while being incongruent in several aspects (Garnery et al., 1992, 1993; Franck et al., 1998), leading, unfortunately, to substantial confusion about the number of “true” evolutionary subgroupings and the appropriate way of naming them.
Ruttner (1988) developed and introduced the concept of “branches”, or evolutionary lineages within Apis mellifera, upon the notice that in statistical analyses of morphometric data samples from different subspecies grouped together according to their geographic origin. Based on morphometric analysis results, A. mellifera subspecies can therefore be grouped into four well differentiated lineages: a west Mediterranean and northwest European lineage M (mellifera and iberiensis, but originally also including intermissa, sahariensis, siciliana, as well as the related and later described ruttneri (Sheppard et al., 1997)) that were considered as links between the tropical African and the west Mediterranean subspecies), lineage C from southeastern Europe and the eastern Mediterranean (ligustica, carnica, macedonica, cecropia, cypria, adami), lineage O in the Near East and western Asia (caucasica, anatoliaca, syriaca, meda, armeniaca, jemenitica, and the later described pomonella (Sheppard and Meixner, 2003)), and lineage A (lamarckii, , andansonii, scutellata, monticola, litorea, capensis, unicolor, and similar and newly described simensis (Meixner et al., 2011)) from the African continent.
Following the development of molecular techniques and their application to the intraspecific variation of A. mellifera, it became evident that mitochondrial DNA variation reflected the lineages proposed by Ruttner in some cases, while in others it did not. While subspecies from Africa and north and west Europe grouped together consistently into lineages M and A based on both morphological and mtDNA results (with intermissa, sahariensis and siciliana belonging to a North African group within lineage A), this was not the case for the Mediterranean, east European and western Asian subspecies that formed the morphological lineages C and O. Regarding these groups, mitochondrial DNA patterns did not vary substantially and did not differentiate between them. This result was interpreted in a way that a mitochondrial lineage C was hypothesized inclusive of all subspecies belonging to the morphologically defined C and O lineages, with the existence of lineage O being questioned (Garnery et al., 1992). Thus, in the subsequently published body of papers on mitochondrial variation within A. mellifera, a model of only three major evolutionary lineages (A, M, C) was generally assumed to reflect the evolutionary history of the species (Garnery et al., 1993; Franck et al., 1998), and the morphologically based model of four lineages was mostly rejected.
Later however, populations from the Near East were found to exhibit mitochondrial patterns different from the three lineages found so far, and the existence of a fourth mitochondrial lineage was postulated (Franck et al., 2000; Palmer et al., 2000). Unfortunately however, following the interpretation that the mitochondrial “associate” of the morphological lineage O had finally been detected, this newly found mitochondrial lineage was also named O (Franck et al., 2000), now making confusion complete. Upon the inclusion of additional data and a new analysis, this lineage has recently been identified as a sublineage of the African lineage A and been renamed to Z (Alburaki et al., 2011). Finally, yet another mitochondrial lineage, named Y, has been identified in northeastern Africa (Franck et al., 2001) which also belongs into the context of the African lineages (Meixner et al., unpublished).
Most interestingly, in recent comprehensive analyses based on nuclear markers (SNPs), and including representative samples of 14 subspecies (Whitfield et al., 2006), the resulting groupings largely reflected the traditional four lineages postulated on the basis of morphology.