Arch Virol (2000) 145: 2015–2026
Molecular phylogenetics and the classiﬁcation of honey bee viruses
J. D. Evans and A. C. Hung
USDA-ARS Bee Research Laboratory, Beltsville, Maryland, U.S.A.
Accepted May 17, 2000
Summary. We present the phylogenetic relationships of several picorna-like
RNA viruses found in honey bees, with respect to 13 additional plant and an-
imal positive-strand RNA viruses. Most of the honey bee viruses fall into an
unnamed family of insect RNA viruses typiﬁed by the Drosophila C virus. Dif-
ferent bee viruses are broadly distributed within this group, suggesting either that
the ability to infect honey bees has evolved multiple times, or that these viruses
are generalistic in their abilities to infect insect hosts. At least one major change
in gene order has occurred among the bee viruses, based on their phylogenetic
afﬁliations. At the amino-acid level, the bee viruses differed by 15–28% at three
conserved loci. Most differed by greater than 50% at the RNA level, indicating
that sequence-based methods for bee virus identiﬁcation must be tailored to at
least three different virus clades independently.
In addition to producing honey and other valuable hive products, honey bees (Apis
mellifera), provide important pollinator services worldwide, with a net economic
value in the tens of billions of dollars. Parasites and pathogens can strongly impact
both feral and domesticated honey bee populations . The most visible threats
to honey bees, and arguably the most important, are parasitic mites in the genus
Varroa. Nevertheless, the impact of parasitic mites can be heightened by the
effects of other honey bee parasites and pathogens, including viruses [7, 8, 15].
Thanks to several decades of study, a considerable database exists for the
serological and physical identiﬁcation of honey bee viruses (reviewed by [3, 6])
and for determining the efﬁcacy of several virus species experimentally [2, 4].
Despite these advances, the impacts of speciﬁc virus taxa on honey bee popula-
tions have been notoriously difﬁcult to assess [7, 31]. We feel that these difﬁcul-
ties have arisen from two major barriers. First, it has proved difﬁcult to determine
the presence of viruses in bees from the ﬁeld, except in those (apparently rare)
cases when individual bees manifest diagnostic physiological and behavioral