Biology Reference
In-Depth Information
Þow, they showed that in cages with no CI or unidirectional CI the decay proceeded rapidly; in
contrast, in cages with bidirectional CI there was little to no decay in linkage disequilibria. Thus,
bidirectional CI can limit gene Þow and maintain genetic divergence between populations. Recent
theoretical work by Telschow et al. (2002b) also shows that bidirectional CI can accelerate genetic
divergence among populations even when bacterial transmission is inefÝcient (i.e., infected females
lay some uninfected eggs) and CI levels are incomplete (i.e., not all offspring die in incompatible
crosses). Thus, CI may assist speciation over a broad range of biologically realistic conditions.
Any attempt to understand how bidirectional CI can promote speciation is confronted with the
question of how bidirectional CI arises, that is, how can two or more different CI-
Wolbachia
make
it into the same host species? BrieÞy, there are four paths to the evolution of reciprocal CI. The
Ýrst is independent acquisition, in which two allopatric populations acquire different
Wolbachia
through independent horizontal-transfer events. The phylogenetic evidence largely favors this path
(Table 17.1),
and clear examples can be found in
D. simulans
(Clancy and Hoffmann, 1996) and
Nasonia
(Werren et al., 1995b; Bordenstein and Werren, 1998; Werren and Bartos, 2001). The
second path is codivergence, in which a single, ancestral
Wolbachia
diverges into two CI-
Wolbachia
types within its hostÔs species. There is phylogenetic evidence of codivergence at the genetic level
in the species pair
N. giraulti
and
N. longicornis
(Werren et al., 1995b; Werren and Bartos, 2001).
However, both of these species are infected with strains of A and B
Wolbachia
, and only the B
Wolbachia
has undergone codivergence. Measuring bidirectional CI between the two B strains will
require isolation of single B-infected
N. giraulti
and
N. longicornis
, an effort that has so far proven
difÝcult (S.R. Bordenstein and J.H. Werren, unpublished). Third, segregation of a double infection
(e.g., A and B) can lead to individuals harboring single A and B
Wolbachia
that are bidirectionally
incompatible. However, the establishment of single A- and B-infected populations would be unlikely
because double infections can spread easily against single infections (Sinkins et al., 1995; Perrot-
Minnot et al., 1996). Finally, different host genetic inÞuences on the same
Wolbachia
variant could
possibly lead to bidirectional CI between populations or species. The diversity of incompatibility
types in
C. pipiens
, but lack of
Wolbachia
sequence variation among these CI types, is consistent
with this model (Guillemaud et al., 1997). However, only one
Wolbachia
gene has been surveyed
for genetic variation in this system. Taken together, results thus far indicate that bidirectional CI
typically evolves through independent acquisition: 14 of 19 cases show evidence of independent
acquisition via horizontal transfer (Table 17.1).
Model 2 (CI Coupled with Genetically Based Isolation)
Wolbachia
-induced CI will probably play its most signiÝcant role in speciation when it is coupled
with additional isolating barriers. This view is consistent with at least three lines of evidence. First,
the likelihood that
Wolbachia
will be the only cause of speciation is reduced due to the typically
incomplete levels of CI (Boyle et al., 1993; Breeuwer and Werren, 1993), inefÝcient bacterial
transmission (Turelli and Hoffmann, 1995), and any reproductive isolation that has evolved before
Wolbachia
enter a host system. Second, unidirectional CI between allopatric populations is common,
probably more so than bidirectional CI. Because unidirectional CI is a one-way cross incompatibility
between infected and uninfected populations, gene Þow can still be Þuid through populations via
the compatible cross direction (uninfected male infected female). And third, there is a growing
consensus that speciation rarely occurs due to a single form of reproductive isolation. Rather, as
populations diverge and begin to sustain independent evolutionary fates, reproductive isolation will
gradually evolve due to the accumulation of several (incomplete) isolation barriers (Coyne and Orr,
1997; Sasa et al., 1998; Presgraves, 2002). However, as we will see,
Wolbachia
can still play an
essential role in promoting species formation even when CI is coupled with other isolating barriers.
Consider two allopatric populations, one Ýxed for a
Wolbachia
infection and the other unin-
fected. What would happen if infected individuals migrated into the range of the uninfected
population or vice versa? Would these two populations fuse back into one? Would the
Wolbachia
