Parthenogenesis Associated with Wolbachia - Insect Symbiosis - page 187

Many PI- Wolbachia are distributed in the B group, but a few also occur in the A group (Werren
et al., 1995; Gottlieb et al., 1998; Zhou et al., 1998; Plantard et al., 1999; Vavre et al., 1999; van
Meer et al., 1999; von der Schulenburg et al., 2000). When we review these studies and calculate
the percentage of PI- Wolbachia in both groups, using only those Wolbachia cases where the
phenotype has been established, we Ýnd PI- Wolbachia in 28% (10 of 36) of the Wolbachia in group
A and 54% (20 in 37) in group B.
Only in Trichogramma spp. do PI- Wolbachia show a clear phylogenetic pattern. Independent of
the studied gene, wsp , ftsZ , or 16S rDNA, these Wolbachia always form a separate cluster. Schilthuizen
and Stouthamer (1997) studied the phylogenetic relationships between PI- Wolbachia and their Tri-
chogramma hosts in detail. A comparison of the phylogenetic tree of the host species with the tree
of Wolbachia in these species showed them to be incongruent. The major reason for this incongruence
is most likely horizontal transfer of the Wolbachia among different Trichogramma species.
Do PI- Wolbachia differ in this respect from Wolbachia inducing other effects? At present we
cannot make strong statements concerning this pattern because our sample size is too small. For
parasitoids of the genus Aphytis several PI- Wolbachia -infected lines have been sequenced using
the ftsZ gene (Gottlieb et al., 1998). In this genus, the ftsZ sequence of the Wolbachia is identical,
suggesting that here too we Ýnd a high similarity of PI- Wolbachia within the genus. Similar
patterns are also found to some extent in CI-inducing Wolbachia , for example, the sibling species
of Nasonia spp. carry similar Wolbachia . The same applies to Culex spp. (Zhou et al., 1998).
Wolbachia in Diplolepis , most likely associated with PI, show an extraordinary pattern: the
Wolbachia in Ýve Diplolepis species from France cluster together in the B group, whereas Ýve
other PI- Wolbachia from North American species are distributed over the A and the B group
(Plantard et al., 1999). This distribution would be consistent with the hypothesis that the pecu-
liarities of the host induce the PI effect. The Leptopilina spp. are a very interesting case because
here CI- and PI- Wolbachia exist within the same genus. The pattern we observe is that all CI-
Wolbachia in Leptopilina are distributed over the A group, mostly associated with the Wolbachia
of their host species (Vavre et al., 1999), but the PI- Wolbachia in L. clavipes and L. australis
occur in a separate cluster in the B group.
Besides Schilthuizen and Stouthamer (1997), many other phylogenetic studies detected indirect
evidence of horizontal transfer (OÔNeill et al., 1992; Stouthamer et al., 1993; Rousset and Solignac,
1995; Werren et al., 1995; Vavre et al., 1999). Host-to-parasitoid and parasitoid-to-parasitoid trans-
fer of Wolbachia has been considered mainly as potential transmission routes. Horizontal transmis-
sion was shown for PI- Wolbachia by Huigens et al. (2000): horizontal transfer of PI- Wolbachia
between T. kaykai parasitoids occurs when they share the same food source.
HOST FITNESS: HIGHER COSTS OF CARRYING WOLBACHIA
IN MIXED POPULATIONS
The effects of Wolbachia on the Ýtness of its host are inÞuenced by the following factors: trans-
mission route, presence of a genomic conÞict between Wolbachia and host, and the physiological
cost of carrying bacteria. The relationship between Wolbachia and its host can range from an
adversarial relationship in populations where only a fraction of the females is infected with the PI-
Wolbachia (under these circumstances we expect a genomic conÞict in which the nuclear genes of
the host may try to suppress the Wolbachia or its effect) to a completely mutualistic relationship
in which the infection with Wolbachia in a host population has gone to Ýxation and all females of
a population are infected with Wolbachia .
T RANSMISSION R OUTE
Since the early part of the 20th century, the evolution of parasite virulence has received much
attention. Although PI- Wolbachia are not true parasites but nonobligatory symbionts, theories on