Symbiosis and the Origin of Species - Insect Symbiosis - page 226

were to make contact and hybridize? Hybridization would lead to high levels of F1 hybrid inviability
due to CI, and parents would be selected to mate discriminately with compatible individuals. How
strong must CI be to select for mate-discrimination genes? What levels of migration will restrict
or promote CI-assisted reinforcement? Will the required linkage disequilibrium between cytotype
and genotype be maintained to complete reinforcement, and will the incipient species just fuse or
will one displace the other? These questions await answers from theoretical studies as well as
natural and experimental systems where the conditions that affect reinforcement can be directly
tested. Circumstantial evidence for this process comes from Nasonia . Two species ( N. giraulti and
N. longicornis ) are microsympatrically embedded within the range of the third species, N. vitrip-
ennis . All three species are bidirectionally incompatible, and, consistent with reinforcement, N.
giraulti and N. longicornis females show stronger discrimination against N. vitripennis males than
the reciprocal cross (Bordenstein and Werren, 1998; Drapeau and Werren, 1999; Bordenstein et
al., 2000). However, several postzygotic barriers besides CI exist between these species (Table
17.2), any of which may facilitate reinforcement.
Is there any reason to think that Wolbachia -induced CI may have an unusual effect on rein-
forcement? It seems likely that reinforcement is more probable when driven by Wolbachia -induced
CI rather than intrinsic postzygotic genetically based incompatibilities. Let us assume that the two
scenarios are identical except for the basis of the postmating isolation between the populations.
Thus, for example, the amount of migration and intensity of the postmating isolation are the same.
However, hybrid Ýtness is reduced due to Wolbachia in one scenario and a simple two-locus genetic
incompatibility in the other. Now why might Wolbachia have a higher likelihood of driving
reinforcement? Because CI halts gene Þow at the F1 generation, whereas most genes involved in
early genetic incompatibilities are recessive and limit gene Þow in some F2 genotypes or the
heterogametic F1 genotype (in accordance with HaldaneÔs rule). The upshot of this difference is
twofold. First, the F1 isolation caused by CI reduces more gene Þow by eliminating hybrids
irrespective of their sex or genotype; second, F1 isolation prevents recombination from slashing
the required linkage disequilibria between the incompatibility locus and the mate-discrimination
locus (Felsenstein, 1981; Kirkpatrick and Servedio, 1999). Recessive incompatibilities do not share
this luxury because more Ýt hybrids will be produced and recombination in the previous generations
can break down the required linkage disequilibria. Theoretical treatments of these issues are needed
to evaluate this prediction and the conditions associated with CI-assisted reinforcement.
ASEXUALITY, SEXUAL DEGRADATION,
AND THE ORIGIN OF SPECIES
The process by which an asexual population splits from a sexual population is a form of cladogenesis
that can be termed asexual speciation. This process falls neatly under the Biological Species Concept
because it is concerned with the severing of gene Þow and the evolution of reproductive isolation
between sexual and asexual populations. But Ýrst it is necessary to distinguish between two forms
of asexual reproduction Ð arrhenotoky and thelytoky. The former is the typical mode of reproduc-
tion in haplodiploid insects where males are produced from unfertilized (haploid) eggs and females
are produced from fertilized (diploid) eggs. The latter, in which all unfertilized eggs become
(diploid) females, is the one we are concerned with here and will generally be referred to as
asexuality in the text.
Asexuality may be under symbiotic or genetic control, though there is good reason to believe
that symbiotic bacteria are more often than not the causative agent, at least in haplodiploid
organisms. Over 30 cases within the Hymenoptera have been documented by Luck et al. (1992),
and within the last few years there has been a modest burst of work revealing a bacterial basis of
asexuality. Every case deals with a cytoplasmically inherited bacterium, and while not all the
bacteria have been identiÝed, Wolbachia have historically been the most common culprit
(Stouthamer et al., 1993; Stouthamer, 1997). The mechanism of Wolbachia -induced parthenogenesis