CHROMOSOMAL EVOLUTION AND SPECIATION
119
needed, but a donor chromosome is equally necessary. It is assumed
that the donor undergoes breakage and that the two portions "seal" the
freshly broken chromosome ends.
The donor chromosome may be a regular member of the set (such as
a Y) or it may be a supernumerary element. In the former case we
assume that the unbroken donor persists in the species together with
the translocated portions, i.e., that the "fragmentation" results in hyperploidy for the genetic material contained in the donor. If the unbroken
donor does not persist, we are dealing with a simple translocation which
does not involve an increase in chromosome number. Where the donor
is a supernumerary it can, however, be impermanent.
Evolutionary "fragmentation" of chromosomes is hence not the simple,
single-step process it has generally been supposed to be, but results, like
fusion, from reciprocal translocations of a special kind. But whereas a
"fusion" involves the loss of small chromosome sections containing a
centromere and two telomeres, a "fragmentation" leads to hyperploidy
of small sections or (if the donor is a supernumerary) to the addition
of some extra genetic material to the regular karyotype.
Not all chromosomes are suitable for functioning as donors. A potential donor must as a rule be small, and it must not lead to inviability
when trisomic or tetrasomic—in fact, the fragmentation is unlikely to
establish itself unless trisomy for the donor regions confers a heterotic
advantage. And, in the case of groups with monocentrie chromosomes,
the centromere supplied by the donor must be a fully competent and
efficient one (the centromeres of many supernumeraries seem to be inefficient, so that mitotic nondisjunction occurs frequently). The efficiency of a centromere may, of course, depend in part on the rest of the
chromosome, so that a centromere which is inefficient in one chromosome
may become more efficient when transferred to another chromosome.
The above are rather special requirements, and help us to understand
why chromosome fragmentation is not a frequent process in evolution.
An acrocentric chromosome from which almost all the genetically active
material has been removed by a deletion would seem to be a suitable
type of donor in most instances.
It is now generally accepted that paracentric inversions in Drosophila populations function as mechanisms for perpetuating adaptive
genie polymorphism, although there is not complete unanimity as to
just how these mechanisms operate in natural populations (Dobzhansky, 1951; Dobzhansky and Pavlovsky, 1953; Wallace, 1953; da Cunha
and Dobzhansky, 1954; Epling et ah, 1953; Cain and Sheppard, 1954).
Such inversions are found in wild populations of most but not all species
of
Drosophih.
119
needed, but a donor chromosome is equally necessary. It is assumed
that the donor undergoes breakage and that the two portions "seal" the
freshly broken chromosome ends.
The donor chromosome may be a regular member of the set (such as
a Y) or it may be a supernumerary element. In the former case we
assume that the unbroken donor persists in the species together with
the translocated portions, i.e., that the "fragmentation" results in hyperploidy for the genetic material contained in the donor. If the unbroken
donor does not persist, we are dealing with a simple translocation which
does not involve an increase in chromosome number. Where the donor
is a supernumerary it can, however, be impermanent.
Evolutionary "fragmentation" of chromosomes is hence not the simple,
single-step process it has generally been supposed to be, but results, like
fusion, from reciprocal translocations of a special kind. But whereas a
"fusion" involves the loss of small chromosome sections containing a
centromere and two telomeres, a "fragmentation" leads to hyperploidy
of small sections or (if the donor is a supernumerary) to the addition
of some extra genetic material to the regular karyotype.
Not all chromosomes are suitable for functioning as donors. A potential donor must as a rule be small, and it must not lead to inviability
when trisomic or tetrasomic—in fact, the fragmentation is unlikely to
establish itself unless trisomy for the donor regions confers a heterotic
advantage. And, in the case of groups with monocentrie chromosomes,
the centromere supplied by the donor must be a fully competent and
efficient one (the centromeres of many supernumeraries seem to be inefficient, so that mitotic nondisjunction occurs frequently). The efficiency of a centromere may, of course, depend in part on the rest of the
chromosome, so that a centromere which is inefficient in one chromosome
may become more efficient when transferred to another chromosome.
The above are rather special requirements, and help us to understand
why chromosome fragmentation is not a frequent process in evolution.
An acrocentric chromosome from which almost all the genetically active
material has been removed by a deletion would seem to be a suitable
type of donor in most instances.
It is now generally accepted that paracentric inversions in Drosophila populations function as mechanisms for perpetuating adaptive
genie polymorphism, although there is not complete unanimity as to
just how these mechanisms operate in natural populations (Dobzhansky, 1951; Dobzhansky and Pavlovsky, 1953; Wallace, 1953; da Cunha
and Dobzhansky, 1954; Epling et ah, 1953; Cain and Sheppard, 1954).
Such inversions are found in wild populations of most but not all species
of
Drosophih.