To the uninitiated, chromosome number may appear to reflect genome size — more chromosomes would mean a larger genome. This is not necessarily the case if we measure genome size by the number of base pairs in a genome. There are two primary ways to change chromosome number: chromosomal duplications and chromosomal fusions/fissions. Chromosomal duplications (either through polyploidization or aneuploidy) do change the size of the genome, either by creating a second copy of a single chromosome or duplicating an entire genome. Fusions and fissions, on the other hand, merely rearrange the content of the genome — either decreasing chromosome number by combining two chromosomes into one or increasing chromosome number by dividing one chromosome into two — but the actual size of the genome does not change.
A lot of people do recognize the difference between fusions and chromosomal duplications. But they struggle to explain how fusions could persist in population or fix between them, inventing elaborate explanations for chromosomal rearrangements. The struggle to understand the fitness consequences of chromosomal rearrangements is exemplified in this email over the Bionet listserve:
The Robertsonsian fusion that formed human chromosome number two (from ancestral 2A and 2B, as it is preserved in the other apes) should have caused a serious reproductive barrier, overcome only by consanguity of the highest order; mating amongst first-degree relatives. Any breeding outside the immediate family would have lead to unacceptable chromosomal imbalances.
The emailer goes on to say that this is an important unresolved issue in human genome evolution. Was there dramatic inbreeding during early human evolution to prevent heterokaryotypic individuals (those that are heterozygous for the two different chromosomal arrangements) from arising? The fitness consequences of heterokaryotypic individuals are addressed in this reply:
Floating polymorphism for Robertsonian fusions is quite common, and in many cases apparently causes little or no reductiuon in fertility (examples from rodents, spiders, grasshoppers, bovids, cockroaches….). There is no clear pattern as to when fusions will result in nondisjunction vs when they won’t, but fusion trivalents are more likely to be stable when the arms of the fusion are of similar length (there is a difference in arm length on our chromosome 2, but it isn’t extreme) and crossovers are relatively distal (don’t know if that is the case here) – the 14/21 fusion in humans is an example of very unequal arm lengths and it often missegregates, although people with this fusion still produce offspring and most are normal (5-10% of offspring are tri 21, presumably many more lost through spontaneous abortion, in addition to the other nondisjunction products). So fixation of the chromosome 2 fusion is unlikely to have required the very extreme circumstances you have outlined.
The big concern here is how chromosomal fusions and fissions affect meiosis. If they interfere with proper segregation of chromosomes, then heterokaryotypic individuals will suffer a fitness cost because they won’t be able to produce viable gametes. But fusions, on their own, don’t make proper segregation impossible, as shown below:
While these rearrangements may lead to problems during meiotic segregation, they do not make meiosis impossible. But one often notices chromosomal rearrangements that differentiate two species (ie, the fusion that occurred after humans diverged from chimpanzees). One might think that selection against heterokaryotypic individuals may lead to the reproductive isolation between karyotypes. But it’s not the rearrangements themselves that lead to reproductive isolation; it’s the genes that are contained within the rearrangements. For example, inversions prevent genetic exchange (recombination) between karyotypes which allows for mutations that cause reproductive isolation between karyotypes to accumulate within those inverted regions.