Theoretically it can not be more than 50% because of the assumption which is
made. It is assumed there is no interference among crossovers. If this
assumption holds then "even" numbers of crossovers will be equal to no
crossovers, hence, reducing the total number of observed recombinants.
BUT, biologically, it may not be true. I give you an example. Consider a
region of a chromosome in which one crossover happens every time a meiosis
takes place. Also assume this crossover prevents occurance of any other
crossover at that region. Now if we know the linkage phase of an individual,
we will see that all gametes are recombinants, giving a recombination rate
of 100%:
Linkage phase: AB/ab
Gametic frequencies: AB=0 ab=0 Ab=50 aB=50
If we did not know the linkage phase given above, one would conclude that the
actual phase is Ab/aB with zero recombination rate.
I do not have any observed data to support this example but it is possible
to happen biologically. Therefore, biologically, we can not restrict the
recombination rate to under 50%. Even if the interference is not that strong,
it may still lead to recombination rates of more that 50%.
Another case in which recombination rate of more that 50% may be observed is
due to sample size (the coin example given in this thread).
Ardeshir Nejati