It has been suggested that ratios of coupling to repulsion-phase linked mar
kers can be used to distinguish between allopolyploids and autopolyploids,
because repulsion-phase linkages are much more difficult to detect in autop
olyploids with polysomic inheritance than allopolyploids with disomic inher
itance. In this report, we analyze the segregation pattern of repulsion-pha
se linked markers in polyploids without complete preferential pairing. The
observed repulsion-phase recombination fraction (R) in such polyploids is c
omposed of a fraction due to crossing-over (R-c) and another fraction due t
o independent assortment (R-i). R-i is the minimum distance that can be det
ected between repulsion-phase linked markers. Because Ri is high in autopol
yploids (0.3373, 0.4000, 0.4286 and 0.4444) for autopolyploids of 2n=4x. 6x
. 8x and 10x), large population sizes are required to reliably detect repul
sion linkages. In addition, the default linkage used in mapping-programs mu
st be greater than the corresponding R-i to determine whether a polyploid i
s a true autopolyploid. Unfortunately, much lower default linkages than the
R(i)s have been used in recent polyploid studies to determine polyploid ty
pe, and markers have been incorporated into polyploid maps based on the R v
alues. Herein. we describe how mapping repulsion linkages can result in spu
rious results. and present methods to accurately detect the degree of prefe
rential pairing in polyploids using repulsion linkage analysis.