The problem of mutagen-induced sterility has been especially studied in barley by GAUL, and in Arabidopsis by MUELLER and by VAN DER VEEN. Their results and conclusions are essentially confirmed and extended by the present Arabidopsis experiment, which compares EMS with X-rays.
For a given treatment, M,-fertility is correlated with Ml-fertility, but mutant frequency in M, is equal for the different Ml-fertility classes. It follows that efficiency in mutation breeding can be increased by pre-selection for fertility in M,, and by using doses which do not make fully fertile Ml-sectors too rare, EMS is more efficient than X-rays by a factor 4 at equal Ml-fertility, and a factor 3 at equal M,-fertility. As a result of difference in fertility spectrum, fertility improvement from M, to M, is better for X-rays than for EMS.
A considerable load of diplontic sterility and of embryonic lethals (non-germinating seeds) is simply the toll one has to pay for the higher mutant frequency after EMS.
Fortunately, part of the diplontic sterility can be pre-selected against in M,, and maternally conditioned embryonic lethals can be selected against by discarding seedlots from individual Ml-sectors which show reduced germination. (M,-line selection may also be practised against recessive embryonic lethals and seedling lethals.