Elsewhere we have noted Brachychiton bidwillii as an attractive and characteristic tree in Bukkulla Conservation Park ( see Cover of Queensland Naturalist 42 4-6 2004). This is a small tree or shrub, often multistemmed from the ground up. It is totally unlike the bottle trees ( B rupestris , B. australe. ) or other Brachychitons that tend to have stout stems.
However when growing plants from Bukkulla seed I noticed a remarkable developmental process. Germination initially and unsurprisingly involved production of a simple radicle (root initial ) for about 40 mm. This then developed a swelling at its junction with the seed. The cotyledons did not deploy as seedling leaves but remained bound in the seed coat. Green formed on the exposed part of the swelling and there was a considerable delay until the plumule (stem initial) emerged. Stem development was then rapid and the first leaves had the form of those in the mature plant. The swollen base was conspicuous at first but became insignificant as the plant developed.
Everything suggests to me that the development pattern is an adaptation for the seedling to store water at the first opportunity, perhaps allowing the seedling to become established over a series of disjunct rainfall events, instead of requiring a major wet period.
The other interesting aspect is its evolutionary implications. Was this feature of seedling development the path through which the bottle trees got their shape ? Did the ancestral bottle trees start off like this ?