Today, we have the third in the series on Hamamelidaceae guest-written and photographed by Martin Deasy. Martin writes:
The sight of Hamamelis virginiana with its tangle of bare twigs covered in yellow blossom is one of the most striking sights of the late autumn and early winter. Widely distributed in the deciduous forests of eastern North America, it forms a small, spreading tree with small branches. It is well known as the source of the witch hazel extract widely used as an astringent, obtained from a decoction of the stems.
Assigned its own tribe (Hamamelideae) within subfamily Hamamelidoideae, the genus Hamamelis is characterized by strictly 4-merous, hermaphrodite flowers, with long, ribbon-like petals that are circinate (rolled like a fire-hose) in bud. A whorl of 4 staminodes secrete a nectar reward for pollinators. The anthers have only one sporangium per theca, and each theca opens outwards by a single valve, rather as if the anther were releasing its pollen through a pair of car doors.
Unlike other Hamamelis species, which flower from late winter into early spring, Hamamelis virginiana flowers in the autumn, from October until Christmas (in the northern hemisphere), and is pollinated by insects (mainly small flies). Following pollination, the pollen tube penetrates downwards towards the base of the carpel, at which stage it ceases development and overwinters before growth is recommenced–and the ovule fertilized–in late spring. There is thus a delay between pollination and fertilization of up to 7 months.
A possible explanation for Hamamelis virginiana‘s eccentric phenology (flowering period) has been suggested by observations from the Ozark mountains, where the species overlaps in range with the later flowering Hamamelis vernalis. In unusual years in which both taxa flowered simultaneously, it was observed that, given the choice, insect pollinators strongly favoured Hamamelis vernalis. This raises the possibility that the displacement of its flowering period into the late autumn might represent an adaptive strategy allowing Hamamelis virginiana to avoid having to compete for pollinators with its more appealing relation. Even then, relying on insect pollinators during the coldest parts of the year proves very inefficient , one study finding the rate of fruit set to be less than 1% (ref: Anderson and Hill 2002).
The issue of the numerous disjunct distributions characteristic of the Hamamelidaceae has already been partly treated in yesterday’s post on Trichocladus crinitus. Hamamelis offers another particularly good example. Of the five Hamamelis species, three are from eastern North America, while two are east Asian–the classic “Tertiary Relict” disjunct distribution that has fascinated and teased botanists since it was first noticed in the 18th century. As the climate cooled in the Oligocene (ca. 35mya-23mya), the high-latitude Tertiary flora seems gradually to have been forced southwards, ultimately into its present-day refuges of either eastern Asia or southeastern North America.
The original geographical area occupied by the northern Tertiary flora included territory now submerged beneath the Pacific, Atlantic and Arctic oceans–the so-called “land bridges” (though they should not be thought of as mere bridges: they constituted fairly permanent land in their own right). Asia and North America have historically been connected by a substantial landmass known as Beringia, and it seems plausible that the extant Hamamelis species may have differentiated from an ancestral population present in or around Beringia. In this context, it is striking that the Japanese hamamelis (Hamamelis japonica, from the islands of southern Japan) is more closely related to the American species than to the Chinese Hamamelis mollis.