53 - Polygonaceae

Summary

Herbs, shrubs or climbers, rarely trees. Leaves usually alternate and usually with sheathing stipules (ochreae). Flowers bisexual or unisexual. Perianth segments 3–6, sepaloid or petaloid, free or connate, persistent, imbricate in bud. Stamens usually 6–9; anthers 2-celled, opening lengthways. Ovary superior, syncarpous, unilocular, with a solitary basal orthotropous ovule. Fruits indehiscent, hard, trigonous or lenticular, usually enveloped in the perianth.

Contains about 40 or more genera and 800 species throughout the world, but mainly in temperate regions.

There has been much dispute over how many genera there are in the Polygonaceae. Our examination of the details of the flowers has shown that there is so much variation in size and shape, even on the same plant, that they cannot be used to any extent in this classification. On the other hand an examination on an almost daily basis over the last few years of many species in Cambridge Botanic Garden has shown that the type of inflorescence and general habit divides them into a number of natural groups. Either the whole lot except Rumex and related genera can be put in one genus, Polygonum, or each of these groups can be regarded as a distinct genus. Any intermediate position, which is what has mostly been adopted, brings about problems with intermediates and the use of keys. Believing that the general tendency is to split, we have followed the latter course. Galasso et al. (2009) agree with some of our decisions.

In the case of species the same principles have been applied and we have split rather than lumped. Selfpollination appears to be widespread, but when hybrids do occur they are often perennials and spread vegetatively. In the genus Polygonum sensu stricto we have followed the principles of Alexis Jordan, who grew these plants for many years; those that we have grown support his findings. Arthur Chater has collected these plants extensively in Cardiganshire and has found almost exactly the same number of taxa as we have found around Cambridge. Rumex is completely different and all taxa appear to hybridise with one another, except those that we have removed to Acetosella and Acetosa.