HAMAMELIDAE

subclass of woody or herbaceous dicotyledonous flowering plants. The subclass Hamamelidae consists of 11 orders, with most species belonging to the orders Urticales and Fagales. The subclass as a whole consists of 24 families and about 3,400 species. The flowers of plants belonging to the subclass are generally wind-pollinated and mostly small, with a reduced or absent perianth and small sepals when present at all. The stamens are generally not laminar but number at least two or in some cases many. The carpels form a compound ovary. The smaller orders in the subclass are Trochodendrales, Daphniphyllales, Didymelales, Eucommiales, Hamamelidales, Leitneriales, Juglandales, Myricales, and Casuarinales. subclass of woody or herbaceous plants, the smallest of the six subclasses of the class Magnoliopsida. Among its members are the witch hazel family (Hamamelidaceae), the hemp family (Cannabaceae), the nettle family (Urticaceae), the walnut family (Juglandaceae), the bayberry family (Myricaceae), the beech family (Fagaceae), and the birch family (Betulaceae). Because of the presence of catkins, or aments, many members of the Hamamelidae, plus a number of unrelated families, were previously classified in an artificial group called the Amentiferae, or Amentaceae. The chief features of the members of the Amentiferae were staminate flowers, and frequently also pistillate flowers, in catkins, the sepals and petals reduced or absent, and a general trend toward wind pollination (anemophily). The taxa included in the Amentiferae varied with individual usage but often included the Piperaceae (Magnoliidae); the Saururaceae, Myricaceae, Juglandaceae, and Leitneriaceae (Hamamelidae); the Salicaceae (Dilleniidae); the Garryaceae, and Julianiaceae (Rosidae); and other groups. The amentiferous syndrome exhibited by these disparate families is now considered to have arisen through convergent evolution and thus does not indicate close relationships between the families. The members of Hamamelidae are woody or herbaceous, usually with simple leaves, less often with the leaves pinnately or palmately compound. The stamens usually consist of well-defined filaments and anthers and are not thin and flat (laminar) as in some members of the Magnoliidae, the subclass containing the greatest number of species with features considered to be primitive within the angiosperms. The Hamamelidae produce tannins (proanthocyanins, ellagic acids, and gallic acids), presumably as a defense mechanism against the attack of insects and other pests, and rarely alkaloids. The flowers in the Hamamelidae are frequently very small, often unisexual and with the perianth reduced or absent, and are often borne in catkins. Wind pollination is common in the group. The seeds are never numerous and are not borne on parietal placentas. There is currently no consensus on the composition and classification of the subclass Hamamelidae. Based on various kinds of studies, especially on leaf architecture for example, some researchers are even of the opinion that the group has more than one origin (polyphyletic). Arthur Cronquist and Robert Thorne, two American botanists who have devoted much of their careers to interpreting the evolutionary history of the angiosperms, have developed systems of classification that take into account all available biological evidence. They disagree, however, in major ways on the contents and classification of the Hamamelidae. In Cronquist's scheme (which is followed in this article), which is basically the same as the system proposed by the Armenian botanist Armen Takhtajan, the Hamamelidae contain 11 orders (Casuarinales, Daphniphyllales, Didymelales, Eucommiales, Fagales, Hamamelidales, Juglandales, Leitneriales, Myricales, Trochodendrales, and Urticales). Within the Hamamelidae, Cronquist's Hamamelidales is basal to all other orders except the Trochodendrales, which appears to be a dead-end side branch that arose directly from the line leading from the ancestors of the subclass Hamamelidae to the present-day order Hamamelidales. The approximately 3,400 species in about 26 families in the Hamamelidae are distributed unequally among 11 orders. The Urticales with roughly 2,200 to 2,300 species and the Fagales with about 1,100 species are by far the 2 largest orders. Within the remaining 9 orders are many small, isolated families that have only 1 or 2 species. Among these are the Tetracentraceae, Eucommiaceae, Cercidiphyllaceae, and Leitneriaceae, all thought to be primitive and suggesting links to the most primitive group of flowering plants, the Magnoliidae. The Hamamelidae are thought by some to have arisen along with the Rosidae from among magnoliid ancestors during the later part of the Early Cretaceous (about 100 million years ago) and to have branched off and become fully distinct during the earliest part of the Late Cretaceous (97.5 to 66.4 million years ago). Thorne's classification of the Hamamelidae differs considerably from the system proposed by Cronquist, not so much in that it treats the entire Angiospermae as a single class with two subclasses, the Dicotyledoneae and Monocotyledoneae, but in the placement and contents of the orders included in the Hamamelidae. At the level of classification equivalent to Cronquist's subclass, Thorne uses the category superorder, and it is within the superorder Rosanae that he places the orders Casuarinales, Fagales, Hamamelidales (including the Eucommiales and Trochodendrales), Pittosporales (including the Daphniphyllales and Didymelales), and Rosales. The Juglandales are placed in the superorder Juglandanae and the Urticales in the superorder Malvanae. A third system, proposed by the Swedish botanist Rolf Dahlgren, is similar to Thorne's in postulating close relationships between the Urticales, Malvales, and Euphorbiales. Proponents of the systems of Cronquist and Thorne are about evenly divided. Making it even more difficult to arrive at a clear circumscription of the Hamamelidae is the fact that reasonable evidence, seemingly proving or disproving one point of view or the other, can be presented by either side. Until conclusive evidence becomes available to support the recognition of a single system of classification, it seems best to allow that each has merits. David E. Boufford Additional reading A good, all-inclusive work with many important references is Peter R. Crane and Stephen Blackmore (eds.), Evolution, Systematics, and Fossil History of the Hamamelidae, 2 vol. (1989), containing a collection of papers by specialists, including one by Robert F. Thorne, Hamamelididae': A Commentary, vol. 1, pp. 916, an alternative to Cronquist's system of classification, less detailed, slightly more technical, and less descriptive. Useful journal articles include B.H. Tiffney, Fruit and Seed Dispersal and the Evolution of the Hamamelidae, Annals of the Missouri Botanical Garden, 73:394416 (1986); and Robert F. Thorne, The Amentiferae' or Hamamelidae as an Artificial Group: A Summary Statement, Brittonia, 25(4):395403 (1973). Also of interest is Peter K. Endress, Evolutionary Trends in the Hamamelidales-Fagales Group, in Plant Systematics and Evolution, Supplementum, vol. 1 (1977), pp. 321347, a very good account of similarities, with many clarifying illustrations.