The Case for Keeping Cornus in the Broad Linnaean Sense

Abstract

The concept we call Cornus sensu lato was milleniums old when Carl Linnaeus wrote Species plantarum. Following tradition, he viewed Cornus as embracing the cornelian cherry and the bractless dogwoods; following Dillenius and Plukenet, he added dwarf cornels and flowering dogwood. Today, in eastern Europe and in Asia, steadfast splitters view all those as genera-Cornus sensu stricto, Swida, Chamaepericlymenum, Cynoxylon-and make four more from dogwoods that Linnaeus did not know-Afrocrania, Bothrocaryum, Dendrobenthamia, Discocrania. No one's thoughts on evolution mandate splitting-all agree the segregates are closer to each other than to any group outside of Cornus-and there are the following reasons for retaining Cornus in its broad sense: some splits were made on weak grounds (Chamaepericlymenum) or on arguments now overweighed by other evidence (Dendrobenthamia crosses with Cynoxylon). In England, Japan, and North America, experiments with splitting won few converts. Splitting upsets nomenclature, making it depend on how one views the names of Rafinesque or where one sees divergence of the dogwood subgroups. Indeed, splitting hinders thinking on divergence (those who split off Afrocrania, for instance, lose sight of its kinship with cornelian cherry), and misaligning subgroups within Cornus hinders circumscription of Cornaceae, a most unruly family.