植生史研究 Jpn. J. Histor. Bot. Vol. 19 Nos. 1–2 p. April 2011 ©2011 Japanese Association of Historical Botany Introduction The identification of the living representatives of the Taxodiaceae (now subsumed within the Cupres- saceae) is based on a whole-plant concept, utilizing an approach that recognizes significant morphologi- cal features such as the shape, architecture, and age of the tree; leaf arrangement on the leading and lateral branches; leaf morphology; and the size and shape of the seed cones (Silba, 1986; Rushforth, 1987; Wu & Raven, 1999; Farjon, 2005). Regardless of the fea- tures used, perhaps the single most useful feature that facilitates species identification, description, and clas- sification among the conifers is seed cone morphology (Matsumoto et al., 1997; Aulenback & LePage, 1998; LePage, 2001, 2003). Whereas leaf morphology alone can sometimes be valuable, the identification and clas- sification of isolated leaves that are preserved as com- pression fossils is generally difficult in taxa where leaf polymorphism is prevalent. Among the genera of the Taxodiaceae, leaf polymorphism is present in Glyp- tostrobus Endlicher, Metasequoia Hu et Cheng, Se- quoiadendron Buchholz, Sequoia Endlicher, Taiwania Hayata, and Taxodium Richard (Brown, 1936; Bůžek, 1971; Christophel, 1976; Farjon, 2005). Although this is not necessarily problematic when working with living representatives, variation in leaf morphology together with preservation quality further contribute to the difficulty of correctly identifying fos- sil leaves that are not associated with reproductive or- gans. Such difficulties are further amplified when sever- al taxa that are known to produce polymorphic leaves are preserved as part of the same floristic assemblage. As a result, the paleobotanical literature is rife with morphospecies names that are based on isolated leaves that were incorrectly identified (Heer, 1868–1878; Chaney, 1951; LePage et al., 2005; LePage, 2007). This however does not mean that misidentified specimens have no taxonomic and phylogenetic value. On the contrary, the reassessment of fossil leaves and their ex- tant counterparts may often reveal morphological and anatomical features that were previously overlooked as being diagnostic (Matsumoto et al., 1994, 1995; Stockey & Ko, 1986; Stockey & Atkinson, 1993). The problem of correctly identifying fossil taxodia- ceous leaves is not unique to any generation of paleo- botanists and the uncertainty of identifying isolated leaves extends to the earliest plant classification sys- tems. Brongniart established the genus Taxites Brong- niart for fossil leaves that were “analogous in part to the yews and Podocarpus, while in other cases they could not be separated from the yews or Taxodium” (Brongniart, 1828, p. 101), while Presl created the ge- nus Taxodites Presl for “distichous twigs and leaves strongly analogous to those of Taxodium” (Presl in Sternberg, 1838, p. 204). Chaney (1951, p. 174) also recognized this problem and stated that “As far back as the middle of the past century, fossil foliage of Sequoia has been confused with that of Taxodium. During all 109–116 Ben A. LePage 1 : A classification system to separate leaves of the Cephalotaxaceae, Taxaceae, and Taxodiaceae Abstract The identification and classification of isolated fossil leaves of the Taxodiaceae is generally difficult when the remains are not associated with reproductive organs. These problems are further exacerbated when more than one genus of the Taxodiaceae and/or genera of the Taxaceae or Cephalotaxaceae are preserved in the same deposits or the quality or mode of preservation precludes confident identification. To help facilitate the identification and separation of isolated taxodiaceous fossil leaves, the leaves of extant representatives of Cepha- lotaxus Siebold et Zuccarini ex Endlicher, Glyptostrobus Endlicher, Metasequoia Hu et Cheng, Sequoia Endli- cher, Taxodium Richard, Taxus L., and Torreya Arnott were examined. The results of this investigation provided sufficient data about external morphological features to establish a system of classification for leaves that could be used to assist in the identification and classification of isolated fossil leaves. 1 Academy of Natural Sciences, 1900 Benjamin Franklin Parkway, Philadelphia, Pennsylvania, 19103 USA and PECO Energy Company, 2301 Market Street, S7-2, Philadelphia, Pennsylvania, 19103 USA (e-mail: ben.lepage@exeloncorp.com)