In plants, meristems are responsible for generating new organs while also maintaining totipotency (Wolff, 1759). The organization of the shoot apical meristem and its cytohistochemical zonation is directly related to the growth and development of all lateral organs in vascular plants. Surprisingly, all plant meristems were once thought to be composed of a single apical cell (Nägeli, 1845, 1846), but subsequent cytohistochemical studies demonstrated that seed plant meristems are multicellular with zonation (Popham, 1951; Gifford & Corson, 1971). This concept of the seed plant meristem has permeated the field and has been integral for unraveling the molecular genetics of its function (Gaillochet et al., 2015). Cytohistochemical analyses of fern meristems have also shown that the fern meristem is multicellular with zonation (McAlpin & White, 1974; Stevenson, 1976b, 1978; Gifford, 1983; White & Turner, 1995). However, the classical idea that the fern meristem is composed of a single or a few apical cell initials remains a prevailing view in books and developmental genetic studies (Harrison et al., 2005; Sano et al., 2005; Tomescu, 2010; Schneider, 2013; Banks, 2015; Frank et al., 2015; Plant Ontology Consortium (POC), 2015; Fig. 1a). The propagation of this unicellular meristem model can be best illustrated by examining different scientific sources from the last 10 years. Unlike older references that focused on the anatomy of fern meristems (McAlpin & White, 1974; Bierhorst, 1977; Gifford, 1983), these more recent references do not go into careful detail in their definition of the meristem. However, many of the citations that do include a definition simply refer to the large apical initial or initials and do not discuss the fern meristem as multicellular with zonation. For instance, in defining apical cells the Plant Ontology Consortium – which is a current effort to produce a controlled vocabulary describing plant anatomy, morphology and stages of development for all plants – clearly embraces the idea of a fern meristem consisting of a single apical initial: 'Apical cells occur only at the tip of a shoot axis apex, leaf apex, root apex, thallus apex, or protonema in bryophytes and some pteridophytes. Apical growth results from division of a single meristematic cell located at the tip of an apical meristem or plant organ, rather than from a population of meristematic cells located at the tip of an apical meristem.' (Plant Ontology Consortium (POC), 2015). Furthermore, in a paper investigating leaf development, Harrison et al. (2005) when discussing the meristem of the lycophyte Selaginella kraussiana, the only cells that they discussed are the apical cells: 'a strip of large cells on the surface of the apex' (Harrison et al., 2005, p. 511). Then, talking about the fern Osmunda cinnamomea, they state, 'As in Selaginella, the meristem has distinct apical cells on the surface, and bifurcates during branching.' (Harrison et al., 2005, p. 512). Similarly in another paper about the function of Class I KNOX genes in ferns, Sano et al. (2005) explain that: 'The shoot apical meristem of Ceratopteris sporophyte contains a single tetrahedral apical cell.' Moreover, in reviews on the morphological features of ferns, authors again evoke the classical view by stating that: 'The growing tips of seed-free vascular plants are characterized by unstratified apical meristems in which one or several apical initials, present at the surface of the meristem, are responsible for tissue generation. This plesiomorphic vascular plant character is contrasting with the stratified nature of most seed plant apical meristems' (Tomescu, 2010, pp. 74–75) and 'Ferns show rather simple organized apical meristems' (White & Turner, 1995; Imaichi 2008); and, finally, 'Usually, they contain a single apical cell. In a few taxa, such as Marattiaceae and Osmundaceae, the larger size of the apical meristem results in the formation of an apical cell group instead of a single one. At the sporophyte level, three kinds of apical meristems can be differentiated. The shoot apical meristem (SAM) has usually a prominent apical cell of a tetrahedral shape in most ferns …' (Schneider, 2013, p. 123). However, a correctly defined fern meristem as multicellular is integral for interpreting broad comparative developmental genetic studies. An example of the definite need to recognize the fern apical meristem as a multicellular zoned structure is illustrated by a recent publication by Frank et al. (2015). In this paper, the authors use transcriptomic sequencing of the shoot apex to compare the molecular signatures of anatomically distinct zones of a lycophyte and a fern. Although one of its discussion sections is labeled 'The core domain is transcriptionally similar to the angiosperm peripheral zone' (Frank et al., 2015, p. 902), the authors never use their results to discuss the definition of the meristem that is made in their introduction: 'the SAMs of many seedless plants typically house a single, pyramidal, initial called the apical cell (AC)' (Frank et al., 2015, p. 893). Furthermore, the Banks (2015) Commentary that highlighted Frank et al.'s (2015) paper reiterates that the fern meristem is composed of apical initials only; 'Ferns (monilophytes) and some lycophytes, including Selaginella moellendorffii, have apical meristems that are unusual compared to the apical meristems of angiosperms. Fern and Selaginella apical meristems have a single distinct apical, pie-shaped cell whereas the angiosperm meristem is multicellular with distinct zones and layers and no apical cell …' (Banks, 2015, p. 486). All meristems must balance distinct cellular processes: the maintenance of totipotent cells that rarely divide, and the rapidly dividing cells that will later differentiate to build the plant body. Anatomical and histochemical analyses have been used to characterize the multicellular seed plant meristem. For example, angiosperm meristems are considered to be composed of a central zone of quiescent totipotent cells, and of a peripheral and a rib zone both with highly dividing cells (Gifford & Corson, 1971; Steeves & Sussex, 1989). Developmental genetic analyses have uncovered several mechanisms for the balance of cellular processes that must occur in a multicellular meristem (Gaillochet et al., 2015). For example, the noncell autonomous action of SHOOTMERISTEMLESS (STM) is necessary to maintain shoot apical meristem function in Arabidopsis thaliana (Kim et al., 2003, 2005). Moreover, proper communication between cells with different mitotic activities is necessary; if there is an imbalance in the cellular processes then either the quiescent totipotent cells are not maintained and the shoot terminates, or there is an overproliferation of cells and additional lateral organs are generated (Gaillochet et al., 2015). Cytohistochemical analyses have indicated that fern meristems are also multicellular with apical initial(s) that rarely divide, subapical initials and a cup-shaped region of rapidly dividing cells (McAlpin & White, 1974; Stevenson, 1976a, 1978; Bierhorst, 1977; White & Turner, 1995; Fig. 1b). Similar results were obtained from experimental analyses that investigated the integration of the meristem in angiosperms and ferns. When the meristem of either Impatiens (angiosperm) or Dryopteris (fern) was punctured, more than two apices were formed (Wardlaw, 1949; Steeves & Sussex, 1989). These results suggest that both fern and seed plant meristems are multicellular and there is communication among the cells of the meristem. Accordingly, a wider knowledge of the fern meristem as multicellular is integral not only for investigations of meristems, but also for understanding the evolution of the lateral organs they produce. In order to provide molecular evidence for the structure and function of the fern meristem, we investigated the expression of Class I KNOX orthologs in the leptosporangiate fern Elaphoglossum peltatum f. peltatum (Vasco et al., 2013). Class I KNOX genes, such as STM, are expressed throughout the shoot apical meristem in angiosperms and are necessary for meristem maintenance (Endrizzi et al., 1996; Long et al., 1996). Here we show that the meristem maintenance Class I KNOX orthologs are expressed throughout the shoot apical dome of a leptosporangiate fern. These expression results clearly show that the fern meristem is a multicellular meristem composed of an apical initial and its surrounding cells that are cytoplasmically dense and actively dividing. Class I KNOX sequences of Elaphoglossum peltatum (Sw.) Urb. f. peltatum were isolated by PCR with degenerate primers (01KNOXf5′CCBGARCTBGACMABTTYATGG and 02KNOXr5′CCAGTGSCKYTTCCKYTGRTTDATRAACC and by 5′RACE (Clontech Laboratories Inc., Mountain View, CA, USA) according to the manufacturer's protocol. For the phylogenetic analyses Class I KNOX sequences for all major groups of land plants were obtained from GenBank. A list of all sampled species is provided in Supporting Information Table S1. Bayesian phylogenetic analyses were performed as in Floyd & Bowman (2007) using unambiguously aligned nucleotides of the homeodomain with MrBayes v.3.2.6 (Ronquist & Huelsenbeck, 2003) on CIPRES (http://www.phylo.org; Miller et al., 2010). (Alignment available in Fig. S1.) For the histological analyses, Safranin O and Astra Blue staining of tissue sections was performed. For the in situ hybridization, tissues were fixed in formaldehyde acetic acid for 2–4 h and then dehydrated through a graded ethanol series to 100% ethanol. Tissue was embedded in paraplast x-tra (Fisher brand) and sectioned as previously described (Ambrose et al., 2000). Gene-specific fragments of EppC1KNOX1 and EppC1KNOX2 were amplified using primers designed for this study. Digoxigenin (DIG)-labeled gene-specific probes were generated according to the manufacturer's instructions (Roche Applied Science, Indianapolis, IN, USA). Five independent experiments each for EppC1KNOX1 and EppC1KNOX2 expression were performed. Hybridizations were performed as in Ambrose et al. (2000). We cloned two KNOX genes from E. peltatum and demonstrated through phylogenetic analyses that they are orthologous to seed plant Class I KNOX genes (Fig. 1c). We investigated the expression of both E. peltatum Class I KNOX by in situ hybridization. We found that both E. peltatum Class I KNOX orthologs were expressed throughout the entire apical dome of the vegetative shoot similar to STM from seed plants (Fig. 1d,e). In 40% of our experiments, expression of either E. peltatum Class I KNOX ortholog was not detected in the apical initial but was detected in the surrounding cells. These molecular results provide further support that the fern meristem includes apical initials and surrounding cells of the shoot apex, as previously indicated based on anatomical analyses (McAlpin & White, 1974; Stevenson, 1976a; White & Turner, 1995; Fig. 1b). Although expression analyses of Class I KNOX orthologs have previously been studied in the ferns Osmunda regalis (Harrison et al., 2005), Anogramma chaerophylla (Bharathan et al., 2002), and Ceratopteris richardii (Sano et al., 2005), these were focused on other developmental processes and therefore did not concentrate on expression in the fern meristem. In lycophytes, the Class I KNOX ortholog in Selaginella kraussiana has been shown to be expressed throughout the meristem (Harrison et al., 2005) similar to our results here for ferns. SkKNOX1 expression throughout the apical dome supports previous anatomical studies indicating that lycophyte meristems are multicellular (Freeberg & Wetmore, 1967; Ogura, 1972; Stevenson, 1976b). Angiosperm Class I KNOX proteins have been shown to act noncell-autonomously and to move through the plasmodesmata (Kim et al., 2003). It has long been known that plasmodesmata formation differs amongst vascular plants (Imaichi & Hiratsuka, 2007). In ferns and lycophytes with prominent apical initials plasmodesmata are formed based on cell lineage, while in seed plants plasmodesmata are interface-specific (Imaichi & Hiratsuka, 2007). Differences in plasmodesmata function may hold the key for understanding the structure of meristems, and the relative contribution of the prominent apical initials and derivatives to the integrity and physiological activity of the multicellular fern meristems. The regions of the multicellular fern meristem have been variously termed; apical initials, pro-meristem, cup-shaped zone, central zone, subapical initials, surface initials, surface layer, subsurface initials and peripheral zone (Ogura, 1972; McAlpin & White, 1974; Stevenson, 1976a, 1978; White & Turner, 1995). Based on cellular activities and Class I KNOX expression we propose to simply name the regions of the fern multicellular meristems as composed of apical initial(s) that rarely divide and a peripheral zone composed of cytoplasmically dense, rapidly dividing cells (Fig. 1f). This definition is supported by recent Next-generation results, which indicated that similar genes were expressed in the peripheral zone of fern meristems (Fig. 1f) as in the peripheral zone of seed plants (Frank et al., 2015). There have been several recent advances in knockout technologies in ferns (Muthukumar et al., 2013; Plackett et al., 2014; Bui et al., 2015) and transcriptome analyses (Frank et al., 2015). The firm recognition of the fern and lycophyte meristems as multicellular with cytohistochemical zonation is integral for the interpretation of broad comparative developmental analyses. We apologize to all authors whose work we could not cite due to space constraints. We thank Tynisha Smalls for her excellent technical assistance, Dennis Stevenson for lively discussions of this work, and Robbin C. Moran, Natalia Pabón-Mora, Favio Gonzales, David Barrington and Michael Sundue for helpful comments. We thank Charles Alford (http://www.rareferns.com) for the E. peltatum f. peltatum plants and staff of NYBG Nolen glasshouses for propagating them. B.A.A. acknowledges funding from the National Science Foundation (DEB-1020443). A.V. acknowledges partial funding from Consejo Nacional de Ciencia y Tecnología (CONACyT). B.A.A. and A.V. conceived the study, performed the experiments, analyzed the data and prepared the manuscript. Please note: Wiley Blackwell are not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office. Fig. S1 Aligned matrix of Class I KNOX sequences. Table S1 Class I KNOX sequences used in this study Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.