New Species and Notes on the Natural History of Markea (Solanaceae) from Colombia and Ecuador Sandra KnappDepartment of Botany, The Natural History Museum, Cromwell Road, London SW7 5BD, U.K.ABSTRACT. Three new species of Markea, M. an-tioquiensis, M. epifita, and M. pilosa, are described,having been discovered as a result of revision ofmaterial for the Checklist of the Vascular Plants ofEcuador. The relationships of these new taxa to thepreviously described Colombian endemic Markealopezii Hunziker and the recently described M.spruceana Hunziker, and the status of the genusMarkea s. str., are discussed. An amplified descrip-tion is provided for M. spruceana with notes on itsnatural history. The three new taxa are illustrated,and a distribution map with all four taxa is provid-ed. In the course of preparing the treatment of theSolanaceae for the Checklist of the Vascular Plantsof Ecuador (Jergensen & Le6n Y. (editors), inprep.), several unusual specimens of epiphytic So-lanaceae were encountered. I have noted that thesespecimens were in existence (Knapp et al., 1997;noted without names or citations), but have not un-til now had enough material to permit completespecies descriptions. Recent collecting in the mid-dle elevations of both Andean slopes in Ecuadorhas resulted in more complete material. All of the taxa described here are allied withMarkea lopezii Hunziker of Andean Colombia, shar-ing with it membranous leaves often with trichomeson both surfaces, elongate inflorescences, tubularfunnelform corollas, winged pedicel apices, andlarge, membranous calyx lobes. Markea lopezii wastreated as taxon incertae sedis in Knapp et al.(1997), and its true relationships are not clear. Inthe cladistic analysis in Knapp et al. (1997), it wasthe sister taxon to the genus Solandra, due in largepart to its possession of large, orange flowers andto the large amount of missing data for it (and itsrelative M. aff. lopezii, described recently as M.spruceana Hunziker) in the data matrix. Missingdata can seriously affect the placement of taxa incladistic analyses, so the placement of M. lopeziiand M. spruceana should be treated with caution.Markea lopezii, M. spruceana, plus the species de-scribed here clearly form a natural group, butwhether they in fact represent a new genus alliedNovoN 8: 152-161. 1998.to Markea awaits analysis of further collections ofboth flowering and fruiting material. Current generic limits in Markea are unclear,with several genera having been split from it (e.g.,Hawkesiophyton Hunziker and SchultesianthusHunziker; see Hunziker, 1979), and with the genusbeing defined in its traditional sense on the basisof its strictly basifixed anthers, imbricate or coch-lear imbrication, and often laciniate calyx lobes(Hunziker, 1979), all of which can be variable inthe taxa concerned (see Knapp et al., 1997). Knappet al. (1997) identified a clade with several seed-based synapomorphies consisting of Markea in thetraditional sense plus the segregate genus Hawk-esiophyton. The seeds are not known from M. lopeziior from two of the four species treated here. Hun-ziker (1997), in describing M. spruceana, noted thatfruit and seeds were not known. As I have on loanto me from Ecuadorian herbaria several specimensof this species in fruit, I provide a complete de-scription of this little known taxon here (also seespecies discussion). The fruit and seeds of Markeas. str. (as defined in Knapp et al., 1997) are dis-tinctive: the fruit is a berry with thin, chartaceouspericarp that dries translucent, and the seeds arepale tan or orange with elongate, rectangular testalcells. Pollen of Markea s. str. is quite diverse, withthree main pollen types being present in the genus(Persson et al., 1994). The pollen of M. lopezii withits spiny supratectal processes is completely anom-alous in the tribe Juanulloeae and may be relatedto pollination syndrome (see Ferguson & Pearce,1986). Hunziker's (1997) amplified concept of M.lopezii included taxa with and without what he calls"equinada" (echinate: more correctly the tectumbears spiny supratectal processes; truly echinatepollen is not found in the Solanaceae) exine, a re-sult that needs confirmation and documentation.Preliminary examination of the pollen of two of thenew species described here, M. antioquiensis andM. pilosa, indicates that pollen of M. antioquiensisis superficially similar to the Markea sessilifloratype and that of M. pilosa similar to the Markeacamponoti type (pollen types sensu Persson et al.,
BioStor
