Need for Research in Neotropics

While the paleotropics has more genera than the neotropics (61 versus 36), the neotropics is proportionately much richer in species with South America alone having roughly two-thirds of the 3,200 species in the family. Croat's studies in Central and South America show that future priorities for taxonomic research with Araceae are clearly for systematic studies of the large and medium-sized genera in the neotropics, especially in South America, where the new and poorly known species often outnumber those having known names. In many areas and for most genera investigated, large numbers of novelties occur. For example, for Anthurium of Panama, 54% of the taxa were new to science (Croat, 1986a); for Anthurium sect. Pachyneurium 42% of taxa occurring in Central and South America were new (Croat, 1991a), for the revision of Philodendron subg. Philodendron of Central America (Croat, 1997, in press), 62% of taxa were new, as were 40% of Philodendron subg. Pteromischum (Grayum, 1996), and 47% of Dracontium (Zhu, pers. comm.). Thus it seems that more emphasis and manpower and energy must be applied to research with the Araceae of the neotropics.

Our level of knowledge of the systematics of the neotropical Araceae also varies greatly from area to area, due largely to recent revisionary work or to the interest and area focused on by particular workers, e.g., G. S. Bunting in Mexico (Bunting, 1965) and Venezuela, Croat in Panama and Central America (Croat, 1978a, 1983a, 1986a, 1986b, 1988a, 1991a), and Croat & Grayum in Costa Rica. Central America is, in general, less species-rich than South America with species diversity generally increasing as one approaches South America (Croat, 1986a, 1986b). Though some parts of Central America, especially Panama, have shown unprecedented increases in the known aroid flora (Croat, 1985a), it is still much more well known than South America largely not only from the more prolonged effort by aroid taxonomists in the region but also due to the fact that some parts of Central America are much less rich in species per unit area than many parts of South America; most notably the species-rich northwestern region of that continent (Croat, 1992a).

Central America

Most of the earlier work in Central America was undertaken by P. C. Standley in a series of floristic works (Standley, 1927, 1928, 1933, 1937, 1944). Others who contributed to floristic surveys of Central America include W. B. Hemsley (Hemsley, 1885), and for Mexico, Eizi Matuda (Matuda, 1954; Williams, 1981; Espejo & Lopez, 1993), [Veracruz] (Sosa & Gómez-Pompa, 1994).

The larger genera of Araceae in Central America have already been revised. These are Anthurium (Croat, 1983a, 1986a, 1991a) and Philodendron (Croat, in press). Other middle-sized genera have recently been revised or at least have modern revisions. These include: Syngonium (Croat, 1981b), Monstera (Madison, 1977a), and Spathiphyllum (Bunting, 1960a). Revisions of Rhodospatha for the neotropics and Dieffenbachia for Central America are being prepared by Croat, and one for Rhodospatha is nearing completion. However, even in Central America some recently revised genera such as Monstera (Madison, 1977a) are now inadequate. New species, though fewer in number, are also now known for Spathiphyllum, Syngonium, and Anthurium, the latter two having been revised within the last 15 years (Croat, 1981b, 1983a, 1986a, 1991a). Croat is committed to complete the entire family for the Flora of Mesoamerica which will constitute a revision of virtually all the Araceae of Mexico and Central America.

West Indies

The flora of the West Indies is much less species-rich and is in general well known. Other general non-flora papers that deal with West Indian taxa include: Philodendron, (Mayo, 1981), Xanthosoma (Stehlé, 1946), and the ecology and taxonomy of Trinidad Araceae (Simmonds, 1950a, 1950b; Mayo, 1986a). There are still significant taxonomic problems with some species of Anthurium in the Lesser Antilles. One of the most troublesome aspects of taxonomic studies in the West Indies is that the type specimens are either inadequate (need epitypification) or completely lacking all together (need neotypification), due to the fact that this region was among the first areas in the neotropics that was botanized.

South America

While the aroid floras from some parts of South America, especially the Amazon basin, are reasonably well known, the species from the lower slopes on either side of the Andes and especially those species along the western slopes of the Andes in northwestern South America, are very poorly known and 50% or more of their species are new to science (Croat, 1985e, 1992a, 1995b). Some areas for which floristic surveys have been conducted in the past 30 years are relatively well known. This is especially true of Venezuela, where extensive work has been done, especially by G. S. Bunting (Bunting, 1975, 1979, 1986a, 1988, 1988a, 1989) and also by Croat & Lambert (1987). The Venezuelan flora contains 266 species and an additional 25 subspecies or varieties.

The Guiana region is relatively well known at least in part because it is relatively species-poor rather than because of the extent of the collecting efforts. Suriname was, until recent years, the only part of the Guianas that received much attention in regard to Araceae, and largely due to the work of Jonker-Verhoef & Jonker (1953a, 1953b, 1966, 1968a, 1968b) in Suriname. Recently, the whole region is receiving more attention because of work on the flora of the Guianas project and to the Araceae treatment being carried out by Croat. There are an estimated 121 species in that flora. In addition, Bunting (1995) has completed the Araceae treatment for the Flora of the Venezuelan Guyana, the Venezuelan counterpart of the Guianas flora. This flora treats 19 genera (including Urospathella G. S. Bunting considered by some as a synonym of Urospatha) and 177 species.

Another example of a relatively well known area is the state of Bahía in Brazil where Simon Mayo and other members of the Kew Garden staff, especially R. Harley, have made a number of expeditions and are heavily devoted to the floristics of the state (Harley & Mayo, 1980; Mayo, 1984). Mayo has also prepared a revision of the Araceae of Bahía (Mayo, manuscript) and a checklist for all of Brazil. Any reference to the number of species in Brazil for any genus discussed in this paper relies heavily on this unpublished work. Mayo has also worked closely with many Brazilian botanists to encourage their participation in work with Araceae of Brazil (Mayo & Nadruz, 1992; Catharino & Olaio, 1990).

Parts of southern South America are by now also well known and floristic treatments have been prepared for Argentina (Crisci, 1968, 1968a, 1971) and Paraguay (Croat & Mount, 1988). A floristic treatment has also been completed for the state of Santa Catarina (Reitz, 1957).

The Flora of Peru (Macbride, 1936), though falling short of giving an accurate picture of the species count for Peru, does come close to indicating the number of species actually described for Peru, since, except for Anthurium sect. Pachyneurium (Croat, 1991a), few groups have had many species described from Peru since MacBride's publication (Macbride, 1936). A more accurate accounting of the number of species of Araceae in Peru is published in the Catalogue of the Flowering Plants and Gymnosperms of Peru (Croat, 1993). Though not a thorough revision of the species occurring in the country, this list takes into account all species of plants described for Peru as well as all species represented only by herbarium specimens that were able to be verified by experts for each family. The checklist contains 210 species of Araceae for Peru but does not include any unpublished names. Many new species remain to be described.

While there is no completed Araceae treatment for the flora for Ecuador, a recently published checklist for the Amazonian lowlands (Renner et al., 1990) listed 92 species of Araceae (a few of them undescribed) and gave some indication of the species diversity of that part of Ecuador. Unfortunately the Amazonian lowlands represent one of the most species-poor portions of the country if its area is taken into account, due to the widespread nature of the species in that zone. A checklist for the entire flora of Ecuador is being prepared by Peter Jorgenson at the Missouri Botanical Garden. Croat will be responsible for editing the checklist of the Araceae treatment.

The lowland Amazon basin is also relatively well known, principally because of the fact that the species inhabiting the Amazon lowlands are normally wide-ranging and often common species. The vast Amazonian region lying between the Atlantic coast and the foothills of the Andes has moderately few, mostly wide-ranging species. Species diversity increases dramatically as one approaches the foothills of the Andes in the west. Species occurring on the lowermost slopes of the Andes tend to range widely in a north-south direction, often from Colombia to Bolivia and thus tend not to be endemic. However, some of the species of this region are currently believed to be endemic. The degree of endemism increases as elevation rises on the slopes of the Andes and as the terrain becomes more dissected with river valleys (Croat, 1994c).

To the east of the Amazon basin, especially in the Guiana Highlands and in eastern Brazil, from the state of Bahia south almost to Uruguay, the rate of endemism is much higher. Nearly all the species occurring in this region are endemic to eastern Brazil, and few range into the Amazon basin.

Though many species were described from eastern Brazil, by early aroid specialists, including Schott, K. Koch, and Engler, based on the early collecting efforts by botanists such as Glaziou, A. F. Regnell, Riedel, and others, these areas remain poorly known, especially because of the taxonomic complexity in such groups as Anthurium sect. Urospadix Engl., which dominates the area. Perhaps the only group of aroids well known in the region are members of the recently revised Philodendron subg. Meconostigma (Mayo, 1991a).

The truly temperate parts of the continent are devoid of aroids and the subtropical portions of the continent, while containing a number of small, frequently endemic genera in the tribe Spathicarpeae, are also relatively species-poor.

Species diversity is high throughout the South American Andes but especially along the northwestern slope extending from Chocó Province in Colombia and on both the eastern and western slopes in the Andes in the region of the equator and on the eastern slopes of the Andes in Peru. Species diversity is also relatively high in the Cordillera de Merida of western Venezuela but remarkably less so in the northern part of the Eastern Cordillera of Colombia and on the entire western flank of the Eastern Cordillera in Colombia (Croat, 1992a). Species diversity is also relatively low in the Central Cordillera of Colombia. High species diversity is correlated with high precipitation and with the absence of prolonged dry seasons.

Species richness is greatest between sea level and middle elevations to about 1,500 meters. While some species may range to about 3,750 m, diversity drops off dramatically above 2,000 m. Seasonally dry areas, such as the central plateau of eastern Brazil and the lower Amazon basin, are relatively species-poor as are the generally treeless llanos of Venezuela.

Endemism is also especially high in the Andes of western South America, including the eastern range of the Andes that extends into Venezuela. Endemism is also high in the Guiana Highlands and in parts of North America, especially in Mexico and in lower Central America, in Costa Rica, and Panama. For example, Mexico, with 41 taxa of Anthurium has 26 endemic taxa. Guatemala has only three endemic species. Honduras and Nicaragua each have a single endemic species. Costa Rica has 68 taxa with 22 species endemic, and Panama has 150 species of which 82 species, 55% of the total, are considered endemic.

Because of the high rate of endemism and the very high speciation in many parts of the Andes, our taxonomic knowledge of all but a few areas of the Andes is poor. Although selected areas of the Colombian Andes, such as the region of Popayán in Cauca Department, the departments of Antioquia, and the department of Cundinamarca, especially around Bogota, were well collected in the late nineteenth century by collectors such as Lehmann, early enough to have their material included in the revisions of both Schott (1860) or Engler (1905a), many areas had not been collected until recent times.

Pichincha Province and a few other areas of Pacific coastal Ecuador were well collected by L. Sodiro (Sodiro, 1901a, 1901b, 1902-1903, 1903, 1905a-c, 1906, 1907, 1908a, 1908b) and 257 species (including 281 taxa) were described. Despite this, the region remains poorly known, largely because of the inability to locate and study his widely scattered and poorly documented collections. Some Sodiro specimens are deposited in European herbaria (Croat, 1991) but most collections are deposited in the poorly curated herbarium (QPLS) of the monastery in Quito, where Sodiro originally worked. The collections may not be borrowed and the conditions in the herbarium make their study there very difficult (Croat, 1991). Nevertheless some recent attempts at revisionary work have taken place in Ecuador in Pichincha Province. Floristic surveys of the Araceae have been made of the Reserva ENDESA on the western slopes of Volcán Pichincha (Croat & Rodríguez, 1995). This work, begun by Jimena Rodriguez de Salvador while a student at the Universidad Catolica in Quito, was subsequently augmented by investigations by Croat. Other florulas are also being prepared as well and a comparison of six different florulas in Ecuador has been completed (Croat, 1995b).

A genus by genus account of the taxonomic status of neotropical genera and a discussion of poorly known floristic regions in the neotropics has been published elsewhere (Croat, 1994c).