Brittonia, 31(1), 1979, pp. 26-38. �9 1979, by the New York Botanical Garden, Bronx, NY 10458

NOTES ON THE VEGETATION OF AMAZONIA III. THE TERMINOLOGY OF AMAZONIAN FOREST TYPES

SUBJECT TO INUNDATION

GHILLEAN T. PRANCE

Prance, Ghillean T. (New York Botanical Garden, Bronx, NY 10458). Notes on the vegetation of Amazonia III. The terminology of Amazonian forest types subject to inundation. Brittonia 31: 26-38. 1979.--The types of Amazonian forests subject to inundation can be organized into seven categories which are herewith named and described. This classification is intended to set in order the confusion of terminology used in the past. The types are: (1) seasonal vdrzea~forest flooded by regular annual cycles of white-water rivers; (2) seasonal igapd--forest flooded by regular annual cycles of black- and clear-water rivers; (3) mangrove forests flooded twice daily by salt-water tides; (4) tidal vdrzea--forest flooded twice daily by fresh water backed up from tides; (5) floodplain forest---on low lying ground flooded by irregular rainfall, generally in upper reaches of rivers; (6) permanent white-water swamp forest; (7) permanent igap6--black-water forest. The first five types are periodically inundated and the last two are permanently waterlogged. This terminology is closer to that used by lim- nologists by restricting the use of igapd to forest inundated by black and clear water.

Floodplain areas have received increasing attention in recent years with the de- velopment of Amazonia. Various recent workers and also the Brazilian govern- ment agency SUDAM (Superintend~ncia de desenvolvimento da Amaz6nia) have stressed the need to use the floodplain for agriculture. Camargo (1968), Moreira (1970), Goodland and Irwin (1975, 1977) and many other recent authors have stressed the suitability of vdrzea forest areas for intensive agriculture because of their annual soil enrichment by flooding. Agricultural exploitation of vdrzea could relieve considerable pressure from the forest on unflooded terra firme, which is char- acteristically on soils not well suited for agriculture. As the floodplain forest receives more attention it is important to define accurately the terminology of the natural forest in order to carry out further studies of the vegetation.

There has been so much confusion in the botanical literature resulting from the inconsistent terminology applied to the different types of Amazonian forest subject to permanent or temporary inundation that it is necessary to adopt a uniform termi- nology that accounts for vegetation cover, the type of water, and the duration of flooding. The terms vdrzea and igap6 in Brazil and tahuampa in Peru have been used in so many different ways by botanists, ecologists and scientists in other disci- plines as to often be meaningless. I have, therefore, proposed herewith a simple classification of the major inundated vegetation types of Amazonia in the hope that they can be employed consistently and uniformly in the future in both systematic and ecological literature.

Inundated forest represents the various types of climax of the hydrosere, as discussed by Richards (1952). The classification presented in this paper is based on the major climax forest types and does not attempt to define all the types of successional forest on recently formed banks of alluvium.

BRITTONIA 31: 26-38. January-March, 1979.

1979] PRANCE: AMAZONIAN FOREST TYPES 2 7

Igap6 and Vfirzea

The two terms most frequently applied to the flooded forests of Brazilian Amazonia are vdrzea and igapr. They have been employed in many different ways and their definitions have consequently been obscured. The usage of the terms by limnologists has been different from that of botanists. Vdrzea has usually been applied to the ground seasonally flooded by the rivers of Amazonia. Almost all workers recognize a seasonal variation of flooding in their use of the term vdrzea. Sioli (1956) and other limnologists use the term only for the areas flooded by white-water rivers. This is the definition which I advocate here. This application of the term vdrzea excludes the areas flooded by black-water rivers and also by clear- water rivers, such as the Rio Tapaj6s. At present the areas flooded by black- and clear-water rivers are termed igap6 in upper Amazonia, and vdrzea by many of the people of lower Amazonia. In lower Amazonia the term igap6 has most generally been applied to permanently waterlogged swamp forest, a vegetation type which is much more common near the Amazon estuary than in the region of the main black- water rivers of upper Amazonia. Richards (1952, pp. 285, 443) equates the term igap6 with peaty swamp or moor forest as he terms it, and applies vdrzea to forest "liable to flooding." The confusion of the terms igap6 and vdrzea is apparent in the literature. Gessner (1968), Moreira (1970) and many other workers included forests flooded by both black and white water in their concept of the term igapd.

Ducke and Black (1953, 1954) in what stands as the classical work on Ama- zonian phytogeography, also use vrrzea for all periodically flooded forest and igap6 for forest on ground which never becomes dry. However they concentrated their attention on the forest occurring on terra firme and did not describe in detail the inundated types.

Aubrrville (1961) , in his book on the vegetation types of Brazil, also used the term vdrzea for all periodically flooded forest and igap6 for all permanently flooded forest. He pointed out the important distinction between "high vfirzea," which is flooded only for a short period when the rivers crest, and "low vfirzea," which is flooded for a much longer period. He also clearly drew attention to the differences in vegetational composition between areas flooded by the different water types, but did not give details. He presented the results of an inventory of an area of tidal vdrzea near Belrm and one of the seasonal igap6 on the Rio Negro. The latter inventory was done by William A. Rodrigues, who used the term vdrzea for the riverine vegetation of the black-water Rio Negro.

Takeuchi (1962) applied igap6 to forests growing on very low ground with only a very short dry period, regardless of water type. For example, the area described by Takeuchi at Lago Janauaca is clearly subject to white-water flooding with many species of plants that do not grow in association with black-water, and the Rio Cuieiras area is one of black water with its characteristic vegetation. In Prance (1975) I also followed the commonest usage of these terms in botanical literature by defining temporarily flooded areas as vdrzea and permanently flooded ones as igapr.

The most satisfactory vegetational classification of inundated forest to date is that of Pires (1961) who applied the term igap6 to permanent swamp areas in- undated by either black or white water, and to seasonally inundated areas along either black- or clear-water rivers.

Earlier Sioli (1951) had proposed a revised terminology with the term vdrzea restricted to white-water flooded areas and igap6 for black-water flooded areas. Irmler (1977) also followed Sioli's terminology and clearly applied vdrzea to forest flooded by white water and igap6 to black water. In addition, Irmler recognized

28 BRITTONIA [VOL. 31

three types of vdrzea and two of igap6 based on the degree of flooding and the distribution of the benthic macrofauna. Irmler helped greatly to clarify the differ- ence between vdrzea and igap6, but did so primarily by differentiating the benthic fauna and did not relate these differences closely to new data about the vegetation. He used the vegetation data of other workers who had not distinguished clearly the different types of flood-zone vegetation by their phytosociology. In conclusion, we clearly see that the present usage of the terms vdrzea and igap6 or their English equivalents is obviously confused and there is a need for redefinition of the inundated forest types.

Similarly, in Amazonian Peru, forests which are seasonally flooded by black- and white-water rivers are not distinguished, both types being called tahuampa.

The Types of Water in Amazonia

An important factor in the composition of the inundated forest vegetation is the water type. There are three main types of water in the rivers of Amazonia. Certain riverine plants grow only in one water type, mainly because of differences in acidity and nutrients. For example the famous water-lily, Victoria amazonica (Poepp.) Sowerby, only grows in white-water areas.

White waters, which are actually muddy brown, carry a heavv suspension of alluvium. The Rio Solim6es, a white-water river, has a pH of 6.9-7.4 (Schmidt, 1972b).

Black waters appear dark blue to black and contain clear water which is stained dark brown by a colloidal suspension of plant compounds. The black water of the Rio Negro has a pH of 4.6-5.2 (Anon., 1972).

The clear or crystal water is neither muddy nor stained by suspended material. Clear-water rivers usually appear blue or green owing to sky reflections. The largest is the Rio Tapaj6s.

In addition to the marked difference in pH among the water ~ypes there are con- siderable variations in nutrients and humic matter. Schmidt (1972a, 1972b) gives the humic contents of the three main Amazon water types as follows:

Rio Negro (black water) 26.6 rag/liter Rio Solim6es (white water) 14.1 mg/liter Rio Tapaj6s (clear water) 2.26 mg/liter

Further details about Amazonian water contents are given in the various studies of the Max Planck Limnology Institute in P16n, for example in Anonymous (1972), Schmidt (1972a, 1972b), Sioli (1965, 1967), and in Williams et al. (1972). It is obvious that such great differences in pH, nutrients and humic matter will have an influence on vegetation type, yet current definitions of the flooded forest types largely ignore the water characters.

The Vegetation Types of Inundated Forest

There are seven main types of vegetation on inundated ground in Amazonia, five of which are periodically inundated and two of which are permanent swamp forests. These may be defined both by vegetation cover and by the type of water or duration of flooding. These seven types of flooded vegetation are defined in the key below, and the associated vegetation discussed in some detail. The classification below differs from previous vegetational definitions in that it follows the practice adopted by limnologists which restricts the term igap6 to black- and clear-water areas.

1979] PRANCE: AMAZONIAN FOREST TYPES 29

Key to the principal types of Amazonian forests subject to inundation

1. Periodically inundated forest. A. Flooded by regular annual cycles of rivers.

i. White water. 1. SEASONAL VARZEA ii. Black and clear water. 2. SEASONAL IGAPO

B. Flooded by tidal movements. i. Salt water. 3. MANGROVE ii. Fresh water backup. 4. T IDAL VARZEA

C. Flooded by irregular rainfall (flash floods). 5. FLOODPLAIN FOREST

2. Permanently inundated forest.

i. White water. 6. P E R M A N E N T SWAMP FOREST ii. Black and clear water. 7. P E R M A N E N T IGAPO

Apart from the seven major types of inundated forest distinguished above, there are obviously many additional subdivisions based on the duration of flooding, variation in the pH and nutrient content of the different rivers, and other factors. Some plant species are confined to regions of one water type, others are widespread and occur in various types of inundated forest. For example, two tree species char- acteristic of both black- and white-water inundated forests are Virola elongata (Benth.) Warb. and Caryocar microcarpum Ducke. The riverside tree Allantoma lineata (Mart. ex Berg) Miers ranges from the upper rlos Orinoco and Negro in Venezuela to the vicinity of Bel6m. In southern Venezuela in the Rio Negro region

Fi~. 1. Aspect of seasonal vdrzea (swamp) forest, Riozinho, Pardi, Brazil. In flood s.eason the trees in this photograph will be about 2 meters deep in water.

30 BRITTONIA [VOL. 31

Fro. 2. Seasonal vdrzea (swamp) forest flooded annually by a small white-water river, Riozinho, Par~. Tall buttressed trees are common.

A. lineata is always a black-water igap6 plant, downstream it is frequent in white- water vdrzea and near the coast around Bel6m it is frequent in tidal vdrzea.

The purpose here is to define only the major categories, as adequate comparative ecological studies do not yet exist by which further divisions might be distinguished. There is a great need for detailed vegetation analysis of areas in which each type

1979] PRANCE; AMAZONIAN FOREST TYPES

TABLE I SOME CHARACTERISTIC TREES OF SEASONAL V/~RZEA FOREST

31

Species Family

Bombax munguba Mart. Bothriospora corymbosa Hook. Calycophyllum spruceanum (Benth.) K. Schum. Carapa guianensis Aubl.* Cecropia spp. Ceiba pentandra Gaertn.* Couroupita subsessilis Pilg. Cratavea tapia L. Gustavia augusta L. Hevea brasiliensis Muell. Hura crepitans Muell.* Euterpe oleraeea Mart. Piranhea trifoliata Baill. Vitex cymosa Bert.

Bombacaceae Rubiaceae Rubiaceae Meliaceae Moraceae Bombacaceae Lecythidaceae Capparidaceae Lecythidaceae Euphorbiaceae Euphorbiaceae Arecaceae Euphorbiaceae Verbenaceae

* Also widespread outside vdrzea.

occurs to find out whether there are any association communities definable in terms of their vegetation, as Irmler (1977) found in terms of the benthic fauna.

1. Seasonal vfirzea (swamp) forest = Mata de vdrzea

This is forest that is flooded annually by the rise of the white-water rivers (Figs. 1 & 2). Seasonal vdrzea supports a high biomass with many large trees and lianas, but fewer arthropods than igap6 (see Adis, 1977). Buttress roots (Fig. 2) and pneumatophores are common. There is considerable variation in the seasonal vdrzea forest because of the length of the annual flood period. On sloping river banks the duration of inundation decreases with distance from the river bank. This is the most extensive and widespread of all types of flooded forest in Amazonia and occurs along all the major white-water rivers, often extending several kilometers back from the river bank. The herbaceous zone of the forest is rich in individuals of Heliconia (Musaceae) and Costus (Zingiberaceae). Some of the many species of trees characteristic of this forest are listed in Table I. Many of the predominantly vdrzea species in Brazil are not confined to vdrzea in Amazonian Peru, where the climate is wetter, or in extra-Amazonian tropical America. Pires (1961) made a geographic division of the vdrzea into "vdrzea fores t of lower Amazonia" and "vdrzea forest of upper Amazonia." This division is not used here since it is preferable to wait until there is a firm basis for subdividing the various types of seasonal swamp forest according to association communities.

2. Tidal v~irzea (swamp) forest = Mata de vdrzea estuario

This forest is flooded and drained twice daily by the tidal movements, because high tides temporarily block the flow of rivers in the estuarine region and cause them to flood the adjacent forest. Curiously this more consistently wet type of forest is very similar to the seasonal vdrzea both in species composition and in physiognomy. It is forest of relatively high biomass and flooded by the white water, the predominant type in rivers and creeks of the estuarine region. As there is con-

32 BRITTONIA [VOL. 31

FIG. 3. Black-water igap6, Rio Negro near Manaus, further down river than in Fig. 5 where there is much less species diversity and only a short dry non-flooded period.

siderable variation within the tidal vdrzea, it could probably be divided into various vegetational subdivisions. The estuarine areas subject to daily tides are obviously more swamp-like than land flooded only by spring tides, the latter tending to be similar to the seasonal vdrzea.

Areas dominated by palms are frequent in the tidal vdrzea as well as the prevalent mixed forest. The commonest palms, Mauritia flexuosa L.f. and Euterpe oleracea Mart., often found in pure stands, are also found in other habitats, whereas Raphia taedigera Mart. and Manicaria saccifera Gaertn. are two abundant species restricted to tidal areas throughout their range.

Common trees of the tidal vdrzea forests include Virola surinamensis Warb., Cedrelinga castanaeformis Ducke, Ceiba pentandra Gaertn. and Mora paraensis Ducke. One of the most characteristic species of the forest margin and succession is Machaerium lanatum Tul. Other species of the coastal area include Pithe- colobium huberi Ducke, Derris latifolia Prain, Hymenaea comosa (Sw.) Miers, Inga bourgoni DC., and Tabebuia aquatilis (E. Mey.) Sprague & Sandwith.

3. Seasonal igap6

This term is restricted here to the forest which is annually flooded by black- or clear-water rivers. Igap6 usually has sandy soil which sustains a much poorer vegetation than the vdrzea forest of white-water rivers. In some places igap6 experi- ences desert-like conditions when it dries out, consequently it has fewer species and displays xeromorphic adaptations, such as sclerophyllous leaves. The forest is often interspersed with sandy beaches. There is much less species diversity, and the trees tend to be low and tortuous. Sclerophyllous leaves are characteristic and there is a greater number of insects (Adis, 1977). However, denser and physi-

1979] PRANCE: AMAZONIAN FOREST TYPES 33

FIG. 4. Black-water igap6, Rio Negro near Manaus. Even at the low-water level there are shrubs which spend only 1-3 months above the water. This photo is taken below Figs. 3 & 5.

ognomically more "vdrzea-like" forest also occurs in some areas of igap6, especially in upper Amazon ia in Colombia and Peru, as well as in the A m a z o n delta region, where it occurs on richer soil rather than the usual impoverished sand. The forest often descends a lmost to the low-water level, and consequently some of it has only a two-month dry period (Figs. 3, 4, 5 & 6) . M a n y species of trees and shrubs are endemic to this habitat; a few also grow in seasonal vdrzea. Some of the charac- teristic plants of igap6 are listed in Table II . One of the characteristic unders tory palms of this forest is Leopoldinia pulchra Mart. (Fig. 5) . Ano the r c o m m o n palm of igap6 is Mauritia aculeata H.B.K.

TABLE I I SOME CHARACTERISTIC PLANTS OF IGAPO

Taxon Family

Aldina latifolia Benth. var. latifolia Couepia paraensis (Mart. & Zucc.) Benth. Eugenia inundata DC.* Eugenia spp. Eschweilera ( lugastrum ) eoriaeeum (DC.) Miers* Licania apetala (E. Mey.) Fritsch Macrolobium acaeiae[olium (Benth.) Benth.* Parkia peetinata (H.B.K.) Benth. Crudia amazoniea Benth. Tabebuia barbata (E. Mey.) Sandw.

Caesalpiniaceae Chrysobalanaceae Myrtaceae Myrtaceae Lecythidaceae Chrysobalanaceae Caesalpiniaceae Mimosaceae Caesalpiniaceae Bignoniaceae

* Also occur less abundantly in white- or mixed-water areas.

34 BRITTONIA [VOL. 31

FIG. 5. Black-water igap6, Rio Negro near Manaus. The palm Leopoldinia pulchra Mart. is characteristic of the upper part of the igap6 that is flooded for only a short time and has a more diverse vegetation than the areas flooded for longer periods in Figs. 3 & 4.

4. M a n g r o v e forest = Mangue

M a n g r o v e forest occurs abundan t ly in the es tuar ine region of A m a z o n i a in areas which are f looded dai ly by brack ish water . The mangrove fo rma t ion is well known and easi ly recognized. I t is inc luded here only because it is one of the inunda ted

1979] PRANCE: AMAZONIAN FOREST TYPES 35

forest types of Amazonia. Where it extends into fresh water it is often part of the succession colonizing newly formed banks, rather than a climax forest type. This typical and uniform vegetation of tidal salt-water areas around the tropics is char- acterized in the Amazon delta by Red Mangrove (Rhizophora mangle L.) , Siriuba (Avicennia nitida Jacq.) and Languncularia racemosa Gaertn.) . The mangrove Rhizophora racemosa G. F. Mey. extends as far upstream as the tides in some places and thus is not confined to brackish water.

5. Floodplain forest

This term refers to forest that is flash-flooded by irregular rainfall. It exists mainly in the upper reaches of Amazonia and beside small creeks. Floodplain forest is subject to flash flooding by heavy rainfall at any time of year rather than by the regular seasonal flooding of large rivers. It is physiognomically similar to seasonal vdrzea and shares many of the same species. Where flooding is usually of short duration only, various species characteristic of terra firme forest also occur.

6. Permanent swamp forest (permanent igap6 and permanent white-water swamp forest)

Permanent swamp forest, on ground that is continuously waterlogged, is rarer in Amazonia than the periodically flooded forest types. It is often of the "back swamp" type and occurs behind the main river banks in depressed areas that never fully drain out in the dry season and that are created by a bank or levee of riverine alluvium which forms between them and the river. In lower Brazilian Amazonia the permanent swamp forest has more generally been termed igap6, leading to con- fusion of terminology with black-water areas. In this paper the term igalM is restricted to the black- and clear-water areas since the two vegetation types are quite different. There are very few black-water permanent swamps or permanent igap6s in Amazonia. All permanent swamp forests have relatively few species, but the trees are often large and physiognomically the forest often resembles vdrzea. In the delta area there is extensive lower permanent swamp forest under tidal in- fluence, with trees 10-15 m tall with straight boles. The trees have many adapta- tions such as scleromorphic leaves suggesting that they are under water stress. Adventitious rooting (Fig. 6) is common in black-water igapds. The canopy is more open than in the vdrzea and the ground flora is rich in Cyperaceae. This type of swamp forest was termed "igap6 do estudrio" by Pires (1961) and is best referred to as tidal permanent swamp forest. All the permanent swamp forests of the Amazon basin proper are fresh-water swamps in the sense of Richards (1952) and Whitmore (1975) , overlying gleyed soil. In the coastal region of the Guianas peat swamp forest occurs, locally termed pegass swamp. The conditions which lead to the formation of a particular kind of permanent swamp forest are not fully under. stood, and there has been considerable controversy over this issue in the tropica' Far East where both kinds occur extensively (Whitmore 1975).

7. Mixed water inundated forests

In the region of Manaus, Brazil, where the large black-water Rio Negro and the large white-water Rio Solim6es come together, and in other river confluences with differing water types, the forest falls under the influence of both types of water since the rivers crest at slightly different times. Another example of this is seen in the northern part of the Anavilhanas Archipelago in the Rio Negro, where the influenr/

36 BRITTONIA [VOL. 31

Fic. 6. Adventitious rooting is a characteristic of black-water igap6, photographed here beside the Rio Negro, above Manaus.

of the large white-water river, the Rio Branco, influences the igap6 vegetation of the Rio Negro. In the northern part of the Anavilhanas the physiognomy of the igap6 is more similar to that of white-water vdrzea.

This brief review of the major types of inundation forest points to the lack of good data on plant association communities by which to define the various vegeta-

1979] PRANCE: AMAZONIAN FOREST TYPES 37

tion types. There is an obvious need for further study of the inunda ted forest types to differentiate the subdivisions of the categories defined here according to their vegetation and physiognomy rather than by the degree of f looding and type of water.

Acknowledgments

Recent work in Amazon ia has been supported by Nat ional Science F o u n d a t i o n grant I N T 75-19282. I am especially grateful to Diplo. Biol. Joachim Adis, Drs. Spencer C. H. Barrett , Alwyn Gentry, Scott A. Mori, J. Mur~a Pires, Wil l iam A. Rodrigues and Timothy C. Whi tmore for reading a draft of this paper and for offering many helpful suggestions.

Literature Cited

Adis, J. 1977. Programa mlnimo para anfilises de ecosistemas: Antr6podos terrestres em florestas inundfiveis da Amaz6nia Central. Acta Amazonica 7: 223-229.

Anonymous. 1972. Die lonenfracht der Rio Negro, Staat Amazonas, Brasilien, nach unter- suchungen yon Dr. Harold Ungemach. Amazoniana 3 (2): 175-185.

Aubr6ville, A. 1961. I~tude 6cologique des principales formations v6g&ales du Br6sil et contribution 5. la connaissance des for~ts de 1'Amazonie Br6silienne. Centr. Techn. Forest. Trop. (Paris). 268 pp.

Camargo, F .E . 1968. Recursos naturais e humanos da Amaz6nia. Revista Brasil. Polit. Internatl. XI (41/42): 84-100.

Ducke, A. & G. A. Black. 1953. Phytogeographical notes on the Brazilian Amazon. Anais Acad. Brasil. Ci. 25(1): 1-46.

1954. Notas s6bre a fitogeografia da Amaz6nia Brasileira. Bol. T6cn. Inst. Agron. N. 29: 1-62.

Gessner, F. 1968. Zur 6kolgischen Problematik der Uberschwemmungsw~ilder das Amazonas. Int. Rev. gesampten Hydrobiol. 53: 525-527.

Goodland, R. J. A. & H. S. Irwin. 1975. Amazon jungle: Green hell to red desert? Elsevier. Amsterdam & New York. 155 pp.

1977. Amazonian forest and cerrado: development and environmental conserva- tion. Pp. 214-233. In: G.T. Prance & T. S. Elias (eds.). Extinction is forever. New York Bot. Gard.

Irmler, U. 1977. Inundation--Forest types in the vicinity of Manaus. Biogeographica 8: 17-29.

Moreira, E. 1970. Os Igapds e seu aproveitamento. Imprensa Universitaria, Bel6m-Par(l. Pires, J .M. 1961. Esb6~o fitogeogrfifico da Amaz6nica. Revista Soc. Agron. Patti 7: 3-8. Prance, G.T. 1975. Flora and vegetation. Pp. 101-111. In: R. J. A. Goodland & H. S. Irwin.

Amazon jungle: Green hell to red desert? Elsevier. Amsterdam & New York. Richards, P.W. 1952. The tropical rain forest. Cambridge Univ. Press. 450 pp. Schmidt, G .W. 1972a. Chemical properties of some waters in the tropical rain forest of

Central Amazonia along the new road Manaus-Caracaral. Amazoniana 3 (2): 199-207. 1972b. Amounts of suspended and dissolved substances in the middle reaches of

the Amazon over the course of one year. Amazoniana 3(2): 208-223. Sioli, H. 1951. Zum Alterungsprozess von Fltissen und Fliisstypen im Amazonasgebiet.

Arch. Hydrobiol. 45: 267-283. 1956. t3ber Natur und Mensch im brasilianischen Amazonasgebeit. Erkunde 10(2):

89-109. 1965. Bemerkungen zur Typologie amazonischer Fliisse. Amazoniana 1: 74-83. 1967. Studies in Amazonian waters. Atas do Simp6sio s6bre a Biota Amaz6nica 3:

9-50. Takeuchi, M. 1962. The structure of the Amazonian vegetation. VI. Igap6. J. Fac. Sci. Univ.

Tokyo, Sect. 3, Botany 8: 297-304.

38 BRITTONIA [VOL. 31

Whitmore, T. C. 1975. Tropical rain forests of the Far East. Clarendon Press, Oxford. 282 pp.

Williams, W. A., R. S. Loomis & P. T. Alvim. 1972. Environments of evergreen rain forests of the lower Rio Negro, Brasil. Tropical Ecology 13: 65-78.

BOOK REVIEW

Polygonaceae (Buckwheat Family) of New York State. Richard S. Mitchell and J. Kenneth Dean. Contributions to a Flora of New York State I, edited by Richard S. Mitchell. Bulletin 431, 79 pp. + index, illus., maps. New York State Museum, Albany. 1978. $2.50.

This account of the Polygonaceae of New York state, floristic in scope but essentially monographic in depth and originality, is the first contribution to a state flora which, when completed in the same style, promises to become a classic in its field. The Buckwheat family is represented in New York by the common rhubarb, 15 docks (Rumex), no less than 32 knotweeds (PoIygonum, sens. lat.), 2 true buckwheats (Fagopyrum) and the annual jointweed (Polygonella articulata), 51 species in all, of which nearly two-fifths are adventive or fully naturalized in the area. At the outset the authors have tackled a group notorious for its com- plexity, which involves in Rumex hybridization and polyploidy and in Polygonum the vexed status of the amphibious species and of the cosmopolitan weeds. It is an impressive beginning.

Physically the book is a soft-cover pamphlet of 80 generous pages (28 • 21 cm), clearly printed on paper of good quality, faultlessly proofed, a pleasure to read and to handle. It consists of a short preface, the floristic substance, appendices listing fungi and some insects associated with Polygonaceae in New York, an impressive bibliography, and an index to Latin names. The descriptive matter includes a detailed description of each genus, keys to the species, and for each of these an original account, an illustration (habit-sketch with significant organs appropriately enlarged), a map of dispersal in the state, followed by paragraphs on taxonomic variation and economic uses. A box paragraph immediately following the name of each species encapsulates several of its leading features: common names, taxonomic synonyms, protolog, geographic origin, habitat, habit of growth and duration, season of flowering, and dispersal outside the state. A peculiarity of the generic and specific descriptions is the inversion of the traditional order of presentation which starts with flower and fruit and proceeds, as it were backward, by way of inflorescence, leaves and stems to the root or rhizomes. The virtue of this inversion is to place emphasis on the organs of production which furnish the basis of classification. While the evaluation of the species is conservative, care has been taken to note internal complexity and unsolved problems; no dust is swept under the rug. The discussion of Polygonum amphibium, clearly derived from original observations of a sort detectable throughout the contribution, is especially noteworthy.

Dr. Mitchell, Dr. Dean, the council of the Flora Committee and the State Museum are all to be congratulated on initiating a project worthy of the state itself, a flora intelligible to all that does not sacrifice scientific excellence to popular accessibility, a textbook that will be used and enjoyed no less by technical botanists than by plain people. We look forward with anticipation to further contributions and ultimately to a comprehensive account of New York's rich and varied flora. --RUPERT C. BARNEBY, New York Botanical Garden, Bronx, NY 10458.