...

Clr

%‘ The Paim I Comunity’in a Forest of Central Amazonia, ’

Francis Kahn2 O.R;STO.M.-Institut Français de Recherche Scientifique pour le Développement en Coopération (Convênio ORSTOM- CNRS, FranesANPA-CNPq, Brazil), C.P. 478, Manaus, Brazil

and y. Aline de Castro 0.R.S.T.O.M.-Instituto Nacional de Pesquisas da Amazônia, Dept. Ecologia, C.P. 478, Manaus, Brazil

ABSTRACT The palm community in a forest of central Amazonia dearly shows three zones according to the hydromorphic condition of the soil: we!!-&ned soils-of the upland forest, poorly-drained soils of a transition zone, and water-logged soils of the seasonal swamp forest. The community i., remarkable in three aspects: its size (with 2 122 palms/ha, and the highest density on water-logged soi!$; its gear diversity (32 specirs/l.2 ha), Wiririir also depends on the hydromorphic condition of the scii; and its variety of biologcd foms with rharxtxistic acaulescent palms in the understory, monocaulous and multiple-stemmed palr.s in the upper understory, znd axborescempdms only reaching the canupy in the seasonal swamp foreit.

PALMSARE AN ABUNDANT and characteristic component of dance of the most important species in relation to topog- the forests of central Amazonia. They are found in all raphy and soils. levels of the forest, from the understory to the canopy, m- ail soils and topogqbic &tes, arid exhibit a vâtiety of METHODS growth forms. Liowevec, a Moore (1973a) lamented, OUT ignorance of the biology 2nd ecology of palms is almost The survey areas were located in the Tropical Sylviculrure cmplete. Camer’s -w.e?aus book (1966) represents the Experimental Station of INPA, 45 km from Manaus, cornerstoor.=.of our bawledge of the family in these ways. between the BR-174 (Manaus-Boa Vista highway) md Several eariy works dealt >~?jeci.ficallywith the pal= .fio= the Rio Cuieiras, a tributary of the Rio Negro (2”35‘- of Âinamnia. In f.2~:since the mlv studies of this snbjiecr WO’latitude S; 60”00’-60”20‘ longitude W). Rainfdl B&=TJ’s j18Z) h:&q .$ F& ir, yr& z& k about 2.5 m pcr year. The rairi7 sasï= extends from . ~ --- Wallwe’s famous :book:. (ï,STfjj: Spmc5.s WÜ& i.E ;L, December to Ma!: the dry season from -Tu.. to Novem- 1998); ~~ro~p&&&3ns of Mz& (ikx+ ber. The water defi& is low or nil and the potential 1853)): %de (!87i-i908i:,Barbosa-?.o&pes i18ï5- evapotranspiradon is regularly distributed tfiroilghout the 190Tj:. and Burret (i.328-1956)’, only’ILiackde’s work year (Ribeiro, 1976). (19hCi) h. P-JU and Wessels-Sûer’s (1965; ad(19;lli The survey areas, 6 km apart, were selected :o include in and iienezuela, respectively, haye arrtrib- a catena of five topographic sites: plateau, crest, slope, a uted significant!y to che of amazonian pdms. transidon zone ar the base of the slope, and the seasonal Recendy. Balick and de Silva (1982) established the sciz flocdeci “igarapé” (stream) bed (Fig. i). According to of systematic coiiecci-cns in regional herbaria and &û-x& Ramani and de Podestá (1984), the soils of placeau cor- che neczssity to intensify them. respond to ”Latossolo hatelo hco, textura muito ar- Works relating to the ecology of palms are even,rarer gilosa” (heavy yellow clay latosols), the.soils of crest and (see Discussion). Thus, we initiated a series of ecological slope to “Podzólico Vermelho Amarelo Álico, textura ar- investigations en ,hriazonian palas in 1950 at the Na- gilosa” (heavy day red yellow Podzolic soils), the soils of tional Insdrutt for Amazonian Research (INPA). This traiition zone CO hydromorphic Podzols, and the soils of paper presa= a communicy-level study of palms in cen- the “igarapé” bottom to Gley Pouco Húmico (Humic =a! Fimazonia and describes the distribution and abun- Gley). These works dearly showed that the main variable dong the catena is hydromorphic condition. In this paper we shall qualify the latossols of the plateau and the pod- zolic soils of the slope as well-drained soils, the soils of ‘ Received 24 Jul.; 1983. revisiqn accepted 14 June 1981. ’Address during 1985-1986: I:AP/JH, Cailla 783, Iquitos, transition zone as poorly-drained soils and the soils of

,lope 7RESULTS In the 1.2 ha surveyed, 32 species of palms, in 12 genera, were encountered. All species were collected, and vouchers are now deposited at the department of Ecology, INPA, Manaus. Identification was to species, except when floris- tic material was not completely collected (flowers were espeady infrequent). The species (with voucher no.'s) were: htrocatyum acaule Mart. (AC 590), A. sociale FIGURE 1. Vegetation, topographic sites, and soils Barbosa Rodriaues (FK 587), A. mirnbaca Mart. (AC 522), Attalea artaleoides (Barbosa Rodr.) W. Boer (AC 615), A. specctabilis Mart. (FK 618), precatoria Mart. (AC 591), Inartea exowhiza Mart. (FK 581), lr-

-15 .-

10

--r3 -l

2E

I water logged soils -poorly drained soils -well drained soils

FIGURE 2. Distribution of the palm community along the catena. Aa = Attalea attaleoides, Asp = Attalea spectabilir, Asa = htrocaryum acaule, Asj = htroraryum rociale, Asm = htrocaryum munbaca, Ep = Euterpe precatoria, Ir = exowbiza, Jb = Jerrenia bataua, Ma = Mauritia aculeata, Mm = Manicaria martiana, Ob = , Om = Oenocarpus minor, and Si = SyagruJ ìnajaì. spp., Geonoma spp. and are not drawn.

Forest Palm Community 272

.a 1

TABLE 1. Presence ofpalm species in relaiion to soils.

Mature Architectural WDS PDS WLS height model Arborescent palms Oenocarpus bacaba + 15 m Corner Iriartea exorrbiza + + + 15 m Corner Jessenia bataua + f 20-25 m Corner htaiiritia aculiæia + + 15 m Tomlibson Euterpe precatoria (+Y + + 20-25 m Corner Understory palms Syagtus inajai f 8-10 m Corner Oenocarpus minor c + 8-10 m Tomlinson htrocatyum munbaca + 3-10 m Tomlinson Iriartella setìgera + i- 7-10 m Tomlinson Attalea attaleoides i- 3-5 m Corner (acaulescent) Artrocaryum sociale + 3-5 m Corner (acaulescent) Attaled spectabilis , + + 3-5 m Corner (acaulescent) Astrocatyum acaule + 3-5 m Corner (acaulescent) Manicaria martiana f 5m Corner Bactris ~pp.~ 11 1 1’ 0.5-5 m Tomlinson Geonoma ~pp.~ G 1 0.5-5 m Tomlinson a Only seedlings and juveniles of Euterpe precatoria were found on well-drained soils. 12 species of Bactris were found on well-drained soils (WDS). One of these (Bactris Jimpljc$rons Mart., FK 5 18) penetrated the transition zone on poorly-drained soils (PDS). Only one other species (FK 62 1) occurred in seasonal swamp forest on water-logged soils (WLS). The 6 species of Geonotna were restricted to well-drained soils. Only one of them (AC 525) reached the transition zone on poorly-drained soils.

ìartella setigera (Mart,) Wendl. (FK 598), Jessenia ba- 5 m by the of two acaulescent species, AJerocaryum taua [Mart.) Burret (FK 505), Manicayid inartiuna BLU- sociale and Abaie a attaleoideJ. Eleven and six species, ret (EK 5191, Afaariitia aculeatu Mart. (FK 648), respectively, of the genera Bactris and Geonoma were also Oenocarpus bacabu Mart. (FK 645), O. minor Mart. (FK found here but their frequency varied and their total num- 612), Syaps inajai (Spruce) Beccari (FK dol), Bactris ber per plot was often reduced to a few individuals. These spp. (FK 621), (AC 548), (AC 552), (FK XE), {AC species formed hpsof two to ten axes. 549), (AC 5471, (FK 3I8), (FK 624), (FK 626), (FK In the upper level of the understory, between six and 6271, (FK 6131, (FK 5971, Gonoma spp. (AC 5931, ten meters, Sydgrtrs inajai (a monoaxial palm with a stem (AC 5351, (FK 6281, (FK 589), (AC 524), (AC 525). of 4-5 m) was frequent, as were htrocaryrrm muncubd The distribution of these species was strongly related and Oenocarpus minor. The latter two species formed to soils and three distinct “palm zones” are recognized dumps of two to four axes. Less fcequently encountered {Fig. 2): well-drained soils (plateau, crest, slope), poorly- was lrìartella setigera, a small palm, seven to ten meters drained soils (transition zone), and water-logged soils in height, characterized by the productioE of creeping (“ippé” bottom). Each of these “palm zones” was rhizomes. In the forest mid-msy, Oenocarprrs burubu characterized by a number of typical species, bur there ’ reached a height of I5 m. It was the dominkt palm in were also a few, less abundant species which wecom- height and was only found below &continuities of the mon to two or three ZOE (Table 1). upperstory canopy, indicacng thüì k comperes only in -Qnancitímv.- An+? sie *mredby topographic site, dmings with gtcugh iight. li.s&g the mmiq“~ specks. Figures from kh sur- -qareas are cuidad adtherefore total umber of PMS ON POORLY-DRAINED SOILS (TRANSITION ZONE).- palms are given for 2400 rn: (Table 2). Two acaulescent palms, Attaleu spertabilis and &rotar- yrrm amuie, overran the undemory of the forest. They Pahs ON WELL-DRAINED seis (PLATEAU, CREST, SLOPE).- were very abundant and theii lea~esspread up, funnel- In the 7200 mz of forest on well-drained soils we sur- like, to five meters. These two species were strictly limited ‘eyed, 1346 palms between one and ten meters in height to the transition zone at the base of the slope and dis- were recorded, and of these 676 belong to two acaulescent appeared in the vicinity of the “igarapé” when the soil species. No palms reached the canopy in forest on well- becomes waterlogged. There, they were replaced by stem- drained soils which we studied. The understory of the less juveniles of Jessenia baturra. This transition zone does forest on well-drained soils was dominated to a height of indeed stand out; in 2400 rn2, 343 acaulescent palms 212 Kahn and de Castro . \

TABLE 2. Population of most abundanl pdfm species (per 2400 and Benson 198U), seems quite independent of the nature m2) IF: plateau; C: crest; S: slope; TZ: fransitioti of the soil, but requires light and was only found below zone; IB: “igarapé” bottom). openings of the forest canopy. DRY - iMOIST Form DISCUSSION Species P C S TZ IB Acaulescent The palm community of the forest of central Amazonia Astrocaryum sociale 198 206 162 O O dearly shows three zones according to che hydromorphic At talea artafeoides 86 91 75 O O condition of the soil: well-drained soils of the upland Astmcaryrrm acaule O O O 142 O forest, poorly-drained soils of che transition zone, and Attalea spectabilis O O O 385 19 water-logged soils of the seasonal swamp forest. The com- Upper undersrory munity is remarkable for its size, its great diversity, its Oenocarpus minor 6 30 24 19 O variety of biological forms. Syagrus inajai 16 66 8 O O The discussion of community size and diversity will Astrocaryuin mirnbaca 37 60 49 O O be limited due to scarcity of comparable Amazonian lic- Arborescent erature in chis field. Works relating to the ecology of Oenocarp-pus ltacaba 211 231 205 O O palms generally only take into consideration one species: Euterpe precatoria 11 18 31 255 820’ ecology of Raphia palm swamps in Costa Rica (Anderson Jessenia bataua O O O 291’ 1592b and Mori, 1967); demography of Artrocqw.n mexiran- Orher species (22) 135 204 166 28 92 um (Piñero et al. 1977, 1982; Piero and Sarukhán Tocal/2400 m’ 700 906 720 1120 2523 1982); intraspecific competicion in, Souatea durisJima (Yeaton, 1979); and autoecolûgj? of Euterpe giobosa of a With 5 palms above 10 m in height. With respectively 6 and 15 palms above 10 m in he.in &e forests of ?;e;to Pdco (Bannister 1970, Van Valen “igarapi“ borrom, the number of seedlings and juveniles below 1975). Some informaton of an ecologica! name is given 1 m in heighr was estimated hm200 m2 surveyed. i~ various Tor& mnceming Sou& American vegetation (Myers 1933, Davis and Richards 1934, Beard 1955, greater than one meter in height were recorded; of cbc, Cldeíían 1971) oz concerning useful (Cavalcante 259 were Attaiea rpectabiiis and 74 were Astrocaryum 1977, Anderson 1978, &lick 1981). Msc~re(1973b) acaule. Also, 158 young Jessenia bataua (from one to six regroups che taxa in terms of the main neotropical dimax nìeteis in height) were found. formations. Specifidy regarding the Amazon region, . Bouillenne (1930) describes ttre Ataauritia formations on PALMSON WATER-LOGGED SOILS (“IGARAPÉ”BOTTOM).- poorly-drdned sa&ï; Cideman (1969) tacides the ecology The forest on water-logged soils was almost deyoid of the of the Euterpe oleracea formation or “pinouïres” of French acaulescent palms (only 19 Attalea spectabilis), but over- Guiana, and Granville (1974, 1977, 1978) presents sev- run by young Jesrenia bataua (472, already defined). One eral studies on the biology and ecology of the palms of species of Bactrir formed clumps, its axes not exceeding . In fact, quantitative data on palms were &ve meters in height; Mauritia aculeata were distributed found in studies relating to structural analysis of vegm- in isolated patches. Whereas arborescent palms were ab- tins, phytomass evaluatians, and histic inventaies. Sent fram the forest canopy on well-drained soils, they were an important compment of the forest canopy on ABUNDANCE AND DIVERSITY OF ?ALMS.-& are the fOreStS poorly-drained and water-logged soils: Jessenia bataua, of Malaya-n which Whitmore (1973) wrote: “No- Mauritia flexaosa (not encountered in our plots, although where in the primary jungles which cover much of the very common in the “igarapé” bottoms), Mauritia acu- country is one far from a palm”-the forest of central leata, Euterpe precatoria. Amazonia is literally overrun by palms, especially in understory. Klinge and Rodrigues (197 1) established that CROSS TOPOGRAPHY PALMS.-sOme Species Of the forest on palms greater than 1.5 m in 2000 m2 of forest on well- well-drained soils (such as Oenocarpus minor, Inartella drained soil near Manaus represented 17 percent of the setigera, Bactris simplicifrons, and Gonoma spp.) pene- foliar phytomass. Takeuchi (1960) counted 112 palms trated the transition zone at the base of the slope. Euterpe per 1600 m2 on high “terra fir”” forest and 176 per precatoria, frequent on water-logged soils, was occasion- 2200 m2 on low “terra firme” forest. Lechtaler’s (1956) ally found on well-drained soils, but usually only as seed- and Aubréville’s (1961) surveys in Ducke Reserve near lings and juveniles that have germinated and started their Manaus, and Prance et al. (1976) inventory of one hect- growth in gaps; very few reached maturity in the transi- are at kilometer 30 on the Manaus-Itacoatiara road tion zone. Iriartea exowbiza, an arborescent palm with showed, also, the importance of palm populations in the typical stilt roots (Bouillenne 1924, Kahn 1977, BodIey Amazonian forests. Forest Palm Community 213

c TABLE 3. Number of palm Jpecies per plor (1200 mz).

Habitat Catena 1 Catena 2 Plateau 16 16 Crest 19 20 Slope 19 19 Transition zone 5 10 ' ' Igarapë ' bottom 7 6

Palm densities were highest on the well-drained soils of the crest: for the combined area of 2400 m2, 906 individuals were recorded for the crest, as opposed to 700 and 720 for the plateau and slope, respectively (Table 2). This is most likely a result of greater exposure to wind and subsequent higher frequency of gap formation. The forest here thus tends to be less well developed architec- FIGURE 3. Acaulescent palms in upland forest understory on turally, permitting greater penetration of light, which fa- well-drained soils (ArtroraryumjauarenJe). vours the regeneration of palms. Physical and chemical differences exist between the soils of the plateau and those of the slope. Slope soils Mauritia acdeata, which occurs in seasonal swamp for- have a higher density Orr macropores and a lower mgmic esr, is the only species forming dumps. matter content than plateau soils (Ranhand de P3destL . The absence of bentmultiple-stemmed palms

19e4). These differences, however, do not appeax to afht ' &om upland forest was noted and accounted for by Gran- the size of the palm popuhtion on these mo distinct ville (1978). All arborescent palms require high light topographic sites. ievels during the stage of stem growth and thus in the Hydromorphic condition afiects the population den- forest their regeneration tends to be restricted to gaps. As sity of palms. We counted 3643 palms on 4800 m2 (or a gap doses, palm growth must keep pace with pioneer 7590.ha-l) on poorly-drained and water-logged soils, and species (Cecropia, Inga, Pourouma, Protium, etc.) in order 2326 palms on 7200 m2 (or 3231.ha-I) on well-drained to compete effectively for light and continue development soils. The greater size of palm populatians on hydro- af its steta It follows then that such gaps are not suirable morphic soils is essentially due to greater intekj of li& for palms of Tomlinson's growth model (defined by Halle which penetrates into the understory of the open seasonal et al., 1978: basal ramification and possible formation of swamp forest. dumps), since the second axis initiated at the lower part The mean of all plots gives 2122 palms (above one of &e stem will be shaded out, from the moment it meter in height) per hectare. in the farests of Tocan&s emerges, by the pioneer species then maturing in gap. Valley, Pará. Kahn (1983) counted a mean of 602 palms Arborescent mdriple-stemmed palms, however, can (above one meter in height) per hercay, from 10.56 ix deve!op E the sssczxi samq forest on water-logged surveyed- The abundance of palms in this central Ama- soils. Heq dx &rest is more apwith fewer md zonian forest is particularly high. zmhient light Levels XE 5gkLex-&an in upland forest and The species richness of palms in this forest is &o s&5ent to maintaln stem grad. A-borescent palms of high: 32 species in 1.2 ha. This can be compared with Comer's model (monoadous palms with lateral inflores- the parallel study in the forests of the Tocac&?s Valley cences, see Halle et al., 1978) are also more abundant which recorded 21 species in 10.56 ha (Kxhn, 1983). on water-logged soils: 26 palms greater than ten meters The total number of species of palms reaches a maximum in height were recorded in 4800 m' on poorly-drained in upland forest on well-drained soils and decreases in and watet-logged soils compared to four or! well-drained seasonal swamp forest on water-logged soils (Table 3). soils in 7200 m'. Further. on well-drained soils, arbores- Even more striking is the variation of the number of cent palms are rarely exceeding 15 m in height (as also understory species which in plots of 1200 m' decreased noted by Klinge and Rodrigues, 1973), whereas on water- from 14-17 species on well-drained soils to three species logged soils they often reach 25-30 meters. on water-logged soils. In understory species, Tomlinson's growth form is as common as Corner's form (Table 1). Thus, in the case VARIETYOF FoRMs.-The great majority of arborescent of multiple-stemmed palms, htrocaryzrnì niunbaca and palms in the forest of central Amazonia are monocaulaus. Oenocarpus minor occupy the upper levels of the under-

214 Kahn and de Castro

.-. Le!1.;

. - . ___ ”

4 B story and various species of the genera Bactris and Gen- ACKNOWLEDGMENTS om2 are distributed between 0.5 and five meters. Simi- This study was supported by Conselho Nacional de Desenvol- larlY, with monocaulous palms, sYagrus inajai vimento Científico e Tecnológico (CNPq), Instituto Nacional the upper level of the understory, and between WOand de Pesquisas da Amazõnia (INPA) and by the Projeto Multi- five “ers the leaves of the four acaulescent palms (As- laceral .“Ecologia da Floresta Tropical” of the Organisation of tuocag~amsociale, A. acaule, Attalea attaleoides, A. spec- thewAmerican States (OAS/OEA). We are indebted to Cath- tabili.r) constitute an almost continuous cover. While the erine Mackenzie for her helpful assistance on the English manu- script. We would Funher like LO thank Prof. Francis Halle, Jean- Stem Of these species are reduced to a short Louis GuiIlaumet and Charles Clement for providing valuable subterranean axis, the leaves maintain significant sue and comments upon critical reading of the manuscript. We also give the understory of these central Amazonian forest its thank José Palheta for his assistance during field work. charamristic appearance (Fig. 3).

LITERATURE CITED ANDERSON,A. B. 1978. The names and uses of palms among a tribe of Yanomama indians. Principes 22(1): 30-41. ANDERSON,R., AND S. MORI. 1967. A preliminary investigation of Raphia palm swamps, Puerto Viejo, Costa Rica. Turrialba 17: 221-224. AUBRÉVILLE,A. 1961. Érude écologique des principales formations végétales du Brésil. C.T.F.T., Nogent sur Marne. 268 pp. BALICK,M. J. 1981. lemnia bataua and Oenorarpus species: native Amazonian palms as new sources of edible oil. In E. H. Pride et al. (Ed.). New sources of fats and oils, pp. 141-155. American Oil Chem. SOC. , AND M F: DA SILVA. 1982. Palm taxonomy in Brazilian Amazônia: the stare of systematic collections in Regional Herbaria. Brittonia 34: 463-477. BANNISTER,3. A. 1970. Ecological life cyde of Euterpe gIobosa Gam. In E Odum (Ed.). A mpical rain forest, pp. B299- B3 14. U.S. Atomic Energf Cwr~,3ak Ridge, Temes=. BEARD,J. S. 1955. The classification of tropical american ve-ptation types. Ecology 36: 89-100. BODLEY,J. H., Ahm F. C. BENSON. 1980. S&-mot dagby ia ïciamid phin che Peruvian Amazon. Biorropica 12(1): 67-2 1. BOUILLENNE,R. 1924. Les racines échasses de lriarteu exorrhìza (Palmiers) et de Pa12danur div. sp (Pandanacées). Mém. CI. Sci. Acad. R. Belg., 2ème ser., T. VIII. 4jp. . 1930. Un voyage botanique dans le bas-Amazone. Arch. Inst. Bot. LTn-iv. Liège 8: 1-1.85, pls. 1-34. CAVALCANTE,P. B. 1977. Edible Palm Fruits of the Brazilivl Amazoe Prinüpes 21: 91-102. CORNER,E. J. H. 1966. The natural history of palms. Weidenfeld and Nicohn, London. 393 pp., 133 fip., 24 PIS. DAVIS,T. A. W., AND P. W. RICHARDS.1934. The vegetation of Moraballi Ge& Britirh Guiana: an ecological study of a limited area of tropical rain forest, Part II. J. Ecol. 22: 106-155. GLASSMAN,S. F. 1972. A revision of B. E. Dahigren’s Index of ame&xï paiins. J. Cramer, Germany, 294 pp. GRANVILLE,J. J. DE. 1974. Aperçu sur la structure des pneumatophores de deux espèces de sols hydromorphes en Guyane. Cah ORSTOM, sér. Biol. 23: 3-22. . 1977. Notes biologiques sur quelques palmiers guyanais. Cah. ORSTOM, sir. Biol. 12(4): 347-353. . 1978. Recherches sur la flore et la végétation guyanaises. Thèse Doctorat d’Etat Univ. Montpellier. 272 pp. HALLÉ,F., R. A. A. ~DEMAN,AND P. B. TOMLINSON.1978. Tropical ~eesand forests: an architectural analysis. Springer Verlag, Berlin, Heidelberg and New York. 441 pp. KAHN,F. 1977. Analyse structurale des systèmes racinaires des plantes iigncuses de la forêt +cale dense humide. Candollea 32: 321-358. . 1983. Architecture comparée de forêts tropicales humides et dynamique.& la rhizospkère. ’&&e Doctorat &Etat Univ. Montpellier. 426-pp. KLINGE,H., AND W. A. RODRIGUES.1971. Matédorgânica e nutrientes na mata dererra firmeperto de Manaus. AmAmazonica l(1): 69-72. --. 1973. Biomass estimation in a central Amazonian rain farest. Acta Cient. Venezolana 24: 225-237. LECHTHALER,E. 1956. Inventário das hres de um hectare de terra firme da Zona de -Reserva-Florestal Ducke, Municipio de Manaus. Amazonia, Rio de Janeiro Bot. 3, 10 pp. MACBRIDE,J. F. 1960. Flora of . Palms: Field Mus. Nat. Hist., Bot. series, XIII, Pan 1, No. 2: 321-418. MOORE,H. E. JR. 1973a. The major groups of Palms and their distribution. Gentes Herbarum 1 l(2): 27-141. . 1973b. Palms in the tropical forest ecosystems of Africa and . In B. J. Meggers, E. S. Ayensy, and W. D. Duckworth (Eds.). Tropical forest ecosystems in Africa and South America: a comparative review, pp. 63-88. Smith- sonian Institution Press, Washington, D.C. MYERS,J. G.- 1933. Notes pn the vegetation of the Venezuelan llanos. J. Ecol. 21: 335-349. OLDEMAN,R. A. A. 1969. Etude biologique des pinotières de la Guyane française. Cah. ORSTOM, Ser. Bioi. 10: 1-18. . 1974. L’architecture de la forêt guyanaise. Mém. ORSTOM, 74. 204 pp. PiÑmo, D., AND J. SARUKHÁN.1982. Reproductive behavior and its individual variability in a tropical palm Artroraryum mexiranum. J. Ecol. 70: 461-472. --, AND E. GONZALEZ.1977. Estudios demográficos en plantas. Artroraryum mexicanum Liebm. I. Estructura de las poblaciones. Bol. Soc. Bot. Mex. 37: 69-1 18.

Forest MICol7.;munity 215 .* --, AND P. ALBERDI.1982. The costs of reproduction in a tropical palm Artrocatyum mexicanum. J. Ecol. 70: I ' ,e 473-482. POITEAU,A. 1822. Histoire des palmiers de la Guyane Française. Mém. Mus. Hist. Nat. Paris 9: 385-393. PRANCE,G. T., W. A. RODRIC~Q,AND M. F. DA SILVA.1976. Inventkio Florestal de um hectare de mata de terra firme, Km 30 da estrada Manauslyfiacoatiara. Acta Amazonica 6(1): 7-35. RANZANI,G., AND J. A. DE PODESTÁ.1984. Alguns atributos de solos dispostos em uma topossequéncia na bacia hidrográíica III - RegiÕ do Baixo Rio Negro. Acta Amazonica (in press). RIBEIRO,M. DE N. G. 1976. Aspectos climatológicos de Manaus. Acta Amazonica 6(2): 229-233. SPRUCE,R. 1871. Palmae Amazonicae, sive Enumeratio Palma" in Itinere suo per regiones Americae aequatoriales lectarum. Joum. Linn. Soc. 11: 65-175. . 1908. Notes of botanist on the Amazon and Andes, 1: 1-518, 2: 1-542. London. TAWUCHI,M. 1760. A estrutura da vegetação na Amazonia. I. A mata pluvial tropical. Bol. Mus. Paraense E. Goeldi, Bot., 6: 1-17. VAN VALEN,L. 1975. Life, death and energy of a . Biotropica 7(4): 260-267. WALLACE,A. R. 1853. Palm trees of Amazon and their uses. pp. 1-129, pl. 1-48. London. WESSELS-BOER,J. G. 1965. Indigenous palms of Suriname. E. J. Brill, Leiden, 172 pp. . 1971. Clave descriptiva de las palmas de . Acta Bot. Venezuelica 6(1, 2, 3 y 4): 299-362. WHITMORE,T. C. 1973. Palms of Malaya. Oxford Univ. Press, London. 132 pp. * YUTON,R. I. 1979. Intraspecific competition.in a population of stilt root palms, Socratea durhima (Gerst.) Wend. Biotropica ll(2): 155-158.

For a teview of current research on tropical forest clearings (gaps) for the Annual Review of Ecology and Systematics, I won!c! appreciate receiving reprints of papers published or in press. JuiieS. Dwfslow New York Botanical Garden Brom, NY :E4355 USA

216 Kahn and de Castro