A Model for the Importance of Large Arborescent Palms in the Dynamics of Seasonally-Dry Amazonian Forests

A MODEL FOR THE IMPORTANCE OF LARGE ARBORESCENT PALMS IN THE DYNAMICS OF SEASONALLY-DRY AMAZONIAN FORESTS

Rodolfo Salm1, Euphly Jalles-Filho2 & Cynthia Schuck-Paim3

Biota Neotropica v5 (n2) – http://www.biotaneotropica.org.br/v5n2/pt/abstract?article+BN02705022005

Date Received 04/13/2005 Revised 08/08/2005 Accepted 09/12/2005

1School of Environmental Sciences, University of East Anglia, Norwich, Norfolk NR4 7TJ, England. Universidade Federal de São Carlos, Programa de Pós-graduação em Ecologia e Recursos Naturais, Rododovia Washington Luís, km 235, Monjolinho. CEP 13565-905, São Carlos, SP, Brazil. e-mail: [email protected] 2University of São Paulo, Department of Physiology (IB), Rua do Matão, Travessa 14, 05508-030, São Paulo, SP, Brazil e-mail: [email protected] 3University of São Paulo, Department of Experimental Psychology (IP), Av. Prof. Mello Moraes 1721, 05508-030, São Paulo, SP, Brazile-mail: [email protected]

Abstract In this study we propose a model that represents the importance of large arborescent palms in the dynamics of seasonally-dry Amazonian forests. Specifically, the model is aimed at guiding the investigation of the role of large arborescent palms on forest regeneration and succession. Following disturbance, the high level of luminosity reaching recently formed forest gaps favors the quick proliferation of shade-intolerant lianas that, by casting shade on the crowns of mature forest trees and increasing tree-fall probability, suppress forest succession. Due to their columnar architecture palm trees are, however, not severely affected by vines. As the palms grow, the canopy at the gaps becomes gradually higher and denser, progressively obstructing the passage of light, thus hindering the growth of shade-intolerant lianas and enabling late-successional tree development and forest regeneration. Owing to the long time associated with forest regeneration, the model cannot be tested directly, but aspects of it were examined with field data collected at an Attalea maripa-rich secondary forest patch within a matrix of well-preserved seasonally-dry forest in the Southeastern Amazon. The results indicate that (1) forest disturbance is important for the recruitment of large arborescent palms species, (2) these palms can grow rapidly after an event of disturbance, restoring forest canopy height and density, and (3) secondary forest dominated by palm trees species may be floristically similar to nearby undisturbed forests, supporting the hypothesis that the former has undergone regeneration, as purported in the model. Key words: Amazon, Attalea maripa, forest dynamics, palm trees, regeneration.

Resumo Neste estudo propomos um modelo que representa a importância de palmeiras arborescentes de grande porte na dinâmica das florestas amazônicas sazonalmente secas. Especificamente, o modelo visa guiar a investigação do papel destas palmeiras na regeneração da floresta. Após um evento de perturbação, a alta luminosidade das clareiras recentes favorece a rápida proliferação de lianas dependentes de insolação que, sombreando as copas das árvores e aumentando sua probabilidade de queda, tendem a interromper o processo de sucessão. Devido a sua arquitetura colunar, as palmeiras arborescentes, no entanto, não são severamente afetadas por lianas. À medida que estas palmeiras crescem, o dossel das clareiras se torna gradualmente mais alto e denso, limitando desta forma o crescimento das lianas e permitindo o desenvolvimento das árvores de estádios sucessionais tardios e, consequentemente, a regeneração da floresta. Devido ao longo período associado à regeneração da floresta o modelo não pode ser testado diretamente, mas aspectos deste foram examinados com dados de campo coletados em uma mancha de floresta secundária rica em Attalea maripa em uma matriz de florestas primárias bem preservadas no sudeste da Amazônia. Os resultados revelam que (1) eventos de perturbação florestal são importantes para o recrutamento das palmeiras arborescentes de grande porte, (2) estas palmeiras podem crescer rapidamente depois de eventos de perturbação, restaurando a altura e a densidade do dossel, e (3) florestas secundárias dominadas por espécies de palmeiras podem ser floristicamente semelhantes a florestas primárias adjacentes, apoiando a hipótese de que as primeiras avançam na regeneração, como previsto pelo modelo. Palavras-chave: Amazônia, Attalea maripa, dinâmica florestal, palmeiras arborescentes, regeneração. http://www.biotaneotropica.org.br Salm, R.; Jalles-Filho, E. & Schuck-Paim, C. - Biota Neotropica, v5 (n2) - BN02705022005 2 Introduction “early succession” species are able to colonize the site im- The understanding of succession processes in natural mediately following perturbation, modifying the environ- communities as well as their effects upon community stability ment so that it is more suitable for later succession species and organization was among the major concerns of early ecolo- to invade and grow to maturity — the original conception gists (Cowles 1899). Initially, succession was described by a of succession (Clements 1916). The other two models, con- relatively rigid sequence of species successively invading a versely, assume that any arriving species, including those site (Clements 1916), a notion subsequently expanded to in- which usually appear later in the succession process, can clude predictable changes in characteristics of forest struc- colonize a disturbed site. The tolerance model assumes that ture other than species composition, such as biomass, pro- the modification brought about by earlier colonists neither ductivity, diversity and niche breath (Odum 1969). Although increases nor reduces the recruitment rate and probability in these studies ecosystems were viewed as highly predict- of growth to maturity of later colonists. Finally, the inhibi- able and organized systems, objections to this concept arose tion model proposes that once earlier colonists secure space early, reaching a critical point in the early seventies (Connell or other resources, succession is suppressed. 1972, Drury & Nisbet 1973) due to a lack of clearly defined and Seasonally-dry forests, consisting of an assemblage of testable hypotheses about the mechanisms underlying suc- ecologically diverse vegetation formations and generally char- cession events and relative scarcity of direct evidence of late acterized by high discontinuity of their canopy, variable level succession stages (Connell & Slatyer 1977). of deciduousness and strong spatial variation in floristic To contribute to this matter Connell & Slatyer (1977) macromosaics, cover a substantial area of the Amazon basin. These forests are distributed mostly along rainfall gradients described three models to explain the succession process defining the limits between the Amazonian forest and external after an event of disturbance: the facilitation model, the tol- dryer vegetation biomes (Pires 1984, Gentry 1988, Daly & erance model and the inhibition model. All models purport Prance 1989), an area now often referred to as ‘Deforestation that certain species can become established before others Arc of Amazonia’ (Fearnside 1993, Fearnside 1995). In this due to the presence of “colonizing” characteristics such as study we investigate the role played by large arborescent rapid growth and a high level of investment on reproduc- palms in the dynamics of seasonally-dry Amazonian forests. tion. The three models differ, however, in the mechanisms Specifically, we propose a simplified model of the system, in whereby new species colonize the area later in the succes- which the role of palm trees in the dynamics of forest succession cycle: the facilitation model assumes that only certain sion and regeneration is explored (Figure 1).

Figure1. Pictorial model of the importance of large arborescent palm on the regeneration dynamics of seasonally dry Amazonian forests. The negative sign indicates an inhibition effect, whereby the growth of one functional plant group is inhibited, e.g. by another group (for instance, in the gap phase, the presence of lianas inhibit the growth of trees). For a definition of the plant groups and regeneration phases, see the text. http://www.biotaneotropica.org.br Salm, R.; Jalles-Filho, E. & Schuck-Paim, C. - Biota Neotropica, v5 (n2) - BN02705022005 3

The influence of large arborescent palms in the dy- higher and denser, progressively obstructing the passage namics of succession proposed here, incorporates elements of light, thus hindering the growth of shade-intolerant li- of the three models of Connell & Slatyer (1977), as we shall anas and consequently favoring tree development and for- explain as follows. Before proceeding, some explanations est regeneration. We also propose that palm trees are self- are however necessary. Our model considers three func- limiting because, as other early successional species, the tional groups of plants: “shade-intolerant lianas”, “large development of juveniles is hampered by the shadow pro- arborescent palms” and “mature-forest trees”. Although duced by adult palms. With growth of mature forest trees dissimilar in many aspects, one difference of special rel- and palm senescence, the gap phase progresses to matu- evance to the present discussion is the distinct response of rity, and forest structure stabilizes again. the three mentioned groups of plants to light: mature-forest Similarly to Connell & Slatyer’s facilitation model trees have highest survival likelihood under low light inten- (1977), the model just described above is fundamentally fa- sity, shade-intolerant lianas show the opposite response, cilitative, as it states that once a disturbance opens a rela- having the lowest capacity to survive light shortage (and tively large space releasing resources (light and space), large highest growth rate at maximal light intensity), with large arborescent palms are among the early succession species arborescent palms being intermediate. to become established, modifying the environment so that Additionally, forest succession is heuristically parti- it becomes less suitable for the subsequent recruitment of tioned here into three phases: a “gap phase”, a “building early succession species, but more suitable for the recruit- phase” and a “mature phase” (Aubréville 1938). The gap- ment of late succession species. It also incorporates a com- mosaic dynamics, first proposed by Aubréville, have be- ponent of “tolerance” when it assumes that shade-intoler- come a fundamental concept in tropical forest ecology ant lianas are replaced by palm trees which, in turn, are (Richards 1996). The idea is currently widely used in analy- replaced by mature-forest trees, so that the sequence of sis of tree species distribution (Sheil & Ducey 2002) and the species is determined, not solely but fundamentally, by life- assessment of the impact of economic activities upon tropi- history characteristics that enable survival and growth de- cal forests (Fredericksen & Putz 2003). The first phase cor- spite the shade cast by early succession species. Inhibition responds to that immediately following an event of distur- is also considered in our model, since shade-intolerant li- bance, which leads as a by-product to the opening of clear- anas limit mature-forest trees and, as long as they persist, ings in the forest, hence the terminology ‘gap’. Regenera- continue to exclude or suppress subsequent colonists of tion leads to maturity, namely the reestablishment of canopy the mature forest. height and density and the re-colonization of the area by We do not propose that should large arborescent species of trees typical of mature forests. Once at the ma- palms be removed from the ecosystem, regeneration would ture phase, only a new disturbance event can bring the not proceed on the seasonally-dry Amazonian forests. The forest back to the gap phase. balance between forest building and regression depends The effect of large arborescent palms on forest re- on a wide range of factors affecting tree growth and sur- generation can be then understood by considering the fol- vival. In these highly seasonal forests, both drought and lowing dynamics. At the mature phase, the canopy is com- flooding contribute for disturbance, thereby explaining the posed mainly by the group of mature-forest trees, being high frequency of open forests areas with liana-rich forma- dense to a point that the reduced light conditions below it tions (Radam 1974). Due to their columnar architecture, prevents palm (Chazdon 1986, Kahn 1986, Tomlinson 1990, which hinders the spread of lianas, palm trees are however Kahn & de Granville 1992) and liana growth (Schnitzer et al. able to shift this balance in favor of denser vegetations. 2000, Alvira et al. 2004). Such an equilibrium can be broken It is important to highlight that the influence of palms by an event of disturbance, which by opening clearings in on forest succession as proposed here is specifically attrib- the forest enables solar radiation to reach the soil level with uted to the large arborescent forms of palms. Neotropical a higher intensity. Shade-intolerant lianas are then able to palms range widely from shade tolerant to needing high proliferate almost immediately, becoming dominant in this levels of light (Svenning 2001). Because the stem of palms gap phase (Radam 1974). Although lacking self-support is entirely primary, with no addition of secondary vascular capacity, lianas can climb the trees by using their structure or thickening tissues, the establishment phase imposes con- as support, thus severely suppressing tree growth and sur- siderable limitation on the overall habit of the plant vival (see e.g. Radam 1974 and Richards 1996 for examples (Tomlinson 1990). Palm trees compensate for the increasing in seasonally-dry forests). Additionally, they cast shade on mechanical support requirements during height growth by tree crowns and increase tree-fall probability (Schnitzer et a combination of initial development of a stem that has suf- al. 2000, Alvira et al. 2004), thus suppressing forest succes- ficient diameter, sustained cell expansion and increase of sion. Due to their columnar architecture palm trees are, how- stiffness and strength of the stem tissue with age. In some ever, not severely affected by vines (Richards 1996). As the cases, there is also cell division within the stem (Rich 1987). palm trees grow, the canopy at the gaps becomes gradually Due to the growth of their large stem, arborescent palms http://www.biotaneotropica.org.br Salm, R.; Jalles-Filho, E. & Schuck-Paim, C. - Biota Neotropica, v5 (n2) - BN02705022005 4 become increasingly light-demanding with increasing size ests are colonized by pioneer dicotyledonous species of (de Granville 1992) and generally depend on large gaps for the genus Cecropia or Vismia (Mesquita et al. 2001). Palm recruitment to the adult stage (Kahn 1986), whereas small trees indeed appear to be of much less important for regen- palms are widespread at the understorey of tropical forests, eration both in the rainier areas of the Amazon and on the even under the energetically restricting conditions of dense Atlantic forest. One reason behind the difference in the in- shade cast by the canopy (Chazdon 1986). fluence of palm trees on regeneration among ecosystems Indeed, ecologists soon related the developmental might be associated to the relative competitive advantage constraints imposed by the establishment phase of large of shade-intolerant lianas over mature forest trees in sea- arborescent palms to the general rarity of palm trees in areas sonally dry areas, due to the vines’ ability to produce deeper of dense forest with closed canopy, when compared to open roots than trees with the energy that these climbers save by forests (Kahn 1986, de Granville 1992, Kahn & de Granville not producing a self-sustained stem. Such deep roots allow 1992). While most species of large arborescent palms are the vines to reach the phreatic water for a longer period at generally rare in pristine, well-drained areas, they often domi- the dry-season—an advantage that is relatively less impor- nate secondary forests in the Amazon (Spruce 1871, Kahn tant in wetter forests (Nepstad et al. 1994). & Castro 1985, Kahn et al. 1988, Kahn & de Granville 1992, Ballée 1988, 1989). An exception is the palm Iriartea Materials and methods deltoidea, remarkably abundant in western Amazonia (Pit- man et al. 2001). This species does not seem to depend on Some aspects of the model are examined with field large gaps for recruitment (Svenning 1999). Such an excep- data collected at an Attalea maripa-rich secondary forest tion might be explained by the growth of stilt roots that give patch within a matrix of well-preserved seasonally-dry for- structural support to the development of a mature sized est in the Southeastern Amazon. To this end, we studied a trunk aboveground (Uhl & Dransfield 1987), and the possi- patch of secondary forests where palm trees were abun- bility of these palms to grow under dense canopy (Terborgh dant, in a matrix of primary forests, a few hundred meters o o & Davenport 2001). from the Pinkaití research station (7 46’18"S; 51 57’42"W), at the Kayapó Indigenous Lands, South-eastern Amazon, For example, large areas dominated by palm trees in Brazil. A palm grove that, according to the Kayapó indig- the Brazilian State of Maranhão are related to a ancient colo- enous population at the area, was cleared for the plantation nization frontier at this region (Kahn & de Granville 1992), of subsistence crops and subsequently abandoned in the and forest patches at the State of Pará were associated first half of the last century, was adjacent to a large isolated through archaeological evidence to abandoned Indian vil- hillside, the mirante, which provided a panoramic view of lages (Ballée 1988, 1989). Palm trees are also often dominant the study area. The study area was formed by a forest of in cleared areas around roads across the seasonally dry several hectares largely dominated by palm trees, being Amazonian forests. It is important to take into account that, adjacent to a denser forest, where few palm trees can be although many large arborescent palms have ecological re- seen. In a 16 ha grid system, established at the limit between quirements as described by the model, this is not true for all these areas, adult palms were exhaustively searched and (Svenning 1999, Araus & Hogan 1994), as described above 569 A. maripa palms were mapped. Within this grid, and at a in relation to the silt-palm I. deltoidea (Svenning 1999, Uhl surrounding undisturbed forest, studies were conducted & Dransfield 1987, Terborgh & Davenport 2001). on tree species diversity (Salm 2004a), palm growth (Salm Also of importance is the observation that the pro- 2004b) and the importance of forest disturbance for the re- posed model is, in its present form, restricted to seasonally- cruitment of large arborescent palms (Salm 2005). dry Amazonian forests, since the dynamics of regeneration is likely to follow different routes, involving mainly early successional dicotyledonous trees, under conditions other Results and discussion than those found in these biomes. In the seasonally-dry The results of these studies show that the large Amazonian forests, the alternating annual periods of strong arborescent palm A. maripa are generally rare at undis- rainfall and drought create a high level of natural distur- turbed areas of the Pinkaití (2 palms ha-1), but abundant bance, suitable for the growth of gigantic trees like mahoga- in naturally disturbed (12 palms ha-1) (Salm 2005) and nies (Swietenia macrophylla) and Brazil-nuts trees dominant at a patch of secondary forests within the study (Bertholletia excelsa), which create enormous gaps upon site (53 palms ha-1) (Salm 2004a). These findings support their deaths further increasing disturbance, thus establish- the proposed notion of large arborescent palms as light ing the ideal circumstances for the recruitment and growth demanding successional species that depend on large of large arborescent palms (Kahn 1986). But regeneration events of disturbance to become ecologically dominant not necessarily involving palms as ecologically important - a view that, although disputed (Svenning 1999), is colonizing-trees seem to prevail wetter areas elsewhere in largely accepted (Kahn 1986, de Granville 1992, Kahn & the Amazon. In Central Amazon, for example, secondary for- de Granville 1992), and essential to our model. http://www.biotaneotropica.org.br Salm, R.; Jalles-Filho, E. & Schuck-Paim, C. - Biota Neotropica, v5 (n2) - BN02705022005 5

The relatively fast rate of palm growth in height as- ronmental reconstitution in a likely future of large-scale sumed in the model, which quickly leads to the develop- deforestation in the Amazon. It is our hope that it contrib- ment of a relatively dense canopy, will keep the sun from utes for the development of further studies applied to the directly reaching the soil and therefore limit the growth of conservation of the highly threatened seasonally-dry for- shade-intolerant lianas. Supporting this assumption, growth ests of the Amazon, and to the sustainable exploitation and rate estimates showed that, once the palm’s stem of adult management of these ecologically important and economi- diameter is developed, a palm-dominated canopy could be cally valuable palm species. formed within a decade (Salm 2004b). This finding also highlights the potential importance of large arborescent palms as facilitative “tools” in programs of regeneration of defor- References ested areas within the largely destroyed and endangered ALVIRA, D, PUTZ, F.E. & FREDERICKSEN T.S. 2004. (Zimmerman et al. 2001) seasonally-dry Amazonian forests. Liana loads and post-logging liana densities after liana cutting in a lowland forest in Bolivia. For. Ecol. Additionally, the studied patch of secondary forest Manage. 190: 73-86. seems to be in an advanced stage of regeneration, having senescent palm populations (Salm 2004b) and floristically re- ARAUS, J. L. & HOGAN, K. P. 1994. Leaf structure and sembling the pristine areas of the Pinkaití Research Station patterns of photoinhibition in two Neotropical palms in (Salm 2004a). The successional nature of the palm grove is in clearings and forest understory during the dry season. fact evident on the structure of this forest patch. Its basal area Am. J. Bot. 81: 726-738. is extremely low (15 m2.ha-1), significantly lower than that of AUBREVILLE, A. 1938. La foret coloniale: les forets de the adjacent pristine forest (Salm 2004a). Such basal area is l’Afrique occidentale francaise. An. Acad. Sci. Co- even lower than the limits (18-24 m2.ha-1) considered for liana lon. Paris, 9: 1-245. forests (Pires & Prance 1985). The average height of the palm BALLÉE, W. 1988. Indigenous adaptation to amazonian palm grove forest, on the other hand, is not significantly different forests. Principes 32: 47-54. from that of the adjacent forest (Salm 2004a). These observa- BALLÉE, W. 1989. The Culture of Amazonian Forests. Adv. tions support the model’s proposed role for arborescent palms Econ. Bot. 7: 1-21. on the regeneration of ecosystem structure, given its recon- BONGERS, F., SCHNITZER, S.A. & TRAORE, D. 2002. stituting role of canopy height and density. The importance of lianas and consequences for for- Taken together, these results support the proposed est management in West Africa. BioTerre, Special role or arborescent palms on the regeneration of seasonally edition: 59 - 70. dry Amazonian forests. A final remark is however neces- CHAZDON, R.L. 1986. Light variation and carbon gain in sary. Despite its importance in the model, the negative ef- rainforest understory palm. J. Ecol. 74: 995-1012. fects of lianas on the growth of mature forest trees as well CLEMENTS F.E. 1916. Plant succession: an analysis of as on palms were not directly studied here due to the prac- the development of vegetation. Washington Publ, tical difficulties to measure vine’s biomass and, perhaps Washington. most importantly, the limitations associated to the temporal CONNELL, J.H. 1972. Community interactions on marine scale needed for such an investigation. Still, their suppres- rocky intertidal shores. Ann. Rev. Ecol. Syst. 3: 169-192. sive effects on mature forest trees, by shading their crowns CONNELL, J.H. & SLATYER, R.O. 1977. Mechanisms of and increasing tree fall probability, is recognized by the succession in natural communities and their role in com- Kayapó Indians (who cut the vines that climb useful tree munity stability and organization. Am. Nat. 111: 119-1144. species) and well-known in the forestry literature (Alvira et COWLES, H.C. 1899. The ecological relations of the al. 2004, Bongers et al. 2002). In Central America (Barro Colo- vegetation on the sand dunes of Lake Michigan. Bot. rado Island), vines dominate gap-phase regeneration and Gaz. 27:97-117. seem to inhibit non-pioneer tree survival (Schnitzer et al. 2000) but are significantly more abundant in younger than DALY, D. C. & PRANCE, G. T. 1989. Brazilian Amazon. In in older forests (Dewalt et al. 2000). Floristic inventory of tropical countries (D. G. Campbell & H. D. Hammond eds.) New York Botanical Garden, In conclusion, our model provides the basis for new New York, p. 401-426. lines of investigation that will allow the ecology of palm DE GRANVILLE, J. 1992. Life Forms and Growth Strategies trees to be better understood within the context of Amazo- of Guianan Palms as Related to their Ecology. Bull. I. Fr. nian forests. Field investigation in other regions, aided by Étud. And. 21: 533-548. the use of high-definition satellite imagery enabling the assessment of forest structure over large areas, should fos- DEWALT, S.J., STEFAN, A.C., SCHNITZER, A. & ter the development of the model together with its ability to DENSLOW, J.S. 2000. Density and diversity of lianas make quantitative predictions. Nevertheless, as it stands, along a chronosequence in a central Panamanian low- it highlights the potential role of palms as tools for envi- land forest. J. Trop. Ecol. 16:1-19. http://www.biotaneotropica.org.br Salm, R.; Jalles-Filho, E. & Schuck-Paim, C. - Biota Neotropica, v5 (n2) - BN02705022005 6 DRURY, W.H. & NISBET, I.C.T. 1973. Succession. J. Arnold SALM, R. 2004a. Tree species diversity in a seasonally-dry Arboretum 54: 331-368. forest: the case of the Pinkaití site, in the Kayapó Indig- FEARNSIDE, P.M. 1993. Deforestation in Brazilian enous Area, Southeastern limits of the Amazon. Acta Amazonia: the effect of population and land tenure. Amaz. 34: 435-443. Ambio 22: 537-545. SALM, R. 2004b. Densidade do caule e crescimento de Attalea maripa e Astrocaryum aculeatum: implicações FEARNSIDE, P.M. 1995. Potential Impacts of Climatic- para a distribuição de palmeiras arborescentes na floresta Change on Natural Forests and Forestry in Brazilian Amazônica. Biota Neotropica. 4 (1): http:// Amazonia. For. Ecol. Manag. 78: 51-70. www.biotaneotropica.org.br/v4n1/pt/ FREDERICKSEN, T. S, PUTZ, F. E. 2003. Silvicultural inten- abstract?article+BN00104012004 sification for tropical forest conservation. Biod. Conserv. SALM, R. 2005. The importance of forest disturbance for 12: 1445 – 1453. the recruitment of the large arborescent palm Attalea GENTRY, A.H. 1988. Changes in plant community diversity maripa in a seasonally-dry Amazonian forest. Biota and floristic composition on environmental and geo- Neotrop. 5:1. http://www.biotaneotropica.org.br/v5n1/pt/ graphical gradients. Ann. Missouri Bot. Gard. 75: 1-34. abstract?article+BN00305012005 KAHN, F. & CASTRO, A. 1985. The palm community in a SHEIL, D. & DUCEY, M. 2002. An extreme-value approach forest of central Amazonia, Brazil. Biotropica 20: 266-269. to detect clumping and an application to tropical forest gap-mosaic dynamics. J. Trop. Ecol. 18: 671-686. KAHN, F. 1986. Life forms of Amazonian palms in relation to SCHNITZER, S.A., DALLING, J.W. & WALTER, P. 2000. forest structure and dynamics. Biotropica 18: 214-218. The impact of lianas on tree regeneration in tropical for- KAHN, F., K. MEJIA, & CASTRO, A. 1988. Species rich- est canopy gaps: evidence for an alternative pathway of ness and density of palms in terra firme forests of gap-phase regeneration. J. Ecol. 88: 655-666. Amazonia. Biotropica 20: 266-269. SPRUCE, R. 1871. Palmae Amazonicae. J. Linn. Soc. Bot. KAHN, F. & GRANVILLE, J. 1992. Palms in forest Ecosys- 11: 65-183. tems of Amazonia. Springer Verlag, Heidelberg. SVENNING, J.C. 1999. Recruitment of tall arborescent palms MESQUITA, R.C.G., ICKES, K., GANADE, G. & in the Yasuni National Park, Amazonian Ecuador: are large treefall gaps important? J. Trop. Ecol. 15: 355-366. WILLIAMSON, G. B. 2001. Alternative successional path- ways in the Amazon basin. J. Ecol. 89: 528-537. SVENNING, J.C. 2001. Environmental heterogeneity, recruitment limitation and the mesoscale distribution of palms NEPSTAD, D.C., DE CARVALHO, C.R. & VIEIRA, S. 1994. in a tropical montane rain forest (Maquipucuna, Ecua- The role of deep roots in the hydrological and car- dor). J. Trop. Ecol. 17: 97-113. bon cycles of Amazonian forests and pastures. TERBORGH, J. & DAVENPORT, L. 2001. Endogenous and Nature 372: 666. exogenous control of leaf morphology in Iriartea ODUM, E.P. 1969. The strategy of ecossistem development deltoidea (Palmae). J. Trop. Ecol. 17:695-703. Science 164: 262-270. TOMLINSON, P.B. 1990. The structural biology of palms. PIRES, J.M. 1984. The Amazonian forest. In The Amazon: Clarendon Press, Oxford. limnology and landscape ecology of a mighty tropical UHL, C. & DRANSFIELD, J. 1987. Genera Palmarum. Allen river and its basin (H. Sioli, ed.). Dr. W. Junk Publishers, Press, Kansas. Dordrecht, p. 581-601. ZIMMERMAN, B., PERES, C.A., MALCOLM, J.R. & PIRES, J.M. & PRANCE, G.T. 1985. Vegetation types of the TURNER, T. 2001. Conservation and Development Alli- Brazilian Amazonia. Amazonia. G. T. Prance and T. E. ances with the Kayapó of South-eastern Amazonia, a Tropi- cal Forest Indigenous People. Env. Conserv. 28: 10-22. Lovejoy. Oxford, Pergamon Press. PITMAN, N.C.A., TERBORGH, J.W., SILMAN, M.R., Title: A model for the importance of large arborescent palms NÚNEZ V.P., NEILL, D.A., CERÓN, C.E., PALACIOS, in the dynamics of seasonally-dry Amazonian forests. W.A., & AULESTIA, M. 2001. Dominance and distribution of tree species in upper Amazonian terra firme for- Authors: Rodolfo Salm, Euphly Jalles-Filho & Cynthia ests. Ecology, 82: 2101-2117. Schuck-Paim RADAM 1974. Projeto RadamBrasil. Folha SB22 Araguaia e parte da folha SC 22 Tocantins. Instituto Brasileiro de Biota Neotropica, Vol. 5 ( number 2): 2005 Geografia e Estatística, Rio de Janeiro. http://www.biotaneotropica.org.br/v5n2/pt/ RICH, P.M. 1987. Mechanical structure of the stem of ar- abstract?article+BN02705022005 borescent palms. Bot. Gaz. 148: 42-50. Date Received 04/13/2005 - Revised 08/08/2005 Accepted 09/12/2005 RICHARDS, P.W. 1996. The tropical rain forest. Cambridge ISSN 1676-0611 University Press, Cambridge. http://www.biotaneotropica.org.br