793

BIOTROPICA 32(4b): 793810 2000

Patterns of Floristic Differentiation among Atlantic Forests in
Southeastern Brazil and the Inuence of Climate
1
Ary T. Oliveira-Filho
2
, and Marco Aure lio L. Fontes
Departamento de Cie ncias Florestais, Universidade Federal de Lavras, 37200-000, Lavras, MG, Brazil
ABSTRACT
Thetreeoraof southeastern Brazilian Atlantic forestswasinvestigated according to two main aspects: (a) thevariations
in oristic composition of both rain and semi-deciduous forests were analyzed in terms of geographic and climatic
variablesby performing multivariateanalyseson 125 existing oristic checklists; and (b) thelinksof both rain and semi-
deciduousforeststo Amazonian forestsand Cerrados(woody savanna) wereassessed. All analyseswereperformed at the
species, genus, and family levels. Theinformation obtained for the125 forest areaswasorganized into an environmental
databasecontaining geographic and climatic records, and aoristic databasecontaining binary presencerecordsfor 2532
species, 520 genera, and 106 families. Canonical correspondence analyses (CCA) wereutilized to assesstherelationship
between geographic and climatic variables, and tree ora composition. Venn diagramsand cluster analyseswereused to
assess theoristic linksto Amazonian forestsand Cerrados. Thefollowing patternsweredetected at all threetaxonomic
levels: (a) the differentiation between rain and semi-deciduous forests is oristically consistent and strongly correlated
with rainfall regime, although transitionsmay beabrupt to gradual; (b) anorthsouth differentiation existsfor both rain
and semi-deciduousforests, probably caused by variationsin both temperatureand rainfall regime; (c) Theoraof semi-
deciduous forests also changes with increasing distance from the ocean and theassociated increasing rainfall seasonality;
and (d) elevation and associated temperatures are strongly correlated with the internal differentiation of both rain and
semi-deciduous forests. To a considerable extent, the tree ora of semi-deciduous forests is a subset of the rain forest
ora, probably extracting speciesthat areableto copewith alonger dry season. Thereisgreater oristic similarity at the
species level between Atlantic rain and semi-deciduous forests than between any of these and either Amazonian rain
forests or Cerrados. Nevertheless, semi-deciduous forests and Cerrados show stronger links, particularly at the generic
and familial levels. Therefore, there is little oristic ground for viewing Atlantic rain forests as being closer to their
Amazonian counterparts than to the adjacent semi-deciduous forests. The most appropriate view of rain and semi-
deciduousforestsin southeastern Brazil isthat of acontinuum in treespeciesdistribution. Wesuggest that thedenition
of Atlantic forestsshould be as comprehensiveas that of Amazonian forests.
RESUMO
A ora arborea das orestas Atlanticas do sudeste do Brasil e investigada sob dois aspectos principais: (a) as variacoesem
composicao or stica de orestas ombrolas e semidec duas sao analisadas sob a otica de variaveis geogracas e climaticas
por meio de analisesmultivariadasde125 listagensor sticasexistentesna literatura; e(b) oslacosdasorestasombrolas
esemidec duascom asorestasAmazonicasecerradossao avaliados. Todasanalisesforamfeitasnosn veisdeespecie, genero
e fam lia. A informacao obtida para as 125 areas de oresta foi organizada em um banco de dados ambientais, contendo
registrosgeogracoseclimaticos, eumbanco dedadosor sticos, contendo registrosbinariosdepresencapara2532especies,
520 generos e 106 fam lias. Analises de correspondencia canonica (ACC) foram utilizadas para avaliar as relacoes entre
variaveis geogracas e climaticas e a composicao da ora arborea. Diagramas de Venn e analises de agrupamento foram
utilizados para avaliar os lacos or sticos com orestas Amazonicase cerrados. Osseguintespadroesforam detectadospara
todos os n veis taxonomicos: (a) A diferenciacao entre orestas ombrolas e semidec duas e oristicamente consistente e
fortemente correlacionada com o regime de chuvas, embora as transicoes possam ser abruptas a graduais; (b) Ha uma
diferenciacao norte-sul tanto paraorestasombrolascomo semidec duas, provavelmentecausadapor variacoesemtemper-
aturaeregimedechuvas; (c) A oradasorestassemidec duastambemmudacomadistanciado oceanoeocorrespondente
aumento da duracao da estacao seca; e (d) Altitude e suas correspondentesvariacoes detemperatura sao fortementecorre-
lacionadas com a diferenciacao interna tanto das orestas ombrolas como semidec duas. A ora arborea das orestas
semidec duase, emboamedida, um sub-conjunto daoradasorestasombrolas, provavelmenteextraindoespeciescapazes
deenfrentar umaestacao secamaisprolongada. Ha maissimilaridadeor stica, no n vel deespecies, entreorestasAtlanticas
ombrolas e semidec duas do que entre qualquer destas e as orestas Amazonicas, ou mesmo os cerrados. No entanto,
orestassemidec duasecerradosmostramlacosor sticosmaisfortes, particularmentenosn veisdegeneroefam lia. Portanto,
ha poucofundamentoor sticoparasepensar nasorestasombrolasAtlanticascomomaisproximasdesuascorrespondentes
Amazonicas do que de suas vizinhas semidec duas. A abordagem mais correta para orestas ombrolas e semidec duasdo
sudestebrasileiro e deum cont nuo dedistribuicao deespecies. Sugerimos, portanto, queadenicao deorestasAtlanticas
deveser tao abrangentequanto a dasorestasAmazonicas.
Key words: Atlantic forests; Brazil; climate; multivariateanalysis; phytogeography; treeora; tropical rain forests, tropical
seasonal forests.
1
Received 13 August 1998; revision accepted 20 March 1999.
2
Correponding author.
794 Oliveira-Filho and Fontes
THE BRAZILIAN ATLANTIC FORESTS ORIGINALLY COV-
ERED an area of ca 1.1 million km
2
, corresponding
to 12 percent of the land surface of the country,
stretchingfor 3300 kmalongtheeastern Brazilian
coast between the latitudes of 6 and 30S (SOS
Mata Atlantica & INPE 1993). Atlantic forests
makeup thesecond largest tropical moist forest area
of South America, after thevast Amazonian domain.
The two forests are separated by the so-called di-
agonal of open formations, a corridor of seasonal
and open formationsthat includesthesemiaridCaa-
tingas of northeastern Brazil, the Cerrado (woody
savanna) of central Brazil, and the Chaco of Para-
guayArgentinaBolivia(Prado & Gibbs1993). At-
lantic forestsareamongthemost threatenedtropical
forests in the world, because their range largely co-
incides with the most populated areas of Brazil
where the settlement of European pioneersand Af-
rican slavesstarted four centuriesago. They noware
reduced to only ca5 percent of their original cover
and most remnants are either small and disturbed
fragments or wider areas sheltered on steep moun-
tain slopes(Viana& Tabanez 1996, SOSMataAt-
lantica1998).
As might be expected for a vast and diversied
vegetation province, the Atlantic forests have been
labeled with several names and their classication
and geographic distribution are still controversial
(Leitao Filho 1987, 1993; Camara 1990). The ac-
ademic discord about the inland extent of Atlantic
forests became a bitter political dispute after they
weregranted strict legal protection by the1988 Bra-
zilian Constitution and declared aBiosphereReserve
by UNESCO. Fixing inland limits to the Atlantic
forests, however, isnoeasy task, sincetheir transition
to the hinterland open formations is very complex
and more or less gradual. This transition can be
classied into threeregionsaccordingtotheadjacent
open formation.
A relatively abrupt transition to the semiarid
Caatingasoccursin northeastern Brazil whereanar-
row strip (50 km) of coastal rain forests is bor-
dered by an equally narrow inland belt of seasonal
semi-deciduousforests, which also occursashinter-
land montane forest enclaves, the brejos (Andrade-
Lima1982). Thetransition between coastal rain for-
ests and Cerrados in southeastern Brazil involves a
much larger extent of semi-deciduous forests that
becomes increasingly wider toward the south and
forms complex mosaics with Cerrado vegetation to
the west (Fig. 1). These semi-deciduousforestsalso
stretch southward along theParana River basin into
eastern Paraguay and northeastern Argentinawhere
they aretransitional totheChaco. In thissubtropical
region, largeextentsof Araucariamixed forestssep-
arate the coastal rain forests and the western semi-
deciduousforests.
The many denitions of Atlantic forests found
in theliteraturecan beclassied intotwomain views
that wecall heresensustricto(ss) and sensulato(sl).
Strictly speaking, Atlantic forests(ss) compriseonly
thecoastal rain forestsup to 300 km inland, where
rainfall islocally boosted by oceanic windsand sea-
sidemountain ranges, particularly in thesouth. This
isthemost traditional and widespread denition of
Atlantic forests since Azevedo (1950) coined the
term Mata Atlantica. For instance, the IBGE clas-
sication system for Brazilian vegetation (Veloso et
al. 1991) includesall Atlantic forestsin thecategory
denserain forest, alongwith most Amazonian for-
ests. Some authors have suggested that the physi-
ognomicsimilarity between Amazonian and Atlantic
rain forestswould show correspondingstrongoris-
tic links reinforced by numerous disjunct species
(Rizzini 1963, Andrade-Lima 1964, Mori et al.
1981).
For Atlantic forests (ss) supporters, the neigh-
boring semi-deciduous forests are a distinct vegeta-
tion formation, often called matasdeplanalto (pla-
teau forests), due to their distribution on the hin-
terland highlands. Cabrera and Willink (1973)
raised thisdifferentiation to thebiogeographiclevel,
merging the semi-deciduous forests of the Parana
River basin with theAraucariamixed foreststo con-
stitute the Paranense Biogeographic Province. The
oraof semi-deciduousforestsisconsideredaseither
transitional between that of rain forests and Cerra-
dos(e.g., Leitao Filho 1987) or part of acontinuum
of forest species distribution that includes central
Brazilian gallery foreststhat eventually linksAtlantic
to Amazonian forests (Oliveira-Filho & Ratter
1995).
The broad view of Atlantic forests (sl) attaches
semi-deciduous forests and Araucaria mixed forests
to coastal rain forests, pushing thelimitsof thedo-
main up to 700 km inward from thecoast (Fernan-
des & Bezerra 1990). This denition largely corre-
sponds to therst attempt to classify Brazilian veg-
etation (von Martius1840), which named all eastern
extra-Amazonian forests Dryads. To avoid misun-
derstandingscaused by thelong-termassociationbe-
tween rain forests and Atlantic forests, authors
adopting thebroad view usually added theadjective
moist to encompassboth rain and semi-deciduous
forests (e.g., Viana et al. 1997). This denition of
Atlantic forests has recently become widespread in
theliterature, although it wascaused lessby achang-
ing scientic approach than by the current policies
Flora and Climate of Atlantic Forests 795
FIGURE1. Map of southeastern Brazil showingthepredominant vegetation formations, adapted fromIBGE(1993),
and Walter climatic diagramsfor selected localities.
of both environmental conservation and research
funding. Sound quantitative assessments of oristic
variation in Atlantic forests (sl), in association with
underlying environmental factors, are important to
understanding the geographic patternsof thesefor-
estsfor both scientic and conservation purposes.
We sought patterns of oristic differentiation
among Atlantic forests (sl) that could be associated
with geographic and climatic variables, and assessed
the oristic links among Atlantic forests (both rain
and semi-deciduous), Amazonian forests, and Cer-
rados. We chose the arboreal component and fo-
cused on southeastern Brazil (transition toCerrados)
dueto dataavailability, and becausewebelievedthat
thethreescenariosof transition to open formations
required separateanalyses. Weaddressed thefollow-
ing questions: (a) is the traditional treatment of
coastal Atlantic rain forestsastheonly Atlantic for-
estsconsistent in termsof oristic composition; i.e.,
how strongly differentiated are semi-deciduousand
rain forests?; (b) to what extent is the tree species
composition of semi-deciduous forests transitional
796 Oliveira-Filho and Fontes
between thoseof rain forestsand Cerrados?; (c) are
coastal Atlantic rain forestscloser to Amazonian rain
forests in terms of oristic composition than they
are to neighboring semi-deciduous forests?; (d) do
both rain and semi-deciduous forests change their
oristic composition with theclimatic variationsoc-
curring within their range?; are rainfall regime and
elevation as important for oristic patterns as they
are for traditional physiognomic classication sys-
tems?; and (e) howaretheabovequestionsanswered
at thespecies, genusand family levels?
Themajor patternsemerging from our analyses
add an important contribution to discussion of the
oristic composition of Atlantic forests. In addition,
the results also contribute information on the pat-
terns of plant species diversity and distribution in
theregion, which isimportant for planningresearch
and conservation policies.
METHODS
FLORISTIC, GEOGRAPHIC, AND CLIMATIC DATA.Wese-
lected from the literature a total of 144 oristic
checklists produced by surveys of the tree ora in
125 areas of the Atlantic Forest (sl) in southeastern
Brazil (contact authors for data). The geographic
range was dened between the eastern Brazilian
coast and 55 longitude, and between 1400 and
2630Slatitude(Fig. 2). Vegetation physiognomies
included coastal rain forestsand semi-deciduousfor-
ests, both coastal and hinterland. Forest areas of
edaphic climax were not included (e.g., those on
coastal sand dunes, swamps, and calcareous out-
crops). Survey methodsvaried widely: most authors
used plot or plotless (PCQ) sampling; only six sur-
veyswerebased on intensivecollectingof plant spec-
imens; and many combined both procedures. Forest
areas were dened arbitrarily within a maximum
rangeof 20 km width and 150 m elevation, includ-
ing sections of large continuous forest tracts (e.g.,
Uba, Prd), assemblagesof forest fragments(e.g., Vic,
Ter), and areasseparated by altitude(e.g., P1-P2-P3,
Itt-Vma). A few areaswith greater elevational range
(e.g., Mci) required consultingtheauthorsabout the
predominant altitude of plant collecting activities.
We obtained the following geographic information
for each area: latitudeand longitudeat thecenter of
thearea, median altitude, and shortest distancefrom
theocean. Wealso extracted from theliteraturethe
annual and monthly meansfor thetemperatureand
rainfall of each forest area or the nearest meteoro-
logical stations. When thesamesourceof thecheck-
list did not provide the climatic records, they were
obtained from DNMet (1992). Most areasrequired
interpolation and/or standard correction for altitude
(Thorntwaite 1948). We obtained themean dura-
tion of thedry season for each areafrom thenum-
ber of days of water shortage given by Walter dia-
grams (Walter 1985). We also calculated a rainfall
distribution ratio from theproportion between the
mean precipitation of the dry (JuneAugust) and
rainy (DecemberFebruary) seasons. Summarized
information for the 125 areas is available from au-
thor upon request.
CLASSIFICATION OF THE FOREST AREAS.Traditional
classication systems for Brazilian vegetation (e.g.,
IBGE System; Veloso et al. 1991) usually classify
forest physiognomiesbased essentially on rainfall re-
gime and temperature, the latter inferred from lati-
tude and altitude. To assess this approach from the
oristic perspective, weclassied the125 forest areas
as either rain forests or semi-deciduous forests, and
divided them into four elevational classes, resulting
in eight main forest formations(Fig. 2). Areasnorth
of 2320S (tropical climates) or above 700 m
(mountain climates) wereclassied asrain forestsin
which thedry season lastsfor 30 daysor less; others
(40160 days) were semi-deciduous forests, follow-
ing Borhidi (1991). Areassituated south of 2320S
and below 700 m (non-montane subtropical cli-
mates) were classied as semi-deciduous and rain
forests with total annual rainfall between 1500 and
2000 mm and 2000 and 3600 mm, respectively
(moist and wet subtropical forests of Holdridge et
al. [1971]). Theelevational categorieswere: lowland,
300 m; submontane, 300700 m; lower montane,
7001100 m; and upper montane, 1100 m. The
classication criteriawerearbitrary and not intended
to represent any objective limit between forest for-
mations; however, thegeographicdistribution of the
classied areas coincided almost perfectly with the
vegetation units of the map produced by IBGE
(1993; compareFigs. 1 and 2).
PREPARATION AND REVISION OF THE DATABASES.We
entered the information from the 125 checklists
onto spreadsheetsusing Microsoft Excel 97 in order
to producetwo databases. Therst consisted of ba-
sic information about each areaforest formation,
geographic coordinates, median altitude, distance
fromtheocean, annual and monthly meansfor tem-
perature and rainfall, mean duration of dry season,
rainfall distribution ratio, and literaturesource. The
second databasewasessentially amatrix of treespe-
ciespresencein the125 forest areas. Beforereaching
itsnal form, theinformation contained in thisda-
Flora and Climate of Atlantic Forests 797
FIGURE 2. Map of southeastern Brazil showing thelocation of the125 areasof Atlantic Forest used in theoristic
analyses. The forest areas are identied by three-letter codes (contact authors for details), and classied aseither rain
or semi-deciduous forest, and into four elevational categories: lowland, 300 m; submontane, 300700 m; lower
montane, 7001100 m; and upper montane, 1100 m.
tabaseunderwent adetailed revision tocheck all spe-
ciesnamescited in thechecklistsfor synonymy and
geographic distribution. Thistask required consult-
ing 337 monographs and revisions, as well as spe-
cialists from the Universities of Sao Paulo, Campi-
nas, Rio de Janeiro, Belo Horizonte, and Bras lia,
and the Botanical Gardens of Rio de Janeiro, Sao
Paulo, Kew, New York, and Missouri. When the
monographs referred to herbarium specimens un-
equivocally collected in any of the 125 areas, the
specieswasadded tothedatabase. Thenal database
contained 2532 species, 520 genera, and 106 fami-
lies. To compare the tree ora of southeastern Bra-
zilian Atlantic foreststo that of Amazonian rain for-
ests and Cerrados, we analyzed two additional da-
tabases. The rst was extracted from Oliveira-Filho
and Ratter (1994) and consisted of 22 areasof terra
rme (upland) Amazonian rain forests that con-
tained 1530 species. Thesecond consisted of 98 ar-
easof Cerrado (Ratter et al. 1996) that included528
species. We also derived databases containing the
number of species per genus and family in each of
the 125 forest areas, as well as in the Amazonian
forestsand Cerrado checklists.
798 Oliveira-Filho and Fontes
MULTIVARIATE ANALYSES.We used canonical corre-
spondence analysis, CCA (ter Braak 1987, 1995)
processed by the program PC-ORD 3.0 (McCune
& Mefford 1997) to investigatetherelationshipsbe-
tween tree ora and geographic and climatic (here-
after geo-climatic) variables of the 125 forest areas.
CCA seekspatternsof datastructurefrom amatrix
of oristic records correlated to a set of variables
provided by an environmental matrix. Weprepared
oristic matricesat threetaxonomic levelsspecies,
genus, and family. The species matrix consisted of
binary data of species presence per forest area and
included the 1002 species present on ve or more
areas. The generic matrix was comprised of the
number of speciesper genusin the125 forest areas
and contained all 520 genera. The familial matrix
consisted of thenumber of speciesper family in the
125 forest areas and incorporated all 106 families.
The patterns emerging from each of these oristic
matrices were related by CCA to 13 variables con-
tained in ageo-climatic matrix. Thesewerethegeo-
graphic variables: (a) latitude; (b) longitude; (c) me-
dian altitude; (d) distance from the ocean and the
climatic variables (e) mean annual temperatureand
mean monthly temperaturesin (f ) July and (g) Jan-
uary; (h) mean temperaturerange(JanuaryJuly); (i)
mean annual rainfall and mean monthly rainfall of
the(j) dry (JuneAugust) and (k) rainy (December
February) seasons; (l) rainfall distribution ratio; and
(m) mean duration of thedry season. After prelim-
inary analyses, six variables were eliminated due to
either high redundancy (variance ination factor
20) or poor correlation (intra-set correlationswith
axes1, 2, or 3 0.4). Theeliminated variableswere:
(a) longitude, highly redundant with distancefrom
the ocean; (b) latitude, poorly correlated with the
axes; and themonthly temperatures(f, g) and season
rainfall (j, k), all redundant with the respectivean-
nual means. TheMonteCarlo permutation test was
performed to assess the signicance of the correla-
tionsfound.
CONDENSED FLORISTIC DATA.Because the CCAs
demonstrated that the forest classication system
adopted was highly consistent, we eventually con-
densed theoristic information contained in theda-
tabases by clustering therecordswithin main forest
formations; however, as the CCAs also detected an
additional northsouth differentiation for rain and
semi-deciduous forests of lower altitudes and an
eastwest pattern for semi-deciduousforests, there-
spectivesubcategoriesweresplit accordingly. A con-
secutive upper level clustering operation, with both
rain and semi-deciduous forests, merged lowland
and submontane forests as low altitude forests,
and lower and upper montane forests as simply
high altitudeforests. Weused thecondensed data
to perform a direct quantitative assessment of the
oristic linksbetween theforest formationsby plot-
ting the number of shared and exclusive species in
Venn diagrams. Wealso performed hierarchical clas-
sications (Kent & Coker 1992) of the condensed
matrices using the program PC-ORD 3.0. Cluster
analysesused Sorensen distancesfor speciespresence
data and euclidian distancesfor generaand families
(number of species as abundance data); thelinkage
method wasgroup average. A TWINSPAN analysis
of the condensed species matrix was performed to
identify indicator species for the forest formations.
Thegeneraand familieswith thehighest number of
speciesin each forest formation wereextracted from
thematrices.
RESULTS
MULTIVARIATE ANALYSES.CCA resultsfor thethree
taxonomic levelsaresummarized in Table1. Eigen-
values were relatively high for species (0.3), indi-
cating considerable species turnover along the gra-
dients summarized in axes 1 and 2. On the other
hand, they were increasingly small for genera and
families (0.2), indicating short gradients (i.e,
there were large proportions of genera and families
occurring throughout the gradients, varying essen-
tially in abundance; ter Braak 1995). The cumula-
tivepercentagevariancesaccounted for by CCAaxes
werealso small, indicating that considerablenoise
remained unexplained. As might be expected for
higher and more widely distributed taxa, the per-
centage variance was higher for families (i.e., the
noise was lower). Small eigenvalues and low ex-
plained variancesarenormal in vegetation dataand
do not impair thesignicanceof taxaenvironment
relations (ter Braak 1988). In fact, Pearsons corre-
lationsfor taxaenvironment wereconsistentlyhigh.
In addition, MonteCarlo permutation testsindicat-
ed that taxadataand geo-climaticvariablesweresig-
nicantly correlated for therst threecanonical axes
of all ordinations.
Therst canonical axisin thespeciesordination
was correlated best with distance from the ocean,
followed by rainfall distribution ratio, altitude and
duration of thedry season (Table2). Also, duration
of thedry season washighly correlated with distance
from the ocean (positively) and rainfall distribution
ratio (negatively). Thisindicatesthat thedatastruc-
turesummarized by therst axisprimarily reected
ageographic gradient based on penetration into the
Flora and Climate of Atlantic Forests 799
TABLE 1. Summaryof canonical correspondenceanalyses(CCAs) andMonteCarlopermutationtest of thetreeoraand
geo-climatic variablesfor 125 areasof theAtlanticForest performed for binarydata of speciespresence, and
for generaandfamiliesusingthenumber of speciesasabundancedata.
Axis 1 Axis 2 Axis 3
Eigenvalue (species)
Cumulative percentagevariance of species presence data
Pearsons correlation for speciesenvironment
Signicanceof speciesenvironment correlation (Monte Carlo test)
0.405
6.4
0.954
0.01
0.290
10.9
0.934
0.01
0.180
13.7
0.829
0.01
Eigenvalue (genera)
Cumulative percentagevariance of genera abundance data
Pearsons correlation for generaenvironment
Signicanceof generaenvironment correlation (Monte Carlo test)
0.200
6.4
0.933
0.01
0.138
10.9
0.917
0.01
0.092
13.8
0.784
0.05
Eigenvalue (families)
Cumulative percentagevariance of families abundancedata
Pearsons correlation for familiesenvironment
Signicanceof familiesenvironment correlation (Monte Carlo test)
0.081
10.5
0.886
0.01
0.046
16.4
0.822
0.01
0.028
20.0
0.741
0.01
continents interior which corresponded to a major
climatic gradient (i.e., an increasingly longer drysea-
son). Altitude, another important geographic gradi-
ent in the rst canonical axis, was highly and neg-
atively correlated with both annual temperatureand
rainfall seasonality. The second canonical axis had
stronger correlations with annual temperature and
temperature range. The CCA for genera showed
correlation patterns very similar to those found for
species, although inverting the relative weight of
somevariableson therst two axes. Unlikethefor-
mer ordinations, the CCA for families related the
distancefrom theocean and rainfall variablesto axis
2, and altitudeand temperaturevariablesto axis1.
The relationship between the geo-climatic vari-
ables and the species composition of the125 forest
areas is illustrated by the CCA biplot in Figure 3.
Theapplication of theforest classication categories
to the ordinated areas helped the interpretation of
the biplot. The dichotomy between rain and semi-
deciduous forestsappearsasa diagonal perpendicu-
lar to thearrow of distancefrom theocean. Toward
the right side of the diagram, the pattern summa-
rized for rain forests (closed symbols) involves de-
creasingaltitudeand temperaturerange, andincreas-
ingannual temperatureand rainfall distributionuni-
formity. Lower and upper montane rain forests are
clearly discriminated at thetop of thediagram. The
areas of low altitude rain forests are found in two
distinct groups: the northern group (bottom-right)
composed of theareasin Bahiaand Esp rito Santo,
and the southern group (top right), which includes
the areas in Rio de Janeiro, Sao Paulo, and Parana.
The rainiest areas of coastal Sao Paulo and Parana,
all with annual rainfall 2000 mm, aremoreclear-
ly discriminated in axis 3 (not shown), which is
strongly correlated with annual rainfall.
Three main patterns are summarized for semi-
deciduousforests(open symbols) on theCCAbiplot
(Fig. 3). The areas of coastal lowland semi-decidu-
ous forests of northern Rio de Janeiro (Cfr, Cgo,
Imb) appear on the bottom right, not far from the
northern lowland rain forests, and arehighly corre-
lated with elevated temperatures. They arefollowed
toward the center left by the submontane semi-de-
ciduous forests of the Rio Doce basin in eastern
Minas Gerais (Prd, Crt, Cng, Vic, Pnv, Imd, Sbr),
all of northern distribution. Most submontanesemi-
deciduousforestsof southern distribution (Sao Pau-
lo and Parana) appear on thecenter-left of thedia-
gram and arefollowed toward theleft by lower and
upper montanesemi-deciduousforests. Thestrongly
seasonal areasof western distribution (Goias, Distri-
to Federal, northern Sao Paulo, and western Minas
Gerais) areaccommodated on thebottomleft of the
diagram and also show an altitudinal gradient.
Ashappened with theCCAfor species, thesame
forest groupsaredistinguished in thebiplotsyielded
by the CCAs for genera and families (Fig. 4). This
made clear that the same geographic and climatic
variables underlie the oristic patterns at all three
taxonomic levels. Theordination of species, genera,
and familiesby CCA isnot shown in Figures3 and
4 becausethebiplotswould becrowded by countless
taxa dots.
ANALYSES OF CONDENSED FLORISTIC INFORMATION.
As they were extracted from oristic checklists for
particular forest areas, the condensed information
must be regarded as a means of quantitatively as-
8
0
0
O
l
i
v
e
i
r
a
-
F
i
l
h
o
a
n
d
F
o
n
t
e
s
TABLE 2. Canonical correspondenceanalyses(CCAs) of thetreeoraandgeo-climaticvariablesfor 125areasof theAtlanticForest performedfor binarydataof speciespresence, and
for generaandfamiliesusingthenumber of speciesasabundancedata: intra-set andinter-set correlationsfor therst twoordinationaxes, andweightedcorrelationmatrix
for thegeo-climaticvariablessupplied. Correlationswithabsolutevalues0.5 areenhancedin bold.
Geo-climatic variables
Intra-set
correlations
Axis 1 Axis 2
Inter-set
correlations
Axis 1 Axis 2
Geo-climatic variables
Altit.
Dist.
ocean
Ann.
temp.
Temp.
range
Ann.
rainf.
Rainf.
distr.
CCA for species
Altitude
Distance from the ocean
Annual temperature
Temperature range
Annual rainfall
Rainfall distribution ratio
Duration of dry season
0.76
0.82
0.42
0.06
0.28
0.81
0.73
0.31
0.50
0.75
0.75
0.43
0.29
0.41
0.72
0.78
0.40
0.06
0.27
0.78
0.70
0.29
0.47
0.70
0.71
0.40
0.27
0.38

0.41
0.77
0.08
0.17
0.62
0.35

0.06
0.35
0.27
0.47
0.72

0.49
0.11
0.40
0.07

0.26
0.03
0.39

0.18
0.44

0.65
CCA for genera:
Altitude
Distance from the ocean
Annual temperature
Temperature range
Annual rainfall
Rainfall distribution ratio
Duration of dry season
0.88
0.53
0.76
0.34
0.09
0.83
0.46
0.16
0.74
0.57
0.44
0.56
0.18
0.68
0.82
0.50
0.71
0.32
0.08
0.77
0.43
0.15
0.68
0.52
0.40
0.51
0.16
0.62

0.42
0.80
0.13
0.15
0.63
0.35

0.02
0.31
0.27
0.48
0.73

0.51
0.11
0.44
0.04

0.26
0.10
0.35

0.17
0.44

0.63
CCA for families:
Altitude
Distance from the ocean
Annual temperature
Temperature range
Annual rainfall
Rainfall distribution ratio
Duration of dry season
0.82
0.10
0.95
0.41
0.26
0.39
0.12
0.49
0.78
0.17
0.03
0.58
0.71
0.78
0.72
0.09
0.84
0.37
0.23
03.5
0.11
0.40
0.66
0.14
0.03
0.48
0.59
0.64

0.42
0.80
0.13
0.15
0.63
0.35

0.02
0.31
0.27
0.48
0.73

0.51
0.11
0.44
0.04

0.26
0.10
0.35

0.17
0.44

0.63
Flora and Climate of Atlantic Forests 801
FIGURE 3. Biplot yielded by canonical correspondence analysis showing the ordination of 125 areas of Atlantic
Forest on the rst two axes, based on the presence of 1002 tree species, and their correlation with geo-climatic
variables, shown asarrows(scaled 2.5 for clarity). Forest areasareidentied by their three-letter codes(Fig. 2) and
forest formationsare indicated by symbols.
sessing the oristic links between the main forest
formations and not as actual gures for number of
species, either total or in common.
The Venn diagrams in Figure 5 give a direct
assessment of the composition and patterns for the
tree ora in some main Atlantic Forest formations.
It isevident from therst diagram that theoraof
rain forests is much richer in tree species than that
of semi-deciduous forests. Rain forests had 31 per-
cent more species than semi-deciduous forests, al-
though they were represented by a smaller number
of surveyed areas. It isalso clear that rain and semi-
deciduous forests shared a high proportion of tree
species: 50 and 66 percent, respectively. Further-
more, the proportion ishigher on thesemi-decidu-
ous side, suggesting that their arboreal ora is, to a
great extent, afraction of themuch richer rain forest
ora.
The second and third Venn diagramsin Figure
5 show thedistribution of thenumber of specieson
two altitudinal classes of both rain and semi-decid-
uousforests, thelatter also showing therelationship
802 Oliveira-Filho and Fontes
FIGURE 4. Biplots yielded by canonical correspon-
dence analyses showing the ordination of 125 areas of
Atlantic Forest on therst two axesbased on thenumber
of tree species per genus (A) and family (B), and their
correlation with geo-climatic variables, shown as arrows
(scaled 2.5 for clarity). Forest areas are identied by
their three-letter codes (Fig. 2) and forest formationsare
indicated by symbols (see Fig. 3).
FIGURE5. Venn diagramsshowingthenumber of tree
speciesshared by 125 areasof Atlanticforest, merged into
main Atlantic Forest formations, 22 areas of Amazonian
Rain Forests, and 98 areas of Cerrado. N number of
checklists; S number of species.
to the western forests. Although the total number
of species decreased with altitudein both cases, the
number of surveyed areas also decreased, and this
may have accentuated the decline. The proportion
of species shared by the two altitudinal classes was
similar: 37 and 35 percent for rain and semi-decid-
uous forests, respectively. Western semi-deciduous
forestsshared 63 and 59 percent of specieswith low
and high altitudesemi-deciduousforests.
Thefourth and fth Venn diagramsstrongly re-
inforcethedifferentiation between thenorthern(Ba-
hia, Esp rito Santo, and eastern Minas Gerais) and
southern (Rio de Janeiro, Sao Paulo and Parana)
sections of both low altitude rain and semi-decidu-
ousforests. They shared only 26 and 33 percent of
thespecies, whilenorthern rain and semi-deciduous
forestsshared 29 percent (sixth diagram) and south-
ern rain and semi-deciduous forests shared 24 per-
cent (seventh diagram) of thespecies. Therefore, the
differentiation between northern and southern
groupsof thesameforest formation wereapparently
of asimilar magnitudein termsof thedifferentiation
between different formations within the same geo-
graphic region.
Cerrados shared a much larger proportion
(55%) of their ora with Atlantic forests than did
Amazonian forests (20%; eighth and ninth dia-
grams). Also, the Cerrado ora was much more
closely related to Atlantic semi-deciduous forests
than to the Atlantic rain forests, while the species
shared with Amazonian forests were distributed
moreevenly.
The classication dendrograms (Fig. 6) show
somecommon patternsfor thethreetaxonomiclev-
els. Most rain and semi-deciduous Atlantic forests
tend to cluster together. The ora of both Cerrado
and Amazonian forests were highly differentiated
Flora and Climate of Atlantic Forests 803
FIGURE 6. Dendrograms from group averaging of
Srensensoristic similarity for speciesand squared euclid-
ian distances for genera and families of the arboreal ora
of 125 areas of Atlantic Forests, merged into nine main
forest formations, 23 areas of Amazonian Forests, and 98
areasof Cerrado. RF rain forests; SF Semi-deciduous
Forests. Distanceon thedendrogram scaleisWishartsob-
jectivefunction, not a raw dissimilarity measure.
from the Atlantic forests at the species and generic
levels. At thefamilial level, however, Cerradomerged
at alow level with montanesemi-deciduousforests,
while Amazonian forests remained drastically sepa-
rated.
Themost important generaand familiesof each
main forest formation are given in Tables 3 and 4,
and aselection of indicator speciesgiven by TWIN-
SPAN for some formations is provided in Table 5.
Thefrequency of thesespeciesor therelativeabun-
dance of some genera and families in a survey may
help classify the forest area and allow inferences
about thelocal climate. Someinterestingtrendscan
beobserved with increasingaltitudein thevemain
Atlantic Forest formations (Table 3). The relative
importancedecreased for somegenerasuch asMar-
lierea, Pouteria, Trichilia, Erythroxylum, Swartzia, Fi-
cus, and Machaerium, and increased for otherssuch
asMiconia, Mollinedia, Nectandra, Myrsine, Tibouch-
ina, Solanum, Ilex, and Gomidesia. The same hap-
pened with somefamilies: therelativeimportanceof
Sapotaceae, Chrysobalanaceae, Rutaceae, and Mor-
aceae decreased, while that of Melastomataceae,
Compositae, Solanaceae, and Myrsinaceaeincreased
with increasing altitude. Western semi-deciduous
forestsapparently combineelementsof bothlowand
high altitude.
Therank of generaand familieswith thehighest
number of species given in Table 4 reinforces the
patterns shown in the dendrograms. A few genera
and familiesshowed marked differencesin their rank
position between Atlantic rain and semi-deciduous
forests (e.g., Pouteria, Gomidesia, Ilex, Machaerium,
Sapotaceae, and Chrysobalanaceae). Many genera
ranked high in both Atlantic forests were less im-
portant either in Cerrado (e.g., Ocotea, Inga, Ficus,
and Trichilia) or Amazonian rain forests (e.g., Myr-
cia, Inga, Machaerium, and Solanum). Likewise, im-
portant familiesin Atlantic forestswereranked in a
lower position in Cerrados(e.g., Lauraceaeand Eu-
phorbiaceae) and Amazonian rain forests (e.g., Me-
lastomataceaeand Myrtaceae). In addition, Cerrado
had particularly important genera (e.g., Byrsonima,
Kielmeyera, Vochysia, and Alibertia) and families(e.g.,
Malpiphiaceae and Vochysiaceae). The same is ob-
served in Amazonian forests for some genera (e.g.,
Licania, Protium, Eschweilera, Virola, and Mouriri)
and families (e.g., Sapotaceae, Chrysobalanaceae,
Burseraceae, and Lecythidaceae). An important link
between Amazonian and Atlantic rain forests, how-
ever, liesin thenumerousspeciesof Sapotaceaeand
Chrysobalanaceae, and in thehigher diversity of Le-
cythidaceae, compared to semi-deciduousforests.
DISCUSSION
Most oristic patterns and their correlation with
geo-climaticvariables, particularlyrainfall seasonality
and temperature, were consistent at all three taxo-
nomic levels, suggesting that thesefactorshavehad
a long time inuence on the evolution and specia-
tion of treetaxain southeastern Brazil. Thiswasnot
a surprise because rainfall and temperature are the
chief factors determining the distribution of the
worlds vegetation formations, and the history of
vegetation and climatein southeastern Brazil during
the Quaternary shows dramatic shifts in both tem-
perature and rainfall regime (Ledru 1993, Ledru et
al. 1998).
8
0
4
O
l
i
v
e
i
r
a
-
F
i
l
h
o
a
n
d
F
o
n
t
e
s
TABLE 3. Generaandfamilieswiththehighest number of species(S) on thetreeorafor 102 areasof theAtlanticForest classiedintovemainforest formations. N number of
areas.
Rain forests(N 48)
Low altitude
(N 27)
S
1475
High altitude
(N 21)
S
1280
Semi-deciduousforests (N 77)
Low altitude
(N 36)
S
1124
High altitude
(N 16)
S
709
Western
(N 25)
S
699
Eugenia
Myrcia
Ocotea
Miconia
Pouteria
Marlierea
Erythroxylum
Inga
Licania
Ficus
70
37
36
35
26
22
21
20
19
19
Eugenia
Miconia
Ocotea
Myrcia
Mollinedia
Inga
Solanum
Gomidesia
Tibouchina
Psychotria
73
48
40
32
27
24
21
17
16
16
Eugenia
Ocotea
Myrcia
Miconia
Machaerium
Ficus
Inga
Casearia
Maytenus
Erythroxylum
44
32
26
25
21
19
15
14
12
12
Miconia
Myrcia
Ocotea
Eugenia
Inga
Ilex
Nectandra
Tabebuia
Tibouchina
Trichilia
26
24
22
16
11
10
10
9
9
9
Miconia
Eugenia
Ficus
Myrcia
Machaerium
Aspidosperma
Inga
Erythroxylum
Casearia
Nectandra
25
19
14
14
12
11
10
9
9
9
Tibouchina
Psychotria
Tabebuia
Trichilia
Gomidesia
Machaerium
Casearia
Guatteria
Cordia
Mollinedia
Nectandra
18
17
16
16
16
15
14
13
13
13
12
Maytenus
Marlierea
Myrsine
Myrceugenia
Pouteria
Casearia
Machaerium
Trichilia
Rudgea
Symplocos
Ficus
15
15
13
13
13
12
12
12
12
11
11
Trichilia
Mollinedia
Solanum
Guatteria
Tabebuia
Swartzia
Pouteria
Cordia
Rudgea
Nectandra
Campomanesia
12
12
12
11
11
10
10
9
9
9
9
Solanum
Machaerium
Ficus
Psychotria
Guatteria
Aspidosperma
Maytenus
Erythroxylum
Casearia
Senna
Mollinedia
9
8
8
8
7
7
7
7
7
7
6
Ocotea
Trichilia
Symplocos
Ilex
Maytenus
Byrsonima
Guapira
Hirtella
Campomanesia
Pouteria
Vochysia
9
8
8
7
7
7
7
6
6
6
6
Myrtaceae
Fabaceae
Rubiaceae
Lauraceae
Melastomataceae
Caesalpiniaceae
Mimosaceae
Sapotaceae
Chrysobalanaceae
Euphorbiaceae
203
91
84
78
76
58
49
49
43
42
Myrtaceae
Melastomataceae
Lauraceae
Rubiaceae
Fabaceae
Mimosaceae
Euphorbiaceae
Caesalpiniaceae
Monimiaceae
Solanaceae
203
82
81
76
58
49
35
35
35
31
Myrtaceae
Fabaceae
Rubiaceae
Lauraceae
Euphorbiaceae
Mimosaceae
Caesalpiniaceae
Moraceae
Melastomataceae
Annonaceae
129
83
66
61
47
46
40
35
34
30
Myrtaceae
Lauraceae
Melastomataceae
Fabaceae
Rubiaceae
Mimosaceae
Euphorbiaceae
Caesalpiniaceae
Asteraceae
Annonaceae
83
44
38
31
31
28
25
23
22
19
Myrtaceae
Fabaceae
Rubiaceae
Melastomataceae
Mimosaceae
Euphorbiaceae
Lauraceae
Caesalpiniaceae
Moraceae
Annonaceae
61
52
35
30
27
26
25
24
19
16
Flora and Climate of Atlantic Forests 805
T
A
B
L
E
3
.
C
o
n
t
i
n
u
e
d
.
R
a
i
n
f
o
r
e
s
t
s
(
N

4
8
)
L
o
w
a
l
t
i
t
u
d
e
(
N

2
7
)
S
1
4
7
5
H
i
g
h
a
l
t
i
t
u
d
e
(
N

2
1
)
S
1
2
8
0
S
e
m
i
-
d
e
c
i
d
u
o
u
s
f
o
r
e
s
t
s
(
N

7
7
)
L
o
w
a
l
t
i
t
u
d
e
(
N

3
6
)
S
1
1
2
4
H
i
g
h
a
l
t
i
t
u
d
e
(
N

1
6
)
S
7
0
9
W
e
s
t
e
r
n
(
N

2
5
)
S
6
9
9
M
o
r
a
c
e
a
e
A
n
n
o
n
a
c
e
a
e
C
l
u
s
i
a
c
e
a
e
M
e
l
i
a
c
e
a
e
A
p
o
c
y
n
a
c
e
a
e
4
1
3
8
2
7
2
6
2
5
A
s
t
e
r
a
c
e
a
e
S
a
p
o
t
a
c
e
a
e
A
n
n
o
n
a
c
e
a
e
C
l
u
s
i
a
c
e
a
e
S
a
p
i
n
d
a
c
e
a
e
3
0
2
9
2
5
2
3
2
3
R
u
t
a
c
e
a
e
F
l
a
c
o
u
r
t
i
a
c
e
a
e
S
a
p
o
t
a
c
e
a
e
S
a
p
i
n
d
a
c
e
a
e
B
i
g
n
o
n
i
a
c
e
a
e
2
8
2
4
2
3
2
1
1
9
B
i
g
n
o
n
i
a
c
e
a
e
C
l
u
s
i
a
c
e
a
e
M
e
l
i
a
c
e
a
e
R
u
t
a
c
e
a
e
S
a
p
i
n
d
a
c
e
a
e
1
4
1
4
1
4
1
4
1
4
A
p
o
c
y
n
a
c
e
a
e
F
l
a
c
o
u
r
t
i
a
c
e
a
e
V
o
c
h
y
s
i
a
c
e
a
e
M
e
l
i
a
c
e
a
e
A
r
e
c
a
c
e
a
e
1
5
1
5
1
5
1
4
1
3
S
a
p
i
n
d
a
c
e
a
e
B
i
g
n
o
n
i
a
c
e
a
e
R
u
t
a
c
e
a
e
F
l
a
c
o
u
r
t
i
a
c
e
a
e
E
r
y
t
h
r
o
x
y
l
a
c
e
a
e
A
r
e
c
a
c
e
a
e
M
o
n
i
m
i
a
c
e
a
e
2
5
2
4
2
3
2
2
2
1
2
0
1
7
F
l
a
c
o
u
r
t
i
a
c
e
a
e
M
e
l
i
a
c
e
a
e
M
y
r
s
i
n
a
c
e
a
e
M
o
r
a
c
e
a
e
V
o
c
h
y
s
i
a
c
e
a
e
B
i
g
n
o
n
i
a
c
e
a
e
C
h
r
y
s
o
b
a
l
a
n
a
c
e
a
e
1
9
1
9
1
9
1
8
1
8
1
7
1
7
M
e
l
i
a
c
e
a
e
S
o
l
a
n
a
c
e
a
e
C
l
u
s
i
a
c
e
a
e
M
o
n
i
m
i
a
c
e
a
e
N
y
c
t
a
g
i
n
a
c
e
a
e
A
p
o
c
y
n
a
c
e
a
e
M
y
r
s
i
n
a
c
e
a
e
1
8
1
8
1
7
1
7
1
7
1
6
1
4
F
l
a
c
o
u
r
t
i
a
c
e
a
e
M
o
r
a
c
e
a
e
S
o
l
a
n
a
c
e
a
e
M
o
n
i
m
i
a
c
e
a
e
V
e
r
b
e
n
a
c
e
a
e
A
q
u
i
f
o
l
i
a
c
e
a
e
V
o
c
h
y
s
i
a
c
e
a
e
1
3
1
3
1
3
1
1
1
1
1
0
1
0
B
i
g
n
o
n
i
a
c
e
a
e
A
s
t
e
r
a
c
e
a
e
C
h
r
y
s
o
b
a
l
a
n
a
c
e
a
e
C
l
u
s
i
a
c
e
a
e
R
u
t
a
c
e
a
e
S
a
p
i
n
d
a
c
e
a
e
A
n
a
c
a
r
d
i
a
c
e
a
e
1
3
1
2
1
2
1
1
1
1
1
1
1
0
The tree species composition of both rain and
semi-deciduousforestsalso washighly inuenced by
altitudeand associated temperaturesat all threetax-
onomic levels, awell-known fact for mountain veg-
etation worldwide (Hugget 1995). Some oristic
patternsfound with increasingaltitudealso coincid-
ed with those cited by Gentry (1995) for Andean
and Central American forests (i.e., the increasing
contribution of Asteraceae, Melastomataceae, and
Solanaceae to the tree ora and the decreaseof Le-
guminosae, with the exception of Inga species).
Many genera strongly correlated with high altitudes
in southeastern Brazil such as Drymis, Hedyosmum,
Weinmannia, Clethra, Podocarpus, Meliosma, Meri-
ania, Ilex, Clusia, Myrsine, Miconia, Prunus, Roupala,
and Trichipteris, are also considered diagnostic of
Neotropical cloud forests (Webster 1995). At the
species level, many listed ashigh altitudeindicators
such as Drymisbrasiliensis, Weinmanniadiscolor, W.
paullinifolia, Podocarpus lambertii, P. sellowii, and
Hedyosmumbrasiliense, are already known to follow
an upper montane distribution pattern along Bra-
zilian mountain ranges(Giulietti & Pirani 1988).
The importance of altitude in the oristic dif-
ferentiation of semi-deciduousforestshasbeen doc-
umented for thestateof Sao Paulo (Saliset al. 1995,
Torres et al. 1997), and southeastern Brazil in gen-
eral (Oliveira-Filho et al. 1994). Occasional frosts
havebeen mentioned by Oliveira-Filho et al. (1994)
as an important factor limiting species distribution
toward both higher elevations and latitudes. Resis-
tanceto frostsisakey factor determiningtreespecies
distribution in Australian forests(Read & Hill 1989,
Read & Hope 1989). The inuence of altitudeon
climate, however, isfar morecomplex than creating
temperature gradients and frost events. Rising ele-
vation also decreasesatmospheric pressure, increases
solar radiation, accelerates air masses, promotes
higher cloudiness, and boostsrainfall (Jones1992).
Thewell established differentiation between rain
and semi-deciduous Atlantic forests based on phys-
iognomy was oristically consistent at all threetax-
onomic levels, and to some extent correlates with
the coastlandhinterland dichotomy. This hasbeen
demonstrated for the state of Sao Paulo (Torres et
al. 1997) wherethereisastrong oristic separation
between coastal hygrophyllous forests (annual
rainfall 2000 mm and no dry season) and hinter-
land mesophyllousforests (annual rainfall ca1400
mm and a dry season). Sao Paulo and Parana are
wherethisdichotomy ismost pronounced in south-
eastern Brazil due to the relatively abrupt vegeta-
tional transition at the Serra do Mar, the coastal
mountain range that helps create two sharply dis-
806 Oliveira-Filho and Fontes
TABLE 4. Genera and families with thehighest number of species(S) on thetreeora of Atlantic rain forests, semi-
deciduousforests, Amazonianrain forests, andCerrados. N number of areas.
Atlantic
rain forests
(N 48)
S
2012
Atlantic
semi-deciduous
forests
(N 77)
S
1533
Amazonian
rain forests
(N 22)
S
1530
Cerrado
(N 98)
S
528
Eugenia
Miconia
Myrcia
Ocotea
Mollinedia
Inga
Tibouchina
Erythroxylum
Marlierea
Pouteria
Gomidesia
Solanum
Psychotria
Licania
Ficus
Machaerium
Trichilia
Maytenus
Rudgea
Tabebuia
Casearia
108
59
57
49
30
28
28
27
27
27
25
25
23
21
21
20
19
18
18
17
16
Eugenia
Miconia
Myrcia
Ocotea
Machaerium
Ficus
Ilex
Erythroxylum
Inga
Casearia
Solanum
Guatteria
Maytenus
Senna
Tibouchina
Mollinedia
Aspidosperma
Nectandra
Trichilia
Campomanesia
Pouteria
59
46
46
37
24
22
20
19
19
15
15
14
14
14
14
14
13
13
13
12
12
Pouteria
Licania
Inga
Protium
Miconia
Swartzia
Ocotea
Eschweilera
Sloanea
Casearia
Eugenia
Ficus
Virola
Duguetia
Mouriri
Cordia
Pourouma
Hirtella
Aniba
Micropholis
Iryanthera
44
38
37
35
26
25
24
19
18
18
18
17
17
15
15
14
14
14
14
14
13
Myrcia
Miconia
Byrsonima
Eugenia
Aspidosperma
Psidium
Kielmeyera
Vochysia
Cordia
Bauhinia
Machaerium
Alibertia
Symplocos
Licania
Erythroxylum
Casearia
Vernonia
Ocotea
Qualea
Myrsine
Solanum
24
20
18
12
11
9
8
8
7
7
7
7
7
6
6
6
5
5
5
5
5
Myrtaceae
Melastomataceae
Rubiaceae
Fabaceae
Lauraceae
Caesalpiniaceae
Mimosaceae
Euphorbiaceae
Sapotaceae
Chrysobalanaceae
Annonaceae
Moraceae
Monimiaceae
Clusiaceae
Solanaceae
Asteraceae
Sapindaceae
Rutaceae
Meliaceae
Bignoniaceae
Erythroxylaceae
Flacourtiaceae
308
117
116
111
106
68
67
53
50
49
48
43
40
39
38
37
33
31
29
28
27
27
Myrtaceae
Fabaceae
Rubiaceae
Lauraceae
Melastomataceae
Mimosaceae
Caesalpiniaceae
Euphorbiaceae
Moraceae
Annonaceae
Rutaceae
Asteraceae
Sapindaceae
Flacourtiaceae
Sapotaceae
Clusiaceae
Solanaceae
Apocynaceae
Bignoniaceae
Nyctaginaceae
Monimiaceae
Myrsinaceae
187
100
84
76
67
57
55
54
38
37
30
29
29
26
26
25
24
23
23
22
21
21
Fabaceae
Sapotaceae
Lauraceae
Mimosaceae
Annonaceae
Caesalpiniaceae
Chrysobalanaceae
Moraceae
Rubiaceae
Burseraceae
Euphorbiaceae
Lecythidaceae
Apocynaceae
Myrtaceae
Arecaceae
Melastomataceae
Myristicaceae
Clusiaceae
Flacourtiaceae
Meliaceae
Sapindaceae
Bombacaceae
93
88
79
76
74
73
68
62
52
51
49
47
46
40
36
35
33
32
32
30
29
27
Myrtaceae
Rubiaceae
Melastomataceae
Caesalpiniaceae
Asteraceae
Fabaceae
Malpighiaceae
Vochysiaceae
Apocynaceae
Annonaceae
Mimosaceae
Clusiaceae
Euphorbiaceae
Anacardiaceae
Arecaceae
Chrysobalanaceae
Verbenaceae
Bignoniaceae
Lauraceae
Boraginaceae
Combretaceae
Myrsinaceae
51
30
27
25
24
24
22
18
17
16
16
13
12
10
10
10
9
8
8
7
7
7
tinct climates. Theseaward sideof theSerrado Mar
has the highest mean annual rainfall (up to 3600
mm) of the entire Atlantic Forest range, while the
inland sidehastypical seasonal climateswith annual
rainfall between 1300 and 1600 mm. Thisreinforc-
estheinaccurateimageof two sharply distinct forest
formations.
Thetransition from rain to semi-deciduousfor-
estsmay begradual and complex, and not necessar-
ily linked to the coastalhinterland climatic gradi-
ents. Coastal climates vary dramatically in south-
eastern Brazil. Annual rainfall declines from south-
ern Sao Paulo toward the north of Rio de Janeiro
state where semi-deciduous forests reach the coast
near Campos dos Goitacazes and give rise to a gap
in rain forest distribution (Fig. 1). Thedrier coastal
climate of this region is caused by thecold oceanic
upwelling of Cabo Frio (Araujo 1997). Coastal rain
Flora and Climate of Atlantic Forests 807
TABLE 5. Selection of tree species associated with major groups of Atlantic Forest formations based on TWINSPAN
classicationof 125 forest areasand1002 species.
Northern lowaltituderain forests: Bactrissetosa, Bathysa nicholsonii, Brosimumguianensis, Byrsonimasericea, Cam-
pomanesia guaviroba, Carpotrochebrasiliensis, Caryocar edule, Caseariaulmifolia, Cedrelaodorata, Chrysophyllumlucen-
tifolium, Cupania emarginata, Dalbergia nigra, Diplotropis incexis, Ecclinusa ramiora, Esenbeckia leiocarpa, Eugenia
brasiliensis, E. moraviana, E. stictosepala, E. sulcata, E. umbelliora, E. verrucosa, Ficusorganensis, Gallesiaintegrifolia,
Geissospermumlaeve, Heliocarpusamericanus, Himatanthuslancifolius, Hortiaarborea, Hydrogaster trinervis, Hymenolob-
iumjaneirense, Ingaedulis, I. agelliformis, I. striata, Ixoragardneriana, Jacarandapuberula, Jacaratiaheptaphylla, Joan-
nesia princeps, Lecythispisonis, Lonchocarpuscampestris, Machaeriumscleroxylum, Marliereaobscura, M. sylvatica, May-
tenus robusta, Melanoxylon brauna, Metrodorea nigra, Micropholis gardnerianum, Mouriri chamissoana, Myrocarpus
frondosus, Ocotea elegans, O. indecora, O. puberula, O. velutina, Paratecomaperoba, Platymisciumoribundum, Pogon-
ophora schomburgkiana, Pourouma guianensis, Protiumwarmingianum, Pseudobombax grandiorum, Pseudopiptadenia
contorta, P. leptostachya, Pterocarpusrohrii, Pterygotabrasiliensis, Rinoreabahiensis, Rustiaformosa, Schizolobiumparahyba,
Simaroubaamara, Sloaneaguianensis, Solanumswartzianum, Sparattospermaleucanthum, Stryphnodendronpulcherrimum,
Swartzia acutifolia, S. aemingii, Tabebuia roseo-alba, Tetrastylidiumgrandifolium, Thyrsodiumspruceanum, Trichilia
casaretti, T. elegans, T. lepidota, Vataireopsisararoba, Virola gardneri, V. ofcinalis, V. oleifera, Vismia baccifera, Vitex
megapotamica, Zanthoxylummonogynum, Zollerniailicifolia.
Southernlowaltituderainforests:Allophyluspetiolulatus, Alseisoribunda, Anibarmula, Astrocaryumaculeatissimum,
Attalea dubia, Balizia pedicellaris, Bathysa meridionalis, Brosimumglazioui, Calyptranthes grandifolia, Chrysophyllum
exuosum, Citronella megaphylla, Coussapoa microcarpa, Cryptocarya moschata, Cupania racemosa, Ecclinusa ramiora,
Eriothecapentaphylla, Eugeniacerasiora, Euterpeedulis, Ficusorganensis, Galipeamultiora, Gomidesiaanacardiaefolia,
G. spectabilis, Guatteria australis, Heisteria silvianii, Hymenolobiumjaneirense, Ilex dumosa, Inga capitata, I. edulis,
Jacaranda puberula, Licaria armeniaca, Lonchocarpusmuehlbergianus, Malouetia arborea, Marlierea suaveolens, M. to-
mentosa, Micropholis crassipedicellata, Mollinedia schottiana, M. uleana, Myrcia pubipetala, Nectandra membranacea,
Neomytranthes glomerata, Nephelea sternbergii, Ocotea brachybotra, O. dispersa, O. divaricata, O. puberula, Parinari
excelsa, Pilocarpus pauciorus, Posoqueria acutifolia, Pouteria caimito, P. venosa, Psychotria carthagenensis, Pterocarpus
rohrri, Quiinaglaziovii, Ruprechtialaxiora, Sclerolobiumdenudatum, Sebastianiabrasiliensis, Stylogyneambigua, Sweetia
fruticosa, Tabebuiaheptaphylla, Talaumaovata, Tetrastylidiumgrandifolium, Tetrorchidiumrubrivenium, Virolagardneri,
V. oleifera.
High altituderain and semi-deciduousforests: Bathysa meridionalis, Byrsonima laxiora, Calyptranthesclusiaefolia,
Caseariaobliqua, Cecropiaglazioui, Cinnamomumglaziovii, Clethrascabra, Clusiacriuva, Connarusregnellii, Cryptocarya
saligna, Daphnopsisfasciculata, Drimysbrasiliensis, Eremanthusincanus, Eugenia blastantha, Euplassa incana, Ficuslu-
schnatiana, F. mexiae, Geonoma schottiana, Gomidesia eriocalyx, G. spectabilis, Gordonia fruticosa, Guatterianigrescens,
Hedyosmumbrasiliensis, Heisteriasilvianii, Hieronymaferruginea, Leucochlorumincuriale, Maytenusglazioviana, M. sal-
icifolia, Meliosmasellowii, Miconiabrunnea, M. chartacea, M. cinnamomifolia, M. pepericarpa, M. theaezans, Mollinedia
argyrogyna, Myrcialaruotteana, Myrsinelancifolia, Nectandragrandiora, N. lanceolata, N. nitidula, N. puberula, Ocotea
brachybotra, O. silvestris, Ouratea semiserrata, Picramnia glazioviana, Pimenta pseudocaryophyllus, Protiumwidgrenii,
Psychotria suterella, Qualea jundiahy, Quiina glaziovii, Schefera angustissima, S. calva, Salacia elliptica, Siphoneugena
widgreniana, Solanumbullatum, Symplocoscelastrinea, Tabebuiachrysotricha, Tibouchinastenocarpa, Trembleyaparviora,
Trichiliaemarginata, Trichipteriscorcovadensis, Vanillosmopsiserythropappa, Vismiabrasiliensis, Weinmaniapauliniaefolia,
Xylosmaciliatifolium.
Eastern lowaltitudesemi-deciduousforests: Acacia glomerosa, A. polyphylla, Albizia niopoides, Aloysiavirgata,
Apuleia leiocarpa, Aspidosperma polyneuron, Astroniumfraxinifolium, Balfourodendron riedelianum, Bastardiopsis
densiora, Chrysophyllumgononocarpum, Copaiferalangsdorfi, Cordiatrichotoma, Cupaniaoblongifolia, Duguetia
lanceolata, Enterolobiumcontortisiliquum, Esenbeckialeiocarpa, Eugeniainvolucrata, E. moraviana, E. sulcata, Ficus
guaranitica, F. insipida, F. obtusifolia, Guareaguidonea, G. kunthiana, Himatanthuslanceifolius, Holocalyxbalansae,
Lonchocarpus cultratus, Machaeriumparaguariensis, Maytenus aquifolia, Metrodorea stipularis, Myrcia multiora,
M. rostrata, M. tomentosa, Nectandra cissiora, Ocotea puberula, Ouratea castaneifolia, Parapiptadeniarigida, Pa-
tagonulaamericana, Picramniasellowii, Pisoniaambigua, Prockiacrucis, Prunussellowii, Pterogynenitens, Siparuna
guianensis, Solanum granuloso-leprosum, Swartzia apetala, Sweetia fruticosa, Tabernaemontana hystrix, Trichilia
casaretti, T. catigua, T. claussenii, T. elegans, T. hirta, Xylopia sericea, Zanthoxylumcaribaeum, Z. riedelianum,
Zeyheriatuberculosa.
Western montaneand submontanesemi-deciduousforests: Acosmiumdasycarpum, Acrocomia aculeata, Aegi-
phila lhotzkiana, Agonandra brasiliensis, Albizia niopoides, Alibertia concolor, A. macrophylla, Anadenantheraper-
egrina, Apeiba tibourbou, Astroniumfraxinifolium, Bastardiopsisdensiora, Callisthenemajor, Dalbergiamiscolob-
ium, Diospyroshispida, Eugenia punicifolia, Faramea cyanea, Genipa americana, Gomidesia lindeniana, Guatteria
sellowiana, Hedyosmumbrasiliense, Inga alba, Ixora warmingii, Lueheapaniculata, Machaeriumacutifolium, Mar-
garitaria nobilis, Miconia chamissois, M. sellowiana, Myrcia tomentosa, Nectandra cissiora, Platypodiumelegans,
Pouteria gardnerii, Protiumspruceanum, Pseudobombax tomentosum, Pseudolmedia guaranitica, Pterogynenitens,
Qualea dichotoma, Rudgea viburnoides, Salacia elliptica, Siparunaguianensis, Siphoneugenadensiora, Styraxcam-
porum, Sweetia fruticosa, Symplocus nitens, Terminalia argentea, Tibouchina candolleana, Virola sebifera, Vismia
guianensis, Xylopia aromatica.
808 Oliveira-Filho and Fontes
TABLE 5. Continued.
Supertramp species: Aegiphila sellowiana, Alchorneaglandulosa, A. triplinervea, Amaiouaguianensis, Andirafraxini-
folia, Aspidospermaparvifolium, Cabraleacanjerana, Calophyllumbrasiliense, Carinianaestrellensis, Caseariadecandra, C.
sylvestris, Cecropia pachystachya, Cedrella ssilis, Celtisiguanaea, Copaifera langsdorfi, Cordia sellowiana, Croton ori-
bundus, Cupania vernalis, Dendropanax cuneatum, Endlicheria paniculata, Erythroxylumcitrifolium, Eugenia orida,
Guapiraopposita, Guareaguidonia, G. macrophylla, Guazumaulmifolia, Hymenaeacourbaril, Ingavera, Jacaratiaspinosa,
Lueheadivaricata, Mabeastulifera, Macluratinctoria, Mataybaelaeagnoides, M. guianensis, Maytenuscommunis, Myrcia
rostrata, Myrciaria oribunda, Myrsineumbellata, Nectandra oppositifolia, Ocotea corymbosa, Pera glabrata, Piptadenia
gonoacantha, Protiumheptaphyllum, Roupalabrasiliensis, Sapiumglandulatum, Soroceabonplandii, Tabebuiaserratifolia,
Tapiriraguianensis, Tremamicrantha, Trichiliacatigua, Zanthoxylumrhoifolium.
forests reappear in Esp rito Santo state, as annual
rainfall increases and seasonality decreases, until
reaching the warm and hyper-humid hylaea of
southern Bahiastate. Asdemonstrated by theoris-
tic analyses, theseclimatic variationscorrelated with
a strong oristic differentiation between northern
and southern coastal rain forests, a fact already de-
tected by Siqueira (1994) and Oliveira-Filho and
Ratter (1995). Thesetwo forest blocksrepresent dif-
ferent and gradual oristic transitions from rain to
semi-deciduous forests as rainfall amountsdecrease,
both endingat theCamposdosGoitacazesgap. The
northern transition occurs in warmer climate and
includes declining mean temperatures toward the
south. The southern transition starts in cooler sub-
tropical climateand includesrisingtemperaturesto-
ward the north. Mean temperature is probably the
chief factor determining the northsouth oristic
differentiation of coastal rain forests, while both
temperature and rainfall regime account for much
of theinternal variation within thetwoforest blocks.
Another important change toward the north is
that the mountain ranges are progressively farther
from the coast and become lower in altitude, par-
ticularly north of Rio Doce. Thisopensspacefor a
wider band of coastal lowlandsknown astabuleiros
that extend from northern Rio deJaneiro to north-
eastern Brazil. Changesin rainfall regimetoward the
interior are more gradual and have a smaller range
in Esp rito Santo, eastern MinasGerais, and south-
ern Bahiathan in Sao Paulo and neighboringstates.
Likewise, the transition between rain and semi-de-
ciduousforestsisrelatively gentler in thenorth. Ad-
ditionally, lowaltitudespenetratedeep into thecon-
tinent along the valleys of the Rio Doce, Mucuri,
and Jequitinhonha, allowing typical lowland rain
forest speciesto expand their distribution towardthe
interior. Thisisknown for many rain forest species,
both Amazonian and Atlantic, that are able to ex-
pand their distribution into areasof strongly season-
al climatesviariverineforests(Oliveira-Filho& Rat-
ter 1995). Thismay explain why theareasof semi-
deciduous forest in the Rio Doce River basin show
strong oristic links with the tabuleiro rain forests
of Esp rito Santo and southern Bahia.
Rainfall seasonality wasapparently moreimpor-
tant than annual rainfall in distinguishing rain and
semi-deciduous forests, and the 30-day duration of
thedry season wasan effectivelimitingcriterionthat
produced a geographic distribution largely coinci-
dent with that of IBGE (1993). Rain forests are
found in areas with annual rainfall as low as 1173
mm (Rio de Janeiro), provided that the rains are
well distributed. Only among subtropical forestsof
low altitudes does annual rainfall prevail over sea-
sonality in separating rain and semi-deciduousfor-
ests, probably becauselow winter temperaturesplay
an additional role in forest deciduousness as sug-
gested by Holdridgeet al. (1971). Increasingrainfall
seasonality with increasing distance from the ocean
also wasaleading factor determiningoristicdiffer-
entiation among hinterland semi-deciduousforests.
The forest enclaves occurring within the Cerrado
domain (e.g., Bras lia), where dry seasons may last
for up to 160 days, were the most differentiated
semi-deciduousforests; they also had greater oristic
links with the Cerrado tree ora. These facts may
reinforcetheviewof rain and semi-deciduousforests
in southeastern Brazil asacontinuumof treespecies
distribution determined basically by rainfall regime
that also leadsto theCerrado treeora, assuggested
by Leitao Filho (1987); however, one must bear in
mind that other important factorsalso intervenein
forestsavanna transitions, particularly re and soil
fertility (Furley et al. 1992).
To a considerableextent, thetreeoraof semi-
deciduous forests is a fraction of the much richer
rain forest ora, and probably iscomposed of species
ableto copewith relatively longer dry seasons. Tree
speciesdiversity ishighly correlated with water con-
sumption and energy uptake, resourcesthat arepar-
titioned among species and limit their number in
forest communities (Hugget 1995). Water shortage
probably playsthechief rolein reducingspeciesrich-
Flora and Climate of Atlantic Forests 809
nessof semi-deciduousforestscompared to rain for-
ests. Thesimpler structureof semi-deciduousforests
also favors a comparatively reduced number of un-
derstory species (Gentry & Emmons 1987). There
wasgreater oristic similarity at all taxonomic levels
between Atlantic rain and semi-deciduous forests
than between any of thelatter and Amazonian rain
forests. Therefore, there is little oristic ground for
viewing Atlantic rain forestsasbeing closer to Am-
azonian rain forests than to their adjacent semi-de-
ciduousforests. Amazonian and Atlantic rain forests
are more similar in physiognomic characteristics
than oristic aspects(Silva& Shepherd 1986). It is
not incorrect to describe rain and semi-deciduous
forests as physiognomic and oristic expressions of
a singlegreat Atlantic Forest domain if theconcept
of theAmazonian forest domain includesclosedrain
forests(both upland and oodplain), open rain for-
ests(monsoon forests), heath forests(campinarana),
and incidentally, semi-deciduous forests (Veloso et
al. 1991). The denition of Atlantic forestsshould
beascomprehensiveasthat of Amazonian forestsin
order to encompassall forest physiognomieseast of
thedry corridor, from northeastern Brazil to south-
ern Brazil, eastern Paraguay, and northeastern Ar-
gentina. Thisaddsnot only thesemi-deciduousfor-
estsbut also thesouthern subtropical Araucariafor-
estsand thenortheastern enclavesof brejo foreststo
theSouth American Atlantic forests.
ACKNOWLEDGMENTS
ATOF thanks the CNPq for research grant no. 301644/
88-8. We were helped during the taxonomic revision of
the database by Carolyn Proenca (Myrtaceae) from the
University of Bras lia; Jose Rubens Pirani (Simaroubaceae
and Rutaceae), Marina Tereza Campos (Rubiaceae), and
Lucia Lohman (Bignoniaceae) from the University of Sao
Paulo; Maria Lucia Kawazaki (Myrtaceae) and Ines Cor-
deiro (Euphorbiaceae) fromtheSao Paulo BotanicGarden;
Carlos Reynel (Zanthoxylum) from theMissouri Botanical
Garden; Washington Marcondes-Ferreira (Aspidosperma)
from the University of Campinas; Douglas Daly (Burser-
aceae) from the New York Botanical Garden; and Gwil
Lewis (Leguminosae), Terry Pennington (Inga and Sapo-
taceae), and Daniela Zappi (Cactaceae and Rubiaceae)
from theRoyal Botanic Gardens, Kew. TerezaSposito, Ro-
sangela Tristao Borem, Maryland Sanchez, Fernando Ped-
roni, Jeanini Felli, Dorothy SueDunn deAraujo, Wagner
Lopes, ArianePeixoto, AlexandreFrancisco daSilva, Maria
Teresa Zugliani Toniato, Bruno Senna Correa, Veronica
Ulup Andersen, Sebastiao do Amaral, and Marcelo Trin-
dadeNascimento helped to obtain checkliststhat weredif-
cult to access. We thank Brenda Mackey and Garry
Hartshorn for their valuablecommentson themanuscript.
LITERATURE CITED
ANDRADE-LIMA, D. 1964. Contribuicao a dinamicada ora do Brasil. Arq. Inst. Cie n. Terra2: 1520.
. 1982. Present-day forest refuges in northeastern Brazil. In G. T. Prance (Ed.). Biological diversication in the
tropics, pp. 220245. Plenum Press, New York, New York.
ARAUJO, D. S. D. 1997. Mata Atlantica: CPD siteSA14, Cabo Frio Region, south-eastern Brazil. InS. D. Davis, V. H.
Heywood, O. Herrera-MacBryde, J. Villa-Lobosand A. C. Hamilton. (Eds.). Centresof plant diversity: aguide
and strategy for their conservation, pp. 373375. World Wildlife Fund and The World Conservation Union,
London, England.
AZEVEDO, A. 1950. Regioesclimato-botanicasdo Brasil. Estudo togeograco eorestal. Anuario Brasileiro deEconomia
Florestal 11: 201232.
BORHIDI, A. 1991. Phytogeography and vegetation ecology of Cuba. Akademiai Kiado, Budapest, Hungary.
CABRERA, A. L., AND A. WILLINK. 1973. Biogeograa de America Latina. Secretaria General de la Organizacion de los
EstadosAmericanos, Washington, D.C.
CAMARA, I. G. 1990. Plano de acao para a Mata Atlantica. Fundacao SOSMata Atlantica, Sao Paulo, Brasil.
DNMET. 1992. Normais climatologicas (19611990). Departamento Nacional deMeteorologia, Ministerio daAgricul-
tura. Bras lia, Brasil.
FERNANDES, A., AND P. BEZERRA. 1990. Estudo togeograco do Brasil. StylusCom., Fortaleza, Brasil. 205 pp.
FURLEY, P. A., J. PROCTOR, AND J. A. RATTER. 1992. Nature and dynamicsof forestsavannaboundaries. Chapman and
Hall, London, England.
GENTRY, A. H. 1995. Patterns of diversity and oristic composition in Neotropical montane forests. In S. P. Churchill,
H. Balslev, E. Forero, and J. L. Luteyn (Eds.). Biodiversity and conservation of Neotropical montane forests.
Neotropical Montane Forest Biodiversity and Conservation Symposium 1, pp. 103126. New York Botanical
Garden, New York, New York.
, AND L. H. EMMONS. 1987. Geographical variation in fertility, phenology, and composition of theunderstory of
Neotropical forests. Biotropica19: 216227.
GIULIETTI, A. M., AND J. R. PIRANI. 1988. Patternsof geographic distribution of somespeciesfromtheEspinhacoRange,
Minas Gerais and Bahia, Brazil. In P. E. Vanzolini and W. R. Heyer (Eds.) Proceedings of a Workshop on
Neotropical Distribution Patterns, pp. 3969. Academia Brasileira de Ciencias, Rio de Janeiro, Brasil.
HOLDRIDGE, L. R., W. C. GRENKE, W. H. HATHEWAY, T. LIANG, AND J. A. TOSI JR. 1971. Forest environmentsin tropical
life zones: a pilot study. Pergamon, Oxford, England.
HUGGET, R. J. 1995. Geoecology, an evolutionary approach. Routledge, London, England.
IBGE. 1993. Mapa de vegetacao do Brasil. Instituto Brasileiro de Geograae Estat stica. Rio de Janeiro, Brasil.
810 Oliveira-Filho and Fontes
JONES, H. G. 1992. Plants and microclimate: a quantitative approach to environmental plant physiology 2. Cambridge
University Press, Cambridge, England.
KENT, M., AND P. COKER. 1992. Vegetation description and analysis: a practical approach. Belhaven Press, London,
England.
LEDRU, M. P. 1993. Late Quaternary environmental and climatic changesin central Brazil. Quat. Res. 39: 9098.
, AND M. L. SALGADO-LABOURIAU, AND M. L. LORSCHEITER. 1998. Vegetation dynamics in southern and central
Brazil during the last 10,000 yr. B. P. Rev. Palaeobot. Palyn. 99: 131142.
LEITA

O FILHO, H. F. 1987. Consideracoes sobre a or stica de orestas tropicaisesub-tropicaisdo Brasil. IPEF 35: 41
46.
. 1993. Ecologiada Mata Atlanticaem Cubatao. Ed. UNESP/UNICAMP, Sao Paulo, Brasil.
MCCUNE, B., AND M. J. MEFFORD. 1997. PC-ORD. Multivariateanalysisof ecological data, version 3.0. MjM Software
Design, Glaneden Beach, Oregon.
MORI, S. A., B. M. BOOM, AND G. T. PRANCE. 1981. Distribution patternsand conservation of eastern Brazilian coastal
forest tree species. Brittonia33: 233245.
OLIVEIRA-FILHO, A. T., AND J. A. RATTER. 1994. Database: woody ora of 106 forest areas of eastern tropical South
America. Royal Botanic Garden Edinburg, Edinburgh, Scotland.
, AND . 1995. A study of theorigin of central Brazilian forestsby theanalysisof plant speciesdistribution
patterns. Edinb. J. Bot. 52: 141194.
, E. A. VILELA, M. L. GAVILANES, AND D. A. CARVALHO. 1994. Comparison of the woody ora and soils of six
areas of montanesemi-deciduousforest in southern MinasGerais, Brazil. Edinb. J. Bot. 51: 355389.
PRADO, D. E., AND P. E. GIBBS. 1993. Patternsof speciesdistribution in thedry seasonal forestsof South America. Ann.
Mo. Bot. Gard. 80: 902927.
RATTER, J. A., S. BRIDGEWATER, R. ATKINSON, AND J. F. RIBEIRO. 1996. Analysis of the oristic composition of the
Brazilian cerrado vegetation. Edinb. J. Bot. 53: 153180.
READ, J., AND R. S. HILL. 1989. Theresponseof someAustralian temperaterain forest treespeciestofreezingtemperatures
and its biological signicance. J. Biogeogr. 16: 2127.
, AND G. S. HOPE. 1989. Foliar resistance of some evergreeen tropical and extratropical Australasian Nothofagus
species. Aust. J. Bot. 37: 361373.
RIZZINI, C. T. 1963. Nota previa sobre a divisao togeogracado Brasil. Revta. Brasil. Geogr. 25: 164.
SALIS, S. M., G. J. SHEPHERD, AND C. A. JOLY. 1995. Floristic comparison of mesophytic semi-deciduous forests of the
interior of thestateof Sao Paulo, southeast Brazil. Vegetatio 119: 155164.
SILVA, A. F., AND G. J. SHEPHERD. 1986. Comparacoes or sticas entre algumas matas brasileiras utilizando analise de
agrupamento. Revta. Brasil. Bot. 9: 8186.
SIQUEIRA, M. F. 1994. Analise or stica e ordenacao de especies arboreas da Mata Atlantica atraves de dados binarios.
MSc. thesis, UniversidadeEstadual de Campinas, Campinas, Brasil.
SOS MATA ATLA

NTICA. 1998. Atlas da evolucao dos remanescentes orestais e ecossistemas associados no dom nio da
Mata Atlanticano per odo 19901995. Fundacao SOSMata Atlantica, Sao Paulo, Brasil.
, AND INPE. 1993. Evolucao dosremanescentesorestaiseecossistemasassociadosdo dom nio daMataAtlantica.
Fundacao SOSMata Atlantica, Instituto Nacional de PesquisasEspaciais, Sao Paulo, Brasil.
TER BRAAK, C. J. F. 1987. The analysis of vegetationenvironment relationships by canonical correspondence analysis.
Vegetatio 69: 6977.
. 1988. CANOCOa FORTRAN program for canonical community ordination by (partial) (detrended) (ca-
nonical) correspondenceanalysis, principal componentsanalysis, and redundancy analysis, version 2.1. Technical
report LWA-88-02, TNOInstituteof Applied Computer Science, Wageningen, Germany.
. 1995. Ordination. In R. H. G. Jongman, C. J. F. ter Braak, and O. F. R. van Tongeren. Data analysis in
community and landscapeecology, pp. 91173. CambridgeUniversity Press, Cambridge, England.
THORNTWAITE, C. W. 1948. An approach toward a rational classication of climate. Geogr. Rev. 38: 5594.
TORRES, R. B., F. R. MARTINS, AND L. S. K. GOUVEA. 1997. Climate, soil, and tree ora relationships in forests in the
stateof Sao Paulo, southeastern Brazil. Revta. Brasil. Bot. 20: 4149.
VELOSO, H. P., A. L. R. RANGEL FILHO, AND J. C. A. LIMA. 1991. Classicacao da vegetacao brasileira, adaptada a um
sistema universal. Instituto Brasileiro de Geograae Estat stica, Rio de Janeiro, Brasil.
VIANA, V. M., AND A. A. J. TABANEZ. 1996. Biology and conservation of forest fragmentsin theBrazilian Atlantic moist
forest. In J. Schelhas and R. Greenberg (Eds.). Forest patchesin tropical landscapes, pp. 151167. Island Press,
Washington, DC.
, , AND J. L. F. BATISTA. 1997. Dynamics and restoration of forest fragments in the Brazilian Atlantic
moist forest. In W. F. Laurance and R. O. Bierregaard (Eds.). Tropical forest remnantsEcology, management,
and conservation of fragmented communities, pp. 351365. TheUniversity of Chicago Press, Chicago, Illinois.
VON MARTIUS, C. F. P. 1840. Tabulphysiognomie. Brasiliregionis iconibus express. In C. F. P. von Martius, S.
Endlicher, A. G. Eichler, and J. Urban (Eds.). Florabrasiliensis, vol. 1, pp. 1110, Tomus1. Lipsaeaapud Frid.
Fleischer in Comm., Munich, Germany.
WALTER, H. 1985. Vegetation of the earth and ecological systems of the geo-biosphere. Third edition Springer-Verlag,
Berlin, Germany.
WEBSTER, G. L. 1995. The panorama of Neotropical cloud forests. In S. P. Churchill, H. Balslev, E. Forero, and J. L.
Luteyn (Eds.). Biodiversity and conservation of Neotropical montane forests. Neotropical MontaneForest Bio-
diversity and Conservation Symposium 1, pp. 5377. New York Botanical Garden, New York, New York.