universidade estadual de campinas instituto de...
TRANSCRIPT
Classified - Internal use
UNIVERSIDADE ESTADUAL DE CAMPINAS
INSTITUTO DE BIOLOGIA
IVAN JEFERSON SAMPAIO DIOGO
BIOGEOGRAPHY AND DIVERSITY OF HUMID MOUTAIN
FORESTS IN NORTHEASTERN, BRAZIL
BIOGEOGRAFIA E DIVERSIDADE DE FLORESTAS
SERRANAS ÚMIDAS DO NORDESTE DO BRASIL
CAMPINAS
2017
Classified - Internal use
IVAN JEFERSON SAMPAIO DIOGO
BIOGEOGRAPHY AND DIVERSITY OF HUMID MOUTAIN FORESTS IN NORTHEASTERN, BRAZIL
BIOGEOGRAFIA E DIVERSIDADE DE FLORESTAS SERRANAS ÚMIDAS DO NORDESTE DO BRASIL
Thesis presented to the Institute of Biology of the University of Campinas in partial fulfillment of the requirements for the degree of Doctor in Plant Biology.
Tese apresentada ao Instituto de Biologia da Universidade Estadual de Campinas como parte dos requisitos exigidos para a obtenção do Título de Doutor em Biologia Vegetal.
Orientador: FLAVIO ANTONIO MAËS DOS SANTOS Co-Orientador: ITAYGUARA RIBEIRO DA COSTA
CAMPINAS
2017
ESTE ARQUIVO DIGITAL CORRESPONDE À
VERSÃO FINAL DA TESE DEFENDIDA PELO
ALUNO IVAN JEFERSON SAMPAIO DIOGO E
ORIENTADA PELO PROF. DR. FLAVIO
ANTONIO MAËS DOS SANTOS
Classified - Internal use
Agência(s) de fomento e nº(s) de processo(s): CNPq, 140822/2013-5; CNPq, 234983/2014-0 ORCID: http://orcid.org/0000-0002-5440-9851
Ficha catalográfica Universidade Estadual de Campinas
Biblioteca do Instituto de Biologia Mara Janaina de Oliveira - CRB 8/6972
Diogo, Ivan Jeferson Sampaio, 1988- D621b DioBiogeography and diversity of humid mountain forests in Northeastern,
Brazil / Ivan Jeferson Sampaio Diogo. – Campinas, SP : [s.n.], 2017.
DioOrientador: Flavio Antonio Maës dos Santos. DioCoorientador: Itayguara Ribeiro da Costa. DioTese (doutorado) – Universidade Estadual de Campinas, Instituto de
Biologia.
Dio1. Topografia. 2. Carbono. 3. Zoocoria. 4. Pólen. 5. Biodiversidade -
Brasil, Nordeste. I. Santos, Flavio Antonio Maës dos, 1958-. II. Costa,
Itayguara Ribeiro. III. Universidade Estadual de Campinas. Instituto de
Biologia. IV. Título.
Informações para Biblioteca Digital
Título em outro idioma: Biogeografia e diversidade de florestas serranas úmidas
do Nordeste do Brasil Palavras-chave em inglês: Topography Carbon Zoochory Pollen Biodiversity - Brazil, Northeast Área de concentração: Biologia Vegetal Titulação: Doutor em Biologia Vegetal Banca examinadora: Flavio Antonio Maës dos Santos [Orientador] Leonardo Dias Meireles Pedro Vasconcellos Eisenlohr Ingrid Koch Gustavo Hiroaki Shimizu Data de defesa: 22-11-2017 Programa de Pós-Graduação: Biologia Vegetal
4
Classified - Internal use
Campinas, 22 de novembro de 2017.
COMISSÃO EXAMINADORA
Prof. Dr. Flavio Antonio Maës dos Santos
Prof. Dr. Leonardo Dias Meireles
Prof. Dr. Pedro Vasconcelos Eisenlohr
Prof.(a). Dr.(a). Ingrid Koch
Dr. Gustavo Hiroaki Shimizu Os membros da Comissão Examinadora acima assinaram a Ata de defesa, que se encontra no processo de vida acadêmica do aluno.
5
Classified - Internal use
AGRADECIMENTOS
Não é nada fácil começar, cursar e (principalmente!) terminar um doutorado. A gente encontra um bilhão
de percalços nessa estrada. No entanto, conhecemos e reconhecemos pessoas maravilhosas que nos dão
força e carinho na trajetória (e que trajetória!). Tentei estabelecer uma regra cronológica de
agradecimento, embora as lembranças tenham me feito choramingar um pouco.
Aos meus pais, Sr. Diogo e Sra. Cristina, por todos os ensinamentos, dedicação e apoio
que me deram durante toda a minha vida acadêmica. Sem vocês não seria a pessoa que
sou hoje, suas virtudes me fizeram um homem melhor. Obrigado, pois mesmo sem
acreditar, vocês me permitiram voar. Ainda aqui, a minha irmã, pela companhia, força
e desejos de sucesso.
Sem dúvida, uma das pessoas que mais me estimulou durante todo o processo de
doutorado, foi o meu eterno orientador, Itayguara Ribeiro. A você, pela força,
compreensão, discussões, trabalhos, tabelas, exsicatas, campos, estufas, identificações,
envio de folhas, aquelas cervejinhas e conversas aleatórias, muitíssimo obrigado por
manter viva essa chama em mim.
Ao meu orientador de doutorado, Flavio Mäes, por ter me recebido de braços abertos na
UNICAMP. Por aceitar um novo desafio científico e estar sempre disponível para
conversas e questionamentos.
À Karin Santos, porque sem seu apoio de campo e amizade, eu não conseguiria ter feito
esse doutorado. Além do abrigo em Estocolmo. Muito obrigado!
Ao CNPq, pela bolsa de doutorado e doutorado sanduíche concedidas.
Aos docentes e pesquisadores afiliados à UNICAMP, Jorge Tamashiro, Fernando
Roberto Martins, Simone Aparecida Vieira, Ricardo Ribeiro Rodrigues, Luciana
Alves, Valéria Martins, pelas contribuições e ensinamentos durante o curso.
Aos meus colegas de curso e laboratório, Carol Signori, Ana Liboni, Cris Yuri, Raquel
Mourura, Fernanda Barros, Lu, Anninha, Cinthia, Júnior, Isa, Azul, Gus, João,
Nathi, Rafa Guimarães, Caio, Thaís, Magda, pela companhia, ajudas acadêmicas,
farras, cervejas, churras e IFCH. Em especial à Fernanda Cabral, por sempre me apoiar
e acreditar em mim, me ajudando sempre que possível, pelos vinhos, conversas e o amor
(até na Suécia); e à Gigi, por me auxiliar nas identificações, sempre dizer palavras lindas
de carinho e cozinhar aquele maravilhoso cheesecake.
Aos meus companheiros da rep nuvem, em especial Ruan Mariano e Rafhael Parintins,
por me acolherem super bem, darem as melhores festas, compartilhar todas as refeições,
conversar besteira e me fazerem me sentir em casa. Às minhas companheiras da melhor
casinha, Mônica, Thaís pelas pizzas, festinhas, composteira, e melhor uso do quintal de
casa; em especial, à Carol Potas, pelas lindas lembranças morando juntos, os vinhos
tomados, os apoios dados, tristezas e alegrias compartilhas, os papos cabeça e a grande
amizade.
Aos meus amigos de Campinas, Pira e Rafael, por me apresentar aos locais bonitos da
cidade e sempre me levarem pra beber com o objetivo de esquecer um pouco do
doutorado. Magnum, Rafa Guimarães e Neto, pelas lindas sexta-feiras pelas quais eu
6
Classified - Internal use
esperava ansiosamente. Vizinho e Antonielle pela companhia maravilhosa, festinhas
usuais e conversas mais loucas. Abel, pela amizade inesperada e acolhida tão boa.
Ao prof. Dr. Alexandre Antonelli, por me fazer se sentir tão bem acolhido em uma
instituição como a Universidade de Gotemburgo. Agradeço a oportunidade, o empenho,
a ajuda e, principalmente, o fato de acreditar em mim.
To the AntonelliLab team, for being so lovely and kind with me in the new life abroad.
In special, to Anna and Ylva, who helped me a lot in different ways. To Alex Zizka,
Tobi, Eric, Romina, Fernanda, Thaís, Bruno, Angela, Kemal, Victor and Josué for
receive me in a such a nice group (with discussion and parties). To Božo, for being more
than a lab mate, but a friend for the entire life. To my special Swedish friends, Mia, Juan
and Nicklas, for the support out of the university and the great birthday celebration. In
special to Camila, Victor Gonçalez and Daniela to show me the meaning of partnership,
friendship and happiness in other country.
Aos meus amigos cearenses que sempre me deram apoio nessa conquista, Lilian, por
sempre me encorajar com palavras de carinho e positividade quando eu mais precisava.
Fernando, por estar sempre ao meu lado e não largar essa amizade ao ouvir as lamúrias
da vida acadêmica. Ao Alexandre Holanda, por compartilhar as amarguras do
doutorado, mas também as alegrias e os artigos. Ao Hilton Galvão, por sempre me
colocar pra cima e ajudar com piadinhas infames e por Copenhagen. Ao Luiz Filho, por
ser essa fonte de doçura e força. Ao Vinícius Bonjerbar, por ser mais do que
companheiro nos meus momentos tensos quando estive distante. Ao paraibano mais fofo,
Marcelo Ramos, por me dar força e alegria quando mais precisei, pelas horas de
conversas afáveis e pelo inigualável sentimento desenvolvido.
Muito obrigado, de coração!
7
Classified - Internal use
RESUMO
Os eventos de glaciação e deglaciação durante o Pleistoceno aceleraram a disjunção das
florestas úmidas da América do Sul, ocasionando retração e expansão de formações
vegetais e alta diversidade biológica dos Neotrópicos. As florestas serranas localizadas
no semiárido nordestino são consideradas um dos ambientes mais sensíveis às variações
climáticas e uma das melhores áreas para estudar a dinâmica e os efeitos da (de)glaciação
mundial. As perguntas norteadoras desse trabalho foram: 1) Qual a composição florística,
diversidade e como se estrutura a Serra de Maranguape?, 2) Como a flora da Serra de
Maranguape se relaciona com as outras fitofisionomias do estado do Ceará?, 3) Qual a
composição palinológica das Serras de Maranguape, Baturité e Pacatuba?, 4) Como a
distribuição da vegetação está relacionada com a chuva polínica atual?, 5) Qual a relação
entre as florestas serranas úmidas do Noredeste brasileiro, 6) Quais são e como as
variáveis ambientais explicam a distribuição de espécies em escalas local e regional?, 7)
Como essas florestas estão relacionadas com os outros domínios brasileiros? e 8) Estas
florestas podem ser classificadas como uma única unidade florística? A Serra de
Maranguape possui uma clara distinção florística entre barlavento, sotavento e topo, com
a prevalência de Myrtaceae em locais mais úmidos e Fabaceae nos mais secos. Diferença
que também foi encontrada nas florestas de Baturité e Pacatuba, tanto na vegetação
quanto na chuva polínica. Ao compararmos a chuva polínica com a vegetação atual, o
predomínio de Myrtaceae é característico dessas áreas, no entanto, observamos pólens
que não estão presentes na vegetação atual e vice-versa. As áreas próximas do topo
apresentam características de florestas nebulares e condições edáficas distintas. Sotavento
é a área mais rica e diversa, com uma flora distinta e compartilhada com as diversas
fitofisionomias locais, entretanto, sua composição varia significativamente de uma serra
pra outra. Analisando as 17 florestas serranas úmidas nordestinas, verificamos uma
distinção clara entre as florestas mais ao norte (Ceará) e as mais ao sul (Pernambuco,
Paraíba e Sergipe). Temperatura e pluviosidade inluenciam a distribuição local de
espécies e pólen nas florestas serranas. Por outro lado, quando consideramos a escala
regional, verificamos que fatores edáficos (carbono orgânico) e ecológicos (zoocoria) são
responsáveis pela distribuição de espécies nas florestas serranas nordestinas. As florestas
serranas nordestinas são uma bioregião única e distinta com características próprias e
espécies indicadoras, como Guettarda angelica e Manilkara rufula.
Palavras-chave: temperatura, precipitação, sotavento, carbono orgânico, zoocoria,
pólen, bioregião.
8
Classified - Internal use
ABSTRACT
The events of glaciation and deglaciation during the Pleistocene accelerated the the South
American rainforest disjunction, leading to the retraction and expansion of plant
formations and Neotropics’ high biological diversity. The semi-arid montane forests are
greatly affected by climatic variations, thus, one of the best areas to study the dynamics
and effects of the (de)glaciation. As guiding questions of this study: 1) What is the
floristic composition, diversity, and structure of Serra de Maranguape? 2) How is the flora
of Serra de Maranguape related to other phytophysiognomies in the Ceará state? 3) What
is the palynological composition of the Serra de Maranguape, Baturité and Pacatuba? 4)
How is the distribution of vegetation related to the current pollen rain? 5) What is the
relationship between the humid mountain forests of the Brazilian Northeast? 6) Which
are an how the environmental variables explain the distribution of species at local and
regional scales? 7) How those mountain forests are related to the other domains? and 8)
Do they represent a single floristic unit? The Serra de Maranguape has a clear floristic
distinction among windward, leeward and top, with a prevalence of Myrtaceae in humid
places and Fabaceae in the driest. This difference was also found in Baturité and Pacatuba
forests, both in the vegetation and in the pollen rain. When we compared the pollen rain
with current vegetation, the predominance of Myrtaceae is a clear characteristic of these
areas, however, we observe pollen that is not present in the current vegetation and vice
versa. The areas near to the top have nebular forests characteristics and different edaphic
conditions, with pollen and species of heliophilous and pioneer plants (Alchornea,
Miconia and Clusia). Leeward is the richest and most diverse area with a distinct flora,
which is shared with the various local phytophysiognomies, however, its composition
varies significantly from one mountain range to another. Analyzing the 18 northeastern
mountain forests, we found a clear distinction between the northernmost forests (Ceará)
and the southern (Pernambuco, Paraíba e Sergipe). Temperature and rainfall affect the
species and pollen local distribution in the mountain forests of Ceará. On the other hand,
when we consider the regional scale, we verified that edaphic (organic carbon) and
ecological factors (zoochory) are responsible for the distribution of species in the
northeastern mountain forests. The Brazilian Northeastern Mountain Forests are a distinct
and unique bioregion with its own chacteristic and indicator species, such Guettarda
angelica and Manilkara rufula.
Key words: temperature, precipitation, leeward, organic carbon, zoochory, pollen rain,
bioregion.
9
Classified - Internal use
SUMÁRIO
INTRODUÇÃO GERAL..……......................................................................................10
CHAPTER I - VEGETATION STRUCTURE, SPECIES COMPOSITION AND
DISTRIBUTION IN A TOPOGRAPHIC GRADIENT OF A MOUNTAIN FOREST,
NORTHEASTERN BRAZIL ……………......................................................................18
CHAPTER II- POLLEN-BASED CHARACTERIZATION OF MONTANE FOREST
TYPES IN NORTH-EASTERN BRAZIL.......................................................................47
CHAPTER III- DISPERSAL AND EDAPHIC FACTORS DRIVING PLANT SPECIES
COMPOSITION IN MOUNTAIN FORESTS IN A SEMIARID REGION
OF BRAZIL.....................................................................................................................77
CHAPTER IV- ELEVATIONAL GRADIENTS OF MOIST FOREST IN BRAZILIAN
SEMIARID: A DISTINCT BIOREGION……………………..…………………….....97
CONSIDERAÇÕES FINAIS ........................................................................................116
RFERÊNCIAS...............................................................................................................118
ANEXOS.......................................................................................................................163
10
Classified - Internal use
INTRODUÇÃO GERAL
Períodos glaciais e interglaciais e o surgimento de zonas de refúgio
A distribuição da vegetação por paisagens ecológicas depende em grande parte da
variação no clima, demonstrada pela translocação de espécies por diferentes gradientes
topográficos e geográficos, um processo consolidado durante o último evento de
deglaciação (Betancourt et al., 1990). Os períodos glaciais e interglaciais que se
alternaram sobre todas as regiões da Terra durante o Terciário (desde o Eoceno, há 23
milhões de ano) e Quaternário (Pleistoceno Tardio e Holoceno, há 1,6 milhão de anos)
causaram modificações no nível dos oceanos, na temperatura e na quantidade de gelo das
calotas polares (Bigarella et al., 1975; Bigarella & Andrade-Lima, 1982; Allen &
Breshears, 1998).
Os eventos de acréscimo de precipitação foram intensos e sincrônicos por
apresentarem um padrão de longas fases áridas intercaladas por fases úmidas (Hoorn et
al., 2010) e retração e expansão de formações vegetais (Whitmore & Prance, 1987; Prado
& Gibbs, 1993; Oliveira-Filho & Ratter, 1995). Esses eventos paleoclimáticos e
geomorfológicos foram responsáveis pela geração de alta diversidade biológica e
endemismos nos Neotrópicos (Antonelli & Sanmartín, 2011).
Estudos que buscam explicar a diversidade biológica neotropical ainda estão em
fase inicial (Mello-Martins, 2011). Historicamente, muito mais atenção tem sido dedicada
à diversificação da região amazônica por meio da teoria dos refúgios (Haffer, 1969), que
sugere que as flutuações climáticas do Pleistoceno fizeram com que as formações
florestais tenham sido fragmentadas por formações secas e abertas, como as savanas.
Dessa forma, os fragmentos florestais remanescentes seriam isolados e, nesses refúgios,
11
Classified - Internal use
novas espécies surgiriam pelo isolamento de espécies ancestrais amplamente distribuídas
(especiação por vicariância).
Embora a teoria dos refúgios tenha recebido um amplo apoio inicialmente
(Vanzolini & Williams, 1981), ela não foi corroborada por dados empíricos na área da
Amazônia, com evidências congruentes e diferentes hipóteses de diversificação (da Silva
& Patton, 1998; Marks et al., 2002; Aleixo, 2004; Zink, Klicka & Barber, 2004; Rull,
2008). Por outro lado, estudos paleoecológicos, de modelagem e filogeográficos sugerem
a ocorrência de refúgios florestais do Pleistoceno para outras áreas tropicais (Hugall et
al., 2002), como a Floresta Atlântica, que foi fragmentada por áreas abertas, gerando
manchas florestais isoladas em toda sua extensão (Behling & Lichte, 1997; Ledru et al.,
1998; Behling & Negrelle, 2001; Behling, 2002).
Diferentemente da disposição interiorana dos refúgios amazônicos, as áreas de
refúgio na Mata Atlântica geralmente ocorrem próximas do litoral (Ab’sáber, 1977;
Brown, 1982; Prance, 1982; Haffer, 1987; Sant'Anna-Neto & Nery, 2005; Carnaval &
Moritz, 2008; Mello-Martins, 2011), relacionadas às encostas das serras, onde existiram
condições para a permanência de florestas pelas massas de ar úmidas vindas do Oceano
Atlântico.
Carnaval & Moritz (2008) sugerem um modelo, baseado em estudos de fauna, que
mostra que houve uma grande mancha florestal contínua entre o Rio Doce e o Rio São
Francisco, que foi fragmentada em diversas manchas de refúgio, correspondendo aos
centros atuais de endemismo de vários táxons e de padrões de diversidade do DNA
mitocondrial de algumas espécies. Em diversos grupos taxonômicos, têm sido
encontrados padrões filogeográficos que corroboram a hipótese de refúgios do
12
Classified - Internal use
Pleistoceno na Floresta Atlântica (Cabanne et al., 2008; Carnaval et al., 2009; D’Horta et
al., 2011; Pavan et al., 2011).
Zonas ecotonais e o Nordeste Brasileiro
As respostas da vegetação às variações climáticas são mais rápidas e extremas nos
limites entre diferentes ecossistemas, denominadas ecótonos (Gosz, 1992; Risser, 1995).
Nas últimas décadas, diversos estudos abordaram eventos de glaciação e deglaciação
durante o Terciário e Quaternário em ambientes ecotonais no mundo: em regiões
semiáridas mexicanas (Allen & Breshears, 1998), regiões alpinas nos EUA (Walsh et al.,
1994), chaco boliviano (Latrubesse et al., 2013), sudeste da Tasmânia (Macphail, 1979),
bacias montanhosas na Suíça (Köplin et al., 2013) e montanhas subalpinas bulgarianas
(Tonkov et al., 2012).
No Brasil, os estudos de ecótonos são mais recentes e estão em constante processo
de desenvolvimento (Behling, 2001; Behling & Costa, 2001; Behling et al., 2001;
Latrubesse & Franzinelli, 2002; Vidotto et al., 2007; Santos et al., 2007; Diogo et al.,
2013). Os ecótonos localizados no semiárido são considerados um dos mais sensíveis às
variações climáticas (Allen & Breshears, 1998). Desse modo, uma das melhores áreas
para estudar a dinâmica e efeitos da glaciação é o Nordeste brasileiro. A cobertura vegetal
dessa região é modulada pelo clima (Juaréz & Liu, 2001; Ferreira et al., 2001), sendo a
caatinga seu tipo de vegetação predominante.
A configuração atual do relevo do Nordeste brasileiro deriva das forças de
distensão oriundas da separação das placas tectônicas da América do Sul e da África, que
provocaram o soerguimento de bacias sedimentares e de maciços cristalinos residuais
durante o período Cretáceo (Claudino-Sales; Peulvast, 2007). Devido à presença de
fósseis característicos (mamíferos pleistocênicos associados a climas úmidos e registros
13
Classified - Internal use
polínicos de plantas ombrófilas), rochas calcárias e datações de recifes marinhos,
diferentes autores afirmam a existência de períodos de alta umidade no Terciário e
Quaternário em regiões que atualmente são consideradas semiáridas (Arz et al., 1999;
DeOliveira et al., 1999; Behling et al., 2000; Auler et al., 2004; Wang et al., 2004;
Ximenes, 2008).
Floresta Amazônica, Floresta Atlântica e as Florestas Serranas Nordestinas
As flutuações no clima foram supostamente suficientes para promover a retração
e expansão da Floresta Amazônica e da Mata Atlântica, gerando a hipótese de que elas
podem ter sido conectadas, permitindo o intercâmbio de espécies (Cole, 1960; Andrade-
Lima, 1966, 1982; Ab’saber, 1977; Behling et al., 2000; Auler et al., 2004, Wang et al.,
2004). Após intervalos entre períodos glaciais e interglaciais, as florestas tropicais
atlântica e amazônica retornaram a sua distribuição original, gerando ilhas vegetacionais
(Andrade-Lima, 1982) que permaneceram em locais de mesoclima favorável.
A Floresta Amazônica abrange as bacias amazônicas desde o Alto Orinoco,
situado no estado do Amazonas, até o Baixo Tocantins, no Pará, extrapolando as
fronteiras brasileiras para atingir os países vizinhos como: Peru, Bolívia, Equador,
Venezuela, Colômbia, Suriname, Guiana e Guiana Francesa (Pires, 1953; Terborgh &
Andresen, 1998), além de regiões pré-amazônicas (extremo oriental), como o estado do
Maranhão (Muniz, 1996). A floresta Atlântica estendia-se do Rio Grande do Sul até o
nordeste brasileiro (Rio Grande do Norte), tendo como região nuclear entre o rio São
Francisco e o Rio Doce (no Estado da Bahia) e São Paulo, ocupando 1 milhão de km²
(Joly et al., 1991; Veloso et al., 1991; Rizzini, 1997; Scudeller et al., 2001).
Dentro do domínio Atlântico e Amazônico são encontradas diversas
fitofisionomias florestais e formações associadas ora adentrando no continente, ora se
14
Classified - Internal use
restringindo a uma estreita faixa litorânea de planície, ocorrendo também os encraves
florestais no Nordeste (Schäffer & Prochnow, 2002), conhecidos como brejos de altitude
ou florestas serranas. A origem da vegetação encontrada nas florestas serranas é baseada
na teoria de refúgios, onde esses encraves florestais úmidos localizados no semiárido
teriam funcionado como refúgios ecológicos (Haffer, 1969; Bigarella et al., 1975;
Andrade-Lima, 1982; Prance, 1982) para diferentes espécies vegetais e animais,
constituindo abrigo natural de diversas espécies ameaçadas de extinção e de espécies
novas ainda não descritas (Schneider et al., 1999).
As florestas serranas são ilhas de floresta úmida estabelecidas na região semi-
árida, sendo cercadas por uma vegetação de caatinga (Andrade-Lima 1982), classificadas
como “áreas de exceção” dentro do domínio do nordeste semi-árido (Lins 1989). A
existência dessas ilhas de floresta em uma região onde a precipitação média anual varia
entre 240 - 900 mm (Lins 1989) está associada à ocorrência de planaltos e chapadas entre
500 - 1.100 m altitude (por exemplo, Borborema, Chapada do Araripe, Chapada de
Ibiapaba), onde as chuvas orográficas garantem níveis de precipitação superiores a 1.200
mm/ano (Andrade-Lima 1960, 1961). Quando comparados às regiões semiáridas, os
brejos possuem condições privilegiadas quanto à umidade do solo e do ar, temperatura e
cobertura vegetal (Andrade-Lima 1966).
As florestas serranas são caracterizadas por apresentar plantas com distribuição
amazônica e algumas espécies típicas das florestas serranas do sul e sudeste do Brasil
(Tabarelli, 2001). Santos (2002) reforçou essa hipótese ao encontrar um padrão de
distribuição disjunta da flora lenhosa na Amazônia e em 12 localidades de floresta serrana
nordestina de Pernambuco que se enquadra em um modelo de separação sequencial e
gradativa de uma condição preexistente (retração da floresta úmida). Dessa forma, o
15
Classified - Internal use
processo de separação continuou até que essas florestas atingissem o número e a
conformação espacial atual (Tabela 1).
Tabela 1. Número e área florestal das florestas serranas ocorrentes no Nordeste (sensu Andrade-Lima,
1982).
Estados No de florestas serranas Área florestal (km2) %
Ceará 11 6.596,50 34,2
Rio Grande do Norte 5 1.147,50 6
Paraíba 8 6.760 35,1
Pernambuco 23 4.850 25,2
Sergipe 2 670 3,5
Total 49 19.259 100
No estado do Ceará, as serras úmidas são encontradas nos topos das serras e
chapadas e nas vertentes a barlavento (que recebem as chuvas orográficas), como é o caso
das serras de Aratanha, Maranguape, Meruoca, Uruburetama, Planalto da Ibiapaba do
Norte, Chapada do Araripe e Baturité (Figueiredo & Barbosa, 1990). Tais serras
representam apenas 3,71% da superfície do Estado (Xavier, 2007, Figura 1) e, dentre
estas, a Serra de Baturité merece destaque por ser a mais extensa, uma das mais altas e
mais úmidas do Ceará (Araújo et al., 2007).
As melhores condições climáticas das florestas serranas atraíram e continuam
atraindo pecuaristas e agricultores, que, através da criação de bovinos e do plantio de
lavouras permanentes (banana, café e citros) e temporárias (hortaliças, mandioca, milho
e feijão) constituem a base da estrutura sócio-econômica nessas áreas (Lins 1989). Silva
& Tabarelli (2000) demonstraram que as florestas serranas nordestinas necessitam de
ações conservacionistas mais intensas, já que estas são muito ameaçadas, tanto pela
grande distância existente entre elas, como pela exploração antrópica. Considerando que
esses encraves florestais constituem abrigo natural de diversas espécies ameaçadas de
extinção e que a maior parte da biodiversidade encontra-se hoje localizada em pequenos
fragmentos florestais pouco estudados e marginalizados por iniciativas conservacionistas,
16
Classified - Internal use
o conhecimento da estrutura e da composição das comunidades vegetais e animais e de
suas interações com o ambiente é de suma importância para a implementação de ações
visando à diminuição da perda de biodiversidade e à conservação ambiental dessas
florestas.
Poucos estudos nas florestas úmidas serranas no Nordeste brasileiro enfocam
biogeografia por meio de composição taxonômica. As análises fitogeográficas abordam
relações entre o Centro Pernambucano de Endemismo e outras fitofisionomias florestais
(Santos et al., 2007), relações de florestas serranas com florestas de terras baixas em
Pernambuco (Cavalcanti & Tabarelli, 2004) e comparações florísticas e estruturais entre
florestas estacionais e úmidas em Pernambuco (Lopes et al., 2008).
Uma vez que os táxons ombrófilos provenientes das florestas tropicais geralmente
não toleram condições de aridez, a hipótese principal desse estudo é de que as florestas
serranas configuram-se como áreas únicas e distintas de tensão ecológica com grande
diferença da caatinga circundante, já que os sucessivos ciclos climáticos no Nordeste
durante o Pleistoceno foram responsáveis pelo isolamento de populações de espécies
vegetais nessas florestas. Dentro deste contexto, conhecer a vegetação das florestas
serranas é de suma importância para entender a origem desses encraves de vegetação
úmida. A segunda hipótese é de que a distribuição de espécies nessas florestas é
dependente da altitude e das condições edáficas, diferenciando-se portanto entre as
vertentes das serras e entre os diversos tipos vegetacionais localizados no semiárido.
Admitindo-se que as florestas serranas são áreas ecotonais ilhadas no semiárido,
a pergunta geral que norteia o trabalho é: Qual a relação entre as florestas serranas úmidas
e os diferentes domínios brasileiros, e como se dá a distribuição de espécies nessas áreas?
Desse modo, estudamos uma serra específica, comparamos com os estudos já existentes,
elaboraramos listas das espécies ocorrentes nas florestas serranas e verificamos suas
17
Classified - Internal use
características principais que as delimitam enquanto áreas de refúgio. A partir daí,
denominamos perguntas específicas: A) Qual a composição florística, diversidade e como
se estrutura a Serra de Maranguape, Ceará, comparando vertentes e topo? B) Quais são e
como as variáveis ambientais explicam a distribuição de espécies na Serra de
Maranguape? C) Como a flora da Serra de Maranguape se relaciona com as outras
fitofisionomias do estado do Ceará? (Capítulo 1, onde a Serra de Maranguape foi
escolhida como um objeto de estudo das florestas serranas úmidas em escala local). D)
Qual a composição palinológica das Serras de Maranguape, Baturité e Pacatuba,
comparando vertentes e topo? E) Como a distribuição da vegetação está relacionada com
a atual chuva polínica? F) Qual e como as variáveis ambientais explicam a distribuição
da chuva polínica nessas serras? (Capítulo 2, que mostra a importância da chuva polínica
das florestas serranas cearenses para entender a distribuição atual da vegetação). G) Qual
a relação entre as serras úmidas do Nordeste brasileiro? H) Qual as variáveis bióticas e
abióticas (processos ecológicos) e como elas explicam a atual distribuição de espécies
nessas serras nordestinas? (Capítulo 3, que acrescenta uma visão regional das florestas
serranas úmidas nordestinas) I) Qual a relação das florestas serranas úmidas com
Amazônia, Floresta Atlântica, Cerrado e Caatinga? J) As florestas serranas do Nordeste
brasileiro podem ser classificadas com uma única unidade florística? (Capítulo 4, que
relaciona os domínios brasileiros com as florestas serranas nordestinas e traz uma nova
classificação dessas áreas).
18
Classified - Internal use
CHAPTER I
VEGETATION STRUCTURE, SPECIES COMPOSITION AND DISTRIBUTION
IN A TOPOGRAPHIC GRADIENT OF A MOUNTAIN FOREST, NORTHEASTERN
BRAZIL *
SPECIES DISTRIBUTION IN A TOPOGRAPHIC GRADIENT
*MANUSCRIPT SUBMITTED TO JOURNAL OF SYSTEMATICS AND EVOLUTION
IVAN JEFERSON SAMPAIO DIOGO1*, KARIN DOS SANTOS2, ITAYGUARA RIBEIRO
DA COSTA 3 & FLAVIO MÄES DOS SANTOS 1
¹University of Campinas - UNICAMP, Institute of Biology, Department of Plant Biology, 13083-970.
Campinas, SP, Brazil *[email protected]
²Swedish Museum of Natural History, Botany Department, P.O. Box 50007 SE-104 05 Stockholm, Sweden
³Federal University of Ceará - UFC, Departament of Biology, Sciences Center, 60455-760. Fortaleza,
CE, Brazil
19
Classified - Internal use
Abstract
The humid mountain forests located in Brazilian semiarid region are diverse and
endangered areas located in the semiarid region. We aimed to examine the
phytosociological structure, richness, diversity and species distribution of woody plant of
the mountain forest at Serra de Maranguape, Brazil. Data collection went from 2013 to
2015 and we divided the whole area into three topographic categories: windward (600-
800 m, WMA), leeward (600-800 m, LMA) and top (above 800 m, TMA). We calculated
different phytosociological parameters and floristic diversity of each category. A
correspondence analysis was performed to analyze the indirect ordination of forest sites
by species abundance. A canonical correspondence analysis was conducted to verify
which variables were driving species distribution. We used non-metric multidimensional
scaling distance and average linkage method to investigate the similarity among mountain
forests and thorny woodlands. A total of 1536 individuals belonging to 144 tree species
distributed in 44 families and 93 genera were recorded. Myrtaceae, Fabaceae and
Rubiaceae were the most species rich families. Myrcia splendens had the highest IV
followed by Guapira nitida and Mollinedia ovata. The leeward slope showed the highest
richness and diversity index, the windward showed the highest density and the top showed
the highest basal area. The PCA ordination indicated a greater similarity between TMA
and WMA than LMA. Temperature and precipitation are the climatic factors driving
species distribution in Maranguape. Crystalline origin mountains are closer than
sedimentary ones, but the leeward slope from Baturité was close to thorny woodland
areas. Our results will be useful for conservation or restoration purposes on the poor
known mountain forests of Northeast, Brazil.
Key words: Myrtaceae, Fabaceae, temperature, precipitation, leeward, windward and
top.
20
Classified - Internal use
1 Introduction
The Brazilian Atlantic Forest (BAF) represents the second largest tropical forest
in the world and may contain between one and eight percent of the world’s species (Silva
and Casteleti, 2003). However, BAF has been explored and degraded for more than 500
years, one of the most threatened and fragmented ecosystem in the world. Nowadays,
only around 11.4% to 16% of its original cover remains (Ribeiro et al., 2009, 2011).
Forzza et al. (2012) estimated that approximately 20,000 species (40% endemics) of
vascular plants are found in the BAF, which make it one of the most distinctive
biogeographical units in the entire Neotropical region.
Through the last century, the Atlantic forest has been extensively deforested
because of coffee and sugarcane monocultures to the economic development of Brazil
(Gibbs et al., 2010). Nevertheless, BAF is recognized as one of the 35 biodiveristy
hotspots for conservation priorities (Myers et al., 2000; Mittermeier et al., 2004; Zachos
and Habel, 2011). Some forest fragments still exist and are concentrated on the top of
mountains and steep slopes, where is difficult to access and the anthropogenic activities
is unfeasible.
The northernmost BAF stretches 1,500 km along the atlantic coast from Bahia to
Rio Grande do Norte states. In Ceará state, the original Atlantic Forest is distributed in
small and isolated fragments, the mountain forests, such as Baturité (Araújo et al., 2007).
These fragments may play an important role as stepping stones for diversity, dispersal
and conservation of species (Anderson and Jenkins, 2006). Due to their history, these
small enclaves on mountain tops are considered as interglacial microrefugia, i.e., relics
of a past expansion of the Atlantic forest (Ledru et al., 2007; Montade et al., 2014).
21
Classified - Internal use
These humid forest enclaves located in the semiarid region have functioned as
ecological refuges for flora and fauna, providing natural shelter of several endangered
species and new species not yet described (Andrade-Lima, 1982). They are characterized
by having plants with Amazon, Atlantic and Caatinga distribution (Santos et al., 2007).
There are 49 mountain forests, distributed in the states of Ceará, Rio Grande do Norte,
Paraiba and Pernambuco, which cover 19 259 km2; however, this number is actually
reduced by half. (Vasconcelos Sobrinho, 1971). In Ceará state, these humid forests are
found on the tops of mountains and plateaus and on the windward side (receiving
orographic rainfall), which cover 3.71% of the state surface (Xavier, 2007).
Few studies and inventories are located on these areas and it demands further
attention especially given the extensive loss of habitat. We aimed to examine the floristic
composition, diversity and phytosociological structure of woody plant of the mountain
forest at Serra de Maranguape, Ceará state, at different topographic areas: leeward,
windward and top. We conducted statistical comparisons between Serra de Maranguape
and other mountain forests from Ceará State in order to verify if they differ in terms of
topographical, horizontal and vertical phytosociological structure.
2 Material & Methods
2.1 Study area
The study took place in the mountainous rainforest of the protected area of Serra
de Maranguape (Área de Proteção Ambiental da Serra de Maranguape; SEMA). The
SEMA is one of the high elevation areas of Ceará State, reaching 980 m a.s.l. (3°53′40.32″
S, 38°43′13.56″ W), covers an area of 71 km2 and is located 30 km from the Atlantic
Ocean (Fig. 1). The SEMA has a very steep crystalline relief and different types of plant
physionomies and is a protected área (APA).The local climate is characterized by a hot
22
Classified - Internal use
semiarid, with rainy summer (December to May) and dry winter (mainly from September
to November), BSh in Köppen-Geiger system (Peel et al., 2007). In opposite to the
semiarid climate, the SEMA has a tropical and sub-humid climate (Aw), with avarage
rainfall above 1300 mm year-1 and average temperature between 23°C and 26°C. The soil
classes vary between Alisol, Luvisol and Neosol (Arruda, 2001).
While the lower areas of the mountain forests are dominated by deciduous thorny
woodland (caatinga), upper areas are dominated by semideciduous and ombrophilous
forests. The position of the massif in relation to the wind exposure provides a higher
humidity on the windward slopes due to the rain-forced convection. This fact results in
different gradients among the slopes and, in general, higher precipitation than the
surrounding semi-arid region, where the average precipitation rate is 700 mm per year
(Funceme, 2005).
2.2 Data collection
Data collection started in July 2013 and completed in January 2015 and was based
on the point-centered quarter method (Cottam & Curtis, 1956). This procedure consists
of dividing each sampling point into four quarters by a pair of perpendicular lines. We
divided the SEMA into three topographic categories: top (TMA) from 880 m to 934 m
(3o53'48''S; 38o43'15"W), leeward (LMA) from 654 m to 702 m (3°54'30.9''S;
38°43'57.0''W) and windward (WMA) from 678 m to 796 m (3o54'31''S; 38o43'01"W).
We established two areas of 0.5ha in each one (separated by a minimum distance
of 200m and a maximum of 1km; Fig. 2A). In each area, we sampled five parallel
transects with 100m length, 10 points quadrats and 50m between them (Fig. 2B), totaling
3ha, 30 transects and 300 point quadrats. We attempted to sample a standardized area of
each category, in order to isolate the effect of area from other variables, such as edge and
23
Classified - Internal use
anthropogenic effects. In each quarter, we recorded the nearest tree of 15 cm perimeter at
breast height (PBH) or larger and measured their circumference (cm), height (m) and
distance from the point to the tree (m). Multi-stemmed individuals at breast height were
sampled if at least one stem showed this minimal PBH criterion. When these individuals
had a PBH < 30cm, we sampled another individual in the same quadrat with PBH ≥ 30
cm (Fig. 2C).
We identified specimens by consulting the literature and specialists and by
comparison in herbaria. The plant families were listed according to the Angiosperm
Phylogeny Group IV guidelines (APG IV, 2016). Vouchers of specimens were deposited
in the EAC (Federal University of Ceará, Brazil), (University of Campinas, Brazil) and S
(Swedish Museum of Natural History, Stockholm, Sweden) herbaria.
2.3 Data analysis
We analyzed the following phytosociological parameters: absolute and relative
density (based on the distance from the point to the tree), frequency and dominance; cover
value, importance value (IV) and basal area of each topographic category and of whole
SEMA (Mueller-Dombois & Ellenberg, 1974). We calculated floristic diversity by
Shannon-Wiener (H’), Pielou’s evenness (J’) and Simpson index (D). A non-parametric
Kruskal-Wallis test was applied to verify statistical differences among forest sites on
average tree height and diameter. A Chi-square test was performed to verify differences
in the proportion of multi-stemmed individuals at breast height. These analyzes were
performed using the software Fitopac 2.1 (Shepherd, 2009) and PC-ORD 6.0 (McCune
& Mefford, 2011).
A correspondence analysis (CA) (Hill, 1973) was performed to analyze the
indirect ordination of forest sites by species abundance in order to verify the horizontal
24
Classified - Internal use
structure similarities among the different sites. Those species observed at only one site
were eliminated in this analysis because the CA is very sensitive to their presence
(McCune & Grace, 2002).
To test for differences in horizontal and vertical structure between leeward,
windward and top, the structural parameters of species richness (S), Shannon-Wiener
index (H'), total density per hectare (DE), total dominance per hectare (DO) were listed
for Baturité and Maranguape mountains and compared. Furthermore, in order to
determine correlations between the spatial distribution of vegetation groups and
environmental data from each area, Canonical Correspondence Analysis (CCA) was
carried out. Environmental parameters such as slope angle and elevation were calculated
using QGIS software (QGIS Development Team, 2015) running the ASTER Global
Digital Elevation Model (from METI and NASA). Mean annual precipitation and
temperature values for the period 1950–2000 were obtained using the WorldClim
database for each location of the surface samples with a spatial resolution of 1 km2 (Fick
and Hijmans, 2017). To test the correlation among the environmental variables, we did a
Pearson correlation test (r).
We used non-metric multidimensional scaling (nMDS) with Jaccard index and
average linkage method (UPGMA) using the Jaccard index to test for floristic similarities
among Cearás’s mountain forests and caatinga (Table 1). These analyses were performed
in PC-ORD 6.0 (McCune & Mefford, 2011).
3 Results
A total of 1536 individuals belonging to 144 tree species distributed in 44 families
and 93 genera were recorded in the 6 samples (Table 2). Myrtaceae (21 spp.), Fabaceae
(17 spp.), Rubiaceae (15 spp.), Lauraceae and Sapotaceae (6 spp.), and Malpighiaceae
25
Classified - Internal use
and Euphorbiaceae (5 spp.) were the most species rich families, which accounted for
52,1% of the species observed at the study site. Eugenia and Myrcia (07 spp.), and Inga,
Senna, Ocotea and Byrsonima (4 spp.) were the richest genera. Many families and genera
were represented by only one species each. Myrtaceae, Nyctaginaceae, Fabaceae,
Lauraceae, Sapindaceae and Rubiaceae showed the highest IV and represented almost
50% of the total IV.
Although Myrcia splendens had the highest IV (7.5% of the total IV), number of
individuals, relative density and relative frequency, its relative dominance was lower than
that of Guapira nitida and Mollinedia ovata, which together represented 12.1% of the
total IV (Table 2). Myrcia splendens, G. nitida, M. ovata, Cupania impressinervia,
Handroanthus serratifolius, Nectandra cuspidata, Inga bollandii, G. opposita,
Pilocarpus spicatus, C. oblongifolia, Margaritaria nobilis, Cinnamomum triplinerve had
more than 30 individuals and were the species with highest IV (43% of the total, Table
2). However, Neea floribunda had 19 individuals and presented a big IV value (5.47).
The standing dead biomass was represented by 52 individuals and 3.5% of the total IV.
The total density for the three forest sites was 1,275 individuals.ha−1, the total
dominance was 27.39 m2.ha−1, the total basal area was 26.53 m2.ha−1. The Shannon-
Wiener index was 4.182, evenness was 0.841 and Simpson index was 0.027. The LMA
site showed the highest species-level richness (95 spp.) and the WMA showed the highest
density (522 individuals.ha−1). However, the WMA site showed the lowest dominance
(29.15 m2.ha−1) and the TMA showed the highest basal area (28.374 m2.ha−1), while the
LMA presented the highest Shannon-Wiener and evenness index (Table 3).
From the 144 species sampled, 20 (13.9%) were present in the three sample areas,
16 (11.1%) between TMA and WMA, 5 (3.47%) between TMA and LMA, and 13
26
Classified - Internal use
(9.03%) between WMA and LMA slopes. Thus, 12 (8.33%) species were found only in
TMA, 17 (11.8%) only in WMA and 41 (28.5%) only in LMA. The CA ordination of
forest sites by species abundance indicated a greater similarity between TMA and WMA.
The inertia explained by the first two axes was 70.8% (Fig. 3). M. splendens was abundant
in all sample areas, but was strongly associated with TMA and WMA, as well as G. nitida.
M. ovata. and I. bolandii were more associated with TMA areas, while C. impressinervia
and Neoraputia magnifica with WMA areas. On the other hand, P. spicatus, H.
serratifolius and Senegalia poliphilla were more associated with LMA areas (Table 3 and
Fig. 3).
The tree stems in SEMA had an average height of 11.45m ± 4.58 and an average
diameter of 58.67cm ± 43.59. The average height had no significant difference among the
sites, although the diameter (H = 187.4521, P < 0.001) in TMA were significantly higher
than WMA and LMA. The WMA and LMA sites were similar in terms of the proportion
of multi-stemmed individuals at breast height, 13.36% and 12.34% respectively, whereas
in TMA showed a higher proportion (23.75%) of stems per individual than the other sites
(χ2 = 18.766, P = 0.0005, DF = 2). The Shannon–Wiener (H’) and evenness (J) were
higher in LMA, followed by WMA and TMA respectively, whereas the Simpson index
(D) was higher in TMA (Table 3).
The LMA showed distinct horizontal structure when compared to WMA and
TMA, whereas these two are quite similar. LMA had higher richness (S) (T = 4.679, P =
0.0003, GL = 9), and higher diversity according to the Shannon-Wiener index (H') (U =
0.00, Z(U) = 3.607, P = 0.002); however, the total density per hectare was significantly
lower (T = −3.478, P = 0.0017, GL = 9). TMA had a significantly higher total dominance
per hectare (T = 1.067, P = 0.00012, GL = 9). WMA had higher total density per hectare
(T = 2.149, P = 0.0023, GL = 9). On the other hand, the leeward slope of Baturité
27
Classified - Internal use
mountain had the lower richness and diversity per hectare, whereas the top presented
higher density, richness and diversity per hectare (see Araújo et al., 2007). However,
Baturité and Maranguape mountains have one similarity: the both leeward slopes have
the lowest density.
The distribution of samples in the CCA diagram clearly separates four groups,
which displays windward and top of each mountain in two different groups and each
leeward as a unique group, which reflects the vegetation variation from the seasonal semi-
deciduous montane forest to the dense ombrophilous forest. Temperature and
precipitation were the main ecological parameters driving the species distribution in these
mountains. TMA and WMA had greater values for temperature and precipitation,
whereas TB and WB had the same greater values for precipitation, but lower for
temperature. LMA had greater values for temperature, but lower for precipitation, and LB
had lower values for both variables (Fig. 4).
The environmental parameters are mainly correlated with axis 2. The main
correlation concerning axis 1 is with the temperature (-0.715), followed by precipitation
(-0.580). The closest environmental parameters related to axis 2 are precipitation (0.802),
temperature (0.415) and slope (0.190) (Table 4).
The cluster dendrogram showed two different groups: one with TB, WB, TMA,
WMA, LMA, MF1, MF2 and MF3, and another one with LB and CA 1, CA2, CA3 and
CA4 (17.5% of information remaining, Fig. 5). In a higher resolution, the leeward slope
of Baturité was grouped with all Caatinga areas, except for CA2. Maranguape, MF1, TB
and WB were grouped together. MF2 and MF3 formed the last group (Fig. 5). The
correlations between the cophenetic similarities and the Jaccard original similarities were
strong (0.78), confirming the reliability of the UPGMA analyses.
28
Classified - Internal use
The NMDS ordination provided a two-dimensional solution, with both axes being
significant (P = 0.01), and explained 79% of the correlations between the distances in the
original space and the ordination distances. After 38 iterations, the final stress value was
13.71, which is a satisfactory result (McCune and Grace, 2002). The NMDS diagram
showed a clear separation among all the mountains and caatinga areas with three groups
formed (Fig. 6). The nMDS analysis ordered the same groups from the UPGMA,
revealing a great correspondence between them.
There was no correlation between the geographical distance and the floristic
among topographical sides (r = 0.12; p = 0.32). On the other hand, there was a strong
correlation between the geographical distance and the floristic among mountain forests (r
= 0,63; p < 0,01).
4 Discussion
The results of our study were largely inconsistent with those of several other
floristics studies in Ceará state, mainly because of the Caatinga areas, where the species
richness is lower and Fabaceae is the richest family (Queiroz, 2006; Lima et al., 2012a;
Moro et al. 2014). On the other hand, we found some similarities among the mountain
forest areas of Ceará, such as the higher number of species, Myrtaceae as the richest
family and Myrcia splendens as the most common species (Araújo et al., 2007; Ribeiro-
Silva et al., 2012). Maranguape and Baturité mountains share many species, including
similar richness, H’, IV, density, frequency and dominance because of their geographical
proximity and same crystalline origin. Moreover, those similarities also occur with
mountain forests located in other states from Northeast, Brazil (Diogo et al., In press).
Considering the slopes, it is worth noting that, although they are geographically
close, each one has a distinctive flora with some exclusive species. This result
corroborates the hypothesis that the plant distribution is more strongly associated with the
29
Classified - Internal use
forest type and environmental heterogeneity than with the geographical distance (Lima et
al., 2012a; Gonzalez-Caro et al., 2014; Sabatini et al., 2014). The greater richness and
diversity of species in the leeward slope of SEMA do not agree with Gentry (1982, 1988),
who pointed that a greater diversity would be expected in wetter areas. Araújo et al.
(2006) and Fernández-Palacios et al. (1995) also found a positive correlation between
precipitation and species richness and diversity, with the windward areas being the most
diverse. On the other hand, our findings agree with some studies with Fabaceae in
Baturité, Brazil (Lima et al., 2012b), with floristic composition in a rain forest in Taiwan
(Chen et al., 1997) and with altitudinal pattern of vegetation on Tenerife (Fernández-
Palacios and Nicolás, 1995).
Although we found a greater richness and diversity for total flora in the drier areas
of the Maranguape Mountain Range, this result was not observed when the analysis was
carried out at the family level. For instance, Myrtaceae has higher richness and diversity
in wetter areas as the Atlantic rainforest (Murray-Smith et al., 2009; Lucas & Bünger,
2015), Amazonia (Nigel et al., 2002) and the windward slope, whereas Fabaceae has the
opposite pattern, with higher richness and diversity in the drier areas as the Caatinga
(Moro et al., 2014) and the leeward slope. The richness and diversity of species vary
according to the taxonomic group analyzed and its evolutionary history (Murray-Smith
et al., 2008; Schouten et al., 2009), each family responds differently according to the
climate and topography. For instance, temperature is the most important climatic variable
for Fabaceae, but for other families, such as Myrtaceae and Bignoniaceae it is
precipitation (Punyasena et al., 2008). The forest composition change found when
comparing windward and leeward was also observed in the corresponding pollen
assemblages of those slopes (Montade et al., 2016). The windward slope is an area which
is under the influence of orographic rainfalls and has features of ombrophilous forest as
30
Classified - Internal use
its dominant type of vegetation, as we see an increase of Myrtaceae, Lauraceae and
Nyctaginaceae families. The leeward slope is an area with lower precipitation and higher
temperatures and has semideciduous seasonal forest as vegetation type with an increase
of Fabaceae, Bignoniaceae and Rutaceae families.
The type of forest located closer to the mountain top differs partly in their structure
and composition from the forests located just below, which demonstrates typical features
of forests under influence of clouds, such as low canopy height, low species richness,
high proportion of multi-stemmed individuals at breast height, high tree diameter and
high dominance (Hamilton et al., 1995; Falkenberg and Voltolini, 1995; Meireles et al.,
2008). Among the sampled species, an increase of pioneer and heliophilous trees is
observed, such as Alchornea glandulosa, Pourouma guianensis and Miconia mirabilis
(Souza et al., 2006; Pessoa et al., 2012), as well as an increase of narrow geographical
distribution species, which occurred only in forest formations of high-altitude mountain
ranges, such as Clusia melchiorii and Clusia paralicola (Araujo & Scarano, 2007). The
representativeness increase of those taxa in both vegetation and pollen rain (Montade et
al., 2016), may be consequence of the existence of specific environmental conditions on
the mountain tops or areas located close to it, such as different soil conditions related to
the rocks. Those local conditions can explain the difference in plant community, which is
more dependent on atmospheric moisture under the influence of fogs and/or clouds and
is dominated by species ecologically adapted (Kitayama, 1992). Moreover, although
rainfalls would reach their maxima close to the mountain tops, local environmental
conditions induce lower diversity and richness and higher density when compared to
forests located a few tens of meters below. The higher Simpson index for top surveys
suggests a higher concentration of individuals within a few species as altitude increases
(Martins & Santos, 1999).
31
Classified - Internal use
The lower richness, diversity index and density in leeward slope of Baturité are
related with more restrictive environmental conditions, such as high temperatures and
small amount of rainfall (Mantovani, 2006). Furthermore, LB is greatly influenced by
caatinga flora (Cereus jamacaru, Croton blanchetianus, Jatropha mollissima and
Mimosa caesalpiniifolia), which suggests that those areas share some characteristics, such
as shallow depth soils with high pH, maximum and minimum temperatures above the
average for mountain forests, and annual rainfall concentrated in a few months of the year
and below the average for mountain forests (Araújo et al., 2007). Although the leeward
slope of Maranguape has the lower density, it presents a greater mix of species with some
Caatinga species (Anadenanthera colubrina and Poincianella bracteosa), Amazon
rainforest (Abarema jupunba and Coccoloba latifolia) and Atlantic rainforest (Alseis
floribunda and Celtis spinosa). Besides sharing species with windward and top areas,
more than 20% of the species of Maranguape mountain occurs only in LMA.
This study indicates that many of these species seem to prefer a given area, and
those divergences might be a response to climate conditions: temperature and
precipitation changes induced by the altitudinal gradient. Both faces have different mean
annual temperature and precipitation forming specific communities growing on distincts
environments (Chen et al., 1997; Mantovani, 2006; Punyasena, 2008). Those climate
variables are the chief factors determining the distribution of the vegetation around the
world and are usually associated with rapid transition among ombrophilous forest and
semideciduous seasonal forests (Oliveira-Filho et al., 2006). Tree species diversity is
highly correlated with water consumption and energy uptake, resources that are
partitioned among species, and could limit their number in forest communities (Hugget,
1995). The precipitation changes are controlled by the elevation of air masses on the
windward side, which act as a barrier generating orographic precipitation. The leeward
32
Classified - Internal use
side is not reached by many of the air masses, which decrease rainfalls and increase
evaporation. This mechanism is called rain shadow effect and is commonly observed in
tropical mountainous islands (Giambelluca et al., 2013; Duarte et al. 2005). The variation
in the floristic composition among different plant formations in Maranguape mountain
range is also observed in a larger scale, as in the southeastern and northeastern regions of
Brazil (Oliveira-Filho & Fontes, 2000; Ferraz et al. 2004) and even in South America
(Oliveira-Filho et al., 2006). Furthermore, biogeographical studies show that both
temperature and rainfall regime suffered dramatic shifts during the Quaternary (Ledru,
1993, Ledru et al., 1998).
Slope angle could also influence vegetation distribution, because leads to a
gradual removal of dissolved materials from linear slope and the accumulation of these
materials near the creek site. However, it is a relevant factor only at values over than 40o.
(Fernández-Palacios and Nicolás, 1995; Montade et al., 2016). Significant progress has
also been made in analysing consequences of plant–soil feedbacks for biodiversity-
functioning relationships and plant distribution in mountain forests (Liu et al., 2003; John
et al. 2007; Diogo et al., In Press). Nevertheless, the distribution of individuals in a given
environment is not only determined by abiotic factors, but may also result from biological
interactions such as competition, mutualisms herbivorism, dispersal and pollination
(Wisz et al., 2013; Diogo et al., 2016).
When the flora of SEMA was compared to the main plant formations of Ceará
state, there was a grouping varying from drier to wetter areas. It is worth noting that the
leeward of Baturité is clearly more similar to thorny woodland areas and deciduous forests
(Santos et al., 2007 and Araújo et al., 2007 found similar results). For instance, 34% and
32% of its Fabaceae exclusive species can be found in seasonal forests areas and thorny
woodland (Bauhinia cheilanta and Libidibia ferrea), respectively, but only 4% occur in
33
Classified - Internal use
ombrophilous areas (Copaifera duckei) (Lima et al., 2012b). Whereas the leeward of
Maranguape is more similar to semideciduous and ombrophilous forest, with species that
can be found in both forest formations, including the other mountain forests, such as
Anadenanthera colubrina, Ixora brevifolia and Randia armata, it still has thorny
woodland species, Poincianella bracteosa for example. This difference between leeward
slopes can be explained by the distance from the coast, while Maranguape is only 30 km
far, Baturité is 80 km. The precipitation generated by maritime trade winds is distributed
along a gradient: decrease from the coast to the interior, thus, leeward slope of
Maranguape receives a greater amount of rainfall and, consequentely, has better
conditions to different species grow.
On the contrary, windward and top areas of Baturité and Maranguape are more
related to Pacatuba, Araripe and Ubajara, sharing some species, such as Eugenia florida,
Inga bolandii, I. ingoides, Miconia prasina, Myrcia splendens, Ocotea longifolia,
Psychotria carthagenensis and Vismia guianensis. Those forests have similar climatic
conditions and are clearly more humid and receive higher precipitation amount, which
contribute positively to species distribution there. In addition, Maranguape, Baturité and
Pacatuba are closer because of the same crystalline origin and the geographical
approximation. Whereas, Araripe and Ubajara are closer because of the sedimentary
origin. During the Quaternary, the semi-arid region of Brazil experienced episodes of
climatic fluctuations (Behling et al., 2000) and the forested vegetation could have
established in mountain forests as a humid refugia on the most dry periods (Auler et al.,
2004), which could explain the similarity among these mountains.
5 Conclusion
The huge environmental and climatic heterogeneity (temperature and
precipitation) observed in the Maranguape mountain can explain the high diversity and
34
Classified - Internal use
variation in floristic composition among slopes and top because it enables the coexistence
of different species, and such pattern are also observed on a regional scale. Our study
provides a better knowledge of flora, species distribution, forest relationships of the
mountainous massifs in northeastern Brazil. However, there are still areas in northeastern
mountain forests where no floristic or quantitative surveys have been conducted. As these
areas can be considered as a refugia of tropical rainforest in the middle of semiarid, our
results will be useful for conservation or restoration purposes. The knowledge of the flora
in these regions can bring new insights to the species distribution of forests under the
influence of temperature and precipitation in altitudinal gradients.
6 Acknowledgements
Financial support was provided by Royal Swedish Academy of Sciences, KS was
benefited. IJSD was benefited from PhD position funded by CNPq (Brazil). FAMS was
benefited by CNPq (number 306595/2014-1). We thank the Swedish Museum of Natural
History for host IJSD during the taxonomy process. We thank Fernanda Cabral and
Nállarett Dávila for the help on the identifications.
7 Conflict of interest
No potential conflict of interest was reported by the authors.
35
Classified - Internal use
Table 1. Floristic surveys of 6 areas of mountain forests and 4 areas of caatinga located
in Ceará State used in this study.
Locality Code Geographic coordinates
(S;W)
References
Baturité windward
Baturité leeward
Baturité top
Pacatuba
Ubajara
Araripe
Crateús
Aiuaba
Iguatu
Quixadá
WB
LB
TB
MF1
MF2
MF3
CA1
CA2
CA3
CA4
4º17’54”; 38º56’10’’
4º15’32”; 39º00’04’’
4º12’22”; 38º55’55’’
3°59'04"; 38°37'12"
3º51’16”; 40º55’16”
7º16’32”; 39º27’13”
5°7’; 40°2’22.79”
6°36’01”; 40°19’19”
6°19’46”; 39°22’38’’
4°49′34″, 38°59′09″
Araújo et al. (2007)
Araújo et al. (2007)
Araújo et al. (2007)
Not published
Not published
Ribeiro-Silva et al. (2012)
Costa & Araújo (2012)
Sousa et al. (2007)
Lima et al. (2012)
Costa et al. (2007)
36
Classified - Internal use
Table 2. Phytosociological parameters of species sampled in the three topographic areas
of the Serra of Maranguape, northeastern Brazil. NInd: number of individuals, RelDe:
relative density, NOQ: number of occurrences per quadrant, AbsFr: absolute frequency,
RelFr: relative frequency, RelDo: relative dominance, MaxH: maximum height, MaxD:
maximum diameter, IV: importance value, CV: coverage value.
Species NInd RelDe NOQ AbsFr RelFr RelDo MaxH MaxD IV CV
Myrcia splendens 165 10,74 28 93,33 4,15 7,59 33,00 69,07 22,48 18,33
Guapira nitida 87 5,66 25 83,33 3,70 12,78 25,00 75,76 22,15 18,44
Mollinedia ovata 40 2,60 9 30,00 1,33 10,09 27,00 208,81 14,03 12,69
Cupania impressinervia 80 5,21 27 90,00 4,00 2,11 20,00 32,15 11,32 7,32
Handroanthus serratifolius 47 3,06 19 63,33 2,81 3,96 20,00 97,40 9,83 7,02
Nectandra cuspidata 33 2,15 15 50,00 2,22 3,64 25,00 85,94 8,01 5,79
Inga bollandii 30 1,95 11 36,67 1,63 3,78 26,00 127,32 7,36 5,73
Guapira opposita 34 2,21 15 50,00 2,22 2,85 30,00 49,97 7,29 5,06
Pilocarpus spicatus 42 2,73 8 26,67 1,19 1,92 19,00 72,89 5,84 4,65
Cupania oblongifolia 38 2,47 15 50,00 2,22 1,08 18,00 32,79 5,78 3,55
Neea floribunda 19 1,24 8 26,67 1,19 3,05 19,00 102,81 5,47 4,29
Margaritaria nobilis 34 2,21 11 36,67 1,63 1,60 21,00 51,57 5,45 3,82
Cinnamomum triplinerve 31 2,02 11 36,67 1,63 1,78 20,00 60,16 5,43 3,80
Campomanesia dichotoma 25 1,63 12 40,00 1,78 1,74 20,00 54,11 5,15 3,37
Roupala montana 25 1,63 15 50,00 2,22 0,85 20,00 35,33 4,70 2,47
Miconia mirabilis 30 1,95 10 33,33 1,48 1,11 18,00 53,16 4,54 3,06
Psychotria carthagenensis 28 1,82 15 50,00 2,22 0,46 16,00 21,96 4,50 2,28
Maytenus obtusifolia 21 1,37 8 26,67 1,19 1,77 20,00 65,89 4,32 3,14
Eugenia flavescens 23 1,50 10 33,33 1,48 1,23 18,00 63,66 4,21 2,72
Ixora cf. brevifolia 26 1,69 10 33,33 1,48 0,94 18,00 32,79 4,12 2,64
Neoraputia magnifica 23 1,50 6 20,00 0,89 1,73 20,00 62,39 4,11 3,22
Eugenia florida 23 1,50 10 33,33 1,48 0,79 18,00 45,84 3,77 2,29
Banara guianensis 20 1,30 11 36,67 1,63 0,80 23,00 38,20 3,73 2,10
Zanthoxylum rhoifolium 17 1,11 7 23,33 1,04 1,10 30,00 39,15 3,25 2,21
Senegalia polyphylla 18 1,17 9 30,00 1,33 0,58 12,00 35,65 3,09 1,75
Inga marginata 18 1,17 8 26,67 1,19 0,73 20,00 33,74 3,08 1,90
Guatteria pogonopus 18 1,17 8 26,67 1,19 0,65 18,00 34,06 3,00 1,82
Ficus arpazusa 14 0,91 9 30,00 1,33 0,74 21,00 34,06 2,98 1,65
Manilkara rufula 13 0,85 6 20,00 0,89 1,22 22,00 56,66 2,96 2,07
Ocotea puberula 10 0,65 9 30,00 1,33 0,89 18,00 52,84 2,87 1,54
Casearia sylvestris 15 0,98 11 36,67 1,63 0,23 15,00 19,42 2,84 1,21
Inga laurina 14 0,91 7 23,33 1,04 0,84 22,00 36,61 2,79 1,75
Myrciaria tenella 15 0,98 8 26,67 1,19 0,59 30,00 37,56 2,75 1,57
Miconia prasina 17 1,11 9 30,00 1,33 0,30 12,00 22,60 2,74 1,41
Alchornea glandulosa 12 0,78 7 23,33 1,04 0,81 22,00 41,38 2,63 1,59
Pourouma mollis 12 0,78 7 23,33 1,04 0,78 22,00 40,74 2,60 1,56
Eugenia ligustrina 18 1,17 8 26,67 1,19 0,22 14,00 20,37 2,58 1,40
Geonoma pohliana 13 0,85 9 30,00 1,33 0,40 13,00 26,10 2,58 1,24
Astronium fraxinifolium 9 0,59 7 23,33 1,04 0,91 22,00 79,58 2,54 1,50
Byrsonima nitidifolia 18 1,17 5 16,67 0,74 0,61 19,00 39,79 2,52 1,78
Hyeronima oblonga 13 0,85 6 20,00 0,89 0,73 15,00 33,42 2,47 1,58
Eugenia aff. florida 11 0,72 4 13,33 0,59 0,86 16,00 63,66 2,17 1,57
Coccoloba latifolia 11 0,72 5 16,67 0,74 0,70 11,00 51,25 2,15 1,41
37
Classified - Internal use
Faramea sp. 14 0,91 5 16,67 0,74 0,43 11,00 29,60 2,08 1,34
Randia armata 9 0,59 7 23,33 1,04 0,38 15,00 31,83 2,00 0,96
Ilex sapotifolia 9 0,59 6 20,00 0,89 0,49 20,00 37,88 1,97 1,08
Coutarea hexandra 11 0,72 7 23,33 1,04 0,14 15,00 15,92 1,89 0,86
Campomanesia sp. 7 0,46 5 16,67 0,74 0,66 15,00 44,88 1,85 1,11
Eugenia sp. 11 0,72 5 16,67 0,74 0,37 14,00 34,06 1,83 1,09
Clusia panapanari 5 0,33 2 6,67 0,30 1,10 14,00 70,66 1,73 1,43
Vismia guianensis 10 0,65 4 13,33 0,59 0,45 21,00 36,61 1,69 1,10
Myrcia cf. fenestrata 7 0,46 5 16,67 0,74 0,44 14,00 32,79 1,64 0,90
Solanum maranguapense 8 0,52 5 16,67 0,74 0,37 18,00 28,65 1,63 0,89
Esenbeckia grandiflora 8 0,52 5 16,67 0,74 0,34 18,00 36,61 1,60 0,86
Artocarpus heterophyllus 4 0,26 3 10,00 0,44 0,77 18,00 50,93 1,48 1,03
Pourouma guianensis 7 0,46 3 10,00 0,44 0,42 18,00 31,19 1,32 0,88
Chrysophyllum gonocarpum 6 0,39 5 16,67 0,74 0,16 17,00 22,92 1,29 0,55
Ocotea longifolia 5 0,33 4 13,33 0,59 0,26 17,00 34,06 1,17 0,58
Manihot carthaginensis 7 0,46 3 10,00 0,44 0,27 19,00 24,51 1,17 0,73
Symplocos nitens 5 0,33 4 13,33 0,59 0,24 18,00 27,37 1,16 0,57
Handroanthus impetiginosus 9 0,59 2 6,67 0,30 0,28 16,00 24,83 1,16 0,87
Clusia melchiorii 6 0,39 3 10,00 0,44 0,30 10,00 31,19 1,13 0,69
Clusia paralicola 4 0,26 4 13,33 0,59 0,25 12,00 31,83 1,10 0,51
Myrciaria glazioviana 5 0,33 5 16,67 0,74 0,03 7,00 7,96 1,10 0,36
Sapium glandulosum 4 0,26 2 6,67 0,30 0,53 22,00 42,97 1,09 0,79
Bunchosia cf. acuminata 6 0,39 4 13,33 0,59 0,10 12,00 15,92 1,08 0,49
Ardisia guianensis 4 0,26 4 13,33 0,59 0,21 13,00 38,83 1,06 0,47
Byrsonima crispa 5 0,33 3 10,00 0,44 0,25 18,00 31,51 1,02 0,58
Ficus americana subsp. guianensis 6 0,39 3 10,00 0,44 0,17 20,00 29,28 1,00 0,56
Miconia amacurensis 5 0,33 4 13,33 0,59 0,08 14,00 14,64 1,00 0,40
Inga ingoides 4 0,26 3 10,00 0,44 0,28 10,00 37,56 0,98 0,54
Aspidosperma subincanum 6 0,39 3 10,00 0,44 0,12 12,00 20,05 0,95 0,51
Ocotea cf. indecora 3 0,20 3 10,00 0,44 0,30 8,00 47,75 0,94 0,49
Pseudobombax marginatum 4 0,26 3 10,00 0,44 0,23 16,00 30,56 0,93 0,49
Ficus calyptroceras 2 0,13 2 6,67 0,30 0,50 22,00 63,66 0,92 0,63
Myrsine guianensis 4 0,26 4 13,33 0,59 0,06 11,00 15,28 0,92 0,32
Chrysophyllum rufum 3 0,20 2 6,67 0,30 0,41 20,00 40,11 0,90 0,60
Senna macranthera 2 0,13 2 6,67 0,30 0,47 13,00 60,48 0,89 0,60
Triplaris gardneriana 4 0,26 2 6,67 0,30 0,30 16,00 41,06 0,85 0,56
Senna quinquangulata 3 0,20 2 6,67 0,30 0,35 18,00 40,43 0,85 0,55
Myrcia decorticans 3 0,20 3 10,00 0,44 0,20 12,00 32,47 0,84 0,39
Mangifera indica 1 0,07 1 3,33 0,15 0,60 15,00 70,35 0,81 0,66
Myrcia multiflora 1 0,07 1 3,33 0,15 0,57 20,00 69,07 0,79 0,64
Cordia trichotoma 3 0,20 2 6,67 0,30 0,29 19,00 39,79 0,78 0,48
Cordia cf. taguahyensis 3 0,20 2 6,67 0,30 0,26 18,00 33,74 0,75 0,45
Pouteria cf. reticulata 3 0,20 3 10,00 0,44 0,11 16,00 26,10 0,75 0,30
Trischidium molle 5 0,33 2 6,67 0,30 0,08 12,00 17,83 0,70 0,41
Ocotea cf. notata 4 0,26 2 6,67 0,30 0,15 18,00 23,55 0,70 0,41
Senna pendula 3 0,20 3 10,00 0,44 0,06 13,00 14,01 0,70 0,26
Schweiggeria floribunda 2 0,13 2 6,67 0,30 0,24 15,00 41,38 0,67 0,37
Styrax camporum 3 0,20 1 3,33 0,15 0,27 12,00 41,38 0,61 0,46
Anadenanthera colubrina 2 0,13 2 6,67 0,30 0,15 17,00 31,83 0,58 0,28
Oreopanax capitatus 2 0,13 2 6,67 0,30 0,13 10,00 28,01 0,56 0,26
Thysordium spruceanum 2 0,13 2 6,67 0,30 0,11 20,00 26,10 0,53 0,24
Dulacia guianensis 4 0,26 1 3,33 0,15 0,11 8,00 24,83 0,52 0,37
Alseis floribunda 1 0,07 1 3,33 0,15 0,29 10,00 49,02 0,50 0,35
38
Classified - Internal use
Syagrus cearensis 2 0,13 2 6,67 0,30 0,08 16,00 18,78 0,50 0,21
Apeiba tibourbou 2 0,13 2 6,67 0,30 0,07 11,00 22,92 0,50 0,20
Ouratea cuspidata 2 0,13 2 6,67 0,30 0,06 12,00 16,87 0,49 0,19
Coccoloba obtusifolia 4 0,26 1 3,33 0,15 0,07 9,00 18,46 0,48 0,33
Matayba guianensis 2 0,13 2 6,67 0,30 0,04 10,00 17,51 0,47 0,17
Cabralea canjerana 2 0,13 1 3,33 0,15 0,18 15,00 31,51 0,46 0,31
Heisteria blanchetiana 2 0,13 2 6,67 0,30 0,01 7,00 7,00 0,44 0,14
Abarema jupunba 3 0,20 1 3,33 0,15 0,09 14,00 18,78 0,44 0,29
Byrsonima stipulacea 2 0,13 1 3,33 0,15 0,16 22,00 29,28 0,43 0,29
Jacaratia spinosa 1 0,07 1 3,33 0,15 0,21 11,00 42,02 0,43 0,28
Psychotria sp. 2 0,13 1 3,33 0,15 0,14 17,00 32,47 0,41 0,27
Celtis spinosa 3 0,20 1 3,33 0,15 0,04 9,00 11,46 0,38 0,24
Clusia melchioriii 1 0,07 1 3,33 0,15 0,15 8,00 35,65 0,37 0,22
Poincianella bracteosa 2 0,13 1 3,33 0,15 0,08 16,00 24,83 0,36 0,21
Myrtaceae sp. 2 0,13 1 3,33 0,15 0,06 15,00 18,46 0,34 0,19
Acnistus arborescens 1 0,07 1 3,33 0,15 0,11 17,00 29,60 0,32 0,17
Andira inermis 1 0,07 1 3,33 0,15 0,09 18,00 27,06 0,30 0,15
Faramea multiflora 2 0,13 1 3,33 0,15 0,02 8,00 10,19 0,30 0,15
Cassia grandis 1 0,07 1 3,33 0,15 0,08 14,00 26,42 0,30 0,15
Prunus myrtifolia 1 0,07 1 3,33 0,15 0,06 10,00 21,96 0,27 0,12
Eugenia aff. culicina 1 0,07 1 3,33 0,15 0,06 20,00 21,65 0,27 0,12
Ficus americana 1 0,07 1 3,33 0,15 0,06 15,00 21,65 0,27 0,12
Guarea macrophylla subsp.
tuberculata
1 0,07 1 3,33 0,15 0,05 10,50 20,05 0,26 0,11
Byrsonima gardneriana 1 0,07 1 3,33 0,15 0,05 13,00 19,74 0,26 0,11
Pereskia grandifolia 1 0,07 1 3,33 0,15 0,04 7,00 19,10 0,26 0,11
Luetzelburgia auriculata 1 0,07 1 3,33 0,15 0,04 9,00 18,46 0,25 0,11
Pouteria sp. 1 0,07 1 3,33 0,15 0,04 8,00 18,46 0,25 0,11
Myrcia spectabilis 1 0,07 1 3,33 0,15 0,04 8,00 17,51 0,25 0,10
Myrcia sp. 1 0,07 1 3,33 0,15 0,03 9,00 16,55 0,25 0,10
Brosimum gaudichaudii 1 0,07 1 3,33 0,15 0,02 13,00 14,32 0,24 0,09
Mimosa arenosa 1 0,07 1 3,33 0,15 0,02 12,00 14,01 0,24 0,09
Eugenia sp.2 1 0,07 1 3,33 0,15 0,02 15,00 12,73 0,23 0,08
Posoqueria longiflora 1 0,07 1 3,33 0,15 0,02 8,00 12,73 0,23 0,08
Guettarda viburnoides 1 0,07 1 3,33 0,15 0,01 10,00 11,14 0,23 0,08
Simaba guianensis 1 0,07 1 3,33 0,15 0,01 9,00 11,14 0,23 0,08
Eugenia acutata 1 0,07 1 3,33 0,15 0,01 11,00 10,82 0,23 0,08
Palicourea rudgeoides 1 0,07 1 3,33 0,15 0,01 2,00 9,55 0,22 0,08
Micropholis guianensis 1 0,07 1 3,33 0,15 0,01 9,00 8,59 0,22 0,07
Acalypha villosa 1 0,07 1 3,33 0,15 0,01 7,50 7,96 0,22 0,07
Garcinia gardeneriana 1 0,07 1 3,33 0,15 0,01 7,00 7,32 0,22 0,07
Gustavia augusta 1 0,07 1 3,33 0,15 0,00 6,50 6,05 0,22 0,07
Myrcia aff. racemosa 1 0,07 1 3,33 0,15 0,00 6,00 5,41 0,22 0,07
Myrciaria sp. 1 0,07 1 3,33 0,15 0,00 5,00 4,77 0,22 0,07
Senna hirsuta 1 0,07 1 3,33 0,15 0,00 4,00 4,77 0,22 0,07
39
Classified - Internal use
Table 3. Richness, Shannon-Wiener (H’), Simpson index (D), Pielou’s evenness (J’),
species and families with highest IV (importance value) among each topographic
category and the whole mountain.
Richness
Species/Families
H’ D J’ Species highest IV Families highest IV
Top
(TMA)
54/25 3.233 0.062 0.810 Mollinedia ovata
Myrcia splendens
Guapira nitida
Nyctaginaceae
Myrtaceae
Monimiaceae
Windward
(WMA)
78/35 3.859 0.031 0.886 Myrcia splendens
Guapira nitida
Cupania impressinervia
Myrtaceae
Nyctaginaceae
Rubiaceae
Leeward
(LMA)
95/34 4.033 0.024 0.887 Handroanthus serratifolius
Pilocarpus spicatus
Senegalia poliphylla
Bignoniaceae
Rutaceae
Fabaceae
SEMA 144/44 4.182 0.027 0.841 Myrcia splendens
Guapira nitida
Mollinedia ovata
Myrtaceae
Nyctaginaceae
Fabaceae
40
Classified - Internal use
Table 4. Pearson correlation test (r) of passive ecological variables with axes 1 and 2 of
correspondence analysis. *significant at the 0.01 level.
Ecological variables Axis 1 Axis 2
Temperature - 0.715* 0.415*
Precipitation - 0.580* 0.802*
Elevation - 0.086 - 0.058
Slope - 0.095 0.190*
41
Classified - Internal use
Figure legends
Figure 1. Map of Ceará state, Brazil showing the mountain forests locations (in gray) and
detail of the studied area with altitudinal scale, where the delimited region comprises the
protected area of the Maranguape Mountain (SEMA). Datum, SAD 69, Date, March
2017.
Figure 2. Sample design: A the distribution of the areas with transects in each topographic
category, B an example area with transects and its distances measurements and C the
method of sampling trees in each point-centered quarter, until two tree by quarter. *If the
first sampling individual has less than 30cm of PBH (perimeter at breast height), we
sampled another one with more than 30cm.
Figure 3. Correspondence analysis by the abundance of species for the three sample areas
at Serra de Maranguape, northeastern Brazil. Eigenvalues: Axis 1 = 50.5%, and Axis 2 =
20.3%. P value = 0.001 in both axes.
Figure 4. Canonical correspondece analysis between environmental parameters and
species distribution of leeward, windward and top sides from Baturité and Maranguape
mountains. Eigenvalues: Axis 1 = 68.8% , and Axis 2 = 60% . P value = 0.025 in both
axes.
Figure 5. Unweighted pair group method with arithmetic mean with Jaccard coefficient
as distance measure, illustrating the similarity in floristic patterns at the species level
among Mountain Forests and Caatingas of Ceará state. Cophenetic coefficient = 0.78
See table 1 for locality names.
Figure 6. Non-metric multidimensional scaling yielded by the species data showing the
floristic connections among Mountain Forests and Caatingas of Ceará state. Eigenvalues:
Axis 1 = 55.7%, and Axis 2 = 23.3%. P value (proportion of randomized runs with
stress ≤ observed stress) = 0.01 in both axes. Final stress = 13.71. See table 1 for locality
names.
42
Classified - Internal use
Figures
Figure 1
43
Classified - Internal use
Figure 2
44
Classified - Internal use
Figure 3
45
Classified - Internal use
Figure 4
46
Classified - Internal use
Figure 5
Figure 6
47
Classified - Internal use
CHAPTER II
POLLEN-BASED CHARACTERIZATION OF MONTANE FOREST TYPES IN
NORTH-EASTERN BRAZIL *
*MANUSCRIPT PUBLISHED IN REVIEW OF PALAEOBOTANY AND PALYNOLOGY
(https://doi.org/10.1016/j.revpalbo.2016.07.003)
Vincent Montade1,2*, Ivan Jeferson Sampaio Diogo3, Laurent Bremond1,2, Charly Favier1,
Itayguara Ribeiro da Costa4, Marie-Pierre Ledru1, Laure Paradis1, Eduardo Sávio Passos
Rodrigues Martins5, Julien Burte6, Francisco Hilder Magalhães e Silva7, Christiano Franco
Verola4
1Institut des Sciences de l’Evolution de Montpellier, Université de Montpellier, CNRS, IRD,
EPHE, Place Eugène Bataillon, 34095 Montpellier, Cedex, France
2Ecole Pratique des Hautes Etudes, 4-14 rue Ferrus, 75014 Paris, France
3Institute of Biology, Department of Plant Biology, State University of Campinas, UNICAMP,
13083-970, Campinas, SP, Brazil
4Departamento de Biologia, Centro de Ciências, Universidade Federal do Ceará, Pici, CEP
60455-760 Fortaleza, CE, Brazil
5Fundação Cearense de Meteorologia e Recursos Hídricos, Aldeota, Fortaleza CEP 60115-221,
CE, Brazil
6Centre de Coopéeration Internationale en Recherche Agronomique pour le Développement,
Tunis 1080, Tunisia
7Departamento de Educação, Universidade do Estado da Bahia, CEP 48970-000 Senhor do
Bonfim, BA, Brazil
48
Classified - Internal use
ABSTRACT
Surface soil samples were collected in three mountainous massifs in north-eastern Brazil
to characterize the different vegetation types according to their respective pollen assemblages.
Complementary approach between pollen and vegetation data shows that the pollen rain
accurately reflects the following three main forest types: i) a dense ombrophilous forest (or
tropical moist broadleaf forest) characterized by Myrtaceae associated with high percentages of
Miconia, Guapira, Ilex, Moraceae-Urticaceae undif. or Byrsonima, ii) a seasonal semi-deciduous
montane forest characterized by an increase of Arecaceae associated with Fabaceae-Mimosideae,
Myrtaceae, Piper, Cecropia, Urera and Mitracarpus, and iii) a seasonal deciduous forest
dominated by Fabaceae-Mimosideae and Arecaceae tree taxa associated with Alternanthera,
Cyperaceae and Mitracarpus. Using of botanical data from several plots of ombrophilous forest,
in which several surface soil samples have been collected, allows to roughly estimate the over-
and underrepresentation of pollen taxa relative to their floristic abundance. Furthermore,
distribution of surface soil samples at different altitude and mountain sides also allows to
characterize vegetation variation according to several environmental parameters. The
precipitation increase with altitude is confirmed as the main environmental factor controlling
vegetation distribution. However, the forests located close to the crest with a proportion increase
of pollen taxa characteristic of heliophilous and pioneer trees (Alchornea, Miconia, Clusia), are
also influenced by changes of edaphic conditions. In addition to provide useful information in
understanding of fossil pollen records, this approach improves our understanding of the
ecosystem functioning in mountainous massifs in north-eastern Brazil, an useful knowledge for
conservation or restoration purposes.
Keywords: South America, North-eastern Brazil, Montane rainforest, Tropical forest,
Vegetation distribution, Modern pollen spectra
49
Classified - Internal use
1. INTRODUCTION
Tropical mountains which harbor a high species diversity are considered as the
world's most important diversity hotspots (Barthlott et al., 2005). Indeed, interactions of
an extraordinary variety of wet and dry habitats in close proximity due to local altitudinal
climate gradients allows the coexistence of different vegetation types, contributing to the
high species richness in tropical mountains (Richter, 2008). In particular, these complex
landscapes generate climatically suitable habitats, that can shelter animals and plants from
hostile climates, which is especially important for species conservation in periods of
climate change (Shoo et al., 2011;Williams et al., 2003). In north-eastern Brazil,
mountainous massifs (or “brejo de altitude”), which form islands of moisture, where
tropical rainforests contrast with the dry forest (Caatinga) of the surrounding plain, could
be considered as such a type of habitat. Closely related to the Caatinga in terms of floristic
composition (Rizzini, 1963), these montane rainforests could be result of the evolution
from this dry to a wet forest. On the other hand, these montane rainforests harbor endemic
species of Atlantic and Amazonian forests showing a past connection between these two
biomes (Andrade-Lima, 1982), which thus illustrates the capacity of such areas to shelter
species from long-term climate changes. Several botanical surveys performed in
mountainous massifs in north-eastern Brazil allowed describing composition, density and
diversity of different forest types (e.g., Andrade et al., 2006; Araújo et al., 2007; Costa-
Junior et al., 2008; Ferraz et al., 2004). The development of these forests is related to
altitude increase, generating high orographic precipitation.However, no studies focused
on the specific impact of environmental parameters (such as precipitation) on the
distribution of these forests. Although botanical data from this region provide a
representative set of different vegetation types, they arise from discontinuous sampling,
and consequently they do not allow to link the observed vegetation changes to the closely
50
Classified - Internal use
altitude-related climatic gradient. In comparison to extratropical mountains, vegetation
boundaries in tropical mountains are generally discrete (e.g., Ashton, 2003; Duarte et al.,
2005; Fernández-Palacios and de Nicolás, 1995; Hemp, 2005; Martin et al., 2007).
Discontinuities in vegetation composition may offer insights into the factors controlling
the assembly of plant communities and thus may have significance for management and
nature conservation. The present study aims to fill this gap by using pollen samples from
different plant communities along an altitudinal gradient with discontinuous botanical
data from several mountainous massifs in north-eastern Brazil. Studying modern pollen
samples in the tropics allows an accurate characterization of the different vegetation types
(e.g., Burn et al., 2010; Gosling et al., 2009; Jones et al., 2011). In addition, variations in
the pollen assemblages along altitudinal gradients are sensitive to altitudinal climate
changes (Cárdenas et al., 2014; Schüler et al., 2014; Urrego et al., 2011; Weng et al.,
2004). Here we focus on several mountainous massifs from northeastern Brazil to try to
answer the following questions: Does the modern pollen rain represent the composition
of different forest types? Using modern pollen data, do the present-day environmental
conditions explain the spatial distribution of different vegetation types? By providing a
better understanding of ecosystem functioning, the present study will be helpful in
defining conservations strategies for the rainforest of north-eastern Brazil with growing
human activities.
2. ENVIRONMENTAL SETTINGS AND METHODS
2.1 Study location
Our study was conducted in the state of Ceará of north-eastern Brazil (“Nordeste”)
in three isolated mountainous massifs named Pacatuba, Maranguape and Baturité, that
reach respectively up to 735 m, 920 m and 1115 m asl (Fig. 1). Located between 80 km
(Baturité) and 30 km (Pacatuba and Maranguape) from the coast, mineralogy of basement
51
Classified - Internal use
rocks mainly consists in gneissic facies interpreted as the erosional remnants of an Early
Cretaceous rift shoulder (Peulvast and de Claudino Sales, 2004). Characterized by steep
slopes and sinuous scarps, these mountainous massifs are contrasted with the topography
and climate of the surrounding plains. Climate is defined as hot and humid in mountainous
massifs, semi-arid in lowlands and sub-humid close to coast (FUNCEME;
http://www.funceme.br/). While the temperature with annual mean of c. 27 °C in the
lowlands shows no seasonal variations, rainfalls, mainly from February to May, show a
strong seasonality. Precipitation generated by maritime trade winds is distributed along
two main gradients: (1) a precipitation decrease with the distance from the coast, and (2)
a precipitation increase with the elevation. Annual precipitation ranges from c. 1400 mm
in the coastal area to 700 mm in the interior and reach values higher than 1600 mm in the
mountainous massifs. High evapotranspiration (generally between 2000 and 3000 mm yr
−1) causes an important water deficit resulting in a semi-arid climate. In mountainous
massifs, altitude increase that lowers temperature, results in cloud formation and
precipitation, enabling the development of a highly diversified tropical forest. From the
mountain top downwards, the following vegetation types are observed: a tropical moist
broadleaf forest corresponding to dense ombrophilous forest (sub-montane) according to
Veloso (2012), a seasonal semi-deciduous montane forest, and a seasonal deciduous
forest (Caatinga).
2.2 Field collection and data processing
Fieldwork was carried out in 2013 and 2014 and the modern pollen samples were
collected from surface soils. Each pollen sample consists of 10 sub-samples collected in
the upper 2 cm layer of surface soil and distributed on a plot between 1/2 and 1 ha under
homogeneous vegetation cover. In order to simultaneously obtain a representative set of
each vegetation type and a representative sampling of the climatic gradient, surface
52
Classified - Internal use
samples were located at different altitudes and mountainsides (Fig. 1 and Table 1). A
complete altitudinal transect on the lee- and windward sides was sampled in the Pacatuba
Massif with an altitude interval between each sample of c. 100 m (Fig. 1c). As the Baturité
and Maranguape massifs are more disturbed by human activity, surface soil samples were
collected above 600 m asl (Fig. 1b and d). To study the pollen-vegetation relationships,
samples were collected close to the locations of the botanical surveys performed by
Araújo et al. (2007) in Baturité Massif. In Maranguape Massif four surface soil samples
(MB9, MB7, MS7a and MS7b) were collected in botanical plots performed at the same
time. Each botanical plot consists of ten parallel transects of 100 m length. Following the
point-centered quarter method (Cottam and Curtis, 1956), the nearest trees with at least
15 cm perimeter at breast height were identified along these transects. Surface samples
were collected between two or three transects depending on the topography. For MS7,
because of slight difference in altitude, species composition and vegetation structure, the
plot was split in two sub-groups: a group of eight transects below 700 m asl corresponding
to the surface sample MS7a and a group of two transects above 700 m asl corresponding
to the surface samples MS7b.
After sampling, surface soil samples were transported to the laboratory and stored
in a cold room at 5 °C. To extract pollen grains, an aliquot of 2 cm3 for each sample was
chemically treated, adapting the method of Faegri and Iversen (1975). Surface samples
were processed by five successive KOH (10%) at 70 °C to remove humic acids followed
by HF (70%) to eliminate silicates and by the standard acetolysis method. Prior to
chemical treatment, a calibrated Lycopodium tablet was added to each sample in order to
calculate the pollen concentrations (Stockmarr, 1972). Palynomorphs were counted and
identified using a light microscope (Leica) at 1000×magnification after mounting slides
with residues and glycerine. At least 300 pollen grains were counted for each sample and
53
Classified - Internal use
119 pollen and spore taxa were identified using the reference collection of M.-P. Ledru
(IRD pollen collection) and several specialized publication on pollen morphology
(Colinvaux et al., 1999; Leal et al., 2011; Roubick and Moreno, 1991; Rull, 2003).
Fig. 1. General locationmaps of (a)mountainousmassifs in north-eastern Brazilwith the locations of their
respectivemodern pollen samples (b, c and d). (e) and (f) represent the transect location performed on
themean annual precipitation values based onWorldClimdatabase (Hijmans et al., 2005). Satellite image
romBingmaps) and elevation maps (fromASTER Global Digital Elevation Model from METI and NASA)
were plotted using QGIS software (QGIS Development Team, 2015).
All statistical analyses were performed on the pollen taxa with percentages ≥1%
and percentages were calculated on a pollen sum excluding fern spores. Pollen spectra
(Fig. 2) were plotted using PSIMPOLL 4.10 software (Bennett, 1994). Pollen-vegetation
54
Classified - Internal use
representativeness ratios were calculated for the four surface samples collected in the
botanical plot of the Maranguape Massif by calculating the p/v values (% of arboreal
pollen/%of total stems of all taxa inplot). Only for p/v values, pollen percentages were
calculated on the arboreal pollen sum(Appendix A). As described by Gosling et al. (2009)
this ratio is intended to assess the relative pollen productivity and dispersal of different
pollen taxa (Fig. 3). In order to determine correlations between the spatial distribution of
vegetation groups and environmental data from pollen assemblages, Correspondence
Analysis (CA) was carried out (Figs. 4, 5 and Appendix B). Correspondence Analysis is
a classically used ordination method to summarize the patterns of variations among a
collection of pollen assemblages, as it is well suited to contingency tables. Different or
more refined commonly used ordination methods were also tested (Principal
Correspondence Analysis, square-root transform of percentages, Detrended
Correspondence Analysis) but they lead to similar results. Environmental parameters
such as slope angle and distance to the crest were calculated using QGIS software (QGIS
Development Team, 2015) running the ASTER Global Digital Elevation Model (from
METI and NASA). Mean annual precipitation (MAP) values for the period 1950–2000
were obtained using the WorldClim database for each location of the surface samples.
As the Maranguape and Pacatuba massifs occupy relatively small areas, without
any meteorological stations, the climatic altitudinal gradient is not well resolved at
WorldClim resolutions. Contrariwise, from the Baturité Massif, being larger than the
Maranguape and Pacatuba massifs, more precise meteorological data are available, which
allow obtaining a realistic pattern of precipitation changes according to elevation. A
second estimation of MAP (hereafter MAP2) was then obtained by assuming that
altitude/precipitation patterns are the same over the three massifs. Altitude and MAP were
related along an altitudinal transect between thewind- and leeward side of the Baturité
55
Classified - Internal use
Massif and two non-linear regression fits were performed to model altitudinal variations
of MAP wind- and leeward (Fig. 1e and f).The obtained equations were applied to all
surface sample locations. Then, an increase of 200 mm was added for all samples of
Pacatuba and Maranguape in order to take into account the observed difference in MAP
recorded in rainfall stations located in the lowland at a greater proximity to the seashore.
(Hijmans et al., 2005). The first estimates of MAP is denoted hereafter MAP1 (Table 1).
3. RESULTS
3.1 Modern pollen assemblages
According to their respective altitudinal ranges and their main pollen taxa based
on percentage values, the different vegetation types can be characterized by the following
pollen assemblages (Fig. 2 and Table 2).
Table 1
Names and locations of surface samples with their ecological and environmental parameters. Lat S and
LonW: latitude south and west;MAP: Mean Annual Precipitation; Veg: vegetation type; DF: Deciduous
Forest; SDF: Semi-Deciduous Forest; OF: Ombrophilous Forest (*close to the crest).
Site Massif Lat
S
Lon
W
Altitude
m asl
Side Distance to
the crest m
Slope
°
MAP 1
mm
MAP 2
mm
Veg Human
disturbance
PS11 Pacatuba -3.98 -38.62 130 Windward 2529 10 1296 1228 DF Low
PS9 Pacatuba -3.98 -38.63 158 Windward 2514 26 1272 1269 DF Low
PS8 Pacatuba -3.99 -38.63 263 Windward 2350 16 1272 1401 SDF High
PS10 Pacatuba -3.97 -38.62 312 Windward 2056 32 1320 1451 SDF Low
PS7 Pacatuba -3.99 -38.63 383 Windward 2079 27 1272 1514 SDF Moderate
PS1 Pacatuba -3.97 -38.63 411 Windward 1720 17 1440 1536 SDF Moderate
PS6 Pacatuba -3.98 -38.63 569 Windward 1633 18 1416 1635 OF Low
PS5 Pacatuba -3.98 -38.63 631 Windward 1432 22 1428 1664 OF Low
PS2 Pacatuba -3.97 -38.63 650 Windward 633 3 1428 1672 OF Low
PS4 Pacatuba -3.98 -38.64 724 Windward 446 19 1428 1700 OF Undisturbed
PS3 Pacatuba -3.97 -38.64 759 Windward 204 10 1428 1711 OF Low
PC2* Pacatuba -3.97 -38.64 765 Windward 34 14 1464 1713 OF* Undisturbed
PS12 Pacatuba -3.98 -38.66 575 Leeward 414 18 1428 1544 SDF Moderate
PS13 Pacatuba -3.99 -38.66 489 Leeward 657 32 1296 1459 SDF Moderate
PS14 Pacatuba -3.99 -38.67 395 Leeward 986 23 1296 1353 DF Moderate
PS15 Pacatuba -3.99 -38.67 279 Leeward 1374 21 1284 1199 DF Moderate
PS16 Pacatuba -3.99 -38.67 160 Leeward 2022 5 1284 1010 DF Moderate
Bsal Baturité -4.26 -38.98 735 Leeward 580 14 1548 1476 DF Low
Bjar Baturité -4.29 -39.00 760 Leeward 1411 3 1512 1494 SDF Low
56
Classified - Internal use
Btav Baturité -4.30 -38.92 600 Windward 433 15 1500 1450 OF Low
Bsin Baturité -4.29 -38.93 647 Windward 840 17 1524 1470 OF Low
Barv Baturité -4.24 -38.93 816 Windward 350 12 1560 1528 OF Low
Blag* Baturité -4.21 -38.97 940 Windward 0 31 1584 1557 OF* Low
MB7 Maranguape -3.91 -38.72 730 Windward 800 16 1464 1702 OF Low
MS4* Maranguape -3.90 -38.72 850 Windward 37 21 1524 1737 OF* High
MB9* Maranguape -3.90 -38.72 934 Windward 30 17 1524 1756 OF* Undisturbed
MS5* Maranguape -3.89 -38.72 950 Windward 8 19 1524 1759 OF* Undisturbed
MS7a Maranguape -3.91 -38.73 700 Leeward 1260 6 1476 1650 OF Low
MS7b Maranguape -3.91 -38.73 659 Leeward 1500 12 1392 1617 OF Low
The dense ombrophilous forest (sub-montane): in this forest type, Myrtaceae
remains among the most frequent tree pollen taxa associated with high percentages of
Miconia (Melastomataceae), Guapira (Nyctaginaceae), Ilex (Aquifoliaceae), Moraceae-
Urticaceae undif. or Byrsonima (Malpighiaceae). The dominance of these pollen taxa is
generally observed from the upper limit of the semi-deciduous montane forest upwards.
However, pollen assemblages of ombrophilous forests located close to the crest or to the
mountain top reveal some differences. Indeed, in these samples we recorded a percentage
increase of Miconia, Alchornea (Euphorbiaceae) or Mitracarpus (Rubiaceae) and a slight
increase of Clusia (Clusiaceae). The altitude in which such assemblages are observed,
changes according to the maximum elevation of mountainous massifs, ~900 m asl at
Baturité and Maranguape and ~700 m asl at Pacatuba.
57
Classified - Internal use
Fig. 2. Percentage pollen diagram of surface soil samples of the three mountainous massifs studied in the
north-eastern Brazil. Pollen taxa in bold represent arboreal elements and the other correspond to shrubs or
herbs. Samples located on the mountain top are indicated by “*”.
The seasonal semi-deciduous montane forest: in this forest type, Arecaceae, the
most frequent tree pollen taxon, is associated with Fabaceae-Mimosideae, Myrtaceae,
Piper (Piperaceae), Cecropia (Urticaceae), Urera (Urticaceae) and Mitracarpus. These
assemblages are recorded up to ~500 m asl on the windward side and up to ~600 m asl
on the leeward side. The seasonal deciduous forest: the tree pollen signature is mainly
characterized by Fabaceae-Mimosideae and Arecaceae and the percentages of herbaceous
pollen taxa, such as Alternanthera (Amaranthaceae), Cyperaceae and Mitracarpus,
increase. This pollen assemblage is observed from the base up to ~200 m asl on the
windward side and up to ~400 m asl on the leeward side. In Baturité Massif, located more
58
Classified - Internal use
inland than massifs of Pacatuba and Maranguape, this pollen assemblage is observed up
to ~700 m asl on the leeward side.
Table 2
Main pollen taxa ordered by their respective percentage values for each surface soil samples. Arboreal
pollen taxa are represented in bold. Surface soil samples are ordered according to axis 1 values of the
correspondence analysis (CA) except for the five underlined samples removed from the CA. Color scale of
surface samples from black to light gray represents the three main vegetation types, fromthe ombrophilous,
semi-deciduous to deciduous forest (samples with “*” indicate ombrophilous forest located close to the
crest).
Names Main pollen taxa
MS5* Miconia 50, Guapira 15, Serjania-Cupania type 4, Myrsine 4, Myrtaceae 3, Alchornea 2
MS4* Alchornea 63, Miconia 11, Piper 5, Poaceae 3, Mitracarpus 2
MB9* Moraceae-Urticaceae undif. 21, Miconia 13, Asteraceae tubuliflorae 8, Myrtaceae 5, Guapira 4, Alchornea 4, Solanum
3
MB7 Myrtaceae 26, Guapira 19, Miconia 7, Mitracarpus 5, Borreria 5, Tetrapteris 4, Ilex 3, Moraceae-Urticaceae undif. 2
PS4 Myrtaceae 40, Guapira 24, Mitracarpus 4, Cyperaceae 3, Poaceae 3
PS3 Guapira 29, Mitracarpus 12, Myrtaceae 9, Protium 7, Piper 4, Clusia 4, Moraceae-Urticaceae undif. 4, Miconia 3
PS2 Guapira 16, Zanthoxylum type 14, Anthurium type tetragonum 14, Mitracarpus 7, Cyperaceae 6, Myrtaceae 4,
Arecaceae 4, Miconia 3
Blag* Guapira 17, Mitracarpus 16, Zanthoxylum type 7, Myrsine 6, Myrtaceae 6, Clusia 5, Ilex 5
PS5 Ilex 40, Mitracarpus 12, Myrtaceae 9, Arecaceae 5, Cyperaceae 4, Moraceae-Urticaceae undif. 3
MS7a Myrtaceae 16, Ilex 9, Fabaceae-Mimosideae 7, Mitracarpus 6, Piper 5, Miconia 4, Cyperaceae 4, Gallesia 4, Byrsonima
4
MS7b Myrtaceae 22, Mitracarpus 13, Guapira 9, Zanthoxylum type 6, Tetrapteris 5, Moraceae-Urticaceae undif. 4,
Fabaceae-Mimosideae 3, Piper 3
PC2* Mitracarpus 35, Ilex 10, Guapira 9, Myrtaceae 7, Zanthoxylum type 3, Asteraceae tubuliflorae 3, Clusia 3
PS6 Arecaceae 19, Mitracarpus 16, Borreria 6, Protium 6, Piper 4, Myrtaceae 4, Coutarea 4, Guapira 4
Barv Byrsonima 81, Mitracarpus 6, Myrtaceae 3, Miconia 2
Bsin Byrsonima 53, Mitracarpus 8, Miconia 7, Myrtaceae 3
Btav Byrsonima 23, Miconia 21, Mitracarpus 14, Poaceae 10, Sida 2, Moraceae-Urticaceae undif. 2
PS1 Mitracarpus 14, Borreria 10, Arecaceae 8, Piper 7, Tetrapteris 6, Cecropia 4, Moraceae-Urticaceae undif. 4, Guapira 4,
Fabaceae-Mimosideae 4
PS12 Cecropia 31, Arecaceae 9, Piper 8, Mitracarpus 8, Moraceae-Urticaceae undif. 6, Myrtaceae 5, Astronium 4,
Fabaceae-Mimosideae 4, Miconia 3
PS7 Arecaceae 29, Cecropia 20, Piper 7, Miconia 7, Fabaceae-Mimosideae 4, Zanthoxylum type 3
PS13 Arecaceae 28, Protium 12, Tetrapteris 11, Piper 6, Cecropia 4, Zanthoxylum type 4, Mitracarpus 4, Fabaceae-
Mimosideae 3
PS8 Urera 43, Arecaceae 17, Cecropia 7, Mitracarpus 5, Myrtaceae 4, Fabaceae-Mimosideae 3, Poaceae 2
PS10 Arecaceae 31, Myrtaceae 14, Fabaceae-Mimosideae 8, Mitracarpus 5, Zanthoxylum type 5, Moraceae-Urticaceae
undif. 4
Bsal Mitracarpus 34, Piper 14, Alternanthera 12, Fabaceae-Mimosideae 7, Moraceae-Urticaceae undif. 5, Arecaceae 4,
Myrtaceae 3, Cecropia 2
PS11 Fabaceae-Mimosideae 25, Mitracarpus 20, Cyperaceae 17, Althernanthera 11, Myrtaceae 8, Arecaceae 4, Poaceae 3
Bjar Mitracarpus 32, Alternanthera 24, Fabaceae-Mimosideae 8, Arecaceae 8, Sapium 4, Myrtaceae 3
PS16 Clidemia 22, Mitracarpus 20, Alternanthera 13, Fabaceae-Mimosideae 12, Cecropia 6, Arecaceae 5, Cyperaceae 3
PS14 Fabaceae-Mimosideae 36, Arecaceae 24, Senna 5, Cecropia 4, Piper 3, Moraceae-Urticaceae undif. 3
PS15 Fabaceae-Mimosideae 50, Alternanthera 12, Mitracarpus 9, Arecaceae 8, Cecropia 4
PS9 Fabaceae-Mimosideae 68, Arecaceae 8, Cyperaceae 7, Mitracarpus 3, Poaceae 2
59
Classified - Internal use
3.2. Representativeness of pollen taxa in the ombrophilous forest
Within the botanical plots of the ombrophilous forest studied in the Maranguape
Massif, in which surface soil samples have been collected (MB9, MB7, MS7a, MS7b),
between thirteen and seventeen woody taxa are present in both the vegetation and pollen
rain (see Appendix A). Principal among them is Myrtaceae, which is the most common
taxon (~20% stems) in all the plots. In order to compare the relative representativeness of
the pollen rain to the vegetation, we selected the p/v values for the twelve taxa that are
most significant in the vegetation and pollen rain (Fig. 3). Although the p/v values can
vary between plots for a same taxon, we recognized three main groups of important taxa.
The taxa of the first group, characterized by Moraceae-Urticaceae undif. (mean p/v = 5.1),
Miconia (3.9), Alchornea (2.3) and Areaceae (2), are highly overrepresented in the pollen
rain relative to their abundance in the vegetation. Except for Arecaceae, the p/v values for
all the plots are higher than 2. The second group, that is also overrepresented, is
characterized by taxa with mean p/v values between 1 and 2: Guapira (mean p/v= 1.6),
Myrtaceae (1.3) and Byrsonima (1.2). The taxa Rubiaceae undif. (mean p/v= 0.9), Clusia
(0.8), Fabaceae-Mimosideae (0.8), Zanthoxylum (Rutaceae) type (0.7) and Serjania-
Cupania type (0.2) that correspond to the third group, are generally underrepresented in
the pollen rain relative to their abundance in the vegetation. Several taxa, Bignoniaceae,
Lauraceae, Monimiaceae and Phyllanthaceae, including significant woody species such
as, Handroanthus serratifolius, Nectandra cuspidata, Cinnamomum triplinerve,
Mollinedia ovata, Margaritaria nobilis, are completely absent in the pollen rain.
60
Classified - Internal use
Fig. 3. p/v values according to arboreal taxa present in both the pollen rain and vegetation of the botanical
plots from Maranguape Massif (in black). The p/v ratio corresponds to “% of arboreal pollen / % of total
stems of all taxa in plot”. Values higher and lower than 1 respectively characterize the pollen taxa over-
and underrepresented. Taxa in gray are absent in the pollen rain.
3.3. Multivariate data analysis
Among the 91 different pollen taxa identified, 64 pollen taxa occurring with a
frequency higher than 1% have been used to perform the CA. A first CA analysis clearly
separates a group that includes all samples except five outlier samples (see Appendix B).
These samples characterize the ombrophilous forest of the Baturité Massif with high
frequencies of Byrsonima (Btav, Bsin and Barv) or highly disturbed vegetation (PS8 and
MS4). In order to improve ordination results and provide a better understanding of the
relationships between pollen assemblages and natural vegetation distribution, the outlier
samples were removed and a new CA was performed (Fig. 4). Eigenvalues for the first
and second axes represent respectively 17.8% and 11.7% of the total variation. The
distribution of samples in the CA diagram (Fig. 4a) displays a continuous pattern along
axis 1 which reflects the vegetation variation from the seasonal deciduous forest, the
seasonal semi-deciduous montane forest to the dense ombrophilous forest (sub-montane).
61
Classified - Internal use
Along this gradient we observe a clear distinction between these three main vegetation
types. Ombrophilous forests located close to the crest or to the mountain top are not well
defined by the CA except for two samples from the Maranguape Massif (MS5 and MB9).
Several passive environmental parameters (Altitude, MAP1, MAP2, Slope and Distance
to the crest) have been projected in the axes 1–2 bi-plot of the CA. As shown in Fig. 4b
and by the correlation coefficient (Table 3), the environmental parameters are mainly
correlated with axis 1. The closest environmental parameters related to axis 1 are MAP2
(0.87), altitude (0.81), MAP1 (0.67) and distance to the crest (−0.66). The main
correlation concerning axis 2 is with the Slope (0.19).
Table 3
Pearson correlation test (r) of passive ecological variables with axes 1 and 2 of correspondence analysis.
Fig. 4. Bi-plot of correspondence analysis for axes 1 and 2 with (a) distribution of surface soil samples and
(b) projection of passive environmental parameters. Empty circles represent samples of ombrophilous forest
located close to the crest.
Ordination of the pollen taxa according to axis 1 values reveals a distribution
pattern similar to the pollen taxa distribution observed in our synthetic pollen diagram
according to the altitudinal gradient (Fig. 5a). Altitude and precipitation values of surface
62
Classified - Internal use
samples plotted according to axis 1 values reflect the general structure of the three main
vegetation types (Fig. 5b and c). We observe that the samples of the seasonal deciduous
forest remain below 400 m asl with an annual precipitation lower than 1400 mm and the
samples of the seasonal semi-deciduous montane forest remain below 600 m asl with an
annual precipitation of 1400–1600 mm. Two samples of the Baturité Massif (Bjar and
Bsal) differ from this general pattern showing altitude values higher for their respective
forest type than at the Pacatuba Massif. For the ombrophilous forest, all samples (except
PS6), are located above 600 m asl with an annual precipitation higher than 1600 mm.
63
Classified - Internal use
Fig. 5. (a) The distribution of main pollen taxa percentages discussed in the text, (b) the altitude values of
surface soil samples and (c) the mean annual precipitation of surface soil samples according to axis 1 values
of the correspondence analysis.
4. Discussion
4.1. Pollen-vegetation relationship and characteristic pollen taxa
One of the most typical features of the dense ombrophilous forests (sub-montane)
of the mountainous massifs from north-eastern Brazil, is the dominance of Myrtaceae
which represents the most diversified tree family (Lima et al., 2009; Siqueira et al., 2001).
A dominance of Myrtaceae is evident in the massifs of Baturité and Maranguape, for
instance, only in Baturité more than forty species of Myrtaceae have been identified
(Araújo et al., 2007). In the pollen rain, although Myrtaceae is not always the most
abundant pollen taxon, it remains among the main arboreal pollen percentages in the
ombrophilous forest in all mountainous massifs (Fig. 2 and Table 2). However, in both
the vegetation and pollen rain, the association of Myrtaceae with other dominant taxa
allows to characterize different communities within the ombrophilous forests. In the
Baturité Massif, among the most dominant trees, species of Myrtaceae such as Myrcia
splendens are frequently associated with Byrsonima sericea, Clusia nemorosa, Miconia
cecidophora or Ilex sapotifolia (Araújo et al., 2007; Cavalcante et al., 2000). Pollen
contents of this vegetation type (Barv, Bsin and Btav) from surface soil samples collected
in Baturité are characterized by a high amount of Byrsonima in the pollen rain associated
with Miconia and Myrtaceae (Fig. 2).
In the Pacatuba and Maranguape massifs, Byrsonima is no longer among the main
pollen taxa and Myrtaceae is frequently associated with Guapira, Miconia and Moraceae-
Urticaceae undif. in surface soil samples from the ombrophilous forest (Fig. 2 and Table
2). This difference in pollen contents observed between the Baturité and Pacatuba/
Maranguape massifs is also supported by botanical data. Indeed, at Maranguape, while
64
Classified - Internal use
species of Myrtaceae remain the most abundant and diversified tree family (see Appendix
A), the most common trees identified are M. splendens, Guapira nitida and Cupania
impressinervia. On the leeward side (MS7a and MS7b), the forest composition changes
slightly and the latter trees are frequently associated with Pilocarpus spicatus, H.
serratifolius and several species of Fabaceae-Mimosideae such as Senegalia polyphylla.
These forest composition changes are also partly observed in the corresponding pollen
assemblages of the leeward side with a slight increase of Fabaceae-Mimosideae and
Zanthoxylum type (including Pilocarpus) (Fig. 2).
The forests located closer to the crest or on the mountain top differ partly in their
structure and composition from the ombrophilous forest located just below. This forest
type is generally characterized by a lower canopy and a high proportion of epiphytic
plants. Among the trees, an increase of Alchornea glandulosa, Miconia mirabilis, C.
nemorosa and C. melchiorii is observed in the Baturité and Maranguape massifs (Araújo
et al., 2007 and Appendix A). In the tropical rainforest, these species are considered as
pioneer and heliophilous trees (Pessoa et al., 2012; Souza et al., 2006). In addition, several
species of Clusia (e.g., C. nemorosa) can facultatively apply crassulacean acid
metabolism (CAM) or C3 metabolism according to water availability (Lüttge, 2006;
Vaasen et al., 2006). Species characterized by this a useful adaptation have the ability to
adjust to dry conditions by switching to the CAM mode, but also back to the C3 mode
under humid conditions. The pollen rain from surface soil samples located close to the
crest also reveals these changes in composition of trees with high frequencies of
Alchornea, Miconia (MS5, MS4) or a slight percentage increase of Clusia (Blag).
Representativeness increase of such taxa in both the pollen rain and vegetation, probably
highlights specific environmental conditions prevailing on the mountain tops or areas
located close to the crest.
65
Classified - Internal use
The forest composition of seasonal semi-deciduous and deciduous montane
forests was studied in two plots on the leeward side of the Baturité Massif (Bjar and Bsal,
Araújo et al., 2007). In this zone, the Myrtaceae, which is no longer the structural family,
is replaced in abundance and diversity by the Euphorbiaceae (i.e. Sebastiania
macrocarpa, Manihot carthaginensis, Croton blanchetianus, Croton argyrophylloides,
Sapium obovatum), Fabaceae-Mimosideae (i.e. Mimosa caesalpiniifolia, Mimosa
arenosa) and Fabaceae-Caesalpinioideae (i.e. Bauhinia cheilantha). Species of Fabaceae-
Mimosideae and Euphorbiaceae are frequent components of the Caatinga growing
generally under low humidity conditions prevailing in the lowlands of northeastern Brazil
(Ferraz et al., 1998). Their associated pollen assemblages are characterized by a decrease
of Myrtaceae at the expense of Fabaceae-Mimosideae and by high frequencies of
herbaceous pollen taxa (Mitracarpus or Alternanthera). This change in pollen signature
highlights an opening of the forest canopy, as shown in other semi-arid regions in north-
eastern Brazil (dos Santos et al., 2015; Gomes et al., 2014). Most of the surface soil
samples of the semi-deciduous montane forest were collected at Pacatuba. They differ
from the deciduous forest samples by an increase in Arecaceae pollen frequencies (Fig.
2). Moreover, the pollen rain of the semi-deciduous forests shows a significant increase
of Cecropia, a pioneer tree taxon (Santo-Silva et al., 2013), and Piper indicating
secondary successional shrubs (Pearson et al., 2002). Recurrent droughts could favor the
development of these two taxa in comparison with the ombrophilous forest where higher
altitude and moisture could limit drought impact on vegetation. In addition, more frequent
human disturbance through this altitudinal range (e.g., banana crop) could also explain
the development of these taxa characteristic of forest fragmentation and regeneration.
4.2. Over- and underrepresented pollen taxa in the ombrophilous forest
66
Classified - Internal use
In order to compare directly the botanical with pollen data for the dense
ombrophilous forest, four surface soils samples have been collected in the same locations
as the botanical plots (MB7, MB9, MS7a and MS7b). In particular, calculating of p/v
values allowed us to obtain a rough assessment of the over- and underrepresentation of
each pollen taxon relative to their floristic abundance (Fig. 3 and see Appendix A).
Among the most significant tree taxa represented in the vegetation and pollen rain,
anemophilous taxa such as Moraceae-Urticaceae undif. and Alchornea are highly
overrepresented in all the plots. Characterized by a high pollen productivity and
dispersion, such anemophilous taxa are frequently among the most overrepresented in the
pollen rain (Burn and Mayle, 2008; Bush and Rivera, 2001; Gosling et al., 2005). Miconia
and Arecaceae which are not typical anemophilous taxa are also overrepresented in all
the plots. Generally pollinated by bees (Ishara and Maimoni-Rodella, 2011), species of
Miconia as M. mirabilis frequently identified in the Maranguape Massif are characterized
by flowers with a large number of well exposed anthers. This could favor the pollen
representation increase and can be attributed to a ‘messy pollination’ (Bush and Rivera,
2001; Horn and B, 1990).Well represented in pollen spectra, Guapira, Myrtaceae and
Byrsonima indicate mean p/v values N1 showing probably messy pollination syndromes.
Although taxa such as Myrtaceae have already been reported as overrepresented (Gosling
et al., 2009; Grabandt, 1980), the p/v values for these taxa are also variable and can reach
values b1 in several plots. The floristic composition changes between each plot, which
could partly explain the observed variability of p/v values. Other significant tree taxa
represented in the vegetation, Rubiaceae undif., Clusia, Fabaceae-Mimosideae, the
Zanthoxylum type and the Serjania-Cupania type, are rather underrepresented;
Bignoniaceae, Lauraceae, Monimiaceae and Phyllanthaceae are completely absent in the
pollen rain. Several of these taxa have already been reported as underrepresented or silent
67
Classified - Internal use
taxa, e.g., Clusia (Grabandt, 1980), Rubiaceae, Bignoniaceae, and the Serjania-Cupania
type (Gosling et al., 2009). Otherwise, our p/v values are also not always consistent with
previous trends observed in other regions. Among the most striking, the Zanthoxylum
type has been reported as overrepresented, while Miconia, generally included in
Melastomataceae, has been reported as underrepresented (Bush and Rivera, 2001;
Gosling et al., 2009). Apart from the fact that species included in pollen taxa are not
necessarily the same between the different regions, which certainly generates differences
in terms of pollen representation, our study focused on surface soil samples which also
could generate differences in terms of pollen preservation and representation in
comparison with studies focused on pollen traps. Although this approach based on p/v
values, including different biases, is not perfect, values for some taxa are in accordance
with predictions associated with their pollination strategy (e.g., Moraceae-Urticaceae
undif. and Alchornea) as it has been shown in different tropical regions (Bush, 1995; Bush
and Rivera, 2001; Gosling et al., 2005).
4.3. Impact of environmental factors on forest distribution
Modern pollen samples from the Maranguape, Pacatuba and Baturité massifs
closely reflect the different vegetation types along the altitudinal gradient. On the
windward side of Maranguape and Pacatuba, close to the coast, the lowland seasonal
deciduous forest is replaced by a seasonal semi-deciduous montane forest at 200m asl,
which in turn is replaced by a dense ombrophilous forest (sub-montane) at 500 m asl. The
same pattern is observed on the leeward side although each ecotone is located between
100 and 200 m higher than on the windward side. The spatial distribution of the vegetation
in mountainous areas is commonly related to precipitation and temperature changes
induced by the altitudinal gradient. The studied massifs from north-eastern Brazil are not
high enough to result in temperatures that could be considered as a direct limiting factor
68
Classified - Internal use
on the vegetation, as was already described from other tropical areas (Duarte et al., 2005;
Sarthou et al., 2009). Consequently, precipitation (MAP2) highly correlated with the
distribution of modern pollen samples on axis 1 of the CA represents the main factor
controlling the distribution of the different forest types (Table 3 and Fig. 4). Such a pattern
of precipitation changes is controlled by the elevation of air masses on the windward side
generating orographic precipitation. On the leeward side, the altitude decrease, warms up
and dries the air masses, which dissipate fogs, decrease rainfalls and increase evaporation.
This mechanism, the “foehn effect”, is frequently observed in tropical mountainous
islands like Hawaii (Giambelluca et al., 2013) and Cape Verde (Duarte et al., 2005). A
high correlation coefficient (r=0.82) between altitude and precipitation (MAP2) changes
of surface soil samples shows the influence of the altitude gradient on precipitation
increase and vegetation distribution.
While altitude increase represents one of the main factors influencing
precipitation, the distance from the coast is also an important factor for explaining
changes in precipitation rates. In particular, rainfall decrease is observed when the
distance from the coast increases. For example, at Baturité (80 km inland) the deciduous
forest grows up to 700 m asl on the leeward side (Bsal), while it remains below 400 m asl
on the same side at Pacatuba Massif (30 km from the coast). Slope angle could also
influence vegetation distribution. However, the slope angle has been generally shown as
a relevant factor at values N40° (Fernández-Palacios and de Nicolás, 1995). None of our
surface samples reaches such values, which could explain that no correlation is observed
between the slope angle values and the vegetation distribution revealed by the CA axes
(Table 3). Distance to the crest is inversely correlated with axis 1 of the CA, because
when this distance increases, altitude generally decreases, thus reflecting the decreasing
precipitation.
69
Classified - Internal use
In areas located close to the crest or on the mountain top, pollen assemblages from
surface soil samples reflect a different community of the ombrophilous forest, with
abundant heliophilous and pioneer tree taxa such as Alchornea, Clusia or Miconia.
Different edaphic conditions related to the rocky soil of the summit reduce the water-
holding capacities. Such local conditions could explain the development of plant
communities that are more dependent on atmospheric moisture and thus more strongly
correlate with changes therein. This is particularly well illustrated by a greater abundance
of epiphytic plants near the crest. In addition, associated with strong winds, the drying
effect and falling trees are more important in these areas. The strong development of
pioneer trees in these forests confirms strong dynamics and probably a high resilience
related to frequent natural disturbances. Consequently, this different forest community is
more disturbed and affected by the frequent droughts of this region than the dense
ombrophilous forest (Hastenrath and Heller, 1977). Moreover, although rainfalls would
reach their maxima close to the mountain tops, local environmental conditions induce
drier moisture conditions than in the dense ombrophilous forest, only located a few tens
of meters below. At this point, additional and exhaustive botanical surveys would be
essential to better describe such a forest community, preferably focusing on epiphytic
plant or liana diversity, which seems tobe important structural plant components at higher
elevation.
5. Conclusions
Our complementary approach between pollen analysis of surface soil samples and
present-day vegetation in mountainous massifs in north-eastern Brazil shows that the
pollen rain accurately reflects the different vegetation types. Based on the different pollen
assemblages, the following tropical forest types can be recognized: i) a dense
ombrophilous forest (or tropical moist broadleaf forest) characterized by Myrtaceae
70
Classified - Internal use
associated with high percentages of Miconia, Guapira, Ilex, Moraceae-Urticaceae undif.
or Byrsonima, ii) a seasonal semi-deciduousmontane forest characterized by an increase
of Arecaceae associated with Fabaceae-Mimosideae, Myrtaceae, Piper, Cecropia, Urera
and Mitracarpus, and iii) a seasonal deciduous forest dominated by Fabaceae-
Mimosideae and Arecaceae tree taxa associated with Alternanthera, Cyperaceae and
Mitracarpus. In areas located close to the crest, a different ombrophilous forest
community has been identified by the increase of Alchornea, Miconia or Clusia. Being
typical heliophilous and pioneer trees, these taxa indicate drier moisture conditions,
related to local environmental factors than in the dense ombrophilous forest just located
a few tens of meters below. Concerning the dense ombrophilous forest, pollen rain has
also been directly compared with data from several botanical plots. Although we observed
some differences with previous studies performed in tropical regions, pollen
representativeness for several taxa are conform to the predictions inferred from their
pollination strategies. In particular, anemophilous pollen taxa are always overrepresented
relative to their floristic abundance.
In addition to characterize each vegetation type, our ordination of pollen data,
based on surface soil samples distributed at different altitudes and mountain sides, allows
the reconstruction of the spatial distribution of different vegetation types. Unlike previous
studies, this approach combining vegetation and pollen data also allows studying the
influence of environmental factors on vegetation distribution. In particular, we
characterize the high importance of precipitation changes (associated with increasing
altitude) that represents the main environmental factor controlling vegetation distribution,
except for the forest located close to the crest in which changes of edaphic conditions also
seem to be important. Our study provides a better understanding of the ecosystem
functioning in mountainous massifs in north-eastern Brazil. As these areas can be
71
Classified - Internal use
considered as “islands” of tropical rainforest, this knowledge will be useful for
conservation or restoration purposes. Furthermore, development of this type of
comparison between pollen and vegetation data will be helpful, to improve our
understanding of fossil pollen records for reconstructing past vegetation dynamics,
climate changes and diversity processes of tropical rainforest in a semi-arid environment.
Acknowledgements
Financial support was provided by FUNCAP (CI1-0080-000170100/ 13),
FAPESP (BIOTA 2013/50297-0), NSF (DOB 1343578) and NASA and IRD. V.M.
benefited from a postdoctoral position funded by FUNCAP at FUNCEME (Brazil) and
EPHE at ISEM (France) I.J.S.D benefited from PhD position funded by CNPq (Brazil).
We thank William D. Gosling and the Editor of RPP for their very constructive and
helpful comments which have greatly contributed to improve this article. We also thank
Daniel Sabatier for the stimulating discussions and Claire-Line Montade for the help
during the fieldwork. This is Institut des Sciences de l’Evolution de Montpellier
publication n°2016-138 SUD.
72
Classified - Internal use
Appendix A. List of pollen taxa present in both the pollen rain and vegetation of the
botanical plots from Maranguape Massif including botanical and pollen data. Taxa
in gray represent plant families absent in the pollen rain. pbh: perimeter at breast
height
MB9
Pollen taxa Vegetation data Pollen data
No. of stems ≥ 10
cm dbh
% of total stems
(v) No. of
pollen % of pollen
(p) p/v
Myrtaceae 114 23.8 15 6.5 0.3
Guapira 62 12.9 13 5.7 0.4
Serjania-Cupania type 54 11.3 8 3.5 0.3
Miconia 28 5.8 39 17.0 2.9
Rubiaceae undif. 15 3.1 10 4.3 1.4
Clusia 13 2.7 3 1.3 0.5
Moraceae-Urticaceae undif. 13 2.7 64 27.8 10.3
Roupala 12 2.5 8 3.5 1.4
Alchornea 10 2.1 11 4.8 2.3
Ficus 6 1.3 2 0.9 0.7
Byrsonima 5 1.0 2 0.9 0.8
Solanum 3 0.6 9 3.9 6.2
Symplocos 3 0.6 4 1.7 2.8
Ilex 2 0.4 1 0.4 1.0
Senna 1 0.2 3 1.3 6.2
Pouteria 1 0.2 1 0.4 2.1
Acalypha 1 0.2 1 0.4 2.1
Total 343 71.6 194 84.3
Lauraceae 44 9.2
Monimiaceae 38 7.9
Fabaceae-Mimosideae 28 5.8
Bignoniaceae 2 0.4
Total 112 23.4
73
Classified - Internal use
MB7
Pollen taxa Vegetation data Pollen data
No. of stems ≥ 10
cm dbh
% of total stems
(v) No. of
pollen % of pollen
(p) p/v
Myrtaceae 100 18.7 86 37.9 2.0
Rubiaceae undif. 60 11.2 8 3.5 0.3
Guapira 46 8.6 62 27.3 3.2
Serjania-Cupania type 37 6.9 2 0.9 0.1
Fabaceae-Mimosideae 28 5.2 2 0.9 0.2
Zanthoxylum type 24 4.5 7 3.1 0.7
Miconia 20 3.7 24 10.6 2.8
Ficus 14 2.6 1 0.4 0.2
Moraceae-Urticaceae undif. 8 1.5 8 3.5 2.4
Ilex 7 1.3 10 4.4 3.4
Solanum 5 0.9 1 0.4 0.5
Arecaceae 4 0.7 3 1.3 1.8
Clusia 4 0.7 2 0.9 1.2
Sapium 4 0.7 1 0.4 0.6
Myrsine 3 0.6 2 0.9 1.6
Alchornea 2 0.4 2 0.9 2.4
Gallesia 1 0.2 2 0.9 4.7
Total 367 68.6 223 98.2
Bignoniaceae 21 3.9
Lauraceae 30 5.6
Phyllanthaceae 20 3.7
Monimiaceae 2 0.4
Total 73 13.6
74
Classified - Internal use
MS7a_L1to7
Pollen taxa Vegetation data Pollen data
No. of stems ≥ 10 cm dbh
% of total stems
(v) No. of
pollen % of pollen
(p) p/v
Myrtaceae 76 21.2 48 23.1 1.1
Fabaceae-Mimosideae 25 7.0 22 10.6 1.5
Zanthoxylum type 24 6.7 10 4.8 0.7
Guapira 22 6.1 8 3.8 0.6
Serjania-Cupania type 13 3.6 3 1.4 0.4
Rubiaceae undif. 12 3.3 7 3.4 1.0
Byrsonima 9 2.5 12 5.8 2.3
Arecaceae 8 2.2 5 2.4 1.1
Astronium 5 1.4 4 1.9 1.4
Aspidosperma 5 1.4 1 0.5 0.3
Moraceae-Urticaceae undif. 4 1.1 6 2.9 2.6
Miconia 3 0.8 12 5.8 6.9
Myrsine 1 0.3 3 1.4 5.2
Senna 1 0.3 2 1.0 3.5
Total 208 57.9 143 68.8
Bignoniaceae 21 5.8
Lauraceae 8 2.2
Phyllanthaceae 8 2.2
Total 37 10.3
75
Classified - Internal use
MS7bL8to10
Pollen taxa Vegetation data Pollen data
No. of stems ≥ 10
cm dbh
% of total stems
(v) No. of
pollen % of pollen
(p) p/v
Myrtaceae 33 20.2 68 33.0 1.6
Zanthoxylum type 19 11.7 17 8.3 0.7
Serjania-Cupania type 17 10.4 1 0.5 0.0
Byrsonima 11 6.7 8 3.9 0.6
Guapira 10 6.1 28 13.6 2.2
Fabaceae-Mimosideae 10 6.1 8 3.9 0.6
Rubiaceae undif. 9 5.5 10 4.9 0.9
Roupala 4 2.5 7 3.4 1.4
Astronium 4 2.5 2 1.0 0.4
Senna 3 1.8 1 0.5 0.3
Arecaceae 2 1.2 8 3.9 3.2
Ficus 2 1.2 1 0.5 0.4
Miconia 1 0.6 4 1.9 3.2
Total 125 76.7 163 79.1
Bignoniaceae 12 7.4
Phyllanthaceae 6 3.7
Lauraceae 4 2.5
Total 22 13.5
76
Classified - Internal use
Appendix B. Bi-plot of correspondence analysis with all modern pollen samples for
axes 1 and 2 (a) and for axes 1 and 3 (b). Only the names of outlier samples have
been indicated
77
Classified - Internal use
CHAPTER III
DISPERSAL AND EDAPHIC FACTORS DRIVING PLANT SPECIES COMPOSITION
IN MOUNTAIN FORESTS IN A SEMIARID REGION OF BRAZIL *
*MANUSCRIPT ACCEPTED BY PLANT ECOLOGY AND DIVERSITY
Ivan Jeferson Sampaio Diogoa*, Fernando Roberto Martinsa, Itayguara Ribeiro da
Costab and Flavio Antonio Maës dos Santosa
a Department of Plant Biology, University of Campinas – UNICAMP, Campinas, Brazil
b Department of Biology, Federal University of Ceará– UFC, Fortaleza, Brazil
78
Classified - Internal use
Abstract
Background: Species distribution may result from several factors, including the ability
to disperse, biotic and abiotic factors.
Aims: To understand the patterns of abiotic and biotic influence on species distribution
in the Brazilian northeastern mountain forests.
Methods: We related topography, climate, soil and biotic predictors to species
composition in 17 forest stands in northeastern Brazil, using non-metric multidimensional
scaling and cluster analysis. We built minimum adequate models to select the variables,
and a canonical correspondence analysis to relate environmental and biotic variables to
species composition.
Results: The NMDS showed a clear but gradual discrimination gradient from southern to
northern sites in the distribution and abundance of the surveyed 1427 species, belonging
to 216 genera of woody plants at 17 locations. The Zoochory and Soil Organic Carbon
model explained 43% of the correlation between the response variables and explanatory
variables. The CCA analysis indicated four different groups based on zoochory and SOM.
Conclusions: The zoochory and soil organic carbon predictors, with their associated
ecological processes, determine the niche of mountain forests’ woody species and may
be the decisive factors for the distribution of tropical trees in Brazilian Northeastern
mountain forests.
Keywords: model selection, soil organic carbon, zoochoric dispersion.
79
Classified - Internal use
Introduction
A fundamental issue in ecology is the understanding of the mechanisms that
generate the spatial distributions of species and their abundances at different spatial
scales. Disturbance processes and local site factors interact to regulate patterns and
processes in biological communities that lead to species coexistence and maintenance of
species diversity (Chesson 2000).
Much research has discussed the role of historical processes and environmental
factors in shaping community vegetation in tropical forests (Boulangeat et al. 2012; Chust
et al. 2006; Phillips et al. 2003). However, factors posited as the most important have
varied widely across studies, especially when considering different spatial scales and
areas of study (Normand et al. 2006; Tuomisto and Ruokolainen 2008).
Three main drivers could interact independently or simultaneously and influence
the observed spatial distribution of a given species: the abiotic environment, dispersal
limitation and biotic interactions (Austin 2007; Götzenberger et al. 2011; Soberon 2007).
The abiotic constraints may delimit the niche of a species, given its physiological limits
(Chase & Leibold 2003). Dispersal limitations may influence the distribution range of a
species by allowing or prevent it from reaching a suitable site; dispersal limitation is
linked to dispersal capability (Vellend et al. 2007). Finally, biotic interactions may affect
resource availability, with consequences for abundance (e.g. competition,
Lortie et al. 2004, Pons and Pausas 2006).
Climate and soil variables have been shown to be the most relevant variables
when predicting plant species distributions at regional to continental scales
(Thuiller et al. 2004). Biotic interactions are assumed to limit distribution at local scales
(Soberon 2007; Wiens 2011; Godsoe et al. 2017). Dispersal limitations act at larger
extents (Boulangeat et al. 2012). The mechanisms involved in the changes in species
80
Classified - Internal use
distribution by dispersal are still unknown for montane areas. So far, species distribution
has been modelled using species distribution models (SDMs), which ignore the effects of
dispersal and biotic interactions (VanDerWal ET AL. 2009).
Although these drivers of species distributions could act at different spatial scales,
we aimed to quantify how dispersal limitation and/or abiotic factors are related to the
floristic composition of mountain forests in north-eastern, Brazil. These humid forest
enclaves located in the semiarid region have functioned as ecological refuges for flora
and fauna since the Pleistocene, providing natural shelter of several endangered species
and new species not yet described and therefore should be priorities for conservation
(Andrade-Lima, 1982). Few studies and inventories are located on these areas and it
demands further attention especially given the extensive loss of habitat.
To understand the species composition and distribution in those endangered and
unique areas, we addressed the following questions: Is the distribution of species in the
northeastern mountain forests follow a gradient? Which variables are most important to
species distribution in those forests? Which is the relative role of the variables in shaping
floristic distribution? and How this influence vary among different sites? We
hypothesized that (1) geographically nearby areas are most correlated, (2) elevation and
rainfall are the environmental factors that have more influence, (3) dispersal limitation is
extremely important to shape the mountain forests community.
Material and methods
Study sites
Montane forests in north-eastern Brazil are defined as those forests that occur
between 600 and 1100 m a.s.l. Bordering or completely surrounded by caatinga, those
81
Classified - Internal use
mountain forests are locally named as brejo forests and have 1200 mm of average annual
rainfall due to an orographic effect (Tabarelli and Santos 2004, Silva and Casteletti,
2003). Steep slopes and sinuous scarps, contrasting with the topography of the
surrounding plains, characterize the mountainous massifs. Climate is defined as hot and
humid, with privileged conditions regarding soil and air humidity and temperature when
compared to the surrounding semiarid region (Tabarelli and Santos 2004).
We made a presence/absence matrix of woody plant species from all checklist
published for 17 montane forests sites in north-eastern Brazil, (Figure 1, Table 1).
Selection of abiotic and biotic variables
Environmental predictors were selected with the understanding of known
ecological, ecophysiological and biophysical processes, and were used as our explanatory
variables. Based on literature (Meier et al. 2010, Boulangeat et al. 2012), we selected 10
types of predictors and divided them into four classes: topographic (elevation and slope),
climatic (rainfall and average temperature), edaphic (pH, cation exchange capacity, total
clay, organic carbon and nitrogen) and biotic (dispersal syndrome). We used dispersal
syndrome as a proxy for dispersal limitation. The response variable was the species
ocurrence data that were available from the surveys. We then constructed a matrix using
the geographic data (the plot coordinates), response variable selected and explanatory
variables.
The climate data were compiled from WorldClim (Hijmans et al. 2005) and the
topography data was downloaded from SoilGrids (http://soilgrids.org) at a spatial
resolution of ca. 1 km. When available, we used the variables data from the paper itself.
Dispersion syndrome according to Pijl (1982) was classified as (1) anemochory, (2)
zoochory, (3) autochory or barochory, and (4) hydrochory, using literature reviews and
82
Classified - Internal use
analysis of herbarium specimens. The proportion of dispersion syndromes per community
was based on species richness.
Data analysis
In order to compare the relationships of species composition among the sites, we
carried out an unweighted pair group method with arithmetic mean (UPGMA)
agglomerative cluster analysis with Jaccard coefficient as measure of distance. To assess
how accurately the cluster dendrograms preserved the information from the original
matrices, we calculated the cophenetic correlation coefficient.
To reduce the dimensionality of the community data, we carried out a non-metric
multidimensional scaling (NMDS), also using the Jaccard coefficient. NMDS allows
floristic patterns to be captured without being constrained by a set of predictors (McCune
and Grace 2002). Each NMDS axis has a coefficient of determination (r²) resulting from
the correlation between the distances in the original space and the distances in the
ordination and tested its significance (P) after 999 Monte Carlo permutations. Those
distances were compared, and by iterative procedure, the differences between these arrays
were minimized by using the statistical stress (S) (Clarke and Warwick 2001).
To answer the question as to which variables were most strongly related to species
distribution, we built minimum adequate models (MAMs). Since autocorrelation in
explanatory and response variables violate the assumption of data independence, we first
tested the spatial autocorrelation of the variables before beginning to select the MAMs.
We calculated Moran’s I coefficient, which varies between +1 and -1, where positive
values indicate spatial aggregation (Legendre and Legendre 1998).
We used the first axis scores obtained by the NMDS as response variables to carry
out a multiple regression. Then, we tested the autocorrelation of the residuals obtained by
constructing a spatial correlogram with the Moran’s I coefficient for 10 variables classes
83
Classified - Internal use
and 999 permutations. We tested the correlogram significance with Bonferroni
corrections (Legendre and Legendre 1998).
If the residuals were not autocorrelated, we tested whether or not the explanatory
variables predicted the species abundance and distribution by fitting ordinary least
squares (OLS). The best models are generally those with the lowest Akaike information
criterion (AIC) values. However, for the MAMs, the best models are considered when the
difference between the AIC value of the model and the minimum AIC value of all models
(∆i) is lower than 2 (Diniz-Filho et al. 2008).
After fitting OLS, we tested the collinearity of the variables in the models with
the variance inflation factor (VIF). We also tested for autocorrelation in the residuals of
these models by constructing correlograms with Moran’s I coefficient, and then selected
the MAM that best adjusted to the assumptions of spatial independence and lack of
collinearity among the explanatory variables.
To see the correlation between the sites and the environmental variables selected
by the MAM, we made two canonical correspondence analyses (CCA, the first with the
variables selected by all models with ∆i <2, and the second with the variables of the model
selected). We made 999 permutations and used a Monte Carlo test to evaluate whether
the correlation between the biotic and abiotic matrices was different from that expected
by random chance. We carried out the analyses using the PC-ORD 6.0 software (McCune
and Mefford 2011), and the spatial correlograms and model selection using the SAM 4.0
software (Rangel et al. 2010).
Results
A total of 1427 species in 216 genera of woody plants were compiled for the
species data. The cluster dendrogram showed two different groups: A with all BNMFs
from Ceará and Paraíba and two from Pernambuco, and B with six BNMFs from
84
Classified - Internal use
Pernambuco and one from Sergipe (Figure 2, 87.5% of information remaining).
Moreover, in a finer resolution, group A was separated into three sub-groups: one with a
single site (BNMF 16), another formed by the CE samples (BNMFs 6, 9, 13 and 17) and
the third one formed by two PB samples (BNMF 1 and 3) and three PE samples (BNMF
5, 8 and 10). The correlations between the cophenetic similarities and the Jaccard original
similarities were strong (0.94), confirming the reliability of the UPGMA analyses.
The NMDS ordination provided a two-dimensional solution, with both axes being
significant (P = 0.01), and explained 60% of the correlations between the distances in the
original space and the ordination distances (Figure 3). After 49 iterations, the final stress
value was 13.74, which is a satisfactory result (McCune and Grace 2002). There was a
clear but gradual discrimination gradient from southern to northern BNMFs (Figure 3).
Furthermore, we found four different groups within both axis, in which the closest areas
had more similarity. The NMDS ordination identified the same groups as the UPGMA.
We found no autocorrelation for the explanatory and response variables, or for the
residuals. The AIC sorted a total of 4023 OLS models, from which four had ∆i < 2, with
four explanatory variables: zoochory, organic carbon, precipitation and nitrogen (Table
2).
In the first CCA (Figure 4, axis 1 = 48% and axis 2 = 31%, P > 0.05), two pairs
of variables were formed: zoochory and precipitation had a similar influence on the
species distribution inside the sites (Pearson R between variables = 0.87), and nitrogen
and organic carbon have the same behaviour (Pearson R = 0.92). This can be clearly
observed in the ordination diagram (Figure 4), in which a strong pattern of correlation
among the variables can be observed along the diagonals from the first and fourth
quadrants.
85
Classified - Internal use
The final model was selected because of the independence among data (smallest
VIF) among variables, the smallest ∆i and P < 0.05, included the explanatory variables
zoochory and organic carbon (Table 3), resulting in the following OLS final equation 1:
Y = 1.934 – 1.301zoochory + 0.753organic carbon (1)
The model Zoochory-OC explained 44% of the correlation between the response
variables and explanatory variables (∆i = 1.813). The other models (2, 3 and 4) presented
VIF greater than the zoochory-OC model.
The second CCA was based on the Zoochory-OC model (Figure 5A, axis 1 = 52%
and axis 2 = 27%). The eigenvalues indicated that the probability that is explained by the
environmental or explanatory variables is larger than would expected by random chance
(P = 0.003 to Monte Carlo randomisation test).
The CCA scatterplot indicated the formation of four distinct groups of sites
(Figure 5B). Group 1 – BNMFs 1-5 was found to be directly correlated with OC and
indirectly correlated with zoochory. Group 2 – BNMFs 7, 8 and 14 was directly correlated
with both variables. Group 3 – BNMFs 6, 9, 13 and 17 was directly correlated with
zoochory and indirectly correlated with OC. Group 4 – BNMFs 10, 11, 12, 15 and 16 was
wispy correlated to both variables. The first axis separated the effect of zoochory and OC
into two sides, while the second one highlighted the negative and positive effect of both
variables.
There was a remarkable transition among the groups, suggesting a gradual overlap
of the species. The similarity between CCA and nMDS showed a correlation between the
floristic composition and environmental-dependent distribution of species.
Discussion
Distribution of species across the Brazilian Northeastern Mountain Forests
86
Classified - Internal use
The variation in species composition across the BNMFs can be significantly
predicted by environmental factors and the connections among these forest types indicate
that they are quite divided accordingly to the geographical location. The existence of
certain differences between groups (Figures 2 and 3) does not exclude the possibility of
connections among them. This is demonstrated by the difference between the northern
and southern part of the BNMFs that were shown to be connected in terms of species
composition by the cluster dendrogram, despite the consistent identity of each group (the
same pattern is found in Oliveira-Filho et al. 2005, 2006, Eisenlohr and Oliveira-Filho
2015).
Furthermore, these floristic connections were strongly supported by the ordination
analysis. The transition between these groups is gradual, however it is possible to detect
some changes in species composition between the groups along the gradients. These
findings disagree with the validity about the northern part of Atlantic Forest (see Oliveira-
Filho et al. 2006; Carnaval et al. 2008), showing that even those north-eastern mountain
forests are quite different among them. Andrade et al. (2007) have shown this difference
in genetic level for populations of Monstera adansonii var. klotschiana from these areas.
Although we agree with the conservation implications suggested by Santos and
Tabarelli (2004) and Santos et al. (2007) to the BNMFs, our study sheds light on a number
of additional interesting phytogeographic patterns. We establish a new viewpoint from
which the Brazilian Northeastern Mountain Forests should no longer be considered as a
single group. In fact, we still have a lot to explore and to address from this
phytogeographic region.
Variables and their role in shaping floristic distribution
In most studies of Neotropical region, a large portion of the floristic variance is
not explained by the variables under consideration, the variables can show redundancy in
87
Classified - Internal use
explain species distribution and there are a lot of indirect effects of variables (Ter Braak
1987, Austin 2007, Meier et al. 2010). We found that both soil and dispersal syndrome
were related to the distribution of tree species. Climate, topography, and spatial forces,
among other factors have been well-established for the Atlantic Forest (Oliveira-Filho &
Fontes 2000, Scudeller et al. 2001, Oliveira-Filho et al. 2005, Marques et al. 2011,
Eisenhlor and Oliveira-Filho 2015). However, our study is the first to analyse them in the
north-eastern mountain forests context and to test dispersal limitation.
Conversely to our expectation that elevation and rainfall would be the main factors
to explain differences in species composition (see Eisenlohr and Oliveira-Filho 2015), we
found that zoochory and organic soil carbon content were the most important variables
related to differences in species composition. The fact that these predictors passed a
statistical selection process, through which the relationships between the candidate
predictors and the species distributions were evaluated and all collinearities were
discarded, supports the interpretation of them as a factor that affect floristic patterns. It
has been studied that soils components are very important to plant distribution throughout
the world (Jobbágy and Jackson 2000, John et al. 2007, Coudon et al. 2006, Vicent and
Meguro 2008, Nieto-Lugilde 2014, Horn et al. 2015). However there are still few studies
concerning dispersal limitation as a driver in floristic distribution (Primack and Miao
1992, Pearson and Dawson 2003, Boulangeat et al. 2012).
The soil organic carbon influencing plant distribution has been substantiated by
other studies (Jobbágy and Jackson 2000, Motta et al. 2002, Freschet et al. 2013, Roy et
al. 2013). The main content of OC is the organic matter, located at the litterfall or in the
top 20 or 30 cm of soil, which has a great contribution to the plant allocation (carbon
cycle). Furthermore, we observed that the different groups formed by the CCA (Figure 5)
have similar type of soils and similar amount of soil components, which indicate the
88
Classified - Internal use
grouping. Most of the BNMFs located in PE, PB and SE has eutrophic soils where the
organic matter produced is accumulated, and they belong to groups 1 and 2 (higher
amounts of OC). The opposite is found for CE’s mountain forests, which have distrophic
soils.
On the other hand, few studies have used functional traits such as dispersion
syndrome as a proxy for dispersal limitation (Calba et al. 2014, Mair et al. 2014).
However, it is already known that the dispersion mode has an influence on the
establishment of species (Guisan and Zimmermann 2000; Seidler and Plotkin 2006).
Zoochory is the most common dispersion category in different ecosystems (Howe and
Smallwood 1982). The differential prevalence of this category is most probably the result
of ecological adaptations to successional niches (Kupfer and Malanson 1993). Most of
CE’s mountain forests and three from PE are less disturbed areas and consequently have
greater amount of zoochoric species when compared to the others (Brown Jr. and Brown
1992, Drezner 2001, Diogo et al. 2016).
Conclusions
The patterns described here suggest that the BNMFs species not only withstand
soil constraints but also exhibit variations in their dispersion strategies across the
mountain forests, resulting in a highly species-rich flora. Furthermore, a considerable
number of Atlantic forest tree species are known to be flexible with respect to habitat
(Linares- Palomino et al. 2010).
As part of the variance remains unexplained (what is common in plant ecology,
Ter Braak 1986), the main challenge in the future is to explain floristic patterns with
additional variables (spatials, for instance) and to assess the role of stochastic event in
space and time. This study is among the first to apply selection model to the region,
89
Classified - Internal use
providing an idea for the use of this tool in biodiversity and conservation of restricted and
endangered areas.
Acknowledgements
We thank Luciana C. Franci for the help with statistical analysis.
Disclosure statement
No potential conflict of interest was reported by the authors.
Funding
This work was supported by the Conselho Nacional de Desenvolvimento
Científico e Tecnológico [grant number #479263/2011-6 and #563537/2010-8) and by
the Fundação Cearense de Apoio ao Desenvolvimento Científico e Tecnológico [grant
number #PP1-0033-00025.01.00/10]. The research missions between UFC and
UNICAMP was supported by Coordenação de Aperfeiçoamento de Pessoal de Nível
Superior [grant number 552213/2011-0].
Notes on contributors
Ivan J. S. Diogo is a Ph.D. student whose research interests include the biogeography,
diversity and composition of mountain forests
Itayguara R. da Costa is a professor with interest in the evolution and systematics of
plants..
Fernando R. Martins is a professor with research interests in community ecology
Flavio A. M. dos Santos is a professor with interests in plant population ecology.
90
Classified - Internal use
Tables
Table 1. Floristic surveys of 17 mountain forests used in this study. Code: BNMF –
Brazilian northeastern mountain forests.
Municipality/State Code Geographic
coordinates (S;W)
References
Areia/PB BNMF1 6º57’48”; 35º41’29” Andrade et al. (2006)
São Vicente Férrer/PE BNMF2 7º34’60”; 35º30’00” Ferraz and Rodal (2008)
Maturéia/PB BNMF3 7º10’07”; 37º23’04” Cunha (2010)
Bonito/PE BNMF4 8º29’40’’; 35º41’45’' Rodal et al. (1998)
Catende/PE BNMF5 8º40’01’’; 35º43’00” Costa-Junior et al. (2008)
Guaramiranga/CE
Caruaru/PE
Triunfo/PE
Baturité/CE
Araripe/CE
Pesqueira/PE
Itaparica/PE
Maranguape/CE
Madre de Deus/PE
Serra da Guia/SE
Ubajara/CE
Pacatuba/CE
BNMF6
BNMF7
BNMF8
BNMF9
BNMF10
BNMF11
BNMF12
BNMF13
BNMF14
BNMF15
BNMF16
BNMF17
4º15’48”; 38º55’59”
8º16’60’’; 35º57’59”
7º50’18”; 38º06’05”
4º19’43”; 38º53’05”
7º16’32”; 39º27’13”
8º20’27”; 36º46’59”
9º04’60”; 38º17’59”
4º00’33”; 38º48’49”
8º11’14”; 36º24’60”
9°58’55”; 37°52’06”
3º51’16”; 40º55’16”
3°59'04"; 38°37'12"
Cavalcante et al. (2000)
Tavares et al. (2000)
Ferraz et al. (2003)
Araújo et al. (2007)
Ribeiro-Silva et al. (2012)
Pereira et al. (2010), Pinto et al. (2012)
Rodal & Nascimento (2002)
Diogo et al. In press
Rodal & Nascimento (2008)
Machado et al. (2012)
Not published
Not published
91
Classified - Internal use
Table 2. Ordinary least squares (OLS) model selection, sorted by Akaike information
criterion (AIC), coefficient of determination (r²) and minimum AIC value (∆i) of
variables.
OLS Models Variables r² AIC ∆i
Model 1 Zoochory, Organic Carbon 0.440 38.564 1.673
Model 2 Precipitation, Organic Carbon 0.399 39.359 1.695
Model 3 Zoochory, Nitrogen 0.371 39.412 1.813
Model 4 Precipitation, Nitrogen 0.359 39.498 1.867
Table 3. Non-standard coefficients of the explanatory variables selected by ordinary least
squares model selection, sorted by Akaike information criterion and the respective
variance inflation factors (VIF). (T test, P < 0.05)
Explanatory
variable
Coefficient
VIF
Standard
error
T
P
Zoochory -1.301 1.094 0.087 -1.667 0.126
Organic carbon 0.753 1.094 0.016 1.491 0.167
92
Classified - Internal use
Figures
Figure 1. Geographic distribution of Brazilian Northeastern Mountain Forests, with the
location of the floristic and phytosociological checklists used in this study in black
squares: six in Ceara (CE), eigth in Pernambuco (PE), three in Paraiba (PB) and one in
Sergipe (SE).
93
Classified - Internal use
Figure 2. Unweighted pair group method with arithmetic mean with Jaccard coefficient
as distance measure, illustrating the similarity in floristic patterns at the species level
among Brazilian Northeastern Mountain Forests. Cophenetic coefficient = 0.94. See table
1 for locality names.
94
Classified - Internal use
Figure 3. Non-metric multidimensional scaling yielded by the species data showing the
floristic connections among the Brazilian Northeastern Mountain Forests. R² values: Axis
1 = 44.17%, and Axis 2 = 16.36%. P value (proportion of randomized runs with
stress ≤ observed stress) = 0.02 in both axes. Final stress = 13.74. BNMF16 is a unitary
group. See table 1 for locality names.
95
Classified - Internal use
Figure 4. Ordination diagram produced by Canonical Correspondence Analysis for sites
are represented by circles with their identifications, and environmental variables by
vectors. OC = Organic Carbon, Precipit = Precipitation. See table 1 for locality names.
96
Classified - Internal use
Figure 5. A. Ordination diagram produced by Canonical Correspondence Analysis for
sites vs. the explanatory variables selected by MAM. OC = Organic Carbon. B. CCA
Scatterplot, where the explanatory variables are the axis of the analysis, showing four
different groups. See table 1 for locality names.
97
Classified - Internal use
CHAPTER IV
ELEVATIONAL GRADIENTS OF MOIST FOREST IN BRAZILIAN SEMIARID: A
DISTINCT BIOREGION
*MANUSCRIPT FORMATED TO APPLIED VEGETATION SCIENCE
IVAN DIOGO1*, KARIN SANTOS2, ITAYGUARA COSTA 3, ALEXANDRE ANTONELLI4
& FLAVIO SANTOS 1
¹University of Campinas - UNICAMP, Institute of Biology, Department of Plant Biology, 13083-970.
Campinas, SP, Brazil *[email protected]
²Swedish Museum of Natural History, Botany Department, P.O. Box 50007 SE-104 05 Stockholm, Sweden
³Federal University of Ceará - UFC, Departament of Biology, Sciences Center, 60455-760. Fortaleza,
CE, Brazil
4University of Gothenburg - Department of Biological and Environmental Sciences, Carl Skottsbergs
gata 22B - P.O. Box 461 - SE 405 30 - Göteborg - Sweden
98
Classified - Internal use
Abstract
There is a general agreement that the history of the Neotropical forests is complex, as
they have passed for several changes over evolutionary time. The separation of Amazon
and Atlantic Forest and the increasing number of isolated forest patches occurred during
the last Terciary and Quaternary. Assuming that the Brazilian Northeastern Mountain
Forests (BNMF) are considered microrefugia we tested three hypotheses: 1) Amazon and
Atlantic forests share the larger number of species with BNMF, 2) Caatinga has a great
influence on BNMF flora and 3) BNMF are an unique and interconnected bioregion. We
extracted all floristic dataset of Amazon, Atlantic forest, Cerrado and Caatinga from
NeoTropTree and summed with 17 studies of BNMF. To test the relationship among all
domains, we used Venn Diagrams, WPGMA, MRPP and NMDS analysis. To test if the
BNMF is a distinct bioregion, we performed the Infomap Bioregion Analysis. Amazon
presented more than half (5798 spp., 50.23%) of exclusive species, followed by Atlantic
(2089spp., 18,1%), Cerrado (183 spp., 1,59%), BNMF (168 spp., 1,46%) and Caatinga
(64 spp., 0.55%). The other amount of species (3240 spp., 28.77%) is shared by two,
three, four or five domains, for instance 185 species were accounted for occurring in all
domains. The cluster dendogram and NMDS grouped Cerrado, Atlantic Forest and
Caatinga together and Amazon and BNMF separately. We identified 26 bioregions of
woody plants in South America region and BNMF as a distinct one. The most common
species in BNMF were Manilkara rufula, Wedelia villosa and Guettarda angelica. The
most indicative species were Guettarda angelica and Manilkara rufula. Our results
indicates that past connections between the Neotropical rainforests resulted from specific
climatic conditions and support BNMF as an unique biogeographical region as they were
shaped by different climate conditions.
Key words: Neotropical, Amazon, Atlantic forest, Caatinga, biogeography.
99
Classified - Internal use
Introduction
During the last Terciary/Quaternary, successive climate cyles and
Pleistocene/Holocene glacial and interglacial fluctuations have played an important role
in determining the origin and distribution of living organisms in the world (Hewitt 2000).
The temperature descent through Last Glacial Maximum (LGM, 23~18 kyr BP), and the
warmer climate on the Last Interglacial (LIG, 130~116 kyr BP) had a great impact on
retraction and expansion of vegetation and accounted for the high biological diversity and
biogeographical history of plant formations (Ledru et al. 1996). These climatic
fluctuations would cause the forest formations to be fragmented by open dry formations,
such as savannahs, and the forest fragments would be isolated in refugia areas, where
species accumulate new mutations and new species would arise from the widely
distributed ancestral species (Refuge theory, Haffer 1969). Throughout the years, the most
discussed mechanism of diversification and endemism in the Neotropics is the existence
and effects of forest refugia, mainly in South America Neotropical area (Prance 1982,
Colinvaux et al. 2000, Ledru 2002, Bush & de Oliveira 2006, Graham et al. 2006, Carnal
& Moritz 2008, Rull 2008, Mello-Martins 2011, Bueno et al. 2016).
The Brazilian region was covered by Amazon and Atlantic rainforests, which
together formed a big interconnected and continuous forest (Morley 2000). However, this
continuum was historically compressed and fragmented by an open area (semi-arid
formations) due to LGM and LIG climatic fluctuations, which was responsible by the
separation of Amazon and Atlantic and the increasing number of isolated forest patches
(Bigarella et al. 1975, Ledru et al. 1998, Behling & Negrelle 2001, Behling 2002, Hoorn
et al. 2010, Mello-Martins 2011). Paleovegetation, pollen data and biogeographical
studies have shown evidences of past connections between northern Atlantic and eastern
Amazon (Cole 1960, Andrade-Lima 1966, 1982, Behling et al. 2000, Auler & Smart 2001,
100
Classified - Internal use
Auler et al. 2004, Wang et al. 2004, Santos 2007, Pellegrino et al. 2011, Fouquet et al.
2012). The open and drier area is delimited by Caatinga, Cerrado and Chaco (Ab’Saber
1977).
Despite the expansion of drier physiognomies in lowland areas, the tropical
rainforest persists in historically stable areas (Carnaval & Moritz 2008), for instance:
gallery forests across Cerrado in the center region of Brazil and humid patches of
Caatinga in Northeast (Rizzini 1963, Bigarella et al. 1975, Redford & Fonseca 1986,
Oliveira-Filho & Ratter 1995, Coimbra-Filho & Câmara 1996, Ledru et al. 2002). Many
rainforest tree taxa are supposed to have occurred across Caatinga during the wet phases
of LGM and LIG cycles, suggesting the existence of ecological corridors connecting
Amazon and Atlantic, which is also reinforced by the disjunct distribution of several taxa
in both rainforests (Martini et al. 2007, Turchetto-Zolet et al. 2013, Fouquet et al. 2014).
Specifically, the mountainous regions in the middle of Brazilian semiarid have
been considered as exceptionally rich centers of biodiversity and endemism (Tabarelli
2001, Pôrto et al. 2004). Given their complex topography and habitat heterogeneity, there
is likely to be a high genetic diversity within most montane species and signals of the
genetic and evolutionary process that occurred as a consequence of the Pleistocene glacial
cycles (Haffer 1969, Gentry 1982), and therefore should be priorities for conservation.
There have been studies to verify the effects of LGM and LIG on the diversity of
Neotropical montane species (e.g., Aguirre-Planter et al. 2000, Santos et al. 2007,
Jaramillo-Correa et al. 2008, Koscinski et al. 2008, Carnaval et al. 2009, Hensen et al.
2011). However, only a few of them focused in northern part of Northeast region,
revealing lowland expansion of the tropical rainforest in this subregion that lasted for
4000 years during the last deglaciation (Behling et al. 2000, Ledru et al. 2002, 2007,
Montade et al. 2014).
101
Classified - Internal use
Assuming that the Brazilian Northeastern Mountain Forests (hereafter called by
BNMF) are considered microrefugia areas for showing climatic conditions to support
rainforest species, that bioregion is a geographically distinct assemblages of species and
communities (Vilhena & Antonelli, 2015) and that domain is an area which share the
same phytogeographical, morfoclimatic and geomorphological characteristics
(Ab’Saber, 2003), the aim of our research is to check changes in plant communities along
five domains in order to test three hypotheses: 1) Amazon and Atlantic forests share a
higher number of species with BNMF, 2) Caatinga has a great influence on BNMF flora
and 3) BNMF are an unique and interconnected bioregion.
Material & Methods
Study area
Throughout the semiarid (caatinga) region, the annual rainfall ranges from 240 to
900 mm and falls within 3–7 months (IBGE, 1985; Lins, 1989). On the other hand, there
are “islands” of rainforest vegetation in the middle of the semiarid associated with the
occurrence of plateaus and mountains between 600 and 1100 m (for instance, Borborema,
Araripe, Baturité and Ibiapaba). Bordering or completely surrounded by caatinga, those
mountain forests are locally named as brejo forests, where the annual rainfall exceeds
1200 mm (Andrade-Lima, 1982). They have privileged conditions regarding soil and air
humidity, temperature and vegetation cover when compared to caatinga and present
species with Amazon and Atlantic distribution (Tabarelli and Santos 2004).
There are 49 mountain forests, distributed in the states of Ceará, Rio Grande do
Norte, Paraiba, Sergipe and Pernambuco, which cover 19 259 km2, however, this number
is actually reduced by half due to the anthropization and loss of habitat (Vasconcelos-
Sobrinho, 1971). However, only 17 (concerning plants) were studied until now: two in
102
Classified - Internal use
Paraíba state (PB), eight in Pernambuco state (PE), six in Ceará state (CE) and one in
Sergipe state (SE) (Table 1).
Table 1. Floristic surveys of 17 mountain forests located in Northeast, Brazil.
City/State Geographic
coordinates (S;W)
References
Areia/PB 6º57’48”; 35º41’29” Andrade et al. (2006)
São Vicente Férrer/PE 7º34’60”; 35º30’00” Ferraz & Rodal (2008)
Maturéia/PB 7º10’07”; 37º23’04” Cunha (2010)
Bonito/PE 8º29’40’’; 35º41’45’' Rodal et al. (1998)
Catende/PE 8º40’01’’; 35º43’00” Costa-Junior et al. (2008)
Guaramiranga/CE
Caruaru/PE
Triunfo/PE
Baturité/CE
Araripe/CE
Pesqueira/PE
Itaparica/PE
Maranguape/CE
Madre de Deus/PE
Serra da Guia/SE
Ubajara/CE
Pacatuba/CE
4º15’48”; 38º55’59”
8º16’60’’; 35º57’59”
7º50’18”; 38º06’05”
4º19’43”; 38º53’05”
7º16’32”; 39º27’13”
8º20’27”; 36º46’59”
9º04’60”; 38º17’59”
4º00’33”; 38º48’49”
8º11’14”; 36º24’60”
9°58’55”; 37°52’06”
3º51’16”; 40º55’16”
3°59'04"; 38°37'12"
Cavalcante et al. (2000)
Tavares et al. (2000)
Ferraz et al. (2003)
Araújo et al. (2007)
Ribeiro-Silva et al. (2012)
Pereira et al. (2010), Pinto et al. (2012)
Rodal & Nascimento (2002)
Diogo et al. In press
Rodal & Nascimento (2008)
Machado et al. (2012)
Not published
Not published
These plant surveys represented the best-published information available on
woody plant species composition at the local level and they yet covered a small part of
the area occupied by mountain forests. Plant surveys were conducted based on different
methods, some of them used plant sample criteria (from 5 cm DBH to all woody plant
species) and another part used only trees and shrubs over 3 m in height.
Dataset
103
Classified - Internal use
The floristic dataset of Amazon, Atlantic forest, Cerrado and Caatinga was
extracted from NeoTropTree (Oliveira-Filho, 2014), a database that consists of tree
(defined as free-standing woody plants > 3 m in height) species checklists for > 2000 geo-
referenced sites compiled from the literature and herbarium specimen records. In addition
to that, we updated the list with recent studies available from the literature. A
NeoTropTree site is defined by a single vegetation type founded in a circular area with a
5-km radius. If there are two or more vegetation types occurring in one area, we
considered this as overlapping sites and excluded from the dataset. The NeoTropTree only
includes occurrence records with an indication or evidence of vegetation type and with
complete species lists, which is important to avoid bias caused by different sampling
efforts across the same sites. The species were checked regarding their taxonomy and
synonymies by using two different databases (available
at http://floradobrasil.jbrj.gov.br/ and http://tropicos.org/).
We extracted species from all domains: 7907 of Amazon, 4298 of Atlantic forest,
2916 of Cerrado and 1078 of Caatinga. Those species were summed to 733 species
recorded to BNMF (Fig. 1) to build a presence/absence matrix. The plant families and
species were listed according to the Angiosperm Phylogeny Group IV guidelines (APG
IV, 2016).
Data analysis
Both the numbers of shared and exclusive species at the five domains were
demonstrated using Venn diagrams, in which we compared dry domains (Cerrado and
Caatinga) and wet domains (Amazon and Atlantic) with BNMF. In order to evaluate the
floristic relationships among the domains, a dendrogram displaying the hierarchical
clustering was employed with the Jaccard coefficient as a measure of distance, and the
WPGMA linking method as the cluster algorithm. To assess how accurately the cluster
104
Classified - Internal use
dendrograms preserved the information from the original similarity matrix, we calculated
the cophenetic correlation coefficient.
We tested the hypothesis that all five domains are heterogeneous in terms of
community composition (distance and similarity) by multiresponse permutation
procedure (MRPP). In order to determine the floristic patterns among the domains
resulting from the correlation between the distances in the original space and the floristic
distances, we performed non-metric multidimensional scaling (NMDS, two dimensions
multivariate analysis), with Jaccard coefficient as a measure of similarity. Although a
common practice, we did not exclude rare species from these analyses as they have little
influence in NMDS (Muotka et al. 2002). The adequacy of the ordination for
interpretation was evaluated by using a coefficient of determination (r²), a statistical stress
(S), which its significance (P) was tested after 1,000 Monte Carlo permutations (Clarke
& Warwick 2001).
105
Classified - Internal use
Figure 1. Geographic Distribution of the sites recorded in this study with the location and domain.
Note that BNMF sites are located in the middle of Caatinga domain.
All analyses were performed using the R (R Development Core Team 2013). The
VennDiagram package (Chen & Boutros 2011) was used to create the Venn diagram and
the Vegan (Oksanen et al. 2009) and Cluster (Maechler et al. 2013) packages were used
for multivariate analyses.
In order to test if the BNMF can be classified as a bioregion, we downloaded geo-
referenced ocurrences of the recorded species from the global biodiversity information
facility (GBIF) by the speciesgeocodeR package (Töpel et al. 2017). We only included
106
Classified - Internal use
collections without known coordinate issues, based on observations and specimens, and
used the ‘CleanCoordinateLarge’ function of the sampbias package (<
https://github.com/azizka/sampbias_beta/tree/master>) in R to remove potentially
erroneous coordinates.
We cleaned this dataset using the R functions of the same package, checking for
empty coordinates, species reported in the sea and coordinates assigned to country or
province centroids, and then we uploaded on the Infomap Bioregion software (Edler et
al. 2017). The software displays different bioregions on maps and provides a list of the
most common and indicative species for each bioregion. We used the following
parameters: maximum cell capacity = 500; minimum cell capacity = 100; maximum cell
size = 8°; minimum cell size = 2°. See < http://bioregions.mapequation.org/ > for an
explanation of these settings and further software documentation.
Results
A total of 11 542 species, 1 141 genera and 159 plant families were encountered
in the five domains studied (Appendix 1). Amazon presented more than half (5798 spp.,
50.23%) of exclusive species, followed by Atlantic (2089spp., 18,1%), Cerrado (183 spp.,
1,59%), BNMF (168 spp., 1,46%) and Caatinga (64 spp., 0.55%). The other amount of
species (3240 spp., 28.77%) is shared by two, three, four or five domains, for instance
185 species were accounted for occurring in all domains (Fig. 2). From them, we can
name Casearia grandiflora Cambess., Inga ingoides (Rich.) Willd. and Tapirira
guianensis Aubl., which are spread all over Brazil. A great amount of species is
distributed only between Amazon and Cerrado (898 spp.); Atlantic and Cerrado (455
spp.) Amazon, Atlantic and Cerrado (419 spp.); Atlantic, Cerrado and Caatinga (236
spp.). On the other hand, few species are shared only by Amazon, Cerrado and Caatinga
(9 spp.); Amazon, Atlantic, Caatinga and BNMF (7 spp.); Caatinga and BNMF (4 spp.,
107
Classified - Internal use
for example Cordia oncocalyx Allemão); Cerrado and BNMF (3 spp., for example
Senegalia monacantha (Willd.) Seigler & Ebinger); and Amazon, Cerrado, Caatinga and
BNMF (only Commiphora leptophloeos (Mart.) J.B.Gillett). Whereas, no species are
shared only by Amazon and Caatinga nor by Amazon, BNMF and Caatinga (Fig. 2,
Appendix 1).
Figure 2. Venn Diagram with species distribution in five domains. AM = Amazon, AT =
Atlantic, BNMF = Brazilian Northeastern Mountain Forests, CA = Caatinga, CE =
Cerrado.
When we analysed the singular contribution of each domain on BNMF flora, we
found that Atlantic and Amazon were the domains that shared more species (61 and 34
spp., respectively), followed by Caatinga (4 spp.) and Cerrado (3 spp.). A different view,
considering the relationships among dry and wet domains separately, showed that the
savannas shared a lower number of species with BNMF when compared to the forests
(Figure 3). Although Cerrado and Caatinga shared together 255 species with BNMF, the
contribution of each one was very discrepant, with Cerrado being the main contributor
(Figure 3a). The same was found for wet domains, with Atlantic being the best contributor
(Figure 3b).
108
Classified - Internal use
Comparing the Venn diagrams, we found that when we considered all domains
sharing species, the number of species shared by a single domain was decreased. For
instance, Cerrado, Caatinga and Atlantic had a greater number of shared species when
were evaluated into dry and wet domains (50, 10 and 2 times more). However, Amazon
and BNMF shared 34 species in both analyses, which showed that those species were
shared exclusively by the two domains.
Figure 3. Venn Diagram with species distributions in a) dry domains and b) wet domains.
AM = Amazon, AT = Atlantic, BNMF = Brazilian Northeastern Mountain Forests, CA =
Caatinga, CE = Cerrado.
The cluster dendrogram grouped BNMF, Atlantic, Cerrado and Caatinga together,
but Amazon alone (Fig. 4). In a bigger resolution, BNMF formed a single group (33% of
information remaining), same as Caatinga (73%). Atlantic and Cerrado are very similar
in flora. The correlations between the cophenetic similarities and the Jaccard original
similarities were strong (0.91), showing the reliability of the WPGMA analyses.
109
Classified - Internal use
Figure 4. Weighted pair group method with arithmetic mean with Jaccard coefficient as
distance measure, illustrating the similarity in floristic patterns at the species level among
BNMF, Amazon, Atlantic, Cerrado and Caatinga. Cophenetic coefficient = 0.91.
The MRPP method proved that the five domains presented a heterogeneity of
species, with a significant variation in floristic composition in the different sites (A =
0.53, p <0.001). This result gives support to employ comparisons and ordinations with
the domains, since the species composition differs among the groups as expected
stochastically.
The NMDS ordination provided a two-dimensional solution, with both axes being
significant (P = 0.01), and explained almost 50% of the correlations between the
distances in the original space and the ordination distances. After 15 iterations, the final
stress value was 14.49, which is satisfactory according to McCune and Grace (2002). The
NMDS diagram showed a clear separation of all domains with some connections among
Atlantic, Cerrado and Caatinga (Figure 5). The NMDS analysis ordered the same groups
from the WPGMA, revealing a great correspondence between them. It is worth noting
that BNMF was found to be isolated in both analyses.
110
Classified - Internal use
Figure 5. Non-metric multidimensional scaling yielded by the species data showing the
floristic connections among the five domains. R² values: Axis 1 = 25.34%, and Axis
2 = 23.27%. P value (proportion of randomized runs with stress ≤ observed stress) <
0.0001 in both axes. Final stress = 14.49.
We identified 26 bioregions of woody plants in South America region as
illustrated in Figure 6. Most of the species belong to the larger bioregions, however we
verified smaller bioregions, such as in the western Amazon where species turnover is high
and there are many species in just a few cells. We identified clearly the eastern Amazon,
a mix between Cerrado and Caatinga, two bioregions (northern and southern portion) of
Atlantic forest and the BNMF as an unique bioregion (Fig. 6). Besides that, we also
identified Pantanal, Chaco, Pampas and other South American bioregions, although they
are not the focus of this study. The most common and exclusive species in BNMF were
111
Classified - Internal use
Solanum rhytidoandrum, Manilkara rufula, Wedelia villosa and Guettarda angelica. The
most indicative species were Guettarda angelica and Manilkara rufula (Score = 16.0).
Figure 6. Bioregion map of the woody plants occurring in South America generated with
Infomap Bioregions, using the GBIF species range distribution. Different coloured areas
represent a bioregion. Amazon in different colors, mainly blue, Cerrado and Caatinga in
dark gray, Atlantic in brown and pink, Pantanal in dark purple and BNMF in light purple
are the bioregions identified in Brazil.
Discussion
Our findings provided clues about the species dynamics of Neotropical rainforests
through the LGM and LGI, corroborating past links and indicating spatial displacements
that may have affected species distribution around BNMF areas. Furthermore, we found
evidences of great connections among different domains that give ideas about
biogeographical islands, isolated fragments and bioregions.
Paleoclimate models for those domains suggest at least three pathways for the
connections between Amazon and Atlantic forests: the central Brazil bridge, based on
species interchange between eastern Amazon and south-eastern Atlantic Forest.
112
Classified - Internal use
(Oliveira-Filho & Ratter 1995); the Northeast one that would have crossed the Caatinga
at certain periods since the late Tertiary (Andrade-Lima 1964); and the major southern
route through the Paraná River basin (Por 1992). The high number of species shared by
Cerrado with Amazon and Atlantic forests support the hypothesis of the first pathway
with interchange during the Quaternary, and those species are currently restricted to
multiple interglacial refugia (Collevatti et al. 2009, Barbosa et al. 2012, Bonatelli et al.
2014). Méio et al. (2003) also found a great number of plant species (58.9%) that occur
in Cerrado are also occurring in Atlantic and Amazon or both. For instance, the tree
Plathymenia reticulata shows a pattern of recent expansion from the central Cerrado to
northeastern Brazil via eastern (Atlantic coast) and western (inland) colonization routes
(Novaes et al., 2010). In addition, the species shared by Caatinga and Cerrado are part of
the coalescence of dry forests in peripheral depressions during the dry periods of
Quaternary (Ab'Saber 1983, Pennington et al. 2000).
The second pathway is related to the interchange along Caatinga domain.
Although the Caatinga domain is one of the less studied and vegetation shifts during the
Pleistocene are unclear, the evidence suggests that Atlantic forest have been shaping the
Caatinga diversity (Thomé et al., 2016). For instance, the extensive haplotype of the
orchid Epidendrum cinnabarinum indicates connections between populations of the
Atlantic Forest and Caatinga inselbergs (Pinheiro et al., 2014). This pattern is the
opposite found for Amazon and Caatinga, they shared no species, and there is no
connections addressed on literature (Leal et al. 2016). On the other hand, the mountain
forests (BNFM) that presently exists with the Caatinga domain are actual relicts of an
ancient and wider rainforest.
Studies on endemism and inventories of the BNMF are still scarce, however, we
found a great number of exclusive trees and shrubs in these areas (almost the same as
113
Classified - Internal use
Cerrado, see appendix 1). They have been functioned as ecological refuges and stepping
stones providing natural shelter of several endangered species and new species not yet
described (Andrade-Lima 1982, Carnaval & Bates 2007, Sobral-Souza et al. 2015) and,
consequently, often present endemic taxas (Loebmann & Haddad, 2010; Camardelli &
Napoli, 2012; Guedes et al., 2014). Furthermore, we accept our first hypothesis, the
lineages diversity of Atlantic and Amazon had been shaped along recurrent connections
between them during LGM, what contributed to the main BNMF diversity. Although
Amazon has less contribution when compared to the Atlantic forest, it has the same
amount of exclusive species when compared among all domais and only among humid
domains, which indicates that connections between Amazon and BNMF are better
explained by vicariance events. Oppositely, the long-distance dispersal events are more
common between Atlantic forest and BNMF because of the geographic proximity.
We reject our second hypothesis, the geographic proximity does not influence on
the interchange between Caatinga and BNMF, where we accounted only four species
shared. The same occurs between Cerrado and BNMF, with three species shared. The
reason for such few number is probably the difference in environmental conditions, as
soil properties, temperature and humidity, that is related to the species niche. We can see
this pattern, when we compare analyses along all domain and only between dry and wet
domains, both Cerrado and Caatinga have more species shared when analyzed separately
because they are sharing species with other domains rather than BNFM (Figs. 2 and 3).
In terms of the affinities, we can see a slight trend towards a closer association
between Cerrado, Atlantic forest and Caatinga as revealed by the analyses of WPGMA
and NMDS. The same pattern is described by Costa (2003) for small mammals and by
Leal et al. (2016) in a review for phylogeographic studies on plants. The analyses also
indicate a clear separation of Amazon and BNMF from the other domains, mainly because
114
Classified - Internal use
of their exclusive species. This separation is also confirmed by the bioregion analysis,
where we accept our third hypothesis: the BNMF constitute a new biogeographical
region.
The portioning of Amazon into several highly divergent phylogeographical areas
was already found by Cracraft & Prum, (1988) and Patton et al. (2000). The northern and
southern portion of Atlantic forest was described by Carnaval & Moritz (2008) and
confirmed by phylogeographical studies (Sobral-Souza et al. 2015, Peres et al. 2017). The
Caatinga and Cerrado are together part of the dry South American diagonal (Werneck et
al. 2011). The bioregions identified are consistent with those found by Vilhena and
Antonelli (2015) for amphibian species, except for some smaller bioregions. Considering
the Neotropics, our results are in accordance to the regionalization proposed by Morrone
(2006; 2014), for instance the Amazonian, the Paraná, the Colombian and the Chaco
provinces. We also compared our data with the World Wildlife Fund (WWF) ecoregions
(Olson et al. 2001), which is very similar. Furthermore, our results announces the
Brazilian Northeast Mountain Forests as a new bioregion never described before with its
exclusive species and characteristics.
Nevertheless, bioregionalization based on species distribution needs to deal with
the quantity and quality of occurrence data because they clearly contain spatial,
taxonomic and temporal biases. For instance, Amazon has the biggest remaining area and
consequently the major exclusive species proportion when compared to the other
domains. On the contrary, despite of the smaller remaining area, Atlantic has more
exclusive species than Cerrado. Additionally, the NeoTropTree has surveys from
different times (decades) and concentrated in some areas (around the state capital, for
example). So, the detection of bioregions is impacted by how we treat and revise the data
available.
115
Classified - Internal use
Conclusion
To understand better the emergence of those new bioregions or even accept their
existence, it is necessary to develop phylogeographical studies focuse on species or
groups of species sampled in different domains. Our results indicates that past
connections between the Neotropical rainforests resulted from specific climatic
conditions in this area and support that subregions of Amazon and Atlantic Forest, such
as Brazilian Northeastern Mountain Forest, should be considered as distinct and unique
biogeographical regions as they were shaped by different climate conditions.
Appendix 1 – List of species of all domains studied. Available online:
https://docs.google.com/spreadsheets/d/1g2lfhKGII_p-SMG_aPM0XKC_j-ZsrDKw-
az4DtMofRg/edit?ts=59847e39#gid=0
116
Classified - Internal use
CONSIDERAÇÕES FINAIS
Apresentamos uma síntese de estudos para as Florestas Serranas do Nordeste
Brasileiro, explorando características físicas, climáticas e bióticas, padrões gerais de
similaridade florística, relações com os diferentes domínios brasileiros, distribuição de
espécies e composição palinológica. É patente que a história das florestas serranas
nordestinas vem sendo modificada, embora ainda haja grande lacunas, os resultados aqui
encontrados facilitam a compreensão do funcionamento ecossistêmico em gradientes
altitudinais de florestas serranas do Nordeste brasileiro.
Chamamos atenção para cinco destaques da tese. O primeiro destaque é o fato da
vertente à sotavento da Serra de Maranguape apresentar maior diversidade e riqueza de
espécies e flora bem diferente das áreas de Caatinga. O segundo é que as áreas de topo
das florestas serranas cearenses ostentam características de florestas nebulares. O terceiro
é a reestruturação constante das florestas serranas, uma vez que há espécies da chuva
polínica não encontradas na vegetação e vice-versa. O quarto é a incidência de zoocoria
como processo chave na distribuição de espécies nas florestas serranas nordestinas (escala
regional). O quinto e mais importante é a classificação das florestas serranas como uma
bioregião única e distinta do Brasil.
Desse modo, econtramos resultados novos e significativos para a biodiversidade,
biogeografia e distribuição de espécies nas florestas serranas do Nordeste brasileiro.
Como essas florestas são restritas e constantemente ameaçadas, os resultados serão
bastante úteis para conservação e restauração ambiental. Esse estudo foi o primeiro a
utilizar a ferramenta de seleção de modelos e análises de bioregião em florestas do
semiárido, fornecendo dados e ideias para pesquisas vindouras.
117
Classified - Internal use
Além disso, nosso estudo traz dados que colaboram para a conservação e
preservação das florestas serranas nordestina, fornecendo listas de espécies, espécies
indicativas e ameaçadas de extinção, critérios para elaborar planos de manejo e novas
Unidades de Conservação. Os próximos passos e pesquisas futuras irão focar nas origens
e centros de dispersão de espécies-chaves destas florestas (filogeografia) e respostas da
distribuição destas espécies frente às mudanças climáticas (modelagem ecológica). Esses
dados, quando levados em conta por novos estudos, potencialmente maximizarão
resultados, metodologias e conhecimento para o semiárido brasileiro, na medida em que
permitem direcionar novos esforços para onde há mais carência de dados.
118
Classified - Internal use
REFERÊNCIAS
Ab’saber A.N. 1977. Espaços Ocupados pela Expansão dos Climas secos na América do
Sul por ocasião dos períodos Glaciais Quaternários. Paleoclimas, 3: 1-9.
Allen, C. D. & Breshears, D. D. 1998. Drought-induced shift of a forest–woodland
ecotone: Rapid landscape response to climate variation. Proceedings of the
National Academy of Sciences, 95: 14839 – 14842.
Andrade-lima, D. 1966. Contribuição ao paralelismo da flora amazônico-nordestina. Pp.
1-30. In: Boletim Técnico IPA (Boutton, T.W., eds). Recife.
Andrade-lima, D. 1982. The Caatinga Dominium. Revista Brasileira de Botânica, 4: 149-
153.
Araujo, F.S., Gomes, V.S., Silveira, A.P., Figueiredo, M.A., Oliveira, R.S., Bruno,
M.M.A., Lima-verde, L.W., Silva, E.F., Otutumi, A.T. & Ribeiro, K.A. 2007.
Efeito da variação topoclimática e estrutura da vegetação da serra de Baturité,
Ceará. Pp. 75-162 In: Diversidade e conservação da Biota da serra de Baturité,
Ceará. (Oliveira, T.S. & Araújo, F.S., eds.) Edições UFC/COELCE: Fortaleza.
Arz, H.W.; Pätzold, J.; Wefer, G. 1999. Climatic changes during the last deglaciation
recorded in sediment cores from the northeastern Brazilian Continental Margin.
Geo-Marine Letters, 19: 209-218.
Auler, A.S.; Wang, X.; Edwards, R.L.; Cheng, H.; Cristalli, P.S.; Smart, P.L.; Richards,
D.A. 2004. Quaternary ecological and geomorphic changes associated with
rainfall events in presently semi-arid northeastern Brazil. Journal of Quaternary
Science, 19(7): 693-701.
119
Classified - Internal use
Behling H. & Lichte M. 1997. Evidence of dry and cold climatic conditions at glacial
times in tropical southeastern Brazil. Quaternary Research 48: 348-358.
Behling, H., Arz, W.H., Pätzold, J., Wefer, G., 2000. Late quaternary vegetational and
climate dynamics in northeastern Brazil, inferences from marine core GeoB3104-
1. Quaternary Science Reviews, 19: 981- 994.
Behling, H.; Costa M.L. 2001. Holocene vegetational and coastal environmental changes
from the Lago Crispim record in northeastern Pará State, eastern Amazonia.
Review of Palaeobotany and Palynology, 114: 145-155.
Behling, H. 2001. Late Quaternary environmental changes in the Lagoa da Curuça region
(eastern Amazonia, Brazil) and evidence of Podocarpus in the Amazon lowland.
Vegetation History and Archeobotany, 10: 175-183.
Behling H. & Negrelle R.R.B. 2001. Tropical Rain Forest and Climate Dynamics of the
Atlantic Lowland, Southern Brazil, during the Late Quaternary. Quaternary
Research 56: 383-389.
Behling, H.; Keim, G.; Irion, G.; Junk, W.; Mello, J.N. 2001. Holocene environmental
changes in the Central Amazon Basin inferred from Lago Calado (Brazil).
Palaeogeography, Palaeoclimatology, Palaeoecology, 173: 87-101.
Betancourt, J. L. & Van Devender, T. R., eds. (1990) Packrat Middens: The Last 40,000
Years of Biotic Change (Univ. of Arizona Press, Tucson, AZ).
Bigarella, J.J. & Andrade-Lima, D. (1982) Paleoenvironmental changes in Brazil.
Biological diversification in the tropics (ed. por G.T. Prance), pp. 27-40.
Columbia University Press, New York.
120
Classified - Internal use
Bigarella, J.J.; Andrade-Lima, D. & Riehs, P.J. (1975) Considerações a respeito das
mudanças paleoambientais na distribuição de algumas espécies vegetais e animais
no Brasil. Anais da Academia Brasileira de Ciência, 47: 411-464.
Cavalcanti, D. & Tabarelli, M. 2004. Distribuição das plantas Amazônico-Nordestinas no
Centro de Endemismo Pernambuco: Brejos de Altitude vs. Floresta de Terras
Baixas. Pp. 285-296. In: Brejos de Altitude em Pernambuco e Paraíba. (Pôrto,
K.C.; Cabral, J.J.P. & Tabarelli, M., eds.). Ministério do Meio Ambiente, Brasília.
Claudino-Sales, V.; Peulvast, J.P. 2007. Evolução morfoestrutural do relevo da margem
continental do estado do Ceará, Nordeste do Brasil. Caminhos de Geografia (on-
line), 8(20): 1-21.
Cole, M.M. 1960. Cerrado, caatinga and pantanal: The distribution and origin of the
savana vegetation of Brazil. The Geographical Journal, 126(2): 168-179.
DeOliveira, P.E.; Barreto, A.M.F.; Suguio, K. 1999. Late Pleistocene/Holocene climatic
and vegetational history of the Brazilian caatinga: the fossil dunes of the middle
São Francisco River. Palaeogeography, Palaeoclimatology, Palaeoecology, 152:
319-337.
Diogo, I. J. S. 2013. Aspectos biogeográficos e autoecológicos de encraves florestais
úmidos no semiárido. Dissertação de mestrado, Universidade Federal do Ceará.
98p.
Ferreira, N. J.; Laçava, C. I. V.; Sobral, Z. R. 2001 A climatological study of convective
cloudbands in northeastern Brazil Part I: Preliminary analysis. Australian
Meteorological Magazine, 50: 105-113.
121
Classified - Internal use
Figueiredo, M. A.; Barbosa, M. A. 1990. A vegetação e flora na serra de Baturité. Coleção
Mossoroense. Série B, n.747.
Gosz, J. R. 1992. Gradient analysis of ecological change in time and space: implications
for forest management. Ecological Applications, 2: 248-261.
Haffer, J. 1969. Speciation in Amazonian Forests Birds. Science, 165: 131-137.
Joly, C. A.; Leitão-Filho, H. F.; Silva, S. M. 1991. O patrimônio florístico da Mata
Atlântica. Pp. 96-128. In: (Câmara, I. G., Coord.). Editora Index e Fund. SOS
Mata Atlântica: São Paulo.
Juaréz, R. I. N.; Liu, W. 2001. FFT analysis on NDVI annual cicle and climatic
regionality in Northeast Brazil. International Journal of Climatology, 21:1803-
1820.
Köplin, N.; Schädler, B.; Viviroli, D.; Weingartner, R. 2013. The importance of glacier
and forest change in hydrological climate-impact studies. Hydrology and Earth
System Sciences, 17: 619-635.
Latrubesse, E. M. & Franzinelli, E. 2002. The holocene alluvial plain of the middle
amazon river, Brazil. Geomorphology, 44: 241–257.
Latrubesse, E. M.; Stevaux, J. C.; Cremon, E. H.; May, J-H; Tatumi, S. H.; Hurtado, M.
A.; Bezada, M.; Argollo, J. B. 2013. Late Quaternary megafans, fans and fluvio-
aeolian interactions in the Bolivian Chaco, Tropical South America.
Palaeogeography, Palaeoclimatology, Palaeoecology, 356: 75-88.
Ledru M.P.; Salatino, M.L.F.; Ceccantini, G.; Salatino, A.; Pinheiro, F.; Pintaud, J.C.
2007. Regional assessment of the impact of climatic change on the distribution of
122
Classified - Internal use
a tropical conifer in the lowlands of South America. Diversity and Distributions,
13: 761- 771.
Lins, R.C. 1989. As áreas de exceção do agreste de Pernambuco. Sudene, Recife.
Lopes, C.G.M., Ferraz, E.M.N., Araújo, E.L. (2008) Physiognomic-structural
characterization of dry- and humid-forest fragments (Atlantic Coastal Forest) in
Pernambuco State, NE Brazil. Plant Ecology, 198, 1–18.
Macphail, M. K. 1979. Vegetation and climates in southern Tasmania since the last
glaciation. Quaternary Research, 11:306-341.
Santos, A. M., Cavalcanti, D. R., Silva, J. M. C. D., Tabarelli, M. 2007. Biogeographical
relationships among tropical forests in north-eastern Brazil. Journal of
Biogeography, 34: 437–446.
Nimer, E. 1972. Climatologia da Região Nordeste do Brasil: subsídios à geografia
regional do Brasil, Revista Brasileira de Geografia, 34: 5-51,
Oliveira-Filho, A.T. & Ratter, J.A. 1995. A study of the origin of Central Brazilian forests
by the analysis of plant species distribution patterns. Edinburgh Journal of Botany,
52: 141-194.
Pires, J.M.; Dobzhansky, T. & Black, G.A. 1953. An estimate of the number of species
of trees in an Amazonian forest community. Botanical Gazette, 114: 467-77.
Prado, D.E. & Gibbs, P.E. 1993. Patterns of species distributions in the dry seasonal
forests of South America. Annals of Missouri Botanic Gardens, 80: 902-927.
123
Classified - Internal use
Prance, G.T. 1982. Forest refuges: evidences from woody angiosperms. Pp. 137-158. In:
Biological diversification in the tropics. (Prance, G.T., Ed.) Columbia University
Press: New York.
Risser, P. G. 1995. The status of the science examining ecotones. BioScience, 45:318-
325.
Rizzini, C.T. 1997. Tratado de fitogeografia do Brasil: aspectos ecológicos, sociológicos
e florísticos. 2 ed. Âmbito Cultural Edições, Rio de Janeiro. 427p.
Santos, A.M.M. 2002. Distribuição de plantas lenhosas e relações históricas entre a
floresta Amazônica, a floresta Atlântica costeira e os brejos de altitude do nordeste
brasileiro. Dissertação de Mestrado, Universidade Federal de Pernambuco,
Recife.
Schäffer, W. B. & Prochnow, M. 2002. A Mata Atlântica e você: como preservar,
recuperar e se beneficiar da mais ameaçada floresta brasileira. Apremavi, Brasília,
Brasil, 156 p.
Scudeller, V. V.; Martins, F.R. & Shepherd, G.J. 2001. Distribution and abundance of
arboreal species in the atlantic ombrophilous dense forest in Southeastern Brazil.
Plant Ecology, 152:185-199.
Stute, M.; Forster, M.; Frischkorn, H.; Serejo, A.; Clark, J.F.; Schlosser, P.; Broecker,
W.S.; Bonani, G. 1985. Cooling of tropical Brazil (5 ºC) during the last glacial
maximum. Science, 269: 379–383.
Tabarelli, M. 2001. Integridade e ameaças aos brejos da Paraíba e Pernambuco. Pp. 82-
91.In: Plano de Conservação dos Brejos de Paraíba e Pernambuco. Relatório
Técnico do Subprojeto Recuperação e Manejo dos Ecossistemas Naturais de
124
Classified - Internal use
Brejos de Altitude de Pernambuco e Paraíba. (Tabarelli, M., Ed.) Projeto
PROBIO, Ministério do Meio Ambiente: Recife.
Pôrto, K. C., Cabral, J. J. P., Tabarelli, M. 2004. Brejos de altitude em Pernambuco e
Paraíba: história natural, ecologia e conservação. Ministério do Meio Ambiente:
Recife.
Silva, J.M.C. & Tabarelli M. 2000. Tree species impoverishment and the future flora of
the Atlantic forest of northeast Brazil. Nature 404:72-74.
Terborgh, J. & Andresen, E. 1998. The composition of Amazonian forests: patterns at
local and regional scales. Journal of Tropical Ecology, 14: 645-664.
Tonkov, S.; Bozilova, E.; Possnert, G. 2012. Postglacial vegetation history as recorded
from the subalpine Lake Ribno (NW Rila Mts), Bulgaria. Central European
Journal of Biology, 8:64-77.
Veloso, H.P., Rangel-Filho, A.L.R. & Lima, J.C.A. 1991. Classificação da vegetação
brasileira adaptada a um sistema universal. IBGE, Rio de Janeiro.
Vidotto, E.; Pessenda, R. C. L.; Ribeiro, A. S.; Freitas, H. A.; Bendassolli, J. A. 2007.
Dinâmica do ecótono floresta-campo no sul do estado do Amazonas no Holoceno,
através de estudos isotópicos e fitossociológicos. Acta amazonica, 37: 385-400.
Whitmore, T.C. & Prance, G.T. 1987. Biogeography and quaternary History in Tropical
America. Clarendon Press, Oxford.
Xavier, F.A.S.; Oliveira, T. S.; Araújo, F. S.; Gomes, V. S. 2007. Manejo da vegetação
sob linhas de transmissão de energia elétrica na Serra de Baturité. Ciência
Florestal, 17: 351-364.
125
Classified - Internal use
Ximenes, C.L. Tanques Fossilíferos de Itapipoca, CE - Bebedouros e cemitérios de
megafauna pré-histórica. 2008. In: Sítios Geológicos e Paleontológicos do Brasil.
(Winge, M.; Schobbenhaus, C.; Souza, C.R.G.; Fernandes, A.C.S.; Berbert-Born,
M.; Queiroz, E.T.; Eds.). Disponível em
<http://www.sigep.cprm.gov.br/sitio014/sitio014.pdf>.
Walsh, S. J.; Butler, D. R.; Allen, R. T.; Malanson, G. P. 1994. Influence of snow patterns
and snow avalanches on the alpine treeline ecotone. Journal of Vegetation
Science, 5: 657-672.
Wang, X.; Auler, A.S.; Edwards, R.L.; Cheng, H.; Cristalli, P.S.; Smart, P.L.; Richards,
D.A.; Shen, C.C. 2004. Wet periods in northeastern Brazil over the past 210 kr
linked to distant climate anomalies. Nature, 432: 740-743.
Zemlak, T.S.; Habit, E.M.; Walde, S.J.; Battini, M.A.; Adams, E.D.M.; Ruzzante, D.E.
2008. Across the southern Andes on fin: glacial refugia, drainage reversals and a
secondary contact zone revealed by the phylogeographical signal of Galaxias
platei in Patagonia. Molecular Ecology, 17: 5049–5061.
References – Chapter I
Anderson AB, Jenkins CN. 2006. Applying nature’s design: corridors as a strategy for
biodiversity conservation. New York: Columbia University Press.
Andrade-lima D. 1982. The Caatinga Dominium. Revista Brasileira de Botânica 4: 149-
153.
APG IV. 2016. An update of the Angiosperm Phylogeny Group classification for the
orders and families of flowering plants: APG IV. Botanical Journal of the Linnean
Society 181: 1–20.
126
Classified - Internal use
Araujo FS, Gomes VS, Silveira AP, Figueiredo MA, Oliveira RS, Bruno MMA, Lima-
verde LW, Silva EF, Otutumi AT, Ribeiro KA. 2007. Efeito da variação
topoclimática e estrutura da vegetação da serra de Baturité, Ceará. In: Oliveira TS,
Araújo FS eds. Diversidade e conservação da Biota da serra de Baturité, Ceará.
Fortaleza: Edições UFC/COELCE. 137-162.
Arruda LV. 2001. Serra de Maranguape-CE: ecodinâmica da paisagem e implicações
socioambientais. Dissertação. PRODEMA: Universidade Federal do Ceará.
Auler AS, Wang X, Edwards RL, Cheng H, Cristalli PS, Smart PL, Richards DA. 2004.
Quaternary ecological and geomorphic changes associated with rainfall events in
presently semiarid northeastern Brazil. Journal of Quaternary Science 19: 693–
701.
Behling H, Arz HW, Pätzold J, Wefer G. 2000. Late Quaternary vegetational and climate
dynamics in northeastern Brazil, inferences from marine core GeoB 3104–1.
Quaternary Science Reviews 19: 981–994.
Costa RC, Araújo FS 2012. Physiognomy and structure of a caatinga with Cordia
oncocalyx (Boraginaceae), a new type of community in Andrade-Lima’s
classification of caatingas. Rodriguésia 63: 269-276.
Costa RC, Araújo FS, Lima-Verde LW. 2007. Flora and life-form spectrum in an area of
deciduous thorn woodland (caatinga) in northeastern, Brazil. Journal of Arid
Environments 68: 237-247.
Cottam G, Curtis JT. 1956. The use of distance measures in phytossociological sampling.
Ecology 37: 451-460.
127
Classified - Internal use
Diogo IJS, Martins FR, Verola CF, Costa, IR 2016. Variation in plant-animal interactions
along an elevational gradient of moist forest in a semiarid area of Brazil. Acta
Botanica Brasilica 30: 27-34.
Diogo IJS, Martins FR, Costa IR, Santos FAM 2017. Dispersal and edaphic factors
driving plant species composition in mountain forests in a semiarid region
of Brazil. Plant Ecology and Diversity In Press.
Duarte MC, Rego F, Moreira I. 2005. Distribution patterns of plant communities on
Santiago Island, Cape Verde. Journal of Vegetation Science 16: 283–292.
Falkenberg DB, Voltolini JC. 1995. The montane cloud forest in Southern Brazil. In:
Hamilton LS, Juvik JO, Scatena FN. eds. Tropical montane cloud forests. New
York: Springer-Verlag. 138-149.
Fernández-Palacios JM, de Nicolás JP. 1995. Altitudinal pattern of vegetation variation
on Tenerife. Journal of Vegetation Science 6, 183–190.
Ferraz EMN, Araújo EL, Silva SI. 2004. Floristic similarities between lowland and
montane areas of Atlantic Coastal Forest in Northeastern Brazil. Plant Ecology
174: 59–70.
Fick SE, Hijmans RJ. 2017. Worldclim 2: New 1-km spatial resolution climate surfaces
for global land areas. Internatioal Journal of Climatology.
Forzza RC, Baumgratz JFA, Bicudo CEM, Canhos DAL, Carvalho Jr AA, Coelho MAN,
Costa AF, Costa DP, Hopkins MG, Leitman PM, Lohmann LG, Lughadha EN,
Maia LC, Martinelli G, Menezes M, Morim MP, Peixoto AL, Pirani JR, Prado J,
Queiroz LP, Souza S, Souza VC, Stehmann JR, Sylvestre LS, Walter BMT, Zappi
128
Classified - Internal use
DC. 2012. New Brazilian floristic list highlights conservation challenges.
Bioscience 62: 39–45.
Funceme. Fundacão Cearense de Meteorologia e Recursos Hidrícos. 2005. Relatório de
pluviometria por faixa de anos – estado do Ceará [online]. Available from
www.funceme.br/DEPAM/index.htm [Accessed 22 January 2015].
Gentry A.H. 1982. Neotropical floristic diversity: phytogeographical connections
between Central and South America, pleistocene climatic fluctuations, or an
accident of the andean orogeny? Annals of Missouri Botanical Garden 69:557-
593.
Gentry, A.H. 1988. Changes in plant community diversity and floristic composition on
environmental and geographical gradients. Annals of Missouri Botanical Garden
75:1-34.
Giambelluca TW, Chen Q, Frazier AG, Price JP, Chen Y-L, Chu P-S, Eischeid JK,
Delparte DM. 2013. Online rainfall atlas of Hawai’i. Bulletin of the American
Meteorological Society 94, 313–316.
Gibbs HK, Ruesch AS, Achard F, Clayton MK, Holmgren P, Ramankutty N, Foley JA.
2010. Tropical forests were the primary sources of new agricultural land in the
1980s and 1990s. Proceedings of the National Academy of Sciences of the United
States of America 107: 16732–16737.
Gonzalez-Caro S, Natalia UM, Alvarez E, Stevenson PR, Swenson NG. 2014.
Phylogenetic alpha and beta diversity in tropical tree assemblages along regional-
scale environmental gradients in northwest South America. Journal of Plant
Ecology 7: 145-153.
129
Classified - Internal use
Hamilton LS, Juvik JO, Scatena FN. 1995. The Puerto Rico tropical cloud forest
symposium: introduction and workshop synthesis. In: Hamilton LS, Juvik JO,
Scatena FN. eds. Tropical montane cloud forests. New York, Springer-Verlag. 1-
23.
Hugget RJ. 1995. Geoecology, an evolutionary approach. London: Routledge.
John R, Dalling JW, Harms KE, Yavitt JB, Stallard RF, Mirabello M, Hubbell SP,
Valencia R, Navarrete H, Vallejo M, Foster RB. 2007. Soil nutrients influence
spatial distributions of tropical tree species. Proceedings of the National Academy
of Sciences USA 104: 864-869.
Kitayama K. 1992. An altitudinal transect study of the vegetation on Mont Kinabalu,
Borneo. Vegetatio 102: 149-171.
Ledru MP 1993. Late Quaternary environmental and climatic changes in central
Brazil. Quaternary Research 39:90–98.
Ledru MP, Salgado-Labouriau ML, Lorscheiter ML. 1998. Vegetation dynamics in
southern and central Brazil during the last 10,000 yr. Review of Paleobotany and
Palynology 99:131–142.
Ledru M-P, Salatino MLF, Ceccantini GT, Salatino A, Pinheiro F, Pintaud J-C. 2007.
Regional assessment of the impact of climatic change on the distribution of a
tropical conifer in the lowlands South America. Diversity and Distributions 13:
761–771.
Lima BG, Coelho MFB, Oliveira OF. 2012. Caracterização florística de duas áreas de
caatinga na Região Centro-Sul do Ceará, Brasil. Bioscience Journal 28: 277-296.
130
Classified - Internal use
Lima JR, Mansano VF, Araújo FS. 2012a. Coexistence and geographical distribution of
Leguminosae in an area of Atlantic forest in the semi-arid region of Brazil.
Journal of Systematics and Evolution 50: 25–35.
Lima JR, Mansano VF, Araújo FS. 2012b. Richness and diversity of Leguminosae in an
altitudinal gradient in thetropical semi-arid zone of Brazil. Journal of Systematics
and Evolution 50: 433–442.
Liu SL, Ma KM, Fu BJ, Kang YX, Zhang JY, Zhang YX. 2003. The relationship between
landform,soil characteristics and plant community structure in the donglingshan
mountain region, Beijing. Chinese Journal of Plant Ecology 27, 496–502.
Mantovanni, W. 2006. Conservação de biodiversidade: importância das serras úmidas no
nordeste semi-árido brasileiro. In: Oliveira TS, Araújo FS. eds. Diversidade e
Conservação da Biota na Serra de Baturité, Ceará. Fortaleza: Edições UFC,
COELCE. 3-15.
Martins FR, Santos FAM. 1999. Técnicas usuais de estimativa da biodiversidade. Revista
Holos ed. especial: 236-267.
McCune B, Mefford MJ. 2011. PC-ORD 6.0 for windows: multivariate analysis of
ecological data. Gleneden Beach: MjM Software.
Meireles LD, Shepherd GJ, Kinoshita LS. 2008. Variações na composição florística e na
estrutura fitossociológica de uma floresta ombrófila densa alto-montana na Serra
da Mantiqueira, Monte Verde, MG. Revista Brasileira de Botânica 31: 559-574.
Mittermeier R, Gil PR, Hoffmann M, Pilgrim J, Brooks T, Mittermeier CG, Lamoreux J,
Fonseca GAB, Seligmann PA. 2004. Hotspots Revisited – Earth’s biologically
131
Classified - Internal use
richest and most endangered terrestrial ecoregions. Washington: Conservation
International.
Montade V, Ledru M-P, Burte J, Martins E, Verola CF, da Costa IR, Magalhaes H. 2014.
Stability of a Neotropical microrefugium during climatic instability. Journal of
Biogeography 41: 1215–1226.
Montade V, Diogo IJS, Bremond L, Favier C, Costa IR, Ledru M-P, Paradis L, Martins
ESPR, Burte J, Magalhães e Silva FH, Verola CF 2016. Pollen-based
characterization of montane forest types in north-eastern Brazil. Review of
Palaeobotany and Palynology 234: 147-158.
Moro MF, Nic Lughadha E, Filer DL, Araújo FS, Martins FR. 2014. A catalogue of the
vascular plants of the Caatinga Phytogeographical Domain: a synthesis of floristic
and phytosociological surveys. Phytotaxa 160: 1–118.
Moro MF, Macedo MB, Moura-Fé MM, Castro ASF, Costa RC. 2015. Vegetação,
unidades fitoecológicas e diversidade paisagística do estado do Ceará.
Rodriguésia 66: 717–743.
Mueller-Dombois D, Ellenberg H. 1974. Aims and methods of vegetation ecology. New
York: John Wiley & Sons.
Murray-Smith C, Brummitt NA, OliveirA-Filho AT, Bachman S, Moat J, Lughadha
EMN, Lucas EJ. 2009. Plant Diversity Hotspots in the Atlantic Coastal Forests of
Brazil. Conservation Biology 23: 151–163.
Myers N, Mittermeier RA, Mittermeier CG, Da Fonseca GA, Kent J. 2000. Biodiversity
hotspots for conservation priorities. Nature 403: 853-858.
132
Classified - Internal use
Oliveira-Filho AT. 2006. TreeAtlan 1.0. Tree flora of the South American Atlantic Forest:
A database involving geography, diversity and conservation [online]. Available
from http://www.icb.ufmg.br/treeatlan/ [accessed 21 April 2015].
Oliveira-Filho AT, Fontes MA. 2000. Patterns of floristic differentiation among Atlantic
Forests in Southeastern Brazil and the influence of climate. Biotropica 32: 793–
810.
Peel MC, Finlayson BL, McMahon TA. 2007. Updated world map of the Köppen-Geiger
climate classification. Hydrology and Earth System Sciences 11: 1633-1644.
Pessoa MDS, Vleeschouwer KMD, Talora DC, Rocha L, Amorim AMA, 2012.
Reproductive phenology of Miconia mirabilis (Melastomataceae) within three
distinct physiognomies of Atlantic Forest, Bahia, Brazil. Biota Neotropica 12: 49–
56.
Punyasena SW. 2008. Estimating neotropical palaeotemperature and palaeoprecipitation
using plant family climatic optima. Palaeogeology, Palaeoclimatology,
Palaeoecology 265: 226–237.
QGIS DT. 2015. QGIS Geographic Information System Developers Manual. Open
Source Geospatial Foundation Project [online]. Available from
http://qgis.osgeo.org/ [accessed 14 August 2016].
Queiroz LP. 2006. The Brazilian caatinga: Phytogeographical patterns inferred from
distribution data of the Leguminosae. In: Pennington RT, Lewis GP, Ratter
JA eds. Neotropical savannas and dry forests: Plant diversity, biogeography and
conservation. Boca Raton : Taylor & Francis. 121–157.
133
Classified - Internal use
Ribeiro MC, Martensen AC, Metzger JP, Tabarelli M, Scarano FR, Fortin M-J. 2011. The
Brazilian Atlantic forest: A shrinking biodiversity hotspot. In: Zachos FE, Habel
JC eds. Biodiversity hotspots: Distribution and protection of conservation priority
areas. Berlin: Springer-Verlag. 405–434.
Ribeiro MC, Metzger JP, Martensen AC, Ponzoni FJ, Hirota MM. 2009. The Brazilian
Atlantic Forest: How much is left, and how is the remaining forest distributed?
Implications for conservation. Biological Conservation 142: 1141–1153.
Sabatini FM, Burrascano S, Tuomisto H, Blasi C. 2014. Ground layer plant species
turnover and beta diversity in Southern-european old-groeth forests. PLoS ONE
9: e95244.
SantosAMM, Cavalcanti DR, Silva JMC, Tabarelli M. 2007. Biogeographical
relationships among tropical forests in northeastern Brazil. Journal of
Biogeography 34: 437–446.
Schouten MA, Verweij PA, Barendregt A, Kleukers RMJC, Kalkman VJ, Ruiter
PC. 2009. Determinants of species richness patterns in the Netherlands across
multiple taxonomic groups. Biodiversity and Conservation 18: 203–217.
Shepherd GJ 2009. FITOPAC 2.1. Departamento de Biologia Vegetal, Universidade
Estadual de Campinas, Campinas [online] Available from
http://pedroeisenlohr.webnode.com.br/fitopac/ [accessed 3 May 2015].
Silva JMC, Casteleti CHM. 2003. Status of the biodiversity of the Atlantic Forest of
Brazil. In: Galindo-Leal C, Câmara IG. eds. The Atlantic Forest of South America:
Biodiversity status, trends, and outlook. Washington: Center for Applied
Biodiversity Science and Island Press. 43–59.
134
Classified - Internal use
Sousa JT, Araújo PGM, Sousa JR, Silva MAM, Lima AS, Souza MMA. 2007.
Caracterização de uma caatinga arbórea no município de Aiuaba-CE. Cadernos
de Cultura e Ciência (URCA) 2: 1-10.
Souza JP, Araújo GM, Schiavini I, Duarte PC. 2006. Comparison between canopy
treesand arboreal lower strata of urban semideciduous seasonal forest in Araguari
-MG. Brazilian Archives of Biologyand Technology 49: 775–783.
Tabarelli M. 2001. Integridade e ameaças aos brejos da Paraíba e Pernambuco. In:
Tabarelli M. eds. Plano de Conservação dos Brejos de Paraíba e Pernambuco.
Relatório Técnico do Subprojeto Recuperação e Manejo dos Ecossistemas
Naturais de Brejos de Altitude de Pernambuco e Paraíba. Recife: Ministério do
Meio Ambiente. 82-91.
Vasconcelos-Sobrinho, J. 1971. Os brejos de altitude e as matas serranas. In: J.
Vasconcelos-Sobrinho (ed.). As regiões naturais do Nordeste, o meio e a
civilização. Conselho de Desenvolvimento de Pernambuco, Recife, pp. 79-86.
Xavier FAS, Oliveira TS, Araújo FS, Gomes VS. 2007. Manejo da vegetação sob linhas
de transmissão de energia elétrica na Serra de Baturité. Ciência Florestal 17: 351-
364.
Wisz MS, Pottier J, Kissling WD, Pellissier L, Lenoir J, Damgaard CF, Dormann CF,
Forchhammer MC, Grytnes J-A, Guisan A, Heikkinen RK, Høye TT, Kühn I,
Luoto M, Maiorano L, Nilsson M-C, Normand S, Öckinger E, Schmidt NM,
Termansen M, Timmermann A, Wardle DA, Aastrup P, Svenning J-C. 2013. The
role of biotic interactions in shaping distributions and realised assemblages of
species: implications for species distribution modelling. Biological Reviews 88:
15–30.
135
Classified - Internal use
Zachos F, Habel J. 2011. Biodiversity hotspots: distribution and protection of
conservation priority areas. Heidelberg/Dordrecht/ London/New York: Springer.
References Chapter II
Andrade, L.A., Oliveira, F.X., Nascimento, I.S., Fabricante, J.R., Sampaio, E.V.S.B.,
Barbosa, M.R.V., 2006. Análise florística e estrutural de matas ciliares ocorrentes
em brejo de altitude no município de Areia. Paraíba. Rev. Bras. Ciênc. Agrár. 1,
31–40. http://dx.doi.org/10.5039/agrária.v1i1.9.
Andrade-Lima, D., 1982. Present-day forest refuges in northeastern Brazil. In: Prance,
G.T. (Ed.), Biological Diversification in the Tropics. Columbia University Press,
New York, pp. 245–251.
Araújo, F.S., Gomes, V.S., Silveira, A.P., Figueiredo, M.A., Oliveira, R.F., Bruno,
M.M.A., Lima- Verde, L.W., Silva, E.F., Otutumi, A.T., Ribeiro, K.A., 2007.
Efeito da variação topoclimática e estrutura da vegetação da serra de Baturité,
Ceará. In: Oliveira, T.S., Araújo, F.S. (Eds.), Diversidade E Conservação Da Biota
Da Serra de Baturité. Ceará. UFC/COELCE, Fortaleza, pp. 73–136.
Ashton, P.S., 2003. Floristic zonation of tree communities on wet tropical mountains
revisited. Perspect. Plant Ecol. Evol. Syst. 6, 87–104.
http://dx.doi.org/10.1078/1433-8319-00044.
Barthlott, W., Mutke, J., Rafiqpoor, D., Kier, G., Kreft, H., 2005. Global centers of
vascular plant diversity. Nova Acta Leopold. NF 92 (342), 61–83.
Bennett, K.D., 1994. “psimpoll” version 2.23: a C program for analysing pollen data and
plotting pollen diagrams. INQUA Comm. Study Holocene Work. Group Data-
Handl. Methods Newlett. 11, 4–6.
Burn,M.J., Mayle, F.E., 2008. Palynological differentiation between genera of the
Moraceae family and implications for Amazonian palaeoecology. Rev. Palaeobot.
Palynol. 149, 187–201. http://dx.doi.org/10.1016/j.revpalbo.2007.12.003.
Burn, M.J., Mayle, F.E., Killeen, T.J., 2010. Pollen-based differentiation of Amazonian
rainforest communities and implications for lowland palaeoecology in tropical
South America. Palaeogeogr. Palaeoclimatol. Palaeoecol. 295, 1–18.
http://dx.doi.org/10.1016/j.palaeo.2010.05.009.
Bush, M.B., 1995. Neotropical plant reproductive strategies and fossil pollen
representation. Am. Nat. 145, 594–609. http://dx.doi.org/10.1086/285757.
136
Classified - Internal use
Bush, M.B., Rivera, R., 2001. Reproductive ecology and pollen representation among
neotropical trees. Glob. Ecol. Biogeogr. 10, 359–367.
http://dx.doi.org/10.1046/j.1466-822X.2001.00247.x.
Cárdenas, M.L., Gosling,W.D., Pennington, R.T., Poole, I., Sherlock, S.C., Mothes, P.,
2014. Forests of the tropical eastern Andean flank during the middle Pleistocene.
Palaeogeogr. Palaeoclimatol. Palaeoecol. 393, 76–89.
http://dx.doi.org/10.1016/j.palaeo.2013.10.009.
Cavalcante, A., Soares, J.J., Figueiredo, M.A., 2000. Comparative phytosociology of tree
sinusiae between contiguous forests in different stages of succession. Rev.
Brasileira Biol. 60, 551–562.
Colinvaux, P., De Oliveira, P.E., Patiño, J.E.M., 1999. Amazon Pollen Manual and Atlas.
Harwood Academic Publisher, Amsterdam.
Costa-Junior, R.F., Ferreira, R.L.C., Rodal, M.J.N., Feliciano, A.L.P., Marangon, L.C.,
Silva,W.C., 2008. Estrutura fitossociológica do componente arbóreo de
umfragmento de Floresta Ombrófila Densa na mata sul de Pernambuco, nordeste
do Brasil. Ciênc. Florest. 18, 173–183. http://dx.doi.org/10.5902/19805098455.
Cottam, G., Curtis, J.T., 1956. The use of distance measures in phytosociological
sampling. Ecology 37, 451–460. http://dx.doi.org/10.2307/1930167.
Duarte, M.c., Rego, F., Moreira, I., 2005. Distribution patterns of plant communities on
Santiago Island, Cape Verde. J. Veg. Sci. 16, 283–292.
http://dx.doi.org/10.1111/j.1654-1103.2005.tb02366.x.
Faegri, K., Iversen, J., 1975. Textbook of Pollen Analysis. John Wiley & Sons, London.
Fernández-Palacios, J.M., de Nicolás, J.P., 1995. Altitudinal pattern of vegetation
variation on Tenerife. J. Veg. Sci. 6, 183–190. http://dx.doi.org/10.2307/3236213.
Ferraz, E.M.N., Rodal, M.J.N., Sampaio, E.V.S.B., Pereira, R.d.C.A., 1998. Composição
florística em trechos de vegetação de caatinga e brejo de altitude na região do Vale
do Pajeú. Pernambuco. Braz. J. Bot. 21, 7–15. http://dx.doi.org/10.1590/S0100-
84041998000100002.
Ferraz, E.M.N., Araújo, E.L., Silva, S.I., 2004. Floristic similarities between lowland and
montane areas of Atlantic Coastal Forest in Northeastern Brazil. Plant Ecol. 174,
59–70. http://dx.doi.org/10.1023/B:VEGE.0000046062.77560.f5.
Giambelluca, T.W., Chen, Q., Frazier, A.G., Price, J.P., Chen, Y.-L., Chu, P.-S., Eischeid,
J.K., Delparte, D.M., 2012. Online rainfall atlas of Hawai'i. Bull. Am. Meteorol.
Soc. 94, 313–316. http://dx.doi.org/10.1175/BAMS-D-11-00228.1.
137
Classified - Internal use
Gomes, J.M.d.S., Lima e Lima, L.C., Santos, F.dA.R., Silva, F.H.M.e., 2014. First records
of pollen rain in bromeliad tanks in an area of Caatinga in northeastern Brazil.
Acta Bot. Bras. 28, 176–183. http://dx.doi.org/10.1590/S0102-
33062014000200004.
Gosling,W., Mayle, F.E., Tate, N.J., Killeen, T.J., 2005. Modern pollain-rain
characteristics of tall terra firme moist evergreen forest, Southern Amazonia.
Quat. Res. 64, 284–297.
Gosling,W.D., Mayle, F.E., Tate, N.J., Killeen, T.J., 2009. Differentiation between
Neotropical rainforest, dry forest, and savannah ecosystems by their modern
pollen spectra and implications for the fossil pollen record. Rev. Palaeobot.
Palynol. 153, 70–85. http://dx.doi.org/10.1016/j.revpalbo.2008.06.007.
Grabandt, R.A.J., 1980. Pollen rain in relation to arboreal vegetation in the Colombian
cordillera oriental. Rev. Palaeobot. Palynol. 29, 65–147.
http://dx.doi.org/10.1016/0034-6667(80)90043-3.
Hastenrath, S., Heller, L., 1977. Dynamics of climatic hazards in northeast Brazil. Q. J.
R. Meteorol. Soc. 103, 77–92. http://dx.doi.org/10.1002/qj.49710343505.
Hemp, A., 2005. Continuum or zonation? Altitudinal gradients in the forest vegetation of
Mt. Kilimanjaro. Plant Ecol. 184, 27–42. http://dx.doi.org/10.1007/s11258-005-
9049-4.
Hijmans, R.J., Cameron, S.E., Parra, J.L., Jones, P.G., Jarvis, A., 2005. Very high
resolution interpolated climate surfaces for global land areas. Int. J. Climatol. 25,
1965–1978. http://dx.doi.org/10.1002/joc.1276.
Horn, S.P., B, W.R., 1990. On the occurrence of Ficus pollen in neotropical Quaternary
sediments. Palynology 14, 3–6.
Ishara, K.L., Maimoni-Rodella, R.d.C.S., 2011. Pollination and dispersal systems in a
Cerrado remnant (Brazilian Savanna) in Southeastern Brazil. Braz. Arch. Biol.
Technol. 54, 629–642. http://dx.doi.org/10.1590/S1516-89132011000300025.
Jones, H.T., Mayle, F.E., Pennington, R.T., Killeen, T.J., 2011. Characterisation of
Bolivian savanna ecosystems by their modern pollen rain and implications for
fossil pollen records. Rev. Palaeobot. Palynol. 164, 223–237.
http://dx.doi.org/10.1016/j.revpalbo.2011.01.001.
Leal, A., Berrío, J.C., Raimúndez, E., Bilbao, B., 2011. A pollen atlas of premontane
woody and herbaceous communities from the upland savannas of Guayana,
138
Classified - Internal use
Venezuela. Palynology 35, 226–266.
http://dx.doi.org/10.1080/gspalynol.35.2.226.
Lima, J.R., Sampaio, E.V.S.B., Rodal, M.J.N., Araújo, F.S., 2009. Composição florística
da floresta estacional decídua montana de Serra das Almas, CE. Brasil. Acta Bot.
Bras. 23, 756–763.
Lüttge, U., 2006. Photosynthetic flexibility and ecophysiological plasticity: questions and
lessons from Clusia, the only CAM tree, in the neotropics. New Phytol. 171, 7–
25. http://dx.doi.org/10.1111/j.1469-8137.2006.01755.x.
Martin, P.H., Sherman, R.E., Fahey, T.J., 2007. Tropical montane forest ecotones: climate
gradients, natural disturbance, and vegetation zonation in the Cordillera Central,
Dominican Republic. J. Biogeogr. 34, 1792–1806.
http://dx.doi.org/10.1111/j.1365-2699.2007.01726.x.
Pearson, T.R.H., Burslem, D.F.R.P., Mullins, C.E., Dalling, J.W., 2002. Germination
ecology of neotropical pioneers: interacting effects of environmental conditions
and seed size. Ecology 83, 2798–2807. http://dx.doi.org/10.1890/0012-
9658(2002)083[2798:GEONPI]2.0.CO;2.
Pessoa, M.d.S., Vleeschouwer, K.M.D., Talora, D.C., Rocha, L., Amorim, A.M.A., 2012.
Reproductive phenology of Miconia mirabilis (Melastomataceae) within three
distinct physiognomies of Atlantic Forest, Bahia, Brazil. Biota Neotrop. 12, 49–
56. http://dx.doi.org/10.1590/S1676-06032012000200006.
Peulvast, J.-P., de Claudino Sales, V., 2004. Stepped surfaces and palaeolandforms in the
northern Brazilian “Nordeste”: constraints on models of morphotectonic
evolution. Geomorphology 62, 89–122.
http://dx.doi.org/10.1016/j.geomorph.2004.02.006.
QGIS Development Team, 2015. QGIS Geographic Information System Developers
Manual. Open Source Geospatial Foundation Project.
Richter, M., 2008. Tropical mountain forests – distribution and general features. In:
Gradstein, S., Homeier, J., Gansert, D. (Eds.), The Tropical Mountain Forest –
Patterns and Processes in a Biodiversity HotspotBiodiversity and Ecology Series.
Göttingen Centre for Biodiversity and Ecology, pp. 7–24.
Rizzini, C.T., 1963. Nota prévia sobre a divisão fitogeográfica do Brasil. Rev. Bras.
Geogr. 1, 1–64.
Roubick, D.W., Moreno, J.E., 1991. Pollen and Spores of Barro Colorado Island.
Missouri Botanical Garden, Saint Louis, MO.
139
Classified - Internal use
Rull, V., 2003. An illustrated key for the identification of pollen from Pantepui and the
Gran Sabana (eastern Venezuelan Guayana). Palynology 27, 99–133.
http://dx.doi.org/10.2113/27.1.99.
Santo-Silva, E.E., Almeida, W.R., Melo, F.P.L., Zickel, C.S., Tabarelli, M., 2013. The
nature of seedling assemblages in a fragmented tropical landscape: implications
for forest regeneration. Biotropica 45, 386–394.
http://dx.doi.org/10.1111/btp.12013.
Santos, D.A.D., Lima e Lima, L.C., Santos, F.d.A.R., Silva, F.H.M.e., 2015. First report
of modern pollen deposition in moss polsters in a semiarid area of Bahia Brazil.
Acta Bot. Bras. 29, 534–544.
Sarthou, C., Kounda-Kiki, C., Vaçulik, A., Mora, P., Ponge, J.-F., 2009. Successional
patterns on tropical inselbergs: a case study on the Nouragues inselberg (French
Guiana). Flora - Morphol. Distrib. Funct. Ecol. Plants 204, 396–407.
http://dx.doi.org/10.1016/j.flora.2008.05.004.
Schüler, L., Hemp, A., Behling, H., 2014. Relationship between vegetation and modern
pollen-rain along an elevational gradient on Kilimanjaro, Tanzania. The Holocene
24, 702–713. http://dx.doi.org/10.1177/0959683614526939.
Shoo, L.P., Storlie, C., Vanderwal, J., Little, J., Williams, S.E., 2011. Targeted protection
and restoration to conserve tropical biodiversity in a warming world. Glob.
Change Biol. 17, 186–193. http://dx.doi.org/10.1111/j.1365-2486.2010.02218.x.
Siqueira, D.R., Rodal, M.J.N., Lins-e-Silva, A.C.B., Melo, A.L., 2001. Physiognomy,
structure and floristic in an area of Atlantic Forest in northeast Brazil. In:
Gottsberger, G., Liede, S. (Eds.), Life Forms and Dynamics in Tropical Forest.
Gebr. Borntraeger, Berlin, pp. 11–27.
Souza, J.P., Araújo, G.M., Schiavini, I., Duarte, P.C., 2006. Comparison between canopy
trees and arboreal lower strata of urban semideciduous seasonal forest in Araguari
-MG. Braz. Arch. Biol. Technol. 49, 775–783. http://dx.doi.org/10.1590/S1516-
89132006000600012.
Stockmarr, J., 1972. Tablets with spores used in absolute pollen analysis. Pollen Spores
13, 615–621.
Urrego, D.H., Silman, M.R., Correa-Metrio, A., Bush,M.B., 2011. Pollen–vegetation
relationships along steep climatic gradients in western Amazonia. J. Veg. Sci. 22,
795–806. http://dx.doi.org/10.1111/j.1654-1103.2011.01289.x.
140
Classified - Internal use
Vaasen, A., Begerow, D., Hampp, R., 2006. Phosphoenolpyruvate carboxylase genes in
C3, crassulacean acid metabolism (CAM) and C3/CAM intermediate species of
the genus Clusia: rapid reversible C3/CAM switches are based on the C3
housekeeping gene. Plant Cell Environ. 29, 2113–2123.
http://dx.doi.org/10.1111/j.1365-3040.2006.01583.x.
Veloso, H.P., 2012. Manual técnico da vegetação brasileira, IBGE. ed. 2th ed. IBGE, Rio
de Janeiro.
Weng, C., Bush,M.B., Silman,M.R., 2004. An analysis of modern pollen rain on an
elevational gradient in southern Peru. J. Trop. Ecol. 20, 113–124.
http://dx.doi.org/10.1017/S0266467403001068.
Williams, S.E., Bolitho, E.E., Fox, S., 2003. Climate change in Australian tropical
rainforests: an impending environmental catastrophe. Proc. R. Soc. Lond. B Biol.
Sci. 270, 1887–1892. http://dx.doi.org/10.1098/rspb.2003.2464.
Rerences Chapter III
Andrade LA, Oliveira FX, Nascimento,IS, Fabricante JR, Sampaio EVSB, Barbosa
MRV. 2006. Análise florística e estrutural de matas ciliares ocorrentes em brejo de
altitude no/ município de Areia, Paraíba. Revista Brasileira de Ciências Agrárias 1: 31-
40.
Araújo FS, Gomes VS, Silveira AP, Figueiredo MA, Oliveira RS, Bruno MMA, Lima-
Verde LW, Silva EF, Otutumi AT, Ribeiro KA. 2007. Efeito da variação topoclimática e
estrutura da vegetação da serra de Baturité, Ceará. In: Oliveira TS. and Araújo FS. (eds),
Diversidade e conservação da Biota da serra de Baturité, Ceará. Edições UFC/COELCE,
pp. 73-136.
Austin MP. 2005. Vegetation and environment: discontinuities and continuities. — In:
van der Maarel E. (eds), Vegetation Ecology. Blackwell Publishing, pp. 52–84.
Austin MP. 2007. Species distribution models and ecological theory: a critical
assessment and some new approaches. Ecological Modelling 200, 1–19.
141
Classified - Internal use
Baker HG. 1972. Seed weight in relation to environmental conditions in California.
Ecology 53:997-1010.
Bolker BM, Pacala SW, Neuhauser C. 2003. Spatial dynamics in model plant
communities: what do we really know? American Naturalist 162:135-148.
Boulangeat I, Gravel D, Thuiller W. 2012. Accounting for dispersal and biotic
interactions to disentangle the drivers of species distributions and their
abundances. Ecology Letters 15: 584–593.
Brown Jr. KS, Brown GG. 1992. Habitat alteration and species loss in Brazilian forests.
In: Whitmore TC, Sayer JA. (Eds.), Tropical Deforestation and Species Extinction.
Chapman & Hall, London, pp. 129–142.
Bruno JF, Stachowicz JJ, Bertness MD. 2003. Inclusion of facilitation into ecological
theory. Trends in Ecology & Evolution 18:119-125.
Calba S, Maris V, Devictor V. 2014. Measuring and explaining large-scale distribution
of functional and phylogenetic diversity in birds: separating ecological drivers from
methodological choices. Global Ecology and Biogeography 23: 669–678.
Cavalcante AMB, Soares JJ, Figueredo MA. 2000. Comparative phytosociology of tree
sinusiae between contiguous forests in different stages of succession. Revista Brasileira
de Biologia 60: 551–562.
Cavender-Bares J, Bazzaz FA. 2000. Changes in drought response strategies with
ontogeny in Quercus rubra: implications for scaling from seedlings to mature trees.
Oecologia 124:8–18
142
Classified - Internal use
Chust G, Chave J, Condit R, Aguilar S, Lao S., Perez R. 2006. Determinants and spatial
modeling of tree beta-diversity in a tropical forest landscape in Panama. Journal of
Vegetation Science 17: 83–92.
Clarke KR, Warwick RM. 2001. A further biodiversity index applicable to species lists:
variation in taxonomic distinctness. Marine Ecology Progress Series 216: 265–278.
Comita LS, Condit R, Hubbell SP. 2007. Developmental changes in habitat associations
of tropical trees. Journal of Ecology 95:482-492.
Condit R, Ashton PS, Baker P, Bunyavejchewin S, Gunatilleke S, Gunatilleke N,
Hubbell SP, Foster RB, Itoh A, LaFrankie JV, Lee HS, Losos E, Manokaran N,
Sukumar R, Yamakura T. 2000. Spatial patterns in the distribution of tropical tree
species. Science 288:1414-1418.
Costa-Júnior RF, Ferreira RLC, Rodal MJN, Feliciano ALP, Marangon LC, Silva WC.
2008. Estrutura fitossociológica do componente arbóreo de um fragmento de floresta
ombrófila densa na Mata Sul de Pernambuco, Nordeste do Brasil. Ciência Florestal
18(2): 173-183.
Coudon C, Gegout J, Piedallu C, Rameau J. 2006. Soil nutritional factors improve
models of plant species distribution: an illustration with Acer campestre (L.) in France.
Journal of Biogeography 33: 1750–1763.
Cunha MCL. 2010. Comunidades de árvores e o ambiente na floresta estacional
semidecidual montana do Pico do Jabre, PB [PhD thesis]. [Brasília (DF)]: Universidade
de Brasília, Brasília.
Diogo IJS, Martins FR, Verola CF, Costa IR. 2016. Variation in plant-interactions along
143
Classified - Internal use
an elevational gradient of moist forest in a semiarid area of Brazil. Acta Botanica
Brasilica 30(1): 27-34.
Diniz-Filho, JAF, Rangel TLVB, Bini LM. 2008. Model selection and information
theory in geographical ecology. Global Ecology and Biogeography 17:479-488.
Drezner TD, Fall PL, Stromberg JC. 2001. Plant distribution and dispersal mechanisms
at the Hassayampa River Preserve, Arizona, USA. Global Ecology and Biogeography
10:205-217.
Ejrnæs R, Bruun HH, Graae BJ. 2006. Community assembly in experimental
grasslands: suitable environment or timely arrival? Ecology 87:1225-1233.
Ferraz EMN, Rodal MJN. 2008. Floristic characterization of a remnant ombrophyllous
montane forest at São Vicente Férrer, Pernambuco, Brazil. Memoirs of the New York
Botanical Garden 100: 468-510.
Ferraz EMN, Rodal MJN, Sampaio, EVSB. 2003. Physiognomy and structure of
vegetation along an altitudinal gradient in the semi-arid region of northeastern
Brazil. Phytocoenologia 33(1): 71-92.
Freschett GT et al. 2013. Linking litter decomposition of above- and below-ground
organs to plant–soil feedbacks worldwide. Journal of Ecology 101: 943–952.
Getzin S, Wiegand T, Wiegand K, He F. 2008. Heterogeneity influences spatial patterns
and demographics in forest stands. Journal of Ecology 96:807-820.
Godsoe W, Jankowski J, Holt RD, Gravel D. 2017. Integrating Biogeography with
Contemporary Niche Theory. Trends in Ecology & Evolution.
144
Classified - Internal use
Götzenberger L, d Bello F, Bråthen KA, Davison J, Dubuis A, Guisan A, Lepš J,
Lindborg R, Moora M, Pärtel M, Pellissier L, Pottier J, Vittoz P, Zobel K, Zobel M.
2011. Ecological assembly rules in plant communities — approaches, patterns and
prospects. Biological Reviews.
Guisan A, Zimmerman NE. 2000. Predictive habitat distribution models in
ecology. Ecological Modelling 135: 147–186.
Hall JS, McKenna JJ, Ashton PMS, Gregoire TG. 2004. Habitat characterization
underestimate the role of edaphic factors controlling the distribution of
Entandrophragma. Ecology 85:2171-2183.
Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A. 2005. Very high resolution
interpolated climate surfaces for global land areas. International Journal of Climatology
25: 1965-1978.
Horn S, Hempel S, Ristow M, Rillig MC, Kowarik I, Caruso T. 2015. Plant community
assembly at small scales: Spatial vs. environmental factors in a European grassland.
Acta oecologica 63: 56–62.
Howe HF, Smallwood J. 1982. Ecology of seed dispersal. Annual Review of Ecology
and Systematics 13, 201–228.
Hubbell SP. 2001. The unified neutral theory of biodiversity and biogeography.
MPB32. Princenton (NJ): Princeton University Press.
Jenny H. 1950. Causes of high nitrogen and organic matter content in temperate and
tropical soils. Soil Science 69: 63–69.
145
Classified - Internal use
Jobbágy EG and Jackson RB. 2000. The vertical distribution of soil organic carbon and
its relation to climate and vegetation. Ecological applications 10: 423–436.
John R, Dalling JW, Harms KE, Yavitt JB, Stallard RF, Mirabello M, Hubbell SP,
Valencia R, Navarrete H, Vallejo M, Foster RB. 2007. Soil nutrients influence spatial
distribution of tropical tree species. Proceedings of the National Academy of Sciences
of the United States of America 104:864-869.
Kupfer JA, Malanson GP. 1993. Structure and composition of a riparian forest
edge. Physical Geography 14: 154–170.
Legendre P, Legendre L. 1998. Numerical Ecology. Amsterdam (BV): Elsevier Science.
Leishman MR, Westoby M. 1994. The role of large seed size in shaded conditions:
experimental evidence. Functional Ecology 8:205-214.
Linares-Palomino RA, Oliveira-Filho T, Pennington RT. 2010. Neotropical seasonally
dry forests: Diversity, endemism and biogeography of woody plants. In R. Dirzo, H.
Mooney, G. Ceballos, and H.Young (Eds.). Seasonally dry tropical forests: Biology and
conservation, pp. 3–21. Island Press, Washington, USA.
Machado WJ, Prata APN, Mello AA. 2012. Floristic composition in areas of Caatinga
and Brejo de Altitude in Sergipe state, Brazil. Check List 8(6): 1089–1101.
Mair L, Hill JK, Fox R, Botham M, Brereton T, Thomas CD. 2014. Nature Climate
Change 4:127–131.
McCune B, Mefford MJ. 2011. PC-ORD. Multivariate Analysis of Ecological Data.
Oregon.
146
Classified - Internal use
Meier ES, Kienast F, Pearman PB, Svenning J-C, Thuiller W, Araújo MB, Guisan A
and Zimmermann NE. 2010. Biotic and abiotic variables show little redundancy in
explaining tree species distributions. Ecography 33: 1038–1048.
Mollison, B, Slay R. 2000. Introduction to permaculture. 2nd ed. Tyalgum,, Australia:
Tagari Publications.
Motta AC, Reeves DW, Touchton JT. 2002.Tillage intensity effects on chemical
indicators of soil quality in two coastal plain soils. Communications in Soil Science and
Plant Analysis 33:913-932.
Nieto-Lugilde D et al. 2014. Tree cover at fine and coarse spatial grains interacts with
shade tolerance to shape plant species distributions across the Alps. Patterns and process
in Ecology 38: 578–589.
Normand S, Vormisto J, Svenning, JC, Grández C, Balslev, H. 2006. Geographical and
environmental controls of palm beta diversity in paleo-riverine terrace forests in
Amazonian Peru. Plant Ecology, 186:161–176.
Oades JM. 1988. The retention of organic matter in soils. Biogeochemistry 5:33–70.
Paruelo JM, Jobbágy EG, Sala OE, Lauenroth WK, Burke IC. 1997. Functional and
structural convergence of temperate grassland and shrubland ecosystems. Ecological
Applications 8:194–206.
Pearson RG and Dawson TP. 2003. Predicting the impacts of climate change on the
distribution of species: are bioclimate envelope models useful? Global Ecology an
Biogeography 12(5): 361–371.
147
Classified - Internal use
Pereira RCA, Silva JA, Barbosa JIS. 2010. Flora de um “brejo de altitude” de
Pernambuco: Reserva ecológica da Serra Negra. Anais da Academia Pernambucana de
Ciência Agronômica 7:286-304.
Perry GLW, Enright NJ, Miller BP, Lamont BB. 2009. Nearest-neighbour interactions
in species-rich shrublands: the roles of abundance, spatial patterns and resources. Oikos
118:161-174.
Phillips OL, Vargas PN, Monteagudo AL, Cruz AP, Zans MEC, Sanchez WG, Yli-Hall,
M, Rose, S. 2003. Habitat association among Amazonian tree species: a landscape-scale
approach. Journal of Ecology 91: 757–775.
Pijl LV. 1982. Principles of dispersal in higher plants. 3.ed. Berlim: Springer Verlag.
Pinto MSC, Sampaio EVSB, Nascimento LM. 2012. Florística e estrutura da vegetação
de um brejo de altitude em Pesqueira, PE, Brasil. Revista Nordestina de Biologia 21(1):
47-79.
Pons J, Pausas JG. 2006. Oak regeneration in heterogeneous landscapes: the case of
fragmented Quercus suberforests in the eastern Iberian Peninsula. Forest Ecology &
Management 231: 196–204.
Potts MD, Ashton PS, Kaufman LS, Plotkin JB. 2002. Habitat patterns in tropical
forests: A comparison of 105 plots in Northwest Borneo. Ecology 83: 2782–2797.
Rangel TF, Diniz-Filho JAF, Bini LM. 2010. SAM: a comprehensive application for
Spatial Analysis in Macroecology. Ecography 33:46-50.
Primack RB and Miao SL. 1992. Dispersal Can Limit Local Plant Distribution.
Conservation Biology 6: 513-519.
148
Classified - Internal use
Ribeiro-Silva S, Medeiros MB, Gomes BM, Seixas ENC, Silma MAP. 2012.
Angiosperms from the Araripe National Forest, Ceará, Brazil. Check List 8(4): 744–
751.
Rodal MJN, Nascimento LM. 2002. Levantamento florístico da floresta serrana da
Reserva Biológica de Serra Negra, Itaparica-PE. Acta botanica brasílica 16:481-500.
Rodal MJN, Nascimento LM. 2008. .Fisionomia e estrutura de uma floresta estacional
montana do maciço da Borborema, Pernambuco – Brasil. Revista Brasileira de Botânica
31:27-39.
Rodal MJN, Sales MF, Silva MJ, Silva AG. 2005. Flora de um Brejo de Altitude na
escarpa oriental do planalto da Borborema, PE, Brasil. Acta botanica brasilica 19(4):
843-858.
Roy M, Rochet J, Manzi S, Jargeat P, Gryta H, Moreau P, Gardes M. 2013. What
determines Alnus-associated ectomycorrhizal community diversity and specificity? A
comparison of host and habitat effects at a regional scale. New phytologist 198: 1228–
1238.
Russo SE, Davies SJ, King DA, Tan S. 2005. Soil-related performance variation and
distributions of tree species in a Bornean rain forest. Journal of Ecology 93:879-889.
Santos, A.M.M & Tabarelli, M. 2004. Integridade, esforço e diretrizes para
conservação dos brejos de Paraíba e Pernambuco. Brejos de altitude em Pernambuco e
Paraíba (ed. por K.C. Porto, J.J.P. Cabral & M. Tabarelli), pp 309-318. Ministério do
Meio Ambiente – MMA, Brasília-DF.
Schlesinger WH. 1977. Carbon balance in terrestrial detritus. Annual Review of
Ecology, Evolution and Systematics 8: 51–81.
149
Classified - Internal use
Seidler TG, Plotkin JB. 2006. Seed dispersal and spatial pattern in tropica trees. PLoS
Biology 4:2132-2137.
Silva JMC, Casteletti CHM. 2003. Status of the biodiversity of the Atlantic Forest of
Brazil. In: Galindo-Leal C., Câmara IG. The Atlantic Forest of South America:
biodiversity status, threats, and outlook. Washington (DC): Center for Applied
Biodiversity Science and Island Press, pp. 43–59.
Tabarelli M, Santos AMM. 2004. Uma breve descrição sobre a história natural dos
brejos nordestinos. In: Pôrto KC, Cabral JJP, Tabarelli M. Brejos de altitude em
Pernambuco e Paraíba: história natural, ecologia e conservação. Brasília (DF):
Ministério do Meio Ambiente, pp. 17–24.
Tavares MCG, Rodal MJN, Melo AL, Araújo MF 2000. Fitossociologia do componente
arbóreo de um trecho de Floresta Ombrófila Montana do Parque Ecológico João
Vasconcelos Sobrinho, Caruaru, Pernambuco. Naturalia 25: 17-32.
Ter Braak, CJF. 1986. Canonical correspondence analysis: a new eigenvector technique
for multivariate direct gradient analysis. Ecology 67: 1167–1179.
Terborgh J. 1973. On the notion of favorableness in plant ecology. The American
Naturalist 107:481-501.
Thuiller W. 2004. Patterns and uncertainties of species’ range shifts under climate
change. Global Change Biology 10: 2020–2027.
Tiessen H, Cuevas E, Chacon P. 1994. The role of soil organic matter in sustaining soil
fertility. Nature 371:783– 785.
150
Classified - Internal use
Tilman D. 2004. Niche tradeoffs, neutrality, and community structure: a stochastic
theory of resource competition, invasion, and community assembly. Proceedings of
National Academic Science 101:10854–10861.
Tuomisto H, Ruokolainen K. 2008. Analyzing or explaining beta diversity? Reply.
Ecology 89:3244-3256.
VanDerWal J, Shoo LP, Johnson CN, Williams SE. 2009. Abundance & the
Environmental Niche: environmental Suitability Estimated from Niche Models Predicts
the Upper Limit of Local Abundance. The American Naturalist 174:282–291.
Vincent RC, Meguro M. 2008. Influence of soil properties on the abundance of plant
species in ferruginous rocky soils vegetation, southeastern Brazil. Brazilian Journal of
Botany 31: 377-388.
Wiens JJ. 2011. The niche, biogeography and species interactions. Philosophical
Transactions of the Royal Society of London B: Biological Sciences 366:2336-2350.
References Chapter IV
References
Ab’saber A.N. 1977. Espaços Ocupados pela Expansão dos Climas secos na América do
Sul por ocasião dos períodos Glaciais Quaternários. Paleoclimas, 3: 1-9.
Ab’saber A.N. 1983. O domínio dos cerrados: Introdução ao conhecimento. Revista do
Servidor 111:41-55.
Ab’Sáber AN. 2003. Os domínios de natureza no Brasil: potencialidades paisagísticas.
São Paulo: Ateliê Editorial.
151
Classified - Internal use
Aguirre-Planter E., Furnier G.R., Eguiarte L.E. 2000. Low levels of genetic variation
within and high levels of genetic differentiation among populations of species of
Abies from southern México and Guatemala. American Journal of Botany, 87,
362–371.
Andrade-lima, D. 1966. Contribuição ao paralelismo da flora amazônico-nordestina. Pp.
1-30. In: Boletim Técnico IPA (Boutton, T.W., eds). Recife.
Andrade-lima, D. 1982. The Caatinga Dominium. Revista Brasileira de Botânica, 4: 149-
153.
APG IV. 2016. An update of the Angiosperm Phylogeny Group classification for the
orders and families of flowering plants: APG IV. Botanical Journal of the Linnean
Society 181: 1–20.
Auler, A.S.; Wang, X.; Edwards, R.L.; Cheng, H.; Cristalli, P.S.; Smart, P.L.; Richards,
D.A. 2004. Quaternary ecological and geomorphic changes associated with
rainfall events in presently semi-arid northeastern Brazil. Journal of Quaternary
Science, 19(7): 693-701.
Barbosa, A. R., Fiorini, C. F., Silva-Pereira, V., Mello-Silva, R., and Borba, E.
L. 2012. Geographical genetic structuring and phenotypic variation in
the Vellozia hirsuta(Velloziaceae) ochlospecies complex. American Journal of
Botany 99: 1477–1488.
Batalha-Filho, H., Fjeldså, J., Fabre, P. H., and Miyaki, C. Y. 2013. Connections
between the Atlantic and the Amazonian forest avifaunas represent distinct
historical events. Journal of Ornithology 154: 41–50.
152
Classified - Internal use
Bonatelli, I. A. S., Perez, M. F., Peterson, A. T., Taylor, N. P., Zappi, D.
C., Machado, M. C., Koch, I., Pires, A. H., and Moraes, E. M. 2014. Interglacial
microrefugia and diversification of a cactus species complex: phylogeography and
palaeodistributional reconstructions for Pilosocereus aurisetus and
allies. Molecular Ecology 23: 3044–3063.
Behling, H., Arz, W.H., Pätzold, J., Wefer, G., 2000. Late quaternary vegetational and
climate dynamics in northeastern Brazil, inferences from marine core GeoB3104-
1. Quaternary Science Reviews, 19: 981- 994.
Behling H. & Negrelle R.R.B. 2001. Tropical Rain Forest and Climate Dynamics of the
Atlantic Lowland, Southern Brazil, during the Late Quaternary. Quaternary
Research 56: 383-389.
Behling, H. 2002. South and southeast Brazilian grasslands during Late Quaternary times:
a synthesis. Palaeogeography, Palaeoclimatology, Palaeoecolgy 177: 19-27.
Bigarella, J.J.; Andrade-Lima, D. & Riehs, P.J. (1975) Considerações a respeito das
mudanças paleoambientais na distribuição de algumas espécies vegetais e animais
no Brasil. Anais da Academia Brasileira de Ciência, 47: 411-464.
Bueno, M., Pennington, R. T., Dexter K. G., Kamino, L. H. Y., Pontara, L. Neves, D. R.
M., Ratter, J. A., Oliveira-Filho, A. T., 2016. Effects of Quaternary Climatic
Fluctuations on the Distribution of Neotropical Savanna Tree Species. Ecography
39:1-12.
Bush, M. B. and de Oliveira, P. E. 2006. The rise and fall of the refugial hypothesis of
Amazonian speciation: a paleoecological perspective. Biota Neotropica 6: 1–17.
153
Classified - Internal use
Camardelli, M. & Napoli, M.F. (2012) Amphibian conservation in the Caatinga biome
and semiarid region of Brazil. Herpetologica, 68, 31–47.
Carnaval, A. C. and Moritz, C. 2008. Historical climate modelling predicts patterns of
current biodiversity in the Brazilian Atlantic forest. Journal of Biogeography 35:
1187–1201.
Carnaval, A.C. & Bates, J.M. (2007) Amphibian DNA shows marked genetic structure
and tracks Pleistocene climate change in northeastern
Brazil. Evolution, 61, 2942–2957.
Carnaval, A.C., Hickerson, M.J., Haddad, C.F.B., Rodrigues, M.T. & Moritz,
C. 2009. Stability predicts genetic diversity in the Brazilian atlantic forest
hotspot. Science 323: 785–789.
Chen H. & Boutros P.C. 2011. VennDiagram: a package for the generation of highly-
customizable Venn and Euler diagrams in R. BMC Bioinformatics 12:35.
Clarke K.R., Warwick R.M. 2001. A further biodiversity index applicable to species lists:
variation in taxonomic distinctness. Marine Ecology Progress Series 216: 265–
278.
Coimbra-Filho A.F, Câmara, I.G. 1996. Os limites originais do bioma Mata Atlaˆntica na
Região do Nordeste do Brasil. FBCN.
Cole, M.M. 1960. Cerrado, caatinga and pantanal: The distribuition and origin of the
savana vegetation of Brazil. The Geographical Journal, 126(2): 168-179.
154
Classified - Internal use
Costa, L.P. 2003. The historical bridge between the Amazon and the Atlantic Forest of
Brazil: a study of molecular phylogeography with small mammals. Journal of
Biogeography 30: 71-86.
Colinvaux, P.A., De Oliveira, P.E., Moreno, J.E., Miller, M.C., Bush, M.B. (1996) A
long pollen record from lowland Amazonia: forest cooling in glacial
times. Science 274: 85–87.
Collevatti, R. G., Rabelo, S. G., and Vieira, R. F. 2009. Phylogeography and disjunct
distribution in Lychnophora ericoides (Asteraceae), an endangered cerrado shrub
species. Annals of Botany 104: 655–664.
Cracraft, J. & Prum, R.O. 1988. Patterns and processes of diversification: speciation and
historical congruence in some Neotropical birds. Evolution, 42, 603–620.
Edler, D., Guedes, T., Zizka, A., Rosvall, M. & Antonelli, A. 2017. Infomap bioregions:
interactive mapping of biogeographical regions from species
distributions. Systematic Biology 66(2):197–204.
Fouquet A., Loebmann D., Castroviejo-Fisher S., Padial J.M, Orrico, V.G.D., Lyra M.L.,
Roberto, I.J., Kok, F.J.R, Haddad, C.F.B, Rodrigues, M.T. 2012. From Amazonia
to the Atlantic forest: molecular phylogeny of Phyzelaphryninae frogs reveals
unexpected diversity and a striking biogeographic pattern emphasizing
conservation challenges. Molecular Phylogenetics Evolution 65:547–556.
Fouquet A., Cassini C.S., Haddad C.F.B., Pech N., Rodrigues, M.T. 2014. Species
delimitation, patterns of diversification and historical biogeography of the
Neotropical frog genus Adenomera (Anura, Leptodactylidae). Journal of
Biogeography 41: 855–870.
155
Classified - Internal use
Gentry A.H. 1982. Neotropical floristic diversity: phytogeographical connections
between Central and South America, pleistocene climatic fluctuations, or an
accident of the andean orogeny? Annals of Missouri Botanical Garden 69:557-
593.
Guedes, T.B., Sawaya, R.J. & Nogueira, C. (2014) Biogeography, vicariance and
conservation of snakes of the neglected and endangered Caatinga region, north-
eastern Brazil. Journal of Biogeography, 41, 919–931.
Haffer, J. 1969. Speciation in Amazonn forest birds. – Science 168: 131–137.
Hensen I., Teich I., Hirsch H., von Wehrden H., Renison D. 2011. Range-wide genetic
structure and diversity of the endemic tree line species Polylepis australis
(Rosaceae) in Argentina. American Journal of Botany 98: 1825 – 1833
Hewitt G. M. 2000 The genetic legacy of the Quaternary ice ages. Nature 405: 907–913.
Hoorn C., Wesselingh F.P., terSteege H., Bermudez M.A., Mora A., Sevink J.,Sanmarti
n I. 2010. Amazonia through time: Andean uplift, climate change, landscape
evolution, and biodiversity. Science 330: 927–931.
Instituto Brasileiro de Geografia e Estatística (IBGE). 1995. Fundação Instituto Brasileiro
de Geografia e Estatística. Anuário Estatístico do Brasil. Rio de Janeiro: IBGE.
Jaramillo-Correa J.P., Aguirre-Planter E., Khasa D.P., Eguiarte L.E., Pinero D, Furnier
G.R. 2008. Ancestry and divergence of subtropical montane forest isolates:
molecular biogeography of the genus Abies (Pinaceae) in southern Mexico and
Guatemala. Molecular Ecology 17: 2476–2490.
156
Classified - Internal use
Koscinski, D., Handford, P., Tubaro, P.L., Sharp, S. & Loug- heed, S.C. (2008)
Pleistocene climatic cycling and diversifi- cation of the Andean treefrog,
Hypsiboas andinus. Molecular Ecology, 17, 2012-2025.
Leal, B.S.S., Silva, C.P., Pinheiro F. 2016. Phylogeographic Studies Depict the Role of
Space and Time Scales of Plant Speciation in a Highly Diverse Neotropical
Region. Critical reviews in Plant Science 35:215-230.
Ledru, M.P.; Braga, P.I.S; Soubiès, F.; Martin, L.; Suguio, K. & Turcq, B. 1996. The Last
50,000 years in the neotropics (Southern Brazil): evolution of vegetation and
climate. Palaeogeography, Palaeoclimatology, Palaeoecology 123:239-259.
Ledru M.P., Salgado-Labouriau M.L., Lorscheiter M.L. 1998. Vegetation dynamics in
southern and central Brazil during the last 10,000 yr. Review of Paleobotany and
Palynology 99:131–142.
Ledru, M. P. 2002. Late Quaternary history and evolution of the Cerrados as revealed by
palynological records. – In: Oliveira, P. S. and Marquis, R. J. (eds), The Cerrados
of Brazil: ecology and natural history of a Neotropical savanna. Columbia Univ.
Press, pp. 33–50.
Ledru M.P., Salatino M.L.F., Ceccantini G.T., Salatino A., Pinheiro F., Pintaud J.C. 2007.
Regional assessment of the impact of climatic change on the distribution of a
tropical conifer in the lowlands South America. Diversity and Distributions 13:
761–771.
Lins, R.C. 1989. As áreas de exceção do agreste de Pernambuco. Sudene, Recife.
157
Classified - Internal use
Loebmann, D. & Haddad, C.F.B. (2010) Amphibians and reptiles from a highly diverse
area of the Caatinga domain: composition and conservation implications. Biota
Neotropica, 10, 227–256.
Maechler, M., P. Rousseeuw, A. Struyf, M. Hubert, and K. Hornik. 2013. cluster: Cluster
Analysis Basics and Extensions. R package version 1.14.4.
Martini A.M.Z., Fiaschi P., Amorim A.M., Paixão, J.L. 2007. A hot-point within a hot-
spot: a high diversity site in Brazil’s Atlantic Forest. Biodiversity and
Conservation 16: 3111–3128.
Méio, B. B., Freitas, C V., Jatobá, L., Silva, M. E. F., Ribeiro, J. F., & Henriques, R. P.
B. 2003. Influência da flora das florestas Amazônica e Atlântica na vegetação do
cerrado sensu stricto. Brazilian Journal of Botany 26: 437-444.
Mello-Martins, F. Historical biogeography of the Brazilian Atlantic forest and the
Carnaval–Moritz model of Pleistocene refugia: what do phylogeographical
studies tell us? Biological Jorunal of the Linnean Society 104: 499-509.
Montade V, Ledru M-P, Burte J, Martins E, Verola CF, da Costa IR, Magalhaes H. 2014.
Stability of a Neotropical microrefugium during climatic instability. Journal of
Biogeography 41: 1215–1226.
Morley, R.J. (2000). Origin and Evolution of Tropical Rain Forests. John Wiley and
Sons, Chichester.
Muotka T., Laasonen P. 2002. Ecosystem recovery in restored headwater streams: the
role of enhanced leaf retention. Journal of Applied Ecology 39: 145–156.
158
Classified - Internal use
Novaes, R. M. L., Lemos-Filho, J. P., Ribeiro, R. A., and Lovato, M.
B.2010. Phylogeography of Plathymenia reticulate (Leguminosae) reveals
patterns of recent range expansion towards northeastern Brazil and southern
Cerrados in Eastern Tropical South America. Molecular Ecology 19: 985–998.
Oksanen, J., F. G. Blanchet, R. Kindt, P. Legendre, P. R. Minchin, R. B. O'Hara, G. L.
Simpson, P. Solymos, M. H. H. Stevens, and H. Wagner. 2012. vegan:
Community ecology package: R package version 2.1-13/r2115. http://CRAN.R-
project.org/package=vegan
Oliveira-Filho, A.T. 2014. NeoTropTree, Flora arbórea da Região Neotropical: Um
banco de dados envolvendo biogeografia, diversidade e
conservação. Universidade Federal de Minas Gerais. Available from
http://www.icb.ufmg.br/treeatlan/ [Accessed 30 May 2016].
Oliveira-Filho A.T., Ratter J.A. 1995. A study of the origin of Central Brazilian forests
by the analysis of plant species distribution patterns. Edinburgh Journal of Botany
52:41–194.
Olson D.M., Dinerstein E.,Wikramanayake E.D., Burgess N.D., Powell G.V.,
Underwood E.C., D’amico J.A., Itoua I., Strand H.E., Morrison J.C., Loucks C.J.,
Allnutt T.F., Ricketts T.H., Kura Y., Lamoreux J.F., WettengelW.W., Hedao P.,
Kassem K.R. 2001. Terrestrial ecoregions of the world: a new map of life on earth
a new global map of terrestrial ecoregions provides an innovative tool for
conserving biodiversity. BioScience 51:933–938.
Patton, J.L., Da Silva, M.N.F.& Malcolm, J.R. 2000. Mammals of the Rio Juruá and the
evolutionary and ecological diversification of Amazonia. Bulletin of the
American Museum of Natural History, 244, 1–306.
159
Classified - Internal use
Pellegrino K.C.M., Rodrigues M.T., Harris D.J., Yonenaga-Yassuda Y., Sites J.W. 2011.
Molecular phylogeny, biogeography and insights into the origin of
parthenogenesis in the Neotropical genus Leposoma (Squamata:
gymnophthalmidae): Ancient links between the Atlantic Forest and Amazonia.
Molecular Phylogenetics Evolution 61:446–459.
Pennington, R.T., Prado, D.E., Pendry, C.A. 2000. Neotropical seasonally dry forests and
Quaternary vegetation changes. Journal of Biogeography 27: 261-273.
Peres, E.A., Silva, M.J., Solferini, V.N. 2017. Phylogeography of the spider Araneus
venatrix (Araneidae) suggests past connections between Amazon and Atlantic
rainforests. Biological Journal of the Linnean Society XX,:1–15.
Pinheiro, F., Cozzolino, S., Draper, D., de Barros, F., Félix, L. P., Fay, M. F.,
and Palma-Silva, C. 2014. Rock outcrop orchids reveal the genetic connectivity
and diversity of inselbergs of northeastern Brazil. BMC Evolutionary
Biology 14: 49.
Por, F.D. (1992) Sooretama: the Atlantic rain forest of Brazil, p. 130. SPB Academic
Publishing, The Hague.
Pôrto, K. C., Cabral, J. J. P., Tabarelli, M. 2004. Brejos de altitude em Pernambuco e
Paraíba: história natural, ecologia e conservação. Ministério do Meio Ambiente:
Recife.
Prance, G.T. 1982. Forest refuges: evidences from woody angiosperms. Pp. 137-158. In:
Biological diversification in the tropics. (Prance, G.T., Ed.) Columbia University
Press: New York.
160
Classified - Internal use
R Developement Core Team. 2013. R: A language and environment for statistical
computing. R Foundation Statistical Computing, Vienna, Austria.
Redford K.H., Fonseca G.A. 1986. The role of gallery forests in the zoogeography of the
Cerrito’s non-volant mammalian fauna. Biotropica 18:126–135.
Rizzini, C.T. 1963. Nota prévia sobre a divisão fitogeográfica do Brasil. Revista
Brasileira de Geografia 1, 1–64.
Rull, V. 2008. Speciation timing and neotropical biodiversity: the Tertiary–Quaternary
debate in the light of molecular phylogenetic evidence. Molecular Ecology 17:
2722–2729.
Santos A.M.M., Cavalcanti D.R., Silva J.M.C., Tabarelli, M. 2007. Biogeographical
relationships among tropical forests in north-eastern Brazil. Journal of
Biogeography 34: 437-446.
Sobral-Souza, T., Lima-Ribeiro, M.S. & Solferini, V.N. (2015) Biogeography of
Neotropical Rainforests: past connections between Amazon and Atlantic Forest
detected by ecological niche modeling. Evolutionary Ecology, 29, 643–655.
Tabarelli, M. 2001. Integridade e ameaças aos brejos da Paraíba e Pernambuco. Pp. 82-
91.In: Plano de Conservação dos Brejos de Paraíba e Pernambuco. Relatório
Técnico do Subprojeto Recuperação e Manejo dos Ecossistemas Naturais de
Brejos de Altitude de Pernambuco e Paraíba. (Tabarelli, M., Ed.) Projeto
PROBIO, Ministério do Meio Ambiente: Recife.
Tabarelli M, Santos AMM. 2004. Uma breve descrição sobre a história natural dos brejos
nordestinos. In: Pôrto KC, Cabral JJP, Tabarelli M. Brejos de altitude em
161
Classified - Internal use
Pernambuco e Paraíba: história natural, ecologia e conservação. Brasília (DF):
Ministério do Meio Ambiente, pp. 17–24.
Thomé, M. T. C., Sequeira, F., Brusquetti, F., Carstens, B. C., Haddad, C. F.
B., Rodrigues, M. T., and Alexandrino, J. 2016. Recurrent connections between
Amazon and Atlantic forests shaped diversity in Caatinga four-eyed frogs. Journal
of Biogeography. 43: 1045–1056.
Töpel M., Zizka A., Calió M.F., Scharn R., Silvestro D., Antonelli A. 2017.
Speciesgeocoder: fast categorization of species occurrences for analyses of
biodiversity, biogeography, ecology, and evolution. Systematic Biology 66: 145-
151.
Turchetto-Zolet A.C., Pinheiro F., Salgueiro F. Palma-Silva C. 2013. Phylogeographical
patterns shed light on evolutionary process in South America. Molecular Ecology
22: 1193–1213
Urrego, D.H., Silman, M.R. & Bush, M.B. 2005. The last glacial maximum: stability and
change in a western Amazonian cloud forest. Journal of Quaternary
Science 20: 693–701.
Vasconcelos-Sobrinho, J. 1971. Os brejos de altitude e as matas serranas. In: J.
Vasconcelos-Sobrinho (ed.). As regiões naturais do Nordeste, o meio e a
civilização. Conselho de Desenvolvimento de Pernambuco, Recife, pp. 79-86.
Vilhena D.A., Antonelli A. 2015. A network approach for identifying and delimiting
biogeographical regions. Nature Communications 6:6848.
162
Classified - Internal use
Wang, X.; Auler, A.S.; Edwards, R.L.; Cheng, H.; Cristalli, P.S.; Smart, P.L.; Richards,
D.A.; Shen, C.C. 2004. Wet periods in northeastern Brazil over the past 210 kr
linked to distant climate anomalies. Nature, 432: 740-743.
Werneck, F. P., Costa, G. C., Colli, G. R., Prado, D. E., & Sites, J. W. Jr. 2011. Revisiting
the historical distribution of seasonally dry tropical forests: New insights based on
palaeodistribution modelling and palynological evidence. Global Ecology and
Biogeography 20: 272–288.
163
Classified - Internal use
ANEXOS
DECLARAÇÃO
Em observância ao §5º do Artigo 1º da Informação CCPG-
UNICAMP/001/15, referente a Bioética e Biossegurança, declaro que o
conteúdo de minha Tese de Doutorado, intitulada “BIOGEOGRAPHY AND
DIVERSITY OF HUMID MOUTAIN FORESTS IN NORTHEASTERN,
BRAZIL”, desenvolvida no Programa de Pós-Graduação em Biologia Vegetal
do Instituto de Biologia da Unicamp, não versa sobre pesquisa envolvendo
seres humanos, animais ou temas afetos a Biossegurança.
Assinatura:
Nome do(a) aluno(a): Ivan Jeferson Sampaio Diogo
Assinatura:
Nome do(a) orientador(a): Flavio Antonio Mäes dos Santos
Data: 30 de novembro de 2017.
COORDENADORIA DE PÓS-GRADUAÇÃO INSTITUTO DE BIOLOGIA Universidade Estadual de Campinas Caixa Postal 6109. 13083-970, Campinas, SP, Brasil Fone (19) 3521-6378. email: [email protected]
164
Classified - Internal use
Declaração
As cópias de artigos de minha autoria ou de minha co-autoria, já publicados ou
submetidos para publicação em revistas científicas ou anais de congressos
sujeitos a arbitragem, que constam da minha Dissertação/Tese de
Mestrado/Doutorado, intitulada BIOGEOGRAPHY AND DIVERSITY OF
HUMID MOUTAIN FORESTS IN NORTHEASTERN, BRAZIL, não infringem
os dispositivos da Lei n.° 9.610/98, nem o direito autoral de qualquer editora.
Campinas, 30 de novembro de 2017
Assinatura:
Nome do(a) aluno(a): Ivan Jeferson Sampaio Diogo
Assinatura:
Nome do(a) orientador(a): Flavio Antonio Mäes dos Santos