Heloisa Dantas Brum

Ecologia e economia de açaí (Euterpe precatoria) em

comunidades ribeirinhas na Amazônia Central

Tese submetida ao programa de Pós-

Graduação em Ecologia da Universidade
Federal do Rio Grande do Norte para a
obtenção do Grau de Doutora em Ecologia.

Orientador: Alexandre Fadigas de Souza

Co-orientadora: Adriana Rosa Carvalho

Natal - RN
2019
 Universidade Federal do Rio Grande do Norte - UFRN
Sistema de Bibliotecas - SISBI
Catalogação de Publicação na Fonte. UFRN - Biblioteca Setorial Prof. Leopoldo Nelson
- Centro de Biociências - CB

Brum, Heloisa Dantas.

Economia e ecologia de açaí (Euterpe precatoria) em
comunidades ribeirinhas na Amazônia Central / Heloisa Dantas
Brum. - Natal, 2019.
156 f.: il.

Tese (Doutorado) - Universidade Federal do Rio Grande do

Norte. Centro de Biociências. Programa de Pós-graduação em
Ecologia.
Orientador: Prof. Dr. Alexandre Fadigas de Souza.
Coorientadora: Profa. Dra. Adriana Rosa Carvalho.

1. Conservação - Tese. 2. Manejo - Tese. 3. Desenvolvimento

sustentável - Tese. I. Souza, Alexandre Fadigas de. II.
Carvalho, Adriana Rosa. III. Universidade Federal do Rio Grande
do Norte. IV. Título.

RN/UF/BSE-CB CDU 502/504

Elaborado por KATIA REJANE DA SILVA - CRB-15/351

 Sumário

AGRADECIMENTOS 5

RESUMO 9

ABSTRACT 11

INTRODUÇÃO GERAL 13

REFERÊNCIAS 20

SOCIO-ECONOMIC AND ENVIRONMENTAL DRIVERS OF VULNERABILITY IN RESOURCE

DEPENDENT COMMUNITIES IN THE AMAZON TROPICAL FOREST 27

ABSTRACT 27
INTRODUCTION 28
METHODS 33
RESULTS 42
DISCUSSION 47
REFERENCES 53

DISENTANGLING THE SUPPLY CHAIN OF AÇAÍ PALM FRUIT IN THE AMAZON AND ITS ROLE ON
LOCAL WELLBEING 64

ABSTRACT 64
INTRODUCTION 65
METHODS 68
RESULTS 74
DISCUSSION 86
REFERENCES 92

FLOODING DISTURBANCE AND SHADE STRESS SHAPE POPULATION STRUCTURE OF AÇAÍ

PALM, THE MOST ABUNDANT AMAZON SPECIES 101

ABSTRACT 102
INTRODUCTION 103
METHODS 108
RESULTS 113
DISCUSSION 120
ACKNOWLEDGEMENT 126
REFERENCES 127

CONCLUSÃO GERAL 141

REFERÊNCIAS 144

ANEXOS 146

ROTEIRO DE ENTREVISTAS 146

 AGRADECIMENTOS

Agradeço, primeiramente, à Fundação de Amparo à Pesquisa do Estado do Amazonas

(FAPEAM) pela bolsa de doutorado. O presente trabalho foi realizado com apoio da

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brasil (CAPES) - Código

de Financiamento 001. Ao
o Departamento de Educação e Treinamento do Governo

Australiano pela bolsa durante o período sanduiche. As atividades de campo deste

estudo foram possíveis com o financiamento da Rufford Small Grant Foundation (RSGF)

e da Conservation, Food and Health Fou

Foundation (CFH), e com o apoio imprescindível da

Associação Amigos do Peixe

Peixe-Boi (AMPA) e do Instituto Piagaçu (IPi).
(IPi) Agradeço ao

Departamento de Mudanças Climáticas e Unidades de Conservação (DEMUC/SDS) que

forneceu a autorização para o desenvolvimento das aatividades

tividades na Unidade
U de

Conservação,, e ao Departamento de Ecologia da Universidade Federal do Rio Grande do

Norte (UFRN) pelo apoio institucional e financeiro.

Ao Alexandre, pela orientação impecável, pela paciência, pelo bom humor e pelo

conhecimento infinito.
nito. Por ser essa pessoa incrível e admirável, que trabalha

diariamente para que o mundo melhore. Foram anos de aprendizado contigo, e espero

que essa parceria se estenda por um longo tempo, pois ainda há muito pra aprender

com você!

A Adriana, que complementou

mentou esse trabalho na medida exata e que foi responsável por

ampliar nossos horizontes para além da ecologia

ecologia.. Obrigada pelo apoio e incentivo, pelas

conversas e amizade.. À Nadine e Ingrid, pela paciência e amizade durante os meses

incríveis na Austrália. Espero revê

revê-las em breve!

5
Este trabalho é um dos resultados dos esforços de centenas de pessoas que, há anos,

vêm desenvolvendo e aprimorando o conceito de sustentabilidade na Amazônia, e

construído um modelo de Unidades de Conservação que seja justo para bichos, plantas,

paisagens e para os povos da floresta. A coleta de dados deste estudo foi feita em

algumas dezenas de dias, mas só foi possível após algumas dezenas de anos de atuação

de diversas instituições na região, como o Instituto Piagaçu, IDSM, WCS,INPA,

WC UFAM,

AMPA, dentre outras ONGs e Instituições de pesquisa e conservação da Amazônia.

Juntamente com as populações tradicionais,, são os grandes responsáveis por ainda

termos a maior floresta tropical do planeta em pé! Um agradecimento à essas

instituiçõesque
uiçõesque sempre serviram de exemplo e inspiração. E como instituições são feitas

por pessoas, um agradecimento a cada um que se dedica a essa causa.

Agradeço às pessoas incríveis que sempre colaboram com as pesquisas desenvolvidas

na RDS Piagaçu-Purus, em
m especial, Assis e Socorro (e Felipe e Pedrinho) por tomarem

conta da gente e de tudo em todos os momentos! Aos meninos que fazem com que a

coleta de dados aconteça, sem vocês nada disso existiria: Mário, Lucas, Queven,

Marquinhos, a família Moreira intei

inteira
ra que está sempre do nosso lado. Um especial

agradecimento também aos “amigos da cidade” Marcos, Isa e Débora, que também

foram importantíssimos nessa fase de campo. Aos moradores da RDS Piagaçu Purus,

pelos ensinamentos, pelas histórias vividas e contada

contadas,
s, pela receptividade e confiança,

serei eternamente grata!

Um agradecimento especial à minha querida mãe, pelo apoio incondicional

incondicional, sempre e

para tudo o que precisasse

sse,, mesmo quando eu achava que não... À minha família

6
maravilhosa, os de sangue e os de co
coração,
ração, por não me obrigar a abandonar os grupos

de WhatsApp antes, durante e depois das eleições!

Aos amigos que são a razão de tudo na vida, afinal. Eli, por me ensinar que na vida as

vezes a gente só tem que dizer “sim, é isso aí!”, pela façanha de conseguir fazer o açaí

ser mais bonito num desenho do que ao vivo! Por estar ao meu lado (dia e noite,

inclusive, inseparável!) em uma das viagens mais incríveis da minha vida, identifica
identificando a

fauna australiana como ninguém! À Eugenia pela criatividade inf

infinita
inita (e desenhos

também lindos!), pelo carinho (mais com a Preta, eu sei!) e pela sinceridade em todos

os momentos! Por me ensinar que francês, português e inglês podem ser falados numa

mesma frase, e que – o mais importante de tudo, e o que sempre fica – por me ensinar

que “passa! tudo passa...”. A Julia (que vem me perseguindo desde a Amazônia,

Piagaçu-Purus,
Purus, Natal e Austrália, tudo bem, eu também te amo!), que tem me ensinado

que ser legal comigo é muito mais legal para todos e que apesar de nenhum talento
tale

artístico, é uma inspiração para o planejamento de viagens! Gracias, queridas!

À Isa, um agradecimento especial, repleto de orgulho de te ver na Amazônia e no Purus!

Construindo tua caminhada agora em solo amazônico, e sendo essa pessoa doce e

querida, sempre,, e cada vez mais forte!

Aos amigos da Amazônia, que assim como essa floresta incrível, nunca sairão do meu

coração: Zé Luiz,Isolde,
Isolde, Bráulio, Mariana, Carlota, Flavinha, Flávia Baiana, Manô, Ana,

Rosinha, Leo, Pardal, Alemão, Diogo

Diogo, Minhoca, Dani, Marininha.. Aos amigos amazônicos

que foram, como eu, viver perto do mar, Dri

Dri, Guiga, Carol, Bernardo, Carolzinha,
Carol Helder,

Jotinha e Álvaro. Ao Duka, pelos anos de parceria, pelas risadas, pelo carinho, pelo

7
apoio e pela confiança de
desempre! Aos meus irmãos de coração, Ximba e Jone, que

mesmo quando vocês somem, meu amor por vocês continua, de forma incondicional
incondicional!

Aos melhores vizinhos do planeta, Má, Dadão, Chico e Analu, pelo carinho, pela

amizade, pelas risadas, pelos cuidados infinitos, e por conseguir nos tranquilizar até

quando alguém subia no telhado... E também aos bichinhos queridos, que deixam

qualquer problema da vida pequeno, Lampião, Preta, Xuja, Bonsai

Bonsai-de-Tigre, Foca,

Batman e Xujinha.

Ao Juanca, amigo querido, um especial agradecimento por me salvar em todas as crises

dos computadores e por ser meu fotógrafo preferido de Amazônia (Isa, você é a

fotógrafa preferida, viu?). Aos tantos outros amigos queridos, Nati, Paty, Rodrigo, Pichi,

Rafa, Miriam, Carlinhos, Gis, Liana, CB, Paulo, Marília, Dard

Dardo, Dani, Pocas, Nico, Fer,

Álvaro, Ana Helena, Laurinha, Caramujo, Carol, May, Bianca, André e Bruno. Aos

professores e alunos do Decol, que além da ciência de altíssima qualidade (balbúrdia

que chama, né?), trabalham para que o departamento tenha esse cli
clima tranquilo e

colaborativo.

Aos amigos queridos que Natal me deu, Mica, Berg, Caio, Lu

Luiza,
a, Igor, Gabis (as duas) e

Gabriel! Ao Beto, por me ouvir e me entender mais do que eu mesma, e por deixar isso

lindamente registrado pra sempre em mim.

Agradeço às mudanças e reviravoltasd

reviravoltasda vida que, como ondas gigantes, com maresia e

a brisa do mar dessa cidade do Sol

Sol,assustam um pouco, masvêm
vêm cheias de vida!
vid

8
 RESUMO

A interdependência histórica entre populações humanas e o uso tradicional da

biodiversidade e ecossistemas moldaram a Amazônia como um dos mais ricos e
complexos sistemas socioecológicos do mundo. A Floresta Amazônica ainda enfrenta
uma série de ameaças com o desmatamento para pecuária e produção de soja,
abertura de estradas, além dos inúmeros imp
impactos
actos em curto e longo prazo das
construções de barragens e sobre
sobre-exploração
exploração de recursos naturais. Todas essas
ameaças têm um impacto global, já que muitos estudos demonstraram o papel chave
da maior floresta tropical do planeta na regulação do clima. As U
Unidades de
Conservação são uma das mais eficientes estratégias para proteger a biodiversidade,
mas a simples criação não garante sua efetividade. A conservação depende da
integração entre os aspectos ecológicos e sociais das pessoas e instituições envolvid
envolvidas
com o uso e proteção da biodiversidade. Produtos florestais não
não-madeireiros
madeireiros são um
componente cultural importante de populações tradicionais e podem representar uma
fonte significativa de renda. Especificamente a palmeira arbórea açaí ((Euterpe
precatoria,, Arecaceae) é uma espécie de planta icônica da Amazônia com um enorme
valor cultural desde a Amazônia Central e Ocidental até a América Central.
Recentemente tem ganhado atenção nos mercados nacional e internacional por conta
de seu conteúdo enérgico e n
nutricional
utricional e suas propriedades antioxidantes. Este estudo
foi desenvolvido com o intuito de fornecer informações sobre os aspectos sociais e
ecológicos sobre uso e comercialização de recursos naturais por comunidades
ribeirinhas na Amazônia Central, e espe
especificamente sobre o açai (Euterpe
Euterpe precatoria)
precatoria em
diferentes habitats.No
No primeiro capítulo, através de análises de aspectos relacionados à
vulnerabilidade socioeconômica em comunidades ribeirinhas, nós identificamos
identifica dois
grupos distintos de pessoas, um em situação de maior vulnerabilidade social que outro.
outro
A vulnerabilidade é desencadeada por características sócio demográficas,
demográficas mas também
por aspectos econômicos como o número de atividades extrativistas desenvolvidas e a
renda proveniente de atividades espe
específicas, como o extrativismo da castanha
astanha-do-Brasil
e do açaí. O segundo
egundo capítulo descreveu a cadeia produtiva do açaí, avaliando as

9
conexões e elementos-chave
chave através de diferentes métricas, com o intuito de identificar
possíveis gargalos sociais e econômi
econômicos
cos para o extrativismo e comercialização do açaí.
No terceiro capítulo nós descrevemos os estágios de vida de E. precatoria e observamos
que a relação entre altura e diâmetro mudou ao longo do desenvolvimento
ontogenético e houve uma marcante dependência de habitat tanto na densidade
quanto na distribuição de tamanhos da população. As populações da terra firme são
dominadas por juvenis, en
enquanto que as populações das várzeas são dominadas por
indivíduos reprodutivos.
s. A proximidade com assentamentos humanos não foi
relacionada com os parâmetros associados com estrutura populacional. Nós
pretendemos que este estudo seja um primeiro passo para fornecer conhecimento
científico e apoio para a organização social aos moradores locais para uma maior
participação ao longo da cadeia produtiva do açaí e nas decisões de manejo, buscando
alcançar os objetivos das Reservas de Desenvolvimento Sustentável.

Palavras-chave:
chave: Conservação, Manejo, Desenvolvimento Sustentável, Amazônia,
Produtos florestais não-madeireiros.
madeireiros.

10
 ABSTRACT
The historical interdependence between human populations and the traditional use of
biodiversity and ecosystems shaped the Amazon as one of the most rich and complex
social-ecological
ecological systems. Amazon forest still facing a huge set of threats through
deforestation to cattle ranching, soy production and road paving, besides the several
damages
ges on short and long
long-term of dams’ construction and over-exploration
exploration of natural
resources. All these threats have a global impact, as innumerous studies already show
the key role of the largest world tropical forest for climate regulating. Protected areas
are one of the most efficient way to protect biodiversity but the simple creation is not a
guarantee of their effectiveness. Conservation is dependent on the integration of
ecological and social aspects of stakeholders responsible for use and protection o
of
biodiversity. Non-timber
timber forest resources are an important cultural component of
traditional people and can represent a significative source of household income.
Specifically, the palm tree açai ((Euterpe precatoria,, Arecaceae) is the most iconic
Amazon plant
lant species with a huge cultural value in Central and Westerns Amazon till
Central America. Recently, it is gaining attention of national and international markets
because of its nutritious and energetic contents and antioxidant properties. This study
was developed to provide information on social and ecological aspects of the use and
commercialization of natural resources in riverine communities in Central Amazon, and
specifically on açaí palm tree ((Euterpe precatoria)) in different habitats. In the first
chapter, through the analysis of aspects of social and economic vulnerability in riverine
communities, wee were able to identified two distinct groups of respondents
respondents, one more
disadvantaged and socially vulnerable than the other. Vulnerability
ulnerability is triggered by socio-
socio
demographic features and also by economic aspects such as the number of extractive
activities developed and the income generated from some specific activities, like Brazil-
nut and Açai harvesting. The second chapter descr
described
ibed the açaí supply chain, evaluating
connections and key elements through different metrics, in order to identify possible
social and economic bottlenecks for açai harvesting and commercialization. IIn the third

11
chapter, we described life stages of
ofE. precatoria and evaluated that the height:diameter
relationship changed along the ontogenetic development and tthere
here was a marked
habitat dependency in both the density and population size distribution, with
populations in upland forests dominated by juveniles w
while
hile populations in the
floodplains were dominated by reproductive palms. Nearness of human settlements
was not related to population structure parameters.
parameters.We
We intend this study to be the first
step to provide ecological knowledge and support for social orga
organization
nization to assist locals
in a larger participation along the supply chain and management decisions, aiming to
achieve the goals of the Sustainable Development Reserve.

Key-words:
words: Conservation, Management, Sustainable development, Amazon, Non
Non-timber
forest resources.

12
 INTRODUÇÃO GERAL

A Bacia Amazônica compreende a maior floresta tropical remanescente do planeta,

abrigando a maior biodiversidade de espécies e fitofisionomias (Hopkins 2007).

2007) A

Floresta Amazônica abriga uma vasta diversidade cultural, com diferentes idiomas

indígenas e grupos étnicos coexist

coexistindo com a natureza (Lima and Pozzobon
Poz 2006;

Cámara-Leret
Leret et al. 2019; IBGE 2019a)
2019a).. A interdependência histórica entre populações

humanas e o uso tradicional da biodiversidade e ecossistemas (Levis et al. 2012)

moldaram a Amazônia como uma dos mais ricos e complexos sistemas socioecológicos

do mundo (Berkes and Folke 1998)

1998).. Ao mesmo tempo, a Floresta Amazônica está

enfrentando uma série de ameaças com o desmatamento para pecuária e produção de

soja (Nepstad et al. 2006b) e abertura de estradas (Nepstad et al. 2001),

2001) além dos

inúmeros impactos em curto e longo prazo das construções de barragens (Benchimol

and Peres 2015; Lees et al. 2016; Fearnside 2019; Gerlak et al. 2019)
2019),, caça (Peres et al.

2016) e exploração madeireira (Partial et al. 1999; Nepstad et al. 2004a; Darrigo et al.

2016).. Todas essas ameaças têm um impacto global, já que muitos estudos

demonstraram o papel chave da maior floresta tropical do planeta na regulação do

clima (Carvalho et al. 2004; Fearnside 2005; Malhi et al. 2008; Nepstad et al. 2008; Lees

et al. 2016).. Diferentes estratégias para protege

protegerr a biodiversidade e manter os modos

de vida tradicionais das comunidades locais estão sendo desempenhadas, umas melhor

sucedidas que outras.

13
Uma das formas mais eficientes para proteger a biodiversidade e o modo de vida

tradicional é a criação de Unidade

Unidades de Conservação (UC) (Leverington et al. 2010).
2010)

Mesmo assim, há um padrão global de que um considerável número de UCs são criadas

em áreas residuais, áreas marginais com nenhum iinteresse

nteresse econômico e,

consequentemente, não planejadas estrategicamente para proteger a biodiversidade

local (Venter et al. 2017; Vieira et al. 2019)

2019).. Além disso, essa iniciativa geralmente

carece de recursos financeiros,

anceiros, infraestrutura e pessoal (Watson et al. 2014; Campos-
Campos

Silva et al. 2015),, mesmo se os benefícios de preservar os serviços ecológicos e a

biodiversidade sejam enormes (Costanza et al. 1997).. Unidades de Conservação de Uso

Sustentável, em que o uso comercial e para subsistência de recursos naturais é

permitido (equivalente às categorias IV a VI da IUCN; Dudley 2013),, são uma abordagem

promissora para efetivamente promover o uso e a conservação da biodiversidade,

juntamente com governança e engajamento das partes locais envolvidas (Nepstad and

Mcgrath 2002).

As Unidades de Conservação de Uso Sustentável foram inicialmente idealizadas ccomo

uma forma de conciliar o modo de vida tradicional com a conservação da

biodiversidade (Brasil 2000)

2000).. Na década de 1990 o conceito de desenvolvimento

sustentável estava sendo construído de forma multifacetada pelo mundo. No Brasil,

esse conceito teve o apoio de duas pessoas com papel chave na história da biologia da

conservação. Chico Mendes, um soldado da borracha, assassinado em 1988 por

fazendeiros locais, que expôs para o mundo a ideia de proteger a floresta através o uso

tradicional dos seus recursos; e Márcio Ayres, um primatólogo e o principal responsável

pela criação da primeira Reserva de Desenvolvimento Sustentável (RDS) – RDS

14
Mamirauá – em 1996, um marco histórico e orientador para as futuras ações de

conservação da Amazônia Brasileira (Peralta 2012).. Ao mesmo tempo, o movimento

ambientalista estava espalhando seu conceito de desenvolvimento sustentável através

da Cúpula da Terra (Rio 92) e da Convenção da Diversidade Biológica em 1992, que

levou a diversos outros protocolos, convenções e um comprometimento global com a

proteção do planeta. Em 2000 o Governo Federal criou o Sistema Nacional de Unidades

de Conservação (SNUC; Brasil 2000) que, juntamente com o Programa Áreas Protegidas

da Amazônia (ARPA), foram responsáveis pelo fortalecimento e proteção de cerca de 60

milhões de hectares em 117 Unidades de Conservação na Amazônia (MMA 2019).

2019) Uma

das UCs criadas durante esse período extremamente favorável foi a RDS Piagaçu Purus

(RDS-PP),
PP), na região do Baixo Rio Purus, em uma área responsável por fornecer mais de

30% do pescado que abastecia a capital Manaus (Batista and Petrere 2015) e que vinha

sofrendo intensa exploração de recursos naturais há décadas (Loureiro

iro 1986) e sem o

respeito aos direitos sociais das populações locais (Martins 2011).. Infelizmente, a

simples criação de uma RDS não é garantia de proteção da natureza e bem

bem-estar da

população local.

O engajamento da sociedade é crucial para a conservação efetiva (Peralta 2012),

especialmente se é possível combinar a ação conjunta de múltiplas instituições (Berkes

2009) de forma participativa (Pimm et al. 2001; Nepstad et al. 2006a; Soares-Filho
Soares et al.

2006; Peralta 2012),, através de iniciativas promissoras como o mane

manejo de base

comunitária (Naidoo and Adamowicz 2005; Naidoo et al. 2011; Cinner et al. 2012b;

Campos-Silva
Silva and Peres 2016)
2016), o manejo adaptativo (Holling 1978) ou ainda o co-

15
manejo adaptativo (Berkes 2009)
2009).. Esses conceitos são extremamente interdependentes

e compreendem a ideia do “a
“aprender-fazendo”,
fazendo”, vinda do manejo adaptativo com a

colaboração intrínseca do conceito de co

co-manejo (Folke et al. 2002).
2002) Na Amazônia

Brasileira o manejo do pirarucu ((Arapaima gigas,, que está entre os maiores peixes de

água doce do mundo) é um dos melhores exemplos de iniciativas bem sucedidas de co

co-

manejo adaptativo (Campos

(Campos-Silva
Silva and Peres 2016; Petersen et al. 2016),
2016) com inúmeros

efeitos positivos em outros grupos de espécies e ecossistema (Campos-Silva

(Campos et al.

2018a) e benefícios sociais (Castello et al. 2009).. Um dos gargalos desse tipo de

programa, muitas vezes, é o retorno econômico para as populações locais, que é o fator

limitante para manter o engajamento e comprometimento dos envolvidos. Juntamente

com dados ecológicos e com o conhecimento tradicional para embasar as dec

decisões de

manejo (Castello et al. 201

2010),, a estruturação de cadeias de produtos da

sociobiodiversidade pode ser crucial para engajar as pessoas e fornecer as ferramentas

necessárias para conciliar a proteção da biodiversidade, com o retorno econômico e o

bem-estar
estar das populações locais.

Produtos florestais não-madeireiros

madeireiros são um componente cultural importante de

populações tradicionais (Ticktin 2004) e podem representar uma fonte significativa de

renda (Walter 2001).. As palmeiras são os componentes mais importantes de vários

habitats amazônicos e estão fortemente conectadas com o arcabouço cultural de

populações rurais e urbanas (Goulding and Smith 2007).. A palmeira arbórea açaí

(Euterpe precatoria,, Arecaceae) é uma espécie de planta icônica da Amazônia com um

enorme valor cultural desde a Amazônia Central e Ocidental até a América Central

16
(Goulding and Smith 2007)
2007).. Recentemente tem ganhado atenção nos mercados

nacional e internacional por conta de seu conteúdo en

energético
ico e nutricional (Yuyama et

al. 2011) e suaspropriedades antioxidants (Pacheco-Palencia et al. 2009).

2009) O açaí é um

dos mais importantes produtos florestais do Brasil (Nobre 2017).. Dados de 2017

indicam que 93,1% da produção brasileira de açaí é fornecida pela espécie E. oleraceae

(“açaí-do-Pará”),
Pará”), restrita à Amazônia Oriental (IBGE 2018),, mas o suprimento de açai

pela espécie E. precatoria vem crescendo exponencialmente ano a ano (IBGE 2018).

Curiosamente até 2018 a espécie não era considerada uma fonte de polpa de açaí pela

legislação brasileira,
ileira, que reconhecia a polpa de açaí apenas a extraída de E.

oleraceae(BRASIL
(BRASIL 2000; Brasil 2018).. O Brasil produz cerca de 220.000 toneladas de açaí

anualmente (cerca de 50.000 toneladas de E. precatoria)) e a atividade gera uma renda

de pouco mais de US$ 150 milhões/ano (IBGE 2018).. A grande maioria da produção de

açaí do Estado do Amazonas é coletada em áreas de floresta nativa (AMAZONAS 2005)

e os moradores locais relataram que o fruto destas áreas tem maior qualidade em

relação ao fruto de áreas de plantio de açaí. De fato, o uso e a comercialização do açaí

podem incentivar a preservação da floresta nativa e se tornar um símbolo do uso

sustentável
tentável da floresta e, como sugerido por M. Goulding e N. Smith, ser uma sentinela

para a conservação da Floresta Amazônica (Goulding and Smith 2007).

Este estudo foi desenvolvido com o intuito de fornecer informações sobre os aspectos

sociais e ecológicos sobre uso e comercialização de recursos naturais por comunidades

ribeirinhas na Amazônia
zônia Central, e especificamente sobre o açai ((Euterpe
Euterpe precatoria)
precatoria em

diferentes habitats. Nós pretendemos que este estudo seja um primeiro passo para

17
fornecer conhecimento científico e apoio para a organização social aos moradores

locais para uma maior participação

rticipação ao longo da cadeia produtiva do açaí e nas decisões

de manejo, buscando alcançar os objetivos das Reservas de Desenvolvimento

Sustentável.

No primeiro capítulo, “Fatores socioeconômicos e ambientais da vulnerabilidade em

comunidades dependentes de recursos na Floresta Amazônica” (

(Socio-

economicandenvironmental drivers ofvulnerability in resourcedependentcommunities in

theAmazon Forest)) nós descrevemos aspectos sociais da população local associados

com a vulnerabilidade social e a sensibilidade à mudanças, identificando fatores que

podem impedir o desenvolvimento de atividades econômicas lucrativas e relacionadas

com o uso racional e sustentável da sociobiodiversidade da região. Neste capítulo

abordamos um conceito extremamente amplo, a vulnerabil

vulnerabilidade socioeconômica
oeconômica, que

pode ser abordada de acordo com diferentes métricas, e em diferentes dimensões

(ecológica e/ou socioecológica), de acordo com Marshall et al. (2013), de forma

ilustrada no modelo conceitual (Figura 1). Para esse estudo, optamos po

por considerar a

vulnerabilidade socioeconômica associada a alguns aspectos principais, como a renda

doméstica e a dependência de recursos naturais

naturais,, que se mostraram como métricas

consistentes com descrição da realidade local estudada

estudada,, e como fatores fundamentais
fundame

para compreender a capacidade adaptativa e a sensibilidade ààs mudanças.

No segundo capítulo “Compreendendo a cadeia produtiva do açai na Amazônia e seu

papel no bem-estar
estar local” ((Disentanglingthesupplychainof
Disentanglingthesupplychainof açaí palmfruit in

theAmazonand its role on local wellbeing

wellbeing)) nós descrevemos os elementos e as conexões

da cadeia produtiva do açaí e analisamos a conectividade e os elementos

elementos-chave a partir

18
de métricas quantitativas, avaliando de que forma os aspectos econômicos da atividade

com açaí podem favorecer

ecer a segurança alimentar, o retorno econômico juntamente

com a promoção do bem-estar

estar para as comunidades locais.

Figura 1.. Modelo conceitual identificando os aspectos associados com a vulnerabilidade

socioeconômica em suas diferentes dimensões (ecológ
(ecológica
ica e socioeconômica). Extraído
de Marshall et al. (2013).

Por fim, no terceiro capítulo “Distúrbio por inundação e estresse pela sombra moldam a

estrutura populacional do açaí, a mais abundante espécie da Amazônia”

(Floodingdisturbanceandshade
Floodingdisturbanceandshade stress shapepopulationstructureof açaí palm,

themostabundantAmazonspecies
themostabundantAmazonspecies)) nós analisamos aspectos ecológicos da estrutura

populacional de E. precatoria em habitats distintos (várzea e terra firme), o que

19
forneceu informações sobre os fatores ambientais responsáv
responsáveis
eis pela distribuição da

espécie mais abundante da Amazônia (ter Steege et al. 2013) indicando uma estratégia

ecológica de E. precatoria como uma especialista intermediária de pequenas clareiras

(Denslow 1980).

Referências

AMAZONAS 2005. Secretaria de Desenvolvimento Sustentável do Estado do Amazonas

(SDS): Cadeia produtiva do açaí no estado do Amazonas. MENEZES, M.

Batista, V. da S. and Petrere, M. 2015. Characterization of the commercial fish

production landed at Manaus, A

Amazonas State, Brazil. - Acta Amaz. 33: 53–66.
53

Berkes, F. 2009. Evolution of co

co-management:
management: role of knowledge generation, bridging

organizations and social learning. - J. Environ. Manage. 90: 1692–702.

702.

Berkes, F. and Folke, C. 1998. Linking social ecologic

ecological
al systems: Management Practices

and Social Mechanisms for Building Resilience.

Brasil 2000. Lei Federal n 9.985 de 18/07/2000. - SNUC - Sist. Nac. unidades Conserv. in

press.

Brasil 2018. Instrução Normativa no 49, de 26 de setembro de 2018. in press.

SIL 2000. INSTRUÇÃO NORMATIVA No 01, DE 7 DE JANEIRO DE 2000.

BRASIL

Cámara-Leret,
Leret, R. et al. 2019. Indigenous knowledge networks in the face of global

change. - Proc. Natl. Acad. Sci. 116: 9913

9913–9918.

Campos-silva,
silva, J. V. and Peres, C. A. 2019. Brazil ’ s policies stuck in the mud. - Science

(80-. ). 363: 1046.

20
Campos-Silva,
Silva, J. V. and Peres, C. A. 2016. Community
Community-based
based management induces rapid

recovery of a high-value
value tropical freshwater fishery. - Sci. Rep. 6: 1––13.

Campos-Silva,
Silva, J. V. et al. 2015. Policy reversals do not bode well for conservation in

Brazilian Amazonia. - Nat. e Conserv. 13: 193

193–195.

Campos-Silva,
Silva, J. V. et al. 2018. Unintended multispecies co
co-benefits
benefits of an Amazonian

community-based
based conservation programme. - Nat. Sustain. 1: 650––656.

Carvalho, G. ett al. 2004. AN AMAZON PERSPECTIVE ON THE FOREST – CLIMATE

CONNECTION : : OPPORTUNITY FOR CLIMATE MITIGATION , CONSERVATION AND

DEVELOPMENT? - Environ. Dev. Sustain. 6: 163

163–174.

Castello, L. et al. 2009. Lessons from integrating fishers of arapaima in small

small-scale

fisheries management at the mamirauá reserve, amazon. - Environ. Manage. 43:

197–209.

Castello, L. et al. 2010. Participatory Conservation and Local Knowledge in the Amazon

Várzea: The Pirarucu Management Scheme in Mamirauá. - In: The Amazon Várzea:

The
he decade past and the decade ahead. pp. 261
261–275.

Cinner, J. E. et al. 2012. Comanagement of coral reef social

social-ecological
ecological systems. - Proc.

Natl. Acad. Sci. 109: 5219

5219–5222.

Costanza, R. et al. 1997. The value of the world ’ s ecosystem services and natural

capital. - Nature in press.

Darrigo, M. R. et al. 2016. Effects of reduced impact logging on the forest regeneration

in the central Amazonia. - For. Ecol. Manage. 360: 52–59.

Denslow, J. 1980. Gap Partitioning among Tropical Rainforest Trees. - Biotropica 12: 47–

55.

21
Dudley, N. (Ed) 2013. Guidelines for Applying Protected Area Management Categories

(IUCN, Ed.).

Fearnside, P. M. 2005. Deforestation in Brazilian Amazonia: History, Rates, and

Consequences. - Conserv. Biol. 19: 680–688.

Fearnside, P. 2019. Represas

sas hidroeléctricas en la Amazonia brasileña: impactos

ambientales y sociales. - Rev. Estud. Bras. 6: 123–138.

Folke, C. et al. 2002. Resilience and sustainable development: building adaptive capacity

in a world of transformations. - Ambio 31: 437–440.

Gerlak,
lak, A. K. et al. 2019. Dams, Chinese investments, and EIAs: A race to the bottom in

South America? - Ambio in press.

Goulding, M. and Smith, N. 2007. Palmeiras: sentinelas para a conservação da

Amazônia. - Amazon Conservation Association (ACA) / Sociedade Civil Mamirauá.

Holling, C. S. 1978. Adaptive Environmental Assessment and Management (Pitman

Press, Ed.).

Hopkins, M. J. G. 2007. Modelling the known and unknown plant biodiversity of the

Amazon Basin. - J. Biogeogr. 34: 1400–1411.

IBGE 2018. Produção da Extração Vegetal e da Silvicultura – PEVS. Acesso em Oct/2018,

in: www.ibge.gov.br.

IBGE 2019. Censo demografico. - Inst. Bras. Geogr. e Estatística

Lees, A. C. et al. 2016. Hydropower and the future of Amazonian biodiversity. -

Biodivers. Conserv. 25: 45

451–466.

Leverington, F. et al. 2010. A global analysis of protected area management

effectiveness. - Environ. Manage. 46: 685

685–698.

22
Levis, C. et al. 2012. Historical Human Footprint on Modern Tree Species Composition in

the Purus-Madeira
Madeira Interfluve, Central Amazonia. - PLoS One in press.

Lima, D. and Pozzobon, J. 2006. Amazônia socioambiental: sustentabilidade ecológica e

diversidade social. - Estud. Avançados 19: 45

45–76.

Loureiro, A. 1986. A grande crise (1908

(1908-1912).

Malhi, Y. et al. 2008. Climate change, def

deforestation,
orestation, and the fate of the Amazon. -

Science (80-.. ). 319: 169

169–172.

Marshall, N. A. et al. 2013. Social Vulnerability of Marine Resource Users to Extreme

Weather Events. - Ecosystems 16: 797

797–809.

Martins, R. D. A. 2011. Fair trade practices in the Northwest Brazilian Amazon. - BAR -

Brazilian Adm. Rev. 8: 412

412–432.

MMA 2019. http://arpa.mma.gov.br.

Naidoo, R. and Adamowicz, W. L. 2005. Biodiversity and nature

nature-based
based tourism at forest

reserves in Uganda. - Environ. Dev. Econ. 10: 159

159–178.

Naidoo, R. et al. 2011. Namibia’s community

community-based
based natural resource management

programme: An unrecognized payments for ecosystem services scheme. - Environ.

Conserv. 38: 445–453.

453.

Nepstad, D. C. and Mcgrath, D. G. 2002. Frontier Governance in Amazonia. - Science

(80-. ). in press.

Nepstad, D. et al. 2001. Road paving, fire regime feedbacks, and the future of Amazon

forests. - For. Ecol. Manage. 154: 395

395–407.

Nepstad, D. et al. 2004. Managing the amazon timber industry. - Conserv. Biol. 18: 575–
575

577.

23
Nepstad, D. C. et al. 2006a.
006a. Globalization of the Amazon Soy and Beef Industries
Industries :

Opportunities for Conservation. - Conserv. Biol. 20: 1595–1603.

Nepstad, D. et al. 2006b. Inhibition of Amazon deforestation and fire by parks and

indigenous lands. - Conserv. Biol. 20: 65

65–73.

Nepstad,
stad, D. C. et al. 2008. Interactions among Amazon land use, forests and climate:

prospects for a near-term

term forest tipping point. - Philos. Trans. R. Soc. Lond. B. Biol.

Sci. 363: 1737–1746.

Nobre, C. A. 2017. Brazil ’ s Fruitful Example: Açaí. - World Res.. Inst.: 2017–2020.
2017

Pacheco-Palencia,
Palencia, L. A. et al. 2009. Phytochemical composition and thermal stability of

two commercial açai species, Euterpe oleracea and Euterpe precatoria. - Food

Chem. 115: 1199–1205.

1205.

Partial, K. et al. 1999. Large

Large-scale impoverishment
nt of Amazonian forests by logging and

fire. - Nature 1405: 1997

1997–2000.

Peralta, N. 2012. “Toda Ação De Conservação Precisa Ser Aceita Pela Sociedade”

Manejo Participativo Em Reserva De Desenvolvimento Sustentável.

Peres, C. A. et al. 2016. Dispersal limita

limitation induces long-term
term biomass collapse in

overhunted Amazonian forests. - PNAS in press.

Petersen, T. A. et al. 2016. Recovery of Arapaima sp. populations by community

community-based

management in floodplains of the Purus River, Amazon. - J. Fish Biol. 89: 241–248.
241

Pimm, S. L. et al. 2001. Can We Defy Nature ’ s End

End ? - Science (80-.. ). 293: 2207–2208.
2207

Soares-Filho,
Filho, B. S. et al. 2006. Modelling conservation in the Amazon basin. - Nature

440: 520–523.

ter Steege, H. et al. 2013. Hyperdominance in the Amazonian tree flora. - Science 342:

24
 1243092.

Ticktin, T. 2004. The ecological implications of harvesting non

non-timber
timber forest products. -

J. Appl. Ecol. 41: 11–21.

21.

Venter, O. et al. 2017. Bias in protected - area location and its effects on long - term

aspirations of biodiversity
versity conventions. - Conserv. Biol. 32: 127–134.
134.

Vieira, R. R. S. et al. 2019. The residual nature of protected areas in Brazil. - Biol.

Conserv. 233: 152–161.

161.

Walter, S. 2001. Non-Wood

Wood Forest Products in Africa: A regional and national Overview.

- Work. Pap. Trav. FOPW/01/1 Food Agric. Organ.

Watson, J. E. M. et al. 2014. The performance and potential of protected areas. - Nature

515: 67–73.

Yuyama, L. K. O. et al. 2011. Caracterização físico

físico-química
química do suco de açaí de Euterpe

precatoria Mart. oriundo de diferentes ecossistemas amazônicos. - Acta Amaz. 41:

545–552.

25
1
26
 2 Socio-economic and environmental drivers of vulnerability in
3 resource dependent communities in the Amazon Tropical Forest
4

5 Heloisa D. Brum; Alexandre F. Souza; Adriana R. Carvalho; Ingrid van Putten; Nadine

6 Marshall.

7 Abstract
8 Social-ecological systems represent an important concept for managing the use and
9 conservation of natural resources. Several social factors can act in different directions
10 improving or decreasing natural resource dependency and sensitivity to change.
11 Traditional communities in rural areas usually have a strong occupational identity, a
12 high connection with the local environment and a high social and economic
13 dependency on natural resources. Assessing dependency on natural resources and
14 evaluating how people can be sensitive to change can help to provide deeper insights
15 into the nature of social vulnerability, which is important for adaptation planning. The
16 goals of this study are to 1) identifying groups of people with different levels of social-
17 economic vulnerability; 2) identifying socio-demographic, economic, geographic and
18 environmental factors that affect income sources and levels for local residents; 3)
19 determine the most important source of income for local residents and 4) determine
20 the relationship between occupational identity and the major source of income. The
21 study was carried out in the Piagaçu-Purus Sustainable Development Reserve, in Central
22 Amazon, Brazil through semi-structured interviews. We were able to identified two
23 distinct groups of respondents with regards to socio-demographic and economic
24 vulnerability, one more disadvantaged and socially vulnerable than the other. Also, we
25 uncovered that the vulnerability is triggered by economic aspects such as the number
26 of extractive activities developed and the income generated from some specific
27 activities, but also by socio-demographic features such as family size and schooling. The
28 extractive activity which earn higher income is generally not the one associated to the
29 occupational identity. The identification of these factors could help understand possible
30 drivers of vulnerability and guide future actions to build flexibility and adaptive capacity
31 for people in these remote areas.

27
32 Introduction

33 Social-ecological systems represent an important concept for managing the use and

34 conservation of natural resources (Berkes and Folke 1998; Folke 2004, 2006). The social

35 and ecological components are intrinsically linked and inherently complex, where

36 changes, instability and uncertainty are integral to them (Berkes and Folke 1998). How

37 much perturbation a system can absorb and still maintain its structure and function

38 depends on its resilience (Holling 1973), which in turn, is dependent on the system’s

39 sensitivity to change and adaptive capacity (Folke et al. 2002). Sensitivity to change is a

40 multi-faceted concept and within the social components of the system, the focus of this

41 paper, it effectively describes how dependent people are on natural resources.

42 Resource dependency is strongly influenced by, among other things, family ties,

43 attachment to place, occupational identity and the proportion of family income derived

44 from a resource (Marshall et al. 2009).

45 Social factors such as these can act in different directions improving or decreasing

46 resource dependency and sensitivity to change (Tunsdall 1969; Sonn and Fisher 1998;

47 Adger et al. 2011; Marshall et al. 2013a). Occupational identity, for example, where

48 individuals strongly identify as “a fisher” or “a carpenter”, can enhance resource

49 dependency (Tunsdall 1969), through decreasing the flexibility that a resource

50 dependent person has to find a new occupation when needed. On the other hand, a

51 strong occupational identity can be associated with a higher adaptive capacity if it

52 provides resource dependent people with a deeper understanding of the system’s

53 functioning, facilitating them to cope with possible changes in the system (Marshall et

54 al. 2013a). The same occupational identity can be negatively associated with the

28
55 capacity to transform, which means that, even when promoting a positive effect on

56 adaptive capacity, occupational identity can hamper the ability to change and

57 adapt(Marshall et al. 2012).

58 Traditional communities in rural areas usually have a strong occupational identity, a

59 high connection with the local environment and a high social and economic

60 dependency on natural resources (Lane and Rickson 1997; Silva et al. 2007; Pangging et

61 al. 2011). In Amazonian riverine communities, people usually have various sources of

62 income through developing different natural resources activities (Perz 2010; Vasco

63 Pérez et al. 2015). One possible reason for this diversification is their expose to the

64 annual variation in water levels in local water bodies and rivers (Junk 1989; Parolin et al.

65 2004; Wittmann et al. 2010). These highly variable and dynamic processes (Ferreira

66 2000; Parolin et al. 2004) determine the potential of the environment to be productive

67 for the communities dependent upon them (Schöngart and Junk 2007). However,

68 recently, local communities have become especially vulnerable to ecological changes in

69 the systems as a result of resource over-exploitation and climate change (Rosenzweig

70 and Parry 1994; Marengo et al. 2008, 2011; World Bank 2010; Cinner et al. 2012a;

71 Pinho et al. 2015).

72 During the last decades the Amazon has experienced an increase in the frequency of

73 “once in a century” drought and flood events (Marengo et al. 2008, 2011, 2013; Pinho

74 et al. 2015). Extreme drought causes major mortality in fishes and trees (Borma et al.

75 2013), isolates people from the city or their agricultural land, and nullifies agricultural

76 production (Marengo et al. 2008). Bushfires also become a problem when drought is

77 followed by a lack of rain (Nepstad et al. 2004b; Xu et al. 2011). In extreme flooding

29
 78 events, people can lose their house and furniture, roads are closed, food and water

79 supplies are affected and medical assistance efforts are hampered (Marengo et al.

80 2013). Local people, who are already in vulnerable situations of poverty and lack of

81 access to public policies, must deal with these events, recover and continue earning a

82 livelihood, which not always is possible without government assistance (Cutter 1995;

83 Brulle and Pellow 2005; Mohai et al. 2009; Marengo et al. 2013).

84 The Purus river region, in the Brazilian Central Amazon, is marked by a long history of

85 exploitation of natural resources (Ferrarini 2009) and was one of the most sought-after

86 places for hunting manatees, alligators, turtles and different fish species (Ferrarini 2009;

87 Antunes et al. 2011; Batista and Petrere 2015). The high level of pressure on natural

88 resources and the high dependency of local riverine and indigenous households on

89 natural resources in this region (Antunes et al. 2016; Ríos-Villamizar et al. 2017; Silvério

90 et al. 2019) led civil society to pressure government into creating Protected Areas (Deus

91 et al. 2003). Considering the social and biological diversity of the Purus river, this action

92 resulted in two distinct levels of protection; No-Take Protected Areas, and Sustainable

93 Development Protected Areas, the latter which has the main goal to reconcile

94 biodiversity conservation with the maintenance of the traditional way of life of local

95 people (Brasil 2000).

96 Local communities can have several sources of income based on sometimes a large

97 range of local products. Understanding the multiplicity of income sources and the

98 multiple dependencies on natural resources is crucial if we intend to build and maintain

99 sustainability and local people’s life quality, in order to accomplish the goal of a

100 Sustainable Development Reserve. The lower Purus River region within the Central

30
101 Amazon region, provides an excellent case study example to better understand the

102 consequences of being dependent on a range of natural products. Because of its

103 proximity to Manaus (Amazonas state capital and the largest city of the Amazon Basin),

104 the region has recently suffered a massive harvesting and over-exploitation of several

105 animal and plant species (Deus et al. 2003, Antunes et al. 2011, Vieira et al. 2016).

106 During the 1950s there was significant migration to this area specifically to collect

107 Brazil-nut (Bertholletiaexcelsa, Lecythidaceae) (Ferrarini 2009; Antures et al. 2011). At

108 this time the rubber (HeveabrasiliensisEuphorbiaceae) harvesting sector had collapsed

109 and Brazil-nut was the new economic activity of interest (Loureiro 1986; Antunes et al.

110 2011). For some decades now the palm berry açaí (Euterpe precatoria Arecaceae) has

111 been another important forest product for the local economy (Muniz-miret et al. 1996).

112 Recently, this species has become very popular in international markets and is gaining

113 more attention in the Amazon (Brondízio 2008). Fishing and agriculture are also

114 important activities, as fish and cassava flour form the basis of food for local people

115 (Adams et al. 2006). The area is also one of the places where participatory management

116 of pirarucu(Arapaima gigas) is developed (Castello et al. 2009). This species is highly

117 endangered by past-exploration, and fishing is regulated by the Brazilian Federal

118 Government and well described in the literature (Castello 2004; Castello et al. 2009;

119 Peralta 2012; Campos-Silva and Peres 2016).

120 Assessing dependency on natural resources and evaluating how people can be sensitive

121 to change can help to provide deeper insights into the nature of social vulnerability

122 (Marshall et al. 2007), which is important for adaptation planning (Engle 2011). One

123 way is evaluate the flexibility associated with being resource dependent (Marshall et al.

124 2007; Cinner et al. 2018). Identifying more or less vulnerable groups of people is

31
125 possible through the identification of socio-demographic and/or economic factors

126 associated with vulnerability (Marshall et al. 2007), which can vary depending on

127 biophysical conditions, history and cultural aspects (Berrouet et al. 2018). An important

128 consideration for identifying the key drivers of vulnerability is to quantify household

129 income, especially when it is at low levels, because it hampers the access to basic needs

130 (Reyes-García et al. 2016) and directly influences well-being (Diener and Biswas-Diener

131 2002; Easterlin 2003). Other factors are also likely to be important drivers of

132 vulnerability. Considering a social-ecological system with a high number of natural

133 resource activities and different levels of dependency on each one, it will be important

134 to identify the factors describing household income sources and levels and the natural

135 resource activities with higher economic importance. The association between socio-

136 demographic (like family size and schooling), economic (as different sources of

137 household income) and cultural variables (as occupational identity) can indicate what

138 are the drivers of sensitivity to change and vulnerability, contributing to the decision-

139 making process regarding people life quality and natural resource management and

140 conservation.

141 The goals of this study are to 1) identifying groups of people with different levels of

142 social-economic vulnerability; 2) identifying socio-demographic, economic, geographic

143 and environmental factors that affect income sources and levels for local residents (not

144 considering the previous formed groups); 3) determine the most important source of

145 income for local residents and 4) determine the relationship between occupational

146 identity (cultural aspect) and the major source of income.

147

32
148 Methods

149 Study area

150 The study was carried out in the Piagaçu-Purus Sustainable Development Reserve,

151 hereafter referred as the ‘reserve’. It is a Protected Area of the State of Amazonas,

152 located in the lower Purus river region between the Purus-Madeira and Purus-Juruá

153 interfluves (Figure 1). Is a sustainable use protected area, equivalent to the

154 International Union for Conservation of Nature (IUCN) V and VI categories (Dudley

155 2013). The reserve includes the municipalities of Beruri, Coari, Tapauá and Anori,

156 comprising an area of 834,245 hectares. Nowadays the lower Purus river region has

157 almost 2 million hectares under a mosaic of no-take and sustainable use protected

158 areas (Deus et al. 2003).

159 The region is covered by terra firme or upland forest, which is located in the higher

160 portions of the Basin (Sioli 1985), and by the varzea, which is the forest reached by

161 flooded rivers and remains underwater annually from December to June (Ayres 1993).

162 The annual flooding or flood pulse (Junk et al. 1989) is responsible for major changes in

163 the landscape (Wittmann et al. 2004), as well as the dynamics of soil use by human

164 populations (Junk & Piedade 2004). The average river height difference between times

165 of flooding and dry periods reaches up to 10 m (Ronchail et al. 2006; Satyamurty et al.

166 2013). Seasonal flooding in the study area reaches up to 3 m of height and lasts less

167 than 50 days/year (Wittmann et al. 2010).

168 The entire reserve currently comprises around 65 villages and over 5000

169 people(AMAZONAS 2019). Many residents in this region have multiple skills they apply

170 to a number of resource use activities such as fishery, agriculture, horticulture,

33
171 carpentry, and forest resource harvesting. The flood pulse regulates all these activities,

172 including migration between the villages and the city. Many fishers from outside the

173 area will visit the reserve to fish during the year. However, the villages included in this

174 study are somewhat isolated from visitation because they at a distance from the Purus

175 waterway, and they are instead located near lakes and small rivers. The relative

176 geographic isolation decreases outsiders fishing pressure as well as hunting or

177 harvesting other natural resources. At the same time this isolation makes it more

178 difficult for local people to get access to the nearest town (to buy food supplies or for

179 health reasons). Importantly it may mean that it is more expensive or difficult to

180 transport harvested products to trade centers. In two of the five villages there is an

181 initiative in progress of participatory management of pirarucu (Arapaima gigas),

182 hereafter referred as ‘Arapaima management’.

183
184 Figure 1. The Sustainable Development Reserve Piagaçu Purus (SDR-PP) in the lower
185 Purus River, Central Amazon. White circles represent the two local villages in Ayapuá
186 sector (close to Terra firme) and three local villages in the Uauaçu sector (close to
187 Varzea) (Image by V.Vazquez).

34
188

189 Data gathering phase

190 The data sample included quantitative and qualitative information collected by field-

191 based interviews through semi-structured surveys. Before interviewing we informed

192 potential participants about the purpose of the project and obtained written consent to

193 participate. Everyone approached by interviewers agreed to participate. Participants

194 were interviewed in their houses or in public spaces of the village. Each interview lasted

195 around 1 hour and was performed by two researchers over 15 days.

196 Questions comprised information on personal data and on the economic activities

197 developed by the interviewee that relied on natural resources. Specifically regarding to

198 economics, data collected informed on the income earning activities developed, the

199 resources exploited, income provided, time spent at each activity, amount

200 commercialized and also information of others no-resource based sources of income.

201 Even though people in rural areas in the Amazon develop different activities to provide

202 income, they usually keep one activity as the main job, which is how they usually

203 identify themselves, and how the others identify them. We asked respondents to

204 identify the occupation that best described his/her main activity and we named this

205 information as the occupational identity. This identity represents the person’s

206 attachment to a job or a place (Twigger-Ross and Uzzell 1996; Wynveen et al. 2010). To

207 describe natural resource dependency we used variables (Table 1) based on the most

208 relevant predictors described in literature (Marshall et al. 2007) and we considered

209 dependency through an individual scale (Marshall et al. 2010) focusing on the

210 proportion of natural resources activities that complement household income. At each

35
211 of the five villages being studied the goal was to interview at least a 30% of the total

212 number of households. As many individuals in the same family are engaged in the

213 economic activities that supply the home, the household was the sample unit.

214

215 Data analysis

216 To separate groups of people we performed a non-hierarchical cluster analysis, using

217 distance matrix based on Gower’s index, after standardization of continuous variables

218 using the set of variables quoted on goal 1 (Table 2) and described in Table 1. In order

219 to create consistent groups, we visually choose the number of clusters based on a

220 graphical representation, selecting the number of groups with lower overlap. We

221 obtained the cluster membership for each of our respondents and tested the

222 differences between the cluster groups for a range of demographic and economic

223 variables.

224 Given that the respondent’s occupational identity did not necessarily correspond to the

225 activity providing them with their main source of revenue, we created an index of

226 identity strength (described in Table 1) to investigate this discrepancy. For individuals

227 who did not identify their main source of income to be the same as their occupational

228 identity, the identity strength was set to zero. To investigate if the occupational identity

229 job provides higher income, we performed a linear model using the Identity Strength

230 (IS) as response variable and a set of socio-economic variables as predictors (See

231 description of variables in Table 1 and equation in Table 2).

36
232 We performed three general linear models (detailed in Table 2) to identify factors

233 affecting income source and levels, to determine the most important source of income

234 and to determine the relationship between occupational identity and the major source

235 of income. To identify socioeconomic factors affecting income source and levels, weran

236 the linear model using a set of socio-demographic and economic variables as predictors

237 (see description of variables in Table 1 and equation in Table 2) and the individual

238 income from natural resource activities as response variable.

239 To identify environmental factors affecting income source and levels, we analyzed the

240 overlap between the flood pulse and income provided by natural resources activities

241 developed by respondents along the year. We used two different response variables:

242 average income by month and the intensity of activity (described in Table 1). The flood

243 pulse was estimated through the average water level for each month for the last 10

244 years registered in the hydrological station of Beruri-AM, the closest town to the study

245 area. This station is one of the several meteorological stations where the Brazilian

246 Water Agency (ANA) monitor the flood levels along the year.

247 In order to understand the most important activity to total household income, we

248 considered as predictors the income per month from each activity developed by

249 respondents. In this case, income was divided by the number of months respondents

250 spent in each activity, leading to a comparable measure of income per effort. The

251 variable ‘distance’ represent the distance from the village to the closest town and was

252 measured by the river (not in a straight line) using the distance tool ‘measure line’

253 (QGIS Development Team 2009) and the variable ‘dependency on fishing’ was chosen

254 because fishing (see description of variables in Table 1 and equation in Table 2) is a

37
255 traditional activity in the Amazon and is considered as the most important activity

256 (Barthem et al. 1997; Victoria and Morgana 2011).

257 All linear models were developed following the same procedure. From the full model

258 we selected and validated the most parsimonious model using the Akaike’s Information

259 Criterion corrected for small sample size (AICc) (Burnham and Anderson 2002), followed

260 by hierarchical partition of each explanatory variable (Barton 2013). We used the

261 comparisons of the beta estimates to measure the size effect of each predictor in each

262 response variable. Where the response variable lacked normality (InrandTi) we used the

263 logarithmic form. All continuous predictors were standardized prior to the analysis. All

264 statistical analysis was carried out using R (R Core Team 2017).

38
265 Table 1. Variables used to represent resource dependency for local riverine inhabitants of the Piagaçu-Purus Sustainable Development Reserve,
266 in lower Purus River, Central Amazon. N=49.
267
Variable Acronym Category What it means Type Mean SD
Family size FS Number of children plus parent 7.2 3.86
Age divided by the number of years living in the village,
Family ties FT 0.84 0.27
representing attachment to place
Age Age Age (years) 44.59 14.88
Socio-
Schooling Sch Number of years in formal school Continuous 2.67 3.73
demographic
Calculated from the formula: IS = I1st - I2nd / Dact (it was
only measured for respondents for whom there was a
Identitystrength IS 32.79 80.24
correspondence between the occupational identity and
the major source of income)
The way the respondents described their own identities
Occupationalidentity IDjob Cultural Categorical
according to their employment
Distance Dis Geographic Distance to the closest town (by the river, in km) 135.6 13.32
Sum of household income from natural resources
Total householdincome Ti activities, alternative sources of income, and arapaima 1714.63 1604.93
management (in US dollars/year)
Individual income from Income (in US dollars/year) from activities using natural
Inr 152.19 198.75
natural resources activities resources divided by the family size
Continuous
Number of economic activities reliant on natural
resources based on the reported income for each
Labor effort Eff 50.26 34.8
natural resource activity multiplied by the sum of
Economic months spent on that activity
Sum of the number of citations of people who work
Intensityofactivity Iact NA NA
each month with each activity (Brazil-nut harvesting,

39
 açai harvesting, agriculture, fishing, carpentry and
arapaima management).

Average income for all activities developed in each

Averageincome/month Ainc 608.87 193.72
month along the year (in US dollars/year).
Subsidies received from the federal and/or state
government (in US dollars/year) per family. It can
include retirement, pension, bolsa-floresta (a public
state policy that provide financial aid in return to non-
Household social
Ben opening new areas of cultivation in native forest), closed 660.08 959.23
assistancebenefit
benefit (a payment for fishers during the closed
reproduction season) and “bolsa-familia” (social welfare
program that provide financial aid to poor Brazilian
families)
Income received from the federal and/or state
Social assistancebenefit Ib government – ‘Ben’ – (in US dollars/year) divided by the 114.55 184.00
family size
Dependency on multiple Proportion of income that comes from multiple natural
Depnr 0.51 0.32
natural resources activities resources activities
Dependency on fishing
Depfis Proportion of income that comes from fishing 0.23 0.29
(supply chain access)
Activity that provide the highest household income
Mostimportanteconomicact
Act1st (agriculture, fishing, açai, brazil nut, management, Categorical NA NA
ivity
carpentry)
Householdincome per Income from each activity (x) divided by the number of
MIx NA NA
effort months that each person spent on that activity
Income from alternative sources like commerce and Continuous
Alternativeincome Ialt other jobs not related to natural resources activities (in 186.18 375.18
US dollars/year)

40
 Income from the most
I1st Major income (in US dollars/year) 103.26 199.55
important activity
Income from the second
I2nd Second major income (in US dollars/year) 33.44 83.40
most important activity
Number of activities developed, varying from 1 to 5
Diversityofactivities Dact (agriculture, açai harvesting, brazil-nut harvesting, 2.86 0.87
carpentry and fishery)
268
269
270
271 Table 2. Variables used in the analysis to achieve each goal. In goal 3, MI refers to mean income from the following activities: carp: income from
272 carpentry; bnut: income from brazil-nut; agri: income from agriculture; acai: income from açai; fish: income from fishing; alt: income from
273 alternative sources plus government social benefits; man: income from arapaima management. Otheracronyms for thevariables are described in
274 Table 1.

Goal Response variable Transformationof data Variables used and/or equations to achieve each goal

standardization (continuous predictors) /

1 N/A FS, FT, Sch, Eff, Inr,Ib, Depnr,Depfis,Act1st
Gower index of distance
Individual income from natural log (response variable) / standardization
2b Inr ~ FS + FT + Dis + Sch + Eff + Act1st
resources activities (continuous predictors)
3 Total householdincome log (response variable) Ti ~ MIcarp + MIbnut + MIagri + MIaçai + MIfish+ MIalt+ MIman
log + 1 (response variable) / standardization
4 Identitystrength IS ~ FS + Age + Sch + Dis + Eff + Inr + Ialt + Ben + IDjob
(continuous predictors)
275

41
276
277 Results

278 Socio-demographic and economic vulnerability

279 An overview of the socio-economic characteristics of the villages indicated an average

280 age of 44.6 years old (ranging from 19 to 84) and an average family size of 7.2 people

281 with 2.7 years of schooling. Total individual income is extremely variable (US$ 323.07 

282 327.51/year), but is very low on average. From 49 respondents, only seven are above

283 the poverty line and earn more than $1.90/day (World Bank 2018). Each respondent

284 develops, on average, 2.86 activities for household income (varying from 1 to 5). The

285 cluster examination clearly separates two distinct groups of respondents with regards

286 to socio-demographic and economic vulnerability (Figure 2).

287

288
289
290 Figure 2. A. Cluster plot showing two groups of people (n=49), with multiple natural
291 resource dependent group (in light gray), with 26 people and fishing dependent group
292 (in dark grey) with 23 people; B – F: Box-plots comparing the differences between the
293 groups regarding B) Family size; C) Schooling; D) Individual income from natural

42
294 resources activities; E) Dependency on multiple natural resources and F) Dependency
295 on fishing. Vertical bars represent total amplitude of the distribution (minimum and
296 maximum values). Black dots represent outliers. Each box represents 50% of the central
297 values of the distribution, restricted by the first (lower) and third quartiles (top).
298 Notches in the box represent the 95% confidence interval around the median.
299

300 The first cluster, on the left side of Figure 2A comprises 26 individuals, of whom 16 self-

301 identified as Brazil-nut harvesters. The individuals in this cluster generally have smaller

302 family size, more years in formal school, higher individual income, higher financial

303 dependency on natural resources activities but lower dependency on fishing (Figure 2B-

304 F; Table 3). We labeled this group of respondents as “multiple natural resource

305 dependent”.

306 The second group has 23 individuals, of whom six are fishers (from 8 fisher in the whole

307 sample). This group has larger family and lower levels of schooling, lower individual

308 income and lower financial dependency on multiple resources (Figure 2B-F; Table 3).

309 We labeled this group as “fishing dependent”, as they are more financial dependent on

310 fishing. Eleven individuals were set in the overlapping area (Figure 2A) and only one

311 interviewee (a local trader) was detached from the two groups but set closer to fishing

312 dependent group, probably due to the higher income from trading.

313 The distance of each village to the closest town did not contribute to distinguish socio-

314 demographic and economic features among villages. Conversely, even though age was

315 not included as a variable for cluster analysis, its difference was tested between groups,

316 showing that the “fishing dependent” group is older than the “multiple natural resource

317 dependent” group (Kruskal-Wallis test = 6.5819, df = 1, p = 0.0103).

43
318

319 Table 3. Differences between the groups generated by cluster analysis according to
320 mean, standard deviation and Kruskal-Wallis test (1Significant differences). Group 1 has
321 26 individualsandgroup 2 has 23 individuals.
322
Multiple natural Kruskal-
Variables Acronym Fishing dep.
resource dep. Wallis test
Mean SD Mean SD
Family size FS 5.54 3.24 9.09 3.65 p = 0.0011
Family ties FT 0.86 0.25 0.81 0.29 p = 0.400
Schooling Sch 4.12 3.97 1.04 2.67 p = 0.0011
Labor effort Eff 60 38.49 39.26 26.83 p = 0.070
Individual income Inr 253.5 228.99 37.67 27.55 p < 0.001 1
Social 106.3 123.8
assistancebenefit Ib 5 212.76 3 149.19 p = 0.182
Dependency on
multiple natural
resources Depnr 0.67 0.27 0.34 0.29 p < 0.001 1
Dependencyonfishin
g Depfis 0.08 0.11 0.39 0.35 p = 0.0011
323

324

325 Family and individual income

326 The difference in income between the two groups was quite evident. Because of that,

327 the next step in the analysis was to explain what could predict the family and individual

328 income. As expected, the income (both individual income from natural resources

329 activities and total income per family) is positively affected by the brazil-nut and açai

330 harvesting (Figure 3A and 3B). The labor effort is an important predictor for the

331 individual income (Figure 3A) while carpentry presented a great significant effect only

332 on income per family (Figure 3B).

44
333 Schooling, although not significant according to the comparisons of the beta estimates,

334 was present in the majority of the best parsimonious models and positively correlated

335 to the number of activities develop by locals (R2=0.18, F=9.933, p< 0.01).Family size,

336 which was also marginally significant, has a negative effect on income levels, meaning

337 that bigger families generally have lower income per family member (Figure 3A).

338 Agriculture was not selected in the most parsimonious models in the model selection

339 process.

340

Individual incom e (nat . res.) Tot al incom e per fam ily

Açai Açai
Brazil-nut
Brazil-nut
Labor effort
Pr ed i ct o r s

Schooling Carpentry

Carpentry Alt er native

Agricultur e
Arapaim a
Fam ily ties m anagem ent

Fam ily size Fishing

0 1 2 3 0.000 0.004 0.008

Regression coe ficients ( +- 9 5 % CI) Regression coe ficient s ( +- 9 5 % CI)
341

342 Figure 3. Coefficient estimates (±95% confidence intervals) showing the magnitude and
343 direction of effects for different predictors on A) the household income from natural
344 resources activities per family member and B) the total household income per family in
345 five villages into the Sustainable Development Reserve Piagaçu Purus, Central Amazon.
346 Confidence intervals crossing the zero vertical line are not significant. Blue lines
347 represent the predictors with positive and significant effect on the response variable.
348

349 The activities of lesser economic importance were Arapaima management and fishing. A

350 clear pattern emerged between the average income (Figure 4A) and the intensity of

351 natural resources activities with the months of rising water (Figure 4B). Most activities

45
352 are developed during the drought season (Figure 4B), but most part of the income is

353 earned during the flood season, especially because of Brazil-nut and açaí harvesting

354 (Figure 4A).

355

Activity 2100
60
60
Brazil-nut
Açai
Agriculture
Fishing
Carpentry
Arap. manag. 1800
Intensity of activity 40
40

Flood pulse (cm)

variable
Bnut_aver
Acai_a ver
Agr i_aver

1500Fish_a ver
Carp_aver
Manag_a ver

20
20

1200

0 0

J F M A M J J A S O N D
2.5 5.0
Month
7.5 10.0 12.5

356 Month

357
358 Figure 4. Relationship between the flood pulse (dark blue line in both plots) with the A)
359 average income from natural resources activities and with B) the intensity of each
360 activity along the year (for variables details, see Table 1).
361
362

363 Occupational identity

364 Occupational identity was not related to the major source of income for 65% of

365 harvesters interviewed(Figure 5). Most people identified themselves as a Brazil-nut

366 harvesting, but the major source of income, on average, is provided by government

367 benefits (Figure 5). But, it is worth mentioning that it was not expected that someone

368 identify themselves as a “benefit receiver” and, excluding this source of income, people

369 that self-identify as a Brazil-nut harvester actually have this activity as the main income

46
370 source (Figure 5). We were not able to explain the strength of respondent identity on

371 the basis of any variable tested, because all confidence intervals crossed the zero

372 vertical line through beta estimates comparisons.

373

374

375 Figure 5.Proportion of people that self-identify with each activity (right) and therelative
376 importance (average) of each source of income on total income for each activity (left).
377 Note that ‘Benefits’ and ‘Arapaima management’ are not shown in the leftplot, because
378 no one self-identify with these activities.
379
380

381 Discussion

382 In this study, inside a protected area in Central Amazon, two different groups of people

383 could be identified, one more disadvantaged and socially vulnerable than the other.

384 Also, we uncovered that the vulnerability is triggered by economic aspects such as the

385 number of extractive activities developed and the income generated from some specific

386 activities, but also by socio-demographic features such as family size and schooling.

387 Resources are constantly affected by the dynamics of the ecological system. Harvesters

388 are constrained by the access of communities to the market and by the likely for

389 obtaining income from other sources. The extractive activity which earn higher income

390 is generally not the one associated to the occupational identity. The identification of

47
391 these factors could help understand possible drivers of vulnerability and guide future

392 actions to build flexibility and adaptive capacity for people in these remote areas

393 (Marshall et al. 2007; Cinner et al. 2018).

394

395 Identifying vulnerable groups

396 The fishing dependent group is more disadvantaged by being focused in only one

397 extractive activity, by having lower income and lower schooling and for forming larger

398 families. Fishing dependency is an important factor with regards to vulnerability, since

399 this group is almost half of the population assessed, which we believe, is a good

400 representation of all lower Purus riverine communities. Even to Amazonian

401 communities, which pursue wide knowledge on environment and natural resources

402 (Posey 1983, 1984, 1986, 1987; Anderson and Ioris 1992; Anderson et al. 1995; Silva et

403 al. 2007; Hallwass et al. 2019) enabling flexibility to cope with changes (Aswani et al.

404 2018), formal education is crucial to positively complement the traditional knowledge

405 and improve life quality (Reyes-García et al. 2010). In this sense, schooling could pave

406 the way to flexibility and built the skills to adapt in many ways, if a change in the system

407 requires it, influencing on health outcomes and access to adequate housing facilities

408 (Catalano et al. 2004; Gounder and Xing 2012).

409 Instead, since larger family size is associated with poverty (Orbeta-Jr 2006) and less

410 flexibility to change (Lobao and Meyer 2001), income and family size of fishing

411 dependent group hamper changes and preclude the adaptive capacity (Cinner et al.

412 2018) mainly because parent’s responsibility and effort to sustain the family is greater.

48
413 Conversely, sometimes, in rural areas, as in the Central Amazon communities, children

414 up from 8 years old help parents in several natural resources activities, especially for

415 subsistence purpose (AMAZONAS 2019). Then, in some cases, child labor could provide

416 an important contribution to family income, even helping to pay for school costs when

417 it is necessary (Suryahadi et al. 2005). Notwithstanding, users dependent of multiple

418 natural resource are younger respondents who have smaller family’s group. This may

419 simply be due to the fact that young people are still building their families or may

420 indicate that they are having fewer children than in the past, as the Brazilian fecundity

421 rate is decreasing (IBGE 2019a).

422

423 Predictors of income source and levels

424 According to our data, the higher total income is dependent on forest resources and in

425 general, agrees with the assertion that natural resources activities are deeply

426 dependent on flood pulse (Wittmann et al. 2004). Among diversity of activities

427 performed in the area assessed, Brazil-nut and açai harvest were the main drivers of

428 higher household income, earning the largest portion of income during the flood

429 season, followed by carpentry. As less people is working as carpenter, this may partly

430 explain why they achieve the second highest average income. Carpentry is the only

431 activity that does not follow any flood pulse pattern and may be developed all over the

432 year but is less frequent likely because of the special skills and specialized ability

433 required. These skills are quite different from those needed for açaí and brazil-nut

434 harvesting, which are performed during part of the year (during rising waters). All other

49
435 activities are developed predominantly during the drought season and generate much

436 lower income.

437 Fishery, for instance, is performed mainly during dry season, when waters are back to

438 the main channel of the river. At this season, buyers are constantly traveling to the

439 villages. However, this is not the production season for forest resources fruits and

440 seeds, specifically açai and Brazil-nut, and may seriously hamper the outflow of forest

441 products gathered by the communities. This is important to consider when evaluating

442 possible impact on these communities from anthropogenic modifications in the

443 environment or climate change, as changing in the intensity and duration of flood pulse

444 can drastically affect economic activities responsible for generating large portion of

445 household income for local people (Borma et al. 2013; Marengo et al. 2013; Pinho et al.

446 2015).

447 Household income produced by açai and Brazil-nut was positively affected by the labor

448 effort, a measure composed by the diversity of income activities and time spent

449 performing each one. Considering the income divided by family members, there is a

450 clear decrease in individual income provision. It means that as family increase, the

451 increase in labor effort is not enough to provide the equivalent income per family

452 member (Orbeta-Jr 2006), leading us to conclude that vulnerability increases together

453 with family size.

454 Even though there was no clear difference detected between the two groups regarding

455 labor effort, the diversity of activities is positively correlated with schooling. So, the

456 greater income returns observed in the multiple natural resources dependent group

50
457 could be, partly, a consequence of higher schooling. This may agree with the idea that

458 formal education contextualized in local reality may improve income returns and

459 biodiversity conservation (Garmezy 1991; Becker and Ghimire 2003). It also expands the

460 diversity of activities carried out by people and enhance learning and flexibility through

461 new skills and knowledge (Castagno and Brayboy 2009).

462 To a certain extent, the relationship between the level of household income and

463 flexibility can be positive or negative (Marshall et al. 2013a). Those people involved with

464 bigger business and income may have less flexibility (the feeling that there is “more to

465 lose”). On the other hand, if less money is involved, feeling that there is “nothing to

466 lose” can provide flexibility and, as consequence, build adaptive capacity (Cinner et al.

467 2018). Probably the key question may be related to the amount of money involved, and

468 what exactly is a low or a high income. In the lower Purus river, the relative isolation,

469 the few achievement of basic services in health and supplies, and the lower income

470 below the poverty line for most people (World Bank 2018) is a factor decreasing

471 flexibility and increasing vulnerability. As a result, people have less options to cope with

472 basic problems (Garmezy 1991; Benzie 2014; Preston et al. 2014).

473

474 Occupational identity

475 Our findings suggested that in general, the extractive activity, which earns higher

476 income is not the one associated with the occupational identity. Most respondents still

477 identified as a Brazil-nut harvester, which is the activity that the first generation of

51
478 migrants in the area came to do during the last 50 - 70 years (Ferrarini 2009). Many

479 different activities have been developed by people since then, but the sense of identity

480 has not changed. This may be the reason why there is not a strong connection between

481 occupational identity and the most important income earning activity. A sense of

482 identity is an important factor that can influence resource dependency and adaptive

483 capacity (Marshall and Marshall 2007; Marshall et al. 2010, 2012). If a person’s sense of

484 identity is strong (their attachment to their job is strong), it can decrease flexibility in a

485 way that, if the system changes, people are adapt due to their desire to hold on to their

486 identity. On the other hand, a strong identity can also be a driver of a strong sense of

487 the need for adaptation (Marshall et al. 2013a). For instance, if a person has intimate

488 knowledge about a job and possesses a complete domain of techniques this can allow

489 him/her to deal with the changes in the system and adapt perhaps better. Identity is still

490 an important future area of research because ensuring people that maintain this

491 identity might be very important for their cultural wellbeing.

492 This study highlights the importance of an integrated approach between ecological,

493 socio-demographic and economic issues related to traditional use and trade of

494 biodiversity products in Central Amazon. It is insufficient to evaluate sustainable use of

495 natural resources without this broad approach, especially in developing countries,

496 where poverty and inequality sometimes are the main drivers to social vulnerability

497 (Adger 2006; Lemos et al. 2013), even in a resource-rich ecological system. Natural

498 resources are subject to high variability already (i.e. flood levels) and although local

499 people in the Amazon have adapted to this variability to a large extent, this variability

52
500 might increase with climate change (Lemos et al. 2013; World Bank 2018). People in

501 remote areas are subject to a larger risk (Brooks et al. 2005), especially when the

502 isolation is accompanied by the lack of basic facilities and social organization (Adger et

503 al. 2006; Lemos et al. 2013), which makes people lose the ability to act collectively

504 (Adger 2009). Increasing income source diversity can be effective in building flexibility

505 (Bailey and Pomeroy 1996) and decreasing sensitivity to change (Adger 2000), more

506 than just achieving a higher income. Future actions should consider efforts to increase

507 income diversity (Cinner and Bodin 2010) through a better understanding of forest

508 resources productive chains, which were the most profitable activities.

509

510 References
511

512 Adams, C. et al. 2006. Sociedades caboclas amazônicas: modernidade e invisibilidade

513 (EA São Paulo: FAPESP, Ed.).

514 Adger, W. N. 2000. Social and ecological resilience : are they related ? - Prog. Hum.

515 Geogr. 24: 347–364.

516 Adger, W. N. 2006. Vulnerability. - Glob. Environ. Chang. 16: 268–281.

517 Adger, W. N. 2009. Social Capital, Collective Action, and Adaptation to Climate Change.

518 - Econ. Geogr. 79: 387–404.

519 Adger, W. N. et al. 2006. Fairness in Adaptation to Climatic Change (WN Adger, A

520 Paavola, S Huq, and M Mace, Eds.). - MIT Press.

521 Adger, W. N. et al. 2011. This Must Be the Place : Underrepresentation of Identity and

522 Meaning in Climate Change Decision-Making Research Articles This Must Be the

53
523 Place : Underrepresentation of Identity and Meaning in Climate Change Decision-.

524 - Glob. Environ. Polit. 11: 1–25.

525 AMAZONAS 2019. Plano de gestão da Reserva de Desenvolvimento Sustentável

526 Piagaçu Purus.

527 Anderson, A. B. and Ioris, E. M. 1992. Valuing the rain forest: Economic strategies by

528 small-scale forest extractivists in the Amazon estuary. - Hum. Ecol. 20: 337–369.

529 Anderson, A. et al. 1995. Forest management pattern in the floodplain of the Amazon

530 estuary. - Conserv. Biol. 9: 47–61.

531 Antunes, A. P. et al. 2011. Rio, floresta e gente no baixo Purus: saber e uso da

532 biodiversidade na RDS Piagaçu Purus. - In: Santos, G. M. dos (ed), Album Purus.

533 EDUA, pp. 167–195.

534 Antunes, A. P. et al. 2016. Empty forest or empty rivers? A century of commercial

535 hunting in Amazonia. - Sci. Adv. in press.

536 Antures, A. P. et al. 2011. Rio, floresta e gente no baixo Purus: saber e uso da

537 biodiversidade na RDS Piagaçu Purus. - In: Santos, G. M. (Org. . (ed), Álbum Purus.

538 1st ed.n. EDUA, pp. 338.

539 Aswani, S. et al. 2018. Global trends of local ecological knowledge and future

540 implications. - PLoS One 13: 1–19.

541 Ayres, J. M. 1993. As matas de várzea do Mamirauá, médio rio Solimões.

542 Bailey, C. and Pomeroy, C. 1996. Resource dependency and development options in

543 coastal Southeast Asia. - Soc. Nat. Resour. 9: 191–199.

544 Barthem, R. B. et al. 1997. A Pesca na Amazonia: Problemas e Perspectivas para o seu

545 manejo. - In: MCT/ CNPq/ Sociedade Civil Mamirauá (ed), Manejo e conservação

54
546 de vida silvestre no Brasil. pp. 173–185.

547 Barton, K. 2013. MuMIn: multi-model inference. R package version 1. in press.

548 Batista, V. da S. and Petrere, M. 2015. Characterization of the commercial fish

549 production landed at Manaus, Amazonas State, Brazil. - Acta Amaz. 33: 53–66.

550 Becker, C. D. and Ghimire, K. 2003. Synergy between traditional ecological knowledge

551 and conservation science supports forest preservation in Ecuador. - Conserv. Ecol.

552 in press.

553 Benzie, M. 2014. Social justice and adaptation in the UK. - Ecol. Soc. in press.

554 Berkes, F. and Folke, C. 1998. Linking social ecological systems: Management Practices

555 and Social Mechanisms for Building Resilience.

556 Berrouet, L. M. et al. 2018. Vulnerability of socio—ecological systems: A conceptual

557 Framework. - Ecol. Indic. 84: 632–647.

558 Borma, L. S. et al. 2013. Impactos dos eventos extremos de seca e cheia sobre os

559 recursos hídricos amazônicos e ações da Defesa Civil. - In: Secas na Amazônia:

560 causas e consequências. Oficina de textos, pp. 367.

561 Brasil 2000. Lei Federal n 9.985 de 18/07/2000. - SNUC - Sist. Nac. unidades Conserv. in

562 press.

563 Brondízio, E. 2008. The Amazonian Caboclo and the Açaí palm: Forest Farmers in the

564 Global Market.

565 Brooks, N. et al. 2005. The determinants of vulnerability and adaptive capacity at the

566 national level and the implications for adaptation. - Glob. Environ. Chang. 15:

567 151–163.

568 Brulle, R. J. and Pellow, D. N. 2005. ENVIRONMENTAL JUSTICE: Human Health and

55
569 Environmental Inequalities. - Annu. Rev. Public Health 27: 103–124.

570 Burnham, K. P. and Anderson, D. R. 2002. Model selection and multimodel inference.

571 Campos-Silva, J. V. and Peres, C. A. 2016. Community-based management induces

572 rapid recovery of a high-value tropical freshwater fishery. - Sci. Rep. 6: 1–13.

573 Castagno, A. E. and Brayboy, B. M. J. 2009. Culturally Responsive Schooling for

574 Indigenous Youth: A Review of the Literature. - Rev. Educ. Res. 78: 941–993.

575 Castello, L. 2004. A Method to Count Pirarucu Arapaima gigas: Fishers , Assessment ,

576 and Management. - North Am. J. Fish. Manag. 24: 379–389.

577 Castello, L. et al. 2009. Lessons from integrating fishers of arapaima in small-scale

578 fisheries management at the mamirauá reserve, amazon. - Environ. Manage. 43:

579 197–209.

580 Cinner, J. E. and Bodin, Ö. 2010. Livelihood diversification in tropical coastal

581 communities: A network-based approach to analyzing “livelihood landscapes.” -

582 PLoS One in press.

583 Cinner, J. E. et al. 2012. Vulnerability of coastal communities to key impacts of climate

584 change on coral reef fisheries. - Glob. Environ. Chang. 22: 12–20.

585 Cinner, J. E. et al. 2018. Building adaptive capacity to climate change in tropical coastal

586 communities. - Nat. Clim. Chang. 8: 117–123.

587 Cutter, S. L. 1995. Race, class and environmental justice. - Prog. Hum. Geogr. 19: 111–

588 122.

589 Deus, C. P. et al. 2003. Piagaçu - Purus: Bases Científicas para a criação de uma Reserva

590 de Desenvolvimento Sustentável (IDSM, Ed.).

591 Diener, E. and Biswas-Diener, R. 2002. Wil money increase subjective well-being?

56
592 Dudley, N. (Ed) 2013. Guidelines for Applying Protected Area Management Categories

593 (IUCN, Ed.).

594 Easterlin, R. A. 2003. Explaining happiness. - PNAS 100: 11176–11183.

595 Ferrarini, S. A. 2009. Rio Purus: história, cultura e ecologia (FTD, Ed.).

596 Ferreira, L. V. 2000. Effects of flooding duration on species richness, floristic

597 composition and forest structure in river margin habitat in Amazonian blackwater

598 floodplain forests: Implications for future design of protected areas. - Biodivers.

599 Conserv. 9: 1–14.

600 Folke, C. 2004. Traditional Knowledge in Ecological Systems. - Ecol. Soc. 9: 7.

601 Folke, C. 2006. Resilience: The emergence of a perspective for social-ecological

602 systems analyses. - Glob. Environ. Chang. 16: 253–267.

603 Folke, C. et al. 2002. Resilience and sustainable development: building adaptive

604 capacity in a world of transformations. - Ambio 31: 437–440.

605 Garmezy, N. 1991. Resiliency and vulnerability to adverse developmental outcomes

606 associated with poverty. - Am. Behav. Sci. 34: 416–430.

607 Gounder, R. and Xing, Z. 2012. Impact of education and health on poverty reduction:

608 Monetary and non-monetary evidence from Fiji. - Econ. Model. 29: 787–794.

609 Hallwass, G. et al. 2019. Fishers’ knowledge indicates temporal changes in composition

610 and abundance of fishing resources in Amazon protected areas. - Anim. Conserv.:

611 acv.12504.

612 Holling, C. S. 1973. RESILIENCE AND STABILITY OF ECOLOGICAL SYSTEMS. -

613 Annu.Rev.Ecol.Syst. 4: 1–23.

614 IBGE 2019. Censo demografico. - Inst. Bras. Geogr. e Estatística

57
615 Junk, W. J. 1989. Flood tolerance and tree distribution in central Amazonian

616 floodplains. - ACADEMIC PRESS LIMITED.

617 Junk, W. J. et al. 1989. The flood pulse concept in river-floodplain systems. - Can. Spec.

618 Publ. Fish. Aquat. Sci. 106: 110–127.

619 Lane, M. B. and Rickson, R. E. 1997. Resource development and resource dependency

620 of indigenous communities: Australia’s Jawoyn aborigines and mining at

621 coronation hill. - Soc. Nat. Resour. 10: 121–142.

622 Lemos, M. C. et al. 2013. Climate Science for Serving Society. - Clim. Sci. Serv. Soc.:

623 437–457.

624 Lobao, L. and Meyer, K. 2001. The Great Agricultural Transition : Crisis , Change , and

625 Social Consequences of Twentieth. - Annu. Rev. Sociol. 27: 103–124.

626 Loureiro, A. 1986. A grande crise (1908-1912).

627 Marengo, J. a. et al. 2008. The drought of Amazonia in 2005. - J. Clim. 21: 495–516.

628 Marengo, J. a. et al. 2011. The drought of 2010 in the context of historical droughts in

629 the Amazon region. - Geophys. Res. Lett. 38: 1–5.

630 Marengo, J. A. et al. 2013. Recent Extremes of Drought and Flooding in Amazonia:

631 Vulnerabilities and Human Adaptation. - Am. J. Clim. Chang. 02: 87–96.

632 Marshall, N. A. and Marshall, P. A. 2007. Conceptualizing and operationalizing social

633 resilience within commercial fisheries in northern Australia. - Ecol. Soc. in press.

634 Marshall, N. A. et al. 2007. How Resource Depedency Can Influence Social Resilience

635 within a Primary Resource Industry. - Rural Sociol. 72: 359–390.

636 Marshall, N. A. et al. 2009. A framework for social adaptation to climate change:

637 Sustaining tropical coastal communities e industries.

58
638 Marshall, N. A. et al. 2010. The links between resource dependency and attitude of

639 commercial fishers to coral reef conservation in the red sea. - Ambio 39: 305–313.

640 Marshall, N. A. et al. 2012. Transformational capacity and the influence of place and

641 identity. - Environ. Res. Lett. in press.

642 Marshall, N. A. et al. 2013. Social Vulnerability of Marine Resource Users to Extreme

643 Weather Events. - Ecosystems 16: 797–809.

644 Mohai, P. et al. 2009. Environmental Justice. - Annu. Rev. Environ. Resour. 34: 405–

645 430.

646 Muniz-miret, N. et al. 1996. The economic value of managing the açaí palm (Euterpe

647 oleraceae Mart.) in the floodplains of the Amazon estuary, Pará, Brazil. - For. Ecol.

648 Manage. 87: 163–173.

649 Nepstad, D. et al. 2004. Amazon drought and its implications for forest flammability

650 and tree growth: a basin-wide analysis. - Glob. Chang. Biol. 10: 704–717.

651 Orbeta-Jr, A. C. 2006. Poverty, vulnerability and family size: evidence from the

652 Philippines. - In: Poverty Strategies in Asia: A Growth Plus Approach. pp. 326.

653 Pangging, G. et al. 2011. Traditional management practices of natural resources of

654 forest dependent communities in Arunachal Pradesh - A case study of fringe

655 villages in Banderdewa forest range. - Indian J. Tradit. Knowl. 10: 269–275.

656 Parolin, P. et al. 2004. Central Amazonian Floodplain Forests: Tree Adaptations in a

657 Pulsing System. - Bot. Rev. 70: 357–380.

658 Peralta, N. 2012. “Toda Ação De Conservação Precisa Ser Aceita Pela Sociedade”

659 Manejo Participativo Em Reserva De Desenvolvimento Sustentável.

660 Perz, S. G. 2010. From Sustainable Development to “Productive Conservation:” Forest

59
661 Conservation Options and Agricultural Income and Assets in the Brazilian

662 Amazon*. - Rural Sociol. 66: 93–112.

663 Pinho, P. F. et al. 2015. Complex socio-ecological dynamics driven by extreme events in

664 the Amazon. - Reg. Environ. Chang. 15: 643–655.

665 Posey, D. 1983. Indigenous knowledge and development: An ideological bridge to

666 future. - Cienc. Cult. 35: 977–994.

667 Posey, D. 1984. Ethnoecology as applied anthropology in Amazonian development. -

668 Hum. Organ. 43: 95–106.

669 Posey, D. 1986. Manejo de floresta secunda´ria, capoeiras, campos e cerrados

670 (Kayapo´). - In: Suma etnológica Brasileira. pp. 173–185.

671 Posey, D. 1987. Etnobiologia: teoria e prática. - In: Suma etnológica Brasileira. pp. 15–

672 25.

673 Preston, I. et al. 2014. Climate change and social justice: an evidence review.

674 QGIS Development Team 2009. QGIS Geographic Information System. in press.

675 R Core Team 2017. R: A Language and Environment for Statistical Computing. in press.

676 Reyes-García, V. et al. 2010. Schooling and local environmental knowledge: Do they

677 complement or substitute each other? - Int. J. Educ. Dev. 30: 305–313.

678 Reyes-García, V. et al. 2016. Subjective Wellbeing and Income: Empirical Patterns in

679 the Rural Developing World. - J. Happiness Stud. 17: 773–791.

680 Ríos-Villamizar, E. A. et al. 2017. Surface water quality and deforestation of the Purus

681 river basin, Brazilian Amazon. - Int. Aquat. Res. 9: 81–88.

682 Ronchail, J. et al. 2006. Impact of the Amazon tributaries on major flood in Óbidos. -

683 Clim. Var. Chang. - Hydrol. Impacts 2001: 220–225.

60
684 Rosenzweig, C. and Parry, M. L. 1994. Potential Impacts of Climate change on world

685 food supply. - Nature in press.

686 Satyamurty, P. et al. 2013. A quick look at the 2012 record flood in the Amazon Basin. -

687 Geophys. Res. Lett. 40: 1396–1401.

688 Schöngart, J. and Junk, W. J. 2007. Forecasting the flood-pulse in Central Amazonia by

689 ENSO-indices. - J. Hydrol. 335: 124–132.

690 Silva, A. L. et al. 2007. Ethnobotany of riverine populations from the Rio Negro,

691 Amazonia (Brazil) Article. - J. Ethnobiol. 27: 46–72.

692 Silvério, D. V. et al. 2019. Fire, fragmentation, and windstorms: A recipe for tropical

693 forest degradation. - J. Ecol. 107: 656–667.

694 Sioli, H. 1985. Amazônia: Fundamentos da Ecologia da maior região de Florestas

695 Tropicais. - Stuttgart: 73p.

696 Sonn, C. C. and Fisher, A. T. 1998. Sense of community: Community resilient responses

697 to oppression and change. - J. Community Psychol. 26: 457–472.

698 Suryahadi, A. et al. 2005. Child Labor in Indonesia during the Economic Crisis : The

699 Trade-off between School and Work. - Dev. Change 36: 351–373.

700 Tunsdall, J. 1969. The fishermen: the sociology of an extremeoccupation. - MacGibbon

701 and Kee.

702 Twigger-Ross, C. L. and Uzzell, D. L. 1996. Place and identity processes. - J. Environ.

703 Psychol. 16: 205–220.

704 Vasco Pérez, C. et al. 2015. Income diversification of migrant colonists vs. indigenous

705 populations: Contrasting strategies in the Amazon. - J. Rural Stud. 42: 1–10.

706 Victoria, J. I. and Morgana, C. de A. 2011. El Consumo de pescado en la Amazonía

61
707 brasileña. COPESCAALC Documento Ocasional. No 13.

708 Vieira, M. A. R. M. et al. 2016. Participatory Monitoring and Management of

709 Subsistence Hunting in the Piagaçu-Purus Reserve, Brazil. - Conserv. Soc. 13: 254.

710 Wittmann, F. et al. 2004. The várzea forests in Amazonia: Flooding and the highly

711 dynamic geomorphology interact with natural forest succession. - For. Ecol.

712 Manage. 196: 199–212.

713 Wittmann, F. et al. 2010. Amazonian Floodplain Forests - Springer.

714 World Bank 2010. Development and Climate Change: Second Generation Bioenergy

715 Potential. - World Dev. Rep. 2010

716 World Bank 2018. Poverty and shared prosperity 2018: Piecing together the poverty

717 puzzle. - Piecing Together the Poverty Puzzle

718 Wynveen, C. J. et al. 2010. Place meanings ascribed to marine settings: The case of the

719 great barrier reef marine park. - Leis. Sci. 32: 270–287.

720 Xu, L. et al. 2011. Widespread decline in greenness of Amazonian vegetation due to the

721 2010 drought. - Geophys. Res. Lett. 38: 2–5.

722
723

62
1

63
 2 Disentangling the supply chain of açaí palm fruit in the Amazon and
3 its role on local wellbeing
4
5 Heloisa Dantas Brum, Alexandre Fadigas de Souza, Adriana Rosa Carvalho,Ingrid van
6 Putten, João Vitor Campos-Silva.
7

8 Abstract

9 Protected Areas (PAs) represent the most widespread strategy to ensure biodiversity
10 conservation and local livelihoods. There is strong evidence around the world that
11 conservation actions are more effective when combined the joint action of multiple
12 agencies with the participation of local people through community based-
13 management, adapting management or co-management. Sustainable-use PAs
14 sometimes do not accomplish their primary goalkeeping many local residents below
15 the poverty line, especially in tropical developing countries often experiencing
16 shortages in human and funding resources. Supply chain of socio-biodiversity products
17 has a high potential to conciliate conservation and income generation to local people,
18 but it may represent a complex network. Here, we conducted a comprehensive
19 assessment on the importance of açaí in an Amazonian social-ecological system.
20 Specifically, our aims were: (1) To evaluate connectance and relative importance of the
21 açai supply chain elements and identify its key elements; (2) To model the
22 sociodemographic and economic factors driving thelocal açaí income and (3) To
23 estimate the minimum price for açai to ensure a local fair trade. Our results elucidate
24 the harvesting operation of a biodiversity asset with a huge cultural value and great
25 importance to food security, which is fundamental to achieve the goals of Sustainable
26 Use Protected Area. Socio-biodiversity products can contribute to biodiversity
27 conservation together with economic return and rescuing of cultural pride. Açai can be
28 the symbol of this process, as a catalyzer of social transformation in riverine
29 communities in the Amazon.

64
30 Keywords: Community-based management, co-management, conservation, nom-
31 timber products, Euterpe precatoria.
32

33 Introduction

34 Protecting biodiversity is of one of the most challenging tasks of our time, given the

35 difficulty in reconciling these actions with the growing market demands of natural

36 resources (Machovina et al. 2015). Different strategies to protect biodiversity and

37 natural resources are being experienced (Berkes and Folke 1998). Although there is no

38 “one favorite cure-all solution” (Ostrom 2007; Ostrom et al. 2007 and others),

39 Protected Areas (PAs) represent the most widespread strategy to ensure biodiversity

40 conservation and local livelihoods (Leverington et al. 2010). Sustainable-use PAs,

41 including categories IV-VI of IUCN (Dudley 2013), can host good examples of efficient

42 use of natural resources, promoting the alignment of biodiversity conservation local

43 well-being (Berkes 2007). Even not completely avoiding external impacts of the

44 surrounding landscape, such as deforestation in tropical forest (de Marques et al.

45 2016), Sustainable Use PAs can be more effective in achieve its goals when allied with

46 governance and stakeholders engagement (Nepstad and Mcgrath 2002).There is

47 strong evidence around the world that conservation actions are more effective when

48 combined the joint action of multiple agencies (Berkes 2009) with the participation of

49 local people through community based-management (Naidoo and Adamowicz 2005;

50 Naidoo et al. 2011; Cinner et al. 2012b; Campos-Silva and Peres 2016), adapting

51 management (Holling 1978) or adaptive co-management (Berkes 2009).

65
52 Traditional communities has a strong relationship with natural resources (Sunderlin et

53 al. 2005), which in turns can create a complex knowledge network of cultural and

54 biological heritage, responsible for resilience of traditional communities (Cámara-Leret

55 et al. 2019). Consequently, local ecological knowledge (LEK) can enhance health and

56 well-being of local people (Mcdade et al. 2007), provide economic returns (Balick and

57 Mendelsohn 1992; Willis 2017) and increase the efficiency of resource use and

58 conservation (Baldauf et al. 2015; Schmidt et al. 2015). Moreover, participatory

59 conservation can reduce the overall costs of conservation (Somanathan et al. 2009),

60 this is very important especially in developing countries often experiencing shortages

61 in financial resources, infrastructure and people (Bruner et al. 2004; Watson et al.

62 2014; Campos-Silva et al. 2015).

63 Brazilian Amazon has been experiencing effective arrangements where local residents

64 are ensuring the protection of natural systems aligned with socioeconomic

65 improvements. The co-management of arapaima (the largest freshwater scaled fish on

66 earth) for instance represent one of the most well-succeeded initiative to conciliate

67 the conservation of floodplains, population recovery of arapaima and local wellbeing

68 (Castello et al. 2009). This is a useful example to rethink the management of natural

69 resources in the Amazon, because it ensures the harvesting zoning, including the

70 establishment of no take areas and promote engagement, commitment and

71 compliance of local communities at the same time (Campos-Silva and Peres 2016). This

72 model could inspire the co-management of a set of biodiversity assets, ensuring the

73 protection of this resources and the local involvement in conservation practices.

66
74 One common bottleneck in maintaining co-management actions is the lack of financial

75 return for locals, which can compromises the whole initiative (Campos-Silva et al.

76 2018b). Protected Areas of sustainable use sometimes do not accomplish their primary

77 goal, keeping many local residents below the poverty line (Specht et al. 2019). Supply

78 chain of socio-biodiversity products has a high potential to conciliate conservation and

79 income generation to local people, but it may represent a complex network. Food

80 supply chain, especially, are usually composed of small-scale producers in rural areas,

81 who often does not have the possibility to be responsible to transporting the products

82 to urban centers (Hernandez and Pedroza 2016), making the presence of the

83 middleman crucial in the network (Pedroza 2013). At the same time, food supply

84 chains usually requires a broad approach in order to understand the factors affecting

85 not only the importance of natural resources to local economy, but to local

86 subsistence as well (Martinot et al. 2017).

87 The palm tree açaí (Euterpe precatoria, Arecaceae) is the most abundant plant species

88 of the Amazon (ter Steege et al. 2013), and has a great importance in social, ecological

89 and cultural aspects. Recently, it has been increasing its demand on the international

90 market (Brondízio 2008), because of the use of fruits to produce a nutritious and high

91 energy beverage associated with health and sports activities. It is an important forest

92 resource to help engage local communities in a profitable economic activity together

93 with the valuation of biodiversity. In this paper we aim to investigate the importance

94 of açaí on Amazonian social-ecological system and specifically: (1) To evaluate

95 connectance and relative importance of the açai supply chain elements and identify its

67
 96 key elements; (2) To model the sociodemographic and economic factors that affect

97 açaí local income and (3) To estimate the minimum price for a açai fair trade.

98

99 Methods

100 The species

101 Euterpe precatoria, popularly known as açai, assai or açaí-da-mata (meaning “forest

102 açaí”) is a single-stemmed palm that may reach up to 20 m height occurring from

103 Central America to Bolivia (Stoian 2004). The species is considered domesticated or

104 semi-domesticated by indigenous people since pre-Columbian times (Clement et al.

105 2015; Levis et al. 2017), and can form oligarchic forest patches, in a way similar to,

106 although not as dense as E. oleracea, the related species that occur in Eastern Amazon

107 (Peters et al. 1989). Açaí is a useful plant, with economic and cultural value for

108 indigenous and riverine people. Recently, it has gaining attention of national and

109 international markets because of its nutritious and energetic contents (Yuyama et al.

110 2011) and antioxidant properties (Pacheco-Palencia et al. 2009). Data until 2017

111 indicate that 93.1% of Brazilian açaí production is supplied by E. oleraceae(IBGE 2018).

112 Brazil produces around 220,000 tons of açaí annually (IBGE 2018) and the activity

113 revenue is about US$ 150 million/year. The great majority of açai production in the

114 state of Amazonas comes from native forest (AMAZONAS 2005).

115

116

68
117 Study area

118 The study was carried out in the Piagaçu-Purus Sustainable Development Reserve,

119 hereafter referred as the ‘reserve’. It is a Protected Area of the State of Amazonas,

120 located in the lower Purus river region between the Purus-Madeira and Purus-Juruá

121 interfluves (Figure 1). It is a sustainable use protected area (equivalent to the V and VI

122 categories of IUCN, Dudley 2013). The reserve includes the towns of Beruri, Coari,

123 Tapauá and Anori, comprising an area of 834,245.00 hectares. Nowadays the lower

124 Purus river region has almost 2 million hectares under a mosaic of protected areas for

125 direct and indirect use (Deus et al. 2003).

126 The region is covered by terra firme or upland forest, which is located in the higher

127 portions of the Basin (Sioli 1985), and by the varzea, which is the forest reached by

128 flooded rivers and remains underwater annually from December to June (Ayres 1993).

129 The annual flooding or flood pulse (Junk et al. 1989) is responsible for major changes in

130 the landscape (Wittmann et al. 2004), as well as the dynamics of soil use by human

131 populations (Junk & Piedade 2004). The average river height difference between times

132 of flooding and dry periods reaches up to 10 m (Ronchail et al. 2006; Satyamurty et al.

133 2013). Seasonal flooding in the study area reaches up to 3 m of height and lasts less

134 than 50 days/year (Wittmann et al. 2010).

135 The entire reserve currently comprises around 65 villages and over 5000 people. This

136 reserve represents a huge social-ecological system composed by a set of livelihoods,

137 including fishery, agriculture, horticulture, carpentry, arapaima management and

69
138 forest resource harvesting. The intensity of each activity varies with the flood pulse.

139 Açaí and Brazil-nut

nut harvesting are seasonal activities, that take place during the

140 flooding season, when tree prod

produces mature fruits.

141

B C
A

142

143 Figure 1.The

The Sustainable Development Reserve Piagaçu Purus (SDR (SDR-PP)
PP) in the lower
144 Purus River, Central Amazon. White circles represent the two local villages in Ayapuá
145 sector (close to Terra firme) and three local villages in the Uauaçu sector (close to
146 Varzea) (Image by V.Vazquez). B. Purus river (photo: Paulo Santos); C. Açaí tree
147 (Euterpe precatoria)) in native forest (photo: H. Brum); D. Açaí fruits attached to the
148 bunch (photo: Isabela Oliveira); E. Açaí beverage being manually processed (photo:
149 Isabela Oliveira).

150

151 Data collection

152 The data gathering was realized in five communities distributed in two sectors, being

153 two communities in Ayapu

Ayapuá and three in Uauaçu sectors. Sectors represents a

70
154 geographic and political division, especially for management actions in Protected

155 Areas.

156 The data sample included quantitative and qualitative information collected by field-

157 based interviews through semi-structured survey. Before interviewing we informed on

158 the project purpose and requested written consent to participate. Everyone

159 approached by interviewers agreed to participate. Participants were interviewed in

160 their houses or in public spaces of the village. Each interview lasted around 1 hour and

161 was performed by two researchers over 15 days.

162 Questions comprised information on personal and demographic data (like age, number

163 of children, place of birth) and on the economic activities about açai harvesting, as the

164 income provided, time spent during the season (days of work), amount

165 commercialized and consumed and cost to harvest (in material and equipment, if it

166 was necessary). Each person was classified according to the level of the açaí supply

167 chain it occupies (collector, processor, distributor or consumer). At each of the five

168 villages being studied the goal was to interview at least a 30% of the total number of

169 households. As many individuals in the same family are engaged in the economic

170 activities that supply the home, the household was the sample unit.

171

172 Data analysis

173 The number of links and elements of the açaí supply chain connections in the

174 Sustainable Development Reserve Piagaçu Purus (SDR-PP) was evaluated through the

71
175 Supply Chain Index (SCI), develop by Plagányi and colaborators (2014), which

176 incorporate a measure of connectance and importance of flow between supply chain

177 elements (Plagányi et al. 2014). The index comprises two metrics. The first, sji,

178 represents the proportion of total product that receiver j receives from supplier i

179 relative to all product flowing into that element j. The second, pj, measures the

180 proportion of the total product in the supply chain that flows into receiver j. Each area

181 (sector Ayapuá and Uauaçu) has its own açaí supply chain, with specific elements in

182 each one. A lower value of SCI can indicate a higher connectance and resilience to

183 external actions, while higher SCI can indicate a higher dependency on few elements.

184 However, the interpretation depends on local context, and a low SCI can be associated

185 with high resilience and ability to cope and adapt with changes (Plagányi et al. 2014).

186 The sum of SCI of each supply chain element is denoted SCIT (supply chain index total)

187 and it can be parameterized by the number of links of the supply chain.

188 We also calculated simple metrics to the supply chain, such as the number of links (L)

189 and nodes (n) and the connectance (L/n2). The economic multiplier was calculated

190 based on the average price that açai (fruits or pulp) is sold in each level of the supply

191 chain. It represents the value by which the purchase price must be multiplicated to

192 give the sale price, with no quantification of costs. It represents a measure of added

193 value along the supply chain(Dyck and Sumaila 2010).

194 To evaluate what affects the household income provided by açai activity, we

195 performed generalized linear models considering as predictors a set of socioeconomic

196 variables (years of schooling, family size, distance from the closest city, time spent in

72
197 açaí activity, occupational identity and the açaí supply chain level occupied by the

198 harvester) that composed the full model. The response variable was the individual

199 income from açai activity (in logarithmic form), which was the income divided by

200 family member. From the full model we selected and validated the most parsimonious

201 model using the Akaike’s Information Criterion corrected for small sample size (AICc)

202 (Burnham and Anderson 2002). We also performed a model average after model

203 selection, considering the beta average of all variables within parsimonious models.

204 Finally, we compared the relative effect size of all variables, once the variables were

205 standardized through z standardization. All assumptions were respected, according

206 with Zuuret al. (2010), and all statistical analysis was carried out using R (R Core Team

207 2017).

208 We estimated the costs along the productive chain for collectors, processors and for

209 people that develop more than one stage of the activity. Distributors, who composed

210 only three people in the sample, were not able to provide information about costs and

211 earning from all activities they develop (as they usually work with different natural

212 resources in this area). As is expected that residents develop multiple resource

213 activities, we estimated the costs assuming that the relative importance of each source

214 of income is the same for the cost, leading to a value of cost proportional to the

215 income. We quantified the minimum price for açaí (fruits and pulp) considering an

216 equivalent of earning at least US$12.88/day, and comparing with the current value. All

217 costs and prices were calculated in Brazilian Reais and then converted to US dollar,

73
218 considering the conversion rate of April 12 of 2019 for US$ 1.00 = R$ 3.88

219 (https://www.bcb.gov.br/).

220

221 Results

222 We interviewed 53 local residents, seven women and 46 men. Açai harvesting is an

223 activity predominantly male but women help during the processing stage, mainly. We

224 identified 26 people involved in collections of fruits, 11 in processing, 4 in distribution

225 and 21 consumers (which did not develop any economic activity with açaí production).

226 Nine people developed more than one activity (as collection and processing, or

227 processing and distribution), which explains why the sum of people involved in the

228 supply chain is higher than 53. Respondents presented age between 19 and 84 years

229 (45.38 years on average), schooling from zero to 13 years (2.84 years on average) and

230 family size varying from 1 to 17 people (7.13 people on average).

231

232 The productive chain

233 The açai production in a local scale comprises four stages: collection, processing,

234 distribution and consume. The connections’ diversity and strength between the supply

235 chain elements vary between levels and localities, and the costs and profits are also

236 extreme variable between elements. The characterization of elements and stages of

237 the supply chain in both areas and the time and methods used during the activity are

238 basically the same.

74
239 In the Lower Purus River region, the fruit production season occurs from December to

240 June, and the season peak from February to April. The fruit production varies between

241 flooded and unflooded forest. Usually, açaí production begin in unflooded forest

242 (December) and then starts in flooded forest in February, during the flood season. The

243 activity is concomitant with Brazil-nut (Bertholletia excels Lecythidaceae) harvesting,

244 which represent one of the most important forest products to household economy in

245 the Lower Purus river region and a great importance in the occupational identity of

246 locals (Brum et al., in preparation). Most açaí collectors also work with Brazil-nut,

247 almost at the same time or shortly after. It was common the report of a substitution of

248 Brazil-nut by açai in the last years (specially in 2016), when the Brazil-nut fruit

249 production was extremely low in the Brazilian Amazon (IBGE 2018). The production of

250 açai and Brazil-nut are often flowed to four towns nearby the reserve (Table 1).

251

252 Table 1. Number of people, and production of açaí and Brazil-nut in the four towns
253 that compose the SDR-PP. The percentage of population and production is related to
254 the state of Amazonas. Data were extracted from the Brazilian Institute of Geography
255 and Statistics website (IBGE 2019b).
256
Açaí Brazil-nut
Town Population
(tons) (tons)
Anori 12,731 1,657 5
Beruri 13,220 169 3,000
Codajás 27,303 12,000 600
Coari 84,277 1,896 300
Total 137,531 15,722 3,905
% 3.37 31.13 30.54
257

75
258 Below we will describe each one of the supply chain elements present in the local scale

259 of açaí production, in the North area of SDR-PP.

260

261 1. Collector

262 Collectors are all men, between 19 and 60 years (37.94 years on average), responsible

263 to harvest the açaí fruit from the nature. To climb the palm trees, they use a bell

264 composed by palms leaves or a plastic fiber bag (locally called “peconha”), which they

265 attach to their feet to climb. Additionally, they also use a machete to cut the açaí

266 bunch and bags to carry the fruits (usually the same type of bag they use to climb the

267 tree). The trade unit is a bag, which has, on average, 54.5 Kg, and can produce, on

268 average, 37.4 liters of açaí pulp. To fill a bag, collectors must gather, on average, 7.78

269 açai bunches. Collector work, on average, 52 days/season (from 36 to 72 days) on açaí

270 activity. In a working day, a collector spends, on average, 6.5 hours (from 4 to 12 h),

271 searching and climbing açaí trees on native forest. One collector can gather from 1.5 to

272 4 bags (2.25 bags on average) in one day. They usually collect more than that, working

273 in partnerships with other collectors.

274 Collector sells the bags to processors (the next level of supply chain) by US$ 12.88. This

275 price can be a little bit higher in the beginning and end of the season. Collectors work

276 by themselves, but at the moment they go to a collecting day, they usually have the

277 trade settled. This is important as açaí fruit cannot be storage, and they must be

278 processed in no more than 24 hours after collecting from the trees. Collectors don’t

279 have formal agreements with processors, but they usually have a trust relationship,

76
280 and processor usually have some “favorite collectors”, whom they trust will collect

281 high quality fruits.

282

283 2. Processor

284 Is the person who process the fruits to produce the açai beverage (locally called as

285 “açai wine”, which is not alcoholic). They are all men, from 33 to 60 years (46.55 years

286 on average). They usually have their own mixer, which is an equipment crucial to an

287 efficient use of the fruits and it was considered as an artisanal process (Figure 2). The

288 activity is usually developed in the processor’s house, and this is the part of the supply

289 chain with more women participation, especially during packing. They can process up

290 to 30 bags in one day during the season peak, but on average is four. This stage

291 requires a higher spent with equipment, material and fuel (this will be described in

292 cost evaluation session), besides an infrastructure to process and storage the pulp.

293 Processors usually get water from the rain, collected in tanks or directly from the river,

294 and they make a treatment with chemicals for the removal of sediments (through a

295 flocculating agent) and sterilization.

296 Two processors sold part of its production to an industry, which is an ice cream factory

297 that were doing some quality tests to evaluate the possibility to settle a contract with

298 local producers in the area. The industrial process was not detailed in this paper, as we

299 did not quantify the amount processes and sold by the industry (Figure 2).

300

77
301 3. Distributor

302 Is the person who transport the açai wine or the fruits to the closest city, directly to

303 consumers or (in the case of fruits) to another processor or industry that use açaí (like

304 ice cream industry, for example). This person is also responsible to buy and transport

305 other natural resources products, like Brazil-nut, fish and agriculture products (mainly

306 cassava floor), which represent the most important source for household income for

307 local residents. The distributor usually have some family relationship with locals, and

308 have a trust relationship, not necessarily based on a fair trade with everyone, but

309 following the “aviamento” systems, common practice in the Amazon since the rubber

310 harvesting, where commercial relationships with locals lead to a constant debt (Silva

311 2010; Antunes et al. 2011). This situation is not as cruel and unequal as was in the past,

312 but still happens, especially in fish activities, which represent the most profitable

313 activity to the distributor. Fish is the most important economic activity to Amazon

314 residents and, according to distributors “is what pays the trip”. For local residents,

315 however, fishing is not the most profitable activity, as the price per kilo is extremely

316 low and the cost with fuel and ice avoid a fair income. Even so, locals must fish to

317 attract the distributor, and be able to sell the important products for their household

318 income which is, basically, Brazil-nut, açaí and cassava flour (Brum et al., in

319 preparation).

320

321

322

78
323 4. Consumer

324 We interviewed only local consumers, located in the five studied villages. Açai is very

325 appreciated by locals and some people declare a consumption of up to 1 liter per day

326 per person, which can occur during the peak of the season (around February and

327 March). During one season, we estimated an average consumption of 47.5 liters of açai

328 per person. Many people have the ability to climb the tree and collect açaí, even if it is

329 just for subsistence (with no commerce), and the knowledge to prepare the beverage

330 manually, if they do not have the mixer. But, since some residents acquired the mixer

331 few years ago, is more common people buy açaí from local processors. One liter of açai

332 pulp can be purchased, in local communities, by US$ 1.03 on average (varying from

333 US$ 0.77/liter to US$ 1.29/liter).

334 We did not interview people in other consumers markets, but is widely known that

335 açai is very appreciated in all Amazonas state and each small town can represent an

336 important consumer market. Beruri, which is the town that receive açai production

337 from Ayapuá communities, is the closest town to the reserve and most respondents

338 use this town to healthy assistance and food supplies. It is also the most important

339 economic center, being the first stop of production flow from inside the reserve.

340 Codajás (the town that receive açai production from Uauaçu communities) is consider

341 the “açaí capital” of Amazonas state and concentrates the largest part of the

342 Amazonas production (Table 1). Coari and Anori are towns that, together with Beruri,

343 compose the Northern region of the SDR-PP. Codajás is in its vicinity and it is an

79
344 important center for production flow, health assistance and food supplies for some

345 people from the Uauaçu area, but does not have an area inside the reserve.

346

347 Supply chain analysis

348 Supply chain was characterized for Ayapuá and Uauaçu, and each one presented some

349 visual distinct characteristics (Figure 2). The two supply chains differ slightly in the

350 number of nodes and links, which provides a similar score for some simple metrics as

351 link per node and the connectance (Table 2). The evaluation of magnitude and

352 connectance weighted by the quantity of product flowing through SCI identified some

353 key elements and it is observed different values of SCI, SCIT and SCIst (Figure 2, Table

354 2). Supply chain of Ayapuá has more nodes and links, and the product flow is highly

355 spread between elements. The lower value of SCI indicates a more connected supply

356 chain and less dependency on few elements. In supply chain of Uauaçu there is a lower

357 number of links and nodes, and connectance is slightly higher, being marked by an

358 extremely high flow between few elements (Figure 2, Table 3). The higher value of SCI

359 indicates a network more dependent in few elements which is, in this case, processor

360 number 2, responsible for carry a considerer amount of açai fruits to Codajás town (the

361 most important açai production center of the state of Amazonas, Table 1).

362 We did not identify products flow between areas, but there are two elements that are

363 the same, which is the city (Manaus, capital of Amazonas state, a consumer market)

364 and the ice cream industry, which we considered as industrial processor in a large-

80
365 scale (Figure 2). It is worth noting that the distributed quoted by respondents can

366 travel to both sectors and buy products in different areas. They usually have more

367 strong relationships with specific and few harvesters, but there is no official contract.

368 Even so we analyzed the supply chains separately.

369

370 Table 2. Metrics of supply chain for each area. Number of nodes (n) and links (L), links
371 per node (L/n) and connectance (L/n2) are simple metrics. SCI total (SCIT) and SCI
372 standardized (SCIst) are based on individual SCI estimated for each supply chain
373 element (for each level). Key elements are the top three elements with the higher SCI
374 individual score and are show in bold.

Links/nod Connectanc
Area Level SCI SCIT SCIst n L
e e (L/n2)
Processor 1 0.008
Processor 2 0.023
Distributor 1 0.017
Distributor 2 0.051 2 2
Ayapuá 0.280 0.01 1.14 0.05
Local community 0.010 1 4
Town 1 (Beruri) 0.092
Industry 0.010
City (Manaus) 0.069
Processor 1 0.0093
0.091
Processor 2
1
0.426
Distributor
0 1 1
Uauaçu 1.261 0.07 1.19 0.07
Local community 0.0015 6 9
0.728
Town 2 (Codajás)
1
Industry 0.0010
City (Manaus) 0.0036
375
376

377

378

81
379

380

381 Figure 2. Supply chains of Ayapuá and Uauaçu, showing the connectance between
382 elements, based on the flow of açai fruits and pulp. For both areas, elements are
383 represented by one element, but in fact collectors are a set of 11 and 9 elements,
384 processors are a set off 5 and 2 elements and distributors are a set of 2 and 1 element

82
385 in Ayapuá and Uauaçu, respectively. Purple circles indicate key elements (the top three
386 individual SCI scores). Number in boxes indicates the amount of açaí sold (in bags for
387 fruit and liters for açaí processed. The first box – from Collector to Processor indicate
388 the equivalent amount in liters, but the product is sold in bags) and the economic
389 multiplier (M). Figure produced by Eugenia C. Schmidt.
390

391 Effect of açai harvesting on household income

392 The model selection procedure selected seven variables to predict açaí individual

393 income, but only years of schooling and the position occupied in the supply chain,

394 specifically as a processor, had a positive effect on individual household income from

395 açai activity (Figure 3A). Household income from açai activity did not differ between

396 each level occupied in the supply chain (Kruskall-wallis: 4.6009, df = 3, p= 0.2035;

397 Figure 3B).

398

399

400 Figure 3. A) Coefficient estimates (±95% confidence intervals) showing the magnitude
401 and direction of effects for different predictors on açai household income in five
402 villages into the Sustainable Development Reserve Piagaçu Purus, Central Amazon.

83
 403 Confidence intervals crossing the zero vertical line are not significant. Blue lines
404 represent the predictors with positive and significant effect on the response variable.
405 B) Distribution of individual income from açai activity for each supply chain level in the
406 Sustainable Development Reserve Piagaçu Purus, Central Amazon.
407

408

409 Cost, net profit and açai fair trade

410 Cost of açai collection and processing is variable but, in general, is higher for people

411 who processes then only collects (Table 3). People responsible for collection and

412 processing works more days during the crop (Table 5). The net profit is extremely

413 variable as well, and higher for who is responsible for processing. It is important to

414 note that these people are not dependent only on açai. They work with several natural

415 resources (Brum et al., in preparation) and because of that, the negative net profit

416 does not mean, necessarily, a debt. The activity probably was realized with other

417 natural resources funds. Our sample included only 3 distributors, and they could not

418 provide the information about costs and days of work.

419

420 Table 3. Mean and standard deviation of the income per day, number of working days,
421 variable and total cost and net profit for açaí production for each level of the açai
422 supply chain during one crop.
1 2 3
Variable Cost Total Cost Income per day Daysofwork Net profit
(US$/day) (US$/day) (US$/day) (US$/day)
Mean SD Mean SD Mean SD Mean SD Mean SD
Collector 66.61 56.47 279.46 171.22 8.85 7.82 39.06 20.86 162.12 204.59
Collector-processor 278.62 267.87 2564.84 1998.87 8.40 8.20 60.00 13.15 23.83 359.23
Processor 1755.79 1513.18 2738.17 1342.06 44.50 9.50 50.00 33.04 -28.99 675.20
423 1
This metric was calculated considering the proportional cost (see methods).
424 2
This metric was calculated adding variable cost to fixed cost (as infrastructure like boat and/or mixer).

84
425 3
This metric was calculated subtracting the variable cost from the income provided by açai activity.
426

427 The cost of açai production is negatively affected by the fishing activity and positively

428 affected by days of work and the income provided by açai activity (Figure 4). Assuming

429 that açai activity should provide to local residents a daily income of, at least, US$ 12.88

430 (equivalent to R$50.00, which represents a common daily payment for services

431 between residents in the SDR-PP), and maintaining the average number of days of

432 work (Table 3), we suggest that a minimum price to accomplish a fair trade for açaí

433 harvesting would be US$ 18.74/bag. Maintaining the current average total revenue of

434 processors, they should sell the açai pulp for US$ 1.51/liter, instead of US$ 1.03/liter

435 (current average price). Some respondents informed that a good change in the açai

436 trade would be the commercialization in kilos, not in bags, as the amount of fruits (and

437 consequently, the weight) varies if the bags is more or less full. If they continued to

438 work, on average, 39-40 days/crop (Table 3), the fair trade for açai collectors would be

439 at least $0.34/kg (instead of the current price around $0.24/Kg), which represent an

440 increasing of 41% in the current price.

441

85
442

443 Figure 4. Coefficient estimates (±95% confidence intervals) showing the magnitude and
444 direction of effects for different predictors on açai production costs in five villages into
445 the Sustainable Development Reserve Piagaçu Purus, Central Amazon. Confidence
446 intervals crossing the zero vertical line are not significant. Blue lines represent the
447 predictors with positive and significant effect on the response variable.
448

449 Discussion

450 Considering the increasing of markets demand (zu Ermgassen et al. 2014) and climate

451 change, the assessment of supply chain resilience and the identification of its key

452 elements can be crucial to an advance planning and to provide the tools needed to

453 build adaptive capacity and decrease sensitivity to change (Lim-Camacho et al. 2017),

454 especially in tropical social-ecological systems, where local communities are often

455 under social and economic vulnerability (Brum et al., in preparation). Our results

456 elucidates the harvesting operation of a biodiversity asset with a huge cultural value

457 (Balick and Mendelsohn 1992; Mcdade et al. 2007) and great importance to food

86
458 security (Martinot et al. 2017), which can play a fundamental role in the achievement

459 of the goals of Sustainable Use Protected Area (Brasil 2000).

460 The açai supply chain has similar elements in both areas in the Lower Purus river

461 (Communities Ayapuá and Uauaçu), and the overall description of supply chain

462 elements is similar to other locations in Central Amazon (Martinot et al. 2017).

463 Evaluation of simple metrics such as number of links, nodes and connectance, showed

464 similar scores for both areas, while the comparison through weighted metrics such as

465 the supply chain index, as suggested by Plagányi et al (2014) indicated some distinct

466 features, which reinforces the importance in consider not only simple metrics to

467 supply chain assessment. The amount of product that flows through each element

468 indicated that, at least one distributor (or middlemen) in each area has a key role,

469 being basically the only connection to the urban center, which can have positive or

470 negative consequences to the supply chain resilience (Plagányi et al. 2014). Few

471 elements and connections can indicate stronger relationships, which can provide

472 resilience and flexibility to face changes in the systems (Plagányi et al. 2014). On the

473 other hand, more diffuse supply network, with many links, can also be efficient, as

474 provide different possibilities of connections (Nagurney 2010). In this communities in

475 Lower Purus River, few distributors can promote a lower harvesting pressure on

476 natural resources but, at the same time, it can limit commercial relationships and

477 competition, hampering the achievement of a fair trade and creating a strong

478 dependency on an external element. The distributor can take an advantage of the

479 relative isolation of these communities to influence products trade, paying a lower

87
480 price (Bayma et al. 2014), and promoting the “aviamento” system, which consist in

481 exchange of natural resources (fishes, Brazil-nut, cassava flour or açaí) for food and

482 other basic supplies from the city, in an extremely disadvantageous practice to local

483 residents (Silva 2010).

484 In Lower Purus River, the supply chains of different products are highly connected but,

485 in general, they are all dependent on fishing, which is the most important resource for

486 the Amazonian economy, but provide lower financial return to local residents,

487 comparing with forest resources (Brum et al., in preparation). The Supply Chain Index

488 (SCI) score does not indicate what section of the supply chain needs to be change

489 (Plagányi et al. 2014). Ayapuá communities presented a lower SCIT score, which can

490 indicate higher connectance and resilience to external shocks, while the higher SCIT of

491 Uauaçu means a higher dependency on few elements (Plagányi et al. 2014). In fact,

492 Uauaçu communities have a great proportion (around 85% of açai collected) that is

493 sold to Codajás town before being processed, which means that the most significant

494 added value is incorporated outside the Protected Area. To Ayapuá communities,

495 around 76% of açaí production is benefited inside the Protected Area, which means

496 that a great proportion of the added value is absorbed by local processors, before

497 being transported to urban centers, even not considering the processing costs.

498 Providing to local residents the possibility to be part of other levels of the açaí supply

499 chain through local associations or cooperatives can help them in achieve a fair trade

500 (Bayma et al. 2014). Even so, it is important to highlight that the processing stage has

501 the higher cost, and it demands a great concern about the quality of water used in the

88
502 process and all sanitary issues, important to be consider when dealing with food

503 products. The commercialization of NTFP can provide a “emergency net” besides a

504 “daily net” (Shackleton and Shackleton 2004) which can improve flexibility and

505 decrease social and economic vulnerabilities through diversification of activities

506 (Cinner and Bodin 2010).

507

508 Household income and costs

509 Positive outcomes related to the development of local supply chain of biodiversity

510 products involves the generation of household income and maintenance of local

511 subsistence (Marshall et al. 2003). Considering the positive effect of schooling in the

512 household income, a previous study in the area found that schooling is positively

513 related to the diversity of activities developed by locals, and the diversity of activity is

514 positively correlated to household income (Brum et al., in preparation). Formal

515 education is associated with several social benefits that can provide, besides the

516 learning of multiple skills, return in household income (Catalano et al. 2004; Reyes-

517 García et al. 2010). The diversification of activities with natural resources is especially

518 important in these areas, where seasonal flood pulse is responsible for complete

519 change in the landscape and the availability of natural resources along the year.

520 Therefore, people with a higher diversity of activities would present higher income, as

521 they can develop some activity along the whole year.

89
522 According to household income model, work as a processor is advantageous to

523 income, even with higher costs. The açai processing is the stage with the biggest

524 economic multiplier, indicating that this stage provides the higher economic benefit

525 along the supply chain, reaching more than three times the earning from the collection

526 stage. Even if the costs of processing the fruits are higher for processors, the income

527 generated by each level of the supply chain is extremely variable and does not vary

528 between then. The development of other natural resources activities does not affect

529 the cost of açai harvesting, which means that cost usually is not completed shared

530 between different activities.

531 Considering that açai is an important part of subsistence food consumption, the fair

532 price for açai producers should not avoid local residents of purchase the processed

533 product. Our suggestion in increasing the price for local collectors should promote an

534 increase in the price of processed pulp, as consequence. In this case, it could be

535 reasonable to think that açai pulp should have two different prices, one to local

536 consumers (in riverine communities) and other to distributors and consumers outside

537 the Protected Area. The CONAB (National Company of Provision - Companhia Nacional

538 de Abastecimento in Portuguese), a federal agency that supports, among others

539 products, the commercialization of non-timber forest products in Brazil, estimated the

540 minimum price of açaí in US$ 0.42/Kg(CONAB 2019). Therefore, the minimum price for

541 a bag of fruits, which has, on average, 54.5 Kg, should be US$22.89. In our analysis we

542 estimated the price for açai in US$ 0.34/Kg. And maintaining the current average total

90
543 revenue of processors, they should sell the açai pulp for US$ 1.83/liter for distributors

544 and external consumers.

545

546 Recommendations and policy implications

547 The Lower Purus River region is one of the most important area for açaí production of

548 Amazonas state and, because of the growing demand for açai in national and

549 international markets, it may need a support for locals to assist zoning of areas of use

550 and good practices to ensure a fair trade. Besides this, we proposed a set of three

551 recommendations in order to contribute to a more efficient and fair supply chain for

552 local residents. The first one represents a local demand from açai harvesters, and

553 consist in the commercialization of açai in kilos instead bags. It is important to ensure a

554 higher accuracy in estimating costs and revenues for collectors and processors. A

555 second recommendationwould be to promote an increasing of açai price, to

556 strengthen the benefits of local traders and distributors in order to encourage a

557 distribution network more specific to forest resources. The fact that distributors are

558 conducted to the Protected Area because of the profit fish market, their interest in

559 forest resources and consequently, their willing to pay a fair price is considerably

560 lower. The cultural value and the importance of açaí for food security and local

561 wellbeing lead us to our third recommendation, which is to encourage the structuring

562 of an açai co-management, involving collectors and local processors working together

563 to share the costs and benefits, and providing a fair distribution of açai benefits

564 between local residents and to external markets. It could be made through the

91
565 structuring of local associations or cooperatives in which residents can work together

566 with a clear division of functions, costs and profits.

567 There is not only one solution or general rule to protect biodiversity, but all actions

568 should consider the structuring of local governance (Berkes 2007) and the importance

569 of multiple agencies contribution (Berkes 2009). Tropical environments usually

570 combine rich biodiversity with social vulnerable people (Goklany 2007) in a complex

571 socioecological system. Social vulnerability is strongly associated with lower income

572 (Preston et al. 2014), but economic return is not the only aspect of poverty and basic

573 needs of developing country people (World Bank 2018). To conservation planning it is

574 necessary to consider socio-cultural values together with economic sustainability

575 (Laurila-Pant et al. 2015). Encouraging diversity of activities and the developing of

576 subsistence livelihoods can contribute to decrease poverty and improve local well-

577 being (Krishna 2006). Beyond that, local stakeholders’ engagement is key to ensure

578 efficiency (Campos-Silva and Peres 2016) and legitimacy on conservation actions

579 (Peralta 2012). The development of supply chain of socio-biodiversity products can

580 promote biodiversity conservation together with economic return and rescuing of

581 cultural pride and well-being. Açai can be one of the symbols of this process, as a

582 catalyzer of social transformation in riverine communities in the Amazon.

583

584 References

585 AMAZONAS 2005. Secretaria de Desenvolvimento Sustentável do Estado do Amazonas

586 (SDS): Cadeia produtiva do açaí no estado do Amazonas. MENEZES, M.

92
587 Antunes, A. P. et al. 2011. Rio, floresta e gente no baixo Purus: saber e uso da

588 biodiversidade na RDS Piagaçu Purus. - In: Santos, G. M. dos (ed), Album Purus.

589 EDUA, pp. 167–195.

590 Ayres, J. M. 1993. As matas de várzea do Mamirauá, médio rio Solimões.

591 Baldauf, C. et al. 2015. Moving from the ecological sustainability to the participatory

592 management of janaguba (Himatanthus drasticus; Apocynaceae). - In: Ecological

593 Sustainability for Non-timber Forest Products. in press.

594 Balick, M. J. and Mendelsohn, R. 1992. Assessing the Economic Value of Traditional

595 Rain Forests Medicines from Tropical. - Conserv. Biol. 6: 128–130.

596 Bayma, M. M. A. et al. 2014. Aspectos da cadeia produtiva da castanha-do-brasil no

597 estado do Acre, Brasil Brazil nut productive chain aspects in the state of Acre,

598 Brazil. - Bol. Mus. Para. Emílio Goeldi. Cienc. Nat 9: 417–426.

599 Berkes, F. 2007. Community-based conservation in a globalized world. - Proc. Natl.

600 Acad. Sci. 104: 15188–15193.

601 Berkes, F. 2009. Evolution of co-management: role of knowledge generation, bridging

602 organizations and social learning. - J. Environ. Manage. 90: 1692–702.

603 Berkes, F. and Folke, C. 1998. Linking social ecological systems: Management Practices

604 and Social Mechanisms for Building Resilience.

605 Brasil 2000. Lei Federal n 9.985 de 18/07/2000. - SNUC - Sist. Nac. unidades Conserv. in

606 press.

607 Brondízio, E. 2008. The Amazonian Caboclo and the Açaí palm: Forest Farmers in the

608 Global Market.

609 Bruner, A. G. et al. 2004. Financial Costs and Shortfalls of Managing and Expanding

93
610 Protected-Area Systems in Developing Countries. - Bioscience 54: 1119–1126.

611 Burnham, K. P. and Anderson, D. R. 2002. Model selection and multimodel inference.

612 Cámara-Leret, R. et al. 2019. Indigenous knowledge networks in the face of global

613 change. - Proc. Natl. Acad. Sci. 116: 9913–9918.

614 Campos-Silva, J. V. and Peres, C. A. 2016. Community-based management induces

615 rapid recovery of a high-value tropical freshwater fishery. - Sci. Rep. 6: 1–13.

616 Campos-Silva, J. V. et al. 2015. Policy reversals do not bode well for conservation in

617 Brazilian Amazonia. - Nat. e Conserv. 13: 193–195.

618 Campos-Silva, J. V. et al. 2018. Unintended multispecies co-benefits of an Amazonian

619 community-based conservation programme. - Nat. Sustain. 1: 650–656.

620 Castello, L. et al. 2009. Lessons from integrating fishers of arapaima in small-scale

621 fisheries management at the mamirauá reserve, amazon. - Environ. Manage. 43:

622 197–209.

623 Catalano, R. F. et al. 2004. Findings from the Social Development Research Group. - J.

624 Sch. Health 74: 252–261.

625 Cinner, J. E. and Bodin, Ö. 2010. Livelihood diversification in tropical coastal

626 communities: A network-based approach to analyzing “livelihood landscapes.” -

627 PLoS One in press.

628 Cinner, J. E. et al. 2012. Comanagement of coral reef social-ecological systems. - Proc.

629 Natl. Acad. Sci. 109: 5219–5222.

630 Clement, C. R. et al. 2015. The domestication of amazonia before european conquest. -

631 Proc. R. Soc. B Biol. Sci. in press.

632 CONAB 2019. Política de garantia de preços mínimos para produtos da

94
633 sociobiodiversidade. - Companhia Nacional do Abastecimento.

634 de Marques, A. A. B. et al. 2016. Human population and socioeconomic modulators of

635 conservation performance in 788 Amazonian and Atlantic Forest reserves. - PeerJ:

636 e2206.

637 Deus, C. P. et al. 2003. Piagaçu - Purus: Bases Científicas para a criação de uma Reserva

638 de Desenvolvimento Sustentável (IDSM, Ed.).

639 Dudley, N. (Ed) 2013. Guidelines for Applying Protected Area Management Categories

640 (IUCN, Ed.).

641 Dyck, A. J. and Sumaila, U. R. 2010. Economic impact of ocean fish populations in the

642 global fishery. - J. Bioeconomics 12: 227–243.

643 Goklany, I. M. 2007. Integrated strategies to reduce vulnerability and advance

644 adaptation, mitigation, and sustainable development. - Mitig. Adapt. Strateg.

645 Glob. Chang. 12: 755–786.

646 Hernandez, J. M. and Pedroza, C. 2016. The influence of the network topology on the

647 agility of a supply chain. - arXiv: 1–24.

648 Holling, C. S. 1978. Adaptive Environmental Assessment and Management (Pitman

649 Press, Ed.).

650 IBGE 2018. Produção da Extração Vegetal e da Silvicultura – PEVS. Acesso em

651 Oct/2018, in: www.ibge.gov.br.

652 IBGE 2019. IBGE.: www.ibge.gov.br.

653 Junk, W. J. et al. 1989. The flood pulse concept in river-floodplain systems. - Can. Spec.

654 Publ. Fish. Aquat. Sci. 106: 110–127.

655 Krishna, A. 2006. Pathways out of and into poverty in 36 villages of Andhra Pradesh,

95
656 India. - World Dev. 34: 271–288.

657 Laurila-Pant, M. et al. 2015. How to value biodiversity in environmental management?

658 - Ecol. Indic. 55: 1–11.

659 Leverington, F. et al. 2010. A global analysis of protected area management

660 effectiveness. - Environ. Manage. 46: 685–698.

661 Levis, C. et al. 2017. Persistent effects of pre-Columbian plant domestication on

662 Amazonian forest composition. - Science (80-. ). 931: 925–931.

663 Lim-Camacho, L. et al. 2017. Complex resource supply chains display higher resilience

664 to simulated climate shocks. - Glob. Environ. Chang. 46: 126–138.

665 Machovina, B. et al. 2015. Biodiversity conservation: The key is reducing meat

666 consumption. - Sci. Total Environ. 536: 419–431.

667 Marshall, E. et al. 2003. Commercialisation of non-timber forest products: first steps in

668 analysing the factors influencing success. - Int. For. Rev. 5: 128–137.

669 Martinot, J. F. et al. 2017. Coletar ou Cultivar : as escolhas dos produtores de açaí-da-

670 mata (Euterpe precatoria) do Amazonas. - RESR 55: 751–766.

671 Mcdade, T. W. et al. 2007. Ethnobotanical knowledge is associated with indices of child

672 health in the Bolivian Amazon. - PNAS in press.

673 Nagurney, A. 2010. Optimal supply chain network design and redesign at minimal total

674 cost and with demand satisfaction. - Int. J. Prod. Econ. 128: 200–208.

675 Naidoo, R. and Adamowicz, W. L. 2005. Biodiversity and nature-based tourism at forest

676 reserves in Uganda. - Environ. Dev. Econ. 10: 159–178.

677 Naidoo, R. et al. 2011. Namibia’s community-based natural resource management

678 programme: An unrecognized payments for ecosystem services scheme. -

96
679 Environ. Conserv. 38: 445–453.

680 Nepstad, D. C. and Mcgrath, D. G. 2002. Frontier Governance in Amazonia. - Science

681 (80-. ). in press.

682 Ostrom, E. 2007. A diagnostic approach for going beyond panaceas. - PNAS 104:

683 15181–15187.

684 Ostrom, E. et al. 2007. Going beyond panaceas. - PNAS 104: 15176–15178.

685 Pacheco-Palencia, L. A. et al. 2009. Phytochemical composition and thermal stability of

686 two commercial açai species, Euterpe oleracea and Euterpe precatoria. - Food

687 Chem. 115: 1199–1205.

688 Pedroza, C. 2013. Middlemen, informal trading and its linkages with IUU fishing

689 activities in the port of Progreso, Mexico. - Mar. Policy 39: 135–143.

690 Peralta, N. 2012. “Toda Ação De Conservação Precisa Ser Aceita Pela Sociedade”

691 Manejo Participativo Em Reserva De Desenvolvimento Sustentável.

692 Peters, C. M. et al. 1989. Oligarchic Forests of Economic Plants in Amazonia: Utilization

693 and Conservation of an Important Tropical Resource. - Conserv. Biol. 3: 341–349.

694 Plagányi, É. E. et al. 2014. A quantitative metric to identify critical elements within

695 seafood supply networks. - PLoS One in press.

696 Preston, I. et al. 2014. Climate change and social justice: an evidence review.

697 R Core Team 2017. R: A Language and Environment for Statistical Computing. in press.

698 Reyes-García, V. et al. 2010. Schooling and local environmental knowledge: Do they

699 complement or substitute each other? - Int. J. Educ. Dev. 30: 305–313.

700 Ronchail, J. et al. 2006. Impact of the Amazon tributaries on major flood in Óbidos. -

701 Clim. Var. Chang. - Hydrol. Impacts 2001: 220–225.

97
702 Satyamurty, P. et al. 2013. A quick look at the 2012 record flood in the Amazon Basin. -

703 Geophys. Res. Lett. 40: 1396–1401.

704 Schmidt, I. B. et al. 2015. Sustainability of golden grass flower stalk harvesting in the

705 Brazilian savanna. - In: Ecological Sustainability for Non-timber Forest Products. in

706 press.

707 Shackleton, C. and Shackleton, S. 2004. The importance of non-timber forest product

708 in rural livelihood security and as safety nets: a review of evidence from South

709 Africa. - S. Afr. J. Sci. 100: pp 658-664.

710 Silva, S. 2010. Estado e políticas públicas no mercado de castanha-dobrasil no Estado

711 do Acre: uma análise pela abordagem do desenvolvimento local. - IDeAS: 103–

712 128.

713 Sioli, H. 1985. Amazônia: Fundamentos da Ecologia da maior região de Florestas

714 Tropicais. - Stuttgart: 73p.

715 Somanathan, E. et al. 2009. Decentralization for cost-effective conservation. - Proc.

716 Natl. Acad. Sci. in press.

717 Specht, M. J. et al. 2019. Socioeconomic differences among resident, users and

718 neighbour populations of a protected area in the Brazilian dry forest. - J. Environ.

719 Manage. 232: 607–614.

720 Stoian, D. 2004. Todo lo que sube tiene que bajar : la economía del palmito (Euterpe

721 precatoria Mart.) en el norte amazónico de Bolivia. - In: Productos forestales,

722 medios de subsistencia y conservación: estudios de caso sobre sistemas de

723 manejo de productos forestales no maderables. Vol 3. pp. 117–140.

724 Sunderlin, W. D. et al. 2005. Livelihoods, forests, and conservation in developing

98
725 countries: An Overview. - World Dev. 33: 1383–1402.

726 ter Steege, H. et al. 2013. Hyperdominance in the Amazonian tree flora. - Science 342:

727 1243092.

728 Watson, J. E. M. et al. 2014. The performance and potential of protected areas. -

729 Nature 515: 67–73.

730 Willis, K. J. 2017. State of the World’s Plants 2017.

731 Wittmann, F. et al. 2004. The várzea forests in Amazonia: Flooding and the highly

732 dynamic geomorphology interact with natural forest succession. - For. Ecol.

733 Manage. 196: 199–212.

734 Wittmann, F. et al. 2010. Amazonian Floodplain Forests - Springer.

735 World Bank 2018. Poverty and shared prosperity 2018: Piecing together the poverty

736 puzzle. - Piecing Together the Poverty Puzzle

737 Yuyama, L. K. O. et al. 2011. Caracterização físico-química do suco de açaí de Euterpe

738 precatoria Mart. oriundo de diferentes ecossistemas amazônicos. - Acta Amaz. 41:

739 545–552.

740 zu Ermgassen, E. K. H. J. et al. 2014. Livestock: tackle demand and yields. - Nature 508:

741 186–186.

742

743

744

745

746

99
1

100
 2 Manuscrito submetido para revista Botany em Maio de 2019

4 Flooding disturbance and shade stress shape population structure

5 of açaí palm, the most abundant Amazon species
6

7 Authors

8 Heloisa Dantas Brum1*, Alexandre F. Souza2

9 1 - Programa de Pós-Graduação em Ecologia, CB, Universidade Federal do Rio Grande

10 do Norte, Campus Universitário, Lagoa Nova, Natal 59072-970, RN, Brazil. E-mail:
11 hdbrum@gmail.com

12

13 2 - Departamento de Ecologia, CB, Universidade Federal do Rio Grande do Norte,

14 Campus Universitário, Lagoa Nova, Natal 59072-970, RN, Brazil. E-mail:
15 alexsouza.cb.ufrn.br@gmail.com

16

17 *Corresponding author

18 Programa de Pós-Graduação em Ecologia, CB, Universidade Federal do Rio Grande do

19 Norte, Campus Universitário, Lagoa Nova, Natal 59072-970, RN, Brazil.

20 Telephone number: +55 (84) 99915-4336

21 hdbrum@gmail.com

101
22 Abstract

23 Euterpe precatoria (açaí) is the most abundant plant species in the Amazon basin and
24 one of the main non-timber forest products of the continent. A thorough understanding
25 of the ecology of this species is needed to support sustainable management initiatives.
26 Resource availability, disturbance regime, and human management are some of the
27 main factors influencing population structure. Here we described life stages of E.
28 precatoria, evaluated their allometric relationships, and assessed the effects of habitat
29 type (floodplain and unflooded upland) and proximity of human settlements on
30 population size distribution in Central Amazon near the Purus River. The
31 height:diameter relationship increased from Seedlings to Juvenile 2 but decreased from
32 Juvenile 2 to Reproductive 2, indicating changing height investment for any given
33 diameter along these life stages. There was a marked habitat dependency in both the
34 density and population size distribution, with populations in upland forests dominated
35 by juveniles while populations in the floodplains were dominated by reproductive
36 palms. Nearness of human settlements was not related to population structure
37 parameters. The patterns we uncovered have implications for our interpretation of
38 widespread Amazon forest species that inhabit habitats with contrasting disturbance
39 regimes and resource levels like flooded várzea and upland terra firme.

40

41 Key-words: Várzea, Terra firme, Population structure, Forest Management, Skewness,

42 Allometry.

43

44

45

102
46 Introduction

47 Euterpe precatoria Mart. is an iconic palm tree popularly known as açaí or assaí in South

48 America. It is the most abundant species in the Amazon basin, being one of the few

49 hyperdominant species in four of the broad five forest types in the region (ter Steege et

50 al. 2013). It is also one of the main non-timber forest products of the continent (Stoian

51 2004). The species is traditionally used for the extraction of fruits, heart of palm, and

52 timber by indigenous peoples (Anderson 1977; Albert and Tourneau 2007). It produces

53 large quantities of dark single-seeded fruits that, as in other Euterpe species, is

54 dispersed by many birds and mammals (Leite et al. 2012). The fruit is commonly

55 collected and a regional market developed around its use to produce a creamy

56 beverage known as ‘açaí wine’ that is very nutritious and bears superior antioxidant and

57 anti-inflammatory properties (Kang et al. 2012) that have been shown to bring health

58 benefits like memory protection when used as a dietary supplementation (Carey et al.

59 2017). Euterpe precatoria fruit consumption may be a sustainable alternative to this

60 species’ use in countries like Bolivia, where the traditional consumption of palm hearth

61 implies that palms are killed for resource extraction (Rocha 2004; Velarde and Moraes

62 2008). Most recently prices for E. precatoria products have increased dramatically due

63 to the global commodization of palm hearts and raw, processed, or lyophilized fruit pulp

64 as nutritional supplement that are increasingly consumed in Brazil as well as exported

65 to countries outside South America, mainly Europe, Canada and USA (Bussmann and

66 Zambrana 2012). Market sales for açaí berry (E. oleracea and E. precatoria combined)

67 amount USD 126.3 millon/yr in Brazil alone (Martinot et al. 2017). With increasing trade

103
68 volumes, E. precatoria and many other exploited native species cannot meet demand in

69 a sustainable manner (Mostacedo and Fredericksen 1999; Stoian 2004; Vallejo et al.

70 2016). A thorough understanding of the biology and ecology of this species is thus

71 needed, in order to support sustainable management initiatives (Velarde and Moraes

72 2008).

73 One of the most easily measurable population parameter in the field is population

74 structure, which integrates in itself a wealth of demographic information (Avalos et al.

75 2013; Peltzer et al. 2014). Yet, it may be a puzzling tool for the assessment of natural

76 populations, because of two reasons. The first is that the interpretation of population

77 structure may be easier if biologically relevant classes are clearly identified, e.g. through

78 the identification of life stages instead of arbitrary size classes (Gatsuk et al. 1980; Souza

79 et al. 2000, 2003; Caswell 2001). The second is that the patterns of relative abundances

80 of juveniles and adults they portray are no reliable proxy of future population growth

81 and, therefore, of population persistence in any given habitat or set of environmental or

82 human-imposed conditions (Johnson et al. 1994; Condit et al. 1998; Souza 2007; Virillo

83 et al. 2011; Bin et al. 2012). Attempts to infer population persistence from static

84 population structure and the relative abundance of juveniles relative to adults are,

85 therefore, flawed. In fact, population size structures result from the realized growth,

86 mortality, and fecundity rates across different size classes (Caswell 2001; Wright et al.

87 2003). A more fruitful use of population structure is the comparison of population size

88 distributions across habitats with contrasting disturbance histories or between co-

89 occurring species in the same habitat, which allows the detection of functional groups

104
 90 of species with similar life histories (Swaine et al. 1990; Poorter et al. 1996; Wright et al.

91 2003; Souza 2007; Souza et al. 2008, 2010). Following the fast-growing resource

92 acquiring vs slow growth stress tolerant life history gradient repeatedly found

93 worldwide (Grime and Pierce 2012; Reich 2014; Díaz et al. 2016), left-skewed size

94 distributions dominated by large individuals characterize chronic recruitment failure of

95 gap-dependent species with large fecundities, high seedling mortality, and high sapling

96 growth, while right-skewed size distributions dominated by small individuals

97 characterize shade-tolerant species with the opposite traits (Lorimer and Krug 1983;

98 Swaine et al. 1990; Poorter et al. 1996; Wright et al. 2003; Souza 2007; Souza et al.

99 2008; Vlam et al. 2014).

100 Among the factors that influence population structure are the effects of resource

101 availability, the disturbance regime, and human management (Mostacedo and

102 Fredericksen 1999; Souza and Martins 2004; Souza 2007; Avalos et al. 2013; Peltzer et

103 al. 2014). Habitat differentiation between várzea floodplains and upland terra firme are

104 among the main drivers of both light resource and disturbance levels in the Amazon

105 basin. The Amazon basin harbors a great variety of floodplains that cover nearly

106 500,000 Km2 (Junk et al. 2011) that includes a great variety of habitats such as

107 seasonally inundated forests and swamps, with plant communities adapted to differing

108 flooding regimes (Junk et al. 2012). These hydro-edaphic conditions exert strong control

109 on tree species distributions, connecting floodplain floras even when rivers drain

110 distinct climatic regions (Wittmann and Junk 2003). Environmental variation linked to

111 flooding regime affects demographic parameters of neighboring populations, and may

105
112 be detectable even over short geographic distances (Otárola and Avalos 2014). The

113 disturbance caused by seasonal flooding involves oxygen deprivation, sedimentation,

114 and mechanic damage that frequently kills younger plants (Parolin 2008). It has

115 triggered the evolution of metabolic pathways promoting either endurance or escape

116 strategies (Parolin 2008; Voesenek and Bailey-serres 2015). Contrary to flooded forests,

117 which present lower and more irregular canopies allowing more light penetration

118 (Souza and Martins 2005; Sawada et al. 2015)⁠, in uplands the main limitation to plant

119 growth and establishment is the deep shade cast by tall canopies, which severely

120 reduce understory light levels (Svenning 2001, 2002; Myster 2016).

121 The shady understory environments have induced the evolution of ecological strategies

122 that fall along the low growth/size shade stress tolerant – fast growth/height light

123 acquisition trade-off (Grime 1977; Forgiarini et al. 2013; Reich 2014). In palms, the lack

124 of secondary meristems precludes exploitation of light through lateral growth,

125 restricting possible plant responses to shading to increases in allometric height:

126 diameter increments (Avalos and Otárola 2010). The height growth of Euterpe

127 precatoria is supported by stilt roots whose size scales with plant height rather than

128 with topographic variation (Avalos and Otárola 2010), but the effects of habitat

129 variation between upland and floodplains on palm shape and allometry are poorly

130 known. Even less studied are the effects of human management on E.

131 precatoriapopulation ecology. These effects area potentially sizable because the species

132 is considered domesticated or semi-domesticated by indigenous people since pre-

133 Columbian times (Clement et al. 2015; Levis et al. 2017)⁠. The intensive management of

106
134 preferred species by indigenous and current locals has taken place mainly along the

135 floodplain margins of large rivers (McMichael et al. 2012), and is thus expected to

136 influence the abundance of the species across different habitat types (Ticktin et al.

137 2012; Baldauf and Maës dos Santos 2013).

138 Here we describe the macromorphological life stages of E. precatoria, evaluate their

139 allometric relationships, and assess the effects of habitat type (floodplain and

140 unflooded upland) and proximity of human settlements on population size distribution

141 in Central Amazon. Specifically, we tested the following hypotheses: 1) Due to the

142 higher and more closed canopies in upland than in floodplain forests (Myster 2016),

143 established palms should present taller and slenderer trunks in upland forests.

144 Therefore, there should be a significant interaction between height vs. diameter

145 allometric relationship and habitat; 2) Because E. precatoria forms seedling banks in

146 upland shaded understory (Peña-Claros and Zuidema 2000; Rocha 2004; Isaza et al.

147 2017), we infer that the species present a moderate degree of shade tolerance (Condit

148 et al. 1998; Mostacedo and Fredericksen 1999) that allows it to wait for canopy opening

149 increases (Avalos et al. 2013; Otárola and Avalos 2014). We thus expect right-skewed

150 size distributions in upland habitats where adult recruitment and productivity are light-

151 limited. Considering that in the floodplains flooding disturbs seedling recruitment by

152 killing seedlings (Parolin 2008), we also expect reduced size distribution skewness or

153 even symmetrical size distributions. This would reflect the prevalence of larger plants in

154 brighter and more productive floodplains, which would be more favorable to adult

155 growth and survivorship once the juvenile recruitment bottleneck is surpassed (Avalos

107
156 et al. 2013; Otárola and Avalos 2014). 3) Given the rationale above, we expect to find

157 greater seedling and juvenile densities in upland but greater adult densities in

158 floodplains (Rocha 2004; Velarde and Moraes 2008; Otárola and Avalos 2014); 4) Due to

159 negative effects of fruit harvesting on population recruitment (Mostacedo and

160 Fredericksen 1999), we expect that proximity to human settlements depress the

161 number of juveniles. Hence, we expect to find a positive relationship between the

162 distance from the nearest human settlement and the skewness of population size

163 distribution (where more positive skewness reflects increased juvenile abundance). The

164 outcomes of these tests can help stakeholders and local users to deal with management

165 decisions.

166

167 Methods

168 Study area and species

169 Euterpe precatoria is a single-stemmed palm (but see Avalos and Schneider 2011)that

170 may reach up to 20 m height occurring from Central America to Bolivia (Stoian 2004).

171 Individuals are obligate outcrossing monoecious and present a long flowering time that

172 provide easily accessed resources to thousands of insect flower visitors (Kuchmeister et

173 al. 1997). It’s large seeds (ca. 11 cm diameter Aguiar and Mendonça 2003) germinate

174 immediately after harvesting (Costa et al. 2018). The ability of E. precatoria to

175 germinate in darkness and its preference for relatively mild germination temperatures

176 (20ºC) may favor its establishment in seasonally flooded habitats, while its tolerance to

108
177 moderate desiccation (Costa et al. 2018) allows germination in upland forests. As show

178 for other Euterpe species, germination is likely facilitated by frugivorous dispersal (Leite

179 et al. 2012). Seedlings tolerate shade and do not respond promptly to increased

180 irradiance (Coelho et al. 2015). The species attains higher densities, sizes, and fruit

181 productivity in floodplains than in unflooded uplands, where it shows etiolation signs

182 (Velarde and Moraes 2008).

183 The study was carried out in the Piagaçu-Purus Sustainable Development Reserve (SDR-

184 PP), Amazonas State, located in the Purus-Madeira and Purus-Juruá interfluve (Figure

185 1). The reserve covers 834,245 ha and allows human habitation and the sustainable use

186 and commercialization of natural resources. It harbors 65 riverine communities and

187 more than 5000 people, who live mainly from family agriculture, fishing, hunting, and

188 the extraction of non-timber forest products like açaí. The SDR-PP is located in the

189 municipalities of Coari and Codajás, which together account for 96% of the 50,000 tons

190 of açaí produced by the Amazonas state annually (IBGE 2018). Açaí is collected in the

191 study region non-destructively by climbing. Although the Purus river presents up to 10

192 m annual level variation (ANA 2018), the várzea we studied is located at a high várzea

193 area, with seasonal flooding of up to 3 m height and during less than 50 days.year-1

194 (Wittmann et al. 2010).

195

109
196

197 Figure 1. Lower Purus river area in Central Amazon, showing the location of riverine
198 communities and the 10 blocks where individuals of E. precatoria (açaí) were marked
199 and measured.

200

201 Data collection

202 Sampling was directed at sites selected after a quick participatory mapping with local

203 residents, who indicated areas used for açaí fruit harvesting with differing intensities.

204 Data were collected in 200 20 x 20m permanent plots (total 8 ha) distributed in 10

205 blocks, five in the floodplain and five in the upland (Figure 1). In each block, plots were

206 arranged in four 100-m long transects, each one subdivided into five plots, and

207 stratified into two topographic positions. In each block, two transects were located in

208 low-lying terrain, thus subjected to more frequent and long-lasting flooding, and two

110
209 were located in elevated terrain, thus subjected to less frequent and short-lasting

210 flooding (approximately with one month of difference). Because the number of newly

211 germinated seedlings was very large in the floodplain plots, we subsampled this life

212 stage only through four 1 x 1 m subplots located at the corners of each 20 x 20m plot.

213 From February to March 2016, all individuals of E. precatoria were tagged and

214 measured for diameter at soil level, total height, and number of leaves in all plots.

215 Individuals in early development that lacked the aerial stem had diameter and height

216 measured at the base, which correspond to the group of leaf sheaths. For seedlings,

217 tags were tied to a wooden (on uplands) or metal (in the floodplains) stick, in order to

218 avoid damaging the plants. For plants with stem base enlargement or with aerial roots,

219 diameter measurement was made above these structures. Presence and height of

220 reproductive structures were also measured. We used the distance to the nearest

221 human settlement as a proxy to the effects of human activities (Duvall 2007; Vandam et

222 al. 2013; Sumarga 2017). Distance to the nearest human settlement was measured as

223 sum of linear distances actually traveled by land and water by the inhabitants of each

224 settlement to reach the location of each transect using a Garmin GPSmap 62s and the

225 QGIS software (QGIS Development Team 2009).

226 Statistical analysis

227 Euterpe precatoria life stages were established based on macromorphological traits

228 (Gatsuk et al. 1980). Further subdivisions within life stages were established based on

229 breakpoints in the relationship between stem diameter and height. This was assessed

230 through a loess regression between diameter and height after iterative adjustment of

111
231 alpha (α) and lambda (λ) parameters (Jacoby 2000). All statistical analyses were

232 performed in the open software R (R Core Team 2017). We performed an analysis of

233 covariance to assess the effect of life stages on the allometric relationship between

234 diameter and height. We used the natural logarithm of height to quantify palm size, and

235 the coefficient of skewness (g1, Zar 1996) to summarize the skewness of the height

236 distributions and, thus, population structure (Wright et al. 2003) at the transect (i.e.,

237 blocks of 5 20 x 20 m plots) scale (n = 40). When g1 > 0, it indicated right-skewed height

238 distributions with few tall and many short palms, g1 = 0 indicated symmetrical height

239 distributions, and g1 < 0 indicated left-skewed height distributions with many tall and

240 few short palms (Zar 1996). Seedlings were excluded from population structure

241 analyses because their short-lived duration, mainly in the floodplains, could distort the

242 results (personal observation).

243 We used Generalized Linear Models (GLM) to evaluate habitat (floodplain x upland) and

244 human (distance to the nearest human settlement) on the skewness of population

245 height distribution. Following Zuur et al. (2009) and Plant (2012), different models were

246 fitted in order to decide which model structure best described the error structure in the

247 data. In order to decide whether a random term was necessary, we used the second-

248 order Akaike Information Criterion (AICc, used for small sample sizes) and ANOVA to

249 compare models. Models whose ΔAIC < 2 were regarded as equally plausible model

250 (Burnham and Anderson 2002). We fitted a GLM without a random term (i.e.,

251 containing fixed terms only), a Generalized Linear Mixed Model (GLMM) using blocks as

252 a random intercept term, and a GLMM using both blocks and habitat types as random

112
253 terms (Zuur et al. 2009). All models were fit by maximizing the restricted log-likelihood

254 using the functions gls or lme of the nlme package (Pinheiro et al. 2017). The random

255 effect in the GLMM accounted for the possible lack of independence of the skewness

256 values estimated for the blocks. The 20 x20 m plot-level effects of habitat type and

257 distance to human settlement on palm density were assessed through a GLMM using

258 Poisson error and blocks and transects as nested random terms (Zuur et al. 2009).

259

260 Results

261 We marked and mapped 3613 individuals of açaí in different life stages in both the

262 floodplains and upland habitats. Observation of external macromorphological structures

263 allowed the distinction of two pre-reproductive stages: seedlings and juveniles.

264 Seedlings were defined as individuals without the presence of an aerial stem. Seedling

265 leaves present fan-shaped leaves (Figure 2). Juveniles were all individuals with an aerial

266 stem and diameter at soil level below the minimum diameter found for reproductive

267 palms, which was 10.0 cm (Figure 2). Juveniles presented a marked discontinuity in their

268 height vs. diameter relationship at ca. 1.5 m height (Figure 3A) and were thus

269 subdivided into Juvenile 1 (height < 1.5 m) and Juvenile 2 (1.5 ≤ height < 10 m). The

270 diameter x height relationship was nonlinear (loess regression: α = 0.35, λ = 2, Figure 3)

271 with two marked slope changes. The first change was located at the transition between

272 Juveniles 1 and 2, and the second among reproductive palms at ca. 23.0 cm diameter, a

273 threshold from which higher diameter values were not accompanied by higher height

113
274 values (Figure 3). Reproductive palms were thus divided into Reproductive 1 (height ≥10

275 m and diameter < 23 cm) and Reproductive 2 (height ≥ 10 m and diameter ≥ 23 cm).

276 The dependency of the relationship between diameter and height on life stages was

277 further confirmed through an ANCOVA (R2 = 0.99; F = 2.64× 104; df = 3598; P = 2.2 ×

278 10-16) (Figure 3B). On the one hand, the declivity of the linear fit increased from

279 Seedlings to Juvenile 1, and from these to Juvenile 2, indicating increasing heights for

280 any given diameter in these life stages. On the other hand, the same declivity decreased

281 from Juvenile 2 to Reproductive 1, and from these to Reproductive 2, indicating

282 decreasing height investment for any given diameter in these life stages. A second

283 ANCOVA tested for the effects of habitat type (upland x flooded forest) on the

284 height:diameter relationship, this time excluded both seedlings and Reproductive 2

285 palms because these two life stages did not occur in upland forests. Habitat type,

286 however, did not alter the height:diameter relationship significantly either as an

287 isolated factor (F = 2.66, df = 1, P = 0.103) or in interaction with life stages (F = 1.479, df

288 = 1, P = 0.2241).

289

114
290
291 Figure 2. Life stages of Euterpe precatoria (Arecaceae). A. Fruit bunch; B. Detail of fruits;
292 C. Detail of the seed with initial development of seedling; D. Seedling; E. Juvenile 1; F.
293 Reproductive. See text to description of stages characteristics. Drawing by Eliziane
294 Garcia.

295

296

297 Figure 3. A) Overall relationship between height and diameter of 3613 açaí palms
298 (Euterpe precatoria) measured in upland and floodplain forests, Central Amazon. The

115
299 fitted line is a Loess regression. B) Height × diameter relationship on a log scale with
300 fitted linear model adjusted by life stage through ANCOVA.

301

302 Seedlings were by far the denser life stage, with average 470.75 ± 486.21 plants ha-1 in

303 the unflooded areas and 4025.0 ± 5086.31 plants ha-1 in the floodplains. Seedlings

304 apart, overall palm density was greater in unflooded forests (66.88 ± 129.38 plants ha-1)

305 than in flooded forests (31.13 ± 79.65 plants ha-1). Palm density changed significantly

306 with habitat as well as with life stage. The best GLMM model included the interaction

307 between habitat and life stage, as well as blocks and transects as random terms (Table

308 1). This indicates that population density varied between transects within blocks as well

309 as between blocks in a random yet sizable way. The habitat vs.× life stage interaction

310 term indicates that the density of different stages varied between flooded and

311 unflooded forests in different ways. Unflooded forests showed to be dominated by

312 juveniles, with Juvenile 1 and Juvenile 2 palms having the highest densities. On the

313 contrary, flooded forests were dominated by Reproductive 1, which was the densest life

314 stage (Fig. 4). Reproductive 2 were scarce in flooded forests, but absent in unflooded

315 ones.

316

317

318

319

116
320

321 Table 1. Models generated to test the factors affecting the density of individuals of the
322 açaí population (Euterpe precatoria) in Central Amazonia.

Model Random DF AIC AIC

factor
1 Habitat * Life stage Block and transect 20 21699 -
2 Habitat + Life stage Block and transect 10 25314 3615
3 Habitat * Life stage Block 19 28198 6499
323

324

325

326

327 Figure 4. Average density of açaí (Euterpe precatoria) palms in unflooded terra firme
328 and flooded várzea in Central Amazon. Error bars are standard errors. J1 =Juvenile 1, J2

329 = Juvenile 2, R1 = Reproductive 1, R2 = Reproductive 2 And R3 = Reproductive 3.

330

331

117
332 Table 2. Generalized Linear Models used to investigate the influence of habitat type
333 (floodplain or unflooded) and distance to the nearest human settlement in explaining
334 variation in the symmetry of Euterpe precatoria height distribution in Central Amazon.
335 Model 4 control for spatial autocorrelation through blocks of plots taken as a random
336 factor. The AIC values were calculated only among the linear models, which can be
337 compared to each other. Models are presented in increasing AICc order. AIC values in
338 bold indicate statistically equivalent plausible models (ΔAIC < 2).

Model Fixed Factors Random DF AIC AIC

Factors
1 GLM Habitat 3 95.74 -
2 GLM Habitat + Distance 4 96.00 0.26
3 GLM Habitat * Distance 5 97.52 1.79
4 GLMM Habitat * Distance Block 6 99.52 -
339

340

341 Population structure as measured through the symmetry of height distribution did not

342 show spatial structure. The AIC of the mixed model (GLMM) using blocks of plots to

343 control for spatial effects on symmetry distribution was larger than the best model’s AIC

344 (Table 2). This indicates that size structure of populations of E. precatoria did not

345 present detectable spatial autocorrelation in the study region. The non-spatial GLM

346 models that included the effect of distance from the nearest human settlement were

347 undistinguishable from the simplest model containing the effect of habitat alone, which

348 also presented the lowest AIC. We can thus conclude that the habitat effects sufficed to

349 explain variation in the symmetry of height distribution, and that distance from the

350 nearest human settlement did not have meaningful explanatory power (Figure 5A).

351 Populations in unflooded forests presented right-skewed height distributions with

118
352 positive symmetries, while populations in flooded forests tended to present uniform or

353 negatively-skewed height distributions with negative symmetries (Figure 5B). Examples

354 of right-skewed and left-skewed populations are given in Figs. 5C and 5D.

355

356

357 Figure 5. Skewness of the population of açai (E. precatoria) in Central Amazon, in one
358 transect in (A)terra firme and 1 transect in(B) varzea; (C) the relationship between
359 skewness and distance from the nearest community (each point represents one
360 transect with five plot of 20 x 20m) and (D) relationship between skewness and habitat.
361 The notches represent the confidence interval around the median.

362

119
363 Discussion

364 We identified for the first time natural life stages of Euterpe precatoria, the most

365 abundant species in the Amazon basin (ter Steege et al. 2013). We registered changes

366 in the allometric relationships of the species along its maturation process and also

367 uncovered a marked habitat dependency in the density of different life stages. These

368 changes were accompanied by changes in population structure in response to flooding

369 regime but not to nearness of human settlements. The patterns we uncovered have

370 implications for our interpretation of widespread Amazon forest species that inhabit

371 habitats with contrasting disturbance regimes and resource levels like flooded várzeas

372 and upland terra firme.

373 The interpretation of population phenomena is facilitated if biologically relevant classes

374 of individuals are clearly identified, as through the identification of life stages instead of

375 arbitrary size classes (Gatsuk et al. 1980; Caswell 2001)⁠. The macromorphological life

376 stages of E. precatoriawe identified fit known developmental phases of palms

377 (Tomlinson and Jeffrey 1990; Souza et al. 2000, 2003). Similarly to Euterpe

378 edulis(Carvalho et al. 1999), but differently from other palm species (Souza et al. 2000,

379 2003), E. precatoria seedlings did not produce entire leaves but fan-shaped pinnatisect

380 leaf blades. The production of partitioned leaf blades by seedlings may reduce damage

381 by herbivores, whose effects are particularly harmful in moist and shaded environments

382 like the tropical forest floor (Carvalho et al. 1999; Baraza et al. 2004)⁠. Seedlings, which

383 lacked an aerial stem, presented a very low height:diameter allometric relationship. This

120
384 corresponded to the stem building phase described for E. precatoriaby Avalos and

385 Otárola (2010). In this phase intensive stem base growth takes place at the expense of

386 height growth, and secure stability for future investment in height increases (Tomlinson

387 and Jeffrey 1990; Kimura and Simbolon 2002)⁠. The increasing allometric relationship

388 that we found from seedlings until Reproductive 1 depicts increasing investment in

389 height growth given an overbuilt stem base, that is further stabilized through

390 adventitious roots. Tall palms are released from the biomechanical constrains

391 considered in traditional allometric models because of enlarged bases, but also because

392 they develop secondary thickening growth and have additional support provided by stilt

393 roots (Avalos and Otárola 2010)⁠. The height:diameter relationship is finally reduced in

394 Reproductive 2 adults, among which larger diameter values correspond to very small

395 height increases. Forest canopy height is likely a key factor mediating E.

396 precatoriaallometric relationships through its ontogeny.

397

398 The treefall gap and flood disturbance regimes

399

400 Contrary to our first hypothesis, we did not find a significant interaction between height

401 vs. diameter allometric relationship and habitat, and palms were not taller and did not

402 present slenderer trunks in upland forests. However, the hypotheses could not be

403 totally discarded because Reproductive 2 palms were only found in flooded forests.

404 Throughout the Amazon basin, flooded forests form a variety of distinct ecosystems

405 from neighboring upland forests (Junk et al. 2011, 2012)⁠. A common factor linking

121
406 different types of flooded forests, however, is lower and more irregular forest canopies

407 than upland forests, caused by flooding-induced restricted tree growth (Souza and

408 Martins 2005; Sawada et al. 2015). The restriction of Reproductive 2 palms, which did

409 not reach the largest heights reached by Reproductive 1, to the shorter flooded forests

410 indicates that part of the reproductive palms may have reached directly sun-lit upper

411 canopy and further height growth is not needed in order to avoid light competition with

412 other tree species. This agrees with Peña-Claros and Zuidema (2000), who found that E.

413 precatoriareached reproductive maturity at smaller sizes in Bolivian flooded forests

414 than in upland forests. This interpretation is also supported by results related to palm

415 density variation and size distribution. In agreement with findings in Bolivia (Stoian

416 2004)⁠, E. precatoriareached higher adult densities in flooded forests. This indicates that

417 the increased light availability in flooded forests sustain larger adult açaí populations, as

418 found in other tropical forests in which increased adult palm density has been found in

419 sites with increased canopy opening (Kahn 1987; Silva Matos et al 1999; Souza and

420 Martins 2006)⁠. Furthermore, in agreement with our second hypothesis, population size

421 distributions in floodplain forests were significantly more left-skewed than in upland

422 forests, indicating a prevalence of tall palms relative to small ones. Left-skewed

423 population structures indicate chronic regeneration failure resulting from low

424 survivorship and/or growth among younger individuals (Condit et al. 1998; Caswell

425 2001)⁠. Flooding is known to produce oxygen deprivation, sedimentation, and mechanic

426 damage that frequently kills younger plants (Parolin 2008; Voesenek and Bailey-serres

427 2015). Given their abundance and prevalence in flood-prone várzeas, adult E.

122
428 precatoriamust possess physiological and/or anatomical traits that enable them to

429 withstand and grow under the deep seasonal flooding of Amazon floodplains, which

430 may reach up to 10 m (Junk et al. 2011)⁠. Our study area, however, can be regarded as

431 high a varzea (sensuAyres 1993)⁠, with flooding limited to no longer than three months

432 and 2 m submersion (H. Brum, personal observation). It remains to be seen, thus what

433 are the effects of low várzea, with deeper and long-lasting flood on E.

434 precatoriapopulation structure.

435 Indeed, juveniles were nearly absent from flooded forests, which were dominated by

436 reproductive palms. Contrary of some species, whose seedlings may survive up to 300

437 days submerged (Parolin 2008)⁠, most of the dense aggregates of recently-germinated

438 seedlings we observed in the floodplains were wiped out shortly after by the rising

439 waters of the Purus river (H Brum, personal observation). The seasonal flooding affects

440 the dynamics of various plant species (Junk et al. 1989; Wittmann and Junk 2003;

441 Wittmann et al. 2004) and is probably a key factor in the population dynamics of E.

442 precatoria in the floodplains. Inter-annual variations in the flooding pulse are known to

443 occur, meaning that certain trees can miss being waterlogged for one or two

444 consecutive years if the water is untypically low (Parolin 2008)⁠. The establishment of

445 new individuals in flood-prone habitats seems thus to be dependent on the occurrence

446 of a recruitment window in dry years in which flooding fails to eliminate young palms.

447 Results from population density and height distribution indicated a different scenario

448 for populations on upland terra firme forests. There, populations of E.

123
449 precatoriaprobably suffer from light-limitation due to competition with tall trees

450 (Myster 2016)⁠. Seedling recruitment of E. precatoria is common in shaded understory,

451 where the species produces a persistent seedling bank (Rocha 2004)⁠. As we assumed in

452 the Introduction, E. precatoriamust present a moderate degree of shade tolerance

453 (Condit et al. 1998; Mostacedo and Fredericksen 1999) in order to withstand shade and

454 wait for canopy openings (Avalos et al. 2013; Otárola and Avalos 2014)⁠. Reduced

455 growth, as happens with young plants in shade, is one of the factors producing right-

456 skewed population structures, since individuals accumulate in early life stages (Condit et

457 al. 1998; Caswell 2001; Souza 2007). Both our density and population structure results

458 therefore agree with the interpretation in which E. precatoriasuffers from light

459 limitation in upland forests, accumulating individuals at early life stages that remain

460 collectively at distinct levels of suppression until light increases allow increased height

461 growth and the positioning of the crown in higher and better-lit canopy levels (Svenning

462 2001, 2002)⁠. A result of such recruitment bottleneck would be the low density of

463 reproductive palms in upland forests relative to palms in smaller life stages we

464 registered.

465 Distinct recruitment bottlenecks in upland and floodplain forests probably promote

466 distinct E. precatoriapopulations structures, but these do not inform us about

467 population growth or viability in either of these two habitat types. This is because

468 measures of population structure like size distribution are no reliable proxy of future

469 population growth and, therefore, of population persistence in any given habitat or set

470 of environmental or human-imposed conditions (Johnson et al. 1994; Condit et al. 1998;

124
471 Virillo et al. 2011; Bin et al. 2012). Otárola and Avalos (2014), suggested that source

472 populations of E. precatoriain more open-canopy favorable habitats with abundant

473 adults could sustain sink populations in closed-canopy upland forests. Hypotheses such

474 as this and the different disturbance regimes that we advanced above will have to wait

475 for studies on the population dynamics of the species in order for us to understand its

476 demographic patterns and long-term dependency of different kinds of disturbances to

477 maintain viable populations.

478 Contrary to our expectation, the distance to the nearest human settlement, used as a

479 proxy of management intensity (Ticktin et al. 2012; Baldauf and Maës dos Santos 2013)⁠,

480 had no effect on neither population density nor population structure. One possible

481 reason for that was that local people harvest and manage E. precatoriain the whole of

482 our study area in a shifting way (Levis et al. 2017)⁠, and we thus were unable to sample

483 truly unused populations. Other possible explanations are that environmental effects

484 overrun management differences or that these were subtle enough so as not to

485 produce detectable changes in the studied population parameters. Finally, the

486 comparison of population size distributions across habitats with contrasting disturbance

487 histories allows the detection of functional groups of species with similar life histories

488 (Swaine et al. 1990; Poorter et al. 1996; Wright et al. 2003; Souza 2007; Souza et al.

489 2008; de Souza et al. 2010)⁠. Considering the resource acquiring vs. stress tolerant life

490 history gradient (Grime and Pierce 2012; Reich 2014; Díaz et al. 2016)⁠, the ecological

491 strategy of E. precatoriacould tentatively be understood as intermediate and small-gap

492 specialist (Denslow 1980)⁠, since it can germinate in the shadow (Rocha 2004)⁠, presents

125
493 enough shade tolerance so as to produce seedling and juvenile banks, but needs

494 increased light resource levels to attain maturity and sustain population growth, what is

495 facilitated in open-canopy floodplains given recruitment windows (Lorimer and Krug

496 1983; Swaine et al. 1990; Poorter et al. 1996; Wright et al. 2003; Vlam et al. 2014)⁠.

497

498 Acknowledgement

499 The study was funded by Rufford Small Grant Foundation (Project number 15827-2) and

500 Conservation, Food and Health Foundation. H. Brum received scholarship from the

501 Fundação de Amparo a Pesquisa do Estado do Amazonas (FAPEAM) and Piagaçu

502 Institute supported field infrastructure. This research is registered in the Sistema

503 Nacional de Gestão do Patrimônio Genético e do Conhecimento Tradicional Associado

504 (SISGEN) under code A6817C1. For SDR-PP access we received autorization from

505 Departamento de Mudanças Climáticas e Gestão de Unidades de Conservação

506 (DEMUC)/Secretaria do Estado de Meio Ambiente (SDS). We thank to V. Vasquez for the

507 map and E. G. Oliveira for the drawing. We also wish to thank the field assistance of

508 Severino G. Brito, Mario P. Souza, Lucas Barbosa, Queven Moreira and Marcos Ramos

509 and all residents of SDR-PP for hospitality.

510

126
511 References

512 Aguiar, M.O., and Mendonça, M.S. de. 2003. Morfo-anatomia da semente de Euterpe

513 precatoria Mart. (Palmae). Rev. Bras. Sementes 25(1): 37–42. doi:10.1590/s0101-

514 31222003000100007.

515 Albert, B., and Tourneau, F. Le. 2007. Ethnogeography and Resource Use among the

516 Yanomami: toward a model of “reticular space.” Curr. Anthropol. 48(4): 584–592.

517 ANA. 2018. Agência Nacional da Águas - ANA. Sistema de informações hidrológicas.

518 Anderson, A.B. 1977. Os nomes e usos de palmeiras entre uma tribo de índios

519 Yanomama. Acta Amaz. 7(1): 5–13. doi:10.1590/1809-43921977071005.

520 Avalos, G., Fernandez Otarola, M., and Engeln, J.T. 2013. Successional stage,

521 fragmentation and exposure to extraction influence the population structure of

522 Euterpe precatoria (Arecaeae). Rev. Biol. Trop. 61(3): 1415–1424.

523 Avalos, G., and Otárola, M.F. 2010. Allometry and stilt root structure of the Neotropical

524 palm Euterpe Precatoria (Arecaceae) across sites and successional stages. Am. J.

525 Bot. 97(3): 388–394. doi:10.3732/ajb.0900149.

526 Avalos, G., and Schneider, R.A. 2011. Quantification of Ramet production in the

527 neotropical palm Euterpe Precatoria (Arecaceae) in Costa Rica. Ecotropica 17(2):

528 31–38.

529 Ayres, J.M. 1993. As matas de várzea do Mamirauá, médio rio Solimões.

127
530 Baldauf, C., and Maës dos Santos, F.A. 2013. Ethnobotany, Traditional Knowledge, and

531 Diachronic Changes in Non-Timber Forest Products Management: A Case Study of

532 Himatanthus drasticus (Apocynaceae) in the Brazilian Savanna1. Econ. Bot. 67(2):

533 110–120. doi:10.1007/s12231-013-9228-5.

534 Baraza, E., Gómez, J.M., Hódar, J.A., and Zamora, R. 2004. Herbivory has a greater

535 impact in shade than in sun: response of Quercus pyrenaica seedlings to

536 multifactorial environmental variation . Can. J. Bot. 82(3): 357–364.

537 doi:10.1139/b04-004.

538 Bin, Y., Ye, W., Muller-Landau, H.C., Wu, L., Lian, J., and Cao, H. 2012. Unimodal Tree

539 Size Distributions Possibly Result from Relatively Strong Conservatism in

540 Intermediate Size Classes. PLoS One 7(12). doi:10.1371/journal.pone.0052596.

541 Burnham, K.P., and Anderson, D.R. 2002. Model selection and multimodel inference.

542 doi:10.1016/j.ecolmodel.2003.11.004.

543 Bussmann, R.W., and Zambrana, N.Y.P. 2012. Facing global markets - Usage changes in

544 Western Amazonian plants: The example of Euterpe precatoria Mart. and E.

545 oleracea Mart. Acta Soc. Bot. Pol. 81(4): 257–261. doi:10.5586/asbp.2012.032.

546 Carey, A.N., Miller, M.G., Fisher, D.R., Bielinski, D.F., Gilman, C.K., Poulose, S.M., and

547 Shukitt-Hale, B. 2017. Dietary supplementation with the polyphenol-rich açaí pulps

548 (Euterpe oleracea Mart. and Euterpe precatoria Mart.) improves cognition in aged

549 rats and attenuates inflammatory signaling in BV-2 microglial cells. Nutr. Neurosci.

550 20(4): 238–245. Taylor & Francis. doi:10.1080/1028415X.2015.1115213.

128
551 Carvalho, R.M., Martins, F.R., and Santos, F. a M. 1999. Leaf ecology of pre-reproductive

552 ontogenetic stages of the palm tree Euterpe edulis Mart. (Arecaceae). Ann. Bot.

553 83(3): 225–233. doi:10.1006/anbo.1998.0810.

554 Caswell, H. 2001. Matrix population models: construction, analysis, and interpretation.

555 Edited ByS. Associates.

556 Clement, C.R., Denevan, W.M., Heckenberger, M.J., Junqueira, A.B., Neves, E.G.,

557 Teixeira, W.G., and Woods, W.I. 2015. The domestication of amazonia before

558 european conquest. Proc. R. Soc. B Biol. Sci. 282(1812).

559 doi:10.1098/rspb.2015.0813.

560 Coelho, C.F., Miranda, I., Melo, Z., and Barbosa, E. 2015. Physiological Behave of Acai

561 Seedlings (Euterpe precatoria MART.) under Increasing Levels of Irradiance. J.

562 Agric. Sci. 7(3): 236–242. doi:10.5539/jas.v7n3p236.

563 Condit, R., Sukumar, R., Hubbell, S.P., and Foster, R.B. 1998. Predicting population

564 trends from size distributions: a direct test in a tropical tree community. Am. Nat.

565 152(4): 495–509. doi:10.1086/286186.

566 Costa, C.R.X., Pivetta, K.F.L., de Souza, G.R.B., Mazzini-Guedes, R.B., Pereira, S.T.S., and

567 Nogueira, M.R. 2018. Effects of Temperature, Light and Seed Moisture Content on

568 Germination of Euterpe precatoria Palm. Am. J. Plant Sci. 09(01): 98–106.

569 doi:10.4236/ajps.2018.91009.

570 Denslow, J. 1980. Gap Partitioning among Tropical Rainforest Trees. Biotropica 12(2):

571 47–55.

129
572 Díaz, S., Kattge, J., Cornelissen, J.H.C., Wright, I.J., Lavorel, S., Dray, S., Reu, B., Kleyer,

573 M., Wirth, C., Colin Prentice, I., Garnier, E., Bönisch, G., Westoby, M., Poorter, H.,

574 Reich, P.B., Moles, A.T., Dickie, J., Gillison, A.N., Zanne, A.E., Chave, J., Joseph

575 Wright, S., Sheremet Ev, S.N., Jactel, H., Baraloto, C., Cerabolini, B., Pierce, S.,

576 Shipley, B., Kirkup, D., Casanoves, F., Joswig, J.S., Günther, A., Falczuk, V., Rüger, N.,

577 Mahecha, M.D., and Gorné, L.D. 2016. The global spectrum of plant form and

578 function. Nature 529(7585): 167–171. Nature Publishing Group.

579 doi:10.1038/nature16489.

580 Duvall, C.S. 2007. Human settlement and baobab distribution in south-western Mali. J.

581 Biogeogr. 34(11): 1947–1961. doi:10.1111/j.1365-2699.2007.01751.x.

582 Forgiarini, C., Souza, A.F., Longhi, S.J., and Oliveira, J.M. 2013. In the lack of extreme

583 pioneers: Trait relationships and ecological strategies of 66 subtropical tree

584 species. J. Plant Ecol. 8(4): 359–367. doi:10.1093/jpe/rtu028.

585 Gatsuk, L.E., Smirnova, O. V, Vorontzova, L.I., Zaugolnova, L.B., and Zhukova, L. a. 1980.

586 Age States of Plants of Various Growth Forms: A Review. J. Ecol. 68(2): 675.

587 doi:10.2307/2259429.

588 Grime, J.P. 1977. Evidence for the Existence of Three Primary Strategies in Plants and Its

589 Relevance to Ecological and Evolutionary Theory. Am. Nat. 111(982): 1169–1194.

590 Grime, J.P., and Pierce, S. 2012. The Evolutionary Strategies that Shape Ecosystems. In

591 The Evolutionary Strategies that Shape Ecosystems. doi:10.1002/9781118223246.

130
592 IBGE. 2018. Produção da Extração Vegetal e da Silvicultura – PEVS. Acesso em Oct/2018,

593 in: www.ibge.gov.br.

594 Isaza, C., Bernal, R., Galeano, G., and Martorell, C. 2017. Demography of Euterpe

595 precatoria and Mauritia flexuosa in the Amazon: application of integral projection

596 models for their harvest. Biotropica 49(5): 653–664. doi:10.1111/btp.12424.

597 Jacoby, W.G. 2000. Loess: A nonparametric, graphical tool for depicting relationships

598 between variables. Elect. Stud. 19(4): 577–613. doi:10.1016/S0261-

599 3794(99)00028-1.

600 Johnson, E.A., Miyanishi, K., and Kleb, H. 1994. The Hazards of Interpretation of Static

601 Age Structures as Shown by Stand Reconstructions in a Pinus Contorta -- Picea

602 Engelmannii Forest. J. Ecol. 82(4): 923–931. doi:10.2307/2261455.

603 Junk, W.J., Bayley, P.B., and Sparks, R.E. 1989. The flood pulse concept in river-

604 floodplain systems. doi:10.1371/journal.pone.0028909.

605 Junk, W.J., Piedade, M.T.F., Schöngart, J., Cohn-Haft, M., Adeney, J.M., and Wittmann, F.

606 2011. A classification of major naturally-occurring amazonian lowland wetlands.

607 Wetlands 31(4): 623–640. doi:10.1007/s13157-011-0190-7.

608 Junk, W.J., Piedade, M.T.F., Schöngart, J., and Wittmann, F. 2012. A classification of

609 major natural habitats of Amazonian white-water river floodplains (várzeas). Wetl.

610 Ecol. Manag. 20(6): 461–475. doi:10.1007/s11273-012-9268-0.

611 Kahn, F. 1987. The distribution of palms as a function of local topography in Amazonian

612 terra-firme forests. Experientia 43(3): 251–259. doi:10.1007/BF01945548.

131
613 Kang, J., Thakali, K.M., Xie, C., Kondo, M., Tong, Y., Ou, B., Jensen, G., Medina, M.B.,

614 Schauss, A.G., and Wu, X. 2012. Bioactivities of açaí (Euterpe precatoria Mart.) fruit

615 pulp, superior antioxidant and anti-inflammatory properties to Euterpe oleracea

616 Mart. Food Chem. 133(3): 671–677. Elsevier Ltd.

617 doi:10.1016/j.foodchem.2012.01.048.

618 Kimura, M., and Simbolon, H. 2002. Allometry and life history of a forest understory

619 palm Pinanga coronata (Arecaceae) on Mount Halimun, West Java. Ecol. Res. 17:

620 323–338.

621 Kuchmeister, H., Silberbauer-Gottsberger, I., and Gottsberger, G. 1997. Flowering,

622 pollination, nectar standing crop, and nectaries ofEuterpe precatoria (Arecaceae),

623 an Amazonian rain forest palm. Plant Syst. Evol. 206(1–4): 71–97.

624 doi:10.1007/bf00987942.

625 Leite, A. de B., Brancalion, P.H.S., Guevara, R., and Galetti, M. 2012. Differential seed

626 germination of a keystone palm (Euterpe edulis) dispersed by avian frugivores. J.

627 Trop. Ecol. 28(46): 615–618. doi:10.1017/S02.

628 Levis, C., Costa, F.R.C., Bongers, F., Peña-Claros, M., Clement, C.R., Junqueira, A.B.,

629 Neves, E.G., Tamanaha, E.K., Figueiredo, F.O.G., Salomão, R.P., Castilho, C. V.,

630 Magnusson, W.E., Phillips, O.L., Guevara, J.E., Sabatier, D., Molino, J.-F., Cardenas

631 Lopez, D., and ter Steege, H. 2017. Persistent effects of pre-Columbian plant

632 domestication on Amazonian forest composition. Science (80-. ). 931(March): 925–

633 931.

132
634 Lorimer, C.G., and Krug, A.G. 1983. Diameter Distributions in Even-aged Stands of

635 Shade-tolerant and Midtolerant Tree Species. Am. Midl. Nat. 109(2): 331–345.

636 Martinot, J.F., Pereira, H. dos S., and Silva, S.C.P. da. 2017. Coletar ou Cultivar : as

637 escolhas dos produtores de açaí-da-mata (Euterpe precatoria) do Amazonas. RESR

638 55(4): 751–766.

639 McMichael, C.H., Piperno, D.R., Bush, M.B., Silman, M.R., Zimmerman, A.R., Raczka,

640 M.F., and Lobato, L.C. 2012. Sparse pre-Columbian human habitation in Western

641 Amazonia. Science (80-. ). 336(6087): 1429–1431. doi:10.1126/science.1219982.

642 Mostacedo, C.B., and Fredericksen, T.S. 1999. Regeneration status of important tropical

643 forest tree species in Bolivia: Assessment and recommendations. For. Ecol.

644 Manage. 124(2–3): 263–273. doi:10.1016/S0378-1127(99)00076-6.

645 Myster, R.W. 2016. The Physical Structure of Forests in the Amazon Basin: a Review.

646 Bot. Rev. 82(4): 407–427. The Botanical Review. doi:10.1007/s12229-016-9174-x.

647 Otárola, M.F., and Avalos, G. 2014. Demographic variation across successional stages

648 and their effects on the population dynamics of the neotropical palm Euterpe

649 precatoria. Am. J. Bot. 101(6): 1023–1028. doi:10.3732/ajb.1400089.

650 Parolin, P. 2008. Submerged in darkness: Adaptations to prolonged submergence by

651 woody species of the Amazonian floodplains. Ann. Bot. 103(2): 359–376.

652 doi:10.1093/aob/mcn216.

133
653 Peltzer, D. a., Allen, R.B., Bellingham, P.J., Richardson, S.J., Wright, E.F., Knightbridge,

654 P.I., and Mason, N.W.H. 2014. Disentangling drivers of tree population size

655 distributions. For. Ecol. Manage. 331: 165–179. doi:10.1016/j.foreco.2014.06.037.

656 Peña-Claros, M., and Zuidema, P. 2000. Limitaciones demográficas para el aprovecha

657 miento sostenible de Euterpe precatoria para producción de palmito en dos tipos

658 de bosque de Bolivia. Ecol. en Boliv. 34(April): 7–25.

659 Pinheiro, J., Bates, D., DebRoy, S., and Sarkar. 2017. nlme: Linear and nonlinear mixed

660 effects models. Version 3.1-131. Available from https://cran.r-

661 project.org/package=nlme.

662 Plant, R. 2012. Spatial Data Analysis in Ecology and Agriculture Using R. In Spatial Data

663 Analysis in Ecology and Agriculture Using R. doi:10.1201/b11769.

664 Poorter, L., Bongers, F., and van Rompaey, R. 1996. ScienceDirect.com - Forest Ecology

665 and Management - Regeneration of canopy tree species at five sites in West

666 African moist forest. For. Ecol. … 84: 61–69. doi:10.1016/0378-1127(96)03736-X.

667 QGIS Development Team. 2009. QGIS Geographic Information System. Open Source

668 Geospatial Foundation.

669 R Core Team. 2017. R: A Language and Environment for Statistical Computing. R

670 Foundation for Statistical Computing, Vienna, Austria.

671 Reich, P.B. 2014. The world-wide “fast-slow” plant economics spectrum: A traits

672 manifesto. J. Ecol. 102(2): 275–301. doi:10.1111/1365-2745.12211.

134
673 Rocha, E. 2004. Potencial ecológico para o manejo de frutos de açaizeiro (Euterpe

674 precatoria Mart.) em áreas extrativistas no Acre, Brasil. Acta Amaz. 34(2): 237–250.

675 doi:10.1590/S0044-59672004000200012.

676 Sawada, Y., Suwa, R., Jindo, K., Endo, T., Oki, K., Sawada, H., Arai, E., Shimabukuro, Y.E.,

677 Celes, C.H.S., Campos, M.A.A., Higuchi, F.G., Lima, A.J.N., Higuchi, N., Kajimoto, T.,

678 and Ishizuka, M. 2015. A new 500-m resolution map of canopy height for Amazon

679 forest using spaceborne LiDAR and cloud-free MODIS imagery. Int. J. Appl. Earth

680 Obs. Geoinf. 43(April): 92–101. Elsevier B.V. doi:10.1016/j.jag.2015.04.003.

681 Silva Matos et al. 1999. The role of density dependence in the population dynamics of a

682 tropical palm.

683 Souza, A.F. 2007. Ecological interpretation of multiple population size structures in

684 trees: The case of Araucaria angustifolia in South America. Austral Ecol. 32(5): 524–

685 533. doi:10.1111/j.1442-9993.2007.01724.x.

686 Souza, A.F., Forgiarini, C., Longhi, S.J., and Brena, D. a. 2008. Regeneration patterns of a

687 long-lived dominant conifer and the effects of logging in southern South America.

688 Acta Oecologica 34(2): 221–232. doi:10.1016/j.actao.2008.05.013.

689 Souza, A.F., and Martins, F.R. 2004. Aspectos da dinâmica populacional de uma palmeira

690 clonal na floresta paludícola da Reserva Municipal de Santa Genebra (Campinas,

691 SP). Dep. Ecol. Doutorado: 164.

135
692 Souza, A.F., and Martins, F.R. 2005. Spatial variation and dynamics of flooding, canopy

693 openness, and structure in a Neotropical swamp forest. Plant Ecol. 180(2): 161–

694 173. doi:10.1007/s11258-004-7811-7.

695 Souza, A.F., and Martins, F.R. 2006. Demography of the clonal palm Geonoma

696 brevispatha in a Neotropical swamp forest. Austral Ecol. 31(7): 869–881.

697 doi:10.1111/j.1442-9993.2006.01650.x.

698 Souza, A.F., Martins, F.R., and Bernacci, L.C. 2003. Clonal growth and reproductive

699 strategies of the understory tropical palm Geonoma brevispatha : an ontogenetic

700 approach. Can. J. Bot. 112: 101–112. doi:10.1139/B03-002.

701 Souza, A.F., Martins, F.R., and Ecology, S.P. 2010. Spatial Distribution of an Undergrowth

702 Palm in Fragments of the Brazilian Atlantic Forest Stable URL :

703 http://www.jstor.org/stable/20146351 Spatial distribution Atlantic Forest of an

704 undergrowth palm in fragments of the Brazilian. 164(2): 141–155.

705 Souza, A.F., Martins, F.R., and Matos, D.M.S. 2000. Detecting ontogenetic stages of the

706 palm Attalea humilis in fragments of the Brazilian Atlantic forest. Can. J. Bot. 78(9):

707 1227–1237. doi:10.1139/cjb-78-9-1227.

708 de Souza, I.F., Souza, A.F., Pizo, M.A., and Ganade, G. 2010. Using tree population size

709 structures to assess the impacts of cattle grazing and eucalypts plantations in

710 subtropical South America. Biodivers. Conserv. 19(6): 1683–1698.

711 doi:10.1007/s10531-010-9796-y.

136
712 ter Steege, H., Pitman, N.C. a, Sabatier, D., Baraloto, C., Salomão, R.P., Guevara, J.E.,

713 Phillips, O.L., Castilho, C. V, Magnusson, W.E., Molino, J.-F., Monteagudo, A., Núñez

714 Vargas, P., Montero, J.C., Feldpausch, T.R., Coronado, E.N.H., Killeen, T.J.,

715 Mostacedo, B., Vasquez, R., Assis, R.L., Terborgh, J., Wittmann, F., Andrade, A.,

716 Laurance, W.F., Laurance, S.G.W., Marimon, B.S., Marimon, B.-H., Guimarães

717 Vieira, I.C., Amaral, I.L., Brienen, R., Castellanos, H., Cárdenas López, D.,

718 Duivenvoorden, J.F., Mogollón, H.F., Matos, F.D.D.A., Dávila, N., García-Villacorta,

719 R., Stevenson Diaz, P.R., Costa, F., Emilio, T., Levis, C., Schietti, J., Souza, P., Alonso,

720 A., Dallmeier, F., Montoya, A.J.D., Fernandez Piedade, M.T., Araujo-Murakami, A.,

721 Arroyo, L., Gribel, R., Fine, P.V. a, Peres, C. a, Toledo, M., Aymard C, G. a, Baker,

722 T.R., Cerón, C., Engel, J., Henkel, T.W., Maas, P., Petronelli, P., Stropp, J., Zartman,

723 C.E., Daly, D., Neill, D., Silveira, M., Paredes, M.R., Chave, J., Lima Filho, D.D.A.,

724 Jørgensen, P.M., Fuentes, A., Schöngart, J., Cornejo Valverde, F., Di Fiore, A.,

725 Jimenez, E.M., Peñuela Mora, M.C., Phillips, J.F., Rivas, G., van Andel, T.R., von

726 Hildebrand, P., Hoffman, B., Zent, E.L., Malhi, Y., Prieto, A., Rudas, A., Ruschell, A.R.,

727 Silva, N., Vos, V., Zent, S., Oliveira, A. a, Schutz, A.C., Gonzales, T., Trindade

728 Nascimento, M., Ramirez-Angulo, H., Sierra, R., Tirado, M., Umaña Medina, M.N.,

729 van der Heijden, G., Vela, C.I. a, Vilanova Torre, E., Vriesendorp, C., Wang, O.,

730 Young, K.R., Baider, C., Balslev, H., Ferreira, C., Mesones, I., Torres-Lezama, A.,

731 Urrego Giraldo, L.E., Zagt, R., Alexiades, M.N., Hernandez, L., Huamantupa-

732 Chuquimaco, I., Milliken, W., Palacios Cuenca, W., Pauletto, D., Valderrama

733 Sandoval, E., Valenzuela Gamarra, L., Dexter, K.G., Feeley, K., Lopez-Gonzalez, G.,

137
734 and Silman, M.R. 2013. Hyperdominance in the Amazonian tree flora. Science

735 342(6156): 1243092. doi:10.1126/science.1243092.

736 Stoian, D. 2004. Todo lo que sube tiene que bajar : la economía del palmito (Euterpe

737 precatoria Mart.) en el norte amazónico de Bolivia. In Productos forestales, medios

738 de subsistencia y conservación: estudios de caso sobre sistemas de manejo de

739 productos forestales no maderables, Vol 3. pp. 117–140.

740 Sumarga, E. 2017. Spatial indicators for human activities may explain the 2015 fire

741 hotspot distribution in central Kalimantan Indonesia. Trop. Conserv. Sci. 10.

742 doi:10.1177/1940082917706168.

743 Svenning, J.-C. 2001. On the role of microenvironmental heterogeneity in the ecology

744 and diversification of neotropical rain-forest palms (Arecaceae). Bot. Rev. 67(1): 1–

745 53. doi:10.1007/BF02857848.

746 Svenning, J. 2002. Crown Illumination Limits the Population Growth Rate of a

747 Neotropical Understorey Palm growth rate of a neotropical. Plant Ecol. 159(2):

748 185–199.

749 Swaine, M.D., Lieberman, D., and Hall, J.B. 1990. Structure and dynamics of a tropical

750 dry forest in Ghana. Vegetatio 88(1): 31–51. doi:10.1007/BF00032601.

751 Ticktin, T., Ganesan, R., Paramesha, M., and Setty, S. 2012. Disentangling the effects of

752 multiple anthropogenic drivers on the decline of two tropical dry forest trees. J.

753 Appl. Ecol. 49(4): 774–784. doi:10.1111/j.1365-2664.2012.02156.x.

138
754 Tomlinson, P.T., and Jeffrey, E.C. 1990. The Structural Biology of Palms. Clarendon Press

755 Oxford, Oxford.

756 Vallejo, M.I., Galeano, G., Valderrama, N., and Bernal, R. 2016. Consumers, the market

757 and the socio-ecological background of Euterpe oleracea palm heart production in

758 Colombia. Bot. J. Linn. Soc. 182(2): 526–535. doi:10.1111/boj.12451.

759 Vandam, R., Kaptijn, E., and Vanschoenwinkel, B. 2013. Disentangling the Spatio-

760 Environmental Drivers of Human Settlement: An Eigenvector Based Variation

761 Decomposition. PLoS One 8(7). doi:10.1371/journal.pone.0067726.

762 Velarde, M.J., and Moraes, M. 2008. Densidad de individuos adultos y producción de

763 frutos del asaí (Euterpe precatoria, Arecaceae) en Riberalta, Bolivia. Ecol. en Boliv.

764 43(2): 99–110.

765 Virillo, C.B., Martins, F.R., Tamashiro, J.Y., and Santos, F.A.M. dos. 2011. Is size structure

766 a good measure of future trends of plant populations? an empirical approach using

767 five woody species from the Cerrado (Brazilian savanna). Acta Bot. Brasilica 25(3):

768 593–600. doi:10.1590/s0102-33062011000300012.

769 Vlam, M., Baker, P.J., Bunyavejchewin, S., Mohren, G.M.J., and Zuidema, P.A. 2014.

770 Understanding recruitment failure in tropical tree species: Insights from a tree-ring

771 study. For. Ecol. Manage. 312: 108–116. Elsevier B.V.

772 doi:10.1016/j.foreco.2013.10.016.

773 Voesenek, L. a C.J., and Bailey-serres, J. 2015. Flood adaptive traits and processes : an

774 overview. New Phytol. 206: 57–73. doi:10.1111/nph.13209.

139
775 Wittmann, F., and et al. 2004. The várzea forests in Amazonia: Flooding and the highly

776 dynamic geomorphology interact with natural forest succession. For. Ecol.

777 Manage. 196(2–3): 199–212. doi:10.1016/j.foreco.2004.02.060.

778 Wittmann, F., and Junk, W.J. 2003. Sapling communities in Amazonian white-water

779 forests. J. Biogeogr. 30(10): 1533–1544. doi:10.1046/j.1365-2699.2003.00966.x.

780 Wittmann, F., Schöngart, J., and Junk, W.J. 2010. Amazonian Floodplain Forests -

781 Springer. doi:10.1007/978-90-481-8725-6.

782 Wright, S.J., Muller-landau, H.C., Condit, R., and Hubbell, S.P. 2003. Gap-Dependent

783 Recruitment , Realized Vital Rates , and Size Distributions of Tropical Trees. Ecology

784 84(12): 3174–3185.

785 Zar, J.H. 1996. Biostatistical Analysis. In 3ed editio. Edited ByU.S.R. Prentice Hall, Inc.

786 Zuur, A.F., Ieno, E.N., Walker, N.J., Saveliev, A.A., and Smith, G.M. 2009. Mixed effects

787 models and extensions in ecology with R. doi:10.1017/CBO9781107415324.004.

788
789

140
 CONCLUSÃO GERAL

O presente trabalho abordou aspecto

aspectoss sócio demográficos, econômicos e ecológicos a

respeito de uma das mais importantes espécies florestais da Amazônia. Num contexto

mais geral, identificamos fatores que aumentam a vulnerabilidade de populações

humanas, diminuindo flexibilidade e a sensibil

sensibilidade
idade às mudanças, a partir de

informações sobre uso e retorno econômico de recursos naturais florestais, aquáticos e

agrícolas em comunidades ribeirinhas no interior de uma Reserva de Desenvolvimento

Sustentável (RDS) na Amazônia Central. Os produtos flor

florestais
estais não-madeireiros
não

representam uma fundamental fonte de renda (Newton et al. 2016)sendo,

2016) para a

maioria dos moradores entrevistados, mais significativo do que a pesca, que é tida

como a atividade econômica mais importante da Amazônia

Amazônia(Santos
(Santos and Santos 2005;

Batista and Petrere 2015).O

.O açaí, a castanha
castanha-do-Brasil,
Brasil, e outros produtos florestais são

produtos secundários na cadeia produtiva da pesca, e os moradores locais dependem

da presença de um agente externo, o distribuidor, que além de ser responsável pela

compra de recursos naturais,

ais, também é quem faz uma série de prestação de serviços

importantes, como transporte de passageiros e encomendas e a venda de suprimentos

industrializados e medicamentos. Ainda que a RDS Piagaçu Purus seja um ponto focal

para a coleta de açaí do Amazonas por moradores do município de Codajás (capital do

açaí do Amazonas), é possível que a área ainda esteja muito aquém de seu potencial

produtivo de açaí. Análises da estrutura populacional do açai indicam que a espécie

responde de forma antagônica em relaçã

relaçãoo às características do ambiente, o que já pode

pressupor estratégias diferentes de manejo para as áreas de várzea e terra firme.

141
As Unidades de Conservação de Uso Sustentável são um modelo de gestão

desenvolvido para conciliar a conservação da biodiversi

biodiversidade
dade com bem-estar
bem

social(Brasil 2000),, o que exige que a criação e implementação da UC seja parte de um

processo participativo envolvendo múltiplos atores (Berkes 2007),, motivado por

interesses conservacionistas e demandas sociais legítimas

legítimas(Peralta
(Peralta 2012)
2012). O

funcionamento ideal de uma RDS muitas vezes é limitado por uma série de fato
fatores,

dentre eles a limitação de recursos financeiros (Watson et al. 2014; Campos-Silva

Campos et al.

2015) e os conflitos
onflitos existente
existentes entre os diferentes grupos com interesses divergentes

em relação ao uso dos recursos naturais. A renda doméstica é apenas um dos fatores

relacionados com vulnerabilidade social, mas em níveis muito baixos, pode ser o

principal fator a reduzir

eduzir flexibilidade e aumentar a sensibilidade às mudanças,

impedindo que as pessoas alcancem necessidades básicas (Reyes-García

García et al. 2016) e

qualidade de vida (Diener and Biswas

Biswas-Diener 2002; Easterlin 2003).. A associação entre

diferentes níveis de renda e a flexibilidade podem explicar algum

algumas
as decisões e reações

frente à situações de mudanças e crises (Marshall et al. 2013b).Uma

.Uma renda doméstica

elevada, frente à uma mudança iminente, pode levar as pessoas a não arriscar uma

nova atividade, ou empreender em um novo negócio, com medo de perder o que foi

conquistado (Marshall et al. 2013b)

2013b). Por outro lado,
do, uma baixa renda pode

proporcionar uma maior flexibilidade, pela sensação de que “não há nada a perder”,

permitindo a busca por algo novo e o desenvolvimento de novas atividades. É esperado

que cada realidade individual exija uma abordagem específica pa

para
ra ações de gestão e

conservação de recursos naturais (Berkes 2007; Ostrom et al. 2007) mas, ao mesmo

tempo, avaliando a realidade

dade de países em desenvolvimento, onde concentram
concentram-se

142
tanto a alta biodiversidade como a baixa renda em grande parte da população, é

evidente que a busca pela redução da pobreza pode ser considerada a meta

orientadora das ações (World Bank 2018)

2018). Dessa forma, faz-se
se necessário incorporar

nos processos de gestão e conservação a participação ativa da sociedade nas decisões,

o empoderamento
ramento das comunidades locais e o resgate cultural do conhecimento

tradicional.

A estruturação da cadeia produtiva através do co

co-manejo
manejo do açaí pode fortalecer a

geração de renda, garantindo uma valoração da floresta em pé e unindo conhecimentos

tradicionais
ais com o científico e garantindo a participação comunitária em todas as

etapas do processo, inclusive no desenvolvimento da pesquisa científica. A

continuidade dos estudos ecológicos pode fornecer informações mais precisas sobre a

variação da produtividadee de açaí entre diferentes áreas e ambientes, contribuindo

com o planeamento das coletas e zoneamento das áreas de uso. O monitoramento de

indivíduos de açaí em áreas de floresta nativa poderá fornecer informações valiosas

sobre a dinâmica populacional, pe

permitindo
rmitindo a estimativa de níveis sustentáveis de

exploração em cada ambiente, várzea e terra firme. A organização dos extrativistas em

cooperativas ou associações pode aumentar a eficiência em todas as etapas do

processo produtivo, desde a coleta até o proc

processamento,
essamento, com redução de custos e

aumento da confiabilidade na entrega de produção mínima, o que se torna um fator

importante no momento de estabelecimento de contratos e relações de fornecimento

de açaí para indústrias ou distribuidores em larga escala. O açaí como uma das espécies

mais importantes da Amazônia, pode ser o símbolo dessa transformação social tão

143
necessária para a realidade socioeconômica de populações dependentes de recursos

naturais na maior floresta tropical do planeta.

Referências

Batista, V. da S., and Petrere, M. 2015. Characterization of the commercial fish

production landed at Manaus, Amazonas State, Brazil. Acta Amaz. 33(1): 53–66.

doi:10.1590/1809-4392200331066.
4392200331066.

Berkes, F. 2007. Community

Community-based conservation in a globalized world. Proc. Natl. Acad.

Sci. 104(39): 15188–15193.

15193. doi:10.1073/pnas.0702098104.

Brasil. 2000. Lei Federal n 9.985 de 18/07/2000. SNUC - Sist. Nac. unidades Conserv.

Campos-Silva,
Silva, J.V., da Fonse
Fonseca
ca Junior, S.F., and da Silva Peres, C.A. 2015. Policy reversals

do not bode well for conservation in Brazilian Amazonia. Nat. e Conserv. 13(2):

193–195.
195. Associação Brasileira de Ciência Ecológica e Conservação.

doi:10.1016/j.ncon.2015.11.006.

Diener, E., and Biswas-Diener,

Diener, R. 2002. Wil money increase subjective well
well-being? In

Social Indicators Research. doi:10.1023/A.

Easterlin, R.A. 2003. Explaining happiness. PNAS 100(19):

(19): 11176
11176–11183.

doi:10.1007/s00534-009
009-0125-3.

Marshall, N.A., Tobin, R.C., Marshall, P.A., Gooch, M., and Hobday, A.J. 2013. Social

Vulnerability of Marine Resource Users to Extreme Weather Events. Ecosystems

16(5): 797–809.
809. doi:10.1007/s10021
doi:10.1007/s10021-013-9651-6.

Newton, P., Miller, D.C., Byenkya, M.A.A., and Agrawal, A. 2016. Who are forest
forest-

144
 dependent
ndent people? A taxonomy to aid livelihood and land use decision
decision-making in

forested regions. Land use policy 57: 388–395.

395. Elsevier Ltd.

doi:10.1016/j.landusepol.2016.05.032.

Ostrom, E., Janssen, M.A., and Anderies, J.M. 2007. Going beyond panaceas. PNAS

104(39): 15176–15178.
15178.

Peralta, N. 2012. “Toda Ação De Conservação Precisa Ser Aceita Pela Sociedade”

Manejo Participativo Em Reserva De Desenvolvimento Sustentável. Belo Horizonte,

MG: UFMG.

Reyes-García,
García, V., Babigumira, R., Pyhälä, A., Wunder, S., Zorondo
Zorondo-Rodríguez,
odríguez, F., and

Angelsen, A. 2016. Subjective Wellbeing and Income: Empirical Patterns in the

Rural Developing World. J. Happiness Stud. 17(2): 773–791.

791. doi:10.1007/s10902-
doi:10.1007/s10902

014-9608-2.

Santos, G.M. dos, and Santos, A.C.M. dos. 2005. Sustentabilidade da pe

pesca na

Amazônia. Estud. Avançados 19(54): 165–182.

Watson, J.E.M., Dudley, N., Segan, D.B., and Hockings, M. 2014. The performance and

potential of protected areas. Nature 515(7525): 67–73.

73. doi:10.1038/nature13947.

145
 ANEXOS

Roteiro de entrevistas

QUESTIONÁRIO SOCIOECONÔMICO (PARA TODOS)

Entrevistador:___________________ Comunidade:___________________

Data: _______

INFORMAÇÕES PESSOAIS
Nome:__________________________________________________________

Idade: sexo (M) (F) Anos de estudo:

Estado civil: Numero de filhos:

Onde nasceu?

Há quanto tempo mora aqui?

Antes morava onde?

Quantos familiares moram na casa? Mulheres:______ Homens:______

Qual sua principal atividade de renda/profissão?

Quanto ganha (mês/semana/dia) com este trabalho?

Quais as demais atividades de renda/profissão?

Quanto ganha (mês/semana/dia) com este trabalho? (para cada atividade listada)

Trabalha com açaí?

Quando começou?
Que atividade desenvolve (coleta, beneficiamento, comércio, etc)?
Quanto tempo por ano trabalha com açaí?
Alguém da família trabalha junto? Se sim, em que parte?

146
(Nível de acessibilidade/abertura
acessibilidade/abertura)

Trabalha hoje no Instituto Piagaçu Trabalhou Instituto Piagaçu

Em qual atividade: Em qual atividade:
Há quanto tempo: Por quanto tempo:
Desde quando: Durante quais anos:
Porque foi trabalhar com o IPi: Porque foi trabalhar no IPi:
Remuneração: Remuneração:
Achou bom, ruim ou indiferente Achou bom, ruim ou indiferente

Participa de alguma atividade de manejo? ( ) não

( ) sim. Qual:__________________________________ Há quanto
tempo:______________
Por que foi trabalhar nisto?
Qual a remuneração?

147
COLETOR – quem sobe na árvo
árvore
Nome:

Informações específicas sobre o açaí

Coleta por contrato com
2016 2017
beneficiador/distribuidor
Contratante
Quanto paga
De quanto tempo é o contrato
Quanto recebeu
Obs

Vende fruto sem contrato 2016 2017

A quem vende
Por quanto vende
Quanto vende/safra
Quanto recebeu
Obs

Vende fruto batido 2016 2017

A quem vende
Por quanto vende
Quanto vende/safra
Tem batedeira
Gasto de energia/semana
Quanto tempo gasta batendo
açaí - consumo
Quanto tempo gasta batendo
açaí

Para consumo 2016 2017

Quanto coleta por semana
para consumo
Quanto consome com amigos
e vizinhos
Quanto consome com a
família
Quanto doapara outros
Obs

Qual a época de coleta do açaí?

148
A época muda dependendo da cheia ou da seca?
Coleta em área de ( ) várzea ( ) de terra firme ( ) nas duas?
Onde coleta mais?
Você tem quantos locais de coleta? Quais?
Trabalha sozinho?
Outras pessoas coletam na mesma área? ( ) não ( ) sim
Se sim, de onde são? ( ) da mesma comunidade ( ) de outra(s):
Quanto tempo por dia você passa coletando açaí?
Quantos cachos você coleta em um dia?
Quantos cachos precisa para encher uma saca?
Quantos litros rende uma saca de fruto? (Quantos Kg)
Como/onde você transporta os frut
frutos coletados?
Como armazena?
Quanto do total apanhado que perde: em transporte: ____________
No armazenamento: _____________
Quais as características do melhor fruto para venda?
Quais materiais que você precisa para coletar açaí e quanto custa cada um?

Material Preço Quanto tempo dura Quanto compra (mês/ano)

Corda (m)
Saca (?)
Terçado (un.)
Despolpadeira
Canoa
Gasolina
Bota

Há conflito com outras pessoas nas áreas em que você coleta do açaí ou em outras
áreas?
Desde quando há conflito? Com quem? Por que? Como você acha que poderia ser
resolvido?

149
Alguém zela (cuida) das áreas em que tem conflito?
E nas áreas sem conflito_ há alguém que cuide da área?
Alguma destas áreas tem manejo? Em qual área?
Qual o tipo de manejo
Quem realiza/coordena?
Quais as maiores dificuldades do trabalho com açaí?
O que ajudaria a melhorar a atividade?

Como é o processo de coleta? Descrição de um dia de trabalho (logística, condição

climática, época, material, cuidados com segurança, etc)

150
BENEFICIADOR

Quando começa a época de beneficiamento do açaí?

Como é feito o beneficiamento? (mecanizado ou manual, descrever o processo)
Quantas horas por dia você passa hoje beneficiando o açaí?
Por quantas horas no dia a batedeira funciona?
Se tem despolpadeira: quantas tem: ________ Quando adquiriu a primeira:________
Qual a potência? Quantas horas fica ligada por dia?
Antes de ter a despolpadeira, trabalhava com açaí? ( ) não ( ) sim: o que fazia?
(vendia, coletava, comprava e repassava
repassava,, outro:_________________________)
Hoje, vende os frutos ou apenas o açaí batido?
Qual a quantidade que vende de vinho por mês durante a safra?
Quantos litros (polpa grossa) rende uma saca de fruto (quantos Kg de açaí)?
Qual a quantidade que vende de fr
fruto?
Vende a produção de açaí para quem?
Seu comprador é fixo? (sempre o mesmo)
Qual o valor da polpa/suco:
Qual o valor do fruto:
Quem põe o preço no açaí sendo vendido: ( ) você; ou ( ) o comprador?
A qualidade do fruto traz diferença no preço do fru
fruto
to e da polpa? De quanto?
O que determina a qualidade do fruto/suco?
Quanto vendeu nos últimos 2 anos?
Açaí é armazenado em polpa apenas? ( )sim ( )não:___________
Onde você armazena o açaí: ( ) freezer/geladeira só para isto
( ) Freezer/geladeira da família
O açaí que chega eé beneficiado, normalmente fica armazenado com você por quanto
tempo até ser vendido?
No final da estação, sobra açaí para venda que você guarda? ( ) não ( ) sim: quanto
tempo leva até vender tudo?
Como você
cê consegue o açaí que beneficia:
( ) a família coleta ( ) o próprio beneficiador coleta

151
 ( ) compra dos comunitários ( ) compra de distribuidor
( ) contrata pessoas para fazer a coleta ( ) outro:
Secompra, de quem compra
ra (quantos são) e quanto paga?
Se contrata, quantas pessoas contrata por safra para coletar o açaí para você?
Contrata sempre as mesmas pessoas? Por que?
Faz ( ) contrato; ( ) pagamento diário por saca ( ) pagamento diário por cachos (
) ( ) paga semanal por produção (especificar o que é produção) outro:____
O pagamento é ( ) diário; ( ) semanal; ( ) mensal
Como faz o pagamento: ( ) em dinheiro ( ) em açaí
( ) com outros produtos. Quais?
( ) faz divisão de benefícios ( ) outro
Se há divisão de benefícios, como ocorre?
Os coletores auxiliam em outra parte do processo? Qual? Recebem por isso?
No beneficiamento do açaí, que serviços são feitos:
por homens:
quais são feitos por mulheres:
Quanto paga para:
Coletores –
Operadores de despolpadeira –
Embalador –
Limpeza dos frutos –

Quanto gasta de energia elétrica/mês durante a safra do açaí?_______ E fora da safra?

Tem gerador próprio? ( ) Sim ( ) Não
Se sim, quanto gasta de gasolina por dia durante o trabalho com açaí?
Quanto pagou pelo gerador?
Quanto gasta com manutenção?
Gastos da comunidade com gerador:
Quantas horas fica ligado?
Quanto gasta (por dia/mês)?
Qual a potência do gerador?

152
Quanto de embalagem gasta durante a safra?
Quanto custa a embalagem?
Tem
em geladeira para armazenar o açaí ou compra gelo?
Quanto de gelo compra por dia durante a safra do açaí?
Quanto custa o gelo?
Vende a produção na comunidade? ( ) sim. Compradores:
Leva produção para venda em outros locais? Quais?
Tem custo de transporte? De quanto/semana?
Quantos intermediários você tem que vem aqui e buscam o açaí para revender?
Quem são seus intermediários? São sempre os mesmos?
De onde vem a água para lavar o açaí?
Precisa de algum cuidado a mais para usar a água para lavar os frutos?
Sabe se existe alguma doença que pode ser transmitida com o consumo do açaí?
Já ouviu falar no barbeiro? e o que ele pode causar?

153
INTERMEDIÁRIO/DISTRIBUIDOR
Nome: Idade: Origem:

Quais comunidades visita durante a safra para comprar açaí?

Vende produtos
odutos nas comunidades? O que? Por quanto?

O que compra em cada uma delas?

Qual o produto mais importante?

Quanto de açaí compra por safra?

Que porcentagem deste total é de ( ) fruto; e que % é de ( ) açaí batido?

Quanto paga pelo litro/saca/kg do açaí em fruto?

Quanto paga pelo litro/saca/kg do açaí batido?

Compra sempre das mesmas pessoas ou de quem estiver vendendo?

Há algum controle de qualidade do açaí? Qual?

Liste seus fornecedores/comunidade:

Tem algum tipo de contrato ou acerto?

Se sim, como e quanto paga por pessoa?

Vende para quem?

Por quanto?

Onde vende?

Quanto vende de açaí por safra?

Quanto a safra do açaí te rende por mês/semana?

Tem preço diferente para cada comprador? Descrever.

Se tivesse mais açaí para compra, conseguiria vender mais?

Quanto gasta por dia por viagem para compra de açaí?

Qual o mínimo que precisa comprar por viagem para compensar a viagem?

154
Faz quantas viagens por mês/ano para safra do açaí?

Tem funcionários? Quantos? Quanto paga/mês para cada um?

Quanto vendeu de açaí esse ano?

Quanto vendeu no último ano? Por quanto?

Tem alguma perda? Açaí que estraga, por exemplo. Porque?

Qual sua maior fonte de renda anual?

Como era no passado?

- Antes da criação da RDS
RDS-PP (pré-2003)
- Fundação da comunidade (verificar a data)

155