Universidade de Brasília

Instituto de Ciências Biológicas

Programa de Pós-Graduação em Ecologia

Implicações e aprendizados do manejo integrado do fogo no

Cerrado: estudo de caso no Parque Nacional da Chapada das
Mesas (PNCM)

Lívia Carvalho Moura

Brasília- DF
2018
 Universidade de Brasília
Instituto de Ciências Biológicas
Programa de Pós-Graduação em Ecologia

Implicações e aprendizados do manejo integrado do fogo no

Cerrado: estudo de caso no Parque Nacional da Chapada das
Mesas (PNCM)

Aluna: Lívia Carvalho Moura

Orientador: Aldicir Osni Scariot

Tese apresentada ao Programa de Pós

Graduação em Ecologia da Universidade
de Brasília, como pré-requisito para
obtenção do título de Doutora em Ecologia

Brasília
2018
 Lívia Carvalho Moura

Implicações e aprendizados do manejo integrado do fogo

no Cerrado: estudo de caso no Parque Nacional da
Chapada das Mesas (PNCM)

Banca examinadora:

Orientador: Prof. Dr. Aldicir Osni Scariot

Presidente
Embrapa/Cenargen

Prof. Dr. Marcelo F. Simon Prof. Dr. Vínicius L. Dantas

Membro titular Membro titular
Embrapa/Cenargen IG/UFU

Prof. Dr. Heloísa S. Miranda Prof. Dr. Cássia B. Munhoz

Membro titular Membro suplente
IB/UnB IB/UnB

Brasília, DF
Junho de 2018

iii
"Não é o mais forte que sobrevive,
nem o mais inteligente, mas o que
melhor se adapta às mudanças."
Charles Darwin

iv
 Agradecimentos

Agradeço ao Programa de Pós-Graduação em Ecologia, Departamento de Ecologia e

Laboratório de Ecologia Vegetal da UnB por me dar a oportunidade de desenvolver esta tese
de pesquisa e por todos os aprendizados e infraestrutura.

Àqueles que financiaram este trabalho e a apresentação desta pesquisa em eventos

científicos: Capes e CNPq (GD140118/2017-9), DPP/UnB (Edital Mobilidade n. 01/2015),
DIBIO/ICMBio (Projeto PNUD-BRA/08/023), GIZ/FINATEC (83198593), FAP/DF
(0193.000563/2016) e FAPESP (2015/06743-0).

Esta pesquisa foi realizada com o apoio e esforço de muitas pessoas, que acreditaram e se
desempenharam para desenvolver esta tese.

Agradeço à minha mãe (Ivone Carvalho), eterna fonte de inspiração e dedicação; ao meu pai
(Ronaldo Moura) por todo amparo e apoio técnico; à minha amada avó Isaura Carvalho por
todo carinho e acolhida; aos meus primos Eduardo Carvalho, Carolina Carvalho e minhas tias
(Ione C. Pinto e Rose Teodoro) por todo suporte e paciência. E à todos os meus familiares
que sempre foram a minha base, sem vocês eu não teria chegado até aqui.

À professora Isabel Schmidt por ter acreditado no meu sonho e ter investido muita energia
em colocá-lo em prática. Agradeço-a em especial pela paciência, disposição e ensinamentos.

Ao meu orientador (Aldicir Scariot) por sempre estar disponível e ter aceitado fazer parte
dessa empreitada.

Aos professores e pesquisadores Alessandra Fidelis, Heloísa Miranda, Vânia Pivello,

Ludivine Eloy, Ricardo Machado, John Hay, James Welch e Clay Trauernicht por todas as
ideias, inspirações e orientações para que este trabalho florescesse.

Às Instituições ambientais ICMBio, IBAMA, MMA e à GIZ através dos colaboradores Paulo
Adriano Dias, Christian Berlinck, Kátia Ribeiro, Alexandre Sampaio, Anja Hoffmann,
Philipp Buss, Marcelo Segalerba, Eliana Jorge, Robin Beatty, Mariovaldo Santana, Maria
Carolina Camargo, Ângela Garda por toda ajuda e apoio logístico.

Aos funcionários do PNCM e às Brigadas de Incêndio dos anos 2014 à 2017, em especial ao
Edvan Barros, seu Ântonio, Jordson, Deijacy Rego, Thais Souza, Christiane Silva, Ronan

v
Brito, Renato, seu Raimundo, Edmar e Manoel por terem me ajudado na execução dos
trabalhos de campo.

Aos moradores do PNCM por toda colaboração, participação, interesse e acolhida.

Aos técnicos, voluntários e colaboradores Clara Baringo, Maxmiller Ferreira, Naomi Sato,
Leandro Garcia, Daniel Chaves, Fernando Murad, André Coutinho, Waira, Carla Faleiro,
Alessandro Moreira, Ana Carla, Samuel Montenegro, Fabiana Souza, Ana Clara, Ana
Beatriz, Victor Ferreira, Gabriella Damasceno e Yuri Feitosa por toda ajuda, amizade, risadas
e paciência ao longo dessa jornada.

Ao Jeremy Russell-Smith, Brett Murphy, Alan Andersen, Garry Cook, Cameron Yates,
Kamaljit Sangha, Jay Evans, Dominique Lynch e Andrew Edwards pela orientação,
inspiração e acolhida durante parte do meu doutorado sanduíche desenvolvido na Austrália.

Aos amigos Moara Canova, Layon Demarchi, Jéssica Cruz, Fernanda Catenacci, Tatiane
Basconi, Stefanie Pereira, Bábara Sarti, Ariadne Pallas, Andresa Figueiredo, Talita Cantuária,
Paulo Sérgio, Luís Otávio, Maristela, Danielle Lana, Gustavo Mariano, Pamela Moser,
Monique Alves, Tamiris, Raissa Ribeiro, Marco Túlio, Natacha Silva, Sílvia Laine, Vanusa
Babaçu, Carlos Eduardo, Lílian Bernardes, Claire Webbs, Rose, Elly, Mike, Maria, Shelly,
Wendy, Sofia, Amanda, Mark, Grant, Gabriela Arcoverde por toda a ajuda e suporte
emocional durante essa empreitada.

vi
 Sumário

Lista de figuras............................................................................................................. viii

Lista de tabelas................................................................................................................ x
Resumo ............................................................................................................................ 1
Abstract ........................................................................................................................... 2
Introdução geral ............................................................................................................. 4
Capítulo 1 - The legacy of colonial fire management policies on traditional
livelihoods and ecological sustainability in three fire-prone continental savanna
settings: impacts, consequences, new directions ........................................................ 10
Abstract........................................................................................................................... 11
1. Introduction ................................................................................................................ 11
2. An overview of traditional use of fire in savannas ..................................................... 12
3. Traditional fire management systems and fire policies .............................................. 14
3.1 Southern African savanna ......................................................................................... 14
3.2 Northern Australian savanna .................................................................................... 17
3.3 Brazilian savanna (cerrado) ...................................................................................... 19
4. Savanna fire management policies: contributions from traditional systems and
implications for institutional change .............................................................................. 22
Capítulo 2 - How does the Cerrado burn? Informing fire management in the
Brazilian savanna ......................................................................................................... 27
1. Introduction ................................................................................................................ 27
2. Material and methods ................................................................................................. 28
2.1 Study area ................................................................................................................. 28
2.2 Fire treatments .......................................................................................................... 30
2.3 Fire measurements and environmental conditions ................................................... 31
2.4 Statistical analysis..................................................................................................... 32
3. Results ........................................................................................................................ 32
3.2 Fire intensity ............................................................................................................. 32
3.2 Fine fuel consumption .............................................................................................. 39
4. Discussion ................................................................................................................... 43
4.1 Season, weather and fire interval .............................................................................. 43
4.2 Fire behaviour predictions for management systems ............................................... 45
Capítulo 3 - How Cerrado woody species respond to different fire management
systems? ......................................................................................................................... 48
1. Introduction ................................................................................................................ 48

vi
2. Material and method ................................................................................................... 50
2.1 Study areas ................................................................................................................ 50
2.2 Experimental design ................................................................................................. 52
2.3 Tree response attributes ............................................................................................ 55
2.4 Tree community assessment ..................................................................................... 56
2.5 Data analysis ............................................................................................................. 57
3. Results ........................................................................................................................ 58
3.1 Stem density and recruitment ................................................................................... 58
3.2 Stem survival and topkill .......................................................................................... 61
3.3 Tree growth rates ...................................................................................................... 62
4. Discussion ................................................................................................................... 64
4. 1 Short-term, implications of fire management systems to tree communities ........... 64
4.2 Fire management applicability and ecological precautions ...................................... 68
Conclusão ...................................................................................................................... 72
Referências .................................................................................................................... 75

vii
 Lista de figuras

Capítulo 1

Fig. 1. Diagram summarizing key points related to different fire management policies in
savannas, their implications to the environment and population and overall future perspectives.
...................................................................................................................................................26

Capítulo 2

Fig. 1. Chapada das Mesas National Park location in regional, national and continental scales.29
Fig. 2. Fire intensities in 2015 and 2017, measured in different seasons, weather conditions and
fire intervals in CMNP ...............................................................................................................33
Fig. 3. Fire intensity in different fire intervals, seasons and weather conditions in a woodland
savanna in the CMNP.................................................................................................................36
Fig. 4. Fire intensity under two fire intervals and two weather conditions in a woodland savanna
(CMNP) in May 2015.. ..............................................................................................................39
Fig. 5. (a) Pre-fire biomass in a LDS treatment; (b) Post-fire biomass in a LDS treatment; (c)
Pre-fire biomass in a biennial, EDS evening treatment; and (d) Post-fire biomass in a biennial,
EDS evening treatment...............................................................................................................40
Fig. 6. Fine fuel consumption in 2015 and 2017, sampled in different seasons, weather
conditions and fire intervals in CMNP. ......................................................................................41
Fig. 7. Fine fuel consumption in different fire intervals, seasons and weather conditions in a
woodland savanna in the CMNP ................................................................................................42
Fig. 8. Fine fuel consumption in two fire intervals and two weather conditions in a woodland
savanna (CMNP) in May 2015 ...................................................................................................43

Capítulo 3

Fig. 1. Map with the location of Chapada das Mesas National Park in regional, national and
continental scale, where the experiment for this study took place ..............................................51
Fig. 2. (a) The LDS plot (LF) before experimental fire; (b) LDS fire (September, LF), carried
out at mid-day; (c) LDS plot (LF) after experimental fire. .........................................................53
Fig. 3. (a) Triennial, high-intensity EDS plot (High-EF) before experimental fire; (b) triennial,
high-intensity EDS fire, carried out at mid-day (May, High-EF); (c) biennial, high-intensity
EDS plot (High-EF) before experimental fire; (d) biennial, high-intensity EDS fire, carried out
at mid-day (May, High-EF). .......................................................................................................54
Fig. 4. (a) Triennial, low-intensity EDS fire, carried out in the evening (May, Low-EF); (b)
triennial, low-intensity EDS plot (Low-EF) after experimental fire; (c) biennial, low-intensity
EDS fire, carried out in the evening (May, Low-EF); (d) biennial, low-intensity EDS plot (Low-
EF) after experimental fire. ........................................................................................................54
Fig. 5. (a) Plot protected from fire for five years (NF); (b) Plot protected from fire for four years
(NF)............................................................................................................................................55

viii
Fig. 6. Stem recruitment per hectare from 2015 to 2017 in experimental plots submitted to
different fire frequencies and treatments ………………………………………………….. 59
Fig. 7. Change in stem density per hectare in different diameter classes submitted to different
fire frequencies and treatments.. .................................................................................................60
Fig. 8. Stem survival within four diameter classes submitted to different fire frequencies and
treatments.. .................................................................................................................................61
Fig. 9. Stem topkill within four diameter classes submitted to different fire frequencies and
treatments.. .................................................................................................................................62
Fig. 10. Change in basal area under two fire intervals (two and three years) submitted to
different fire treatments ..............................................................................................................63
Fig. 11. Change in tree height within different classes submitted to different fire frequencies and
treatments.....................................................................................................................................64

ix
 Lista de tabelas

Capítulo 2

Table 1. Number and characteristics of the 31 prescribed, experimental fires carried out in a
open savanna, located in the CMNP, in 2015 and 2017 .............................................................31
Table 2. Weather conditions and fire behaviour in 2015 and 2017 for each treatment and fire
frequency. ..................................................................................................................................35
Table 3. Weather conditions and fire behaviour for each treatment and fire frequency. ............35
Table 4. Model selection to describe fire intensity variability under different fire regimes in a
woodland savanna, northern Cerrado region (CMNP). ..............................................................37
Table 5. Model selection to describe variations in fine fuel consumption, under different fire
regimes in a woodland savanna, northern Cerrado region (CMNP). ..........................................41

Capítulo 3

Table 1. Number of studied areas, plots and prescribed fires within each treatment in the PNCM
in 2015. ......................................................................................................................................53
Table 2. Number of stems per hectare within the diameter classes, fire frequency and
treatments. ..................................................................................................................................56

x
 Resumo

A política de zero tolerância ao fogo no Cerrado desde o período colonial contribuiu,

dentre outros fatores, para o aumento da flamabilidade de ambientes pirofíticos e
conflitos entre as autoridades governamentais e comunidades rurais. A partir de 2014
teve inicio o Programa Piloto de Manejo Integrado do Fogo (MIF), em que queimas
prescritas foram implementadas em algumas unidades de conservação (UC) do Cerrado
na tentativa de diminuir a extensão de áreas atingidas por incêndios, as emissões de
gases de efeito estufa (GEE) e conflitos com as comunidades locais bem como proteger
vegetações sensíveis ao fogo, como matas de galeria, de incêndios. Assim, as queimas
prescritas estão sendo conduzidas no início da estação seca, e em locais estratégicos.
Estas queimas são realizadas sob condições meteorológicas amenas, geralmente no final
da tarde quando a umidade relativa do ar é maior e a temperatura do ar se reduz
gradualmente. Embora outras pesquisas tenham caracterizado o comportamento e efeito
do fogo em comunidades arbóreas no Cerrado no início da estação seca, faltam
informações referentes a queimas de final de tarde voltadas para o manejo, as quais
tendem a apresentar menores intensidades de fogo. Experimentos colaborativos e as
estratégias utilizada por outros países podem subsidiar ações de manejo. Esta tese foi
desenvolvida em uma das três primeiras UC do Cerrado a implementar o MIF, o Parque
Nacional da Chapada das Mesas, e está dividida em três capítulos: (1) revisão
bibliográfica sintetizando as políticas de manejo do fogo nas principais regiões
savânicas do hemisfério sul; (2) caracterização do comportamento de 24 queimas
experimentais no início da estação seca, ao meio-dia e no final da tarde, e de sete
queimas tardias em áreas de cerrado sentido restrito abertas com histórico de queima de
dois e três anos; e (3) quantificação dos efeitos dessas queimas experimentais em
comunidades lenhosas (cerca de 4.900 árvores). Programas de manejo integrativos e
flexíveis, que implementam predominantemente queimas no início da estação seca,
reduziram substancialmente a frequência de incêndios no final da estação seca, as
emissões de GEE e os conflitos com populações locais. De maneira geral, intervalos de
queima maiores (três anos) apresentaram intensidades de fogo maiores do que intervalos
menores (dois anos). O sistema de fogo de fim de tarde no início da estação seca,
mostrou-se como o mais eficiente para atingir os objetivos do MIF em unidades de

1
conservação a curto-prazo. Esta tese foi desenvolvida para aumentar as informações
científicas disponíveis para pesquisadores, instituições ambientais e comunidades rurais
que utilizam o fogo como instrumento para o manejo da terra.

Palavras-chave: conhecimento tradicional, época do fogo, incêndio, política do fogo,

queima prescrita, savana.

Abstract

The zero tolerance fire policy in the Cerrado (Brazilian savanna) since the colonial era
contributed, among other factors, to increase the flammability of fire-prone
environments and conflicts between the authorities and rural communities. Since 2014,
a Integrated Fire Management (IFM) Pilot Program started centered in prescribed burns.
This IFM program started in Cerrado protected areas (PA) aiming to reduce at the same
time wildfires, greenhouse gas (GHG) emissions and conflicts with the local
communities, as well as to protect fire-sensitive vegetation, such as riparian forests,
from wildfires. The prescribed burns are being conducted during early dry season, and
when there is high risk of wildfire spread the burns, they are undertaken under mild
weather conditions, generally in the end of the afternoon when air relative humidity is
higher and air temperature gradually reduces. Although other researches have
characterized fire behaviour and fire effects on Cerrado tree communities in early dry
season, there is a lack of information on the behavior and effects of evening burns such
as those performed for fire management, which tend to be of lower fire intensity.
Collaborative experiments and the strategies used by other countries can subsidize
management actions. This dissertation was developed in one of the first three Cerrado
PA to implement IFM, the Chapada das Mesas National Park, and is divided in three
chapters: (1) literature review synthesizing management policies in the largest savanna
regions in the southern hemisphere; (2) characterization of fire behaviour in 24
experimental burns in early dry season, around mid-day and evening, and of seven late
dry season burns in open typical cerrado areas, with two and three years time since fire;
and (3) quantification of the effects of these experimental burns in tree communities
(around 4,900 trees). Integrative and flexible management programmes, which
implement early dry season burns predominantly, have substantially reduced late dry
season, wildfire frequency, GHG emissions and conflicts with local populations.

2
Generally, longer fire intervals (three years) presented higher fire intensities than
shorter intervals (two years). The fire system of early dry season, evening burns was the
most efficient to reach short-term IFM goals in PA. This dissertation was developed to
enhance the scientific information available for researchers, environmental institutions
and rural communities who use fire as a land management tool.

Keywords: fire policy, fire season, prescribed burn, savanna, traditional knowledge,
wildfire.

3
Introdução geral

A presença do fogo na natureza é datada de aproximadamente 420 milhões de

anos (Scott & Glasspool, 2006), o regime de fogo, assim como outros distúrbios
ecológicos, é determinante no processo de seleção e distribuição de espécies em
diferentes ecossistemas (Bond & Keeley, 2005; Bowman et al., 2009). O regime de fogo
é caracterizado especialmente pela frequência, época e extensão da queima de acordo
com a disponibilidade de combustível e de fatores abióticos, como condições
meteorológicas e geomorfológicas (Whelan, 1995). Diferentes regimes de fogo ao longo
de milhares de anos influenciaram o estabelecimento de ecossistemas dependentes,
adaptados e sensíveis ao fogo (Myers, 2006).

Embora o fogo ocorra de forma natural em quase todos os continentes do

mundo, a partir de 50.000-100.000 anos antes do presente (AP) o uso intensivo do fogo
pelo Homem provocou mudanças nos regimes de fogo e consequentemente
transformações nas paisagens (Bar-Yosef, 2002; Bowman et al., 2011). Assim, o fogo
contribuiu para o estabelecimento e desenvolvimento de populações humanas (Bowman
et al., 2011; Pyne, 1993, 2016), bem como para a manutenção de algumas
fitofisionomias como as campestres e savânicas, em que árvores e gramíneas coexistem
(Accatino et al., 2010; Bond, 2016; Bond et al., 2005; Higgins et al., 2000; Scholes &
Archer, 1997a). Por outro lado, o fogo descontrolado - incêndio - pode gerar perdas
ambientais e materiais, altos custos de combate e ameaça a saúde e a vida das pessoas
(Dombeck et al., 2004; Moritz et al., 2014; Stephens et al., 2009).

A savana é considerada o bioma mais pirofítico do mundo, responsável por

cerca de 86% de todas as queimas (Mouillot & Field, 2005), onde 30% de toda a
produtividade primária do planeta é produzida graças a elevada taxa de precipitação,
geralmente acima de 700 mm por ano, que após um período de estiagem bem marcado
se transforma em combustível disponível para queima (Accatino et al., 2010; Grace et
al., 2006; Sankaran et al., 2005). De forma geral, a biomassa fina - graminóides, demais
ervas e subarbustos com ramos até 0,6 mm de diâmetro (Luke & McArthur, 1978) - é
produzida mais rapidamente em formações savânicas e campestres, levando a intervalos
menores de queimas do que em formações florestais (Bond & Keeley, 2005; Fidelis et

4
al., 2010b; Miranda et al., 2009). Fisionomias campestres e savânicas, em geral
apresentam substrato graminoso-herbáceo contínuo, o que favorece fogos de superfície
com maiores velocidades de propagação do fogo, e são compostas por espécies mais
adaptadas ao fogo quando comparadas a formações florestais (Higgins et al., 2000;
Hoffmann et al., 2012).

Embora o uso do fogo como ferramenta de manejo tenha possibilitado o

processo histórico de ocupação das savanas, o interesse econômico voltado para
ecossistemas florestais e a falta de compreensão dos europeus da dinâmica ecológica de
ecossistemas savânicos no período colonial, levaram a políticas de exclusão do fogo em
regiões como a África subsariana (Archibald, 2016; Laris & Wardell, 2006; Pyne,
1997a; Wardell et al., 2004), América do Sul (Durigan & Ratter, 2016; Mistry et al.,
2016; Pivello, 2006; Sletto, 2008) e norte da Austrália (Edwards et al., 2015; Ritchie,
2009; Russell-Smith et al., 2007). Estas políticas de exclusão do fogo contribuíram para
(i) o aumento da flamabilidade de ambientes pirofíticos, devido a ocorrência de amplas
áreas (milhares de hectares) com histórico de queima semelhante e grandes quantidades
de material seco combustível acumulados com a ausência de queimas, aumentando
assim a probabilidade de propagação de incêndios (Bond & Parr, 2010; Durigan &
Ratter, 2016; Pivello, 2006; Russell-Smith et al., 2007), (ii) conflitos entre as
autoridades governamentais e comunidades tradicionais (Batista et al., 2018; Eriksen,
2007; Kull, 2002; Mistry & Bizerril, 2011; Morrison & Cooke, 2003), e/ou (iii) ao
adensamento da vegetação lenhosa em locais onde as condições climáticas e edáficas
propiciam ao adensamento de espécies lenhosas (Archibald, 2016; Honda & Durigan,
2016; O’Connor et al., 2014).

Países como Austrália (Russell-Smith et al., 2013b), Estados Unidos (Freeman

et al., 2017b; Ryan et al., 2013) e Canadá (Lee et al., 2002; Stocks & Martell, 2016) têm
implementado ações de manejo de fogo nas últimas décadas, que incluem queimas
prescritas na tentativa de diminuir as emissões de gases de efeito estufa (GEE) e a
extensão de áreas queimadas por incêndios. Na Austrália, o reconhecimento e a
incorporação do conhecimento ecológico tradicional (CET) dos aborígenes nas técnicas
de queima de áreas silvestres tem resultado em maior efetividade de manejo, e
possibilitado a geração de renda através da venda de créditos de carbono pela redução

5
das emissões de GEE gerada pela redução de incêndios no final da estação seca,
resultantes do manejo de fogo e da realização de queimas prescritas especialmente no
início da estação seca (Preece, 2007; Price et al., 2012; Russell-Smith, 2016; Russell-
Smith et al., 2013a). Experiências como estas estão incentivando queimas prescritas e
manejo integrado de fogo (MIF) em outras regiões pirofíticas, como em algumas zonas
rurais e reservas ambientais africanas (Goldammer & de Ronde, 2004; Kepe, 2005) e
recentemente (2014) em algumas áreas protegidas brasileiras (Schmidt et al., 2018).

Diferentes sistemas de manejo do fogo resultam em regimes de fogo distintos,

em que o comportamento do fogo de cada queima é determinado pelas características
bióticas (biomassa combustível) e abióticas (temperatura, velocidade do vento, umidade
relativa, relevo) de ecossistemas locais (Rothermel & Deeming, 1980; Whelan, 1995).
No Cerrado, estudos indicam que práticas de queima de comunidades rurais e indígenas
favorecem a criação de mosaicos de vegetação queimada e não queimada, que podem
impedir a propagação de incêndios no auge da estação seca, e beneficiar a reprodução,
dispersão e forrageamento de algumas espécies (Mistry & Bizerril, 2011; Mistry et al.,
2005; Pivello, 2006; Welch et al., 2013). Os incêndios, que são comuns no final da
estação seca (agosto-setembro) (França, 2010), tendem a atingir grandes extensões de
áreas (muitas vezes > 50.000 ha) que incluem vegetações resistentes e sensíveis ao fogo
(Barradas, 2017; Batista et al., 2018; Pereira Júnior et al., 2014), e causam maior taxa de
mortalidade ou topkill de espécies arbóreas do que as queimas no início da estação seca
(Hoffmann & Solbrig, 2003; Sato & Miranda, 1996).

Neste contexto, esta pesquisa teve por objetivos avaliar e subsidiar diferentes
sistemas de manejo do fogo no Cerrado e para isto foi dividida em três capítulos: (1)
investigação das mudanças nas políticas de manejo do fogo em diferentes savanas
tropicais do mundo e suas respectivas consequências ecológicas e sociais; (2)
caracterização do comportamento do fogo em áreas de cerrado sentido restrito aberto,
considerando diferentes épocas e frequências de queimas; e (3) avaliação dos efeitos
ecológicos dos regimes de fogo estabelecidos por sistemas de exclusão do fogo,
incêndio e queimas prescritas na vegetação arbórea em áreas de cerrado sentido restrito
aberto.

No capítulo um desta tese foram avaliadas as mudanças nas abordagens de

6
queima em três regiões savânicas - Cerrado, norte da Austrália e sul da África - do
período pré-colonial até a atualidade. Ao contextualizar política e economicamente os
diferentes sistemas de manejo do fogo e pontuar as consequência sociais e ambientais
destes sistemas, foi possível identificar os aprendizados e desafios associados a políticas
de fogo sustentáveis e viáveis para regiões pirofíticas.

O capítulo dois apresenta as características do comportamento do fogo em

diferentes sistemas de manejo em uma das três primeiras unidades de conservação (UC)
a implementar o MIF no Cerrado: Parque Nacional da Chapada das Mesas (PNCM),
localizado no sul do Maranhão. Estes sistemas foram avaliados em áreas com histórico
de queima entre dois e três anos em fisionomias de cerrado sentido restrito e foram
representados pelas seguintes queimas experimentais: (1) queimas no final da estação
seca (setembro), simulando o regime de fogo predominante em UC do Cerrado
regulamentadas pela política de exclusão do fogo; (2) queimas no início da estação seca
(maio) próximas do meio-dia, quando a temperatura do ar atinge o máximo e a umidade
relativa é menor que 50%; e (3) queimas no início da estação seca (maio) no final da
tarde, quando a temperatura do ar diminui e a umidade relativa é maior que 50%. As
queimas experimentais realizadas no início da estação seca representam as queimas
implementadas pelo programa do MIF e, por isso, foram utilizados os mesmos
parâmetros estabelecidos para as tomadas de decisão antes de cada queima prescrita.

Finalmente, no capítulo três, foram avaliados os efeitos destes sistemas de

manejo do fogo em comunidades de espécies arbóreas em relação a: porcentagem de
sobrevivência e topkill; número de recrutamentos; mudança na densidade de indivíduos,
área basal e altura. Parcelas permanentes de pesquisa foram estabelecidas em
fisionomias de cerrado sentido restrito no PNCM para possibilitar o monitoramento das
comunidades arbóreas no período de 2015 a 2017. Para testar as implicações dos
diferentes sistemas de manejo, o experimento foi dividido em: (a) parcelas sob regime
bienal de queimas no final da estação seca (setembro); (b) parcelas sob regime bienal e
trienal de queimas no início da estação seca (maio) iniciadas próximo ao meio-dia com
a umidade relativa do ar abaixo de 50%; (c) parcelas sob regime bienal e trienal de
queimas no início da estação seca (maio) iniciadas no final da tarde com a umidade
relativa do ar acima de 50%; e (d) parcelas protegidas do fogo por quatro e cinco anos

7
fogo.

Esta tese apresenta os primeiros resultados do acompanhamento científico de

atividades de manejo de fogo no Cerrado. O delineamento experimental realizado,
assim como todas as medidas realizadas em campo, contaram com a participação de
gestores ambientais, moradores e brigadistas do Parque Nacional da Chapada das Mesas
bem como outros funcionários do Instituto Chico Mendes de Conservação da
Biodiversidade (ICMBio), responsável pela gestão de UC federais no Brasil. O
financiamento para realização deste trabalho ocorreu por meio dos seguintes projetos:
(i) "Acompanhamento técnico-científico da implementação do Manejo Integrado do
Fogo em Unidades de Conservação do Cerrado" (GIZ/FINATEC, número 83198593)
no âmbito do Projeto Cerrado-Jalapão, uma parceria entre os governos brasileiro e
alemão (MMA, 2015; Schmidt et al., 2016a; b, 2018); (ii) "Proposta de manejo
integrado do fogo para o Parque Nacional da Chapada das Mesas, MA" (identificação
Projeto PNUD - BRA/08/023), subsidiado pela Diretoria de Pesquisa, Avaliação e
Monitoramento da Biodiversidade (DIBio) do ICMBio e equipe gestora do PNCM; (iii)
"Integrating social and ecological reponses in a scientific evaluation of fire
management: Lessons learnt from Australian fire-prone ecosystems", em parceria com
o Darwin Centre for Bushfire Research, da Universidade de Charles Darwin na
Austrália, financiado pela Fundação de Apoio a Pesquisa do Distrito Federal (FAP-DF,
número 0193.000563/2016); e (iv) "Como a época do fogo afeta a vegetação do
Cerrado?", em parceria com a Universidade Estadual Paulista (UNESP) e financiado
pela Fundação de Amparo a Pesquisa do Estado de São Paulo (FAPESP, número
2015/06743-0).

O papel da pesquisa aplicada no auxílio à gestão de UC e

elaboração/implementação de políticas públicas na área ambiental é amplamente
reconhecido. No entanto, a contribuição de resultados de pesquisa para a gestão podem
muitas vezes ser limitados (Arruda et al., 2018; Dickinson & Ryan, 2010; Driscoll et al.,
2010; Gomes et al., 2018; Russell-Smith et al., 2003). Para aumentar a contribuição da
ecologia aplicada à gestão é recomendada a realização de experimentos conjuntos.
Assim, a criação e a manutenção de ambientes de aprendizagem, que agreguem
diferentes atores sociais, como pesquisadores, gestores e comunidades locais, são

8
essenciais para a integração entre pesquisa, gestão e conhecimento tradicional (Berkes
et al., 2000; Christensen, 2005; Perry et al., 2018; Petty et al., 2015; van Wilgen &
Biggs, 2011). Neste contexto, além das contribuições científicas e da formação da
autora, os experimentos realizados para a execução desta tese de doutorado
contribuíram também para integrar conhecimentos técnicos, científicos e tradicionais
voltados para o Parque Nacional da Chapada das Mesas, MA, uma das primeiras UC do
Cerrado a implementar ações de manejo de fogo, quebrando o paradigma da Política de
“fogo-zero” em vigência no Brasil por séculos.

Esta tese foi escrita em forma de artigos na língua inglesa, em que cada capítulo
corresponde a um artigo. Para facilitar a leitura e evitar repetições, as referências citadas
em cada capítulo foram colocadas no final da tese.

9
Capítulo 1 - The legacy of colonial fire management policies on traditional
livelihoods and ecological sustainability in three fire-prone continental savanna
settings: impacts, consequences, new directions

O legado das políticas de manejo coloniais nos modos de vida tradicionais e na

sustentabilidade ecológica em três cenários de savana continentais pirofíticos:
impactos, consequências, novos rumos

Moura, Livia C. a, *; Scariot, Aldicir O. b; Schmidt, Isabel B. a; Beatty, Robin c; Russell-

Smith, Jeremy d
a
Departamento de Ecologia, University of Brasília, Brazil
b
Laboratório de Ecologia e Conservação, Embrapa Recursos Genéticos e Biotecnologia,
Brasília, Brazil
c
321Fire, Inhambane, Mozambique
d
Darwin Centre for Bushfire Research, Charles Darwin University, Darwin, Australia

* Corresponding author.
Email addresses: [email protected] (L.C. Moura), [email protected]
(A. O. Scariot), [email protected] (I.B. Schmidt), [email protected] (R.
Beatty), [email protected] (J. Russell-Smith).

Artigo submetido a revista "Journal of Environmental Management"

10
Abstract

Land occupation and management systems have defined fire regimes and
intercontinental landscapes for millennia. The savanna biome is responsible for 86% of
all fire events, contributes to 10% of the total carbon emissions annually and is home to
10% of the human population. From the 16th century, European colonization instituted
fire suppression policies in many tropical savannas, markedly disrupting traditional fire
management practices and transforming ecosystems. In this paper we assess savanna
burning approaches from pre-colonial to contemporary eras in three regions: northern
Australia, southern Africa and Brazil. In these three regions, fire suppression policies
have led to (i) conflicts between government authorities and local communities; (ii)
frequent late dry season wildfires and/or (iii) woody encroachment. Such consequences
have required changes to historical fire management policies, including recognition and
incorporation of traditional ecological knowledge in contemporary savanna fire
management contexts. The application of integrative and flexible fire management
programmes, mostly implementing prescribed early dry season fires, has substantially
reduced the frequency of more severe late dry season fires and greenhouse gas
emissions, as well as generating income opportunities for rural and traditional
communities. We present a brief history of fire management policies in these three
important fire-prone savanna regions, and identify ongoing challenges for
implementation of culturally and ecologically sustainable, integrated fire management
policies.

Keywords: cerrado; fire regime; fire suppression; integrated fire management;

prescribed fire; traditional ecological knowledge.

1. Introduction

Fire has contributed to the evolution of Homo erectus species since the lower
Pleistocene, from at least 1.6 million years ago (Bellomo, 1994). Over the past 50,000
years the recurrent use of fire for foraging and agriculture has shaped modern fire
regimes characterized by human-influenced fire frequency and seasonality (Bowman et
al., 2011). The savanna is the most fire-prone biome in the world, due to the
characteristic continuous grassy fuel layer that becomes increasingly flammable with
the progression of the lengthy (5-8 month) dry season (Giglio et al., 2013; van der Werf
et al., 2017). Worldwide, savannas produce 30% of the earth's net primary production
(Grace et al., 2006), accounts for 86% of all fire events (Mouillot & Field, 2005),
resulting in ~10% contribution of total carbon emissions annually (van der Werf et al.,
2010). Africa is responsible for 71% of all savanna CO2 emissions, South America for
12% and Australia for 7.3% (UNU, 2015). The biome is home to approximately 10% of
the human population (White et al., 2000), many of whom rely on the use of fire for

11
their livelihoods (Laris, 2002; Mistry et al., 2005, 2016; Russell-Smith et al., 2009b;
UNU, 2015; Welch et al., 2018).

The European colonization of savanna regions since the 16 th Century was

marked by the official prohibition in the use of fire (Pyne, 1997b). Fire suppression
policies neglected and disrupted local agriculture and landscape management practices
and led to profound ecosystem changes, ranging from woody encroachment (of native
and exotic species) due to reduced fire frequency (O’Connor et al., 2014), to extensive
accumulation of flammable ground fuels and increased incidence of wildfires (Bond &
Parr, 2010; Pyne, 1997b; Ryan et al., 2013; UNU, 2015). Because of these unwanted
consequences, in recent decades there has being growing recognition of the value and
effectiveness of traditional (Indigenous, Aboriginal) fire management approaches for
addressing contemporary fire management issues in fire-prone savannas —at least in a
few southern African countries, parts of north America (Ryan et al., 2013), more
recently in Brazil (Schmidt et al., 2018), and especially in northern Australia (Preece,
2007; Russell-Smith et al., 2013b; UNU, 2015). However, as a legacy of colonial
history, traditional savanna fire management practices are still officially stigmatized
and/or prohibited in most African (Laris & Wardell, 2006), Asian (van Vliet et al.,
2013), and South American savannas (FAO, 2007; Russell-Smith et al., 2013b; UNU,
2015; Welch et al., 2018).

In this contribution we assess the antecedents of savanna burning approaches in

three fire-prone continental savanna settings – northern Australia, southern Africa and
Brazil. Our focus is especially concerned with: (i) the history of fire management in
these regions; (ii) the contributions of traditional ecological knowledge (TEK) systems
to contemporary savanna fire management policies; and (iii) the challenges associated
with incorporating and maintaining traditional practices in legal frameworks to better
achieve sustainable fire management in these and other fire-prone regions.

2. An overview of traditional use of fire in savannas

Humans have used fire as a tool and influenced fire regimes in Africa over the
past 400,000 years (Bird & Cali, 1998), Australia’s tropical savannas over at least
60,000 years (Clarkson et al., 2017; Roberts et al., 1993), and in Brazil over at least

12
32,000 years (Guidon & Delibrias, 1986). The use of fire is recognized as an essential
tool underpinning the development and expansion of hunting and foraging societies
(Bird & Cali, 1998; Jones, 1969). In the Holocene, fire was essential to the development
of cattle domestication 10,000-8,000 years BP (before present) in the eastern Sahara in
Africa (Marshall & Hildebrand, 2002), plant domestication in Northern Brazil around
8,000 years BP (Clement et al., 2010), and 5,000 years BP in western Africa
(Phillipson, 2005). Overall, ancient anthropogenic burning has markedly contributed to
human evolution (Bowman et al., 2011) and, at the same time, contributed to the
creation of high-diversity landscapes, where the formation of small-scale successional
mosaics increased local-scale biodiversity and prevented losses of important habitat
resources (Bliege Bird et al., 2008; Bowman et al., 2016; Trauernicht et al., 2015a).

In present times, fire continues to be used worldwide by traditional societies for

multiple purposes, including: land clearing for agriculture, hunting, stimulating plant
flowering/fruiting for harvesting purposes, regenerating pasture species, reducing fuels,
increasing visibility, extending habitat zones, avoiding and selecting species, pest
elimination, communication, protection, ceremonies and rituals, (e.g. Garde et al., 2009;
Mistry et al., 2016; Shaffer, 2010). The frequency and the selection of areas to burn has
depended on management goals, environmental conditions and resource availability
(Bowman et al., 2011; Welch et al., 2018).

“Slash and burn" (swidden) agriculture is one of the most ancient cultivation
systems (Iversen, 1956), which gradually became commonly practiced among Brazilian
Indigenous groups (Dean, 1996; Piperno, 2011) and some African societies (Harlan,
1976). To implement this itinerant cultivation system in Brazil, small (usually < 1ha)
forested patches are cut down during the end of the dry season so the biomass is ready
to burn as soon as the rainy season starts, releasing nutrients from ashes to cultivated
plants (Pedroso Júnior et al., 2008). North Australian Aborigines by contrast practiced
"fire stick farming", which consisted of setting either small savanna fires with a
smoldering fire-stick to hunt small game and gather plant resources while they traversed
their traditional estates (Jones, 1969), or highly organized fire-drives to hunt larger
game (Altman, 2009).

13
 Prior to European colonization, repeated burning in small (typically hectare-
scale) contiguous patches was a common strategy used by African, Brazilian and
Australian Indigenous people to create mosaics with different fire histories and burnt
extents (Bliege Bird et al., 2008; Garde et al., 2009; Laris, 2002; Welch et al., 2013,
2018). In these respective regions, burning would commence as fuels dried (cured) in
the early dry season (EDS), from April to May, and extend throughout the dry season
until the onset of the rainy season (September to March). Strategically burnt patches—
for example reinforcing natural features such as wetlands, streams, rocky areas—served
also as effective firebreaks thereby reducing the risk of extensive late dry season (LDS)
wildfires (generally August – October) (e.g. Garde et al., 2009; Laris, 2002; Welch et
al., 2018).

3. Traditional fire management systems and fire policies

3.1 Southern African savanna

During the 18th and 19th centuries gradual transference of land ownership from
traditional local authority to the state domain enabled European colonial authorities to
consolidate exploitation of African lands, labour and resources (Roe et al., 2009).
Natural resource management policies during the colonial era were a central component
of extending European political control into rural African landscapes (Neumann, 1998).
Resources, such as wildlife and timber, were progressively placed under central
regulatory authority with the rights of local people alienated over time (Roe et al.,
2009). Disruption of Indigenous cultural traditions and practices was highly successful
with, amongst others, "civilizing missions" that promoted racial segregation to divide
and control local populations (Conklin, 1998).

Traditional fire management systems were discouraged or prohibited based on

the view that Indigenous burning regimes threatened resources, property, the social
hierarchies of rigid colonial societies, and were environmentally degrading (Laris, 2002;
Mistry, 2000, 2002; Pearce, 2000; Pyne, 1993, 1997a, b). Fire suppression policies
reduced numbers of controlled burns changing traditional fire regimes across southern
Africa to almost exclusive LDS fires (Shaffer, 2010). Extensive LDS wildfires are
frequent (often annual-biennial) and extensive (100 – 300km2) in remote sparsely

14
populated landscapes (e.g. protected areas - PAs), greatly reduced in areas utilized by
cattle, and further reduced in areas of cultivation and settlement (Archibald et al., 2009,
2010).

A history of frequent LDS fires has resulted in a reduction of tree density and
canopy cover (Frost & Robertson, 1987; Furley et al., 2008; Higgins et al., 2000),
damage to fire-prone ecosystems and loss of biodiversity (Archibald, 2016; Laris &
Wardell, 2006), significant greenhouse gas (GHG) emissions (Korontzi, 2005; Lehsten
et al., 2009; Scholes et al., 1996; van der Werf et al., 2010) or conversely facilitated
woody encroachment (Archibald, 2016; Bond & Parr, 2010; O’Connor et al., 2014).
Fire prohibition led to conflicts between governmental officials and Indigenous or rural
communities, given impacts of LDS wildfires on people’s livelihoods, and often
resulted in protest arson fires (Eriksen, 2007; Kull, 2002).

Starting in the late 1950’s, newly independent nations typically adopted

colonially-derived political structures and, in many instances, extended central authority
over lands and resources further alienating local rights (Alden Wily, 2008; Mamdani,
1996). Colonial approaches to fire management were among those retained and have
largely continued to the present via the emphasis on ‘green conditionality’ in the
disbursement of aid and loans to developing nations (Bryant & Bailey, 1997; Mistry,
2002; Suchet, 2002). Policy for commercial agriculture, timber and PAs paralleled
developments in ecological knowledge and has oscillated between no-burn to prescribed
burns and lightning strike regimes (Kepe, 2005; van Wilgen et al., 2000). Fire regimes
developed to conserve and manage southern African savannas are primarily based on
natural, experimental or archaeological evidence (Govender et al., 2006; Hall, 1984;
Roques et al., 2001; van Wilgen et al., 2003, 2004), consistently ignoring TEK and
burning practices in contemporary rural communities (Shaffer, 2010). The traditional
fire management practices that have persisted, in varying forms, throughout southern
Africa have generally been unsanctioned.

The emergence of Community-based Natural Resource Management (CBNRM)

in the 1980s initiated a dramatic shift in natural resource use rights of local people on
communal lands (Jones & Murphree, 2001; Jones, 2004). Driven by political reform
(independence), legislated decentralized approaches devolved substantial rights over

15
wildlife and wildlife-based revenue to local people in Zimbabwe and Namibia.
Innovative joint venture tourism enterprises, based on traditional ecological knowledge,
developed into Zimbabwe’s CAMPFIRE program (1989-2001) covering over 40,000
km2 of communal land and generating $20 million in revenue, and Namibia’s
Communal Area Conservancies currently cover more than 14% of the country, involve
over 200,000 people and earn $2.5 million p.a. (NACSO, 2012). These have played a
key role in the development of CBNRM throughout the region and across sub-Saharan
Africa (Suich et al., 2009).

Community-based Fire Management (CBFiM) and Integrated Fire Management

(IFM) concepts developed alongside the ‘community forestry’ branch of CBNRM in the
late 1990s to reduce conflicts, firefighting costs and environmental or forest losses
(Goldammer & de Ronde, 2004; Pricope & Binford, 2012; Walters, 2015). Early
initiatives tended to reproduce colonial approaches to fire management through training
community work forces to implement externally driven prevention and suppression
activities in communal areas, for example the Namibia-Finland Forestry Programme,
and Working on Fire in South Africa.

Traditional fire knowledge was promoted as a central component of CBFiM and

IFM programs in the late 2000s through growing recognition of contemporary
Indigenous African, Australian and South American communities elsewhere using
controlled burns to expand or maintain subsistence livelihood activities and resources,
and reduce wildfires (Kepe, 2005; Kepe & Scoones, 1999; Kull, 2002; Laris, 2002;
Mistry et al., 2005; Sheuyange et al., 2005; Yibarbuk et al., 2001). Programs in
Botswana (UNDP, 2010), Mozambique (CARE/WWF, 2013), Namibia (FAO, 2011;
UNU, 2015), Swaziland, Tanzania (Sills et al., 2014), Zambia (TNC, 2018) and
Zimbabwe have significantly improved community livelihoods, reduced wildfires and,
to varying degrees, been adopted into national and international fire management
policies.

At present, such programs rely upon international development funding and

remain relatively limited in duration and extent. In recent years, significant attempts
have been made to incentivize community-based EDS fire management approaches to
reduce GHG emissions reduction. A Verified Carbon Standard methodology was

16
approved recently for quantifying GHG emissions reductions from the implementation
of EDS burning in miombo woodlands in Tanzania, Mozambique and Malawi (Sills et
al., 2014; VCS, 2015).

3.2 Northern Australian savanna

European colonization of northern Australia commenced in the mid-19th Century

and was associated with the establishment of large beef cattle pastoral leases and
incipient mining activities. The colonizers effectively gave no recognition to Aboriginal
interests in land under a mythical legal concept termed ‘terra nullius’, which ignored
the reality of Aboriginal occupation now known to span at least sixty millennia
(Clarkson et al. 2017). Catastrophic population declines are well documented by the
start of the 20th Century associated mostly with disease, translocation and, in some
regions, massacres (Levitus, 2009; Ritchie, 2009; Roberts, 2005).

In consequence, traditional fire regimes across the northern savannas were

substantially interrupted and replaced with a new regime dominated by extensive (often
>1000 km2) and frequent (>0.3 fires p.a.) LDS wildfires (Edwards et al., 2015;
Morrison & Cooke, 2003; Russell-Smith et al., 2007). These contemporary fire regimes
have had, and in many locations continue to have, very significant deleterious impacts
on a variety of ecosystem attributes, including: fire-vulnerable plant taxa and
assemblages (Bowman & Panton, 1993; Freeman et al., 2017a; Russell-Smith et al.,
2012), granivorous birds (Franklin et al., 2005), small mammals (Lawes et al., 2015;
Woinarski et al., 2011), and greenhouse gas emissions (Edwards et al., 2015, 2018). In
some locations, traditional management practices continue to the present day, most
notably in the Arnhem Land region of the Northern Territory (Garde et al., 2009;
Yibarbuk et al., 2001).

In 1976 and 1993 the Australian Government instituted legislation giving further
opportunities for Aboriginal people to reclaim land—if claimants could provide
evidence of their traditional association with that land. The 1976 legislation, which
applies only to the 1.42 million km2 Northern Territory, has allowed Aboriginal people
to reclaim over 50% of that jurisdiction as freehold title; the 1993 legislation has
enabled Aboriginal people nationally to claim lands under a joint ‘Native Title’

17
arrangement which does not extinguish pre-existing land use titles. Collectively, these
recent developments have enabled Aboriginal people to currently reclaim 56% of the
1.2 million km2 north Australian savannas region receiving mean annual rainfall of at
least 600 mm (Archer et al., 2018). Despite being relatively “land rich”, regional
Aboriginal people remain severely economically and socially disadvantaged (Russell-
Smith et al., 2013b).

Land reclamation by Aboriginal claimants has also been associated with a strong
desire to reinstitute cultural practices, including responsibilities for fire management of
traditional estates. This intent has not been thwarted by colonial legislative instruments
(e.g. bushfire management Acts of the three north Australian jurisdictions—
Queensland, Northern Territory, Western Australia) given tacit understanding that the
risks of LDS wildfire need to be managed by various means, including allowance for
the lighting, under permitted conditions, of preventative fires (Preece, 2007). General
acceptance of the pragmatic wisdom of Aboriginal landscape-scale fire management
approaches are enshrined in regional biodiversity conservation applications—for
example, underpinning fire management in the World Heritage property, Kakadu
National Park (Director of National Parks, 2016).

Since the early 2000s, direct support for the implementation of Aboriginal
landscape-scale savanna fire management initiatives has been afforded through
development of Australia’s formal savanna burning GHG emissions abatement
programme - as part of meeting Australia’s Kyoto reporting targets. Foremost amongst
those developments has been the West Arnhem Land Fire Abatement (WALFA)
project, formally established in 2006 to deliver effective fire management and
associated industrial-scale GHG emissions abatement over a 28,000 km 2 region, under
an offset arrangement with a multi-national corporate (Russell-Smith et al., 2013a).
Over the past decade WALFA has substantially exceeded the contracted 100,000 t.CO 2-
e annual abatement, transformed the former LDS-dominated fire regime to one where
annual fire extent is reduced and the great majority of burning is delivered through
prescribed EDS management, and is providing broader ecosystem health and
community benefits (Price et al., 2012; Russell-Smith et al., 2013a).

18
 In 2013 the Australian Government formally accredited a savanna burning GHG
emissions accounting methodology (CoA, 2013), as part of its then Carbon Farming
Initiative (CFI) programme. Further refinements to that emissions abatement method
have continued (CoA, 2015) under the revised Emissions Reduction Fund (ERF)
programme, and new complementary savanna burning carbon sequestration components
are in preparation (e.g. Cook et al., 2016). Under the ERF, the Australian Government
purchases carbon credits (where a credit = 1 t.CO2-e) from contracted project suppliers
currently at ~USD 10 per credit. Additionally there is a nationally strong voluntary
market where business entities buy savanna-burning credits at significantly higher
prices to meet their corporate social responsibility requirements.

As at early 2018 there were 78 formally ERF-registered savanna burning

projects of which 32 are on Aboriginal lands, encompassing 25% of the entire 1.2
million km2 savanna region above the 600 mm mean annual rainfall isohyet.
Importantly, while the application of savanna burning projects may adequately
complement the cultural requirements of Aboriginal landowners (Garde et al., 2009;
Russell-Smith et al., 2009b), contemporary regulatory and administrative requirements
add levels of complexity which have the potential to disenfranchise full landowner
participation (Perry et al., 2018).

This Australian experience is encouraging the development of similar

approaches in other savanna regions, especially southern Africa and most recently in
Brazil (Russell-Smith et al., 2013b; Schmidt et al., 2018; UNU, 2015).

3.3 Brazilian savanna (cerrado)

European colonization started in the Atlantic rainforest along the Brazilian coast
in the 1500’s, focusing on Brazilwood (Paubrasilia echinata) extraction, and sugarcane
and coffee plantations. Fire was used to implement all three economic activities,
incorporating and adapted from traditional Indigenous techniques. Approximately 100
years after Europeans arrived, the use of fire became less tolerated, due to its damaging
effects on sensitive rainforest vegetation (Dean, 1996). In the 1700’s, livestock and
agricultural activities expanded inland through the Brazilian savanna (Silva et al.,
2012). Europeans progressively occupied Indigenous territories and consequently large

19
extents of native savanna landscape were transformed to pastures, monoculture crops
and settlements. Indigenous populations have been significantly reduced to less than
10% of their pre-Columbian number, due to diseases and conflicts (Pivello, 2006).

As well as native vegetation being replaced by exotic grasses and extensive

monocultures, Indigenous traditional burning systems and activities were significantly
reduced (Nevle & Bird, 2008). However, some of their techniques were adapted and
incorporated in agricultural and livestock production systems (Pivello, 2011). Major
differences between traditional Indigenous and farmers’ burning practices included the
size of the areas managed, fire frequency, vegetation type, and purposes. Moreover, the
indiscriminate use of fire in monocultures and pastures, as well as the accumulation of
unmanaged fuel loads in remote areas, led to changes in the fire regime maintained by
Indigenous people; the Brazilian savanna became mostly burnt by extensive LDS
wildfires rather than small patchy burns (Pereira Júnior et al., 2014; Pivello, 2011). The
recurrence of large wildfires resulted in a zero fire tolerance policy in the whole
country, regardless of the vegetation type or seasonal period (Durigan & Ratter, 2016).
Consequences of this policy for the savannas included forest encroachment in less
frequently burnt locations (Abreu et al., 2017), landscape homogenization (Durigan &
Ratter, 2016), changes to animal community composition and structure (Briani et al.,
2004; Frizzo et al., 2011), and increases in GHG emissions due to frequent and
extensive LDS wildfires (Pinto & Bustamante, 2010).

Throughout its history, Brazil has pursued a development policy where

agriculture and livestock have been identified as the main economic activities (Silva et
al., 2012), resulting in deforestation and immeasurable biodiversity losses, including
within the savanna region (Klink & Machado, 2005). Nevertheless, in 1961, the
establishment of the first three PAs in central Brazil was an important step towards
recognizing the environmental values of savanna ecosystems. However, these PAs
forbade human settlement and the direct use of natural resources even when these
overlapped with Indigenous peoples’ and small farmers’ territories (MMA, 2006).
Consequently, PAs have become a source of conflict (Diegues, 2000), frequently
involving the use of fire by local communities, for livelihood purposes or protest
(Batista et al., 2018; Mistry & Bizerril, 2011). The natural tendency of unmanaged fire-

20
prone savanna ecosystems to develop highly flammable fuel load (continuous,
homogeneous, and dry), together with illegal fire ignitions, has rendered Brazilian
savanna PAs increasingly vulnerable to LDS wildfires (Durigan & Ratter, 2016;
Pivello, 2006).

After decades of frustrated attempts to reconsider zero-fire policies, and

scientific evidence of the ecological role of fire in the Brazilian savannas (Maravalhas
& Vasconcelos, 2014; Miranda, 2010; Pivello, 2011; Simon et al., 2009), the national
legislation was changed in 2012 to explicitly allow for fire management for
conservation purposes in fire-prone environments (Schmidt et al., 2018). A binational
Brazil-German project collaboration) implemented the first Integrated Fire Management
program in the Brazilian savanna, 2012 - 2016 (the “Cerrado-Jalapão project”). This
was the first politically sanctioned trial for using prescribed burns to reduce wildfires,
reduce GHG emissions, and support biodiversity conservation outcomes in Brazilian
savanna PAs. Other goals of the project were to: (1) enhance improved dialogue
between environmental managers, scientists and local communities, and; (2) reduce
fire-fighting costs (Schmidt et al., 2018).

The Cerrado-Jalapão IFM pilot programme trialled the implementation of low

intensity, EDS (April to July) fires, commencing in 2014 at three PAs. Subsequently,
the trial was extended and in 2017 included eight PAs and 11 Indigenous reserves. The
burning program was annually planned, implemented, monitored and evaluated jointly
by policymakers and stakeholders, including representatives from international, national
and state environmental institutions, researchers, and local communities. The Cerrado-
Jalapão IFM project resulted in 57% less LDS fires over three years of implementation
(Schmidt et al., 2018), stimulating broader acceptance of fire management and
specifically prescribed fires in both public and private lands.

The Cerrado-Jalapão IFM programme has contributed significantly to retraction

of the zero-fire paradigm in Brazilian fire-prone ecosystems that will likely spread to
other Brazilian and South American regions. The programme has also helped to push
further legislative changes regarding the hiring of wildland fire-fighters and fire-
managers—promoting the training and hiring of community-based fire brigades, and

21
encouraging local communities’ participation in landscape and fire management
decisions and activities in communal, public and private areas.

Preliminary data indicate that fire management following this model may
contribute to reduced GHG fire emissions. Further GHG monitoring will help quantify
resultant benefits and pave the way for future financing of enhanced fire management
actions through carbon abatement programs.

4. Savanna fire management policies: contributions from traditional systems and

implications for institutional change

Indigenous fire practices, socio-economic conditions and land management

policies have rarely been constructively aligned throughout the long colonial history of
savanna occupation. This mismatch can be attributed to the prioritization of economic
development policies detrimental to broader socio-ecological values. Fire suppression
policies, initially focused on conserving forest resources, were introduced to uphold
centralized economic interests, resulting in broad-scale impacts on local livelihoods and
land use practices, conflicts between state authorities and local communities, and
variable environmental changes. Changing these colonial fire management paradigms,
particularly policies emphasizing fire suppression, is a long-term social and political
process, which must overcome institutional resistance and lack of investment in
community engagement, governance and technological infrastructure (Russell-Smith et
al., 2013b; UNU, 2015).

In recent decades there has been growing support from at least some savanna
state authorities to formally recognize traditional fire knowledge systems in
contemporary management contexts. In substantial part this recognition has been driven
by the magnitude of recurrent fire management issues associated particularly with the
impacts of extensive LDS wildfires on both rural livelihoods and ecological values. As
well, there has been growing understanding of the dynamic role which fire regimes play
in maintaining healthy savanna ecosystems. As illustrated by our inter-continental
comparative assessment, traditional savanna fire knowledge systems and practices
contribute a number of valuable common practical lessons pertinent to contemporary
management contexts. For example, application of fine-scale strategic EDS fire
management can: (1) significantly reduce the extent and impact of LDS savanna fires;

22
hence (2) allow for the undertaking of prescribed LDS management applications under
safely contained conditions (e.g. using relatively intense fires to address unwanted
woody thickening and encroachment); (3) significantly reduce GHG emissions, and
impacts on biodiversity assets and values; hence (4) generate income and opportunities
for regional traditional communities through GHG emissions abatement and
community-based fire management (CBFiM) schemes (Russell-Smith, 2016; Schmidt et
al., 2018; UNU, 2015).

Effective fire management policies in fire-prone ecosystems are dependent on

adequate legal frameworks, institutional capacities, and broad community support and
engagement to protect human lives and property, manage destructive wildfires, and
conserve natural ecosystems (Moritz et al., 2014; Pereira et al., 2012; Stephens et al.,
2009). While national fire management policy frameworks typically reflect top-down
control, recognition of traditional management knowledge and practice in formal policy
frameworks has obvious benefits for enhancing fire management effectiveness
(Dombeck et al., 2004; Foster et al., 2003; Mistry et al., 2016; Myers, 2006; Silva et al.,
2010). However, for many developing countries in the post-colonial era, transitioning
from state-centric fire suppression policies to local-level inclusive approaches poses
significant challenges, including lack of local-level governance and well defined policy
processes, and institutional and community capacity to deal with socially contentious
fire management issues (Brown, 2003; Mistry et al., 2016; Russell-Smith et al., 2013b;
Tacconi et al., 2006; UNU, 2015).

The practical benefits of co-opting and adapting traditional savanna fire

management approaches has been demonstrated by local or regional-scale examples in
each of the three fire-prone continental settings. In the sparsely settled and fire-prone
savannas of northern Australia, regional fire management policies have tacitly accepted
(while not fully acknowledging) the pragmatic utility of traditional Aboriginal seasonal
fire management practices (Preece, 2007). Positive experience with that policy approach
has, over the past decade, encouraged development of a formal (state-sanctioned) GHG
emissions abatement program which, at the time of writing, is now being implemented
across 300,000 km2 of Australia’s northern, higher rainfall (>600 mm p.a.) region,
mostly on Aboriginal-owned lands (Russell-Smith et al., 2018).

23
 Although most savanna regions in South America and sub-Saharan Africa are
still regulated by suppressive fire management measures, recent integrative approaches
(e.g. CBFiM, IFM) are increasingly delivering optimistic results and expanding official
fire acceptance of rural activities and Indigenous practices—including support for
sustaining local livelihoods, and incorporation into PA management practices
(Goldammer & de Ronde, 2004; Schmidt et al., 2018). Ironically, savanna PAs present
challenging settings for implementing strategic fire management programmes since they
currently comprise the most extensively burnt tenures in sub-Saharan African and South
American continental contexts. As noted previously such conditions can be attributed to
regionally high and continuous fuel loads associated with generally limited landscape
fragmentation, fire use restrictions, and abundant sources of ignition from surrounding
densely settled agricultural holdings (Archibald, 2016; Batista et al., 2018; Pereira
Júnior et al., 2014; Schmidt et al., 2018). Inclusive policies which support the
sustainable fire management practices of rural communities (including those whose
livelihoods are reliant on PAs), and enhanced understanding of the dynamic role of fire
regimes in maintaining healthy savanna ecosystems, are mandatory for eliminating
conflicts, reducing wildfire issues, and conserving biodiversity.

As documented here, recent decades have witnessed the emergence (albeit

tentative) of a profound shift to official acceptance and recognition of more community-
based fire management approaches in three fire-prone continental savanna settings. Key
points arising from that discussion can be summarized as a schematic (Fig. 1),
illustrating fundamental differences in practical societal, livelihood, and environmental
outcomes emanating from the implementation of repressive or inclusive fire
management policy approaches, respectively. Importantly, the latter inclusive approach
inherently requires ongoing refinement through adaptive monitoring and management
processes (Gómez-Baggethun et al., 2013; Huffman, 2013; Roos et al., 2016), and, as
witnessed recently on all three continents, it has helped energize development of CBFM
and IFM approaches (Goldammer & de Ronde, 2004; Schmidt et al., 2018), and novel
incentivized GHG emissions abatement and carbon sequestration schemes (Russell-
Smith et al., 2013a; UNU, 2015). After a long colonial history of repressive fire
management policies, the importance of locality- and context-specific Indigenous and

24
traditional fire management approaches is at last being recognized to help address
global savanna fire management problems.

25
Fig. 1. Diagram summarizing key points related to different fire management policies in savannas, their implications to the environment and
population and overall future perspectives.

26
Capítulo 2 - How does the Cerrado burn? Informing fire management in the
Brazilian savanna

Como o Cerrado queima? Informando o manejo do fogo na savana brasileira

1. Introduction

Fire-prone ecosystems are characterized by the annual occurrence of ideal

conditions for fire to propagate across the landscape, provided by continuous layer of
fine fuel, lengthy dry season and systematic ignition source (Bowman et al., 2009;
Whelan, 1995). In the Anthropocene era, human ignition sources have increased fire
occurrence and burned extents (Bowman et al., 2011; Pyne, 1997b; Roos et al., 2014).
Which tends to enhance vegetation traits of flammability, e.g. favouring graminoids
over woody plants and keeping opened physiognomies (Accatino et al., 2010; Bond et
al., 2005; Miranda et al., 2002), where the main source of fuel lies in the standing
structure of dead biomass (Kauffman et al., 1994; Pausas & Moreira, 2012). Recurrent,
large wildfires have become a global problem threatening human lives and assets, as
well as environmental sustainability, such wildfires are especially common in
unmanaged fire-prone environments (Edwards et al., 2015; Hantson et al., 2017; Moritz
et al., 2014; Scott et al., 2016).

Savanna is the most fire-prone biome on earth (Giglio et al., 2013; van der Werf
et al., 2017), responsible for the production of around 30% of the world’s net primary
production (Grace et al., 2006) and approximately 86% of all fire incidents (Mouillot &
Field, 2005). Remote regions are often hit by wildfires because of the combination of
unmanaged, highly flammable, fuel load (Archibald, 2016; Russell-Smith et al., 2013b;
UNU, 2015) and extreme fire weather conditions (Cheney & Sullivan, 2008; Gill et al.,
1996; Moritz, 2003). In Tropical savannas, wildfires commonly occur in the late dry
season (LDS, from August to October in southern hemisphere and from December to
February in the northern hemisphere) with high frequency (annual-triennial), burning
large areas (>100,000 ha, Archibald, 2016; Cahoon Jr et al., 1992; Pereira Júnior et al.,
2014; Pivello, 2006; Russell-Smith et al., 2007; Shaffer, 2010; Yates et al., 2008).
Weather conditions and cured biomass in the LDS increase the chances of high-
intensity fires that can be more severe and difficult to control (Whelan, 1995; Williams
27
et al., 1998). In tropical humid savannas (rainfall >1,100 mm), grasses grow fast
providing enough cured fuel to burn annually by the end of the dry season (Accatino et
al., 2010; Batmanian & Haridasan, 1985; Williams et al., 2002; Yates et al., 2008).

The Cerrado region, in central Brazil, is a tropical humid savanna where LDS
frequent and large wildfires have prevailed due to the fire suppression policy established
since the colonial era (~1500), especially in protected areas and reserves (PAs)
(Barradas, 2017; Batista et al., 2018; França, 2010; Pereira Júnior et al., 2014; Pivello,
2006; Schmidt et al., 2018). This fire regime significantly impacts plant and animal
communities (Durigan & Ratter, 2016; Frizzo et al., 2011; Maravalhas & Vasconcelos,
2014; Pivello, 2011; Sato et al., 2010; Silveira et al., 1999) and can produce more CO2
than the vegetation can reabsorb in short periods of time (Sato et al., 2010). Commonly,
these wildfires are difficult to extinguish and generate large governmental expenses in
firefighting operations (Schmidt et al., 2018). Since 2014, prescribed early dry season
(EDS, April-July) fires have been implemented within a pilot Integrated Fire
Management (IFM) programme. The main goals of this pilot IFM programme are to
change LDS fire regime in Cerrado PAs and Indigenous reserves where it has been
implemented, avoiding large LDS wildfires (Schmidt et al., 2018).

Fire management is yet to be intensively implemented in tropical savanna

ecosystems, once changing fire exclusion paradigms is a long-term process. However,
local and scientific ecological knowledge have been guiding this transition in many
management systems (Driscoll et al., 2010; Goldammer et al., 2004; Mistry et al., 2018;
Myers, 2006; Russell-Smith et al., 2009b; Schmidt et al., 2018). Understanding fire
behaviour under different fire regimes in the Cerrado can help predicting management
systems and also prevent biodiversity loss. Under this perspective, we characterized
prescribed experimental fires within the current fire regime (LDS fires) and the regime
proposed by the IFM (EDS fires), and identified the parameters which determine
patterns on fire behaviour, based on fire intensity and fine fuel consumption.

2. Material and methods

2.1 Study area

Chapada das Mesas National Park is located in the northern Cerrado (CMNP,
7°19’0”S and 47°20’6”W), in the northeast of Brazil (Fig. 1). The CMNP was created in

28
2005 with 160,000 ha aiming to conserve the natural landscape and protect the region
from being converted into forestry, monoculture or intensive livestock farms. However,
the Brazilian fire exclusion policy has increased burned extents by wildfires during the
dry season throughout the park since its creation. The combination of fire-prone
vegetation, high air temperature (average 35°C) and average annual rainfall of 1,500-
1,800 mm, concentrated in seven months (October-April, INMET 2017), makes the
region highly flammable and, consequently, is frequently burned by wildfires. The park
is predominantly covered by savanna physiognomies, mostly open savanna on well-
drained, sandy soils. Most of the fire-sensitive vegetation consist in riparian and
swampy forests and grow along the watercourses.

Fig. 1. Chapada das Mesas National Park location in regional, national and continental scales.
Font: Victor Ferreira, 2015.

The CMNP was one of the first three PAs to implement the IFM pilot program
in 2014, especially because of its excessive wildfire occurrences and the managers'
interest in participating. As the implementation of fire management is still very new to
Brazilian PA managers, environmental authorities are still seeking parameters and
indicators that can help to determine when to start prescribed fires on the field (Schmidt
et al., 2018). Thus, our experimental design was previously discussed with managers
and focused on questions that may support management decisions, such as how fire

29
behaves under different weather conditions in two periods of the day, two seasons (early
vs. late dry season) and fire intervals (two and three years).

2.2 Fire treatments

To manage fire and implement prescribed burns the managers divided the PA in
Fire Management Zones (FMZ), according to land occupation and use, ecological
aspects and accessibility. During the fire management implementation EDS prescribed
fires were carried out within predetermined FMZ from April until June of each year by
the PA's staff. The managers predetermined priority management zones based on fuel
age and wildfire risks, as well as the type of fire to be lit in each setting. However, the
IFM implementation is precursory in Brazil (Schmidt et al., 2018) and, therefore,
several types of fire are being tested and still little is known about their behaviour
during these burns. Time of day was used as an indicator of weather conditions, where
by mid-day the highest temperature and lowest air relative humidity are reached and in
the evening the air temperature is lower and humidity is higher. We conducted LDS
fires to simulate general wildfire conditions and compare responses between both
seasons. In the attempt to monitor the IFM implementation in CMNP we projected this
system into experimental research areas, where we carried out prescribed fires
according to the following parameters: weather conditions, seasonality and fire interval.

All areas were chosen by the CMNP's staff in open savanna physiognomies to
represent main environmental conditions and in fragments previously isolated by
firebreaks. The areas were unburnt for two and three years, the same fire interval as
most areas targeted for management with prescribed fires by managers. Areas with two
years since last fire (last burned in 2013) we called biennial fire frequency and carried
out two prescribed fires, 2015 and 2017; areas with three years since last fire (last
burned in 2012) we called triennial fire frequency and burned only once in 2015. Within
these areas we implemented the following experimental fire treatments: (i) low-intensity
EDS fires (Low-EF): evening fires (17:30-19:00), when air relative humidity was >50%
in May; (ii) high-intensity, EDS fires (High-EF): mid-day fires (11:00-16:30), when air
relative humidity was <50% also in May; (iii) LDS fires (LF): mid-day fires (12:00-
14:30) carried out in September (Table 1). The LF treatment was not conducted in areas
with three years since last fire for security reasons, and in 2017 one of the LF
experimental area was burned by a wildfire. Each prescribed fire was undertaken in

30
areas of at least one hectare. Fire lines in LDS and EDS mid-day treatments were pulled
downwind to forward-burn, while evening EDS fire lines were pulled upwind to back-
burn, since these are the way prescribed fires were implemented by the PA managers.

Table 1. Number and characteristics of the 31 prescribed, experimental fires carried out in a
open savanna, located in the CMNP, in 2015 and 2017. LF = high-intensity, LDS mid-day fires
(September); High-EF = high-intensity, EDS mid-day fires (May); and Low EF = low-intensity,
EDS evening fires (May).

Fire type/ Time since N study N fires N fires

treatments fire (years) areas 2015 2017
Low-EF 2 8 4 4
Low-EF 3 4 4 0
High-EF 2 8 4 4
High-EF 3 4 4 0
LF 2 7* 4 3*

*One area was burned by a wildfire before we carried the experimental fire in 2017

2.3 Fire measurements and environmental conditions

For each experimental fire we recorded: weather conditions (air temperature,

relative humidity and wind speed), rate of fire front spread and fuel consumption to
calculate fire intensity. We calculated fire intensity according to the following equation
(Byram, 1959): I = h × w × r; where I = fire intensity (kW m -1); h = heat released from
the fuel (kJ kg-1), w = weight of fuel consumed per unit area (kg m-2) and r = rate of fire
spread (m s-1). We considered heat released from the fuel as 15,500 kJ kg-1 (Griffin &
Friedel, 1984; Miranda et al., 2010; Pivello et al., 2010; Schmidt et al., 2017). We
calculated fine fuel consumption by collecting all fine fuels (<6mm in diameter) from
the ground, within 12 plots (0.5 x 0.5 m) randomly allocated in each area, before and
after the fires. The samples of fine fuel were dried at temperature between 60-70°C for
three days before being weighted. The rate of spread was estimated visually, using trees,
rocks or wood sticks as references. Air temperature and air humidity were measured
using a portable weather station (Lutron EM-9000) to ensure that the weather conditions
attended the prerequisite established for each treatment. For all experimental fires, the
fine fuel consumption was calculated from the proportion: (wi - wf)/wi, where wi is the
pre-fire, fine biomass (fuel load) and wf is post-fire, fine biomass. To calculate annual
precipitation we used the online dataset from the Brazilian National Institute of
Meteorology (INMET), where we accounted for all rainfall from August in the previous

31
year until July in the burned year. Since in both study years the precipitation in June and
July (between our two fire seasons) was below 32 mm, we used the same precipitation
value as variable for EDS and LF fires.

2.4 Statistical analysis

To compare fire intensities in different fire treatments (High-EF, Low-EF and

LF), fire intervals (two and three years) and burned years (2015 and 2017) we
transformed fire intensity in logarithms and ran one and two-way analysis of variance
(ANOVA, Gotelli and Ellison, 2004), considering significantly different when p value
was below 0.1. We used general linear model (GLM) to test fine fuel consumption in
relation to the same independent variables as used for fire intensity (treatment, time
since fire [TSF] and burned year). We transformed fine fuel consumption values into the
inverse logistic, given by the formula: log[p/(1 - p)] , where p is the proportion value
between 0 and 1 (Crawley, 2002). These analyses were followed by Tukey post hoc
tests and the results are present with mean and standard error (±) values.

We built GLMs with the independent variables we believed to be influencing

fire intensity and fine fuel consumption, and applied Akaike’s information criterion for
small sample sizes (AICc) to calculate the likelihood of each model to explain the data,
and from the lowest ΔAICc and highest AICc weight we selected the most
parsimonious models (Burnham & Anderson, 2002; Symonds & Moussalli, 2011). For
ΔAICc <2 the models were considered equally explanatory. We considered air
humidity, air temperature, wind speed, fuel load, annual precipitation, fire treatment and
TSF- when applicable as variables to understand which of them better explained fire
intensity and fine fuel consumption values. The models included different sets of fires:
all fires (n=31); biennial fires (n=23); all EDS fires (n=24); and all triennial fires (n=8).
We used the R program version 3.5.0 (R-Core-Team, 2017) to run all analyses and
produce all our graphics.

3. Results

3.2 Fire intensity

The most intense fires were registered in areas unburned for three years with
experimental fires undertaken during the day in 2015 (triennial High-EF, 4,400 ± 942

32
kWm-1, Table 2). The intensity of these burnings was similar to the LDS experimental
fires in areas unburned for two years in 2017 (LF, 3,717 ± 754 kWm-1) and significantly
more intense than EDS fires undertaken during the evening in 2015 and 2017 (Low-EF,
640 ± 223 and 749 ± 220 kWm-1, respectively, Fig 2). Both treatments in triennial areas
had higher fire intensities than the corresponding treatments in biennial areas. Within
biennial fires, Low-EF were significantly less intense than LF in both study years,
where fire intensity of High-EF was intermediate. In triennial fires, Low-EF tended to
be less intense than High-EF (Figure 2).

Fig. 2. Fire intensities in 2015 and 2017, measured in different seasons, weather conditions and
fire intervals in CMNP. Legend presents TSF (two and three years) and fire treatment,
respectively, where: LF = high-intensity, LDS mid-day fires (September); High-EF = high-
intensity, EDS mid-day fires (May); and Low-EF = low-intensity, EDS evening fires (May).
The boxes encompass the first and third quartile, lines inside the boxes show the median, bars
indicate minimum and maximum values, and colored circles are the outliers. Different lower
case letters show significant difference (p ≤ 0.1), according to a posteriori contrasts between
TSF, fire treatment and burned year.

Differences between fire intensities in 2015 and 2017 were not significant within
the same treatment (p > 0.40), in contrast, there were significant differences among fires
carried out in different fire intervals (two and three years) and treatments (p < 0.01 and
p < 0.001, respectively). When experimental fires were grouped by fire treatment
regardless of the year (Table 3), the lowest average fire intensity were registered in the
biennial areas submitted to the Low-EF (694 ± 158 kW m-1, Fig 3), where fire rate of

33
spread (0.16 ± 0.04 m s-1), fine fuel consumption (70.8 ± 8 %) and air temperature (29.4
± 0.64 °C) were also the lowest recorded, and air relative humidity (64.9 ± 1.14 %) was
the highest. Conversely, the highest fire intensity, largest volume of fuel load (0.583 ±
0.02 kg m2), second highest fire rate of spread (0.57 ± 0.11 m s-1) and wind speed (1.5 ±
0.25 m s-1) was registered in triennial High-EF areas.

34
Table 2. Weather conditions and fire behaviour in 2015 and 2017 for each treatment and fire frequency. TSF=time since fire; LF = high-intensity, LDS mid-
day fires (September); High-EF = high-intensity, EDS mid-day fires (May); and Low-EF = low-intensity, EDS evening fires (May). Values represent the mean
and ± standard error.

Air Annual Fine fuel

Fire Air relative Wind speed Fuel load Fire rate of Fire intensity
TSF Year temperature precipitation consumption
treatment humidity (%) (m/s) (kg/m2) spread (m/s) (kW/m)
(°C) (mm) (%)
LF two 2015 35.6 ± 0.38 19.1 ± 0.65 1.7 ± 0.12 1466 0.37 ± 0.01 88.5 ± 1.6 0.53 ± 0.12 2833 ± 721
High-EF two 2015 34.2 ± 0.62 48 ± 0.89 1.8 ± 0.4 1466 0.437 ± 0.05 81.5 ± 3.75 0.26 ± 0.07 1570 ± 456
Low-EF two 2015 28.5 ± 0.12 64.2 ± 0.14 0.7 ± 0.26 1466 0.46 ± 0.01 87 ± 2.3 0.11 ± 0.04 640 ± 223
High-EF three 2015 34.1 ± 0.2 44.7 ± 1.55 1.5 ± 0.25 1466 0.582 ± 0.02 84.5 ± 2.8 0.57 ± 0.11 4400 ± 942
Low-EF three 2015 29.9 ± 0.31 57.8 ± 1.81 0.5 ± 0.14 1466 0.565 ± 0.04 85.5 ± 2.45 0.18 ± 0.05 1339 ± 294
LF two 2017 36.8 ± 0.27 29.2 ± 0.47 2.4 ± 0.05 1395 0.357 ± 0.02 89.3 ± 1.91 0.75 ± 0.14 3717 ± 754
High-EF two 2017 33.8 ± 0.37 46.1 ± 0.57 1 ± 0.18 1395 0.374 ± 0.02 80.7 ± 2.65 0.33 ± 0.03 1570 ± 141
Low-EF two 2017 30.2 ± 1.13 65.5 ± 2.24 0.6 ± 0.04 1395 0.372 ± 0.02 54.5 ± 1.1 0.22 ± 0.04 749 ± 220

Table 3. Weather conditions and fire behaviour for each treatment and fire frequency. TSF=time since fire; LF = high-intensity, LDS mid-day fires
(September); High-EF = high-intensity, EDS mid-day fires (May); and Low-EF = low-intensity, EDS evening fires (May). Values represent the mean and ±
standard error.

Air Annual
Air relative Wind speed Fuel load Fine fuel Fire rate of Fire intensity
Treatments TSF temperature precipitation
humidity (%) (m/s) (kg/m2) consumption (%) spread (m/s) (kW/m)
(°C) (mm)
High-EF three 34.1 ± 0.2 44.7 ± 1.55 1.5 ± 0.25 1466 0.583 ± 0.02 84.5 ± 2.8 0.57 ± 0.11 4400 ± 942
Low-EF three 29.9 ± 0.31 57.8 ± 1.81 0.5 ± 0.14 1466 0.565 ± 0.04 85.5 ± 2.46 0.18 ± 0.05 1340 ± 294
LF two 36.1 ± 0.33 23.4 ± 1.94 2 ± 0.15 1431± 13.4 0.364 ± 0.01 88.9 ± 1.24 0.63 ± 0.1 3212 ± 549
High-EF two 33.9 ± 0.37 47.1 ± 0.63 1.4 ± 0.26 1431 ± 8.49 0.406 ± 0.02 81.1 ± 2.31 0.3 ± 0.04 1570 ± 239
Low-EF two 29.4 ± 0.64 64.9 ± 1.14 0.6 ± 0.13 1431 ± 12.53 0.416 ± 0.02 70.8 ± 8 0.16 ± 0.04 694 ± 158

35
 The model that best explained variations in fire intensity when considering all
experimental fires included TSF and air temperature (AICc = 70.1, wAICc = 0.32,
Table 4a). The second-best model (AICc = 71.7, ΔAICc = 1.5, wAICc = 0.15) included
TSF, air temperature and air relative humidity. Notably, fire interval (TSF) was
associated to the variations in fuel load, as the areas unburned for three years had larger
volumes of fine fuel (>0.565 ± 0.04 kg m -2) than areas unburned for two years (<0.416
± 0.02 kg m-2).

Fig. 3. Fire intensity in different fire intervals, seasons and weather conditions in a woodland
savanna in the CMNP. Legend presents fire treatments, where: LF = high-intensity, LDS mid-
day fires (September); High-EF = high-intensity, EDS mid-day fires (May); and Low-EF = low-
intensity, EDS evening fires (May). The boxes encompass the first and third quartile, lines
inside the boxes show the median, bars indicate minimum and maximum values, and colored
circles are the outliers. Different lower case letters show significant difference (p ≤ 0.08),
according to a posteriori contrasts between time since fire and fire treatment.

36
Table 4. Model selection to describe fire intensity variability under different fire regimes in a
woodland savanna, northern Cerrado region (CMNP). Lower AICc value indicates more
parsimonious model compared to higher one. ΔAICc shows the difference between the model
and the lowest AICc value, and wAICc is the weight attributed to the model, informing the
relative likelihood of the model to explain fire intensity variability. In models: (a) all fires
(n=31) were included and the variables: TSF, treatment, fuel load, wind speed, air temperature,
air relative humidity and annual precipitation; (b) only biennial fires (n=23) were considered
and the variables: treatment, fuel load, wind speed, air temperature, air relative humidity and
annual precipitation; and (c) only EDS fires (n=24) were considered and the variables: TSF,
treatment, fuel load, wind speed, air temperature, air relative humidity and annual precipitation.

Models (a) AICc ΔAICc wAICc

TSF + air temperature 70.1 0 0.317
TSF + air temperature + air relative humidity 71.7 1.5 0.147
TSF + treatment 72.4 2.3 0.102
TSF + air temperature + wind speed 72.9 2.8 0.08
TSF + fuel load + air temperature 73 2.8 0.077
TSF + annual precipitation + air temperature 73 2.8 0.076
TSF + air relative humidity 73.2 3.1 0.068
Fuel load + air temperature 73.8 3.6 0.051
TSF + air temperature + treatment 74.1 4 0.043
TSF + annual precipitation + air relative humidity + air temperature 74.3 4.2 0.039
Models (b) AICc ΔAICc wAICc
Air temperature 53.6 0 0.32
Air temperature + air relative humidity 55.4 1.8 0.127
Air relative humidity 55.6 1.9 0.12
Air temperature + wind speed 56 2.4 0.095
Air temperature + fuel load 56.5 2.9 0.076
Air temperature + annual precipitation 56.5 2.9 0.073
Treatment 56.8 3.2 0.066
Air relative humidity + annual precipitation 57.1 3.5 0.056
Air temperature + air relative humidity + wind speed 57.9 4.2 0.038
Air temperature + air relative humidity + annual precipitation 58.3 4.7 0.03
Models (c) AICc ΔAICc wAICc
Air relative humidity 44.5 0 0.176
Air relative humidity + TSF 44.7 0.3 0.153
TSF + treatment 44.9 0.4 0.144
Air relative humidity + fuel load 44.9 0.5 0.138
Air relative humidity + wind speed 45.1 0.6 0.127
Air temperature 46.1 1.6 0.079
Air relative humidity + fuel load + wind speed 46.8 2.4 0.054
Treatment + fuel load 46.9 2.4 0.053
Air relative humidity + air temperature 47.5 3 0.039
Air relative humidity + treatment 47.5 3.1 0.038

37
 According to the GLM, fire intensities in the biennial areas were similar
between LF (3212 ± 549 kW m-1) and High_EF (1570 ± 239 kW m-1, p = 0.37), and
fires in both treatments were significantly more intense than Low-EF (694 ± 158, p <
0.1). Among these biennial fires, air temperature alone best explains the differences in
fire intensities across treatments (AICc = 53.6, wAICc = 0.32), the second-best model,
similar to the first one, included the combination of air temperature and air relative
humidity (AICc = 55.4, ΔAICc = 1.8, wAICc = 0.13, Table 4b). During LF, air
temperature reached the highest values and the lowest air relative humidity (36.1 ± 0.33
°C and 23.4 ± 1.94 %, respectively), while Low-EF were carried out under the lowest
air temperature and highest air relative humidity (29.4 ± 0.64 °C and 64.9 ± 1.14 %).
Among the triennial fires, High-EF tended to be more intense (with the highest intensity
among all experimental fires: 4,400 ± 942 kWm-1) than Low-EF (1340 ± 294 kWm-1),
but there was no significant difference across treatments (p = 0.16). Considering only
EDS fires, there is no significant difference in fire intensities between Low-EF and
High-EF within the same TSF interval (Fig 4). Whereas the comparisons between the
two TSF showed that the triennial High-EF were significantly more intense than
biennial Low-EF (p = 0.01), where air relative humidity was the variable that most
explained this difference (AICc = 44.5, wAICc = 0.18), and the second-best model
included also air relative humidity and TSF (AICc = 44.7, ΔAICc = 0.3, wAICc = 0.15,
Table 4c).

38
Fig. 4. Fire intensity under two fire intervals and two weather conditions in a woodland savanna
(CMNP) in May 2015. Legend presents fire treatments, where: High-EF = high-intensity, EDS
mid-day fires; and Low-EF = low-intensity, EDS evening fires. The boxes encompass the first
and third quartile, lines inside the boxes show the median, bars indicate minimum and
maximum values, and colored circles are the outliers. Different lower case letters show
significant difference (p ≤ 0.1), according to a posteriori contrasts between time since fire and
fire treatment.

3.2 Fine fuel consumption

Experimental LF consumed the highest proportion of fine fuel (LF, 89.3 ± 1.91
% in 2017 and 88.5 ± 1.6 % in 2015, Fig 5a, b) and the biennial Low-EF in 2017
consumed the lowest proportion of fine fuel (54.5 ± 1.1 %, Fig 5c, d). The consumption
of fine fuel in the Low-EF in 2017 varied significantly from the other types of
experimental fires, fire intervals and year (Fig 6), including from the same treatment in
2015 (p = 0.001). The first three best models contained annual precipitation as a
variable influencing fuel consumption variation (Table 5). Annual precipitation
combined separately with air relative humidity (AICc = 69.2, wAICc = 0.222), wind
speed (AICc = 70.1, ΔAICc = 0.9, wAICc = 0.14) and air temperature (AICc = 70.3,
ΔAICc = 1.2, wAICc = 0.123) were in the first, second and third best models,
respectively. In 2016-2017, annual precipitation (1,395 mm) was lower than in 2014-
2015 (1,466 mm), accordingly, the Low-EF biennial areas produced lower volume of
fuel load in 2017 (0.372 ± 0.02 kg m-2) than in 2015 (0.460 ± 0.01 kg m-2), when fuel
consumption was much higher (87 ± 2.3 %). For 2017, the biennial Low-EF were
carried out in highest air relative humidity (65.5 ± 2.24 %), lowest air temperature (30.2

39
± 1.13 °C) and lowest wind speed (0.6 ± 0.04 m s-1) when compared to the other
treatments. Such mild weather conditions at the time of burning, as well as the longer
rainy season in 2017 have probably contributed to lowering both fire intensity and fine
fuel consumption were.

Fig. 5. (a) Pre-fire biomass in a LDS treatment; (b) Post-fire biomass in a LDS treatment; (c)
Pre-fire biomass in a biennial, EDS evening treatment; and (d) Post-fire biomass in a biennial,
EDS evening treatment. All pictures were taken in the survey of 2017.

40
Fig. 6. Fine fuel consumption in 2015 and 2017, sampled in different seasons, weather
conditions and fire intervals in CMNP. Legend presents TSF (two and three years) and fire
treatment, respectively, where: LF = high-intensity, LDS mid-day fires (September); High-EF =
high-intensity, EDS mid-day fires (May); and Low-EF = low-intensity, EDS evening fires
(May). The boxes encompass the first and third quartile, lines inside the boxes show the median,
bars indicate minimum and maximum values, and colored circles are the outliers. Different
lower case letters show significant difference (p ≤ 0.1), according to a posteriori contrasts
between TSF, fire treatment and burned year.

Table 5. Model selection to describe variations in fine fuel consumption, under different fire
regimes in a woodland savanna, northern Cerrado region (CMNP). Lower AICc value indicates
more parsimonious model compared to higher one. ΔAICc shows the difference between the
model and the lowest AICc value, and wAICc is the weight attributed to the model, informing
the relative likelihood of the model to explain fuel consumption variability.

Models AICc ΔAICc wAICc

Annual precipitation + air relative humidity 69.2 0 0.222
Annual precipitation + wind speed 70.1 0.9 0.14
Annual precipitation + air temperature 70.3 1.2 0.123
Air relative humidity + fuel load 70.3 1.2 0.123
Annual precipitation + air relative humidity + wind speed 70.9 1.7 0.093
Annual precipitation + air relative humidity + fuel load 71.2 2 0.082
Annual precipitation + air temperature + wind speed 71.9 2.7 0.057
Annual precipitation + air relative humidity + air temperature 71.9 2.7 0.056
Annual precipitation + air relative humidity + TSF 71.9 2.8 0.056
Air relative humidity 72.2 3 0.049

When grouping the fine fuel consumption from both burned years (2015 and
2017) and analyzing only fire type and interval, the consumption among treatments

41
showed significant difference (p = 0.03), but not among fire intervals (p = 0.12, Fig 7).
Fine fuel consumption was significantly higher in biennial LF (88.9 ± 1.24 %) than in
biennial Low-EF (70.8 ± 8 %). Between these two treatments, air temperature (Low-EF
= 29.4 ± 0.64 °C and LF = 36.1 ± 0.33), air relative humidity (Low-EF = 64.9 ± 1.14 %
and LF = 23.4 ± 1.94 %) and wind speed (Low-EF = 0.6 ± 0.13 m s-1 and LF = 2 ± 0.15
m s-1) reached the maximum and minimum range, where LF presented more severe
weather conditions and Low-EF milder ones. When considering only EDS fires, the fine
fuel consumption is similar between Low-EF and High-EF (p = 0.3), as well as between
areas unburned for two and three years (p = 0.8, Fig 8).

Fig. 7. Fine fuel consumption in different fire intervals, seasons and weather conditions in a
woodland savanna in the CMNP. Legend presents fire treatments, where: LF = high-intensity,
LDS mid-day fires (September); High-EF = high-intensity, EDS mid-day fires (May); and Low-
EF = low-intensity, EDS evening fires (May). The boxes encompass the first and third quartile,
lines inside the boxes show the median, bars indicate minimum and maximum values, and
colored circles are the outliers. Different lower case letters show significant difference (p ≤ 0.1),
according to a posteriori contrasts between time since fire and fire treatment.

42
Fig. 8. Fine fuel consumption in two fire intervals and two weather conditions in a woodland
savanna (CMNP) in May 2015. Legend presents fire treatments, where: High-EF = high-
intensity, EDS mid-day fires; and Low-EF = low-intensity, EDS evening fires. The boxes
encompass the first and third quartile, lines inside the boxes show the median, bars indicate
minimum and maximum values, and colored circles are the outliers. Different lower case letters
show significant difference (p ≤ 0.1), according to a posteriori contrasts between time since fire
and fire treatment.

4. Discussion

4.1 Season, weather and fire interval

Many studies related to fire behaviour in savanna ecosystems focus on the

influence of seasonality and frequency on fire intensity (Gomes et al., 2018; Govender
et al., 2006; Miranda et al., 2010; Rissi et al., 2017; Schmidt et al., 2017; Williams et
al., 1998). However, the interaction of fire behaviour and different times of the day and
weather conditions in the same period of the year is poorly documented in the literature
for savanna physiognomies (Gill et al., 1996; Hoffmann et al., 2012). Our results show
that mid-day fires in the EDS have similar fire intensities to LDS fires, but they are both
different from EDS evening fires. This pattern is mostly related to variations in air
temperature, air relative humidity and wind speed, which in our experiment varied
along time of the day and season of the year. Other experiments in a savanna shrubland
(campo sujo) in Central Brazil, based on fire seasonality - early, middle and late dry
season - registered the average fire intensity of 503 ± 119/ 2,707 ± 1,899 and 3,009 ±
1,408 kWm-1, respectively, which were not find significantly different from each other

43
(Rissi et al., 2017). Higher fire intensities are commonly reached in the LDS (August-
October) in open savanna physiognomies, in the central Cerrado from 2,842 to 14,396
kWm-1 (Kauffman et al. 1994; Sato 2003), in northern Australian savanna from 2,700 to
7,700 kWm-1 (Russell-Smith et al., 2003; Williams et al., 1998) and in southern African
savanna from 2,566 to 12,912 kWm-1 (Govender et al., 2006; Trollope & Trollope,
2002). However, the period of the year alone is not a reliable predictor for fire intensity
(Bowman et al., 1988; Gomes et al., 2018; Russell-Smith & Edwards, 2006), since other
variables such as fuel moisture, fine fuel load (Fidelis et al., 2010a) and dead fuel
percentage are known to influence fire behaviour (Rissi et al., 2017; Trollope &
Trollope, 2002), and more generally air temperature, relative humidity, wind and slope
(Brown & Davis, 1973; Cheney & Sullivan, 2008; Luke & McArthur, 1978; Trollope &
Trollope, 2002; Whelan, 1995; Wright & Bailey, 1982). Our study showed that fire
season is not the main determinant of fire intensity, TSF and weather conditions are
better predictors. Time since last fire is directly associated with fuel age, and can
influence dead fuel percentage and fuel load (Batmanian & Haridasan, 1985; Gill et al.,
1996; Wagner, 1978). While weather conditions determine the air relative humidity and
fuel moisture of alive plants, air temperature and wind speed may influence fire rate of
spread (Cheney & Sullivan, 2008; Luke & McArthur, 1978; Rothermel, 1983) and
combustion rate (Brown & Davis, 1973).

Fine fuel consumption is another important parameter that should be considered

to evaluate how the vegetation burns. From our experiments, low-intensity, EDS
evenings fires in 2017 registered low fuel consumption (54.5%) compared to EDS mid-
day (80.7-84.5%) and LDS fires (88.5-89.3%), when shorter fire interval led to lower
consumption. Aboveground fuel consumption varies between 54 and 97% in savanna
physiognomies in central Brazil (Castro & Kauffman, 1998; Gomes et al., 2018;
Kauffman et al., 1994; Miranda et al., 2010). In northern Australian savanna, 74% of
the fine fuel is consumed within EDS fires and 86% in LDS fires (Russell-Smith et al.,
2009a), and, accordingly, in southern African savanna between 88-95% is consumed in
the LDS (Shea et al., 1996; Ward et al., 1996). Most fuel consumed is part of the
graminous herbaceous stratum and is exposed to microclimate features to burn
(Beerling & Osborne, 2006; Hoffmann et al., 2012; Miranda et al., 2002).

44
 Even though total annual precipitation was lower in 2016-2017 than in 2014-
2015, the rainy season extended longer in 2017 and consequently, Low-EF were carried
out in milder conditions in 2017 compared to 2015. These milder weather conditions
(lower temperatures and higher relative humidity) at the time of burning, resulted in
lower fuel consumption in the EDS evening fires for that calendar year. Fuel moisture is
correlated to weather features, such as relative humidity, temperature, precipitation and
wind speed (Hoffmann et al., 2012; Miranda et al., 2002). Previous investigations show
that these characteristics influence fire behaviour in Cerrado (Castro & Kauffman,
1998; Miranda et al., 2002; Rissi et al., 2017), Australian savanna (Cheney & Sullivan,
2008; Gill et al., 1996), African savanna (Govender et al., 2006; Savadogo, 2007). Dead
fuel is especially influenced by local weather conditions that change along the day, e.g.
in tropical savannas after 3 pm in the EDS air relative humidity is >60%, whereas in the
LDS is <30% (Gill et al., 1996), with solar radiation reinforcing fuel drying process
(Cheney & Sullivan, 2008). In our experimental areas, the EDS evening fires were lit
under >53% relative humidity and 28-31 °C air temperature; EDS mid-day fires
between 41.5-49 % and 33-36 °C; and LSD mid-day fires under <30% and >35 °C.

Additionally, TSF is associated to progressive curing of grassy layers reaching

the maximum dead fuel percentage two to three years after fire (Coutinho, 2002;
Miranda et al., 2010), each year after burned the herbaceous layer can recover 70% of
the total biomass it had before fire and can recover 100% two years after burned in
Cerrado physiognomies (Batmanian & Haridasan, 1985; Neto et al., 1998). However,
extreme fire weathers can override fuel characteristics, whereas in mild weather
conditions fuel load and fuel moisture are usually the main drivers of fire spread
(Moritz, 2003).

4.2 Fire behaviour predictions for management systems

Wildfires are recurrent in the Cerrado PAs due to the combination of unmanaged
fuel, extreme fire weather and increased anthropogenic source of ignition (Barradas,
2017; Batista et al., 2018; França, 2010; França et al., 2007; Pereira Júnior et al., 2014;
Pivello, 2006). In any fire-prone environment, fire management can only be
successfully implemented when the effect of weather, slope, fuel load, type of fuel,
natural barriers and local communities' activities are acknowledged and are part of the

45
planning and evaluation process. As Cerrado PAs aim to conserve local biodiversity and
landscapes, fire must be used as tool within the IFM pilot programme to help achieving
these goals. In this context, EDS evening fires can be applied in more flammable areas,
with longer fire intervals (>3 years) and hazardous volumes of fine fuel load, since the
mild weather conditions will help fires to extinguish themselves during the night (Gill et
al., 1996; Schmidt et al., 2016b; Williamson et al., 2016). These evening fires are less
likely to run out of control and burn unwanted areas with fire-sensitive species, and at
the same time they will work as firebreaks to LDS wildfires.

The EDS mid-day fires reach higher intensities and fuel consumption and,
therefore, can be implemented in strategic management zones, where fire interval is not
higher than two years in savanna physiognomies, and at the same time patch-mosaic
burnings will help to: generate heterogenic fire histories across the landscape; avoid
wildfires in fire-sensitive areas; and prevent woody encroachment in areas with
favorable soil conditions. Considering weather forecast to perform such higher-intensity
fires are indispensable, including measurements in the field just before burning takes
place. Finally, LDS fires happen under the most hazardous fire weather and are usually
avoided, once they put at risk extensive areas with homogeneous and unmanaged
flammable fuels (Krawchuk et al., 2009; Moritz et al., 2014; Yates et al., 2008).
However, for several management reasons, LDS fires might be desirable and can be
safely performed if adequate planning and EDS fires are undertaken to create barriers to
fire spread (Johnson et al., 2001; Keeley, 2009; Laris et al., 2016; Stephenson et al.,
1991). Systematic fire management, monitoring and mapping will likely reduce areas
burned by wildfires (Price et al., 2012; Russell-Smith et al., 2013a; van Wilgen &
Biggs, 2011) and can eventually help creating landscape heterogeneity, including areas
from which fire is excluded and increasing regional biodiversity (Bliege Bird et al.,
2008; Bowman et al., 2016; Perry et al., 2011; Trauernicht et al., 2015b; van Wilgen et
al., 2003).

Acknowledging these biotic and abiotic local and regional conditions can lower
human and equipment resources required for fire management, as well as better achieve
conservation goals (Dickinson & Ryan, 2010; Driscoll et al., 2010; Moritz et al., 2014;
Price et al., 2012; van Wilgen et al., 2007). However, long-term studies focused on the

46
effects of the different fire regimes upon the biota will be much more complementary
and qualified to inform and help evaluating the outcomes of the implementation of these
management systems.

47
Capítulo 3 - How Cerrado woody species respond to different fire management
systems?
Como as espécies lenhosas do Cerrado respondem a diferentes sistemas de manejo
do fogo?

1. Introduction

The ecological patterns found in different landscapes are the result of the
interaction between ecosystems and socio-temporal processes (Vellend, 2010).
Disturbances, such as fire, are important determinants of these ecological processes and
influence the dynamics of populations, communities and ecosystems (Bond, 2008). As a
natural disturbance, fire has shaped landscapes in almost every continent for millions of
years (Scott & Glasspool, 2006), where fire regime is responsible for major or minor
ecological impacts. Fire regimes are established by the combination of fuel type,
intensity, seasonality, frequency and extent (Whelan, 1995); however, the significance
of each variable depends on local conditions and ecosystem features (Archibald et al.,
2013; Bowman et al., 2009; Collins, 1992; Lehmann et al., 2014; Roos et al., 2014;
Williamson et al., 2016).

The savanna is the most fire-prone biome and the most extensively burnt on
Earth (Giglio et al., 2013; van der Werf et al., 2017). Accordingly, the dynamic of
savanna vegetation such as grasses and trees interactions (Scholes & Archer, 1997b) is
regulated and driven by fire regimes, mostly anthropogenic and occasionally natural
(Skarpe, 1992). Fire regime has proven to control tree densities, biomass and fuel load
as well as species composition in tropical savannas (Bond et al., 2005; Higgins et al.,
2007; Hoffmann & Moreira, 2002) by influencing the demography of woody species
and fuel accumulation (Bond, 2008). Concurrently, fuel accumulation and woody
biomass increases according to moisture availability in different savannas (Bond et al.,
2005; Lehmann et al., 2014; Sankaran et al., 2005). The structure and composition of
tree stratum becomes less complex in systems subjected to frequent fires (annual-
biennial, Williams et al., 1999), whereas, grasses expand due to faster regrowth rate
suppressing juvenile woody plants and increasing fine fuel load (Bond, 2008;
Hoffmann, 1999; Scholes & Archer, 1997b). When frequent, fire hinders tree
recruitment by topkilling (complete loss of aboveground biomass) individuals in a

48
critical range of stem diameter (Hoffmann & Solbrig, 2003) that prevent seedlings and
saplings from escaping the grass stratum (Bond, 2008). Demographic parameters of
woody plant individuals after fire events are also related to initial height in northern
Australian (Werner & Prior, 2013) and Afrotropical savannas and bark-diameter
correlation in Neotropical savannas (Dantas & Pausas, 2013).

The effects of fire on the vegetation also depends on fire behavior characterized
by fire intensity, flame height, fuel quality and quantity, burn efficiency, heat released
and weather conditions (Cheney & Sullivan, 2008; Kauffman et al., 1994; Miranda et
al., 2010; Rothermel, 1983; Whelan, 1995). Seasonality has shown to affect fire
intensity by changing fuel moisture content in southern African (Govender et al., 2006)
and northern Australian savannas (Russell-Smith et al., 2003; Williams et al., 1998),
where late dry season (LDS) fires in Eucalypt-dominated woodlands had significantly
higher intensities than in Eucalypt-dominated open forests and early dry season (EDS)
fires (Russell-Smith et al., 2003). In the central Cerrado region grassland
physiognomies carry out more intense fires and higher combustion efficiency compared
to woody vegetation (Castro & Kauffman, 1998; Miranda et al., 1996), which cause
high rates of top-kill by eliminating the aerial portion of trees and shrubs (Hoffmann et
al., 2009; Kauffman et al., 1994). On the other hand, fire parameters (such as fire
intensity) are not always significantly affected by fire season (EDS vs. LDS), being
rather determined by the proportion of dead fuel in fuel load in Cerrado shrublands
(Rissi et al., 2017). Fire intensity has proven to affect stem survival showing a linear
negative correlation (Hoffmann & Solbrig, 2003; Ryan & Williams, 2011; Williams et
al., 1999). Moreover, EDS (May-June) fires tend to result in fires with lower intensity
and severity compared to LDS (September-October) fires in northern Australian
savanna. In this case, severity is classified from low to high according to leaf-scorch
height, vegetation structure and burn patchiness (Russell-Smith & Edwards, 2006).

Fire management systems have driven fire regime, where the frequency and
extent of wildfires and prescribed fires are the result of different policies. Many fire-
prone environments have experienced fire suppression policies, commonly creating
hazardous flammable surfaces with fuel accumulation that increased wildfire incidence
(Bond & Parr, 2010; Steel et al., 2015; Varner et al., 2005), or, when soil conditions

49
allow for, promoting woody encroachment that reduced fire frequency (Bond et al.,
2005; Honda & Durigan, 2016; O’Connor et al., 2014; Ryan et al., 2013). In the
Brazilian savanna (Cerrado region), attempts to exclude fire lead to frequent (biennial-
triennial), extensive, high-intensity LDS (August-September) wildfires as the prevalent
modern fire regime, affecting especially protected areas (PA) (Batista et al., 2018;
Pereira Júnior et al., 2014; Pivello, 2011). Nevertheless, successful savanna fire
management (SFiM) approaches have inspired the integrated fire management (IFM)
pilot programme initially established in three Cerrado PAs since 2014. The
implementation of EDS (May-July) prescribed fires is part of the programme, and aims
to reduce the extent of wildfires, protect fire sensitive ecosystems and convert the
regime to patchy, low-intensity and severity, EDS fires (Schmidt et al., 2018).
Accordingly, these prescribed fires are being tested for feasibility, meeting conservation
goals and at the same time the PA's local specificities. Positive outcomes, such as 40-
57% reduction in LDS fires in the first three years, already have triggered the
implementation of the IFM programme in other eight PAs and 11 Indigenous
Territories, and in the next few years it is likely to expand to all Cerrado PAs (Schmidt
et al., 2018).

As a flexible management strategy to better guide and inform decision-making

processes, the IFM programme has been scientifically monitored in Chapada das Mesas
National Park (CMNP) since 2014 to provide ecological feedbacks and better achieve
management goals, following examples from Kruger and Kakadu National Parks
(Edwards et al., 2003; Parr et al., 2009; van Wilgen & Biggs, 2011). We have
established permanent plots in CMNP to assess potential changes in the structure of
open savanna physiognomy as a result of the implications of different fire management
systems. Our research is addressed to identifying the patterns of mortality, topkill and
recruitment among woody plants subjected to different fire regimes.

2. Material and method

2.1 Study areas

The CMNP was one of the first PAs to conduct prescribed burns for
management purposes within the IFM pilot program. The park is located in northern

50
Cerrado, northeast of Brazil in the state of Maranhão (7°19’0”S and 47°20’6”W, Fig 1).
It was created in 2005 with 160,000ha, aiming to conserve the natural landscape -
including its distinctive large vegetation range, water sources and characteristic mesa
geological formations - and local biodiversity (Ab’Sáber, 2003; Castro & Martins,
1999; da Silva et al., 2017; ICMBio, 2018; Medeiros et al., 2008).

Fig. 1. Map with the location of Chapada das Mesas National Park in regional, national and
continental scale, where the experiment for this study took place. Font: Victor Ferreira, 2015.

Most of CMNP is covered by well-drained sandy soils and open savanna

vegetation. The park is located in a transition zone between savanna, Amazonian
tropical forest and Caatinga vegetation formations. Within the region, fire-sensitive
(such as swampy forest, riparian and gallery forests) and fire-adapted vegetation (e.g.
grasslands and savannas) types occur in mosaics (Aquino et al., 2007; Martins &
Oliveira, 2011; Medeiros et al., 2008). The region receives annually from 1,500 to 1,900
mm of rainfall, with variations in mean annual air temperature from 25 to 31°C and
relative humidity (RH) from 44 to 83% over the last ten years (INMET, 2017). The
rainy season is concentrated between October and April, receiving from 1,300 to 1,600
mm rainfall, while the dry season (May-September) usually receives less than 200 mm
(INMET, 2017). In this high rainfall tropical region, open savanna areas produce large

51
volume of biomass ~3.16 ± 0.06 t/ha (mean ± standard error) during the first year after a
fire (Moura, unpublished data), which combined with fire suppression efforts
commonly propagated extensive wildfires, burning up to 70% of the park (CMNP
manager, personal communication, 2014).

2.2 Experimental design

To assess how different management strategies and fire regimes can affect the
dynamics of woody species, we established 56 plots (50 x 50 m) distributed in 12
different areas in open savannas inside the CMNP to perform experimental fires (Tab.
1). To facilitate the description of our findings, we defined as biennial fire frequency the
plots established in areas that had burnt two years before our experimental fires, and as
triennial fire frequency the ones in areas burnt three years prior to experimental
burnings. We carried out prescribed fires in the months of May and September of 2015
to simulate: EDS fires applied as part of the fire management policies recently
introduced in Cerrado PAs and LDS fires which are a common result of fire suppression
policies implemented in Brazilian PAs for decades. Within the EDS fires set as part of
the new fire management policy implementation, PA managers can opt for different
environmental conditions that can result in different fire behavior (Moura et al. in prep.,
chapter 2 of this dissertation). Therefore we carried out two types of EDS fires,
resulting in four experimental treatments: (i) LDS fires (LF) - carried out at mid-day
(12:00-14:30) with relative humidity <21% and average air temperature >35ºC in
September, simulating LDS wildfires (Fig 2); (ii) high-intensity, EDS fires (High-EF) -
carried out at mid-day (11:00-16:30) with relative humidity <50% and average air
temperature >32ºC in May (Fig 3); (iii) low-intensity, EDS fires (Low-EF) - carried out
during the evening (17:30-19:00) with relative humidity >50% and average air
temperature <31ºC in May (Fig 4); and (iv) no fire - NF (Fig 5). All treatments were
applied to two experimental plots (50x50m) within each experimental area. Biennial
plots received the four treatments, whereas there was no LDS experimental fires in
areas unburned for three years, i.e. there were no LF triennial plots, since park managers
considered wildfire risks too high.

52
Table 1. Number of studied areas, plots and prescribed fires within each treatment in the PNCM
in 2015. Study plots established in 2015 are 50 x 50 m; LF = high-intensity, LDS fires, carried
out at mid-day with average relative humidity <21% and average air temperature >35ºC; High-
EF = high-intensity, EDS fires, carried out at mid-day with average relative humidity <50% and
average air temperature >32ºC; Low-EF = low-intensity, EDS fires, carried out in the evening
with average relative humidity >50% and average air temperature <31ºC; NF = no fire. Both
Low and High-EF were carried out in May and LF in early September. TSF – time since fire in
years.
Fire type/ N study N study N fires
TSF
treatments areas plots 2015
LF 2 4 8 4
High-EF 2 4 8 4
High-EF 3 4 8 4
Low-EF 2 4 8 4
Low-EF 3 4 8 4
NF 2 4 8 0
NF 3 4 8 0
Total - 12 56 20

Fig. 2. (a) The LDS plot (LF) before experimental fire; (b) LDS fire (September, LF), carried
out at mid-day; (c) LDS plot (LF) after experimental fire.

53
Fig. 3. (a) Triennial, high-intensity EDS plot (High-EF) before experimental fire; (b) triennial,
high-intensity EDS fire, carried out at mid-day (May, High-EF); (c) biennial, high-intensity
EDS plot (High-EF) before experimental fire; (d) biennial, high-intensity EDS fire, carried out
at mid-day (May, High-EF).

Fig. 4. (a) Triennial, low-intensity EDS fire, carried out in the evening (May, Low-EF); (b)
triennial, low-intensity EDS plot (Low-EF) after experimental fire; (c) biennial, low-intensity
EDS fire, carried out in the evening (May, Low-EF); (d) biennial, low-intensity EDS plot (Low-
EF) after experimental fire.

54
Fig. 5. (a) Plot protected from fire for five years (NF); (b) Plot protected from fire for four years
(NF).

We established four areas as replicates for each treatment and TSF in 2015,
however, in 2016, two areas from the biennial NF treatment and one from the triennial
NF treatment were accidentally burned, and one from the LF treatment was burned in
2017. The areas that burned in 2016 were withdrawn from all stem response analysis,
and the LF area burned in 2017 was withdrawn from the analysis that considered
changes within two years after fire (2015-2017).

2.3 Tree response attributes

All trees inside the plots (50 x 50 m) with diameter at 30 cm from the ground
≥5cm were tagged, identified to species level and had all their stems measured
(diameter and estimated height), in 2015, 2016 and 2017, always between April-May.
We sampled approximately 3,250 tree stems per year from all our permanent plots.
Additionally, inside each plot we established a subplot (15 x 15 m) in one of the plot
edges, where around 1,650 tree stems with diameter at ground height ≥1cm were
measured (diameter and height), identified and tagged in the three study years. Each

55
year, we recorded stem response to the treatment into one of the following options: (1)
alive without topkill, here refered as survival (2) dead (without any resprouting), (3)
topkilled (complete loss of aboveground biomass with basal or underground
resprouting). We recorded new recruits (stems that achieved measurement conditions)
yearly in study plots and subplots.

2.4 Tree community assessment

We evaluated tree stem survival and topkill, number of recruitments and

changes in live basal area, tree height and stem density. For stem survival and topkill
and change in stem density we divided the stems in four diameter classes: 1-5 cm, 5-10
cm, 10-15 cm and >15 cm, since stem response is linked to stem diameter (Dantas &
Pausas, 2013; Hoffmann & Solbrig, 2003; Lawes et al., 2011; Souchie et al., 2017). The
average number of stems per hectare in the first survey (2015) decreased from smaller
to larger diameter classes in all treatments and fire intervals (Table 2). Hence, we
aggregated these stems in the last diameter class (15-49 cm) because very few stems
were larger than 15 cm in our plots. We counted the number of stem recruitment for two
diameter classes: < 5 cm in the subplots and ≥ 5 cm in the plots. To evaluate change in
tree height we considered four classes: regenerant (median ≈ 1.8 m), median (median ≈
2.6 m), crown (median ≈ 4) and emergent (median ≈ 8.1 m). These classes were defined
by the first quartile, median, third quartile and maximum height observed in both plots
from each area that received the same treatment.

Table 2. Number of stems per hectare within the diameter classes, fire frequency and
treatments. Values were calculated from the average of stem density per area in 2015, before
experimental fires in the CMNP.

Diameter class Two years without fire Three years without fire
(cm) LF High-EF Low-EF NF High-EF Low-EF NF
0-5 1815±570 1194±451 967±245 1022±47 1100±332 1372±298 1104±293
5-10 83±11 112±15 117±20 84±13 76±8 103±11 121±21
10-15 24±4 39±6 49±7 40±3 67±10 58±9 71±6
15> 24±3 27±5 26±9 28±0 53±9 49±2 59±9

To compare stem survival and topkill we used relative frequency (%) from the
absolute frequency of all response attributes per area. We calculated basal area, stem
density and number of recruits for each area (two plots = 0.5 ha and two subplots =

56
0.045 ha) per treatment and extrapolated the results to 1 hectare to facilitate
comparisons to other studies. We used 2016 data to calculate stem response to 2015
experiment fires (survival and topkill), since fire effects would be clearer to detect. We
analysed the number of recruits, change in live basal area, tree height and stem density
from the data sampled in 2017, two years after the first survey, since a larger period of
time would better indicate changes in tree growth. To calculate changes in basal area
and tree height we included all stems sampled in 2015 and stems considered alive or
recruitment in 2017 and excluded those that died, were topkilled or lost. For changes in
stem density we counted all stems in 2015 and only the live stems and recruitments for
2017.

2.5 Data analysis

We tested how tree communities responded to different fire regimes -

represented by different fire frequencies and seasonality and weather conditions - using
generalized linear model (GLM) and analysis of variance (ANOVA). Absolute and
relative frequencies of stems were used as our response variable for assessing stem
recruitment, survival and topkill. Stem density, tree diameter and height were accounted
as continuous response variables for assessing changes in stem density, live basal area
and height. All four of our predictor variables were categorical: time since fire - TSF
(two and three years), treatments (NF, LF, High-EF and Low-EF), tree diameter class
(1-5 cm, 5-10 cm, 10-15 cm and >15 cm) and tree height classes (regenerant, median,
crown and emergent). We used GLM with normal distribution to test differences in live
basal area change within treatments and TSF; GLM with Poisson distribution to test
variation on stem survival, topkill and recruitment within treatments, TSF and diameter
classes; and three-way ranked ANOVA for non-parametric, negative data to test if
changes in stem density and tree height were associated with treatments, TSF and
diameter or height classes. Pair-wise comparisons of means were tested a posteriori to
contrast least significant differences (p ≤ 0.05) among the predictions in each model
(library "emmeans", "lsmeans" package (R), Lenth, 2017). All analysis and graphics
were run by the statistical software R (R-Core-Team, 2017) using libraries "car",
"dplyr", "plyr", "gstat", "emmeans", "ggpubr" and "gridGraphics" (Fox & Weisberg,
2011; Kassambara, 2017; Lenth, 2016; Paul Murrell, 2018; Pebesma, 2004; Wickham,

57
2011). Stem density and recruitment, survival and topkill percentages are given in
median values, while growth rates, such as basal area and height, are given in mean and
± standard error.

3. Results

3.1 Stem density and recruitment

Number of stem recruitment in smaller diameter size (<5 cm) was

approximately ten times higher than the recruitment of stems of larger diameter sizes
(>5 cm, Fig 2). The recruitment of stems >5 cm was lower in all high-intensity fire
treatments (medians: LF=4 stems/ha, biennial High-EF=5 stems/ha and triennial High-
EF=6 stems/ha) and shorter fire interval (biennial Low-EF=11stems/ha) compared to
triennial fires (Low-EF=16 stems/ha, Fig 6a). In biennial LF plots, larger stems
recruited significantly less than in triennial Low-EF plots (median: 16 stems/ha),
biennial NF (20 stems/ha) and triennial NF (46 stems/ha), whereas the stem recruitment
in triennial Low-EF did not differ from the biennial NF. Among our biennial plots, stem
recruitment to small diameter sizes (<5 cm) had little variation, not differing from NF
treatment (Fig 6b). As for the triennial areas, small diameter stems in High-EF recruited
significantly less stems than in NF treatment (610 stems/ha and 977 stems/ha,
respectively), but not so different from Low-EF (899 stems/ha).

58
Fig. 6. Stem recruitment per hectare from 2015 to 2017 in experimental plots submitted to
different fire frequencies and treatments. (a) Stems with diameter >5 cm; and (b) Stems with
diameter <5 cm. Legend presents TSF (two or three years) and treatment, respectively, where:
LF = high-intensity, LDS fires (September); High-EF = high-intensity, EDS fires (May); Low-
EF = low-intensity, EDS fires (May); and NF = no fire. The boxes encompass the first and third
quartile, lines inside the boxes show the median, bars indicate minimum and maximum values,
and colored circles are the outliers. Different lower case letters show significant difference (P ≤
0.05) per diameter class, according to a posteriori contrasts between TSF and treatments.

Stem density had the largest variation range within the first diameter class (1-5
cm), and areas protected from fires for four and five years (biennial and triennial NF
treatments) were the only ones that had stem increment (122 stems/ha and 711 stems/ha,
respectively, Fig 7a). However, stem density in the first diameter class did not present
significant difference across all experimental fire treatments and biennial NF, except
from triennial NF. These variations among small stems are probably linked to the larger
density of stems in this diameter class prior to burnings in 2015 and high percentages of
topkill (>60%) after experimental fires. Even though stem recruitment to diameters <5
cm was high (>422 stems/ha) in all areas, it did not recover the initial number of stems
in 2015, before the experimental fires. There was an increase of 36 stems/ha in the
second diameter class (5-10cm) in areas protected from fires for five years (triennial
NF), whereas in areas protected for four years (biennial NF) this increase was of 5
stems/ha. Among the experimental fire treatments, stem density of individuals in the
second diameter class increased in the triennial Low-EF (3 stems/ha) and decreased in
the other treatments, losing the maximum number of stem in the LF treatment (22

59
stems/ha). High and Low-EF treatments in the triennial areas had similar change in stem
density to areas protected from fire for four and five years (Fig 7b). Stem recruitment to
diameters >5 cm was the highest in triennial Low-EF (16 stems/ha) among all fire
treatments, leading to the increase in stem density. There was a reduction in stem
density of larger trees (> 10 cm, encompassing our third and fourth diameter classes) in
all treatments during our study, including those protected from fires (NF, Fig 7c, d).

Fig. 7. Change in stem density per hectare in different diameter classes submitted to different
fire frequencies and treatments. Diameters <5 cm were measured at ground level and diameters
>5 cm were measured 30 cm from the ground. (a) Stems from 1-5 cm; (b) stems from 5-10 cm;
(c) stems from 10-15 cm; and (d) stems >15 cm in diameter. Legend presents TSF (two or three
years) and treatment, respectively, where: LF = high-intensity, LDS fires (September); High-EF
= high-intensity, EDS fires (May); Low-EF = low-intensity, EDS fires (May); and NF = no fire.
The boxes encompass the first and third quartile, lines inside the boxes show the median, bars
indicate minimum and maximum values, and colored circles are the outliers. Different lower
case letters show significant difference (P ≤ 0.05) per diameter class, according to a posteriori
contrasts between TSF and treatments.

60
3.2 Stem survival and topkill

The stems in the first diameter class (1-5 cm) submitted to biennial LF and to
triennial High-EF had the lowest survival (i.e. persistence of the aerial plant structure)
percentage (3% and 8%, respectively), with no significant difference between them (Fig
8). Survival percentages of these small stems were also low after biennial High-EF
(10%) and Low-EF (11%), which were similar to survival in response to triennial High-
EF (8%) and lower than after Low-EF (16%). Stem survival was high in both no-fire
treatments, biennial (83%) and triennial areas (97%), and were significantly higher than
all experimental burned areas. Stems in the second diameter class (5-10 cm) survived
notably more than the ones in the first class. Yet, LF presented the lowest stem survival
(72%), which was significantly lower than both NF areas. The survival of stems >10 cm
in diameter did not significantly differ across treatments and fire frequencies, but stems
in LF and High-EF plots tended to survive less.

Fig. 8. Stem survival within four diameter classes submitted to different fire frequencies and
treatments. Legend presents TSF (two or three years) and treatment, respectively, where: LF =
high-intensity, LDS fires (September); High-EF = high-intensity, EDS fires (May); Low-EF =
low-intensity, EDS fires (May); and NF = no fire. The boxes encompass the first and third
quartile, lines inside the boxes show the median, bars indicate minimum and maximum values,
and colored circles are the outliers. Different lower case letters show significant difference (P ≤
0.05) per diameter class, according to a posteriori contrasts between TSF and treatments.

Topkill (loss of all aerial tissue followed by resprout) was higher among stems
from the first diameter class in fire treatments, where significant differences laid

61
between both NF (3-4%) areas and the other treatments, Low-EF (56%) and High-EF
(77%) from the triennial areas, and the triennial Low-EF and LF (72%, Fig 9). In the
second diameter class, LF treatment presented the highest stem topkill (20%), and
biennial Low-EF (6%) had more stem topkills than biennial High-EF (2%) and both NF
(0) areas. Conversely, topkills in biennial Low-EF were similar to triennial Low and
High-EF (both 3%), with no significant difference. For the other two diameter classes
(>10 cm) stem topkill was significantly higher in LF (4%) when compared to the other
treatments that burned in 2015, but not different from NF (0-5%).

Fig. 9. Stem topkill within four diameter classes submitted to different fire frequencies and
treatments. Legend presents TSF (two or three years) and treatment, respectively, where: LF =
high-intensity, LDS fires (September); High-EF = high-intensity, EDS fires (May); Low-EF =
low-intensity, EDS fires (May); and NF = no fire. The boxes encompass the first and third
quartile, lines inside the boxes show the median, bars indicate minimum and maximum values,
and colored circles are the outliers. Different lower case letters show significant difference (P ≤
0.05) per diameter class, according to a posteriori contrasts between TSF and treatments.

3.3 Tree growth rates

There was an increment in total basal areas in all study areas and treatments,
except in LF and biennial High-EF treatments. The change in basal area was similar
across all experimental burned plots, but different from the ones protected from fire for
5 years (triennial NF, Fig 10). The largest loss in basal area was in the LF treatment
(0.375 ± 0.04 m²ha-1), which was significantly different from biennial and triennial NF

62
(0.544 ± 0.17 and 0.930 ± 0.18 m²ha-1, respectively), where we found the largest basal
area increment (Fig 6).

Fig. 10. Change in basal area under two fire intervals (two and three years) submitted to
different fire treatments. Legend presents TSF (two or three years) and treatment, respectively,
where: LF = high-intensity, LDS fires (September); High-EF = high-intensity, EDS fires (May);
Low-EF = low-intensity, EDS fires (May); and NF = no fire. The boxes encompass the first and
third quartile, lines inside the boxes show the median, bars indicate minimum and maximum
values, and colored circles are the outliers. Different lower case letters show significant
difference (P ≤ 0.05) per diameter class, according to a posteriori contrasts between TSF and
treatments.

Almost all the trees in our plots from all treatments and areas grew in height
two years after the experimental fires (Fig 11). The trees that grew the least in height
were the regenerant class (median height = 1.8 m), where trees subjected to the LF
treatment had the smallest height increment (0,02 ± 0.09 m, Fig 11a) of all classes.
Larger increment in height was measured in the emergent class (median height = 8.1),
and the highest increase was observed in the triennial Low-EF (2,62 ± 0.31 m, Fig 11d).
None of the differences observed in height change were significant among the
treatments or areas in each tree height class, but generally trees in the triennial areas
grew more in height than biennial ones.

63
Fig. 11. Change in tree height within different classes submitted to different fire frequencies and
treatments. Height classes were distributed in: (a) regenerant, with median height of 1.8 m; (b)
median, with median height of 2.6 m; (c) crown, with median height of 4 m; and (d) emergent,
with median height of 8.1 m. Legend presents TSF (two or three years) and treatment,
respectively, where: LF = high-intensity, LDS fires (September); High-EF = high-intensity,
EDS fires (May); Low-EF = low-intensity, EDS fires (May); and NF = no fire. The boxes
encompass the first and third quartile, lines inside the boxes show the median, bars indicate
minimum and maximum values, and colored circles are the outliers. Different lower case letters
show significant difference (P ≤ 0.05) per diameter class, according to a posteriori contrasts
between TSF and treatments.

4. Discussion

4. 1 Short-term, implications of fire management systems to tree communities

In our experiment, stem recruitment, and density (among diameters <10 cm)
decreased under higher fire frequency (biennial) and more extreme fire weather
conditions (LDS, mid-day fires). Tree recruitment responded to fire seasonality in the
central Cerrado region, where within a period of 10 years there were 480, 148 and 60
recruitments per hectare (to diameters >5 cm) after fires with gradually increasing
intensities in June, August and September, respectively (Sato et al., 2010). However, the

64
seed bank and vegetative reproduction of several species can benefit from fire passage
in different periods of the year and can stimulate the recruitment of less competitive
species (Keeley & Fotheringham, 2000; Whelan, 1995) in savannas (Andrade &
Miranda, 2010; Williams et al., 2005), woodlands (Enright & Lamont, 1989; Keeley,
1987; Roche et al., 1998), shrublands (Martin, 1995; Pierce & Cowling, 1991; Zammit
& Zedler, 1988) and grassland vegetation (Benson & Hartnett, 2006; le Stradic et al.,
2015). Several shrub species recruit more under frequent, high-intensity fires in the
semi-arid woodland in eastern Australia (Hodgkinson, 1991; Knox & Clarke, 2006), as
well as exotic fire-adapted species in fire-prone ecosystems (Brooks et al., 2004;
Haidinger & Keeley, 2017).

Frequent and LDS fire regimes have resulted in the reduction of tree abundance
in other Cerrado ecosystems (Hoffmann, 1999; Lima et al., 2009; Medeiros & Miranda,
2005; Ribeiro et al., 2012; Sato, 2003), northern Australian savannas (Braithwaite &
Estbergs, 1985; Russell-Smith & Edwards, 2006; Russell-Smith et al., 2003; Williams
et al., 1999; Yates et al., 2008), southern African savannas (Eckhardt et al., 2000;
Higgins et al., 2007; Roques et al., 2001; van Wilgen et al., 2003) and oak savannas in
North America (Peterson & Reich, 2001; Tester, 1989). Additionally, frequent fires may
limit species composition (Durigan et al., 1994; Silva & Batalha, 2010) and favor those
with vegetative reproduction (Clarke et al., 2013; Hoffmann, 1998; Setterfield, 2002),
and the combination of high-intensity and severity, frequent fires can lead to the
dominance of low stature, multi-stemmed trees (Bellingham & Sparrow, 2000;
Govender et al., 2006).

Severe fire weather conditions, as those presented in the LDS and triennial mid-
day, EDS in the CMNP, combined with fine fuel load are likely to increase fire intensity
(Cheney et al., 1993; Hoffmann et al., 2012; Whelan, 1995). Such combination may
increased tree stem mortality, especially in smaller diameter class stems (< 10 cm
diameter at breast height - DBH), in Cerrado physiognomies (Kauffman et al., 1994;
Medeiros & Miranda, 2005; Ribeiro et al., 2012; Sato & Miranda, 1996) and other
savanna ecosystems (Gill et al., 1990; Higgins et al., 2000; Williams et al., 2005, 1999).
In our experiment, the survival - maintenance of aboveground biomass - of small stems
(diameter <5 cm) after all types of fire was very low compared to stems that did not

65
burn for four and five years, whereas small-medium stems (diameter between >5 and
<10 cm) were more tolerant to the experimental fires, and only stems submitted to LDS
fires survived significantly less than the ones unburned. The losses in stem density of
larger trees (>10 cm) might be better explained by other factors influencing trees
survival in our experiment rather than fire, such as lower annual rainfall and prolonged
drought, once water availability also influences savanna dynamics (Medina & Silva,
1990; Skarpe, 1992), e.g. in 2015 annual precipitation (1,243 mm) was much lower than
the average of the five previous years (1,885 mm, INMET, 2017).

Generally, plants with small stems did not die one year after fire passage but
resprouted from the base or ground after losing their aboveground biomass, i.e.
topkilled. Previous studies in Cerrado indicate that a significant number of small stems
are able to recover to pre-fire size one year after burnt (Hoffmann & Solbrig, 2003;
Medeiros & Miranda, 2005), but frequent fires keep them from growing into
reproductive sizes (Hoffmann, 1998). Since our experimental fires were undertaken in
short fire intervals (two and three years after previous fires), they have probably
influenced the negative fluctuations of stem density in larger diameter classes (>5 cm)
in all biennial fire treatments. Although, short-to-medium term outcomes of a single
high-intensity wildfire can result in high mortality of medium-to-large diameter trees in
northern Australian savannas (Williams et al., 1999), larger stems (diameters >10 cm)
subjected to LDS fires survived much better than the smaller ones (>93%) in the
CMNP, as reported for other Cerrado regions in central Brazil (Medeiros & Miranda,
2005; Ramos, 1990; Rocha-Silva, 1999; Sato & Miranda, 1996). Topkill observed in
larger stems (>10 cm) in NF treatments may be a consequence of previous severe fires
that could have weakened tree structure or other elements, such as low rainfall rates,
storms that brought down trees or parasites that could have also affected our plots rather
than only our experimental fire.

Small Cerrado trees (diameter <5 cm) are more vulnerable to fire than larger
trees because larger stems tend to present thicker barks that protect living tissues from
heat or impacts (Dantas & Pausas, 2013; Guedes, 1993; Keeley et al., 2011; Miranda et
al., 2002). However, fire-adapted species are capable of resprouting after repeated
topkill by fires (Bond & Midgley, 2001; Hoffmann et al., 2009; Lawes et al., 2011). The

66
small stems subjected to the triennial mid-day, EDS fires were the most topkilled (77%)
of our experiment followed by biennial LDS fires, showing similar result to other
typical Cerrado areas, where longer fire interval resulted in increased topkill (Rocha-
Silva, 1999) and higher intensity fires caused more topkill than lower intensity fires,
especially in small trees (Hoffmann & Solbrig, 2003; Sato, 2003). In all experimental
fires, the highest stem survival (16%) and lowest topkill percentages (56%) occurred
among small trees subjected to triennial evening, EDS fires. Yet, other studies indicated
that the combination of short fire interval and low-intensity fires, commonly presenting
lower rate of spread and higher residence time at temperatures above 60ºC, can lead
small stems (with bark thickness <6 mm) to higher tree damage than fires with higher
rate of spread, such as high-intensity fires (Guedes, 1993; Uhl & Kauffman, 1990).

Long-term site productivity is affected by frequent and widespread fires in

Cerrado ecosystems (Batmanian & Haridasan, 1985), especially when more volume of
biomass is consumed by fire and greater quantities of nutrients (C, N and S) are lost
(Kauffman et al., 1994). In our experiment, the total basal area of woody plants reduced
in areas subjected to more severe fire weather conditions (mid-day EDS and LDS) and
shorter fire intervals (biennial frequency), as reported previously in other Cerrado areas
(Batmanian & Haridasan, 1985; Hoffmann, 1999). In northern Australian savannas, the
growth (in height and diameter) of eucalyptus adult tree is limited by fires in the LDS
(Murphy et al., 2010; Russell-Smith et al., 2003; Williams et al., 1999), and accelerated
by EDS fires when compared to unburned trees (Murphy et al., 2010; Prior et al., 2006;
Werner, 2005). These studies were unable to account for slow-growing species (Werner
& Prior, 2013) and local-scale variability in resources and ecological interactions
(Lehmann et al., 2009), such competition and predation, which could have masked tree
growth and survival in the experiment. Because Cerrado adult trees usually have thick
barks that protect them from fire (Coutinho, 2002; Guedes, 1993; Ramos, 1990; Simon
& Pennington, 2012; Walter & Ribeiro, 2010), high-intensity fires with high scorching
height can remove or consume parts of the dead tissue and under-estimate diameter
growths (Felfili et al., 2005), feature also reported in other savanna environments (Gill
et al., 1986; Prior et al., 2006). Stem growth is remarkably different between species
and among larger and smaller individuals (Felfili, 1995; Felfili et al., 2005; Miranda et
al., 2007; Prior et al., 2006; Werner & Prior, 2013), as well as their capacity to survive

67
fire events and keep their aboveground structures (Grady & Hoffmann, 2012; Hoffmann
& Solbrig, 2003; Lawes et al., 2011). The overall increase in basal area in our
experiment, is probably explained by diameter growth among larger stems (>10 cm),
since topkill percentage was <5% and survival percentage >83% among large trees
(third and fourth diameter classes) in all fire frequencies and treatments.

In our experiment, trees increased in height in all fire treatments, although trees
that burned in biennial LDS increased less in height than those submitted to biennial
and triennial, EDS fires. Annual fire frequencies have resulted in height loss (average of
10 cm after two years) in woody communities within the Cerrado (Armando, 1994).
Fire intensity is correlated with scorch height in African and Australian savannas, where
high-intensity fires are likely to produce greater flame heights and topkill stems and
branches, reducing trees to coppice from the stem collar region (Luke & McArthur,
1978; Trollope & Trollope, 2002; Werner & Prior, 2013; Williams et al., 1998, 1999).
Frequently in fire-prone ecosystems, woody species present strategic traits that keeps
them from being topkilled and protect their buds from fire (Bond & Keeley, 2005;
Lehmann et al., 2014; Simon & Pennington, 2012). In Brazilian and Australian
savannas bark thickness results in less mortality and biomass losses (Coutinho, 1990;
Guedes, 1993; Hoffmann & Solbrig, 2003; Lawes et al., 2011; Werner & Prior, 2013),
while in African and north American savannas escaping the "fire trap" height presents a
better strategy for trees to overcome fire and animal predation (Dantas & Pausas, 2013;
Grady & Hoffmann, 2012; Higgins et al., 2000; Staver et al., 2009). Increment in height
was reduced among regenerating trees (median height of 1.8 m) in the CMNP, with less
than 0.16 m in areas burned biennially and twice this increment in areas burned
triennially and unburned for four and five years, however, taller trees (> 4 m) are more
susceptible to height estimation errors than smaller trees. Seedlings and regenerating
trees commonly present low height increment one year after fire passage since the
process of resprouting after being topkilled slows stem growth (Braz et al., 2000;
Hoffmann & Solbrig, 2003; Miranda et al., 2009; Oliveira & Silva, 1993).

4.2 Fire management applicability and ecological precautions

The coexistence of trees and grasses in savanna ecosystems is regulated by

fireclimate and soil conditions (Accatino et al., 2010; Bond et al., 2005; Bowman et al.,

68
2009; Gardner, 2006; Higgins et al., 2000; Lehmann et al., 2014; Sankaran et al., 2005).
To guarantee the conservation of landscapes, fire sensitive ecosystems and local
biodiversity in savanna PAs, managers must permanently assess the implemented
management systems and, when necessary incorporate new strategies. The Kakadu and
Kruger National Parks are great examples of successful fire implementation in savanna
PAs that were subjected to long-term research monitoring and evaluation to help
improve their management systems (Bond & Archibald, 2003; Edwards et al., 2003;
McGregor et al., 2010; Parr et al., 2009; van Wilgen et al., 2004; Werner, 2005;
Williams, 1995). Although IFM is a pilot programme and pioneer in using fire in
Cerrado PAs (Schmidt et al., 2018), fire researchers have long been struggling to
experimentally burn study areas and advance in ecological knowledge (Coutinho, 1981;
Gomes et al., 2018; Miranda et al., 1993; Miranda, 2010; Pivello & Coutinho, 1996). In
our study, we monitored Cerrado woody communities subjected to different
management strategies, including the burning protocols established by the IFM that
consists in EDS prescribed fires.

Overall in the CMNP, biennial, LDS experimental fires that mimic the frequent
LDS wildfire regime most Cerrado PAs are subjected to, reduced tree density,
recruitment, survival and growth, as well as increased stem topkills in the open savanna
vegetation. In the same experiments, we found that mid-day, EDS fires in a biennial
frequency caused almost the same results, followed by triennial, mid-day EDS fires.
Our results indicate that evening, EDS fires are likely to have less impact on woody
communities, especially within three years interval. Therefore, prescribed fires in
savannas should not rely only upon seasonality, the weather conditions and fire interval
must be part of management plans (Cheney et al., 1993; Gill et al., 1996; Govender et
al., 2006; Hoffmann et al., 2012; Price et al., 2014), although the weather will need to
be verified in locus before burning.

Importantly, management of PAs is not exclusively dedicated in conserving tree

communities, but also to keep landscapes and local biodiversity, which may include
opening dense woody physiognomies or fragmenting the fuel load in highly flammable
unburned areas. In such situations, evening fires in the EDS are recommended since the
weather conditions - air relative humidity above 50% and cooler temperatures in mesic

69
savannas - will keep fire intensity and rate of spread at lower levels and will probably
self-extinguish when fuel moisture increase during the night or when reaching
moisturized environments, such as gallery forests or denser vegetation fragments (Gill
et al., 1996; McGuffog et al., 2001; Schmidt et al., 2018; Williamson et al., 2016). Thus
this type of fire: (i) offers more safety in hazardous flammable areas (Edwards et al.,
2015; Govender et al., 2006; Haney & Power, 1996); (ii) lower management costs and
losses (Moritz et al., 2014; Penman et al., 2011; Schmidt et al., 2018); (iii) reduces
GHG emissions by avoiding frequent wildfire incidents (Murphy et al., 2015; Russell-
Smith et al., 2009a, 2013a; UNU, 2015; van der Werf et al., 2017); (iv) creates espatial
mosaics of patchy burned vegetation (Bliege Bird et al., 2008; Laris, 2002; Oliveira et
al., 2015; Weir et al., 2000); and (v) protects fire-sensitive ecosystems from wildfires
(Bradstock et al., 2005; Maravalhas & Vasconcelos, 2014; Price et al., 2012; Russell-
Smith et al., 1998; Schmidt et al., 2018; Trauernicht et al., 2015a).

Where grassland fuel load imposes high risk to fire spread, management actions
with prescribed fires should be prioritized and burning under low relative humidity
(<50%), high air temperatures and windy weather conditions should be avoided
(Cheney et al., 1993; Just et al., 2016; Rothermel, 1972; Trollope & Trollope, 2002).
Nonetheless, high-intensity fires also plays an important role in the natural dynamics of
ecosystems (Bond & Keeley, 2005; Johnson et al., 2001; Meyn et al., 2007; Stephenson
et al., 1991; Turner & Dale, 1998). High intensity fires can be strategic to certain
management goals, such as to fragment homogeneous landscapes (Govender et al.,
2006; Marcoux et al., 2015; Roques et al., 2001; Scott et al., 2012), opening woody
encroached areas or maintaining grassland dominated landscapes (Bond & Midgley,
1995; Pausas & Moreira, 2012), eliminate invasive species (Brooks et al., 2004) and
favour several plants' life stage or trait (Armstrong & Phillips, 2012; Bellingham &
Sparrow, 2000; Keeley & Zedler, 2009; Werner, 2005; Whelan et al., 2002). In this
scenarios, the interaction between LDS fires or fires carried out in more severe weather
conditions during the dry season and flammable vegetation structure can create higher
intensity fires and result in these desired, environmental outcomes. Notably, this is only
applicable to fragmented areas, where natural or anthropogenic barriers will prevent
fires from spreading to larger extents turning into uncontrolled wildfires.

70
 Finally, every management system is susceptible to environmental changes and
failing to acknowledge them will rise the risks of initiating wildfires (Fernandes, 2013;
Moritz et al., 2012; Roos et al., 2016), therefore, managers and researchers' feedbacks
will constantly enhance the systems' reliability (Christensen, 2005; Driscoll et al., 2010;
van Wilgen et al., 2007). We point out that these guidelines have been addressed to
inform how tree communities respond to several fire management arrangements,
whereas local and regional characteristics may require readjustments or even different
evaluative protocols to deliver better results (Dickinson & Ryan, 2010; Driscoll et al.,
2010; Gomes et al., 2018; Groffman et al., 2006). Cerrado woody species present slow-
growing patterns, therefore, our results are limited to a short-term monitoring period
and further lengthier studies and continuous monitoring in PAs will probably provide a
more complete assessment.

71
Conclusão

A partir desta tese foi possível: (i) sintetizar políticas de manejo do fogo nas
principais regiões savânicas do hemisfério sul, e fazer recomendações para a
recentemente implementada política de manejo do fogo no Cerrado; (ii) caracterizar o
comportamento de queimas de manejo no início da estação seca e de queimas tardias,
que simulam os frequentes incêndios no Cerrado e (iii) quantificar os efeitos de
diferentes regimes de queima em comunidades lenhosas em áreas de cerrado sentido
restrito no Parque Nacional da Chapada das Mesas. A recente institucionalização do uso
do fogo no manejo de unidades de conservação (UC) do Cerrado, possibilitou a
investigação científica em escala de paisagem deste trabalho, e a realização dos
experimentos de acodo com parâmetros utilizados em campo para se estabelecer as
queimas prescritas do manejo integrado do fogo (MIF) desde 2014. Com o
entendimento do processo de mudança da política do fogo, bem como dos regimes de
queima estabelecidos por cada sistema, foi possível elaborar experimentos de pesquisa
que subsidiem todas as etapas do manejo (planejamento, implementação e avaliação,
Schmidt et al., 2016), incluindo alguns parâmetros que podem ser utilizados nesta
transição.

Uma das metas de manejo mais frequente em formações savânicas é diminuir o

número de ocorrência e a extensão de incêndios através da redução de material
combustível (Goldammer & de Ronde, 2004; Govender et al., 2006; Russell-Smith,
2016; Schmidt et al., 2018). Embora quando esta meta é exclusiva seus resultados sejam
limitados e questionáveis, reduzir o risco de incêndios tem sido um dos maiores
incentivos para a expansão do programa de MIF no Cerrado. Outros países também
priorizaram o manejo do combustível com queimas controladas em ecossistemas
pirofíticos ao quebrarem o paradigma de exclusão do fogo em atividades de manejo, e
com o tempo conseguiram aprimorar as técnicas e incluir novas metas de conservação
(Finney, 2001; Laris, 2002; Penman et al., 2011; Russell-Smith et al., 2009b; Stephens
& Ruth, 2005; Wardell et al., 2004) e até gerar renda para populações rurais (Murphy et
al., 2015; Russell-Smith et al., 2007). Neste sentido, na fase piloto do MIF as
instituições ambientais, como o Instituto Chico Mendes da Biodiversidade (ICMBio) e
o Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais (IBAMA), estão

72
priorizando as ações de manejo que tenham o menor impacto possível na conservação e
funcionamento dos ecossistemas manejados. Dentro dessas ações de manejo, as UC
participantes do MIF estão pela primeira vez conduzindo queimas prescritas no início
da estação seca em horários estratégicos para que o combustível seja consumido, as
comunidades presentes não sofram grandes impactos e o fogo não queime áreas
indesejadas, especialmente aquelas de vegetação sensíveis ao fogo como as matas de
galeria.

Queimas no início da estação seca no final da tarde, em que as condições

meteorológicas são mais amenas (temperatura do ar e velocidade do vento mais baixa e
umidade relativa mais alta), tendem a ter velocidade de propagação e intensidade do
fogo mais baixas e a consumir menor quantidade de combustível do que queimas de
meio-dia no início da estação seca ou queimas no final da estação seca. Além de serem
mais fáceis de controlar, muitas vezes se auto-extinguindo durante a noite, estas
queimas de menor intensidade causam menor dano nas comunidades de árvores do
cerrado típico. De maneira geral, intervalos de queima maiores (três anos) apresentaram
intensidades de fogo maiores do que intervalos menores (dois anos), em que queimas no
início da estação seca ao meio-dia em áreas com histórico de fogo de três anos geraram
fogos de intensidade maior (estatisticamente equivalente) do que queimas em áreas com
histórico de dois anos sem fogo no final da estação seca. Embora estas queimas de
maior intensidade em condições meteorológicas mais severas resultem em taxa de
sobrevivência e topkill das árvores similares em ambos os intervalos de queima, as
queimas do início da estação seca no final da tarde em áreas sem queimar há três anos
resultaram nos menores impactos, do que as áreas queimadas durante o dia. Enquanto
que mesmo as queimas em condições mais amenas causaram aumento de mortalidade e
perda de parte aérea (topkill) quando comparadas a áreas protegidas do fogo por quatro
e cinco anos.

De maneira preliminar e a curto-prazo, foram observadas potenciais mudanças

na dinâmica e estrutura da vegetação arbórea do Cerrado no PNCM, como: a
diminuição do número de indivíduos de lenhosos, o atraso no crescimento de plantas,
especialmente das plantas menores (diâmetro <5cm) como resultado de queimas no
final da estação seca, que caracterizam o regime de fogo predominante na maioria das

73
UC do Cerrado (França, 2010; Pereira Júnior et al., 2014; Pivello, 2011). Apesar das
queimas de fim-de-tarde no início da estação seca também resultarem na redução de
indivíduos arbóreos, especialmente daqueles com fustes pequenos (<5 cm de diâmetro),
esta redução é significativamente menor do que queimas de meio dia na mesma época
ou no final da estação seca e a taxa de crescimento das árvores é comparável a de áreas
protegidas do fogo por quatro anos. Assim, este sistema de fogo de fim de tarde no
início da estação seca, embora fracamente descrito na literatura voltada para o Cerrado
(Hoffmann et al., 2012; Schmidt et al., 2016b, 2018), mostrou-se, pelo menos a curto
prazo, como o mais eficiente para atingir os objetivos atuais do MIF em UC.

Uma vez que as queimas prescritas são comumente implementadas por

moradores locais em zonas rurais brasileiras ao longo da estação seca em diferentes
horários (Batista et al., 2018; Borges et al., 2016; Eloy et al., 2018; Mistry et al., 2018;
Moura, 2013), os resultados obtidos com esta pesquisa também podem contribuir na
investigação sobre o comportamento e consequências deste diferentes tipos de fogo na
vegetação arbórea. Esta tese foi desenvolvida para melhorar as informações disponíveis
na academia científica, nas instituições ambientais e comunidades que utilizam o fogo
como instrumento para o manejo da terra. Desta forma, esta pesquisa contribuiu: (i) para
informar moradores do PNCM, gestores e brigadistas sobre a ecologia do fogo; (ii) na
formação de alunos de graduação e pós-graduação que acompanharam o
desenvolvimento do projeto em laboratório e nas práticas de campo; e (iii) no fomento
de discussões voltadas tanto para o conhecimento técnico-científico como para o
conhecimento tradicional/local-empírico.

Pesquisas voltadas para o manejo podem auxiliar na tomada de decisão, e na

avaliação e adaptação de sistemas implementados por órgãos ambientais, desde que:
sejam guiadas pelos praticantes; levem em consideração o conhecimento tradicional e
científico e a experiência profissional dos gestores envolvidos; e considerem a realidade
e especificidades da região estudada (Christensen, 2005; Driscoll et al., 2010; van
Wilgen et al., 2007). As recomendações sobre os tipos de queimas que deve ser
implementadas pelo MIF depende integralmente dos objetivos de cada UC e,
consequentemente, das metas estabelecidas para cumpri-los. Após definir as metas, é
importante ressaltar que fatores como material combustível, intervalo de queima,

74
precipitação anual, condições meteorológicas (determinadas também pelo horário do
dia), época do ano e aceiros e barreiras naturais ou artificiais podem ser usados como
parâmetros para realizar as queimas prescritas e atingir os resultados desejados.

Embora esta pesquisa tenha sido desenvolvida em uma escala local, foi possível
acompanhar vários tipos de queimas em escala de paisagem, abrangendo diversas
comunidades vegetais e suas interações com as queimas, bem como dar retorno a equipe
do PNCM e produzir informações valiosas para a continuidade e expansão do programa
de MIF em UC. Os resultados aqui presentes estão limitados a um período de
monitoramento curto (três anos) e ao acompanhamento exclusivo de espécies arbóreas
como os indicadores ecológicos dos experimentos. Portanto, pesquisas de longo-prazo
que incluam outros indicadores ambientais e envolvam maiores escalas espaciais,
embora difíceis, poderão contribuir ainda mais para aumentar a eficiência do manejo do
fogo no Cerrado.

Referências

Ab’Sáber, A. N. (2003). Os domínios de natureza no Brasil: potencialidades

paisagísticas (2nd ed.). São Paulo: Ateliê.
Abreu, R. C. R., Hoffmann, W. A., Vasconcelos, H. L., Pilon, N. A., Rossatto, D. R., &
Durigan, G. (2017). The biodiversity cost of carbon sequestration in tropical
savanna. Science Advances, 3(8), e1701284.
https://doi.org/10.1126/sciadv.1701284
Accatino, F., De Michele, C., Vezzoli, R., Donzelli, D., & Scholes, R. J. (2010). Tree-
grass co-existence in savanna: Interactions of rain and fire. Journal of Theoretical
Biology, 267(2), 235–242. https://doi.org/10.1016/j.jtbi.2010.08.012
Alden Wily, L. (2008). Custom and commonage in Africa: rethinking the orthodoxies.
Land Use Policy, 25, 43–52.
Altman, J. (2009). Manwurrk (fire drive) at Namilewohwo. In Culture, ecology and
economy of savanna fire management in northern Australia: Rekindling the Wurrk
tradition (pp. 165–180). Melbourne: CSIRO Publications.
Andrade, L. A. Z., & Miranda, H. S. (2010). O fator fogo no banco de sementes. In H.
S. Miranda (Ed.), Efeitos do regime do fogo sobre a estrutura de comunidades de
cerrado: Resultados do Projeto Fogo (pp. 103–120). Brasília, DF: MMA,
IBAMA.
Aquino, F. D. G., Walter, B. M. T., & Ribeiro, J. F. (2007). Woody community
dynamics in two fragments of “cerrado” stricto sensu over a seven-year period
(1995-2002), MA, Brazil. Revista Brasileira de Botânica, 30(1), 113–121.
https://doi.org/10.1590/S0100-84042007000100011

75
Archer, R., Russell-Smith, J., Kerins, S., Costanza, R., Edwards, A., & Sangha, K.
(2018). Change and continuity: the North Australia cultural landscape. In J.
Russell-Smith, H. Pedersen, G. James, & K. Sangha (Eds.), Sustainable land
sector development in northern Australia: Indigenous rights, aspirations, and
cultural responsibilities. USA: CRC Press.
Archibald, S. (2016). Managing the human component of fire regimes: lessons from
Africa. Philosophical Transactions of the Royal Society B: Biological Sciences,
371(1696), 20150346. https://doi.org/10.1098/rstb.2015.0346
Archibald, S., Lehmann, C. E. R., Gómez-dans, J. L., & Bradstock, R. A. (2013).
Defining pyromes and global syndromes of fire regimes. Proceedings of the
National Academy of Sciences, 110(16), 6442–6447.
https://doi.org/10.1073/pnas.1211466110/-
/DCSupplemental.www.pnas.org/cgi/doi/10.1073/pnas.1211466110
Archibald, S., Roy, D. P., Wilgen, B. W. van, & Scholes, R. J. (2009). What limits fire?
An examination of drivers of burnt area in Southern Africa. Global Change
Biology, 15, 613–630. https://doi.org/10.1111/j.1365-2486.2008.01754.x
Archibald, S., Scholes, R. J., Roy, D. P., Roberts, G., & Boschetti, L. (2010). Southern
African fire regimes as revealed by remote sensing. International Journal of
Wildland Fire, 19(7), 861–878. https://doi.org/10.1071/WF10008
Armando, M. S. (1994). O impacto do fogo na rebrota de algumas espécies de árvores
do Cerrado. Universidade de Brasília.
Armstrong, G., & Phillips, B. (2012). Fire history from life-history: Determining the
fire regime that a plant community is adapted using life-histories. PLoS ONE, 7(2),
1–8. https://doi.org/10.1371/journal.pone.0031544
Arruda, F. V. de, Sousa, D. G. de, Teresa, F. B., Prado, V. H. M. do, Cunha, H. F. da, &
Izzo, T. J. (2018). Trends and gaps of the scientific literature about the effects of
fire on Brazilian Cerrado. Biota Neotropica, 18(1), 1–6.
https://doi.org/10.1590/1676-0611-bn-2017-0426
Bar-Yosef, O. (2002). The upper Paleolithic revolution. Annual Reviews of
Anthropology, 31, 363–393.
Barradas, A. C. S. (2017). A gestão do fogo na Estação Ecológica Serra Geral do
Tocantins. Escola Nacional de Botânica Tropical, Rio de Janeiro.
Batista, E. K. L., Russell-Smith, J., França, H., & Figueira, J. E. C. (2018). An
evaluation of contemporary savanna fire regimes in the Canastra National Park,
Brazil: Outcomes of fire suppression policies. Journal of Environmental
Management, 205, 40–49. https://doi.org/10.1016/j.jenvman.2017.09.053
Batmanian, G. J., & Haridasan, M. (1985). Primary production and accumulation of
nutrients by the ground layer community of cerrado vegetation of central Brazil.
Plant and Soil, 88(3), 437–440. https://doi.org/10.1007/BF02197500
Beerling, D. J., & Osborne, C. P. (2006). The origin of the savanna biome. Global
Change Biology, 12(11), 2023–2031. https://doi.org/10.1111/j.1365-
2486.2006.01239.x

76
Bellingham, P. J., & Sparrow, A. D. (2000). Resprouting as a life history strategy in
woody plant communities. Oikos, 89(2), 409–416.
Bellomo, R. V. (1994). Methods of determining early hominid behavioral activities
associated with the controlled use of fire at FxJj 20 Main, Koobi Fora, Keny.
Journal of Human Evolution, 27, 173–195.
Benson, E. J., & Hartnett, D. C. (2006). The role of seed and vegetative reproduction in
plant recruitment and demography in tallgrass prairie. Plant Ecology, 187(2), 163–
177. https://doi.org/10.1007/s11258-005-0975-y
Berkes, F., Colding, J., & Folke, C. (2000). Rediscovery of traditional ecological
knowledge as adaptive management. Ecological Applications, 10(5), 1251–1262.
Bird, M. I., & Cali, J. A. (1998). A million-year record of fire in sub-Saharan Africa.
Nature, 394(6695), 767–769. https://doi.org/10.1038/29507
Bliege Bird, R., Bird, D. W., Codding, B. F., Parker, C. H., & Jones, J. H. (2008). The
‘fire stick farming’ hypothesis: Australian aboriginal foraging strategies,
biodiversity, and anthropogenic fire mosaics. Proceedings of the National
Academy of Sciences, 105(39), 14796–14801.
Bond, W. J. (2008). What Limits Trees in C 4 Grasslands and Savannas? Annual
Review of Ecology, Evolution, and Systematics, 39(1), 641–659.
https://doi.org/10.1146/annurev.ecolsys.39.110707.173411
Bond, W. J. (2016). Ancient grasslands at risk. Science, 351(6269), 120–122.
https://doi.org/10.1126/science.aad5132
Bond, W. J., & Archibald, S. (2003). Confronting complexity: fire policy choices in
South African savanna parks. International Journal of Wildland Fire, 12(3–4),
381–389. https://doi.org/10.1071/wf03024
Bond, W. J., & Keeley, J. E. (2005). Fire as a global ‘ herbivore ’: the ecology and
evolution of flammable ecosystems, 20(7).
https://doi.org/10.1016/j.tree.2005.04.025
Bond, W. J., & Midgley, J. J. (1995). Kill thy neighbour : an individualistic argument
for the evolution of flammability. Oikos, 73, 79–85.
Bond, W. J., & Midgley, J. J. (2001). Ecology of sprouting in woody plants : the
persistence niche, 16(1), 45–51.
Bond, W. J., & Parr, C. L. (2010). Beyond the forest edge: Ecology, diversity and
conservation of the grassy biomes. Biological Conservation, 143(10), 2395–2404.
https://doi.org/10.1016/j.biocon.2009.12.012
Bond, W. J., Woodward, F. I., & Midgley, G. F. (2005). The global distribution of
ecosystems in a world without fire. New Phytologist, 165, 525–538.
Borges, S. L., Eloy, L., Schmidt, I. B., Barradas, A. A. S., & Santos, I. A. (2016). Fire
management in veredas (palm swamps): new perspectives on traditional farming
systems in Jalapão, Brazil. Ambiente & Sociedade, 19(3), 269–294.
https://doi.org/10.1590/1809-4422ASOC20150020R1V1932016
Bowman, D. M. J. S., Balch, J., Artaxo, P., Bond, W. J., Cochrane, M. A., D’Antonio,
C. M., … Swetnam, T. W. (2011). The human dimension of fire regimes on Earth.

77
 Journal of Biogeography, 38(12), 2223–2236. https://doi.org/10.1111/j.1365-
2699.2011.02595.x
Bowman, D. M. J. S., Balch, J. K., Artaxo, P., Bond, W. J., Carlson, J. M., Cochrane,
M. A., … Pyne, S. J. (2009). Fire in the Earth system. Science, 324, 481–484.
Bowman, D. M. J. S., & Panton, W. J. (1993). Decline of Callitris intratropica R.T.
Baker & H.G. Smith in the Northern Territory: Implications for pre- and post-
European colonization fire regimes. Journal of Biogeography, 20(4), 373–381.
Bowman, D. M. J. S., Perry, G. L. W., Higgins, S. I., Johnson, C. N., Fuhlendorf, S. D.,
& Murphy, B. P. (2016). Pyrodiversity is the coupling of biodiversity and fire
regimes in food webs. Philosophical Transactions of the Royal Society B:
Biological Sciences, 371(1696), 20150169. https://doi.org/10.1098/rstb.2015.0169
Bowman, D. M. J. S., Wilson, B. A., & Hooper, R. J. (1988). Response of eucalyptus
forest and woodland to four fire regimes at Munmarlary, Northern Territory,
Australia. The Journal of Ecology, 76(1), 215. https://doi.org/10.2307/2260465
Bradstock, R. a., Bedward, M., Gill, a. M., & Cohn, J. S. (2005). Which mosaic? A
landscape ecological approach for evaluating interactions between fire regimes,
habitat and animals. Wildlife Research, 32(1), 409–423.
https://doi.org/10.1071/WR02114
Braithwaite, R. W., & Estbergs, J. A. (1985). Fire patterns and woody vegetation trends
in the Alligator Rivers Region of northern Australia. In J. C. Tothill & J. J. Mott
(Eds.), Ecology and management of the world’s Savannas (pp. 359–364).
Canberra, Australia: AUstralian Academy of Science.
Braz, V. S., Kanegae, M. F., & Franco, A. C. (2000). Estabelecimento e
desenvolvimento de Dalbergia miscolobium Benth. em duas fitofisionomias típicas
dos cerrados do Brasil Central. Acta Botanica Brasilica, 14(1), 27–35.
https://doi.org/10.1590/S0102-33062000000100004
Briani, D. C., Palma, A. R., Vieira, E. M., & Henriques, R. P. (2004). Post-fire
succession of small mammals in the Cerrado of central Brazil. Biodiversity and
Conservation, 13(5), 1023–1037.
https://doi.org/10.1023/B:BIOC.0000014467.27138.0b
Brooks, M. L., Antonio, C. M. D., Richardson, D. M., Grace, J. B., Keeley, J. O. N. E.,
Tomaso, J. M. D. I., … Pyke, D. (2004). Effects of invasive alien plants on fire
regimes. BioScience, 54(7), 677–688.
Brown, A. A., & Davis, K. . (1973). Forest fire: control and use. New York: McGraw
Hill Book Co.
Brown, K. (2003). Three challenges for a real people-centred conservation. Global
Ecology and Biogeography, 12, 89–92.
Bryant, R. L., & Bailey, S. (1997). Third world political ecology. London, UK:
Routledge.
Burnham, K. P., & Anderson, D. R. (2002). Model selection and multi-model inference:
a practical information-theoretic approach (2nd ed.). New York: Springer-Verlag.
Retrieved from https://books.google.com.br/books?hl=pt-BR&lr=&id=fT1Iu-h6E-

78
 oC&oi=fnd&pg=PR7&dq=Model+selection+and+multi-
model+inference:+a+practical+information-
theoretic+approach.&ots=tftt5YXzp9&sig=Mn813ptUgRHSfajvlK5kr7Ln-
Ug#v=onepage&q=Model selection and multi-mod
Byram, G. M. (1959). Combustion of forest fuels. In K. P. Davis (Ed.), Forest fire:
control and use (pp. 61–89). New York: Hill Book Co.
Cahoon Jr, D. R., Stocks, B. J., Levine, J. S., Cofer, W. R., & O’Neill, K. P. (1992).
Seasonal distribution of African savanna fires. Nature, 359(6398), 812.
CARE/WWF. (2013). Third controlled burning campaign draws to a close this week.
Retrieved April 6, 2018, from https://primeirasesegundas.net/2013/08/12/third-
controlled-burning-campaign-draws-to-a-close-this-week/
Castro, A. A. J. F., & Martins, F. R. (1999). Cerrados do Brasil e do Nordeste:
caracterização, área de ocupação e considerações sobre a sua fitodiversidade.
Pesquisa Em Foco, 7(9), 147–178.
Castro, E. A. de, & Kauffman, J. B. (1998). Ecosystem structure in the Brazilian
Cerrado: a vegetation gradient of aboveground biomass, root mass and
consumption by fire. Journal of Tropical Biology, 14, 263–283.
Cheney, N., Gould, J., & Catchpole, W. (1993). The Influence of Fuel, Weather and
Fire Shape Variables on Fire-Spread in Grasslands. International Journal of
Wildland Fire, 3(1), 31. https://doi.org/10.1071/WF9930031
Cheney, N. P., & Sullivan, A. (2008). Grassfires: fuel, weather and fire behaviour. (N.
P. Cheney & A. Sullivan, Eds.). Collingwood, Australia: CSIRO Publishing.
Retrieved from https://books.google.com.br/books?hl=pt-
BR&lr=&id=dcvsgwQ3xt4C&oi=fnd&pg=PP1&dq=Grassfires:+fuel,+weather+an
d+fire+behaviour&ots=tU_cH0BhxT&sig=1KY-
q4dDVb0_s6_4Ac7HUtNs7ho#v=onepage&q=Grassfires%3A fuel%2C weather
and fire behaviour&f=false
Christensen, N. L. (2005). Fire in the Parks: A Case Study for Change Management.
The George Wright Forum, 22, 12–31.
Clarke, P. J., Burrows, G. E., Enright, N. J., & Knox, K. J. E. (2013). Tansley review
Resprouting as a key functional trait : how buds , protection and resources drive
persistence after fire, 19–35.
Clarkson, C., Jacobs, Z., Marwick, B., Fullagar, R., Wallis, L., Smith, M., … Pardoe, C.
(2017). Human occupation of northern Australia by 65,000 years ago. Nature,
547(7663), 306–310. https://doi.org/10.1038/nature22968
Clement, C. R., de Cristo-Araújo, M., d’Eeckenbrugge, G. C., Pereira, A. A., &
Picanço-Rodrigues, D. (2010). Origin and domestication of native Amazonian
crops. Diversity (Vol. 2). https://doi.org/10.3390/d2010072
CoA. (2013). Carbon credits (Carbon Farming Initiative) reduction of greenhouse gas
emissions through early dry season savanna burning—1.1. Methodology
determination 2013. Retrieved February 25, 2018, from
http://www.comlaw.gov.au/Series/F2013L01165

79
CoA. (2015). Carbon credits (Carbon Farming Initiative—emissions abatement
through savanna fire management). Methodology determination 2015.
Commonwealth of Australia. Canberra. Retrieved from
http://www.comlaw.gov.au/Details/F2015L00344
Collins, S. L. (1992). Fire fequency and community heterogeneity in tallgrass prairie
vegetation. Ecology, 73(6), 2001–2006.
Conklin, a L. (1998). Colonialism and human rights, a contradiction in terms? The case
of France and West Africa, 1895-1914. The American Historical Review, 103(2),
419–442. https://doi.org/10.1086/ahr/103.2.419
Cook, G. D., P., M. C., Muepu, M., & Liedloff, A. C. (2016). Dead organic matter and
the dynamics of carbon and greenhouse gas emissions in frequently burnt
savannas. International Journal of Wildland Fire, 25, 1252–1263.
Coutinho, L. M. (1981). Aspectos ecológicos do fogo no cerrado: Notas sobre a
ocorrência e datação de carvões vegetais encontrados no interior do solo, em Emas,
Pirassunga, S.P. Rev. Bras. Bot., 4, 115–117.
Coutinho, L. M. (1990). Fire in the Ecology of the Brazilian Cerrado. In J. G.
Goldammer (Ed.), Fire in the Tropical Biota: Ecossystem Processes and Global
Challenges (pp. 82–105). Berlin: Springer-Verlag.
Coutinho, L. M. (2002). O bioma do cerrado. In A. L. Klein (Ed.), Eugen Warming e o
cerrado brasileiro: um século depois. (pp. 77–91). São Paulo, Brasil: Edusp e
Imprensa Oficial do Estado.
Crawley, M. J. (2002). Statistical Computing: an Introduction to Data Analysis using S-
Plus. Chischester, UK: Wiley.
da Silva, M. L. A., Paiva, L. S., Araújo, M. F. V, & da Conceição, G. M. (2017).
Environmental perception of residents of the Parque Nacional da Chapada das
Mesas, in the phytogeographic domain in the forests from Brazilian Cerrado.
Espacios, 38(22), 33. Retrieved from
https://www.scopus.com/inward/record.uri?eid=2-s2.0-
85019044047&partnerID=40&md5=5a6f81b5ffead8fdd9cee214ba9e0353
Dantas, V. de L., & Pausas, J. G. (2013). The lanky and the corky: fire-escape strategies
in savanna woody species. Journal of Ecology, 101(5), 1265–1272.
https://doi.org/10.1111/1365-2745.12118
Dean, W. (1996). A ferro e fogo: a história e a devastação da Mata Atlântica
brasileira. São Paulo: Companhia das Letras.
Dickinson, M. B., & Ryan, K. C. (2010). Introduction: strengthening the foundation of
wildland fire effects prediction for research and management. Fire Ecology, 6(1),
1–24. https://doi.org/10.4996/fireecology0601001
Diegues, A. C. (2000). O mito moderno da natureza intocada. São Paulo:
NUPAUB/USP.
Director-of-National-Parks. (2016). Kakadu National Park Management Plan 2016-
2026. Retrieved from www.environment.gov.au/topics/na onal-parks/parks-
australia/publicatons

80
Dombeck, M. P., Williams, J. E., & Wood, C. A. (2004). Wildfire policy and public
lands: Integrating scientific understanding with social concerns across landscapes.
Conservation Biology, 18(4), 883–889. https://doi.org/10.1111/j.1523-
1739.2004.00491.x
Driscoll, D. a., Lindenmayer, D. B., Bennett, A. F., Bode, M., Bradstock, R. a., Cary, G.
J., … York, A. (2010). Fire management for biodiversity conservation: Key
research questions and our capacity to answer them. Biological Conservation,
143(9), 1928–1939. https://doi.org/10.1016/j.biocon.2010.05.026
Durigan, G., de Freitas Leitao Filho, H., & Ribeiro Rodrigues, R. (1994).
Phytosociology and structure of a frequently burnt cerrado vegetation in SE-Brazil.
Flora, 189(2), 153–160. https://doi.org/10.1016/S0367-2530(17)30582-0
Durigan, G., & Ratter, J. A. (2016). The need for a consistent fire policy for Cerrado
conservation. Journal of Applied Ecology, 53(1), 11–15.
https://doi.org/10.1111/1365-2664.12559
Eckhardt, H. C., van Wilgen, B. W., & Biggs, H. C. (2000). Trends in woody vegetation
cover in the Kruger National Park, South Africa, between 1940 and 1998. African
Journal of Ecology, 38(2), 108–115.
Edwards, A. C., Russell-Smith, J., & Maier, S. W. (2018). A comparison and validation
of satellite-derived fire severity mapping techniques in fire prone north Australian
savannas: Extreme fires and tree stem mortality. Remote Sensing of Environment,
206(May 2017), 287–299. https://doi.org/10.1016/j.rse.2017.12.038
Edwards, A., Kennett, R., Price, O., Russell-Smith, J., Spiers, G., & Woinarski, J.
(2003). Monitoring the impacts of fire regimes on vegetation in northern Australia:
an example from Kakadu National Park. International Journal of Wildland Fire,
12(4), 427–440. https://doi.org/10.1071/WF03031
Edwards, A., Russell-Smith, J., & Meyer, M. (2015). Contemporary fire regime risks to
key ecological assets and processes in north Australian savannas. International
Journal of Wildland Fire, 24(6), 857–870. https://doi.org/10.1071/WF14197
Eloy, L., A. Bilbao, B., Mistry, J., & Schmidt, I. B. (2018). From fire suppression to fire
management: Advances and resistances to changes in fire policy in the savannas of
Brazil and Venezuela. The Geographical Journal, (October 2017), 1–13.
https://doi.org/10.1111/geoj.12245
Enright, N. J., & Lamont, B. B. (1989). Seed banks, fire season, safe sites and seedling
recruitment in five co-occurring banksia species A. Journal of Ecology, 77(4),
1111–1122.
Eriksen, C. (2007). Why do they burn the ‘bush’? Fire, rural livelihoods, and
conservation in Zambia. Geographical Journal, 173(3), 242–256.
FAO. (2007). Fire management: global assessment 2006. Rome.
FAO. (2011). Community-based fire management: A review. Rome, Italy: FAO.
Felfili, J. M. (1995). Growth, recruitment and mortality in the Gama gallery forest in
central Brazil over a six - year period ( 1985 – 1991 ). Journal of Tropical
Ecology, 11(1), 67–83. https://doi.org/10.1017/S0266467400008415

81
Felfili, J. M., Carvalho, F. A., & Haidar, R. F. (2005). Manual para o monitoramento de
parcelas perma- nentes nos biomas cerrado e pantana. Brasília: Universidade de
Brasília, Departamento de Engenharia Florestal.
Fernandes, P. M. (2013). Fire-smart management of forest landscapes in the
Mediterranean basin under global change. Landscape and Urban Planning, 110(1),
175–182. https://doi.org/10.1016/j.landurbplan.2012.10.014
Fidelis, A., Delgado-Cartay, M. D., Blanco, C. C., Müller, S. C., Pillar, V. D., &
Pfadenhauer, J. (2010a). Fire intensity and severity in Brazilian Campos
grasslands. Interciencia, 35, 739–745.
Fidelis, A., Müller, S. C., Pillar, V. D., & Pfadenhauer, J. (2010b). Population biology
and regeneration of forbs and shrubs after fire in Brazilian Campos grasslands.
Plant Ecology, 211, 107–117. https://doi.org/10.1007/s11258-010-9776-z
Finney, M. A. (2001). Design of regular landscape fuel treatment patterns for modifying
fire growth and behavior. Forest Science, 47(2), 219–228.
Foster, D., Swanson, F., Aber, J., Burke, I., Brokaw, N., Tilman, D., & Knapp, A.
(2003). The importance of land-use legacies to ecology and conservation.
BioScience, 53(1), 77–88. https://doi.org/10.1641/0006
Fox, J., & Weisberg, S. (2011). An {R} Companion to Applied Regression, Second
Edition. Thousand Oaks CA: Sage. Retrieved from
http://socserv.socsci.mcmaster.ca/jfox/Books/Companion
França, H. (2010). Os incêndios de 2010 nos parques nacionais do Cerrado. São José
do Rio Preto.
França, H., Ramos-Neto, M. B., & Setzer, A. (2007). O fogo no Parque Nacional das
Emas. Brasília: MMA.
Franklin, D. C., Whitehead, P. J., Pardon, G., Matthews, J., McMahon, P., & McIntyre,
D. (2005). Geographic patterns and correlates of the decline of granivorous birds in
northern Australia. Wildlife Research, 32(5), 399–408.
https://doi.org/10.1071/WR05052
Freeman, J., Edwards, A., & Russell-Smith, J. (2017a). Fire-Driven Decline of Endemic
Allosyncarpia Monsoon Rainforests in Northern Australia. Forests, 8(12), 481.
https://doi.org/10.3390/f8120481
Freeman, J., Kobziar, L., Rose, E. W., & Cropper, W. (2017b). A critique of the
historical-fire-regime concept in conservation. Conservation Biology, 31(5), 976–
985. https://doi.org/10.1111/cobi.12942
Frizzo, T. L. M., Bonizário, C., Borges, M. P., & Vasconcelos, H. L. (2011). Uma
revisão dos efeitos do fogo sobre a fauna de formações savânicas do Brasil.
Oecologia Australis, 15(02), 365–379. https://doi.org/10.4257/oeco.2011.1502.13
Frost, P. G. H., & Robertson, F. (1987). The ecological effects of fire in savannas. In B.
H. Walker (Ed.), Determinants of tropical savannas (pp. 93–140). Miami, FL,
USA: ICSU Press.
Furley, P., Rees, R. M., Ryan, C. M., & Saiz, G. (2008). Savanna burning and the
assessment of long-term fire experiments with particular reference to Zimbabwe.

82
 Progress in Physical Geography, 32, 611–634.
Garde, M., Nadjamerrek, B. L., Kolkkiwarra, M., Kalarriya, J., Djandjomerr, J.,
Birriyabirriya, B., … Biless, P. (2009). The language of fire: seasonality, resources
and landscape burning on the Arnhem Land Plateau. In J. Russell-Smith, P.
Whitehead, & P. Cooke (Eds.), Culture, ecology and economy of fire management
in north Australian savannas: Rekindling the Wurrk tradition (pp. 85–164).
Collingwood, Australia: CSIRO.
Gardner, T. a. (2006). Tree-grass coexistence in the Brazilian cerrado: demographic
consequences of environmental instability. Journal of Biogeography, 33(3), 448–
463. https://doi.org/10.1111/j.1365-2699.2005.01420.x
Giglio, L., Randerson, J. T., & van der Werf, G. R. (2013). Analysis of daily, monthly,
and annual burned area using the fourth-generation global fire emissions database
(GFED4). Journal of Geophysical Research: Biogeosciences, 118(1), 317–328.
https://doi.org/10.1002/jgrg.20042
Gill, A. M., Cheney, N. P., Walker, J., & Tunstall, B. R. (1986). Bark losses from two
eucalypt species following fires of different intensities. Australian Forest
Research. CSIRO Australia, 16(1), 1–7.
Gill, A. M., Hoare, J. R. L., & Cheney, N. P. (1990). Fires and Their Effects in the Wet-
Dry Tropics of Australia. In J. G. Goldammer (Ed.), Fire in the Tropical Biota.
Ecological Studies (Analysis and Synthesis) (pp. 159–178). Berlin, Heidelberg:
Springer.
Gill, A. M., Moore, P. H. R., & Williams, R. J. (1996). Fire weather in the wet-dry
tropics of the World Heritage Kakadu National Park, Australia. Austral Ecology,
21(3), 302–308. https://doi.org/10.1111/j.1442-9993.1996.tb00612.x
Goldammer, J. G., & de Ronde, C. (2004). Wildland fire management handbook for
sub-Sahara Africa. (J. G. Goldammer & C. de Ronde, Eds.). Freiburg:
Oneworldbooks.
Goldammer, J. G., Frost, P., Jurvelius, M., Kammigna, E., & Kruger, T. (2004).
Community participation in integrated forest fire management: some experiences
from Africa. (J. G. Goldammer & C. de Ronde, Eds.). Frieburg: Global Fire
Monitoring Center.
Gomes, L., Miranda, H. S., & Bustamante, M. M. da C. (2018). How can we advance
the knowledge on the behavior and effects of fire in the Cerrado biome? Forest
Ecology and Management, 417(February), 281–290.
https://doi.org/10.1016/j.foreco.2018.02.032
Gómez-Baggethun, E., Corbera, E., & Reyes-García, V. (2013). Traditional ecological
knowledge and global environmental change: Research findings and policy
implications. Ecology and Society, 18(4), art72. https://doi.org/10.5751/ES-06288-
180472
Gotelli, N. J., & Ellison, A. M. (2004). A primer of ecological statistics. Sunderland,
United States: Sinauer Associates Publishers.
Govender, N., Trollope, W. S. W., & Van Wilgen, B. W. (2006). The effect of fire
season, fire frequency, rainfall and management on fire intensity in savanna

83
 vegetation in South Africa. Journal of Applied Ecology, 43(4), 748–758.
https://doi.org/10.1111/j.1365-2664.2006.01184.x
Grace, J., Jose, J. S., Meir, P., Miranda, H. S., & Montes, R. a. (2006). Productivity and
carbon fluxes of tropical savannas. Journal of Biogeography, 33(3), 387–400.
https://doi.org/10.1111/j.1365-2699.2005.01448.x
Grady, J. M., & Hoffmann, W. a. (2012). Caught in a fire trap: Recurring fire creates
stable size equilibria in woody resprouters. Ecology, 93(9), 2052–2060.
https://doi.org/10.1890/12-0354.1
Griffin, G. F., & Friedel, M. H. (1984). Effects of fire in Central Australia rangelands:
Fire and fuel characteristics and change in herbage and nutrients. Australiam
Journal of Ecology, 9, 381–393.
Groffman, P. M., Baron, J. S., Blett, T., Gold, A. J., Goodman, I., Gunderson, L. H., …
Wiens, J. (2006). Ecological thresholds: The key to successful environmental
management or an important concept with no practical application? Ecosystems,
9(1), 1–13. https://doi.org/10.1007/s10021-003-0142-z
Guedes, D. M. (1993). Resistência das árvores do Cerrado ao fogo: papel da casca
como isolante térmico. Universidade de Brasília.
Guidon, N., & Delibrias, G. (1986). Carbon-14 dates point to man in the Americas
32,000 years ago. Nature, 321(6072), 769–771. https://doi.org/10.1038/321769a0
Haidinger, T. L., & Keeley, J. E. (2017). Role of high fire frequency in destruction of
mixed chaparral. California Botanical Society, 40(3), 141–147.
Hall, M. (1984). Man’s historical and traditional use of fire in southern Africa. In P.
Booysen & N. Tainton (Eds.), Ecological effects of re in South African ecosystems
(pp. 40–52). Berlin, Germany: Springer-Verlag.
Haney, A., & Power, R. L. (1996). Adaptive management for sound ecosystem
management. Environmental Management, 20(6), 879–886.
https://doi.org/10.1007/BF01205968
Hantson, S., Scheffer, M., Pueyo, S., Xu, C., Lasslop, G., van Nes, E. H., …
Mendelsohn, J. (2017). Rare, Intense, Big fires dominate the global tropics under
drier conditions. Scientific Reports, 7(1), 14374. https://doi.org/10.1038/s41598-
017-14654-9
Harlan, J. R. (1976). Origins of African plant domestication. (J. R. Harlan, Ed.). Walter
de Gruyter.
Higgins, S. I., Bond, W. J., February, E. C., Bronn, A., Douglas, I., Enslin, B., …
Elizabeth, P. O. B. P. (2007). Effects of four decades of fire manipulation on
woody vegetation structure in savanna. Ecological Society of America, 88(5),
1119–1125.
Higgins, S. I., Bond, W. J., & Trollope, W. S. W. (2000). Fire, resprouting and
variability: a recipe for grass - tree coexistence in savanna. Journal of Ecology, 88,
213–229.
Hodgkinson, K. C. (1991). Shrub recruitment response to intensity and season of fire in
a semi-arid woodland. Journal of Applied Ecology, 28(1), 60–70.

84
Hoffmann, W. A. (1998). Post-burn reproduction of woody plants in a neotropical
savanna: The relative importance of sexual and vegetative reproduction. Journal of
Applied Ecology, 35(3), 422–433. https://doi.org/10.1046/j.1365-
2664.1998.00321.x
Hoffmann, W. A. (1999). Fire and population dynamics of woody plants in a
neotropical savanna: matrix model projections. Ecology, 80(4), 1354–1369.
Hoffmann, W. a., Adasme, R., Haridasan, M., De Carvalho, M. T., Geiger, E. L.,
Pereira, M. a B., … Franco, A. C. (2009). Tree topkill, not mortality, governs the
dynamics of savanna-forest boundaries under frequent fire in central Brazil.
Ecology, 90(5), 1326–1337. https://doi.org/10.1890/08-0741.1
Hoffmann, W. a., Jaconis, S. Y., Mckinley, K. L., Geiger, E. L., Gotsch, S. G., &
Franco, A. C. (2012). Fuels or microclimate? Understanding the drivers of fire
feedbacks at savanna-forest boundaries. Austral Ecology, 37(6), 634–643.
https://doi.org/10.1111/j.1442-9993.2011.02324.x
Hoffmann, W. A., & Moreira, A. G. (2002). The role of fire in population dynamics of
woody plants. In P. S. Oliveira & R. J. Marquis (Eds.), The Cerrados of Brazil.
Ecology and natural history of a neotropical savanna (pp. 159–177). New York,
NY: Columbia University Press.
Hoffmann, W., & Solbrig, O. T. (2003). The role of topkill in the differential response
of savanna woody species to fire. Forest Ecology and Management, 180(1–3),
273–286. https://doi.org/10.1016/S0378-1127(02)00566-2
Honda, E. A., & Durigan, G. (2016). Woody encroachment and its consequences on
hydrological processes in the savannah. Philosophical Transactions of the Royal
Society of London. Series B, Biological Sciences, 371(1703), 115–133.
https://doi.org/10.1098/rstb.2015.0313
Huffman, M. R. (2013). The many elements of traditional fire knowledge: Synthesis,
classification, and aids to cross-cultural problem solving in fire-dependent systems
around the world. Ecology and Society, 18(4), 3. https://doi.org/10.5751/ES-
05843-180403
ICMBio. (2018). Parque Nacional da Chapada das Mesas. Retrieved May 15, 2018,
from http://www.icmbio.gov.br/portal/unidadesdeconservacao/biomas-
brasileiros/cerrado/unidades-de-conservacao-cerrado/2079-parna-da-chapada-das-
mesas
INMET. (2017). Banco de dados meteorológicos para ensino e pesquisa. Retrieved
April 11, 2017, from http://www.inmet.gov.br/portal/index.php?r=bdmep/bdmep
Iversen, J. (1956). Forest clearence in the stone age. Scientific American, 194(3), 36–41.
Johnson, E. A., Miyanishi, K., & Bridge, R. J. (2001). Wildfire regime in the boreal
forest and the idea of suppression and fuel buildup. Conservation Biology, 15(6),
1554–1557. Retrieved from http://doi.wiley.com/10.1046/j.1523-
1739.2002.16502.x
Jones, B., & Murphree, M. W. (2001). The evolution of policy on community
conservation in Namibia and Zimbabwe. In D. Hulme & M. W. Murphree (Eds.),
African wildlife and livelihoods: The promise and performance of community

85
 conservation (pp. 38–58). Oxford, UK: James Currey.
Jones, B. T. B. (2004). CBNRM, poverty reduction and sustainable livelihoods:
Developing criteria for evaluating the contribution of CBNRM to poverty
reduction and alleviation in southern Africa. Harare (Zimbabwe) and Cape Town
(South Africa). Retrieved from
http://www.plaas.org.za/sites/default/files/publications-pdf/CBNRM 07.pdf
Jones, R. (1969). Fire-stick farming. Australian Natural History, 16(7), 224–228.
Just, M. G., Hohmann, M. G., & Hoffmann, W. A. (2016). Where fire stops: vegetation
structure and microclimate influence fire spread along an ecotonal gradient. Plant
Ecology, 217(6), 631–644. https://doi.org/10.1007/s11258-015-0545-x
Kassambara, A. (2017). ggpubr: “ggplot2” Based Publication Ready Plots.
Kauffman, J. B., Cummings, D. L., & Ward, D. E. (1994). Relationships of fire,
biomass and nutrient dynamics along a vegetation gradient in the Brazilian
Cerrado. The Journal of Ecology, 82(3), 519. https://doi.org/10.2307/2261261
Keeley, J. . E. (1987). Role of fire in seed germination of woody taxa in California
chaparral. Ecology, 68(2), 434–443.
Keeley, J. E. (2009). Fire intensity, fire severity and burn severity: a brief review and
suggested usage. International Journal of Wildland Fire, 18, 116–126.
Keeley, J. E., & Fotheringham, C. J. (2000). Role of fire in regeneration from seed.
Seeds: The Ecology of Regeneration in Plant Communities, 2, 311–330.
https://doi.org/10.1079/9780851994321.0311
Keeley, J. E., Pausas, J. G., Rundel, P. W., Bond, W. J., & Bradstock, R. A. (2011). Fire
as an evolutionary pressure shaping plant traits. Trends in Plant Science, 16(8),
406–411. https://doi.org/10.1016/j.tplants.2011.04.002
Keeley, J. E., & Zedler, P. H. (2009). Large, high-intensity fire events in southern
California shrublands: debunking the fine- grain age patch model. Ecological
Applications, 19(1), 69–94.
Kepe, T. (2005). Grasslands ablaze: vegetation burning by rural people in Pondoland,
South Africa. South African Geographic Journal, 87(1), 10–17.
Kepe, T., & Scoones, I. (1999). Creating grasslands: social institutions and
environmental change in Mkambati Area, South Africa. Human Ecology, 27(1),
29–53.
Klink, C. A., & Machado, R. B. (2005). Conservation of the Brazilian Cerrado.
Conservation Biology, 19(3), 707–713.
Knox, K. J. E., & Clarke, P. J. (2006). Fire season and intensity affect shrub recruitment
in temperate sclerophyllous woodlands. Oecologia, 149(4), 730–739.
https://doi.org/10.1007/s00442-006-0480-6
Korontzi, S. (2005). Seasonal patterns in biomass burning emissions from southern
African vegetation fires for the year 2000. Global Change Biology, 11, 1680–1700.
Krawchuk, M. A., Moritz, M. A., Parisien, M.-A., Van Dorn, J., & Hayhoe, K. (2009).
Global Pyrogeography: the Current and Future Distribution of Wildfire. PLoS

86
 ONE, 4(4), e5102. https://doi.org/10.1371/journal.pone.0005102
Kull, C. A. (2002). Madagascar aflame: landscape burning as peasant protest,
resistance, or a resource management tool? Political Geography, 21, 927–953.
Laris, P. (2002). Burning the seasonal mosaic : Preventative burning strategies in the
wooded savanna of Southern Mali. Human Ecology, 30(2), 155–186.
Laris, P., Koné, M., Dadashi, S., & Dembele, F. (2016). The early/late fire dichotomy:
Time for a reassessment of Aubreville’s savanna fire experiments. Progress in
Physical Geography, 41(1), 68–94. https://doi.org/10.1177/0309133316665570
Laris, P., & Wardell, D. A. (2006). Good, bad or ‘necessary evil’? Reinterpreting the
colonial burning experiments in the savanna landscapes of West Africa.
Geographical Journal, 172, 271–290.
Lawes, M. J., Adie, H., Russell-Smith, J., Murphy, B., & Midgley, J. J. (2011). How do
small savanna trees avoid stem mortality by fire? The roles of stem diameter,
height and bark thickness. Ecosphere, 2(4), art42. https://doi.org/10.1890/ES10-
00204.1
Lawes, M. J., Murphy, B. P., Fisher, A., Woinarski, J. C. Z., Edwards, A. C., & Russell-
Smith, J. (2015). Small mammals decline with increasing fire extent in northern
Australia: evidence from long-term monitoring in Kakadu National Park.
International Journal of Wildland Fire, 24, 712–722.
https://doi.org/10.1071/WF14163
le Stradic, S., Silveira, F. A. O., Buisson, E., Cazelles, K., Carvalho, V., & Fernandes,
G. W. (2015). Diversity of germination strategies and seed dormancy in
herbaceous species of campo rupestre grasslands. Austral Ecology, 40(5), 537–546.
https://doi.org/10.1111/aec.12221
Lee, B. S., Alexander, M. E., Hawkes, B. C., Lynham, T. J., Stocks, B. J., & Englefield,
P. (2002). Information systems in support of wildland fire management decision
making in Canada. Computers and Electronics in Agriculture, 37, 185–198.
Lehmann, C. E. R., Anderson, T. M., Sankaran, M., Higgins, S. I., Archibald, S.,
Hoffmann, W. A., … Bond, W. J. (2014). Savanna vegetation-fire-climate
relationships differ among continents. Science, 343(6170), 548–552.
https://doi.org/10.1126/science.1247355
Lehmann, C. E. R., Prior, L. D., & Bowman, D. M. J. S. (2009). Fire controls
population structure in four dominant tree species in a tropical savanna. Oecologia,
161(3), 505–515. https://doi.org/10.1007/s00442-009-1395-9
Lehsten, V., Tansey, K., Baltzer, H., Thonicke, K., Spessa, A., Weber, U., … Arneth,
A. (2009). Estimating carbon emissions from African wildfires. Biogeosciences, 6,
349–360.
Lenth, R. V. (2016). Using lsmeans. Journal of Statistical Software, 69(1), 1–33.
Retrieved from https://cran.r-project.org/web/packages/lsmeans/vignettes/using-
lsmeans.pdf
Levitus, R. (2009). Change and catastrophe: adaptation, re-adaptation and fire in the
Alligator Rivers region. In J. Russell-Smith, P. Whitehead, & Peter Cooke (Eds.),

87
 Culture, ecology and economy of fire management in north Australian savannas:
Rekindling the Wurrk tradition (pp. 41–67). Collingwood, Australia: CSIRO
Publishing.
Lima, S. E. de, Lima, H. S., & Ratter, J. A. (2009). Mudanças pós-fogo na estrutura e
composição da vegetação lenhosa, em um cerrado mesotrófico, no período de
cinco anos (1997-2002) em Nova Xavantina-MT. Cerne, 15(4), 468–480.
Luke, R. H., & McArthur, A. G. (1978). Bushfires in Australia. Camberra, Austrália:
Australian Government Publising Service.
Mamdani, M. (1996). Citizen and subject: contemporary Africa and the legacy of late
colonialism. Princeton, New Jersey: Princeton University Press.
Maravalhas, J., & Vasconcelos, H. L. (2014). Revisiting the pyrodiversity – biodiversity
hypothesis : long-term fire regimes and the structure of ant communities in a
Neotropical savanna hotspot. Journal of Applied Ecology, 51(6), 1661–1668.
https://doi.org/10.1111/1365-2664
Marcoux, H. M., Daniels, L. D., Gergel, S. E., Da Silva, E., Gedalof, Z., & Hessburg, P.
F. (2015). Differentiating mixed- and high-severity fire regimes in mixed-conifer
forests of the Canadian Cordillera. Forest Ecology and Management, 341, 45–58.
https://doi.org/10.1016/j.foreco.2014.12.027
Marshall, F., & Hildebrand, E. (2002). Cattle Before Crops: The Beginnings of Food
Production in Africa. Journal of World Prehistory, 16(2), 99–143.
Martin, B. D. (1995). Postfire reproduction of Croton californicus (Euphorbiaceae) and
associated perennials in coastal sage scrub of southern California. Crossosoma,
21(2), 41–56.
Martins, M. B., & Oliveira, T. G. de. (2011). Amazônia Maranhense: diversidade e
conservação. Museu Paraense Emílio Goeldi. Belém: MPEG.
McGregor, S., Lawson, V., Christophersen, P., Kennett, R., Boyden, J., Bayliss, P., …
Andersen, A. N. (2010). Indigenous wetland burning: Conserving natural and
cultural resources in Australia’s world heritage-listed Kakadu National Park.
Human Ecology, 38, 721–729. https://doi.org/10.1007/s10745-010-9362-y
McGuffog, T., Dyer, R., & Williams, B. (2001). Burning operations. In R. J. W. R.
Dyer, P. Jacklyn, I. Partridge, J. Russell-Smith (Ed.), Savanna Burning:
Understanding and Using Fire inNorthern Australia (pp. 81–101). Darwin,
Australia: TropicalSavannas Co-operative Research Centre.
Medeiros, M. B. De, & Miranda, S. (2005). Mortalidade pós-fogo em espécies lenhosas
de campo sujo submetido a três queimadas prescritas anuais, 19(3), 493–500.
Medeiros, M. B., Walter, B. M. T., & Silva, G. P. (2008). Phytosociology of cerrado
stricto sensu in Carolina County, MA, Brazil. Cerne, Lavras, 14(4), 285–294.
Retrieved from http://www.scopus.com/inward/record.url?eid=2-s2.0-
70349649490&partnerID=40&md5=5e3f4e624535987ff34b904842368c35
Medina, E., & Silva, J. F. (1990). Savannas of Northern South America: A Steady State
Regulated by Water-Fire Interactions on a Background of Low Nutrient
Availability. Journal of Biogeography, 17(4/5), 403.

88
 https://doi.org/10.2307/2845370
Meyn, A., White, P. S., Buhk, C., & Jentsch, A. (2007). Environmental drivers of large,
infrequent wildfires: The emerging conceptual model. Progress in Physical
Geography, 31(3), 287–312. https://doi.org/10.1177/0309133307079365
Miranda, A. C., Miranda, H. S., Dias, I. de F. O., & Dias, B. F. de S. (1993). Soil and
air temperatures during prescribed cerrado fires in Central Brazil. Journal of
Tropical Ecology, 9(3), 313–320.
Miranda, H. S. (2010). Efeitos do regime de fogo sobre a estrutura de comunidades de
Cerrado: Resultados do projeto fogo. Brasília, DF: MMA, IBAMA.
Miranda, H. S., Bustamante, M. M. C., & Miranda, A. C. (2002). The Fire Factor. In P.
S. Oliveira & R. J. Marquis (Eds.), The Cerrados of Brazil: Ecology and Natural
History of a Neotropical Savanna (pp. 51–68). New York: Columbia University
Press.
Miranda, H. S., Neto, W. N., & Neves, B. M. C. (2010). Caracterização das queimadas
de Cerrado. In H. S. Miranda (Ed.), Efeitos do regime do fogo sobre a estrutura de
comunidades de cerrado: Resultados do Projeto Fogo (pp. 23–34). Brasília, DF:
MMA, IBAMA.
Miranda, H. S., Rocha-Silva, E. P., & Miranda, A. C. (1996). Comportamento do fogo
em queimadas de campo sujo. In H. S. Miranda, C. H. Saito, & B. F. S. Dias
(Eds.), Impactos de queimadas em áreas de cerrado e restinga (pp. 1–10). Brasília,
DF: ECL/UnB.
Miranda, H. S., Sato, M. N., Neto, W. N., & Aires, F. S. (2009). Fires in the Cerrado,
the Brazilian savanna. In M. A. Cochrane (Ed.), Tropical fire ecology: Climate
Change, Land Use and Ecosystem Dynamics. (pp. 427–450). Heidelberg,
Germany: Springer-Praxis.
Miranda, S. do C., Silva-Júnior, M. C. da, & Salles, L. A. (2007). A comunidade
lenhosa de cerrado rupestre na Serra Dourada, Goiás. Heringeriana, 1, 43–53.
Mistry, J. (2000). Savannas. Progress in Physical Geography, 24(2), 273–279.
Mistry, J. (2002). Savannas and development. In V. Desai & R. B. Potter (Eds.), The
companion to development studies (3rd ed., pp. 305–310). New York: Routledge.
Mistry, J., Berardi, A., Andrade, V., Krahô, T., Krahô, P., & Leonardos, O. (2005).
Indigenous fire management in the cerrado of Brazil: The case of the Krahô of
Tocantíns. Human Ecology, 33(3), 365–386. https://doi.org/10.1007/s10745-005-
4143-8
Mistry, J., Bilbao, B. A., & Berardi, A. (2016). Community owned solutions for fire
management in tropical ecosystems: case studies from Indigenous communities of
South America. Philosophical Transactions of the Royal Society B: Biological
Sciences, 371(1696), 20150174. https://doi.org/10.1098/rstb.2015.0174
Mistry, J., & Bizerril, M. (2011). Por que é importante entender as inter-relações entre
pessoas, fogo e áreas protegidas? Biodiversidade Brasileira, 1(2), 40–49.
Mistry, J., Schmidt, I. B., Eloy, L., & Bilbao, B. (2018). New perspectives in fire
management in South American savannas: The importance of intercultural

89
 governance. Ambio, 1–8. https://doi.org/10.1007/s13280-018-1054-7
MMA. (2006). Sistema nacional de unidades de conservação da natureza (6th ed.).
Brasília: Ministério do Meio Ambiente.
MMA. (2015). Projeto Cerrado-Jalapão. Retrieved October 10, 2017, from
http://cerradojalapao.mma.gov.br/
Moritz, M. A. (2003). Spatiotemporal analysis of controls on shrubland fire regimes:
age dependency and fire hazard. Ecology, 84(2), 351–361.
Moritz, M. A., Batllori, E., Bradstock, R. A., Gill, A. M., Handmer, J., Hessburg, P. F.,
… Syphard, A. D. (2014). Learning to coexist with wildfire.
https://doi.org/10.1038/nature13946
Moritz, M. A., Parisien, M.-A., Batllori, E., Krawchuk, M. A., Van Dorn, J., Ganz, D.
J., & Hayhoe, K. (2012). Climate change and disruptions to global fire activity.
Ecosphere, 3(6), art49. https://doi.org/10.1890/ES11-00345.1
Morrison, J. H., & Cooke, P. M. (2003). Caring for country : Indigenous people
managing country using fire with particular emphasis on Northern Australia. In
Proceedings of National Landcare Conference (p. 10). Darwin, Australia.
Mouillot, F., & Field, C. B. (2005). Fire history and the global carbon budget: A 1° × 1°
fire history reconstruction for the 20th century. Global Change Biology, 11(3),
398–420. https://doi.org/10.1111/j.1365-2486.2005.00920.x
Moura, L. C. (2013). Fogo no Parque Nacional da Serra da Canastra/MG: Abordagem
dos aspectos fisiográficos e humanos na concepção de uma proposta de manejo
integrado de fogo. Universidade Estadual Paulista.
Murphy, B. P., Russell-Smith, J., & Prior, L. D. (2010). Frequent fires reduce tree
growth in northern Australian savannas: Implications for tree demography and
carbon sequestration. Global Change Biology, 16(1), 331–343.
https://doi.org/10.1111/j.1365-2486.2009.01933.x
Murphy, B., Russell-Smith, J., Edwards, A., Meyer, M., & Meyer, C. M. (2015).
Carbon accounting and savanna fire management. Melbourne: CSIRO Publishing.
Myers, R. L. (2006). Living with fire: sustaining ecosystems and livelihoods through
integrated fire management. Tallahassee, USA: The Nature Conservancy, Global
Fire Initiative.
NACSO. (2012). Namibian Community-Based Natural Resource Management Support
Organisations. Retrieved November 20, 2012, from http://www.nacso.org.na/index
Neto, W. N., Andrade, S. M. A., & Miranda, H. S. (1998). The dynamics of herbaceous
layer following prescribed burning: a four year study in the Brazilian savanna. In
International Conference on Forest Fire Research, 3 (p. 1785–1792.). Coimbra:
Adai.
Neumann, R. (1998). Imposing wilderness: Struggles over livelihood and nature
preservation in Africa. Berkeley, Los Angeles and London: University of
California Press.
Nevle, R. J., & Bird, D. K. (2008). Effects of syn-pandemic fi re reduction and
reforestation in the tropical Americas on atmospheric CO 2 during European

90
 conquest. Palaeogeography, Palaeoclimatology, Palaeocology, 264, 25–38.
https://doi.org/10.1016/j.palaeo.2008.03.008
O’Connor, T. G., Puttick, J. R., & Hoffman, M. T. (2014). Bush encroachment in
southern Africa: Changes and causes. African Journal of Range and Forage
Science, 31(2), 67–88. https://doi.org/10.2989/10220119.2014.939996
Oliveira, P. E., & Silva, J. C. S. (1993). Reproductive biology of two species of
Kielmeyera (Guttiferae) in the cerrados of Central Brazil. Journal of Tropical
Ecology, 9(1), 67–79. https://doi.org/https://doi.org/10.1017/S0266467400006970
Oliveira, S. L. J., Campagnolo, M. L., Price, O. F., Edwards, A. C., & Russell-Smith, J.
(2015). Ecological implications of fine-scale fire patchiness and severity in tropical
savannas of Northern Australia. Fire Ecology, 11(1), 10–31.
Parr, C. L., Woinarski, J. C. Z., & Pienaar, D. J. (2009). Cornerstones of biodiversity
conservation? Comparing the management effectiveness of Kruger and Kakadu
National Parks, two key savanna reserves. Biodiversity and Conservation, 18(13),
3643–3662. https://doi.org/10.1007/s10531-009-9669-4
Paul Murrell, Z. W. (2018). gridGraphics: Redraw Base Graphics Using “grid”
Graphics. Retrieved from https://cran.r-
project.org/web/packages/gridGraphics/index.html
Pausas, J. G., & Moreira, B. (2012). Flammability as a biological concept. New
Phytologist, 194, 610–613.
Pearce, F. (2000). Inventing Africa. Which is more authentic: a game-rich wilderness or
cattle pastur. New Scientist, 167(2251), 30.
Pebesma, E. J. (2004). Multivariable geostatistics in S: the gstat package. Computers &
Geosciences, 30(7), 683–691.
Pedroso Júnior, N. N., Murrieta, R. S. S., & Adams, C. (2008). A agricultura de corte e
queima: um sistema em transformação. Boletim Do Museu Paraense Emílio
Goeldi. Ciências Humanas, 3(2), 153–174.
Penman, T. D., Christie, F. J., Andersen, A. N., Bradstock, R. A., Cary, G. J.,
Henderson, M. K., … York, A. (2011). Prescribed burning: How can it work to
conserve the things we value? International Journal of Wildland Fire, 20(6), 721–
733. https://doi.org/10.1071/WF09131
Pereira Júnior, A. C., Oliveira, S. L. J., Pereira, J. M. C., & Turkman, M. A. A. (2014).
Modelling fire frequency in a Cerrado savanna protected area. Plos One, 9(7),
e102380. https://doi.org/10.1371/journal.pone.0102380
Pereira, P., Mierauskas, P., Úbeda, X., Mataix-solera, J., & Cerda, A. (2012). Fire in
protected areas: the effect of protection and importance of fire management.
Environmental Research, Engeneering and Management, 59(1), 52–62.
Perry, D. a., Hessburg, P. F., Skinner, C. N., Spies, T. a., Stephens, S. L., Taylor, A. H.,
… Riegel, G. (2011). The ecology of mixed severity fire regimes in Washington,
Oregon, and Northern California. Forest Ecology and Management, 262(5), 703–
717. https://doi.org/10.1016/j.foreco.2011.05.004
Perry, J. J., Sinclair, M., Wikmunea, H., Wolmby, S., Martin, D., & Martin, B. (2018).

91
 The divergence of traditional Aboriginal and contemporary fire management
practices on Wik traditional lands, Cape York Peninsula, Northern Australia.
Ecological Management & Restoration, 19(1), 24–31.
https://doi.org/10.1111/emr.12301
Peterson, D. W., & Reich, P. B. (2001). Prescribed fire in oak savanna: fire frequency
effects on stand structure and dynamics. Ecological Applications, 11(3), 914–927.
Petty, A. M., Isendahl, C., Brenkert-Smith, H., Goldstein, D. J., Rhemtulla, J. M.,
Rahman, S. A., & Kumasi, T. C. (2015). Applying historical ecology to natural
resource management institutions: Lessons from two case studies of landscape fire
management. Global Environmental Change, 31, 1–10.
https://doi.org/10.1016/j.gloenvcha.2014.11.004
Phillipson, D. W. (2005). African archaeology. (D. W. Phillipson, Ed.) (3rd ed.).
Cambridge: Cambridge University Press.
Pierce, S. M., & Cowling, R. M. (1991). Dynamics of soil-stored seed sanks of six
shrubs in fire-prone dune fynbos. Journal of Ecology, 79(3), 731–747.
Pinto, A. S., & Bustamante, M. M. C. (2010). Emissões de gases-traço (CO2, CO, NO e
N2O) do solo para a atmosfera em fitofisionomias de Cerrado sob diferentes
regimes de queima. In H. S. Miranda (Ed.), Efeitos do regime do fogo sobre a
estrutura de comunidades de cerrado: Projeto Fogo (pp. 47–57). Brasília, DF:
MMA, IBAMA.
Piperno, D. R. (2011). The Origins of Plant Cultivation and Domestication in the New
World Tropics. Current Anthropology, 52(S4), S453–S470.
https://doi.org/10.1086/659998
Pivello, V. R. (2006). Fire management for biological conservation in the Brazilian
Cerrado. In J. Mistry & A. Berardi (Eds.), Savannas and dry forests: linking people
with nature (pp. 129–154). Aldershot: Ashgate.
Pivello, V. R. (2011). The use of fire in the Cerrado and Amazonian rainforests of
Brazil: Past and present. Fire Ecology, 7(1), 24–39.
https://doi.org/10.4996/fireecology.0701024
Pivello, V. R., & Coutinho, L. M. (1996). A qualitative successional model to assist in
the management of Brazilian cerrados, 87, 127–138.
Pivello, V. R., Oliveras, I., Miranda, H. S., Haridasan, M., Sato, M. N., & Meirelles, S.
T. (2010). Effect of fires on soil nutrient availability in an open savanna in Central
Brazil. Plant and Soil, 337(1–2), 111–123. https://doi.org/10.1007/s11104-010-
0508-x
Preece, N. (2007). Traditional and ecological fires and effects of bushfire laws in north
Australian savannas. International Journal of Wildland Fire, 16(4), 378–389.
https://doi.org/10.1071/WF05079
Price, O. F., Borah, R., & Maier, S. W. (2014). Role of weather and fuel in stopping fire
spread in tropical savannas. Austral Ecology, 39(2), 135–144.
https://doi.org/10.1111/aec.12021
Price, O. F., Russell-Smith, J., & Watt, F. (2012). The influence of prescribed fire on

92
 the extent of wildfire in savanna landscapes of western Arnhem. International
Journal of Wildland Fire, 21(3), 297–305.
Pricope, N. G., & Binford, M. W. (2012). A spatio-temporal analysis of fi re recurrence
and extent for semi-arid savanna ecosystems in southern Africa using moderate-
resolution satellite imagery. Journal of Environmental Management, 100, 72–85.
https://doi.org/10.1016/j.jenvman.2012.01.024
Prior, L. D., Brook, B. W., Williams, R. J., Werner, P. A., Bradshaw, C. J. A., &
Bowman, D. M. J. S. (2006). Environmental and allometric drivers of tree growth
rates in a north Australian savanna. Forest Ecology and Management, 234(1–3),
164–180. https://doi.org/10.1016/j.foreco.2006.06.034
Pyne, S. J. (1993). Keeper of the flame: a survey of anthropogenic fire. Fire in the
environment: its ecological, climatic, and atmospheric chemical importance. New
York: John Wiley and Sons.
Pyne, S. J. (1997a). Vestal fire: an environmental history, told through fire, of Europe
and Europe’s encounter with the World. Seattle and London: University of
Washington Press.
Pyne, S. J. (1997b). World fire: the culture of fire on Earth. Seattle and London:
University of Washington Press.
Pyne, S. J. (2016). Fire in the mind: changing understandings of fire in Western
civilization. Philosophical Transactions of the Royal Society B: Biological
Sciences. https://doi.org/10.1098/rstb.2015.0166
R-Core-Team. (2017). R: A language and environment for statistical computing.
Vienna, Austria: R Foundation for Statistical Computing. Retrieved from
http://www.r-project.org/
Ramos, A. E. (1990). Efeitos da queima sobre a vegetação lenhosa do Cerrado.
Ribeiro, M. N., Sanchez, M., Pedroni, F., & Karine da Silva Peixoto. (2012). Fogo e
dinâmica da comunidade lenhosa em cerrado sentido restrito, Barra do Garças,
Mato Grosso. Acta Botanica Brasilica, 26(1), 203–217.
Rissi, M. N., Baeza, M. J., Gorgone-Barbosa, E., Zupo, T., & Fidelis, A. (2017). Does
season affect fire behaviour in the Cerrado? International Journal of Wildland
Fire, 26(5), 427–433. https://doi.org/10.1071/WF14210
Ritchie, D. (2009). Things fall apart: the end of an era of systematic Indigenous fire
management. In J. Russell-Smith, P. Whitehead, & P. Cooke (Eds.), Culture,
ecology and economy of fire management in north Australian savannas:
Rekindling the Wurrk tradition (pp. 23–40). Collingwood, Australia: CSIRO.
Roberts, R., Jones, R., & Smith, M. A. (1993). Optical dating at Deaf Adder Gorge,
Northern Territory, indicates human occupation between 53,000 and 60,000 years
ago. Australian Archaeology, 31, 58–59.
Roberts, T. (2005). Frontier justice: A history of the Gulf country to 1900. Brisbane:
University of Queensland Press.
Rocha-Silva, E. P. (1999). Efeito do regime de queima na taxa de mortalidade e na
estrutura da vegetação lenhosa de campo sujo de Cerrado. Universidade de

93
 Brasília.
Roche, S., Dixon, K. W., & Pate, J. S. (1998). For everything a season: Smoke-induced
seed germination and seedling recruitment in a Western Australian Banksia
woodland. Austral Ecology, 23(2), 111–120. https://doi.org/10.1111/j.1442-
9993.1998.tb00709.x
Roe, D., Nelson, F., & Sandbrook, C. (2009). Community management of natural
resources in Africa: Impacts, experiences and future directions. (D. Roe, F.
Nelson, & C. Sandbrook, Eds.), International Institute for Environment and
Development (Natural Re). London, UK: IIED.
Roos, C. I., Bowman, D. M. J. S., Balch, J. K., Artaxo, P., Bond, W. J. ., Cochrane6,
M., … Swetnam, T. W. (2014). Pyrogeography, historical ecology, and the human
dimensions of fire regimes. Journal of Biogeography, 41, 833–836.
Roos, C. I., Scott, A. C., Belcher, C. M., Chaloner, W. G., Aylen, J., Bird, R. B., …
Steelman, T. (2016). Living on a flammable planet: interdisciplinary, cross-scalar
and varied cultural lessons, prospects and challenges. Philosophical Transactions
of the Royal Society B: Biological Sciences, 371(1696), 20150469.
https://doi.org/10.1098/rstb.2015.0469
Roques, K. G., O’Connor, T. G., & Watkinson, A. R. (2001). Dynamics of shrub
encroachment in an African savanna: relative influences of fire, herbivory, rainfall
and density dependence. Journal of Applied Ecology, 38(2), 268–280.
https://doi.org/10.1046/j.1365-2664.2001.00567.x
Rothermel, R. C. (1972). A mathematical model for predicting fire spread in wildland
fuels. USDA Forest Service Research Paper INT USA, 48. Retrieved from
http://www.srs.fs.usda.gov/pubs/32533
Rothermel, R. C. (1983). How to predict the spread and intensity of forest and range
fires. Boise, Idaho.
Rothermel, R. C., & Deeming, J. E. (1980). Measuring and interpreting fire behaviour
for correlation with fire effects. Ogden, Utah.
Russell-Smith, B. J. (2016). Fire management business in Australia’s tropical savannas:
Lighting the way for a new ecosystem services model for the north? Ecological
Management and Restoration, 17(1), 4–7. https://doi.org/10.1111/emr.12201
Russell-Smith, J., Cook, G. D., Cooke, P. M., Edwards, A. C., Lendrum, M., Meyer, C.
P. (Mick), & Whitehead, P. J. (2013a). Managing fire regimes in north Australian
savannas: Applying Aboriginal approaches to contemporary global problems.
Frontiers in Ecology and the Environment, 11(Figure 1), 55–63.
https://doi.org/10.1890/120251
Russell-Smith, J., & Edwards, A. C. (2006). Seasonality and fire severity in savanna
landscapes of monsoonal northern Australia. International Journal of Wildland
Fire, 15(4), 541. https://doi.org/10.1071/WF05111
Russell-Smith, J., Edwards, A. C., & Price, O. F. (2012). Simplifying the savanna: the
trajectory of fire-sensitive vegetation mosaics in northern Australia. Journal of
Biogeography, 39, 1303–1317. https://doi.org/10.1111/j.1365-2699.2012.02679.x

94
Russell-Smith, J., Monagle, C., Jacobsohn, M., Beatty, R. L., Bilbao, B., Millán, A., …
Sánchez-Rose, I. (2013b). Can savanna burning projects deliver measurable
greenhouse emissions reductions and sustainable livelihood opportunities in fire-
prone settings? Climatic Change, 140(1), 47–61. https://doi.org/10.1007/s10584-
013-0910-5
Russell-Smith, J., Murphy, B. P., Meyer, C. M., Cook, G. D., Maier, S., Edwards, A. C.,
… Brocklehurst, P. (2009a). Improving estimates of savanna burning emissions for
greenhouse accounting in northern Australia : limitations , challenges ,
applications. International Journal OfWildland Fire, 18, 1–18.
Russell-Smith, J., Ryan, P. G., Klessa, D., Waight, G., & Harwood, R. (1998). Fire
regimes, fire-sensitive vegetation and fire management of the sandstone Arnhem
Plateau, monsoonal northern Australia. Journal of Applied Ecology, 35, 829–846.
https://doi.org/10.1111/j.1365-2664.1998.tb00002.x
Russell-Smith, J., Sangha, K. K., Costanza, R., Kubiszewski, I., & Edwards, A. (2018).
Towards a sustainable diversified land sector economy for North Australia. In J.
Russell-Smith, H. Pedersen, G. James, & K. Sangha (Eds.), Sustainable land
sector development in Northern Australia: Indigenous rights, aspirations, and
cultural responsibilities. Boca Raton USA: CRC Press.
Russell-Smith, J., Whitehead, P., & Cooke, P. (2009b). Culture, ecology and economy
of fire management in north Australian savannas rekindling the Wurrk tradition.
(J. Russell-Smith, P. Whitehead, & P. Cooke, Eds.). Collingwood, Australia:
CSIRO Publications.
Russell-Smith, J., Whitehead, P. J., Cook, G. D., & Hoare, J. L. (2003). Response of
eucalyptus-dominated savanna to frequent fires: lessons from munmarlary, 1973–
1996. Ecological Monographs, 73(3), 349–375.
Russell-Smith, J., Yates, C. P., Whitehead, P. J., Smith, R., Craig, R., Allan, G. E., …
Gill, a. M. (2007). Bushfires “down under”: Patterns and implications of
contemporary Australian landscape burning. International Journal of Wildland
Fire, 16, 361–377. https://doi.org/10.1071/WF07018
Ryan, C. M., & Williams, M. (2011). How does fire intensity and frequency affect
miombo woodland tree populations and biomass? Ecological Applications, 21(1),
48–60.
Ryan, K. C., Knapp, E. E., & Varner, J. M. (2013). Prescribed fire in North American
forests and woodlands: history, current practice, and challenges. Frontiers in
Ecology and the Environment, 11(1), 15–24. https://doi.org/10.1890/120329
Sankaran, M., Hanan, N. P., Scholes, R. J., Ratnam, J., Augustine, D. J., Cade, B. S., …
Zambatis, N. (2005). Determinants of woody cover in African savannas. Nature,
438(8), 846–849. https://doi.org/10.1038/nature04070
Sato, M. N. (2003). Efeito a longo prazo de queimadas prescritas na estrutura de
comunidade de lenhosas da vegetação do cerrado sensu stricto. Universidade de
Brasília, Brasília, DF.
Sato, M. N., & Miranda, H. S. (1996). Mortalidade de plantas lenhosas do cerrado sensu
stricto submetidas a diferentes regimes de queima. In H. S. Miranda, C. H. Saito,

95
 & B. F. S. Dias (Eds.), Impactos de queimadas em áreas de cerrado e restinga (pp.
102–111). Brasília, DF: ECL/UNB.
Sato, M. N., Miranda, H. S., & Maia, J. M. F. (2010). O fogo e o estrato arbóreo do
Cerrado: efeitos imediatos e longo prazo. In H. S. Miranda (Ed.), Efeitos do regime
do fogo sobre a estrutura de comunidades de cerrado: Resultados do Projeto
Fogo. (pp. 77–91). Brasília: IBAMA/MMA.
Savadogo, P. (2007). Dynamics of Sudanian savanna-woodland ecosystem in response
to disturbances. Swedish University of Agricultural Sciences.
Schmidt, I. B., Fidelis, A., Miranda, H. S., & Ticktin, T. (2017). How do the wets burn?
Fire behavior and intensity in wet grasslands in the Brazilian savanna. Brazilian
Journal of Botany, 40(1), 167–175. https://doi.org/10.1007/s40415-016-0330-7
Schmidt, I. B., Fonseca, C. B., Ferreira, M. C., & Sato, M. N. (2016a). Experiências
internacionais de manejo integrado do fogo em áreas protegidas: recomendações
para implementação de manejo integrado de fogo no Cerrado. Biodiversidade
Brasileira, 6(2), 41–54.
Schmidt, I. B., Fonseca, C. B., Ferreira, M. C., & Sato, M. N. (2016b). Implementação
do programa piloto de manejo integrado do fogo em três unidades de conservação
do Cerrado. Biodiversidade Brasileira, 6(2), 55–70.
Schmidt, I. B., Moura, L. C., Ferreira, M. C., Eloy, L., Sampaio, A. B., Dias, P. A., &
Berlinck, C. N. (2018). Fire management in the Brazilian savanna: First steps and
the way forward. Journal of Applied Ecology, Policy Dir, 1–8.
https://doi.org/10.1111/1365-2664.13118
Scholes, R. J., & Archer, S. R. (1997a). Tree-Grass Interactions in Savannas. Annual
Review of Ecology and Systematics, 28(1), 517–544.
https://doi.org/10.1146/annurev.ecolsys.28.1.517
Scholes, R. J., & Archer, S. R. (1997b). Tree-Grass Interactions in Savannas. Annual
Review of Ecology and Systematics, 28(1), 517–544.
https://doi.org/10.1146/annurev.ecolsys.28.1.517
Scholes, R. J., Ward, D. E., & Justice, C. O. (1996). Emissions of trace gases and
aerosol particles due to vegetation burining in southern hemisphere Africa. Journal
of Geophysical Research, 101(D19), 23,623-677,682.
Scott, A. C., Chaloner, W. G., Belcher, C. M., & Roos, C. I. (2016). The interaction of
fire and mankind. Philosophical Transactions of the Royal Society B: Biological
Sciences, 371(1696), 20160149. https://doi.org/10.1098/rstb.2016.0149
Scott, A. C., & Glasspool, I. J. (2006). The diversification of Paleozoic fire systems and
fluctuations in atmospheric oxygen concentration, 103(29).
https://doi.org/0604090103
Scott, K., Setterfield, S. A., Douglas, M. M., Parr, C. L., Schatz, J., & Andersen, A. N.
(2012). Does long-term fire exclusion in an Australian tropical savanna result in a
biome shift? A test using the reintroduction of fire. Austral Ecology, 37(6), 693–
711. https://doi.org/10.1111/j.1442-9993.2012.02379.x
Setterfield, S. A. (2002). Seedling establishment in an Australian tropical savanna:

96
 effects of seed supply, soil disturbance and fire. Journal of Applied Ecology, 39(6),
949–959. https://doi.org/10.1046/j.1365-2664.2002.00772.x
Shaffer, L. J. (2010). Indigenous fire use to manage savanna landscapes in southern
Mozambique. Fire Ecology, 6(2), 43–59.
https://doi.org/10.4996/fireecolgy.0602043
Shea, R. W., Shea, B. W., Kauffman, J. B., Darold E., W., Haskins, C. I., & Scholes, M.
C. (1996). Fuel biomass and combustion factors associated with fires in savanna
ecosystems of South Africa and Zambia. Journal of Geophysical Research,
101(D19), 551–568. https://doi.org/http://dx.doi.org/10.1029/95JD02047;
doi:10.1029/9
Sheuyange, A., Oba, G., & Weladji, R. B. (2005). Effects of anthropogenic fire history
on savanna vegetation in northeastern Namibia. Journal of Environmental
Management, 75(3), 189–198. https://doi.org/10.1016/j.jenvman.2004.11.004
Sills, E. O., Atmadja, S. S., de Sassi, C., Duchelle, A. E., Kweka, D. L., Resosudarmo,
I. A. P., & Sunderlin, W. D. (2014). REDD+ on the ground: A case book of
subnational initiatives across the globe. Bogor, Indonesia: CIFOR.
Silva, I. A., & Batalha, M. A. (2010). Woody plant species co-occurrence in Brazilian
savannas under different fire frequencies. Acta Oecologica, 36(1), 85–91.
https://doi.org/10.1016/j.actao.2009.10.004
Silva, J. S., Rego, F., Fernandes, P., & Rigolot, E. (2010). Towards integrated fire
management: Outcomes of the European project fire paradox. Joensuu.
Silva, M. C., Boaventura, V. M., & Fioravanti, M. C. S. (2012). História do
povoamento bovino no Brasil central. Revista UFG, 13, 34–41.
Silveira, L., Rodrigues, F. H. G., Jácomo, A. T. de A., & Diniz Filho, J. A. F. (1999).
Impact of wildfires on the megafauna of Emas National Park, central Brazil. Oryx,
33(2), 108–114. https://doi.org/10.1046/j.1365-3008.1999.00039.x
Simon, M. F., Grether, R., Queiroz, L. P., Skema, C., Pennington, R. T., & Hughes, C.
E. (2009). Recent assembly of the Cerrado, a neotropical plant diversity hotspot,
by in situ evolution of adaptations to fire. Proceedings of the National Academy of
Sciences, 106. https://doi.org/20359-20364
Simon, M. F., & Pennington, T. (2012). Evidence for adaptation to fire regimes in the
tropical savannas of the Brazilian Cerrado. International Journal of Plant Sciences,
173(6), 711–723. https://doi.org/10.1086/665973
Skarpe, C. (1992). Dynamics of savanna ecosystems. Journal of Vegetation Science,
3(3), 293–300. https://doi.org/10.2307/3235754
Sletto, B. (2008). The Knowledge that Counts: Institutional Identities, Policy Science,
and the Conflict Over Fire Management in the Gran Sabana, Venezuela. World
Development, 36(10), 1938–1955. https://doi.org/10.1016/j.worlddev.2008.02.008
Souchie, F. F., Pinto, J. R. R., Lenza, E., Gomes, L., Maracahipes-Santos, L., &
Silvério, D. V. (2017). Post-fire resprouting strategies of woody vegetation in the
Brazilian savanna. Acta Botanica Brasilica, 31(2), 260–266.
https://doi.org/10.1590/0102-33062016abb0376

97
Staver, A. C., Bond, W. J., Stock, W. D., Rensburg, S. J. Van, & Waldram, M. S.
(2009). Herbivory and fire interact to suppress tree density Browsing in an African
savanna. Ecological Applications, 19(7), 1909–1919. https://doi.org/10.1890/08-
1907.1
Steel, Z. L., Safford, H. D., & Viers, J. H. (2015). The fire frequency-severity
relationship and the legacy of fire suppression in California forests. Ecosphere,
6(1), 1–23.
Stephens, S. L., Adams, M. A., Handmer, J., Kearns, F. R., Leicester, B., Leonard, J., &
Moritz, M. A. (2009). Urban-wildland fires: How California and other regions of
the US can learn from Australia. Environmental Research Letters, 4, 1–5.
https://doi.org/10.1088/1748-9326/4/1/014010
Stephens, S. L., & Ruth, L. W. (2005). Federal forest-fire policy in the United States.
Ecological Applications, 15(2), 532–542. https://doi.org/10.1890/04-0545
Stephenson, N. L., Parsons, D. L., & Swetnam, T. W. (1991). Restoring natural fire to
the sequoia-mixed conifer forest: should intense fire play a role? Proceedings of
the 17th Tall Timbers Fire Ecology Conference, 17, 321–337.
Stocks, B. J., & Martell, D. L. (2016). Forest fire management expenditures in Canada:
1970 – 2013. The Forestry Chronicle, 92(3), 298–306.
Suchet, S. (2002). “Totally Wild”? Colonising discourses, indigenous knowledges and
managing wildlife. Australian Geographer, 33(2), 141–157.
https://doi.org/10.1080/00049180220150972
Suich, H., Child, B., & Spenceley, A. (2009). Evolution and innovation in wildlife
conservation: Parks and game ranches to transfrontier conservation areas.
London and Sterling: Earthscan.
Symonds, M. R. E., & Moussalli, A. (2011). A brief guide to model selection,
multimodel inference and model averaging in behavioural ecology using Akaike’s
information criterion. Behavioral Ecology and Sociobiology, 65(1), 13–21.
https://doi.org/10.1007/s00265-010-1037-6
Tacconi, L., Moore, P. F., & Kaimowitz, D. (2006). Fires in tropical forests – what is
really the problem? lessons from Indonesia. Mitigation and Adaptation Strategies
for Global Change, 12(1), 55–66. https://doi.org/10.1007/s11027-006-9040-y
Tester, J. R. (1989). Effects of fire frequency on oak savanna in east-central Minnesota.
Bulletin of the Torrey Botanical Club, 116(2), 134.
https://doi.org/10.2307/2997196
TNC. (2018). Five ways we’ve improved Zambia’s fire management. Retrieved March
20, 2018, from https://www.nature.org/ourinitiatives/regions/africa/explore/five-
ways-weve-improved-zambias-fire-management-in-five-years.xml
Trauernicht, C., Brook, B. W., Murphy, B. P., Williamson, G. J., & Bowman, D. M. J.
S. (2015a). Local and global pyrogeographic evidence that indigenous fire
management creates pyrodiversity. Ecology and Evolution.
https://doi.org/10.1002/ece3.1494
Trauernicht, C., Brook, B. W., Murphy, B. P., Williamson, G. J., & Bowman, D. M. J.

98
 S. (2015b). Local and global pyrogeographic evidence that indigenous fire
management creates pyrodiversity. Ecology and Evolution, 5(9), 1908–1918.
https://doi.org/10.1002/ece3.1494
Trollope, L. A., & Trollope, W. S. W. (2002). Fire behaviour a key factor in the fire
ecology of African grasslands and savannas. In D. X. Viegas (Ed.), Forest fire
research and wildland fire safety. Coimbra, Portugal: Millpress Rotterdam
Netherlands. Retrieved from http://www.comets-uavs.org/papers/RAMIRO-
ICFFR-2002.pdf
Turner, M. G., & Dale, V. H. (1998). Comparing large, infrequent disturbances: What
have we learned? Ecosystems, 1(6), 493–496.
https://doi.org/10.1007/s100219900045
Uhl, C., & Kauffman, J. B. (1990). Deforestation, fire susceptibility, and potential tree
responses to fire in the eastern Amazon. Ecology, 71(2), 437–449.
UNDP. (2010). Mababe Zokotsama community development trust. Retrieved April 6,
2018, from
https://sgp.undp.org/index.php?option=com_sgpprojects&view=projectdetail&id=
14985&Itemid=272
UNU. (2015). The global potential of indigenous fire management: findings of the
regional feasibility assessments. Canberra, Australia.
van der Werf, G. R., Randerson, J. T., Giglio, L., Collatz, G. J., Mu, M., Kasibhatla, P.
S., … Van Leeuwen, T. T. (2010). Global fire emissions and the contribution of
deforestation, savanna, forest, agricultural, and peat fires (1997-2009).
Atmospheric Chemistry and Physics, 10(23), 11707–11735.
https://doi.org/10.5194/acp-10-11707-2010
van der Werf, G. R., Randerson, J. T., Giglio, L., Van Leeuwen, T. T., Chen, Y.,
Rogers, B. M., … Kasibhatla, P. S. (2017). Global fire emissions estimates during
1997-2016. Earth System Science Data, 9(2), 697–720.
https://doi.org/10.5194/essd-9-697-2017
van Vliet, N., Mertz, O., Birch-Thomsen, T., & Schmook, B. (2013). Is there a
continuing rationale for swidden cultivation in the 21st century? Human Ecology,
41(1), 1–5. https://doi.org/10.1007/s10745-013-9562-3
van Wilgen, B., Biggs, H. C., O’Regan, S. P., & Mare, N. (2000). A fire history of the
savanna ecosystems in the Kruger National Park, South Africa, between 1941 and
1996. South African Journal of Science, 96, 167–178. Retrieved from
http://www.ncbi.nlm.nih.gov/pubmed/3962
van Wilgen, B., Trollope, W., Biggs, H., Potgeiter, A., & Brockett., B. (2003). Fire as a
driver of ecosystem variability. In J. T. du Toit, K. H. Rogers, & H. C. Biggs
(Eds.), The Kruger experience: ecology and management of savanna heterogeneity
(pp. 149–170). Washington, Covelo and London: Island Press.
van Wilgen, B. W., & Biggs, H. C. (2011). A critical assessment of adaptive ecosystem
management in a large savanna protected area in South Africa. Biological
Conservation, 144(4), 1179–1187. https://doi.org/10.1016/j.biocon.2010.05.006
van Wilgen, B. W., Govender, N., & Biggs, H. C. (2007). The contribution of fire

99
 research to fire management: a critical review of a long-term experiment in the
Kruger National Park , South Africa. International Journal of Wildland Fire, 16,
519–530.
van Wilgen, B. W., Govender, N., Biggs, H. C., Ntsala, D., & Funda, X. N. (2004).
Response of savanna fire regimes to changing fire-management policies in a large
African national park. Conservation Biology, 18(6), 1533–1540.
https://doi.org/10.1111/j.1523-1739.2004.00362.x
Varner, J. M., Gordon, D. R., Putz, F. E., & Kevin Hiers, J. (2005). Restoring fire to
long-unburned Pinus palustris ecosystems: Novel fire effects and consequences for
long-unburned ecosystems. Restoration Ecology, 13(3), 536–544.
https://doi.org/10.1111/j.1526-100X.2005.00067.x
VCS. (2015). Verified Carbon Standard. Methodology for avoided forest degradation
through fire management.
Vellend, M. (2010). Conceptual synthesis in community ecology. The Quarterly Review
of Biology, 85(2), 183–206. https://doi.org/10.1086/652373
Wagner, C. V. (1978). Age-class distribution and the forest fire cycle. Canadian
Journal of Forest Research, 8(2), 220–227.
Walter, B. M. T., & Ribeiro, J. F. (2010). Diversidade fitofisionômica e o papel do fogo
no bioma Cerrado. In H. S. Miranda (Ed.), Efeitos do regime de fogo sobre a
estrutura de comunidades de Cerrado: Projeto Fogo (pp. 59–76). Brasília, DF:
IBAMA.
Walters, G. M. (2015). Changing fire governance in Gabon′s Plateaux Bateke savanna
landscape. Conservation and Society, 13(3), 275–286.
https://doi.org/10.4103/0972-4923.170404
Ward, D. E., Hao, W. M., Susott, R. a., Babbitt, R. E., Shea, R. W., Kauffman, J. B., &
Justice, C. O. (1996). Effect of fuel composition on combustion efficiency and
emission factors for African savanna ecosystems. Journal of Geophysical
Research: Atmospheres, 101(D19), 23569–23576.
https://doi.org/10.1029/95JD02595
Wardell, D. A., Nielsen, T. T., Rasmussen, K., & Mbow, C. . (2004). Fire history, fire
regimes and fire management in West Africa: an overview. In J. Goldammer & C.
de Ronde (Eds.), Wildland fire management handbook for sub-Sahara Africa (pp.
350–381). Freiburg: Oneworldbooks.
Weir, J. M. H., Johnson, E. A., & Miyanishi, K. (2000). Fire frequency and the spatial
age mosaic of the mixed-wood boreal forest in Western Canada. Ecological
Applications, 10, 1162–1177.
Welch, J. R., Brondízio, E. S., Hetrick, S. S., & Coimbra Jr, C. E. A. (2013). Indigenous
Burning as Conservation Practice : Neotropical Savanna Recovery amid
Agribusiness Deforestation in Central Brazil. Plos One, 8(12).
https://doi.org/10.1371/journal.pone.0081226
Welch, J. R., LeCompte, J. K., Butz, R. J., Steward, A. M., & Russell-Smith, J. (2018).
Anthropogenic fire history, ecology, and management in fire-prone landscapes: An
intercontinental review. In C. T. Fowler & J. R. Welch (Eds.), Fire Otherwise:

100
 Ethnobiology of Burning for a Changing World. Salt Lake City: The University of
Utah Press.
Werner, P. A. (2005). Impact of feral water buffalo and fire on growth and survival of
mature savanna trees: An experimental field study in Kakadu National Park,
northern Australia. Austral Ecology, 30(6), 625–647.
https://doi.org/10.1111/j.1442-9993.2005.01491.x
Werner, P. A., & Prior, L. D. (2013). Demography and growth of subadult savanna
trees: Interactions of life history, size, fire season, and grassy understory.
Ecological Monographs, 83(1), 67–93. https://doi.org/10.1890/12-1153.1
Whelan, R. J. (1995). The ecology of fire. Cambridge: Cambridge University Press.
Whelan, R. J., Rodgerson, L., Dickman, C. R., & Sutherland, E. F. (2002). Critical life
cycles of plants and animals: developing a process-based understanding of
population changes in fire-prone landscapes. In R. A. Bradstock, J. E. Williams, &
M. A. Gill (Eds.), Flammable Australia: the fire regimes and biodiversity of a
continent (pp. 94–124). Cambridge: Cambridge University Press.
White, R., Murray, S., & Rohweder, M. (2000). Pilot analysis of global ecosystems:
grassland ecosystems. World Resources Institute. Washington: World Resources
Institute. https://doi.org/10.1021/es0032881
Wickham, H. (2011). The split-apply-combine strategy for data analysis. Journal of
Statistical Software, 40(1), 1–29.
Williams, J. R., Griffiths, A. D., & Allan, G. E. (2002). Fire regimes and biodi- versity
in the wet–dry tropical landscapes of northern Australia. In R. A. Bradstock, J. E.
Williams, & A. M. Gill (Eds.), Flammable Australia: the fire regimes and
biodiversity of a continent (pp. 281–304). Cambridge: Cambridge University Press.
Williams, P. R., Congdon, R. A., Grice, A. C., & Clarke, P. J. (2005). Germinable soil
seed banks in a tropical savanna: seasonal dynamics and effects of fire. Austral
Ecology, 30(1), 79–90. https://doi.org/10.1111/j.1442-9993.2004.01426.x
Williams, R. J. (1995). Tree mortality in relation to fire intensity in a tropical savanna of
the Kakadu region, Northern Territory, Australia. CALMScience Supplement, 4,
77–82.
Williams, R. J., Cook, G. D., Gill, A. M., & Moore, P. H. R. (1999). Fire regime, fire
intensity and tree survival in a tropical savanna in northern Australia. Australian
Journal of Ecology, 24, 50–59.
Williams, R. J., Gill, A. M., & Moore, P. H. R. (1998). Seasonal changes in fire
behaviour in a tropical savanna in northern Australia. International Journal of
Wildland Fire, 8(4), 227–239. https://doi.org/doi:10.1071/WF9980227
Williamson, G. J., Prior, L. D., Jolly, W. M., Cochrane, M. A., Murphy, B. P., &
Bowman, D. M. J. S. (2016). Measurement of inter- and intra-annual variability of
landscape fire activity at a continental scale: The Australian case. Environmental
Research Letters, 11(3). https://doi.org/10.1088/1748-9326/11/3/035003
Woinarski, J. C. Z., Legge, S., Fitzsimons, J. A., Traill, B. J., Burbidge, A. A., Fisher,
A., … Ziembicki, M. (2011). The disappearing mammal fauna of northern

101
 Australia: Context, cause, and response. Conservation Letters, 4(3), 192–201.
https://doi.org/10.1111/j.1755-263X.2011.00164.x
Wright, H. A., & Bailey, A. . (1982). Fire ecology: United States and southern Canada.
New York: John Wiley & Sons.
Yates, C. P., Edwards, A. C., & Russell-Smith, J. (2008). Big fires and their ecological
impacts in Australian savannas: size and frequency matters. International Journal
of Wildland Fire, 17(6), 768–781. https://doi.org/10.1071/WF07150
Yibarbuk, D., Whitehead, P. J., Jackson, D., Godjuwa, C., Fisher, A., Cook, P., …
Bowman, D. M. J. S. (2001). Fire ecology and Aboriginal land management in
central Arnhem Land , northern Australia: a tradition of ecosystem management.
Journal of Biogeography, 28, 325–343.
Zammit, C. A., & Zedler, P. H. (1988). The influence of dominant shrubs, fire, and time
since fire on soil seed banks in mixed chaparral. Vegetatio, 75(3), 175–187.
https://doi.org/10.1007/BF00045632

102