Escolar Documentos
Profissional Documentos
Cultura Documentos
TESE DE DOUTORADO
Natal/RN
Maio de 2020
i
Comissão Examinadora:
Natal/RN
Maio de 2020
Universidade Federal do Rio Grande do Norte - UFRN
Sistema de Bibliotecas - SISBI
Catalogação de Publicação na Fonte. UFRN - Biblioteca Setorial Prof. Ronaldo Xavier de Arruda - CCET
A minha esposa Alyne Marques, cuja companhia me faz o homem mais feliz do
mundo, sendo também a minha melhor amiga. Além disso, agradeço pela compreensão,
palavras de amor e motivação nos momentos mais difíceis do doutorado.
A minha mãe Francisca Valentim da Silva Bezerra e ao meu pai Ilo Marques
Bezerra (in memoriam ) pela formação, amor e incentivo aos meus estudos.
Agradeço aos discentes do PPGCC pelo companheirismo nas aulas e nos eventos
do DCAC. Em especial aos colegas da sala 38 que compartilharam momentos de alegria
e aprendizado durante o doutorado.
Por m, agradeço ao Instituto Nacional do Semiárido pelos dados da torre de uxo,
ao Instituto Nacional de Meteorologia pelas informações meteorológicas e a NASA pelos
dados de sensoriamento remoto.
Tocando em frente
RESUMO
As Florestas Tropicais Sazonalmente Secas (SDTF, do inglês Seasonally Dry Tropical
Forests ) estão entre os biomas mais importantes em termos de uxos hidrológicos e de
carbono locais/globais e sua vulnerabilidade. O Bioma Caatinga é uma das maiores
SDTF do mundo que ocupa uma área contígua de 844.453 km2 , possui uma vasta
biodiversidade endêmica e é conhecido como uma importante área da vida selvagem
do planeta. Assim, o objetivo do presente estudo foi avaliar a variabilidade diária,
sazonal e interanual da evapotranspiração (ET ) e dos controles biofísicos e características
(a condutância da superfície - Gs ; fator de desacoplamento - Ω; e o coeciente de
Priestley-Taylor - α) de controle em um ambiente preservado do Bioma Caatinga durante
dois anos de seca na região Nordeste do Brasil, com um nível de detalhamento ainda
não desenvolvido neste bioma. As medições foram realizadas utilizando o sistema eddy
covariance durante o período de 1 de janeiro de 2014 a 31 de dezembro de 2015. Na
escala sazonal, os menores valores da ET foram observados durante a estação seca nos
dois anos do experimento (0,3 e 0,2 mm dia-1 ) como consequência da baixa umidade do
solo, que favorecem o fechamento parcial dos estômatos e reduzem os valores do Gs (0,22
e 0,13 mm s-1 ). O contrário ocorre no período mais chuvoso, cujas médias da ET (2,6 e
1,7 mm dia-1 ) e Gs (3,74 e 2,13 mm s-1 ) atingem valores mais elevados. Em relação aos
valores anuais, a diferença dos totais anuais de precipitação observados entre os anos é a
causa mais provável das diferenças observadas nos valores anuais da ET . Em 2014 a ET
foi 473,3 mm, em 2015 283,4 mm, o que implicou em uma diminuição dos valores de Gs ,
Ω e α. Os resultados sugerem que na estação seca a ET foi controlada pela umidade do
solo e, na estação chuvosa, foi parcialmente controlada pela energia disponível. Com o
presente estudo, houve um avanço no conhecimento sobre a variabilidade diária, sazonal e
interanual da ET e das suas características de controle. Essas informações são benécas,
por exemplo, na calibração de modelos para fazer representações mais realísticas sobre
o Bioma Caatinga. Como perspectivas futuras, pretende-se analisar o comportamento
da ET em anos contrastantes e avaliar a umidade do solo em profundidades múltiplas
durante uma próxima campanha experimental no Bioma Caatinga.
ABSTRACT
Seasonally Dry Tropical Forests (SDTF) are among the most important biomes regarding
regional/global hydrological and carbon uxes. The Caatinga Biome is one of the
largest STDF in the world that occupies a contiguous area of 844,453 km 2 , has a vast
endemic biodiversity and is known an important wildlife area on the planet. Thus, the
objective of this study was to evaluate the diurnal, seasonal and interannual variability
of evapotranspiration (ET ) and its biophysical control and characteristics (surface
conductance - Gs ; decoupling coecient - Ω; e Priestley-Taylor coecient - α) in a
preserved Caatinga Biome environment during two dry years in the Brazilian northeastern
region. A study on this subject with this level of detail in this biome is unprecedent.
Measurements were carried out using an eddy covariance system during the period from
1st January 2014 to 31st December 2015. The lowest ET values were observed in the
dry season of both experiment years (0.3 and 0.2 mm day-1 ) as a consequence of low
soil moisture, which favored partial stomatal closure and reduced Gs values (0.22 and
0.13 mm s-1 ). The opposite occured in the wet season, when ET (2.6 and 1.7 mm day-1 )
and Gs (3.74 and 2.13 mm s-1 ) means reached higher values. Regarding annual values,
dierences between total annual rainfall in both years is the most probable cause for the
dierences observed in annual ET values. In 2014, ET was of 473.3 mm while in 2015
it was of 283.4 mm, which incurred in an overall decrease in Gs , Ω and α values. The
results suggest that in the dry season, ET was controlled by soil moisture and, in the
rainy season, it was partially controlled by available energy. With the present study,
there was an advance in knowledge about the daily, seasonal and interannual variability
of ET and its control characteristics. This information is benecial, for example, in the
calibration of models to make more realistic representations about the Caatinga Biome.
As future perspectives, we intend to analyze the ET behavior in contrasting years and
to evaluate soil moisture at multiple depths during an upcoming experimental campaign
in the Caatinga Biome.
3. MATERIAL E MÉTODOS 32
3.1. Área de estudo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.2. Dados de superfície . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.3. Processamento e pós-processamento dos dados de baixa e alta frequência . . 35
3.4. Cálculo dos parâmetros de superfície . . . . . . . . . . . . . . . . . . . . . . 38
3.5. Condições da vegetação . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3.6. Análises estatísticas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3.7. Análise sazonal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4. RESULTADOS E DISCUSSÃO 42
4.1 Condições meteorológicas . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.2. Footprint e direção do vento . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.3. Variações diurnas e sazonais . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.4. Análise da variabilidade do fator de desacoplamento e do coeciente de
Priestley-Taylor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.5. Controles siológicos via condutância da superfície . . . . . . . . . . . . . . 52
5. CONCLUSÃO 59
6. PERSPECTIVAS FUTURAS 61
REFERÊNCIAS 62
APÊNDICE A 79
APÊNDICE B 84
APÊNDICE C 87
1
Lista de Figuras
1 Localização do Bioma Caatinga em relação ao Brasil e também aos
continentes americano e africano. . . . . . . . . . . . . . . . . . . . . . . . 12
2 Classicação climática do Brasil de acordo com a classicação proposta por
Köppen. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3 Diagrama simplicado do ciclo do carbono na superfície. As setas que
apontam para a atmosfera indicam que o carbono foi emitido e as setas
que apontam para a superfície indicam que houve sequestro de carbono. . . 15
4 Evolução da quantidade de dióxido de carbono (CO2 ) na atmosfera
terrestre, em partes por milhão (ppm) de 1959 até 2017. As medições
foram realizadas no observatório de Mauna Loa - Havaí, EUA. . . . . . . . 16
5 Componentes do balanço de radiação na superfície da Terra. Rn é o saldo
de radiação, RS↓ é a radiação de onda curta incidente, RS↑ é a radiação de
onda curta reetiva pela superfície, RL↓ é a radiação de onda longa emitida
pelas nuvens, partículas e gases, RLre↑ é a radiação de onda longa reetida
pela superfície e RLem↑ é a radiação de onda longa emitida pela terra. . . . 19
6 Componentes do balanço de energia à superfície. λE é o uxo de calor
latente, H é o uxo de calor sensível, G é o uxo de calor no solo, S é a
energia armazenada na biomassa e no ar dentro do dossel vegetativo e F é
a densidade do uxo de energia associada com o uxo de dióxido de carbono. 20
7 Média diária dos uxos de energia em 2014 (esquerda) e 2015 (direta)
durante o período de transição seco-chuvoso (a e b); período chuvoso (c
e d); período de transição chuvoso-seco (e e f); e o período seco (g e h).
Os termos Rn, λE , H , G e ε representam, respectivamente, o saldo de
radiação, o uxo de calor latente, o uxo de calor sensível, o uxo de calor
no solo e o resíduo (ou décit). . . . . . . . . . . . . . . . . . . . . . . . . 21
8 Efeitos do desmatamento no balanço hídrico. H , λE e ET indicam o
uxo de calor sensível, o uxo de calor latente e a evapotranspiração real,
respectivamente. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
9 Ilustração do processo de evapotranspiração. As setas de cor verde indicam
o processo de transpiração das plantas e as setas de cor laranja indicam o
processo de evaporação da água presente na superfície. . . . . . . . . . . . 26
10 Fluxo de água da folha para a atmosfera. As células-guarda regulam a
abertura estomática e, portanto, a taxa de transpiração. O duplo papel dos
estômatos como condutores de vapor de água e CO2 garante o acoplamento
da transpiração com a dinâmica do carbono e da energia. . . . . . . . . . . 27
11 Representação esquemática das resistências da superfície (rs ) e
aerodinâmica (ra ) para a evapotranspiração. . . . . . . . . . . . . . . . . . 28
12 Localização geográca da torre micrometeorológica instalada na
ESEC-Seridó. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
13 Parte da saída dos dados brutos de alta frequência processados em uxos
médios de meia-hora. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
14 Plataforma de processamento dos dados de uxo desenvolvida pelo
Instituto Max Planck de Biogeoquímica. . . . . . . . . . . . . . . . . . . . 37
2
3
4
LISTA DE ABREVIATURAS E SIGLAS
BE Balanço de Energia
IC Intervalo de Conança
5
LISTA DE SÍMBOLOS
ET Evapotranspiração real
Fc Fluxo de carbono
fr Frequência relativa
gs Condutância estomática
q Umidade especíca do ar
6
7
Rg Radiação Global
Rn Saldo de radiação
Ta Temperatura do ar
Ts Temperatura do solo
UR Umidade relativa do ar
VI Valor de importância
Ω Fator de desacoplamento
α Coeciente de Priestley-Taylor
γ Constante psicrométrica
ρa Densidade do ar
INTRODUÇÃO 1
O aumento das emissões dos gases do efeito estufa para a atmosfera nas
últimas décadas, com destaque para o dióxido de carbono (CO2 ), decorrente das
atividades humanas e sua inuência no clima, tornou-se um grande desao ecológico
e político (MONDANI et al., 2017). Isso pode ser exemplicado pelo que tem sido
evidenciado nos relatórios do Painel Intergovernamental de Mudanças Climáticas
(IPCC, do inglês Intergovernmental Panel on Climate Change ) e pela Administração
Oceânica e Atmosférica Nacional (NOAA, do inglês National Oceanic and Atmospheric
Administration ), indicando que a concentração de CO2 na atmosfera já ultrapassou 400
partes por milhão (ppm), sendo que estimativas para o ano de 1750 indicaram uma
concentração média de 280 ppm (LE TREUT et al., 2007).
Embora a maior parte das emissões de CO2 seja absorvida pelos oceanos, uma
parcela considerável da absorção desse gás é realizada pela biosfera terrestre. Por outro
lado, o sequestro de carbono pelos oceanos tem aumentado nos últimos anos, mas a
eciência dessa absorção têm sido menor. Dessa forma, avaliar o sequestro desse gás pela
biosfera terrestre se torna cada vez mais importante. Trocas gasosas com a atmosfera são
realizadas como resultado de processos biológicos (fotossíntese e respiração) nos quais
os ecossistemas terrestres atuam como sumidouro de carbono, chegando a compensar de
10% a 60% das emissões anuais de combustíveis fósseis (HOUGHTON, 2007; MONSON e
BALDOCCHI, 2014; PIAO et al., 2009; SCHIMEL et al., 2001). Desse modo, os biomas
terrestres desempenham um papel muito importante no ciclo global do carbono e são
considerados aliados em potencial na elaboração de estratégias de mitigação dos efeitos
das mudanças climáticas (LUYSSAERT et al., 2007; SCHIMEL et al., 2001).
9
10
entre regiões vegetadas por diferentes culturas (LIU et al., 2011; SCOTT, 2010).
Os efeitos das secas nas SDTF estão recebendo cada vez mais atenção de
pesquisas. Na Caatinga, estudos observacionais sobre a condutância estomática no
nível foliar (FALCÃO et al., 2015; MENDES et al., 2017; SANTOS et al., 2014),
uxos de energia e CO2 (CAMPOS et al., 2019; OLIVEIRA et al., 2006; PIRES et
al., 2017; SILVA et al., 2017; SOUZA et al., 2015; TEIXEIRA et al., 2008), Índice de
Vegetação da Diferença Normalizada (NDVI, do inglês Normalized Dierence Vegetation
Index ) (BARBOSA et al., 2006; CUNHA et al., 2015; MUTTI et al., 2019; SOUZA
et al., 2016) tornaram-se mais relevantes considerando as projeções climáticas sobre
mudanças climáticas para regiões semiáridas, que enfatizam cenários com aumentos da
11
Nesse contexto, o objetivo geral deste estudo foi avaliar a variabilidade diurna,
sazonal e interanual da evapotranspiração e das características de controle (condutância da
superfície, fator de desacoplamento e o coeciente de Priestley-Taylor) em um ambiente
preservado do Bioma Caatinga. Para isso, os seguintes objetivos especícos foram
propostos: (i) investigar os padrões diários, sazonais e anuais da evapotranspiração
do Bioma Caatinga e das características de controle; e (ii) analisar a sensibilidade da
condutância da superfície ao décit de pressão de vapor e temperatura do ar.
REVISÃO DE LITERATURA 2
Este capítulo visa apresentar os aspectos gerais do Bioma Caatinga, o ciclo do
carbono e os impactos nas mudanças climáticas, o balanço de energia na superfície e a
evapotranspiração e seus mecanismos de controle. Será apresentada uma conceituação
objetiva e uma fundamentação teórica abrangente em relação ao que tem sido pesquisado
na literatura cientíca sobre a interação biosfera-atmosfera em diversos biomas ao redor
do mundo.
1
Região biogeográca que compreende a América Central, incluindo a parte sul do México e da península
da Baja Califórnia, o sul da Flórida, todas as ilhas do Caribe e a América do Sul.
2
Ecossistema que precisa de uma entrada para manter seus processos vitais e, na maioria dos casos, um
meio de exportar energia, materiais processados e resíduos gerados.
12
13
A altitude média está entre 400 e 500 metros, porém a Caatinga também está
presente em parte do litoral dos estados do Rio Grande do Norte e Ceará, também
cobrindo grandes planaltos acima de 1000 metros de altitude (MENEZES et al., 2012).
Com relação a granulometria, as texturas do solo variam dos muito arenosos (Neossolos
Quartzarênicos, 4%) aos muito argilosos (Vertissolos, 1%). Os primeiros com menor
capacidade de retenção de água, mas de fácil percolação. Por exemplo, nos solos com
menos de um metro de profundidade, a água retida é suciente para suprir as plantas
apenas por poucas semanas. Após acabar o estoque de água, se não houver novas chuvas,
inicia-se um período de deciência hídrica. Nos solos com vários metros de profundidade,
o armazenamento de água pode durar meses e as plantas podem não ter deciência se
suas raízes conseguirem explorar um volume grande (MENEZES et al. 2012; SAMPAIO,
2010). Quatro ordens dominantes de solo formam um mosaico espacial cobrindo 66%
da área (Latossolos, 19%; Litossolos, 19%; Argissolos, 15% e Luvisolos, 13%) e mais de
80% da área tem alguma limitação do solo em relação ao uso agrícola (MENEZES et al.
14
2012).
Os climas que predominam na região são (i) semiárido caracterizado por baixas
altitudes e latitudes e (ii) Tropical com o inverno seco, denotados, respectivamente, por
BSh e As e acordo com a classicação climática de Köppen (Figura 2) (ALVARES et
al., 2014). A região é caracterizada por elevadas temperaturas médias (23-27 ◦ C), baixa
amplitude térmica, umidade relativa do ar por volta de 50%, insolação próxima a 2800
horas ano-1 e alta evapotranspiração potencial (acima de 1500 mm ano-1 ). As precipitações
são escassas (em média inferiores a 800 mm) e concentradas em poucos meses do ano
sendo seguidos de um prolongado período seco (ALVARES et al., 2014; MENEZES et al.,
2012; OLIVEIRA et al., 2017; SAMPAIO, 2010). A região apresenta uma variabilidade
interanual marcante na precipitação o que implica em secas de intensidade e períodos
variados (MARENGO et al., 2016; MARES et al. 1985; OLIVEIRA et al., 2017; RIBEIRO
et al., 2015; SAMPAIO et al. 1995). Projeções do Painel Brasileiro de Mudanças
Climáticas apontam para um aumento em até 4,5 ◦ C na temperatura do ar e agravamento
no décit hídrico da região até o nal do século 21 (PBMC, 2014).
λE + H = Rn − G − S − F. (2)
Rn = λE + H + G. (3)
aos uxos diurnos, exceto para G (ARYA, 2001). Por exemplo, a Figura 7 apresenta um
gráco clássico da partição do saldo de radiação no período de 24 horas em um estudo
realizado por Campos et al. (2019) em área de Caatinga preservada ao norte do semiárido
brasileiro.
Figura 7: Média diária dos uxos de energia em 2014 (esquerda) e 2015 (direta) durante
o período de transição seco-chuvoso (a e b); período chuvoso (c e d); período de transição
chuvoso-seco (e e f); e o período seco (g e h). Os termos Rn, λE , H , G e ε representam,
respectivamente, o saldo de radiação, o uxo de calor latente, o uxo de calor sensível, o
uxo de calor no solo e o resíduo (ou décit).
Segundo Foley et al. (2003) e Costa e Foley (2000), alterações no uso e na cobertura
da terra podem afetar os uxos e parâmetros biofísicos de superfície de várias maneiras.
O aumento do albedo (fração da radiação de onda curta incidente que é reetida) tende
a esfriar a superfície, porém a rugosidade supercial, a área foliar e a profundidade das
raízes, por serem mais baixas em pastagens do que em orestas, implicam em uma redução
drástica na ET , aumentando substancialmente a temperatura e diminuindo a precipitação
(Figura 8).
25
26
Estômatos são estruturas dos órgãos aéreos da maioria das plantas (Figura 10).
O termo estômato indica uma fenda microscópica (ou ostíolo) através da superfície do
órgão vegetal, permitindo a comunicação entre o seu interior e o ambiente externo e
um par de células especializadas, denominadas de células-guarda, que circundam essa
fenda. As células-guarda respondem a sinais ambientais, alterando as suas dimensões e
regulando o tamanho da abertura do ostíolo. Os estômatos são regulados por estímulos que
maximizam a absorção de CO2 e minimizam as perdas de vapor de água pela transpiração,
respondendo de maneira rápida a estímulos abióticos, sendo, portanto, um mecanismo de
controle das interações planta-atmosfera (ASSMANN, 1999; TAIZ et al., 2017). Altas
taxas de radiação, temperaturas e concentrações de CO2 no interior dos estômatos podem
acelerar o seu fechamento. Quando o potencial hídrico das folhas diminui, os estômatos
reduzem a abertura sendo este efeito preponderante sobre os demais fatores abióticos,
mesmo em condições de muita radiação solar, temperatura e CO2 . Os estômatos poderão
abrir-se quando for restituído o uxo de água da folha e superado o décit hídrico
(MARENCO e LOPES, 2009).
3
Dessecação é o estado de extrema secura, ou o processo de extrema secagem.
29
No Cerrado Campo Sujo, localizado no Mato Grosso, que é caracterizado por uma
vegetação herbácea e arbustiva apresentando baixa densidade vegetativa, Rodrigues et al.
(2014), durante um ano de estudo, identicaram que a variação sazonal na precipitação
e/ou a disponibilidade de água no solo foi a variável mais importante que controlou os
uxos de energia e o Gs . Durante a estação seca, os valores diários de Ω caram, em geral,
abaixo de 0,12 indicando um forte acoplamento superfície-atmosfera. Foi evidenciado
que a condutância estomática pode ter sido o controle mais importante sobre as taxas de
Gc e Ω, especialmente durante o estação seca quando o estresse hídrico limita e maiores
valores do VPD limitam a condutância estomática. Neste estudo, a estimativa do α não
foi obtida. Silva et al. (2017) objetivaram estudar os uxos de CO2 , água e energia na
Caatinga e pastagem no município de Serra Talhada em Pernambuco. Foi constatado
que em ambos os tipos de vegetação a ET na estação seca foi controlada pela vegetação
e na estação chuvosa foi controlada pelas condições atmosféricas. A ET sobre Caatinga
foi maior que sobre pastagem. Foi concluído que mudanças no uso da terra ao converter
áreas de Caatinga em pastagens tenderá a aumentar H e a diminuir a ET . Todavia, não
calcularam os valores de Gs para investigar o controle biofísico da evapotranspiração.
5
As plantas do tipo C4 (por exemplo, cana-de-açúcar) apresentam taxa fotossintética praticamente
constante com o aumento da temperatura (no limiar da temperatura ótima) . As do tipo C3 (por
exemplo, o trigo) apresentam um decrescimento dessa taxa com o aumento da temperatura (no limiar
da temperatura ótima).
Tabela 2: Comparações entre os valores da precipitação acumulada, evapotranspiração real (ET ), média do uxo diário de calor latente
(λE ), média da condutância da superfície média diária (Gs ), média diária do fator de desacoplamento (Ω) e média diária coeciente de
Priestley-Taylor (α) entre os estudos apresentados na revisão de literatura da ET e seus mecanismos de controle.
Precipitação ET total λE (b) Gs
Tipo de Vegetação e região Período Ω α Pesquisa
Total (mm) (mm) (W m-2 ) (mm s-1 )
Floresta tropical, Amazônia 1 de julho - 30 de junho,
2200,0 1281,2 99,5 12,7 − − Rocha et al. (2004)
Central 2000-2001
Floresta temperada 1 de janeiro - 31 de dezembro,
4362,0 1701,6 44,1 15,6 0,26 0,72 Wilson e Baldocchi (2000)
decídua, Estados Unidos 1995-1997(a)
Pastagem em prado alpino 11 de novembro - 10 de novembro,
394,7 347,8 27,0 9,8 0,42 0,66 Gu et al. (2005)
Planalto do Tibet 2002-2003
Pastagem em prado alpino 16 de outubro - 15 de outubro,
337,6 358,2 27,8 3,5 0,26 0,35 Ma et al. (2015)
Planalto do Tibet 2011-2013(a)
1 de janeiro - 31 de dezembro,
Pastagem úmida, Japão 1194,2 808,5 62,8 11,7 0,65 1,07 Li et al. (2005)
1999
Pastagem mediterrânea, 1 de janeiro - 30 de abril,
3366,0 1914,0 27,8 4,1 0,27 0,45 Ryu et al. (2008)
Califórnia/EUA 2002-2007(a)
Cerrado - Campo Sujo, 1 de março - 1 março,
1030,0 927,5 72 4,0(c) 0,34(c) − Rodrigues et al. (2014)
Mato Grosso 2011-2012
Caatinga preservada, 1 de março - 28 de fevereiro,
430,2 526,6 40,9 − 0,40 − Silva et al. (2017)
Semiárido Central 2014-2015
Pastagem, Semiárido 1 de março - 28 de fevereiro,
430,2 277,7 21,6 − 0,15 − Silva et al. (2017)
Central 2014-2015
(a) Valores observados em mais de um ano de estudo.
(b) Os valores das médias diárias do λE foram derivados da ET total e vice-versa.
(c) Os valores das médias diárias λE , Gs e Ω foram estimados pelos informações contidas no gráco.
Fonte: Elaborado pelo autor.
31
MATERIAL E MÉTODOS 3
Os procedimentos metodológicos que foram utilizados nesta pesquisa estão
detalhados neste capítulo. Sendo assim, o material e métodos são apresentados em
sete seções, a saber: (i) área de estudo; (ii) dados de superfície; (iii) processamento e
pós-processamento dos dados de baixa e alta frequência; (iv) cálculo dos parâmetros de
superfície; (v) condições da vegetação; (vi) análises estatísticas e (vii) análise sazonal.
Fonte: Elaborado pelo autor com dados disponibilizados pela Pesquisa Geológica dos Estados
Unidos em <https://earthexplorer.usgs.gov/>.
32
33
0, 408∆(Rn − G) + γ (Ta900 u · V PD
+273) 2
ET0 = , (4)
∆ + γ(1 + 0, 34u2 )
6
Publicação de número 56 da Organização das Nações Unidas para a Alimentação e a Agricultura (do
inglês Food and Agriculture Organization )
35
Rn = H + λE + G, (5)
H = ρ a cp w 0 T 0 (6)
λE = ρa λw0 q 0 , (7)
Figura 13: Parte da saída dos dados brutos de alta frequência processados em uxos
médios de meia-hora.
por Papale et al., (2006). A rotina desse algoritmo no software estatístico R (Apêndice
B) foi desenvolvida pelo Grupo de Estudos Observacionais e de Modelagem da Interação
Biosfera-Atmosfera (GEOMA) da Universidade Federal do Rio Grande do Norte (UFRN).
Neste experimento, ocorreram duas explicações para dados faltantes: (i) ausência
de dados gerada pela identicação de spikes, mas com os demais dados meteorológicos
disponíveis; (ii) ausência total dos dados de uxo e/ou meteorológicos. O preenchimento
das lacunas geradas pela ausência dos dados de uxos turbulentos foi realizada utilizando
a metodologia descrita por Reichstein (2005) e a estimativa do ltro da velocidade de
fricção (u∗ ) por Papale et al. (2006) com a colaboração da plataforma de estudos em
Biogeoquímica desenvolvida no Instituto Max Planck (<http://www.bgc-jena.mpg.de/
~MDIwork/eddyproc/>) (Figura 14).
Figura 14: Plataforma de processamento dos dados de uxo desenvolvida pelo Instituto
Max Planck de Biogeoquímica.
Na planilha eletrônica gerada pelo pós-processamento dos dados, cada valor que
preencheu uma lacuna gerada tem um número que representa a qualidade desse dado,
sendo os números 1, 2 e 3 para os dados originais, mais conáveis, médio e menos
conáveis, respectivamente. Vale ressaltar que todos os dados que tiveram qualidade
3 foram descartados (CAMPOS et al., 2019).
48
X 1800 · λEi
ET = , (8)
i=1
λ
Ga · γ · λE
Gs = , (9)
∆(Rn − G) + ρa · cp · Ga · V P D − λE(∆ + γ)
1 1 1
= + (10)
ra raM rb
ou ainda,
Ga = GaM + Gb . (11)
u2∗
GaM = (12)
u
ou ainda,
u
raM = , (13)
u2∗
em que u é a velocidade do vento acima do dossel que é estimada pelo anemômetro sônico.
∆ + 1
γ
Ω= . (16)
∆ + 1 + Ga
γ G s
ET ∆+γ λE
α= = · , (17)
ETeq ∆ H + λE
A correlação de Spearman (ρ) foi calculada para se vericar as relações entre duas
variáveis, sendo testada a um nível de signicância de 10% sob a hipótese de nulidade.
Modelos de regressões lineares e não-lineares, além da correlação de Spearman, foram
utilizados para avaliar a relação entre duas variáveis e os coecientes presentes nesses
modelos foram testados a um nível de signicância de 5% também sob a hipótese de
nulidade. As análises estatísticas foram realizadas com o software R (R CORE TEAM,
2018).
41
Figura 15: Precipitação média climatológica mensal e precipitação acumulada mensal nos
anos de 2014 e 2015 na região da ESEC-Seridó
200
Climatologia
2014
175
2015
150
Precipitação (mm)
125
100
75
50
25
0
JAN FEV MAR ABR MAI JUN JUL AGO SET OUT NOV DEZ
Meses
Fonte: Elaborado pelo autor a partir das informações do Departamento de Ciências Atmosféricas da
Universidade Federal de Campina Grande e do Banco de Dados Meteorológicos para Ensino e Pesquisa
do Instituto Nacional de Meteorologia.
42
43
80
(a) Período Período Período Período
Precipitação (mm)
chuvoso seco chuvoso seco
8 10
60
ET 0 (mm)
ET0
40
6
Precipitação
4
20
2
0
0
(b)
30
Index
R g (MJ m−2)
25
20
15
10
(c) Ta Ts
36
T (°C)
32
28
24
(d)
3
V P D (kPa)
2
1
0
J F M A M J J A S O N D J F M A M J J A S O N D
2014 2015
Mês
Fonte: Elaborado pelo autor.
Pela natureza persistente do período de seca, pode-se supor que esse foi um período
em que oscilações de baixa frequência, notadamente a Oscilação Decadal do Pacíco, que
associada aos regimes da Oscilação do Atlântico Norte podem induzir uma diferença na
posição climatológica da Zona de Convergência Intertropical, direcionando-a mais para
Norte, conforme mostrado por Kayano e Andreoli (2004), o que reduz signicativamente
a chuva sobre a maior parte da região Nordeste do Brasil. O fato da distribuição das
chuvas ter sido mais homogênea em 2014 e, em contraste, 2015 ter apresentado menos da
metade dos dias com chuva do ano anterior e dois eventos acima da média dos eventos
de precipitação intensa da região (OLIVEIRA et al., 2017), sugere-se que no ano mais
seco, as chuvas ocorrem de maneira isolada e intensa e o mecanismo de convecção local
é mais eciente durante os meses mais quentes do ano, o que concorda com análises
observacionais prévias (ZHOU e LAU, 2001).
44
Tabela 5: Médias diárias ± desvio padrão da radiação global (Rg , MJ m-2 dia-1 ), décit
de pressão de vapor (VPD, kPa), temperatura do ar (Ta , ◦ C), temperatura do solo (Ts ,
◦
C) e evapotranspiração de referência (ET0 , mm); totais sazonais e anuais da precipitação
(mm) e da evapotranspiração de referência (ET0 total, mm)
Trans. seco-chuvoso Período chuvoso Trans. chuvoso-seco Período Seco Anual
Variável
2014 2015 2014 2015 2014 2015 2014 2015 2014 2015
Rg 23,5±3,1 24,3±2,9 22,0±3,8 22,5±3,2 18,0±3,4 18,5±33,3 23,0±3,3 24,8±2,2 22,0±3,9 22,6±3,8
VPD 2,1±0,4 2,3±0,4 1,1±0,5 1,6±0,6 1,5±0,3 1,6±0,4 2,1±0,3 2,1±0,4 1,7±0,6 1,9±0,6
Ta 30,1±0,9 30,6±0,8 28,8±1,3 29,3±1,1 27,5±1,0 28,6±1,4 28,9±1,1 29,6±1,3 28,9±1,4 29,5±1,4
Ts 32,8±1,9 35,1±2,3 30,1±2,0 31,4±1,9 29,0±0,7 30,3±2,2 33,2±1,7 34,8±2,7 31,4±2,4 32,9±3,1
Precipitação 65,6 28,5 419,6 381,5 21,6 56,0 6,2 0,0 513,0 466,0
ET0 8,6±0,6 8,9±0,7 7,6±0,7 7,8±0,6 7,2±0,4 7,9±0,6 8,4±0,6 8,6±0,7 8,0±0,8 8,3±0,8
ET0 total 788,1 551,6 910,8 689,8 438,6 730,2 771,6 1043,7 2909,1 3015,3
Foi observado que existe uma consistência na análise anual do VPD, de forma que
o ano com menor precipitação (2015) foi mais quente e seco, resultando em maiores
valores do VPD. Souza et al. (2015), em área de Caatinga sob condições de seca
intensa, obtiveram valor médio anual do VPD (1,95 kPa) semelhante aos obtidos nesse
estudo. As variações sazonais foram semelhantes aos valores relatados para o Cerrado
(RODRIGUES et al., 2014) e oresta tropical decídua no norte da Tailândia (IGARASH
et al., 2015), que apresentam vegetação submetidas a uma estação seca prolongada. Os
valores apresentados na Tabela 5 também foram superiores aos obtidos por Silva et al.
(2017), entretanto, o fato das vegetações apresentarem características diferentes pode
impactar na evapotranspiração e, por consequência, na umidade relativa do ar.
Figura 17: Direção do vento e frequência acumulada durante os anos de 2014 (painel
esquerdo) e 2015 (painel direito) na área de estudo.
todas as estações, exceto durante o período seco, quando os seus valores foram quase
lineares durante todo o período diário. O padrão diurno da ET (Figuras 18c e 18d) foi
relativamente consistente com o comportamento da curva do Gs , entretanto os valores
máximos ocorreram entre 10:00 e 14:00 horas.
O padrão do ciclo diurno do Gs foi consistente para cada ano e estação observados.
Entretanto, a amplitude do ciclo diurno de Gs foi maior na estação chuvosa (Figuras 18a
e 18b). Os maiores valores de Gs pela manhã do que a tarde é consistente com vários
outros estudos que observaram máxima condutância da superfície e do dossel entre 8:00
e 12:00 horas e um mínimo ao entardecer. Por exemplo, no Cerrado (GIAMBELLUCA
et al., 2009; RODRIGUES et al., 2014, 2016) e na Amazônia (VOURLITIS et al., 2008).
Nesse sentido, o fechamento dos estômatos nos horários de maior demanda evaporativa
é um mecanismo de ajuste às condições de estresse para manter o status hídrico das
plantas (FLEXAS et al., 2004; PINHO-PESSOA et al. 2018).
(a) (b)
5
2014 2015
4
G s (mm s−1)
3
2
1
0
(c) (d)
0.15
0.10
E T (mm)
0.05
0.4 0.00
transição chuvoso−seco
0.2
Ω
0.1
0.0
0 3 6 9 12 15 18 21 24 0 3 6 9 12 15 18 21 24
Horário Local Horário Local
ET 1,0±0,7 0,6±0,3 2,6±1,1 1,7±0,9 0,9±0,4 0,8±0,6 0,3±0,2 0,2±0,1 1,3±1,2 0,8±0,8
Gs 0,85±0,83 0,50±0,36 3,74±2,19 2,13±1,49 1,00±0,65 0,87±0,67 0,22±0,21 0,13±0,13 1,66±2,00 0,87±1,13
Ω 0,08±0,07 0,04±0,03 0,29±0,14 0,17±0,11 0,09±0,06 0,08±0,06 0,02±0,02 0,01±0,01 0,14±0,15 0,07±0,09
α 0,27±0,21 0,18±0,15 0,72±0,30 0,50±0,27 0,37±0,19 0,30±0,19 0,09±0,06 0,05±0,04 0,39±0,33 0,24±0,25
NDVI 0,46±0,13 0,35±0,08 0,65±0,11 0,63±0,11 0,46±0,08 0,50±0,08 0,31±0,03 0,29±0,04 0,49±0,16 0,44±0,16
LAI 1,21±0,62 0,78±0,23 1,94±0,63 1,57±0,62 1,07±0,43 1,09±0,47 0,53±0,16 0,68±0,12 1,26±0,74 1,02±0,53
menor absorção de CO2 (FLEXAS e MEDRANO, 2002), que leva à perda de folhas
(senescência) na Caatinga sob condições de estresse hídrico. Assim, os baixos valores
da ET no período seco podem ser atribuídos à redução da área foliar (evidenciada pela
diminuição do LAI), senescência foliar das espécies da Caatinga, como um mecanismo
de tolerância à seca devido à baixa precipitação e limitação da capacidade de retenção
de água nos solos da região do estudo, classicados como cristalino, ou seja, solo com
pouca profundidade (FEITOSA e FEITOSA, 2011). Além da senescência foliar, o Gs
nas poucas folhas remanescentes das espécies semidecíduas foi drasticamente reduzido
com a diminuição da umidade do solo, reetindo valores extremamente baixos de ET .
Esses dados implicam diferentes interações planta-ambiente afetando ET nos diferentes
anos de estudo, evidenciando o papel do Gs no controle das taxas de evapotranspiração
da Caatinga atrelados a distribuição das chuvas e alterações da cobertura vegetal.
Apesar de 2015 ter sido mais seco (maior demanda por água na atmosfera), os
valores da ET foram menores, indicando o eciente mecanismo de resiliência à seca
que as espécies do Bioma Caatinga possuem, pois as mesmas fecham seus estômatos
no período de seca, diminuindo a perda de água por transpiração. Essa resiliência
é vericada pela variação sazonal dos valores diários de Ω que será discutido maia
adiante. Ao mesmo tempo, a fenologia do bioma responde rapidamente à ocorrência
de chuvas como pode ser vericado na variabilidade sazonal do NDVI e LAI (Figura
20a), que aumenta gradativamente da estação seca para a estação chuvosa. Esse
comportamento mostra que durante o período chuvoso, a vegetação (predominantemente
espécies decíduas) intensica as suas atividades siológicas e metabólicas, migrando
do estágio de dormência para o estágio de plena atividade fotossintética (BARBOSA
et al., 2006; BARBOSA e KUMAR, 2016; DOMBROSKI et al., 2011; SILVA et al.,
2017). Os resultados obtidos nesta pesquisa foram consistentes aos estudos anteriores
das regiões semiárida do Brasil. Por exemplo, Silva et al. (2017) mediram a ET na
Caatinga durante os anos de 2014 e 2015 e observaram maiores valores durante a estação
chuvosa e menores valores na estação seca. Em comparação com outras orestas, os
resultados obtidos no presente estudo estão de acordo com os padrões observados por
Ma et al. (2015) em ambiente semiárido nas estepes alpinas do Planalto Tibetano,
porém a restrição hídrica dessa região está relacionada ao período de congelamento do solo.
Os grácos de dispersão entre Gs e VPD para os anos de 2014 e 2015 são mostrados
nas Figuras 22a-b. Nota-se que o Gs decresceu exponencialmente com aumento do VPD.
No período seco, VPD > 1, 5 kPa ocasionou valores de Gs próximos de zero. No entanto,
o maior efeito do VPD no Gs foi observado no ano de 2014 quando comparado ao ano de
2015 (R2 = 0, 63 e 0, 19, respectivamente). Os grácos de dispersão entre Gs e Ta para os
anos de 2014 e 2015 são mostrados nas Figuras 22c-d. O melhor ajuste entre essas duas
variáveis só foi possível para o período chuvoso de 2014, sendo essa relação exponencial
(R2 = 0, 48). Nesse período, Ta > 31, 7◦ C (percentil 90%) implicou máximo de Gs igual
a 0,91 mms-1 . Esse valor de Gs é menor do que a média dos seus respectivos valores para
Ta ≤ 31, 7◦ C (4,08 mms-1 ). Por outro lado, no período chuvoso de 2015 não foi possível
observar um ajuste claro para os dados (R2 = 0, 27). Além disso, Ta > 32, 4◦ C (percentil
90%) implicou máximo de Gs igual a 2,1 mms-1 . Esse valor foi relativamente próximo
da média do Gs para Ta ≤ 32, 4◦ C (2,25 mms-1 ). Esses resultados são muito relevantes,
pois provavelmente o baixo Gs possa ser explicado pela respiração das plantas no nível
mínimo, apenas para se manterem vivas, ou seja, essa relação do Gs com Ta evidencia a
vulnerabilidade do bioma Caatinga diante do cenário de aumento das temperaturas no
contexto das mudanças climáticas.
53
Figura 20: (a) Índice de vegetação da diferença normalizada (NDVI) e índice de área
foliar (LAI); médias diárias (8:00-17:00 horas) da (b) condutância da superfície (Gs ),
(c) evapotranspiração real (ET ), (d) coeciente de Priestley-Taylor (α), e (e) fator de
desacoplamento (Ω) no período do estudo na Caatinga (ESEC-Seridó).
1.2
4
LAI (m2 m−2)
chuvoso seco chuvoso seco
0.8
3
NDVI
NDVI
2
LAI
0.4
1
0.0
0
(b)
8
G s (mm s−1)
Index
6
4
2
0
(c)
5
4
E T (mm)
3
2
1
1.5 0
(d)
Index
1.0
α
0.5
0.6 0.0
(e)
0.4
Ω
0.2
0.0
J F M A M J J A S O N D J F M A M J J A S O N D
2014 2015
Mês
55
56
1.2
2014 2015 Transição seco−chuvoso
Transição chuvoso−seco
1.0
base1$et_eteq
base2$et_eteq
0.8
α
0.6
0.4
R2 = 0.94 R2 = 0.88
valor−p < 0.001 valor−p < 0.001
0.0
0 2 4 6 8 0 2 4 6 8
G s (mm s−1)
base1$gs_mm G s (mm s−1)
base2$gs_mm
Figura 22: Relação entre a condutância média diária (Gs ) (08:00-17:00 horas) e o décit
de pressão de vapor (VPD ); e entre Gs e a temperatura do ar (Ta ) (08:00-17:00 horas)
na Caatinga (ESEC-Seridó) nos anos de 2014 (a e c) e 2015 (b e d), respectivamente. A
curva de cor preta foi ajustada para todos os dados de 2014 e 2015 e a curva de cor azul
foi ajustada apenas para o período chuvoso de 2014 e 2015.
(a) G s = 10.774exp(− 1.29VPD) (b) G s = 4.097exp(− 0.898VPD)
8
R2 = 0.63 R2 = 0.19
7
base2$gs_mm
G (mms−1)
5
4
2014 2015
s
3
2
1
0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
VPD (kPa) VPD (kPa)
R = 0.48
2 Transição seco−chuvoso
Transição chuvoso−seco
valor−p< 0.001
6
G s = 19.023 − 0.5496T a
G s (mms−1)
R2 = 0.27
valor−p< 0.001
4
2014 2015
2
0
26 28 30 32 34 36 26 28 30 32 34 36
T a (°C) T a (°C)
Figura 23: Relação entre a média de 16 dias da condutância média diária (Gs ) e os valores
de NDVI nos anos de 2014 (a) e 2015 (b) na Caatinga (ESEC-Seridó).
(a) (b)
6
Período chuvoso
Período seco
2014 2015 Transição seco−chuvoso
5
dados_16dias_2014$gs_mm
dados_16dias_2015$gs_mm
Transição chuvoso−seco
4
R2 = 0.97 R2 = 0.58
3
2
1
0
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.2 0.3 0.4 0.5 0.6 0.7 0.8
NDVI NDVI
gs4dias_2015[, 6]
G s = 0.502exp(0.838LAI)
G (mm s−1)
gs4dias_2014[−c(17,
R2 = 0.52
4
2
0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0
2 −2 2 −2
LAI (m m ) LAI (m m )
O Gs atingiu o pico por volta das 09:00 horas, antes do VPD atingir o seu valor
máximo. A antecipação do horário durante o qual o Gs atinge o seu pico máximo em
relação ao horário durante o qual o VPD atinge o seu máximo valor é um comportamento
recorrente em orestas. No entanto, em orestas tropicais, que incluem savanas tropicais,
orestas de transição, orestas tropicais sazonais e orestas tropicais perúmidas, o (Gs )
tem atingido o seu pico no início da manhã após o nascer do sol.
59
60
61
REFERÊNCIAS
AHLSTRÖM, Anders et al. The dominant role of semi-arid ecosystems in the trend and
variability of the land CO2 sink. Science, v. 348, n. 6237, p. 895-899, 2015.
AIRES, L. M.; PIO, C. A.; PEREIRA, J. S. The eect of drought on energy and
water vapour exchange above a mediterranean C3/C4 grassland in Southern Portugal.
Agricultural and Forest Meteorology, v. 148, n. 4, p. 565-579, 2008.
ALLEN, R. G. et al. Crop Evapotranspiration - Guidelines for computing crop water
requirements. FAO Irrigation and drainage paper 56, Rome, v. 156, p. 178, 1998.
ALTHOFF, Tiago Diniz et al. Climate change impacts on the sustainability of the
rewood harvest and vegetation and soil carbon stocks in a tropical dry forest in Santa
Teresinha Municipality, Northeast Brazil. Forest Ecology and Management, v. 360,
p. 367-375, 2016.
ALVARES, Clayton Alcarde et al. Köppen's climate classication map for Brazil.
Meteorologische Zeitschrift, v. 22, n. 6, p. 711-728, 2013.
ANDEREGG, William RL et al. Pervasive drought legacies in forest ecosystems and
their implications for carbon cycle models. Science, v. 349, n. 6247, p. 528-532, 2015.
ANDREOLI, Rita Valéria et al. Seasonal anomalous rainfall in the central and eastern
Amazon and associated anomalous oceanic and atmospheric patterns. International
Journal of Climatology, v. 32, n. 8, p. 1193-1205, 2012.
ARAÚJO, A. C. et al. The spatial variability of CO2 storage and the interpretation of
eddy covariance uxes in central Amazonia. Agricultural and Forest Meteorology,
v. 150, n. 2, p. 226-237, 2010.
62
63
BAI, Bing et al. Extensive translational regulation during seed germination revealed by
polysomal proling. New Phytologist, v. 214, n. 1, p. 233-244, 2017.
BALA, Anju et al. Estimation and validation of actual evapotranspiration for wheat
crop using SEBAL model over Hisar district, Haryana, India. Current Science, v. 113,
n. 1, 2017.
BALDOCCHI, Dennis et al. FLUXNET: A new tool to study the temporal and spatial
variability of ecosystem-scale carbon dioxide, water vapor, and energy ux densities.
Bulletin of the American Meteorological Society, v. 82, n. 11, p. 2415-2434, 2001.
BALDOCCHI, Dennis D. Assessing the eddy covariance technique for evaluating carbon
dioxide exchange rates of ecosystems: past, present and future. Global change biology,
v. 9, n. 4, p. 479-492, 2003.
BALDOCCHI, Dennis D.; VOGEL, Christoph A.; HALL, Brad. Seasonal variation of
energy and water vapor exchange rates above and below a boreal jack pine forest canopy.
Journal of Geophysical Research: Atmospheres, v. 102, n. D24, p. 28939-28951,
1997.
BALDOCCHI, Dennis D.; XU, Liukang. What limits evaporation from Mediterranean
oak woodlands-The supply of moisture in the soil, physiological control by plants or
the demand by the atmosphere? Advances in Water Resources, v. 30, n. 10, p.
2113-2122, 2007.
64
BURBA, George; ANDERSON, Dan. A brief practical guide to eddy covariance ux
measurements: principles and workow examples for scientic and industrial applications.
Li-Cor Biosciences, 2010.
CABRAL, Osvaldo MR et al. Fluxes of CO2 above a plantation of Eucalyptus in
southeast Brazil. Agricultural and Forest Meteorology, v. 151, n. 1, p. 49-59, 2011.
CABRAL, Osvaldo MR et al. Water and energy uxes from a woodland savanna
(cerrado) in southeast Brazil. Journal of Hydrology: Regional Studies, v. 4, p.
22-40, 2015.
CAMPOS, Suany et al. Closure and partitioning of the energy balance in a preserved area
of a Brazilian seasonally dry tropical forest. Agricultural and Forest Meteorology,
v. 271, p. 398-412, 2019.
CHEN, Shiping et al. Energy balance and partition in Inner Mongolia steppe ecosystems
with dierent land use types. Agricultural and Forest Meteorology, v. 149, n. 11,
p. 1800-1809, 2009.
DOMBROSKI, J. L. D. et al. Water relations of Caatinga trees in the dry season. South
African Journal of Botany, v. 77, n. 2, p. 430-434, 2011.
DOUGLAS, E. M. et al. The impact of agricultural intensication and irrigation on
land-atmosphere interactions and Indian monsoon precipitation-A mesoscale modeling
perspective. Global and Planetary Change, v. 67, n. 1-2, p. 117-128, 2009.
EAMUS, Derek et al. Carbon and water uxes in an arid-zone Acacia savanna woodland:
An analyses of seasonal patterns and responses to rainfall events. Agricultural and
Forest Meteorology, v. 182, p. 225-238, 2013.
EFRON, Bradley. Computers and the theory of statistics: thinking the unthinkable.
SIAM review, v. 21, n. 4, p. 460-480, 1979.
EICHELMANN, Elke et al. Evapotranspiration, water use eciency, and energy
partitioning of a mature switchgrass stand. Agricultural and forest meteorology, v.
217, p. 108-119, 2016.
FLEXAS, Jaume et al. Keeping a positive carbon balance under adverse conditions:
responses of photosynthesis and respiration to water stress. Physiologia Plantarum,
v. 127, n. 3, p. 343-352, 2006.
FOLEY, Jonathan A. et al. Green surprise? How terrestrial ecosystems could aect
earth's climate. Frontiers in Ecology and the Environment, v. 1, n. 1, p. 38-44,
2003.
GAO, Z. et al. Seasonal and diurnal variations in moisture, heat and CO 2 uxes over
a typical steppe prairie in Inner Mongolia, China. Hydrology and Earth System
Sciences, v. 13, n. 7, p. 987-998, 2009.
GAO, Z. et al. Seasonal and diurnal variations in moisture, heat and CO 2 uxes over
a typical steppe prairie in Inner Mongolia, China. Hydrology and Earth System
Sciences, v. 13, n. 7, p. 987-998, 2009.
GIAMBELLUCA, Thomas W. et al. Evapotranspiration and energy balance of Brazilian
savannas with contrasting tree density. Agricultural and forest meteorology, v. 149,
n. 8, p. 1365-1376, 2009.
GILLILAND, Joshua M.; KEIM, Barry D. Position of the South Atlantic anticyclone
and its impact on surface conditions across Brazil. Journal of Applied Meteorology
and Climatology, v. 57, n. 3, p. 535-553, 2018.
GIONGO, V. Balanço de carbono no Semiárido brasileiro; perspectivas e desaos.
Embrapa Semiárido-Capítulo em livro técnico (INFOTECA-E), 2011.
GRANIER, A. et al. The carbon balance of a young beech forest. Functional ecology,
v. 14, n. 3, p. 312-325, 2000.
HAO, Yanbin et al. Seasonal and interannual variation in water vapor and energy
exchange over a typical steppe in Inner Mongolia, China. Agricultural and Forest
Meteorology, v. 146, n. 1-2, p. 57-69, 2007.
HARRISON, Giles. Meteorological measurements and instrumentation. John Wiley &
Sons, 2015.
67
HE, Huachun et al. Assessment on the energy ow and carbon emissions of integrated
steelmaking plants. Energy Reports, v. 3, p. 29-36, 2017.
HEILMAN, J. L. et al. Soil and canopy energy balances in a west Texas vineyard.
Agricultural and Forest Meteorology, v. 71, n. 1-2, p. 99-114, 1994.
HILLER, Rebecca; ZEEMAN, Matthias J.; EUGSTER, Werner. Eddy-covariance ux
measurements in the complex terrain of an Alpine valley in Switzerland. Boundary-
Layer Meteorology, v. 127, n. 3, p. 449-467, 2008.
HOUGHTON, R. A. Balancing the global carbon budget. Annual Review of Earth
and Planetary Sciences, v. 35, p. 313-347, 2007.
IGARASHI, Yasunori et al. Environmental control of canopy stomatal conductance
in a tropical deciduous forest in northern Thailand. Agricultural and Forest
Meteorology, v. 202, p. 1-10, 2015.
IPCC, CLIMATE CHANGE. The Physical Science Basis.Contribution of Working
Group I to the Fifth Assessment Report of the Intergovernmental Panel on
Climate Change, p. 33-118, 2013.
JARVIS, Paul G.; STEWART, John B.; MEIR, Patrick. Fluxes of carbon dioxide at
Thetford Forest. Hydrology and Earth System Sciences Discussions, v. 11, n. 1,
p. 245-255, 2007.
KABAT, Pavel et al. (Ed.). Vegetation, water, humans and the climate: A new
perspective on an internactive system. Springer Science & Business Media, 2004.
LEI, HUI-MIN; YANG, DA-WEN. Seasonal and interannual variations in carbon dioxide
exchange over a cropland in the North China Plain. Global Change Biology, v. 16,
n. 11, p. 2944-2957, 2010.
LEI, HUI-MIN; YANG, DA-WEN. Seasonal and interannual variations in carbon dioxide
exchange over a cropland in the North China Plain. Global Change Biology, v. 16,
n. 11, p. 2944-2957, 2010.
LI, S.-G. et al. Evapotranspiration from a wet temperate grassland and its sensitivity to
microenvironmental variables. Hydrological Processes: An International Journal,
v. 19, n. 2, p. 517-532, 2005.
LUTGENS, Frederick K.; TARBUCK, Edward J. The atmosphere. Upper Saddle River,
NJ: Prentice Hall, 2001.
LUYSSAERT, Sebastiaan et al. CO2 balance of boreal, temperate, and tropical forests
derived from a global database. Global change biology, v. 13, n. 12, p. 2509-2537,
2007.
MA, Ning et al. Environmental and biophysical controls on the evapotranspiration over
the highest alpine steppe. Journal of Hydrology, v. 529, p. 980-992, 2015.
MAGRIN, G. O. et al. Central and South America in Climate Change 2014: Impacts,
Adaptation, and Vulnerability. Part B: Regional Aspects. Contribution of
Working Group II to the Fifth Assessment Report of the Intergovernmental
Panel of Climate Change (eds. Barros, VR et al.), v. 1499, p. 1566, 2014.
MALHI, Yadvinder et al. Comprehensive assessment of carbon productivity, allocation
and storage in three Amazonian forests. Global Change Biology, v. 15, n. 5, p.
1255-1274, 2009.
MEIR, Pedro et al. The fate of assimilated carbon during drought: impacts on respiration
in Amazon rainforests. Philosophical Transactions of the Royal Society B: Biological
Sciences, v. 363, n. 1498, p. 1849-1855, 2008.
MENDES, Keila R. et al. Croton blanchetianus modulates its morphophysiological
responses to tolerate drought in a tropical dry forest. Functional plant biology, v. 44,
n. 10, p. 1039-1051, 2017.
METCALFE, Daniel B. et al. Shifts in plant respiration and carbon use eciency at a
large-scale drought experiment in the eastern Amazon. New Phytologist, v. 187, n. 3,
p. 608-621, 2010.
MILES, Lera et al. A global overview of the conservation status of tropical dry forests.
Journal of Biogeography, v. 33, n. 3, p. 491-505, 2006.
MIRANDA, A. C. et al. Fluxes of carbon, water and energy over Brazilian Cerrado: an
analysis using eddy covariance and stable isotopes. Plant, Cell & Environment, v.
20, n. 3, p. 315-328, 1997.
PIAO, Shilong et al. The carbon balance of terrestrial ecosystems in China. Nature, v.
458, n. 7241, p. 1009, 2009.
PIELKE, Roger A. et al. Interactions between the atmosphere and terrestrial ecosystems:
inuence on weather and climate. Global change biology, v. 4, n. 5, p. 461-475, 1998.
PIRES, Wilis Neiva et al. Fluxos de radiação, energia, CO2 e vapor d'água em uma área
de caatinga em regeneração. Agrometeoros, v. 25, n. 1, 2017.
QIN, Zhong et al. Identication of important factors for water vapor ux and CO2
exchange in a cropland. Ecological Modelling, v. 221, n. 4, p. 575-581, 2010.
QUANSAH, Emmanuel et al. Carbon dioxide uxes from contrasting ecosystems in the
Sudanian Savanna in West Africa. Carbon balance and management, v. 10, n. 1, p.
1, 2015.
QUANSAH, Emmanuel et al. Carbon dioxide uxes from contrasting ecosystems in the
Sudanian Savanna in West Africa. Carbon balance and management, v. 10, n. 1, p.
1, 2015.
R Core Team, 2018. R: a language and environment for statistical computing. In: R
Foundation for Statistical Computing, Vienna, Austria.
RÄSÄNEN, Matti et al. Carbon balance of a grazed savanna grassland ecosystem in
South Africa. Biogeosciences, 2017.
RAJAN, Nithya; MAAS, Stephan J.; CUI, Song. Extreme drought eects on summer
evapotranspiration and energy balance of a grassland in the Southern Great Plains.
Ecohydrology, v. 8, n. 7, p. 1194-1204, 2015.
REICHSTEIN, Markus et al. On the separation of net ecosystem exchange into
assimilation and ecosystem respiration: review and improved algorithm. Global
Change Biology, v. 11, n. 9, p. 1424-1439, 2005.
73
ROCHA, Humberto R. et al. Seasonality of water and heat uxes over a tropical forest
in eastern Amazonia. Ecological applications, v. 14, n. sp4, p. 22-32, 2004.
SÁNCHEZ, J. M.; CASELLES, V.; RUBIO, E. M. Analysis of the energy balance closure
over a FLUXNET boreal forest in Finland. Hydrology and Earth System Sciences,
v. 14, n. 8, p. 1487-1497, 2010.
SAMPAIO, E.V.S.B. Overview of the Brazilian Caatinga. In: Bullock, S.H., Mooney,
H.A., Medina, E. (Eds.). Seasonally Dry Tropical Forests, University Press, Cambridge,
p. 35-63, 1995.
SANTOS, Mauro G. et al. Caatinga, the Brazilian dry tropical forest: can it tolerate
climate changes?. Theoretical and Experimental Plant Physiology, v. 26, n. 1, p.
83-99, 2014.
SANTOS, Rubens M. et al. Identity and relationships of the Arboreal Caatinga among
other oristic units of seasonally dry tropical forests (SDTFs) of north-eastern and
Central Brazil. Ecology and Evolution, v. 2, n. 2, p. 409-428, 2012.
SCOTT, Russell L. Using watershed water balance to evaluate the accuracy of eddy
covariance evaporation measurements for three semiarid ecosystems. Agricultural and
Forest Meteorology, v. 150, n. 2, p. 219-225, 2010.
SELLERS, Piers et al. The Boreal Ecosystem-Atmosphere Study (BOREAS): an
overview and early results from the 1994 eld year. Bulletin of the American
Meteorological Society, v. 76, n. 9, p. 1549-1577, 1995.
SELLERS, William D. Physical climatology. University of Chicago Press, 1965.
SILVA, Paulo Ferreira et al. Seasonal patterns of carbon dioxide, water and energy uxes
over the Caatinga and grassland in the semi-arid region of Brazil. Journal of Arid
Environments, v. 147, p. 71-82, 2017.
75
SCANLON, Todd M.; KUSTAS, William P. Partitioning carbon dioxide and water vapor
uxes using correlation analysis. Agricultural and Forest Meteorology, v. 150, n.
1, p. 89-99, 2010.
SOTTA, Eleneide Do et al. Eects of an induced drought on soil carbon dioxide (CO2 )
eux and soil CO2 production in an Eastern Amazonian rainforest, Brazil. Global
Change Biology, v. 13, n. 10, p. 2218-2229, 2007.
SOUZA, Leonardo Queiroz et al. How much nitrogen is xed by biological symbiosis in
tropical dry forests? 1. Trees and shrubs. Nutrient Cycling in Agroecosystems, v.
2, n. 94, p. 171-179, 2012.
TAGESSON, Torbern et al. Dynamics in carbon exchange uxes for a grazed semi-arid
savanna ecosystem in West Africa. Agriculture, Ecosystems & Environment, v.
205, p. 15-24, 2015.
TOMASELLA, Javier et al. Desertication trends in the Northeast of Brazil over the
period 2000-2016. International Journal of Applied Earth Observation and
Geoinformation, v. 73, p. 197-206, 2018.
TUCCI, M. L. S. et al. Diurnal and seasonal variation in photosynthesis of peach palms
grown under subtropical conditions. Photosynthetica, v. 48, n. 3, p. 421-429, 2010.
VON RANDOW, Celso et al. Inter-annual variability of carbon and water uxes in
Amazonian forest, Cerrado and pasture sites, as simulated by terrestrial biosphere
models. Agricultural and Forest meteorology, v. 182, p. 145-155, 2013.
VON RANDOW, Celso et al. Scale variability of atmospheric surface layer uxes
of energy and carbon over a tropical rain forest in southwest Amazonia 1. Diurnal
conditions. Journal of Geophysical Research: Atmospheres, v. 107, n. D20, p.
LBA 29-1-LBA 29-12, 2002.
VOURLITIS, G. L. et al. Spatial and temporal variations in hectare-scale net CO2 ux,
respiration and gross primary production of Arctic tundra ecosystems. Functional
Ecology, v. 14, n. 2, p. 203-214, 2000.
VOURLITIS, George L. et al. Energy balance and canopy conductance of a tropical
semi-deciduous forest of the southern Amazon Basin. Water Resources Research, v.
44, n. 3, 2008.
77
WEVER, Linda A.; FLANAGAN, Lawrence B.; CARLSON, Peter J. Seasonal and
interannual variation in evapotranspiration, energy balance and surface conductance in a
northern temperate grassland. Agricultural and Forest meteorology, v. 112, n. 1,
p. 31-49, 2002.
ZERI, Marcelo et al. Variability of carbon and water uxes following climate extremes
over a tropical forest in southwestern Amazonia. PloS one, v. 9, n. 2, p. e88130, 2014.
ZHA, Tianshan et al. Controls of evapotranspiration and CO2 uxes from scots pine by
surface conductance and abiotic factors. PloS one, v. 8, n. 7, p. e69027, 2013.
ZHOU, Jiayu; LAU, K.-M. Principal modes of interannual and decadal variability of
summer rainfall over South America. International Journal of Climatology: A
Journal of the Royal Meteorological Society, v. 21, n. 13, p. 1623-1644, 2001.
ZHU, Gaofeng et al. Energy ux partitioning and evapotranspiration in a sub-alpine
spruce forest ecosystem. Hydrological Processes, v. 28, n. 19, p. 5093-5104, 2014.
APÊNDICE A
O MÉTODO EDDY COVARIANCE
O método EC é uma técnica utilizada para quanticar em alta frequência
o transporte de gases, uxo de calor latente, uxo de calor sensível e momento
entre a superfície terrestre e a atmosfera (AUBINET et al., 2012; BURBA, 2013;
FOKEN, 2008). Nesse sentido, o método se tornou padrão na estimativa das trocas
líquidas de CO2 entre os ecossistemas e a atmosfera para longos períodos (HILLER
et al., 2008). A proliferação global de locais que fazem essa estimativa fornece uma
contribuição para estudo de controles ambientais, biológicos e climatológicos da troca
líquida entre a vegetação e a atmosfera (BALDOCCHI et al., 2001; WILSON et al., 2002).
Figura 24: Localização das 562 torres ativas da FLUXNET até março de 2016 e cobertura
terrestre da classicação MODIS IGBP.
79
80
Figura 25: Fluxo de ar no ecossistema. A seta para direita indica a direção predominante
do vento.
Os uxos turbulentos verticais (Ft ) são denidos como a média do produto das
propriedades contidas e transportadas em um vórtice. Essas propriedades são a sua massa
(a qual considerando uma unidade de volume é determinada por sua densidade ρa ), sua
velocidade vertical (w ) e o conteúdo volumétrico de qualquer outra variável de interesse
(s), conforme a Equação 17 (BURBA e ANDERSON, 2010; OKE, 1987):
Ft = ρa ws. (18)
A unidade do Ft será a quantidade transferida por unidade de área por unidade de tempo
(STULL, 1988), sendo geralmente representado em µmol m−2 s−1 .
81
x = x + x0 , (19)
(i) x0 = 0
(ii) xy = x y + x0 y 0
(iii) x y = x y
(iv) ax = ax
(v) x + y = x + y
82
F = ρ a w 0 s0 (23)
sws = w0 s0 . (26)
83
Sabendo que o uxo de carbono (Fc ) é a soma dos uxos gerados pela atividade
fotossintética e pela respiração das espécies presentes na superfície (SCANLON e
KUSTAS, 2010), então H , λE e Fc podem ser estimados baseando-se na Equação 25
(ARYA, 2001; BURBA e ANDERSON, 2003; HARTMANN, 1994; OKE, 1987; STULL,
1988).
H = ρ a cp w 0 T 0 (27)
λE = ρa λw0 q 0 (28)
F c = ρ a w 0 c0 (29)
#########################################################################
# #
# REMOVER OS DADOS ESPÚRIOS - SCRIPT R #
# #
# Copyright: GEOMIBA Ano: 2017 #
# #
# Grupo e Estudos Observacionais e de Modelagem da Interação #
# Biosfera-Atmosfera #
#########################################################################
dados = read.table("nome_do_arquivo.txt",head=F,dec=".")
########################################################################
########################################################################
################# #################
################# RODAR O PROGRAMA SPIKES #################
################# #################
########################################################################
########################################################################
spikes = function(N){
84
85
x=dados
y=na.omit(x)
n=floor(nrow(y)/(48))
M = matrix(y[1:(48*n),1],48,n)
m1=nrow(M)
n1=ncol(M)
medianas=round(apply(M,2,median),4)
for(i in 1:N){
max_m = max(medianas[(i+1):(N+1)])
min_m = min(medianas[(i+1):(N+1)])
A[i,1] = (max_m - min_m)
for(i in (N+1):(n1-N)){
max_ant = max(medianas[(N+1):(i+1)])
min_ant = min(medianas[(N+1):(i+1)])
max_dep = max(medianas[(i-N):(i-1)])
min_dep = min(medianas[(i-N):(i-1)])
max_aux = max(max_ant,max_dep)
min_aux = min(min_ant,min_dep)
A[i,1] = (max_aux-min_aux)
}
for(i in (n1-N+1):n1){
max_m = max(medianas[(i-N):(i-1)])
min_m = min(medianas[(i-N):(i-1)])
A[i,1] = (max_m-min_m)
}
for(l in 1:n1){
for(ind in 2:m1){
repmat = function(X,m,n){
##R equivalent of repmat (matlab)
mx = dim(X)[1]
nx = dim(X)[2]
matrix(t(matrix(X,mx,nx*n)),mx*m,nx*n,byrow=T)
}
86
A_aux = repmat(t(A),m1-1,1)
aux[2:m1,]=comp
saida[1:(m1*n1)]=matrix(aux,m1*n1,1)
s=which(saida==1)
s1=which(saida==0 | saida==1)
par(mar=c(4.2,4.2,1,1))
pdf("gráfico.pdf",width=14,height=8)
plot(s1,y[s1,1],col="grey20",type="l",axes=F,xlab="",ylab="")
par(new=TRUE)
plot(s,y[s,1],col="red",pch="*",cex=1.5,
xlab="Posição no banco de dados",ylab="Valor da variável")
legend("bottomright",legend=c("spike"),pch="*",col="red",cex=1.0)
dev.off()
tabela=data.frame(y,saida)
tabela[which(tabela[,2]==1),] = NA
write.table(tabela[,1],"saida.txt",sep="\t",row.names=FALSE)
########################################################################
########################################################################
################# #################
################# FIM DO PROGRAMA #################
################# #################
########################################################################
########################################################################
87
Agricultural and Forest Meteorology 287 (2020) 107957
A R T I C LE I N FO A B S T R A C T
Keywords: Seasonally dry tropical forests are among the most important biomes regarding regional and global hydrological
Evapotranspiration and carbon fluxes. Thus, the objective of this study was to evaluate the seasonal and interannual variability of
Bulk surface conductance evapotranspiration (ET) and its biophysical control and characteristics (surface conductance—Gs; decoupling
Decoupling factor coefficient—Ω; ratio between actual evapotranspiration and equilibrium evapotranspiration—ET/ETeq) in a
Caatinga Biome
preserved Caatinga Biome environment during two dry years in the Northeast Brazil region. A study on this
Brazilian Semiarid
subject with this level of detail in this biome is unprecedent. Measurements were carried out using an eddy
covariance system during the period from 1st January 2014 to 31st December 2015. The lowest ET values were
observed in the dry season of both experiment years (0.3 and 0.2 mm day−1) as a consequence of poor water
availability, which favored partial stomatal closure and reduced Gs values (0.22 and 0.13 mm s−1). The opposite
occurred in the wet season, when ET (2.6 and 1.7 mm day−1) and Gs (3.74 and 2.13 mm s−1) means reached
higher values. Regarding annual values, differences between total annual rainfall in both years is the most
probable cause for the differences observed in annual ET values. In 2014, annual ET was of 473.3 mm while in
2015 it was 283.4 mm, which incurred in an overall decrease in Gs, Ω and ET/ETeq values. Leaf senescence and
extremely low Gs values during the dry season suggest that the trees of the Caatinga Biome are more resilient
regarding the use of water and are avoiding water stress caused under low water availability.
⁎
Corresponding author
E-mail address: thiago.valentim@ifrn.edu.br (T.V. Marques).
https://doi.org/10.1016/j.agrformet.2020.107957
Received 15 February 2019; Received in revised form 27 February 2020; Accepted 2 March 2020
0168-1923/ © 2020 Published by Elsevier B.V.
T.V. Marques, et al. Agricultural and Forest Meteorology 287 (2020) 107957
the year (Oliveira et al., 2017; Mutti et al., 2019). The Caatinga has Mutti et al., 2019; Santos et al., 2020). On the other hand, a detailed
been identified as one of the most important wildlife regions of the analysis of the characteristics of ET control in the Caatinga Biome has
globe and most biodiverse dry forests (Mittermeier et al., 2003; not yet been developed. Although Teixeira et al., (2008) briefly com-
Pennington et al., 2006; Santos et al., 2014; Koch et al., 2017). mented the characteristics of surface resistance (reciprocal of Gs), the
Nevertheless, only 1% of this Biome has been converted into protected analysis of environmental controls was not carried out, since it was not
areas and the Caatinga and other SDTF have received less attention the objective of said study. In this sense, more effort is needed in order
than tropical rainforests regarding research efforts (Koch et al., 2017; to better comprehend the environmental and biophysical controls of ET
Tomasella et al., 2018). and in which way they affect heat and mass transfer in the Caatinga.
The CO2 transfer rate from the atmosphere to the carboxylation sites Thus, long term detailed studies on the energy and water fluxes be-
during the photosynthesis process is closely related to water lost to the tween vegetation and the atmosphere are needed due to the vulner-
atmosphere through leaf transpiration. These exchanges are controlled ability of this environment to anthropic activities and climate change.
by the interaction of various environmental factors (such as solar ra- Furthermore, the accuracy of climate scenarios and their impacts on
diation, air temperature, vapor pressure deficit—VPD and soil water climate, biodiversity and the population of these region is still deba-
content), besides biological processes inherent to the vegetation such as table, and uncertainties are high (Magrin et al., 2014; Marengo and
leaf emergence and development and stomatal conductance (Zha et al., Bernasconi, 2015). Such a challenge reinforces the critical need for
2013). At the plant level, physiological control of CO2 absorption and studies on how responds to environmental variables and how it influ-
transpiration is carried out by stomata and the modulation of such ences energy and mass fluxes, which is particularly important for un-
processes is quantified in terms of leaf stomatal conductance derstanding future biosphere-atmosphere interactions in the Brazilian
(Fanourakis et al., 2013; Lin et al., 2015; Wehr et al., 2017). At the Semiarid region.
ecosystem level, the control of both evapotranspiration (ET) and CO2 In this context, the objective of this study was to evaluate the sea-
absorption are quantified in terms of surface conductance (Gs) and its sonal variability of evapotranspiration and its control mechanisms in a
relationship with environmental and biological factors. preserved Caatinga Biome environment during two dry years. We also
The quantification of Gs is important not only for understanding the analyzed the daytime pattern of evapotranspiration and compared the
mechanisms that control ET and CO2 exchange, but also for calibrating sensitivity of surface conductance to VPD.
the Penmann-Monteith equation for future applications under the
conditions in which Gs was calculated. According to Tan et al. (2019), if 2. Material and methods
there is a viable method to reliably obtain Gs values and its environ-
mental controls, then ET can be easily calculated at local and global 2.1. Site description
scales using only usually observed meteorological variables. However,
the parameterization of Gs is challenging, since it is regulated by the The study was conducted in a preserved fragment of the Caatinga
physical environment but also varies between species, especially in Biome, in the Rio Grande do Norte State, Brazilian Semiarid. An 11-
tropical forests (Tan et al., 2019). Usually Gs has been described by the meters tall flux tower equipped with an eddy covariance system was
Penman-Monteith equation in its big-leaf version (Tan et al., 2019) and installed in the Seridó Ecological Station (ESEC-Seridó; 6°34’42”S,
its controls have been analyzed through the relationship between it and 37°15’05”W, 205 m, above sea level), a conservation unit of the
different biophysical parameters, such as VPD, vegetation indices and Caatinga Biome located between the Serra Negra do Norte and Caicó
the ratio between actual and equilibrium evapotranspiration (ET/ETeq). cities (Fig. 1). The ESEC-Seridó area is managed by the Chico Mendes
These analyses have been performed in different biomes around the Institute for Biodiversity Conservation (ICMBio). The flux tower was
world (Ryu et al., 2008; Zha et al., 2013; Ma et al., 2015; Tan et al., acquired by the Brazilian National Institute of Semiarid (INSA) and is
2019). part of the National Observatory of Water and Carbon Dynamics in the
Recently, observational studies in the Caatinga have been developed Caatinga Biome (NOWCDCB) network.
in order to better understand the soil-plant-atmosphere relationship in The ESEC-Seridó comprises an area of 1163 ha of remnant Caatinga,
this biome. It has been observed that the closure of the energy balance characterized by a dry and xerophyte forest composed by deciduous
is better during the wet season and under very unstable conditions, and and semi-deciduous plant species, predominantly sparsely distributed
most of the net radiation is converted into sensible heat flux small trees and shrubs besides herb patches which thrive only during
(Teixeira et al., 2008; Campos et al., 2019). Mutti et al. (2018) showed the wet season (Freitas et al., 2010; Souza et al., 2012; Althoff et al.,
that during a wet year, ET differences between land cover classes were 2016; Tavares-Damasceno et al., 2017). The canopy is around 6 meters
less noticeable due to soil saturation and the urgency of vegetated high. According to Table 1, which shows the relative frequency (RtF)
surfaces to meet their physiological needs. In a dry year, however, the and the Importance Value (IV) of species which occur in the study area
differences were more evident, with bare soil showing lower ET rates (Santana et al., 2016), one can note that the Leguminosae and Eu-
and vegetation classes showing higher ET values in a Semiarid Brazil phorbiaceae families have the highest count of individuals and there is a
watershed predominantly covered by the Caatinga. Da Silva balance between the number of tree and shrub species, although the
et al. (2017) measured the water and energy fluxes over the Caatinga first is dominant (RtF higher than 50%). Shrub species have a RtF of
and found that the ET in the dry season was controlled by the vegeta- around 23%. Among the four dominant species, three are arboreous
tion and in the wet season it was controlled by the atmospheric con- (Caesalpinia pyramidalis Tul., Aspidosperma pyrifolium Mart., and Ana-
ditions. Studies of this nature are important because they allow a better denanthera colubrina (Vell.) Brenan) and one is shrub (Croton blan-
understanding of the effects of projected climate change scenarios for chetianus Baill.). Altogether, the three arboreous species have a total RtF
this biome, which indicate increasing trends in air temperature, higher higher than 40%.
evapotranspiration rates, decreased rainfall and, consequently, ag-
gravation of water deficit (Magrin et al., 2014; Marengo et al., 2017). The region's soil is predominantly Entisol with mainly sandy loam
Given the large area of the Caatinga Biome, one can infer that it and sandy clay loam textures. It is a shallow (approximately 40 cm)
plays an important role in regional (or even global) processes related to rocky soil, with low fertility and moderately drained (Althoff et al.,
the biosphere-atmosphere interactions (Moura et al., 2016). Previous 2016). The region's climate according to the Köppen's climate clas-
studies in this biome have clarified some uncertainties about key sub- sification is Bsh (semiarid low longitude and altitude), with the wet
jects such as the role of seasonal rainfall on the closure and partitioning season established between January and May, mean annual rainfall
of the energy balance, evapotranspiration and CO2 exchange of around 700.0 mm, mean annual temperature of 25.0 °C and mean
(Teixeira et al., 2008; da Silva et al., 2017; Campos et al., 2019; annual air humidity of around 60% (Alvares et al., 2013;
2
T.V. Marques, et al. Agricultural and Forest Meteorology 287 (2020) 107957
Fig. 1. Geographical location of the micrometeorological tower installed in the ESEC-Seridó, Brazil.
Campos et al., 2019). The observation site is mainly flat with slopes canopy in the region). On the other hand, heat plates for the mea-
varying from 1° to 3° and the distance from the flux tower to the surement of G were installed at a 0.05 m depth in the soil. All data were
border of the continuous Caatinga area is higher than 300 m in the stored in a datalogger model CR3000 (Campbell Scientific, Inc., Logan,
prevailing wind direction. UT, USA). A summary of the variables which compose the high fre-
quency and low frequency datasets and their respective measuring in-
2.2. Measurements and data struments is shown in Table 2.
Additionally, we calculate the daily reference evapotranspiration
The experiment was conducted during two years, from January (mm day−1) (ET0), which is a variable that depicts the potential water
2014 until December 2015. Flux tower measurements originated two lost from the surface to the atmosphere and was calculated according to
datasets: i) high frequency data (10 Hz sampling): CO2 and H2O(v) the method described in the FAO56 (Allen et al., 1998):
concentration, the three wind speed components ux, uy and uz, sonic 900
temperature, high frequency temperature and atmospheric pressure; ii) 0.408·Δ·(Rn − G ) + γ· T ·u2·VPD
a + 273
ET0 = ,
low frequency data (5s sampling stored as 30 minutes averages): the Δ + γ ·(1 + 0.34·u2) (1)
four components of the radiation balance, air temperature, relative
humidity in air, soil heat flux density (G). All sensors were installed at where Δ is the slope of the saturation vapor pressure versus air tem-
an 11.0 m height above the soil (4.0 m above the average vegetation perature curve (kPa°C−1), Rn is the surface net radiation (MJ m−2
Table 1
Species, family, common name, life-form, relative frequency (RtF, %) and importance value (IV) sampled in the study area by Santana et al (2016). All scientific
names of species were obtained in The Plant List platform*.
Species Family Common name Life-form RtF IV
⁎
Available on http://www.theplantlist.org/. Accessed in Oct 12, 2018.
3
T.V. Marques, et al. Agricultural and Forest Meteorology 287 (2020) 107957
Table 2
Measured variables, description of the instruments installed in the flux tower of the preserved Caatinga area in the Seridó Ecological Station, Rio Grande do Norte,
Brazil, and their sampling frequency.
Variable Instrument (Manufacturer) Sampling
CO2 and H2O concentrations CO2/H2O Open-Path Gas Analyzer EC150 (Campbell Scientific, Inc., Logan, UT, USA) 10 Hz
ux, uy, uz, sonic temperature 3D Sonic Anemometer model CSAT3 (Campbell Scientific, Inc., Logan, UT, USA) 10 Hz
Atmospheric pressure (hPa) Enhanced Barometer model PTB110 (Vaisala Corporation, Helsink, Finland). 10 Hz
High frequency temperature (°C) HMP155A probe (Vaisala Corporation, Helsinki, Finland) 10 Hz
Net radiation components CNR4 net radiometer (Kipp & Zonen BV, Delft, The Netherlands) 5s
Air temperature (°C) and relative humidity (%) Temperature and Relative Humidity Probe model HMP45C (Vaisala Corporation, Helsink, Finland) 5s
Soil heat flux (W/m2) Soil heat Flux Plates model HFP01SC (Hukseflux Thermal Sensors, Delft, The Netherlands) 5s
day−1), G is the soil heat flux (MJ m−2 day−1), γ is the psychrometric direction according to the equipment user guide (Campbell Scientific
constant (kPa°C−1), Ta is the mean air temperature (°C), u2 is the wind Inc., 2012), there were few data in the shadow zone direction.
speed at 2 m height (m s−1), and VPD is the vapor pressure deficit (kPa) Physically inconsistent values (spikes) caused by changes in the
or the difference between the saturation pressure (es) and partial footprint or imprecise measurements were identified as sudden gaps in
pressure (ea). The methods used to calculate Δ, Rn, G, γ, u2, es and ea are the time series of the 30 minutes energy fluxes averages. For the
described in detail by Allen et al. (1998). identification of other spikes, we used an algorithm based on the
comparison between the difference of each half-hourly value and the
2.3. Data processing range of the moving median daily cycle, as described by
Papale et al. (2006). Regarding data quality, only high and medium
The energy balance (EB) equation depicts the conversion of net quality data were analyzed (flags 1 and 2), while low quality data (flag
energy in energy and mass fluxes between the canopy and the atmo- 3) were discarded.
sphere. Neglecting the energy stored in the canopy and the energy used In this experiment there were two main situations where data were
by plants in the photosynthesis and respiration processes, which re- lacking: i) data gaps caused by spikes identification and removal, but no
present less than 2% of total net radiation (Heilman et al., 1994; gaps in the meteorological dataset; ii) data gaps both in the flux mea-
Allen et al., 1998), the EB equation can be written as: surements and in the meteorological dataset. Turbulent fluxes data gaps
were filled using the method described by Reichstein et al. (2005)
Rn = H + λE + G, (2) where the friction velocity (u*) filtering was estimated according to the
where λE and H are the latent and sensible heat flux densities, re- method described by Papale et al. (2006) with the aid of the online
spectively. All components of the previous equation were analyzed as platform developed by the Max Planck Institute for Biogeochemistry
half-hourly averages in W m−2. (http://www.bgc-jena.mpg.de/∼MDIwork/eddyproc/).
Sensible and latent heat turbulent fluxes were determined using
high frequency data measured by the eddy covariance system based on 2.6. Footprint calculation
the equations:
The flux footprint was calculated using the two-dimensional para-
H = ρair ·cp·w′T ′ and (3) meterization model called Flux Footprint Prediction, as described by
Kljun et al. (2015). This model requires the following data: flux mea-
λE = ρair ·λ·w′q′, (4) surement height (z m = 11 m) , zero-plane displacement (d), surface
where ρair is the air density (1.2 kg m−3), λ is the latent heat of water friction velocity (u*), vertical wind velocity deviation (σw) and rough-
vaporization (2.45 × 106 J kg−1), w′ is the deviation from the mean of ness length (z0). In our study site the canopy height (h) was of 6 m.
vertical wind velocity field (m s−1), q′ is the deviation from the mean of However, this parameterization is valid for moderate friction velocity
specific humidity (kg of water/kg of air), cp is the specific heat of moist values (u* > 0.1 m s−1) and for a limited range of boundary layer sta-
air at constant pressure (1004 J kg−1 °C−1) and T′ is the deviation from bility conditions (−15.5 ≤ z m / L) , where L is the Monin-Obukhov length
the mean of air temperature (K). (Kljun et al. (2015). In our study area, Campos et al. (2019) previously
In order to obtain half-hourly average fluxes, 10 Hz raw data were considered d = (2/3)·h and z 0 = 0.123·h .
processed in the EdiRe software (Rob Clement, University of Edinburgh,
Scotland). Raw data in the TOB3 (table-oriented-binary 3) binary 2.7. Bulk surface parameters
format was read and converted into TOB1 (table-oriented-binary 1)
using the LoggerNet 4.3 software (Campbell Scientific, Inc., Logan, UT, The ability of a given surface to transfer water vapor to the atmo-
USA) in order to allow data processing. The output data file consisted of sphere is what defines Gs. Obtaining Gs and aerodynamic conductance
a spreadsheet with the turbulent variables used in this study indexed as (Ga) is crucial to understand the dynamics and mechanisms involved in
48 half-hourly data per day. mass exchange (ET and CO2) between terrestrial ecosystems and the
The procedures carried out for the processing of data have been atmosphere, as reported by Rocha et al. (2004), Zha et al. (2013), and
thoroughly described in Campos et al. (2019), which used the same Ma et al. (2015). Gs values in the Caatinga Biome were calculated
dataset of the present study. Data processing included the detection of through the inverted Penman–Monteith equation (Stewart, 1988):
spikes, delay correction of H2O / CO2 in relation to vertical wind Ga·γ ·λE
component, coordinates correction (2D rotation) using planar fit Gs = .
Δ·(Rn − G ) + ρair ·cp·Ga·VPD − λE ·(Δ + γ ) (5)
method, correction of spectral loss, sonic virtual temperature correc-
tion, corrections for flux density fluctuations (Webb et al., 1980), as In the present study, the final values of Gs were transformed from
well as the incorporated frequency response correction derived from m s−1 to mm s−1 to agree with other studies. The term Ga is the
Moore (1986) and Massman (2000). We also applied corrections re- aerodynamic conductance which is reciprocal to the aerodynamic re-
garding the reduction of wind velocity or the increase in turbulence sistance to water vapor transport (ra), which in turn varies according to
caused by the shadow of the tower and the sensor. As the sensor was surface roughness and wind speed. The ra parameter comprises the
installed in order to remain pointed towards the predominant wind aerodynamic resistance for momentum (raM) and the quasi-laminar
4
T.V. Marques, et al. Agricultural and Forest Meteorology 287 (2020) 107957
layer resistance (rb), connected in series according to the following indices data as 16-day composites by selecting the best values in this
equation (Thom, 1972; Tan et al., 2019): period according to clear sky conditions and the imaging angle
1 1 1 (Didan, 2015). For the LAI we used the MCD15A3H (v006)—Terra
Ga = GaM + Gb or = + . +Aqua LAI product, which consists of 4-day composites by selecting
ra raM rb (6)
the best values in the period from both satellites (Myneni et al., 2015).
From τ = ρu = ρuGaM , where τ is the vertical flux of horizontal
2
*
momentum, some previous studies such as Szeicz et al. (1969), have 2.9. Statistical analysis
shown that:
The daily means and totals of the meteorological variables, fluxes
GaM = u 2/ u or raM = u/ u 2, (7)
* * and ET control mechanisms were bootstrapped over seasonal intervals
where u is the wind speed measured at a height z, and u* is the friction for the estimation of random variance ( ± 95% of confidence interval –
velocity, both measured in m s−1. According to Tan et al. (2019), Eq. CI) about the mean according to the methodology presented by
(7) provides the best and most reliable GaM estimates over tropical Zanella de Arruda et al. (2016). Statistically significant differences
forests. (p < 0.05) in the mean seasonal value for a given meteorological
According to Monteith and Unsworth (2013) the values of rb are variable, flux or ET control mechanism were determined by the degree
rarely determined directly. However, they can be estimated from values of overlap in the 95% bootstrapped CI (Zanella de Arruda et al., 2016).
of parameters observed in field experiments. Also according to Linear and non-linear regression models were used to evaluate the re-
Monteith and Unsworth (2013), Thom's empirical equation (Eq. (8)) is lationship between two given variables. Additionally, the model coef-
an adequate approximation of rb, at least for u* values typically in the ficients were tested under the null hypothesis at a 5% significance level.
range between 0.1 and 0.5 m s−1, which is the case of observational All statistical analysis was carried out using the R software (R Core
values in our experimental site (Campos et al., 2019): Team, 2018).
where Ω ranges from 0 to 1. According to Baldocchi and Xu (2007) and Annual rainfall in 2014 and 2015 was of 513 mm and 466 mm,
Ma et al. (2015), when Ω → 1 the surface is completely decoupled from respectively while the 30-year average in the region is of 758 mm. Daily
overhead conditions. In this case, ET is mainly controlled by available rainfall rates in 2014 and 2015, as shown in Fig. 2a, highlight the
energy. On the other hand, when Ω → 0, ET is mainly controlled by marked rainfall seasonal variability in the study area. In the course of
bulk surface conductance, indicating a strong influence of the vegeta- the experiment years, roughly 82% of total rainfall was registered
tion in the control of ET and the atmospheric humidity deficit. during the wet season (Table 3) and in 2014 rainfall was more homo-
In order to determine whether atmospheric demand or surface geneously distributed, with 71 rain days (Fig. 2a) and a daily maximum
moisture supply was the limiting factor, we calculated the ratio be- of 45.4 mm. In 2015, on the other hand, only 34 rain days were re-
tween ET and equilibrium evapotranspiration (ETeq). The daytime mean gistered (Fig 2a.) with two events (65 mm and 60 mm) above the heavy
(08:00–17:00 local time) was obtained through an adapted version of rainfall events threshold, which is of 50 mm in the study area
the λEeq equation proposed by Priestley and Taylor (1972): (Oliveira et al., 2017).
ET Δ+γ λE Besides rainfall, Fig. 2a also shows daily ET0 values, which were
= · .
ETeq Δ H + λE (11) predominantly above 6 mm with a daily average of 8.0 mm day−1 in
2014 and 8.3 mm day−1 in 2015 (p < 0.05; Table 3). Fig 2b–d shows
Using the ET/ETeq ratio to compare different observational sites is daily global radiation (Rg in MJ m−2 ), Ta, soil temperature (Ts in °C) and
reasonable since they are normalized by the equilibrium rates mainly VPD values, respectively, in which a clear seasonal variation can be
determined by Rn (Wilson and Baldocchi, 2000). observed. In 2014, mean Rg was of 22.0 MJ m−2 day−1 (Table 3),
ranging from 11.4 to 27.6 MJ m−2 day−1 with maximum values oc-
2.8. Vegetation condition curring mainly during the dry-wet transition season, while minimum
values were observed in the wet-dry transition season. In 2015, Rg had
In order to assess the behavior of the vegetation during the ex- the same pattern observed in 2014, with maximum and minimum va-
periment we used remote sensing data obtained through the Moderate lues observed in the same periods and a mean value of 22.6 MJ m−2
Resolution Imaging Spectroradiometer (MODIS) sensor aboard the day−1 ranging from 11.0 to 27.5 MJ m−2 day−1 (Table 3 and Fig. 2b).
Terra and Aqua satellites and made available by the United States The seasonality of Ta (and Ts) was consistent with Rg, with a mean
Geological Survey (USGS) (https://earthexplorer.usgs.gov). We used value of 28.9 °C (31.4 °C) in 2014 (Table 3) ranging from 24.7 °C
both Normalized Difference Vegetation Index (NDVI) and Leaf Area (27.4 °C) in the wet season to 32.2 °C (37.0 °C) in the dry-wet season
Index (LAI) data. The NDVI was retrieved from the MOD13A2 (Fig. 2c), while in 2015 the mean value was of 29.5 °C (32.9 °C), ran-
(v006)—Terra Vegetation Indices product, which provides vegetation ging from 25.5 °C (27.3 °C) in the wet-dry season to 32.4 °C (38.5 °C) in
5
T.V. Marques, et al. Agricultural and Forest Meteorology 287 (2020) 107957
Fig. 2. Accumulated rainfall and daily means for (a) reference evapotranspiration (ET0), (b) global incident radiation (Rg), (c) air and soil temperature (Ta and Ts) and
(d) vapor pressure deficit (VPD) during the studied period in the Caatinga (ESEC-Seridó). Blue and red shades represent wet and dry periods, respectively.
the dry-wet season (Table 3 and Fig. 2c). The VPD mean value was 1.7 reported by previous studies on the climatology of the region
kPa in 2014 and 1.9 kPa in 2015 (Table 3). The highest daily VPD (Santos and Santos e Silva, 2013; Gilliland and Klein, 2018).
means were registered in the dry season of 2014 (2.7 kPa) and 2015
(3.0 kPa) while the minimum means (0.2 kPa and 0.3 kPa) were re- 3.4. Diurnal and seasonal changes
gistered in the wet seasons of 2014 and 2015 (Table 3 and Fig. 2d).
Overall, the variables presented higher annual means if compared to Fig. 4 shows the mean diurnal cycle of ET and its controlling factors
the 30-year average (21.6 MJ m−2 day−1, 26.8 °C and 1.3 kPa, re- (Gs and Ω) in 2014 (left panel) and 2015 (right panel). According to
spectively for Rg, Ta and VPD). Fig. 4a and b, Gs increased sharply from sunrise (06:00 local time) until
it reached its maximum values between 08:00 and 10:00 local time.
3.3. Footprint and wind direction From this time on, Gs values decreased almost linearly until they've
become stationary during the night (0.01 mm s−1) in both years and in
The footprint calculated from the location of the eddy covariance all seasons, except during the dry season when its values were almost
tower ranged from 150 meters (wet season) to 200 meters (dry season). constant during daytime. The diurnal pattern of ET (Fig. 4c–d) was
One should notice that the distance from the eddy covariance tower to relatively consistent with the Gs curve but with higher values between
the border of the conservation unit is larger than 300 m, so the footprint 10:00 and 14:00 local time.
calculated is adequate for this study (Campos et al., 2019). The pre- One can notice in Fig. 4e–f that the behavior of the diurnal cycle of
dominant local wind direction during the study period was southeast Ω was similar during all seasons and in both years; after sunrise (about
(Fig. 3) due to the influence of the South Atlantic Subtropical High as 06:00 local time) there was a sharp increase until it reached its
Table 3
Daily mean ± standard deviation for global radiation (Rg, MJ m−2 day−1), vapor pressure deficit (VPD, kPa), air temperature (Ta, °C), soil temperature (Ts, °C), and
reference evapotranspiration (ET0, mm); seasonal and annual rainfall (Rainfall, mm) and reference evapotranspiration (Total ET0, mm) totals.
Variable Dry-wet transition Wet season Wet-dry transition Dry season Anual
2014 2015 2014 2015 2014 2015 2014 2015 2014 2015
Rg 23.5 ± 3.1 24.3 ± 2.9 22.0 ± 3.8 22.5 ± 3.2 18.0 ± 3.4 18.5 ± 3.3 23.0 ± 3.3 24.8 ± 2.2 22.0 ± 3.9 22.6 ± 3.8
VPD 2.1 ± 0.4 2.3 ± 0.4 1.1 ± 0.5 1.6 ± 0.6 1.5 ± 0.3 1.6 ± 0.4 2.1 ± 0.3 2.1 ± 0.4 1.7 ± 0.6 1.9 ± 0.6
Ta 30.1 ± 0.9 30.6 ± 0.8 28.8 ± 1.3 29.3 ± 1.1 27.5 ± 1.0 28.6 ± 1.4 28.9 ± 1.1 29.6 ± 1.3 28.9 ± 1.4 29.5 ± 1.4
Ts 32.8 ± 1.9 35.1 ± 2.3 30.1 ± 2.0 31.4 ± 1.9 29.0 ± 0.7 30.3 ± 2.2 33.2 ± 1.7 34.8 ± 2.7 31.4 ± 2.4 32.9 ± 3.1
Rainfall 65.6 28.5 419.6 381.5 21.6 56.0 6.2 0.0 513.0 466.0
ET0 8.6 ± 0.6 8.9 ± 0.7 7.6 ± 0.7 7.8 ± 0.6 7.2 ± 0.4 7.9 ± 0.6 8.4 ± 0.6 8.6 ± 0.7 8.0 ± 0.8 8.3 ± 0.8
Total ET0 788.1 551.6 910.8 689.8 438.6 730.2 771.6 1043.7 2909.1 3015.3
6
T.V. Marques, et al. Agricultural and Forest Meteorology 287 (2020) 107957
and the effect of the forest floor ET in periods of leaf loss. As expected,
maximum evapotranspiration rates were registered during the wet
season, while minimum rates were observed during the dry season.
Peak evapotranspiration in the wet season in 2014 (4.35 mm day−1)
was higher than in 2015 (3.82 mm day−1). On the other hand, most ET
values during the dry season were < 0.1 mm day−1 in 2014 and >
0.5 mm day−1 in 2015. Annual accumulated ET during 2014 was of
473.3 mm while in 2015 it was 283.4 mm (Table 4). The ET/ETeq ratio
mean value was of 0.39 ranging from 0.01 to 1.26 in 2014 and of 0.24
ranging from 0.01 to 1.07 in 2015. Regarding seasonal variability, the
highest and lowest ET/ETeq values were observed, respectively, in the
wet and dry seasons in both years. ET/ETeq > 1 probably occurred due
to moisture advection during the wet season towards the continent
(tower flux site), associated with the confluence of northeasterly and
southeasterly trade winds which are part of the Intertropical Con-
vergence Zone (ITCZ). This confluence is highly associated with the
southernmost positioning of the ITCZ during this period, which mod-
ulates the wet season over the region (Hastenrath, 2006).
The decoupling coefficient (Ω) was calculated in order to determine
the relative importance of Gs to changes in annual, seasonal and daily
evapotranspiration rates. In the dry and transition seasons, Ω diurnal
cycle values were below 0.1 with minimums during the night, sug-
gesting a greater stomatal control in water loss by plants throughout the
day (Fig. 4e–f). Estimates of Ω ranged from 0.01 (dry season) to 0.54
(wet season) with a 0.14 mean in 2014 and from 0.01 (dry season) to
0.48 (wet season) with a 0.07 mean in 2015. Thus, as expected, higher
Ω values were registered during the wet season while lower values were
observed during the dry season (Fig. 4c and Table 4), indicating that
surface conductance increasingly controlled ET with the persistence of
dry conditions.
Fig. 3. Wind direction and cumulative frequency during the experiment years 3.5. Physiological control via surface conductance
in the study area. Colors represent wind speed classes.
The ratio between actual evapotranspiration and equilibrium eva-
maximum daily values around 08:00–9:00 local time. From this time potranspiration (ET/ETeq) and Gs in each season of 2014 and 2015 in
on, it decreased almost linearly until sunset (about 18:00 local time) the Caatinga are shown in Fig. 6. It is possible to observe that there is a
and it remained fairly constant during the night. The highest daily non-linear relationship between ET/ETeq and Gs, with the highest values
variability was observed during the wet season, from 0.1 to almost 0.4 during the wet season. The equilibrium ET/ETeq ≈ 1 is reached when
in 2014 and from 0.1 to 0.3 in 2015. In the transition seasons the Gs ≈ 8 mms−1, which happens only occasionally in both years, but with
maximum daily value was slightly larger than 0.1 and during the dry less frequency during 2015, which was a drier year.
season Ω was almost constant and close to 0.0, indicating a situation of Overall, Gs decreases exponentially with the increase in VPD
permanent decoupling or water deficit. (Fig. 7). In the dry season, a VPD > 1.5 kPa resulted in close to zero Gs
By observing the whole experimental period, one can notice clear values. However, the effects of VPD on Gs were clearer in 2014 than in
seasonal variations in NDVI, Gs, ET, ET/ETeq and Ω daily values (Fig. 5). 2015 (R2 = 0.63 and 0.19, respectively). Fig. 8 shows the relationship
Changes in the Caatinga phenology were evaluated using the LAI and between Gs (08:00–17:00 local time) and the NDVI and LAI. In 2014, Gs
NDVI (Fig. 5a). In the wet season, LAI (NDVI) peaked at 3.3 m2m−2 increased exponentially with the increase in the NDVI (R2 = 0.97 ;
(0.76) in 2014 and 2.5 m2m−2 (0.73) in 2015, gradually decreasing in p < 0.01) while in 2015 the relationship was weaker (R2 = 0.52 ;
response to water deficit and the senescence of vegetation, which is p < 0.05) and could be represented by a first degree polynomial
composed by deciduous and semi-deciduous plant species. Results show (Fig. 8a–b). Regarding the LAI, its relationship with Gs in 2014 is also
that in the dry season the mean LAI decreased as much as 73% in 2014 exponential (R2 = 0.52 ; p < 0.01) while in 2015 it can be represented
and 57% in 2015 (Table 4). Considering annual means, both the LAI by a first degree polynomial (R2 = 0.32 ; p < 0.05) (Fig. 8c–d). Table 5
and NDVI were higher in 2014 (1.26 m2m−2 and 0.49) than in 2015 shows rainfall, total evapotranspiration, daily mean λE, mean surface
(1.02 m2m−2 and 0.44) (Table 4). conductance, daily mean Ω and ET/ETeq ratio observed in our study and
The seasonal patterns of Gs daily means are shown in Fig. 5b. The in different ecosystems with similar climate characteristics.
lowest Gs values were found during the dry season with 95% of them
below 0.4 mm s−1. In the wet-dry transition season Gs values gradually 4. Discussion
decreased as the environment became drier while the opposite was
observed during the dry-wet transition season. The Gs parameter was 4.1. Meteorological aspects
higher in the wet season, when the LAI reached its maximum values.
Thus, increased rainfall and soil moisture contributed to higher Gs va- The annual analysis of the VPD was consistent when comparing the
lues, which peaked at 8.65 mm s−1 in 2014 and 7.30 mm s−1 in 2015 results of 2014 and 2015, that is, the year with less rainfall volume
(Fig. 5b). In 2014, Gs values were 52% (wet season) and 23% (dry (2015) was warmer and drier, which resulted in higher VPD values.
season) higher than in 2015. Seasonal variations were also similar to the values reported in Brazilian
ET fluxes (Fig. 5c) presented a seasonal variation similar to that of tropical savannas (Rodrigues et al., 2014) and deciduous tropical for-
the NDVI and Gs because of the seasonal behavior of leaf physiology ests in northern Thailand (Igarash et al., 2015), which also present
vegetation susceptible to extended drought periods. On the other hand,
7
T.V. Marques, et al. Agricultural and Forest Meteorology 287 (2020) 107957
Fig. 4. Diurnal variations in evapotranspiration (ET), surface conductance (Gs) and decoupling coefficient (Ω) by season during the study period in the Caatinga
(ESEC-Seridó).
VPD values found in this study were higher than the ones found in Additionally, the authors of said study also reported that the morpho-
tropical forest-transitions (Vourlitis et al., 2008) and tropical rainforests physiological response of the C. blanchetianus is modulated mainly by
(da Rocha et al., 2004), because in these biomes annual rainfall can water availability.
reach up to 2,000 mm and the seasonal variability of relative humidity The pattern of the diurnal cycle of Gs was consistent in each studied
in air is not significant. Values presented in Table 3 were also higher year and season, but the range between maximum and minimum half-
than the ones reported by da Silva et al. (2017) during the 2014–2015 hourly values was higher in the wet season (Fig. 4a–b). The higher Gs
drought period, but in a Caatinga region with different characteristics values found during the morning agree with other studies that observed
from the present study, which may influence evapotranspiration rates maximum surface and canopy conductance between 8:00 and 12:00
and therefore relative humidity in air. and minimum values during sunset in different environments such as
Daily mean ET0 values greater than 6 mm (Fig. 2a) are consistent tropical Brazilian savannas (Giambelluca et al., 2009;
with other databases over the region (such as Xavier et al., 2015, for Rodrigues et al. 2014, 2016) and tropical semi-deciduous forest in the
example) and highlights the high atmospheric demand for water in the southern Amazon Basin (Vourlitis et al., 2008). Furthermore, the
studied region. Annual mean values of 8.0 mm day−1 in 2014 and Spearman's correlation coefficient (ρ) indicated a decrease in Gs after
8.3 mm day−1 in 2015 (p < 0.05; Table 3) also reaffirm a more intense noon which may be explained by the increase in air temperature
drought in 2015 than in 2014. ( ρ = −0.67 and − 0.59 in the wet season of 2014 and 2015, respectively;
ρ = −0.28 and − 0.37 in the dry season of 2014 and 2015, respectively),
4.2. Diurnal and seasonal changes decrease in relative humidity of air ( ρ = 0.55 and 0.62 in the wet season
of 2014 and 2015, respectively; ρ = 0.45 and 0.25 in the dry season of
LAI was strongly influenced by rainfall distribution and, conse- 2014 and 2015, respectively) and higher VPD ( ρ = −0.51 and − 0.62 in
quently, soil moisture conditions during the experiment period, which the wet season of 2014 and 2015, respectively; ρ = −0.40 and − 0.32 in
is crucial to leaf coverage seasonality in the Caatinga. The marked the dry season of 2014 and 2015, respectively). Stomatal closure during
seasonality of rainfall and soil moisture in the Caatinga affects phe- the higher evaporative demand periods is a mechanism of adjustment to
nology, stomatal conductance seasonality and photosynthesis stress conditions in order to maintain plant available water
(Santos et al., 2012; Mendes et al., 2017). The study by (Flexas et al., 2004; Pinho-Pessoa et al., 2018). There are a number of
Mutti et al. (2019) reports that the phenological response of the Caa- plant-scale studies on the stomatal closure of Caatinga tree species
tinga (in the same study site), or its vegetation dynamic, is closely re- when leaves are exposed to high VPD at low water availability condi-
lated to rainfall seasonality, which modulates the physiological and tions (Santos et al., 2014; Mendes et al., 2017; Pinho-Pessoa et al.,
metabolic characteristics, and the LAI of plants in this region. Fur- 2018). Such studies reported that stomatal conductance varied during
thermore, Mendes et al. (2017) showed that C. blanchetianus popula- the diurnal cycle at ambient temperature and humidity, with higher
tions, an endemic semi-deciduous species, have high adaptive capacity values between 8:00 and 11:00.
to environmental changes, mainly due to the stomatal closure. In the present study, seasonal increases and decreases in Gs were
8
T.V. Marques, et al. Agricultural and Forest Meteorology 287 (2020) 107957
related to leaf growth and loss, respectively. Thus, a reduction in Gs mainly because even during the dry season there is still some tree cover
values was associated with the leaf senescence mechanism as a response due to the presence of semi-deciduous species. However, a diurnal lag
to water scarcity, avoiding the excessive loss of water by plants and was observed between Gs and ET which can be attributed to the esti-
reducing ET rates. However, Gs peaks at the beginning of the wet season mation of Gs itself. That is, in the early hours of morning stomata are
with the expansion of leaf cover. This indicates that Gs increases might already open (increasing Gs values), but ET is low due to low VPD va-
have been affected by the morphophysiological state at the leaf scale. lues. Throughout the day, VPD increases and ET increases with it, which
By analyzing Fig. 4b one can notice that in the wet season of 2015 explains why diurnal ET peaks a couple of hours after Gs.
(heterogeneous rainfall distribution), Gs behavior varied constantly and At the beginning of the wet season a sharp linear increase in ET (and
25% of the values were under 0.97 mm s−1, which means stomata were Gs) was noticed when leaf cover started to grow as soon as soil moisture
partially closed. The lag between environmental factors and soil water was reestablished. The decrease in ET values at the end of the wet
storage may be responsible for this oscillatory pattern in Gs, although season and during the wet-dry transition occurred mainly due to the
explaining the causes of this phenomenon relies on complex interac- reduction in stomatal opening as a defense mechanism against water
tions between environmental physical factors and the modulation of stress, since diffusive resistance to water vapor reduces transpiration
stomatal opening (Zhang et al., 2014; Bai et al., 2017). These Gs var- (Flexas et al., 2006). It is important to highlight that photosynthesis
iations in time may be related to the effects of soil water availability on decreases with the closing of stomata due to the reduced CO2 absorp-
stomatal opening and/or photosynthetic capacity. tion (Flexas and Medrano, 2002), which leads to leaf loss (senescence)
ET fluxes seasonal behavior was perfectly coupled to Gs due to in the Caatinga under water stress conditions. Thus, low ET values
seasonal variations in leaf physiology and the effect of the forest floor during the dry season can be related to the decrease in NDVI and
ET in periods of leaf loss. It is important to notice that it is difficult to LAI—which represents leaf senescence in the Caatinga—as a me-
individually analyze the effect of forest floor ET and leaf physiology, chanism of resilience to drought, to low rainfall periods and to low
Table 4
Seasonal and annual total evapotranspiration in the Caatinga (total ET, mm) and daily means ± standard deviation for the evapotranspiration (ET, mm), surface
conductance (Gs, mm s−1), decoupling coefficient (Ω), ratio between ET and equilibrium evapotranspiration (ET/ETeq); NDVI and Leaf Area Index (LAI, m2 m−2 ).
Variable Dry-wet transition Wet season Wet-dry transition Dry season Annual
2014 2015 2014 2015 2014 2015 2014 2015 2014 2015
Total ET 87.7 38.6 307.4 152.6 53.6 72.7 24.6 19.5 473.3 283.4
ET 1.0 ± 0.7 0.6 ± 0.3 2.6 ± 1.1 1.7 ± 0.9 0.9 ± 0.4 0.8 ± 0.6 0.3 ± 0.2 0.2 ± 0.1 1.3 ± 1.2 0.8 ± 0.8
Gs 0.85 ± 0.83 0.50 ± 0.36 3.74 ± 2.19 2.13 ± 1.49 1.00 ± 0.65 0.87 ± 0.67 0.22 ± 0.21 0.13 ± 0.13 1.66 ± 2.00 0.87 ± 1.13
Ω 0.08 ± 0.07 0.04 ± 0.03 0.29 ± 0.14 0.17 ± 0.11 0.09 ± 0.06 0.08 ± 0.06 0.02 ± 0.02 0.01 ± 0.01 0.14 ± 0.15 0.07 ± 0.09
ET/ETeq 0.27 ± 0.21 0.18 ± 0.15 0.72 ± 0.30 0.50 ± 0.27 0.37 ± 0.19 0.30 ± 0.19 0.09 ± 0.06 0.05 ± 0.04 0.39 ± 0.33 0.24 ± 0.25
NDVI 0.46 ± 0.13 0.35 ± 0.08 0.65 ± 0.11 0.63 ± 0.11 0.46 ± 0.08 0.50 ± 0.08 0.31 ± 0.03 0.29 ± 0.04 0.49 ± 0.16 0.44 ± 0.16
LAI 1.21 ± 0.62 0.78 ± 0.23 1.94 ± 0.63 1.57 ± 0.62 1.07 ± 0.43 1.09 ± 0.47 0.53 ± 0.16 0.68 ± 0.12 1.26 ± 0.74 1.02 ± 0.53
9
T.V. Marques, et al. Agricultural and Forest Meteorology 287 (2020) 107957
Fig. 6. Relationship between daily mean conductance (Gs) and the ratio between actual evapotranspiration and equilibrium evapotranspiration (ET/ETeq) during
2014 (a) and 2015 (b) in the Caatinga (ESEC-Seridó)
water storage capacity of the shallow soils in the study region. Besides forests, our results agree with the behavior observed by Ma et al. (2015)
senescence of the leaf tissue, Gs drastically reduced in the few re- in the semiarid environment of the alpine steppes of the Tibetan Pla-
maining leaves of semi-deciduous species because of soil moisture teau, although water restrictions in this region are related to the periods
deficits, which incurred in lower ET values. These results indicate dif- of frozen soil.
ferent plant-environment interactions which affected ET in both study Another characteristic of the Caatinga Biome which could be ob-
years, highlighting the role of Gs in controlling evapotranspiration rates served was its response to rainfall pulses with sudden increases in ET. A
in the Caatinga according to rainfall distribution and tree cover few days after the rainfall event ET rates exponentially decrease until
changes. they reach the rate observed before the event. Remarkable increases in
Although 2015 was a drier year (higher atmospheric demand for ET after rainfall pulses were also observed by Eamus et al. (2013)
water), ET values were lower, evidencing the highly efficient resilience during a two-year experiment in Mulga (an Australian SDTF) and by
mechanism of the Caatinga Biome species, which has been previously da Silva et al. (2017) in 365 days of study in a preserved Caatinga area.
discussed and consists of stomatal closure during dry periods, reducing Furthermore, ET interannual patterns in the Mulga were similar to
water lost by transpiration. This resilience can be better explained by those of the present study, with daily maxima of around 3 mm day−1 in
the seasonal variations in Ω diurnal values, which will be discussed in a year with below-average and poorly distributed rainfall and daily
the following sections. At the same time, the phenology of the biome maximums of around 4 mm day−1 in a year with near-average homo-
promptly responds to the occurrence of rainfall, as shown by the NDVI geneously distributed rainfall. In the study conducted by da Silva
and LAI seasonal variability (Fig. 5a), which gradually increases in the et al. (2017) the wet season lasted for six months (88% of annual
transition from the dry season to the wet season. This behavior shows rainfall) and although total annual rainfall (430.2 mm year-1) was
that during the wet season, vegetation (mainly deciduous species) in- lower than in the present study, total annual ET (526.6 mm year-1) and
tensifies its phenological and metabolic activities (Barbosa et al., 2006; maximum daily ET (5.4 mm year-1) were higher. These differences can
Dombroski et al., 2011; Barbosa and Kumar, 2016; da Silva et al., probably be explained by the difference in total annual rainfall between
2017). years.
Regarding ET behavior in the Caatinga, our results were consistent
with previous studies in the semiarid region of Brazil. For example, 4.3. Analysis on the variability of the decoupling factor and ET/ETeq
da Silva et al. (2017) retrieved evapotranspiration in the Caatinga
during 2014 and 2015 and observed higher values during the wet The Caatinga canopy was more decoupled from the atmosphere in
season and lower values during the dry season. In comparison to other the first hours of the morning, represented by high Ω values (Fig. 4e–f)
Fig. 7. Relationship between daily mean conductance (Gs) and the vapor pressure deficit (VPD) during 2014 (a) and 2015 (b) in the Caatinga (ESEC-Seridó)
10
T.V. Marques, et al. Agricultural and Forest Meteorology 287 (2020) 107957
Fig. 8. Relationship between 16-day mean surface conductance (Gs) and the NDVI during 2014 (a) and 2015 (b), and between 4-day mean Gs and the LAI during
2014 (c) and 2015 (d) in the Caatinga (ESEC-Seridó)
and suggesting that ET was controlled mainly by available energy. annual ET in the Caatinga was not limited by available energy, but by
These high Ω values in the beginning of the morning were a con- water scarcity associated with droughts. Table 5 shows a summary of
sequence of high Gs, low VPD and wind speed. Throughout the day, an rainfall and ET parameters in different ecosystems and in the present
accentuated decrease in Ω can be observed, indicating an increasing study. In environments with water restrictions such as the Caatinga and
atmosphere-canopy coupling. The explanation for the decline in Ω was arid drylands, ET is controlled by water available in the soil, which is
the combination of the gradual decrease in Gs after noon, high VPD, the main limiting factor.
high energy availability and relatively strong winds, which usually
occur during the afternoon (McNaughton and Jarvis, 1983). 4.4. Physiological control via surface conductance
The seasonal pattern of Ω was similar to that of Gs. Low Ω values
indicated a strong control of evapotranspiration by surface con- The non-linear relationship between ET/ETeq and Gs is consistent
ductance, particularly in the transition (wet-dry and dry-wet) and dry with studies carried out in different vegetated surfaces such as tempe-
seasons, when Ω was lower than 0.1 (Table 4). Low Ω values are also rate deciduous forest (Wilson and Baldocchi, 2000), alpine meadows in
associated with the seasonal behavior of rainfall and, consequently, the Tibetan Plateau (Gu et al., 2008), grasslands in the United States
with lower water availability in the soil and increasing VPD (Table 3). Ω (Krishnan et al., 2012), corn and wheat croplands in China (Lei and
values found in this study are different from the ones found by da Silva Yang, 2010) and alpine steppe in the Tibetan Plateau (Ma et al., 2015).
et al. (2017) in the Caatinga (average value was 0.4), although rainfall In both years a negative physiological feedback in the canopy was
was better distributed in that study. Wever et al. (2002) also observed a observed, which agrees with the classical analysis by Jarvis and
decline in Ω with the reduction of soil moisture and increasing VPD in McNaughton (1986). Similar results were obtained in pastures by
northern temperate grasslands during the growing season. Regarding Ω Wever et al. (2002), Li et al. (2005), Zhang et al. (2007),
values in the wet season, the averages of 0.29 in 2014 and 0.17 in 2015 Aires et al. (2008), Ryu et al. (2008), Krishnan et al. (2012) and in
indicate that ET in the Caatinga was partially decoupled from the at- forests by Granier and Bréda (1996), Zha et al. (2013) and Zhu el al.
mosphere. Considering Ω variability, one can observe that during the (2014). Because ET and Gs are strongly correlated (r ≥ 0.89 in all study
dry seasons, ET was completely controlled by the vegetation, which period), these results are also consistent with the studies by
partially agrees with da Silva et al. (2017). Given that solar radiation is Teixeira et al. (2008) and da Silva et al. (2017), since high ET values
not a limiting factor for the development of the Caatinga vegetation, the can be explained by low VPD which indicates an increase in the opening
decoupling is due to below-average and poorly distributed rainfall, of stomata as a response to the water vapor transfer gradient. Fig. 7
especially in 2015. shows that, in face of similar energy conditions, the regression curve
Low ET/ETeq and Ω mean annual values in both years indicate that between Gs and VPD was better fitted in 2014 (R2 = 0.63) when
11
T.V. Marques, et al. Agricultural and Forest Meteorology 287 (2020) 107957
Comparison between accumulated rainfall (P), actual evapotranspiration (ET), daily mean latent heat flux (λE), daily mean surface conductance (Gs), daily mean decoupling coefficient (Ω), daily mean ratio between ET compared to 2015 (R2 = 0.19) . In the dry season of both years VPD did
In this study
In this study
logical and metabolic activities.
Previous studies frequently focused on the diurnal responses of
Source
2014 when 71 rain days were registered. On the other hand, in 2015
(34 rain days) no relation between data was found (Fig. 7b). Therefore,
Gs (mm s−1)
9.8
3.5
4.1
New results were found in the present study. These different factors
-
(Fig. 8b–d) they were linearly adjusted. This result showed that even in
scenarios with limited water availability, the different temporal dis-
473.3
283.4
808.5
347.8
358.2
526.6
1914
3366.0
513.0
466.0
394.7
337.6
430.2
and 2015 (22.0 W m−2) are similar to those obtained in other studies
March 1–February 28, 2014–2015
January 1–April 30, 2002–2007a
2017). However, Gs, Ω, and ET/ETeq values found in this study are
considerably lower than values found in the aforementioned studies.
These differences are possibly associated with the extreme and persis-
Period
tent drought condition observed during the experiment, which did not
occur in said studies.
Arid and semiarid biomes cover approximately one third of the
Values observed in more than one study year.
pastures.
Gu et al. (2005) observed seasonal patterns in an alpine meadow
grassland and reported that vegetation phenology and soil water con-
tent were the main factors which affected energy partitioning in that
Table 5
12
T.V. Marques, et al. Agricultural and Forest Meteorology 287 (2020) 107957
during frozen soil and transition periods, Gs values were considerably probable cause for the retrieval of higher ET rates in 2014 (473.3 mm)
lower (0.91 and 2.45 mm s−1, respectively), suggesting that ET was than in 2015 (283.4 mm), which incurred in lower Gs, decoupling
mainly controlled by soil moisture. In contrast, with the increase in coefficient (Ω) and ET/ETeq ratio in 2015 compared to 2014. Leaf se-
water availability during the wet season, ET behavior was coupled with nescence and the drastic reduction in surface conductance during the
available energy while at the end of this season ET gradually decreased dry season allows the Caatinga Biome trees to optimize water use in
in response to water scarcity and plants senescence. Ryu et al. (2008) order to avoid or attenuate water stress, becoming more resilient to the
investigated ET interannual variability and energy exchanges in Medi- reduction of water availability in the soil. Major controlling factors for
terranean annual grassland, observing that ET was most sensitive to the hydrological fluxes in this ecosystem need more attention in future
availability of soil moisture during the transition to the senescence researches, particularly on transpiration via sap flow monitoring.
period rather than at the onset of the greenness period, causing annual
evapotranspiration to be strongly modulated by growing season length. Declaration of Competing Interest
In the Caatinga biome, da Silva et al. (2017) studied CO2, water and
energy fluxes and observed that in 2014 the daily mean Ω value was The authors declare that they have no known competing financial
28% larger than the value found in the present study and ET was also interests or personal relationships that could have appeared to influ-
larger overall. However, they did not present Gs values in order to in- ence the work reported in this paper.
vestigate the biophysical controls of evapotranspiration. It should be
noted that in environments with water scarcity, such as the Caatinga, Acknowledgments
ET is controlled by the availability of soil moisture while in humid and
sub-humid tropical ecosystems available energy plays a crucial role in The authors are thankful to the Brazilian National Institute of Semi-
ET control, given that water is not a limiting factor. Studies using the Arid (INSA) and CNPq/Capes/FACEPE (Process no 465764/2014-2 and
eddy covariance method are important to better understand the beha- 420854/2018-5) for funding the project which originated the eddy
vior of stomatal modulation and its influence on the regulation of water covariance data used in this study which is partially based on the Ph.D.
fluxes in response to extreme environmental conditions, as suggested by Thesis by Thiago Valentim Marques carried out in the Climate Sciences
Ghimire et al. (2014) and corroborated by the findings of the present Graduate Program of the Federal University of Rio Grande do Norte. We
study in the Caatinga Biome. are also thankful to the CNPq for funding the Research Productivity
Grant to the eighth and seventeenth authors (Processes no 309165/
5. Conclusions 2010-5, and 303061/2014-6, respectively) and to the ICMBio (Chico
Mendes Institute for Biodiversity Conservation) for providing the ex-
A seasonal analysis of the biophysical control and characteristics of perimental area and to the ESEC-Seridó (Ecological Station of Seridó)
ET was carried out in the Caatinga Biome with a detail level that is for supporting experimental activities. The authors would also like to
unprecedent for this biome. The physiological controls at the ecosystem honor Dr. Ignacio Hernán Salcedo (in memorian), Emeritus Professor at
level were based on surface conductance (Gs) and its relationships with the Federal University of Pernambuco, former President of the INSA
vegetation and atmospheric parameters (NDVI, LAI and VPD). (2011-2015) and one of the creators of the NOWCDCB network.
Additionally, we analyzed the decoupling coefficient (Ω), and the ratio
between actual evapotranspiration and equilibrium evapotranspiration References
(ET/ETeq). The study was conducted with data measured in a flux tower
during two dry years (2014 and 2015). The present study showed that Aires, L.M., Pio, C.A., Pereira, J.S., 2008. The effect of drought on energy and water
the lowest ET values in both years occurred in the dry season, as a vapour exchange above a Mediterranean C3/C4 grassland in Southern Portugal.
Agric. For. Meteorol. 148 (4), 565–579. https://doi.org/10.1016/j.agrformet.2007.
consequence of leaf loss (low LAI values) or the partial closing of the 11.001.
stomata of the few evergreen species caused by water and rainfall Allen, R.G., Pereira, L.S., Raes, D., Smith, M., 1998. Crop Evapotranspiration – Guidelines
shortage during this period. The highest ET values were registered in for Computing Crop Water Requirements - FAO Irrigation and Drainage Paper 56.
Rome Italy. http://www.fao.org/docrep/X0490E/x0490e00.htm.
the wet season, when Caatinga species physiology and metabolism are Althoff, T.D., Menezes, R.S.C., Carvalho, A.L., Pinto, A.S., Santiago, G.A.C.F., Ometto,
fully active. J.P.H.B., Randow, S.V., Sampaio, E.V.S.B., 2016. Climate change impacts on the
Gs peaked at approximately 09:00 (local time), before VPD reached sustainability of the firewood harvest and vegetation and soil carbon stocks in a
tropical dry forest in Santa Teresinha Municipality. Northeast Brazil. Forest Ecol.
its maximum value. This lag between the time when Gs reaches its Manag. 360, 367–375. https://doi.org/10.1016/j.foreco.2015.10.001.
maximum peak and the time when VPD reaches its maximum value is a Alvares, C.A., Stape, J.L., Sentelhas, P.C., De Moraes Gonçalves, J.L., Sparovek, G., 2013.
recurring feature of forests. However, in tropical forests, which include Köppen's climate classification map for Brazil. Meteorol. Z. 22 (6), 711–728. https://
doi.org/10.1127/0941-2948/2013/0507.
tropical savannas, transition forests, seasonal forests and perhumid
Araújo, E.L., Castro, C.C., Albuquerque, U.P., 2007. Dynamics of brazilian Caatinga e a
rainforests, evidences show that Gs usually peaks in the early morning review concerning the plants, environment and people. Funct. Ecol. Comm. 1, 15–28.
after sunrise. Bai, B., Peviani, A., van der Horst, S., Gamm, M., Snel, B., Bentsink, L., Hanson, J., 2017.
Non-linear relationships between Gs versus VPD and Gs versus ET/ Extensive translational regulation during seed germination revealed by polysomal
profiling. New Phytol 214 (1), 233–244. https://doi.org/10.1111/nph.14355.
ETeq were found, corroborating studies in other ecosystems around the Baldocchi, D.D., Xu, L., 2007. What limits evaporation from Mediterranean oak wood-
world. On the other hand, the relationships between Gs versus NDVI lands – the supply of moisture in the soil, physiological control by plants or the
and Gs versus LAI were non-linear in the year 2014 and linear in 2015. demand by the atmosphere? Adv. Water Resour 30 (10), 2113–2122. https://doi.org/
10.1016/j.advwatres.2006.06.013.
These divergences may be associated with the differences observed in Barbosa, H.A., Huete, A.R., Baethgen, W.E., 2006. A 20-year study of NDVI variability
total annual rainfall. The seasonal variations of Gs, ET/ETeq and Ω were over the Northeast Region of Brazil. J. Arid Environ. 67, 288–307. https://doi.org/
remarkable, which is an expected feature of regions with a prolonged 10.1016/j.jaridenv.2006.02.022.
Barbosa, H.A., Kumar, T.V.L., 2016. Influence of rainfall variability on the vegetation
dry season (Ma et al., 2015; Tan et al., 2019). The analysis of the sea- dynamics over Northeastern Brazil. J. Arid Environ. 124, 377–387. https://doi.org/
sonal variation of Ω highlighted the severe drought conditions to which 10.1016/j.jaridenv.2015.08.015.
the region was exposed, since the surface did not decouple from the Campbell Scientific Inc, 2012. EC150 CO2 and H2O Open-path Gas Analyzer and EC100
Electronics With Optional CSAT3A 3D Sonic Anemometer. User Manual.ftp://ftp.
atmosphere in any day (maximum of 0.54 in the two years of this campbellsci.com/pub/csl/outgoing/uk/manuals/ec150.pdf. (Accessed 15, December
study). 2019).
Thus, results suggest that ET was controlled by the availability of Campos, S., Mendes, K.R., da Silva, L.L., Mutti, P.R., Medeiros, S.S., Amorim, L.B., dos
Santos, C.A.C., Perez-Marin, A.M., Ramos, T.M., Marques, T.V., Lucio, P.S., Costa,
surface water in the dry season and partially controlled by available
G.B., Santos e Silva, C.M., Bezerra, B.G., 2019. Closure and partitioning of the energy
energy during the wet season. When comparing studied years, we ob- balance in a preserved area of Brazilian seasonally dry tropical forest. Agric. For.
served that the larger accumulated rainfall in 2014 was the most Meteorol. 271, 398–412. https://doi.org/10.1016/j.agrformet.2019.03.018.
13
T.V. Marques, et al. Agricultural and Forest Meteorology 287 (2020) 107957
da Rocha, H.R., Goulden, M.L., Miler, S.D., Menton, M.C., Pinto, L.D.V.O., Freitas, H.C., microenvironmental variables. Hydrol. Process. 19 (2), 517–532. https://doi.org/10.
Figueira, A.M.S., 2004. Seasonality of water and heat fluxes over a tropical forest in 1002/hyp.5673.
eastern Amazonia. Ecological Application 14, 22–32. https://doi.org/10.1890/02- Lin, Y.-S., Medlyn, B.E., Duursma, R.A., Prentice, I.C., Wang, H., Baig, S., Eamus, D., de
6001. Dios, V.R., Mitchell, P., Ellsworth, D.S., de Beeck, M.O., Wallin, G., Uddling, J.,
da Silva, P.F., Lima, J.R.S., Antonino, A.C.D., Souza, R., de Souza, E.S., Silva, J.R.I., Alves, Tarvainen, L., Linderson, M.-L., Cernusak, L.A., Nippert, J.B., Ocheltree, T.W., Tissue,
E.M., 2017. Seasonal patterns of carbon dioxide, water and energy fluxes over the D.T., Martin-StPaul, N.K., Rogers, A., Warren, J.M., De Angelis, P., Hikosaka, K., Han,
Caatinga and grassland in the semi-arid region of Brazil. J. Arid Environ. 147, 71–82. Q., Onoda, Y., Gimeno, T.E., Barton, C.V.M., Bennie, J., Bonal, D., Bosc, A., Löw, M.,
https://doi.org/10.1016/j.jaridenv.2017.09.003. Macinins-Ng, C., Rey, A., Rowland, L., Setterfield, S.A., Tausz-Posch, S., Zaragoza-
Didan, K., 2015. MOD13Q1 MODIS/Terra Vegetation Indices 16-Day L3 Global 250m SIN Castells, J., Broadmeadow, M.S.J., Drake, J.E., Freeman, M., Ghannoum, O., Hutley,
Grid V006 [Data set]. NASA EOSDIS LP DAAC. https://doi.org/10.5067/MODIS/ L.B., Kelly, J.W., Kikuzawa, K., Kolari, P., Koyama, K., Limousin, J.-M., Meir, P., Lola
MOD13Q1.006. da Costa, A.C., Mikkelsen, T.N., Salinas, N., Sun, W., Wingate, L., 2015. Optimal
Dombroski, J.L.D., Praxedes, S.C., de Freitas, R.M.O., Pontes, F.M., 2011. Water relations stomatal behaviour around the world. Nat. Clim. Change. 5, 459–464. https://doi.
of Caatinga trees in the dry season. S. Afr. J. Bot. 77 (2), 430–434. https://doi.org/10. org/10.1038/nclimate2550.
1016/j.sajb.2010.11.001. Linares-Palomino, R., Oliveira-Filho, A.T., Pennington, R.T., 2011. Neotropical seasonally
Eamus, D., Cleverly, J., Boulain, N., Grant, N., Faux, R., Villalobos-Vega, R., 2013. Carbon dry forests: diversity, endemism, and biogeography of woody plants. In: Dirzo, R.,
and water fluxes in an arid-zone Acacia savanna woodland: an analyses of seasonal Young, H.S., Mooney, H.A., Ceballos, G. (Eds.), Seasonally dry tropical forests:
patterns and responses to rainfall events. Agr. For. Meteorol. 182-183, 225–238. ecology and conservation. Island Press, Washington, pp. 3–21. https://doi.org/10.
https://doi.org/10.1016/j.agrformet.2013.04.020. 5822/978-1-61091-021-7_1.
Fanourakis, D., Heuvelink, E., Carvalho, S.M.P., 2013. A comprehensive analysis of the Ma, N., Zhang, Y., Guo, Y., Gao, H., Zhang, H., Wang, Y., 2015. Environmental and
physiological and anatomical components involved in higher water loss rates after biophysical controls on the evapotranspiration over the highest alpine steppe.
leaf development at high humidity. J. Plant. Physiol. 170, 890–898. https://doi.org/ Journal of Hydrology 529, 980–992. https://doi.org/10.1016/j.jhydrol.2015.09.013.
10.1016/j.jplph.2013.01.013. Magrin, G.O., Marengo, J.A., Boulanger, J.-P., Buckeridge, M.S., Castellanos, E., Poveda,
Flexas, J., Bota, J., Galmés, J., Medrano, H., Ribas-Carbo, M., 2006. Kepping a positive G., Scarano, F.R., Vicuña, S., 2014. Central and South America. In: Barros, V.R., Field,
carbon balance under adverse conditions: responses of photosynthesis and respiration C.B., Dokken, D.J., Mastrandrea, M.D., Mach, K.J., Bilir, T.E., Chatterjee, M., Ebi,
to water stress. Physiol. Plant. 127, 343–352. https://doi.org/10.1111/jac.12245. K.L., Estrada, Y.O., Genova, R.C., Girma, B., Kissel, E.S., Levy, A.N., MacCracken, S.,
Flexas, J., Bota, J., Loreto, F., Cornic, G., Sharkey, T.D., 2004. Diffusive and metabolic Mastrandrea, P.R., White, L.L. (Eds.), Climate Change 2014: Impacts, Adaptation, and
limitations to photosynthesis under drought and salinity in C3 plants. Plant Biol 6, Vulnerability. Part B: Regional Aspects. Contribution of Working Group II to the Fifth
269–279. https://doi.org/10.1055/s-2004-820867. Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge
Flexas, J., Medrano, H., 2002. Drought-inhibition of photosynthesis in C3 plants: stomatal University Press, Cambridge, United Kingdom, pp. 1499–1566 and New York, NY,
and non-stomatal limitations revisited. Ann. Bot. 89 (2), 183–189. https://doi.org/ USA.
10.1093/aob/mcf027. Marengo, J.A., Alves, L.M., Alvala, R., Cunha, A.P., Brito, S., Moraes, O.L., 2017. Climatic
Freitas, A.D.S., Sampaio, E.V.S.B., Santos, C.E.R.S., Fernandes, A.R., 2010. Biological characteristics of the 2010-2016 drought in the semiarid Northeast Brazil region. An.
nitrogen fixation in tree legumes of the Brazilian semi-arid Caatinga. J. Arid Environ. Acad. Bras. Cienc. 90 (2), 1973–1985.
74, 344–349. https://doi.org/10.1016/j.jaridenv.2009.09.018. Marengo, J.A., Bernasconi, M., 2015. Regional differences in aridity/drought conditions
Ghimire, B., Lee, C., Heo, K., 2014. Leaf anatomy and its implications for phylogenetic over Northeast Brazil: present state and future projections. Clim. Chang. 129 (1–2),
relationships in Taxaceae s. l. J. Plant Res 127 (3), 373–388. https://doi.org/10. 103–115. https://doi.org/10.1007/s10584-014-1310-1.
1007/s10265-014-0625-3. Massman, W.J., 2000. A simple method for estimating frequency response corrections for
Giambelluca, T.W., Scholz, F.G., Bucci, S.J., Meinzer, F.C., Goldstein, G., Hoffmann, W.A., eddy covariance systems. Agric. For. Meteorol. 104, 185–198. https://doi.org/10.
Franco, A.C., Buchert, M.P., 2009. Evapotranspiration and energy balance of 1016/S0168-1923(00)00164-7.
Brazilians savannas with contrasting tree density. Agric. For. Meteorol. 149, McNaughton, K.G., Jarvis, P.G., 1983. Predicting effects of vegetation changes on tran-
1365–1376. https://doi.org/10.1016/j.agrformet.2009.03.006. spiration and evaporation. In: Kozlowski, T.T. (Ed.), Water deficits and plant growth
Gilliland, J.M., Klein, B.D, 2018. Position of the South Atlantic anticyclone and its impact v.7. Academic Press, New York, pp. 1–47.
on surface conditions across Brazil. J. Appl. Meteorol. Climatol. 57 (3), 535–553. Mendes, K.R., Granja, J.A.A., Ometto, J.P., Antonino, A.C.D., Menezes, R.S.C., Pereira,
https://doi.org/10.1175/JAMC-D-17-0178.1. E.C., Pompelli, M.F., 2017. Croton blanchetianus modulates its morphophysiological
Granier, A., Bréda, N., 1996. Modelling canopy conductance and stand transpiration of an responses to tolerate drought in a tropical dry forest. Funct. Plant Biol. 44 (10),
oak forest from sap flow measurements. Annals For. Sci. 53, 537–546. https://doi. 1039–1051. https://doi.org/10.1071/FP17098.
org/10.1051/forest:19960233. Miles, L., Newton, A.C., DeFries, R.S., Ravilious, C., May, I., Blyth, S., Kapos, V., Gordon,
Gu, S., Tang, Y., Cui, X., Du, M., Zhao, L., Li, Y., Xu, S., Zhou, H., Kato, T., Qi, P., Zhao, X., J.E., 2006. A global overview of the conservation status of tropical dry forests. J.
2008. Characterizing evapotranspiration over a meadow ecosystem on the Qinghai- Biogeogr. 33 (3), 491–505. https://doi.org/10.1111/j.1365-2699.2005.01424.x.
Tibetan Plateau. J. Geophys. Res. 113 (D8). https://doi.org/10.1029/2007jd009173. Mittermeier, R.A., Mittermeier, C.G., Brooks, T.M., Pilgrim, J.D., Konstant, W.R., da
Gu, S., Tang, Y., Cui, X., Kato, T., Du, M., Li, Y., Zhao, X., 2005. Energy exchange between Fonseca, G.A.B., Kormos, C., 2003. Wilderness and biodiversity conservation. Proc.
the atmosphere and a meadow ecosystem on the Qinghai-Tibetan Plateau. Agric. For. Natl. Acad. Sci. 100, 10309–10313. https://doi.org/10.1073/pnas.1732458100.
Meteorol. 129 (3–4), 175–185. https://doi.org/10.1016/j.agrformet.2004.12.002. Monteith, J., Unsworth, M., 2013. Principles of Environmental Physics: Plants, Animals,
Hastenrath, S., 2006. Circulation and teleconnection mechanisms of Northeast Brazil and the Atmosphere. Elsevier, Oxford, pp. 422.
droughts. Prog. Oceanogr. 70, 407–415. https://doi.org/10.1016/j.pocean.2005.07. Moore, C.J., 1986. Frequency response corrections for eddy correlation systems. Boud.-
004. Layer Meteorol. 37, 17–35.
Heilman, J.L., McInnes, K., Savage, M., Gesch, R., Lascano, R.J., 1994. Soil and canopy Moura, F., de, B.P., Mendes Malhado, A.C., Ladle, R.J., 2013. Nursing the caatinga back to
energy balances in a west Texas vineyard. Agric. For. Meteorol. 71 (1–2), 99–114. health. J. Arid Environ. 90, 67–68. https://doi.org/10.1016/j.jaridenv.2012.10.010.
https://doi.org/10.1016/0168-1923(94)90102-3 10.1590/0001- Moura, P.M., Althoff, T.D., Oliveira, R.A., Souto, J.S., Souto, P.C., Menezes, R.S.C.,
3765201720170206. Sampaio, E.V.S.B., 2016. Carbon and nutrient fluxes through litterfall at four suc-
Igarashi, Y., Kumagai, T.O., Yoshifuji, N., Sato, T., Tanaka, N., Tanaka, K., Suzuki, M., cession stages of Caatinga dry forest in Northeastern Brazil. Nutr. Cycl. Agroecosyst.
Tantasirin, C., 2015. Environmental control of canopy stomatal conductance in a 105, 25–38. https://doi.org/10.1007/s10705-016-9771-4.
tropical deciduous forest in northern Thailand. Agric. For. Meteorol. 202, 1–10. Mutti, P.R., da Silva, L.L., Medeiros, S.S., Dubreuil, V., Mendes, K.R., Marques, T.V.,
https://doi.org/10.1016/j.agrformet.2014.11.013. Lúcio, P.S., Santos e Silva, C.M., Bezerra, B.G., 2019. Basin scale rainfall-evapo-
Jarvis, P.G., McNaughton, K.G., 1986. Stomatal control of transpiration: scaling up from transpiration dynamics in a tropical semiarid environment during dry and wet years.
leaf to region. Adv. Ecol. Res. 15, 1–49. Int. J. Appl. Earth Obs. Geoinformation. 75, 29–43. https://doi.org/10.1016/j.jag.
Kljun, N., Calanca, P., Rotach, M.W., Schmid, H.P., 2015. A simple two-dimensional 2018.10.007.
parameterisation for Flux Footprint Prediction (FFP). Geosci. Model Dev. 8, Myneni, R., Knyazikhin, Y., Park, T. (2015). MCD15A3H MODIS/Terra+Aqua Leaf Area
3695–3713. https://doi.org/10.5194/gmd-8-3695-2015. Index/FPAR 4-day L4 Global 500m SIN Grid V006. NASA EOSDIS Land Processes
Koch, R., Almeida-Cortez, J.S., Kleinschmit, B., 2017. Revealing areas of high nature DAAC. 10.5067/MODIS/MCD15A3H.006.
conservation importance in a seasonally dry tropical forest in Brazil: Combination of Oliveira, P.T., Santos e Silva, C.M., Lima, K.C., 2017. Climatology and trend analysis of
modelled plant diversity hot spots and threat patterns. J. Nat. Conserv. 35, 24–39. extreme precipitation in subregions of Northeast Brazil. Theor. Appl. Climatol. 130
https://doi.org/10.1016/j.jnc.2016.11.004. (1–2), 77–90. https://doi.org/10.1007/s00704-016-1865-z.
Krishnan, P., Meyers, T.P., Scott, R.L., Kennedy, L., Heuer, M., 2012. Energy exchange Papale, D., Reichstein, M., Aubinet, M., Canfora, E., Bernhofer, C., Kutsch, W., Longdoz,
and evapotranspiration over two temperate semi-arid grasslands in North America. B., Rambal, S., Valentini, R., Vesala, T., Yakir, D., 2006. Towards a standardized
Agric. For. Meteorol. 153, 31–44. https://doi.org/10.1016/j.agrformet.2011.09.017. processing of Net Ecosystem Exchange measured with eddy covariance technique:
Lal, R., 2004. Soil carbon sequestration to mitigate climate change. Geoderma 123 (1–2), algorithms and uncertainty estimation. Biogeosciences 3, 571–583. https://doi.org/
1–22. https://doi.org/10.1016/j.geoderma.2004.01.032. 10.5194/bg-3-571-2006.
Leal, I.R., Silva, J.M.C., Tabarelli, M., Lacher Jr., T.E., 2005. Changing the course of Pennington, R.T., Lewis, G.P., Ratter, J.A., 2006. An overview of the plant diversity,
biodiversity conservation the Caatinga of Northeastern Brazil. Conserv. Biol. 19 (3), biogeography and conservation of neotropical savannas and seasonally dry forests.
701–706. https://doi.org/10.1111/j.1523-1739.2005.00703.x. In: Pennington, R.T., Lewis, G.P., Ratter, J.A. (Eds.), Neotropical Savannas and
Lei, H., Yang, D., 2010. Interannual and seasonal variability in evapotranspiration and Seasonally Dry Forests: Plant Diversity, Biogeography, and Conservation. CRC Press,
energy partitioning over an irrigated cropland in the North China Plain. Agric. For. Boca Raton, pp. 1–29.
Meteorol. 150 (4), 581–589. https://doi.org/10.1016/j.agrformet.2010.01.022. Pennington, R.T., Prado, D.E., Pendry, C.A., 2000. Neotropical seasonally dry forests and
Li, S.G., Lai, C.T., Lee, G., Shimoda, S., Yokoyama, T., Higuchi, A., Oikawa, T., 2005. Quaternary vegetation changes. J. Biogeogr. 27 (2), 261–273. https://doi.org/10.
Evapotranspiration from a wet temperate grassland and its sensitivity to 1046/j.1365-2699.2000.00397.x.
14
T.V. Marques, et al. Agricultural and Forest Meteorology 287 (2020) 107957
Pinho-Pessoa, A.C.B., Mendes, K.R., Jarma-Orozco, A., Pereira, M.P.S., Santos, M.A., Peng, S-S., Liu, X-Y., da Rocha, H.R., Kosugi, Y., Hirano, T., Saleska, S.R., Foulden,
Menezes, R.S.C., Ometto, J.P., Pereira, E.C., Pompelli, M.F., 2018. Interannual M.L., Zeng, J., Ding, F-J., Gao, F., Song, L., 2019. Surface conductance for evapo-
Variation in Temperature and Rainfall can Modulate the Physiological and transpiration of tropical forests: Calculations, variations, and controls. Agric. For.
Photoprotective Mechanisms of a Native Semiarid Plant Species. Indian J. Sci. Meteorol. 275, 317–328. https://doi.org/10.1016/j.agrformet.2019.06.006.
Technol. 11 (42). https://doi.org/10.17485/ijst/2018/v11i41/130972. Tang, J., Bolstad, P.V., Ewers, B.E., Desai, A.R., Davis, K.J., Carey, E.V., 2006. Sap flux-
Priestley, C.H.B., Taylor, R.J., 1972. On the assessment of surface heat flux and eva- upscaled canopy transpiration, stomatal conductance, and water use efficiency in an
poration using large-scale parameters. Mon. Weather Rev. 100, 81–92. https://doi. old growth forest in the Great Lakes region of the United States. J. Geophys. Res. 111
org/10.1175/1520-0493(1972)100<0081:OTAOSH>2.3.CO;2. (G2), 1095–1100. https://doi.org/10.1029/2005JG000083.
R Core Team, 2018. R: a language and environment for statistical computing. In: R Tavares-Damasceno, J.P., Silveira, J.L.G.S., Câmara, T., Stedile, P.C., Macario, P., Toledo-
Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org. Lima, G.S., Pichorim, M., 2017. Effect of drought on demography of Pileated Finch
Reichstein, M., Falge, E., Baldocchi, D., Papale, D., Aubinet, M., Berbigier, P., Bernhofer, (Coryphospingus pileatus: Thraupidae) in northeastern Brazil. J. Arid Environ. 147,
C., Buchmann, N., Gilmanov, T., Granier, A., Grünwald, T., Havrankova, K., 63–70. https://doi.org/10.1016/j.jaridenv.2017.09.006.
Ilvesniemi, H., Janous, D., Knohl, A., Laurila, T., Lohila, A., Loustau, D., Matteucci, Teixeira, A.H.C., Bastiaanssen, W.G.M., Ahmad, M.D., Moura, M.S.B., Bos, M.G., 2008.
G., Meyers, T., Miglietta, F., Ourcival, J.M., Pumpanen, J., Rambal, S., Rotenberg, E., Analysis of energy fluxes and vegetation-atmosphere parameters in irrigated and
Sanz, M., Tenhunen, J., Seufert, G., Vaccari, F., Vesala, T., Yakir, D., Valentini, R., natural ecosystems of semi-arid Brazil. J. Hydrol. 362, 110–127. https://doi.org/10.
2005. On the separation of net ecosystem exchange into assimilation and ecosystem 1016/j.jhydrol.2008.08.011.
respiration: Review and improved algorithm. Glob. Chang. Biol. 11, 1424–1439. Thom, A.S., 1972. Momentum, mass and heat exchange of vegetation. Q. J. R. Meteorol.
https://doi.org/10.1111/j.1365-2486.2005.001002.x. Soc. 98, 124–134. https://doi.org/10.1002/qj.49709841510.
Rodrigues, T.R., Vourlitis, G.L., Lobo, F.D.A., Oliveira, R.G., Nogueira, J.D.S., 2014. Tomasella, J., Vieira, R.M.S.P., Barbosa, A.A., Rodriguez, D.A., Santana, M.O., Sestini,
Seasonal variation in energy balance and canopy conductance for a tropical savanna M.F., 2018. Desertification trends in the Northeast of Brazil over the period
ecosystem of south central Mato Grosso. Brazil. J. Geophys. Res. (Biogeoscience) 119 2000–2016. Int. J. Appl. Earth. Obs. Geoinf. 73, 197–206. https://doi.org/10.1016/j.
(1), 1–13. https://doi.org/10.1002/2013JG002472. jag.2018.06.012.
Rodrigues, T.R., Vourlitis, G.L., Lobo, F.D.A., Santanna, F.B., Arruda, P.H.Z.D., Nogueira, Tucci, M.L.S., Erismann, N.M., Machado, E.C., Ribeiro, R.V., 2010. Diurnal and Seasonal
J.D.S., 2016. Modeling canopy conductance under contrasting seasonal conditions for variation in photosynthesis of peach palms grown under subtropical conditions.
a tropical savanna ecosystem of south central Mato Grosso. Brazil. Agric. For. Photosynthetica 48 (3), 421–429. https://doi.org/10.1007/s11099-010-0055-y.
Meteorol. 218–219, 218–229. https://doi.org/10.1016/j.agrformet.2015.12.060. Vourlitis, G.L., Nogueira, J.S., Lobo, F.A., Sendall, K.M., Paulo, S.R., Dias, C.A.A., Pinto
Ryu, Y., Baldocchi, D.D., Ma, S., Hehn, T., 2008. Interannual variability of evapo- Jr., O.B., Andrade, N.L.R., 2008. Energy balance and canopy conductance of a tro-
transpiration and energy exchange over an annual grassland in California. J. pical semi-deciduous forest of the southern Amazon Basin. Water Resour. Res. 44,
Geophys. Res. 113https://doi.org/10.1029/2007JD009263. D09104. W03412. https://doi.org/10.1029/2006WR005526.
Sampaio, E.V.S.B., 1995. Overview of the Brazilian Caatinga. In: Bullock, S.H., Mooney, Webb, E.K., Pearman, G.I., Leuning, R., 1980. Correction of flux measurements for density
H.A., Medina, E. (Eds.), Seasonally Dry Tropical Forests. University Press, Cambridge, effects due to heat and water vapour transfer. Quart. J. Roy. Meteorol. Soc. 106,
pp. 35–63. 85–100.
Santana, J.A.S., Santana Júnior, J.A.S., Barreto, W.S., Ferreira, A.T.S., 2016. Estrutura e Wehr, R., Commane, R., Munger, J.W., McManus, J.B., Nelson, D.D., Zahniser, M.S.,
distribuição espacial da vegetação da Caatinga na Estação Ecológica do Seridó. RN. Saleska, S.R., Wofsy, S.C., 2017. Dynamics of canopy stomatal conductance, tran-
Braz. J. For. Res. 36, 355–361. https://doi.org/10.4336/2016.pfb.36.88.1002. In spiration, and evaporation in a temperate deciduous forest, validated by carbonyl
Portuguese. sulfide uptake. Biogeosciences 14 (2), 389–401. https://doi.org/10.5194/bg-14-389-
Santos, M.G., Oliveira, M.T., Figueiredo, K.V., Falcão, H.M., Arruda, E.C.P., Almeida- 2017.
Cortez, J., Sampaio, E.V.S.B., Ometto, J.P.H.B., Menezes, R.S.C., Oliveira, A.F.M., Wever, L.A., Flanagan, L.B., Carlson, P.J., 2002. Seasonal and interannual variation in
Pompelli, M.F., Antonino, A.C.D., 2014. Caatinga, the Brazilian dry tropical forest: evapotranspiration, energy balance and surface conductance in a northern temperate
can it tolerate climate changes? Theor. Exp. Plant Physiol. 26, 83–99. https://doi. grassland. Agric. For. Meteor. 112, 31–49. https://doi.org/10.1016/S0168-1923(02)
org/10.1007/s40626-014-0008-0. 00041-2.
Santos, A.S., Santos e Silva, C.M., 2013. Seasonality, Interannual Variability, and Linear Wilson, K.B., Baldocchi, D.D., 2000. Seasonal and interannual variability of energy fluxes
Tendency of Wind Speeds in the Northeast Brazil from 1986 to 2011. The Sci. World over a broadleaved temperate deciduous forest in North America. Agric. For.
J. https://doi.org/10.1155/2013/490857. 490857. Meteorol. 100 (1), 1–18. https://doi.org/10.1016/S0168-1923(99)00088-X.
Santos, C.A.C., Mariano, D.A., Nascimento, F.C.A., Dantas, F.R.C., Oliveira, G., Silva, Xavier, A.C., King, C.W., Scanlon, B.R., 2015. Daily gridded meteorological variables in
M.T., Silva, L.L., Silva, B.B., Bezerra, B.G., Safa, B., Medeiros, S.S., Neale, C.M.U., Brazil (1980–2013). Int. J. Climatol. 36 (6), 2644–2659. https://doi.org/10.1002/
2020. Spatio-temporal patterns of energy exchange and evapotranspiration during an joc.4518.
intense drought for drylands in Brazil. International Journal of Applied Earth Zanella De Arruda, P.H., Vourlitis, G.L., Santanna, F.B., Pinto Jr., O.B., Lobo, F.A.,
Observation and Geoinformation 85, 101982. https://doi.org/10.1016/j.jag.2019. Nogueira, J.S., 2016. Large net CO2 loss from a grass-dominated tropical savanna in
101982. south-central Brazil in response to seasonal and interannual drought. J. Geophys. Res.
Santos, R.M., Oliveira-Filho, A.T., Eisenlohr, P.V., Queiroz, L.P., Cardoso, D.B., Rodal, Biogeosci. 121, 2110–2124. https://doi.org/10.1002/2016JG003404.
M.J., 2012. Identity and relationships of the Arboreal Caatinga among other floristic Zha, T., Li, C., Kellomäki, S., Peltola, H., Wang, K.Y., Zhang, Y., 2013. Controls of eva-
units of seasonally dry tropical forests (SDTFs) of north-eastern and Central Brazil. potranspiration and CO2 fluxes from scots pine by surface conductance and abiotic
Ecol. Evol. 2 (2), 409–428. https://doi.org/10.1002/ece3.91. factors. PLoS One 8. 10.1371/journal.pone.0069027.
Särkinen, T., Iganci, J.R.V., Linares-Palomino, R., Simon, M.F., Prado, D.E., 2011. Zhang, Q., Manzoni, S., Katul, G., Porporato, A., Yang, D.W., 2014. The hysteretic eva-
Forgotten forests - issues and prospects in biome mapping using Seasonally Dry potranspiration–vapor pressure deficit relation. J. Geophys. Res. Biogeo. 119 (2),
Tropical Forests as a case study. BMC Ecol. 11–27. https://doi.org/10.1186/1472- 125–140. https://doi.org/10.1002/2013JG002484.
6785-11-27. Zhang, Y., Kadota, T., Ohata, T., Oyunbaatar, D., 2007. Environmental controls on eva-
Souza, L.Q., Freitas, A.D.S., Sampaio, E.V.D.S.B., Moura, P.M., Menezes, R.S.C., 2012. potranspiration from sparse grassland in Mongolia. Hydrol. Process. 21 (15),
How much nitrogen is fixed by biological symbiosis in tropical dry forests? 1. Trees 2016–2027. https://doi.org/10.1002/hyp.6711.
and shrubs. Nutr. Cycl. Agroecosyst 2 (94), 171–179. https://doi.org/10.1007/ Zhang, Y., Meinzer, F.C., Jin-Hua, Q., Guillermo, G., Kun-Fang, C., 2013. Midday stomatal
s10705-012-9531-z. conductance is more related to stem rather than leaf water status in subtropical de-
Stewart, J.B., 1988. Modelling surface conductance of pine forest. Agric. For. Meteorol. ciduous and evergreen broadleaf trees. Plant Cell Environ 36 (1), 149–158. https://
43 (1), 19–35. https://doi.org/10.1016/0168-1923(88)90003-2. doi.org/10.1111/j.1365-3040.2012.02563.x.
Szeicz, G., Endrödi, G., Tajchman, S., 1969. Aerodynamic and surface factors in eva- Zhu, G., Lu, L., Su, Y., Wang, X., Cui, X., Ma, J., He, J., Zhang, K., Li, C., 2014. Energy flux
poration. Water Resour. Res. 5, 380–394. https://doi.org/10.1029/ partitioning and evapotranspiration in a sub-alpine spruce forest ecosystem. Hydrol.
WR005i002p00380. Process. 28 (19), 5093–5104. https://doi.org/10.1002/hyp.9995.
Tan, Z-H., Zhao, J-F., Wang, G-Z., Chen, M-P., Yang, L-Y., He, C-S., Restrepo-Coupe, N.,
15