Homogeneização biótica em ambientes aquáticos continentais · russas do mundo! Obrigada pelo...
181
Universidade Federal de Goiás Instituto de Ciências Biológicas Programa de Pós-Graduação em Ecologia e Evolução Homogeneização biótica em ambientes aquáticos continentais DANIELLE KATHARINE PETSCH Goiânia 2018
Transcript of Homogeneização biótica em ambientes aquáticos continentais · russas do mundo! Obrigada pelo...
1
Universidade Federal de Goiás
Instituto de Ciências Biológicas
Programa de Pós-Graduação em Ecologia e Evolução
Homogeneização biótica em ambientes
aquáticos continentais
DANIELLE KATHARINE PETSCH
Goiânia
2018
2
3
DANIELLE KATHARINE PETSCH
Homogeneização biótica em ambientes
aquáticos continentais
Tese apresentada ao Programa de Pós-Graduação
em Ecologia e Evolução do Departamento de
Ecologia do Instituto de Ciências Biológicas da
Universidade Federal de Goiás como requisito
parcial para obtenção do título de Doutora em
Ecologia e Evolução.
Orientador: Prof. Dr. Adriano Sanches Melo
Goiânia
2018
4
5
6
7
DEDICATÓRIA
Dedico esse trabalho aos meus pais Osmar e Maria
Helena e ao meu noivo Yuri pelo apoio irrestrito em todas
minhas decisões - mesmo e principalmente por aquelas
que me fizeram estar longe deles.
8
“It's a dangerous business, Frodo, going out your door. You step onto the
road, and if you don't keep your feet, there's no knowing where you might
be swept off to.”
J.R.R. Tolkien - The Lord of the Rings
9
AGRADECIMENTOS
Assim como Bilbo Bolseiro na saga O Senhor dos Anéis, eu ainda não tinha me
aventurado para terras distantes da minha casa até certa altura da minha vida – até o fim
do mestrado. No entanto, a jornada do doutorado que começou em Goiânia me levou
também para outros destinos inesperados onde conheci lugares incríveis e encontrei
pessoas fantásticas que tornaram a caminhada muito mais fácil, feliz e especial – e
também fizeram a seção dos agradecimentos se tornar mais longa!
Agradeço, primeiramente, à minha família por constituir o porto seguro de todas
minhas aventuras. Em especial, agradeço aos meus pais Osmar e Maria Helena e ao meu
noivo Yuri pelo amor e apoio incondicional em todas as decisões que me fizeram chegar
até aqui. A distância foi muitas vezes difícil, mas o amor de vocês sempre me deu forças
para continuar.
Agradeço ao Adriano por ser o melhor orientador do mundo! Um exemplo de
ética e profissionalismo. Foi um orientador maravilhoso desde o primeiro dia de
doutorado até o fim da tese, quando estávamos na mesma cidade ou distantes por cerca
de 8.000 quilômetros. É minha inspiração para o futuro que desejo seguir! Sinto-me
privilegiada por dizer que fui sua aluna.
Agradeço aos amigos de Maringá por me incentivarem e apoiarem a fazer o
doutorado em Goiânia, mas também por sempre me receberem de braços abertos todas
as (muitas) vezes que retornava à UEM: aos amigos da graduação (Lô, Say, Nati,
Barbris, Ju, Nay, Dri e Fer), aos amigos do mestrado (Lô, Nati, Barbris, Jean, Ju, Bia,
Vini, Camis, Herick) e aos amigos do laboratório Zoobentos (Gi, Camis, Flávio, Rafa,
Rê, Ana, Jess, Vá, Alice e Róger).
Agradeço aos amigos de Goiânia, que me fizeram entender o ditado de que o que
faz um lugar ser bom são as pessoas que vivem nele. Encontrei pessoas maravilhosas
10
nessa cidade que me proporcionaram uma estadia imensamente feliz! Agradeço a todos
os colegas do PPG EcoEvol pela convivência diária! Muitos foram tão receptivos,
acolhedores e queridos que se tornaram os “ursinhos carinhosos”. Em especial,
agradeço Barbba, Laris, Flávia, Vini, Cibele, Lilian, Lara, Fernando, Leila, Olívia,
Angélica, André, Tati, Raíssa e Klein. A todos os amigos do LETS, mas especialmente
ao Luciano, conterrâneo parceiro e irmão de orientação, e também Vini, Jaques, Jean,
Lores, Leila, Jesus, Marco Túlio, Júlio, Regata, Alice e Elisa. Aos amigos do residencial
New Orleans, minha casinha goiana por três anos, pelas festinhas juninas, disputas na
dança da cadeira e pelos episódios emocionantes de Game of Thrones nos domingos à
noite: Vini, Thársis, Fabi, Lê, Isaque, Breno, Lara, Rherison, Laris e Kayque. Todos
vocês contribuíram de alguma forma para que eu me sentisse mais em casa na terra do
pequi!
Agradeço aos amigos que fiz nos quatro meses em Rio Claro. Erison, Fer e Cris
pela gentil hospedagem. Agradeço ao Tadeu e aos demais do LaTa e agregados por me
ajudarem tanto no planejamento das coletas, na triagem e identificação do material, mas
também pelos almoços no RU e churrasquinhos com panceta na “casinha”: Xuleta,
Edineusa, Jéssica, Larica, Sayuri e Raul. Finalmente, agradeço imensamente a equipe
mais linda de coleta de riachos desse Brasil: Amá, Carlinhos e Larica, por “meterem o
loko” comigo desde os riachos “padrão Finlândia” até os riachos não lá muito bons.
Tanto a gigante competência de todos vocês como as comidinhas deliciosas, as disputas
do fusca azul e a cantoria ao som da diva Sandy tornaram tudo mais fácil e muito
prazeroso. Coletaria muito mais que 100 riachos ao lado de vocês!
Agradeço às pessoas incríveis que encontrei nos quatro meses que vivi no
Canadá. Ao Karl Cottenie, por me receber e me orientar tão bem. Aos colegas de office,
Michelle e Josh, por serem sempre tão gentis e atenciosos comigo. Ao Elmer e Elvira,
11
por me ensinarem muito mais que inglês, mas também servirem de inspiração como
casal e seres humanos. A Hynnaya, Fabi e Ricardo por compartilharem diversos
momentos canadenses com aquele toque brazuca. À Gabe, por ser a melhor landlady
que esse Canadá já viu e me dar a oportunidade de conviver com sua família linda e de
morar em sua maravilhosa casa “Darling”, que foi um verdadeiro lar por quatro meses.
Só pude chamar de lar por causa dos roommies que tive: Paty, Harley, Lucky (o dog
mais fofo e temperamental que existe) e, principalmente, Bruna e Laura, presentes do
Canadá que carregarei por toda a vida. Esse período incrível que vivenciei nessas terras
canadenses não teria metade da graça se não fosse por vocês. Obrigada pelas comidas
maravilhosas e pela parceria em todas as aventuras que inclui não me deixarem desistir
de esquiar mesmo quando essa parecia ser uma mádeia; pedalar de pijamas; encarar frio
abaixo de 10 graus negativos para passear; e andar em uma das 10 maiores montanhas-
russas do mundo! Obrigada pelo apoio em todos os momentos em todos esses dias que
estive no Canadá – e mesmo fora dele.
Agradeço às pessoas que encontrei na Alemanha que também tornaram essa
jornada germânica mais fácil. Ao Jonathan Chase, por ter sido um orientador ainda mais
maravilhoso do que imaginei. Aos meus filósofos preferidos, Wecio e Newton,
principalmente ao feliz apoio que me deram nos meus primeiros dias alemães. Agradeço
ao Martin, o alemão mais brasileiro que conheço, por ser um excelente anfitrião
“Leipzigiano”. Aos colegas do iDiv por me receberem tão bem e pela agradável
companhia nesses seis meses; em especial Leana, Thore, Lotte, Eduardo, Andros, Sue,
Elena e Amanda. Agradeço também aos iDivianos Felix, Lotte e Dylan por me
ajudarem no projeto sempre que precisei. Agradeço também aos amigos brasileiros
espalhados pela Europa que foram parceiros de diversas viagens que estarão sempre
guardadas nas minhas mais lindas recordações: Gabi, Mari, Barbris, Barbba, Carina,
12
Mayra e Vini. Agradeço ao Jani Heino pela tão gentil recepção durante a semana que
passei em seu laboratório na University of Oulu. Agradeço a Mari (e Mustikka, claro!)
por me hospedar tão bem e me mostrar a melhor experiência finlandês (sauna escaldante
alternada com um rio gelado!).
Agradeço aos maravilhosos colaboradores que tive o privilégio de trabalhar ao
longo dos diferentes capítulos. Em especial, ao Karl Cottenie, Jonathan Chase, Tadeu
Siqueira, Fabiana Schneck, Jani Heino e Juliana Dias.
Agradeço também a todas as pessoas que forneceram dados para a realização da
meta-análise e também às pessoas que me auxiliaram com informações sobre os
atributos funcionais dos insetos aquáticos.
Finalmente, agradeço aos órgãos que proporcionaram suporte financeiro e
oportunidades para que eu pudesse realizar tranquilamente meu doutorado no Brasil
bem como nos períodos fora: à Capes pela bolsa de doutorado no Brasil e na Alemanha
pelo Programa de Doutorado Sanduíche no Exterior (PDSE), e a Global Affairs Canada
– Emerging Leaders in the Americas Program (ELAP) pela bolsa canadense.
Enfim, a todos que me ajudaram de alguma forma nessa jornada do doutorado,
deixo meu sincero muito obrigada!
13
SUMÁRIO
Apresentação da Tese ............................................................................. 14
Resumo ..................................................................................................... 16
Abstract .................................................................................................... 17
Introdução Geral ..................................................................................... 18
Capítulo 1. Causes and consequences of biotic homogenization in
freshwater ecosystems ............................................................................... 26
Capítulo 2. Substratum simplification reduces beta diversity of stream
algal communities …….……………….……..…..……………………… 55
Capítulo 3. Floods homogenize aquatic communities across time but not
across space in a Neotropical floodplain …….…………..………..……. 84
Capítulo 4. Human land-use does not homogenize aquatic insect
communities in boreal and tropical streams ………..………………….. 119
Capítulo 5. Land-use effects on streams biodiversity: a meta-analysis...152
Considerações finais ............................................................................. 180
14
APRESENTAÇÃO DA TESE
Esta tese inclui Introdução Geral, cinco capítulos na forma de artigos e Considerações
Finais. A Introdução Geral apresenta os principais referenciais teóricos e problemas
ecológicos que motivaram a elaboração dessa tese. Cada capítulo representa um
manuscrito científico elaborado com base nas normas da revista em que foi publicado
ou será submetido, embora algumas modificações tenham sido feitas para facilitar a
leitura da tese. O primeiro capítulo, fruto da minha qualificação de doutorado, foi
publicado na revista International Review of Hydrobiology em 2016, e é intitulado
“Causes and consequences of biotic homogenization in freshwater ecosystems”. Ele
trata de uma revisão teórica sobre as principais causas e consequências da
homogeneização biótica em ambientes aquáticos continentais. No segundo capítulo,
intitulado “Substratum simplification reduces beta diversity of stream algal
communities”, utilizei dados de um experimento de campo conduzido pela Profª Dra
Fabiana Schneck para avaliar se a simplificação de habitats causa homogeneização
biótica de algas perifíticas. Ele foi publicado na revista Freshwater Biology em 2017. O
terceiro capítulo, intitulado “Floods homogenize aquatic communities across time but
not across space in a Neotropical floodplain”, foi desenvolvido em colaboração com o
Prof. Dr. Karl Cottenie durante meu doutorado sanduíche na University of Guelph
(Guelph, Canadá), bem como com pesquisadores do PPG em Ecologia de Ambientes
Aquáticos Continentais e Nupelia/UEM que forneceram dados de 16 anos de
monitoramento do projeto PELD (“Pesquisas Ecológicas de Longa duração”) na
planície de inundação do alto rio Paraná. O manuscrito está redigido nas normas da
revista Aquatic Sciences e trata do efeito do pulso de inundação sob a diversidade beta
de macrófitas e zooplâncton no espaço e no tempo. Já o quarto capítulo da tese, “Human
land-use does not homogenize aquatic insect communities in boreal and tropical
15
streams”, está inserido em um projeto maior intitulado “Scaling biodiversity in tropical
and boreal streams: implications for diversity mapping and environmental assessment
(ScaleBio)”, coordenado no Brasil pelo Profº Dr Tadeu Siqueira e na Finlândia pelo
Profº Dr Jani Heino. Visitei os laboratórios coordenados por ambos os professores e
participei das coletas nos 100 riachos brasileiros. O manuscrito trata da comparação da
diversidade beta entre riachos boreais e entre riachos tropicais e da possível
homogeneização biótica em ambas as regiões devido à redução da heterogeneidade
ambiental e aumento da severidade ambiental mediados pelo intensivo uso do solo. Esse
manuscrito está redigido no formato da revista Ecological Indicators. Finalmente, o
quinto e último capítulo da tese cujo título é “Land-use effects on stream biodiversity: a
meta-analysis” corresponde a uma meta-análise e foi desenvolvido em parceria com o
Profº Dr. Jonathan Chase durante meu doutorado sanduíche no German Centre for
Integrative Biodiversity Research (iDiv) (Leipzig, Alemanha). Trata dos efeitos de
diferentes tipos de uso do solo sob a biodiversidade em riachos. Esse manuscrito está
redigido nas normas da seção Reports da revista Ecology. Por fim, a seção
“Considerações finais” sumariza as principais conclusões da tese.
16
RESUMO
O aumento da similaridade entre comunidades é um processo conhecido como
homogeneização biótica. Em ecossistemas aquáticos continentais a homogeneização
biótica pode ser promovida por diversas causas naturais (e.g. pulso de inundação) e
antrópicas (e.g. modificações do uso do solo). No primeiro capítulo, revisei as
principais causas e consequências da homogeneização de biotas aquáticas continentais.
No segundo capítulo, por meio de um experimento, demonstrei que a simplificação de
habitats pode causar homogeneização de algas perifíticas, embora o resultado dependa
da forma como se estima a homogeneização. No terceiro capítulo, usando dados de
zooplâncton e macrófitas, mostrei que as cheias homogeneizaram uma mesma lagoa ao
longo do tempo, mas não tornam lagoas mais similares espacialmente. No quarto
capítulo demonstrei que a diversidade beta taxonômica de insetos aquáticos foi maior
entre riachos tropicais enquanto a diversidade beta funcional foi maior entre riachos
boreais. O aumento da degradação ambiental e redução na heterogeneidade de habitat
relacionados ao uso do solo não causaram homogeneização taxonômica nem funcional
dos insetos aquáticos em riachos tropicais ou boreais. Por fim, no quinto capítulo,
observei em uma meta-análise que riachos modificados possuem menor riqueza e
equitabilidade além de uma diferente composição de espécies em relação aos riachos
mais conservados. No entanto, modificações no uso do solo não causaram
homogeneização biótica. Embora os efeitos de possíveis causas de homogeneização de
biotas aquáticas sejam ainda controversos, recomendamos que estudos sobre
biodiversidade incluam a diversidade beta para uma melhor compreensão dos
mecanismos que estruturam as comunidades frente a distúrbios antrópicos ou naturais.
Palavras-chave: Diversidade beta; Hábitats simples; Planície de inundação; Uso do
solo; Homogeneização funcional; Riachos; Meta-análise.
17
ABSTRACT
The increase in similarity among communities is a process known as biotic
homogenization. In freshwater ecosystems, biotic homogenization may be promoted by
different natural (e.g. flood pulse) and human (e.g. land use) causes. In the first chapter,
I reviewed the main causes and consequences of freshwater homogenization. In the
second chapter, using an experimental approach, I showed that habitat simplification
may cause homogenization of periphytic algae, but the results depended on how
dissimilarity was estimated. In the third chapter, using zooplankton and macrophytes
data, I showed that floods homogenized individual lakes across time but did not make
the lakes spatially more similar. In the fourth chapter, I demonstrated that taxonomic
beta diversity of aquatic insects was higher among tropical streams but functional beta
diversity was higher among boreal streams. The increase of environmental harshness
and decrease of environmental heterogeneity did not cause taxonomic or functional
homogenization of aquatic insects among tropical or boreal streams. Finally, in the fifth
chapter, I found in a meta-analysis that human modified streams have low species
richness and equitability, although a distinct species composition regarding to reference
streams. However, land-use changes did not cause biotic homogenization. Although the
effects of possible biotic homogenization causes are still controversy, we recommend
that biodiversity studies should include beta diversity to better understand mechanisms
structuring communities under pressure of human or natural disturbances.
Keywords: Beta diversity; Habitat simplification; Floodplain; Land use; Functional
homogenization; Streams; Meta-analysis.
18
INTRODUÇÃO GERAL
A biodiversidade está diminuindo em uma taxa nunca vista antes (Butchart et al., 2010).
Impactos antrópicos tais como a introdução de espécies, a simplificação e alteração de
hábitats e as mudanças climáticas têm causado a extinção de muitas espécies (Rahel &
Olden, 2008; McGill et al. 2015) aumento do número de espécies extintas recentemente
é tão alarmante que estimativas comparando as taxas naturais de extinção em fósseis às
taxas apresentadas atualmente e indicam que podemos estar vivenciando um novo
evento de extinção em massa (Barnosky et al., 2011). No entanto, além da redução do
número de espécies, outras complexas consequências podem ser geradas pela intensa
atividade humana, como o favorecimento de espécies generalistas e de ampla
distribuição em detrimento das mais especialistas e raras, tornando as comunidades cada
vez mais parecidas (Elton, 1958; McKinney & Lockwood, 1999). Esse processo de
aumento da similaridade entre comunidades é conhecido como homogeneização biótica
(Olden et al., 2004), e pode ser mensurado pela diversidade beta (i.e. variabilidade entre
as comunidades). A homogeneização biótica é atualmente considerada como um
processo tão preocupante que o período em que vivenciamos tem sido denominado de
“Homogenoceno” ou “A Nova Pangeia” (Olden, 2006).
Um dos primeiros pesquisadores a reconhecer o processo de homogeneização
biótica foi Charles Elton, em seu livro “The ecology of invasions by animals and plants”
publicado em 1958. Charles Elton percebeu que as extinções e as invasões de espécies
em decorrência da exploração humana e da dispersão mediada pelo comércio
intercontinental estavam tornando as biotas, anteriormente distintas, mais parecidas. No
entanto, os pesquisadores responsáveis por consagrarem o termo de homogeneização
biótica foram Michael L. McKinney e Julie L. Lockwood em 1999, quando postularam
que a homogeneização biótica é a “substituição de biotas locais por espécies não-
19
nativas, geralmente introduzidas por humanos”. Embora os primeiros estudos sobre
homogeneização biótica tenham focado principalmente nos efeitos da introdução de
espécies exóticas (e.g. Rahel, 2002; Olden & Poff, 2003), muitas outras causas de
aumento da similaridade entre as comunidades foram posteriormente reconhecidas, tais
como modificações no uso do solo (e.g. Siqueira et al., 2015; Solar et al., 2015),
mudanças climáticas (e.g. Magurran et al., 2015) e eutrofização (e.g. Donohue et al.,
2009).
Ecossistemas aquáticos continentais, que estão entre os mais diversos e ao
mesmo tempo entre os mais ameaçados ecossistemas do globo (Strayer & Dudgeon,
2010), tem tido suas comunidades mais homogêneas principalmente devido a causas
relacionadas a atividades humanas, tais como a introdução de espécies não-nativas, o
barramento fluvial e as modificações no uso do solo (e.g. Beisner et al., 2003; Vitule et
al., 2012; Daga et al., 2015; Siqueira et al., 2015). A conservação de ecossistemas
aquáticos continentais é ainda de especial importância devido às diversas funções e
serviços ecossistêmicos que desempenham, tais como provisão e regulação da água,
pesca, produção primária e ciclagem de nutrientes (Millennium Ecosystem Assessment,
2005; Vörösmarty et al., 2010). Além disso, a homogeneização biótica pode também
tornar as comunidades ainda mais vulneráveis frente aos distúrbios promovidos pela
pressão antrópica por sincronizar as respostas entre as comunidades locais (Olden et al.,
2004).
A conversão de áreas vegetais nativas em áreas utilizadas pelo homem é uma das
principais causas de perda da diversidade biológica em ecossistemas aquáticos e
terrestres, dos polos aos trópicos (Sala et al., 2000). A perda de biodiversidade em
ecossistemas aquáticos promovida por mudanças no uso do solo pode ser mediada por
dois principais mecanismos: aumento da severidade ambiental e redução na
20
heterogeneidade de habitat. A severidade ambiental ocorre quando as condições
abióticas são limitantes para a maioria das espécies (Chase, 2007, 2010), como quando
altas concentrações de nitrogênio e fósforo provindas da agricultura promovem
eutrofização de corpos aquáticos, tornando as condições na água favoráveis apenas a
poucas espécies. Se apenas o mesmo conjunto limitado de espécies ocorre entre os
habitats com condições ambientais mais severas, a dissimilaridade entre essas
comunidades locais é reduzida (Chase, 2007, 2010). Já a redução na heterogeneidade de
habitat em ecossistemas aquáticos pode ocorrer, por exemplo, quando o desmatamento
promove o assoreamento do leito dos riachos reduzindo a diversidade e complexidade
do substrato. A heterogeneidade de habitats aumenta o número de espécies por fornecer
maior disponibilidade de recursos, microhabitats e refúgios (Schneck & Melo, 2013;
Pierre & Kovalenko, 2014; Stein et al., 2014). Tais condições favoráveis a uma maior
gama de espécies podem facilitar a estocasticidade na história de colonização que
associada aos efeitos prioritários (i.e. o efeito dos primeiros colonizadores nos
seguintes), pode tornar as comunidades mais diferentes entre os habitats mais
heterogêneos do que entre os habitats mais homogêneos (Chase, 2010; Vannette &
Fukami, 2014). Além disso, a variabilidade nas condições abióticas físicas e químicas
pode refletir em uma distinta composição de espécies entre os habitats heterogêneos,
causando menor dissimilaridade entre as comunidades com menor heterogeneidade
ambiental. Ambos os processos, i.e., maior severidade ambiental e menor
heterogeneidade de habitat, são bem conhecidos por reduzir riqueza de espécies em
ecossistemas aquáticos (e.g. Allan, 2004; Leal et al., 2016), mas seus efeitos são ainda
controversos em relação à diversidade beta.
Embora a homogeneização biótica seja geralmente considerada como uma
consequência negativa de atividades antrópicas, fenômenos naturais também podem
21
homogeneizar as biotas aquáticas. Por exemplo, em sistemas de rio-planície de
inundação as cheias podem aumentar a similaridade biológica entre os ambientes, pois
tendem a aumentar a conectividade e a similaridade ambiental entre os locais (Thomaz
et al., 2007; Bozelli et al., 2015). Por outro lado, durante a seca, os ambientes se tornam
mais diferenciados, o que permitiria uma maior variabilidade de espécies entre os locais.
Esses mecanismos, responsáveis por uma homogeneização espacial das comunidades
em períodos de cheia (i.e. aumento da similaridade entre os locais em um mesmo
período), poderiam também estar relacionados a uma homogeneização temporal das
comunidades (i.e. aumento da similaridade entre os períodos de cheia em um mesmo
local). Adicionar a dimensão temporal em estudos de homogeneização biótica pode
resultar em uma compreensão mais profunda sobre os mecanismos subjacentes ao
aumento da similaridade entre as comunidades aquáticas.
As principais causas e consequências da homogeneização biótica em ambientes
aquáticos continentais são sumarizadas em uma revisão teórica no Capítulo 1. Algumas
dessas possíveis causas de homogeneização biótica são exploradas mais detalhadamente
por meio de diferentes métodos (i.e. experimento, dados observacionais e meta-análise)
nos capítulos seguintes da tese, como a simplificação de habitats (Capítulo 2), pulso de
inundação (Capítulo 3) e modificações no uso do solo (Capítulo 4 e Capítulo 5). Mais
especificamente, investiguei no segundo capítulo, por meio de uma abordagem
experimental, se a comunidade de algas perifíticas é mais homogênea entre substratos
simples do que entre substratos complexos. No terceiro capítulo, investiguei se as
comunidades aquáticas de um mesmo local são mais similares entre períodos de cheia
do que entre períodos de seca em uma planície de inundação neotropical. Também
investiguei se as comunidades aquáticas são espacialmente mais similares entre si
durante o período de cheia do que durante o período de seca. No quarto capítulo,
22
utilizando 100 riachos amostrados no Brasil e 100 riachos amostrados na Finlândia,
investiguei se a diversidade beta taxonômica e funcional é maior entre riachos tropicais
do que entre boreais e se mudanças no uso do solo, mediadas por degradação e
homogeneidade ambiental, reduzem a diversidade beta taxonômica e funcional em
ambas as regiões climáticas. Finalmente, no quinto capítulo, conduzi uma meta-análise
em riachos para investigar se modificações no uso do solo reduzem a riqueza observada,
extrapolada e a equitabilidade, e ainda se mudam a composição de espécies e reduzem a
diversidade beta acarretando em homogeneização biótica.
Referências
Allan J. D. 2004. Landscapes and riverscapes: the influence of land use on stream
ecosystems. Annual Review of Ecology, Evolution, and Systematics, 35: 257–
284.
Barnosky A. D. et al. 2011. Has the Earth's sixth mass extinction already arrived?
Nature, 471: 51-57.
Beisner B. E., Ives A. R. & Carpenter S. R. 2003. The effects of an exotic fish invasion
on the prey communities of two lakes. Journal of Animal Ecology, 72: 331–342.
Bozelli R. L., Thomaz S. M., Padial A. A., Lopes P. M. & Bini L. M. 2015. Floods
decrease zooplankton beta diversity and environmental heterogeneity in an
Amazonian floodplain system. Hydrobiologia, 753: 233-241.
Butchart S. H. M. et al. 2010. Global biodiversity: Indicators of recent declines.
Science, 328: 1164-1168.
Chase J. M. 2007. Drought mediates the importance of stochastic community assembly.
Proceedings of the National Academy of Sciences USA, 104: 17430–17434.
23
Chase J. M. 2010. Stochastic community assembly causes higher biodiversity in more
productive environments. Science, 328: 1388–1391.
Daga V. S., Skora F., Padial A. A., Gubiani E. A. & Vitule J. R. S. 2015.
Homogenization dynamics of the fish assemblages in Neotropical reservoirs:
comparing the roles of introduced species and their vectors. Hydrobiologia, 746:
327–347.
Donohue I., Jackson A. L., Pusch M. T. & Irvine K. 2009. Nutrient enrichment
homogenizes lake benthic assemblages at local and regional scales. Ecology,
90: 3470–3477.
Elton C. S. 1958. The ecology of invasions by animals and plants. London: Methuen.
Leal C. G. et al. 2016. Multi-scale assessment of human-induced changes to Amazonian
instream habitats. Landscape Ecology, 31: 1725–1745.
Magurran A. E., Dornelas M., Moyes F., Gotelli N. J. & McGill B. 2015. Rapid biotic
homogenization of marine fish assemblages. Nature Communications, 6: 1–5.
McGill B J., Dornelas M., Gotelli N. J. & Magurran A. E. 2015. Fifteen forms of
biodiversity trend in the Anthropocene. Trends in Ecology & Evolution, 30:
104:113.
McKinney M. L. & Lockwood J. L. 1999. Biotic homogenization: a few winners
replacing many losers in the next mass extinction. Trends in Ecology &
Evolution, 14: 450-453.
Millenium Ecosystem Assessment. 2005. Ecosystems and Human Well Being. Island
Press, Washington DC.
Olden J. D. & Poff N. L. 2003. Toward a mechanistic understanding and prediction of
biotic homogenization. The American Naturalist, 162: 442–460.
24
Olden J. D. 2006. Biotic homogenization: a new research agenda for conservation
biogeography. Journal of Biogeography, 33: 2027–2039.
Olden J. D., Poff N. R., Douglas M. R., Douglas M. E. & Fausch K. D. 2004.
Ecological and evolutionary consequences of biotic homogenization. Trends in
Ecology & Evolution, 19: 18-24.
Pierre J. I. S. & Kovalenko K. E. 2014. Effect of habitat complexity attributes on
species richness. Ecosphere, 5: 1–10.
Rahel F. J. 2002. Homogenization of freshwater faunas. Annual Review of Ecology,
Evolution, and Systematics, 33: 291–315.
Rahel F. J. & Olden J. D. 2008. Assessing the effects of climate change on aquatic
invasive species. Conservation Biology, 22: 521–533.
Sala O. E. et al. 2000. Global biodiversity scenarios for the year 2100. Science, 287:
1770-1774.
Schneck F. & Melo A. S. 2013. High assemblage persistence in heterogeneous habitats:
an experimental test with stream benthic algae. Freshwater Biology, 58: 365–
371.
Siqueira T., Lacerda C. G. T. & Saito V. S. 2015. How does landscape modification
induce biological homogenization in tropical stream metacommunities?
Biotropica, 47: 509–516.
Solar R. B. C. et al. 2015. How pervasive is biotic homogenization in human-modified
tropical forest landscapes? Ecology Letters, 18: 1108–1118.
Stein A., Gerstner K. & Kreft H. 2014. Environmental heterogeneity as a universal
driver of species richness across taxa, biomes and spatial scales. Ecology
Letters, 17: 866–880.
25
Strayer D. J. & Dudgeon D. 2010. Freshwater biodiversity conservation: recent progress
and future challenges. Journal of the North American Benthological Society, 29:
344-358.
Thomaz S. M., Bini L. M. & Bozelli R. L. 2007. Floods increase similarity among
aquatic habitats in river floodplain systems. Hydrobiologia, 579: 1-13
Vannette R. L. & Fukami T. 2014. Historical contingency in species interactions:
towards niche-based predictions. Ecology Letters, 17: 115–124.
Vitule J. R. S., Skóra F. & Abilhoa V. 2012. Homogenization of freshwater fish faunas
after the elimination of a natural barrier by a dam in Neotropics. Diversity and
Distributions, 18: 111–120.
Vörösmarty C. J. et al. 2010. Global threats to human water security and river
biodiversity. Nature, 467: 555–561.
26
CAUSES AND CONSEQUENCES OF BIOTIC
HOMOGENIZATION IN FRESHWATER
ECOSYSTEMS1
1 Petsch, D. K. 2016. Causes and consequences of biotic homogenization in freshwater
ecosystems. International Review of Hydrobiology, 101:113–122.
C APÍTULO 1
1
27
Abstract
Biotic homogenization goes beyond the increase in taxonomic similarity among
communities. It also involves the loss of biological differences in any organizational
level (e.g., populations or communities) in terms of functional, taxonomic or genetic
features. There are many ways to measure biotic homogenization, and the results
depend on temporal and spatial scales, the biological group and the richness of the
communities. In freshwater ecosystems, the main investigated causes of biotic
homogenization correspond to the introduction of non-native species, damming, and
changes in land use. However, other natural and anthropogenic causes also increase
similarity among aquatic biota, such as climatic change, changes in productivity, and
flood and drought events. The consequences of biotic homogenization in freshwater
ecosystems are less explored than its causes, despite its severe implications, such as
lesser resistant/resilient communities, loss of ecosystem functions, and higher
vulnerability to diseases. Finally, biotic homogenization is a complex process that
requires attention in conservation strategies, especially because forecasts suggest that
freshwater biotas will continue to become more homogeneous in the future.
Keywords: Beta diversity / Aquatic communities / Similarity / Dams / Biological
invasions
28
1 Overview
Biodiversity is declining at an accelerated rate due to human activity. However, human
influences may not only reduce the number of species, but also increases similarity
among biotas, for instance by losing rare species and spreading common species in a
process recognized as biotic homogenization (McKinney and Lockwood, 1999).
Paleontological records suggest that biotic homogenization events occurred in the past,
such as the Great American Biotic Interchange, when the formation of the Panamanian
land bridge allowed the mixing of species between North and South America (Olden,
2006). However, these past events seem localized and isolated compared to current ones
(Olden and Poff, 2004). Nowadays, the biotic homogenization is considered so alarming
that this contemporary period is recognized as “New Pangea” or “Homogecene” (Olden,
2006).
Charles Elton was probably the first to recognize the process of biotic
homogenization (Olden, 2006). In his book The ecology of invasions by animals and
plants published in 1958, Elton suggested the breakdown of Wallace’s faunal realms
due to human-mediated dispersal among continents. In fact, current evidences suggest
that Wallace’s six classic faunal realms defined by dispersal limitation may be replaced
by only two defined by climate (i.e., temperate or tropical) (Capinha et al., 2015).
However, McKinney and Lockwood (1999) were responsible for the first formal
definition of biotic homogenization, related to the “replacement of local biotas with
nonindigenous species, usually introduced by humans”. According to McKinney and
Lockwood (1999), biotic homogenization occurs when a disturbance promotes the
geographic expansion of some species (“winners”) and the geographic reduction of
others (“losers”).
29
Many studies reviewed different aspects of biotic homogenization, such as its
definition and quantification (Olden and Rooney, 2006), mechanisms (Oden and Poff,
2004), conservation strategies (Olden, 2006) and its ecological, evolutionary (Olden et
al., 2004) and human (Olden et al., 2005) consequences. Particularly for freshwater
ecosystems, Rahel (2002) summarized the main causes of biotic homogenization. He
focused mainly on studies using fish in North America and investigated some
anthropogenic causes of biotic homogenization. This study is intended to fill some gaps
from the Rahel (2002) review. In particular, other causes of biotic homogenization are
reviewed, not only anthropogenic causes, and bias is avoided for a single region or
biological group. An attempt is made to understand how freshwater biotas become more
homogenous and what the consequences of this are. Different types of biotic
homogenization are defined along with ways to measure it. Biotic homogenization
patterns on different spatial and temporal scales are discussed and the main natural and
anthropogenic causes of biotic homogenization in freshwater systems are investigated.
Finally, some possible consequences of biotic homogenization from various
perspectives are discussed.
2 Defining biotic homogenization types
In biotic homogenization, some biological differences are lost (Olden et al., 2011).
Biotas may become more similar in taxonomic, functional, phylogenetic, and genetic
features (see Fig. 1 for a summary of biotic homogenization types). Taxonomic
homogenization is the most common form of biotic homogenization, defined as the
increase of similarity in species composition among communities (Olden and Rooney,
2006). Higher similarity among fish communities in dams compared with in free river
stretches (Clavero and Hermoso, 2011), and among zooplankton communities during
30
flood pulses (Bozelli et al., 2015) are some examples of taxonomic homogenization in
freshwater ecosystems.
Figure 1. Summary of biotic homogenization types among communities and among
populations.
However, the loss and gain of species driving biotic homogenization are not
random and may be influenced by species features (McKinney and Lockwood, 1999).
More sensitive species may be replaced by more tolerant species following
environmental change (McKinney and Lockwood, 1999; Olden and Rooney, 2006;
Olden et al., 2011). This replacement may lead to a functional homogenization (i.e.,
increasing species features similarity). For instance, fish communities were more
functionally similar over the years due to entry of non-native species (Pool and Olden,
2012).
Communities may also become more homogenous in a phylogenetic way.
Phylogenetic homogenization (i.e., increased relatedness among species) may occur, for
instance, by: (i) conservation of traits that provide tolerance to some environmental
change, or (ii) the entry of non-native but phylogenetically related species. One example
in freshwater systems is the hypothesis of phylogenetic homogenization among native
frog communities in ponds invaded by a non-native frog (Both and Melo, 2015).
31
Phylogenetic and taxonomic similarities differ mainly because the last ignore
phylogenetic relatedness (i.e., species as independent units) while the former do
consider it (i.e., species as not independent units). From a phylogenetic perspective, four
different species belonging to same family correspond to a less diverse community than
four different species all belonging to different families. The taxonomic perspective
does not make such distinction.
Additionally, individuals in a population are not identical and may vary in
features related to morphology, behavior, or physiology (Bolnick et al., 2011).
Therefore, the variability of phenotypic features of individuals in a population (e.g.,
body size or mouth morphology) may also be investigated in the biotic homogenization
context. One hypothetical example: individuals of one fish species could have a high
variability of morphological or behavioral traits related to feeding in unimpacted
streams due to the high variety of available resources. However, in modified streams,
the variability of these traits could be reduced due to the low diversity of available
resources.
The decrease of genetic variability within and among populations can also lead
to genetic homogenization (Olden et al., 2004; Olden and Rooney, 2006). The main
mechanisms underlying genetic homogenization involve intentional translocation of
populations, introduction of species outside their original distribution area, and the
bottleneck effect due to drastic reduction of population size (Olden et al., 2004).
Although genetic homogenization is poorly investigated in freshwater ecosystems, this
process can result in significant ecological and evolutionary consequences (see the
section “Concluding remarks and perspectives”).
3 Measuring biotic homogenization
32
Biotic homogenization may be quantified by many ways. One strategy is to quantify
increasing similarity among biotas over time (Olden et al., 2004; Olden and Rooney,
2006). For that, similarity among biotas is calculated at a given time (i.e., historical
period), and after an interval of time (i.e., current period) (e.g., Vitule et al., 2012; Daga
et al, 2015; Miyazono et al., 2015). However, biotic homogenization may also be
measured by comparing the similarity between biotas subject and not subject to some
homogenizing factor at the same time period (e.g., impacted vs. unimpacted streams;
Siqueira et al., 2015).
Biotic homogenization among communities may be quantified by beta diversity
(e.g., variability among communities) in terms of functional, phylogenetic, and
taxonomic composition. Beta diversity may be calculated using different metrics (some
of the most used are Jaccard, Sørensen, and Bray-Curtis). The choice of metric is
important to quantify biotic homogenization because they may capture different aspects
of similarity among communities. For example, Siqueira et al. (2015) investigated
taxonomic homogenization of aquatic insects among modified streams using the
Jaccard, Gower and Manhattan indexes. However, they only found biotic
homogenization using the Manhattan index that emphasized the differences in relative
abundances of species. Therefore, in this case, the highest similarity among modified
streams was attributed to changes in the relative abundances of species rather than the
simple presence or absence of species (Siqueira et al., 2015).
It is also important to take into account richness differences among communities
in order to quantify biotic homogenization. One reason is because the dissimilarity
among communities measured by traditional indices (e.g., Jaccard and Sørensen) may
arise by replacement and species richness difference among communities (Baselga
2010, 2012). For example, beta diversity may remain similar between historical and
33
current periods because an increased difference in species richness between time
periods can obscure the fact that the assemblages have become more similar due to the
loss of unshared species (Baeten et al., 2012; Angeler, 2013). In this way, since
dissimilarity indexes may be affected by richness differences, obviously the detection of
biotic homogenization may also be affected.
While taxonomic beta diversity can be measured by the proportion of shared
species, phylogenetic and functional dissimilarities can be quantified by the proportion
of shared branches in a functional or phylogenetic dendrogram (Graham and Fine,
2008). Indices used to calculate taxonomic beta diversity (e.g., Jaccard and Sørensen)
could be adapted to calculate functional and phylogenetic beta diversity (e.g.,
phyloSør). The decrease of phenotypic trait variability among individuals (e.g.,
morphological and behavioral features) may be quantified by measurements of variance
and the standard deviation of some feature. Finally, genetic homogenization can be
quantified from genetic composition as allelic frequency, percentage of polymorphic
loci or average heterozygosity (Olden and Rooney, 2006).
4 How spatial and temporal scales affect homogenization of freshwater
biotas
A better understanding of community assembly often depends on the spatial or temporal
scales employed, and the scale perception depends on species features (e.g.,
geographical range and life cycle) (Wiens, 1989). The relationship between beta
diversity and spatial scale is dependent on two scale components: spatial extent (i.e.,
total sampled area), and spatial grain (i.e., sample unit size) (Wiens, 1989; Barton et al.,
2013). On the one hand, beta diversity tends to increase with higher spatial extent
(Barton et al., 2013; Spasojevic et al., 2016), mainly due to higher environmental
34
heterogeneity and dispersal limitation (Nekola and White, 1999). On the other hand,
beta diversity tends to decrease with increasing spatial grain (Barton et al., 2013;
Spasojevic et al., 2016) due to high probabilities of recording introductions and lower
probabilities of recording extirpations (Olden, 2006). In sum, we could expect higher
levels of biotic homogenization among coarser spatial grains and lower spatial extents
(Olden, 2006). Moreover, using political units (i.e., states, provinces, countries) as the
observation unit may result in underestimated biotic homogenization because natural
and biogeographic barriers (e.g., mountains and watersheds) that define the historical
distinctiveness of a region are not considered (Olden, 2006).
More specifically for freshwater ecosystems, biotic homogenization at different
spatial scales has resulted in contrasting patterns. Fish communities in reservoirs were
more homogeneous taxonomically among sub-catchments and became more different in
the same sub-catchment over time, indicating that fauna was more concordant in space
than in time (Daga et al., 2015). Fish communities were more similar among Canadian
provinces and more dissimilar among eco-regions of a single province (i.e., more
homogenous in a larger extent and grain) (Taylor, 2010). Finally, communities of
benthic invertebrates were more homogeneous due to eutrophication both at local and
regional scales (i.e., without difference among employed scales) (Donohue et al., 2009).
Investigating biotic homogenization across time may indicate different processes
acting in different periods. For instance, fish communities in reservoirs were more
dissimilar between historical periods but became more homogeneous in a more current
comparison (Petesse and Petrere, 2012). This phenomenon may occur, for example,
when non-native species initially invade only some communities (i.e., tendency to
differentiate biota) but later they are established across all the metacommunity (i.e.,
tendency to homogenize the biota). Indeed, biotic differentiation arising from the entry
35
of non-native species can lead to biotic homogenization (Toussaint et al., 2014), a
process that demands caution because it is only understood through temporal
monitoring. Furthermore, consequences of anthropogenic changes are not always
immediate. Following a disturbance event, local species extinction may take some time,
a delay known as "extinction debt" (Kuussaari et al., 2009). Therefore, the historical
legacy of a disturbance can also influence contemporary patterns of biotic
homogenization in freshwater ecosystems (as demonstrated in terrestrial ecosystems by
the influence of historical agriculture in understory plant beta diversity; Mattingly et al.,
2015).
5 Causes of freshwater biotic homogenization
Biotic homogenization in freshwater systems may derive from anthropogenic and
natural causes. Rahel (2002) reviewed the main causes of freshwater homogenization
related only to anthropogenic activities, such as non-native species introduction,
damming, land use, and urbanization. Here, other possible causes of biotic
homogenization in freshwater systems are added, natural or anthropogenic, such as
productivity, climatic changes, drought, and flood. Different causes of biotic
homogenization may act together (e.g., non-native species establishment favored by
dams (e.g., Johnson et al., 2008) or by climatic changes (Rahel and Olden, 2008)).
Although the causes of biotic homogenization are diverse, the mechanisms that generate
biotic homogenization are species entry and/or extinction, or increase and/or decrease of
species range, usually associated to some natural or anthropogenic environmental
change (Rahel, 2002) (Fig. 2).
5.1 Introduction of non-native species
36
Non-native species introduction seems to be the most studied and widespread cause of
biotic homogenization in freshwater ecosystems. The introduction of non-native species
can either increase similarity when the same species invade communities (i.e., biotic
homogenization) or decrease similarity when different species are established among
communities (i.e., biotic differentiation) (Rahel, 2002). The establishment of non-native
species can also increase the similarity among communities indirectly if the introduction
drives the extinction of native species unshared among communities (e.g., by predation
or competition) (Rahel, 2002; Olden and Poff, 2003).
The introduction of non-native species in freshwater ecosystems is usually
mediated by overcoming geographical barriers at different scales, such as oceans (e.g.,
ballast water of ships) or high waterfalls between locations in a same river (e.g.,
damming) (Rahel, 2007). Many studies have identified biotic homogenization due to the
introduction of non-native species for different biological groups in freshwater
ecosystems, such as fish (Olden & Poff, 2012; Toussaint et al, 2014; Daga et al, 2015),
benthic invertebrates (Sardiña et al., 2011), and floodplain forest understories (Johnson
et al., 2014). The introduction of non-native species is usually facilitated by other
causes of biotic homogenization (see below).
5.2 Damming
One well-known effect of damming is the homogenization of river flow (Poff et al.,
2007). Consequently, communities may also become more homogeneous, as
highlighted in the title of a paper by Moyle and Mont (2007): "Homogeneous rivers,
homogeneous faunas". River flow homogenization may act synergistically with other
changes induced by dams, such as reduction of sediment flow, river bed simplification,
reduction of connectivity among the sub-catchments of a floodplain, changes in thermal
37
regime (Poff et al., 2007), reduction of the intensity and duration of flood pulses (Souza
Filho, 2009), and facilitation of invasion of non-native species (Johnson et al., 2008).
Moreover, many native species are locally extinct due to new environmental conditions
imposed by the reservoirs (e.g., migratory fish or species that do not tolerate lentic
conditions; Agostinho et al., 2016).
The relationship between biotic homogenization and damming was investigated
for different biotas and using different techniques. For instance, fish fauna was found to
be more homogeneous among reservoirs when compared to free river stretches (Clavero
and Hermoso, 2011; Pool and Olden, 2012). Comparing historical and contemporary
periods, fish communities were more homogeneous among stretches of a river above a
dam but more differentiated in sections below the dam (Glowacki and Penczak, 2013).
Dams may reduce connectivity among habitats because they represent a new
barrier to the migration of some species. However, dams may connect habitats that were
originally separated by flooding large natural barriers. For instance, Seven Falls
(Parana, Brazil) was a large barrier to the dispersion of Paraná River fishes;
consequently, fish compositions above and below the falls were very dissimilar (Julio-
Junior et al., 2009). However, after the flooding of the Seven Falls by the Itaipu Dam,
fish communities above and below the dam became more similar than before the
damming (Vitule et al., 2012). Another interesting example is small dams removal.
Kornis et al. (2015) observed that fish fauna between portions upstream and
downstream of the dam became more similar after dam removal, because some
opportunistic species that favored more lentic conditions and the warmer water in the
lower portions colonized the upper portions.
5.3 Land use
38
The detrimental consequences of inadequate land use are not restricted to loss of native
vegetation (Lake et al., 2010). In aquatic ecosystems, land use may increase the
sedimentation and the entry of nutrients (e.g., N and P), cause water pollution by heavy
metals, promote habitat simplification, reduce the shading and consequently increase
water temperature and decrease dissolved oxygen and organic matter input from
riparian vegetation (Allan, 2004). One of the main biological consequences of land use
in freshwater ecosystems is the loss of more sensitive species and the expansion of more
tolerant ones (e.g., Scott & Helfman, 2001; Lougheed et al., 2008), which may
homogenize the biota. Land use can increase similarity among biotas both in lotic (e.g.,
stream macroinvertebrates; Siqueira et al., 2015), and in lentic ecosystems (e.g.,
macrophytes and zooplankton in floodplains; Lougheed et al., 2008). Moreover,
different land uses (e.g., pasture, agriculture, and forestry) can lead to different patterns
of similarity depending on the impact intensity (Siqueira et al., 2015).
As cities are built only to support human needs, they are very similar to each
other and restrictive for most native species (McKinney, 2006). Human settlement
introduces, accidentally or intentionally, many non-native species, and provides
favorable conditions for their establishment (McKinney, 2006). Urbanization may
homogenize aquatic biota via the establishment of cosmopolitan non-native species and
the extirpation of unique native species in water bodies (Rahel, 2000; Marchetti et al.,
2006, but see Barboza et al., 2015). Moreover, urbanization may have effects not only
on a local scale (i.e., loss of species due to deforestation), but also regional (i.e., spread
of pollutants) and even global scales (i.e., urban centers as the most responsible for
greenhouse gas emissions that may increase water temperature) (Grimm, 2008).
5.4 Productivity
39
The effects of productivity (usually measured as an increase of N, P and/or chlorophyll-
a) on similarity in freshwater ecosystems are varied. Chase (2010) found biotic
homogenization among low productivity experimental ponds due to deterministic
processes that allowed only the establishment of a few species in most ponds. However,
an increase of nutrients homogenized benthic invertebrate communities within and
between Irish lakes (Donohue et al., 2009). Fish communities from Danish lakes also
became more similar as a consequence of homogenization of benthic habitats exploited
by fish due to eutrophication (Menezes et al., 2015).
Artificial and rapid increases of nutrients may act as a deterministic filter
allowing only a few species to establish among eutrophic environments (Donohue et al.,
2009). However, very low productivity could also act in the same way as a strong filter.
In this way, initial nutrient content and velocity of eutrophication may explain the
contrasting results found among studies in freshwater ecosystems (Donohue et al.,
2009).
5.5 Climatic changes
Although the global climate has undergone natural changes across geological time,
human actions are accelerating this process, which is predicted as one of the major
threats to biodiversity in near future scenarios (Sala et al., 2000). The main climatic
changes predicted involve, on a local scale, changes in climatic conditions (e.g.,
temperature increases, rainfall modifications), changes in climate extremes (e.g., drastic
droughts and floods), and changes in seasonality (e.g., delay in starting seasons) (Garcia
et al., 2014). Some species may adapt to new environmental conditions by increasing,
decreasing or changing their distributional range, but may also suffer a decrease in their
abundances or even become locally extinct (Ackerly et al., 2010). All these mechanisms
40
could lead to biotic homogenization. Here in this section, I focus in global warming.
Fish from the North Atlantic, for example, became more similar due to changes in their
range driven by the increase in seawater temperature over the period 1986 to 2013
(Magurran et al., 2015).
More specifically for freshwater ecosystems, climatic changes may raise water
temperatures, increase climatic extremes (e.g., drastic flood or drought; see next
sections) and alter the flow of streams (Poff et al., 2007). However, few studies have
investigated the relationship between biotic homogenization and observed and projected
climatic changes in freshwater ecosystems. For example, fish in streams under a global
warming scenario may become more similar taxonomically and functionally due to
increased colonization opportunities due to climatic change (Buisson and Grenouillet,
2009). However, the increase in water temperature in Swedish lakes and rivers over 34-
years did not change the composition of aquatic invertebrates (Burgmer et al., 2007).
5.6 Flood
Floodplains are systems with high environmental heterogeneity and high biodiversity in
terms of aquatic and terrestrial species (Agostinho et al., 2004). Hydrological regimes,
characterized by periods of high and low water, are a key mechanism in these floodplain
river systems (Junk et al., 1989; Thomaz et al., 2004). During the drought period, many
aquatic habitats (lakes, channels, wetlands) remain isolated and local forces (i.e.,
environmental heterogeneity, biotic interactions and water re-suspension in the case of
shallow lakes) may become more evident (Thomaz et al., 2004; Thomaz et al., 2007;
Bozelli et al., 2015). During the flood period, high water levels may connect habitats
and, as a consequence, increase the exchange of water, sediment, nutrients and
organisms promoting more homogenous habitats (Thomaz et al., 2007, Bozelli et al,
41
2015). Flooding is a particular mode of homogenization because it is seasonal in nature
and somewhat predictable.
Many streams and small rivers are subjected to flash floods (i.e., very quick with
an intense increase of discharge). Flash floods may facilitate species downstream drift
or even lead to local extirpations. In this sense, passive emigration after flooding could
be an additional mechanism for biotic homogenization among river reaches regarding
species entry and extinctions. Flash floods facilitated the downstream dispersal of
introduced fishes in the headwater streams of the Atlantic Forest, which could mix the
fish fauna and increase similarity between the headwaters and larger rivers (Magalhães
and Jacobi, 2013).
5.7 Drought
Extreme drought events may also lead to biotic homogenization. Here, the main
mechanisms are related to environmental restrictions imposed by droughts. For
example, communities within experimental ponds subject to a severe drought were
more homogeneous than communities that did not suffer drought (Chase, 2007). The
environmental severity imposed by drought acted as a filter allowing only a subset from
the species pool to survive under such conditions, making these ponds more similar
(Chase, 2007).
Dry-land rivers may suffer desertification and salinization in association with
changes in communities. For example, a tributary in North America had its water
discharge reduced, which resulted in a higher salinity in the contemporary period (2010)
in relation to a historical period (1970) (Miyazono et al., 2015). During the historical
period, the main river above the confluence was more saline. In that period, the tributary
contributed to reduce salinity downstream of the confluence and, thus, to increase
habitat heterogeneity in the basin. With the tributary salinization, the main river below
42
the confluence also became saline in the contemporary period. The fish community
responded to the salinity changes and as a result the portions above and below the
confluence became more similar in the current period (2010) in relation to the historical
one (1970). In this way, fish homogenization did not occur due to the entry of non-
native species, but due to the tolerance of native species to salinity. As the stretch below
the confluence became more saline, species sensitive to salinity had reduced abundance
or were excluded, while species tolerant to salinity from the upper reaches also
colonized the lower reaches.
6 Ecological, evolutionary and social consequences of biotic
homogenization in freshwater ecosystems
Biotic homogenization is detected in many biological groups as a result of various
causes. However, most studies do not investigate the ecological and evolutionary
consequences in populations, communities or ecosystems. In this way, there are few
realistic examples regarding the consequences of biotic homogenization and more
speculations about this topic.
Here, two interesting studies on consequences of biotic homogenization in
freshwater ecosystems are highlighted. In the first, homogenization of one community
also affected associated species in a predator-prey relationship. More specifically, fish
community homogenization among Canadian lakes due to the replacement of different
dominant native fishes by only one non-native predator also homogenized zooplankton
community prey (Beisner et al., 2003). In the second example, biotic homogenization of
a host community also affected a parasite community. The affinity of freshwater clam
larvae (Anodonta anatina) is very low with non-native fishes. Consequently, the
homogenization of the fish community driven by the loss of native species and
43
introduction of non-native species reduced the fish species pool suitable for parasitism
by bivalve larvae (Douda et al., 2013).
Olden et al. (2004) suggest many consequences of biotic homogenization in their
review, which may be applied to terrestrial and aquatic systems. Regarding community
homogenization, Olden et al. (2004) suggest consequences, including: (i) high
vulnerability to environmental changes (e.g., extreme drought or pollution) due to
synchrony among communities; (ii) decrease in resilience and/or resistance after some
disturbance; and (iii) damage in ecosystem functions or services (e.g., nutrient cycling
and fish production, respectively). Concerning genetic homogenization, Olden et al.
(2004) suggest that (i) homogenization by intraspecific hybridization can harm the
fitness of individuals for disrupting local adaptations, and (ii) homogenization by
interspecific hybridization may homogenize two previously distinct species that were
adapted to their own environments (see also Agostinho et al., 2010). These
consequences of genetic homogenization may lead small populations to extinction,
especially in fish, where hybridizations are relatively common due to external
fertilization and weak reproductive isolation mechanisms (Olden et al., 2004). Finally,
regarding evolutionary aspects, Olden et al. (2004) suggest that: (i) high gene flow
between populations could hamper allopatric speciation; (ii) hybridization could
increase diversification if the descendants are fertile; and (iii) non-native species
established in a different environment could differentiate phenotypically from the
original population. For more details about speculations described above, see Olden et
al. (2004).
The consequences of biotic homogenization in freshwater ecosystems also
include an economic aspect of lead to losses for ecotourism and fishing (Olden et al.,
2005). For instance, an amateur fisher who had to travel from one region to other one to
44
fish a particular fish species (thus promoting the tourism industry) may, as a result of
biotic homogenization, opt to catch the fish at a location nearer his/her home. In sum,
biotic homogenization affects tourism because "every place is the same, why go
somewhere?" (Olden et al., 2005 (paraphrasing the words of James Kunstler's book
"The Geography of Nowhere")).
7 Concluding remarks and perspectives
The biotic homogenization promoted by anthropogenic disturbances seems to still be
increasing, since species invasions and extinctions, its main drivers, are not decreasing.
For instance, in 42 simulated scenarios of possible fish invasions and extinctions in
global freshwater systems, the forecast is increases similarity among communities in all
simulated scenarios (Villeger et al., 2015). Although the mitigation of invasions of non-
native species and the loss of native species is difficult, it is not impossible through
governmental actions and dissemination of information on their impacts to the whole
population, to avoid other freshwater communities becoming more homogenous (Olden
et al., 2011).
45
Figure 2. Conceptual model summarizing the main causes, mechanisms and
consequences of biotic homogenization in freshwater ecosystems. (A) and (B) are
different fish communities which become more similar due to entry of same species (1)
or extinction of unshared species (2). Each letter inside fishes indicates a different
species.
In summary, freshwater systems may become more homogenous due to many
natural and anthropogenic causes (Fig. 2). Biotic homogenization in freshwater systems
has consequences for communities (e.g., resistance/resilience decrease), populations
(e.g., genetic variability reduction increases susceptibility to diseases), biological
interactions (e.g., predators homogenize prey or homogenization of hosts affects the
parasites), or even ecosystems (e.g., affect ecosystem functions and services) (Olden et
al., 2004). However, the consequences of biotic homogenization, particularly in
freshwaters ecosystems should be further explored. Moreover, different temporal and
spatial scales and different biological groups may show complex processes of increasing
similarity among freshwater communities. Therefore, as we live in a changing and
46
connected world, it is important that the causes and consequences of biotic
homogenization are further investigated.
Acknowledgments
I am grateful to CAPES (Coordenação de Aperfeiçoamento de Pessoal de Nível
Superior) for granting my PhD scholarship. I am very grateful to Luciano F. Sgarbi,
Jean C. G. Ortega, Louizi S. M. Braghin, Lilian P. Sales, Barbara C. G. Gimenez,
Gisele D. Pinha, Natalia C. Lacerda, Robertson Azevedo, Mario Almeida-Neto, João C.
Nabout and Adriano S. Melo for comments in early versions of this manuscript.
References
Ackerly, D. D., Loarie, S. R., Cornwel, W. K., Weiss, S. B., Hamilton, H., Branciforte,
R., Kraft, N. J. B. 2010: The geography of climate change: implications for
conservation biogeography. Divers Distrib. 16, 476–487.
Agostinho, A. A., Thomaz, S. M., Gomes, L. C. 2004: Threats for biodiversity in the
floodplain of the Upper Paraná River: effects of hydrological regulation by
dams. Ecohydrol Hydrobiol 4, 267-289.
Agostinho, A. A., Pelicice, F. M., Gomes, L. C., Júlio Jr, H. F. 2010: Reservoir fish
stocking: When one plus one may be less than two. Natureza & Conservação 8,
103–111.
Agostinho, A. A., Gomes, L. C., Santos, N. C. L., Ortega, J. C. G., Pelicice, F. M. 2016:
Fish assemblages in Neotropical reservoirs: Colonization patterns, impacts and
management. Fisheries Res. 173, 26–36.
Allan, J. D. 2004: Landscapes and riverscapes: the influence of land use on stream
ecosystems. Annu Rev Ecol Evol Syst 35, 257–284.
47
Angeler, D. G. 2013: Revealing a conservation challenge through partitioned long-term
beta diversity: increasing turnover and decreasing nestedness of boreal lake
metacommunities. Divers Distrib. 19, 772–781.
Baeten, L., Vangansbeke, P., Hermy, M., Peterken, G.V., Verheyen, K. 2012:
Distinguishing between turnover and nestedness in the quantification of biotic
homogenization. Biodivers Conserv. 21, 1399–1409.
Barboza, L. G. A., Mormul, R. P., Higuti, J. 2015: Beta diversity as a tool for
determining priority streams for management actions. Water Science &
Technol. 71, 1429:1435.
Barton, P. S., Cunningham, S. A., Manning, A. D., Gibb, H., Lindenmayer, D. B.,
Didham, R. K. 2013: The spatial scaling of beta diversity. Global Ecol Biogeogr
22, 639–647.
Baselga, A. 2010: Partitioning the turnover and nestedness components of beta
diversity. Global Ecol Biogeogr 19, 134–143.
Baselga, A. 2012: The relationship between species replacement, dissimilarity derived
from nestedness, and nestedness. Global Ecol Biogeogr 21, 1223-1232.
Beisner, B. E., Ives, A. R., Carpenter, S. R. 2003: The effects of an exotic fish invasion
on the prey communities of two lakes. J Anim Ecol. 72, 331–342.
Bolnick, D. I., Priyanga, A., Araújo, M. S., Bürger, R., Levine, J. M., Novak, M.,
Rudolf, V. H. W., Schreiber, S. J., Urban, M. C., Vasseur, D. A. 2011: Why
intraspecific trait variation matters in community ecology? Trends Ecol Evol.
26,183-192.
Both, C., Melo, A. S. 2015. Diversity of anuran communities facing bullfrog invasion in
Atlantic Forest ponds. Biol Inv 17, 1137-1147.
48
Bozelli, R. L., Thomaz, S. M., Padial, A. A., Lopes P. M., Bini, L. M. 2015: Floods
decrease zooplankton beta diversity and environmental heterogeneity in an
Amazonian floodplain system. Hydrobiologia 753, 233-241.
Buisson L., Grenouillet, G. 2009: Contrasted impacts of climate change on stream fish
assemblages along an environmental gradient. Divers Distrib15, 613–626.
Burgmer, T., Hillebrand, H., Pfenninger, M. 2007: Effects of climate driven temperature
changes on the diversity of freshwater macroinvertebrates. Oecologia 151, 93–
103.
Capinha, C., Essl, F., Seebens, H., Moser, D., Pereira, H. M. 2015: The dispersal of
alien species redefines biogeography in the Anthropocene. Science 348, 1248-
125.
Chase, J. M. 2007: Drought mediates the importance of stochastic community assembly.
Proc Natl Acad Sci USA 104, 17430–17434.
Chase, J. M. 2010: Stochastic Community Assembly Causes Higher Biodiversity in
More Productive Environments. Science 328, 1388:1391.
Clavero, M., Hermoso, V. 2011: Reservoirs promote the taxonomic homogenization of
fish communities within river basins. Biodivers Conserv. 20, 41–57.
Daga, V. S., Skora, F., Padial, A. A., Gubiani, E. A., Vitule, J. R. S. 2015:
Homogenization dynamics of the fish assemblages in Neotropical reservoirs:
comparing the roles of introduced species and their vectors. Hydrobiologia 746,
327–347.
Donohue, I., Jackson, A. L., Pusch, M. T., Irvine, K. 2009: Nutrient enrichment
homogenizes lake benthic assemblages at local and regional scales. Ecology 90,
3470–3477.
49
Douda, K., Lopes-Lima, M., Hinzmann, M., Machado, J., Varandas, S., Teixeira, A.,
Sousa, R. 2013: Biotic homogenization as a threat to native affiliate species:
fish introductions dilute freshwater mussel’s host resources. Divers Distrib. 19,
933–942.
Elton, C. S. 1958: The ecology of invasions by animals and plants. London: Methuen.
Garcia, R. A., Cabeza, M., Rahbek, C., Araújo, M. B. 2014: Multiple dimensions of
climate change and their implications for biodiversity. Science 344, 486-498.
Głowacki, L. B., Penczak, T. 2013: Drivers of fish diversity,
homogenization/differentiation and species range expansions at the watershed
scale. Divers Distrib. 19, 907–918.
Graham, C. H., Fine, P. V. A. 2008: Phylogenetic beta diversity: linking ecological and
evolutionary processes across space in time. Ecol Lett. 11, 1265–1277.
Grimm, N. B., Faeth, S. H., Golubiewski, N. E., Redman, C. L., Wu, J., Bai, X., Briggs,
J. M. 2008. Global change and the ecology of cities. Science 319, 756 – 760.
Johnson, P. T. J., Olden, J. D., Zanden, M. J. V. 2008: Dam invaders: impoundments
facilitate biological invasions into freshwaters. Frontiers Ecol Environ. 6, 357–
363.
Johnson, S. E., Mudrak, E. L., Waller, D. M. 2014: Local increases in diversity
accompany community homogenization in floodplain forest understories. J.
Vegetation Science 25, 885–896.
Júlio Jr, H. F., Tós, C. D., Agostinho, A. A., Pavanelli, C. S. 2009: A massive invasion
of fish species after eliminating a natural barrier in the upper river Paraná basin.
Neotropical Ichthyol. 7, 709-718.
50
Junk, W. J., Bayley, P. B., Sparks, R. E. 1989: The flood pulse concept in river
floodplain systems. In: Dodge, D.P. (Ed.). Proceedings of the International
Large River Symposium. Can J Fish Aquat Sci 106, 110-127.
Kornis, M. S., Weidel, B. C., Powers, S. M., Diebel, M. W., Cline, T. J., Fox J. M.,
Kitchell, J. F. 2005: Fish community dynamics following dam removal in a
fragmented agricultural stream. Aquat Sci 77, 465–480.
Kuussaari, M., Bommarco, R., Heikkinen, R. K., Helm, A., Krauss, J., Lindborg, R.,
Ockinger, E., Pärtel, M., Pino, J., Rodà, F., Stefanescu, C., Teder, T., Zobel, M.,
Steffan-Dewenter, I. 2009: Extinction debt: a challenge for biodiversity
conservation. Trends Ecol Evol 24, 564–571.
Lake, P.S., Thomson, J. R., Lada, H., Mac Nally, R., Reid, D., Stanaway, J., Taylor, A.
C. 2010: Diversity and distribution of macroinvertebrates in lentic habitats in
massively altered landscapes in south-eastern Australia. Divers Distrib. 16,
713–724.
Lougheed, V. L., Mcintosh, M. D., Parker, C. A., Stevenson, R. J. 2008: Wetland
degradation leads to homogenization of the biota at local and landscape scales.
Freshwater Biol 53, 2402–2413.
Magalhães, A. L. B, Jacobi, C. M. 2013: Asian aquarium fishes in a Neotropical
biodiversity hotspot: impeding establishment, spread and impacts. Biol
Invasions 15, 2157–2163.
Magurran, A. E., Dornelas, M., Moyes, F., Gotelli, N. J., McGill, B. 2015: Rapid biotic
homogenization of marine fish assemblages. Nature Commun 6, 1-5.
Marchetti, M. P., Lockwood, J. L., Light, T. 2006: Effects of urbanization on
California’s fish diversity: Differentiation, homogenization and the influence of
spatial scale. Biol Conserv 127, 2130–2318.
51
Mattingly, W. B., Orrock, J. L., Collins, C. D., Brudvig, L. A., Damschen, E. I.,
Veldman, J. W., Walker, J. L. 2015: Historical agriculture alters the effects of
fire on understory plant beta diversity. Oecologia 177, 507-518.
McKinney, M. L. 2006: Urbanization as a major cause of biotic homogenization. Biol
Conserv. 127, 247–260.
McKinney, M. L., Lockwood, J. L. 1999: Biotic homogenization: a few winners
replacing many losers in the next mass extinction. Trends Ecol Evol 14, 450-
453.
Menezes, R. F., Borchsenius, F., Svenning, J. C., Davidson, T. A., Søndergaard, M.,
Lauridsen, T. L., Landkildehus, F., Jeppesen, E. 2015: Homogenization of fish
assemblages in different lake depth strata at local and regional scales.
Freshwater Biol 60, 745–757.
Miyazono S., Patiño R., Taylor, C. M. 2015: Desertification, salinization, and biotic
homogenization in a dryland river ecosystem. Science Total Environ 511, 444–
453.
Moyle, P. B., Mount, J. F. 2007: Homogenous rivers, homogenous faunas. Proc Natl
Acad Sci USA 104, 5711–5712.
Nekola, J.C., White, P. S. 1999: The distance decay of similarity in biogeography and
ecology. J Biogeogr. 26, 867–878.
Olden, J. D., Poff, N. L. 2003: Toward a mechanistic understanding and prediction of
biotic homogenization. Am Nat 162, 442–460.
Olden, J. D., Poff, N. L. 2004: Ecological processes driving biotic homogenization:
testing a mechanistic model using fish faunas. Ecology 85, 1867–1875.
52
Olden, J. D., Poff, N. R., Douglas, M. R., Douglas, M. E., Fausch, K. D. 2004:
Ecological and evolutionary consequences of biotic homogenization. Trends
Ecol Evol 19, 18-24.
Olden, J. D., Douglas, M. E., Douglas, M. R. 2005: The human dimensions of biotic
homogenization. Conserv Biol 19, 2036–2038.
Olden, J. D. 2006: Biotic homogenization: a new research agenda for conservation
biogeography. J Biogeogr. 33, 2027–2039.
Olden, J. D., Rooney, T. P. 2006: On defining and quantifying biotic homogenization.
Global Ecol Biogeogr 15, 113–120.
Olden, J. D., Lockwood, J. L., Parr, C. L. 2011: Biological invasions and the
homogenization of faunas and floras. Whittaker, R.J., Ladle, R. J (eds)
Conservation biogeography. pp. 224–243. Wiley-Blackwell, Oxford.
Petesse, M. L., Petrere, M. 2012: Tendency towards homogenization in fish
assemblages in the cascade reservoir system of the Tietê river basin, Brazil.
Ecol Eng 48, 109–116.
Poff, N. L., Olden, J. D., Merritt, D. M., Pepin, D. M. 2007: Homogenization of
regional river dynamics by dams and global biodiversity implications. Proc Natl
Acad Sci USA 104, 5732–5737.
Pool, T. K., Olden, J. D. 2012: Taxonomic and functional homogenization of an
endemic desert fish fauna. Divers Distrib 18, 366–376
Rahel, F. J. 2000: Homogenization of fish faunas across the United States. Science 288,
854–856.
Rahel, F. J. 2002: Homogenization of freshwater faunas. Annu Rev Ecol Evol Syst 33,
291–315.
53
Rahel, F. J. 2007: Biogeographic barriers, connectivity and homogenization of
freshwater faunas: it’s a small world after all. Freshwater Biol 52, 696–710.
Rahel, F. J., Olden, J. D. 2008: Assessing the effects of climate change on aquatic
invasive species. Conserv Biol 22, 521–533.
Sala, O. E., Chapin, F. S., Armesto, J. J., Berlow, E., Bloomfield, J., Dirzo, R.,
HuberSanwald, E., Huenneke, L. F., Jackson, R. B., Kinzig, A., Leemans, R.,
Lodge, D. M., Mooney, H. A., Oesterheld, M., Poff, N., Sykes, M. T., Walker,
B. H., Walker, M., Wall, D. H. 2000: Global biodiversity scenarios for the year
2100. Science 287, 1770–1774.
Sardiña, P., Chaves, E., Marchese, M. R. 2011: J North Am Benthol Soc. 30, 1009-1023.
Scott, M.C., Helfman, G. S. 2001: Native invasions, homogenizations, and the
mismeasure of integrity of fish assemblages. Fisheries 26, 6-15.
Siqueira, T., Lacerda, C. G. T., Saito, V. S. 2015: How Does Landscape Modification
Induce Biological Homogenization in Tropical Stream Metacommunities?
Biotropica 47, 509–516.
Spasojevic, M. J., Turner, B. L., Myers, J. A. 2016: When does intraspecific trait
variation contribute to functional beta-diversity? J Ecol 104, 487–496.
Souza Filho, E. E. 2009: Evaluation of the Upper Paraná River discharge controlled by
reservoirs. Braz J Biol 69, 707-716.
Taylor, E. B. 2010: Changes in taxonomy and species distributions and their influence
on estimates of faunal homogenization and differentiation in freshwater fishes.
Diver Distrib. 16, 676–689.
Thomaz S. M., Pagioro, T. A., Bini, L. M , Roberto, M. C., Rocha, R. R. A. 2004:
Limnological characterization of the aquatic environments and the influence of
hydrometric levels. In: Thomaz, S. M., Agostinho, A. A., Hahn N. S. (eds). The
54
upper Paraná river and its floodplain, physical aspects, ecology and
conservation. Leiden, Netherlands: Backhuys Publishers.
Thomaz, S. M., Bini, L. M., Bozelli, R. L. 2007: Floods increase similarity among
aquatic habitats in river-floodplain systems. Hydrobiologia 579, 1–13.
Toussaint, A., Beauchard, O., Oberdorff, T., Brosse, S., Villéger, S. 2014: Historical
assemblage distinctiveness and the introduction of widespread non-native
species explain worldwide changes in freshwater fish taxonomic dissimilarity.
Global Ecol Biogeogr. 23, 574–584.
Villéger, S., Blanchet, S., Beauchard, O., Oberdorff, T., Brosse, S. 2015: From current
distinctiveness to future homogenization of the world’s freshwater fish faunas.
Divers Distrib. 21, 223–235.
Vitule, J. R. S., Skóra, F., Abilhoa, V. 2012: Homogenization of freshwater fish faunas
after the elimination of a natural barrier by a dam in Neotropics. Divers Distrib.
18, 111–120.
Wiens, J. 1989: Spatial scaling in ecology. Funct Ecol 3, 385–397.
55
C
SUBSTRATUM SIMPLIFICATION REDUCES
BETA DIVERSITY OF STREAM ALGAL
COMMUNITIES2
2 Petsch, D. K., Schneck, F., Melo, A. S. 2017. Substratum simplification reduces beta
diversity of stream algal communities. Freshwater Biology, 62: 205–213.
APÍTULO 2
22
56
SUMMARY
1. Reduced species richness with increased habitat simplification is a well-known
relationship in community ecology. However, habitat simplification can also lead to a
reduction in beta diversity if the loss of species is not random. We tested the hypothesis
that beta diversity of periphytic algae is lower among simple than among complex
substrata.
2. We conducted a field experiment using simple (smooth) and complex (rough)
artificial substrata colonized by periphytic algae to calculate beta diversity among each
substratum type. We initially estimated beta diversity using the Jaccard dissimilarity
index and its turnover component. As species richness differed between substratum
types, we also employed the Raup-Crick dissimilarity index that estimates beta diversity
by resampling from the species pool. We also deconstructed the total dataset into three
functional groups based on the position occupied by each species within the periphytic
matrix (low profile, high profile and motile functional groups).
3. Beta diversity estimated using both Jaccard dissimilarity and its turnover component
was higher among simplified substrata for the all-species dataset and for high profile
and motile groups. However, after taking into account differences in species richness
between substratum types using the Raup-Crick index, beta diversity was higher among
complex substrata than among simple ones for the total dataset and for the low profile
group.
4. We emphasize that differences in species richness must be considered for the
quantification of beta diversity, because this might confound the dissimilarity identified
and, consequently, lead to erroneous conclusions.
5. The higher beta diversity among complex substrata might be the result of priority
effects, in which early colonists constrain the establishment of later arriving species,
57
causing each patch to harbor a distinct species composition. Further, algae life strategies
may be an important driver of beta diversity among simple and among complex
substrata, as periphytic algae position in the biofilm may affect their susceptibility to
shear stress. On the one hand, stochasticity in colonization history on complex substrata
may have driven high beta diversity for the low profile group among this type of
substratum. On the other hand, the reduced set of high profile and motile species on
simple habitats may have driven these species to more occasional and rare occurrence,
increasing beta diversity among this type of substratum and resulting in similar beta
diversity among both types of substrata.
6. Priority effects should be most frequent on complex substrata. However, only a
reduced set of species might survive on simple substrata, occupying most of the
available patches and causing beta diversity reduction.
58
Introduction
Increasing species richness with increasing habitat complexity is a well-known
relationship in community ecology (e.g. Schneck, Schwarzbold & Melo, 2011; Pierre &
Kovalenko, 2014; Stein, Gerstner & Kreft, 2014). In general, complex habitats contain a
higher number of species than simple habitats, as they provide a greater variety and
quantity of resources (Pierre & Kovalenko, 2014), different microhabitats, suitable
reproduction sites (Johnson, 2007) and physical refuges against predation (Palmer,
Menninger & Bernhardt, 2010; Kovalenko, Thomaz & Warfe, 2012). Habitat
complexity may also affect beta diversity, i.e. variability in the species composition
among the communities of a given area (Anderson, Ellingsen & McArdle, 2006),
although scarce empirical support exists for this relationship (Heino et al., 2015).
Beta diversity can arise from both deterministic and stochastic mechanisms
(Chase, 2010). Purely deterministic processes occur when resources or conditions create
distinct environments, which favor different species (Chase, 2010). In contrast, purely
stochastic processes include the extinction/colonization dynamic in ecological drift
(Chase, 2007; 2010). For instance, Chase (2010) found that an increased beta diversity
resulting from stochastic effects was more prevalent in productive ponds than in those
with low productivity (similar to found by Chase, 2007; Vannette & Fukami, 2014).
Only a reduced but widespread set of species was able to colonize the low productive
experimental ponds. In contrast, productive ponds were colonized not only by an
increased set of species, but by many infrequent species that persisted probably due to
priority effects. A possible generalization of the findings of Chase (2007, 2010) is that
beta diversity should be high in species-rich environments, such as in complex habitats
(or productive ponds in the study of Chase, 2010), where colonists are derived from a
59
large set of species and where priority effects are more prevalent (Vannette & Fukami,
2014).
In lotic systems, reduction in beta diversity is generally studied as the result of
processes at broad spatial scales, such as damming (Petesse & Petrere, 2012; Daga et
al., 2015) and landscape modification (Siqueira, Lacerda & Saito, 2015). However, beta
diversity may also be affected at fine spatial scales as a result of the loss of habitat
complexity (Hewitt et al., 2010). For instance, the complexity provided by the surface
roughness of substrata, formed by crevices, pits and small projections, plays an
important role in structuring periphytic algal communities (i.e. algae adhering to or
associated with submerged substrata) in streams (Bergey, 2005; Schneck et al., 2011)
by providing refuges and supporting high biomass (Bergey, 2005). Habitat
simplification in lotic environments at the scale of substratum roughness may be caused,
for instance, by concreted and channelized streambeds (Ferreira et al., 1999) or by
siltation due to removal of riparian vegetation (Casatti, Ferreira & Carvalho, 2009),
reducing substratum roughness and, consequently, reducing the number of species
(Schneck et al., 2011). Additionally, these simplified substrata may restrict species
colonization to a reduced set of species, whereas a much wider species pool can
colonize and survive on complex substrata (Schneck et al., 2011) and lead to a distinct
species composition simply by stochastic colonization history and priority effects
(Chase, 2010; Vannette & Fukami, 2014).
The response of periphytic algae to substratum simplification may depend on
their ability to cope with disturbance and resource depletion (Passy, 2007; Lange,
Townsend & Matthaei, 2016) and may be summarized in algal life strategies according
to the position occupied by each species within the three-dimensional periphytic matrix
(Passy, 2007). For instance, species that live in the low layer of the periphytic biofilm
60
(hereafter low profile group), such as prostrate and short-stature species, are
‘disturbance-free’ (Passy, 2007) and thus may be able to colonize and survive on both
simple and complex substrata (Schneck et al., 2011). Conversely, species that occupy
the high layer of the biofilm (hereafter high profile group), e.g. tall-stature erect,
stalked, and filamentous species (Passy, 2007), and motile species may be more
strongly affected by substratum complexity since both groups may benefit from the
protection provided by rough substrata (Schneck et al., 2011).
Beta diversity may be estimated in different ways and for many purposes (for a
review, see Tuomisto 2010a, 2010b). These different methods can produce contrasting
results. Moreover, beta diversity is quantified using observed alpha and gamma
diversity values, which are prone to sampling bias (Tuomisto, 2010b). In fact,
dissimilarity among communities can arise simply due to differences in local species
richness (Baselga, 2010; Chase et al., 2011). This effect can be minimized by the use of
dissimilarity indices that are intended to measure turnover, but which are unaffected by
differences in species richness among communities (Melo, Rangel & Diniz-Filho, 2009;
Baselga, 2010). However, a null model approach may be more effective than common
dissimilarity indexes (i.e. Jaccard and Sørensen dissimilarities) as the former quantifies
how much pairwise community dissimilarities differ from that which would be expected
by chance (Chase et al., 2011). This null model approach is used in the Raup-Crick
index, which estimates dissimilarity as a probability that a pair of samples is non-
identical in species composition (Chase et al., 2011; Oksanen et al., 2015).
We employed an experimental approach to test the hypothesis that small-scale
habitat simplification leads to reduction of beta diversity of periphytic algae. We
investigated beta diversity among simplified and among complex substrata separately
for three algal functional groups (low profile, high profile, and motile). Moreover, to
61
address concerns about richness differences in beta diversity quantification, we tested
our hypothesis using a traditional dissimilarity index (i.e. Jaccard index), an index that
quantifies true turnover among sites (i.e. turnover component of Jaccard dissimilarity)
and then using an index that controls the influence of richness differences between
treatments (i.e. Raup-Crick index).
Methods
Study area
We performed the experiment in Marco stream (1,100 m a.s.l.; 28º36'S; 49º51'W), state
of Rio Grande do Sul, southern Brazil, which is a fourth-order stream located in a
plateau region composed predominantly of natural grassland vegetation with patches of
Mixed Ombrophilous Forest (Araucaria Forest) occurring scattered throughout the area.
The climate is high-altitude subtropical (Cfb), with uniform precipitation throughout the
year (Behling, 2002). Annual mean rainfall ranges from 1400 to 2200 mm and annual
mean temperature ranges from 12 to 18°C, with negative temperatures in winter
(Behling, 2002). The stream has a stony bottom with oligotrophic (Buckup et al., 2007),
clear, and fast-flowing waters characterized by low electrical conductivity (22 µS cm-1)
and mild acidity (pH 6.6). The mean current velocity in the reaches studied (runs and
riffles) was 26 cm s-1 ± 13 cm s-1 during the study period. Stream width varies from 2 to
5 m, and the depth varies from 0.2 to 0.4 m in the reaches studied. The stream has a
natural open canopy without woody riparian vegetation along most of its length, and all
studied reaches are unshaded. This characteristic makes this stream an excellent system
to study periphytic algae, which are often the most important primary producers in
streams, especially in unshaded systems (Biggs, 1996).
62
Experimental design
We designed an experiment using smooth and rough artificial substrata as proxies for
simple and complex habitats, respectively. We left the substrata to be colonized for 45 d
before the first sampling (sampling occurred from May to July 2009). Then, we took
samples on six occasions (every 15 d) in 11 stream reaches (at least 100 m apart from
each other). At each stream reach, we sampled two substrata units of each type (smooth
and rough), which were then pooled and constituted one experimental unit (n = 132).
Acrylic substrata (5 × 5 cm) had either a smooth surface or one containing
parallel crevices for algal colonization (Figure S1A). Complexity is related to variation
in the abundance/density of physical elements (Tokeshi & Arakaki, 2012), i.e. crevices
in our study. Previous studies have shown that algal assemblages remain protected
within crevices that are less than 2 mm wide (Bergey & Weaver, 2004). Accordingly,
we created nine crevices 1 mm wide and 1 mm deep to create the rough surfaces. All
substrata were glued onto flat basaltic paving stones (50 × 50 × 8 cm) that we placed in
each of the 11 stream reaches (Figure S1B). We arranged the substrata on each paving
stone by alternating smooth and rough surfaces ~2.4 cm apart from each other within
six rows and six columns (only 24 substrata out of 36 were used in this study), such that
if the first substratum in a row had a smooth surface, the first substratum in the next row
had a rough surface. The complex substrata were positioned with crevices
perpendicularly aligned to stream flow. Each paving stone contained all substrata
necessary for the six samplings and, thus, we were able to reduce the influence of the
variation in physical variables between smooth and rough substrata within sites and
sampling occasions, since both substratum types were only dozens of centimeters apart
and thus under similar environmental conditions within each of the 11 stream reaches.
No large disturbance was recorded during the period of study.
63
The same experimental data used in our study has been used in two other studies
but with different aims from ours: i) to investigate if richness, density, nestedness and
composition of periphytic algal communities differ between rough and simple substrata
(Schneck et al., 2011); and ii) to investigate if smooth substrata decrease the temporal
persistence of periphytic algal communities (Schneck & Melo, 2013). A third study
used part of the data we use here as well as additional data to investigate if algal
resistance and resilience to a high-flow disturbance are higher on rough than on smooth
substrata (Schneck & Melo, 2012). These aforementioned papers contain additional
details concerning the field experiment.
Biological analysis
We brushed the upper surface of the substrata with a toothbrush to remove the biofilm
and preserved the samples with 4% formaldehyde. We determined periphytic algal
composition by counting 500 cells or units (each unit corresponded to 10-µm-long fine-
celled cyanobacterial filaments) from each experimental unit with an inverted
microscope at 400× magnification. Algae were identified to the lowest practical
taxonomic level, mostly species. Some closely related species which are discernible
only by their reproductive structures (e.g. Oedogonium and Bulbochaete species) or
species that need to be cultivated in controlled cultures (e.g. Stigeoclonium) (John,
2003) were represented here by genus, but for simplicity we refer hereafter to species.
By counting a fixed number of cells/units, we minimized the possible effect of higher
surface area in the rough substrata. To identify diatom species, we examined acid-
cleaned subsamples mounted onto glass slides (using NaphraxTM as mounting medium)
at 1000× magnification through a light microscope (Lowe & LaLiberte, 2007).
64
Information on species composition has been provided elsewhere (Schneck et
al., 2011). However, in order to provide a context for interpretation we provide the
following brief description. The dataset included 92 taxa of periphytic algae; 79 species
on simple substrata (mean richness = 18 species) and 86 species on complex substrata
(mean richness = 26 species). Diatoms were predominant and consisted of 56 species
and 85% of the total cell density in both treatments. The dominant species in both
treatments were the diatoms Achnanthidium minutissimum (Kützing) Czarnecki,
Cocconeis placentula Ehrenberg and Ulnaria ulna (Nitzsch) P. Compère. We also
classified algae in three functional groups: low profile = 43 species, high profile = 26
species, and motile = 23 species, based on information provided by Passy (2007), Lange
et al. (2011), Wagenhoff et al. (2013) and Law et al. (2014) (see Table S1 in Supporting
Information). For more details about species composition, please see Schneck et al.
(2011) and Table S1.
Data analysis
We calculated beta diversity of periphytic algae among simple and among complex
substrata at the 11 stream reaches. Analyses were repeated for the six sampling
occasions. We estimated the dissimilarity using three different metrics (i.e. Jaccard,
turnover component of Jaccard and Raup-Crick) and used the resulting matrix of each
metric to evaluate the multivariate homogeneity of group dispersions (PERMDISP;
Anderson et al., 2006). PERMDISP tests whether the mean within-group dispersion
(measured by the mean distance of samples to its group centroid/median in the full
dimensional space calculated in a Principal Coordinates Analysis (PCoA)) is similar
among the groups (Anderson & Walsh, 2013). We used a restricted permutation design
by strata (i.e., permControl=strata in permutest function), which took the sampling
65
occasion into account, to test the difference in beta diversity among simple and among
complex substrata in the 11 stream reaches. The test was done using 999 permutations.
We used the default option in betadisper function in the vegan package (Oksanen et al.,
2015) for the R Environment (R Core Team, 2014) which uses medians instead of
centroids, but hereafter we termed it centroid as it is more familiar to ecologists. We
performed the analysis using all algal taxa and then separately for each functional group
(low profile, high profile and motile).
We used three different dissimilarity metrics to calculate beta diversity: Jaccard,
turnover component of Jaccard and Raup-Crick indices. Jaccard dissimilarity index is a
common and traditional metric to calculate beta diversity for presence/absence data.
This dissimilarity can be decomposed in two components: one related only to
replacement of species across sites (i.e. turnover component of dissimilarity) and
another related to differences in species richness across sites (i.e. nestedness component
of dissimilarity) (Baselga, 2010). We used both the Jaccard dissimilarity and its
turnover component of dissimilarity to measure beta diversity.
We also used the Raup-Crick dissimilarity index as it estimates beta diversity
after taking into account differences in species richness. This is of particular importance
in our study, because complex substrata had higher species richness than simple
substrata (see Schneck et al., 2011) and, thus could potentially affect estimates of beta
diversity (e.g. using the Jaccard dissimilarity index). For instance, in a species pool (e.g.
stream) containing 20 species, higher beta diversity (using presence/absence data)
should be found among substrata that harbor, on average, 5 species when compared to
those harboring 15 species. By chance, the proportion of shared species should be
higher in the species richest habitat and, thus, produce low beta diversity values. The
Raup-Crick index is obtained by 1) calculating the number of shared species between
66
two samples and the species richness of each one; 2) tabulating all species present in
samples (the species pool) as well as their species occupancy in all samples; 3)
generating a distribution of the number of shared species according to a null model; and
finally, 4) comparing the observed shared species to the distribution of shared species
produced by a null model (Chase et al., 2011). The Raup-Crick index is obtained as a
probability, i.e. the proportion of shared species richness values produced by the null
model that was smaller or equal than the observed shared species richness. We used a
null model where the probability of selecting species is proportional to the species
frequencies. We used the functions betadisper and raupcrick in the vegan package
(Oksanen et al., 2015), and beta.pair in the betapart package (Baselga et al., 2013) in R
Environment (R Core Team, 2014).
Results
Contrary to our hypothesis, we found higher beta diversity among simple (mean
distance to centroid = 0.434) than among complex (mean distance to centroid = 0.391)
substrata using the traditional Jaccard index (F1,130 = 18.175; P = 0.001; Fig. 1). Using
the turnover component we observed a similar pattern as found when using Jaccard
dissimilarity, with higher beta diversity among simple substrata (mean distance to
centroid = 0.373) than among complex substrata (mean distance to centroid = 0.339),
although the magnitude of the difference between substratum types was lower (F1,130 =
8.715; P = 0.004).
67
Fig. 1 Beta diversity among complex and among simple substrata using Jaccard
dissimilarity index. a) Principal Coordinates Analysis (PCoA) plots of periphytic algal
communities among simple (smooth) and among complex (rough) substrata. We
performed a single PCoA ordination, but plotted the six different sampling occasions
separately for clarity. Accordingly, centroids in each plot not necessarily will be in the
center of the polygon for each sampling occasion. Numbers 1 to 6 indicate the sequence
of each sampling occasion, used as strata in the analysis. Black squares with a
continuous line = complex substrata; gray squares with a dashed line = simple substrata.
b) Average distance to the centroid of periphytic algal communities among complex and
among simple substrata. Lines link the substrata on each sampling occasion. Part A is a
two-dimensional representation of a many-axes ordination. Accordingly, the part A
illustrates the main method but is not very good to represent the differences between
treatments as only two-axes are presented. Part B shows differences between simple and
complex substrata much more clearly as it composes the results of all ordination axes.
However, beta diversity among simple substrata was lower than among complex
substrata using the Raup-Crick index that controls the influence of species richness
differences between substratum types. Using this metric, the mean distance to the
centroid in the multivariate space among simple substrata (mean distance to centroid =
0.185) was lower than among complex substrata (mean distance to centroid = 0.278);
(F1,130 = 14.26; P = 0.001; Fig. 2).
68
Fig. 2 Beta diversity among complex and among simple substrata using the Raup-Crick
metric. a) Principal Coordinates Analysis (PCoA) plots of periphytic algal communities
among simple (smooth) and among complex (rough) substrata. We performed a single
PCoA ordination, but plotted the six different sampling occasions separately for clarity.
Accordingly, centroids in each plot not necessarily will be in the center of the polygon
for each sampling occasion. Numbers 1 to 6 indicate the sequence of each sampling
occasion, used as strata in the analysis. Black squares with a continuous line = complex
substrata; gray squares with a dashed line = simple substrata. b) Average distance to the
centroid of periphytic algal communities among complex and among simple substrata.
We observed significantly higher beta diversity among simple substrata than
among complex substrata for the high profile and motile functional groups and similar
beta diversity between treatments for the low profile group using both the Jaccard
dissimilarity and its turnover component (Table 1). However, using the Raup-Crick
dissimilarity, we observed higher beta diversity among complex substrata for the low
profile group and no difference in beta diversity between treatments for the high profile
and motile groups (Table 1).
Table 1. Beta diversity among complex (rough) and among simple (smooth) substrata
using Jaccard, turnover and Raup-Crick dissimilarity indices for three algal functional
groups. Beta diversity was estimated as the average distance to median in an ordination
space.
Algal functional groups
69
High profile Low profile Motile
JACCARD
Mean distance Complex 0.310 0.428 0.461
to centroid Simple 0.378 0.444 0.520
Statistic F(1,130)=23.49* F(1,130)=1.86 F(1,120)=8.36*
TURNOVER
Mean distance Complex 0.251 0.370 0.342
to centroid Simple 0.309 0.368 0.432
Statistic F(1,130)=13.19* F(1,130)=0.01 F(1,120)=9.37*
RAUP-CRICK
Mean distance Complex 0.347 0.349 0.402
to centroid Simple 0.307 0.265 0.423
Statistic F(1,130)=1.89 F(1,130)=10.41* F(1,120)=0.41
* indicates P < 0.01
Discussion
We corroborated our hypothesis that periphytic algal beta diversity is lower among
simple substrata than among complex substrata using Raup-Crick index that avoids the
influence of richness differences between treatments, but rejected our hypothesis using
Jaccard dissimilarity and its turnover component. Simple substrata had fewer species
and that may tend to increase beta diversity using the Jaccard and its turnover metric.
These results emphasize that differences in species richness should be considered in
studies of beta diversity.
The concern about the influence of species richness on beta diversity has only
recently received attention from ecologists (Baselga, 2010; Chase et al., 2011). Indeed,
Chase et al. (2011) demonstrated that higher dissimilarities among communities with
lower species richness (e.g. influenced by local disturbance, predators, productivity and,
in our case, substratum simplification) can be expected by chance only. Our results
corroborate the idea that the Raup-Crick is a suitable metric to compare beta diversity
values among treatments that differ in species richness (Chase et al., 2011).
70
Beta diversity reduction can result from ecological filters, leading to the
dominance of a similar subset of species able to resist conditions that are unfavorable to
many community members (Chase, 2007). In this way, deterministic processes might be
responsible for decreased richness (see Chase et al. (2011)) and decreased beta diversity
of periphytic algae on simple substrata, allowing mainly the same reduced set of stress
tolerant species to occur on these substrata. Indeed, Schneck et al. (2011) using the
same dataset as ours found that the composition of periphytic algae on simple substrata
was a nested subset of species from complex substrata. Similarly, Chase (2010) found
that lower-productivity ponds were a nested subset of higher-productivity ponds, also
indicating a strong environmental filter on community assembly.
The roughness provided by the crevices on the surface of the substrata was,
therefore, responsible for the higher beta diversity among complex than among
simplified substrata when species richness was accounted for. Indeed, the streambed in
natural lotic systems is composed of various irregularities (Taniguchi & Tokeshi, 2004)
that provide refuges against grazing and physical disturbances such as high-discharge
and desiccation events for periphytic algae (DeNicola & McIntire, 1990; Taniguchi &
Tokeshi, 2004; Schneck, Schwarzbold & Melo, 2013; Tonetto et al., 2014; 2015).
Complex substrata provide suitable conditions to a large set of species and might favor
the occurrence of stochasticity in colonization/establishment history. Priority effects
occur when early colonists constrain the establishment of later arriving species.
Stochasticity in colonization history associated to priority effects on complex substrata
may cause each rough substratum to harbor a distinct species composition. This
interpretation is similar to that of Chase (2007, 2010) for higher beta diversity among
permanent and productive ponds and to that of Vannette & Fukami (2014) for suitable
microcosms when compared to microcosms with low resources.
71
Regarding different algal functional groups, results for the Raup-Crick index
indicated higher beta diversity among complex substrata for the low profile group and
no differences for the high profile and motile groups. The low profile group comprises
species of short stature living in the innermost layer of the biofilm, features that provide
more resistance to physical disturbance (Passy, 2007). It could be expected that the
resistance to disturbance of low profile species would lead them to occur similarly on
both simple and complex substrata (e.g. adnately attached and prostrate species in
Schneck et al., 2011), resulting in similar beta diversity between the two types of
habitat. However, our results suggest that refuges on complex substrata may have
favored low profile early colonists (e.g. Miller, Lowe & Rotenberry, 1987) and, thus,
generated stochasticity in colonization history followed by priority effects in this more
benign habitat (Chase, 2007; 2010), driving high beta diversity. These low profile
colonists (e.g. Achnanthidium and Cocconeis) may reproduce fast enough to
monopolize the surface of substrata, inhibiting the colonization by other species
(Goldsborough & Robinson, 1986; see also Steinman & McIntire, 1990). The same
results could be expected especially for the high profile group, since this group may be
favored by crevices (Schneck et al., 2011) and is composed by some typically early
colonists, such as araphid diatoms (DeNicola & McIntire, 1990). However, the reduced
set of high profile and motile species on simple substrata (as shown in Schneck et al.,
2011 for filamentous, erect/stalked – mostly high profile species – and motile species)
may explain our results of similar beta diversity between both substratum types. That is,
a reduced number of species on simple substrata could drive these species to more
occasional and rare occurrence, increasing beta diversity also among this type of habitat
and resulting in similar beta diversity among both substratum types. These results
72
demonstrate that it is interesting to take into account the different life strategies of algal
species when investigating beta diversity.
We were most interested in beta diversity reduction and in its quantification
using the Raup-Crick index. However, it is worth noting an interesting result apparent in
the ordination used to estimate beta diversity. The distance between centroids of the
simple and complex substrata were further apart using the Jaccard index than using the
Raup-Crick index. In other words, species dissimilarities between simple and complex
substrata were more distinct using the Jaccard index. This reduction in dissimilarity
between treatments using Raup-Crick compared to Jaccard index is also apparent in the
coral reefs example presented by Chase et al. (2011), although not in the example of
freshwater ponds used in the same study. It is possible that this effect of the Raup-Crick
index is due to the removal of the effects of differences in species richness on the
estimation of dissimilarity. If so, this would be a way to study turnover-only
dissimilarities among communities and, thus, be equivalent to the approach using
specific turnover indices in ecological studies (Legendre, 2014).
We concluded that (i) differences in species richness must be considered for the
quantification of beta diversity, because they might confound the dissimilarity patterns
identified. Moreover, (ii) beta diversity among simple and among complex substrata
may change among algal functional groups. We also highlight that (iii) habitat
simplification (at least as shown using substratum complexity in this study), which is
one of the main threats to biodiversity, might not only reduce species richness locally,
but homogenize communities across space.
Acknowledgments
73
Luciano F. Sgarbi provided help with the figures. The Coordenação de
Aperfeiçoamento de Pessoal de Nível Superior (CAPES) provided a student fellowship
to DKP. The Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)
provided research grants (476304/2007-5; 474560/2009-0) and research fellowships
(302482/2008-3, 307479/2011-0 and 309412/2014-5) to ASM and a research grant to
FS (474279/2013-8).
References
Anderson M.J., Ellingsen K.E. & McArdle B.H. (2006) Multivariate dispersion as a
measure of beta diversity. Ecology Letters, 9, 683–693.
Anderson M.J. & Walsh D.C.I. (2013) What null hypothesis are you testing?
PERMANOVA, ANOSIM and the Mantel test in the face of heterogeneous
dispersions. Ecological Monographs, 83, 557–574.
Baselga A. (2010) Partitioning the turnover and nestedness components of beta
diversity. Global Ecology and Biogeography, 19, 134–143.
Baselga A., Orme D., Villeger S., Bortoli J. & Leprieur F. (2013) betapart: Partitioning
beta diversity into turnover and nestedness components. R package version 1.3.
http://CRAN.R-project.org/package=betapart.
Behling H. (2002) South and southeast Brazilian grasslands during Late Quaternary
times: a synthesis. Palaeogeography, Palaeoclimatology and Palaeoecology,
177, 19–27.
Bergey E.A. & Weaver J.E. (2004) The influence of crevice size on the protection of
epilithic algae from grazers. Freshwater Biology, 49, 1014–1025.
Bergey E.A. (2005) How protective are refuges? Quantifying algal protection in rock
crevices. Freshwater Biology, 50, 1163–1177.
74
Biggs B.J.F. (1996) Patterns in benthic algae of streams. In: Stevenson R.J., Bothwell
M.L. & Lowe R.L. (eds) Algal ecology: freshwater benthic ecosystems.
Academic Press, San Diego, pp 31–56.
Buckup L., Bueno A.A.P., Bond-Buckup G., Casagrande M. & Majolo F. (2007) The
benthic macroinvertebrate fauna of highland streams in southern Brazil:
composition, diversity and structure. Revista Brasileira de Zoologia, 24, 294–
301.
Casatti L., Ferreira C.P. & Carvalho F.R. (2009) Grass-dominated stream sites exhibit
low fish species diversity and dominance by guppies: an assessment of two
tropical pasture river basins. Hydrobiologia, 632, 273–283.
Chase J.M. (2007) Drought mediates the importance of stochastic community assembly.
Proceedings of the National Academy of Sciences USA, 104, 17430–17434.
Chase J.M. (2010) Stochastic community assembly causes higher biodiversity in more
productive environments. Science, 328, 1388–1391.
Chase J.M., Kraft N.J.B., Smith K.G., Vellend M. & Inouye B.D. (2011) Using null
models to disentangle variation in community dissimilarity from variation in α-
diversity. Ecosphere, 2, 1–11.
Daga V.S., Skora F., Padial A.A., Abilhoa V., Gubiani E.A. & Vitule J.R.S. (2015)
Homogenization dynamics of the fish assemblages in Neotropical reservoirs:
comparing the roles of introduced species and their vectors. Hydrobiologia, 746,
327–347.
DeNicola D.M. & McIntire D. (1990) Effects of substrate relief on the distribution of
periphyton in laboratory streams. 1. Hydrology. Journal of Phycology, 26, 624–
633.
75
Ferreira M.T., Franco A., Catarino L., Moreira I. & Sousa P. (1999) Environmental
factors related to the establishment of algal mats in concrete irrigation channels.
Hydrobiologia, 415, 163–168.
Goldsborough L.G. & Robinson G.G.C. (1986) Changes in periphytic algal community
structure as a consequence of short herbicide exposures. Hydrobiologia, 139,
177–192.
Heino J., Melo A.S. & Bini L.M. (2015) Reconceptualising the beta diversity-
environmental heterogeneity relationship in running water systems. Freshwater
Biology, 60, 223–235.
Hewitt J., Thrush S., Lohrer A. & Townsend M. (2010) A latent threat to biodiversity:
consequences of small-scale heterogeneity loss. Biodiversity and Conservation,
19, 1315–1323.
John D. M. (2003) Filamentous and plantlike green algae. In: Wehr J.D. & Sheath R.G.
(eds) Freshwater algae of North America: ecology and classification. Elsevier, San
Diego, pp 311–352.
Johnson D.W. (2007) Habitat complexity modifies post–settlement mortality and
recruitment dynamics of a marine fish. Ecology, 88, 1716–1725.
Kovalenko K.E., Thomaz S.M. & Warfe D.M. (2012) Habitat complexity: approaches
and future directions. Hydrobiologia, 685, 1–17.
Lange K., Liess A., Piggott J.J., Townsend C.R. & Matthaei C.D. (2011) Light,
nutrients and grazing interact to determine stream diatom community
composition and functional group structure. Freshwater Biology, 56, 264–278.
Lange K., Townsend C.R. & Matthaei C.D. (2016) A trait-based framework for stream
algal communities. Ecology and Evolution, 6, 23–36.
76
Legendre P. (2014) Interpreting the replacement and richness difference components of
beta diversity. Global Ecology and Biogeography, 23, 1324–1334.
Law R.J., Elliott J.A. & Thackeray S.J. (2014) Do functional or morphological
classifications explain stream phytobenthic community assemblages? Diatom
Research, 29, 309–324.
Lowe R.L. & LaLiberte G.D. (2007) Benthic stream algae: distribution and structure.
In: Hauer F.R., Lamberti G.A. (eds) Methods in Stream Ecology. Elsevier,
London, pp 327–356.
Melo A.S., Rangel T.F.L.V.B. & Diniz-Filho J.A.F. (2009) Environmental drivers of
beta-diversity patterns in New-World birds and mammals. Ecography, 32, 226–
236.
Miller A.R., Lowe R.L. & Rotenberry J.T. (1987) Succession of diatom communities on
sand grains. Journal of Ecology, 75, 693–709.
Oksanen J., Blanchet F.G., Kindt R., Legendre P., Minchin P.R., O'Hara R.B., Simpson
G.L., Solymos P., Henry M., Stevens H. & Wagner H. (2015) vegan:
Community Ecology Package. R package version 2.2-1. http://CRAN.R-
project.org/package=vegan.
Palmer A.M., Menninger H.L. & Bernhardt E. (2010) River restoration, habitat
heterogeneity and biodiversity: a failure of theory or practice? Freshwater
Biology, 55, 205–222.
Passy S.I. (2007) Diatom ecological guilds display distinct and predictable behavior
along nutrient and disturbance gradients in running waters. Aquatic Botany, 86,
171-178.
77
Petesse M.L. & Petrere M. (2012) Tendency towards homogenization in fish
assemblages in the cascade reservoir system of the Tietê river basin, Brazil.
Ecological Engineering, 48, 109–116.
Pierre J.I.S. & Kovalenko K.E. (2014) Effect of habitat complexity attributes on species
richness. Ecosphere, 5, 1–10.
R Core Team (2014) R: A language and environment for statistical computing. R
Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-
project.org/
Schneck F., Schwarzbold A. & Melo A.S. (2011) Substrate roughness affects stream
benthic algal diversity, assemblage composition, and nestedness. Journal of the
North American Benthological Society, 30, 1049–1056.
Schneck F. & Melo A.S. (2012) Hydrological disturbance overrides the effect of
substratum roughness on the resistance and resilience of stream benthic algae.
Freshwater Biology, 57, 1678–1688.
Schneck F. & Melo A.S. (2013) High assemblage persistence in heterogeneous habitats:
an experimental test with stream benthic algae. Freshwater Biology, 58, 365–
371.
Schneck F., Schwarzbold A. & Melo A.S. (2013) Substrate roughness, fish grazers, and
mesohabitat type interact to determine algal biomass and sediment accrual in a
high-altitude subtropical stream. Hydrobiologia, 711, 165–173.
Siqueira T., Lacerda C.G. & Saito V.S. (2015) How does landscape modification induce
biological homogenization in tropical stream metacommunities? Biotropica, 47,
509–516.
78
Stein A., Gerstner K. & Kreft H. (2014) Environmental heterogeneity as a universal
driver of species richness across taxa, biomes and spatial scales. Ecology
Letters, 17, 866–880.
Steinman A.D. & McIntire C.D. (1990) Recovery of lotic periphyton communities after
disturbance. Environmental Management, 14, 589–604.
Taniguchi H. & Tokeshi M. (2004) Effects of habitat complexity on benthic
assemblages in a variable environment. Freshwater Biology, 49, 1164–1178.
Tokeshi M. & Arakaki S. (2012) Habitat complexity in aquatic systems: fractals and
beyond. Hydrobiologia, 685, 27–47.
Tonetto A.F., Cardoso-Leite R., Peres C.K., Bispo P.C. & Branco C.C.Z. (2014) The
effects of habitat complexity and hydraulic conditions on the establishment of
benthic stream macroalgae. Freshwater Biology, 59, 1687–1694.
Tonetto A.F., Cardoso-Leite R., Novaes M.C. & Guillermo-Ferreira R. (2015) The
relationship between macroalgal morphological complexity and hydraulic
conditions in stream habitats. Hydrobiologia, 747, 33–41.
Tuomisto H. (2010a) A diversity of beta diversities: straightening up a concept gone
awry. Part 1. Defining beta diversity as a function of alpha and gamma diversity.
Ecography, 33, 2–22.
Tuomisto H. (2010b) A diversity of beta diversities: straightening up a concept gone
awry. Part 2. Quantifying beta-diversity and related phenomena. Ecography, 33,
23–45.
Vannette R.L. & Fukami T. (2014) Historical contingency in species interactions:
towards niche-based predictions. Ecology Letters, 17, 115–124.
79
Wagenhoff A., Lange K., Townsend C.R. & Matthaei C.D. (2013) Patterns of benthic
algae and cyanobacteria along twin-stressor gradients of nutrients and fine
sediment: a stream mesocosm experiment. Freshwater Biology, 58, 1849–1863.
80
SUPLEMENTARY MATERIAL
Figure S1. Experimental design. Simple and complex substrata (A) were glued onto flat
basaltic paving stones (B).
81
Table S1. List of taxa found in the study and assignation to one of the functional groups
(FG) defined by Passy (2007): L = low profile, H = high profile, M = motile.
Taxa FG
CYANOBACTERIA
Heteroleibleinia sp. H
Merismopedia sp. L
Spirulina sp. H
Synechocystis sp. L
Tolypothrix sp. H
Unidentified filamentous cyanobacteria H
CHLOROPHYCEAE
Aphanochaete sp. L
Chlorella vulgaris Beijerinck (1890) L
Chlorococcales sp. 1 L
Chlorococcales sp. 2 L
Desmodesmus armatus (Chodat) Hegewald (2000) L
Scenedesmus sp. 1 L
Scenedesmus sp. 2 L
Stigeoclonium sp. H
Unidentified filamentous green alga H
OEDOGONIOPHYCEAE
Bulbochaete sp. H
Oedogonium sp. H
ZYGNEMAPHYCEAE
Closterium incurvum Brébisson (1856) L
Closterium sp. 1 L
Closterium sp. 2 L
Cosmarium amoenum Brébisson in Ralfs (1848) L
Cosmarium angulosum Brébisson (1856) L
Cosmarium reniforme (Ralfs) W.Archer (1874) L
Cosmarium sp. 1 L
Cosmarium sp. 2 L
Cosmarium sp. 3 L
Cosmarium sp. 4 L
Cosmarium sp. 5 L
Cosmarium sp. 6 L
Euastrum sp. 1 L
Euastrum sp. 2 L
82
Appendix S1. (continued)
Taxa FG
ZYGNEMAPHYCEAE
Pleurotaenium ehrenbergii (Brébisson) de Bary (1858) L
Staurastrum punctulatum Brébisson (1848) L
Staurastrum sp. L
Staurodesmus sp. L
Unidentified filamentous desmid H
BACILLARIOPHYCEAE
Achnanthes sp. L
Achnanthidium exiguum (Grunow) Czarnecki (1994) L
Achnanthidium minutissimum (Kützing) Czarnecki (1994) L
Achnanthidium sp. L
Cocconeis placentula Ehrenberg (1838) L
Cymbella tumida (Brébisson) van Heurck (1880) H
Cymbella sp. H
Encyonema minutum (Hilse) Mann (1990) H
Encyonema cf. silesiacum (Bleisch in Rabenhorst) Mann (1990) H
Epithemia sp. L
Eunotia bilunaris (Ehrenberg) Souza (1999) H
Eunotia faba (Ehrenberg) Grunow (1881) H
Eunotia incisa W. Smith ex Gregory (1854) H
Eunotia praerupta Ehrenberg (1843) H
Eunotia pseudosudetica Metzeltin, Lange-Bertalot & García-Rodríguez (2005) H
Fragilaria capucina Desmazière (1825) H
Fragilaria capucina Desmazière var. mesolepta Rabenhorst (1864) H
Frustulia crassinervia (Brèbisson) Lange-Bertalot & Krammer (1996) L
Frustulia sp. L
Gomphonema angustatum (Kützing) Rabenhorst (1864) H
Gomphonema parvulum (Kützing) Kützing (1849) L
Gomphonema sp. 1 H
Gomphonema sp. 2 H
Gomphonema sp. 3 H
Gomphonema sp. 4 L
Hantzschia sp. L
Lemnicola hungarica (Grunow) Round & Basson (1997) L
Luticola costei Metzeltin & Lange-Bertalot (1998) M
Meridion circulare (Greville) Agardh (1831) L
Navicula angusta Grunow (1860) M
83
Appendix S1. (continued)
Taxa FG
BACILLARIOPHYCEAE
Navicula cryptocephala Kützing (1844) M
Navicula cryptotenella Lange-Bertalot (1985) M
Navicula sp. 1 M
Navicula sp. 2 M
Navicula sp. 3 M
Navicula sp. 4 M
Neidium sp. M
Nitzschia palea (Kützing) Smith (1856) M
Nitzschia sp. 1 M
Nitzschia sp. 2 M
Nitzschia sp. 3 M
Nitzschia sp. 4 M
Pinnularia cf. microstauron (Ehrenberg) Cleve (1891) M
Pinnularia subcapitata Gregory (1856) M
Pinnularia sp. 1 M
Pinnularia sp. 2 M
Psammothidium subatomoides (Hustedt) Bukhtiyarova & Round (1996) L
Psammothidium sp. L
Stauroneis sp. L
Surirella angusta Kützing (1844) M
Surirella tenera Gregory (1856) M
Surirella sp. 1 M
Surirella sp. 2 M
Synedra acus Kützing (1844) H
Tryblionella sp. M
Ulnaria ulna (Nitzsch) Compère (2001) H
84
C
FLOODS HOMOGENIZE AQUATIC
COMMUNITIES ACROSS TIME BUT NOT
ACROSS SPACE IN A NEOTROPICAL
FLOODPLAIN3
3 Manuscrito a ser submetido para a revista Aquatic Sciences em colaboração com K.
Cottenie, J. D. Dias, C. C. Bonecker, A. A. Padial, S. M. Thomaz e A. S. Melo.
APÍTULO 3
22
85
Abstract
Biotic homogenization is usually investigated as a consequence from anthropogenic
pressure. However, natural causes such as flood pulse also increase similarity among
communities. We assessed whether floods homogenize zooplankton and macrophytes
communities in space and time using a long-term data over 16 years in six lakes in the
Upper Paraná River floodplain. Regarding to spatial homogenization, we did not find
lower beta diversity among lakes during flood than during drought events, neither for
macrophytes nor zooplankton. In contrast, regarding the temporal biotic
homogenization of one lake, we found that aquatic macrophytes were more similar
among flood than among drought events. Littoral rotifers and littoral cladocerans had
lower beta diversity among floods than among droughts, while all pelagic groups had
higher beta diversity among floods. We may have not found spatial biotic
homogenization because only very large floods may homogenize communities and/or
because stochasticity on extinction and dispersal promoted by flood events may increase
beta diversity. Lower beta diversity among floods may be related to species bank able to
recolonize a similar set of species each flood event whereas communities follow a more
stochastic trajectory among droughts across time.
Keywords Beta diversity, Macrophytes, Zooplankton, Flood pulse
86
Introduction
Floods are natural drivers of biotic homogenization (i.e., beta diversity decrease)
supposed to seasonally increase dispersal events (Nabout et al. 2009; Penha et al. 2017)
and decrease environmental variability among sites (Thomaz et al. 2007; Bozelli et al.
2015). Increased river level during flood events may connect floodplain lakes
previously isolated, mixing their water, sediment, nutrients and facilitating the passive
dispersal of organisms among them (Junk et al. 1989; Thomaz et al. 2007). Both
mechanisms, (i) high dispersal and (ii) environmental homogenization, may cause
communities to become less dissimilar to each other during floods. This occurs because
(i) high dispersal facilitates species to reach more sites and increase the number of
shared species among them, while (ii) environmental homogenization may filter similar
sets of species across sites (Chase 2007).
The responses of communities to flood pulses may, however, depend on the
system connectivity (e.g. Lopes et al. 2014). In dry periods, dissimilarity among lakes
more connected in a floodplain (e.g., lakes permanently connected to a main river)
should be lower than among those less connected (e.g., lakes only temporally connected
to a main river – hereafter called as isolated) due to higher dispersal possibilities among
the former (Thomaz et al. 2007; Lopes et al. 2014; Lansac-Tôha et al. 2016). During
floods, however, previously isolated lakes may also be connected and receive river
water and colonizing species (Penha et al. 2017). In this way, higher temporal
dissimilarity should be expected for isolated than for permanently connected lakes,
particularly during dry periods. Indeed, connectivity among sites in a system seems
crucial for beta diversity. Experimental using metacommunities of microorganisms
showed that high connectivity drove lower dissimilarity (Carrara et al. 2012; Seymour
et al. 2015). Observational studies of aquatic macrophytes also found lower
87
dissimilarity among connected than among unconnected lakes in a spatial snapshot
(Akasaka and Takamura 2012) or across time (Thomaz et al. 2009).
In addition to system connectivity, the effects of flood pulses on similarity
among communities may also differ according to species traits (Padial et al. 2014; Dias
et al. 2016). Body size and dispersal mode of organisms are some of the main features
affecting dispersal and, consequently, similarity across freshwater sites (e.g., de Bie et
al. 2012; Padial et al. 2014; Petsch et al. 2017). Small organisms tend to be more
frequently carried out by water than large ones (de Bie et al. 2012; Padial et al. 2014;
Dias et al. 2016; but see Jenkins et al. 2007). Small zooplankton such as rotifers can be
more easily carried out than large ones such as cladocerans and copepods (Dias et al.
2016). Also, organisms not attached to substrates may be more susceptible to passive
dispersal during flood pulses than those associated to some substrate (Algarte et al.
2014). For example, free-floating macrophytes may be transported more frequently by
flood pulses than those rooted in the bottom of waterbodies.
Flood homogenization has usually been investigated concerning spatial
similarity, i.e., decrease of beta diversity among sites or habitats (e.g. Thomaz et al.
2007; Bozelli et al. 2015). However, effects of flood homogenization may also occur in
the temporal dimension. Indeed, natural communities are not constant but can change
over short- and long-time scales (Sarramejane et al. 2017; Van Allen et al. 2017). In this
way, it can be hypothesized, for a single site, that inter-annual community dissimilarity
is lower among flood periods than among drought periods. Similar mechanisms driving
low community dissimilarity among different sites due to flood may also drive
decreased dissimilarity across floods in a single site: lower dissimilarity in a site among
floods would be expected, for example, if floods increase environmental homogeneity
across time (Van Allen et al. 2017) or if the high connectivity during floods allows the
88
occurrence of the same set species across time. Indeed, higher environmental similarity
drove higher similarity of phytoplankton community among summers (more
environmentally homogenous) than among winters (less environmentally homogenous)
in a subtropical reservoir (Schneck et al. 2011). In addition, as the high connectivity
during the flood create more opportunities to dispersal (Nabout et al. 2009), most of the
species from the species pool could reach a particular lake every flood event,
homogenizing it also across time.
We assessed whether floods homogenize aquatic communities in space and time
using a long-term data over 16 years. We hypothesized that aquatic macrophytes and
zooplankton communities are less dissimilar (i) among sites during floods than during
droughts (spatial beta diversity) and (ii) among floods than among droughts in a same
site (temporal beta diversity). However, as dissimilarity can change according to system
connectivity and species traits, we investigated whether (iii) dissimilarity is lower
during floods in isolated lakes (investigated only in the temporal approach); (iv)
dissimilarity of free macrophytes is lower among floods than rooted and emergent
macrophytes (investigated in both temporal as spatial approach); and (v) dissimilarity of
pelagic and small zooplankton species (i.e. rotifers) is lower among floods than littoral
and larger zooplankton species (i.e. cladocerans and copepods) (investigated only in the
temporal approach).
Materials and methods
Study area and sampling design
The Upper Paraná River floodplain includes high heterogeneity of habitats such as
rivers, channels, ponds, connected and isolated lakes as well as high biodiversity of
terrestrial and aquatic organisms (Agostinho et al. 2004). The Upper Paraná River
89
floodplain is located between the mouths of Paranapanema and Ivinhema rivers,
between Paraná and Mato Grosso do Sul States in Brazil (Agostinho et al. 2008) (Fig.
1). The climate of the Upper Paraná River floodplain is tropical-subtropical with mean
annual temperature around 22ºC. The rainy period is usually from October to March,
and the dry period usually from June to September. Since 2000, many aquatic
communities, including zooplankton and aquatic macrophytes, have been monitored by
a long-term ecological research in many habitats of the Upper Paraná River floodplain.
Our study is part of this major project (Brazil LTER - site 6 [http://www.peld.uem.br]).
We selected six lakes associated to three different rivers of the Upper Paraná
River floodplain (Paraná, Baía and Ivinhema rivers). The Paraná River is the most
important to water level variation in the Upper Paraná River floodplain. However, the
Ivinhema and Baía rivers also contribute to the inundation of habitats adjacent to their
margins, making their associated lakes somehow functionally distinct. We used a paired
design consisting in two lakes by river, where one lake is permanently connected and
another is temporally isolated to a river. Garças and Osmar, respectively, are the
connected and the isolated lakes from Paraná River; Guaraná and Fechada, respectively,
are the connected and isolated lakes from Baía River; Patos and Ventura are
respectively the connected and the isolated lakes from Ivinhema River.
90
Fig. 1 Map of the Upper Paraná River floodplain. Lakes Garças (1), Guaraná (3) and Patos (5)
are permanently connected to a river, whereas lakes Osmar (4), Fechada (2) and Ventura (6) are
connected to a river only during floods.
Data collection
Data on water level were measured daily in the left margin of the Paraná River using a
gauge at the field station of Nupelia/Universidade Estadual de Maringá in Porto Rico
city (Paraná, Brazil). We observed a high variation in water level of Paraná River across
the 16 years of study (Fig. 2). When the water level of Paraná River is higher than 400
cm, most of the waterbodies in the Upper Paraná River floodplain become connected
(Souza Filho 2009). We used this water level threshold as our flood definition.
91
Fig. 2 Water level of the Paraná River (cm) during the 16 studied years (2000-2015).
The horizontal line represents the flood definition used in our study.
Limnological variables were obtained concomitant to aquatic macrophytes and
zooplankton sampling. In each lake, the following variables were measured: dissolved
oxygen (mg l-1, portable oximeter), water temperature (ºC, thermometer coupled to the
oximeter), electric conductivity (μScm-1, portable potentiometer), pH (portable
potentiometer), total alkalinity (μEql-1), turbidity (NTU, portable turbidimeter), total
chlorophyll-α (μg l-1, Golterman et al. 1978), depth (m), total suspended matter (μg l-
1), inorganic suspended matter (μg l-1), organic suspended matter (μg l-1), total nitrogen
(μg l-1, Mackereth et al. 1978) and total phosphorus (μg l-1, Golterman et al. 1978).
Aquatic macrophytes were recorded in the Upper Paraná River floodplain lakes
during 11 years of monitoring from March 2002 to December 2012, usually quarterly
sampled, except for 2003 (only two months sampled) and 2011 (only three months
sampled). We obtained a total of 246 samples (41 sampling occasions * 6 lakes
sampled). Presence and absence data of aquatic macrophytes were recorded visually
from a boat moving at a constant slow speed along the entire shoreline of each one of
92
the six sampled lakes. For submerged plants sampling, a rake was used from a boat for
10 min. Macrophytes species were identified to the lowest taxonomic level possible
using specialized literature (Cook 1990; Velasquez 1994; Pott and Pott 2000; Lorenzi
2000).
Zooplankton community was recorded across 16 monitoring years from
February 2000 to December 2015 usually quarterly sampled (except for 2001 and 2003
when only two months were sampled), totalizing 360 samplings (60 sampling occasions
* 6 lakes sampled). Zooplankton was collected in the pelagic zone of each lake at a
depth of 0.5–1.5 m, at mornings. Using a motorized pump, 600 l of water per sample
were filtered through a 68-μm mesh plankton net. The samples were preserved in a
formalin solution (4%) buffered with calcium carbonate. Zooplankton species were
quantified (ind m-3) using subsampling with a Hensen-Stempell pipette and counting at
least 10% of the concentrated sample in Sedgewick-Rafter chambers (Bottrell et al.
1976). Rotifers, cladocerans and copepods were identified in species using an optical
microscope and specialized literature (see Lansac-Tôha et al. 2009).
Defining biological groups
Aquatic macrophytes consist of a very diverse group, including algae, mosses, ferns and
mainly, seed-bearing plants. These different species are usually classified into different
life forms, such as: rooted submerged (i.e., completely submerged plants rooted into the
sediment); free-floating (i.e., floating plants on or under the water surface); floating-
leaved (i.e., plants rooted in the sediment but with leaves floating on the water surface)
and emergent (i.e., plants rooted in the sediment with foliage extending into the air). We
classified our macrophytes data in three groups: (i) rooted macrophytes (including
floating-leaved and rooted submerged life forms), (ii) free macrophytes (free-floating on
93
or under the water) and (iii) emergent. We separated emergent from rooted macrophytes
because we believe the last are more associated to water variables than the former, since
usually only a little portion of an emergent plant is underwater.
Regarding to zooplankton, one of the main ecological differences among species
is body size, which is strictly related to passive dispersal ability (i.e., smaller organisms
can be dispersed to larger distances). Accordingly, rotifers are supposed to be better
dispersers than cladocerans, which in turn are supposed to be better dispersers than
copepods (e.g., Dias et al. 2016). Moreover, some zooplankton species have
morphological adaptations (e.g. body shape and type of feed) which facilitate them to
live closer or adhered to aquatic macrophytes (i.e., littoral zooplankton) than others
species (i.e., pelagic zooplankton). Littoral and pelagic species can usually migrate
between both lakes regions (Meerhoff et al. 2007). Accordingly, we separated
zooplankton in six groups: (i) littoral copepods, (ii) pelagic copepods, (iii) littoral
cladocerans, (iv) pelagic cladocerans, (v) littoral rotifers and (vi) pelagic rotifers.
Macrophytes incidence and zooplankton abundance across space and time
We visually explored the species incidence for the three biological groups of
macrophytes and the species abundance for the six biological groups of zooplankton
across time (highlighting sampling occasions and floods events) and space (highlighting
difference among rivers and lake connectivity). As we recorded many species, we
selected only the five most abundant from each one of the biological groups for closer
examination. We added to the figures vertical lines indicating the flood events. This
visual inspection may help to understand how flood events act in species distribution
and, consequently, in beta diversity. We used ggplot2 (Wickham 2009) package in the
R program (R Core Team 2016) to construct these figures.
94
Variance partitioning
We performed a variance partitioning separately for each biological group. Here, we
used as response variable the principal coordinates extracted from a Jaccard
dissimilarity matrix. We used four different explanatory matrices related to time,
environment, lake connectivity and associated river. We got 15 variance components;
four components representing the pure explanation of each explanatory matrix and 11
shared components resulting from the diverse combinations of explanatory matrices.
The first explanatory matrix, related to time, consisted of axes of a Principal
Coordinate Analysis of Neighbor Matrices (PCNM) built from the data samplings. Each
eigenvector generated by PCNM (axes usually called as PCNMs) represents a distinct
temporal pattern. We selected a subset of the axes using the forward selection procedure
of Blanchet et al. (2008). We also constructed another matrix representing flood and
drought events across time. We built it using four different ways: (i) the maximum
water level in the period comprehended by the sampling day and 15 days before; (ii) the
average of the water level between the sampling day and 5 days before; (iii)
categorizing the samplings in flood and drought, when the water level was higher or
lower than 400 cm, respectively; and finally (iv) counting the number of days from the
data sampling to the last flood event occurred (i.e., when the water level was higher than
400 cm). We obtained PCNMs for each of the four matrices above separately. As the
findings did not change across the different ways to represent the flood and drought
events (see Figure S1), we used only the PCNMs applied on the number of days from
the data sampling to the last flood event.
We used environmental variables to construct the second explanatory matrix.
For macrophytes, we used the following variables: dissolved oxygen, pH, conductivity,
95
turbidity, total material in suspension, inorganic material in suspension, organic material
in suspension, chlorophyll-α, total nitrogen and total phosphorus. For zooplankton, we
used the same set of limnological variables used for macrophtyes as well as the matrix
of aquatic macrophytes composition. In order to add the aquatic macrophytes as
explanatory variables to zooplankton, we also selected the same sampling occasions of
macrophytes to perform the variance partitioning of zooplankton (i.e., 2002 to 2012).
We standardized the environmental variables to avoid a higher importance of only one
variable using decostand function with “standardize” method in R program. We also
selected the environmental variables using the method described in Blanchet et al.
(2008).
The third explanatory matrix was composed by a dummy variable built to
represent which river each lake was associated (i.e., Baia, Ivinhema or Paraná). Finally,
the fourth explanatory matrix indicated if the lake was permanently connected or not to
a main river (i.e., connected or isolated). We tested if the pure components of variance
were important using ANOVA constrained by each river (except when testing the
importance of the rivers). We used mainly the vegan package (Oksanen et al. 2015) in
the R Environment.
Beta diversity across space and time
We used three different dissimilarity indexes to estimate beta diversity: Raup-Crick,
Jaccard and its turnover component (Baselga 2010). Although Jaccard dissimilarity is
one of the most commonly used among ecologists, it is not supposed to minimize
richness differences as Raup-Crick and turnover component. In general, all the indexes
produced similar results (Table S2), so we opted to show only Raup-Crick index
96
because it was designed to minimize richness difference and best attained normality and
homogeneity assumptions.
We built a schematic model to help to understand how we investigated if floods
homogenize communities across space and time (Figure 3). We measured spatial and
temporal beta diversity separately for each biological group (i.e., three biological groups
of macrophytes and six of zooplankton). First, we selected flood years when during the
first sampling occasion (usually February or March, end of the rainy season) the water
level was higher than 400 cm between the data of sampling and 15 days before (i.e.,
2002, 2003, 2007, 2009, 2010 and 2011). Then, we selected the third sampling occasion
(usually September, end of the dry season) from the same flood years described above.
When there was no species (or only one or two) recorded in some combinations of year
and lake we not included it in the analyses.
We estimated spatial beta diversity between the connected and the isolated lake
associated to a same river. We calculated beta diversity among lakes during the flood
and during the drought period of each selected year (Fig. 3). To investigate spatial
homogenization and if may depend on biological group, we used a linear mixed-effects
model (LMM) with beta diversity as response variable, hydroperiod, biological group
and river as the fixed effects and year as the random effects. We checked normality and
homogeneity assumptions and estimated LMM parameters using maximum likelihood
and Gaussian distribution. We reported the conditional coefficient of determination
(R2m; variability explained only by fixed effects) (Nakagawa and Schielzeth 2013).
We calculated the temporal beta diversity separately for each lake between each
pair of closest years (i.e., 2002x2003; 2003x2007; 2007x2009; 2009x2010; 2010x2011)
to minimize possible problems with temporal autocorrelation. We measured beta
diversity between floods and between droughts of each of these pairs of years (Fig. 3).
97
We used a linear model to investigate temporal homogenization and if it may depend on
biological group and connectivity. For that, we used beta diversity as response variable,
the hydroperiod, biological group, connectivity and river as the predictors. We
estimated our models using Gaussian distribution, checked normality and homogeneity
assumptions and reported the R2adj. We used vegan and betapart packages (Baselga et
al. 2013) for beta diversity analysis and lme4 (Bates et al. 2015) and lmerTest
(Kuznetsova et al. 2016) packages for linear models.
Fig. 3 Steps to calculate spatial (a) and temporal beta diversity (b). Each grey circle
represents a sampling occasion. The arrows show how beta diversity is calculated in
each scenario. Black circles and squares represent different lakes.
Results
We found 47 aquatic macrophytes species and 374 zooplankton species. Regarding the
aquatic macrophytes, 20 species were emergent, 14 were of the rooted life form and 13
species presented free life form. Regarding to zooplankton community, we found 45
copepods (24 littoral and 21 pelagic species), 88 cladocerans (55 littoral and 33 pelagic
species) and 241 rotifers (152 littoral and 89 pelagic species). Including only the
samples used to calculate beta diversity (see Figure 3), we recorded 46 macrophytes
98
species (37 species in floods and 40 species in droughts) and 281 zooplankton species
(206 species in floods and 187 species in droughts).
We did not find a clear relationship between the incidence of aquatic
macrophytes/abundance of zooplankton and flood events (Fig. 4). Although we found
some zooplankton species with their abundances peaks coincident with large floods
(e.g., sp5, sp9 and sp14), most of the species abundance peaks/fall or incidences seem
not related to flooding (e.g., Oxycarium cubense and Polygonum ferrugineum).
99
100
Fig. 4 Aquatic macrophytes incidence from 2002 to 2012 (A) and zooplankton
abundance (square-rooted transformed) from 2000 to 2015 (B and C) in six lakes of a
Neotropical floodplain. Baia, Ivinhema and Paraná correspond to the main rivers which
the lakes could be associated. Grey vertical lines indicate flood events (when the water
level was higher than 400 cm). Wider vertical lines indicate longer flood periods.
Dashed lines = isolated lakes; continuous lines = connected lakes.
We recorded higher total explanation in variance partitioning for aquatic
macrophytes (mean approximately 33%) than for zooplankton (mean approximately
11%) (Table 1; Fig. S1). As only the shared component between environment and river
for macrophytes had a high explanation (6% for emergent, 8 % for rooted and 10% for
free life forms; Fig. S1), we did not show others shared components. The pure
component of time was the most important for emergent macrophytes and for all
101
biological groups of zooplankton. Environmental variables had low importance for both
aquatic macrophytes and zooplankton. Despite of this low explanation of environmental
variables for zooplankton, it is worth to note that many aquatic macrophytes species
were selected as important environmental variables to all zooplankton biological groups
(Table S1). The lake connectivity and the river associated to the lakes were more
important for aquatic macrophytes. Indeed, the three main rivers are quite different in
terms of concentration of nutrients, turbidity and damming upstream (Thomaz et al.
2004; Roberto et al. 2009). Based in our variance partitioning findings, we defined the
design of our study to measure spatial and temporal beta diversity, taking into account
rivers and years.
Table 1 Relative contribution (%) of pure components from variance partitioning for
each biological group. Bold values mean significant components. We showed in this
table only the pure components. Only the shared component between environment and
river for macrophytes had a high explanation (6% for emergent, 8% for rooted and 10%
for free life forms).
Time Environment Connectivity River Residual
Macrophytes Emergent 13 3 5 4 67
Rooted 2 4 3 11 65
Free 6 1 3 9 66
Zooplankton Copepoda 4 2 1 1 90
Littoral Cladocera 5 2 1 1 90
Rotifera 5 2 1 1 89
Zooplankton Copepoda 5 2 1 1 89
Pelagic Cladocera 5 2 1 0 90
Rotifera 6 2 0 1 89
We did not corroborate our predictions about spatial biotic homogenization
because we did not find lower beta diversity among lakes in flood events than in
drought events for aquatic macrophytes (R2m = 0.07) and zooplankton (R2
m = 0.12;
Table 2; Fig. 5). We only found importance of the river (Table 2), where Paraná River
was different from the others (t = 3.321; P = 0.005).
102
Fig. 5 Spatial beta diversity (using Raup-Crick dissimilarity) of aquatic macrophytes
(a) and zooplankton (b) among connected and isolated lakes during flood and during
drought events.
In contrast, for the temporal biotic homogenization, we found that aquatic
macrophytes were more similar among flood than among drought events (R2adj = 0.146;
P = 0.001; Table 2; Fig. 6a). For zooplankton, differences in similarities between
drought and flood depend on biological groups (R2adj = 0.227; P < 0.001; Table 2).
Littoral rotifers and littoral cladocerans had lower beta diversity among floods than
103
among droughts (Fig. 6b), whereas all pelagic groups had higher beta diversity among
floods (Fig. 6b). Temporal beta diversity was not different among connected and
isolated lakes neither for macrophytes nor zooplankton (Table 2).
Fig. 6 Temporal beta diversity (using Raup-Crick dissimilarity) of biological groups of
aquatic macrophytes (a) and zooplankton (b) among flood and among drought
hydroperiods.
Table 2 Results from linear models on spatial and temporal beta diversity using Raup-
Crick as dissimilarity metric. DF = degree of freedom regarding to used parameters
Variable
Spatial Temporal
DF F P DF F P
104
Macrophytes Hydroperiod 1 0.52 0.473 1 5.27 0.025
Group 2 1.39 0.273 2 8.87 <0.001
Hydroperiod X Group 2 0.08 0.915 2 0.59 0.557
River 2 6.31 0.003 2 0.55 0.575
Connectivity - - - 1 2.81 0.096
Zooplankton Hydroperiod 1 0.08 0.776 1 5.23 0.02
Group 5 1.31 0.265 5 9.29 <0.001
Hydroperiod X Group 5 0.41 0.834 5 9.19 <0.001
River 2 1.28 0.281 2 1.04 0.354
Connectivity - - - 1 0.46 0.494
Discussion
We did not corroborate our hypothesis of spatial biotic homogenization between lakes
due floods. However, we corroborated our hypothesis of temporal biotic
homogenization among floods in a same site for most of the biological groups studied.
We found for temporal biotic homogenization that: (i) all the three groups of aquatic
macrophytes were more similar across floods; (ii) littoral rotifers and littoral
cladocerans tended to present higher similarity across floods while the others groups
(i.e., littoral copepods and pelagic rotifers, cladocerans and copepods) had the opposite
pattern (i.e., higher beta diversity across floods); (iii) we did not find support for the
hypothesis that connected and isolated lakes are different in terms of temporal similarity
neither for macrophytes nor zooplankton.
Our findings did not corroborate spatial flood homogenization hypothesis
supported in others studies (e.g. Thomaz et al. 2007; Bozelli et al. 2015 but see Lopes et
al. 2014). Differences in sampling design, spatial scale and taxa might explain such
conflicting results. Bozelli et al. (2015), for example, measured beta diversity across
many lakes and did not repeat it across years. We designed our study differently,
comparing three pairs of lakes, one connected and another isolated to a same river, and
then repeated it over time. Moreover, flood events are not equal but vary in amplitude
105
and duration (Neiff 1990). It is possible that only the largest floods, which were rare in
the studied period due to the damming upstream the Upper Paraná River floodplain
(Souza Filho 2009), homogenize studied communities spatially, while moderate floods,
as the most of flood events recorded during ours 16 years of monitoring, are not enough
to homogenize them. Finally, another plausible possibility is that the water carrying of
floods may cause stochastic local colonization and extinction, both processes related to
ecological drift that may contribute to increase beta diversity among sites instead of
decrease (Myers et al. 2015; Catano et al. 2018).
Contrarily to spatial beta diversity, we found temporal biotic homogenization
across floods for many biological groups (i.e., all macrophytes life forms, littoral
rotifers and littoral cladocerans). On the one hand, lower temporal beta diversity among
floods hydroperiods could be related to repeated dispersal events across time driven by
inundation. As floods are supposed to create more opportunities to dispersal, most of the
species from the species pool could reach the same lake every flood event,
homogenizing it across time. For example, floods events may carry more zooplankton
organisms from littoral to pelagic zone (Lansac-Toha et al. 2009; Simões et al. 2013),
which if repeated across time may decrease beta diversity of littoral zooplankton among
floods. Moreover, resistance eggs produced by many zooplankton species (e.g., Lopes
et al. 2014) as well as seeds and fragments of macrophytes may create a species bank
able to colonize the same lake across time, and the colonization might be trigged by
inundation events. On the other hand, communities can follow a more stochastic
trajectory during each drought event (Thomaz et al. 2009), where priority effects can act
(i.e., effect of the first colonizers on the following species [e.g., Chase 2010]), making
drought periods more dissimilar across the years. In addition, mainly for aquatic
macrophytes, floods could act as a strong environmental filter where only the most
106
resistant species to inundation remain in a same lake. If the same resistant species set
remain during each flood event, beta diversity can be decreased across time among
inundation periods compared to among drought periods. We could find temporal biotic
homogenization but not spatial biotic homogenization because the resistant species to
the flood and/or colonizers could be the same across time in a single lake but not the
same across lakes. These mechanisms should act in a similar way in connected and
isolated lakes, since we did not find differences among them. Indeed, we also found low
importance of connectivity in variance partitioning, mainly for zooplankton.
Differences in temporal beta diversity across biological groups may be found
because species vary in dispersal ability (de Bie et al. 2012). Rotifers and cladocerans
were the smallest biological groups of zooplankton and potentially with higher dispersal
ability (de Bie et al. 2012; Padial et al. 2014; Dias et al. 2016). In addition, copepods
reproduce only sexually, while rotifers and cladocerans may also reproduce by
parthenogenesis and produce resistance eggs, which may increase their colonization
opportunities across time (Gray and Arnott 2012; Lopes et al. 2014). Such features may
be related to lower beta diversity among floods for littoral rotifers and cladocerans.
However, we did not find higher similarity across floods for pelagic
zooplankton. The pelagic zone is supposed to have lower refuge availability than littoral
zone due to higher macrophytes abundance and diversity in the littoral (Meerhoff et al.
2007). In addition, fish predation could be higher during drought when zooplankton and
fishes are more concentrated in the lakes than during flood. Moreover, tropical pelagic
zooplankton may be larger than littoral zooplankton (Meerhoff et al. 2007). Therefore,
pelagic zooplankton, especially groups composed by largest species more easily
captured (i.e., copepods and cladocerans), could be more similar across droughts if
107
predation is higher in this hydroperiod. Indeed, predators can capture the same set of
species across years homogenizing the community prey (Van Allen et al. 2017).
We concluded that floods did not homogenize macrophytes and zooplankton
communities across space but homogenized macrophytes and some zooplankton groups
across time. We suggested that time should be included in futures studies addressing
flood homogenization. Moreover, we found that temporal flood homogenization may
depend on species features. Finally, we highlight the importance of negative results (i.e.,
as ours that flood did not spatially homogenize communities) to question the generality
of ecological hypothesis and to motivate further studies.
Acknowledgments
DKP thanks the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior
(CAPES) and the Global Affairs Canada – Emerging Leaders in the Americas Program
(ELAP) for providing student fellowships. Our study was supported by the “Long-Term
Ecological Research” (PELD) from the Conselho Nacional de Desenvolvimento
Científico e Tecnológico (CNPq) in Brazil. We are grateful to all members of
Macrophytes, Zooplankton and Limnology Laboratories (Nupelia/ Universidade
Estadual de Maringá) for providing data. DKP thanks Jean Ortega for the help with
linear mixed-effect model. JDD thanks CNPq to provide post-doctoral scholarship.
ASM received a research fellowship from CNPq (proc. no. 309412/2014-5).
References
Agostinho AA, Thomaz SS, Gomes LC (2004) Threats for biodiversity in the floodplain
of the Upper Paraná River: effects of hydrological regulation by dams.
Ecohydrology & Hydrobiology 4: 267-289
108
Agostinho AA, Pelicice FM, Gomes LC (2008) Dams and the fish fauna of the
Neotropical region impacts and management related to diversity and fisheries.
Braz J Biol 68: 1119-1132
Akasaka M, Takamura N (2012) Hydrologic connection between ponds positively
affects macrophyte α and γ diversity but negatively affects β diversity. Ecology
93: 967–973
Algarte VM, Rodrigues L, Landeiro VL, Siqueira T, Bini LM (2014) Variance
partitioning of deconstructed periphyton communities: does the use of biological
traits matter? Hydrobiologia 722: 279–290
Baselga A (2010) Partitioning the turnover and nestedness components of beta
diversity. Glob Ecol Biogeogr 19: 134–143
Baselga A, Orme D, Villeger S, Bortoli J, Leprieur F (2013) betapart: partitioning beta
diversity into turnover and nestedness components. R package version
1.3.Available at: http://CRAN.R-project.org/package=betapart
Blanchet FG, Legendre P, Borcard D (2008) Forward selection of explanatory
variables. Ecology 89: 2623–2632
Bottrell HH, Duncan A, Gliwicz ZM, Grygierek E, Herzig A, Hillbricht-Ilkowska A,
Kurasawa H, Larsson P, Weglenska T (1976) A review of some problems in
zooplankton production studies. Norw J Zool 24: 419–456
Bozelli RL, Thomaz SM, Padial AA, Lopes PM, Bini LM (2015) Floods decrease
zooplankton beta diversity and environmental heterogeneity in an Amazonian
floodplain system. Hydrobiologia 753: 233-241
Carrara F, Altermatt F, Rodriguez-Iturbe I, Rinaldo A (2012) Dendritic connectivity
controls biodiversity patterns in experimental metacommunities. Proc Natl Acad
Sci USA 109: 5761–5766
109
Catano PC, Dickson TL, Myers JA (2018) Dispersal and neutral sampling mediate
contingent effects of disturbance on plant beta-diversity: a meta-analysis. Ecol
Lett 20: 347–356
Chase JM (2007) Drought mediates the importance of stochastic community assembly.
Proc Natl Acad Sci USA 104: 17430–17434
Chase JM (2010) Stochastic community assembly causes higher biodiversity in more
productive environments. Science 328: 1388–1391
Cook CDK (1990) Origin, autecology, and spread of some of the world´s most
troublesome aquatic weeds. In Pieterse AH and Murphy KJ (Eds.). Aquatic
weeds: the ecology and management of nuisance aquatic vegetation. Oxford:
Oxford Science Publications. p. 31-38
de Bie T, de Meester L, Brendonck L, Martens K, Goddeeris B, et al. (2012) Body size
and dispersal mode as key traits determining metacommunity structure of
aquatic organisms. Ecol Lett 15: 740–747
Dias DJ, Simões NR, Meerhoff M, Lansac-Tôha FA, Velho LFM, Bonecker CC
(2016) Hydrological dynamics drives zooplankton metacommunity structure in
a Neotropical floodplain. Hydrobiologia 781: 109–125
Gray DK, Arnott SE (2012) The role of dispersal levels, Allee effects and community
resistance as zooplankton communities respond to environmental change. J
Appl Ecol 49: 1216–122
Golterman HL, Clymo RS, Ohstad MAM (1978) Methods for physical and chemical
analysis of freshwater. Blackwell, Oxford
Jenkins DG, Brescacin CR, Duxbury CV, Elliott JA, Evans JA, Grablow KR, Hillegass
M, Lyon BN, Metzger GA, Olandese ML, Pepe D, Silvers GA, Suresch HN,
110
Thompson TN, Trexler CM, Williams GE, Williams NC, Williams SE (2007)
Does size matter for dispersal distance? Glob Ecol Biogeogr 16: 415–425
Junk WJ, Bayley PB, Sparks RE (1989) The flood pulse concept in river floodplain
systems. In: Dodge DP (Ed.). Proceedings of the International Large River
Symposium. Can J Fish Aquat Sci 106: 110-127
Kuznetsova A, Brockhoff PB, Christensen RHB (2017) lmerTest Package: Tests in
Linear Mixed Effects Models. J Stat Softw 82: 1–26
Lansac-Tôha FA, Bonecker CC, Velho LFM, Simões NR, Dias JD, Alves GM,
Takahashi EM (2009) Biodiversity of zooplankton communities in the Upper
Parana´ River floodplain: interannual variation from long-term studies. Braz J
Biol 69: 539–549
Lansac-Tôha FM, Meira BR, Segovia BT, Lansac-Tôha FA, Velho LFM (2016)
Hydrological connectivity determining metacommunity structure of planktonic
heterotrophic flagellates. Hydrobiologia 781:81–94
Lopes PM, Bini LM, Declerck SAJ,, Farjalla VF, Vieira LCG, Bonecker CC, Lansac-
Toha FA, Esteves FA, Bozelli RF (2014) Correlates of zooplankton beta
diversity in tropical lake systems. Plos One 9: 1-8
Lorenzi H (2000) Plantas daninhas do Brasil: terrestres, aquáticas, parasitas e tóxicas. 3
ed. Nova Odessa: Instituto Plantarum de Estudos da Flora. 608p
Mackereth FYH, Heron JG, Talling JJ (1978) Freshwater Biological Association
Scientific Publication 36: 1-120
Meerhoff M, Iglesias C, De Mello FT, Clemente JM, Jensen E, Lauridsen TL, Jeppesen
E (2007) Effects of habitat complexity on community structure and predator
avoidance behaviour of littoral zooplankton in temperate versus subtropical
shallow lakes. Freshw Biol 52: 1009–1021
111
Myers JA, Chase JM, Crandall RM, Jiménez I (2015) Disturbance alters beta-diversity
but not the relative importance of community assembly mechanisms. J Ecol 103:
1291–1299
Nabout JC, Siqueira T, Bini LM, Nogueira IS (2009) No evidence for environmental
and spatial processes in structuring phytoplankton communities. Acta Oecol 35:
720-726
Nakagawa S, Schielzeth H (2013) A general and simple method for obtaining R2 from
generalized linear mixed-effects models. Methods Ecol Evol 4:133-142
Neiff JJ (1990) Ideas para la interpretacion ecológica del Paraná. Interciencia 15: 424–
41
Oksanen J, Blanchet FG, Kindt R, Legendre P, Minchin PR, O’Hara RB, Simpson GL,
Solymos P, Stevens MHH, Szoecs E, Wagner H (2015) vegan: Community
Ecology Package. R package version 2.2-1. Available at: http://CRAN.R-
project.org/package=vegan
Padial AA, Ceschin F, Declerck SA, De Meester L, Bonecker CC, Lansac-Tôha FA,
Bini LM (2014) Dispersal ability determines the role of environmental, spatial
and temporal drivers of metacommunity structure. PloS One 9: e111227
Penha J, Landeiro VL, Ortega JCG, Mateus L (2017) Interchange between flooding and
drying, and spatial connectivity control the fish metacommunity structure in
lakes of the Pantanal wetland. Hydrobiologia, 797:115-126
Petsch DK, Pinha GD, Takeda AM (2017) Dispersal mode and flooding regime as
drivers of benthic metacommunity structure in a Neotropical floodplain.
Hydrobiologia 788:131–141
Pott VJ, Pott A (2000) Plantas aquáticas do Pantanal. Brasília: Embrapa. 404p.
112
R Core Team (2016) R: A Language and Environment for Statistical Computing. R
Foundation for Statistical Computing, Vienna. Available at: http://www.R-
project.org/
Roberto MC, Santana NF, Thomaz SM (2009) Limnology in the Upper Paraná River
floodplain: large-scale spatial and temporal patterns, and the influence of
reservoirs. Braz J Biol 69:717–725
Sarremejane R, Cañedo-Arguëles M,Prat N, Mykrä H, Muotka T, Bonada N (2017) Do
metacommunities vary through time? Intermittent rivers as model systems J
Biogeogr 44: 2752–2763
Schneck F, Schwarzbold A, Rodrigues SC, Melo AS (2011) Environmental variability
drives phytoplankton assemblage persistence in a subtropical reservoir. Austral
Ecol 36: 839–848
Seymour M, Fronhofer EA, Altermatt FA (2015) Dendritic network structure and
dispersal affect temporal dynamics of diversity and species persistence. Oikos
124: 908–916
Simões NR, Dias JD, Leal CM, Braghin LSM, Lansac-Tôha FA, Bonecker CC (2013)
Floods control the influence of environmental gradients on the diversity of
zooplankton communities in a Neotropical floodplain. Aquat Sci 75: 607–617
Souza Filho EE (2009) Evaluation of the Upper Paraná River discharge controlled by
reservoirs. Braz J Biol 69: 707-716
Thomaz SM, Pagioro TA, Bini LM, Roberto MC, Rocha RRA (2004) Limnological
characterization of the aquatic environments and the influence of hydrometric
levels. In: Thomaz SM, Agostinho AA, Hahn NS (Ed.) The upper Paraná River
and its floodplain, physical aspects, ecology and conservation. Backhuys
Publishers, Leiden
113
Thomaz SM, Bini LM, Bozelli RL (2007) Floods increase similarity among aquatic
habitats in river floodplain systems. Hydrobiologia 579: 1-13
Thomaz SM, Carvalho P, Padial AA, Kobayashi JT (2009) Temporal and spatial
patterns of aquatic macrophyte diversity in the Upper Paraná River floodplain.
Braz J Biol 69: 617-625
Van Allen BJ, Rasmussen NL, Dibble CJ, Clay PA, Rudolf VHW (2017) Top predators
determine how biodiversity is partitioned across time and space. Ecol Lett 20:
1004–1013
Velasquez J (1994) Plantas acuaticas vasculares de Venezuela. Caracas: Universidade
Central de Venezuela, Consejo de Desarrollo Cientifico y Humanistico. 992p
Wickham H (2009). ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New
York
114
SUPLEMENTARY MATERIAL
Figure S1. Results from variance partitioning for each biological group. Relative
contributions (%) of time (T), environment (E), connectivity (C) and the river associated
(R). U= unexplained component. Values lower than 1% are not shown for better
visualization (except for the pure components). In the time component, flood was
represented in four different ways: (a) counting the number of days from the data
sampling to the last flood event occurred; (b) the average of the water level between the
sampling day and 5 days before; (c) categorizing the samplings in flood and drought,
when the water level was higher or lower than 400 centimeters, respectively; and finally
(d) the maximum water level in the period comprehended by the sampling day and 15
days before
115
116
117
Table S1. Selected time and environmental variables in pRDA for each biological
group. Cond= conductivity, tur= turbidity, pho= total phosphorus, oxy= dissolved
oxygen, oms= organic material in suspension, ims= inorganic material in suspension,
tms= total material in suspension, chloro= α-chlorophyll, Ed= Egeria densa, Ps= Pistia
stratiotes, P= Panicum, E= Echinodorus, Pa= Polygonum acuminatum, Pf= Polygonum
ferrugineum, Cf= Cabomba furcata, Pc= Pontederia cordata, En= Egeria najas, W=
Wolfiella, Tg= Thalia geniculata, Hv= Hydrilla verticillata, Ld= Lindernia, Ha=
Hymenachne amplexicatus, Mb= Myriophyllum brasiliensis, Hr= Hydrocotyle
ranunculoides, Lm= Limnocharis, Oc= Oxycarium cubensis.
Time variables (PCNMs) Environmental variables
Macrophytes Emergent 1,2,6,27,3,4,8,25,9,5,36,30 cond, tur, pho, oxy, oms
Rooted 36,38,27,24,2 tur, cond, oxy, chloro
Free 2,36,1,28,22,5,9,30 cond, tur, pho, pH, chloro
Zooplankton Copepoda 1,2,9,4,8,11,7,32,13,5,15 cond, Ed, tur, pho, Ps, P, E, Pa, Cf, Pc
littoral Cladocera 2,1,3,34,16,7,15,6,12,20,26 En, tur, Ed, cond, pH, E, ims, Pf, Cf
Rotifera 1,3,5,2,4,6,8,27,20,10,12,14,35 En, tur, Ed, Pa, oxy, cond, Mb, Hr, Tg, E, W
Zooplankton Copepoda 2,1,3,4,5,6,8,12,14,10,7,20,31
En, tur, Ed, cond, pH, W, E, oxy, Pc, Tg, Ld, Lm,
Ha
pelagic Cladocera 1,4,6,2,3,10,5,32,3,8,15,12,11,27,7
En, Ed, tur, tms, cond, pH, Mb, Oc, Tg, Ld, Lm,
pho
Rotifera 1,3,5,2,4,6,8,27,12,10,35,20,18,14, cond, pH, W, E, oxy, Pc, Tg, Hv, Ha, pho
13,15,26,33
118
Table S2. Results of generalized linear mixed-effects models on spatial and temporal
beta diversity using Jaccard and turnover as dissimilarity metrics. Df= degree freedom.
Period corresponds to flood or drought; group corresponds to biological group (e.g.
littoral rotifers); connectivity corresponds to connected or isolated lakes.
Spatial Temporal
Variable df F P df F P
Macrophytes Period 1 0.12 0.721 1 3.83 0.052
Jaccard Group 2 4.22 0.019 2 0.65 0.521
Period X Group 2 2.31 0.107 2 0.14 0.868
Connectivity - - - 1 2.22 0.138
Zooplankton Period 1 0.16 0.689 1 0.08 0.766
Jaccard Group 5 2.01 0.081 5 4.04 0.001
Period X Group 5 0.74 0.591 5 3.63 0.003
Connectivity - - - 1 2.02 0.155
Macrophytes Period 1 1.82 0.181 1 3.89 0.051
Turnover Group 2 1.29 0.281 2 0.84 0.433
Period X Group 2 0.08 0.919 2 0.18 0.829
Connectivity - - - 1 1.14 0.287
Zooplankton Period 1 0.01 0.924 1 1.11 0.293
Turnover Group 5 6.31 <0.001 5 14.69 <0.001
Period X Group 5 0.23 0.944 2 7.69 <0.001
Connectivity - - - 1 0.01 0.904
119
C
HUMAN LAND-USE DOES NOT HOMOGENIZE
AQUATIC INSECT COMMUNITIES IN BOREAL
AND TROPICAL STREAMS4
4 Manuscrito a ser submetido para a revista Ecological Indicators em colaboração com
T. Siqueira, J. Heino, V. S. Saito, J. Jyrkänkallio-Mikkola, K. T. Tolonen, L. M. Bini,
V. L. Landeiro, T. S. F. Silva, V. Pajunen, J. Soininen e A. S. Melo.
APÍTULO 4
22
120
ABSTRACT
Biological diversity is not uniformly distributed across the globe, and it can be lost
through land-use intensification. Land-use intensification may decrease habitat
heterogeneity as well as increases environmental harshness and may decrease not only
the number of species but also the taxonomic and functional variability among
communities in space causing biotic homogenization, that is, lowering beta diversity.
We sampled aquatic insects from 100 boreal and 100 tropical streams covering a wide
gradient of land use to address two main questions: (1) Is taxonomic and functional beta
diversity higher in tropical than in boreal streams? (2) Does land use decrease
taxonomic and functional beta diversity in both regions (i.e. through environmental
harshness and/or environmental homogeneity)? We found higher taxonomic beta
diversity but lower functional beta diversity among tropical than boreal streams. Our
results did not corroborate the expectation of taxonomic or functional biotic
homogenization due to environmental harshness nor due to lower variability in local
environmental variables mediated by intensive land-use. However, different land-use
effects may increase among-stream habitat heterogeneity generating distinct species
composition among streams. Forested streams may be more benign and similar to each
other, allowing high stochasticity in the colonization/extinction dynamics, which in turn
could generate comparable levels of beta diversity among modified streams. We
highlighted that different mechanisms acting simultaneously in modified and conserved
habitats may cause similar beta diversity along a disturbance gradient.
Keywords: deforestation, freshwater, latitudinal gradient of beta diversity, Brazil,
Finland
121
1. Introduction
One of the most widely documented patterns in ecology is the latitudinal diversity
gradient, that is, regions closer to the equator often harbor higher species richness than
those closer to the poles (Gaston, 2000; Brown, 2014). Many evolutionary (e.g. higher
diversification and lower extinction rates of taxa in tropics), historic (e.g. lower
influence of glaciation events in tropics), area (i.e. larger surface area of the tropics) and
ecological (e.g. higher environmental heterogeneity in tropics) hypotheses were
formulated to explain it (Hillebrand, 2004; Mittelbach et al., 2007; Cilleros et al., 2016;
Rodrigues et al., 2017). Whereas higher species richness in tropics is a well-known
pattern, it is still controversial if beta diversity (i.e. variability of community
composition among sites) is also higher in low latitudes (see Qian and Ricklefs, 2007;
Kraft et al., 2011; Qian and Song, 2013). For example, Qian and Ricklefs (2007) found
lower beta diversity of plants in higher latitudes due historical and climatic variables,
while Kraft et al. (2011) found no differences in plant beta diversity across latitude after
taking into account gamma diversity. A higher beta diversity in the tropics is expected,
for example, due to faster decline of regional richness than local richness with latitude
(Soininen, 2010) and because according to Rapoport’s rule, species ranges tend to be
larger closer to the poles than closer to the tropics due to higher environmental tolerance
of the species (Stevens, 1989), which may decrease beta diversity at high latitudes
(Soininen et al., 2007). However, most of the evidence comes from terrestrial and
marine systems (e.g. Qian and Ricklefs, 2007; Kraft et al., 2011; Qian and Song, 2013),
whereas less is known about freshwater systems in this regard.
Although biological diversity is not uniformly distributed across the globe, it can
be decreased by some general factors acting widely across continents. Land-use change
is a worldwide phenomenon driving species loss across multiple taxa and ecosystems
122
(Sala et al., 2000; Millennium Ecosystem Assessment, 2005). Land-use change does not
only reduce the numbers of native species, but drives biotic homogenization across
biological communities (i.e. decrease in beta diversity) by promoting the expansion of
tolerant species (“winners”) and/or the range reduction or extinction of sensitive species
(“losers”) to such environmental modifications (McKinney and Lockwood, 1999;
Castro et al., 2018). However, the effect of land-use changes on beta diversity is still
poorly studied across different ecosystems (Rodrigues et al., 2013; Siqueira et al., 2015;
Solar et al., 2015; Gossner et al., 2016), and its effects are more difficult to predict than
land-use effects on species richness (Fugère et al., 2016; Sfair et al., 2016).
The outcome of “losers” and “winners” as a consequence of land-use changes
suggests that species respond differently to environmental changes. This happens in part
because species have different traits (Gossner et al., 2016; Jonason et al., 2017); i.e.,
features related to life-history, morphology, physiology or phenology determining
species performance (Diaz and Cabido, 2001; Violle et al., 2007). If land-use
intensification decreases the variability of species traits composition across
communities in space, communities become functionally homogeneous (Olden and
Rooney, 2006). The incorporation of organisms’ sensitivity in terms of both taxonomic
and functional diversity could thus improve our understanding of the response of
communities to land-use changes (Castro et al., 2018). For example, land-use
intensification can decrease more severely the taxonomic than functional beta diversity
if communities are composed of many functionally redundant species (i.e., species with
similar traits) (Sfair et al., 2016). In addition, functional homogenization is alarming
because it may limit the functions and services provided by the biological communities
as well as their response to anthropogenic impacts (Cardinale et al., 2012; Gámez-
Virués et al., 2015).
123
Land-use change is a particularly strong driver of biodiversity loss in streams
(e.g., Marchetti et al., 2006; Siqueira et al., 2015). Two main mechanisms driven by
land use may reduce stream biodiversity: (i) increase of environmental harshness and
(ii) reduction of environmental heterogeneity among stream sites. On one hand,
environmental harshness may decrease species richness and homogenize communities
by selectively filtering the same set of resistant species across local communities
(Chase, 2007; Catano et al., 2017). Streams surrounded by intensive land use may
become a harsh habitat for most species because of increased input of terrestrial
sediments that can cover the streambed, input of nutrients and contaminants, alteration
of flow regimes, decrease of long-term coarse organic material input and change of
channel structure (Allan, 2004; Leal et al., 2016; Castro et al., 2018). However, the
effects of increased nutrient input are contradictory: it may generate high beta diversity
due to increase of stochasticity with the higher productivity (e.g. Chase, 2010) or
decrease beta diversity if only a few species are able to tolerate eutrophic conditions
(e.g. Donohue et al., 2009).
Likewise, reduction of environmental heterogeneity (i.e. difference in
environmental conditions among streams) may decrease beta diversity if species
occurrences are driven by environment (Costa and Melo, 2008). Land-use
intensification may decrease among-stream variability of local conditions (e.g. by
homogenizing substrate cover and water velocity) and, consequently, reduce beta
diversity. Alternatively, heterogeneity of land use among streams (e.g. rural, urban and
forestry streams in a same watershed) may increase beta diversity if different species are
selected by environmental conditions associated with each land use type. Therefore,
land use may drive beta diversity through different pathways depending on specific
features of the landscape (Figure 1).
124
Fig. 1. A graphical framework on how land-use could change taxonomic and functional
beta diversity in streams. Positive and negative signals represent, respectivaly, positive
and negative relationships between variables.
We conducted intensive surveys of aquatic insects in boreal and tropical streams
covering a wide gradient of land use in each region. Aquatic insects are functionally
important in freshwater food webs, are specious and abundant in streams worldwide and
respond quickly to changes in stream conditions, such as water quality, substratum type
and flow regimes (Jacobsen et al., 2008; Kennedy et al., 2016). We addressed two main
questions related to land use effects on beta diversity of tropical and boreal aquatic
insects: (1) Is taxonomic and functional beta diversity higher in tropical than in boreal
streams? (2) Does land use decrease taxonomic and functional beta diversity in both
regions (i.e. through environmental harshness and/or environmental homogeneity)? We
expect to find overall higher taxonomic and functional beta diversity among tropical
than among boreal streams, but decreased beta diversity along a gradient of intensive
land-use in both regions. We expect to find lower taxonomic and functional beta
diversity of aquatic insects in both regions among harsher (e.g. low native forest cover
or high proportion of sand substrates) and among environmentally similar streams (e.g.
125
low heterogeneity in land use or low heterogeneity in local environmental variables) due
to land-use intensification.
2. Methods
2.1. Study design
We sampled 20 Finnish (boreal region) and 20 Brazilian (tropical region) watersheds in
years 2014 and 2015, respectively. In each watershed, we sampled five 2nd to 3rd order
streams, totalizing 200 streams (20 watersheds * 5 streams * 2 regions = 200 streams).
In Brazil, we sampled streams located in southeastern parts of the country between
latitudes 23°49S and 24°20S (approximately 120 km in north-south direction and 70 km
in east-west direction). In Finland, we surveyed streams located in western between
latitudes 60°27N and 65°01N (approximately 500 km in north-south direction and 300
km in east-west direction). Streams in Brazil and Finland covered a wide variation in
land use from catchments dominated by agriculture to catchments covered almost
entirely by pristine forests (Atlantic rain forests and boreal forests, respectively) (Figure
2).
126
Fig. 2. Maps of the study regions: sampled streams in Finland (a), location of the São
Paulo State in Brazil (b) and sampled streams in Brazil (c). Map from Heino et al. (in
review).
2.2. Biological data
We sampled one riffle site at each stream using a kick-net (net mesh size = 0.5 mm) for
two minutes (consisted of four 30-seconds sampling units). We identified aquatic
insects to genus level and included genera from the following orders: Ephemeroptera,
Plecoptera, Trichoptera, Coleoptera, Odonata and Megaloptera.
We selected six traits that may be affected by human land-use (i.e., response
traits): refuge building, body shape, locomotion, functional feeding group, respiration
and body size (Table 1). For example, the canopy removal could decrease the number of
a)
b)
c)
127
shredders while the bottom siltation promoted by land use intensification could increase
burrowers (functional feeding guild trait) (Castro et al., 2018). Brazilian and Finnish
trait information was gathered mainly from specialized literature and consult to
specialists (see Acknowledgments).
Table 1. Functional traits analyzed for aquatic insects (adapted from Colzani et al.,
2013).
Traits Categories
Refuge building No refuge
Fixed nets and retreats
Portable shelters of sand, debris and/or wood
Portable shelters of leaf parts
Body shape Hydrodynamic
Not hydrodynamic
Locomotion groups Burrowers
Climbers/crawlers
Sprawlers
Clingers
Swimmers
Functional feeding guild Collector-gatheres
Collector-filterers
Herbivores
Predators
Respiration
Shredders
Tegumental respiration
Gill respiration
Body size
Air
Small-sized (<9 mm)
Medium-sized (9-16 mm)
Large-sized (>16 mm)
2.3. Local environmental data
128
After taking biological samples, we measured current velocity (m/s), depth (cm), mean
stream width of the sampling site and shading (i.e. canopy cover) by riparian vegetation
percentage at each sampling site. We also measured pH and conductivity at each site in
the field using YSI device model 556 MPS (YSI Inc., Ohio, USA). We took water
samples to analyze total nitrogen and total phosphorus following protocols for Finland
(Finnish Board of Waters and the Environment, 1981) and Brazil (Golterman et al.,
1978; Mackereth et al., 1978). We visually estimated particle size classes (%) in 0.25
m2 squares at random locations in a riffle site. We used a modified Wenthworth’s
(1922) scale of particle size classes: sand (0.25-2 mm), gravel (2-16 mm), pebble (16-64
mm), cobble (64-256 mm) and boulder (256-1024 mm). We calculated the Shannon
diversity of substratum particle sizes based on mean estimates for each stream.
2.4. Land use data
We obtained land use variables in a similar way in Brazil and Finland. First, we
delimited 500 m stream segments and then generated a 200 m buffer along each stream
segment. For each buffer, we extracted the proportion of land cover classes (i.e., native
forest, secondary forest, exotic planted forests, pasture, agriculture, urban, mining,
wetland, bare soil, water and mixed). We used Google Earth™ high resolution imagery
to extract land cover classes.
2.5. Taxonomic and functional beta diversity
We calculated beta diversity of aquatic insects for each watershed (i.e., among five
streams by watershed), separately for Brazil and Finland (20 Brazilian watersheds + 20
Finnish watersheds = 40 beta diversity values). We used four different dissimilarity
metrics in order to calculate taxonomic beta diversity: Sorensen, Bray-Curtis, turnover
129
component of Sorensen and turnover component of Bray-Curtis. Sorensen and Bray-
Curtis dissimilarities are simple and common metrics for presence/absence and
abundance data, respectively. Both dissimilarity metrics are affected by species richness
differences, which may be estimated by decomposing them in one component related to
replacement of species across sites (i.e. turnover component of dissimilarity) and
another component related to differences in species richness across sites (i.e. nestedness
component of dissimilarity) (Baselga, 2010). We only used turnover component (and
excluded nestedness) because we were interested in species replacement and not species
loss among streams. We log-transformed abundance data before the computation of beta
diversity. We used beta.pair function in the betapart package (Baselga et al., 2013) in R
environment (R Core Team, 2017) to obtain the turnover component of both Sorensen
and Bray-Curtis indices.
In order to calculate functional beta diversity, we first used the Gower distance
on the species-traits matrix (separately for Brazil and Finland) (Podani and Schmera,
2006). To do that, we used gowdis function in FD (Laliberté et al., 2014) package.
Then, we calculated the MPD (mean pairwise distance) using comdist function of
picante (Kembel et al., 2010) package. This function calculates the expected functional
distance separating two individuals or taxa drawn randomly from different
communities. We used this function with and without weights of log-transformed
abundance data.
Finally, we obtained a single beta diversity value for each watershed and for
each dissimilarity metrics using the multivariate homogeneity of group dispersions
(PERMDISP; Anderson et al., 2006). We used separately all dissimilarity matrices
produced by different metrics (i.e. Sorensen, Bray-Curtis, turnover component of
Sorensen and turnover component of Bray-Curtis, functional beta diversity based on
130
abundance and incidence data). However, as all taxonomic indices were similar to each
other as well as the functional dissimilarities (Figure S1), we show only Sorensen and
functional dissimilarity based on incidence in the main text. To analyze the data with
PERMDISP, we used betadisper function in the vegan package (Oksanen et al., 2015).
2.6. Modeling beta diversity along gradients in environmental harshness and
heterogeneity
We obtained explanatory variables regarding environmental harshness and
heterogeneity at watershed scale. We selected the mean of sand, nitrogen, phosphorus
and deforestation as variables supposed to be related to environmental harshness.
Excess of bottom siltation (here measured mixed with sand) and decrease of water
quality in eutrophic conditions by increased nutrient concentration (here nitrogen and
phosphorus) usually result from land-use intensification, affecting species composition
and species traits selected by the environment (Fugère et al., 2016; Leitão et al., 2018;
Castro et al., 2018). Regarding environmental heterogeneity, we measured the
dissimilarity of land-use and dissimilarity of local environmental variables using
PERMDISP. We obtained the latter using a matrix of Euclidean distance of the
following standardized abiotic variables (using decostand function and standardize
method where each value in a column is standardized to a mean of 0 and standard
deviation of 1): stream width, shading, sand, gravel, pebble, cobble, boulders, water
velocity, depth, pH, conductivity, nitrogen and phosphorus. We also calculated the
Euclidean distance of all land use variables together (i.e., proportions of native forest,
secondary forest, exotic planted forests, pasture, agriculture, urban, mining, wetland,
bare soil, water and mixed) to obtain the land use dissimilarity at watershed. We
131
summarized these predictor variables as well as their possible mechanisms of influence
on beta diversity in Fig. 1.
We investigated if land use affects beta diversity of aquatic insect communities
through increase of environmental harshness and/or modifications on environmental
heterogeneity using a Structural Equation Modeling (SEM). SEM is based on regression
analyses and allows testing multivariate and hierarchical relationships as well as
investigating causality between many variables simultaneously (Shipley, 2004).
Furthermore, SEM allows the investigation of latent variables, which are variables that
are difficult or impossible to obtain directly but can be defined using several measurable
variables (Shipley, 2004; Grace, 2006). In our study, environmental harshness (built
using nitrogen, phosphorus, deforestation and sand) and environmental variability (built
using land-use dissimilarity and local environmental variables dissimilarity) were
included as latent variables. We built four different SEMs to evaluate separately the
effects of land-use on each beta diversity metric (i.e. taxonomic/functional) and country
(Brazil/Finland). In order to check if the model as a whole is adequate to represent our
data, we used a goodness-of-fit chi-square (χ2) statistic. An adequate model should have
a low χ2 statistic and a nonsignificant p-value since it tests the difference between the
observed data and the hypothesized model. We showed only SEMs based on incidence
data (except for functional beta diversity because it did not run), but included SEMs for
abundance data in Table S1 and Figure S2. We used sem function in the lavaan package
(Rosseel, 2012) in R.
In addition, we built a linear regression model using beta diversity within a
watershed as the response variable (one model for each beta diversity metric) and the
following predictors: mean of sand, mean of nitrogen, mean of phosphor, log of total
dissimilarity of local environmental variables, log of deforestation, log of total
132
dissimilarity of land (all numeric variables) and country (categorical variable). We log-
transformed some predictors to better reach normality and homoscedasticity
assumptions. We did not select models (e.g. using AIC) because we already chose the
abovementioned variables based on the existing literature information. We checked for
multicollinearity using vif function. We fit the model using the lm function in R.
3. Results
We recorded 16,133 aquatic insects identified in 83 genera across tropical streams and
86,048 aquatic insects identified in 77 genera across boreal streams. Species richness at
the levels of stream sites and watersheds were higher in Brazil than in Finland, while
abundance showed the opposite pattern (more details in Heino et al. in review).
Our SEM models built to test the hypothesis on how land-use would influence
beta diversity through environmental harshness and environmental variability were not
adequate to our data (i.e. our data was different from hypothesized model), neither for
taxonomic (Brazil: χ2 = 44.71; P < 0.001; Finland: χ2 = 36.61; P < 0.001) nor for
functional beta diversity (Brazil: χ2 =42.04 ; P < 0.001; Finland: χ2 = 28.39; P = 0.008)
using incidence data (except for functional beta diversity in Brazil which was weighed
by abundance). We also did not find adequate models using abundance information
(Table S1). SEMs models failure to fit data may have been due to inadequate model
specification, the small number of watersheds or a genuine lack of relationship between
biotic homogenization and land-use intensification meadiated by environmental
harshness or habitat homogeneity (Fig. 3, Fig. S2 and Table S1).
133
Fig. 3. Structural models showing the relationships of the latent variables environmental
harshness (i.e. deforestation, sand, nitrogen and phosphorus) and environmental
heterogeneity (i.e. local dissimilarity and land-use dissimilarity) to taxonomic (A, B)
and functional beta diversity (C, D) in Brazil (A, C) and in Finland (B, D). Solid lines
indicate significant relationships (P < 0.05). Dotted lines indicate nonsignificant
relationships (P > 0.05). Values next to arrows indicate path coefficients. N = nitrogen,
P = phosphorus.
The linear regression model indicated that taxonomic beta diversity was higher
in Brazil than in Finland but were not influenced by variables related to land use
(F(7,32)= 3.246; R2adj = 0.287; P = 0.010; Table 2, Figure 4, Table S2 and Figure S1). We
found higher functional dissimilarity in Finland than in Brazil and also did not find
evidence for the effects of land-use intensification (F(6,33)= 5.284 R2adj = 0.434; P <
0.001; Table 2, Table S2 and Figure S1).
134
Fig. 4. Beta diversity among Brazilian and among Finnish streams within basins using
Sorensen dissimilarity (A) and functional dissimilarity using incidence data (B).
Table 2. Results from linear regression models for taxonomic (using Sorensen
dissimilarity) and functional (using incidence data) beta diversity. Std error = standard
error.
Estimate Std error t value P
Taxonomic dissimilarity
Intercept 0.361 0.065 5.503 <0.001
Country -0.073 0.026 -2.742 <0.001
Dissimilarity of land use 0.035 0.027 1.284 0.208
Deforestation -0.003 0.022 -0.154 0.878
Dissimilarity of local variables 0.047 0.063 0.747 0.460
Mean of sand 0.013 0.025 0.538 0.594
Mean of nitrogen -0.039 0.026 -1.522 0.137
Mean of phosphor 0.038 0.027 1.388 0.174
Functional dissimilarity
Intercept 0.139 0.008 16.915 <0.001
Country 0.015 0.003 4.668 <0.001
Dissimilarity of land use -0.002 0.003 -0.866 0.393
Deforestation 0.002 0.003 0.977 0.336
Dissimilarity of local variables 0.007 0.008 0.958 0.345
Mean of sand -0.004 0.003 -1.515 0.140
Mean of nitrogen -0.003 0.003 -1.097 0.281
Mean of phosphor -0.0003 0.003 -0.090 0.928
135
4. Discussion
Taxonomic and functional beta diversity were not congruent between tropical and
boreal regions: we found higher taxonomic but lower functional beta diversity among
tropical than among boreal streams. In addition, contrary to our expectation, we did not
find taxonomic or functional biotic homogenization neither due environmental
harshness (i.e. higher sand, nitrogen, phosphor or lower native vegetation cover) nor
due reduction of environmental heterogeneity (i.e. lower land-use heterogeneity and
lower local habitat variability) promoted by intensive land-use using SEMs and linear
regression models. Althought we discussed below possible reasons for why beta
diversity did not decrease with land-use intensity, we could not discard that the gradient
in land use, especially in Finland, may be not strong enough to cause biotic
homogenization or that 20 watersheds sampled in each country could be a low number
of replicates. However, we highlighted that 20 watersheds sampled in each country
represents a huge sampling effort (100 streams in each country), and that we sampled in
order to maximize the disturbance gradient existing in each sampled region.
One of the possible explanations for lower taxonomic beta diversity among
boreal streams is the increase of environmental harshness with latitude (e.g. climate)
because beta diversity is supposed to be lower among harsh habitats (e.g. Chase, 2007;
Chase, 2010). The higher taxonomic but lower functional beta diversity among tropical
streams suggest that, although species composition is more variable among tropical
streams, communities are functionally more redundant than among boreal streams.
Higher taxonomic but lower functional beta diversity in tropics is not an unprecedented
finding. For instance, despite of the high taxonomic beta diversity, functional beta
diversity was very low in tropical estuarine fish communities due dominant functional
groups (Villéger et al., 2012). In addition, regions closer to the poles are supposed to
136
contain a large range and high turnover of viable functional strategies of trees due to
climatic seasonality, which may increase functional beta diversity (Lamanna et al.,
2014). However, the causes of such finding (i.e. higher taxonomic but lower functional
beta diversity in tropics) in streams are poorly known. Indeed, the existence of
latitudinal gradients of taxonomic and functional beta diversities is under ongoing
discussion in the literature (e.g. Qian and Ricklefs, 2007; Kraft et al., 2011; Qian and
Song, 2013).
We found that neither taxonomic nor functional biotic homogenization was
promoted by land-use intensity. The lack of effects of land-use intensity may be even
more surprising regarding functional beta diversity because trait composition is
supposed to respond better to environmental changes than only species composition
(Castro et al., 2018). Nonetheless, it is important to point out that our findings may be
limited by the availability and quality of trait data, mainly in tropical streams, where the
information on the natural history and morphology of aquatic insects is still scarce
(Castro et al., 2018). In addition, the non-inclusion of some possible important variables
such as land use history, disturbance time lag and streams connectivity may also affect
our findings. Indeed, beta diversity patterns may be rather poorly predictable in such
highly dynamic systems as headwater streams (Heino et al., 2015).
While negative effects of intensive land use on stream species richness seem to
be more common (e.g. Corbi et al., 2013; Martins et al., 2017), the effects of land use
on streams beta diversity are still contradictory. Some studies found negative effects
(e.g. Passy and Blanchet, 2007; Maloney et al., 2011), some positive effects (e.g.
Hawkins et al., 2015; Fugère et al., 2016), whereas some studies like ours found no
effects of land use on stream beta diversity (e.g. Larsen and Omerod, 2014). One
possible explanation for such controversial results about land-use effects on stream beta
137
diversity is that land use may increase environmental harshness and decrease habitat
heterogeneity within streams (Allan, 1997), but increase habitat heterogeneity among
streams. For example, land-use changes may decrease habitat heterogeneity within
streams due to bottom siltation and deforestation (Castro et al., 2018), but may increase
heterogeneity among streams if they differ in disturbance intensity or land use types
(Barboza et al., 2015; Fugère et al., 2016). The different land-use types or intensities
may be somehow related to high physical and chemical differentiation among streams
and, consequently, causing distinct species and traits composition adapted to such
environmental conditions and increasing beta diversity. For example, Hawkins et al.
(2015) and Fugère et al. (2016) found higher beta diversity of macroinvertebrate
communities among disturbed streams, suggesting among-taxa differences in stress
tolerance as the underlying mechanism. More benign habitats, such as forested streams
with high within-stream heterogeneity, may increase the importance of priority effects
(i.e. the effects of the early colonizers on the following ones) which may increase
stochasticity in species establishment causing distinct species composition, and
consequently, high beta diversity (e.g. Chase, 2010; Petsch et al. 2017) (see Fig. 4).
Such mechanisms, i.e. high habitat heterogeneity among modified streams and high
stochasticity among forested streams, are not mutually exclusive along a land-use
gradient but may act simultaneously resulting in no difference between beta diversity
among forested and among modified streams.
138
Fig. 5. Habitat heterogeneity within and among forested (A, B, C) and modified (D, E,
F) streams. Habitat heterogeneity within forested streams may be higher than within
modified streams, but habitat heterogeneity among-streams may be higher among
modified streams. Modified streams may differ a lot from each other related to different
land-use types or intensity: the streambed in rural streams may be completed silted as
well as covered marginally by few trees (D); streams surrounded by exotic trees
plantation may have high vegetation cover but composed by only one tree species which
may contaminate the water with allelopathic substances (E); streams surrounded by
pasture may also be silted but with less runoff of contaminants than streams surrounded
by agriculture. Grey symbols and figures inside the streams indicate the substrata; N=
nitrogen.
In conclusion, we found high taxonomic and low functional beta diversity
among tropical streams, but the mechanisms driving such patterns are still unclear. We
did not find evidence that land-use intensification drives taxonomic or functional biotic
homogenization, probably due to stochastic and deterministic processes acting
simultaneusly to cause similar beta diversity among modified and among forested
streams. Our findings suggest the effects of different climatic regions as well as the
effect of environmental harshness and habitat heterogeneity mediated by land-use on
stream beta diversity remain to be completely understood.
139
Acknowledgments
We would like to thank Amarilis Brandão, Carlos F. Sanches and Neliton Lara for
Brazilian field assistance. We are also thankful to Frederico Salles, Pitágoras Bispo,
Lucas Lecci, Allan Santos, Rafael Braga, Bruno Sampaio, Rhainer Ferreira, Claudio
Froehlich and Jorge Nessimian for help on species traits of brazilian insects. The
Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) provided a
student fellowship to DKP. This study was funded by the FAPESP-AKA Joint Call on
Biodiversity and Sustainable Use of Natural Resources (grant 2013/50424-1) from São
Paulo Research Foundation (FAPESP) to TS, and grants from the Academy of Finland
(AKA) to JH (no. 273557) and JS (273560). ASM received a research fellowship from
Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq no.
309412/2014-5).
References
Allan, J.D., Erickson, D.L., Fay J., 1997. The influence of catchment land use on stream
integrity across multiple spatial scales. Freshw. Biol. 37, 149–161.
Allan, J.D., 2004. Landscapes and riverscapes: the influence of land use on stream
ecosystems. Annu. Rev. Ecol. Evol. Syst. 35, 257–284.
Anderson, M.J., Ellingsen, K.E., McArdle, B.H., 2006. Multivariate dispersion as a
measure of beta diversity. Ecol. Lett. 9, 683–693.
Barboza, L.G.A., Mormul, R.P., Higuti, J., 2015. Beta diversity as a tool for
determining priority streams for management actions. Water Sci. Technol . 71,
429-1435.
Baselga, A., 2010. Partitioning the turnover and nestednesscomponents of beta
diversity. Glob. Ecol. Biogeogr. 19, 134–143.
140
Baselga, A., Orme, D., Villeger, S., Bortoli, J., Leprieur, F., 2013. betapart: Partitioning
Beta Diversity into Turnoverand Nestedness Components. R package version
1.3. Available at: http://CRAN.R-project.org/package=betapar.
Brown, J.H., 2014. Why are there so many species in the tropics? J. Biogeogr. 41, 8–22.
Cardinale, B.J., Duffy, J.E., Gonzalez, A., Hooper, D.U., Perrings, C., Venail, P.,
Narwani, A., Mace, G.M., Tilman, D., Wardle, D., Kinzig, A.P., Daily, G.C.,
Loreau, M., Grace, J.B., Larigauderie, A., Srivastava, D.S., Naeem, S., 2012.
Biodiversity loss and its impact on humanity. Nature 486, 59–67.
Castro, D.M.P., Dolédec, S., Callisto, M., 2018. Land cover disturbance homogenizes
aquatic insect functional structure in neotropical savanna streams. Ecol. Ind. 84,
573–582.
Catano, C.P., Dickson, T.L., Myers, J.A., 2017. Dispersal and neutral sampling mediate
contingent effects of disturbance on plant beta-diversity: a meta-analysis. Ecol.
Lett. 20, 347–356.
Chase, J.M., 2007. Drought mediates the importance of stochastic community assembly.
Proc. Natl. Acad. Sci. U.S.A. 104, 17430–17434.
Chase, J.M., 2010. Stochastic Community Assembly Causes Higher Biodiversity in
More Productive Environments. Science 328, 1388-1391.
Cilleros, K., Allard, L., Grenouillet, G., Brosse S., 2016. Taxonomic and functional
diversity patterns reveal different processes shaping European and Amazonian
stream fish assemblages. J. Biogeogr. 43, 1832–1843.
Colzani, E., Siqueira, T., Suriano, M.T., Roque, F.O., 2013. Responses of aquatic insect
functional diversity to landscapechanges in Atlantic forest. Biotropica 45, 343–
350.
141
Corbi, J.J., Kleine, P., Trivinho-Strixino, S., 2013. Are aquatic insect species sensitive
to banana plant cultivation? Ecol. Ind. 25, 156–161.
Costa, S.S., Melo, A.S., 2008. Beta diversity in stream macroinvertebrate assemblages:
among-site and amongmicrohabitat components. Hydrobiologia 598, 131–138.
Díaz, S., Cabido, M., 2001. Vive la difference: plant functional diversity matters to
ecosystem processes. Trends Ecol. Evol. 16, 646–655.
Donohue, I., Jackson, A.L., Pusch, M.T., Irvine, K., 2009. Nutrient enrichment
homogenizes lake benthic assemblages at local and regional scales. Ecology 90,
3470–3477.
Fugère, V., Kasangaki, A., Chapma, L.J., 2016. Land use changes in an afrotropical
biodiversity hotspot affect stream alpha and beta diversity. Ecosphere 7, e01355-
10.1002/ecs2.1355.
Gámez-Virués, S., Perović, D.J., Gossner, M.M., Börschig, C., Blüthgen, N., de Jong,
H., Simons, N.K., Klein, A.M., Krauss, J., Maier, G., Scherber, C., Steckel, J.,
Rothenwöhrer, C., Steffan-Dewenter, I., Weiner, C.N., Weisser, W., Werner, M.,
Tscharntke, T., Westphal, C., 2015. Landscape simplification filters species
traits and drives biotic homogenization. Nat. Commun. 6, 8568.
Gaston, K.J., 2000. Global patterns in biodiversity. Nature 405, 220–227.
Golterman, H.L., Climo, R.S., Ohnstad, M.A.M., 1978. Methods for Physical and
Chemical Analysis of Freshwaters. IBP, Oxford.
Gossner, M.M., Lewinsohn, T.M., Kahl, T., Grassein, F., Boch, S., Prati, D., Birkhofer,
K., Renner, S.C., Sikorski, J., Wubet, T., Arndt, H., Baumgartner, V., Blaser, S.,
Blüthgen, N., Börschig, C., Buscot, F., Diekötter, T., Jorge, L.R., Jung, K.,
Keyel, A.C., Klein, A., Klemmer, S., Krauss, J., Langel, M., Müller, J.,
Overmann, J., Pašalić, E., Penone, C., Perović, D.J., Purschke, O., Schall, P.,
142
Socher, S.A., Sonnemann, I., Tschapka, M., Tscharntke, T., Türke, M., Venter,
P.C., Weiner, C.N., Werner, M., Wolters, V., Wurst, S., Westphal, C., Fischer,
M., Weisser, W.W., Allan, E., 2016. Land-use intensification causes
multitrophic homogenization of grassland communities. Nature 540, 266-269.
Grace, J.B., 2006. Structural equation modeling and natural systems. Cambridge
University Press
Hawkins, C.P., Mykrä, H., Oksanen, J., Vander Laan, J.J., 2015. Environmental
disturbance can increase beta diversity of stream macroinvertebrate assemblages.
Global Ecol. Biogeogr. 24, 483–494.
Heino, J., Melo, A.S., Bini, L.M., Altermatt, F., Al-Shami, S.A., Angeler, D.G.,
Bonada, N., Brand, C., Callisto, M., Cottenie, K., Dangles, O., Dudgeon, D.,
Encalada, A., Göthe, E., Grönroos, M., Hamada, N., Jacobsen, D., Landeiro,
V.L., Ligeiro, R., Martins, R.T., Miserendino, M.L., Rawi, C.S.M., Rodrigues,
M.E., Roque, F.O., Sandin, L., Schmera, D., Sgarbi, L.F., Simaika, J.P.,
Siqueira, T., Thompson, R.M., Townsend, C.R., 2015. A comparative analysis
reveals weak relationships between ecological factors and beta diversity of
stream insect metacommunities at two spatial levels. Ecol. Evol. 6, 1235–1248.
Hillebrand, H., 2004. On the Generality of the Latitudinal Diversity Gradient. Am. Nat.
163, 192-211.
Jacobsen, D., Cressa, C., Mathooko, J.M., Dudgeon, D., 2008. Macroinvertebrates:
composition, life histories and production. In: Tropical stream ecology (D.
Dudgeon, ed.). Elsevier, San Diego, p.65-105.
Jonason, D., Ekroos, J., Öckinger, E., Helenius J., Kuusaari, M., Tiainen, J., Smith,
H.G., Lindorg, R., 2017. Weak functional response to agricultural landscape
homogenisation among plants, butterflies and birds. Ecography 40, 1221–1230.
143
Kembel, S.W., Cowan, P.D., Helmus, M.R., Cornwell, W.K., Morlon, H., Ackerly,
D.D., Blomberg, S.P., Webb, C.O., 2010. Picante: R tools for integrating
phylogenies and ecology. Bioinformatics 26, 1463-1464.
Kennedy, T.A., Muehlbauer, J.B., Yackulic, C.B., Lytle, D.A., Miller, S.A., Dibble,
K.L., Kortenhoeven, E.W., Metcalfe, A.N., Baxter, C.V., 2016. Flow
management for hydropower extirpates aquatic insects, undermining river food
webs. BioScience 66, 561–575.
Kraft, N.J.B., Comita, L.S., Chase, J.M., Sanders, N.J., Swenson, N.G., Crist, T.O.,
Stegen, J.C., Vellend, M., Boyle, B., Anderson, M.J., Cornell, H.V., Davies,
K.F., Freestone, A.L., Inouye, B.D, Harrison, S.P., Myers, J.A., 2011.
Disentangling the drivers of β diversity along latitudinal and elevational
gradients. Science 333, 1755–8.
Lamanna, C., Blonder, B., Violle, C., Kraft, N.J.B., Sandel, B., Šímová, I., Donoghue,
J.C., Svenning, J., McGill, B.J., Boyle, B., Buzzard, V., Dolins, S., Jørgensen,
P.M., Marcuse-Kubitza, A., Morueta-Holme, N., Peet, R.K., Piel, W.H., Regetz,
J., Schildhauer, M., Spencer, N., Thiers, B., Wiser, S.K., Enquist, B.L., 2014.
Functional trait space and the latitudinal diversity gradient. Proc. Natl. Acad.
Sci. U.S.A. 111, 13745–13750.
Laliberté, E., Legendre, P., Shipley, B., 2014. FD: measuring functional diversity from
multiple traits, and other tools for functional ecology. R package version 1, 0–
12.
Larsen, S., Ormerod, S.J., 2014. Anthropogenic modification disrupts species co-
occurrence in stream invertebrates. Global Change Biol. 20, 51–60.
Leal, C.G., Pompeu, P.S., Gardner, T.A., Leitão, R.P., Hughes, R.M., Kaufmann, P.R.,
Zuanon, J., Paula, F.R., Ferraz, S.F.B., Thomson, J.R., Nally, R.M., Ferreira, J.,
144
Barlow, J., 2016. Multi-scale assessment of human-induced changes to
Amazonian instream habitats. Lands. Ecol. 31, 1725–1745.
Leitão, R.P., Zuanon, J., Mouillot, D., Leal, C.G., Hughes, R.M., Kaufmann, P.R.,
Villéger, S., Pompeu, P.S., Kasper, D., de Paula, F.R., Ferraz, S.F.B., Gardner,
T.A., 2018. Disentangling the pathways of land use impacts on the functional
structure of fish assemblages in Amazon streams. Ecography 41, 219–232.
Mackereth, F.J., Heron, H.J., Talling, J.F., 1978. Water Analysis: Some Revised
Methods for Limnologists. Freshwater Biological Association, London.
Maloney, K., Munguia, P., Mitchell, R., 2011. Anthropogenic disturbance and
landscape patterns affect diversity patterns of aquatic benthic
macroinvertebrates. J. North Am. Benthol. Soc. 30, 284–295.
Marchetti, M.P., Lockwood, J.L., Light, T., 2006. Effects of urbanization on
California’s fish diversity: Differentiation, homogenization and the influence of
spatial scale. Biol. Conserv. 127, 2130–2318.
Martins, R.T., Couceiro, S.R.M., Melo, A.S., Moreira, M.P., Hamada, N., 2017. Effects
of urbanization on stream benthic invertebrate communities in Central Amazon.
Ecol. Ind. 73, 480–491.
McKinney, M.L., Lockwood, J.L., 1999. Biotic homogenization: a few winners
replacing many losers in the next mass extinction. Trends Ecol. Evol. 14, 450-
453.
Millenium Ecosystem Assessment, 2005. Ecosystems and Human Well Being. Island
Press, Washington DC.
Mittelbach, G.G., Schemske, D.W., Cornell, H.V., Allen, A.P., Brown, J.M., Bush,
M.B., Harrison, S.P., Hurlbert, A.H., Knowlton, N., Lessios, H.A., McCain,
C.M., McCune, A.R., McDade, L.A., McPeek, M.A., Near, T.J., Price, T.D.,
145
Ricklefs, R.E., Roy, K., Sax, D.F., Schluter, D., Sobel, J.M., Turelli, M., 2007.
Evolution and the latitudinal diversity gradient: speciation, extinction and
biogeography. Ecol. Lett. 10, 315–331.
Olden, J.D., Rooney, T.P., 2006. On defining and quantifying biotic homogenization.
Global Ecol. Biogeogr. 15, 113–120.
Oksanen, J., Blanchet, F.G., Kindt, R., Legendre, P., Minchin, P.R., O'Hara, R.B.,
Simpson, G.L., Solymos, P., Stevens, M.H.H., Wagner, H., 2015. vegan:
Community Ecology Package. R package version 2.2-1. Available at:
http://CRAN.R-project.org/package=vegan.
Passy, S.I., Blanchet, F.G., 2007. Algal communities in human-impacted stream
ecosystems suffer beta-diversity decline. Divers. Distrib. 13, 670–679.
Petsch, D.K., Schneck, F., Melo, A.S., 2017. Substratum simplification reduces beta
diversity of stream algal communities. Freshw. Biol. 62, 205-213.
Podani, J., Schmera, D., 2006. On dendrogram-based measures of functional diversity.
Oikos 115, 179–185.
Qian, H., Ricklefs, R.E., 2007. A latitudinal gradient in large-scale beta diversity for
vascular plants in North America. Ecol. Lett. 10, 737–744.
Qian, H., Song, J.S., 2013. Latitudinal gradients of associations between beta and
gamma diversity of trees in forest communities in the New World. J. Plant Ecol.
6, 12–18.
R Core Team, 2017. R: A Language and Environment for Statistical Computing. R
Foundation for Statistical Computing, Vienna. Available at: http://www.R-
project.org/.
Rodrigues, J.L.M., Pellizari, V.H., Mueller, R., Baek, K., Jesus, E.C., Paula, F.S.,
Mirza, B., Hamaoui, G.S., Tsai, S.M., Feigl, B., Tiedje, J.M., Bohannan, B.J.M.,
146
Nüsslein, K., 2013. Conversion of the Amazon rainforest to agriculture results in
biotic homogenization of soil bacterial communities. Proc. Natl. Acad. Sci.
U.S.A. 110, 988–993.
Rodrigues, J.F.M., Olalla-Tárraga, M.A., Iverson, J.B., Akre, T.S.B., Diniz-Filho,
J.A.F., 2017. Time and environment explain the current richness distribution of
non-marine turtles worldwide. Ecography 40, 1-10.
Rosseel, Y., 2012. lavaan: An R package for structural equation modeling. J. Statistical
Software 48, 1–36.
Sala, O. E., Chapin, F.S., Armesto, J.J., Berlow, E., Bloomfield, J., Dirzo, R., Huber-
Sanwald, E., Huenneke, L.F., Jackson, R.B., Kinzig, A., Leemans, R., Lodge,
D.M., Mooney, H.A., Oesterheld, M., Poff, N.L., Sykes, N.T., Walker, B.H.,
Walker, M., Wall, D.H., 2000. Global biodiversity scenarios for the year 2100.
Science 287, 1770–1774.
Sfair, J.C., Arroyo-Rodriguez, V., Santos, B.A., Tabarelli, M., 2016. Taxonomic and
functional divergence of tree assemblages in a fragmented tropical forest. Ecol.
Appl. 26, 1816–1826.
Shipley, B., 2004. Cause and correlation in biology: a user’s guide to path analysis,
structural equations and causal inference. Cambridge University Press,New
York.
Siqueira, T., Lacerda, C.G.T., Saito, V.S., 2015. How does landscape modification
induce biological homogenization in tropical stream metacommunities?
Biotropica 47, 509–516.
Soininen, J., Lennon, J.J., Hillebrand, H., 2007. A multivariate analysis of beta diversity
across organisms and environments. Ecology 88, 2830–2838.
147
Soininen, J., 2010. Species turnover along abiotic and biotic gradients: patterns in space
equal patterns in time? BioScience 60, 433–439.
Solar, R.R.d.C., Barlow, J., Ferreira, J., Berenguer, E., Lees, A.C., Thomson, J.R.,
Louzada, J., Maués, M., Moura, N.G., Oliveira, V.H.F., Chaul, J.C. M.,
Schoereder, J.H., Vieira, I.C.G., Mac Nally, R., Gardner, T.A., 2015. How
pervasive is biotic homogenization in human-modified tropical forest
landscapes? Ecol. Lett. 18, 1108–1118.
Stevens, G.C., 1989. The latitudinal gradient in geographical range: how so many
species coexist in the tropics. Am. Nat. 133, 240–256.
Villéger, S., Miranda, J.R., Hernandez, D.F., Mouillot, D., 2012. Low functional β-
diversity despite high taxonomic β-diversity among tropical estuarine fish
communities. PLoS One 7, e40679.
Violle, C., Navas, M.L., Vile, D., Kazakou, E., Fortunel, C., Hummel, I., Garnier, E.,
2007. Let the concept of trait be functional! Oikos 116, 882–892.
148
SUPPLEMENTARY MATERIAL
Figure S1. Beta diversity among Brazilian and among Finnish streams within basins
using Sorensen dissimilarity (a), turnover component of Sorensen dissimilarity (b),
functional dissimilarity using incidence data (c), Bray-Curtis dissimilarity (d), turnover
component of Bray-Curtis dissimilarity (e) and functional dissimilarity with abundance
weight (f).
149
Fig. S2 Structural model showing the relationships of the latent variables environmental
harshness (i.e. deforestation, sand, nitrogen and phosphor) and environmental
heterogeneity (i.e. local dissimilarity and land-use dissimilarity) to taxonomic (A, B)
and functional beta diversity (C) in Brazil (A) an Finland (B, C). Solid lines indicate
significant relationships (P < 0.05). Dotted lines indicate nonsignificant relationships (p
> 0.05). Values next to arrows indicate path coefficients. N = nitrogen, P = phosphor.
150
Table S1. Results from the structural equation model to taxonomic and functional beta
diversity in Brazil and Finland using abundance information.
χ2 P
Brazil Taxonomic 44.4 <0.001
Finland Taxonomic 32.3 0.020
Finland Functional 27.2 0.012
151
Table S2. Results from linear model for taxonomic (using Bray-Curtis, turnover
component of Bray-Curtis and turnover component of Sorensen) and functional (using
abundance data) beta diversity.
Estimate Std error t value P
Bray-Curtis
Intercept 0.369 0.065 5.67 <0.01
Country -0.067 0.026 -2.538 0.016
Dissimilarity of land use 0.024 0.027 0.884 0.383
Native deforestation 0.003 0.022 0.129 0.897
Dissimilarity of local variables 0.067 0.063 1.065 0.291
Mean of sand 0.015 0.025 0.609 0.546
Mean of nitrogen -0.035 0.026 -1.369 0.180
Mean of phosphor 0.025 0.027 0.924 0.362
R2adj 0.227 0.028
Turnover component of Bray-Curtis
Intercept 0.231 0.057 4.020 <0.01
Country -0.080 0.023 -3.444 <0.01
Dissimilarity of land use -0.008 0.023 -0.349 0.729
Native deforestation 0.024 0.021 1.209 0.235
Dissimilarity of local variables 0.085 0.055 1.534 0.134
Mean of sand 0.012 0.022 0.538 0.594
Mean of nitrogen -0.030 0.022 0.538 0.594
Mean of phosphor 0.002 0.024 0.114 0.909
R2adj 0.246 0.020
Turnover componente of Sorensen
Intercept 0.239 0.060 3.971 <0.01
Country -0.078 0.024 -3.183 <0.01
Dissimilarity of land use 0.002 0.024 0.086 0.932
Native deforestation 0.019 0.021 0.925 0.361
Dissimilarity of local variables 0.057 0.058 0.982 0.333
Mean of sand 0.026 0.023 1.146 0.260
Mean of nitrogen -0.010 0.025 0.406 0.687
Mean of phosphor -0.030 0.024 -1.290 0.206
R2adj 0.279 0.011
Functional (abundance weighed)
Intercept 0.128 0.011 11.480 <0.01
Country 0.013 0.004 2.916 <0.01
Dissimilarity of land use -0.007 0.004 -1.610 0.117
Native deforestation 0.007 0.003 1.844 0.074
Dissimilarity of local variables 0.006 0.010 0.596 0.555
Mean of sand -0.004 0.004 -0.970 0.339
Mean of nitrogen -0.003 0.004 -0.739 0.465
Mean of phosphor <0.001 0.004 -0.138 0.890
R2adj 0.284 0.010
152
C
LAND-USE EFFECTS ON STREAMS
BIODIVERSITY: A META-ANALYSIS5
5 Manuscrito a ser submetido para a seção Reports da revista Ecology em colaboração
com A. S. Melo, L. Korell e J. M. Chase.
APÍTULO 5
22
153
Abstract
It is well-known that land-use intensification is a major driver of biodiversity loss in
streams, but the main mechanisms underlying this pattern are not yet completely
understood. We conducted a meta-analysis using 39 studies to address how land use
affects alpha and beta diversities as well as relative and total abundance in streams. We
found that species composition of modified streams is different, not only a subset from
the species composition of reference streams. Indeed, many biological monitoring
studies found that some species are indicators of good water quality while others species
are indicators of poor water quality. Land-use changes did not cause biotic
homogenization, maybe because modified streams may differ a lot in terms of type and
intensity of disturbance. We also found lower species richness related to changes in
relative abundance probably due increase of dominance of tolerante species and not just
simply caused by a lower number of individuals due land-use alterations. We
highlighted that the sole use of species richness may not be adequate to disentangle the
main mechanisms underlying biodiversity loss due to land-use intensification.
Keywords: Biotic homogenization, deforestation, ENSPIE, beta diversity, lotic
ecosystems
154
INTRODUCTION
Land-use intensification is a major driver of species loss worldwide (Sala et al.
2000, Newbold et al. 2015), but the main mechanisms underlying this decline in
biodiversity are not yet completely understood. One reason behind this is that the sole
quantification of species richness may not be enough to capture all changes in
biodiversity (Chase and Knight 2013, Hillebrand et al. 2017). For example, (1) land-use
intensification could decrease the number of species and change their relative
abundances, yet the total number of individuals in the community may remain constant.
In this scenario, some species would become dominant and other species would become
extinct or rarer according to their tolerance to land use (e.g. Lougheed et al. 2008,
Castro et al. 2018). Alternatively, (2) land-use intensification could decrease the number
of species and total number of individuals, but change proportionally the relative
abundances of the remaining species. In this scenario, the species richness reduction due
to land-use would not be related to increased species dominance but a consequence
from a sampling artifact since the number of observed species is supposed to decrease
with reduced number of sampled individuals (“more individual hypothesis”; Srivastava
and Lawton 1998, Schuler et al. 2014). Although both scenarios result in lower species
richness with changing land-use intensity, they are caused by different mechanisms.
One way to assess which components of biodiversity (i.e. relative abundance
and/or species numbers) are changed by land-use intensity is to use the probability of
interspecific encounter (PIE) (complement version of Simpson’s index; Hurlbert 1971).
PIE corresponds to the probability that two individuals randomly sampled from a
community belong to different species (Blowes et al. 2017). It is the difference of
rarefied richness for two and one individuals. Because PIE describes the slope of the
individual-based rarefaction curve at its origin, this measure can estimate the richness
155
and evenness of the relative abundance distribution (Chase and Knight 2013).
Converting PIE into an ‘effective number of species’ (ENSPIE = 1/∑ 𝑝𝑖2𝑆𝑖=1 , where 𝑆
represents the number of species and 𝑝𝑖 is the proportion of species 𝑖 in the community)
gives the number of relatively common species (Blowes et al. 2017). Thus, the
comparison of ENSPIE goes beyond the simple report of differences in species numbers
between two communities and may be helpful to understand the mechanisms
underlying, e.g., the effects of land use modification on streams biodiversity (e.g.
changes in total and/or relative abundance). Moreover, ENSPIE is independent of
sample-size and scale (Chase and Knight 2013) while species richness is scale-sensitive
and increases with the sampling area or number of sampled individuals (Colwell et al.
2004). Understanding and taking into account the limitation of diversity measures is
important because drivers of biodiversity change, such as land-use or disturbance, may
be for instance stronger at smaller scale (e.g. community or alpha scale) than larger
scale (e.g. communities combined in one region or gamma scale) (e.g. Chase and
Knight 2013, Powell et al. 2013).
In addition to changes in abundance and species richness, land-use can affect
species composition of communities (e.g. Solar et al. 2015, Gossner et al. 2016). The
variation of species composition from site to site, i.e. beta diversity (Whittaker 1960) is
composed by two components: turnover (i.e. species replacement among communities)
and nestedness (i.e. species richness difference among communities) (Baselga 2012).
The partition of beta diversity into its two components allows the assessment of the
main mechanisms generating variation in species composition (e.g. between pristine and
modified communities). On the one hand, species composition between pristine and
modified communities could be different, with some species adapted to local pristine
conditions while other species could be adapted to or tolerate modified conditions
156
(Gutiérrez-Canova 2013). In this scenario, we would expect a higher contribution of
turnover to total beta diversity. On the other hand, the species composition of modified
communities could not be different of pristine communities, but only a nested subset of
the total species composition from pristine communities. In this scenario, land-use may
drive a non-random extinction of sensitive species allowing only the tolerant species to
survive in modified streams (Gutiérrez-Canova 2013). In this case, we would expect an
increased contribution of the nestedness component to beta diversity.
Besides using beta diversity to verify whether species composition is different
between reference and modified streams, we may use beta diversity estimated among
modified and among reference streams to verify if land-use intensification decrease
communities variability causing biotic homogenization (e.g. Siqueira et al. 2015, Castro
et al. 2018). This biotic homogenization (McKinney and Lockwood 1999) might occur
because of a loss of rare and sensitive species and/or a gain of the same tolerant
widespread species across intensively modified communities (McKinney and
Lockwood 1999, Gossner et al. 2016).
Previous studies on land-use change are primarily conducted in terrestrial
ecosystems (Menge et al. 2009, Gerstner et al. 2014). However, freshwater streams are
megadiverse ecosystems also severely threatened by land-use changes related to
agriculture (e.g. Corbi et al. 2013), forestry (e.g. Konopik et al. 2015) and urbanization
(e.g. Martins et al. 2017). Besides the vegetation cover reduction, land-use
intensification may change the streambed, the inputs of organic material, the
concentration of dissolved oxygen, and the channel structure, and, consequently, affect
negatively diverse stream communities (Allan et al. 1997, Leal et al. 2016).
Our main question is how land-use may change stream biodiversity. Indeed, the
effects of different land-use types on stream species richness, abundance and species
157
composition are not yet completely understood and had not been quantitatively
summarized. We conducted a meta-analysis to address this topic of wide interest for
ecologists, conservationists and decision makers. We investigated if land-use changes
decreases number of observed species richness, extrapolated species richness, ENSPIE
and total abundance at local (alpha) and regional (gamma) scales. We verified if land-
use modification cause biotic homogenization in streams. We also investigated whether
land-use modifies the relative contributions of components of beta diversity (i.e.
nestedness and turnover) between modified and pristine streams.
METHODS
Data collection
We systematically searched studies contrasting biodiversity in modified and
reference (or less modified) streams. First, we used the topic search in ISI Web of
Science with the following terms combinations: (("stream" OR "streams") AND ("land
use" OR "logging" OR "agriculture" OR "plantation" OR "crops" OR "forestry" OR
"urban*" OR "rural" OR "farm" OR "silviculture") AND ("biodiversity" OR "richness"
OR "diversity" OR "beta diversit*" OR "ß diversit*")). We did not apply filters and
selected studies from 1990 to 2017. We searched on July 2017 and found 2102 studies.
We also selected 15 studies from other sources (i.e. Google Scholar).
Second, we reviewed titles and abstracts of all articles to select potentially
relevant studies for our meta-analysis. We excluded 1934 studies because they were not
related to our aim, were theoretical-reviews, were duplicated, quantified incidence or
cover data, identification at only family level or represented design problems (e.g. land-
use type was not defined, no reference streams were included or there was no
replication). After these exclusion criteria, we retained 189 studies to analyze their full
158
text. Of these 189 studies we searched for data of species/genus abundance in the full
text or supplementary material and requested data from authors when not provided. 150
studies had to be excluded from the meta-analysis because the authors did not provide
data or because we detected the problems aforementioned, resulting in 39 studies
included in our meta-analysis (see Fig. S1).
From the remaining studies, we extracted information about taxa, sample size,
sites locations and land-use type. We categorized the streams into three main land-use
types: agriculture (e.g. pasture, sugarcane, banana and coffee plantations), forestry (e.g.
exotic trees plantation and logging) and urbanization (e.g. residential and industrial
areas). We used land-use types as moderators as they are suggested to alter stream
communities differently (e.g. Gimenez et al. 2015, Siqueira et al. 2015). We also
considered season, watersheds and substrata of studied streams as covariates in our
analysis (see metadata information in Table S1).
Data analysis
From each study, we recorded the number of observed species and the number
of individuals in each stream (i.e. alpha scale) as well as calculated Chao’s (1984) non-
parametric method for extrapolating the number of species. Finally, we estimated
ENSPIE for each stream. We repeated these analyses performed at alpha scale to gamma
scale. While alpha scale corresponds to samples taken in one stream, we defined gamma
scale by merging samples from streams in each land use type, also taking into account
possible covariates such as season and substrata. For instance, if the study sampled in
dry and rainy seasons, we categorized the same land-use streams into these seasons. We
calculated multiplicative beta diversity (Whittaker 1960) as the ratio between gamma
and mean alpha diversities of each measure (ENSPIE, observed and extrapolated species
159
richness) separately. We resampled to obtain the same number of streams for each land-
use type at gamma scale and obtained similar results to data without resampling that we
are using in the main text (Fig. S2). Calculations of extrapolated species richness and
ENSPIE were performed using the estimateR and the diversity function (index =
"invsimpson") in the vegan package (Oksanen et al. 2010) of the R environment (R
Core Team 2017).
We also calculated beta diversity to investigate whether there was any turnover
or loss of species between reference and modified streams (i.e. agriculture, forestry or
urban land-uses) (among land-use beta diversiry). For this purpose, we merged stream
data of each study according to the land-use type. We then estimated turnover and
nestedness contribution of beta diversity between reference and modified streams using
different dissimilarity indices for incidence (i.e. Sorensen dissimilarity) and abundance
(i.e. Bray-Curtis dissimilarity) data (see Fig. S3) (Baselga 2012). We divided each
component value (i.e. turnover or nestedness) by total beta diversity to obtain their
percentage of contribution. Finally, we estimated the difference between turnover and
nestedness. We also calculated within land-use beta diversity to estimate biotic
homogenization. To do so, we estimated turnover component of beta diversity measured
among streams of each land-use type separately using Sorensen and Bray-Curtis
dissimilarities and then applied PERMDISP (multivariate homogeneity of groups
dispersions). We used beta.pair and beta.pair.abund functions from the betapart
package (Baselga et al. 2015) and betadisper function of vegan package in R.
To determine the magnitude and direction of how land-use intensification
changes biodiversity at alpha scale, we first calculated mean and standard deviation of
abundance, observed species richness, extrapolated species richness and ENSPIE across
streams of each study. From these values, we then quantified effect sizes and variance
160
using log response ratio measure (i.e., log-proportional change between the means of a
land-use and reference group, Lajeunesse 2011). We expect that land-use types would
decrease abundance, observed species richness, extrapolated species richness and
ENSPIE (i.e., negative effect sizes). Because some studies used common reference
streams for different land use types, we need to take into account the non-independence
among these observations. To reduce such bias we estimated the variance of effect size
as suggested by Lajeunesse (2011), i.e. we classified the observations with common-
control as independent. Finally, we fitted a meta-analytic multilevel mixed-effects
model using study as random effect and land use as moderators. We used rma.mv
function in metafor package (Viechtbauer 2010) to fit our model.
We estimated I2, the proportion of the variance that can potentially be explained
by moderators (Borenstein et al. 2009). We checked publication bias by inspection of
funnel plots and by the Orwin Fail-Safe Number (OFSN; Orwin 1983) using,
respectively, funnel and fsn functions in metafor package. We used the OFSN to
estimate the number of studies required to decrease the observed mean effect size 1/2 of
the cumulative effect that we estimated. We did not detect publication bias (see Fig. S4
and Table S2).
At gamma and beta scale we also quantified effect sizes using log-response ratio
between modified and reference streams for total abundance, observed species richness,
extrapolated species richness and ENSPIE (except total abundance in beta scale). We
calculated the difference between the relative contribution of turnover and nestedness to
beta diversity among land-use types. In this case, values lower than zero correspond to a
higher contribution of nestedness while values higher than zero correspond to higher
contribution of turnover to beta diversity. We also estimated turnover component
within-land use types to verify biotic homogenization. Because we could not estimate
161
variance associate to effect sizes for these aforementioned measures, we fitted an
unweighted linear mixed effects model using study as random effect and land-use types
as fixed effects. We used a Gaussian error distribution and the lmer function in lme4
(Bates et al. 2014) package to fit these models. We did all the figures using ggplot2
(Wickham 2009) and maps (Becker et al. 2017) packages.
RESULTS
The 39 included studies originated from all continents, except Antarctica, but
with a strong bias to South America (Fig. 1). Included studies investigated different
organisms: frogs (n = 2), fishes (n = 5), fungi (n = 1), macrophytes (n = 1), algae (n = 1)
and invertebrates (n = 28).
FIG. 1. Locations of studies included in our meta-analysis. Circle size represents the
sample size of each study.
Across organism types, land-use decreased total abundance at the alpha (mean
effect size ± 95% CI = -0.278 ± 0.225; I2= 99.761) and gamma scale (mean effect size ±
95% CI = -0.311 ± 0.228). However, disentangling land-use types, we found that only
162
urbanization indeed decreased total abundance both at alpha (Fig. 2A) and gamma scale
(Fig. 2B).
FIG. 2. Effect sizes (log-response ratio) and 95% confidence interval (vertical bars)
estimated separately for agriculture, forestry and urbanization land-use types using total
abundance at alpha (A) and gamma (B) scale. Note that alpha was weighted by variance
within studies while gamma was not (see methods).
Overall, land-use decreased ENSPIE (mean effect size ± 95% CI = - 0.324 ±
0.141; I2 = 97.852), observed species richness (mean effect size ± 95% CI = - 0.403 ±
0.147; I2 = 94.811) and extrapolated species richness (mean effect size ± 95% CI = -
0.393 ± 0.156; I2 = 94.643) at alpha scale. However, the subgroups analysis indicated
that effect size of extrapolated species richness was not different from zero in
agriculture land-use (mean effect size ± 95% CI = - 0.129 ± 0.164), and urbanization
seems to have a stronger effect (Fig. 3a).
At gamma scale, land-use also decreased ENSPIE (mean effect size ± 95% CI = -
0.364 ± 0.209) and observed species richness (mean effect size ± 95% CI = - 0.288 ±
0.171), but we did not detect changes in extrapolated species richness (mean effect size
± 95% CI = - 0.162 ± 0.228). However, neither were the effect sizes of all biodiversity
measures different from zero in agriculture land-use nor did the effect size of
extrapolated species richness decline in urban streams (Fig. 3b).
163
At the beta scale, there was no effect of land-use on ENSPIE (mean effect size ±
95% CI = - 0.012 ± 0.139) and observed species richness (mean effect size ± 95% CI =
0.082 ± 0.097). However, effect size was higher than zero for extrapolated species
richness (mean effect size ± 95% CI = 0.133 ± 0.122) although not for forestry land-use
(mean effect size ± 95% CI = - 0.173 ± 0.202) (Fig. 3c). We also detected higher effect
size of observed species richness in urbanization land use (mean effect size ± 95% CI =
0.283 ± 0.157).
FIG. 3. Effect sizes (log-response ratio) and 95% confidence interval (vertical bars)
estimated separately for agriculture, forestry and urbanization land-use types using
ENSPIE, extrapolated and observed species richness at alpha (A), gamma (B) and beta
scales (C). Note that alpha was weighted by variance inside studies while gamma and
beta were not (see methods).
We found a higher relative contribution of turnover rather than nestedness
component to total beta diversity measured between reference and land use streams
using Sorensen (mean difference between turnover and nestedness ± 95% CI = 0.208 ±
0.142) and Bray-Curtis dissimilarities (mean difference between turnover and
nestedness ± 95% CI = 0.401 ± 0.153) (Fig. 4A). The only exception was the contrast
between reference and urban streams using Sorensen dissimilarity, when both turnover
and nestedness components contributed equally to total beta diversity (Fig. 4A). We did
164
not find evidence for biotic homogenization drive by land-use intensification since we
did not find difference between reference and land-use beta diversities neither using
Sorensen nor using Bray-Curtis dissimilarity (Fig. 4B).
FIG. 4. (A) Among land-use beta diversity: difference between turnover and nestedness
components of beta diversity measured between reference and land use streams using
Values higher than zero indicate higher contribution of turnover while values lower than
zero indicate higher contribution of nestedness to beta diversity. (B) Within land-use
beta diversity: turnover component of beta diversity measured among streams within the
same land-use type. Dashed line refers to reference beta-diversity of Sorensen and
twodash line refers of Bray-Curtis. We used incidence (Sorensen dissimilarity in blue)
and abundance data (Bray-Curtis dissimilarity in pink).
DISCUSSION
Our meta-analysis clearly demonstrates that land-use decreases stream
biodiversity across different spatial scales. Despite the many observational studies (e.g.
Konopik et al. 2015, Martins et al. 2017, Prudente et al. 2017), to our knowledge our
study is the first quantitative synthetic study about land-use effects on stream
biodiversity so far. Our synthetic approach allows us to identify the mechanisms
underlying the biodiversity decline though land-use, i.e. whether this decline is driven
by a loss of total individuals or changes in their abundance distribution. We found that
land-use not only affected biodiversity in stream communities by decreasing species
165
richness and their relative abundances, but also by changing species composition (i.e.
higher contribution of turnover than nestedness to total beta diversity). However, we did
not find biotic homogenization caused by land-use intensification. Specifically,
urbanization had the strongest negative effect on stream biodiversity compared to
agriculture and forestry.
The lower species richness in conjunction with lower evenness (i.e. S and
ENSPIE) observed in modified streams might be related to species tolerance to land-use
intensity. One of the most consistent patterns in ecology is that communities are not
composed by species with equal relative abundance but by a few common and many
rare species (Preston 1948, Siqueira et al. 2012). However, the lower evenness in a
community may suggest the influence of some ecological driver such as land-use
modification (Chase and Knight 2013). Few species are ecologically adapted to tolerate
stress caused by land-use such as increased input of sediments covering the streambed,
reduced inputs of allochtonous organic material, input of fertilizers, lower concentration
of dissolved oxygen, and the alteration of flow regime or channel structure (Allan et al.
1997, Leal et al. 2016). Species intolerant to such stress may decrease in abundance or
become extinct while tolerant species may increase in abundance and become dominant
(e.g. Gimenez et al. 2015), resulting in decreasing community evenness as observed
here.
We found higher contribution of the turnover than nestedness component to total
beta diversity, i.e. the species composition in modified streams differed from reference
streams and was not just a subset of the former. Indeed, it is well known since a long
time by biologists and environmental managers that the community structure in more
pristine streams is different from the community structure in human-modified streams
(e.g., Lenat and Crawford 1994). One classic example are specific groups of
166
invertebrates such as EPT (Ephemeroptera, Plecoptera and Trichoptera insect orders)
and CO (Chironomidae insect family and Oligochaeta class) which are used as
biological indicators of streams water quality. Pristine streams are supposed to be
composed mainly by these pollution-sensitive taxa (EPT) while human-impacted
streams are mainly composed by the more pollution-tolerant taxa (CO) (Ruaro and
Gubiani 2013).
Despite some observational studies (e.g., Siqueira et al. 2015, Castro et al.
2018), we did not find support in our meta-analysis for biotic homogenization caused by
land-use modification in streams. One possibility for this finding is that modified
streams may differ a lot from each other even in a same study. For instance, an urban
stream may be channeled, but not others. Some may receive treated water while other
sewage. Some may run through parks and others through industrial areas. These factors
may promote high physical and chemical differentiation and reflect in different species
composition causing higher beta diversity among them (Barboza et al. 2015).
Within the land-use types, urbanization had the strongest negative impact on
stream communities: they had lower observed and extrapolated species richness, as well
as lower total abundance than in streams modified by forestry and agriculture.
Furthermore, we observed a comparable contribution of nestedness and turnover
components to beta diversity in pristine vs. urban streams, suggesting a process that
severely decreases species richness and at the same time changes their identities – but
the identity change is higher in forestry and agriculture land-use since their turnover
component is higher. Urban areas discharge high quantities of solid and liquid waste
into the water (Meyer et al. 2005) and large areas are devegetated and paved because of
urbanization (Marzluff and Ewing 2001). In this way, such modifications may drive
167
urban streams to be even harsher habitats than streams modified by agriculture and
forestry land use, where fewer species and individuals can survive (McKinney 2006).
We found land-use modifications decreased species richness in streams.
However, using measures incorporating changes in species abundance as well as the
beta diversity partition, we could infer that land-use decreased species richness by
increasing the dominance of tolerant species and changing the species composition. We
also did not find support to hypothesis of biotic homogenization caused by land-use
alteration in streams. We demonstrated that only using species richness may not be
enough to understanding the main mechanisms underlying biodiversity loss in streams
due land-use intensification and further studies should include additional components of
biodiversity change.
ACKNOWLEDGMENTS
We are thankful to all authors that kindly provided their data to conduct our
meta-analysis (see metadata in Table S1). We also thank F. May and D. Craven for help
with data analysis, and J. D. G. Trujillo for his helpful comments. DKP acknowledges
the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) and the
Programa de Doutorado Sanduíche no Exterior (PDSE) by the scholarships provided in
Brazil and in Germany. ASM received a research scholarship from Conselho Nacional
de Desenvolvimento Científico e Tecnológico (CNPq, no. 309412/2014-5).
LITERATURE CITED
Allan, D., D. Erickson, and J. Fay. 1997. The influence of catchment land use on stream
integrity across multiple spatial scales. Freshwater Biology 37: 149-161.
168
Barboza, L. G. A., R. P. Mormul, and J. Higuti. 2015. Beta diversity as a tool for
determining priority streams for management actions. Water Science & Technology
71: 1429-1435.
Baselga, A. 2012. The relationship between species replacement, dissimilarity derived
from nestedness, and nestedness. Global Ecology and Biogeography 21:1223–1232.
Baselga, A., and C. D. L. Orme. 2015. betapart: an R package for the study of beta
diversity. Methods in Ecology and Evolution 3: 808–812.
Bates, D., M. Maechler, B. Bolker, and S. Walker. 2014. lme4: Linear mixed-effects
models using Eigen and S4. R package version 1.0-6. http://CRAN.R-
project.org/package=lme4
Becker, R. A., and A. R. R. Wilks version by Ray Brownrigg. Enhancements by T. P
Minka and A. Deckmyn. 2017. maps: Draw Geographical Maps. R package version
3.2.0. https://CRAN.R-project.org/package=maps
Blowes, S. A., J. Belmaker, and J. M. Chase. 2017. Global reef fish richness gradients
emerge from divergent and scale-dependent component changes. Proceeding of the
Royal Society B Biological Sciences 284: 1-9.
Borenstein, M., L. V. Hedges, J. P. T. Higgins, and H. R. Rothstein. 2009. Introduction
to meta-analysis, 1st ed. Wiley, Chichester.
Castro, D. M. P., S. Dolédec, and M. Callisto. 2018. Land cover disturbance
homogenizes aquatic insect functional structure in neotropical savanna streams.
Ecological Indicators 84: 573–582.
Chase, J. M., and T. M. Knight. 2013. Scale-dependent effect sizes of ecological drivers
on biodiversity: why standardised sampling is not enough. Ecology Letters 16: 17–
26.
169
Chao, A. 1984. Nonparametric estimation of the number of classes in a population.
Scandinavian Journal of Statistics 11: 265–270.
Colwell, R. K., C. X. Mao, and J. Chang. 2004. Interpolating, extrapolating, and
comparing incidence-based species accumulation curves. Ecology 85:2717-2727.
Corbi, J. J., P. Kleine, and S. Trivinho-Strixino. 2013. Are aquatic insect species
sensitive to banana plant cultivation? Ecological Indicators 25: 156–161.
Gerstner, K., C. F Dormann, A. Stein, A. M. Manceur, and R. Seppelt. 2014. Effects of
land use on plant diversity – A global meta-analysis. Journal of Applied Ecology
51: 1690–1700.
Gimenez, B. C. G., F. A. Lansac-Tôha, and J. Higuti. 2015. Effect of land use on the
composition, diversity and abundance of insects drifting in neotropical streams.
Brazilian Journal of Biology 75: 52-59.
Gossner, M. M., et al. 2016. Land-use intensification causes multitrophic
homogenization of grassland communities. Nature 540: 266-269.
Gutiérrez-Cánovas, C., A. Millán, J. Velasco, I. P. Vaughan, and S. J. Ormerod. 2013.
Contrasting effects of natural and anthropogenic stressors on beta diversity in river
organisms. Global Ecology and Biogeography 22: 796–805.
Hillebrand, H., et al. 2017. Biodiversity change is uncoupled from species richness
trends: Consequences for conservation and monitoring. Journal of Applied Ecology
55: 169–184.
Hurlbert, S. H. 1971. The nonconcept of species diversity: a critique and alternative
parameters. Ecology 52: 577–586.
Konopik, O., I. Steffan-Dewenter, and U. Grafe. 2015. Effects of logging and oil palm
expansion on stream frog communities on Borneo, Southeast Asia. Biotropica 47:
636–643
170
Lajeunesse, M. J. 2011. On the meta-analysis of response ratios for studies with
correlated and multi-group designs. Ecology 92: 2049-55.
Leal, C.G., et al. 2016. Multi-scale assessment of human-induced changes toAmazonian
instream habitats. Landscape Ecology 31: 1725–1745.
Lenat, D. R., and J. K. Crawford. 1994. Effects of land use on water quality and aquatic
biota of three North Carolina Piedmont streams. Hydrobiologia 294: 185–199.
Lougheed, V. L., M. D. Mcintosh, C. A. Parker, and R. J. Stevenson. 2008. Wetland
degradation leads to homogenization of the biota at local and landscape scales.
Freshwater Biology 53: 2402–2413.
Martins, R. T., S. R. M. Couceiro, A. S. Melo, M. P. Moreira, and N. Hamada 2017.
Effects of urbanization on stream benthic invertebrate communities in Central
Amazon. Ecological Indicators 73: 480–491.
Marzluff, J. M., and K. Ewing 2001. Restoration of fragmented landscapes for the
conservation of birds: a general framework and specific recommendations for
urbanizing landscapes. Restoration Ecology 9: 280–292.
McKinney, M. L., and J. L. Lockwood. 1999. Biotic homogenization: A few winners
replacing many losers in the next mass extinction. Trends in Ecology and Evolution
14: 450–453.
McKinney, M. L. 2006. Urbanization as a major cause of biotic homogenization.
Biological Conservation, 127: 247–260.
Menge, B.A., et al. 2009. Terrestrial ecologists ignore aquatic literature: Asymmetry in
citation breadth in ecological publications and implications for generality and
progress in ecology. Journal of Experimental Marine Biology and Ecology 377: 93-
100.
171
Meyer, J. L., M. J. Paul, and W. K. Taulbee. 2005. Stream ecosystem function in
urbanizing landscapes. Journal of the North American Benthological Society 24:
602-612.
Newbold, T., et al. 2015. Global effects of land use on local terrestrial biodiversity.
Nature 520: 45–50.
Orwin, R. G. 1983. A fail-safe N for effect size in meta-analysis. Journal of Educational
Statistics 8:157-159.
Oksanen, J., et al. 2013. vegan: community ecology package. https://cran.r-
project.org/web/packages/vegan.
Powell, K. I., J. M. Chase, and T. M. Knight. 2013. Invasive plants have scale-
dependent effects on diversity by altering species-area relationships. Science 339:
316-318.
Preston, F. W. 1948. The Commonness, And Rarity, of Species. Ecology 29: 254–283.
Prudente, B. S., P. S. Pompeu, L. Juen, and L. F. A. Montag. 2017. Effects of reduced-
impact logging on physical habitat and fish assemblages in streams of Eastern
Amazonia. Freshwater Biology 62: 303–316.
R Core Team. 2017. R: a language and environment for statistical computing. R
Foundation for Statistical Computing, Vienna, Austria. www.r-project.org
Ruaro, R., and E. A. Gubiani. 2013. A scientometric assessment of 30 years of the
index of biotic integrity in aquatic. Ecological Indicators 29: 105–110.
Sala, O. E., et al. 2000. Global Biodiversity Scenarios for the Year 2100. Science 287:
1770-1774.
Schuler, M. S., J. M. Chase, and T. M. Knight. 2014. More individuals drive the species
energy–area relationship in an experimental zooplankton community. Oikos 124:
1065–107.
172
Siqueira, T., C. G. T. Lacerda, and V. S. Saito. 2015. How does landscape modification
induce biological homogenization in tropical stream metacommunities? Biotropica
47: 509–516.
Solar, R. B. C., et al. 2015. How pervasive is biotic homogenization in human-modified
tropical forest landscapes? Ecology Letters 18: 1108–1118.
Srivastava, D. S., and J. H. Lawton. 1998. Why more productive sites have more
species: an experimental test of theory using tree-hole communities. American
Naturalist 152: 510–529.
Viechtbauer, W. 2010. Conducting meta-analyses in R with the metafor package.
Journal of Statistic Software 36:1-48.
Wickham, H. 2009. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New
York.
Whittaker, R. H. 1960. Vegetation of the Siskiyou mountains, Oregon and California.
Ecological Monographs 30: 279–338.
173
SUPPLEMENTARY MATERIAL
TABLE S1. Metadata of 39 studies included in our meta-analysis. Agr= Agriculture,
Urb= Urbanization; For= Forestry.
First author Year Journal Land use type Taxa Country
Astudillo 2016 Hydrobiologia Agr Invertebrates Mexico
Baldi 2015 S Am J Herpetol Agr Frogs Brazil
Baumgartner 2017 Aquatic Sciences Agr e Urb Invertebrates Switzerland
Benstead 2003 Ecol Appl Agr Invertebrates Madagascar
Bere 2011 Water SA Urb Algae Brazil
Bertaso 2015 Revista Bra Ent Agr Invertebrates Brazil
Chakona 2009 Hydrobiologia Agr e For Invertebrates Zimbabwe
Corbi 2013 Ecol Ind Agr Invertebrates Brazil
Corbi 2017 Hydrobiologia Agr Invertebrates Brazil
Dias 2010 Neotropical Ichtyology Agr Fishes Brazil
Dias-Silva 2010 Zoologia Agr Invertebrates Brazil
Ejsmont-Karabin 1998 Hydrobiologia Agr Invertebrates Poland
Fraser
Thesis For Invertebrates New Zealand
Garcia 2017 Hydrobiologia Agr Invertebrates Mexico
Goldschmidt 2016 Limnologica Agr e Urb Invertebrates Panama
Hall 2001 New Zeal J Mar Fresh Agr, Urb e For Invertebrates New Zealand
Iniguez-Armijos 2016 Ecol Evol Agr e Urb Invertebrates Equador
Jingutt 2012 Sci Total Environ Agr e For Invertebrates Borneo
Konopik 2015 Biotropica For Frogs Borneo
Lenat 1994 Hydrobiologia Agr e Urb Fishes USA
Lorion 2009 Freshw Biol Agr Invertebrates Costa Rica
Martins 2017 Ecol Ind Urb Invertebrates Brazil
Mesa 2010 Hydrobiologia Agr Invertebrates Argentina
Morgan 2005 J N Am Benthol Soc Urb Fishes USA
Muehlbauer 2012 Freshw Science Urb Invertebrates USA
Mykra 2016 Ecosphere Agr e For Fungi Finland
Nogueira 2016 Environ Monit Assess For Invertebrates Brazil
Rawi 2013 Aquatic Ecology For Invertebrates Malaysia
Rodriguez 2017 Acta Biol Colombiana Agr e For Macrophytes Colombia
Roque 2015 Neotropical Ent For Invertebrates Brazil
Shearer 2011 New Zeal J Mar Fresh For e Agr Invertebrates New Zealand
Siegloch 2014 Na Acad Bra Cienc For e Agr Invertebrates Brazil
Silva 1995 Amazoniana Urb Fishes Brazil
Siqueira 2015 Biotropica Agr e For Invertebrates Brazil
Song 2009 Aquatic Ecology Agr Invertebrates France
Teresa 2017 Ecol Ind Agr Fishes Brazil
Vandermyde 2015 Freshw Science For Invertebrates USA
Yule 2015 Freshw Science Urb Invertebrates Malaysia
174
TABLE S2. Fail-safe number (Orwin 1983) quantified using observed species richness,
ENSPIE, extrapolated species richness and total abundance at alpha scale.
Fail-safe number
Observed species richness
ENSPIE
Extrapolated species richness
Total abundance
77
76
77
77
175
FIG. S1. Summary of the systematic review steps. (a) PRISMA flowchart (Moher et al.
2009) summarizing study inclusion and exclusion. (b) Study exclusion criteria after full-
text and data screening.
176
FIG. S2. Effect sizes (log-response ratio) and 95% confidence interval (vertical bars)
estimated separately for agriculture, forestry and urbanization land-use types using
ENSPIE, extrapolated and observed species richness and total abundance at gamma
scale. We resampled the streams to obtain the same number of streams across land-use
types within-studies.
177
FIG. S3. Total beta diversity measured between reference and land use streams using
incidence (Sorensen dissimilarity in blue) and abundance data (Bray-Curtis dissimilarity
in pink).
178
FIG. S4. Funnel plot of observed species richness (a), ENSPIE (b), extrapolated species
richness (c) and total abundance (d) at alpha scale.
179
CONSIDERAÇÕES FINAIS
A homogeneização biótica em comunidades aquáticas é um processo que pode ocorrer
em diferentes ecossistemas e ser promovido por diferentes mecanismos. No primeiro
capítulo, por meio de uma revisão teórica, encontrei que são diversas as causas de
homogeneização biótica em ambientes aquáticos continentais (e.g. introdução de
espécies não-nativas, barragem, uso do solo, produtividade, mudanças climáticas,
inundações e também secas). Também encontrei que, embora a homogeneização
taxonômica medida entre comunidades seja a forma mais comum de homogeneização
biótica, as comunidades também podem se tornar mais similares filogeneticamente ou
funcionalmente e as populações também podem se tornar geneticamente mais similares.
Ainda, encontrei que o aumento da similaridade em ecossistemas aquáticos continentais
pode gerar consequências ecológicas (e.g. afetar comunidades de presas e parasitas
associadas), evolutivas (e.g. homogeneização genética pode prejudicar possível
especiação) e até mesmo sociais (e.g. prejuízo na pesca).
No segundo capítulo, encontrei que a simplificação de hábitats pode causar
homogeneização biótica da comunidade de algas perifíticas, embora o resultado
dependa do índice empregado. Sugiro que a maior diversidade beta entre habitats
complexos possa ser devida a maior estocasticidade na história de colonização das algas
em conjunto com efeitos prioritários. Já a menor diversidade beta entre habitats simples
pode estar relacionada a processos determinísticos, em que o mesmo conjunto reduzido
de espécies tolerantes a velocidade de fluxo ocorre entre os habitats, reduzindo a
dissimilaridade. Ainda, encontrei que levar em consideração as diferentes estratégias de
vida das algas é importante, pois podem evidenciar diferentes mecanismos.
No terceiro capítulo, encontrei que cheias não homogeneizaram as comunidades
de macrófitas ou de zooplâncton no espaço. No entanto, mostrei que uma mesma lagoa
180
foi mais parecida ao longo do tempo entre períodos de cheia do que entre períodos de
seca para a maioria dos grupos. No quarto capítulo, encontrei maior diversidade beta
taxonômica entre riachos tropicais, mas maior diversidade beta funcional entre os
riachos boreais. No entanto, não encontrei homogeneização biótica promovida pelo uso
do solo em decorrência do aumento da degradação ambiental ou redução da
heterogeneidade. Sugeri que esse resultado pode ser devido a dois mecanismos atuando
simultaneamente e que podem gerar similar diversidade beta: processos estocásticos
entre os riachos florestados com condições favoráveis a maioria das espécies e
heterogeneidade entre os riachos modificados pelo uso do solo.
Finalmente, no quinto e último capítulo, encontrei menor riqueza e maior
dominância de espécies em riachos modificados pelo uso do solo. Também constatei
que a composição de espécies é diferente entre riachos florestados e modificados,
provavelmente devido à tolerância das espécies ao uso do solo. No entanto, não
encontrei homogeneização biótica causada por mudanças no uso do solo, e especulo que
seja porque os riachos modificados podem diferir muito em termos de tipo e intensidade
de distúrbio, o que pode refletir em distinta composição de espécies entre eles.
Em relação à tese como um todo, uma importante constatação é que o uso de
múltiplos índices de dissimilaridade se faz necessário para descrevermos
adequadamente os padrões de diversidade beta. Embora seja uma questão mais
metodológica, isso pode afetar grandemente as conclusões dos trabalhos ecológicos. Em
especial, incorporar índices que levem em consideração as possíveis diferenças em
riqueza de espécies entre as comunidades é primordial para evitar a confusão entre os
mecanismos que promovem de fato a substituição de espécies dos que promovem
apenas a perda de espécies entre as comunidades sem alterar a composição de espécies.
181
Embora processos estocásticos e determinísticos possam ser difíceis de serem
mensurados, especulei ao longo dos capítulos que eles podem atuar simultaneamente na
estruturação das comunidades aquáticas sujeitas ou não a possíveis causas de
homogeneização biótica e que ambos podem gerar similar diversidade beta. Uma outra
recomendação provinda de alguns capítulos é que levar em consideração as
características das espécies aquáticas, como as estratégias de vida e tamanho, pode
auxiliar no melhor entendimento da estruturação das comunidades.
Por fim, concluo que mesmo que o processo de homogeneização biótica não seja
detectado, a sua investigação é ainda assim válida pois pode resultar no reconhecimento
de outros processos que alteram a biodiversidade (e.g. diferentes tipos ou intensidade de
distúrbios). Tal investigação pode ser experimental, com dados observacionais ou por
meio de uma revisão. Enfim, homogeneização biótica em ecossistemas aquáticos
continentais, sendo um processo natural ou de perda da biodiversidade mediada pela
pressão antrópica, merece atenção em um mundo em mudança que vivenciamos hoje.
