determinantes ecolÓgicos, evolutivos e histÓrico ... · gradiente latitudinal de diversidade, bem...
TRANSCRIPT
UNIVERSIDADE FEDERAL DE GOIÁS
PRÓ-REITORIA DE PESQUISA E PÓS-GRADUAÇÃO
INSTITUTO DE CIÊNCIAS BIOLÓGICAS
PROGRAMA DE PÓS-GRADUAÇÃO EM ECOLOGIA & EVOLUÇÃO
Franciele Parreira Peixoto
DETERMINANTES ECOLÓGICOS, EVOLUTIVOS E
HISTÓRICO-BIOGEOGRÁFICOS DOS PADRÕES DE
DIVERSIDADE DE MAMÍFEROS TERRESTRES EM
DIFERENTES ESCALAS
Orientador: Dr. Marcus Vinícius Cianciaruso
Coorientador: Dr. Crisóforo Fabrício Villalobos
GOIÂNIA - GO
MARÇO – 2017
UNIVERSIDADE FEDERAL DE GOIÁS
PRÓ-REITORIA DE PESQUISA E PÓS-GRADUAÇÃO
INSTITUTO DE CIÊNCIAS BIOLÓGICAS
PROGRAMA DE PÓS-GRADUAÇÃO EM ECOLOGIA & EVOLUÇÃO
Franciele Parreira Peixoto
DETERMINANTES ECOLÓGICOS, EVOLUTIVOS E
HISTÓRICO-BIOGEOGRÁFICOS DOS PADRÕES DE
DIVERSIDADE DE MAMÍFEROS TERRESTRES EM
DIFERENTES ESCALAS
Orientador: Dr. Marcus Vinícius Cianciaruso
Coorientador: Dr. Crisóforo Fabrício Villalobos
Tese apresentada à Universidade Federal de
Goiás, como parte das exigências do Programa de
Pós-graduação em Ecologia e Evolução para
obtenção do título de Doutora.
GOIÂNIA - GO
MARÇO – 2017
ii
iii
iv
v
“It is not the mountain we conquer, but ourselves”.
Edmund Hillary
Dedico ao que representa a escalada em rocha na minha vida.
É mais que um esporte, é um estilo de vida. Fez ressurgir no meu ser
o instinto primata e minha conexão com a natureza. Ao longo do
doutorado, foi um modo de obter autoconfiança e força
(principalmente para minhas costas e braços, que sofriam com horas
na frente do computador) para enfrentar os obstáculos. Aprendi que
tudo é possível, desde que se deseje profundamente, e que a subida é
mais importante que o pico, apesar da vista ser recompensadora.
vi
AGRADECIMENTOS
Em quatro anos de doutorado, muitas pessoas passaram pela minha vida e
contribuíram para a concretização desse projeto. Muitas dessas pessoas não serão citadas
aqui, por falta de espaço ou falha de memória, mas mesmo assim agradeço por fazerem
parte do meu crescimento pessoal e profissional.
Agradeço:
A minha família querida, sem a qual não teria tido a chance de estudar e realizar
meus projetos. Um especial agradecimento aos meus pais, Celso Alves Peixoto e Maria
Aparecida Parreira, e meus avós maternos, Maria Euripa de Assunção Salgado e
Hermógenes Parreira de Assunção.
Ao meu companheiro, Tiago Segantini Zaterka, por ter estado ao meu lado durante
os últimos três anos, me apoiando com muita compreensão e amor, mesmo nos piores
momentos. Por ter ficado comigo em casa por vários finais de semana e ter se privado de
várias oportunidades de viagens e escaladas.
Aos meus mais que amigos, Elisa Barreto, José Hidasi Neto, Pedro Henrique
Pereira Braga, Vinícios Macedo e Viviane Neves, que estiveram do meu lado nesses anos
em todos os tipos de situação, me ajudando a superar os obstáculos impostos pela vida.
Aos amigos Advaldo Carlos de Souza-Neto, José Hidasi Neto, Pedro Henrique
Pereira Braga e Poliana Mendes, que também foram um pouco orientadores e sempre me
ajudaram com análises e discussões.
Aos integrantes da minha família Hogar, por terem me acolhido nesses últimos
três anos com tanta amizade e carinho. Por cederem uma mesinha na sala para que eu
vii
pudesse estudar e por aliviarem a carga de vários finais de semana, que passei estudando
o dia todo, com cafés, conversas, bolos, jogos e filmes.
A fundação de amparo a pesquisa do estado de Goiás (FAPEG) pela bolsa de
estudos concedida.
Ao órgão onde trabalho desde o mestrado, SECIMA, que me concedeu dispensa
de horários para que eu pudesse finalizar a pós-graduação. Um agradecimento especial a
superintendente Gabriela de Val pela compreensão em todos os momentos e aos meus
companheiros do GCP pelo carinho e por tornarem o ambiente de trabalho mais
aconchegante.
Ao meu orientador, Marcus Cianciaruso, por ter me aguentado durante quase 6
anos de parceria e por ter contribuido muito no meu crescimento pessoal e profissional.
Ao meu coorientador Fabrício Villalobos por estar sempre disposto a ajudar e me
ensinar. Aprendi muito com você e só tenho a agradecer pela parceria
Aos meus co-autores que contribuíram para a qualidade do presente trabalho,
principalmente os professores Joaquin Hortal e Adriano Melo.
A todos os integrantes do Laboratório de Ecologia de Comunidades, professores
e alunos, que tornaram o ambiente de trabalho agradável e prazeroso.
A Universidade Federal de Goiás, em especial ao programa de pós-graduação em
Ecologia e Evolução que me propiciou anos maravilhosos, pessoas incríveis e muito
aprendizado.
viii
RESUMO GERAL
O conceito de nicho ecológico descreve o conjunto de condições bióticas e
abióticas nas quais as espécies podem manter populações e define a área de distribuição
das espécies. O nicho ecológico é multidimensional e, desse modo, é função de várias
características das espécies. Essas características podem mudar rapidamente ou podem
variar de forma muito lenta, permanecendo conservadas ao longo do tempo evolutivo, o
que caracteriza a conservação de nicho (i.e., a tendência de as espécies manterem o nicho
ancestral ao longo do tempo evolutivo). Reter o nicho ancestral implica em menor
capacidade do grupo em se adaptar fora dos limites de distribuição definidos pelo nicho
ecológico. Portanto, o modo como as características ligadas ao nicho evoluíram ao longo
do tempo é determinante para gerar padrões diferenciais de diversidade entre táxons. De
fato, a conservação de nicho tem sido levantada como fator relevante para explicar o
gradiente latitudinal de diversidade, bem como a formação da biota de diferentes habitats.
A conservação de tolerâncias ecológicas em algumas linhagens implica na herança de
uma capacidade limitada de colonizar e se estabelecer em diferentes habitats. Por outro
lado, alguns grupos podem ser ecologicamente mais flexíveis e colonizar novos habitats
através da evolução de nicho, provavelmente relacionada a adaptações funcionais
específicas que possibilitam o estabelecimento das espécies. Este trabalho teve como
objetivo geral avaliar padrões de diversidade de mamíferos terrestres para inferir acerca
dos principais processos ecológicos, evolutivos e histórico-biogeográficos atuantes.
ix
SUMÁRIO
INTRODUÇÃO GERAL ............................................................................................... 1
REFERÊNCIAS ............................................................................................................ 4
CAPÍTULO 01: PHYLOGENETIC CONSERVATISM OF CLIMATIC NICHE
IN BATS .......................................................................................................................... 7
ABSTRACT .................................................................................................................. 8
INTRODUCTION ........................................................................................................ 9
METHODS ................................................................................................................. 13
RESULTS ................................................................................................................... 21
DISCUSSION ............................................................................................................. 25
CONCLUSION ........................................................................................................... 30
REFERENCES ........................................................................................................... 31
SUPPORTING INFORMATION ............................................................................... 37
CAPÍTULO 02: GEOGRAPHICAL PATTERNS OF PHYLOGENETIC BETA
DIVERSITY COMPONENTS IN TERRESTRIAL MAMMALS .......................... 45
ABSTRACT ................................................................................................................ 46
INTRODUCTION ...................................................................................................... 47
METHODS ................................................................................................................. 52
RESULTS ................................................................................................................... 55
DISCUSSION ............................................................................................................. 60
CONCLUSION ........................................................................................................... 66
REFERENCES ........................................................................................................... 66
x
SUPPORTING INFORMATION ............................................................................... 71
CAPÍTULO 03: PHYLOGENETIC AND FUNCTIONAL STRUCTURE OF
AFRICAN MAMMAL ASSEMBLAGES: THE IMPRINT OF HISTORICAL
CLIMATIC CHANGES ON HABITAT FORMATION .......................................... 85
ABSTRACT ................................................................................................................ 86
INTRODUCTION ...................................................................................................... 87
METHODS ................................................................................................................. 92
RESULTS ................................................................................................................... 96
DISCUSSION ........................................................................................................... 100
CONCLUSION ......................................................................................................... 107
REFERENCES ......................................................................................................... 108
SUPPORTING INFORMATION ............................................................................. 114
CONCLUSÃO GERAL ............................................................................................. 123
1
INTRODUÇÃO GERAL
O conceito de nicho ecológico (sensu Hutchinson; Holt, 2009) é central em
ecologia e evolução e descreve o conjunto de condições bióticas e abióticas nas quais as
espécies podem persistir e manter populações. Nesse contexto, o nicho das espécies pode
ser classificado em nicho Grinnelliano e nicho Eltoniano (sensu Soberón, 2007). O nicho
Grinnelliano, também chamado não interativo, diz respeito às variáveis ambientais (e.g.
temperatura, precipitação, pH, etc.) que geralmente definem os limites de distribuição das
espécies em escalas amplas. Por outro lado, o nicho Eltoniano está relacionado a
interações bióticas em escalas locais (e.g. competição, predação, parasitismo, etc.) e à
dinâmica de consumo de recursos (Soberón, 2007). O nicho ecológico é o que,
primariamente, define a área de distribuição das espécies, mas os fatores abióticos (nicho
Grinnelliano) parecem ser mais importantes para definir os limites de distribuição das
espécies em grandes escalas e determinar padrões biogeográficos (Pearson & Dawson,
2003; Soberón & Peterson, 2005; Wiens, 2011). Entretanto, é possível que as variáveis
climáticas possam atuar indiretamente em combinação com fatores bióticos, por exemplo,
determinando a distribuição de recursos no espaço (Wiens, 2011).
O nicho ecológico é multidimensional e, desse modo, é função de várias
características das espécies, como tolerância de temperatura, adaptações de
forrageamento, hábito e outras (sensu Hutchinson; Holt, 2009). Essas características
podem mudar rapidamente ou podem variar de forma muito lenta, permanecendo
conservadas ao longo do tempo evolutivo (Hansen, 1997; Losos, 2008; Kozak & Wiens,
2010; Wiens et al., 2010; Peterson, 2011). Quando características ligadas ao nicho
permanecem semelhantes podemos dizer que há “conservação de nicho”, que pode ser
definida como a tendência de as espécies manterem o nicho ancestral ao longo do tempo
evolutivo (Wiens & Graham, 2005). A conservação de nicho pode ocorrer por meio de
2
um processo de estase evolutiva, no qual a evolução da característica é limitada a um
ótimo adaptativo (Hansen, 1997; Kozak & Wiens, 2010). O modo e a velocidade com que
essas características relacionadas ao nicho evoluem varia entre clados e até mesmo dentro
dos clados, como por exemplo entre famílias dentro da mesma ordem (Diniz-Filho et al.,
2010).
As espécies não podem dispersar fora dos limites do seu nicho ecológico, pois o
mesmo define as condições nas quais elas podem sobreviver. Desse modo, as barreiras à
dispersão (e.g. entre regiões biogeográficas) representam áreas fora dos limites do nicho
ecológico, que são espécie-específico (Wiens, 2011). Reter o nicho ancestral implicará
em menor capacidade do grupo em se adaptar fora desses limites (Wiens, 2011). Portanto,
o modo como as características ligadas ao nicho evoluíram ao longo do tempo é
determinante para gerar padrões diferenciais de diversidade entre táxons (Wiens &
Donoghue, 2004). De fato, a conservação do nicho climático (Grinneliano) tem sido
levantada como fator importante para explicar o gradiente latitudinal de diversidade
(conservação de nicho tropical; Wiens & Donoghue, 2004). Essa hipótese prediz que os
grupos com maior diversidade em menores latitudes teriam se originado em regiões
tropicais e que, devido a conservação de nicho, poucas linhagens teriam evoluído
características que as permitissem conquistar climas mais frios ou áridos em zonas
extratropicais (Wiens & Donoghue, 2004). A prevalência de maior diversidade de
espécies em regiões tropicais se daria pelo fato de que a maioria dos grupos se originaram
nesse clima, que prevaleceu por um longo período em grande parte da superfície terrestre
(Behrensmeyer et al., 1992). Pensando na mesma lógica, a conservação de nicho
climático também poderia gerar os padrões de diminuição de diversidade com o aumento
das altitudes (e.g. Kozak & Wiens, 2010). Desse modo, a conservação/evolução de nicho
climático entre diferentes grupos taxonômicos causaria padrões de dissimilaridade de
3
biota nos limites entre zonas climáticas, bem como entre áreas de diferentes elevações
(Wiens & Donoghue, 2004).
O modo de evolução de nicho também pode ter grande influência na formação da
biota de diferentes habitats. A conservação de tolerâncias ecológicas em algumas
linhagens implica na herança de uma capacidade limitada de colonizar e se estabelecer
em diferentes habitats (conservação de habitat), como foi demonstrado para plantas (Crisp
et al., 2009). Por outro lado, alguns grupos podem ser ecologicamente mais flexíveis e
colonizar novos habitats através da evolução de nicho (Wiens & Donoghue, 2004). A
capacidade diferencial dos clados em se estabelecer em outros habitats, somado com a
história climática das regiões, podem influenciar na formação do pool de linhagens entre
biomas e determinar a estrutura de assembleias no presente (Wiens & Donoghue, 2004;
Gerhold et al., 2015). Conservar o nicho ancestral pode comprometer o potencial
evolutivo do grupo se houver uma contração do habitat, por exemplo, em decorrência de
mudanças climáticas (Crisp et al., 2009). A perda de área do habitat ancestral implicaria
em extinção de espécies ou intensa redução das populações, como parece ter ocorrido
para muitas linhagens de florestas tropicais da África (Janis, 1993; Plana, 2004; Lawes et
al., 2007). Algumas linhagens permaneceram confinadas a pequenas áreas de floresta,
comumente chamadas de refúgios (e.g. Anthony et al., 2007; Johnston & Anthony, 2012;
Jacquet et al., 2014). Por outro lado, algumas linhagens podem ter se adaptado aos novos
habitats formados (Kamilar et al., 2009; Johnston & Anthony, 2012; Cantalapiedra et al.,
2014). O continente Africano teve a maior taxa de perda de área de floresta tropical e
manteve a maior área de savana ao longo do Cenozoico devido a mudanças climáticas e
à prevalência de um clima mais árido ao longo da África equatorial (Maley, 1996;
Kissling et al., 2012; Willis et al., 2013). Os limites entre floresta tropical e ambientes
savânicos africanos estão associados a um gradiente ambiental de aridez que,
4
provavelmente, exigiu adaptações funcionais específicas que possibilitassem a
colonização e estabelecimento das espécies (e.g. Cantalapiedra et al., 2014).
Este trabalho teve como objetivo geral avaliar padrões globais, regionais e locais
de diversidade de mamíferos terrestres para inferir acerca dos principais processos
ecológicos, evolutivos e histórico-biogeográficos atuantes. No primeiro capítulo testei a
hipótese de que a distribuição atual das espécies é resultado de um processo de estase na
evolução de nicho ambiental (Grinnelliano) do grupo. Este capítulo está em revisão no
periódico Global Ecology and Biogeography. No segundo capítulo investiguei o
gradiente global de dissimilaridade filogenética e testei hipóteses relacionando esses
padrões (dissimilaridade real e dissimilaridade devido à diferença de diversidade
filogenética) com processos históricos que originaram o padrão latitudinal de diversidade.
Além disso, testei se a variação desses padrões para três ordens de mamíferos com
diferentes histórias evolutivas está relacionada com a capacidades de dispersão. Este
capítulo foi aceito para publicação no periódico Global Ecology and Biogeography. O
último capítulo está mais centrado em padrões locais de diversidade filogenética e
funcional de assembleias de mamíferos no continente africano. Nesse capítulo investiguei
como esses padrões locais refletem a história climática do continente e a formação do
conjunto de linhagens dos biomas nos quais as assembleias estão inseridas, além dos
processos ecológicos determinantes na coexistência das espécies no nível de assembleia.
REFERÊNCIAS
Anthony, N.M., Johnson-Bawe, M., Jeffery, K., Clifford, S.L., Abernethy, K. a, Tutin,
C.E., Lahm, S. a, White, L.J.T., Utley, J.F., Wickings, E.J. & Bruford, M.W. (2007)
The role of Pleistocene refugia and rivers in shaping gorilla genetic diversity in
central Africa. Proceedings of the National Academy of Sciences USA, 104, 20432–
20436.
Behrensmeyer, A.K., Damuth, J.D. & DiMichele, W.A. (1992) Terrestrial ecosystems
through time : evolutionary paleoecology of terrestrial plants and animals,
5
University of Chicago Press, Chicago.
Cantalapiedra, J.L., Fitzjohn, R.G., Kuhn, T.S., Fernández, M.H., DeMiguel, D., Azanza,
B., Morales, J. & Mooers, A.Ø. (2014) Dietary innovations spurred the
diversification of ruminants during the Caenozoic. Proceedings of the Royal Society
B: Biological Sciences, 281, 20132746.
Crisp, M.D., Arroyo, M.T.K., Cook, L.G., Gandolfo, M. a, Jordan, G.J., McGlone, M.S.,
Weston, P.H., Westoby, M., Wilf, P. & Linder, H.P. (2009) Phylogenetic biome
conservatism on a global scale. Nature, 458, 754–6.
Diniz-Filho, J.A.F., Terribile, L.C., Da Cruz, M.J.R. & Vieira, L.C.G. (2010) Hidden
patterns of phylogenetic non-stationarity overwhelm comparative analyses of niche
conservatism and divergence. Global Ecology and Biogeography, 19, 916–926.
Gerhold, P., Cahill, J.F., Winter, M., Bartish, I. V & Prinzing, A. (2015) Phylogenetic
patterns are not proxies of community assembly mechanisms (they are far better).
Functional Ecology, 29, 600–614.
Hansen, T.F. (1997) Stabilizing selection and the comparative analysis of adaptetion.
Evolution, 51, 1341–1351.
Holt, R.D. (2009) Bringing the Hutchinsonian niche into the 21st century : ecological and
evolutionary perspectives. Proceedings of the National Academy of Sciences USA,
106, 19659–19665.
Jacquet, F., Nicolas, V., Colyn, M., Kadjo, B., Hutterer, R., Decher, J., Akpatou, B.,
Cruaud, C. & Denys, C. (2014) Forest refugia and riverine barriers promote
diversification in the West African pygmy shrew (Crocidura obscurior complex,
Soricomorpha). Zoologica Scripta, 43, 131–148.
Janis, C.M. (1993) Tertiary mammal evolution in the context of changing climates,
vegetation, and tectonic events. Annual Review of Ecology and Systematics, 24, 467–
500.
Johnston, A.R. & Anthony, N.M. (2012) A multi-locus species phylogeny of African
forest duikers in the subfamily Cephalophinae: evidence for a recent radiation in the
Pleistocene. BMC Evolutionary Biology, 12, 120.
Kamilar, J.M., Martin, S.K. & Tosi, A.J. (2009) Combining biogeographic and
phylogenetic data to examine primate speciation: an example using cercopithecin
monkeys. Biotropica, 41, 514–519.
Kissling, W.D., Eiserhardt, W.L., Baker, W.J., Borchsenius, F., Couvreur, T.L.P.,
Balslev, H. & Svenning, J.-C. (2012) Cenozoic imprints on the phylogenetic
structure of palm species assemblages worldwide. Proceedings of the National
Academy of Sciences USA, 109, 7379–7384.
Kozak, K.H. & Wiens, J.J. (2010) Niche conservatism drives elevational diversity
patterns in Appalachian salamanders. The American naturalist, 176, 40–54.
Lawes, M.J., Eeley, H.A.C., Findlay, N.J. & Forbes, D. (2007) Resilient forest faunal
communities in South Africa: a legacy of palaeoclimatic change and extinction
filtering? Journal of Biogeography, 34, 1246–1264.
Losos, J.B. (2008) Phylogenetic niche conservatism, phylogenetic signal and the
6
relationship between phylogenetic relatedness and ecological similarity among
species. Ecology Letters, 11, 995–1003.
Maley, J. (1996) The African rain forest - main characteristics of changes in vegetation
and and climate from the Upper Cretaceous to the Quaternary. Proceedings of The
Royal Society of Edinburgh, 104, 31–73.
Pearson, R.G. & Dawson, T.P. (2003) Predicting the impacts of climate change on the
distribution of species: are bioclimate envelope models useful? Global Ecology and
Biogeography, 12, 361–371.
Peterson, A.T. (2011) Ecological niche conservatism: a time-structured review of
evidence. Journal of Biogeography, 38, 817–827.
Plana, V. (2004) Mechanisms and tempo of evolution in the African Guineo-Congolian
rainforest. Philosophical transactions of the Royal Society of London B: Biological
sciences, 359, 1585–1594.
Soberón, J. (2007) Grinnellian and Eltonian niches and geographic distributions of
species. Ecology Letters, 10, 1115–1123.
Soberón, J. & Peterson, A.T. (2005) Interpretation of models models of fundamental
ecological niches and specie’ distributional areas. Biodiversity Informatics, 2, 1–10.
Wiens, J.J. (2011) The niche, biogeography and species interactions. Philosophical
transactions of the Royal Society B: Biological Sciences, 366, 2336–2350.
Wiens, J.J., Ackerly, D.D., Allen, A.P., Anacker, B.L., Buckley, L.B., Cornell, H. V,
Damschen, E.I., Jonathan Davies, T., Grytnes, J.-A., Harrison, S.P., Hawkins, B. a,
Holt, R.D., McCain, C.M. & Stephens, P.R. (2010) Niche conservatism as an
emerging principle in ecology and conservation biology. Ecology letters, 13, 1310–
1324.
Wiens, J.J. & Donoghue, M.J. (2004) Historical biogeography, ecology and species
richness. Trends in ecology & evolution, 19, 639–644.
Wiens, J.J. & Graham, C.H. (2005) Niche conservatism: integrating evolution, ecology,
and conservation biology. Annual Review of Ecology, Evolution, and Systematics,
36, 519–539.
Willis, K.J., Bennett, K.D., Burrough, S.L., Macias-Fauria, M. & Tovar, C. (2013)
Determining the response of African biota to climate change : using the past to model
the future. Philosophical transactions of the royal society, 368, 20120491.
7
CAPÍTULO 01: PHYLOGENETIC
CONSERVATISM OF CLIMATIC NICHE
IN BATS
8
PHYLOGENETIC CONSERVATISM OF CLIMATIC NICHE IN BATS
Franciele P. Peixoto, Fabricio Villalobos, Marcus V. Cianciaruso
Em revisão no periódico Global Ecology and Biogeography
ABSTRACT
Aim: to evaluate the patterns of climatic niche conservatism in bats. Bats have been
regarded as showing strong niche conservatism, mainly by testing predictions on the
effect of niche conservatism in diversity gradients. However, no specific tests have been
conducted to investigate the extent to which bat species niches are evolutionary
conserved. We address this questions at different phylogenetic scales and using
phylogenetic and geographical approaches.
Location: worldwide
Methods: we used nine climatic variables to describe a multivariate representation of bat
climatic niches. We measured niche position, niche breadth and niche overlap between
sister species pairs (using Schoener’s D). We performed a Mantel test to verify if niche
overlap was related to phylogenetic proximity. At the order and family level, we tested
for phylogenetic signal using the K statistic and phylogenetic signal representation curve.
We also compared the fit of four evolutionary models.
Results: 42.8% of sister species pairs comparison showed that niche similarity was higher
than expected. However, we did not find a significant phylogenetic signal for niche
overlap. At deeper evolutionary scales, K values were smaller than one for NP and NB,
thus different than expected under Brownian motion. PSR and the fit of the OU model
supported a slower evolution than expected under BM, suggesting phylogenetic
conservatism, particularly for niche position. Vespertilionidae and Molossidae presented
similar results.
9
Main conclusions: our findings show that climatic niche position of bats presents
evidence for phylogenetic niche conservatism. In addition, we verified that the evolution
of climatic niches is non-stationary across the order Chiroptera, consistent with the
different histories of clades. We stress the importance of taking into account the method
of choice, the niche feature and the phylogenetic scale being evaluated when testing for
phylogenetic niche conservatism at higher taxonomic levels and its influence on
biodiversity gradients.
Keywords: Chiroptera, evolutionary stasis, Grinneallian niche, niche overlap,
phylogenetic signal, PSR, trait evolution models.
INTRODUCTION
Chiroptera is the second most species-rich mammalian order and the main taxon driving
the latitudinal diversity gradient of the whole mammalian class (Kaufman, 1995; Buckley
et al., 2010). Given the importance of bats in determining mammalian diversity gradients,
they have been considered as an ideal group for testing explanatory hypotheses for such
gradients (Willig et al., 2003a; Stevens, 2006; Buckley et al., 2010). One explanation for
the strong diversity gradient exhibited by bats is phylogenetic niche conservatism (PNC)
–the tendency of species to retain their ancestral niches over time (Wiens & Graham,
2005)–particularly in relation to the tropical niche conservatism hypothesis (Wiens &
Donoghue, 2004). For instance, Buckley et al. (2010) argued that the positive
temperature-richness relationship found for Chiroptera supported PNC as higher richness
accumulates in environments resembling the clades’ tropical ancestral niche. In the same
vein, Stevens (2006, 2011) found that bats inhabiting temperate species-poor regions are
more recently originated (i.e. younger) whereas bats in species-rich tropical regions are
more early diverged (i.e. older), thus supporting PNC in driving the bat diversity gradient.
10
Conversely, Pereira & Palmeirim (2013) recently refuted Stevens’ results, finding that
more early diverged species at higher latitudes and more recently derived species at lower
latitudes. All of these studies, however, have evaluated bats PNC by testing predictions
on its effect in diversity gradients and not specifically testing for niche similarity among
related species. Therefore, an important unanswered question is to what extent do bat
species resemble each other niches.
The ‘niche’ concept is a complex term referring to several aspects of species
ecologies, mainly to the factors allowing their existence and their impacts on those factors
(Chase & Leibold, 2003; Peterson, 2011). Based on these two broad aspects, Soberón
(2007) defined two niche classes: Grinnellian and Eltonian niches. The Grinnellian niche
refers to the non-interactive environmental variables determining species distributions at
the geographical scale, whereas the Eltonian niche refers to the resource-consumer
dynamics and biotic interactions determining distributions at the local scale (Soberón,
2007). Given the coarse spatial structure of the environmental variables defining the
Grinnellian niche compared to the local and complex dynamics of Eltonian niches
(Soberón, 2007), the former is often more tractable and have thus being more widely
studied (Peterson, 2011; Pyron et al., 2015; Olalla-Tárraga et al., 2017). For example,
climatic variables have been extensively used to describe Grinnellian niche, mainly
because of their relevance in setting distributional limits of species or the resources they
rely upon (e.g. fruits for frugivore bats) (Fleming, 1973; Wiens, 2011). Indeed, most PNC
studies have focused on Climatic niches as characterized by the geographical distribution
of species (Hutchinson’s duality; Colwell & Rangel, 2009) and suggested that PNC may
be a widespread biogeographical pattern (Wiens & Graham, 2005; Wiens et al., 2010;
Peterson, 2011). Whether PNC prevails or not, the appropriate methods for testing it and
11
even its meaning are still subject to strong debates (Losos, 2008; Wiens, 2008; Cooper et
al., 2010; Münkemüller et al., 2015; Pyron et al., 2015).
Given the relevance of PNC for addressing many important issues in ecology and
evolution, from species distributions and adaptation to diversity gradients (Pyron et al.,
2015), the application of a particular method for testing it depends on the question of
interest (Wiens et al., 2010). In addition, such a question should also explicitly consider
the temporal or phylogenetic scale as the degree of PNC can vary according to the extent
of phylogenetic inclusiveness of a study (Losos, 2008; Peterson, 2011). For example,
studies at shallow phylogenetic scales (e.g. at species level) can either reveal similar,
stronger or weaker PNC patterns compared to more inclusive studies (e.g. at family level)
as a result of variation in rates of niche evolution (i.e. phylogenetic non-stationarity)
(Diniz-Filho et al., 2010; Diniz-filho et al., 2015). In the case of niche similarity among
related species, two different approaches have been mostly applied: those based on niche
variation across a phylogenetic tree and those based on comparing species distributional
patterns (Wiens et al., 2010). Importantly, both of these approaches should be conducted
on an explicit phylogenetic framework (Diniz-Filho et al., 2010).
Phylogenetic tree approaches include tests of ‘phylogenetic signal’ – “tendency
for related species to resemble each other more than they resemble species drawn at
random from the tree” (Blomberg & Garland, 2002)— and the comparison of the relative
fit of evolutionary models to the data (Wiens et al., 2010; Münkemüller et al., 2015).
Although tests of phylogenetic signal have been widely applied to study PNC (Buckley
et al., 2010; Cooper et al., 2010; Hof et al., 2010; Olalla-Tárraga et al., 2011), it has been
shown that such test can be misleading since no signal could either mean random variation
(no PNC) or stasis (strong PNC) (Revell et al., 2008; Münkemüller et al., 2015).
Alternatively, the relative fit of evolutionary models seems to be a more appropriate way
12
to study PNC (Kozak & Wiens, 2010; Münkemüller et al., 2015). The most common
models are the Brownian motion (BM), a stasis model such as the Ornstein-Uhlenbeck
(OU), and a random variation model (white noise; WN) (Wiens et al., 2010). Under a BM
model, in which tests of phylogenetic signal are based, traits evolve under a random walk
with differences accumulating over time (Felsenstein, 1985). The OU model describes
constrained evolution whereby traits are pulled toward an optimal value, thus evolving
slower than in a Brownian motion model. Accordingly, the fit of an OU model can be
interpreted as stronger evidence for PNC compared to the BM model (Kozak & Wiens,
2010).
Geographical approaches for testing PNC comprise several applications of species
distribution models (SDMs) for which the reciprocity between Grinnellian niche and
geographical distribution of species (Hutchinson’s duality; Colwell & Rangel, 2009) is
the basic assumption. Using SDMs, PNC can be tested under a scenario of invasion where
the prediction is that an invasive species must occupy regions that are environmentally
similar to its native range (Pearman et al., 2008; Guisan et al., 2014). Alternatively, SDMs
can be used to evaluate the potential of one species SDM to predict another species
distribution, where a positive finding would support PNC if conducted between sister
species (Peterson et al., 1999; Wiens et al., 2010). Finally, PNC could also be tested by
measuring the degree of overlap in environmental space (e.g. climatic) among pairs of
sister species, expecting a higher overlap than expected by their environmental
backgrounds under PNC (Warren et al., 2008; Broennimann et al., 2012). This latter
approach can be complemented by a correlation test between the degree of environmental
overlap and species phylogenetic distance (i.e. niche vs phylogenetic similarity) (Knouft
et al., 2006; Cardillo & Warren, 2016). Note that these geographical approaches represent
PNC tests at the species level, thus at a shallower phylogenetic scale than the phylogenetic
13
tree approaches mentioned above that are based on entire phylogenies and thus covering
deeper phylogenetic scales. Therefore, the combination of both phylogenetic and
geographical approaches can provide a more robust way of testing PNC (Wiens et al.,
2010) covering different phylogenetic scales within entire clades and thus evaluating
phylogenetic non-stationarity (Diniz-Filho et al., 2010).
Here, we evaluate for the first time the patterns of climatic niche evolution across
the order Chiroptera. We explicitly focus on climatic niches, using a multivariate
representation of a species niche and two of its features (position and breadth), to
investigate the extent to which bat species niches are evolutionary conserved and the
degree of niche similarity among related bat species. In particular, we are interested in
assessing the prevalence of phylogenetic niche conservatism at different phylogenetic
scales. To do so, we apply phylogenetic approaches at the order level and on different bat
families as well as a geographical approach at the species level by correlating niche and
phylogenetic similarity on species pairs. For the purpose of this study, we follow the
‘PNC as a pattern’ perspective (Pyron et al., 2015) and Losos' (2008) view that niches of
closely related species should be more similar than expected under neutral drift to qualify
as PNC. Based on the predictions of PNC over time and space, we expect that (i) bats will
exhibit a higher phylogenetic signal than expected under Brownian motion, (ii) the
distribution of niche values over the bat phylogeny will correspond to an OU model, and
that (iii) sister species will show greater niche overlap than less related species.
METHODS
Phylogenetic information and geographical data
We obtained phylogenetic information on bats relationships from a species-level
supertree of mammals (Fritz et al. (2009), which is an update of Bininda-Emonds et al.
14
(2007) supertree and based on the bats’ supertree originally proposed by Jones et al.
(2002). To perform the analyses, we extracted all bat species from the supertree and
obtained range maps (i.e., extents of occurrence polygons) for these species available
from the IUCN online database (IUCN 2013). We extracted the information of occurrence
for each species from a global raster of 1° × 1° cell resolution. Species were considered
as present in a grid cell if they occurred in at least 50% of it. We recognize that range
maps overestimate species distributions at fine scales but, at the global extent of our
analysis, they are congruent with survey-based data (Hawkins et al., 2008) and provide a
less biased view of distributions than occurrence records (Hurlbert & Jetz, 2007). In
addition, our choice of resolution (1º) may underestimate climatic variability within grid
cells but given the relatively large ranges of bat species (Willig et al., 2003b), this effect
can be negligible as suggested by several studies of niche conservatism at
macroecological scales using similar or even coarser resolutions (e.g. Hof et al., 2010;
Gouveia et al., 2014). From the 1030 bat species on the supertree, 121 species lack
geographical information. Therefore, we conducted subsequent analyses for a set of 909
species with both phylogenetic and geographical data.
Climatic variables and climatic niche
To characterize the climatic niches of bat species, we used a multivariate representation
based on nine climatic variables that are highly related to the ecological and physiological
tolerances of bats and have been suggested as effective predictors of their geographical
distributions (e.g., Ratrimomanarivo et al., 2009; Monadjem et al., 2010; Dixon, 2011;
Flanders et al., 2011; Stoffberg et al., 2012; Schoeman et al., 2013). These variables,
obtained from the Worldclim online database (Hijmans et al., 2005), were the following:
Annual mean temperature (BIO1), isothermality (BIO3), temperature, Seasonality
15
(BIO4), maximum temperature of the warmest month (BIO5), minimum temperature of
the coldest month (BIO6), annual precipitation (BIO12), precipitation of the wettest
month (BIO13), precipitation of the driest month (BIO14) and precipitation seasonality
(BIO15).
Species-level analyses
We were interested in testing PNC at different phylogenetic scales, from species to
families and the whole order. At the level of species (or ‘tips’ in the phylogeny), we
applied an explicitly geographical and environmental test of niche conservatism by
comparing the degree of overlap, similarity and equivalency of climatic niches between
pairs of species (Peterson et al., 1999; Warren et al., 2008). Focusing on species pairs
allowed us a closer inspection of the phylogenetic structure of niche conservatism without
focusing on the entire depth of the phylogeny (Cardillo, 2015). We selected from the
phylogeny those species representing a resolved cherry pair. A cherry is a pair of adjacent
tips on a tree representing two sister species. We identified 161 pairs (322 species), from
which 126 (252 species) had sufficient information for the analyses (i.e. at least five
occurrences for the overlap test).
We follow Broennimann et al. (2012) method that applies kernel smoothers to
species occurrence densities and climatic factors in a gridded space. Such method allows
correcting for niche similarity expected from spatial autocorrelation in the environment,
providing biologically meaningful inferences that are independent of sampling effort and
resolution (Broennimann et al., 2012; Warren et al., 2014). In this case, species climatic
niches were characterized as the envelope of climatic conditions (described by the
selected climatic variables) occupied by its occurrences and their accessible areas. By
considering species accessible areas and their complete occurrences, niche overlap can
16
reliably be evaluated (Barve et al., 2011; Broennimann et al., 2012). We defined
accessible areas as the set of ecoregions where a species occurs (Soberón, 2010; Barve et
al., 2011). We used those ecoregions proposed by Olson et al. (2001). For each species
pair, a PCA ordination test is calibrated on the occupied environments of both species and
use for testing niche overlap, similarity and equivalency along axes of such ordination
(Broennimann et al., 2012). Niche overlap was estimated using Schoener’s D metric,
which varies from 0 (no overlap) to 1 (complete overlap).
To evaluate if the observed patterns of niche overlap were statistically significant,
we conducted tests of niche equivalency and similarity between species pairs. These tests
entertain the possibility of niche overlap being simply determined by the spatial structure
of the climatic spaces occupied by the species (i.e. null expectation). More specifically,
the niche equivalency test evaluates if two niches in two geographical ranges are
indistinguishable in comparison with random occurrences of both species within the two
ranges. In this case, a null distribution of niche overlaps (D metric) is obtained by creating
pseudo-replicates where occurrences from the two ranges are pooled together and
randomly divided into two groups while keeping the original number of occurrences.
Complementarily, the niche similarity test evaluates the more relaxed possibility of the
climatic space occupied in one range being more similar (or different) to the one occupied
in the other range compared to what would be expected by chance. In this case, the null
distribution of D values is obtained from pseudo-replicates generated by randomly
assigning each occurrence in one range to a new location in the other range (Broennimann
et al., 2012). Significant niche equivalency and similarity was determined when the
observed values were outside the 95% confidence limit of a null distribution created by
100 iterations. To determine the existence of niche conservatism, we compared the
percentage of significant niche overlap patterns (similarity and equivalency) between true
17
species pairs and randomly formed species pairs (see below).
Finally, to integrate our niche overlap analysis with the phylogeny of Chiroptera we
performed a Mantel test to evaluate if niche overlap (Schoener’s D) was correlated to
phylogenetic (patristic) distance among species pairs, (e.g. Knouft et al., 2006; Warren et
al., 2008; Diniz-Filho et al., 2010) with significance estimated by 1000 matrix
permutations. A significant negative correlation would imply niche conservatism between
closely related species and niche divergence between distantly related species (Knouft et
al., 2006). During the selection of species pairs to perform the overlap analyses, we
excluded 12 pairs for having less than five occurrences, and 29 pairs for which one species
did not have enough occurrences. Thus, our final set included 281 species from 126
complete pairs plus 29 species who lost their pair but were kept to form random pairs.
Accordingly, such 126 pairs were considered as our true species pairs and random pairs
were formed from obtaining all possible pairwise combinations of those 281 species
(totaling 39340 pairs).
Family- and order-level analyses
To perform analyses at deeper phylogenetic scales, such as families and order, we first
defined two features of the species climatic niches as characterized by the multivariate
climatic space: niche position and niche breadth. To do so, we conducted an ordination
technique known as the outlying mean index (OMI; Dolédec et al., 2000) analysis that
produces uncorrelated niche axes, giving equal weights to sites with high and low richness
and making no assumption on species response curves (Thuiller et al., 2005). OMI results
allow to describe species niches based on niche marginality and niche breadth. Niche
marginality measures how much the mean conditions of a species niche deviates from the
theoretical niche (i.e. an average niche describing the mean conditions within the
18
distribution of the whole clade). Because species having similar distances but in opposite
directions from the average theoretical niche will have similar values of niche
marginality, we used the species scores along the first niche axis of the OMI ordination
as descriptor of niche position (NP) (Gouveia et al., 2014). Niche breadth (NB), or
tolerance, is measured as the dispersion of species occurrences over the multivariate
climatic space, thus, describing the variation of climatic conditions used by the species
(Dolédec et al., 2000; Thuiller et al., 2005). Although niche breadth can be confounded
by range size (Slatyer et al., 2013), our NB metric is based on the variance of species
occurrences along the niche axes. Therefore, species recorded in broad areas with similar
climatic conditions do not necessarily have broader niche breadth/tolerance (Thuiller et
al., 2005). Both NP and NB for each species were later mapped onto the complete
phylogeny for the order-level analysis, which was later pruned to include only species for
each of the studied families (see below).
We considered both niche features (NP and NB) as species traits and tested for their
phylogenetic signal across the family- and order -level phylogenetic trees using the K
statistic (Blomberg et al., 2003). When K = 1, trait evolution follows the expectation
under the Brownian motion model (i.e. trait divergence proportional to time). K > 1
indicates higher trait similarity than expected under Brownian motion, whereas K < 1
indicates lower trait similarity than expected under Brownian motion (Blomberg et al.,
2003). Both K = 1 and > 1 have been used as evidence of PNC (Losos, 2008; Cooper et
al., 2010), whereas K < 1 could mean either stasis (thus PNC) or no phylogenetic structure
(e.g. random variation; Wiens et al., 2010). As stated before, this test can be problematic
but given its ample application, we performed it for the sake of comparison.
Besides evaluating phylogenetic signal, we also applied a graphical method that
provides a means to explore phylogenetic patterns of trait evolution that can be interpreted
19
both in terms of phylogenetic signal and evolutionary models (Bini et al., 2014; Diniz-
Filho et al., 2015). Such method is called Phylogenetic Signal-Representation (PSR)
curves (Diniz-Filho et al., 2012) and it is based on Phylogenetic Eigenvector Regression
(PVR) models where trait variation is modeled by eigenvectors extracted and selected
from a pairwise phylogenetic distance matrix (Diniz-Filho et al., 1998). PSR curves are
built by plotting the R2 from sequential PVR models fitted by successively increasing the
number of eigenvectors against their accumulated eigenvalues (Diniz-Filho et al., 2012).
The shape of the PSR curve allows interpreting trait variation in terms of evolutionary
models such as Brownian motion or an OU process (Diniz-Filho et al., 2012; Bini et al.,
2014). For instance, a 45º line provides evidence that trait evolution follows Brownian
motion (as in K = 1), whereas a PSR curve above the 45º reference line denotes
accelerated trait divergence (i.e. trait evolving faster, less conserved, than expected under
Brownian motion). Conversely, when the PSR curve is below the 45º reference line but
above that of a null model (i.e. absence of phylogenetic signal), there is evidence for
decelerated trait evolution (i.e. trait evolving slower, more conserved, than expected
under Brownian motion like in an OU process) (Diniz Filho et al., 2012). We used the
area between the observed PSR curve and a line with slope = 1 to measure the deviation
from a Brownian motion model. Using permutations, we tested if the PSR area of both
niche features (NP and NB) deviated significantly from Brownian motion (a neutral
model) and null models of trait evolution (Bini et al., 2014; Gouveia et al., 2014; Diniz-
filho et al., 2015).
Given the criticisms on using phylogenetic signal tests for testing PNC, we
followed more recent recommendations (Kozak & Wiens, 2010; Münkemüller et al.,
2015) and compared the relative fit of four evolutionary models to the observed variation
of niche features (NP and NB). The models were the white noise (WN) model of random
20
variation where species trait similarity is phylogenetically independent, a Brownian
motion model of gradual drift, an OU model representing stasis or stabilizing selection
(Butler & King, 2004), and an Early Burst (EB) model describing a slowdown in trait
evolution over time (Harmon et al., 2010). Comparison and selection of best-fitting model
among the four evaluated models was done using the Akaike information criterion
corrected for sample size (Kozak & Wiens, 2010). According to theoretical and more
recent simulation studies, the fit of an OU model can be interpreted as evidence of niche
conservatism (Butler & King, 2004; Münkemüller et al., 2015). This is based on the OU
model basic assumption of trait (niche) evolution under adaptive constraints (Butler &
King, 2004) that, in this case, implies evolutionary stasis associated with climatic regimes
(Kozak & Wiens, 2010). Some of these studies have also suggested fitting more flexible
models of niche evolution, such as OU models considering different optima at different
parts of a phylogeny (Butler & King, 2004; Beaulieu et al., 2012; Münkemüller et al.,
2015). Such models would capture different PNC patterns (i.e. niche conservatism vs
evolution) of particular groups within a larger clade, testing PNC scale-dependency
(Losos, 2008) and thus phylogenetic stationarity (i.e. constant changes throughout the
phylogeny; Diniz Filho et al., 2011). However, such flexible models are computationally
intensive and can produce ill-fitting models resulting in erroneous parameter estimates
(Ho & Ané, 2014; Cooper et al., 2016; J. Beaulieu pers. comm.). Instead, we evaluated
phylogenetic stationarity of niche evolution within Chiroptera by conducting analyses at
different phylogenetic levels: Chiroptera as a whole and families. Considering bat
families separately, which are phylogenetically well-established (Teeling et al., 2005),
we aimed to identify if different bat subclades showed different niche conservatism
patterns, thus accounting for potential non-stationarity. Accordingly, we performed all of
the above analyses separately for each bat family with more than 45 species available for
21
analyses. This approach is akin to fitting OU evolutionary models with multiple optima,
where each group (e.g. family) would have its own optimal niche value (e.g. Kozak &
Wiens, 2010; Wüest et al., 2015). For these analyses, we considered seven bat families:
Emballunoridae (47 species), Hipposideridae (62), Molossidae (88), Phyllostomidae
(141), Pteropodidae (133), Rhinolophidae (64) and Vespertilionidae (330).
RESULTS
Species level
Niche equivalency tests showed that 0.66% of all possible species pairs (39340) exhibited
significant equivalency, from which only 2.3% (six pairs) corresponded to sister species
pairs. Niche similarity tests demonstrated almost the same proportion of higher than
expected niche similarity among both sister and random pairs (sister pairs: 42.8%;
random pairs: 30.82%). The proportion of pairs with lower than expected niche similarity
was 1.04% for random pairs and zero for sister pairs. Schoener’s D values (niche overlap)
among sister species pairs were slightly higher but not significantly different from those
found for random pairs (Appendix S1). Accordingly, the Mantel test ruled out the
hypothesis that niche overlap decreases with phylogenetic distance (Z = 1003780; P =
0.31).
Family and order levels
In general, phylogenetic signal for Chiroptera as a whole and for the individual
families showed Blomberg’s K values smaller than one for both niche features (breadth:
NB; and position: NP). For niche breadth, only the Emballorunidae and Vespertilionidae
families showed significant K values. For niche position (NP), Chiroptera as a whole and
the families Emballorunidae, Molossidae, Pteropodidae and Vespertilionidae showed
22
significant values (Table 1). We found just one case of K value significantly higher than
one, corresponding to the niche position (NP) of the Emballonuridae family (Table 1).
Table 1. The results of K statistic with the respective P values for niche breadth (NB) and
niche position (NP). The results are given for Chiroptera and separately for seven
families.
CLADE NB NP
K value P value K value P value
Chiroptera 0.118 0.277 0.209 0.001
Emballonuridae 0.839 0.001 1.237 0.001
Hipposideridae 0.470 0.626 0.562 0.387
Molossidae 0.460 0.126 0.498 0.019
Phyllostomidae 0.322 0.335 0.322 0.335
Pteropodidae 0.237 0.223 0.353 0.001
Rhinolophidae 0.086 0.679 0.129 0.100
Vespertilionidae 0.302 0.001 0.425 0.001
Conversely, results from phylogenetic signal-representation (PSR) curves showed
that both niche features for most groups (Chiroptera and families) displayed negative PSR
areas, which imply a slower evolution than expected under Brownian motion and thus
higher similarity among closely related species. However, most of these PSR areas were
not significantly different from expected under both null and Brownian motion evolution
models varying greatly among families (Table 2; Appendix S2-S8 in Supporting
Information). For instance, PSR areas of both niche features for Emballorunidae showed
significant differences from a null model but not from Brownian motion, whereas
Rhinolophidae and Hipposideridae showed the opposite with both niche features being
significantly different from Brownian motion but not from the null model.
Vespertilionidae was the only case in which the negative PSR area for both niche features
was significantly different from the null and Brownian motion models (i.e. slower
23
evolution than expected under BM, corresponding to an OU process of evolutionary
stasis). Similarly, Chiroptera and Phyllostomidae showed a significantly negative PSR
area from both the null and Brownian motion models but only for niche position (NP),
thus supporting evolutionary stasis under a OU process. Molossidae niche position
showed significant difference from null but not from Brownian motion. Pteropodidae
PSR area of niche breadth was negative and significantly different from null and
Brownian motion models whereas its niche position differed only from null model (Table
2).
Table 2. Phylogenetic signal representation (PSR) results and their respective P values,
regarding null and Brownian model, for climatic niche features: niche breadth (NB) and
niche position (NP). The results are given for Chiroptera and separately for seven
families.
Clade/variable PSR area P value (null model ) P value (Brownian model)
Chiroptera
NB -0.413 0.111 < 0.001
NP -0.352 < 0.001 < 0.001
Emballonuridae
NB -0.056 < 0.001 0.485
NP 0.089 < 0.001 0.899
Hipposideridae
NB -0.192 0.526 < 0.001
NP -0.109 0.070 0.061
Molossidae
NB -0.214 0.050 < 0.001
NP -0.149 < 0.001 0.031
Phyllostomidae
NB -0.322 0.536 < 0.001
NP -0.268 < 0.001 < 0.001
Pteropodidae
NB -0.295 0.010 < 0.001
NP -0.137 < 0.001 0.091
Rhinolophidae
NB -0.258 0.222 < 0.001
24
NP -0.260 0.292 < 0.001
Vespertilionidae
NB -0.265 < 0.001 < 0.001
NP -0.1803 < 0.001 < 0.001
Results from fitting different evolutionary models were similar to those of the PSR
curves. In general, we found support for an OU evolutionary model in niche position (NP)
for Chiroptera, Molossidae, Pteropodidae and in both niche breath and position for
Vespertilionidae (Table 3). There was no support for the early burst (EB) model in any
of the groups, but there was one case supporting white noise (WN) (Hipposideridae; NB),
and one case supporting Brownian model (BM) (Emballunoridae; NP). Note that the
cases in which we found more support for OU models also presented significantly
negative areas in PSR analysis (Tables 2 and 3), thus supporting phylogenetic niche
conservatism as a result of evolutionary stasis.
Table 3. Fit of alternative evolutionary models for climatic niche features of Chiroptera
and separately for seven families: niche breadth (NB) and niche position (NP). Sample
size–corrected AIC (AICC) and AIC weights (AICW) are reported for each model and for
each variable: WN (white noise); BM (Brownian motion); OU (Ornstein-Uhlenbeck
model) e EB (early burst). Results are highlighted for ∆AIC values ≤ 2.
Clade/variable WN BM OU EB
Chiroptera AICC AICW AICC AICW AICC AICW AICC AICW
NB 3663.7 0.47 4085.83 0.00 3663.49 0.53 4087.85 0.00
NP 3638.3 0.00 3566.38 0.00 3366.82 1.00 3568.39 0.00
Emballonuridae
NB 179.13 0.00 167.50 0.58 169.37 0.23 169.79 0.19
NP 179.51 0.00 148.98 0.61 151.26 0.19 151.21 0.20
Hipposideridae
NB 250.47 0.75 279.48 0.00 242.11 0.25 281.69 0.00
NP 240.37 0.71 262.57 0.00 242.11 0.30 264.78 0.00
25
Molossidae
NB 347.38 0.42 355.07 0.01 346.75 0.57 357.22 0.00
NP 304.96 0.26 316.91 0.00 302.85 0.74 319.06 0.00
Phyllostomidae
NB 676.62 0.55 717.20 0.00 677.01 0.45 719.29 0.00
NP 406.45 0.42 441.75 0.00 405.83 0.58 443.84 0.00
Pteropodidae
NB 540.19 0.60 603.86 0.00 541.01 0.40 605.96 0.00
NP 493.09 0.00 507.66 0.00 471.62 1.00 509.75 0.00
Rhinolophidae
NB 275.07 0.72 309.44 0.00 276.93 0.28 311.65 0.00
NP 251.06 0.44 268.01 0.00 250.61 0.56 270.21 0.00
Vespertilionidae
NB 1146.03 0.00 1183.15 0.00 1125.42 1.00 1185.19 0.00
NP 1364.53 0.00 1288.80 0.00 1269.45 1.00 1290.84 0.00
DISCUSSION
Phylogenetic niche conservatism has been advanced as a major explanation for the
contemporary latitudinal diversity gradient exhibited by bats (Stevens, 2006, 2011;
Buckley et al., 2010; Villalobos et al., 2013). However, no specific evidence for niche
similarity among related bat species, a basic tenet of PNC, has been provided. We aimed
to fill this gap by evaluating patterns of climatic niche conservatism in bats across space
and time. Our findings at different phylogenetic scales support PNC at the order and
family levels but not at the species level. These findings, as well as the variation of PNC
patterns among bat families, imply that the evolution of bat climatic niches has not been
constant over time or similar among different subclades, showing phylogenetic non-
stationarity. Accordingly, our findings highlight the obvious, but usually neglected, role
of the temporal or phylogenetic scale in the consideration of PNC (Losos, 2008; Peterson,
2011). Explicitly considering phylogenetic scales and based on the specific aims of each
of our analyses, we argue that our findings broadly support phylogenetic conservatism of
bat climatic niches.
26
Under PNC, retaining ancestral niches over time means that closely related species
should show higher niche similarity than distantly related species (Losos, 2008; Holt,
2009). Considering the Grinnellian aspect of the niche (Soberón, 2007), such similarity
can be tested by measuring the overlap in the climatic conditions linked to species
geographical distributions (Broennimann et al., 2012). This approach for testing PNC is
common in applications of species distribution modelling, particularly those related to
biological invasions (Guisan et al., 2014) but it has, to the best of our knowledge, never
been applied for testing PNC across species within higher taxa. Our species-level findings
using this niche overlap approach showed that in almost half of the compared species
pairs, niche similarity was higher than expected given their environmental backgrounds.
Taken at face value and for each species pair, these results could be interpreted as
evidence for PNC (Peterson, 2011). However, considering all species pairs, we found no
relationship between niche and phylogenetic similarity, suggesting that closely as well as
distantly related bat species have similar climatic niches. This, in turn, could be
interpreted as absence of PNC particularly in terms of phylogenetic signal since closely
related species do not resemble each other more than distantly related species.
Nevertheless, as pointed out by Wiens et al. (2010), the actual pattern expected
under PNC would be that of no change, meaning that all species within a clade can indeed
show niche similarity (or considerable overlap) regardless of their phylogenetic distance.
In the case of bats, this finding is congruent with the fact that species richness of most bat
families is highest in tropical regions (Procheş, 2005), hence bat species from different
lineages do share their climatic preferences to some extent. Indeed, this has been used as
evidence supporting the tropical niche conservatism hypothesis in bats (Stevens, 2006,
2011; Buckley et al., 2010). In this context our species-level results could be interpreted
as consistent with PNC even though we found no significant phylogenetic signal.
27
Moreover, our order- and family-level results showing phylogenetic non-stationarity and
supporting a stasis evolutionary model could also explain the absence of phylogenetic
signal at the species-level while more strongly supporting PNC at these more inclusive
levels.
Focusing on different phylogenetic scales under different approaches allowed us to
more strongly infer the prevalence of PNC in bat climatic niches. In accordance with the
species-level results, phylogenetic signal analysis at the order and family level did not
support PNC, regardless of its interpretation (i.e., equal or higher than expected under
BM; except for niche position in Emballorunidae) and niche feature (NP and NB),
showing values of K < 1. Nevertheless, as stated before, this finding could actually be
interpreted as evidence for PNC resulting from a stasis evolutionary process as suggested
by recent simulation studies (Revell et al., 2008; Münkemüller et al., 2015). In fact, our
PSR and model-based analyses, supported this interpretation. For Chiroptera as a whole
and regarding niche position (NP), the area of the PSR curve was significantly different
from the null and BM models, just as expected under a stasis or stabilizing selection
model (Diniz-Filho et al., 2012). In agreement, the best fitted model was the OU model,
consistent with a mechanism for PNC (Hansen, 1997; Münkemüller et al., 2015). Taken
together, these findings give evidence that the climatic niches of Chiroptera, at least niche
position, have evolved more slowly than expected under neutral drift (Brownian motion)
thus suggesting PNC at the order level.
We found similar results at the family level but with variation among the studied
families. This phylogenetic non-stationarity shows that PNC patterns are not constant
throughout the Chiropteran phylogeny. For instance, we found support for PNC in the
Molossidae, Vespertilionidae and Emballorunidae but not for the other families. For the
first two families, PSR curves and model fit supported an OU model for niche position,
28
indicating slower evolution than expected under BM (Tables 2,3). For the
Vespertilionidae, we also found support for PNC in niche breadth (i.e., tolerance). These
two families are among the most speciose bat families (representing ~ 46% of all bat
species used in our analyses). Therefore, it is likely that they could have driven the result
found for the whole Chiroptera order. As stated above, Chiroptera has being claimed as
presenting PNC, particularly tropical niche conservatism (TNC) (Buckley et al., 2010;
Stevens, 2011) but also as not presenting it (Pereira & Palmeirim, 2013). Given that most
of bat species are tropical (see Buckley et al., 2010) it seems reasonable to hypothesize
that TNC may have contributed to the latitudinal diversity gradient of bats. However, this
may be true for a couple of families (Phyllostomidae and, perhaps, Emballonuridae) but
not for the Vespertilionidae and Molossidae. The success of molossids and vespertilionids
in the temperate zones is possibly a result of their particular adaptations such as low
metabolic rates and hibernation capacity, that are needful to overcome cold seasons
(McNab, 1969; Stevens, 2004). Therefore, these families could be seen as cases of niche
evolution under the perspective of TNC (Pereira & Palmeirim, 2013). However, this is
not necessarily against the existence of PNC in these families, as suggested by our
findings, especially given their potential origin in the temperate region (Teeling et al.,
2005) and their converse latitudinal gradient (Stevens, et al. 2005), particularly for
Vespertilionidae.
In contrast with previous studies, we did not find strong support for PNC in the
tropical family Phyllostomidae (see Stevens, 2011; Villalobos et al., 2013).
Phyllostomids have been one of the most studied chiropteran families and have served as
a model system given their strong influence in driving the latitudinal gradient of bat
diversity (Willig et al., 2003a; Stevens, 2004). Actually, the strong latitudinal gradient of
phyllostomids, their successful radiation in the tropics and limited distribution in high
29
latitudes are considered a good example of TNC (Stevens, 2004; 2006; 2011). Aside from
the PSR results supporting conservatism of niche position in this family, we rejected the
effect of an OU process (i.e. stabilizing selection) in the evolution of the phyllostomid
climatic niches. In fact, we could not distinguish between the fit of an OU and a White
Noise (WN) evolutionary model for niche position in this family. Importantly, a recent
simulation study showed that a single optimum OU model (like the one applied here) with
strong selection strength can be readily misidentified as a WN model (Münkemüller et
al., 2015). This, in turn, could either mean that phyllostomid climatic niche may have
evolved either under strong selection or randomly (Diniz-Filho et al., 2010). Looking at
the phyllostomid richness and distributional patterns, it is plausible to suppose that their
climatic niches may have evolved under a strong evolutionary stasis (Stevens, 2004;
2011). In fact, recent phylogenetic studies suggest that phyllostomid evolution has been
characterized by ecological and phenotypic stasis following early species’ diversification
(Dumont et al., 2011; Monteiro & Nogueira, 2011; Rojas et al., 2011).
Emballonuridae was the only family where results supported a Brownian-like
evolution of both niche features. Species richness of this family shows a weaker increase
towards the equator compared to the Phyllostomidae, suggesting higher independence of
latitudinally varying resources (Stevens, 2004). Depending on the reference evolutionary
model used to assume PNC, we could consider the support for Brownian evolution of the
emballorunid climatic niche as consistent with PNC (Cooper et al., 2010). In fact, the
emballorunid pattern was the only situation in which Blomberg’s K statistic was higher
than one, providing evidence for PNC in this family also under this debated test (Losos,
2008).
Overall, we found more support for PNC regarding niche position (NP) instead of
niche breadth (NB). This is interesting given that NP provides information about the mean
30
position of the species niche (Thuiller et al., 2005), whereas NB informs about the
variation of climatic conditions that describe species niches (Dolédec et al., 2000). Thus,
PNC in bats seems to be acting through species conserving their niche position instead of
conserving their climatic tolerances. The lack of support for PNC in niche breadth is
consistent with previous findings showing that geographical range size (usually related
with niche breadth; Slatyer et al., 2013) of New World bats is not phylogenetically
structured (Arita, 1993; Villalobos et al., 2013), presenting considerable variation within
and among bat families (Willig et al., 2003b). Finally, recent findings considering the
Eltonian aspect of bat niches (i.e., diet) (Olalla-Tárraga et al., 2017) seem to agree with
our findings on their climatic niches, with support for PNC and scale-dependency of such
pattern. More detailed investigations on the connection between the evolution of different
aspects of species ecological niches and their role in determining diversity patterns are
surely needed. Such endeavor can bring us closer to an evolutionary theory of the niche
(Holt, 2009).
CONCLUSION
Several studies have suggested the presence of phylogenetic niche conservatism in bats
(Stevens, 2006, 2011; Buckley et al., 2010; Villalobos et al., 2013) but none have directly
tested for species niche similarity and across the whole order. Here we have provided
evidence that at least one Climatic niche feature (niche position) of Chiroptera shows
phylogenetic conservatism. Under a ‘PNC as a pattern’ perspective, our findings support
the claim that bat climatic niches have evolved more slowly than expected under neutral
drift (Losos, 2008; Pyron et al., 2015). Moreover, they can also serve as a basis for the
‘PNC as a process’ perspective if, as suggested by the abovementioned studies,
conservation of climatic niches drive bat diversity patterns. Testing for PNC can be a
31
complex endeavor, but comparing among different methodologies and phylogenetic
scales can help us get a clearer picture on the existence of PNC and its role in determining
the observed biodiversity patterns.
REFERENCES
Arita, H.C. (1993) Rarity in neotropical bats: correlations with phylogerny, diet, and body
mass. Ecological Applications, 3, 506–517.
Barve, N., Barve, V., Jiménez-Valverde, A., Lira-Noriega, A., Maher, S.P., Peterson, a.
T., Soberón, J. & Villalobos, F. (2011) The crucial role of the accessible area in
ecological niche modeling and species distribution modeling. Ecological Modelling,
222, 1810–1819.
Beaulieu, J.M., Jhwueng, D.C., Boettiger, C. & O’Meara, B.C. (2012) Modeling
stabilizing selection: Expanding the Ornstein-Uhlenbeck model of adaptive
evolution. Evolution, 66, 2369–2383.
Bini, L.M., Villalobos, F. & Diniz-Filho, J.A.F. (2014) Explorando patrones en rasgos
macroecológicos utilizando regresión secuencial de autovectores filogenéticos.
Ecosistemas, 23, 21–26.
Bininda-Emonds, O.R.P., Cardillo, M., Jones, K.E., MacPhee, R.D.E., Beck, R.M.D.,
Grenyer, R., Price, S. a, Vos, R. a, Gittleman, J.L. & Purvis, A. (2007) The delayed
rise of present-day mammals. Nature, 446, 507–12.
Blomberg, S.P. & Garland, T.J. (2002) Tempo and mode in evolution: phylogenetic
inertia, adaptation and comparative methods. Journal of Evolutionary Biology, 15,
899–910.
Blomberg, S.P., Garland, T.J. & Ives, A.R. (2003) Testing for phylogenetic signal in
comparative data: behavioral traits are more labile. Evolution; international journal
of organic evolution, 57, 717–745.
Broennimann, O., Fitzpatrick, M.C., Pearman, P.B., Petitpierre, B., Pellissier, L., Yoccoz,
N.G., Thuiller, W., Fortin, M.-J., Randin, C., Zimmermann, N.E., Graham, C.H. &
Guisan, A. (2012) Measuring ecological niche overlap from occurrence and spatial
environmental data. Global Ecology and Biogeography, 21, 481–497.
Buckley, L.B., Davies, T.J., Ackerly, D.D., Kraft, N.J.B., Harrison, S.P., Anacker, B.L.,
Cornell, H. V, Damschen, E.I., Grytnes, J.-A., Hawkins, B. a, McCain, C.M.,
Stephens, P.R. & Wiens, J.J. (2010) Phylogeny, niche conservatism and the
latitudinal diversity gradient in mammals. Proceedings of The Royal SocietyThe
Royal Society B: Biological Sciences, 277, 2131–2138.
Butler, M.A. & King, A.A. (2004) Phylogenetic Comparative Analysis : A Modeling
Approach for Adaptive Evolution. The American naturalist, 164, 683–695.
Cardillo, M. (2015) Geographic Range Shifts Do Not Erase the Historic Signal of
Speciation in Mammals. American Naturalist, 185, 343–353.
Cardillo, M. & Warren, D.L. (2016) Analysing patterns of spatial and niche overlap
among species at multiple resolutions. Global Ecology and Biogeography, 25, 951–
32
963.
Chase, J.M. & Leibold, M.A. (2003) Ecological niches: linking classical and
contemporary approaches, University of Chicago Press, Chicago and London.
Colwell, R.K. & Rangel, T.F. (2009) Hutchinson’s duality : the once and future niche.
Proceedings of the National Academy of Sciences, 106, 19651–19658.
Cooper, N., Jetz, W. & Freckleton, R.P. (2010) Phylogenetic comparative approaches for
studying niche conservatism. Journal of Evolutionary Biology, 23, 2529–2539.
Diniz-filho, A.F., Alves, D.M.C.C., Villalobos, F., Sakamoto, M., Brusate, S.L. & Bini,
L.M. (2015) Phylogenetic eigenvectors and nonstationarity in the evolution of
theropod dinosaur skulls. Journal of Evo, 28, 1410–1416.
Diniz-Filho, J.A.F., Bini, L.M., Rangel, T.F., Morales-Castilla, I., Olalla-Tárraga, M.Á.,
Rodríguez, M.Á. & Hawkins, B. a. (2012) On the selection of phylogenetic
eigenvectors for ecological analyses. Ecography, 35, 239–249.
Diniz-Filho, J.A.F., de Sant’Ana, C.E.R. & Bini, L.M. (1998) An Eigenvector Method
for Estimating Phylogenetic Inertia. Evolution, 52, 1247.
Diniz-Filho, J.A.F., Terribile, L.C., Da Cruz, M.J.R. & Vieira, L.C.G. (2010) Hidden
patterns of phylogenetic non-stationarity overwhelm comparative analyses of niche
conservatism and divergence. Global Ecology and Biogeography, 19, 916–926.
Diniz Filho, J.A.F., Rangel, T.F., Santos, T. & Mauricio Bini, L. (2011) Exploring
Patterns of Interspecific Variation in Quantitative Traits Using Sequential
Phylogenetic Eigenvector Regressions. Evolution, 66, 1079–1090.
Dixon, M.D. (2011) Post-Pleistocene range expansion of the recently imperiled eastern
little brown bat ( Myotis lucifugus lucifugus ) from a single southern refugium.
Ecology and Evolution, 1, 191–200.
Dolédec, S., Chessel, D. & Gimaret-Carpentier, C. (2000) Niche separation in community
analysis: a new method. Ecology, 81, 2914–2927.
Dumont, E.R., Dávalos, L.M., Goldberg, A., Santana, S.E., Rex, K. & Voigt, C.C. (2011)
Morphological innovation, diversification and invasion of a new adaptive zone.
Proceedings. Biological sciences / The Royal Society, 279, 1797–805.
Felsenstein, J. (1985) Phylogenies and the Comparative Methods. The American
naturalist, 125, 1–15.
Flanders, J., Wei-Li, Rossiter, S.J. & Zhang, S. (2011) Identifying the effects of the
Pleistocene on the greater horseshoe bat , Rhinolophus ferrumequinum , in East Asia
using ecological niche modelling and phylogenetic analyses. Journal of
Biogeography, 38, 439–452.
Fleming, T.H. (1973) Numbers of Mammal Species in North and Central American
Forest Communities. Ecology, 54, 555–563.
Fritz, S.A., Bininda-Emonds, O.R.P. & Purvis, A. (2009) Geographical variation in
predictors of mammalian extinction risk: big is bad, but only in the tropics. Ecology
Letters, 12, 538–549.
Gouveia, S.F., Hortal, J., Tejedo, M., Duarte, H., Cassemiro, F. a. S., Navas, C. a. &
Diniz-Filho, J.A.F. (2014) Climatic niche at physiological and macroecological
scales: the thermal tolerance-geographical range interface and niche dimensionality.
Global Ecology and Biogeography, 23, 446–456.
33
Guisan, A., Petitpierre, B., Broennimann, O., Daehler, C. & Kueffer, C. (2014) Unifying
niche shift studies: insights from biological invasions. Trends in Ecology &
Evolution, 29, 260–269.
Hansen, T.F. (1997) Stabilizing selection and the comparative analysis of adaptetion.
Evolution, 51, 1341–1351.
Harmon, L.J., Losos, J.B., Davies, T.J., Gillespie, R.G., Gittleman, J.L., Jennings, W.B.,
Kozak, K.H., Mcpeek, M.A., Moreno-roark, F., Near, T.J., Purvis, A., Ricklefs,
R.E., Schluter, D., Ii, J.A.S., Seehausen, O., Sidlauskas, B.L., Torres-carvajal, O.,
Weir, J.T. & Mooers, A.Ø. (2010) Early burst of body size and shape evolution are
rare in comparative data. Evolution, 64, 2385–2396.
Hawkins, B.A., Rueda, M. & Rodríguez, M.Á. (2008) What do range maps and surveys
tell us about diversity patterns? Folia Geobotanica, 43, 345–355.
Hijmans, R.J., Camerons, S.E., Parra, J.L., Jones, P.G. & Jarvis, A. (2005) Very high
resolution interpolated climate surfaces for global land areas. International Journal
of Climatology, 25, 1965–1978.
Ho, L.S.T. & Ané, C. (2014) Intrinsic inference difficulties for trait evolution with
Ornstein-Uhlenbeck models. Methods in Ecology and Evolution, 5, 1133–1146.
Hof, C., Rahbek, C. & Araújo, M.B. (2010) Phylogenetic signals in the climatic niches
of the world’s amphibians. Ecography, 33, 242–250.
Holt, R.D. (2009) Bringing the Hutchinsonian niche into the 21st century : ecological and
evolutionary perspectives. Proceedings of the National Academy of Sciences USA,
106, 19659–19665.
Hurlbert, A.H. & Jetz, W. (2007) Species richness, hotspots, and the scale dependence of
range maps in ecology and conservation. Proceedings of the National Academy of
Sciences USA, 104, 13384–13389.
IUCN (2013) The IUCN Red List of Threatened Species. Version 2013.1,
http://www.iucnredlist.org.
Jones, K.E., Purvis, A., MacLarnon, A., Bininda-Emonds, O.R.P. & Simmons, N.B.
(2002) A phylogenetic supertree of the bats (Mammalia: Chiroptera). Biological
reviews of the Cambridge Philosophical Society, 77, 223–259.
Kaufman, D.M. (1995) Diversity of New World mammals: universality of the latitudinal
gradients of species and bauplans. Journal of Mammalogy, 76, 322–334.
Knouft, J.H., Losos, J.B., Glor, R.E. & Kolbe, J.J. (2006) Phylogenetic analysis of the
evolution of the niche in lizards of the Anolis Sagrei group. Ecology, 87, S29–S38.
Kozak, K.H. & Wiens, J.J. (2010) Niche conservatism drives elevational diversity
patterns in Appalachian salamanders. The American naturalist, 176, 40–54.
Losos, J.B. (2008) Phylogenetic niche conservatism, phylogenetic signal and the
relationship between phylogenetic relatedness and ecological similarity among
species. Ecology Letters, 11, 995–1003.
McNab, B.K. (1969) The economics of temperature regulation in neutropical bats.
Comparative Biochemistry and Physiology, 31, 227–268.
Monadjem, A., Taylor, P.J., Cotterill, F.D.P. & Schoeman, M.C. (2010) Bats of Central
and Southern Africa: A Biogeographic and Taxonomic Synthesis, Wits University
Press., Johannesburg.
34
Monteiro, L.R. & Nogueira, M.R. (2011) Evolutionary patterns and processes in the
radiation of phyllostomid bats. BMC evolutionary biology, 11, 137.
Münkemüller, T., Boucher, F.C., Thuiller, W. & Lavergne, S. (2015) Phylogenetic niche
conservatism - common pitfalls and ways forward. Functional Ecology, 29, 607–
639.
Olalla-Tárraga, M.Á., González-Suárez, M., Bernardo-Madrid, R., Revilla, E. &
Villalobos, F. (2017) Contrasting evidence of phylogenetic trophic niche
conservatism in mammals worldwide. Journal of Biogeography, 44, 99–110.
Olalla-Tárraga, M.Á., McInnes, L., Bini, L.M., Diniz-filho, J.A.F., Fritz, S.A., Hawkins,
B.A., Hortal, J., Orme, C.D.L., Rahbek, C., Rodriguez, M.Á. & Purvis, A. (2011)
Climatic niche conservatism and the evolutionary dynamics in species range
boundaries: global congruence across mammal amphibians. Journal of
Biogeography, 38, 2237–2247.
Olson, D.M., Dinerstein, E., Wikramanayake, E.D., Burgess, N.D., Powell, G.V.N.,
Underwood, E.C., D’amico, J. a., Itoua, I., Strand, H.E., Morrison, J.C., Loucks,
C.J., Allnutt, T.F., Ricketts, T.H., Kura, Y., Lamoreux, J.F., Wettengel, W.W.,
Hedao, P. & Kassem, K.R. (2001) Terrestrial Ecoregions of the World: A New Map
of Life on Earth. BioScience, 51, 933.
Pearman, P.B., Guisan, A., Broennimann, O. & Randin, C.F. (2008) Niche dynamics in
space and time. Trends in ecology & evolution, 23, 149–158.
Pereira, M.J.R. & Palmeirim, J.M. (2013) Latitudinal diversity gradients in new world
bats: are they a consequence of niche conservatism? PloS one, 8, e69245.
Peterson, A.T. (2011) Ecological niche conservatism: a time-structured review of
evidence. Journal of Biogeography, 38, 817–827.
Peterson, A.T., Sober n, J. & Sanchez-Cordero, V. V (1999) Conservatism of ecological
niches in evolutionary time. Science, 285, 1265–1267.
Procheş, Ş. (2005) The world’s biogeographical regions: cluster analyses based on bat
distributions. Journal of Biogeography, 32, 607–614.
Pyron, A.R., Costa, G.C., Patten, M. a. & Burbrink, F.T. (2015) Phylogenetic niche
conservatism and the evolutionary basis of ecological speciation. Biological
Reviews, 90, 1248–1262.
Ratrimomanarivo, F.H., Goodman, S.M., Stanley, W.T., Naidoo, T., Taylor, P.J. & Lamb,
J. (2009) Geographic and Phylogeographic Variation in Chaerephon leucogaster (
Chiroptera : Molossidae ) of Madagascar and the Western Indian Ocean Islands of
Mayotte and Pemba. Acta Chiropterologica, 11, 25–52.
Revell, L.J., Harmon, L.J. & Collar, D.C. (2008) Phylogenetic signal, evolutionary
process, and rate. Systematic biology, 57, 591–601.
Rojas, D., Vale, Á., Ferrero, V. & Navarro, L. (2011) When did plants become important
to leaf-nosed bats? Diversification of feeding habits in the family Phyllostomidae.
Molecular Ecology, 20, 2217–2228.
Schoeman, M.C., Cotterill, F.P.D.W., Taylor, P.J. & Monadjem, A. (2013) Using
potential distributions to explore environmental correlates of bat species richness in
southern Africa : Effects of model selection and taxonomy. Current Zoology, 59.
Slatyer, R.A., Hirst, M. & Sexton, J.P. (2013) Niche breadth predicts geographical range
size: a general ecological pattern. Ecology Letters, 16, 1104–1114.
35
Soberón, J. (2007) Grinnellian and Eltonian niches and geographic distributions of
species. Ecology Letters, 10, 1115–1123.
Soberón, J.M. (2010) Niche and area of distribution modeling: A population ecology
perspective. Ecography, 33, 159–167.
Stevens, R.D. (2006) Historical processes enhance patterns of diversity along latitudinal
gradients. Proceedings of The Royal Society B: Biological Sciences, 273, 2283–
2289.
Stevens, R.D. (2011) Relative effects of time for speciation and tropical niche
conservatism on the latitudinal diversity gradient of phyllostomid bats. Proceedings
of the Royal Society B: Biological Sciences, 278, 2528–2536.
Stevens, R.D. (2004) Untangling latitudinal richness gradients at higher taxonomic levels:
familial perspectives on the diversity of New World bat communities. Journal of
Biogeography, 31, 665–674.
Stoffberg, S., Schoeman, M.C. & Matthee, C.A. (2012) Correlated genetic and ecological
diversification in a widespread southern african horseshoe bat. PLoS biology, 7,
e31946.
Teeling, E.C., Springer, M.S., Madsen, O., Bates, P., O’brien, S.J. & Murphy, W.J. (2005)
A molecular phylogeny for bats illuminates biogeography and the fossil record.
Science (New York, N.Y.), 307, 580–584.
Thuiller, W., Lavorel, S. & Araújo, M.B. (2005) Niche properties and geographical extent
as predictors of species sensitivity to climate change. Global Ecology and
Biogeography, 14, 347–357.
Villalobos, F., Rangel, T.F. & Diniz-Filho, J.A.F. (2013) Phylogenetic fields of species:
cross-species patterns of phylogenetic structure and geographical coexistence.
Proceedings. Biological sciences / The Royal Society, 280, 20122570.
Warren, D.L., Cardillo, M., Rosauer, D.F. & Bolnick, D.I. (2014) Mistaking geography
for biology: Inferring processes from species distributions. Trends in Ecology and
Evolution, 29, 572–580.
Warren, D.L., Glor, R.E. & Turelli, M. (2008) Environmental niche equivalency versus
conservatism: quantitative approaches to niche evolution. Evolution, 62, 2868–2883.
Wiens, J.J. (2008) Commentary on losos (2008): niche conservatism déjà vu. Ecology
Letters, 11, 1004–1005.
Wiens, J.J. (2011) The niche, biogeography and species interactions. Philosophical
transactions of the Royal Society B: Biological Sciences, 366, 2336–2350.
Wiens, J.J., Ackerly, D.D., Allen, A.P., Anacker, B.L., Buckley, L.B., Cornell, H. V,
Damschen, E.I., Jonathan Davies, T., Grytnes, J.-A., Harrison, S.P., Hawkins, B. a,
Holt, R.D., McCain, C.M. & Stephens, P.R. (2010) Niche conservatism as an
emerging principle in ecology and conservation biology. Ecology letters, 13, 1310–
1324.
Wiens, J.J. & Donoghue, M.J. (2004) Historical biogeography, ecology and species
richness. Trends in ecology & evolution, 19, 639–644.
Wiens, J.J. & Graham, C.H. (2005) Niche conservatism: integrating evolution, ecology,
and conservation biology. Annual Review of Ecology, Evolution, and Systematics,
36, 519–539.
36
Willig, M.R., Kaufman, D.M. & Stevens, R.D. (2003a) Latitudinal gradients of
biodiversity: pattern, process, scale, and synthesis. Annual Review of Ecology,
Evolution, and Systematics, 34, 273–309.
Willig, M.R., Patterson, B.D. & Stevens, R.D. (2003b) Patterns of range Size, richness,
and body size in the Chiroptera. Bat Ecology (ed. by T.H. Kunz) and M.B. Fenton),
pp. 580–621. Chicago, The University of Press, Chicago.
37
SUPPORTING INFORMATION
Appendix S1. Boxplot showing the average of niche overlap values (D) between
species pairs formed by random and pairs in cladogenesis (phylogenetically close).
38
Appendix S2. PSR (phylogenetic signal representation) areas for two climatic niche
features of Emballonuridae family: (NP) niche position and (NB) niche breath. Orange
and yellow bands are the confidence intervals for the neutral (Brownian motion) and null
(random) expectations, respectively.
39
Appendix S3. PSR (phylogenetic signal representation) areas for two climatic niche
features of Hipposideridae family: (NP) niche position and (NB) niche breath. Orange
and yellow bands are the confidence intervals for the neutral (Brownian motion) and null
(random) expectations, respectively.
40
Appendix S4. PSR (phylogenetic signal representation) areas for two climatic niche
features of Molossidae family: (NP) niche position and (NB) niche breath. Orange and
yellow bands are the confidence intervals for the neutral (Brownian motion) and null
(random) expectations, respectively.
41
Appendix S5. PSR (phylogenetic signal representation) areas for two climatic niche
features of Phyllostomidae family: (NP) niche position and (NB) niche breath. Orange
and yellow bands are the confidence intervals for the neutral (Brownian motion) and null
(random) expectations, respectively.
42
Appendix S6. PSR (phylogenetic signal representation) areas for two climatic niche
features of Pteropodidae family: (NP) niche position and (NB) niche breath. Orange and
yellow bands are the confidence intervals for the neutral (Brownian motion) and null
(random) expectations, respectively.
43
Appendix S7. PSR (phylogenetic signal representation) areas for two climatic niche
features of Rhinolophidae family: (NP) niche position and (NB) niche breath. Orange and
yellow bands are the confidence intervals for the neutral (Brownian motion) and null
(random) expectations, respectively.
44
Appendix S8. PSR (phylogenetic signal representation) areas for two climatic niche
features of Vespertilionidae family: (NP) niche position and (NB) niche breath. Orange
and yellow bands are the confidence intervals for the neutral (Brownian motion) and null
(random) expectations, respectively.
45
CAPÍTULO 02: GEOGRAPHICAL
PATTERNS OF PHYLOGENETIC BETA
DIVERSITY COMPONENTS IN
TERRESTRIAL MAMMALS
46
GEOGRAPHICAL PATTERNS OF PHYLOGENETIC BETA DIVERSITY
COMPONENTS IN TERRESTRIAL MAMMALS
Aceito para publicação no periódico Global Ecology and Biogeography
F. P. Peixoto, F. Villalobos, A. S. Melo, J. A. F. Diniz-Filho, R. Loyola, T. F. Rangel &
M. V. Cianciaruso
ABSTRACT
Aim: to investigate geographical patterns of phylogenetic beta diversity (PBD) and its
turnover and nestedness-resultant components for terrestrial mammals. We expect an
increase on the contribution of the nestedness-resultant component towards temperate
regions given the historical loss of lineages caused by environmental and spatial
constraints. Analogously, we expect to find a similar increase of the nestedness-resultant
component contribution towards higher altitudes. We expect these patterns to be stronger
for Rodentia because they have poor dispersal ability and may have been less efficient in
recolonizing areas after glaciations.
Location: worldwide.
Methods: we generated terrestrial Mammalia species composition for 200 x 200 km cells
to calculate PBD and its turnover and nestedness-resultant components. All measures
were computed for each cell and the cells in the surrounding radius of one, two or three
adjacent layer cells. We calculated the relative importance of the nestedness-resultant
component as the proportion of the total PBD (PBDratio) and also PBD deviation given
taxonomic beta diversity (PBDdev). PBDdev measures the importance of phylogenetic beta
diversity after factoring out taxonomical beta diversity. We ran simple linear regressions
47
and piecewise regressions to investigate relationships between PBDratio and mean annual
temperature and altitude.
Results: we found a major contribution of nestedness-resultant component linked to
temperate climate, especially for groups with better dispersal capacity. Higher altitudes
were associated with major contribution of turnover-resultant component, particularly for
Rodentia.
Main conclusions: we provide the first global representation of phylogenetic beta
diversity in terrestrial mammals and demonstrated that at higher latitudes PBD is mostly
a result of lineage loss, whereas in highlands it is linked to lineage turnover. By analyzing
global patterns of PBD component contribution, we show demonstrate that dispersal
capacity is essential in determining the response of different lineages to geographical and
environmental barriers.
Keywords: phylobetadiversity, Mammalia, nestedness-resultant component, turnover,
phylogenetic diversity, Carnivora, Chiroptera, Rodentia.
INTRODUCTION
One of the main research goals of both ecologists and biogeographers alike is to
understand the ecological and evolutionary processes determining biodiversity patterns
(Ricklefs, 1987; Gaston, 2000; Wiens & Donoghue, 2004). Among the most ubiquitous
and extensively studied geographical pattern in biodiversity is the latitudinal diversity
gradient (LDG), in which tropical regions hold greater diversity than temperate regions
(Hillebrand, 2004; Mittelbach et al., 2007). Although the main drivers of this pattern are
still in dispute (Brown, 2014), dominant biogeographical theories support historical
processes as responsible for LDG (Mittelbach et al., 2007; Rolland et al., 2014). For
48
example, the Tropical Niche Conservatism hypothesis (TNC, Wiens & Donoghue, 2004)
states that most lineages originated and accumulated in the tropics, given that they have
had more time to speciate. Moreover, dispersal and adaptation to temperate regions have
been constrained for most groups under niche conservatism, thus producing higher
species richness in the tropics. Alternatively, the Out of the Tropics hypothesis (OTT,
Jablonski et al., 2006) predicts that high rates of diversification (i.e. more speciation than
extinction) generated high clade diversity in the tropics. Additionally, low diversity in
temperate zone is hypothesized to be caused by high extinction rates added to abiotic
environmental constraints, hindering colonization of these regions (Jablonski et al.,
2006).
The LDG pattern has been related to an analogous gradient in species beta
diversity, with higher dissimilarity between sites at lower latitudes (Qian & Ricklefs,
2012). Consequently, similar explanations have been put forward to explain both
gradients (Qian & Xiao, 2012). High levels of beta diversity were found to be related to
high energy availability, which indicate an increase in speciation and may have
contributed to LDG pattern. Exploring species dissimilarity patterns across space may
help us to elucidate historical and ecological causes of taxa distribution in a given gradient
(Whittaker, 1972). Beta diversity can be related to a range of processes regarding species
ecology (e.g. traits and niche) and environmental characteristics (e.g. isolation,
topography) (Tuomisto et al., 2003; Potts et al., 2004; Graham et al., 2006; Mcknight et
al., 2007). At global level, the best predictor for beta diversity is the difference in altitude
instead of biotic transitions (Mcknight et al., 2007; Melo et al., 2009), as evidence of
species adaptation to particular environmental conditions (Melo et al., 2009).
Recent developments in estimation and interpretation of beta diversity have
emphasized the need for decomposing between distinct aspects of dissimilarity across
49
space. Lennon et al. (2001) showed that common beta diversity indices are affected not
only by species turnover (or replacements) but also by differences in species richness and
suggested the use of turnover-only indices. This suggestion was fully developed by
Baselga (2010, 2012) who formally proposed a framework for partitioning beta diversity
into two components: dissimilarity due to species turnover and dissimilarity due to
richness differences. Global patterns of beta diversity components revealed a differential
importance of each component in determining the latitudinal diversity gradient,
highlighting the role of distinct processes (Baselga et al., 2012). For example, high
latitudes are generally related to dissimilarity owing to species richness differences
(Baselga et al., 2012), which is interpreted as an imprint of past glaciations and is more
evident for groups with less dispersal capacity (Dobrovolski et al., 2012).
Although beta diversity and its components may provide information about
ecological and evolutionary processes, species composition alone does not consider the
shared ancestry among species (Purvis & Hector, 2000). Even two regions with
completely different species compositions can share much of their evolutionary histories
(Graham & Fine, 2008). Thus, including evolutionary information into the analysis of
compositional and turnover patterns allows for a deeper investigation of the evolutionary
processes responsible for generating biodiversity gradients (Graham & Fine, 2008;
Kubota et al., 2011; Peixoto et al., 2014). Lozupone & Knight (2005), Bryant et al. (2008)
and Graham & Fine (2008) proposed the incorporation of phylogenetic information to
understand beta diversity (BD), which they referred as phylogenetic beta diversity (PBD).
PBD explicitly adds phylogenetic information to describe the difference among species
assemblages, that is, it measures to what extent these assemblages differ in terms of the
evolutionary relationships of its members (Graham & Fine, 2008). Leprieur et al. (2012)
extended Baselga’s partitioning of beta diversity to PBD and proposed its decomposition
50
in components due to lineage turnover (turnover-resultant component; PBDTURN) and due
to differences in phylogenetic diversity (PD) between assemblages (the nestedness-
resultant component; PBDPD) (Leprieur et al., 2012).
Here, we explored for the first time the global patterns of phylogenetic turnover
and nestedness-resultant components for terrestrial mammalians and tested some
biogeographical hypotheses that could explain these patterns (Table 1). In particular, we
are interested in inferring potential historical mechanisms driving PBD geographical
gradient for terrestrial mammalians by extending predictions of current biogeographical
hypotheses for LDG pattern. We hypothesized that historical processes, which have shaped
biodiversity latitudinal patterns, would produce different contributions of PBD components
between tropical and temperate regions. In accordance with TNC and OTT hypotheses, tropics
are both cradles and museums. Therefore, we expect higher lineage replacement (PBDTURN)
towards the equator. On the other hand, we expect higher influence of PBDPD in temperate
regions resulting from lineage loss towards the poles (Baselga, 2012). In the same way as
for climatic zones, we expect that colonization of higher altitudes will be followed by loss
of lineages, displaying major contribution of PBDPD component in mountainous areas.
We also investigated if Carnivora, Chiroptera, and Rodentia have differential
contributions of PBD components due to their efficiency in recolonizing areas that were
affected by climatic changes (Smith & Green, 2005; Dobrovolski et al., 2012).
Considering geographical range size as a proxy for dispersal ability (Gaston, 1996) we
expect that PBDPD contribution in high latitudes will be higher for Rodentia.
51
Table 1 Two hypotheses and their predictions for the phylogenetic beta diversity (PBD)
patterns in Mammalia and its components: the turnover-resultant component (PBDTURN)
and the phylogenetic diversity component (PBDPD).
Hypotheses Predictions References
H1: Historical processes that
shaped biodiversity latitudinal
patterns will produce different
contributions of PBD
components between tropical
and temperate regions. In
accordance with “out of the
tropics” and “tropical niche
conservatism” hypotheses,
tropics are both cradles and
museums. Therefore,
dissimilarity among
communities will be driven
mainly by PBDTURN
component. In opposite,
PBDPD component would be
more important among toward
temperate regions, given loss
of lineages towards the poles.
P1. We expect a negative
relationship between PBDratio
and the annual mean
temperature. Therefore,
regions with temperate climate
will be related to high values
of PBDratio, which means great
contribution of PBDPD.
Otherwise, tropical climate
will be related to lower values
of PBDratio that mean major
contribution of PBDTURN.
Wiens & Donoghue 2004
Jablonski et al., 2006
Baselga, 2010
Baselga et al., 2012
Dobrovolski et al., 2012
H2: In the same way that
dispersal and adaptation to
temperate regions have been
prevented by environmental
constrains, colonization of
higher altitudes will also be
followed by loss of lineages.
P2. We expect high values of
PBDratio being related to high
altitudes, displaying major
contribution of PBDPD.
Wiens & Donoghue 2004
Jablonski et al., 2006
Baselga, 2010
Baselga et al., 2012
Dobrovolski et al., 2012
H3: Differences among the
dispersal capacities among
mammalian orders will result
in distinct contributions of
PBD components in
accordance with their
efficiency in recolonizing
areas that were affected by
climatic changes.
P3. Among the mammalian
orders analysed, we expect
that the contribution of PBDPD
in temperate regions
(prediction 1) will be stronger
for Rodentia, which has
greater frequency of small
range sizes.
Smith & Green, 2005
Harrison et al., 1992
Böhning-Gaese et al., 1998
Dobrovolski et al., 2012
Gaston, 1996
52
METHODS
Data origin
We gathered information on the geographical distribution of 4536 terrestrial
mammalian species (IUCN, 2015). To derive species composition for cells, we
overlapped range maps in a global raster of 200 X 200 km cell size (see Safi et al., 2011).
A grid cell was considered occupied by those species where the centre of the grid cell
intersected with the species ranges. In cases where a species’ range (or isolated sections
of the range distribution) did not intersect with the centre of any grid cell, we considered
the grid cell (or grid cells) intersected by that range as occupied by that species. We used
a dated phylogenetic tree to obtain phylogenetic relationship information among the
species we were studying (4536 mammalian species) (Bininda-Emonds et al., 2007; Fritz
et al., 2009). We collected climate (annual mean temperature) and altitudinal data for
these species from Worldclim online database (Hijmans et al., 2005).
Phylogenetic beta diversity partitioning
We calculated both taxonomic beta diversity (BD) and phylogenetic beta diversity
(PBD) among cells using two families of dissimilarity indices, Sørensen (Sørensen, 1948)
and Jaccard (Jaccard, 1912). PBD differs from BD simply by the replacement of
exclusive/shared ‘species’ by ‘branch lengths’. Thus, PBD captures the proportion of
shared and exclusive branch lengths among assemblages. PBD varies from 0 (when the
assemblage composition and, thus, shared branch lengths is identical) to near 1 (when the
assemblages are composed by distinct species that share no branch in the rooted
phylogenetic tree). We evaluated the deviations of PBD patterns given the taxonomic beta
diversity (BD) in order to identify places where the exchange of lineages was higher or
lower than expected given the species exchange (Graham & Fine, 2008). Such deviations
53
(hereafter PBDdev) were calculated as PBDdev = BD-PBD / BD. When PBD is the same as
the taxonomic beta diversity (i.e. in the case of a star phylogeny, see Melo et al., 2014)
this ratio will be zero. Conversely, if PBD among places is low in relation to BD this ratio
will be close to 1, indicating that there is no lineage exchange. On the other hand, negative
values indicate that lineage exchange is more important than taxonomic beta diversity.
To understand the relative contribution of PBD components in our analysis, we
used an additive partitioning framework (Leprieur et al., 2012) built upon the original
approach of Baselga (2010, 2012). This framework partitions PBD index into one
component representing the turnover-resultant dissimilarity (PBDTURN) and another
component that corresponds to the dissimilarity due to PD differences between regions,
the nestedness-resultant component (PBDPD). In this framework the richness-difference
dissimilarity (PBDPD) accounts for phylogenetic diversity differences derived only from
nestedness (Baselga & Leprieur, 2015).
Data analysis
All measures were computed within a window of cells, using the same procedure of Melo
et al. (2009). To explicitly consider the scale dependence of results, we did the analyses
using three window sizes, defined as radii of one, two, and three cell layers adjacent to
the focal cell. We used both pairwise and multiple-site approaches to calculate PBD. For
the pairwise approach, we obtained dissimilarity values between the focal cell and all
adjacent cells and assigned the average value for the focal cell. In the multiple-site
approach, a single value was obtained from all samples compared (all cells in the window)
following Baselga et al. (2007). A multiple-site index avoids repetition of each site in
several pairs (i.e. lack of independence) and prevents information loss regarding the
number of species shared among the sites that are being compared (Diserud & Ødegaard,
54
2007). The same partitions cited above for the pair-wise indices were calculated for the
multi-site indices. We used functions of the betapart (Baselga et al., 2013) and the
CommEcol (Melo, 2013) packages implemented in R version 3.3.0.
We generated global pattern maps for PBD and their components for terrestrial
mammals and for three mammalian orders separately. We chose Carnivora, Chiroptera,
and Rodentia owing to their global distribution and their dissimilar dispersal capacity.
We tested hypothesis 2 (Table 1) following the rationale that clades with high dispersal
capacity will be more efficient in recolonizing areas that were deglaciated only recently
(e.g. Schloss et al., 2012). As a proxy for dispersal capacity we used the median values
of geographical range size (Gaston, 1996), quantified as the number of occupied cells:
Carnivora (83) > Chiroptera (19) > Rodentia (5). Then, we produced maps showing the
relative importance of the nestedness-resultant PBD component (hereafter PBDratio)
following Dobrovolski et al. (2012): PBDratio = PBDPD / PBD. Values smaller than 0.5
indicate that PBD is mainly determined by turnover-resultant dissimilarity (PBDTURN),
whereas values greater than 0.5 indicate that PBD results from a greater influence of the
dissimilarity component due to PD differences between assemblages (PBDPD). To test
our predictions, we performed simple linear regressions to investigate if PBDratio was
correlated with [H1] annual mean temperature and [H2] altitude. Because Baselga et al.
(2012) demonstrated the existence of a breakpoint in BD geographical pattern for
amphibians, we compared linear regression fits with equivalent piecewise regressions (R
package “segmented” implemented in R version 3.3.1; Muggeo, 2008). We searched for
the combinations of breakpoint values that yielded the lowest global residual sum of
squares and which presented the minimum BIC (Bayesian information criterion) partition
(R package “strucchange” implemented in R version 3.3.0; Zeileis et al., 2002).
55
RESULTS
Jaccard and Sorensen indices produced qualitatively similar results. Also,
although the pairwise approach resulted in absolute values higher than the multiple-site
approach, the overall pattern was the same. Therefore, we present results only for
multiple-site approach with Jaccard index. In general, PBD patterns produced by different
window scale definitions were similar. However, given the higher number of compared
cells and consequently the higher distance regarding the focal cell and its adjacent cells,
well-defined boundaries were easier to identify at larger scale. We present the smallest
scale (1 layer of adjacent cells) map in the main text. Maps for the other scales are
available in Supporting Information (Fig.1; Appendix S1-S2).
We did not detect a clear latitudinal pattern for PBD components. Instead, we
found a complex geographical pattern for all mammalians and for Carnivora, Chiroptera,
and Rodentia (Fig. 1). We did not find support to our expectation that annual mean
temperature would be linearly and negatively related to PBDratio. In general, we found an
interesting non-monotonic relation with U-shaped patterns for almost all groups and
scales. Higher annual mean temperature values were related to wide variation of PBDratio
values. However, at lower temperatures we observed a negative relationship with
PBDratio, denoting a major contribution of PBDPD towards colder climates (Appendix S7-
S10). In all cases, piecewise regressions improved model fit when compared to simple
linear regressions (Appendix S7-S10). The best regression improvement was found to
Carnivora larger scale (from 0.17 to 0.43; p <0.001) and the worst to Carnivora smallest
scale (from 0.05 to 0.08; p <0.001). This result supports the existence of breakpoints on
these relationships, with distinct regions regarding PBD components contribution. Model
fit was higher with the increase of window scale (number of compared cells). The best fit
was found at larger scales for Carnivora (r2=0.43, p <0.001), Rodentia (r2=0.24, p <0.001)
56
and Chiroptera (r2=0.12, p <0.001). Contrary to our prediction, we identified a lower
contribution of PBDPD in colder temperatures for Rodentia rather than for Chiroptera and
Carnivora [Appendix S7-S10 (c)]. Besides, the proportion of nestedness-resultant
component (PBDratio) in overall beta diversity was higher for Carnivora (mean±SD =
0.608±0.285) than for Chiroptera (0.490±0.282) and Rodentia (0.399±0.261).
We observed a tendency towards high PBD values for all groups in mountainous
areas (Fig. 1). That seems to be a common pattern across highlands worldwide, mainly in
regions above 2000 meters (Andes, Nearctic Cordilleran Mountain System, Himalayas,
and Altai Mountains). These regions have a great contribution of lineage turnover in
phylogenetic dissimilarity patterns (PBDTURN) (regions marked in light yellow in Fig. 1).
In addition, turnover-resultant component was more important for Rodentia regarding
other orders (Fig. 1d). Nevertheless, we did not find support to our expectation that
altitude would be positively related to PBDratio. In all cases, both simple linear and
piecewise regressions presented a poor fit (Appendix S11-S14). In general, we observed
a triangular relationship where higher altitudes were related to low PBDratio values (major
contribution of PBDTURN) and a wide PBDratio variation at lower altitudes (Appendix S11-
S14).
57
Figure 1. Maps relating phylogenetic beta diversity (PBD) and relative importance of
PBD components (PBDratio) for (a) 4536 terrestrial mammalian species and, separately,
for (b) Carnivora (c) Chiroptera and (d) Rodentia. Each colour change means a 10%
58
quantile shift in our variables. For example, purple areas present top PBD values with a
great contribution of nestedness resultant component (PBDPD). High values of PBD that
are determined by turnover resultant component (PBDTURN) are represented by light
yellow.
Regions for which PBD were most distinct from the expected based on taxonomic
data (PBDdev) included mountainous and desert regions (blue pattern; Fig. 2). In these
areas, species exchange seems to have less importance in determining PBD patterns than
in other regions. On the other hand, for most of the world’s lowlands species exchange
was similar to lineage exchange (Fig. 2). High PBDTURN values in mountainous areas
were more determined by lineage replacement than by species exchange. In the same way,
major contribution of PBDPD found in transitional and desert areas was more related to
lineage loss than simply to species loss. Moreover, we noticed that the Arabian Peninsula
and the transition surrounding the Sahara Desert presented lower contribution of
taxonomic beta diversity in determining PBD patterns for Chiroptera than for Carnivora
and Rodentia (Fig. 2 c).
59
Figure 2. Maps of phylogenetic beta diversity deviations given the taxonomic beta
diversity (PBDdev) for (a) 4536 terrestrial mammalian species and, separately, for (b)
60
Carnivora (c) Chiroptera and (d) Rodentia. PBDdev identifies places where the exchange
of lineages were higher or lower than expected given the exchange of species. If
phylogenetic beta diversity (PBD) is the same as the taxonomic dissimilarity (BD), this
ratio will be zero. Conversely, if PBD among places is low in relation to BD, this ratio
will be close to 1, indicating that there is no exchange of lineages. On the other hand,
negative values indicate that the exchange of lineages is more important than the
taxonomic beta diversity.
DISCUSSION
Climatic gradient and PBD components
Contrary to what was observed for the taxonomic beta diversity of amphibians,
birds and mammals of the New World (Dobrovolski et al., 2012), at global scale we did
not detect a clear pattern in the relative contribution of PBD components regarding
temperature gradient. On the other hand, we found an interesting non-monotonic relation
with inversion points in the contribution of PBD components regarding global climate
zones, (Appendix S7-S10) in accordance with findings for New World mammal beta
diversity (Castro-Insua et al., 2016). Mechanisms, such as tolerance, are known to be
drivers of non-monotonic responses, since organisms may be deeply affected by severe
environmental changes (Zhang et al., 2015). The U-shaped curve that we found (mainly
for Carnivora) might demonstrate the role of extreme temperatures in extinction events
and, consequently lineage loss, which is supported by high PBDPD values (Appendix S7-
S10).
In some cases, (mainly for Carnivora and Chiroptera) warmer climates were
mostly linked to a wide variation of PBDratio values, while at colder climates the PBDPD
contribution clearly increases towards the poles (Appendix S7-S10). Dobrovolski et al.
61
(2012) demonstrated that high contribution of nestedness-resultant component in
taxonomic beta diversity was related to areas covered by ice during the last glaciations,
as suggested by Baselga (2012). They argued that these areas went through extinction
events and were recolonized since then, missing poor dispersal species. Environmental
constraints caused by low temperatures may have historically influenced spatial lineage
distribution, but we were only able to detect that pattern from a specific breaking point,
which varied for each taxonomic group (Appendix S7-S10). Here, the poorest model fit
was related to Chiroptera, the most temperature-sensitive among the analyzed
mammalian orders. There is a close relationship between temperature and energy costs
for Chiroptera (Stevens, 2006; McCain, 2009) and we did not identify a clear increment
of PBDPD values towards polar climate (glaciation imprints in the lineage dissimilarity
pattern) given that chiropteran distribution do not reach extremely low temperatures.
Dispersal abilities and PBD components
We did not find support to our hypothesis that dispersal abilities of mammalian
orders would determine PBDPD contribution in temperate regions. We expected to find
major PBDPD contribution in colder temperatures for Rodentia, which on average have
small distribution sizes and, therefore, could be less efficient in recolonizing areas that
were affected by climatic changes (e.g. Dobrovolski et al. 2012). Nonetheless, we found
that PBDratio in areas with temperate climate was higher for Carnivora than for Chiroptera
and Rodentia, demonstrating that PBDPD was more important for groups with better
dispersal abilities (Appendix S7-S10). Even not corroborating our hypothesis, dispersal
ability seems to be the main factor determining PBD component contributions to
phylogenetic dissimilarity. We found that, in general, Rodentia presented higher PBD
values [Appendix S4-S6 (a)], which is in accordance with previous findings that less
62
vagile organisms tend to have higher species beta diversity (Buckley et al., 2008; Qian &
Ricklefs, 2012). Moreover, the great contribution of turnover-resultant component for
Rodentia may also be explained by its lower dispersal capacity. Rodentia lineages may
have had more opportunities for in situ speciation, as they tend to have more fragmented
populations that could experience reduced gene flow. On the contrary, high dispersal
abilities should render Carnivora species less sensitivity to barriers, which should inhibit
speciation (Gaston, 1998; Birand et al., 2012).
Dobrovolski et al. (2012) compared taxonomic beta diversity components for
birds, mammals and amphibians, and found that poor dispersal ability increased the
contribution of nestedness-resultant component towards polar regions. In accordance with
our results, dispersal capacity does not act in the same way to generate lineage loss in
colder regions. Nonetheless, Dobrovolski et al. (2012) compared groups with very
distinct evolutionary history whereas here we are working with mammalian orders.
Amphibians are more sensitive to extreme climates because they are ectothermic and they
have very poor dispersal abilities when compared with birds and mammals (Smith &
Green, 2005; Buckley & Jetz, 2007). Therefore, at that taxonomic scale, they were able
to capture distinct imprints of glaciation extinctions among groups. On the other hand,
our findings are in agreement with Castro-Insua et al. (2016) results, which demonstrated
major importance of physiological constraints, compared to dispersal ability, to determine
beta diversity components contribution along latitudinal gradient. However, to infer
deterministic mechanisms it is important to evaluate the contribution of PBD (or BD)
components (PBDratio) taking into account the observed PBD values (Fig. 1). We need to
have in mind that presence or absence of a single species would substantially affect the
contributions of PBD components in areas with extremely low PBD values. For instance,
major contribution of PBDPD in American high latitudes for Carnivora [red cells;
63
Appendix S4 (b)] is also related to low PBD values [blue cells; Fig. 1(b); Appendix S4
(a)].
Altitude and PBD components
The most noticeable pattern we found was high PBD values in mountainous areas,
similar to what has been observed in taxonomic beta diversity (Mcknight et al., 2007;
Melo et al., 2009; Dobrovolski et al., 2012). In the Western Hemisphere, we found the
highest PBD values along the mountainous Pacific edge of the continent, highlighting the
Western Cordilleran System at the Neartic region (Sanmartín et al., 2001) and the Andes
at the Neotropical region (e.g. Mcknight et al., 2007; Melo et al., 2009). In the Eastern
Hemisphere, high PBD values were observed in continental Asia, where a complex of
high mountainous areas are found, including the Altai Mountains (Klinge et al., 2003)
and the Tibetan Plateau (Zheng, 1996). Analogously to temperate zones, we predicted that
highlands represent environmental barriers containing extinction imprints with major
contribution of PBDPD in comparison with areas of low elevation. However, we found
that extreme high altitudes are often related to a major contribution of PBDTURN (low
values of PBDratio; Appendix S11-S14), particularly for Rodentia (light yellow; Fig.1). In
highlands, large differences in altitude and temperature occur over short distances, which
is expected to increase beta diversity (e.g. Mcknight et al., 2007; Melo et al., 2009). High
environmental heterogeneity among short-distance sites may create barriers that
accelerate diversification events (Weir, 2006). In the Andean region, for instance, species
richness is higher than predicted by models of climate and altitude (Weir, 2006). These
barriers could have been even more important to groups with low dispersal capacity, such
as Rodentia, which favored speciation (see Gaston, 1998; Birand et al., 2012) and
contributed to lineage segregation and, consequently, the high PBDTURN values we found.
64
PBDdev patterns
In general, PBDdev patterns highlighted mountainous areas, demonstrating that
lineage exchange in these areas is more important when compared with species exchange.
Indeed, Graham et al. (2009) demonstrated that the strong environmental gradient
produced by elevation differences might produce larger than expected PBD for
hummingbirds, suggesting that members from distinct clades are segregated along these
regions. Additionally, we found that deserts are important areas for chiropteran lineage
exchange, which could be explained for the same reason, since deserts represent a strong
environmental filter for this group (Peixoto et al., 2014).
An interesting pattern arose from negative PBDdev values, which are represented
in dark grey (Fig. 2) and indicate that PBD is more important than taxonomic beta
diversity. It is important to note that our methodological approach restricts comparisons
among neighboring cells, thus, these results reflect local information about phylogenetic
and taxonomic composition. These values resulted from cases where there is low
taxonomic beta diversity among nearby cells and where they did not share particular
species that represent a large branch of phylogeny. In the Carnivora map, for instance,
most dark grey cells are concentrated in extreme north latitudes. In the Eastern
hemisphere, most of these cases were related to the presence of Lynx lynx and Ursus
arctos and in the Western hemisphere to Lynx canadensis.
Transitional areas
Transitional areas found in the South Sahara Desert, Arabian Peninsula, deserts in
Patagonia region, around the Tibetan Plateau, and Panama Isthmus presented high PBD
values with major contribution of nestedness-resultant component (purple patterns;
65
Fig.1). It might be a result of lineage loss produced by extreme environmental conditions
represented by these areas (e.g. Baselga, 2010). Additionally, it is interesting to note how
it phylogenetic beta diversity varies among mammalian orders. The transitions in the
south of Sahara Desert are less pronounced for Rodentia than for Chiroptera and
Carnivora (Fig. 1 c, d). Rodentia is well distributed and adaptable to deserted
environments (Kelt et al., 1996), while the aridity is in fact a great barrier for Chiroptera
(Peixoto et al., 2014). On the other hand, we noticed that Panama Isthmus transitional
area presented low PBD values for Chiroptera, which demonstrates that lineage
composition is more homogeneous for Chiroptera than for Carnivora and Rodentia
(Peixoto et al., 2014). This indicates that the Isthmus was not a strong barrier to several
Chiroptera lineages, probably due to their flight ability as it was to most non-volant
mammal lineages.
An interesting circular-shape pattern appeared in continental Asia on the Qinghai-
Tibetan Plateau, the largest and highest plateau in the world, with an area of 2.5 × 106
km2 and average elevation of ∼4000 m (Zheng, 1996). Around Tibetan Plateau, high PBD
values are mainly determined by differences in phylogenetic diversity (PBDPD) (purple
patterns), particularly for Rodentia, producing a ring circumscribed by areas where
lineage turnover is preponderant (PBDTURN). PBDPD pattern around Tibetan Plateau may
be a result of lineage loss with increasing altitude, which represents a strong
environmental filter (Presley et al., 2012). The pattern on the Qinghai-Tibetan Plateau is
characterized by low PBD values with more contribution of PBDPD for Rodentia,
Carnivora and all mammalian pooled, while Chiroptera is absent. For Chiroptera, the
environmental filter imposed by high altitudes may be stronger owing to energetic costs
linked to low temperatures for this group (Stevens, 2006; McCain, 2009; Presley et al.,
2012). This characteristic contributes to the presence of few lineages in temperate regions
66
and to their absence in colder environments (Lack & Van Den Bussche, 2010), as well as
extreme arid conditions (Fig. 1 c).
CONCLUSION
Previous studies have demonstrated that decoupling beta diversity, or
phylogenetic beta diversity, in its turnover and nestedness-resultant components may be
used to disentangle effects of different processes in shaping diversity distribution
(Baselga, 2012; Dobrovolski et al., 2012; Leprieur et al., 2012; Mouillot et al., 2013;
Peixoto et al., 2014). Here, we identified for the first time global areas for which
mammalian phylogenetic beta diversity was mainly determined by turnover (PBDTURN)
or nestedness-resultant (PBDPD) components. Higher latitudes seem to be mostly
associated with lineage loss, generating stronger PBDPD patterns. On the other hand,
highlands are linked to lineage segregation, highlighted by major contribution of
PBDTURN. Additionally, we found that patterns on PBD components may reveal
peculiarities related to evolutionary history characteristics, such as dispersal capacity,
which is important to understand differential organism responses to geographical or
environmental barriers.
REFERENCES
Baselga, A. (2010) Partitioning the turnover and nestedness components of beta diversity.
Global Ecology and Biogeography, 19, 134–143.
Baselga, A. (2012) The relationship between species replacement, dissimilarity derived
from nestedness, and nestedness. Global Ecology and Biogeography, 21, 1223–
1232.
Baselga, A., Gómez-Rodríguez, C. & Lobo, J.M. (2012) Historical legacies in world
amphibian diversity revealed by the turnover and nestedness components of Beta
diversity. PloS one, 7, e32341.
Baselga, A., Jiménez-Valverde, A. & Niccolini, G. (2007) A multiple-site similarity
measure independent of richness. Biology letters, 3, 642–645.
67
Baselga, A. & Leprieur, F. (2015) Comparing methods to separate components of beta
diversity. Methods in Ecology and Evolution, 6, 1069–1079.
Baselga, A., Orme, D. & Villeger, S. (2013) betapart: Partitioning beta diversity into
turnover and nestedness components. R package version 1.2., https://cran.r-
project.org/package=betapart.
Bininda-Emonds, O.R.P., Cardillo, M., Jones, K.E., MacPhee, R.D.E., Beck, R.M.D.,
Grenyer, R., Price, S. a, Vos, R. a, Gittleman, J.L. & Purvis, A. (2007) The delayed
rise of present-day mammals. Nature, 446, 507–12.
Birand, A., Vose, A. & Gavrilets, S. (2012) Patterns of Species Ranges, Speciation, and
Extinction. The American Naturalist, 179, 1–21.
Böhning-Gaese, K., L. I. Gonz_alez-Guzm_an, and J. H. Brown (1998) Constraints on
dispersal and the evolution of the avifauna of the Northern Hemisphere. Evolutionary
Ecology, 12: 767–783.
Brown, J.H. (2014) Why are there so many species in the tropics? Journal of
Biogeography, 41, 8–22.
Bryant, J.A., Lamanna, C., Morlon, H., Kerkhoff, A.J., Enquist, B.J. & Green, J.L. (2008)
Microbes on mountainsides : Contrasting elevational patterns of bacterial and plant
diversity. Proceedings of the National Academy of Sciences of the United States of
America, 105, 11505–11511.
Buckley, L.B. & Jetz, W. (2007) Environmental and historical constraints on global
patterns of amphibian richness. Proceedings. Biological sciences / The Royal
Society, 274, 1167–1173.
Buckley, L.B., Rodda, G.H. & Jetz, W. (2008) Thermal and energetic constraints on
ectotherm abundance: a global test using lizards. Ecology, 89, 48–55.
Castro-Insua, A., Gómez-Rodríguez, C. & Baselga, A. (2016) Break the pattern :
breakpoints in beta diversity of vertebrates are general across clades and suggest
common historical causes. Global Ecology and Biogeography, 25, 1279–1283.
Diserud, O.H. & Ødegaard, F. (2007) A multiple-site similarity measure. Biology letters,
3, 20–22.
Dobrovolski, R., Melo, A.S., Cassemiro, F. a S. & Diniz-Filho, J.A.F. (2012) Climatic
history and dispersal ability explain the relative importance of turnover and
nestedness components of beta diversity. Global Ecology and Biogeography, 21,
191–197.
Fritz, S.A., Bininda-Emonds, O.R.P. & Purvis, A. (2009) Geographical variation in
predictors of mammalian extinction risk: big is bad, but only in the tropics. Ecology
Letters, 12, 538–549.
Gaston, K.J. (2000) Global patterns in biodiversity. Nature, 405, 220–7.
Gaston, K.J. (1996) Species-range-size distributions: Patterns, mechanisms and
implications. Trends in Ecology and Evolution, 11, 197–201.
Gaston, K.J. (1998) Species-range size distributions: products of speciation, extinction
and transformation. Philosophical Transactions of the Royal Society B: Biological
Sciences, 353, 219–230.
Graham, C.H. & Fine, P.V.A. (2008) Phylogenetic beta diversity : linking ecological and
evolutionary processes across space in time. Ecology Letters, 11, 1265–1277.
Graham, C.H., Moritz, C. & Williams, S.E. (2006) Habitat history improves prediction
of biodiversity in rainforest fauna. Proceedings of the National Academy of Sciences
of the United States of America, 103, 632–636.
Graham, C.H., Parra, J.L., Rahbek, C. & Mcguire, J.A. (2009) Phylogenetic structure in
tropical hummingbird communities. PNAS, 106, 19673–19678.
68
Harrison, S., Ross, S.J. & Lawton, J.H. (1992) Beta diversity on geographic gradients in
Britain. Journal of Animal Ecology, 61, 141–148
Hijmans, R.J., Camerons, S.E., Parra, J.L., Jones, P.G. & Jarvis, A. (2005) Very high
resolution interpolated climate surfaces for global land areas. International Journal
of Climatology, 25, 1965–1978.
Hillebrand, H. (2004) On the generality of the latitudinal diversity gradient. The American
naturalist, 163, 192–211.
IUCN (2015) The IUCN Red List of Threatened Species. Version 2015.1,
http://www.iucnredlist.org.
Jablonski, D., Roy, K. & Valentine, J.W. (2006) Out of the tropics: evolutionary dynamics
of the latitudinal diversity gradient. Science, 314, 102–106.
Jaccard, P. (1912) The distribution of the flora in the alphine zone. The New Phytologist,
XI, 37–50.
Kelt, D.A., Brown, J.H., Heske, E.J., Marquet, P. a, Stephen, R., Reid, J.R.W., Kontantín,
A. & Shenbrot, G. (1996) Community Structure of Desert Small Mammals :
Comparisons Across Four Continents. Ecology, 77, 746–761.
Klinge, M., Böhner, J. & Lehmkuhl, F. (2003) Climate pattern, snow- and timberlines in
the Altai Mountains, Central Asia. Erdkunde, 57, 296–307.
Kubota, Y., Hirao, T., Fujii, S. & Murakami, M. (2011) Phylogenetic beta diversity
reveals historical effects in the assemblage of the tree floras of the Ryukyu
Archipelago. Journal of Biogeography, 38, 1006–1008.
Lack, J.B. & Van Den Bussche, R.A. (2010) Identifying the confounding factors in
resolving phylogenetic relationships in Vespertilionidae. Journal of Mammalogy,
91, 1435–1448.
Lennon, J.J., Koleff, P., Greenwood, J.J.D. & Gaston, K.J. (2001) The geographical
structure of British bird distributions: Diversity, spatial turnover and scale. Journal
of Animal Ecology, 70, 966–979.
Leprieur, F., Albouy, C., Bortoli, J. De, Cowman, P.F., Bellwood, D.R. & Mouillot, D.
(2012) Quantifying phylogenetic beta diversity : distinguishing between “ true ”
turnover of lineages and phylogenetic diversity gradients. PloS one, 7, e42760.
Lozupone, C. & Knight, R. (2005) UniFrac : a New Phylogenetic Method for Comparing
Microbial Communities. Applied and environmental microbiology, 71, 8228–8235.
McCain, C.M. (2009) Vertebrate range sizes indicate that mountains may be “higher” in
the tropics. Ecology letters, 12, 550–560.
Mcknight, M.W., White, P.S., Mcdonald, R.I., Lamoreux, J.F., Sechrest, W., Ridgely,
R.S. & Stuart, S.N. (2007) Putting Beta-Diversity on the Map : Broad-Scale
Congruence and Coincidence in the Extremes. PLoS biology, 5, e272.
Melo, A. (2013) CommEcol - Community Ecology analyses. R package version
1.5.8/r24., https://r-forge.r-project.org/projects/commecol/.
Melo, A.S., Cianciaruso, M. V. & Almeida-Neto, M. (2014) treeNODF: Nestedness to
phylogenetic, functional and other tree-based diversity metrics. Methods in Ecology
and Evolution, 5, 563–572.
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.
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., 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. Ecology
69
Letters, 10, 315–331.
Mouillot, D., De Bortoli, J., Leprieur, F., Parravicini, V., Kulbicki, M. & Bellwood, D.R.
(2013) The challenge of delineating biogeographical regions: Nestedness matters for
Indo-Pacific coral reef fishes. Journal of Biogeography, 40, 2228–2237.
Muggeo, V.M.R. (2008) segmented: An R package to Fit Regression Models with
Broken-Line Relationships. R News, 8, 20–25.
Peixoto, F.P., Braga, P.H.P., Cianciaruso, M.V., Diniz-Filho, J.A.F. & Brito, D. (2014)
Global patterns of phylogenetic beta diversity components in bats. Journal of
Biogeography, 41, 762–772.
Potts, M.D., Davies, S.J., Bossert, W.H., Tan, S. & Nur Supardi, M.N. (2004) Habitat
heterogeneity and niche structure of trees in two tropical rain forests. Oecologia,
139, 446–453.
Presley, S.J., Cisneros, L.M., Patterson, B.D. & Willig, M.R. (2012) Vertebrate
metacommunity structure along an extensive elevational gradient in the tropics: a
comparison of bats, rodents and birds. Global Ecology and Biogeography, 21, 968–
976.
Purvis, A. & Hector, A. (2000) Getting the measure of biodiversity. Nature, 405, 212–
219.
Qian, H. & Ricklefs, R.E. (2012) Disentangling the effects of geographic distance and
environmental dissimilarity on global patterns of species turnover. Global Ecology
and Biogeography, 21, 341–351.
Qian, H. & Xiao, M. (2012) Global patterns of the beta diversity-energy relationship in
terrestrial vertebrates. Acta Oecologica, 39, 67–71.
Ricklefs, R.E. (1987) Community diversity : relative roles of and regional processes
testing predictions of local-process theories. Science, 235, 167–171.
Rolland, J., Condamine, F.L., Jiguet, F. & Morlon, H. (2014) Faster Speciation and
Reduced Extinction in the Tropics Contribute to the Mammalian Latitudinal
Diversity Gradient. PLoS Biology, 12.
Safi, K., Cianciaruso, M. V, Loyola, R.D., Brito, D., Armour-Marshall, K. & Diniz-Filho,
J.A.F. (2011) Understanding global patterns of mammalian functional and
phylogenetic diversity. Philosophical transactions of the Royal Society of London
B: Biological sciences, 366, 2536–2544.
Sanmartín, I., Enghoff, H. & Ronquist, F. (2001) Patterns of animal dispersal, vicariance
and diversification in the Holarctic. Biological Journal of the Linnean Society, 73,
345–390.
Schloss, C. a., Nunez, T. a. & Lawler, J.J. (2012) Dispersal will limit ability of mammals
to track climate change in the Western Hemisphere. Proceedings of the National
Academy of Sciences, 109, 8606–8611.
Smith, M.A. & Green, D.M. (2005) Dispersal and the metapopulation in amphibian and
paradigm ecology are all amphibian conservation: populations metapopulations ?
Ecography, 28, 110–128.
Sørensen, T.A. (1948) A method of establishing groups of equal amplitude in plant
sociology based on similarity of species content, and its application to analyses of
the vegetation on Danish commons. Kongelige Danske Videnskabernes Selskabs
Biologiske Skrifter, 5, 1–34.
Stevens, R.D. (2006) Historical processes enhance patterns of diversity along latitudinal
gradients. Proceedings of The Royal Society B: Biological Sciences, 273, 2283–
2289.
Tuomisto, H., Ruokolainen, K. & Yli-Halla, M. (2003) Dispersal, environment, and
70
floristic variation of western Amazonian forests. Science, 299, 241–244.
Weir, J.A.T.W. (2006) Divergent Timing and Patterns of Species Accumulation in
Lowland and Highland Neotropical Birds. Evolution, 60, 842–855.
Whittaker, R.H. (1972) Evolution and Measurement of Species Diversity. Taxon, 21,
213–251.
Wiens, J.J. & Donoghue, M.J. (2004) Historical biogeography, ecology and species
richness. Trends in ecology & evolution, 19, 639–644.
Zeileis, A., Leisch, F., Homik, K. & Kleiber, C. (2002) strucchange: an R package for
testing for structural change. Journal of Statistical Software, 7, 1–38.
Zhang, Z., Yan, C., Krebs, C.J. & Stenseth, N.C. (2015) Ecological non-monotonicity
and its effects on complexity and stability of populations , communities and
ecosystems. Ecological Modelling, 312, 374–384.
Zheng, D. (1996) The system of physio-geogrphical regions of the Qinghai-Xizan (Tibet)
Plateau. Science in China, 39, 410–417.
71
SUPPORTING INFORMATION
Appendix S1. Maps relating phylogenetic beta diversity (PBD) (second scale: radii of
two cell layers adjacent to the focal cell) and relative importance of PBD components
(PBDratio) for (a) mammalian species and, separately, for (b) Carnivora (c) Chiroptera and
(d) Rodentia. Each colour change means a 10% quantile shift in our variables.
72
Appendix S2. Maps relating phylogenetic beta diversity (PBD) (third scale: radii of three
cell layers adjacent to the focal cell) and relative importance of PBD components
(PBDratio) for (a) mammalian species and, separately, for (b) Carnivora (c) Chiroptera and
(d) Rodentia. Each colour change means a 10% quantile shift in our variables.
73
Appendix S3. Maps of the (a) phylogenetic beta diversity (PBD) and (b) the relative
importance of the PBD components (PBDratio) for 4536 terrestrial mammalian species.
PBD varies from 0 (when the shared branch lengths is identical) to near 1 (when the
assemblages compared share no branch in the rooted phylogenetic tree). PBDratio values
smaller than 0.5 indicate that PBD is mainly determined by the turnover-resultant
dissimilarity, whereas values greater than 0.5 indicate that PBD results from a greater
influence of the component of dissimilarity due to PD differences between assemblages.
74
Appendix S4. Maps of the (a) phylogenetic beta diversity (PBD) and (b) the relative
importance of the PBD components (PBDratio) for Carnivora. PBD varies from 0 (when
the shared branch lengths is identical) to near 1 (when the assemblages compared share
no branch in the rooted phylogenetic tree). PBDratio values smaller than 0.5 indicate that
PBD is mainly determined by the turnover-resultant dissimilarity, whereas values greater
than 0.5 indicate that PBD results from a greater influence of the component of
dissimilarity due to PD differences between assemblages.
75
Appendix S5. Maps of the (a) phylogenetic beta diversity (PBD) and (b) the relative
importance of the PBD components (PBDratio) for Chiroptera. PBD varies from 0 (when
the shared branch lengths is identical) to near 1 (when the assemblages compared share
no branch in the rooted phylogenetic tree). PBDratio values smaller than 0.5 indicate that
PBD is mainly determined by the turnover-resultant dissimilarity, whereas values greater
than 0.5 indicate that PBD results from a greater influence of the component of
dissimilarity due to PD differences between assemblages.
76
Appendix S6. Maps of the (a) phylogenetic beta diversity (PBD) and (b) the relative
importance of the PBD components (PBDratio) for Rodentia. PBD varies from 0 (when the
shared branch lengths is identical) to near 1 (when the assemblages compared share no
branch in the rooted phylogenetic tree). PBDratio values smaller than 0.5 indicate that PBD
is mainly determined by the turnover-resultant dissimilarity, whereas values greater than
0.5 indicate that PBD results from a greater influence of the component of dissimilarity
due to PD differences between assemblages.
77
Appendix S7. Relationship between mean annual temperature and relative contribution
of Mammalia PBD components (PBDratio) for the (a) first, second (b) and third scale (c)
analysed (radii of one, two, and three cell layers adjacent to the focal cell). Results for
linear regression model (lm) and piecewise regression (pw) are displayed. Fitted functions
of piecewise regression are shown (black lines).
78
Appendix S8. Relationship between mean annual temperature and relative contribution
of Carnivora PBD components (PBDratio) for the (a) first, second (b) and third scale (c)
analysed (radii of one, two, and three cell layers adjacent to the focal cell). Results for
linear regression model (lm) and piecewise regression (pw) are displayed. Fitted functions
of piecewise regression are shown (black lines).
79
Appendix S9. Relationship between mean annual temperature and relative contribution
of Chiroptera PBD components (PBDratio) for the (a) first, second (b) and third scale (c)
analysed (radii of one, two, and three cell layers adjacent to the focal cell). Results for
linear regression model (lm) and piecewise regression (pw) are displayed. Fitted functions
of piecewise regression are shown (black lines).
80
Appendix S10. Relationship between mean annual temperature and relative contribution
of Rodentia PBD components (PBDratio) for the (a) first, second (b) and third scale (c)
analysed (radii of one, two, and three cell layers adjacent to the focal cell). Results for
linear regression model (lm) and piecewise regression (pw) are displayed. Fitted functions
of piecewise regression are shown (black lines).
81
Appendix S11. Relationship between altitude and relative contribution of Mammalia
PBD components (PBDratio) for the (a) first, second (b) and third scale (c) analysed (radii
of one, two, and three cell layers adjacent to the focal cell). Results for linear regression
model (lm) and piecewise regression (pw) are displayed. Fitted functions of piecewise
regression are shown (black lines).
82
Appendix S12. Relationship between altitude and relative contribution of Carnivora PBD
components (PBDratio) for the (a) first, second (b) and third scale (c) analysed (radii of
one, two, and three cell layers adjacent to the focal cell). Results for linear regression
model (lm) and piecewise regression (pw) are displayed. Fitted functions of piecewise
regression are shown (black lines).
83
Appendix S13. Relationship between altitude and relative contribution of Chiroptera
PBD components (PBDratio) for the (a) first, second (b) and third scale (c) analysed (radii
of one, two, and three cell layers adjacent to the focal cell). Results for linear regression
model (lm) and piecewise regression (pw) are displayed. Fitted functions of piecewise
regression are shown (black lines).
84
Appendix S14. Relationship between altitude and relative contribution of Rodentia PBD
components (PBDratio) for the (a) first, second (b) and third scale (c) analysed (radii of
one, two, and three cell layers adjacent to the focal cell). Results for linear regression
model (lm) and piecewise regression (pw) are displayed. Fitted functions of piecewise
regression are shown (black lines).
85
CAPÍTULO 03: PHYLOGENETIC AND
FUNCTIONAL STRUCTURE OF
AFRICAN MAMMAL ASSEMBLAGES:
THE IMPRINT OF HISTORICAL
CLIMATIC CHANGES ON HABITAT
FORMATION
86
PHYLOGENETIC AND FUNCTIONAL STRUCTURE OF AFRICAN MAMMAL
ASSEMBLAGES: THE IMPRINT OF HISTORICAL CLIMATIC CHANGES ON
HABITAT FORMATION
Franciele Parreira Peixoto, Joaquín Hortal, Jesús Rodríguez, Marcus Vinicius
Cianciaruso
ABSTRACT
Aim: to investigate the phylogenetic and functional structure of mammal assemblages
distributed over the major African habitats. We addressed the role of historical and
ecological processes in determining the lineage pool formation and local species co-
occurrence.
Location: Afrotropical region
Methods: we explored the phylogenetic and functional structure of mammal local
assemblages and evaluated the phylogenetic and functional structure by habitat to know
in what extent the assemblage structure is a reflection of the habitat lineage/trait pool. We
also assessed the phylogenetic and functional dissimilarities among habitats, and
investigated the contribution of specific clades to the phylogenetic structure and
dissimilarity of the evaluated habitats.
Reuslts: there was a general tendency for increasing phylogenetic overdispersion and
functional clustering towards drier and less productive environments, an evidence that in
harsher environments species from distinct lineages converged to a set of similar
functional traits. On the other hand, closely related species had high trait diversity towards
forested habitats. Additionally, functional turnover was more important than phylogenetic
turnover among habitats.
87
Main conclusions: overall, it seems that a complex history of colonization among
habitats was responsible for lineage pool formation of African biomes. This history of
lineage colonization and diversification was different for specific clades and related to
ecological flexibility and, consequently, the ability to lead with climatic changes, which
have been the main driver of African mammal diversity along the evolutionary time.
Keywords: Afrotropical, historical processes, rainforests, savannas, phylogenetic
diversity, functional diversity.
INTRODUCTION
The assembly of biotic communities has been by far one of the most explored
subjects in ecology (MacArthur & Wilson, 1967; Cody & Diamond, 1975; Ricklefs &
Schluter, 1993; Graves & Rahbek, 2005; Rodríguez et al., 2006; Adler et al., 2013).
Species co-occurrence patterns are shaped by regional and local processes that operate in
a continuum of scales (Ricklefs, 1987). It is hypothesized that historical large-scale
processes (e.g. speciation, extinction, dispersal) shape regional composition, and species
are then filtered by environmental conditions and ecological processes (e.g. competition,
predation) to form local assemblages (Ricklefs, 1987; Ricklefs & Schluter, 1993; Zobel,
1997). Historical processes, such as the time for speciation and past climatic changes,
may drive diversification (Stephens & Wiens, 2003; Wiens, 2011; Kissling et al., 2012)
and shape the available diversity. Moreover, niche evolution and formation of
geographical barriers over geological time may affect dispersion among regions (Renner,
2004; Webb, 2006). Thereby, local patterns rely on historical processes that have
determined which species are available and apt to be present in a given site (Ricklefs,
1987; Rodríguez et al., 2006).
88
By knowing current patterns of community structure, we can understand about the
determinants of community assembly. The phylogenetic structure of a given community
may be used to infer historical processes that shaped the habitat lineage pool in which a
particular community is placed (Hubert et al., 2011; Kissling et al., 2012; Gerhold et al.,
2015). For example, phylogenetic overdispersion might be found in communities from
habitats formed by ancient lineages, and phylogenetic clustering may be driven by recent
diversification in young habitats (Gerhold et al., 2015; Souza-Neto et al., 2016). On the
other hand, by knowing functional community structure associated to niche-related traits
we can better conclude about the acting ecological processes, since they are niche-based
processes (Gerhold et al., 2015). Competition would drive patterns of co-occurrence of
species with different functional traits (resulting in overdispersion), and habitat filtering
would select species with the same characteristics able to deal with stressful environments
(resulting in clustering) (Webb et al., 2002). Thus, by allying functional and phylogenetic
approaches, we should be able to better understand the role of local processes in
structuring community composition and to discriminate it from historical-evolutionary
processes determining lineage pool formation (Safi et al., 2011; Cianciaruso et al., 2012;
Gerhold et al., 2015).
Besides the structure of assemblages, we can investigate the formation of lineage
pool by knowing the phylogenetic beta dissimilarity among sites (Graham & Fine, 2008;
Peixoto et al., 2014, 2017). For example, if distinct habitats are formed by similar
lineages, the phylogenetic dissimilarity between them would be low (Peixoto et al., 2014;
Souza-Neto et al., 2016). The same reasoning could also be followed for functional
diversity (e.g. Swenson et al., 2012). For example, the lineage pool formation of a more
recent habitat/biome might be driven by colonization of lineages from an older habitat
(e.g. Souza-Neto et al., 2016). The colonization of lineages, which characterizes habitat
89
shift, may be accompanied by the development of specific adaptations if the change of
habitat is related to a harsh environmental gradient, such as the aridity gradient (Souza-
Neto et al., 2016). In this case, the lineage composition between habitats would be similar,
but the composition of functional characteristics would be different. On the other hand,
the prevalence of niche conservatism might reduce the exchange of lineages among
habitats because the lineages will conserve environmental tolerances from their ancestral,
rarely shifting of habitat (Wiens & Donoghue, 2004; Crisp et al., 2009).
In the African continent, a unique combination of large-scale historical-
evolutionary processes, such as, continental drift, past climatic changes, extinction,
speciation, and dispersal events have determined the current composition of regional biota
(Janis, 1993; de Vivo & Carmignotto, 2004; Lawes et al., 2007). Continual cycles of
climatic changes have repeatedly caused events of forest contraction and expansion,
which had a significant impact on the diversity and distribution of African flora and fauna
(Janis, 1993; Plana, 2004). The Afrotropics was the continental region with the highest
rate of rainforest loss since Eocene due to a drier climate along most of equatorial Africa
(Maley, 1996; Kissling et al., 2012; Willis et al., 2013). Reduction of rainforest coverage
during glacial maxima confined forest organisms to small areas promoting local
extinctions and in some cases led to the elimination of entire lineages (Vrba, 1992; Janis,
1993; Plana, 2004; Lawes et al., 2007). Conversely, the Afrotropical region has supported
large and stable areas of suitable dry open habitats even during moister periods, when
savannas would have expanded into previously arid areas (Cristoffer & Peres, 2003; de
Vivo & Carmignotto, 2004). That could have turned savannas into a diversification arena
where forest lineages that evolved adaptations would be able to colonize dry open
environments (Cristoffer & Peres, 2003; de Vivo & Carmignotto, 2004). Thus, African
past climactic events led to distinct historical and evolutionary processes determining
90
different habitat lineage-pool, which is supposed to influence community assembly
although it has not been extensively explored (see Gerhold et al., 2015).
Here, we tested the hypothesis that (H1) historical climatic instability and habitat
area loss led to lineage loss in rainforests, which coupled with speciation in forest refuges
and colonization/adaptation of forest lineages to dry open environments (habitat shift)
may have acted shaping contrasting community structure patterns among habitats in a
dryness gradient (Cristoffer & Peres, 2003; de Vivo & Carmignotto, 2004; deMenocal,
2004; Plana, 2004; Lawes et al., 2007; Kissling et al., 2012). From a functional
perspective, we tested the hypothesis (H2) that dryness would represent a strong
environmental filter, tending to select species with similar functional traits that allow
them to colonize and survive in dry open habitats (Ojeda et al., 1999; Cristoffer & Peres,
2003; deMenocal, 2004; Cantalapiedra et al., 2014). On the other hand, competition
would play a major role in more productive environments (e.g. rainforests), where a
higher variety of traits would be found (Schoener, 1982; Weiher & Keddy, 1995) (see
Table 1 for predictions). We explored the phylogenetic and functional structure of
mammal local assemblages distributed over the major African biomes (from desert and
savanna to rainforests). We also evaluated the phylogenetic and functional structure by
habitat to know in what extent the assemblage structure is a reflection of the habitat
lineage/trait pool. We also assessed the phylogenetic and functional dissimilarities among
habitats, and investigated the contribution of specific clades to the phylogenetic structure
and dissimilarity of the evaluated habitats, to test for our hypothesis of dry open habitat
colonization by rainforest lineages (e.g. Leprieur et al., 2012; Souza-Neto et al., 2016).
91
Table 1. Hypotheses and their predictions about phylogenetic and functional patterns for
African mammal assemblages.
Hypotheses Predictions
H1: Historical climatic instability and loss of
rainforest area led to lineage loss in forested
habitats and, at same time, speciation in
forest refuges. On the other hand, the
vastness and stability of dry open habitats
during glacial cycles would have turned them
into diversification sites for several forest
lineages that have evolved adaptations to
colonize these environments. Thus, the
lineages found in dry open habitats will be
the result of habitat shift of rainforest
lineages (Cristoffer & Peres, 2003; de Vivo
& Carmignotto, 2004; deMenocal, 2004).
P1.1. We expect to find progressive levels
of phylogenetic clustering in Tropical
Rainforest assemblages and/or at habitat
level.
P1.2. We expect to find progressive levels
of phylogenetic overdispersion in dry
open habitat assemblages and/or at
habitat level.
P1.3. Phylogenetic beta diversity among
Rainforest and the other habitats will be
lower than expected by chance.
P1.4. Phylogenetic beta diversity among
Rainforest and the other habitats will be
mainly due to phylogenetic nestedness.
P1.5. We expect to find in savanna habitats
overabundance of the same higher taxa
found in forested habitats.
H2: Adaptations to dry open habitats there
were required to make possible the
colonization by forest lineages. Dryness
would represent a strong environmental filter
where only specialized adaptations can
successfully persist (Ojeda et al., 1999;
Cristoffer & Peres, 2003; deMenocal, 2004).
In opposition, in more productive
environments, where competition tend to
play a major role, a higher variety of traits
would be found (Schoener, 1982; Weiher &
Keddy, 1995)
P2.1. We expect to find progressive levels
of functional clustering towards dry open
environments.
P2.2. We expect to find progressive levels
of functional overdispersion towards
rainforest habitat.
P2.3. Functional beta diversity among
Rainforest and the other habitats will be
higher than expected by chance.
P2.4. Functional beta diversity among
Rainforest and the other habitats will be
mainly due to functional turnover.
92
METHODS
Local assemblages
We selected 16 terrestrial non-volant mammal checklists (Chiroptera, Cetacea and
Pinnipedia were excluded) located in all major continental habitats of sub-Saharan Africa
(see Table S1 for references). These checklists were selected due to their completeness
and, especially, because they are located in protected areas covered mainly by a single
biome-type (for example, rainforest or savanna). These sites can be ordered according to
their biome-type in a dryness gradient: desert, savanna, humid savanna, dry forest and
rainforest (Table 2; Figure 1).
Table 2. Description of the sixteen studied communities regarding their identity (ID) and
site names, habitat from Bailey (1989) provinces, species richness, (S) and geographical
coordinates in decimal degrees (Lat, Long).
ID Site Habitat S Lat Long
1 Basse Casamance National Park (Senegal) Humid/semi humid Savanna 45 12.45 -16
2 Parc National de la Pendjari (Benin) Humid/semi humid Savanna 42 11.00 1.50
3 Taï National Park (Ivory Coast) Humid/Semi-humid forests 64 5.46 -6.65
4 Omo Biosphere Reserve (Nigeria) Humid/Semi-humid forests 32 6.50 4.25
5 Dja Faunal Reserve (Cameroon) Humid/Semi-humid forests 95 3.06 13
6 Parc National d'Odzala (Rep. of Congo) Humid/Semi-humid forests 62 0.80 14.88
7 Kibale Forest Corridor National Park (Uganda) Dry Forests/shrub 65 0.5 30.2
8 Amboseli National Park (Kenya) Humid/semi humid Savanna 67 -1.61 37.15
9 West Caprivi Game Reserve (Namibia) Savanna/Shrubland 89 -16.09 22.62
10 Kalahari Gemsbok National Park (South Africa) Savanna/Shrubland 56 -25.68 20.37
11 Naute Dam (Namibia) Desert/Semi-arid 66 -26.97 17.96
12 Malolotjia Nature Reserve (Swaziland) Savanna/Shrubland 68 -26 31.03
13 Mlawula Nature Reserve (Swaziland) Savanna/Shrubland 66 -26 32
14 Augrabies Falls National Park (South Africa) Desert/Semi-arid 46 -27.42 20.35
15 Weenen Game Reserve (South Africa) Humid/semi humid Savanna 39 -28.7 30.4
93
16 Kogelberg Nature Reserve (South Africa) Dry Forests/shrub 55 -33.77 19
Figure 1. Map of Afrotropical region showing the sixteen studied assemblages. The
habitat in which the assemblages are placed is displayed by different symbols (see the
legend). The numbers showed next to the symbols represent the assemblage ID, which
are listed in table 2.
Phylogenetic and functional information
To perform the phylogenetic analyses, we used a dated mammalian phylogenetic
tree (Bininda-Emonds et al., 2007; Fritz et al., 2009). When a given species was not
present in the original phylogeny, we placed it as a polytomy within a sister taxon. After
94
all taxonomic revisions and polytomies insertions (seven species) we obtained a
phylogenetic tree with 750 species (around 80% of all terrestrial non-volant mammals
registered; Galster et al., 2007). To implement functional analyses, we built a dendrogram
using an updated version of the PanTHERIA functional data (Jones et al., 2009; Hidasi-
Neto et al., 2015). We used quantitative data of body mass and binary data of dietary,
habit and activity and missing data were extrapolated by genus (408 species had at least
one of these traits extrapolated by genus).
Assemblage level analyses
One of the greatest challenges in investigating community structure is the species
pool definition (Connor & Simberloff, 1979; Gotelli & McCabe, 2002; Webb et al., 2002;
Lessard et al., 2012a). Results largely depend on how species pools are defined, which
reflects historical processes that shaped species composition at regional scale (see
Carstensen et al., 2013a). We built eleven source pools for each of the 16 studied
assemblages using a geographic database on the distribution of all studied species across
mainland sub-Saharan Africa (South of 20 º N) at the resolution of 1º x 1º grid cells
(Galster et al., 2007; held at the Zoological Museum, University of Copenhagen). The
source pool definitions were: (1) continental, (2) circle unweighted, (3) circle weighted,
(4) biome unweighted, (5) biome weighted, (6) dispersion field unweighted, (7)
dispersion field weighted, (8) cluster unweighted, (9) cluster weighted, (10) module
unweighted and (11) module weighted (see appendix S1 for details).
We calculated standardized effect sizes (SES) of phylogenetic diversity (PD;
Faith, 1992) and functional diversity (FD; Petchey & Gaston, 2002) for each mammal
assemblage. We tested for deviance of community structure values in regards to null
expectation. In all null models species richness was fixed and only species identity was
95
randomized (1000 thousand times) to obtain a standardized effect size (SES) (Gotelli &
Entsminger, 2003). We performed these analyses to each source pool definition (SI
methods). To run these analyses we used the function ses.pd implemented in Picante
package (Kembel et al., 2010).
We calculated Pearson correlation coefficients among SES values (phylogenetic
and functional) obtained from the eleven source pool definitions to test the robustness of
our results. Because we did not find qualitative differences among source pool definitions
(Table S5) we only present the results for “dispersion field weighted”, which is a more
realistic approach in biological terms (Graves & Rahbek, 2005). We performed a linear
regression analysis to test for predictions of our hypotheses H1 and H2 regarding the
relation between phylogenetic and functional SES values and the dryness gradient.
Habitat level analyses
We calculated standardized effect sizes of phylogenetic diversity (PD) and
functional diversity (FD) for each African habitat recorded (i.e. all assemblages of a given
habitat pooled together) with the same null model approach used to assemblage level (see
above). To assess phylogenetic (PBD) and functional beta diversity (FBD) among
habitats, we used an additive partitioning framework (Leprieur et al., 2012) based on the
original approach of Baselga (2010, 2012). This framework allows us to assess the
amount of phylogenetic and functional dissimilarity between assemblages (PhyloSor
index; PBD or FBD) that is owing to real lineages or functional traits turnover, (turnover-
resultant dissimilarity PBDTURN or FBDTURN) or determined by phylogenetic or functional
diversity differences between nested assemblages (PD/FD-resultant dissimilarity
component; PBDPD or FBDPD). PD/FD-resultant component is simply the difference
between PBD/FBD and PBDTURN /FBDTURN i.e. (i.e. PBDPD = PBD – PBDTURN and
96
FBDPD = FBD – FBDTURN). A standardized effect size (SES) was calculated for PBD and
PBDTURN. SES values greater than 1.96 indicate a higher PBD than expected by beta
diversity while SES values below -1.96 indicate lower PBD than expected by beta
diversity (Leprieur et al., 2012). To investigate the taxa responsible for phylogenetic
structure and dissimilarity among habitats, we tested each phylogenetic node for an
overabundance of terminal taxa using a nodesig analysis available in Phylocom 4.2
(Webb et al., 2008). The observed patterns for each sample were compared to those for
random samples (fixing the number of terminal taxa) to find the nodes significantly
overabundant (see Souza-Neto et al., 2016 for a similar approach).
RESULTS
At the assemblage level (using specific species pool definition for each
community), savannas and desert habitats were phylogenetically overdispersed (i.e.
formed by distantly related species) while the phylogenetic diversity of rainforest
assemblages was equal to what one may expect by chance (Table S2). We found a
tendency for decreasing SES values (phylogenetic structure) towards rainforest
assemblages (supporting predictions P1.1. and P1.2; Figure 2a). At habitat scale we found
that rainforest lineage pool was composed by closest related lineages (phylogenetic
clustering) and the phylogenetic structure of the other habitats showed a random pattern
(Table S4), with SES values increasing towards more arid environments (supporting
predictions P1.1. and P1.2; Figure 2c). In general, phylogenetic beta diversity (PBD) among
habitats was low, equal to the expected by chance and mainly determined by turnover
component (contrarily to predictions P1.3. and P1.4.). The highest turnover value was found
between rainforest and savanna (Table 3). Rainforest was mainly related to
overabundance of terminal taxa related to the orders Primate, Cetartiodactyla and
97
Carnivora. The nestedness pattern that we predicted in (P1.5) was only supported for
Cetartiodactyla and Carnivora. The carnivoran taxa overabundance was generally related
to different families for rainforest and the other habitats, while Cetartiodactyla was always
related to Bovidae, but with different subfamilies being represented in rainforest and for
humid savanna (Table 4).
98
Figure 2. Linear regression results relating the SES values and the habitat type, for
assemblage (using Dispersion Field unweighted pool) and habitat level. Results for
phylogenetic structure using (a, c) PD and for functional structure, using (b, d) FD.
Rainforest = Humid/Semi-humid forests; Dry Forest = Dry Forests/shrub; H. Savanna =
Humid/semi humid Savanna; Savanna = Savanna/Shrubland and Desert = Desert/Semi-
arid.
Table 3. Pairwise phylogenetic beta diversity (PBD) values among African habitats are
shown below diagonal and turnover-resultant component of PBD (PBDTURN) is above
diagonal. Values in bold indicate smaller standardized effect size (SES) values than
expected by chance. Rainforest = Humid/Semi-humid forests; Dry Forest = Dry
Forests/shrub; H. Savanna = Humid/semi humid Savanna; Savanna = Savanna/Shrubland
and Desert = Desert/Semi-arid.
Habitat Rainforest Dry forest H. Savanna Savanna Desert
Rainforest --- 0.353 0.367 0.456 0.455
Dry forest 0.379 --- 0.373 0.250 0.348
H. Savanna 0.379 0.386 --- 0.194 0.204
Savanna 0.482 0.316 0.248 --- 0.068
Desert 0.549 0.434 0.326 0.271 ---
Table 4. Clades overrepresented on nodesig analysis by habitat. Rainforest =
Humid/Semi-humid forests; Dry Forest = Dry Forests/shrub; H. Savanna = Humid/semi
humid Savanna; Savanna = Savanna/Shrubland and Desert = Desert/Semi-arid
Habitats Clades overrepresented on nodesig analysis
99
The functional structure of the mammal assemblages showed functional clustering
for dry open environments and increasing SES values (functional structure) towards
rainforest assemblages (corroborating P2.1 and P2.2; Table S3; Figure 2b). Functional
structure at the habitat scale showed the same pattern found to assemblages (Table S4;
Figure 2d). Likewise, supporting predictions P2.3 and P2.4., functional beta diversity (FBD)
Rainforest
Carnivora: families Herpestidae and Nandiniidae, subfamily Viverrinae
(Viverridae) and Genetta genus (Viverridae);
Cetartiodactyla: family Bovidae, subfamily Cephalophinae (Bovidae),
Cephalophus and Tragelaphus genera (Bovidae);
Hyracoidea: Dendrohyrax genus (Procaviidae);
Pholidota: Family Manidae;
Primate: Primate order, Families Cercopithecidae and Galagonidae, subfamily
Perodicticinae and Cercopithecus and Colobus genera;
Rodentia: Family Anomaluridae, Anomalurus and Idiurus genera
(Anomaluridae).
Dry Forest Carnivora: families Mustelidae and Felidae, Herpestes genus (Herpestidae);
H. Savanna Carnivora: families Mustelidae, Felidae, Canidae, Hyaenidae and Canis
genus (Canidae);
Cetartiodactyla: family Bovidae, subfamily Reduncinae (Bovidae), Redunca
and Tragelaphus genera (Bovidae);
Hyracoidea: Dendrohyrax genus (Procaviidae).
Savanna Carnivora: families Herpestidae, Mustelidae, Felidae, Canidae, Hyaenidae
and Felis genus (Felidae);
Cetartiodactyla: family Bovidae;
Desert
Carnivora: families Mustelidae, Felidae, Canidae, Hyaenidae, Panthera
(Felidae) and Galerella (Herpestidae) genus;
Lagomorpha: family Leporidae;
Perissodactyla: Equus genus (Equidae);
Rodentia: Paratomys (Muridae) and Gerbillurus (Muridae) genus.
100
among rainforest and the other habitats were always higher than expected by chance and
mainly determined by turnover (PBDTURN), with functional dissimilarities being higher
when rainforest was compared with savanna and desert (Table 5).
Table 5. Pairwise functional beta diversity (FBD) values among African habitats are
shown below diagonal and turnover-resultant component of FBD (FBDTURN) is above
diagonal. Values in bold indicate higher standardized effect size (SES) values than
expected by chance. Rainforest = Humid/Semi-humid forests; Dry Forest = Dry
Forests/shrub; H. Savanna = Humid/semi humid Savanna; Savanna = Savanna/Shrubland
and Desert = Desert/Semi-arid
Habitat Rainforest Dry forest H. Savanna Savanna Desert
Rainforest --- 0.385 0.423 0.544 0.503
Dry forest 0.420 --- 0.372 0.270 0.372
H. Savanna 0.446 0.383 --- 0.269 0.296
Savanna 0.548 0.317 0.304 --- 0.095
Desert 0.623 0.490 0.440 0.320 ---
DISCUSSION
Phylogenetic structure of habitats and assemblages
In general, results at assemblage and habitat level were congruent, supporting a
major role of African biome formation history in determining lineage co-occurrence
(Gerhold et al., 2015). While the significant phylogenetic structure patterns varied
between the analyzed scales, the same tendency of increasing phylogenetic
overdispersion in a dryness gradient was keep in both, assemblage and habitat level
(Figure 2a, c). It seems that historical processes have determined current lineage pool
101
across African habitats (structured along a dryness gradient) leading to contrasting
phylogenetic structures (Cavender-Bares et al., 2009; Lessard et al., 2012b; Gerhold et
al., 2015). The lineage pool of these habitats have been formed along million years,
reflecting the role of large-scale processes (e.g. climatic changes) in diversification and
extinction patterns (Plana, 2004; Hernández Fernández & Vrba, 2005). Low values of
phylogenetic beta diversity between habitats denotes that even very different habitats
share a great amount of evolutionary history (Leprieur et al., 2012). It is in accordance
with our prediction regarding the supposed colonization history of dry open environments
by African forest lineages (deMenocal, 2004). However, we found that most of
dissimilarity recorded was mainly determined by real lineage turnover. Thus, African
habitats share great amount of mammal lineages but have also keep exclusive lineages
(Table 3; see also Table 4).
It has been suggested that habitat phylogenetic structure could be an imprint of
habitat macroevolutionary diversification, which would result in phylogenetic
overdispersion in ancient habitat and clustering in young habitats (Gerhold et al., 2015;
Souza-Neto et al., 2016). Our results are contrary to this statement, since we found
phylogenetic clustering related to rainforest habitats and a tendency towards
overdispersion for savanna assemblages (Figure 2a, c). Rainforests are much older than
current savanna habitats and prevailed in Afrotropical region during the period when
Mammals passed through a great radiation (i.e. the early Tertiary) (Janis, 1993; Maley,
1996). During climatic cycles, cooling of global temperatures has driven aridification and
replacement of rainforest areas by savanna-type woodlands, followed by grassland
savannas (Cerling, 1992; Cerling et al., 1997; Plana, 2004). In accordance with fossil
evidences, current grassland savanna habitats probably did not exist until the late Miocene
(Janis, 1993). Thus, we could expect that appearance of young clades would coincide
102
with the expansion of new environments, in this case the savannas (e.g. Buckley et al.,
2010). However, it is probably that the patterns we have found may represent the imprints
of a complex history of lineage colonization, speciation and extinction that African
habitats have suffered along the evolutionary time.
In the analysis of overabundant taxa, we found a mixture of shared/exclusive and
ancient/derived clades between forested and dry open habitats (Table 4). In fact, in
accordance with the expected given the rainforest history, the lineage pool was related
with overabundance of ancient taxa, such as Manidae (Pangolins), the only extant family
of the Pholidota order (Janis, 1993; Du Toit et al., 2014). We also found overabundance
related to anomalurid rodents (Anomaluridae), for which the diversification is supposed
to have occurred during the Middle Eocene (Janis, 1993). At this time, vegetation was
primarily tropical forest, which explain why they are forest specific (Maley, 1996).
Overabundance was also found to the order Primates, related to both ancient (e.g. lorisids,
with a Late Eocene in situ evolution) and derived families, such as Cercopithecidae (Janis,
1993; Kamilar et al., 2009; Pozzi et al., 2014). On the other hand, lineage pool structure
of desert, savanna, humid savannas and dry forest were associated with overabundance
of Late Oligocene (around 34 MYA) immigrant taxa (Leporidae family), but mainly
related to Felidae, Hyaenidae and Mustelidae families, which are recognized as
immigrants that only arrived in Africa after the Late Miocene (11 to 5 MYA).
Furthermore, Canidae and Equus genus were also overabundant, which fossil records
recognize as Pliocene (from 5 to 1.9 MYA) immigrants (Vrba, 1992; Janis, 1993).
We predicted that forested and dry open habitats would be related with
overabundance of the same higher taxa, evidencing the lineage colonization history of
African biomes. The overabundance patterns for Carnivora showed different families
being represented in rainforest regarding other habitats. However, Bovidae
103
(Cetartiodactyla) was an overabundant family among almost all African habitats. It has
been proposed that in situ evolution accounted for the appearance of bovids during the
Oligocene-Miocene transition (Janis, 1993; Hernández Fernández & Vrba, 2005;
Cantalapiedra et al., 2014). The Tragelaphus genus (Bovidae) was overabundant in
rainforest and humid savanna and belongs to the most basal among the bovid subfamilies
(Bovinae), which have diversified around 22 MYA as consequence of climatic change
events (Janis, 1993; Hernández Fernández & Vrba, 2005). Specific Bovidae clades were
also related to different habitats, such as Cephalophinae subfamily and its Cephalophus
genus (around 10.8 MYA), which are overabundant in rainforest. This subfamily is
composed by species in general adapted to forest habitats (Pérez-Barbería et al., 2001).
Reduncinae and its Redunca genus (around 6.7 MYA) were overabundant in humid
savanna. Reduncinae is a derived subfamily, which diversified in the Middle Miocene, a
period marked by a global cooling event. Interesting, the diversification of Bovidae and
the distribution of its subfamilies among African habitats help us to understand the
influence of climatic changes in the history of habitat formation (Janis, 1993; Hernández
Fernández & Vrba, 2005; Cantalapiedra et al., 2014).
Most hypotheses on the evolution of the African fauna invoke adaptations to arid
environments, following the emergence of grassland savannas after the late Miocene
(Cerling, 1992; Cerling et al., 1997; deMenocal, 2004). Paleoclimatic records suggest that
shifts in African climatic conditions were accompanied by some changes in African
faunal assemblages towards arid-adapted and grazing species (Cristoffer & Peres, 2003;
deMenocal, 2004). In the latest Miocene, bovids became increasingly predominant and
during the Pliocene the bovid fauna was formed by 50% grazer lineages (Janis, 1993;
Cantalapiedra et al., 2014). Small mammals also experienced a shift to more arid-adapted
forms during the Pliocene, with for example, an increase in abundance of gerbilline
104
rodents (overabundant lineage in desert; Table 4) (Janis, 1993). Although the directional
view of bovid dietary evolution mode (browsing to mixed feeding to grazing), it has been
seen that reversions were common along the clade evolution. It means that dry open
habitat adapted lineages could have recolonized forested environments (Pérez-Barbería
et al., 2001; Cantalapiedra et al., 2014). Another example of the complex history of
lineage pool formation among African habitats is the Genetta genus (Viverridae), which
is a transitional group between rainforest and savanna habitats (Janis, 1993; Gaubert et
al., 2004). Such habitat transition was not directional, with basal and derived species both
adapted to savanna and rainforest habitats (Gaubert et al., 2004).
A lot has been discussed about the influence of climatic events in the evolution of
the African fauna, especially regarding the “Pleistocene forest refuge hypothesis”, which
may have kept ancient lineages and contributed to allopatric speciation during periods of
intense rainforest retraction (e.g. Anthony et al., 2007; Johnston & Anthony, 2012;
Jacquet et al., 2014). On the other hand, hypotheses focused in savannas, such as the
“turnover pulses”, claim that changes in average climate were responsible for periods of
increased speciation and species extinction related to the dynamic of expansion and
retraction/disappearance of habitats (see deMenocal, 2004). In fact, there is evidence that
supports that rainforest refuges were determinant in the diversification of forest
dependent groups (e.g. Cercopithecidae and Cephalophinae; see Table 4), while savannas
were also a cradle for some groups, with intense diversification during expansion periods
(Kamilar et al., 2009; Johnston & Anthony, 2012; Cantalapiedra et al., 2014). Moreover,
African mammal community structure (phylogenetic and functional) was found to be
influenced by modern climate, for primate assemblages, and by paleoclimate (mid-
Holocene and LGM) for Carnivora and ungulate assemblages (Rowan et al., 2016). These
differences were attributed to differential ecological flexibility among mammal groups,
105
corroborated by primate dependence of forested habitats and by the ability to transit
among diets and habitats that has been seen for bovid evolutionary history (Cantalapiedra
et al., 2014; Gouveia et al., 2014; Rowan et al., 2016).
Functional structure of habitats and assemblages
Results at assemblage and habitat level were congruent regarding the tendency of
increasing functional clustering towards dry open habitats. However, at the habitat level,
it seems that a strong filter is acting under desert and savanna habitats causing functional
clustering (Figure 2b, d). We corroborated our hypothesis that different ecological
processes have determined current functional trait pool across major African habitats
(structured along a dryness gradient). It is in accordance with findings that variance in
trait composition is highly explained by climate gradients (Lawing et al., 2016). It is
interesting to notice that at assemblage level, the rainforest communities were related with
the highest values of functional structure, while at habitat level the rainforest SES value
was intermediary between desert/savanna and humid savanna/dry forest values (the
highest SES values). Therefore, it demonstrates that species co-occurrence, at assemblage
level, drives the increasing of functional overdispersion regarding the functional pool
available for rainforest habitat, probably owing to competition (see below). Although
rainforest functional structure was not so overdispersed as found to humid savanna and
dry forest, we corroborated our expectation that rainforest functional space would be more
dissimilar from functional space of other habitats than expected by chance and mainly
determined by real trait turnover. It demonstrates the importance of lineage adaptation
process to gain dry open environments (deMenocal, 2004; Cantalapiedra et al., 2014).
We found functional clustering for assemblages recorded in dry open
environments, despite the coexistence of distantly related lineages. Therefore, even
106
different lineages seem to have converged to required traits for these environments (e.g.
Louys et al., 2011). Our results suggest that ecological processes have played a major role
to select species traits in African habitats instead of historical contingency and
evolutionary history (Lawing et al., 2016). Environmental conditions might result in trait
convergence towards common adaptive forms, even among lineages with different
evolutionary origins and histories (Schluter, 1986; Schluter & Ricklefs, 1993; Moretti &
Legg, 2009). That would be even more likely for extreme environments, (Rodríguez et
al., 2006) which are subject to disturbance, such as fire regimes (Moretti & Legg, 2009).
Our results are consistent with findings that arid conditions are strong filters that require
specific adaptations both to plants and animals (Ojeda et al., 1999). Moreover,
environmental stability hypothesis posits that in variable environments, such as deserts
and savannas, species are likely to converge in their use of abundant and stable resources,
thereby leading to a functional clustering in assemblage scale (Wiens, 1977). It might be
illustrated by the evolution of adaptations to grazing dietary habits in bovid, which
allowed them to gain dry open habitats (Cantalapiedra et al., 2014).
We found that rainforest assemblages support more functionally distinct species
compared to assemblages at the end of the aridity gradient spectrum. In stable habitats,
such as rainforests, it is likely that competition would regulate assemblage structure by
increasing functional diversity through niche segregation (Schoener, 1982). Competition
would be more intense in later-successional communities ( i.e. more productive habitats;
Taylor et al., 1990) leading to divergence of competitive traits (Weiher & Keddy, 1995;
Weiher et al., 1998). This agrees with limiting similarity theory, which posits that
coexisting species need to be sufficiently different to avoid competition (Stubbs &
Wilson, 2004). For example, African rainforest ungulate assemblages can stand very
similar number of co-occurring species with savannas, though the African savannas hold
107
many more species than rainforest, respectively 43 and 29 species (McNaughton &
Georgiadis, 1986). Functional overdispersion may also arise from an increase in
environmental heterogeneity and resource diversification. Increased niche availability
will contribute to the coexistence of higher diversity of functional traits when compared
with homogeneous habitats (Adler et al., 2013). In this context, it has been postulated that
a wider range of foraging methods in ancestral bovids (e.g. Cephalophus genus), that
were forest dwellers, was crucial to their occupancy and diversification in grassy habitats
(Pérez-Barbería et al., 2001).
CONCLUSION
We verified imprints of historical and ecological processes determining the
segregation of mammal lineages and functional traits in the major African habitats. We
found opposite results for phylogenetic and functional structure, which highlights the
need to consider the historical processes in assemblage structure analyses (e.g. Parmentier
et al. 2014). We identified the role of strong environmental filters related to arid systems
promoting trait convergence even among distant related lineages. Rainforests seem to
encompass both recent and ancient lineages and we found a mixture of ancient immigrant
lineages and derived taxa contributing to phylogenetic overdispersion patterns in dry open
habitats (Janis, 1993; Moritz et al., 2000; Plana, 2004). Overall, it seems that a complex
history of lineage colonization among habitats, which was different for specific clades,
was responsible for lineage and functional pool formation of African biomes
(Cantalapiedra et al., 2014; Rowan et al., 2016). The differential diversification history
among African mammal groups is related with their ecological flexibility and
consequently ability to lead with climatic changes that have been the main driver of
108
African mammal diversity along the evolutionary time (deMenocal, 2004; Rowan et al.,
2016).
REFERENCES
Adler, P.B., Fajardo, A., Kleinhesselink, A.R. & Kraft, N.J.B. (2013) Trait-based tests of
coexistence mechanisms. Ecology Letters, 16, 1294–1306.
Anthony, N.M., Johnson-Bawe, M., Jeffery, K., Clifford, S.L., Abernethy, K. a, Tutin,
C.E., Lahm, S. a, White, L.J.T., Utley, J.F., Wickings, E.J. & Bruford, M.W. (2007)
The role of Pleistocene refugia and rivers in shaping gorilla genetic diversity in
central Africa. Proceedings of the National Academy of Sciences USA, 104, 20432–
20436.
Bailey, R.G. (1989) Bailey ecoregions of the continents (reprojected) from the World
Conservation Monitoring Center., NOAA National Geophysical Data Center,
Boulder, CO.
Baselga, A. (2010) Partitioning the turnover and nestedness components of beta diversity.
Global Ecology and Biogeography, 19, 134–143.
Baselga, A. (2012) The relationship between species replacement, dissimilarity derived
from nestedness, and nestedness. Global Ecology and Biogeography, 21, 1223–
1232.
Bininda-Emonds, O.R.P., Cardillo, M., Jones, K.E., MacPhee, R.D.E., Beck, R.M.D.,
Grenyer, R., Price, S. a, Vos, R. a, Gittleman, J.L. & Purvis, A. (2007) The delayed
rise of present-day mammals. Nature, 446, 507–12.
Buckley, L.B., Davies, T.J., Ackerly, D.D., Kraft, N.J.B., Harrison, S.P., Anacker, B.L.,
Cornell, H. V, Damschen, E.I., Grytnes, J.-A., Hawkins, B. a, McCain, C.M.,
Stephens, P.R. & Wiens, J.J. (2010) Phylogeny, niche conservatism and the
latitudinal diversity gradient in mammals. Proceedings of The Royal SocietyThe
Royal Society B: Biological Sciences, 277, 2131–2138.
Cantalapiedra, J.L., Fitzjohn, R.G., Kuhn, T.S., Fernández, M.H., DeMiguel, D., Azanza,
B., Morales, J. & Mooers, A.Ø. (2014) Dietary innovations spurred the
diversification of ruminants during the Caenozoic. Proceedings of the Royal Society
B: Biological Sciences, 281, 20132746.
Carstensen, D.W., Lessard, J.P., Holt, B.G., Krabbe Borregaard, M. & Rahbek, C. (2013)
Introducing the biogeographic species pool. Ecography, 36, 1310–1318.
Cavender-Bares, J., Kozak, K.H., Fine, P.V. a & Kembel, S.W. (2009) The merging of
community ecology and phylogenetic biology. Ecology letters, 12, 693–715.
Cerling, T.E. (1992) Development of grasslands and savannas in East-Africa during the
Neogene. Global and Planetary Change, 97, 241–247.
Cerling, T.E., Harris, J.M., Macfadden, B.J., Leakey, M.G., Quadek, J., Eisenmann, V.
& Ehleringer, J.R. (1997) Global vegetation change through the Miocene/Pliocene
boundary. Nature, 389, 153–158.
Cianciaruso, M. V., Silva, I. a., Batalha, M. a., Gaston, K.J. & Petchey, O.L. (2012) The
influence of fire on phylogenetic and functional structure of woody savannas:
Moving from species to individuals. Perspectives in Plant Ecology, Evolution and
Systematics, 14, 205–216.
Cody, M.L. & Diamond, J.M. (1975) Ecology and evolution of communities, The Belknap
109
Press of Harvard University Press, Cambridge, MA.
Connor, E.F.. & Simberloff, D. (1979) The assembly of species communities : chance or
competition? Ecology, 60, 1132–1140.
Crisp, M.D., Arroyo, M.T.K., Cook, L.G., Gandolfo, M. a, Jordan, G.J., McGlone, M.S.,
Weston, P.H., Westoby, M., Wilf, P. & Linder, H.P. (2009) Phylogenetic biome
conservatism on a global scale. Nature, 458, 754–6.
Cristoffer, C. & Peres, C.A. (2003) Elephants versus butterflies: the ecological role of
large herbivores in the evolutionary history of two tropical worlds. Journal of
Biogeography, 30, 1357–1380.
deMenocal, P.B. (2004) African climate change and faunal evolution during the Pliocene-
Pleistocene. Earth and Planetary Science Letters, 220, 3–24.
Faith, D.P. (1992) Conservation evaluation and phylogenetic diversity. Biological
Conservation, 61, 1–10.
Fritz, S.A., Bininda-Emonds, O.R.P. & Purvis, A. (2009) Geographical variation in
predictors of mammalian extinction risk: big is bad, but only in the tropics. Ecology
Letters, 12, 538–549.
Galster, S., Burgess, N.D., Fjeldså, J., Hansen, L.A. & Rahbek, C. (2007) One degree
resolution databases of the distribution of mammals in Sub-Saharan Africa.
Zoological Museum, University of Copenhagen, On-line data source, version 1.0.
Gaubert, P., Fernandes, C.A., Bruford, M.W. & Veron, G. (2004) Genets (Carnivora,
Viverridae) in Africa: an evolutionary synthesis based on cytochrome b sequences
and morphological characters. Biological Journal of the Linnean Society, 81, 589–
610.
Gerhold, P., Cahill, J.F., Winter, M., Bartish, I. V & Prinzing, A. (2015) Phylogenetic
patterns are not proxies of community assembly mechanisms (they are far better).
Functional Ecology, 29, 600–614.
Gotelli, N.J. & Entsminger, G.L. (2003) Swap algorithms in null model analysis. Ecology,
84, 532–535.
Gotelli, N.J. & McCabe, D.J. (2002) Species co-occurrence: a meta-analysis of J. M.
Diamond’s assembly rules model. Ecology, 83, 2091–2096.
Gouveia, S.F., Villalobos, F., Dobrovolski, R., Beltrão-Mendes, R. & Ferrari, S.F. (2014)
Forest structure drives global diversity of primates. Journal of Animal Ecology, 83,
1523–1530.
Graham, C.H. & Fine, P.V.A. (2008) Phylogenetic beta diversity : linking ecological and
evolutionary processes across space in time. Ecology Letters, 11, 1265–1277.
Graves, G.R. & Rahbek, C. (2005) Source pool geometry and the assembly of continental
avifaunas. Proceedings of the National Academy of Sciences USA, 102, 7871–7876.
Hernández Fernández, M. & Vrba, E.S. (2005) A complete estimate of the phylogenetic
relationships in Ruminantia: a dated species-level supertree of the extant ruminants.
Biological reviews of the Cambridge Philosophical Society, 80, 269–302.
Hidasi-Neto, J., Loyola, R. & Cianciaruso, M. V. (2015) Global and local evolutionary
and ecological distinctiveness of terrestrial mammals: identifying priorities across
scales. Diversity and Distributions, 21, 548–559.
Hubert, N., Paradis, E., Bruggemann, H. & Planes, S. (2011) Community assembly and
diversification in Indo-Pacific coral reef fishes. Ecology and Evolution, 1, 229–277.
Jacquet, F., Nicolas, V., Colyn, M., Kadjo, B., Hutterer, R., Decher, J., Akpatou, B.,
Cruaud, C. & Denys, C. (2014) Forest refugia and riverine barriers promote
diversification in the West African pygmy shrew (Crocidura obscurior complex,
Soricomorpha). Zoologica Scripta, 43, 131–148.
110
Janis, C.M. (1993) Tertiary mammal evolution in the context of changing climates,
vegetation, and tectonic events. Annual Review of Ecology and Systematics, 24, 467–
500.
Johnston, A.R. & Anthony, N.M. (2012) A multi-locus species phylogeny of African
forest duikers in the subfamily Cephalophinae: evidence for a recent radiation in the
Pleistocene. BMC Evolutionary Biology, 12, 120.
Jones, K.E., Bielby, J., Cardillo, M., Fritz, S. a., O’Dell, J., Orme, C.D.L., Safi, K.,
Sechrest, W., Boakes, E.H., Carbone, C., Connolly, C., Cutts, M.J., Foster, J.K.,
Grenyer, R., Habib, M., Plaster, C. a., Price, S. a., Rigby, E. a., Rist, J., Teacher, A.,
Bininda-Emonds, O.R.P., Gittleman, J.L., Mace, G.M. & Purvis, A. (2009)
PanTHERIA: a species-level database of life history, ecology, and geography of
extant and recently extinct mammals. Ecology, 90, 2648.
Kamilar, J.M., Martin, S.K. & Tosi, A.J. (2009) Combining biogeographic and
phylogenetic data to examine primate speciation: an example using cercopithecin
monkeys. Biotropica, 41, 514–519.
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 (Oxford, England), 26, 1463–1464.
Kissling, W.D., Eiserhardt, W.L., Baker, W.J., Borchsenius, F., Couvreur, T.L.P.,
Balslev, H. & Svenning, J.-C. (2012) Cenozoic imprints on the phylogenetic
structure of palm species assemblages worldwide. Proceedings of the National
Academy of Sciences USA, 109, 7379–7384.
Lawes, M.J., Eeley, H.A.C., Findlay, N.J. & Forbes, D. (2007) Resilient forest faunal
communities in South Africa: a legacy of palaeoclimatic change and extinction
filtering? Journal of Biogeography, 34, 1246–1264.
Lawing, A.M., Eronen, J.T., Blois, J.L., Graham, C.H. & Polly, P.D. (2016) Community
functional trait composition at the continental scale: the effects of non-ecological
processes. Ecography, 39, 1–13.
Leprieur, F., Albouy, C., Bortoli, J. De, Cowman, P.F., Bellwood, D.R. & Mouillot, D.
(2012) Quantifying phylogenetic beta diversity : distinguishing between “ true ”
turnover of lineages and phylogenetic diversity gradients. PloS one, 7, e42760.
Lessard, J.-P., Belmaker, J., Myers, J. a, Chase, J.M. & Rahbek, C. (2012a) Inferring local
ecological processes amid species pool influences. Trends in ecology & evolution,
27, 600–607.
Lessard, J.-P., Borregaard, M.K., Fordyce, J. a, Rahbek, C., Weiser, M.D., Dunn, R.R. &
Sanders, N.J. (2012b) Strong influence of regional species pools on continent-wide
structuring of local communities. Proceedings of the Royal Society B: Biological
Sciences, 279, 266–74.
Louys, J., Meloro, C., Elton, S., Ditchfield, P. & Bishop, L.C. (2011) Mammal
community structure correlates with arboreal heterogeneity in faunally and
geographically diverse habitats: Implications for community convergence. Global
Ecology and Biogeography, 20, 717–729.
MacArthur, R.H. & Wilson, E.O. (1967) Theory of Island Biogeography, Princeton
University Press, Princeton, NJ.
Maley, J. (1996) The African rain forest - main characteristics of changes in vegetation
and and climate from the Upper Cretaceous to the Quaternary. Proceedings of The
Royal Society of Edinburgh, 104, 31–73.
McNaughton, S.J. & Georgiadis, N.J. (1986) Ecology of African grazing and browsing
mammals. Annual Review of Ecology and Systematics, 17, 39–66.
111
Moretti, M. & Legg, C. (2009) Combining plant and animal traits to assess community
functional responses to disturbance. Ecography, 32, 299–309.
Moritz, C., Patton, J.L., Schneider, C.J. & Smith, T.B. (2000) DIVERSIFICATION OF
RAINFOREST FAUNAS : An Integrated Molecular Approach. Annual Review of
Ecology and Systematics, 31, 553–563.
Ojeda, R.A., Borghi, C.E., Diaz, G.B., Giannoni, S.M., Mares, M. a. & Braun, J.K. (1999)
Evolutionary convergence of the highly adapted desert rodent Tympanoctomys
barrerae (Octodontidae). Journal of Arid Environments, 41, 443–452.
Parmentier, I., Réjou-Méchain, M., Chave, J., Vleminckx, J., Thomas, D.W., Kenfack,
D., Chuyong, G.B. & Hardy, O.J. (2014) Prevalence of phylogenetic clustering at
multiple scales in an African rain forest tree community. Journal of Ecology, 102,
1008–1016.
Peixoto, F.P., Braga, P.H.P., Cianciaruso, M.V., Diniz-Filho, J.A.F. & Brito, D. (2014)
Global patterns of phylogenetic beta diversity components in bats. Journal of
Biogeography, 41, 762–772.
Peixoto, F.P., Villalobos, F., Melo, A.S., Diniz-Filho, J.A.F., Loyola, R., Rangel, T.F. &
Cianciaruso, M. V. (2017) Geographical patterns of phylogenetic beta-diversity
components in terrestrial mammals. Global Ecology and Biogeography.
Pérez-Barbería, F.J., Gordon, I.J. & Nores, C. (2001) Evolutionary transitions among
feeding style and habitats in ungulates. Evol utionary Ecology Research, 3, 221–
230.
Petchey, O.L. & Gaston, K.J. (2002) Functional diversity (FD), species richness and
community composition. Ecology Letters, 5, 402–411.
Plana, V. (2004) Mechanisms and tempo of evolution in the African Guineo-Congolian
rainforest. Philosophical transactions of the Royal Society of London B: Biological
sciences, 359, 1585–1594.
Pozzi, L., Hodgson, J.A., Burrell, A.S., Sterner, K.N., Raaum, R.L. & Disotell, T.R.
(2014) Primate phylogenetic relationships and divergence dates inferred from
complete mitochondrial genomes. Molecular Phylogenetics and Evolution, 75, 165–
183.
Renner, S. (2004) Plant dispersal across the tropical Atlantic by wind and sea currents.
International Journal of Plant Sciences, 165, S23–S33.
Ricklefs, R.E. (1987) Community diversity : relative roles of and regional processes
testing predictions of local-process theories. Science, 235, 167–171.
Ricklefs, R.E. & Schluter, D. (1993) Species Diversity: Regional and Historical
Influences. Species diversity in Ecological Communities: Historical and
Geographical Perspectives (ed. by R.E. Ricklefs) and D. Schluter), pp. 350–363.
Chicago University Press, Chicago.
Rodríguez, J., Hortal, J. & Nieto, M. (2006) An evaluation of the influence of
environment and biogeography on community structure: the case of Holarctic
mammals. Journal of Biogeography, 33, 291–303.
Rowan, J., Kamilar, J.M., Beaudrot, L. & Reed, K.E. (2016) Strong influence of
palaeoclimate on the structure of modern African mammal communities.
Proceedings of the Royal Society B: Biological Sciences, 283, 20161207.
Safi, K., Cianciaruso, M. V, Loyola, R.D., Brito, D., Armour-Marshall, K. & Diniz-Filho,
J.A.F. (2011) Understanding global patterns of mammalian functional and
phylogenetic diversity. Philosophical transactions of the Royal Society of London
B: Biological sciences, 366, 2536–2544.
Schluter, D. (1986) Ecology similarity and covergence of finch communities. Ecology,
112
67, 1073–1085.
Schluter, D. & Ricklefs, R.E. (1993) Convergence and the regional component of
diversity. Species diversity in eco- logical communities: historical and geographical
perspectives (ed. by R.E. Ricklefs) and D. Schluter), pp. 230–240. University of
Chicago Press, Chicago, Illinois, USA.
Schoener, T.W. (1982) The controvesy over interspecific competition. American
Scientist, 70, 586–590.
Souza-Neto, A.C., Cianciaruso, M. V & Collevatti, R.G. (2016) Habitat shifts shaping
the diversity of a biodiversity hotspot through time : insights from the phylogenetic
structure of Caesalpinioideae in the Brazilian Cerrado. Journal of Biogeography, 43,
340–350.
Stephens, P.R. & Wiens, J.J. (2003) Explaining species richness from continents to
communities: the time-for-speciation effect in emydid turtles. The American
naturalist, 161, 112–128.
Stubbs, W.J. & Wilson, J.B. (2004) Evidence for limiting similarity in a sand dune
community. Journal of Ecology, 92, 557–567.
Swenson, N.G., Erickson, D.L., MI, X., Bourg, N.A., Forero-Montana, J., Ge, X., Howe,
R., Lake, J.K., Liu, X., Ma, K., Pei, N., Thompson, J., URIARTE, M., Wolf, A.,
Wright, S.J., Ye, W., Zhang, J., Zimmerman, J.K. & Kress, W.J. (2012) Phylogenetic
and functional alpha and beta diversity in temperate and tropical tree communities.
Ecology, 93, S112–S125.
Taylor, D.R., Aarssen, L.W. & Loehle, C. (1990) On the relationship between r/K
selection and environmental carrying capacity: a new habitat templet for plant life
history strategies. Oikos, 58, 239–250.
Du Toit, Z., Grobler, J.P., Kotzé, A., Jansen, R., Brettschneider, H. & Dalton, D.L. (2014)
The complete mitochondrial genome of Temminck’s ground pangolin (Smutsia
temminckii; Smuts, 1832) and phylogenetic position of the Pholidota (Weber, 1904).
Gene, 551, 49–54.
de Vivo, M. & Carmignotto, A.P. (2004) Holocene vegetation change and the mammal
faunas of South America and Africa. Journal of Biogeography, 31, 943–957.
Vrba, E.S. (1992) Mammals as a key to evolutionary theory. Journal of Mammalogy, 73,
1–28.
Webb, C.O., Ackerly, D.D. & Kembel, S.W. (2008) Phylocom: Software for the analysis
of phylogenetic community structure and trait evolution. Bioinformatics, 24, 2098–
2100.
Webb, C.O., Ackerly, D.D., McPeek, M. a. & Donoghue, M.J. (2002) Phylogenies and
community ecology. Annual Review of Ecology and Systematics, 33, 475–505.
Webb, S.D. (2006) The great american biotic interchange : patterns and processes. Annals
of the Missouri Botanical Garden, 93, 245–257.
Weiher, E., Clarke, P. & Keddy, P. a. (1998) Community assembly rules, morphological
dispersion, of plant species the coexistence. Oikos, 81, 309–322.
Weiher, E. & Keddy, P.A. (1995) Assembly rules, null models, and trait dispersion: new
questions from old pattern. Oikos, 74, 159–164.
Wiens, J. a. (1977) On competition and variable environments: populations may
experience “ecological crunches” in variable climates, nullifying the assumptions of
competition theory and limiting the usefulness of short-term studies of population
patterns. American Scientist, 65, 590–597.
Wiens, J.J. (2011) The causes of species richness patterns across space, time, and clades
and the role of “ecological limits”. The Quarterly review of biology, 86, 75–96.
113
Wiens, J.J. & Donoghue, M.J. (2004) Historical biogeography, ecology and species
richness. Trends in ecology & evolution, 19, 639–644.
Willis, K.J., Bennett, K.D., Burrough, S.L., Macias-Fauria, M. & Tovar, C. (2013)
Determining the response of African biota to climate change : using the past to model
the future. Philosophical transactions of the royal society, 368, 20120491.
Zobel, M. (1997) The relative role os species pools in determining plant species richness
an alternative explanation of species coexistence? Trends in ecology & evolution,
12, 266–269.
114
SUPPORTING INFORMATION
Table S1. References for each mammal the checklists used in the analyses. MAB; Stands
for the Biological Inventories of the World's Protected Areas Database (UNESCO, Man
and the Biosphere Programme, Information Centre for the Environment, University of
California, Davis; available at http://ice.ucdavis.edu/mab/, Accessed 29 August 2005).
ID Site Ref.
1 Basse Casamance National Park (Senegal) Dupuy (1973)
2 Parc National de la Pendjari (Benin) MAB
3 Taï National Park (Ivory Coast) MAB
4 Omo Biosphere Reserve (Nigeria) MAB
5 Dja Faunal Reserve (Cameroon) MAB
6 Parc National d'Odzala (Rep. of Congo) MAB
7 Kibale Forest Corridor National Park (Uganda) MAB
8 Amboseli National Park (Kenya) MAB
9 West Caprivi Game Reserve (Namibia) MAB
10 Kalahari Gemsbok National Park (South Africa) MAB
11 Naute Dam (Namibia) MAB
12 Malolotjia Nature Reserve (Swaziland) MAB
13 Mlawula Nature Reserve (Swaziland) MAB
14 Augrabies Falls National Park (South Africa) Rautenbach et al. (1979)
15 Weenen Game Reserve (South Africa) Bourquin & Matthias (1995)
16 Kogelberg Nature Reserve (South Africa) MAB
Appendix S1 Methods
Several geographic definitions have been used to determine species membership to a
source pool. The simplest definition includes all species recorded in the biogeographic
region where focal assemblage is located (Hortal et al., 2008). Source pool membership
also has been delimited by distances, including only species recorded in a limited area
size (commonly a circle) around focal assemblage (Graves & Gotelli, 1983; Gotelli &
Graves, 1990). Other definitions have added an additional constraint, including only
species related to specific habitats present in the locality (Gotelli & Graves, 1990). In
addition, Carstensen et al. (2013) have proposed three analytical approaches that can be
used to define source pool, what they called distance-based clustering analysis, network
115
modularity analysis, and assemblage dispersion fields. Indeed, a decay of membership
degree with either increasing geographic distance or decreasing biogeographic affinity
should be used (Graves & Rahbek, 2005).
We used eleven definitions of species pool to perform the analyses, based in these
methodologies cited above: the (1) continental, (2) circle (uw), (3) circle (wd), (4) biome
(uw), (5) biome (wd), (6) dispersion field (uw), (7) dispersion field (ws), (8) cluster (uw),
(9) cluster (wd), (10) module (uw) and (11) module (wd).
Continental pool (AP), defined as all the species present in the Paleotropical
biogeographic region (we avoid using the term ‘regional pool’ to prevent possible
mistakes due to its wide usage in the literature). This assumes that all species
inhabiting the region share a common evolutionary history, and therefore provide a
raw estimate of the species that could have colonized any locality within the region.
Circle pool (CP), defined as all the species present in a circle surrounding the locality,
of 24.49º radius, which is the median size of the area inhabited by at least 10% of the
species present in the locality (see Dispersion field pool below); this assumes that the
assembly of the local biota is made from the species occurring nearby.
Biome pool (BP), defined as all the species present in the continuous area of the same
kind of biome where the assemblage is located, according to an ecoregion
classification followed by Rodríguez et al. (2006). This assumes that the local
assembly is made from species inhabiting ecologically similar habitats.
Dispersion field pool (DP), defined as all species present in the area with overlapping
geographic ranges of, at least, 10% of the species that are present in the locality.
Assemblage dispersion fields are supposed to illustrate the geometry of the source
pool of the local assemblages (Graves & Rahbek, 2005). Therefore, this definition
assumes that there is a strong biogeographical component in the local assembly of the
116
species, and that the geometry of the pool can be assessed from the geographic ranges
of the species present in the locality.
Cluster pool (CLP), defined as the species present in each group formed by the
clumping of the cells in relation to their species composition (Carstensen et al., 2013).
We construct a dissimilarity matrix of the cells in relation to their composition, using
the Sørensen index. The chosen cluster algorithm was the UPGMA method
(unweighted pair-group method using arithmetic averages) (Legendre & Legendre,
2012). We explored visually the k values 2 to 10 to know the number of relevant
groups that were formed inside the cluster. We also performed a K-Means Partitioning
that was coincident with the number of six groups that we chose visually.
Modules pool (MP), defined as the species present in the modules formed by the
network modularity analysis of the cells species composition (Carstensen et al.,
2013). The species and sites are arranged as a two-mode (or bipartite) network, with
sites and species sharing a link if the species is present at the site. Sites and species
that are strongly interconnected are then grouped using a modularity analysis. We
used the Louvain algorithm and obtained four groups.
We evaluate two ways of determining the degree of membership of each species to the
source pool:
Unweighted pool (uw), where all species present in the geographic area defined by
the pool have the same probability of being part of the local assemblage.
Weighted pool (wd or ws), where all species have different degrees of membership to
the source pool according to the geographic location of their ranges, and therefore
have different probabilities of establishing populations in the locality during the
process of assembly. We applied one kinds of weight applied to circle, biome, cluster
and module pools. This was based on the minimum geographic distance between the
117
range of each species and the locality (wd). We applied also another weight kind the
dispersion field pool, based in the maximum number of species present in the locality
that can be found in any of the grid cells of the occupied by the species in question
(ws). Probabilities of pertenence to the pool for each species present in the cell where
then assigned following a sigmoid distribution, giving comparatively higher
probability values to the cells placed near to (or sharing more species with) the
locality.
REFERENCES
Bourquin, O. & Matthias, I. (1995) The vertebrates (excluding birds) of Weenen Reserve,
Kwa-Zulu-Natal. Lammergeyer, 43, 43–58.
Carstensen, D.W., Lessard, J.-P., Holt, B.G., Krabbe Borregaard, M. & Rahbek, C. (2013)
Introducing the biogeographic species pool. Ecography, 36, 1310–1318.
Dupuy, F. (1973) Premiere inventaire des mammifkres du Parc National de Basse
Casamance (Senegal). Bulletin de l’lnstittrt Fondamental D’Afrique Noire (Sciences
Naturelles), 35, 186–197.
Gotelli, N.J. & Graves, G.R. (1990) Body size and the occurrence of avian species on
land-bridge islands. Journal of Biogeography, 17, 315–325.
Graves, G.R. & Gotelli, N.J. (1983) Neotropical land-bridge avifaunas: new approaches
to null hypotheses in biogeography. Oikos, 41, 322–333.
Graves, G.R. & Rahbek, C. (2005) Source pool geometry and the assembly of continental
avifaunas. Proceedings of the National Academy of Sciences USA, 102, 7871–7876.
Hortal, J., Rodríguez, J., Nieto-Díaz, M. & Lobo, J.M. (2008) Regional and
environmental effects on the species richness of mammal assemblages. Journal of
Biogeography, 35, 1202–1214.
Legendre, P. & Legendre, L. (2012) Numerical ecology, Elsevier Science, Amsterdam.
Rautenbach, I.L., Schlitter, D.A. & De Graff, G. (1979) Notes on the mammal fauna of
the Augrabies Falls National Park and surrounding areas with special reference to
regional zoogeographical implications. Koedoe, 22, 157–175.
Rodríguez, J., Hortal, J. & Nieto, M. (2006) An evaluation of the influence of
environment and biogeography on community structure: the case of Holarctic
mammals. Journal of Biogeography, 33, 291–303.
118
Tabela S2. Results of the phylogenetic structure analyzes using PD. We demonstrate the
number of communities that the results were random (Rand), overdispersed (Over) or
clustered (Clust) and their kind of habitats. See the species pool names in methods. The
numbers displayed close to the habitats name corresponds to the ID number of each
community (see Fig.1 and Table 1)
Phylogenetic
Geographic pools Rand. Over. Habitat type Clust. Habitat type
Continental (uw) 15 1 Desert/Semi-arid (11) 0 -
Biome (uw) 15 1 Savanna/Shrubland (12) 0 -
Biome (wd) 14 2 Desert/Semi-arid (11) and
Savanna/Shrubland (12) 0 -
Circle (uw) 15 1 Desert/Semi-arid (11) 0 -
Circle (wd) 15 1 Desert/Semi-arid (11) 0 -
Biogeographic pools
Disp. Field (uw) 14 2 Desert/Semi-arid (11) and
Savanna/Shrubland (12) 0 -
Disp. Field (ws) 14 2 Desert/Semi-arid (11) and
Savanna/Shrubland (12) 0 -
Cluster (uw) 15 1 Desert/Semi-arid (11) 0 -
Cluster (wd) 13 3 Desert/Semi-arid and (11)
Savanna/Shrubland (9,12) 0 -
Module (uw) 14 2 Desert/Semi-arid and (11) and Humid/semi
humid Savanna (2) 0 -
Module (wd) 12 4 Desert/Semi-arid (11), Humid/semi humid
Savanna and (2) Savanna/Shrubland (9,12) 0 -
119
Tabela S3. Results of the functional structure analyzes using FD. We demonstrate the
number of communities that the results were random (Rand), overdispersed (Over) or
clustered (Clust) and their kind of habitats. See the species pool names in methods. The
numbers displayed close to the habitats name corresponds to the ID number of each
community (see Fig.1 and Table 1).
Functional
Geographic pools Rand. Over. Habitat type Clust. Habitat type
Continental (uw) 16 0 - 0 -
Biome (uw) 15 0 - 1 Humid/semi humid Savanna (15)
Biome (wd) 14 0 - 2 Humid/semi humid Savanna (15) and
Savanna/Shrubland (13)
Circle (uw) 13 0 - 3 Humid/semi humid Savanna (15) and
Savanna/Shrubland (10, 13)
Circle (wd) 14 0 - 2 Humid/semi humid Savanna (15) and
Savanna/Shrubland (13)
Biogeographic pools
Disp. Field (uw) 15 1 Humid/Semi-
humid forests (5) 0 -
Disp. Field (ws) 15 0 - 1 Savanna/Shrubland (13)
Cluster (uw) 15 0 - 3 Humid/semi humid Savanna (15)
and Savanna/Shrubland (10, 13)
Cluster (wd) 13 0 - 3 Humid/semi humid Savanna (15) and
Savanna/Shrubland (10, 13)
Module (uw) 14 0 - 2 Savanna/Shrubland (10, 13)
Module (wd) 13 0 - 3 Humid/semi humid Savanna (15) and
Savanna/Shrubland (10, 13)
120
Table S4. Standardized effect size for phylogenetic (SES. PD) and functional diversity
(SES.FD) for each habitat. Rainforest = Humid/Semi-humid forests; Dry Forest = Dry
Forests/shrub; H. Savanna = Humid/semi humid Savanna; Savanna = Savanna/Shrubland
and Desert = Desert/Semi-arid.
Habitat SES.PD P SES.FD P
Rainforest -3.175 0.002* -1.450 0.066
Dry Forest -0.805 0.225 -0.333 0.358
H. Savanna -1.104 0.136 -0.295 0.355
Savanna -0.295 0.381 -2.146 0.016*
Desert -0.329 0.375 -2.125 0.014*
*P < 0.025
Table S5. Pairwise correlations between phylogenetic and functional SES values
obtained for all species pool definitions.
Phylogenetic Functional
Species pool comparisons (SES) r2 P r2 P
Continental Biome (uw) 0.819*** <0.001 0.850*** <0.001
Continental Biome (wd) 0.870* 0.011 0.756*** <0.001
Continental Circle (uw) 0.795*** <0.001 0.747*** <0.001
Continental Circle (wd) 0.805*** <0.001 0.721** 0.001
Continental Disp. Field (uw) 0.996*** <0.001 0.852*** <0.001
Continental Disp. Field (ws) 0.947*** <0.001 0.828*** <0.001
Continental Cluster (uw) 0.790*** <0.001 0.801*** <0.001
Continental Cluster (wd) 0.799*** <0.001 0.734** 0.001
Continental Module (uw) 0.800*** <0.001 0.781*** <0.001
Continental Module (wd) 0.822 0.093 0.705** 0.002
Biome (uw) Biome (wd) 0.922*** <0.001 0.831*** <0.001
Biome (uw) Circle (uw) 0.875*** <0.001 0.798*** <0.001
121
Biome (uw) Circle (wd) 0.823*** <0.001 0.791*** <0.001
Biome (uw) Disp. Field (uw) 0.844*** <0.001 0.944*** <0.001
Biome (uw) Disp. Field (ws) 0.880*** <0.001 0.915*** <0.001
Biome (uw) Cluster (uw) 0.837 0.050 0.703** 0.002
Biome (uw) Cluster (wd) 0.799*** <0.001 0.635** 0.008
Biome (uw) Module (uw) 0.853* 0.025 0.669** 0.004
Biome (uw) Module (wd) 0.792*** <0.001 0.607* 0.012
Biome (wd) Circle (uw) 0.947*** <0.001 0.893*** <0.001
Biome (wd) Circle (wd) 0.956*** <0.001 0.910*** <0.001
Biome (wd) Disp. Field (uw) 0.881** 0.006 0.919*** <0.001
Biome (wd) Disp. Field (ws) 0.946*** <0.001 0.944*** <0.001
Biome (wd) Cluster (uw) 0.930*** <0.001 0.751*** <0.001
Biome (wd) Cluster (wd) 0.871* 0.011 0.703** 0.002
Biome (wd) Module (uw) 0.948*** <0.001 0.739** 0.001
Biome (wd) Module (wd) 0.882*** <0.001 0.702** 0.002
Circle (uw) Circle (wd) 0.975*** <0.001 0.976*** <0.001
Circle (uw) Disp. Field (uw) 0.820 0.097 0.872*** <0.001
Circle (uw) Disp. Field (ws) 0.937*** <0.001 0.942*** <0.001
Circle (uw) Cluster (uw) 0.979*** <0.001 0.893*** <0.001
Circle (uw) Cluster (wd) 0.919*** <0.001 0.864*** <0.001
Circle (uw) Module (uw) 0.991*** <0.001 0.888*** <0.001
Circle (uw) Module (wd) 0.917*** <0.001 0.857*** <0.001
Circle (wd) Disp. Field (uw) 0.818*** <0.001 0.838*** <0.001
Circle (wd) Disp. Field (ws) 0.933*** <0.001 0.921*** <0.001
Circle (wd) Cluster (uw) 0.961*** <0.001 0.892*** <0.001
Circle (wd) Cluster (wd) 0.884** 0.005 0.860*** <0.001
Circle (wd) Module (uw) 0.989*** <0.001 0.898*** <0.001
Circle (wd) Module (wd) 0.907** 0.001 0.872*** <0.001
Disp. Field (uw) Disp. Field (ws) 0.960*** <0.001 0.977*** <0.001
Disp. Field (uw) Cluster (uw) 0.815*** <0.001 0.718** 0.001
Disp. Field (uw) Cluster (wd) 0.82*** 0.095 0.662** 0.005
Disp. Field (uw) Module (uw) 0.822*** 0.091 0.688** 0.003
122
Disp. Field (uw) Module (wd) 0.836*** 0.053 0.633** 0.008
Disp. Field (uw) Cluster (uw) 0.916*** <0.001 0.809*** <0.001
Disp. Field (uw) Cluster (wd) 0.876** 0.008 0.764*** <0.001
Disp. Field (uw) Module (uw) 0.938*** <0.001 0.787*** <0.001
Disp. Field (uw) Module (wd) 0.895*** <0.001 0.749*** <0.001
Cluster (uw) Cluster (wd) 0.952*** <0.001 0.989*** <0.001
Cluster (uw) Module (uw) 0.979*** <0.001 0.988*** <0.001
Cluster (uw) Module (wd) 0.915*** <0.001 0.971*** <0.001
Cluster (wd) Module (uw) 0.916*** <0.001 0.977*** <0.001
Cluster (wd) Module (wd) 0.966*** <0.001 0.982*** <0.001
Module (uw) Module (wd) 0.927*** <0.001 0.983*** <0.001
*P < 0.05; **P < 0.01; ***P < 0.001
123
CONCLUSÃO GERAL
No presente trabalho, avaliei a atuação de processos ecológicos, evolutivos e
histórico–biogeográficos na determinação de padrões de diversidade de mamíferos
terrestres. A partir da observação de padrões de diversidade de Chiroptera, em uma escala
global, investiguei sobre o processo pelo qual se deu a evolução de nicho ambiental.
Concluí que estase evolutiva pode ter sido importante para algumas famílias do grupo, o
que chama a atenção para a importância da não-estacionariedade na evolução de nicho
ambiental e da avaliação de diferentes escalas filogenéticas para concluir sobre
conservação de nicho. Em uma escala ainda global, investiguei o gradiente de diversidade
beta filogenética. Foram identificadas regiões onde a perda de linhagens (diferença na
diversidade filogenética entre locais) é o que determina o gradiente de dissimilaridade
(transições com desertos e limites entre regiões biogeográficas). Foram também
identificadas regiões onde a troca (dissimilaridade real) de linhagens foi mais importante
para gerar esses padrões de diversidade beta filogenética (elevadas altitudes). Esses
padrões foram diferentes entre três ordens de mamíferos analisadas, demonstrando a
importância da capacidade de dispersão e da capacidade de estabelecimento,
possivelmente condicionada à evolução de nicho. Em escala regional, demonstrei que a
história climática do continente Africano foi determinante na formação do pool de
linhagens e características funcionais entre os biomas, o que se refletiu na estrutura
filogenética e funcional de assembleias. A história de colonização e diversificação
diferencial entre grupos de mamíferos parece estar relacionada com a flexibilidade
ecológica e a habilidade de lidar com mudanças climáticas severas que marcaram esse
continente ao longo do Cenozoico. De forma geral, os resultados dos três capítulos
convergiram no sentido de evidenciar a importância da história evolutiva de diferentes
124
grupos taxonômicos (desde ordem até gênero e espécies) para compreender os padrões
de diversidade atuais.