Preprint

Distance decay 2.0 – a global synthesis of taxonomic and functional turnover in ecological communities

Abstract

Understanding the variation in community composition and species abundances, i.e., β-diversity, is at the heart of community ecology. A common approach to examine β-diversity is to evaluate directional turnover in community composition by measuring the decay in the similarity among pairs of communities along spatial or environmental distances. We provide the first global synthesis of taxonomic and functional distance decay along spatial and environmental distance by analysing 149 datasets comprising different types of organisms and environments. We modelled an exponential distance decay for each dataset using generalized linear models and extracted r ² and slope to analyse the strength and the rate of the decay. We studied whether taxonomic or functional similarity has stronger decay across the spatial and environmental distances. We also unveiled the factors driving the rate of decay across the datasets, including latitude, spatial extent, realm, and organismal features. Taxonomic distance decay was stronger along spatial and environmental distances compared with functional distance decay. The rate of taxonomic spatial distance decay was the fastest in the datasets from mid-latitudes while the rate of functional decay increased with latitude. Overall, datasets covering larger spatial extents showed a lower rate of decay along spatial distances but a higher rate of decay along environmental distances. Marine ecosystems had the slowest rate of decay. This synthesis is an important step towards a more holistic understanding of patterns and drivers of taxonomic and functional β-diversity.
Distance decay 2.0 – a global synthesis of taxonomic and functional turnover in ecological
1
communities
2
Caio Graco-Roza1,2,*, Sonja Aarnio2, Nerea Abrego3, Alicia T. R. Acosta4, Janne Alahuhta5,6, Jan
3
Altman7, Claudia Angiolini8, Jukka Aroviita6, Fabio Attorre9, Lars Baastrup-Spohr10, José Juan
4
Barrera-Alba11, Jonathan Belmaker12,13, Idoia Biurrun14, Gianmaria Bonari15, Helge Bruelheide16,17,
5
Sabina Burrascano18, Marta Carboni4, Pedro Cardoso73, José Carlos Carvalho73,74, Giuseppe
6
Castaldelli19, Morten Christensen20, Gilsineia Correa1, Iwona Dembicz21,22, Jürgen Dengler17,22,23,
7
Jiri Dolezal7,24, Patricia Domingos25, Tibor Erös26, Carlos E. L. Ferreira27, Goffredo Filibeck28,
8
Sergio R. Floeter29, Alan Friedlander30, Johanna Gammal31, Anna Gavioli19, Martin M. Gossner32,
9
Itai Granot12, Riccardo Guarino33, Camilla Gustafsson31, Brian Hayden34, Siwen He35,2, Jacob
10
Heilmann-Clausen36, Jani Heino6, John T. Hunter37, Vera Lucia de Moraes Huszar38, Monika
11
Janišová39, Jenny Jyrkänkallio-Mikkola2, Kimmo Kahilainen40 , Julia Kemppinen2, Łukasz
12
Kozub21, Carla Kruk41,42, Michel Kulbiki43, Anna Kuzemko44,45, Peter Christian le Roux46 , Aleksi
13
Lehikoinen47, Domênica Teixeira de Lima48, Angel Lopes-Urrutia49, Balázs A. Lukács50, Miska
14
Luoto2, Stefano Mammola73,75, Marcelo Manzi Marinho1, Luciana da Silva Menezes51, Marco
15
Milardi52, Marcela Miranda53, Gleyci Aparecida Oliveira Moser48, Joerg Mueller54,55, Pekka
16
Niittynen2, Alf Norkko31,56, Arkadiusz Nowak57,58, Jean Ometto53, Otso Ovaskainen59,60, Gerhard E.
17
Overbeck61, Felipe Siqueira Pacheco53, Virpi Pajunen2, Salza Palpurina62, Félix Picazo63,64, Juan
18
Antonio Campos Prieto14, Ivan F. Rodil31, Francesco Maria Sabatini16,17, Shira Salingré12, Michele
19
de Sanctis9, Angel M. Segura65, Lucia Helena Sampaio da Silva66, Zora Dajic Stevanovic67,
20
Grzegorz Swacha68, Anette Teittinen2, Kimmo T. Tolonen69, Ioannis Tsiripidis70, Leena Virta2, 31,
21
Beixin Wang34, Jianjun Wang64, Wolfgang Weisser71, Yuan Xu72, and Janne Soininen2,*.
22
*equal contribution
23
1Laboratory of Ecology and Physiology of Phytoplankton, Department of Plant Biology, State
24
University of Rio de Janeiro, Rua São Francisco Xavier 524, PHLC, Sala 511a, 20550-900 Rio de
25
Janeiro, Brazil
26
2Department of Geosciences and Geography, P.O. Box 65, FI-00014 University of Helsinki,
27
Helsinki, Finland.
28
3Department of Agricultural Sciences, P.O. Box 27, FI-00014 University of Helsinki, Finland.
29
4Department of Science, University of Roma Tre - Viale Guglielmo Marconi 446, I-00146, Rome,
30
Italy.
31
5Geography Research Unit, University of Oulu, P.O. Box 3000, 90014 Oulu, Finland.
32
6Finnish Environment Institute, Freshwater Centre, P.O. Box 413, 90014 Oulu, Finland.
33
7Institute of Botany of The Czech Academy of Sciences, Zámek 1, 25243 Průhonice, Czech
34
Republic.
35
8Department of Life Sciences, University of Siena, Siena, Italy.
36
9Department of Environmental Biology, University Sapienza of Rome, P.le Aldo Moro 5, 00185,
37
Rome, Italy.
38
10Freshwater Biological Laboratory, Department of Biology, University of Copenhagen,
39
Universitetsparken 4 2100 København Ø, Denmark.
40
11Departamento de Ciências do Mar, Instituto do Mar, Universidade Federal de São Paulo. R.
41
Carvalho de Mendonça, 144. Encruzilhada, zip code 11070-100, Santos (SP), Brazil.
42
12School of Zoology, George S. Wise Faculty of Life Sciences, Tel Aviv University, Tel Aviv,
43
Israel.
44
13Steinhardt Museum of Natural History, Tel Aviv University, Tel Aviv, Israel.
45
14Department of Plant Biology and Ecology, University of the Basque Country UPV/EHU, P.O.
46
Box 644, 48080 Bilbao, Spain.
47
15Faculty of Science and Technology, Free University of Bozen-Bolzano, Piazza Università, 5,
48
39100, Bozen-Bolzano, Italy.
49
16Martin-Luther-Universität Halle-Wittenberg, Institut für Biologie. Am Kirchtor 1, 06108 Halle,
50
Germany.
51
17German Centre for Integrative Biodiversity Research (iDiv) Halle-Jena-Leipzig, Deutscher Platz
52
5e, 04103 Leipzig, Germany.
53
18Department of Environmental Biology, Sapienza University of Rome, Rome, Italy.
54
19University of Ferrara, Department of Life Sciences and Biotechnology, Ferrara, V. Luigi Borsari
55
46, 44121, Italy.
56
20 Hvidtjørnevej 1, DK-4180 Sorø, Denmark.
57
21 Department of Ecology and Environmental Conservation, Institute of Environmental Biology,
58
Faculty of Biology, University of Warsaw, Warsaw, Poland.
59
22 Vegetation Ecology, Institute of Natural Resource Sciences (IUNR), Zurich University of
60
Applied Sciences (ZHAW), Grüentalstr. 14, 8820 Wädenswil, Switzerland.
61
23 Plant Ecology, Bayreuth Center of Ecology and Environmental Research (BayCEER),
62
Universitätsstr. 30, 95447 Bayreuth, Germany.
63
24Faculty of Science, Department of Botany, University of South Bohemia, Na Zlaté Stoce 1, 37005
64
České Budějovice, Czech Republic.
65
25Laboratory of Phycology and Environmental Education, Department of Plant Biology, State
66
University of Rio de Janeiro, Rua São Francisco Xavier 524, PHLC, Sala 525/1, 20550-900 Rio de
67
Janeiro, Brazil.
68
26Balaton Limnological Institute, Centre for Ecological Research, Klebelsberg Kuno u. 3., H-8237
69
Tihany, Hungary.
70
27Universidade Federal Fluminense (UFF), Departamento de Biologia Marinha, Niterói, Brasil.
71
28Department of Agricultural and Forest Sciences (DAFNE), University of Tuscia, 01100 Viterbo,
72
Italy.
73
29Marine Macroecology and Biogeography Lab, Depto. de Ecologia e Zoologia, CCB, Universidade
74
Federal de Santa Catarina, Florianopolis, SC, 88040-900, Brazil.
75
30Department of Biology, University of Hawaii, Honolulu, HI 96822, USA.
76
31Tvärminne Zoological Station, University of Helsinki, J.A. Palménin tie 260, 10900, Hanko,
77
Finland.
78
32Forest Entomology, Swiss Federal Research Institute WSL, Zuercherstrasse 111, Birmensdorf
79
CH-8903, Switzerland.
80
33Department STEBICEF - Botanical Unit, via Archirafi, 38, 90123 Palermo.
81
34Biology Department, Canadian Rivers Institute, University of New Brunswick, Fredericton,
82
Canada.
83
35Department of Entomology, College of Plant Protection, Nanjing Agricultural University,
84
Nanjing, China.
85
36Centre for Macroecology, Evolution and Climate, University of Copenhagen, DK-2100
86
Copenhagen Ø, Denmark.
87
37School of Rural and Environmental Sciences, University of New England, Armidale, NSW.
88
38Phycology Laboratory, Botany Department, National Museum, Federal University of Rio de
89
Janeiro, Brasil.
90
39Institute of Botany, Plant Science and Biodiversity Center, Slovak Academy of Sciences,
91
Ďumbierska 1, 974 11 Banská Bystrica, Slovakia.
92
40Lammi Biological Station, University of Helsinki, Pääjärventie 320, FI-16900 Lammi, Finland.
93
41Sección Limnología, IECA, Facultad de Ciencias, Universidad de la República, Uruguay.
94
42Ecología Funcional de Sistemas Acuáticos, CURE-Rocha, Universidad de la República, Uruguay.
95
43IRD (Institut de Recherche pour le Développement), Laboratoire d’Excellence Labex Corail,
96
UMR IRD-UR-CNRS ENTROPIE, Université de Perpignan, 66860 Perpignan Cedex 9, France.
97
44M.G. Kholodny Institute of Botany, National Academy of Sciences of Ukraine, 2,
98
Tereshchenkivska str. 01601, Kyiv, Ukraine.
99
45Department of Botany and Zoology, Masaryk University, Kotlárská 2, 61137 Brno, Czech
100
Republic.
101
46Department of Plant and Soil Sciences, University of Pretoria, Private Bag X20, Pretoria 0028,
102
South Africa.
103
47Finnish Museum of Natural History, PO Box 17, FI-00014 University of Helsinki, Finland.
104
48Departamento de Oceanografia Biológica (DOB), Faculdade de Oceanografia, Universidade do
105
Estado do Rio de Janeiro (UERJ). São Francisco Xavier St. 524- zip code: 2055090, Maracanã, RJ-
106
Brasil.
107
49Instituto Español de Oceanografía (IEO), Centro Oceanográfico de Gijón, 33212 Gijón, Asturias,
108
Spain.
109
50Wetland Ecology Research Group, Centre for Ecological Research, Bem tér 18/C, Debrecen 4026,
110
Hungary.
111
51Laboratory of Grassland Vegetation, Federal University of Rio Grande do Sul, 9500 Bento
112
Gonçalves Av., Porto Alegre, Brazil.
113
52Fisheries New Zealand, Tini a Tangaroa, Ministry for Primary Industries, Manatū Ahu Matua, 34-
114
38 Bowen Street, Wellington, New Zealand.
115
53National Institute for Space Research, Earth System Science Center, CCST/INPE, Av. dos
116
Astronautas, 1758, Jardim da Granja, São José dos Campos, SP 12227-010, Brazil.
117
54Institute for Biochemistry and Biology, University of Potsdam, Maulbeerallee 1, 14469 Potsdam,
118
Germany.
119
55Department of Nature Conservation, Heinz Sielmann Foundation, Unter den Kiefern 9, 14641
120
Wustermark, Germany.
121
56Baltic Sea Centre, Stockholm University, Stockholm, Sweden.
122
57Polish Academy of Sciences Botanical Garden - Center for Biological Diversity Conservation in
123
Powsin, Prawdziwka 2, 02-973 Warsaw, Poland.
124
58University of Opole, Institute of Biology, Oleska 22, 45-052 Opole, Poland.
125
59Organismal and Evolutionary Biology Research Programme, P.O. Box 65, FI-00014 University of
126
Helsinki, Finland.
127
60Centre for Biodiversity Dynamics, Department of Biology, Norwegian University of Science and
128
Technology, N-7491 Trondheim, Norway.
129
61Department of Botany, Universidade Federal do Rio Grande do Sul, Av. Bento Gonçalves 9500,
130
Porto Alegre, RS 91501-970, Brazil.
131
62National Museum of Natural History, Bulgarian Academy of Sciences, 1 Tsar Osvoboditel Blvd,
132
1000 Sofia, Bulgaria.
133
63Department of Biodiversity and Restoration, Pyrenean Institute of Ecology (IPE-CSIC), Avda.
134
Montañana 1005, 50059 Zaragoza, Spain.
135
64State Key Laboratory of Lake Science and Environment, Nanjing Institute of Geography and
136
Limnology, Chinese Academy of Sciences, Nanjing 210008, China.
137
65Modelización y Análisis de Recursos Naturales, CURE-Rocha, Universidad de la
138
República, Uruguay.
139
66Phycologi Laboratory, Botany Department, National Museum, Federal University of Rio de
140
Janeiro, Brazil.
141
67Department of Agrobotany, Faculty of Agriculture, University of Belgrade, Nemanjina 6, 11080
142
Belgrade-Zemun, Serbia.
143
68Botanical Garden, University of Wrocław, ul. Henryka Sienkiewicza 23, 50-335 Wrocław,
144
Poland.
145
69Department of Biological and Environmental Science, University of Jyväskylä, P.O. Box 35,
146
Jyväskylä, FI-40014 Finland.
147
70Department of Botany, School of Biology, Aristotle University of Thessaloniki, 54124
148
Thessaloniki, Greece.
149
71Terrestrial Ecology Research Group, Department of Ecology and Ecosystem Management,
150
Technical University of Munich, Hans-Carl-von-Carlowitz-Platz 2, 85354 Freising, Germany
151
72State Key Laboratory of Estuarine and Coastal Research, East China Normal University, 200241,
152
Shanghai, China.
153
73Laboratory for Integrative Biodiversity Research (LIBRe), Finnish Museum of Natural History
154
Luomus, University of Helsinki, Finland.
155
74Department of Biology, CBMA—Centre for Molecular and Environmental Biology, University of
156
Minho, Campus de Gualtar, Braga, Portugal.
157
75Molecular Ecology Group (MEG), Water Research Institute (IRSA), National Research Council
158
(CNR), Corso Tonolli, 50, 28922 Pallanza, Italy.
159
*corresponding e-mail: caio.roza@helsinki.fi; janne.soininen@helsinki.fi
160
Abstract
161
Understanding the variation in community composition and species abundances, i.e., β-diversity, is
162
at the heart of community ecology. A common approach to examine β-diversity is to evaluate
163
directional turnover in community composition by measuring the decay in the similarity among pairs
164
of communities along spatial or environmental distances. We provide the first global synthesis of
165
taxonomic and functional distance decay along spatial and environmental distance by analysing 149
166
datasets comprising different types of organisms and environments. We modelled an exponential
167
distance decay for each dataset using generalized linear models and extracted r² and slope to analyse
168
the strength and the rate of the decay. We studied whether taxonomic or functional similarity has
169
stronger decay across the spatial and environmental distances. We also unveiled the factors driving
170
the rate of decay across the datasets, including latitude, spatial extent, realm, and organismal features.
171
Taxonomic distance decay was stronger along spatial and environmental distances compared with
172
functional distance decay. The rate of taxonomic spatial distance decay was the fastest in the datasets
173
from mid-latitudes while the rate of functional decay increased with latitude. Overall, datasets
174
covering larger spatial extents showed a lower rate of decay along spatial distances but a higher rate
175
of decay along environmental distances. Marine ecosystems had the slowest rate of decay. This
176
synthesis is an important step towards a more holistic understanding of patterns and drivers of
177
taxonomic and functional β-diversity.
178
Introduction
179
Biodiversity on Earth is shrinking1. Understanding its distribution is therefore paramount to inform
180
conservation efforts, and to evaluate the links between biodiversity, ecosystem functioning,
181
ecosystem services and human well-being2,3. The variation in the occurrence and abundance of
182
species in space and time, i.e., β-diversity, is at the heart of community ecology and biogeography as
183
it provides a direct link between local (α) and regional (γ) diversity4,5. Moreover, β-diversity has
184
become an essential currency in spatial6,7 and temporal8 comparisons of biodiversity patterns and their
185
underlying drivers. β-diversity is also informative in the context of biodiversity conservation and
186
practical management decisions in rapidly changing environments9,10.
187
A common approach to examine spatial β-diversity is to consider directional turnover in community
188
composition with distance, i.e., distance decay 4,11. The similarity among the pairs of biological
189
communities typically decreases (“decays”) with increasing spatial or environmental distance 11,12.
190
This pattern stems mainly from dispersal limitation (related to physical barriers and dispersal
191
ability13) and species-specific responses to spatially structured environmental variation (related to
192
environmental filters and evolutionary processes14) and is well-documented in observational1517 and
193
theoretical studies18 as well as meta-analyses19. Such studies offer interesting insights into the patterns
194
and drivers of spatial taxonomic β-diversity and often provide information about the effects of
195
environmental changes on ecosystem processes and associated functionality. Even if the patterns and
196
drivers of taxonomic β-diversity are relatively well-documented in the biogeographic literature, it is
197
much less understood whether the same patterns occur for functional β-diversity2022. Therefore,
198
functional biogeography emerges as a field to solve questions related to the distribution of forms and
199
functions of individuals, populations, communities, ecosystems, and biomes across spatial scales23.
200
Understanding functional diversity relies on trait-based approaches, which are built on the idea that
201
the environment selects species based on their ecological requirements, and that functional traits
202
capture these requirements better than species identity24. Thus, a trait-based approach should reflect
203
the functional response of biotic communities to environmental gradients better than an approach
204
based on species’ taxonomic identities only, and better predict how biotic communities respond to
205
environmental changes25. Even if functional diversity has been investigated widely at the α-diversity
206
level26,27, our understanding of functional β-diversity is much more limited and fragmented28–32.
207
Comparing the patterns of functional and taxonomic β-diversity across different biotic groups,
208
ecosystems and geographic contexts has the potential to greatly contribute to a better mechanistic
209
understanding of the drivers behind the spatial variation in ecosystem functionality and shed further
210
light on how environmental change may affect ecological communities.
211
Niche filtering along environmental gradients induces coupling of taxonomic and functional diversity
212
patterns because dominant functional strategies dictate along the environmental gradient33,34.
213
However, high taxonomic β-diversity does not necessarily mean high functional β-diversity25,35 (Fig.
214
1a), and the gain or loss of species does not inform about variations in functional β-diversity whenever
215
trait redundancy is high36. For example, taxonomic homogenization does not lead to functional
216
homogenization if the newly introduced species in the assemblages are functionally similar to each
217
other30,37,38. The most pressing question is whether functional features explain more of the distance
218
decay along environmental gradients than species identities, as suggested by some earlier studies39
219
43.
220
Hypotheses
221
Since the emergence of the first comprehensive distance decay meta-analysis19, our understanding of
222
community turnover along spatial and environmental gradients has increased notably. Here, based on
223
existing ecological literature and theory, and as an initial step towards synthesising knowledge, we
224
tested four hypotheses concerning the differences between taxonomic and functional distance decay
225
along the spatial and environmental distances. The master hypothesis is that the distance decay along
226
spatial gradients is stronger for taxonomic similarity than for functional similarity (H1a). This is
227
because spatial factors relate with taxonomic more than functional composition as a result of dispersal
228
processes, dispersal history and speciation42. Such a hypothesis should be valid when functional traits
229
do not comprise dispersal related traits. In contrast, distance decay along environmental gradients is
230
stronger for functional similarity than for taxonomic similarity because functional composition
231
should respond more strongly to environmental variation27,39,40,42 (H1b) (Fig. 1b).
232
Latitudinal gradients
233
We also generalize the effects of major geographic and environmental factors in the three hypotheses,
234
which are tested across the datasets. For example, latitudinal effect has been recognized as a relevant
235
factor in meta-analyses44 and case studies45,46, and these studies suggest that β-diversity should
236
decrease with increasing latitude (Fig. 1c). This is indicated by the faster latitudinal decline in γ-
237
diversity than in α-diversity47,48, and the slopes of the species-area relationships (proxy for turnover)
238
decrease with latitude49. Moreover, Rapoport’s rule50 postulates that species range sizes are larger at
239
high latitudes leading to lower β-diversity. Therefore, we hypothesize that the rate of taxonomic
240
distance decay along spatial gradients is generally slower in the datasets that originate from higher
241
latitudes (H2a). In contrast, functional distance decay may show faster rates in the datasets from higher
242
latitudes. This is because the high diversity of tropical areas stems mainly from niche overlap51, which
243
increases the functional redundancy within communities and reduces the functional turnover52.
244
Regarding the environmental gradients, large-scale environmental heterogeneity tends to increase
245
towards poles19,53,54, leading to a faster rate of functional distance decay along environmental
246
gradients at higher latitudes (H2b). An alternative hypothesis is that extreme climatic conditions at
247
high latitudes decrease functional diversity because abiotic filtering limits the number of possible
248
ecological strategies found in a biotic community55,56, resulting in relatively slow rate of functional
249
distance decay.
250
Spatial extent
251
Distance decay is also likely to be affected by the spatial extent of a given study57. It has been shown
252
that distance decay has a power-law shape at spatial extents that do not exceed regional species pools
253
and exponential shape when extent encompasses multiple species pools12. This suggests that the slope
254
of the relationship becomes flatter with increasing spatial extent11,19, mainly because regional species
255
diversity is limited with a certain upper boundary58. Furthermore, environmental heterogeneity affects
256
the diversity of species59 and functional traits at regional level60,61, but such effects are likely to be
257
scale-dependent6264. To summarize, we hypothesize that the rate of distance decay along spatial
258
gradients is generally slower in the datasets covering larger spatial extent (H3a). In contrast, we
259
hypothesize that the rate of distance decay along environmental gradients is generally faster when
260
spatial extent is larger, especially for functional similarities, which are considered more sensitive to
261
environmental variation (H3b).
262
Realms
263
We also expect that the patterns of distance decay vary among the realms. In general, marine
264
ecosystems are environmentally more homogeneous than terrestrial or freshwater ecosystems, at least
265
in the open ocean65, and typically show weaker dispersal barriers than terrestrial or freshwater
266
ecosystems66. Therefore, we hypothesize that the datasets from marine ecosystems have generally
267
slower rate of taxonomic and functional distance decay than the other ecosystems (H4).
268
Here, we tested these hypotheses using datasets that cover a wide range of biotic groups from
269
unicellular diatoms to vascular plants, fungi, invertebrates, fish, birds, amphibians and mammals, and
270
that originate from marine, terrestrial and freshwater ecosystems spanning broad latitudinal gradients
271
(Fig. 2). To account for major biological differences in biotic groups, we also investigated if distance
272
decay varied among different sized taxa or among taxa with different dispersal mode67,68. By using
273
such a comprehensive, multi-realm and multi-taxon dataset, we will explore patterns at more general
274
level, compared with case studies that have examined both taxonomic and functional β-diversity, but
275
only considered a single or few biotic groups.
276
Material and methods
277
Data collection. We gathered our data by directly contacting data owners or using the existing data
278
sources, such as sPlot69 and CESTES70. We included datasets that provided raw data of species
279
abundances, functional traits, environmental variables and spatial coordinates of the study sites. A
280
few datasets (n = 6) provided only species occurrence rather than abundance information (Appendix
281
S1). The traits included in the datasets were chosen by data owners from a suite of traits that should
282
respond well to environmental variation. For plant datasets compiled from the sPlot database, trait
283
information was commonly derived from the TRY database71. Regarding the CESTES database, we
284
compiled 48 datasets, specifically from: fish communities22,7274, terrestrial vascular plants7586,
285
aquatic macroinvertebrates8789, terrestrial arthropods86,9098, birds83,90,99102, bats102,103, bryophytes85,
286
butterflies98,104, corals105, and foraminifera106. We only included datasets with at least ten sites, two
287
environmental variables and three traits or trait categories. In some cases, more than one dataset
288
representing different taxonomic groups with different responses to environment and dispersal
289
abilities (e.g., stream macroinvertebrates and diatoms) were collected in the same study area. In total,
290
149 datasets representing 17 major biotic groups from terrestrial (n = 87), freshwater (n = 41) and
291
marine (n = 21) environments were assembled amounting to over 17,000 study sites around the globe
292
(Fig. 2). From the 149 datasets, 118 were published in peer reviewed journals (Appendix S1).Taxa
293
were mostly identified to species or morphospecies level but, in a few cases, we used data at genus
294
level if existing taxonomic knowledge did not allow distinguishing individual species. Finally, each
295
dataset included (i) a sites-by-species abundances matrix, (ii) a species-by-traits table, (iii) a sites-by-
296
spatial coordinates table, and (iv) a sites-by-environmental variables table (Fig. 3a). Detailed
297
information about collected datasets can be found in Appendix S1.
298
Data curation. For each dataset, we removed the sites with less than two observed species, and the
299
species with lower than three traits considered. Trait data included ordered, categorical and
300
continuous traits, the latter of which were log transformed (Log10) when needed. Environmental
301
variables were log-transformed (Log10) to approximate normality (except for e.g., temperature, pH
302
and variables given as eigenvectors), and the environmental variables showing strong inter-
303
correlations (pairwise rp < 0.7) were excluded from further analyses107. Spatial coordinates were
304
converted to the World Geodetic System 1984 (WGS84) datum and geographic coordinate system
305
and expressed in decimal degrees with an accuracy up to five decimals. All the data curation and
306
further analyses were performed in the software R v.4.0.2 (ref.108) using the appropriate R packages.
307
We will consistently refer to the functions used and their respective packages from here on.
308
Taxonomic and functional similarities. Pairwise between-site taxonomic and functional similarities
309
were obtained for each dataset following the tree-based approach implemented in the function beta
310
in the package ‘BAT’ v.2.1.0 (ref.109). We used the tree-based approach because it provides an
311
unequivocal comparison of taxonomic and functional similarities110. Community similarity (S) ranges
312
between zero and one and
!
is commonly calculated for the pairs of communities as the sum of the
313
unique features of each community over the sum of the shared features between communities and the
314
unique features of each community. In the tree-based approach, these features are edges, which may
315
have different lengths and be shared by different species that may be present in different
316
communities110. Taxonomic and functional similarities were calculated for species occurrences and
317
abundances based on a Podani family of Sørensen-based indices111. Here, we estimated S between
318
communities j and k as
!𝑆
!" = 1 − #$%
&'$#$%
(1), where a is the sum of the length of the edges shared
319
between the communities j and k, b is the sum of the length of the edges unique to the community j,
320
and c is the sum of the length of the edges unique to the community k.
321
When estimating taxonomic similarities, each species is a unique entity that share no edges with
322
others and, therefore, all the edges of the tree have same length (Fig. 3b). Thus, the sum of the length
323
of the edges equals the sum of the number of the observed species. For functional similarities, the
324
length of the edge shared between two species depends on how similar species are with respect to
325
their traits. To estimate the length of the edges shared by species, we first construct a global (i.e.
326
considering all the species within the dataset) matrix of species similarities by applying the Gower
327
similarity index112 to the species-by-traits table using the function gowdis of the package ‘FD’ v.1.0
328
(ref.113,114). We used a modified version of the Gower index extended to accommodate variables in
329
ordinal scales115. Using the species similarity matrix, we built a global tree of species similarities
330
based on an unweighted pair group method with arithmetic mean (UPGMA) hierarchical cluster using
331
function hclust of package ‘stats’ v.4.0.2 (ref.108). The length of the edge shared by two species was
332
estimated as the distance between the intersection of two species in the global tree to the root of the
333
tree (Fig. 3b). Based on the length of those edges, functional similarities between the pairs of
334
communities were estimated using the equation 1. Therefore, even if two communities do not share
335
any species, taxonomic similarity would be lower than functional similarity in case of the comparison
336
of a continuous functional trait (e.g., body size; Fig. 3b). Note that the calculation of similarities was
337
carried out within each dataset separately. Details of the calculation of similarities using the Sørensen-
338
based indices for occurrence and abundance (i.e., percentage differences index) data can be found in
339
the Appendix S2. We used both occurrence and abundance data because occurrences should be very
340
informative about the drivers and patterns of communities along geographic gradients while
341
abundances should inform well patterns along environmental gradients116. Main results are given for
342
occurrence data in the main text, and abundance-based results can be found in Appendix S3.
343
Spatial and environmental distances. We estimated the spatial and environmental distances
344
between all the pairs of sites separately for each dataset. Spatial distances within each dataset were
345
calculated as the geographic distance in kilometres between the pairs of sites using the function
346
earth.dist of the package ‘fossil’ v.0.4.0 (ref.117; Fig. 3b). To estimate environmental distances, we
347
first standardized the environmental variables to µ = 0 and σ = 1. Then, we calculated the
348
environmental distance between sites as the Euclidean distance using the measured and standardized
349
environmental variables for all the pairs of sites within each dataset (Fig. 3b) using the function
350
vegdist of the package ‘vegan’ v.2.5-6 (ref.118). Because the datasets comprised different number and
351
types of environmental variables, the values of environmental distance were context-dependent and
352
not very informative for comparison across datasets. We therefore assumed that the environmental
353
gradient scaled positively with spatial extent and rescaled the actual environmental distance to range
354
between zero and one in each dataset by dividing actual values by the average environmental distance
355
of the dataset.
356
Distance decay of similarity. We modelled the distance decay of similarity following a negative
357
exponential curve between the community similarity and distance12. This is because maximum spatial
358
distances within our datasets were on average 795.5 kilometres; 95% CI [506.08, 1084.95], and
359
therefore, it is highly likely that many of the datasets encompassed multiple species pools. One of the
360
main assumptions of the distance decay is that Sij > Sjk if the distance between the sites i and j is
361
shorter than the distance between j and k12. That is, the slope of the relationship should be negative,
362
and positive slopes suggest either periodicity in the environmental gradient or a mismatch between
363
the communities and the measured environmental variables11. Here, we calculated distance decay
364
separately for taxonomic and functional similarities along spatial and environmental distance using a
365
generalized linear model (GLM) following a binomial distribution of errors with a log link119 (Fig.
366
3c). Following Latombe et al.120, we included a negative constraint in GLMs such that the slopes are
367
forced to be negative (i.e., slope <= 0). Besides, we included a negative constraint to the intercept of
368
the model such that intercept <= 0. Therefore, because e0 = 1, we avoided intercept values that fall
369
outside the range of taxonomic and functional similarities. We forced the negative coefficients via a
370
non-positive least-square regression121,122 within the iterative re-weighted least-square algorithm123
371
implemented in the function glm.cons of the package ‘zetadiv’ v.1.2.0 (ref.120,124). We estimated a
372
pseudo-R² (hereafter r²) as
𝑟&= 1 − ()*+,-*+./'0%+
()*+,-01,,2*+./'0%+ !(2)
. Because of the pairwise structure of the
373
data, similarities are non-independent, so we performed a leave-one-out Jack knife procedure to
374
obtain the mean and confidence interval of the intercepts and slopes for each model119. Within such
375
framework, the slope represents the rate of decay, that is, the proportion of similarity loss per unit
376
distance, and the r² represents the strength of the relationship between similarity and distance.
377
Although it can be argued that slopes and r² are highly correlated, the correlation between slopes and
378
r² in this study was small (Pearson’s r = 0.10; p-value = 0.240).
379
Statistical analysis. We tested our hypothesis using two different approaches. Firstly, we
380
investigated whether taxonomic or functional distance decay is stronger along spatial and
381
environmental distances (H1) by performing a pairwise t-test to compare r² drawn from GLMs using
382
taxonomic similarity and the GLMs using functional similarity for each dataset (Fig. 3d). Totally, we
383
carried out two pairwise t-tests, one considering the r² from the models using spatial distances, and a
384
second considering the r² from the models using environmental distances.
385
We also investigated the ecological and geographical factors driving the rate of the distance decay
386
across datasets. Each dataset was characterized with respect to (i) latitude, recorded as the absolute
387
mean value of all the sites of the dataset; (ii) spatial extent, expressed as the largest pairwise distance
388
(in km) between study sites; (iii) realm, classified into freshwater, marine and terrestrial
389
environments; (iv) body size, estimated at organism-level as the log transformed fresh weight (g)
390
drawn from literature47,125; (v) dispersal mode, classified as active and passive modes and organisms
391
dispersed by seeds; (vi) taxonomic γ-diversity expressed as the total number of species in the dataset;
392
(vii) functional γ-diversity, measured as the total volume of the union of the n-dimensional
393
hypervolumes estimated within the dataset; (viii) total number of study sites in the dataset and (ix)
394
the number of environmental variables in the dataset. For body sizes, we note that although the size
395
range within the biotic group may be large (up to five orders of magnitude), it is small compared to
396
the overall variation obtained across organism groups (twelve orders of magnitude). For more details
397
on body size approximations, see refs.47,49. The taxonomic γ-diversity was included to study if there
398
is a typical positive relationship between γ-diversity (taxonomic and functional) and β-diversity7,52.
399
Functional γ-diversity was estimated based on geometrical n-dimensional hypervolumes126,127. We
400
used the species functional similarity matrix based on Gower’s index (see the ‘taxonomic and
401
functional similarities’ section) to extract orthogonal synthetic trait axes through a principal
402
coordinate analysis128. Then, the hypervolume of each site within the dataset was calculated using a
403
gaussian kernel density estimate via the function kernel.alpha of the package ‘BAT’129. The
404
hypervolume of all sites were sequentially merged using the function hypervolume_set of the package
405
‘hypervolume’ v.2.0.12 (ref.130), and the united-hypervolume was used to estimate the total amount
406
of functional space occupied by all the species within the dataset using the function get_volume of
407
the package ‘hypervolume’. Because trait dimensionality affects the accuracy of the functional
408
separation of species131,132, we standardized the number of dimensions to seven synthetic traits axes
409
for all datasets. Hypervolumes are expressed in units of SDs to the power of the number of trait
410
dimensions used (i.e., seven). The number of study sites and the number of environmental variables
411
for each dataset were included to explore their potential effect on distance decay.
412
Finally, we used boosted regression trees (BRT) to test the effects of latitude (H2), spatial extent (H3)
413
and realm (H4) on the rate of taxonomic and functional distance decay along spatial and
414
environmental distance across the datasets. In addition, we included dispersal mode, body size,
415
taxonomic and functional γ-diversity, number of sites, and number of environmental variables in the
416
dataset as predictors in the BRTs (Fig. 3d). BRT is a regression modelling technique able to fits
417
nonlinear relationships between predictor and response variables, including interaction among
418
variables by using a boosting strategy to combine results from a large number (usually thousands) of
419
simple regression tree models133. Our BRT outputs included graphs of the shapes of relationships
420
between predictors and the response variable (e.g., linear, curvilinear and sigmoidal response shapes)
421
and a relative importance of predictor variables. We also plotted a LOESS line on these plots to allow
422
for easy visualization of the central tendency of the predicted values. Relative importance is
423
constructed by counting the number of times a variable is selected for splitting in each tree, weighted
424
by the squared improvement of the model as a result of each split, and averaged over all trees (see
425
ref.133135 for more details). BRT parameters were selected to amplify the deviance explained by the
426
model. We tested interaction depth between 2 and 5, and the learning rates of 0.1, 0.01, and 0.001.
427
The best models were the ones with learning rate of 5 and interaction depth of 0.001.We performed
428
a 50–50 cross-validation procedure and estimated the model performance (
𝐷&=429
!!"#$%&'"()*%+,!"#$%&'"-.)//,#%+$0%*$)&
!"#$%&'"()*%+
) following Leathwick et al.107. As the datasets in this study have
430
not always followed the same sampling methodology, and show different functional traits and
431
environmental variables, we fitted the BRT models following a Laplace distribution of the errors to
432
reduce the absolute error loss from the variation among datasets. BRT models were fitted using the
433
function gbm.step of the package ‘dismo’ v.1.1-4 (ref.136).
434
Main results show the distance decay results based on total similarities (equation 1), but we also
435
partitioned the similarities into replacement and richness difference components following the
436
methodology described in the Appendix S2. Replacement gives the variation as a result of the
437
substitution of species (turnover) or functional traits (functional replacement), and richness
438
differences accounts for the variation as a result of net differences induced by the loss/gain of species
439
or traits137. We only show the results of the partitioned components using occurrence data for
440
simplicity. The final figures were prepared using the tools from the tidyverse environment138 in the
441
R software v.4.0.2 (ref.108).
442
Results
443
Strength of the distance decay
444
The taxonomic and functional similarities had a mean correlation of 0.74 (sd ± 0.20) within datasets.
445
The distance decays showed a wide range of shapes, from very steep decays to almost flat
446
relationships (Fig. 4). The average r² using occurrence data for taxonomic similarities was 0.099 (sd
447
± 0.129) and 0.061 (sd ± 0.091) for functional similarities. Spatial distance decays of taxonomic
448
similarities were significantly stronger than the distance decays of functional similarities when
449
considering both occurrence (Fig 4a; t = 6.330, p < 0.001, df = 148) and abundance data (Appendix
450
S3, Fig. S1), supporting H1a spatial distance decay is stronger for taxonomic than functional
451
similarities (Fig. 4a).
452
However, our results did not support H1b as the distance decay for taxonomic similarities (mean r² =
453
0.103, sd ± 0.095) were also, on average, stronger than for functional similarities (mean r² = 0.076,
454
sd ± 0.086) along environmental distances (Fig 4b; t = 6.935, p < 0.001, df = 148). Note, however,
455
that 41 out of 149 datasets had stronger distance decay of functional similarities than taxonomic
456
similarities along environmental gradients. Most of the biotic groups had at least one dataset with a
457
stronger relationship for functional similarities than for taxonomic similarities, except for corals,
458
foraminifera, lichens, amphibians and fungi each of which comprised only one dataset.
459
Rate of the distance decay
460
The mean slope of the spatial distance decay was 0.009 (sd ± 0.027) for taxonomic similarities, and
461
0.004 (sd ± 0.015) for functional similarities (Fig 4a). For environmental distances, the mean slope
462
of the distance decay was 1.073 (sd ± 1.063) for taxonomic similarities and 0.365 (sd ± 0.361) for
463
functional similarities (Fig 4b). Regarding the biotic groups, terrestrial plants had the steepest slopes
464
along spatial distance both for taxonomic and functional similarities (Fig. 5). Along environmental
465
distance, corals had the steepest slopes (Fig. 5). Similar patterns were found for abundance-based
466
similarities, except for the biotic groups, where aquatic plants had the steepest slopes along spatial
467
distances (Appendix S3).
468
Across datasets, BRT explained 36.51% of the deviance of the slopes of the spatial distance decay
469
for taxonomic similarities, and 36.86% for functional similarities using occurrence data. For the
470
distance decay along environmental distances, BRT explained 14.43% of the deviance of the slopes
471
of the decay of taxonomic similarities and 20.40% for functional similarities. Spatial extent and γ-
472
diversity contributed most to the variation in slopes along either spatial or environmental distance
473
using both occurrence and abundance-based similarities (Fig. 6 – 7a, Appendix S3).
474
Latitudinal patterns
475
The slopes of spatial distance decay of both taxonomic and functional similarities were the steepest
476
in datasets centred at ca. 35–45º, partly supporting H2a that distance decay was flatter at high latitudes
477
(Fig. 6a). However, note that taxonomic spatial distance decay sharply decreased towards the poles.
478
The slopes of environmental distance decay were flatter in the datasets from high latitudes (Fig. 6b),
479
providing no support to hypothesis H2b.
480
Spatial extent
481
The distance decay of taxonomic and functional similarities was flatter in the datasets that covered
482
larger spatial extent both for occurrence (Fig. 6a) and abundance data (Appendix S3, Fig. S3a),
483
supporting hypothesis H3a distance decay becomes flatter with increasing spatial extent. For
484
environmental distances, distance decay was steeper in the datasets that covered larger spatial extents
485
for both taxonomic and functional similarities, agreeing thus with H3b that distance decay would
486
become steeper with larger spatial extent.
487
Realms
488
Marine ecosystems had flatter slopes compared to freshwater or terrestrial ecosystems considering
489
environmental distances, but not for spatial distances, thus partly agreeing with H4 (Fig. 6). However,
490
the importance of the realms in BRTs was overall low. A similar pattern emerged for abundance-
491
based similarities (Appendix S3, Fig. S3).
492
Organismal variables and dataset features
493
The slopes of both spatial and environmental distance decays were steeper for larger-bodied
494
organisms in taxonomic and functional similarity (Fig. 7a–b). Organisms relying on seed dispersal
495
had steeper slopes along spatial and environmental distances than other dispersal types, but the overall
496
importance of dispersal mode was low (Fig. 7b). Taxonomic γ-diversity had a U-shaped relationship
497
with slopes for distance decay along spatial and environmental distances (Fig. 7b). Slopes of distance
498
decay had an overall decreasing trend for functional γ-diversity for both spatial and environmental
499
distances (Fig. 7a–b). Generally, slopes were steeper in the datasets where the number of study sites
500
was higher (Fig. 7a), and flatter when datasets comprised only a few environmental variables (Fig
501
7b).
502
Replacement and richness differences
503
The slopes of taxonomic replacement along spatial distance decreased rapidly in the datasets above
504
35º while the functional replacement peaked at ca. 45º (Appendix S4, Fig. S1a). Along environmental
505
distance, the taxonomic replacement increased towards higher latitudes while the functional
506
replacement did not vary notably along latitude (Appendix S4, Fig. S1b). For the richness differences
507
component, the slopes of both taxonomic and functional similarities were the steepest in the datasets
508
at ca. 45º degrees for the spatial distance decay (Appendix S4, Fig. S2a). For environmental distances,
509
slopes became flatter from low to high latitudes up to ca. 50º degrees for taxonomic similarities while
510
for functional similarities, slopes did not vary along latitude (Appendix S4, Fig. S2b). Both
511
replacement and richness differences showed flatter spatial slopes with increasing spatial extent
512
(Appendix S4, Fig S1-S2). In contrast, environmental slopes increased with spatial extent only
513
replacement (Appendix S4, Fig. S1b) while the effects of spatial extent for the slopes of richness
514
differences along environment was very low (Appendix S4, Fig. S2b). Furthermore, marine
515
ecosystems showed the flattest slopes of replacement along environmental gradients (Appendix S4,
516
Fig. S1b) while freshwater ecosystems had the flattest slopes of richness differences (Appendix S4,
517
Fig. S2b). Details about the organismal variables and datasets features can be found in the Appendix
518
S4.
519
Discussion
520
Community ecology and biogeography have lacked a comprehensive evaluation of functional β-
521
diversity across different taxa and ecosystems globally. Earlier studies suggest that functional β-
522
diversity better reflects environmental variability compared with taxonomic β-diversity, and that
523
focusing on functional β-diversity may help, for example, understand how humans impact ecosystems
524
by modifying the local environment33,3941. This is because functional traits should reflect best the
525
ecological requirements of species. Using a comparative analysis across biotic groups, ecosystem
526
types and realms, we show here that (i) taxonomic distance decay is generally stronger along spatial
527
gradients than functional distance decay, and that (ii) the decay of functional similarities along
528
environmental gradients is typically not stronger than the decay of taxonomic similarities, unlike
529
previously suggested.
530
The strength of the distance decay of taxonomic and functional similarities
531
The stronger taxonomic than functional distance decay along space provides empirical evidence for
532
the idea that the taxonomic distance decay is a robust approach for ecological and biogeographical
533
studies, supporting H1a. Compositional differences effectively summarize dispersal-related factors as
534
well as species responses to climatic and other spatially structured environmental variables. However,
535
spatial distance decay of functional similarities may not reflect well geographic differences in biotic
536
communities. This probably stems from the different roles played by deterministic and stochastic
537
drivers when shaping taxonomic and functional composition: functional composition mirrors mostly
538
local environmental filtering and typically does not strongly reflect dispersal limitations or species
539
pool effects that influence stronger taxonomic composition42. Yet, the specific outcomes of any
540
analysis of functional diversity depends on the functional traits included in the analysis139 and how
541
researchers handle individual trait variability140. Also, some morphological or size-related traits with
542
no clear functional meaning may turn out informative when exploring geographic patterns in
543
functional composition42. For example, functional traits rather than species identities explained more
544
variability of tree communities along broad spatial gradients141 or the variation of phytoplankton
545
communities along a large South America gradient142. Such findings point to the fact that the
546
decisions about which functional traits to include in the analysis is critical.
547
Our analysis suggests that, overall, functional distance decay is also somewhat weaker than
548
taxonomic distance decay along environmental gradients. However, this result is likely context-
549
dependent, and the stronger functional than taxonomic distance decay depends on whether the species
550
replaced from one community to another are a random subsample of functionally redundant species
551
from the regional pool or not34. In fact, in 40 datasets, distance decay of functional similarities was
552
stronger than taxonomic similarities along environmental gradients. The datasets with stronger
553
distance decay of functional than taxonomic similarities spanned a broad range of latitudes, number
554
of study sites and environmental variables. Therefore, for using such heterogeneous datasets, we are
555
not able to provide any strict guidance on the choice of functional traits or environmental variables
556
to be measured in future studies. For example, the dataset on grassland arthropods from the
557
Biodiversity exploratories project had standardized traits and environmental variables, but only
558
Homoptera out of four different taxa showed stronger functional than taxonomic distance decay along
559
environmental gradients. One explanation is that the whole organisms are susceptible to
560
environmental filtering, and each species comprises a set of traits that cannot be physically filtered
561
as a response to the environment. Therefore, environmental filtering on a given trait of a species may
562
also filter other traits simultaneously, or a given species may comprise a trait not filtered by the
563
environment, which tends to increase the community similarity among sites. Yet, we emphasize that
564
the variation in the rate of distance decay of functional similarities along environmental gradients
565
across datasets was better explained in BRT than the variation in the rate of the distance decay of
566
taxonomic similarities. This suggests that the taxonomic metrics may be more context dependent than
567
the functional metrics along environmental gradients and that functional features may be more useful
568
to generalize across taxa and ecosystems24. Furthermore, functional distance decay should not be
569
much affected by dispersal effects and regional species pools as compared to taxonomic distance
570
decay.
571
The effects of latitude on the rate of distance decay
572
In addition to our master hypothesis, we investigated whether the rate of distance decay showed
573
consistent variation across ecosystems, along geographic gradients and among major taxonomic
574
groups. We did not find slower rates of decay in the datasets at higher latitudes, but rather, concurring
575
with the recent meta-analysis of species turnover143, we found that taxonomic similarities decayed
576
the fastest at mid latitudes, above which the rate lowered down. Traditionally, this pattern has been
577
explained with the Rapoport’s rule, whereby there is an increase in species range size at higher
578
latitudes144 and hence lower taxonomic turnover. Yet, such finding may also stem from landscape
579
fragmentation that increases β-diversity145, especially at mid latitudes prone to strong human impact
580
and at local spatial scales50. We also observed a faster rate of functional spatial distance decay towards
581
poles, agreeing with our hypothesis. This may reflect the fact that the high species diversity of the
582
tropics is mainly due to niche overlap51, which increases the functional redundancy and reduces the
583
functional turnover52. Furthermore, the latitudinal decrease in the rate of abundance-based functional
584
distance decay (Appendix S3, Fig. S1) suggests an optimal utilization of the functional space, as have
585
been observed earlier exclusively for marine organisms146.
586
Taxonomic and functional distance decay along environmental gradients exhibited a clear minimum
587
in the datasets near 50° while increasing notably from 60° towards the poles especially for taxonomic
588
similarities. This result points to a breakpoint in total similarities that stems from richness differences,
589
as the replacement component did not have similar breakpoints but, rather, had similar replacement
590
levels in the tropics with decreasing trend at mid- and high latitudes. Latitudinal breakpoints in
591
turnover have been found earlier147 in terrestrial vertebrates at ca. 30°, where turnover decreased
592
substantially, while nestedness component increased. Soininen et al.143 found a breakpoint for
593
turnover component at 41°, whereas there was no breakpoint in nestedness component. Present results
594
suggest that the rate of distance decay is relatively similar through the extensive tropical region,
595
whereas it either increases or decreases rapidly at mid latitudes, depending on β-diversity metric or
596
whether this phenomenon is examined along spatial or environmental gradients.
597
The effect of spatial extent on the rate of distance decay
598
The rate of spatial distance decay was slower in the datasets covering larger spatial extent as we
599
hypothesized, perhaps suggesting that regional species pools are limited, and new species are not
600
found constantly at the same frequency when extent is larger. Lower decay rates in larger study areas
601
could also result from repeated patterns in environmental variation, that is, environmental patchiness
602
or natural periodicity in the environment11. Agreeing with our hypothesis, we also found that the rate
603
of decay along environmental distance was higher in the datasets covering larger spatial extent. These
604
findings indicate that spatial distance decay is more affected by species pool effects and dispersal
605
processes than environmental distance decay, possibly because the latter reflects more strongly the
606
level of local deterministic environmental filtering processes. Similar evidence has accumulated from
607
case studies conducted in various ecosystems33,39,41,148. The finding that the rate of distance decay
608
along environmental distance was higher in the datasets covering larger extents indicates the stronger
609
environmental filtering at larger study areas. We also note that, in our BRT models, extent and γ-
610
diversity had by far the largest relative importance, suggesting that their interplay plays a key role in
611
shaping distance decay.
612
The effect of realm on the distance decay
613
We found evidence for a lower rate of distance decay in marine versus terrestrial or freshwater
614
ecosystems. Moreover, we found very comparable distance decay slopes for terrestrial and
615
freshwaters, and the factor ‘realm’ showed low relative importance in the BRT models. Overall, this
616
finding agrees with earlier meta-review on β-diversity19, suggesting that large-scale diversity patterns
617
are generally weaker in marine ecosystems149. However, marine ecosystems would have lower
618
species turnover than freshwater or terrestrial systems49. As connectivity, energy flows, dispersal
619
modes, body size structure and trophic dynamics differ substantially between dry and wet
620
ecosystems150, it would be vital to investigate possible differences in turnover among the realms more
621
closely.
622
Organismal variables and dataset features
623
Organism size did seem to affect taxonomic or functional distance decay along spatial and
624
environmental gradients as the slopes typically increased with organism body size. This may be
625
because β-diversity should be low among the small microbial taxa with efficient passive dispersal19.
626
The rationale behind such idea is that efficient dispersal homogenizes communities among sites
627
resulting in lower β-diversity151. Body size is also a key driver of organisms’ biological complexity152,
628
and it may be that smaller organisms show a much more limited set of trait combinations than
629
macroorganisms, leading to a lower functional redundancy among larger species. Furthermore, our
630
knowledge about the taxonomy and functional traits of organisms is typically size-dependent. For
631
example, the identification of larger species is much easier than that of microorganisms, which also
632
applies to the identification and measurement of soft functional traits153,154. Therefore, the values of
633
β-diversity of small organisms may be typically underestimated.
634
Patterns in environmental distance decay were relatively congruent with spatial distance decay
635
regarding dispersal mode, suggesting that taxa which disperse passively do not seem to track
636
environmental gradients more efficiently compared with less dispersive taxa. It may also be that
637
small-sized taxa were filtered along some unmeasured spatially-structured environmental gradients,
638
and the pattern was thus detected as spatial turnover even if caused by some underlying unmeasured
639
environmental factors. Forthcoming studies would greatly benefit from disentangling the signal of
640
unmeasured environmental variables from true dispersal limitation155.
641
Study design
642
There are also some possibly influential aspects in our study design that should be discussed.
643
Although the study is global in its extent, the availability of datasets was not evenly distributed
644
geographically. This is a well-known problem in biodiversity research156 that calls for
645
complementary studies to verify that these trends hold true in poorly sampled regions.
646
Also, we relied on the suite of traits and environmental variables included in the original datasets
647
and, thus, the collection of traits and environmental variables used differed somewhat among
648
datasets even for the same focal taxonomic groups. This increases the uncertainty on how
649
environmental variables filter the functional structure of communities in different contexts and how
650
strong the taxonomic community-environment relationships are. An alignment of key traits and
651
environmental variables is therefore desirable, but requires a suite of sister studies following the
652
same protocol, which is unfortunately not yet available. Moreover, the fact that some of the biotic
653
groups (e.g., corals, foraminifera) were underrepresented in our analysis with only one dataset
654
included (Fig. 2), or the total lack of some taxa (e.g. aquatic and terrestrial mammals, bacteria),
655
makes it more difficult to generalize distance decay across taxa.
656
Concluding remarks
657
In summary, we believe our analysis is an important step towards a more comprehensive
658
understanding of patterns and drivers of functional β-diversity, particularly in comparison with the
659
patterns and drivers of taxonomic β-diversity that have so far attracted much more research interest
660
compared with functional β-diversity. Here, we found that functional distance decay is scale-
661
dependent and a product of large-scale geographic factors (latitude) and taxonomic and functional
662
g-diversity, but is also driven by organisms’ biology to some degree. In general, taxonomic distance
663
decay provides a better tool for many aspects of biogeographical research, because it reflects
664
dispersal-related factors as well as species responses to climatic and other typically spatially-
665
structured environmental variables. However, functional distance decay may be a cost-effective
666
option for investigating how humans impact ecosystems via modifying the environment. Overall,
667
the present findings and data shed light into the congruence between the functional and taxonomic
668
diversity patterns and provide useful new information to the field of functional biogeography.
669
Author contributions: Caio Graco-Roza and Janne Soininen contributed equally to the original idea,
670
data analysis, and the writing of the first draft. Jani Heino and Otso Ovaskainen advised on the main
671
idea, analysis and commented on the first draft. Francesco Maria Sabatini coordinated the data
672
compilation from the sPlot database and commented on the first draft, Martin Gossner coordinated
673
the compilation of the data from the Biodiversity exploratories project and commented on the first
674
draft. All other authors shown in alphabetical order contributed data and commented on the draft.
675
Acknowledgements: This research was funded by the Coordination for the Improvement of Higher
676
Education Personnel (CAPES), the Carlos Chagas Filho Research Support Foundation (FAPERJ),
677
and the Ella and Georg Erhnrooth foundation. The sPlot project was initiated by sDiv, the Synthesis
678
Centre of the German Centre for Integrative Biodiversity Research (iDiv) HalleJenaLeipzig,
679
funded by the German Research Foundation (DFG FZT 118) and is now a platform of iDiv. The
680
study was supported by the TRY initiative on plant traits (http://www.try-db.org). We are also
681
grateful to Jens Kattge and TRY database. TRY is hosted, developed and maintained at the Max
682
Planck Institute for Biogeochemistry (MPIBGC) in Jena, Germany, in collaboration with the
683
German Centre for Integrative Biodiversity Research (iDiv) HalleJenaLeipzig. The CESTES
684
database of metacommunities is also an initiative of iDiv led by Alienor Jeliazkov. We thank sDiv
685
for supporting the open science initiative.
686
References
687
1. IPBES. Global assessment report on biodiversity and ecosystem services of the
688
Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. (IPBES
689
secretariat, 2019).
690
2. Hooper, D. U. et al. Effects of biodiversity on ecosystem functioning: a consensus of current
691
knowledge. Ecological Monographs 75, 3–35 (2005).
692
3. Cardinale, B. J. et al. Biodiversity loss and its impact on humanity. Nature 486, 59–67
693
(2012).
694
4. Anderson, M. J. et al. Navigating the multiple meanings of β diversity: A roadmap for the
695
practicing ecologist. Ecology Letters 14, 19–28 (2011).
696
5. Mori, A. S., Isbell, F. & Seidl, R. β-Diversity, Community Assembly, and Ecosystem
697
Functioning. Trends in Ecology and Evolution vol. 33 549–564 (2018).
698
6. Qian, H., Ricklefs, R. E. & White, P. S. Beta diversity of angiosperms in temperate floras of
699
eastern Asia and eastern North America. Ecology Letters 8, 15–22 (2004).
700
7. Kraft, N. J. B. et al. Disentangling the drivers of β diversity along latitudinal and elevational
701
gradients. Science 333, 1755–1758 (2011).
702
8. Blowes, S. A. et al. The geography of biodiversity change in marine and terrestrial
703
assemblages. Science 366, 339–345 (2019).
704
9. Gossner, M. M. et al. Land-use intensification causes multitrophic homogenization of
705
grassland communities. Nature 540, 266–269 (2016).
706
10. Socolar, J. B., Gilroy, J. J., Kunin, W. E. & Edwards, D. P. How Should Beta-Diversity
707
Inform Biodiversity Conservation? Trends in Ecology and Evolution 31, 67–80 (2016).
708
11. Nekola, J. C. & White, P. S. The distance decay of similarity in biogeography and ecology.
709
Journal of Biogeography 26, 867–878 (1999).
710
12. Nekola, J. C. & McGill, B. J. Scale dependency in the functional form of the distance decay
711
relationship. Ecography 37, 309–320 (2014).
712
13. Hubbell, S. P. The Unified Neutral Theory of Biodiversity and Biogeography. Monographs
713
in Population Biology vol. 32 (Princeton University Press, 2001).
714
14. Cottenie, K. Integrating environmental and spatial processes in ecological community
715
dynamics. Ecology Letters 8, 1175–1182 (2005).
716
15. Astorga, A. et al. Distance decay of similarity in freshwater communities: Do macro- and
717
microorganisms follow the same rules? Global Ecology and Biogeography 21, 365–375
718
(2012).
719
16. Daws, S. C. et al. Do shared traits create the same fates? Examining the link between
720
morphological type and the biogeography of fungal and bacterial communities. Fungal
721
Ecology 46, 100948 (2020).
722
17. Mammola, S. et al. Local- versus broad-scale environmental drivers of continental β -
723
diversity patterns in subterranean spider communities across Europe. Proceedings of the
724
Royal Society B: Biological Sciences 286, 20191579 (2019).
725
18. Morlon, H. et al. A general framework for the distance-decay of similarity in ecological
726
communities. Ecology Letters 11, 904–917 (2008).
727
19. Soininen, J., McDonald, R. & Hillebrand, H. The distance decay of similarity in ecological
728
communities. Ecography 30, 3–12 (2007).
729
20. Petchey, O. L. On the statistical significance of functional diversity effects. Functional
730
Ecology 18, 297–303 (2004).
731
21. Petchey, O. L. & Gaston, K. J. Functional diversity: Back to basics and looking forward.
732
Ecology Letters 9, 741–758 (2006).
733
22. Villéger, S., Ramos Miranda, J., Flores Hernandez, D. & Mouillot, D. Low functional β-
734
diversity despite high taxonomic β-diversity among tropical estuarine fish communities. PloS
735
one 7, e40679–e40679 (2012).
736
23. Violle, C., Reich, P. B., Pacala, S. W., Enquist, B. J. & Kattge, J. The emergence and
737
promise of functional biogeography. Proceedings of the National Academy of Sciences of the
738
United States of America 111, 13690–13696 (2014).
739
24. McGill, B. J., Enquist, B. J., Weiher, E. & Westoby, M. Rebuilding community ecology
740
from functional traits. Trends in Ecology and Evolution 21, 178–185 (2006).
741
25. Mouillot, D. et al. A functional approach reveals community responses to disturbances.
742
Trends in Ecology and Evolution 28, 167–177 (2013).
743
26. Villéger, S., Mason, N. W. H. H. & Mouillot, D. New multidimensional functional diversity
744
indices for a multifaceted framework in functional ecology. Ecology 89, 2290–2301 (2008).
745
27. Buisson, L., Grenouillet, G., Villéger, S., Canal, J. & Laffaille, P. Toward a loss of
746
functional diversity in stream fish assemblages under climate change. Global Change
747
Biology 19, 387–400 (2013).
748
28. Pool, T. K. & Olden, J. D. Taxonomic and functional homogenization of an endemic desert
749
fish fauna. Diversity and Distributions 18, 366–376 (2011).
750
29. Villéger, S., Grenouillet, G. & Brosse, S. Decomposing functional β-diversity reveals that
751
low functional β-diversity is driven by low functional turnover in European fish assemblages.
752
Global Ecology and Biogeography 22, 671–681 (2013).
753
30. Villéger, S., Grenouillet, G. & Brosse, S. Functional homogenization exceeds taxonomic
754
homogenization among European fish assemblages. Global Ecology and Biogeography 23,
755
1450–1460 (2014).
756
31. Penone, C. et al. Global mammal beta diversity shows parallel assemblage structure in
757
similar but isolated environments. Proceedings of the Royal Society B: Biological Sciences
758
283, 1–9 (2016).
759
32. Heino, J. & Tolonen, K. T. Ecological drivers of multiple facets of beta diversity in a lentic
760
macroinvertebrate metacommunity. Limnology and Oceanography 62, 2431–2444 (2017).
761
33. Sokol, E. R., Benfield, E. F., Belden, L. K. & Maurice Valett, H. The Assembly of
762
Ecological Communities Inferred from Taxonomic and Functional Composition. The
763
American Naturalist 177, 630–644 (2011).
764
34. Swenson, N. G., Anglada-Cordero, P. & Barone, J. A. Deterministic tropical tree
765
community turnover: Evidence from patterns of functional beta diversity along an elevational
766
gradient. Proceedings of the Royal Society B: Biological Sciences 278, 877–884 (2011).
767
35. Leibold, M. A. & McPeek, M. A. Coexistence of the niche and neutral perspectives in
768
community ecology. Ecology 87, 1399–1410 (2006).
769
36. Baiser, B. & Lockwood, J. L. The relationship between functional and taxonomic
770
homogenization. Global Ecology and Biogeography 20, 134–144 (2011).
771
37. Mouton, T. L., Tonkin, J. D., Stephenson, F., Verburg, P. & Floury, M. Increasing climate‐
772
driven taxonomic homogenization but functional differentiation among river
773
macroinvertebrate assemblages. Global Change Biology gcb.15389 (2020)
774
doi:10.1111/gcb.15389.
775
38. Ricotta, C., Laroche, F., Szeidl, L. & Pavoine, S. From alpha to beta functional and
776
phylogenetic redundancy. Methods in Ecology and Evolution 11, 487–493 (2020).
777
39. Meynard, C. N. et al. Beyond taxonomic diversity patterns: how do α, β and γ components
778
of bird functional and phylogenetic diversity respond to environmental gradients across
779
France? Global Ecology and Biogeography 20, 893–903 (2011).
780
40. Spasojevic, M. J., Yablon, E. A., Oberle, B. & Myers, J. A. Ontogenetic trait variation
781
influences tree community assembly across environmental gradients. Ecosphere 5, 1–20
782
(2014).
783
41. Weinstein, B. G. et al. Taxonomic, Phylogenetic, and Trait Beta Diversity in South
784
American Hummingbirds. The American Naturalist 184, 211–224 (2014).
785
42. Soininen, J., Jamoneau, A., Rosebery, J. & Passy, S. I. Global patterns and drivers of
786
species and trait composition in diatoms. Global Ecology and Biogeography 25, 940–950
787
(2016).
788
43. Graco-Roza, C. et al. Functional rather than taxonomic diversity reveals changes in the
789
phytoplankton community of a large dammed river. Ecological Indicators 107048 (2020)
790
doi:10.1016/j.ecolind.2020.107048.
791
44. Soininen, J., Lennon, J. J. & Hillebrand, H. A multivariate analysis of beta diversity across
792
organisms and environments. Ecology 88, 2830–2838 (2007).
793
45. Qian, H. Global comparisons of beta diversity among mammals, birds, reptiles, and
794
amphibians across spatial scales and taxonomic ranks. Journal of Systematics and Evolution
795
47, 509–514 (2009).
796
46. Qian, H., Badgley, C. & Fox, D. L. The latitudinal gradient of beta diversity in relation to
797
climate and topography for mammals in North America. Global Ecology and Biogeography
798
18, 111–122 (2009).
799
47. Hillebrand, H. On the generality of the latitudinal diversity gradient. American Naturalist
800
163, 192–211 (2004).
801
48. Soininen, J. Species Turnover along Abiotic and Biotic Gradients: Patterns in Space Equal
802
Patterns in Time? BioScience 60, 433–439 (2010).
803
49. Drakare, S., Lennon, J. J. & Hillebrand, H. The imprint of the geographical, evolutionary
804
and ecological context on species-area relationships. Ecology Letters 9, 215–227 (2005).
805
50. Rohde, K. Rapoport’s Rule is a local phenomenon and cannot explain latitudinal gradients
806
in species diversity. Biodiversity Letters vol. 3 10–13 (1996).
807
51. May, R. M. & MacArthur, R. H. Niche overlap as a function of environmental variability.
808
Proceedings of the National Academy of Sciences of the United States of America 69, 1109–
809
1113 (1972).
810
52. Lamanna, C. et al. Functional trait space and the latitudinal diversity gradient. Proceedings
811
of the National Academy of Sciences of the United States of America 111, 13745–13750
812
(2014).
813
53. Terborgh, J. On the Notion of Favorableness in Plant Ecology. The American Naturalist
814
107, 481–501 (1973).
815
54. Soininen, J. & Hillebrand, H. Disentangling distance decay of similarity from richness
816
gradients: response to Baselga (2007). Ecography 30, 842–844 (2007).
817
55. Cornwell, W. K. & Ackerly, D. D. Community assembly and shifts in plant trait
818
distributions across an environmental gradient in coastal California. Ecological Monographs
819
79, 109–126 (2009).
820
56. Cornwell, W. K., Schwilk, D. W. & Ackerly, D. D. A trait-based test for habitat filtering:
821
Convex hull volume. Ecology 87, 1465–1471 (2006).
822
57. Steinbauer, M. J., Dolos, K., Reineking, B. & Beierkuhnlein, C. Current measures for
823
distance decay in similarity of species composition are influenced by study extent and grain
824
size. Global Ecology and Biogeography 21, 1203–1212 (2012).
825
58. Triantis, K. A., Guilhaumon, F. & Whittaker, R. J. The island species-area relationship:
826
biology and statistics. Journal of Biogeography 39, 215–231 (2011).
827
59. Pianka, E. R. Latitudinal Gradients in Species Diversity: A Review of Concepts. The
828
American Naturalist 100, 33–46 (1966).
829
60. Questad, E. J. & Foster, B. L. Coexistence through spatio-temporal heterogeneity and
830
species sorting in grassland plant communities. Ecology Letters 11, 717–726 (2008).
831
61. Bergholz, K. et al. Environmental heterogeneity drives fine-scale species assembly and
832
functional diversity of annual plants in a semi-arid environment. Perspectives in Plant
833
Ecology, Evolution and Systematics 24, 138–146 (2017).
834
62. Kadmon, R. & Allouche, O. Integrating the effects of area, isolation, and habitat
835
heterogeneity on species diversity: A unification of island biogeography and niche theory.
836
American Naturalist 170, 443–454 (2007).
837
63. Gazol, A. et al. A negative heterogeneity–diversity relationship found in experimental
838
grassland communities. Oecologia 173, 545–555 (2013).
839
64. Laanisto, L. et al. Microfragmentation concept explains non-positive environmental
840
heterogeneity–diversity relationships. Oecologia 171, 217–226 (2012).
841
65. Clarke, A. Is there a latitudinal diversity cline in the sea? Trends in Ecology & Evolution 7,
842
286–287 (1992).
843
66. Cornell, H. v & Harrison, S. P. What Are Species Pools and When Are They Important?
844
Annual Review of Ecology, Evolution, and Systematics 45, 45–67 (2014).
845
67. Jenkins, D. G. et al. Does size matter for dispersal distance? Global Ecology and
846
Biogeography 16, 415–425 (2007).
847
68. Bie, T. et al. Body size and dispersal mode as key traits determining metacommunity
848
structure of aquatic organisms. Ecology Letters 15, 740–747 (2012).
849
69. Bruelheide, H. et al. sPlot – A new tool for global vegetation analyses. Journal of
850
Vegetation Science 30, 161–186 (2019).
851
70. Jeliazkov, A. et al. A global database for metacommunity ecology, integrating species,
852
traits, environment and space. Scientific data 7, 6 (2020).
853
71. Kattge, J. et al. TRY - a global database of plant traits. Global Change Biology 17, 2905–
854
2935 (2011).
855
72. Brind’Amour, A. et al. Relationships between species feeding traits and environmental
856
conditions in fish communities: A three-matrix approach. Ecological Applications 21, 363–
857
377 (2011).
858
73. Chong-Seng, K. M., Mannering, T. D., Pratchett, M. S., Bellwood, D. R. & Graham, N. A.
859
J. J. The influence of coral reef benthic condition on associated fish assemblages. PLoS ONE
860
7, 1–10 (2012).
861
74. Cleary, D. F. R. R. et al. Variation in the composition of corals, fishes, sponges,
862
echinoderms, ascidians, molluscs, foraminifera and macroalgae across a pronounced in-to-
863
offshore environmental gradient in the Jakarta Bay–Thousand Islands coral reef complex.
864
Marine Pollution Bulletin 110, 701–717 (2016).
865
75. Ribera, I., Dolédec, S., Downie, I. S. & Foster, G. N. Effect of land disturbance and stress
866
on species traits of ground beetle assemblages. Ecology 82, 1112–1129 (2001).
867
76. Pakeman, R. J. Multivariate identification of plant functional response and effect traits in an
868
agricultural landscape. Ecology 92, 1353–1365 (2011).
869
77. Bagaria, G., Pino, J., Rodà, F. & Guardiola, M. Species traits weakly involved in plant
870
responses to landscape properties in Mediterranean grasslands. Journal of Vegetation Science
871
23, 432–442 (2012).
872
78. Frenette-Dussault, C., Shipley, B., Léger, J.-F., Meziane, D. & Hingrat, Y. Functional
873
structure of an arid steppe plant community reveals similarities with Grime’s C-S-R theory.
874
Journal of Vegetation Science 23, 208–222 (2011).
875
79. Fried, G., Kazakou, E. & Gaba, S. Trajectories of weed communities explained by traits
876
associated with species’ response to management practices. Agriculture, Ecosystems and
877
Environment 158, 147–155 (2012).
878
80. Raevel, V., Violle, C. & Munoz, F. Mechanisms of ecological succession: Insights from
879
plant functional strategies. Oikos 121, 1761–1770 (2012).
880
81. Eallonardo, A. S., Leopold, D. J., Fridley, J. D. & Stella, J. C. Salinity tolerance and the
881
decoupling of resource axis plant traits. Journal of Vegetation Science 24, 365–374 (2013).
882
82. Jamil, T., Ozinga, W. A., Kleyer, M. & ter Braak, C. J. F. F. Selecting traits that explain
883
species-environment relationships: A generalized linear mixed model approach. Journal of
884
Vegetation Science 24, 988–1000 (2013).
885
83. Meffert, P. J. & Dziock, F. The influence of urbanisation on diversity and trait composition
886
of birds. Landscape Ecology 28, 943–957 (2013).
887
84. Chmura, D., Żarnowiec, J. & Staniaszek-Kik, M. Interactions between plant traits and
888
environmental factors within and among montane forest belts: A study of vascular species
889
colonising decaying logs. Forest Ecology and Management 379, 216–225 (2016).
890
85. Robroek, B. J. M. M. et al. Taxonomic and functional turnover are decoupled in European
891
peat bogs. Nature Communications 8, 1161 (2017).
892
86. van Klink, R. et al. No detrimental effects of delayed mowing or uncut grass refuges on
893
plant and bryophyte community structure and phytomass production in low-intensity hay
894
meadows. Basic and Applied Ecology 20, 1–9 (2017).
895
87. Díaz, A. M., Alonso, M. L. S. & Gutiérrez, M. R. V.-A. Biological traits of stream
896
macroinvertebrates from a semi-arid catchment: patterns along complex environmental
897
gradients. Freshwater Biology 53, 1–21 (2007).
898
88. Gallardo, B., Gascón, S., García, M. & Comín, F. A. Testing the response of
899
macroinvertebrate functional structure and biodiversity to flooding and confinement. Journal
900
of Limnology 68, 315–326 (2009).
901
89. Jeliazkov, A. et al. Level-dependence of the relationships between amphibian biodiversity
902
and environment in pond systems within an intensive agricultural landscape. Hydrobiologia
903
723, 7–23 (2014).
904
90. Barbaro, L. & van Halder, I. Linking bird, carabid beetle and butterfly life-history traits to
905
habitat fragmentation in mosaic landscapes. Ecography 32, 321–333 (2009).
906
91. Dziock, F. et al. Reproducing or dispersing? Using trait based habitat templet models to
907
analyse Orthoptera response to flooding and land use. Agriculture, Ecosystems and
908
Environment 145, 85–94 (2011).
909
92. Gibb, H. et al. Responses of foliage-living spider assemblage composition and traits to a
910
climatic gradient in Themeda grasslands. Austral Ecology 40, 225–237 (2015).
911
93. Barbosa, E. P. et al. Fine-scale Beta-diversity Patterns Across Multiple Arthropod Taxa
912
Over a Neotropical Latitudinal Gradient. Biotropica 47, 588–594 (2015).
913
94. Krasnov, B. R. et al. Assembly rules of ectoparasite communities across scales: Combining
914
patterns of abiotic factors, host composition, geographic space, phylogeny and traits.
915
Ecography 38, 184–197 (2015).
916
95. Spake, R., Barsoum, N., Newton, A. C. & Doncaster, C. P. Drivers of the composition and
917
diversity of carabid functional traits in UK coniferous plantations. Forest Ecology and
918
Management 359, 300–308 (2016).
919
96. Lowe, E. C., Threlfall, C. G., Wilder, S. M. & Hochuli, D. F. Environmental drivers of
920
spider community composition at multiple scales along an urban gradient. Biodiversity and
921
Conservation 27, 829–852 (2018).
922
97. Gonçalves-Souza, T., Romero, G. Q. & Cottenie, K. Metacommunity versus Biogeography:
923
A Case Study of Two Groups of Neotropical Vegetation-Dwelling Arthropods. PLoS ONE 9,
924
e115137 (2014).
925
98. Bartonova, A., Benes, J., Fric, Z. F., Chobot, K. & Konvicka, M. How universal are reserve
926
design rules? A test using butterflies and their life history traits. Ecography 39, 456–464
927
(2016).
928
99. Cleary, D. F. R. R. et al. Bird species and traits associated with logged and unlogged forest
929
in Borneo. Ecological Applications 17, 1184–1197 (2007).
930
100. Barbaro, L., Brockerhoff, E. G., Giffard, B. & van Halder, I. Edge and area effects on
931
avian assemblages and insectivory in fragmented native forests. Landscape Ecology 27,
932
1451–1463 (2012).
933
101. Barbaro, L. et al. Avian pest control in vineyards is driven by interactions between bird
934
functional diversity and landscape heterogeneity. Journal of Applied Ecology 54, 500–508
935
(2017).
936
102. Charbonnier, Y. M. et al. Bat and bird diversity along independent gradients of latitude and
937
tree composition in European forests. Oecologia 182, 529–537 (2016).
938
103. Farneda, F. Z. et al. Trait-related responses to habitat fragmentation in Amazonian bats.
939
Journal of Applied Ecology 52, 1381–1391 (2015).
940
104. Robinson, N., Kadlec, T., Bowers, M. D. & Guralnick, R. P. Integrating species traits and
941
habitat characteristics into models of butterfly diversity in a fragmented ecosystem.
942
Ecological Modelling 281, 15–25 (2014).
943
105. Rachello-Dolmen, P. G. & Cleary, D. F. R. R. Relating coral species traits to
944
environmental conditions in the Jakarta Bay/Pulau Seribu reef system, Indonesia. Estuarine,
945
Coastal and Shelf Science 73, 816–826 (2007).
946
106. Cleary, D. F. R. R. & Renema, W. Relating species traits of foraminifera to environmental
947
variables in the Spermonde Archipelago, Indonesia. Marine Ecology Progress Series 334,
948
73–82 (2007).
949
107. Leathwick, J. R., Elith, J. & Hastie, T. Comparative performance of generalized additive
950
models and multivariate adaptive regression splines for statistical modelling of species
951
distributions. Ecological Modelling 199, 188–196 (2006).
952
108. R Core Team. R: A Language and Environment for Statistical Computing. (2020).
953
109. Cardoso, P., Mammola, S., Rigal, F. & Carvalho, J. C. BAT: Biodiversity Assessment
954
Tools. (2020).
955
110. Cardoso, P. et al. Partitioning taxon, phylogenetic and functional beta diversity into
956
replacement and richness difference components. Journal of Biogeography 41, 749–761
957
(2014).
958
111. Podani, J., Ricotta, C. & Schmera, D. A general framework for analyzing beta diversity,
959
nestedness and related community-level phenomena based on abundance data. Ecological
960
Complexity 15, 52–61 (2013).
961
112. Gower, J. C. A General Coefficient of Similarity and Some of Its Properties. Biometrics
962
27, 857 (1971).
963
113. Laliberté, E. & Legendre, P. A distance-based framework for measuring functional
964
diversity from multiple traits. Ecology 91, 299–305 (2010).
965
114. Laliberté, E., Legendre, P. & Shipley, B. FD: measuring functional diversity from multiple
966
traits, and other tools for functional ecology. (2014).
967
115. Podani, J. Extending Gower’s general coefficient of similarity to ordinal characters. Taxon
968
48, 331–340 (1999).
969
116. Declerck, S. A. J. J., Coronel, J. S., Legendre, P. & Brendonck, L. Scale dependency of
970
processes structuring metacommunities of cladocerans in temporary pools of High-Andes
971
wetlands. Ecography 34, 296–305 (2011).
972
117. Vavrek, M. J. fossil: palaeoecological and palaeogeographical analysis tools.
973
Palaeontologia Electronica 14, 1T (2011).
974
118. Oksanen, J. et al. vegan: Community Ecology Package. (2019).
975
119. Millar, R. B., Anderson, M. J. & Tolimieri, N. Much ado about nothings: using zero
976
similarity points in distance-decay curves. Ecology vol. 92 (2011).
977
120. Latombe, G., Hui, C. & McGeoch, M. A. Multi-site generalised dissimilarity modelling:
978
using zeta diversity to differentiate drivers of turnover in rare and widespread species.
979
Methods in Ecology and Evolution 8, 431–442 (2017).
980
121. Mullen, K. M. & van Stokkum, I. H. M. nnls: The Lawson-Hanson algorithm for non-
981
negative least squares (NNLS). (2012).
982
122. Lawson, C. L. & Hanson, R. J. Solving Least Squares Problems. Solving Least Squares
983
Problems (1995). doi:10.1137/1.9781611971217.
984
123. Marschner, I. C. glm2: Fitting generalized linear models with convergence problems. The
985
R Journal 3, 12–15 (2011).
986
124. Latombe, G., McGeoch, M. A., Nipperess, D. A. & Hui, C. zetadiv: Functions to Compute
987
Compositional Turnover Using Zeta Diversity. (2020).
988
125. Peters, R. H. The Ecological Implications of Body Size. (Cambridge University Press,
989
1983). doi:10.1017/cbo9780511608551.
990
126. Blonder, B., Lamanna, C., Violle, C. & Enquist, B. J. The n-dimensional hypervolume.
991
Global Ecology and Biogeography 23, (2014).
992
127. Blonder, B. Hypervolume concepts in niche- and trait-based ecology. Ecography 41, 1441–
993
1455 (2018).
994
128. Gower, J. C. Some Distance Properties of Latent Root and Vector Methods Used in
995
Multivariate Analysis. Biometrika 53, (1966).
996
129. Mammola, S. & Cardoso, P. Functional diversity metrics using kernel density n-
997
dimensional hypervolumes. Methods in Ecology and Evolution 11, (2020).
998
130. Blonder, B. & Harris, D. J. hypervolume: High Dimensional Geometry and Set Operations
999
Using Kernel Density Estimation, Support Vector Machines, and Convex Hulls. (2019).
1000
131. Podani, J. & Schmera, D. On dendrogram-based measures of functional diversity. Oikos
1001
115, 179–185 (2006).
1002
132. Maire, E., Grenouillet, G., Brosse, S. & Villéger, S. How many dimensions are needed to
1003
accurately assess functional diversity? A pragmatic approach for assessing the quality of
1004
functional spaces. Global Ecology and Biogeography 24, 728–740 (2015).
1005
133. Friedman, J. H. Greedy function approximation: a gradient boosting machine. Annals of
1006
statistics 1189–1232 (2001).
1007
134. Elith, J., Leathwick, J. R. & Hastie, T. A working guide to boosted regression trees.
1008
Journal of Animal Ecology 77, 802–813 (2008).
1009
135. Hastie, T., Tibshirani, R. & Friedman, J. The elements of statistical learning: data mining,
1010
inference, and prediction. (Springer Science & Business Media, 2009).
1011
136. Hijmans, R. J., Phillips, S., Leathwick, J. & Elith, J. dismo: Species Distribution Modeling.
1012
(2017).
1013
137. Carvalho, J. C., Cardoso, P. & Gomes, P. Determining the relative roles of species
1014
replacement and species richness differences in generating beta-diversity patterns. Global
1015
Ecology and Biogeography 21, 760–771 (2012).
1016
138. Wickham, H. et al. Welcome to the Tidyverse. Journal of Open Source Software 4, 1686
1017
(2019).
1018
139. Zhu, L. et al. Trait choice profoundly affected the ecological conclusions drawn from
1019
functional diversity measures. Scientific reports 7, 3643 (2017).
1020
140. Wong, M. K. L. & Carmona, C. P. Including intraspecific trait variability to avoid
1021
distortion of functional diversity and ecological inference: lessons from natural assemblages.
1022
bioRxiv 2020.09.17.302349 (2020) doi:10.1101/2020.09.17.302349.
1023
141. Siefert, A., Ravenscroft, C., Weiser, M. D. & Swenson, N. G. Functional beta-diversity
1024
patterns reveal deterministic community assembly processes in eastern North American trees.
1025
Global Ecology and Biogeography 22, 682–691 (2013).
1026
142. Kruk, C. et al. Phytoplankton community composition can be predicted best in terms of
1027
morphological groups. Limnology and Oceanography 56, 110–118 (2011).
1028
143. Soininen, J., Heino, J. & Wang, J. A meta-analysis of nestedness and turnover components
1029
of beta diversity across organisms and ecosystems. Global Ecology and Biogeography 27,
1030
96–109 (2018).
1031
144. Stevens, G. C. The Latitudinal Gradient in Geographical Range: How so Many Species
1032
Coexist in the Tropics. Source: The American Naturalist vol. 133 (1989).
1033
145. Jamoneau, A., Chabrerie, O., Closset-Kopp, D. & Decocq, G. Fragmentation alters beta-
1034
diversity patterns of habitat specialists within forest metacommunities. Ecography 35, 124–
1035
133 (2012).
1036
146. Edie, S. M., Jablonski, D. & Valentine, J. W. Contrasting responses of functional diversity
1037
to major losses in taxonomic diversity. Proceedings of the National Academy of Sciences of
1038
the United States of America 115, 732–737 (2018).
1039
147. Castro-Insua, A., Gómez-Rodríguez, C. & Baselga, A. Break the pattern: breakpoints in
1040
beta diversity of vertebrates are general across clades and suggest common historical causes.
1041
Global Ecology and Biogeography 25, 1279–1283 (2016).
1042
148. Zagmajster, M. et al. Geographic variation in range size and beta diversity of groundwater
1043
crustaceans: Insights from habitats with low thermal seasonality. Global Ecology and
1044
Biogeography 23, 1135–1145 (2014).
1045
149. Bierne, N., Bonhomme, F. & David, P. Habitat preference and the marine-speciation
1046
paradox. Proceedings of the Royal Society of London. Series B: Biological Sciences 270,
1047
1399–1406 (2003).
1048
150. Shurin, J. B., Gruner, D. S. & Hillebrand, H. Review All wet or dried up? Real differences
1049
between aquatic and terrestrial food webs. Proceedings of the Royal Society B: Biological
1050
Sciences 273, 1–9 (2006).
1051
151. Mouquet, N. & Loreau, M. Community Patterns in Source‐Sink Metacommunities. The
1052
American Naturalist 162, 544–557 (2003).
1053
152. Heim, N. A. et al. Hierarchical complexity and the size limits of life. Proceedings.
1054
Biological sciences 284, 20171039 (2017).
1055
153. Hodgson, J. G., Wilson, P. J., Hunt, R., Grime, J. P. & Thompson, K. Allocating C-S-R
1056
Plant Functional Types: A Soft Approach to a Hard Problem. Oikos 85, 282 (1999).
1057
154. Martínez, A. et al. Habitat differences filter functional diversity of low dispersive
1058
microscopic animals 1 2. bioRxiv 2020.09.22.308353 (2020)
1059
doi:10.1101/2020.09.22.308353.
1060
155. Stegen, J. C. et al. Quantifying community assembly processes and identifying features
1061
that impose them. The ISME journal 7, 2069–2079 (2013).
1062
156. Titley, M. A., Snaddon, J. L. & Turner, E. C. Scientific research on animal biodiversity is
1063
systematically biased towards vertebrates and temperate regions. PLOS ONE 12, e0189577
1064
(2017).
1065
1066
Figures
1067
1068
Figure 1. (a) Taxonomic and functional distance decay. Two scenarios of distance decay of
1069
taxonomic and functional similarities along spatial and environmental distances. In scenario 1 (for
1070
simplicity, we consider here replacement only), the replacement occurs among species that have
1071
different traits (i.e., colours), which leads to both taxonomic and functional distance decay. In
1072
scenario 2, the replacement occurs among species that have similar traits, which leads to zero
1073
functional distance decay measured by the slope. (b) Master hypothesis: spatial distance decay is
1074
stronger for taxonomic similarities than for functional similarities, while environmental distance
1075
decay is stronger for functional similarities. (c) Specific hypotheses (higher values indicate steeper
1076
slopes) across datasets: Latitude: spatial distance decay is flatter in the datasets from higher latitude
1077
and more notably for taxonomic similarities than for functional similarities. Environmental distance
1078
decay is steeper in datasets from higher latitude for functional similarities, while it does not vary
1079
notably with latitude for taxonomic similarities. Spatial extent: Both taxonomic and functional
1080
spatial distance decay are flatter in the datasets covering larger spatial extent, while environmental
1081
distance decay is steeper in datasets covering larger extent. Realm: Marine ecosystems show flatter
1082
spatial and environmental distance decay than terrestrial and freshwater systems. FRE= freshwater
1083
systems, TER = terrestrial systems, MAR = marine systems.
1084
1085
Figure 2. Study design highlighting (a) map of the study sites coloured according to the realms
1086
(FRE = Freshwater, TER = Terrestrial, MAR = Marine); (b) the number of data sets for major biotic
1087
groups; and (c) the distribution of the datasets with respect to spatial extent, number of study sites,
1088
functional g-diversity (log hypervolume sd7), taxonomic g-diversity (number of species), number of
1089
environmental variables, and latitude.
1090
1091
Figure 3. The analytical framework described step-wisely. The blocks a-c hierarchically describe
1092
the methods performed at dataset level, including the estimation of similarities and distances as well
1093
as the distance decay models of each dataset. The block d describes the tests performed after the
1094
compilation of the metrics from all datasets. The first block (a) shows the four objects used in the
1095
analyses: a species-by-traits table, a sites-by-species matrix, a sites-by-coordinates table and a sites-
1096
by-environment table. The second block (b) illustrates the calculation of taxonomic and functional
1097
similarities, and spatial and environmental distances. In the first example, only species identities are
1098
taken into account and as sites i and j do not share any species, community similarity (blue) equals
1099
zero. In the second example, sites i and j do not share any species, but as two species have same
1100
body size, community similarity (orange) is higher than zero. Similarity is estimated using the
1101
length of the edge of the dendrograms as S = 1-[(b+c)/(2a+b+c)]. The third example shows how
1102
spatial distances were calculated as the geographic distances among sites using spatial coordinates.
1103
The fourth example illustrates how sites far from each other may show similar environmental
1104
conditions and therefore small environmental distance. Environmental distances were calculated as
1105
the Euclidean distances of standardized environmental variables. The third block (c) illustrates the
1106
metrics extracted to study the distance decay across datasets. The strength (r²) and rate (slope) of
1107
decay were extracted from each dataset using log-binomial generalized linear models (GLM). The
1108
models were built separately for each response variable (taxonomic or functional similarity) and
1109
explanatory variables (spatial or environmental distance), totalling four r² values and four slopes.
1110
Also, the data of marine fish from the Mediterranean Sea is shown as an example where the
1111
distance decay of similarity along environmental distance is stronger (higher r²) for functional
1112
similarity than for taxonomic similarity, irrespectively of the rate of decay (slope). The fourth block
1113
(d) describes the analyses used to test the hypotheses and which metrics were considered for each
1114
analysis. The strength (r²) of decay was used to test hypothesis H1 while the rate of decay (slope)
1115
was used to hypotheses H2-H4.
1116
1117
Figure 4. The distance decay along (a) spatial distance, and (b) environmental distance. Each line in
1118
the panels of left and middle columns shows the shape of the distance decay of an individual
1119
dataset. The mean and standard deviation of slopes are given in the plots. The blue lines show the
1120
distance decay of taxonomic similarity while the orange lines show the distance decay of functional
1121
similarity. The panels on the right column show the strength of the distance decay of taxonomic (y-
1122
axis) and functional (x-axis) similarity. The 1:1 line marks the equivalence of r² between taxonomic
1123
and functional similarities. The dots below the line indicate a dataset with stronger decay of
1124
functional than taxonomic similarity, whereas circles above the line indicates stronger decay of
1125
taxonomic than functional similarities.
1126
1127
Figure 5. The average rate of decay of biotic groups using occurrence data along spatial and
1128
environmental distance. The vertical dotted lines highlight the zero rate (absence of decay) and the
1129
horizontal lines indicate the standard deviation of the mean. The blue circles show the rate of decay
1130
of taxonomic similarities while the orange circles show the rate of decay of functional similarities.
1131
1132
Figure 6. Relative effects (%) of geographic factors on the rate of decay along spatial (a) and
1133
environmental (b) distance decay of the total component of taxonomic (TAX - blue) and functional
1134
(FUN - orange) similarities using occurrence data across datasets. Partial dependence plots show the
1135
effects of a predictor variable on the response variable after accounting for the average effects of all
1136
other variables in the model. Semi-transparent lines represent the actual predicted effects; solid lines
1137
represent LOESS fits to predicted values from BRT. We show here only the variables related to the
1138
specific hypotheses, i.e., latitude, spatial extent, and realms (FRE = Freshwater, TER = Terrestrial,
1139
MAR = Marine).
1140
1141
Figure 7. Relative effects (%) of organismal variables and dataset features on the rate of decay
1142
along spatial (a) and environmental (b) distance considering the total component of taxonomic (blue
1143
lines) and functional (orange lines) similarities using occurrence data across datasets. Partial
1144
dependence plots show the effects of a predictor variable on the response variable after accounting
1145
for the average effects of all other variables in the model. Semi-transparent lines represent the actual
1146
predicted effects; solid lines represent LOESS fits to predicted values from BRT. We show here the
1147
organismal variables and the variables related to the dataset features.
1148
... Whilst our ndings highlight spatiotemporal shifts of communities, ecological communities are undergoing compositional and functional homogenization globally 28-30 . As biotic homogenization proceeds, the dissimilarity decreases among different communities across space 31,32 but increases between observations of the same community over time 28 . Indeed, we found cases of high dissimilarities when comparing bird communities in some grid cells in the 1980's to all grid cells in the 2010's, suggesting that those community compositions observed in the 1980's had no corresponding compositions in the 2010's. ...
Preprint
Full-text available
Species’ range shifts and local extinctions caused by global change lead to community composition changes. At large spatial scales, ecological barriers, such as biome boundaries, coastlines, elevation, and temperature gradients, can influence a community's ability to shift. Yet, ecological barriers are rarely considered in global change studies, potentially hindering predictions of biodiversity shifts. We used data from two consecutive European breeding bird atlases to calculate the geographic distance and direction between communities in the 1980's and their nearest compositional equivalent in the 2010’s and modelled their response to barriers. The ecological barriers affected both the distance and direction of bird community composition shifts, with coasts and elevation having the strongest influence. Combining ecological barriers and community shift projections can identify ecological corridors that facilitate shifts of species and communities under global change.
... Stabilising competition is often detected in plant communities (Adler et al., 2018), suggesting our local stabilisation model may provide insight into the roles of seed banks for spatial diversity patterns. For example, terrestrial plant communities tend to exhibit high betadiversity relative to other systems (Graco-Roza et al., 2021), matching model predictions for low to moderate dispersal and moderate survival in the seed bank. Plant systems have been shown to experience dispersal limitation (Myers & Harms, 2009;Tilman, 1997), and separately beta-diversity has been shown to be high for organisms with seed dispersal (Soininen et al., 2007). ...
Article
Full-text available
Dispersal and dormancy are two common strategies allowing for species persistence and the maintenance of biodiversity in variable environments. However, theory and empirical tests of spatial diversity patterns tend to examine either mechanism in isolation. Here, we developed a stochastic, spatially explicit metacommunity model incorporating seed banks with varying germination and survival rates. We found that dormancy and dispersal had interactive, nonlinear effects on the maintenance and distribution of metacommunity diversity. Seed banks promoted local diversity when seed survival was high and maintained regional diversity through interactions with dispersal. The benefits of seed banks for regional diversity were largest when dispersal was high or intermediate, depending on whether local competition was equal or stabilising. Our study shows that classic predictions for how dispersal affects metacommunity diversity can be strongly influenced by dormancy. Together, these results emphasise the need to consider both temporal and spatial processes when predicting multi-scale patterns of diversity.
... These include questions such as what ecological processes and ecosystem services we lose if a species goes extinct or a community change (Cadotte et al., 2011;Cooke et al., 2019), the filtering effect of a given habitat on species traits (Martínez et al., 2021;Micó et al., 2020), and how does the functionality of a community vary through seasons (Rocha et al., 2012). At a broader, macroecological scale, many studies have explored variations in functional richness along gradients of thermal seasonality (latitude; Graco-Roza et al., 2021;Lamanna et al., 2014;Schumm et al., 2019), glacier cover (Brown et al., 2018) or urbanisation (Buchholz et al., 2020;Sol et al., 2020). Mapping the richness of traits can also serve to identify areas of conservation priority based on criteria beyond species richness (Brum et al., 2017;Strecker et al., 2011). ...
Article
Full-text available
The use of functional diversity analyses in ecology has grown exponentially over the past two decades, broadening our understanding of biological diversity and its change across space and time. Virtually all ecological sub‐disciplines recognize the critical value of looking at species and communities from a functional perspective, and this has led to a proliferation of methods for estimating contrasting dimensions of functional diversity. Differences between these methods and their development generated terminological inconsistencies and confusion about the selection of the most appropriate approach for addressing any particular ecological question, hampering the potential for comparative studies, simulation exercises, and meta‐analyses. Two general mathematical frameworks for estimating functional diversity are prevailing: those based on dissimilarity matrices (e.g., Rao entropy, functional dendrograms) and those relying on multidimensional spaces, constructed as either convex hulls or probabilistic hypervolumes. We review these frameworks, discuss their strengths and weaknesses, and provide an overview of the main R packages performing these calculations. In parallel, we propose a way for organizing functional diversity metrics in a unified scheme to quantify the richness, divergence, and regularity of species or individuals under each framework. This overview offers a roadmap for confidently approaching functional diversity analyses both theoretically and practically.
Article
Full-text available
River damming reduces the dispersal rates of biota and filter species by altering the local environmental conditions. Although phytoplankton is a key biotic group to indicate changes in aquatic environments, the effects of river damming on the longitudinal distribution of phytoplankton communities are not widely addressed. Here we investigated the changes on the taxonomic and functional composition of the phytoplankton community in a large tropical dammed river. We expect compositional changes to be higher at taxonomic than functional level especially in the dam-free stretches of the river because running waters select species able to cope with the water turbulence. Furthermore, flow regulation allows species with different traits to colonize the increasingly lentic stretches resulting in higher functional diversity towards dams. Phytoplankton samples were taken in 210 sites, spaced by 5 km, along the whole river channel. Data were split at a reservoir level considering the sites within 30 km before and after each dam to analyze the effects of damming on the community composition. Compositional changes were assessed by estimating the degree to which each community differed from all other communities within the regional pool (community distinctiveness) considering taxonomic and functional composition separately. Functional diversity was estimated as the degree to which species within the same community differed from each other with respect to their trait composition (functional dispersion). We used community-weighted means to test how the different traits were filtered by river damming. Furthermore, we assessed the effects of local environmental conditions on the phytoplankton taxonomic and functional composition using boosted regression trees (BRT). We observed no effects of river damming on species dispersal. Functional turnover was low, suggesting that compositional changes occurred mainly among functionally similar species. However, there was a noticeable reduction of phytoplankton diversity that persisted for more than 100 km after one of the dams. Flow regulation allowed flagellated and mucilaginous species to colonize the dammed areas. After the dams, such species were filtered due to the increase of water mixing downstream, as evidenced by a hump-shaped trend in the functional diversity of the sites within 30 km distance of the dams. BRT analyses highlighted the role of nutrient input in the dammed areas for the community compositional change. This study highlights the importance of considering both taxonomical and functional diversity to describe the effects of human intervention such as damming on aquatic ecosystems.
Article
Full-text available
Global change is increasing biotic homogenization globally, which modifies the functioning of ecosystems. While tendencies towards taxonomic homogenization in biological communities have been extensively studied, functional homogenization remains an understudied facet of biodiversity. Here, we tested four hypotheses related to long-term changes (1991–2016) in the taxonomic and functional arrangement of freshwater macroinvertebrate assemblages across space and possible drivers of these changes. Using data collected annually at 64 river sites in mainland New Zealand, we related temporal changes in taxonomic and functional spatial β-diversity, and the contribution of individual sites to β-diversity, to a set of global, regional, catchment and reach-scale environmental descriptors. We observed long-term, mostly climate-induced, temporal trends towards taxonomic homogenization but functional differentiation among macroinvertebrate assemblages. These changes were mainly driven by replacements of species and functional traits among assemblages, rather than nested species loss. In addition, there was no difference between the mean rate of change in the taxonomic and functional facets of β-diversity. Climatic processes governed overall population and community changes in these freshwater ecosystems, but were amplified by multiple anthropogenic, topographic and biotic drivers of environmental change, acting widely across the landscape. The functional diversification of communities could potentially provide communities with greater stability, resistance and resilience capacity to environmental change, despite ongoing taxonomic homogenization. Therefore, our study highlights a need to further understand temporal trajectories in both taxonomic and functional components of species communities, which could enable a clearer picture of how biodiversity and ecosystems will respond to future global changes.
Preprint
Full-text available
Whereas the study of patterns of distribution of microscopic animals has long been dominated by the ubiquity paradigm, we are starting to appreciate that microscopic animals are not as widespread as previously thought and that habitat preferences may have a strong role in structuring their patterns of occurrence. However, we still ignore to what extent and through which mechanisms the environment selects for specific communities or traits in microscopic animals. This gap is partly due to the lack of data on the relevant traits of many species, and partly because measuring environmental variables at an appropriate resolution may be problematic. We here overcome both issues by analysing the functional space of marine mite communities living in a sea-grass ( Posidonia oceanica ) meadow across two habitats: the leaves and the matte. The strictly benthic lifestyle and the conserved morphology of mites allow for unambiguous characterization of their functional traits, while the discrete nature of the two habitats alleviates the uncertainty in their ecological characterization. Our results show that habitat filters the distribution of certain traits favouring a higher diversity, dispersion, and evenness of functional traits in the matte than in the leaves. We further observed temporal variations in the functional diversity of communities, potentially following the seasonal renovation and decay of seagrass leaves. However, in spite of the stark ecological differences between the two habitats and across seasons, the filtering effect is partial and affects mostly relative species abundances. We conclude that in other microscopic organisms, habitat filtering might appear even more subtle especially if they are capable of long distance dispersal or occur in ecological systems where environmental variables vary continuously or fluctuate through time. Our study therefore emphasises the need of moving from a merely taxonomical toward a functional view of ecological studies of microscopic organisms if we want to achieve a mechanistic understanding of their habitat and distribution patterns. Data availability statement Raw data and R script to generate the analyses will be deposited in a public repository upon acceptance.
Preprint
Full-text available
Functional diversity assessments are crucial and increasingly used for understanding ecological processes and managing ecosystems. The functional diversity of a community is assessed by sampling traits at one or more scales (individuals, populations, species) and calculating a summary index of the variation in trait values. However, it remains unclear how the scale at which traits are sampled and the indices used to estimate functional diversity may alter the patterns observed and inferences about ecological processes. For 40 plant and 61 ant communities, we assess functional diversity using six methods – encompassing various mean-based and probabilistic methods – chosen to reflect common scenarios where different levels of detail are available in trait data. We test whether including trait variability at different scales (from individuals to species) alter functional diversity values calculated using volume-based and dissimilarity-based indices, Functional Richness (FRic) and Rao, respectively. We further test whether such effects alter the functional diversity patterns observed across communities and their relationships with environmental drivers such as abiotic gradients and occurrences of invasive species. Intraspecific trait variability strongly determined FRic and Rao. Methods using only species’ mean trait values to calculate FRic (convex hulls) and Rao (Gower-based dissimilarity) distorted the patterns observed when intraspecific trait variability was considered. These distortions generated Type I and Type II errors for the effects of environmental factors structuring the plant and ant communities. The high sensitivity of FRic to individuals with extreme trait values was revealed in comparisons of different probabilistic methods including among-individual and among-population trait variability in functional diversity. By contrast, values and ecological patterns in Rao were consistent among methods including different scales of intraspecific trait variability. Decisions about where traits are sampled and how trait variability is included in functional diversity can drastically change the patterns observed and conclusions about ecological processes. We recommend sampling the traits of multiple individuals per species and capturing their intraspecific trait variability using probabilistic methods. We discuss how intraspecific trait variability can be reasonably estimated and included in functional diversity in the common circumstance where only limited trait data are available.
Article
Full-text available
1. The use of n‐dimensional hypervolumes in trait‐based ecology is rapidly increasing. By representing the functional space of a species or community as a Hutchinsonian niche, the abstract Euclidean space defined by a set of independent axes corresponding to individuals or species traits, these multidimensional techniques show great potential for the advance of functional ecology theory. 2. In the panorama of existing methods for delineating multidimensional spaces, the R package hypervolume [Glob. Ecol. Biogeogr. (2014) 23:595–609] is currently the most used. However, functions for calculating the standard set of functional diversity (FD) indices—richness, divergence, and regularity—have not been developed within the hypervolume framework yet. This gap is delaying its full exploitation in functional ecology, meanwhile preventing the possibility to compare its performance with that of other methods. 3. We develop a set of functions to calculate FD indices based on n‐dimensional hypervolumes, including alpha (richness), beta (and respective components), dispersion, evenness, contribution, and originality. Altogether, these indices provide a coherent framework to explore the primary mathematical components of FD within a multidimensional setting. These new functions can work either with hypervolume objects or raw data (species presence or abundance and their traits) as input data, and are versatile in terms of input parameters and options. 4. These functions are implemented within BAT (Biodiversity Assessment Tools), an R package for biodiversity assessments. As a coherent corpus of functional indices based on a common algorithm, it opens the possibility to fully explore the strengths of the Hutchinsonian niche concept in community ecology research.
Article
Full-text available
Although it has been commonly observed that fungi and bacteria differ in their regional biogeographic patterns, it is not well understood what traits contribute to these different distributions. Here, we evaluate how morphological type (i.e. unicellular or filamentous growth form) influences the biogeography of soil fungal and bacterial communities across not only Euclidean (i.e. geographic) distances, but also across gradients of climate and edaphic factors and plant community composition. Specifically, we assessed the decay in community similarity over distance (distance-decay relationship) for microbes with unicellular and filamentous morphology in both fungi and bacteria across 40 ecologically diverse sampling sites in Minnesota, USA. Overall, we found that while distance-decay relationships were similar in fungal and bacterial communities over Euclidean distances, there were important differences among morphological groups of fungi and bacteria across gradients of environmental and plant community similarity. Specifically, the distance-decay relationship of unicellular fungi and unicellular bacteria were indistinguishable across environmental similarity. However, as plant community similarity decreased, only filamentous fungi and unicellular bacteria differed significantly in the strength of their distance-decay relationships. Like analyses of other study systems, we also found that pH explained much of the variance in community composition across microbial domains and morphological types and that plant community diversity was more closely correlated with fungal diversity than with bacterial diversity. Collectively, our results suggest that specific ecological traits such as morphological type along with microbial domain are key factors shaping the biogeography of microbial communities.
Article
Full-text available
The use of functional information in the form of species traits plays an important role in explaining biodiversity patterns and responses to environmental changes. Although relationships between species composition, their traits, and the environment have been extensively studied on a case-by-case basis, results are variable, and it remains unclear how generalizable these relationships are across ecosystems, taxa and spatial scales. To address this gap, we collated 80 datasets from trait-based studies into a global database for metaCommunity Ecology: Species, Traits, Environment and Space; “CESTES”. Each dataset includes four matrices: species community abundances or presences/absences across multiple sites, species trait information, environmental variables and spatial coordinates of the sampling sites. The CESTES database is a live database: it will be maintained and expanded in the future as new datasets become available. By its harmonized structure, and the diversity of ecosystem types, taxonomic groups, and spatial scales it covers, the CESTES database provides an important opportunity for synthetic trait-based research in community ecology.
Article
Full-text available
Macroecologists seek to identify drivers of community turnover (β-diversity) through broad spatial scales. However, the influence of local habitat features in driving broad-scale β-diversity patterns remains largely untested, owing to the objective challenges of associating local-scale variables to continental-framed datasets. We examined the relative contribution of local- versus broad-scale drivers of conti- nental β-diversity patterns, using a uniquely suited dataset of cave-dwelling spider communities across Europe (35–70° latitude). Generalized dissimilarity model- ling showed that geographical distance, mean annual temperature and size of the karst area in which caves occurred drove most of β-diversity, with differential contributions of each factor according to the level of subter- ranean specialization. Highly specialized communities were mostly influenced by geographical distance, while less specialized communities were mostly driven by mean annual temperature. Conversely, local-scale habitat features turned out to be meaningless predictors of community change, which emphasizes the idea of caves as the human accessible fraction of the extended network of fissures that more properly represents the elective habitat of the subterra- nean fauna. To the extent that the effect of local features turned to be inconspicuous, caves emerge as experimental model systems in which to study broad biological patterns without the confounding effect of local habitat features.
Article
1. Plot‐level redundancy or alpha redundancy is usually defined as the fraction of species diversity not expressed by functional or phylogenetic diversity. Redundancy is zero when all species in one plot are maximally dissimilar from each other. By contrast, redundancy tends to its maximum if the functional or phylogenetic differences between species tend to be minimal. 2. To explore the ecological drivers of community assembly, ecologists also use dissimilarity measures between pairs of plots (a component of beta diversity). Traditional dissimilarity measures summarize compositional differences between pairs of plots based either on species presence and absence data or on species abundances, thus attributing equal distinctiveness between any two species. 3. In the last decades a number of dissimilarity measures which incorporate information on functional or phylogenetic differences among species have been proposed. Based on such improved measures, we can define an index of beta redundancy for a pair of plots as the fraction of species dissimilarity not expressed by functional or phylogenetic dissimilarity. 4. A necessary condition to get a meaningful index of beta redundancy is that for a given pair of plots, the functional or phylogenetic dissimilarity is always lower or equal to the corresponding species dissimilarity. However, many of the existing indices of functional or phylogenetic dissimilarity can lead to values greater than for species dissimilarity. 5. The aim of this paper is thus to introduce a new family of tree‐based measures of phylogenetic and functional dissimilarity that conform to this requirement. To show the behavior of the proposed measures, a worked example with data on Alpine vegetation is used.