ArticlePDF Available

Partial decoupling between exotic fish and habitat constraints remains evident in late invasion stages

Abstract and Figures

We investigated the relationships between exotic freshwater fish invasions, environmental factors and ecofunctional diversity (i.e. the combination of ecological traits in communities) in streams. We used data from 335 stream sites, belonging to 105 watersheds and 3 basins in Italy, to test whether the exotic species invasion was dominated by species with generalist traits and whether the environment-ecofunctional trait relationships of exotic and native species would differ from each other. We also tested the hypothesis that ecofunctional uniqueness patterns between exotic and native species would be substantially different. We found that generalist traits were widespread in nearly all areas where exotic species occurred, but not all generalist traits were equally abundant in exotic communities. Only temperature tolerant, low oxygen tolerant and eurytopic traits were typically more dominant in exotic communities than native ones, suggesting that not all generalist traits are equally important in the invasion process and that more complex mechanisms of trait selection could take place. Environment-ecofunctional trait relationships of exotic and native species partly differed both in direction and magnitude, suggesting that invasion dynamics could decouple the linkage between environment and biotic communities, but also that this decoupling might decrease at later invasion stages (i.e. > 30 years after major invasions). Finally, site and trait ecofunctional uniqueness differed between exotic and native species. Exotic species ecofunctional diversity hotspots were located in human-disturbed areas, suggesting that human disturbance might play a strong role in invasion patterns. We advocate for a wider use of ecofunctional approaches in conservation studies in the future, as they could be a key to understand complex ecological processes such as exotic invasions.
Content may be subject to copyright.
1 3
Aquatic Sciences (2020) 82:14
Partial decoupling betweenexotic sh andhabitat constraints remains
evident inlate invasion stages
MarcoMilardi1,2· AnnaGavioli1· GiuseppeCastaldelli1· JanneSoininen3
Received: 25 March 2019 / Accepted: 6 December 2019
© Springer Nature Switzerland AG 2019
We investigated the relationships between exotic freshwater fish invasions, environmental factors and ecofunctional diversity
(i.e. the combination of ecological traits in communities) in streams. We used data from 335 stream sites, belonging to 105
watersheds and 3 basins in Italy, to test whether the exotic species invasion was dominated by species with generalist traits and
whether the environment-ecofunctional trait relationships of exotic and native species would differ from each other. We also
tested the hypothesis that ecofunctional uniqueness patterns between exotic and native species would be substantially different.
We found that generalist traits were widespread in nearly all areas where exotic species occurred, but not all generalist traits
were equally abundant in exotic communities. Only temperature tolerant, low oxygen tolerant and eurytopic traits were typi-
cally more dominant in exotic communities than native ones, suggesting that not all generalist traits are equally important in
the invasion process and that more complex mechanisms of trait selection could take place. Environment-ecofunctional trait
relationships of exotic and native species partly differed both in direction and magnitude, suggesting that invasion dynamics
could decouplethe linkage between environment and biotic communities, but also that this decoupling might decrease at later
invasion stages (i.e. > 30years after major invasions). Finally, site and trait ecofunctional uniqueness differed between exotic
and native species. Exotic species ecofunctional diversity hotspots were located in human-disturbed areas, suggesting that
human disturbance might play a strong role in invasion patterns. We advocate for a wider use of ecofunctional approaches in
conservation studies in the future, as they could be a key to understand complex ecological processes such as exotic invasions.
Keywords Biodiversity conservation· Alien species· Species diversity· Functional diversity· Ecofunctional uniqueness·
Native species
The study of geographical distribution of species and of
thespatial variation of biodiversity on Earth has a long
history. Recently, hundreds of studies have underlined the
importance of biodiversity for ecosystem functioning and
resilience to different pressures (e.g. Hooper etal. 2005),
as well as its general worldwide decline (Wiens 2016).
Much effort has been devoted to investigate the dispersal
and evolutionary processes driving taxonomical biodiver-
sity in the past (see e.g. Herrera 1995). However, the need
for a shift from taxonomical to trait-based approaches in
biodiversity research has been acknowledged for over two
decades, because of the close linkage between environment
and ecological traits, as well as their fast response to envi-
ronmental changes (see e.g. Poff etal. 2006; Violle etal.
2014). As a consequence, recent works shifted their attention
towards functional diversity, exploring its spatial patterns
and dynamics.
Functional diversity quantifies the distinctive combina-
tion of morphological, biochemical, physiological, struc-
tural, phenological or behavioral traits that characterizes
living communities and has been advocated to measure
Aquatic Sciences
Electronic supplementary material The online version of this
article (https :// 7-019-0688-2) contains
supplementary material, which is available to authorized users.
* Anna Gavioli
1 Department ofLife Sciences andBiotechnology, University
ofFerrara, via Luigi Borsari 46, 44121Ferrara, Italy
2 Present Address: Fisheries New Zealand, Tini aTangaroa,
Ministry forPrimary Industries, Manatū Ahu Matua, 34-38
Bowen Street, Wellington, NewZealand
3 Department ofGeosciences andGeography, University
ofHelsinki, PO Box64, FI-00014Helsinki, Finland
M.Milardi et al.
1 3
14 Page 2 of 14
ecosystem functioning more reliably than taxonomical
diversity (Cadotte etal. 2011). Ecofunctional diversity, a
subset of functional diversity, identifies the combination of
ecological traits in communities, selected by environmental
conditions and habitat availability (Poff and Ward 1990).
Habitat diversity, geography, land use, soil or water chemis-
try might all play a role in determining which ecofunctional
traits occur in a community (Comte etal. 2016). In this per-
spective, the habitat selects which traits prevail in a given
environment, ultimately linking habitat diversity with spe-
cies ecofunctional diversity. This linkage has been explored
to infer ecological quality from ecofunctional diversity of
biotic communities (Milardi and Castaldelli 2018; Mouil-
lot etal. 2013; Pont etal. 2006; Schmutz etal. 2007), but
exotic species invasions have been recognized as a potential
confounding factor for this linkage.
Exotic species introduction is a global phenomenon,
and one of the main causes of biodiversity loss (Leprieur
etal. 2008; Sala etal. 2000) and community homogeniza-
tion (Rahel 2000). A lot of attention has been devoted to
investigate the consequences of exotic invasions on taxo-
nomical diversity at different geographical scales (Sax and
Gaines 2003). A common notion in invasion biology sug-
gests that successful invaders in environmentally degraded
areas could be favored by their tolerance to a broad range
of environmental conditions and their generalist adapta-
tions (Lurgi etal. 2014; Marvier etal. 2004; Milardi etal.
2018b). If most invaders are generalists, invasion dynamics
would override environmental drivers behind species and
trait distributions and cause a peculiar ecofunctional diver-
sity distribution for exotic communities, compared to that
of native ones which had a longer environment-selection
time and thus clear environment-trait relationships. How-
ever, such difference between exotic and native communities
could decrease with time, as each exotic species becomes
progressively more selected by environmental constraints,
after the initial invasion stage.
The overall ecofunctional structure and distribution of
fish has been poorly studied, so far, and little research exists
on the impact of exotic invasions on the ecofunctional struc-
ture of Mediterranean freshwater fish (Colin etal. 2018;
Villéger etal. 2013). As in many other geographical areas,
exotic invasions are one of the main factors causing biodi-
versity loss, but this is particularly true in the Mediterranean
region and its freshwater ecosystems (Crivelli 1995; Cut-
telod etal. 2009; Médail and Quézel 1999). Only recently, an
ecofunctional characterization of native and exotic fish spe-
cies has been defined for someMediterranean areas (Milardi
and Castaldelli 2018), enabling further research on the link-
ages between biological invasions, ecofunctional diversity
and the environment.
In this study, we focused on the northern Italian penin-
sula, previously highlighted as a good area to investigate
the outcomes of freshwater fish invasions (Castaldelli etal.
2013; Gavioli etal. 2018, 2019; Lanzoni etal. 2018; Milardi
etal. 2018a). We used a spatially-comprehensive dataset,
including several river basins, to explore the spatial distri-
bution of riverine freshwater fish ecofunctional traits and
investigate the relationships between environmental factors
and ecofunctional diversity of native and exotic species. We
tested the hypothesis (H1) that the invasion process would
be largely led by less-specialized species, able to adapt to
a wide range of environmental conditions (i.e. generalist
species). Under this hypothesis, we expected that generalist
ecofunctional traits would be widespread and dominant in
exotic species communities of invaded areas (e.g. Evange-
lista etal. 2008). However, we expected this dominance to
decrease at late-invasion stages (i.e. > 30years after major
invasions, as in our study area). We also tested the hypoth-
esis (H2) that the environment-ecofunctional traits relation-
ships of exotic and native species would differ. In particular,
we expected that the relationships would be overall weaker
for exotic species, as native species traits have been selected
from the regional species pool by environmental factors over
a longer time than the residence time of exotic species. We
ultimately investigated the hypothesis (H3) that ecofunc-
tional traits of exotic and native species would have different
spatial patterns (i.e. between upland and lowland streams),
because lowland areas were highlighted as an invasion hot-
spot (Lanzoni etal.2018; Milardi etal. 2018a). Finally, our
results would help understanding the factors shaping the
ecofunctional diversity of fish communities under exotic
invasions, one of the major threats to native fish conserva-
tion worldwide.
Materials andmethods
Study area
We focused our investigation on northern Italy (Fig.1), an
area hosting more than 17 million inhabitants and impacted
by agricultural activities and livestock farming. The study
region has a Mediterranean continental climate, with an
annual average precipitation of 1036mm and a mean tem-
perature of 12°C. Within this area lies the largest river
basin in Italy, the Po River basin (71,000km2), of which
we included the Po River itself (in all its course), the Oglio
River (one of the most important northern tributaries of
the Po River) and the southern tributaries in the Emilia-
Romagna region. We also included two additional groups
of rivers outside of this basin: the Brenta River (north-east
of the Po River basin) and watercourses of the Emilia-
Romagna (south of the Po River basin). Overall, a total of
335 sampling sites (i.e. 335 stretches of 105 watercourses)
were included in this study, covering heterogeneous habitats
Partial decoupling betweenexotic fish andhabitat constraints remains evident inlate invasion…
1 3
Page 3 of 14 14
(e.g. riffles, pools), and different altitudinal zones and envi-
ronmental conditions (Fig.1).
In the uplands, organic material originating from villages,
small towns and livestock farms is the main source of river
pollution. Conversely, a high degree of urbanization and
intensive agriculture characterize the lowland rivers, where
high nutrient loads have led to eutrophication (Castaldelli
etal. 2013). To support agricultural irrigation, a complex
network of drainage canals has been established in the
lowlands; this system is completely human regulated with
hydrological management directed to both irrigation supply
and drainage (Castaldelli etal. 2013; Milardi etal. 2017).
The study area was in a late invasion stage (Milardi etal.
2018a) at the time of sampling, since loss of native species
and major exotic invasion occurred already prior to 1997
(Castaldelli etal. 2013), before the data analyzed here were
collected. See Supplementary Table1 for detailed informa-
tion about the year of introduction of each exotic species.
Data collection
Fish data were collected within monitoring programs of the
Emilia-Romagna region (Regione Emilia Romagna 2008),
the Padova Province (Provincia di Padova 2010), the Po
River (Autorita’ di bacino del Po 2008) and the Oglio River
(Consorzio dell’Oglio 2016). The monitoring programs were
carried out separately, in different years, over a relatively
long-term period (1999–2010).
Fish sampling was performed by electrofishing, adapting
the standard national monitoring guidelines to the particu-
lar conditions of each watercourse, and using direct current
at 400–600V and 4–5 A (Backiel and Welcomme 1980;
Reynolds 1996). River stretches were sampled once, dur-
ing daylight, in an upstream zigzag direction, ensuring that
the range of present macrohabitats of each site was fully
surveyed (Hankin and Reeves 1988). Electrofishing is con-
sidered the best quantitative method for fish sampling in
shallow waters, up to a maximum of 1m (Zalewski and
Cowx 1990) but its efficacy may be low in deeper waters,
with big and mobile specimens, or with high conductivity.
Such special conditions occurred in almost all the lower
stretches of rivers and in the canals of the lowlands. For this
reason, at these sites, electrofishing was immediately fol-
lowed up with a standard set of trammel nets (with variable
mesh size from 90 to 5mm), with the support of professional
fishermen (Backiel and Welcomme 1980). In order to allow
comparisons between sites (i.e. river stretches), the sampling
effort at each sampling site was standardized based on area
sampled, according to the national fish monitoring guide-
lines (APAT 2007).
Fish species were classified according to Kottelat and
Freyhof (2007), taking into account recent taxonomic
determinations and common names as listed in FishBase
(Froese and Pauly 2017). Each species was categorized as
native or exotic: a species was considered as native when
naturally present in Italian watercourses and as exotic when
human-introduced (IUCN Comitato Italiano 2019; Milardi
etal. 2018a), irrespective of the time elapsed since the intro-
duction (see Supplementary Table1 for further details on
introduction dates). Hybrid specimens or uncertain species
were excluded from this study in order to avoid taxonomic
Abundance of each species was expressed with Moyle
classes (Moyle and Nichols 1973) ranging from 1 (lower
abundance, 1–2 individuals per site) to 5 (higher abun-
dance, more than 50 individuals per site). However, numeri-
cal abundance classes tend to overestimate the ecological
significance of small-bodied species and underestimate
that of large-bodied ones. A weight was thus assigned
to each species based on their average size published in
Fig. 1 Distribution of the 335 sampling sites used in this study (green dots), the watercourses and the three main basins considered in northern
M.Milardi et al.
1 3
14 Page 4 of 14
the literature (1 = small body up to 150 g; 2 = medium
body 150–400 g; 3 = large body over 400g) and mul-
tiplied by Moyle abundance classes, in order to obtain a
more ecologically-balanced representation of the commu-
nity, closer to relative biomasses (i.e. body-mass-corrected
abundances, hereafter referred to simply as abundances,
Milardi and Castaldelli 2018).
Water physicochemical sampling was performed with
standard methods in temporal and spatial proximity with the
fish sampling, by different Regional Environmental Protec-
tion Agencies (ARPAs, in Italian) for the Po, the Brenta and
the rivers in the Emilia-Romagna region. The Oglio River
Water Authority carried out the water sampling in the Oglio
River. Eight physicochemical variables were monitored:
water temperature (T; °C), electrical conductivity (EC; μS
cm−1), chemical oxygen demand (COD; O2 mgl−1), bio-
logical oxygen demand (BOD5; O2 mgl−1), total suspended
solids (TSS; mgl−1), total phosphorus (P; mgl−1), ammo-
nia (NH4
+; mgl−1) and nitrate nitrogen (NO3
; mgl−1). The
geographical position (longitude—Long, latitude—Lat) and
the altitude (Alt) were also considered.
Fish ecofunctional traits
In order to investigate the ecofunctional composition of
fish communities, five different ecological functions were
taken into account: feeding, reproduction, migration, toler-
ance and habitat use. These functions were included because
of their ecological significance and close relationship with
the environment (Noble etal. 2007; Milardi and Castaldelli
2018). Within these ecological functions, all fish species
were classified in guilds, each representing an ecofunctional
trait and therefore referred to as such hereafter [Table1, see
also Noble etal. (2007)].
Ecological functions, guilds and classification for most
species in this study were taken from Milardi and Castaldelli
(2018), where all available information was used to identify
appropriate guilds for each species. Guilds were mutually
Table 1 Ecological functions and guilds (and their abbreviations) considered for each species in this study. Edited from Milardi and Castaldelli
Ecological function Guild Abbreviation Description
Feeding Planktivores P Plankton feeders
Herbivores H Vegetation feeders
Benthivores B Bottom feeders
Invertivores I Invertebrate feeders
Piscivores Pi Fish feeders
Parasite Pa Ematophages
Generalists G Unspecialized feeding
Reproduction Lithophils Lp Spawning on stones or gravel
Phytophils Pp Spawning on submerged vegetation
Phytolithophils PL Spawning on stones or vegetation
Psammophils PS Spawning on sand or mud
Ostracophils O Spawning in molluscs
Pelagophils or live breeding LB Pelagic spawners or live spawners
Polyphils Pps Generalist spawners
Sea spawning Sea Saltwater spawners
Migration Short S Within or close to the site
Intermediate M Up and downstream or into flooded areas
Long L Anadromous and catadromous species
Tolerance Low oxygen tolerants OxT Tolerance/intolerance to low oxygen
(indicatively below 3ppm)
Low oxygen intolerants OxInt
High temperature tolerants HTT Tolerance/intolerance to high temperature
(indicatively above 20°C)
High temperature intolerants HTInt
Habitat use Rheophiles Rhe Preferring fast flowing water
Limnophiles Lim Preferring slow or no current
Eurytopics Eur Having no preference on current velocity
Clear water CW Clear water adapted
Turbid water TW Turbid water adapted
Wide range of conditions WR Adapted to a wide range of water turbidity
Partial decoupling betweenexotic fish andhabitat constraints remains evident inlate invasion…
1 3
Page 5 of 14 14
exclusive (i.e. each species could be assigned to only one
guild within each ecological function) and species were
classified according to their prevalent adult characteristics
(see also Table1 in Milardi and Castaldelli 2018 for a more
detailed explanation), thus partly discounting ontogenetic
shifts (e.g. in diet or habitat) and ecological plasticity. The
same methodology was applied to classify euryhaline spe-
cies that were not included in the previous work (see Sup-
plementary Table1).
Spatial distribution ofexotic species generalist
We focused our analysis on generalist ecofunctional traits,
with the aim to verify their spatial and community domi-
nance within the exotic communities, therefore selecting all
areas with exotic species presence. Additionally the patterns
observed for exotic communities were compared with those
found in native communities over the same areas, to investi-
gate any differences in dominance of generalist traits.
Inside each ecological functions, the ecological guilds
indicative of no preference for, or no adaptation to, specific
environmental conditions (i.e. generalist traits) were selected
(ecological function—ecological guild): Feeding—Gener-
alists (feeding on any source), Reproduction—Polyphils
(spawning on any substrate), Tolerance—Low oxygen toler-
ants and High temperature tolerants (not selective for oxygen
or temperature), Habitat use—Eurytopics (adapted to a wide
range of current velocities) and Wide range of turbidity con-
ditions (idem for turbidity).
To assess the spatial distribution of generalist ecofunc-
tional traits, the number of sites in which generalist traits
were detected was expressed as a share (percentage) of the
total sites with exotic species presence (Table2). To assess
the community dominance of generalist ecofunctional traits,
the share (percentage) of these traits in the communities was
calculated using the abundances of all species (exotic and
native) possessing each generalist trait (Table2). Both dis-
tribution and dominance were also calculated separately for
native and exotic communities, at each site (Table2).
The community dominance of generalist traits of exotic
and native species communities was also investigated in
the area where exotic and native species distributions over-
lapped, testing the null hypothesis that exotic and native
communities would not differ in generalist trait dominance.
Differences in generalist trait dominance were tested using
the paired t test (paired-t, a parametric paired test) and the
Wilcoxon signed-rank test (Wilcoxon, a non-parametric
paired test), which investigate differences in the mean and
median, respectively. Testing was performed using the PAST
3.06 software (Hammer etal. 2001).
The ArcGIS software (ESRI 2011) was used to map the
spatial distribution of ecofunctional traits in exotic commu-
nities, and to investigate the spatial autocorrelation of each
generalist trait, using the Spatial Autocorrelation (Global
Moran’s I) tool. The null hypothesis of the Global Moran’s
I analysis is that the attribute being analyzed is randomly
distributed among the features in the study area. Being an
inferential statistic, positive (and significant) values of the
Global Moran’s I indicate spatial clustering of a generalist
Exotic andnative species
environment‑ecofunctional traits relationships
Land cover data were obtained from the CORINE database
(2012, https ://www.eea.europ
nicus -land-monit oring -servi ce-corin e). In the lowlands,
where estimation of watershed areas is more difficult due to
low slopes and human-regulated flow, the land cover of the
whole river basin or of the administrative province was used.
CORINE land cover classes were merged into five categories
based on the main land use in order to better describe the
study area: urban use (Urban), agricultural use (Agri), forest
(Forest), other natural area (OtherNat), freshwater (Fresh)
and brackish water (Brack). Land cover was expressed as
the share of each of these categories in the watershed of
each site.
Relationships between geographical variables, land
use features, water physico-chemical variables and fish
Table 2 Distribution and average abundance of generalist ecofunctional traits in the area where native and exotic species distribution overlapped
Ecological function Trait Exotic community Native community
Distribution (% sites
with trait presence)
Average dominance
(% of fish community)
Distribution (% sites
with trait presence)
Average dominance
(% of fish community)
Feeding Generalists 80.9 18.6 67.0 22.0
Reproduction Polyphils 31.1 4.9 29.2 8.3
Tolerance Low oxygen tolerants 98.1 95.9 84.2 53.3
High temperature tolerants 97.6 94.0 84.2 49.4
Habitat use Eurytopics 94.7 71.0 91.4 65.7
Wide range of turbidity 79.4 25.3 94.3 71.9
M.Milardi et al.
1 3
14 Page 6 of 14
0306090 120
Abundance classes
1 - 11
12 - 23
24 - 35
36 - 47
48 - 55
Habitat use
Exotic abundance
Exotic abundance
Ecological functions
Generalists Polyphils
Low oxygen tolerants High temperature
Wide range of
water turbidity
(d) (e)
(f) (g)
Partial decoupling betweenexotic fish andhabitat constraints remains evident inlate invasion…
1 3
Page 7 of 14 14
ecofunctional traits were explored using the distance-based
Redundancy Analysis (db-RDA, Legendre and Anderson
1999). This analysis is normally used to analyze species-
by-site data, but here it was used to analyze trait-by-site
data (i.e. the total body-size-corrected abundance of all
species with a certain ecological trait at each site). One of
the advantages of db-RDA is that it can be used with any
non-Euclidean distance measures (Legendre and Anderson
1999). Therefore, taking into account the work of de Bello
etal. (2013), Gower distance measures were used because of
their better suitability in detecting changes in ecofunctional
diversity along environmental gradients. Linear dependen-
cies between variables were assessed through the Variance
Inflation Factors (VIF) and variables showing high collin-
earity (VIF > 10) were removed from the analysis (Borcard
etal. 2011). Explained variance in db-RDA was reported
through adjusted r2 values (correct constrained variance).
Matrices of traits-by-site data were Hellinger transformed
(Legendre and De Cáceres 2013; Legendre and Gallagher
2001) to standardize variations among both species and
community size. Environmental variables expressed as per-
centages were arcsine transformed, while all the other ones
were log-transformed.
These analyses were performed in R software version
3.4.3 (R Core Team 2017): db-RDA analysis used the cap-
scale function within the ‘vegan’ R package (Oksanen etal.
2017) on all species, and also separately for native and exotic
Ecofunctional uniqueness patterns ofnative
andexotic species
Stream order for each site was calculated using a Digital Ele-
vation Model (DEM) (http://www.sinan et.ispra mbien
sia-ispra /downl oad-mais/dem20 /view), resampled into 10m
pixel size to harmonize it. Flow direction and accumulation,
as well as the watershed of each sampling site, were calcu-
lated based on the DEM layer. For the entire river network
generated by flow accumulation, stream order was derived
with the Strahler method (Strahler 1957). This procedure
was reliable for upland streams, but it was less so in the
lowlands, possibly due to the fact that flow direction and
magnitude in the lowlands are not always natural because
of human intervention. The stream order was thus manu-
ally checked and revised when necessary in lowland rivers
and streams. Rivers were grouped into four classes based on
stream order: class 1 (Strahler stream order 1 and 2), class
2 (stream order 3 and 4), class 3 (stream order 5 and 6) and
class 4 (stream order > 6). As drainage and irrigation canals
could not be assigned into any natural class, a separate class
called “Canals” was created. Canals are man-made environ-
ments, usually characterized by low habitat heterogeneity
and controlled hydrology, located in the lowlands southwest
of the Po River, near its delta.
The ecofunctional uniqueness of communities across sites
was investigated with the Local Contribution to Beta Diver-
sity (LCBD) and Species Contribution to Beta Diversity
(SCBD) approaches, developed by Legendre and De Cáceres
(2013), accounting for stream order. These analyses were
originally developed to calculate the total beta diversity from
the total variance of a site, partitioning the total variance into
local contributions to beta diversity (i.e. LCBD), and species
contributions to beta diversity (i.e. SCBD) across the area,
respectively. However, for the first time, the same approach
was applied here to the ecofunctional diversity of communi-
ties: high values of LCBD indicate a unique econfunctional
composition of the local community (arising from a combi-
nation of low ecofunctional diversity, low species richness
and a rare distribution in the sites) and low values of SCBD
highlight those ecofunctional traits that are the most unique
in the data (arising from the uniqueness of traits and their
rare distribution in the dataset).
Uniqueness analyses were performed in R software ver-
sion 3.4.3 (R Core Team 2017) through the ‘vegan’ R pack-
age (Oksanen etal. 2017) and the ‘adespatial’ R package
(Dray etal. 2016) on all species, and also separately for
native and exotic species.
A total of 59 fish species were detected in the study area;
of these, 37 were native and 22 were exotic species. Each
of these species was assigned to guilds within ecological
functions, as defined in Table1 (see also Supplementary
Spatial distribution ofexotic species` generalist
Exotic species were rather widespread in the area (209 out
of 335 sites, 62.4% of the total) and occurred mostly in the
lowlands. Exotic species occasionally reached rather high
abundances (abundance 40 in 12 sites, Fig.2a, and con-
stituting 100% of the fish community in 9 sites). Overall,
generalist traits were widely distributed in the area occupied
by exotics, but not all traits were equally dominant in the
exotic communities. The three most widespread generalist
Fig. 2 Spatial distribution of exotic abundances in the study area (a)
and respective abundances of generalist traits within different ecolog-
ical functions. Feeding–Generalists (b), Reproduction–Polyphils (c),
Tolerance–Low oxygen tolerants (d) and High temperature tolerants
(e), Habitat use–Eurytopics (f) and Wide range of turbidity conditions
M.Milardi et al.
1 3
14 Page 8 of 14
traits were low oxygen tolerance, high temperature tolerance
and eurytopic adaptation to current velocity (98.1%, 97.6%
and 94.7% of the sites, respectively, Fig.2d–f, Table2). The
least widespread traits were polyphily and adaptation to a
wide range of turbidity conditions (31.1% and 79.4% of the
sites, respectively, Fig.2c and g).
Generalist feeder traits were relatively widespread (80.9%
of the sites, Fig.2b), but were a minor component of the
community (18.6% of the total abundance, Table2), on aver-
age. Polyphil traits dominance was also very low, totaling
4.9% of the overall abundance, on average. Generalist traits
were relatively slightly less widespread within the native
species communities present in the same area, with the
exception of adaptations to a wide range of turbidity condi-
tions (94.3% of the sites, Table2). However, rather surpris-
ingly, some generalist traits were in general more dominant
in native communities than exotic ones (H1). Generalists
feeders, polyphils and species adapted to a wide range of
turbidity were all more dominant in native communities than
exotic ones, while high temperature tolerants, low oxygen
tolerants and eurytopics and were less dominant (Table2).
Differences in dominance of generalist traits between
native and exotic communities were statistically significant
for adaptations to a wide range of turbidity, high tempera-
ture tolerance and low oxygen tolerance (paired-t P < 0.01,
Wilcoxon P < 0.01, for all these traits), as well as for polyph-
ily and eurytopic adaptation to current velocity (paired-t
P < 0.05, Wilcoxon P < 0.05), but not for generalist feeders
(paired-t P = 0.19, Wilcoxon P = 0.85). All generalist traits
showed a significant spatial clustering (Global Moran’s
I > 0; P-values < 0.001, for all traits), indicating spatial
Exotic andnative species
environment‑ecofunctional traits relationships
Environmental variables did not show collinearity
(VIF < 10) and they were thus all included in the db-
RDA analysis. The total amount of ecofunctional trait
variance in the communities explained by db-RDA was
41.90% (Fig.3a). The first axis explained 30.13% of the
variance and the second axis explained 4.36% of the vari-
ance (Fig.3a). Most ecofunctional traits were related to
environmental features [e.g. rheophile (Rhe), low oxygen
intolerant (OxInt) and clear water (CW) adapted species at
higher elevations and vice versa), but some traits (e.g. her-
bivore (H) or parasite (Pa) feeding] did not clearly relate
with environmental variables (Fig.3a). Overall, the envi-
ronmental factors most strongly related with ecofunctional
traits were altitude, temperature and stream order class,
which are all closely linked to each other and influenced
both native and exotic species. The db-RDA analysis con-
ducted exclusively on native species (50.12% of variance
explained) showed a clear distribution of ecofunctional
traits (Fig.3b, the first axis explained the 29.43% of the
variance and the second axis explained the 11.63% of the
variance), with clear distinctions related to environmental
-2 -1 01
-1.5 -1.0 -0.5 0.00.5 1.0
CAP1 (30.13% of variance explained)
CAP2 (4.36%
of variance explained
Pi G
Habitat use
-1.5 -1.0 -0.5 0.00.5 1.01.5
-0.5 0.00.5 1.0
CAP1 (29.43% of variance explained)
CAP2 (11.63%
of variance explained
CAP1 (15.23% of variance explained)
CAP2 (7.77%
of variance explained
-1.0 -0.5 0.00.5 1.0
-0.5 0.00.5
(a) All species
(b) Native species
(c) Exotic species
Fig. 3 Plot of distance-based RDA of all species (a), native spe-
cies (b) and exotic species (c) fish ecofunctional traits composition
(colored labels) using geographical variables, land cover classes,
water physico-chemical variables as explanatory variables (black
labels). Fish ecofunctional traits abbreviations are given in Table1
Partial decoupling betweenexotic fish andhabitat constraints remains evident inlate invasion…
1 3
Page 9 of 14 14
features, similar to those found in the general analysis.
Exotic species had an overall lower linkage with environ-
mental variables (35.46% of variance explained) (Fig.3c,
the first axis explained the 15.23% of the variance and
the second axis explained the 7.77% of the variance). In
general, these three analyses revealed similarities along
the first ordination axis in terms of the key environmental
variables (altitude, forests, temperature), whereas results
differed clearly more for the second axis in terms of which
environmental variables drove trait composition.
Ecofunctional uniqueness patterns ofnative
andexotic species
According to the LCBD analysis, the degree of unique-
ness of the communities’ econfunctional composition per
each stream order class showed a clear decreasing pattern
with increasing stream order class (Fig.4a). Communities
in lower stream orders (in the uplands) had high values of
ecofunctional uniqueness, but communities in canals also
showed a surprisingly high degree of uniqueness compared
with other lowland sites (Fig.4a). According to the SCBD
analysis, high ecofunctional uniqueness was associated with
clear water adaptation, intermediate migration and low oxy-
gen tolerance traits (Fig.4b). Herbivorous, planktivorous
and parasitic feeding traits were associated with low unique-
ness (Fig.4b). The econfunctional uniqueness of native com-
munities (LCBD analysis) followed a similar pattern than
that underlined in overall communities (Fig.4c), but a com-
parison with exotic species (Fig.4e) revealed that the latter
contributed the least to uniqueness in higher stream orders
and canals. Agreeing with our hypothesis H3, uniqueness
patterns differed between native and exotic species. Among
native species, litophilic, low oxygen tolerant and clear water
traits contributed the most to uniqueness according to the
SCBD analysis (Fig.4d), while planktivory, pelagic spawn-
ing and parasitism contributed the least. Among exotic spe-
cies, piscivory, phytophily and benthivory were the traits
that contributed the most to uniqueness according to the
SCBD analysis (Fig.4f), while low oxygen tolerance, her-
bivory and pelagic spawning contributed the least.
Our results partly agreed with our initial hypothesis (H1)
that generalist traits were widespread in nearly all areas
with exotic species presence, but not all generalist traits
were equally abundant in the exotic communities. Moreo-
ver, only temperature tolerants, low oxygen tolerants and
eurytopic traits were typically more dominant in exotic
communities than native ones, suggesting that not all
generalist traits are equally determinant in the invasion
process and that more complex mechanisms of trait selec-
tion could be in place. Our analyses underlined strong
linkages between the environment and ecofunctional traits
for both native and exotic species but, as we hypothesized
(H2), environmental variables explained overall a lower
proportion of exotic species traits than of native traits.
This suggests that the linkage between exotic traits and
environmental factors was less evident, i.e. that the envi-
ronment explained less of the trait composition of exotic
species than of native species, even at a late-invasion stage
(i.e. > 30years after major invasions). This suggests that
invasion dynamics could partly override habitat selectiv-
ity resulting in a partial uncoupling of environment and
biotic communities, and that these effects are long-lasting.
Finally, we found some differences in the ecofunctional
uniqueness patterns of native and exotic species and their
respective ecofunctional traits, being in line with our third
hypothesis (H3). Exotic species ecofunctional diversity
hotspots were common in human-disturbed areas, sug-
gesting that human disturbance might play a strong role
in invasion patterns.
Spatial distribution ofexotic species` generalist
It was clear from our results that exotic species presence
was relatively low in upland areas, likely halted by natural
gradients (e.g. temperature and habitat factors) or perhaps
by physical dispersal barriers (Dynesius and Nilsson 1994;
Tockner and Stanford 2002). On the contrary, exotic spe-
cies were widespread in the lowlands, therefore still exert-
ing a significant propagule pressure in the less-invaded
areas in the foothills (Lockwood etal. 2005; Simberloff
2009). Generalist traits were widespread in exotic com-
munities of the lowlands, but not all traits were equally
represented. Furthermore, generalist trait distribution was
not the result of a random process, but different selection
processes operate on different ecological functions (H1).
Some traits related to broad tolerance and flexible habi-
tat use (temperature tolerance, low oxygen tolerance and
adaptation to a wide range of water current conditions)
were dominant among exotic communities (and more
so than in native communities). However, reproductive
polyphily and adaptation to a wide range of turbidity were
more dominant in native communities over the same area,
indicating that perhaps not all generalist traits are equally
relevant to determine invasion success. The timeline of
invasion processes could partly explain this complex
result: early invasion phases could be spearheaded by true
generalists, but in the long run more specialized species
can also be introduced and spread in the area, overriding
M.Milardi et al.
1 3
14 Page 10 of 14
0.002 0.004 0.006 0.008 0.010
Ecofunctional guild contribution to uniqueness
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07
Habitat use
Stream order class
sdliug lanoitcnufocE
seitinummoc evitan fo sseneuqinu lanoitcnufocE
Ecofunctional guild contribution to uniqueness
Stream order class
0.000 0.005 0.010 0.015
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07
sdliug lanoitcnufocE
seitinummoc citoxe fo sseneuqinu lanoitcnufocE
Ecofunctional guild contribution to uniqueness
Stream order class
0.00 0.01 0.02 0.03 0.04
0.00 0.02 0.04 0.06 0.08
sdliug lanoitcnufocE
(a) (b)
(e) (f)
Fig. 4 Ecofunctional uniqueness of fish communities along a stream
order gradient based on Local Contribution to Beta Diversity, LCBD
(a) and contribution of each fish guild to uniqueness based on Spe-
cies Contribution to Beta Diversity, SCBD (b). Uniqueness patterns
were also analyzed separately for native [[LCBD (c), SCBD (d)] and
exotic [LCBD (e), SCBD (f)] species. Guild abbreviations are given
in Table1
Partial decoupling betweenexotic fish andhabitat constraints remains evident inlate invasion…
1 3
Page 11 of 14 14
some of the overall invasion patterns (Nagelkerke etal.
2018). As this area is in a late invasion stage (Milardi
etal. 2018a, 2019), what we observe could be the result
of the complex interactions between species-specific intro-
duction timing and mechanisms of dispersal and interac-
tion with native species, as well as the overall outcome of
longer-term environmental selection of exotic communi-
ties. These mechanisms could not be fully disentangled
through our analysis, and should be further investigated
in future studies, also considering that the spatial scale
covered could influence the results (Taylor etal. 2019).
Exotic andnative species
environment‑ecofunctional traits relationships
Our results confirmed that there are differences in how the
environmental variables influence the distribution of traits for
native and exotic species (H2). The fact that native species eco-
functional traits are more clearly coupled with the environment
confirms our initial hypothesis of a longer selection time. This
tighter coupling of environmental niches was found in other
taxa (Marks and Lechowicz 2005), but very few studies have
dealt with this aspect in fish communities, so far (but see e.g.
Buckwalter etal. 2018). Ecological theory predicts that exotic
species should be equally influenced and selected by the envi-
ronment, eventually, but evidently this selection could take a
much longer time than that elapsed in our study area (Lambri-
nos 2004). The lower influence of the environment on exotic
species ecofunctional traits suggests that invasion dynamics
might still play a role in the observed ecofunctional distribu-
tions (Strayer etal. 2006). Among invasion dynamics, the intro-
duction of new species through fisheries stocking practices and
active dispersal by anglers could confound spatial distribution
patterns and their linkage with the environment, for both exotic
and native species (Hesthagen and Sandlund 2007; Vörösmarty
etal. 2010). Moreover, invasion processes are not driven by
abiotic factors only, and the invasion of one species could facili-
tate the invasions of other species (Simberloff and Von Holle
1999), a mechanism previously suggested to be at play in this
area (Lanzoni etal. 2018; Milardi etal. 2018a). Our results
confirmed that exotic species are less affected by habitat filter-
ing than native species, even at late-invasion stages.
Ecofunctional uniqueness patterns ofnative
andexotic species
Our results (H3) revealed a low ecofunctional trait diversity
in mountain streams, which typically have lower habitat com-
plexity and resource availability but higher habitat quality than
rivers in the lowlands (Aschonitis etal. 2018; Bouska 2018).
This could also be linked to a lower taxonomical diversity in
these areas (Gavioli etal. 2019) as species number is known
to affect uniqueness measures (Legendre 2014). Conversely,
ecofunctional diversity was relatively higher in higher stream
orders and in the lowlands, with the exception of canals,
which had a lower ecofunctional diversity than natural riv-
ers in the same areas. Canals should have low-heterogeneity
habitats, due to their artificial nature, which can partly explain
this result. However, our analysis highlighted that native spe-
cies low ecofunctional diversity played a strong role in shap-
ing this uniqueness pattern. Previous studies have underlined
how freshwater fish invasions have caused a severe decrease
of native taxonomical diversity at the local level, especially
in canals (Castaldelli etal. 2013; Milardi etal. 2018b) and
low species richness could be at the root of the low native
ecofunctional diversity seen in canals. On the contrary, exotic
ecofunctional diversity was highest in the lowlands and par-
ticularly in canals, which could be caused by a high number
of ecofunctionally diverse exotic species. Canals in this area
are indeed hotspots of exotic species diversity (Lanzoni etal.
2018) and host the highest number of exotic species in our
dataset. This is somewhat counterintuitive, given the canals’
simplified habitats, but could be explained by the higher
degree of human disturbance. In the studied canal network,
human intervention continuously causes redistribution of fish
species between different canals, thus likely increasing the
spread and colonization of exotic species (Castaldelli etal.
2013). Our results suggested that exotic invasions might con-
tribute to shape the spatial patterns of ecofunctional diversity
in fish communities (see also Milardi etal. 2019 and Milardi
etal. 2020).
Clear water, rheophilic, intermediate migration, phyto-
phylic, lithophylic and piscivorous species seem to contrib-
ute the most to the overall ecofunctional uniqueness of fish
communities in the area, being typical traits of native species
and most likely belonging to species with intermediate site
occupancy. However, perhaps more interesting are those eco-
functional traits which are rarest in the dataset, as they high-
light the presence of ecofunctionally unique and relatively rare
native marine species (e.g. planktivores), temporarily entering
estuarine and upstream freshwater areas. Rare ecofunctional
traits also highlight the presence of hard-to-detect exotic spe-
cies (e.g. herbivores, such as grass carp, Ctenopharyngodon
idella), which were found only in canals during standard
monitoring, but are also present elsewhere in the Po River
basin (Milardi etal. 2015, 2017). Further investigations are
needed to unravel the exact contribution of species richness,
ecofunctional diversity and its spatial distribution on the eco-
functional uniqueness of communities, as well as the relation-
ship between ecofunctional uniqueness and trait diversity.
M.Milardi et al.
1 3
14 Page 12 of 14
Our results underlined how the decoupling of the environ-
ment-ecofunctional trait relationships of exotic species is
still detectable in late invasions stages, and how invasions
might affect the ecofunctional diversity patterns of fish com-
munities. However, other factors potentially affecting func-
tional diversity (e.g. climate change and habitat loss) should
also be investigated, as they might provide useful insights on
the final outcome of exotic and native species interactions.
Our results also highlighted the importance of extending
the analysis from taxonomical to functional diversity, when
attempting to address the complexity of species interactions
in invaded communities (see e.g. Colin etal. 2018; Godoy
2019; Trivellone etal. 2014). We thus advocate for a wider
use of ecofunctional approaches in the future to evaluate the
consequences of exotic species invasions (Griffiths and Har-
ris 2010; Loiola etal. 2018; Schlaepfer etal. 2011).
Acknowledgements We thank LL.D. V.E. Manduca and Dr. M. Rizzoli
of the Fisheries Bureau of the Emilia-Romagna Region for providing
the Fish Inventories data in the context of a long-term research col-
laboration. The Oglio River Water Authority (Consorzio dell’Oglio, in
Italian) is also acknowledged for providing fish and water quality data
for the Oglio River. We also thank Dr R. Spaggiari and Dr S. Frances-
chini of the Emilia-Romagna Region Environmental Protection Agency
(ARPA-EMR), the Piemonte Region Environmental Protection Agency
(ARPA-Piemonte) and the Veneto Region Environmental Protection
Agency (ARPAV) for providing the water quality database.
Funding No specific funding was received for this work.
APAT (2007) Protocollo di campionamento e analisi della fauna ittica
dei sistemi lotici. ICRAM, Roma
Aschonitis V, Gavioli A, Lanzoni M, Fano E, Feld C, Castaldelli G
(2018) Proposing priorities of intervention for the recovery of
native fish populations using hierarchical ranking of environmen-
tal and exotic species impact. J Environ Manag 210:36–50
Autorita’ di bacino del Po (2008) Monitoraggio dell’ittiofauna e redazi-
one della Carta ittica del Fiume Po. Qualità dell’ittiofauna e del
macrobenthos del fiume Po
Backiel T, Welcomme RL (1980) Guidelines for sampling fish in inland
waters. FAO, Rome, Italy
Borcard D, Gillet F, Legendre P (2011) Numerical ecology with R.
Springer. https ://
Bouska KL (2018) Discontinuities and functional resilience of large
river fish assemblages. Ecosphere 9:e02351
Buckwalter JD, Frimpong EA, Angermeier PL, Barney JN (2018)
Seventy years of stream-fish collections reveal invasions and
native range contractions in an Appalachian (USA) watershed.
Divers Distrib 24:219–232
Cadotte MW, Carscadden K, Mirotchnick N (2011) Beyond species:
functional diversity and the maintenance of ecological processes
and services. J Appl Ecol 48:1079–1087
Castaldelli G, Pluchinotta A, Milardi M, Lanzoni M, Giari L, Rossi
R, Fano EA (2013) Introduction of exotic fish species and
decline of native species in the lower Po basin, north-eastern
Italy. Aquatic Conserv Mar Freshw Ecosyst 23:405–417. https
Colin N, Villéger S, Wilkes M, de Sostoa A, Maceda-Veiga A (2018)
Functional diversity measures revealed impacts of non-native
species and habitat degradation on species-poor freshwater fish
assemblages. Sci Total Environ 625:861–871
Comte L, Cucherousset J, Boulêtreau S, Olden JD (2016) Resource
partitioning and functional diversity of worldwide freshwater
fish communities. Ecosphere 7:e01356
Consorzio dell’Oglio (2016) Progetto di Sperimentazione dei rilasci
per il deflusso minimo vitale nel fiume Oglio sub lacuale: relazi-
one conclusiva 2009–2015. Available athttps ://www .og lio conso menta zione /relaz ione-concl usiva (in italian)
Crivelli A (1995) Are fish introductions a threat to endemic freshwa-
ter fishes in the northern Mediterranean region? Biol Conserv
Cuttelod A, García N, Malak DA, Temple HJ, Katariya V (2009) The
Mediterranean: a biodiversity hotspot under threat Wildlife in
a Changing World–an analysis of the 2008 IUCN Red List of
Threatened Species, p 89
de Bello F, Carmona CP, Mason NW, Sebastià MT, Lepš J (2013)
Which trait dissimilarity for functional diversity: trait means or
trait overlap? J Veg Sci 24:807–819
Dray S, Blanchet G, Borcard D, Clappe S, Guenard G, Jombart T,
Larocque G, Legendre P, Madi N, Wagner HH (2018)adespa-
tial: multivariate multiscale spatial analysis. R package version
0.2-0. https ://CRAN.R-proje ge=adesp atial
Dynesius M, Nilsson C (1994) Fragmentation and flow regulation
of river systems in the northern third of the world. Science
ESRI (2011) ArcGIS Desktop vol 10.1. Environmental Systems
Research Institute, Redlands
Evangelista PH, Kumar S, Stohlgren TJ, Jarnevich CS, Crall AW,
Norman JB III, Barnett DT (2008) Modelling invasion for a
habitat generalist and a specialist plant species. Divers Distrib
Froese R, Pauly D (2017) FishBase. Version 03/2017 World Wide
Web electronic publication, R Froese and D Pauly (eds) Avail-
able at: www fishbase org. Accessed April 2017
Gavioli A, Mancini M, Milardi M, Aschonitis V, Racchetti E, Viaroli
P, Castaldelli G (2018) Exotic species, rather than low flow,
negatively affect native fish in the Oglio River. Res Appl, North-
ern Italy River. https ://
Gavioli A, Milardi M, Castaldelli G, Soininen J (2019) Diversity pat-
terns of native and exotic fish species suggest homogenization
processes, but partly fail to highlight extinction threats. Divers
Distrib 25:983–994
Godoy O (2019) Coexistence theory as a tool to understand biologi-
cal invasions in species interaction networks: implications for
the study of novel ecosystems. Funct Ecol 33:1190–1201. https
Griffiths CJ, Harris S (2010) Prevention of secondary extinctions
through taxon substitution. Conserv Biol 24:645–646
Hammer Ø, Harper D, Ryan P (2001) Past: paleontological statistics
software package for education and data analysis. Palaeontolo-
gia Electronica. http://palae o-elect ronic
Hankin DG, Reeves GH (1988) Estimating total fish abundance and
total habitat area in small streams based on visual estimation
methods. Can J Fish Aquatic Sci 45:834–844
Herrera CM (1995) Dispersal systems in the Mediterranean: ecologi-
cal, evolutionary, and historical determinants. Annu Rev Ecol and
Syst 26:705–727
Hesthagen T, Sandlund O (2007) Non-native freshwater fishes in
Norway: history, consequences and perspectives. J Fish Biol
Partial decoupling betweenexotic fish andhabitat constraints remains evident inlate invasion…
1 3
Page 13 of 14 14
Hooper DU etal (2005) Effects of biodiversity on ecosystem function-
ing: a consensus of current knowledge. Ecol Monogr 75:3–35
IUCN Comitato Italiano (2019) Italian red list of species - Osteich-
thyes. e-ostei chthy es.php. Accessed 3 Apr
Kottelat M, Freyhof J (2007) Handbook of European freshwater fishes.
Publications Kottelat, Cornol
Lambrinos JG (2004) How interactions between ecology and evo-
lution influence contemporary invasion dynamics. Ecology
Lanzoni M, Milardi M, Aschonitis V, Fano EA, Castaldelli G (2018) A
regional fish inventory of inland waters in Northern Italy reveals
the presence of fully exotic fish communities. Eur Zool J 85:1–7.
https :// 263.2017.14153 84
Legendre P (2014) Interpreting the replacement and richness difference
components of beta diversity. Glob Ecol Biogeogr 23:1324–1334
Legendre P, Anderson MJ (1999) Distance-based redundancy analysis:
testing multispecies responses in multifactorial ecological experi-
ments. Ecol Monogr 69:1–24
Legendre P, De Cáceres M (2013) Beta diversity as the variance of
community data: dissimilarity coefficients and partitioning. Ecol
Lett 16:951–963
Legendre P, Gallagher ED (2001) Ecologically meaningful transfor-
mations for ordination of species data. Oecologia 129:271–280
Leprieur F, Beauchard O, Blanchet S, Oberdorff T, Brosse S (2008)
Fish invasions in the world’s river systems: when natural pro-
cesses are blurred by human activities. PLoS Biol 6:e28. https :// al.pbio.00600 28
Lockwood JL, Cassey P, Blackburn T (2005) The role of propagule
pressure in explaining species invasions. Trends Ecol Evol
Loiola PP, de Bello F, Chytrý M, Götzenberger L, Carmona CP,
Pyšek P, Lososová Z (2018) Invaders among locals: alien spe-
cies decrease phylogenetic and functional diversity while increas-
ing dissimilarity among native community members. J Ecol
Lurgi M, Galiana N, López BC, Joppa LN, Montoya JM (2014) Net-
work complexity and species traits mediate the effects of biologi-
cal invasions on dynamic food webs. Front Ecol Evol 2:36
Marks CO, Lechowicz MJ (2005) Alternative designs and the evolution
of functional diversity. Am Nat 167:55–66
Marvier M, Kareiva P, Neubert MG (2004) Habitat destruction, frag-
mentation, and disturbance promote invasion by habitat general-
ists in a multispecies metapopulation. Risk Anal Int J 24:869–878
Médail F, Quézel P (1999) Biodiversity hotspots in the Mediterra-
nean Basin: setting global conservation prioritie. Conserv Biol
Milardi M, Castaldelli G (2018) A novel approach to an ecofunctional
fish index for Mediterranean countries. Ecol Ind 89:376–385.
https :// nd.2018.02.022
Milardi M, Lanzoni M, Kiljunen M, Torniainen J, Castaldelli G
(2015) Natural recruitment contributes to high densities of grass
carp Ctenopharyngodon idella (Valenciennes, 1844) in Western
Europe. Aquat Invasions 10:439–448. https ://
Milardi M, Chapman D, Lanzoni M, Long JM, Castaldelli G (2017)
First evidence of bighead carp wild recruitment in Western
Europe, and its relation to hydrology and temperature. PLoS One
12:e0189517. https :// al.pone.01895 17
Milardi M, Aschonitis V, Gavioli A, Lanzoni M, Fano EA, Castaldelli
G (2018a) Run to the hills: exotic fish invasions and water quality
degradation drive native fish to higher altitudes. Sci Total Environ
624:1325–1335. https :// tenv.2017.12.237
Milardi M, Lanzoni M, Gavioli A, Fano EA, Castaldelli G (2018b)
Long-term fish monitoring underlines a rising tide of temperature
tolerant, rheophilic, benthivore and generalist exotics, irrespective
of hydrological conditions. J Limnol. https ://
jlimn ol.2018.1745
Milardi M, Gavioli A, Soininen J, Castaldelli G (2019) Exotic species
invasions undermine regional functional diversity of freshwater
fish. Sci Rep 9:17921. https :// 8-019-54210
Milardi M, Gavioli A, Soana E, Lanzoni M, Fano EA, Castaldelli
G (2020) The role of species introduction in modifying the
functional diversity of native communities. Sci Total Environ
Mouillot D, Graham NA, Villéger S, Mason NW, Bellwood DR (2013)
A functional approach reveals community responses to distur-
bances. Trends Ecol Evol 28:167–177
Moyle PB, Nichols RD (1973) Ecology of some native and introduced
fishes of the Sierra Nevada foothills in central California. Copeia
Nagelkerke LA, van Onselen E, van Kessel N, Leuven RS (2018)
Functional feeding traits as predictors of invasive success of
alien freshwater fish species using a food-fish model. PLoS One
Noble R, Cowx I, Goffaux D, Kestemont P (2007) Assessing the health
of European rivers using functional ecological guilds of fish com-
munities: standardising species classification and approaches to
metric selection. Fish Manag Ecol 14:381–392
Oksanen J, Blanchet F, Friendly M, Kindt R, Legendre P, Mcglinn D,
Minchin P,O’hara R, Simpson G, Solymos P (2017) Vegan: com-
munity ecology Package,2.4–2
Poff NL, Ward J (1990) Physical habitat template of lotic systems:
recovery in the context of historical pattern of spatiotemporal
heterogeneity. Environ Manag 14:629
Poff NL, Olden JD, Vieira NK, Finn DS, Simmons MP, Kondratieff
BC (2006) Functional trait niches of North American lotic insects:
traits-based ecological applications in light of phylogenetic rela-
tionship. J N Am Benthol Soc 25:730–755
Pont D etal (2006) Assessing river biotic condition at a continen-
tal scale: a European approach using functional metrics and fish
assemblages. J Appl Ecol 43:70–80
Provincia di Padova (2010) Carta ittica: aggiornamento dello stato delle
conoscenze sui popolamenti ittici della Provincia di Padova
Rahel FJ (2000) Homogenization of fish faunas across the United
States. Science 288:854–856
R Core Team (2017) R: a language and environment for statistical
computing. Vienna, Austria; 2014
Regione Emilia Romagna (2008) Carta ittica dell’Emilia Romagna -
Zone A e B. Assessorato Attività Produttive, Sviluppo Economico
e Piano Telematico., Bologna
Reynolds J (1983) Electrofishing. In: Nielsen L, Johnson D (eds) Fish-
eries Techniques. American Fisheries Society, Bethesda Mary-
land, pp 147–163
Sala OE etal (2000) Global biodiversity scenarios for the year 2100.
Science 287:1770–1774
Sax DF, Gaines SD (2003) Species diversity: from global decreases to
local increases. Trends Ecol Evol 18:561–566
Schlaepfer MA, Sax DF, Olden JD (2011) The potential conservation
value of non-native species. Conserv Biol 25:428–437
Schmutz S, Cowx I, Haidvogl G, Pont D (2007) Fish-based methods
for assessing European running waters: a synthesis. Fish Manag
Ecol 14:369–380
Simberloff D (2009) The role of propagule pressure in biological inva-
sions. Annu Rev Ecol Evol Syst 40:81–102
Simberloff D, Von Holle B (1999) Positive interactions of nonindig-
enous species: invasional meltdown? Biol Invasions 1:21–32
Strahler AN (1957) Quantitative analysis of watershed geomorphology.
Eos. Trans Am Geophys Union 38:913–920
M.Milardi et al.
1 3
14 Page 14 of 14
Strayer DL, Eviner VT, Jeschke JM, Pace ML (2006) Understand-
ing the long-term effects of species invasions. Trends Ecol Evol
Taylor CM, Miyazono S, Cheek CA, Edwards RJ, Patiño R (2019) The
spatial scale of homogenisation and differentiation in Chihuahuan
Desert fish assemblages. Freshw Biol 64:222–232
Tockner K, Stanford JA (2002) Riverine flood plains: present state and
future trends. Environ Conserv 29:308–330
Trivellone V, Schoenenberger N, Bellosi B, Jermini M, De Bello F,
Mitchell EA, Moretti M (2014) Indicators for taxonomic and func-
tional aspects of biodiversity in the vineyard agroecosystem of
Southern Switzerland. Biol Conserv 170:103–109
Villéger S, Grenouillet G, Brosse S (2013) Decomposing functional
β-diversity reveals that low functional β-diversity is driven by
low functional turnover in E uropean fish assemblages. Glob Ecol
Biogeogr 22:671–681
Violle C, Reich PB, Pacala SW, Enquist BJ, Kattge J (2014) The emer-
gence and promise of functional biogeography. Proc Natl Acad
Sci 111:13690–13696
Vörösmarty CJ etal (2010) Global threats to human water security and
river biodiversity. Nature 467:555
Wiens JJ (2016) Climate-related local extinctions are already wide-
spread among plant and animal species. PLoS Biol 14:e2001104
Zalewski M, Cowx I (1990) Factors affecting the efficiency of electric
fishing. Fishing with electricity: applications in freshwater fisher-
ies management. Fishing News Books, Oxford, pp 89–111
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
... Furthermore, time elapsed since last introductions was sufficient to analyze distribution patterns after major invasions had already occurred see e.g. 23,48 . Abundance of each species sampled during the monitoring was recorded with Moyle classes (Moyle and Nichols, 1973), which were weighted according to body-size classes in order to obtain a body-mass-corrected abundance, hereafter referred to simply as abundance. ...
Full-text available
We analyzed the large-scale drivers of biological invasions using freshwater fish in a Mediterranean country as a test case, and considering the contribution of single species to the overall invasion pattern. Using Boosted Regression Tree (BRT) models, variation partitioning and Redundancy Analysis (RDA), we found that human factors (especially eutrophication) and climate (especially temperature) were significant drivers of overall invasion. Geography was also relevant in BRT and RDA analysis, both at the overall invasion and the single species level. Only variation partitioning suggested that land use was the second most significant driver group, with considerable overlap between different invasion drivers and only land use and human factors standing out for single effects. There was general accordance both between different analyses, and between invasion outcomes at the overall and the species level, as most invasive species share similar ecological traits and prefer lowland river stretches. Human-mediated eutrophication was the most relevant invasion driver, but the role of geography and climate was at least equally important in explaining freshwater fish invasions. Overall, human factors were less prominent than natural factors in driving the spread and prevalence of invasion, and the species spearheading it.
... Where present, it tends to dominate the community of predator fishes (~ 30% of the whole fish community biomass, M. Milardi, unpublished data). Wels catfish and other introduced fish species are a major problem also for native fish diversity in Italian freshwaters (Milardi et al. 2018(Milardi et al. , 2019a(Milardi et al. , 2020a(Milardi et al. , 2020b(Milardi et al. , 2020c, but to date, little has been done to address this problem. A recent review by Cucherousset et al. (2018) underlined how the species is widespread and abundant also in the rest of its introduced range (e.g. ...
Full-text available
Predatory fish have occasionally been observed preying on birds, sometimes repeatedly, but few studies were able to unravel the overall significance of avian prey in fish diet and the predation impacts on bird populations. We used a control/impact study setup, using a Nature Reserve in northern Italy and a nearby control area, to determine: 1) the contribution of waterbirds to wels catfish diet in the Reserve, 2) the population density of wels catfish in the Reserve and control area and 3) the potential impacts of waterbird depredation by wels catfish on waterbird population trends. Our stable isotope Bayesian mixing model indicated that birds contributed 12.2% (5-27.9%, 50% confidence interval) of the diet of large wels catfish (> 98 cm in total length). Large individuals constituted the majority of the population in the shoreline areas of the reserve in 2013-2019, where the population was stable despite control efforts. Numbers were below detectable levels in the control area. Large wels catfish consumed an average of 224, 148 and 187 kg of birds during the 2019 chick growing period, as estimated through three different bioenergetic models. Compared to the control area, mallard reproductive success was diminished in the Reserve, likely due to higher rates of fish predation, although effects were variable in different years. Overall, our data suggest that high densities of invasive wels catfish might impact waterbird reproductive success through predation on bird chicks, but further studies would be needed to reduce uncertainties related to the intrinsic variability of field ecology data. Our study constitutes a preliminary attempt to assess the potential of introduced wels catfish to affect the conservation value of waterbird protection areas, and should be repeated at broader spatial and temporal scales.
... The influence of habitat quality and diversity on fish functional diversity has rarely been the object of investigation (Barbosa et al., 2020;Milardi et al., 2020a;Teresa and Casatti, 2017), probably due to the lack of any extensive and detailed dataset covering different spatial and structural level of habitat characterization. In this study, we used a single biotic index, based on macroinvertebrates community structure (IBE index), as a proxy for overall habitat quality. ...
Freshwater fish communities are impacted by multiple pressures, determining loss of functional diversity and redundancy. Our aim was to disentangle the roles and relevancies of different pressures in shaping fish communities in small streams of the Po plain (North Italy). Long term trend (1998–2018) of functional diversity of 31 fish communities was assessed and modeled in respect to three potential pressures: temperature increase, intensity of exotic fish invasion, and habitat quality degradation. Ecological traits mostly influenced by the pressures were also identified. Reduction of functional richness mostly due to local extinction or contraction of cold adapted predators, such as salmonids, was linked to increasing temperatures. Warming probably also led to a shift of generalist and dominant species, which became more abundant in streams hosting mixed communities of salmonids and cyprinids, and determined the increase of functional dispersion and uniqueness. Reduction of functional redundancy and increasing functional dispersion were both also related to the introduction of new ecological traits brought by expanding exotic species. Low functional overlap was found among native and exotic species, indicating that the invasion process was mainly controlled by competitive interactions and/or resource opportunism. Functional response to habitat quality was not clearly evident. In conclusion, the impact of temperature increase and exotic species on fish functional diversity was effective, idiosyncratic and mediated by the scale of analysis and by the intensity of pressures.
Full-text available
Exotic species invasions often result in native biodiversity loss, i.e. a lower taxonomic diversity, but current knowledge on invasions effects underlined a potential increase of functional diversity. We thus explored the connections between functional diversity and exotic species invasions, while accounting for their environmental drivers, using a fine-resolution large dataset of Mediterranean stream fish communities. While functional diversity of native and exotic species responded similarly to most environmental constraints, we found significant differences in the effects of altitude and in the different ranking of constraints. These differences suggest that invasion dynamics could play a role in overriding some major environmental drivers. Our results also showed that a lower diversity of ecological traits in communities (about half of less disturbed communities) corresponded to a high invasion degree, and that the exotic component of communities had typically less diverse ecological traits than the native one, even when accounting for stream order and species richness. Overall, our results suggest that possible outcomes of severe exotic species invasions could include a reduced functional diversity of invaded communities, but analyzing data with finer ecological, temporal and spatial resolutions would be needed to pinpoint the causal relationship between invasions and functional diversity.
Full-text available
1.Questions related to understanding which characteristics of the recipient communities make them vulnerable to invasion (i.e. invasibility) are commonly formulated in such a way that they consider ecological communities as whole entities. Yet, species within recipient communities have specific responses to the introduction of exotic species. 2.As a consequence, we know little about the role of species interaction networks within and between trophic levels in determining the invasion likelihood of exotics introduced in previously uninvaded communities, and the temporal dynamics of invaded communities. 3.To obtain a mature understanding of invasion processes and their impacts, I propose here to apply a structural stability approach of species coexistence to the field of biological invasions. The main aim here is to distinguish how direct and indirect species’ responses to invasion in competitive and trophic networks translate to the mechanisms determining the maintenance and stability of species diversity in multispecies and multitrophic assemblages. 4.This approach brings the further opportunity to integrate the field of biological invasions within a broader perspective about the study of novel ecosystems, that is, ecosystems that contain exotic species and face new introductions sometimes under abrupt environmental changes. Moreover, it can help to target specific mechanisms separating winners from losers in novel ecosystems. 5.Synthesis:A solid framework to predict biological invasions and understand community assembly along temporal scales under novel ecosystems can be obtained by framing species’ responses to invasion within recent theoretical and methodological advances stemming from coexistence theory, which seeks to determine how diversity is maintained. This article is protected by copyright. All rights reserved.
Full-text available
Aim Exotic species are a major threat to biodiversity and have modified native communities worldwide. Invasion processes have been extensively studied, but studies on species richness and beta diversity patterns of exotic and native species are rare. We investigate such patterns among exotic and native fish communities in upland and lowland rivers to explore their relationship with environmental drivers. Location Northern Italy. Methods Exotic and native fish beta diversity patterns were investigated separately in lowland and upland sites using Local Contribution to Beta Diversity (LCBD) and Species Contribution to Beta Diversity (SCBD) analyses. To examine the main environmental variables affecting the LCBD, a Boosted Regression Trees (BRT) method was used. Community dispersion among and within stream orders was investigated with the PERMDISP test. Results In lowland sites, exotic species richness was higher than native species richness, especially in large rivers and drainage canals. An opposite trend was found in upland sites, where native species richness was higher than exotic species richness, especially in large rivers. No clear LCBD patterns were found along stream orders in the lowland, whereas higher stream orders in the upland showed the highest LCBD. Its patterns in upland and lowland sites were related to a number of factors, such as total suspended solids and total phosphorus. Community dispersion among stream orders did not show a relationship with environmental heterogeneity. SCBD values were positively correlated with species occupancy in the study area, and native species showed higher SCBD values than exotic species only in the uplands. Main conclusions Large rivers in the uplands are important in maintaining native fish diversity and should be protected against invasive fish. In contrast, most lowland rivers have suffered from biological homogenization. Some rare native species can show low contribution to beta diversity, but still need conservation actions due to their risk of local extinction.
Full-text available
1. We examined riverine desert fish assemblages in the Chihuahuan Desert, U.S.A. at multiple spatial scales of similarity to assess long-term changes to assemblage distinctiveness, identify individual species responsible for changes, and determine the importance of geographic context and species resolution in interpreting patterns of change. 2. We used a well-documented historical data set on fish distribution and abundance, and recent collections of fishes that provided a paired analytical design across 36 localities spanning nearly 3 decades. Patterns of faunal homogenisation and differentiation were assessed at basin-wide, sub-basin and river-reach scales with species occurrence and relative abundance data. Individual species responses were examined to identify the drivers of assemblage change across time. 3. Patterns of assemblage similarity varied across spatial scales and produced seemingly incongruous trends across time. Assemblage distinctiveness depended on the spatial extent of the analyses, the geographical structuring of the fish assemblages, and whether occurrence or relative abundance data were used. These dependencies led to interesting and conflicting patterns of homogenisation and differentiation. The Rio Grande sub-basin showed strong homogenisation with convergence between upstream and downstream reaches that corresponded to declining water quality and quantity from the Rio Conchos in Mexico. In contrast, the Pecos River sub-basin showed strong differentiation between upstream and downstream reaches that corresponded to the successful colonisation and spread of the non-native gulf killifish (Fundulus grandis) in the upper reach, where water quantity and quality have declined. Spatial variability in fish assemblages and their degree of change from historical conditions were largely congruent with anthropogenic modifications to the flow regime and variability in the success of invasive gulf killifish in the basin. 4. The use of species occurrence as opposed to relative abundance data, and the spatial scale of analysis are crucial choices in studies of faunal homogenisation and differentiation, and we have demonstrated how these choices lead to variable results for our study system. Our multi-scale approach and examination of individual species responses identified the ultimate drivers of these differences and illustrated the importance of scale-dependent effects and geographical context on patterns of assemblage distinctiveness, especially with regard to species invasion, species loss and relative abundance shifts.
Full-text available
Functional composition of communities across scales is increasingly used to infer resilience of biotic communities to environmental change. To assess the relevance of these concepts to management of large rivers, analyses were applied to fish community data of the Upper Mississippi River. First, to evaluate whether there was evidence for structural patterns in fish size distributions, a discontinuity analysis was performed. Using long‐term fish data, consistent discontinuities were identified across 14 reaches, suggesting similar structuring processes occur throughout the nearly 1300 river kilometers represented by those reaches. Increased variability in species abundance in relation to proximity to edges of body size aggregations supports the discontinuity hypothesis that body size aggregations are structured by key processes. Functional richness and redundancy were quantified within and across identified scales for each of 14 river reaches. Diversity of trophic and spawning guilds was generally greater in the upstream reaches in comparison with downstream reaches, with the exception of the diversity of large‐bodied spawning guilds. Evidence of functional shifts in the composition of fishes occurred, but differed by size aggregations, likely reflecting scale‐specific resource availability. Redundancy of spawning and trophic groups across body size aggregations suggested downstream reaches of this system may be less resilient to disturbances and were weakly associated with reduced habitat diversity. These findings suggest that discontinuities in large river fish assemblages do occur and may provide indication of shifting resource availability. Further investigation of the underlying processes and scales that support resource availability will be critical to managing for resilience in large river ecosystems.
Full-text available
Invasions of Ponto-Caspian fish species into north-western European river basins accelerated since the opening of the Rhine–Main–Danube Canal in 1992. Since 2002, at least five Ponto-Caspian alien fish species have arrived in The Netherlands. Four species belong to the Gobiidae family (Neogobius fluviatilis, Neogobius melanostomus, Ponticola kessleri, and Proterorhinus semilunaris) and one to the Cyprinidae family (Romanogobio belingi). These species are expected to be potentially deleterious for the populations of four native benthic fish species: Gobio gobio (Cyprinidae), Barbatula barbatula (Nemacheilidae), Cottus perifretum, and C. rhenanus (Cottidae). Invasion success may be dependent on competitive trophic interactions with native species, which are enabled and/or constrained by feeding-related morphological traits. Twenty-two functional feeding traits were measured in nine species (in total 90 specimens). These traits were quantitatively linked to the mechanical, chemical and behavioral properties of a range of aquatic resource categories, using a previously developed food-fish model (FFM). The FFM was used to predict the trophic profile (TP) of each fish: the combined capacities to feed on each of the resource types. The most extreme TPs belonged to three alien species, indicating that they were most specialized among the studied species. Of these three, only P. kessleri overlapped with the two native Cottus species, indicating potential trophic competition. N. fluviatilis and R. belingi did not show any overlap, indicating that there is low trophic competition. The two remaining alien goby species (N. melanostomus and P. semilunaris) had average TPs and could be considered generalist feeders. They overlapped with each other and with G. gobio and B. barbatula, indicating potential trophic competition. This study suggests that both generalist and specialist species can be successful invaders. Since the FFM predicts potential interactions between species, it provides a tool to support horizon scanning and rapid risk assessments of alien species.
Full-text available
The invasion of exotic species is one of the main threats to worldwide biodiversity and can be aided by changes in environmental conditions. We hypothesized that a temporal trend of decreasing discharge and increasing temperature might have favored the invasion of warm-adapted, lentic exotic fish species in the lower Po River, northern Italy. We used presence/absence data over a long-term period (over 20 years) to investigate the dynamics of exotic fish invasion along water temperature and discharge gradients. Mean annual discharge and temperature did not show a clear trend and did not affect exotic fish species invasion, which progressed with time irrespective of these factors. The total number of species fluctuated without a clear trend, which underlined a progressive substitution of native species with exotic ones. Perhaps surprisingly, the community composition changed over time towards more temperature tolerant but also rheophilic, benthivore and generalist fish species. These results highlight how species interactions could be one of the main factors driving the invasion. Furthermore, our data underlines a continuously rising tide of exotics, which questions the success of past control strategies. Considering the current conservation resources limitations, priority should be given to the development of prevention strategies in order to avoid new species introductions.
Although one of the most evident effects of biological invasions is the loss of native taxonomic diversity, contrasting views exist on the consequences of biological invasions on native functional diversity. We investigated this topic using Mediterranean stream, river and canal fish communities as a test case, at 3734 sites in Italy, and distinguishing between exotic and translocated species invasion in three different faunal districts. Our results clearly confirmed that introduced species were widespread and in many cases the invasion was severe (130 communities were completely composed by introduced species). Exotic and translocated fish species had substantially different geographical distribution patterns, perhaps arising from their differences in introduction timing, spread and invasion mechanisms. We also found a clear decreasing trend of functional dispersion along an invasion gradient, confirming our hypothesis that the invasion process can diminish the relative diversity of ecofunctional traits of host fish communities. Furthermore, our results suggested that exotic species might have a greater negative effect than translocated species on the relative diversity of ecofunctional traits of fish communities. This could also be linked to the fact that translocated species are more ecofunctionally similar to native ones, compared to the exotics. Our multivariate analysis of site-specific combinations of ecofunctional traits highlighted some traits characteristic of all invaded communities, while our discriminant analysis underlined how there was a substantial ecofunctional overlap between native, exotic and translocated species groups in most areas.
Rivers worldwide are impacted by human activities such as habitat degradation, habitat fragmentation, waterway flow regulation, and introduction of exotic species, which are responsible for the reduction or the disappearance of native species in many parts of the world. The Oglio River, a tributary of the Po River in Northern Italy, is a good example of a river with a long history of human alteration and where exotic invasions are present. We used data on water parameters and fish communities along the watercourse to investigate whether low flow conditions, degraded water quality, abundant exotic species, and the presence of migration barriers could be a disadvantage for native species. We used ordination methods (redundancy analysis), variance partitioning analysis, and the threshold indicator taxa analysis to explore changes in community composition and ecofunctional traits along an altitude gradient. We found that exotic species affected native ones more than water quality and hydromorphological parameters. Native species were most abundant in the upper reach of the Oglio River, despite low flow and shallow depth. Moreover, rheophilic and clear water native fish decreased rapidly in the lower reach of the river, where exotic species increased. This distribution could be explained by the presence of barriers in the middle reach, which block exotic species migrating upstream from the highly invaded Po River, and by a lower suitability of the upper reach for some exotic species. Our results provide a general description of the fish fauna of a strongly regulated river and can contribute to develop more effective fish and water management practices.
1.Plant invasions can drastically change the structure of native communities, but it is not fully understood whether alien species occupy phylogenetic and functional space within the range occupied by natives, or provide a novel set of evolutionary origins and traits to the invaded communities. 2.Here, we evaluated this open question with data on large number of plant communities from different terrestrial habitats. We used ~27,000 vegetation plots from 26 terrestrial habitats in the Czech Republic and compared phylogenetic and functional diversity (PD and FD, respectively) and community trait means in invaded and non‐invaded plots. We tested for differences (i) between invaded vs. non‐invaded plots, (ii) among natives in invaded vs non‐invaded plots, and (iii) in invaded plots only, with and without aliens. To minimize habitat filtering effects on PD and FD, we ran these tests within the habitat‐specific species pools of the 26 vegetation types. 3.In general, PD, FD and trait mean values changed with invasion, with changes being rather consistent across the habitats considered. Invaded plots were less phylogenetically, but more functionally diverse than non‐invaded plots. The greater FD in invaded plots, compared to non‐invaded ones, was due to greater dissimilarity between natives. In fact, native species in invaded plots showed higher PD and FD than native species in non‐invaded plots, whilst alien species reduced PD and FD in invaded plots. Changes in the trait means with invasion were due to differences in native species in invaded and non‐invaded plots, rather than to an effect of alien species. Within most habitats, the trait means and variance of all aliens were similar to those of all natives, while in some habitats, the variability in traits was greater between aliens that belonged to phylogenetically closer clades. 4.Synthesis. Our results suggest that alien species more often occupy a phylogenetic and functional space within the range formed by the native species in a community. They do so either by filling empty gaps or by excluding natives from the existing phylogenetic and functional space, rather than occupying or creating a phylogenetic and trait space outside of it. This article is protected by copyright. All rights reserved.