Running head
Spatial processes in riparian vegetation
Title
Spatial processes structuring riparian plant communities in agroecosystems: implications for
restoration
BÉRENGER BOURGEOIS,1,2 EDUARDO GONZÁLEZ,3,4 ANNE VANASSE,1 ISABELLE AUBIN,5 AND
MONIQUE POULIN1,2
1 Département de Phytologie, Faculté des Sciences de l’Agriculture et de l’Alimentation,
Université Laval, 2425 rue de l’agriculture, Québec, Québec, G1V 0A6, Canada
2 Québec Centre for Biodiversity Science, Department of Biology, McGill University, Stewart
Biology Building, 1205 Dr. Penfield avenue, Montréal, Québec, H3A 1B1, Canada
3 Department of Biological Sciences, University of Denver, F W Olin Hall, Room 102, 2190 E
Iliff avenue, Denver, Colorado, 80208-9010, USA
4 EcoLab, Université Paul Sabatier, Institut National Polytechnique de Toulouse, Centre
National de la Recherche Scientifique, 118 route de Narbonne Bâtiment 4R1, 31062 Toulouse
Cedex 9, France
5 Natural Resources Canada, Canadian Forest Service, Great Lakes Forestry Centre, 1219
Queen St. East, Sault Ste. Marie, Ontario, P6A 2E5, Canada
Corresponding author: Département de Phytologie, Faculté des Sciences de l'Agriculture et de
l'Alimentation, Université Laval, Pavillon Paul-Comtois, 2425, rue de l'Agriculture, Québec
(Québec) G1V 0A6. [email protected]; [email protected]
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
12
Abstract. The disruption of hydrological connectivity by human activities such as flood
regulation or land-use changes strongly impacts riparian plant communities. However, landscape
scale processes have generally been neglected in riparian restoration projects as opposed to local
conditions, even though such processes might largely influence community recovery. We
surveyed plant composition of field edges and riverbanks in 51 riparian zones restored by tree
planting (565 1-m2 plots) within two agricultural watersheds in southeastern Québec (Canada).
Once the effects of environmental variables (hydrology, soil, agriculture, landscape, restoration)
were partialled out, three models of spatial autocorrelation based on Moran’s Eigenvector Maps
and Asymmetric Eigenvector Maps were compared to quantify the pathways and direction of the
spatial processes structuring riparian communities. The ecological mechanisms underlying
predominant spatial processes were then assessed by regression trees linking species response to
spatial gradients to seed and morphological traits. The structure of riparian communities was
predominantly related to unidirectional spatial gradients from upstream to downstream along
watercourses, which contributed more to species composition than bidirectional gradients along
watercourses or overland. Plant traits selected by regression trees explained 22% of species
response to unidirectional upstream-downstream gradients in field edges, and 24% in riverbanks,
and predominantly corresponded to seed traits rather than morphological traits of the adult
plants. Our study showed that even in agriculturally open landscapes, water flow remains a major
force structuring spatially riparian plant communities by filtering species according to their seed
traits, thereby suggesting long-distance dispersal as a predominant process. Preserving
hydrological connectivity at the watershed-scale and restoring riparian plant communities from
upstream to downstream should be encouraged to improve the ecological integrity of rivers
running through agricultural landscapes.
2
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
34
Key words: agricultural landscapes; ecological restoration; eigenvector maps; hydrochory;
plant traits; spatial autocorrelation; unidirectional vs. bidirectional spatial processes;
watercourse vs. overland spatial processes.
INTRODUCTION
The rapid expansion of agricultural lands worldwide has led to important land-use changes,
habitat fragmentation and flood regulation that are severely degrading riparian habitats. Flow
regulation, which has been estimated to impact over 50% of the world’s large river systems
(Nilsson et al. 2005), cumulatively reduces hydrological connectivity (Ward and Stanford 1995),
affects water runoff (Postel et al. 1996, Allan 2004) and profoundly modifies riparian
communities (Poff and Zimmerman 2010). Dispersal limitation and loss of recruitment sites due
to flow regulation result in impoverished riparian plant communities with reduced cover of
native trees, shrubs and herbs, and higher cover of exotics (Nilsson et al. 1997, Jansson et al.
2000, Catford et al. 2011). Preserving the dynamism, resilience and diversity of riparian plant
communities thus largely depends on the spatio-temporal connectivity of rivers (Poff et al 1997,
Tabacchi et al. 1998, Bendix et Hupp 2000).
The multidimensional gradients structuring rivers and their associated riparian zones have
been outlined in successive theoretical frameworks (Petts and Amoros 1996, Ward 1989, 1998,
Tockner et al. 2010). The River Continuum Concept first identified the longitudinal spatial
processes driving gradual changes in the physical conditions and biotic communities of rivers
from upstream to downstream (Vannote et al. 1980). This spatial structure was extended to
lateral gradients governing exchanges between river channels and floodplains in the Flood Pulse
3
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
56
and Flow Pulse Concepts (Junk et al. 1989, Bayley 1991, Tockner et al. 2000). A vertical
dimension was then added to the conceptualization of river functioning through the Hyporheic
Corridor Concept linking groundwater and rivers (Gibert et al. 1990, Stanford and Ward 1993).
In addition to these three spatial gradients, temporal processes were identified as playing a role in
shaping river communities according to a Patch Dynamics Concept, in which disturbances and
colonization events are related to water flow (Townsend 1989). Hydrological Connectivity, i.e.,
the water-mediated transfer of matter, energy and organisms along these four gradients,
ultimately emerged from these concepts as the key driver of the ecological integrity of rivers
(Ward and Stanford 1995, Ward et al. 1999, Pringle 2003). Even if these theoretical frameworks
have led to a better understanding of river systems in the last three decades, they still remain
conceptual; recently developed statistical tools now makes it possible to quantify the
multidimensional spatial drivers of river ecological processes.
Plant dispersal by water (i.e., hydrochory) strongly structures riparian plant communities
(Nilsson et al. 2010). The composition of riparian plant communities is indeed largely influenced
by the traits of seeds and propagules trapped in river sediments and water column (Boedeltje et
al. 2003, Gurnell et al. 2008, Corenblit et al. 2009), particularly those determining species
hydrochorous ability (Catford and Jansson 2014). For example, high seed buoyancy and seed
release phenology synchronous with flood regimes favor the dispersal of riparian plant species
by river flow (Merritt and Wohl 2002, Boedeltje et al. 2004, González et al. 2015a). However,
the role of hydrochory in structuring the distribution of riparian plants has generally been studied
at the local scale, such as the reach of a river, or for particular species such as invasive species or
Salicaceae riparian trees, rather than for entire communities at large scale (Boedeltje et al. 2003,
Stella et al. 2006, Richardson et al. 2007). Due to these geographic and taxonomic restrictions,
4
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
78
previous studies can only be generalized to a limited extent. Additional seed traits such as mass,
size or longevity have been proposed as relevant plant features for explaining the long-distance
dispersal of riparian plant species (Nilsson et al. 2010), but their contribution to the large-scale
spatial structure of entire plant communities has never been investigated.
The adaptive strategies developed by riparian plant species in response to hydrological
spatial gradients rely on their ability to not only colonize new patches, but also to tolerate flood
disturbances (Lytle and Poff 2004, Bornette et al. 2008, Corenblit et al. 2009). Consequently, the
large-scale spatial structure of riparian plant communities might also depend on morphological
traits that confer tolerance to flood disturbances, such as height, since taller plants can avoid
submergence, or root depth, as greater rooting depth can provide anchorage during flooding and
include storage organs that facilitate resprouting after sediment burial (Lytle and Poff 2004,
Bornette et al. 2008, Catford and Jansson 2014). However, the relative contribution of seed traits
vs. morphological traits to the structure of riparian plant communities at large spatial scales has
never been quantified.
Recognizing the importance of flood regimes for riparian communities has contributed to
the development of hydrological approaches to restoring riparian plant communities that have
been degraded by human activities. These approaches focus on re-creating the physical attributes
of riparian habitats, mainly through renaturalization of flow regimes and manual reconfiguration
of floodplain landforms (Poff et al. 1997, Richter et al. 2003, González et al. 2015b). Sometimes,
nurse species are also introduced during restoration to re-create non-hydrological attributes of
riparian zones (Battaglia et al. 2008, McClain et al. 2011, Harris et al. 2012). Once riparian zones
have been physically reconfigured, plant species are assumed to passively recolonize (Field of
Dreams Hypothesis, “if you build it, they will come”; Palmer et al. 1997). Rivers are indeed
5
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
910
expected to act as ecological corridors that promote the dispersal of target plant species to
restored riparian zones (Naiman et al. 1993). In agricultural landscapes however, multiple large-
scale environmental factors can impact restoration trajectories and outcomes (Allan 2004, Mayer
and Rietkerk 2004, Tscharntke et al. 2005). Such land-use intensification could cause the low
recolonization of herbaceous species through a number of processes including the
impoverishment of the regional species pool (Brudvig 2011), the homogenization of local abiotic
conditions (Vellend et al. 2007), the decrease of landscape connectivity (Tscharntke et al. 2005)
and the limitation of plant dispersal (Hermy and Verheyen 2007). Furthermore, habitat
fragmentation, land clearing and landscape opening associated with agricultural expansion can
further favor the invasion of restored riparian zones by wind-dispersed generalist or exotic
species and contribute to the homogenization of plant communities in these landscapes (Bossuyt
et al. 1999, Vellend et al. 2007, Grashof-Bokdam 2009). In this context, riparian plant
communities are likely to be structured not only by spatial processes occurring along rivers, but
also by those occurring overland. Surprisingly though, riparian restoration projects have
generally neglected to integrate, at least explicitly, the spatial processes fostering the
recolonization process (González et al. 2015b).
The recent development of multivariate multiscale spatial analyses (Dray et al. 2012) now
makes it possible to disentangle the unidirectional (Asymmetric eigenvector maps, Blanchet et
al. 2008) or bidirectional (Moran’s eigenvector maps, Dray et al. 2006) spatial processes that
structure riparian plant communities at the watershed scale. These spatial analyses represent
innovative tools to deepen our understanding of the mechanisms that structure riparian plant
communities and foster their restoration. They also represent an opportunity to formally test river
ecology theoretical frameworks. Such statistical tools can discriminate between allogenic
6
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
1112
processes related to the spatial structure of environmental variables (i.e., spatial dependence) and
autogenic community processes such as dispersal (i.e., pure spatial autocorrelation, Legendre
1993, Dray et al. 2012). These analyses can also identify the spatial scales at which communities
are structured (Dray et al. 2012). In a restoration context, spatial models can help determine the
dispersal vectors that promote the recovery of plant communities, locate disruptions in ecological
connectivity, or assess the impact of land-use on community structure.
This study aims (1) to quantify the contribution of spatial processes to structuring plant
communities in tree-planted riparian zones of agricultural landscapes and (2) to explore the
underlying ecological mechanisms via plant traits. We hypothesized hydrochory to be the main
operative mechanism, as demonstrated by (1) the spatial structure of riparian communities,
principally corresponding to a unidirectional upstream‒downstream spatial model based on
watercourse distances, as opposed to bidirectional processes along watercourses or overland, and
by (2) the spatial distribution of riparian plants, which is predominantly driven by seed traits that
represent plant dispersal.
METHODS
Study area
This study was conducted in two agricultural watersheds in southeastern Québec, Canada.
The Boyer watershed (46°41' N, 70°55' W) extends over an area of 216 km² that includes
scattered forest fragments (24% of the land) but is primarily agricultural land (66% of the land),
of which 26% is farmed with annual crops (principally wheat, corn and soybean). From 1984 to
1992, watercourses were widely channelized over 73% of the watershed’s 215 km of rivers to
improve soil drainage for crop cultivation (OBV Côte-du-Sud/GIRB 2011). The Bélair
7
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
1314
watershed (46°26' N, 70°56' W) covers an area of 43 km² that is predominantly forested (66%),
with 33% agricultural land, of which 18% are annual crops (MAPAQ unpublished data). In
1987, the province banned agricultural practice from riparian zones extending at least 3 m in
width on each side of any stream, in order to reduce soil erosion and improve water quality
(Gouvernement du Québec 1987). This protection was then complemented with extensive tree
planting programs that were conducted in the two studied watersheds from 1995 to 2009. Trees
were planted every 3‒5 m on the flat edge of agricultural fields, and the following species were
mainly introduced: Fraxinus pennsylvanica, Acer saccharum, Picea glauca, Larix laricina,
Quercus rubra and Quercus macrocarpa.
Sampling design
Fifty-one agricultural riparian zones representing a chronosequence from 3 to 17 years
after tree planting were sampled in the two studied watersheds during the summer of 2012. To be
sampled, riparian zones had to meet four conditions: (1) measure at least 40 m long, (2) have
been planted with trees within a single year, (3) be adjacent to an agricultural field with a single
type of crop, (4) have a uniform vegetation structure. Furthermore, sampled riparian zones were
selected in order to be well distributed within the watersheds and associated with rivers of
uniform width (mean ± standard deviation = 3.38 ± 1.60 m) in order to control for hydro-
geomorphologic variability.
Botanical surveys
To account for intra-site variability in plant communities, the number of plant surveys
performed on each site was proportional to site area. At each site, from two to five equidistant
8
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
1516
transects were established perpendicular to the river (i.e., from the field edge to the river bank)
depending on the length of the site. More precisely, for sites shorter than 100 m (min. length 40
m), two transects were used. Additional transects were placed as site length increased by 50 m,
up to a maximum of five transects for sites longer than 200 m (max. length 1150 m). Along these
transects, a minimum of two 1-m2 plots (one in the field edge, one in the riverbank) up to a
maximum of six plots were staked out, depending on transect length, i.e., one plot every 5 m in
the riverbank and one plot every 10 m in the field edge. Species composition of communities
was inventoried within these 1-m² plots by visual estimation of species cover (%) using classes
with a 5% range. Since a preliminary Principal Component Analysis (PCA) revealed distinct
plant compositions in field edge (262 plots) vs. riverbank plots (302 plots, Appendix S1), all
subsequent analyses were performed separately for these two types of communities. Community-
weighted means (see below) were thus calculated separately for field edge and riverbank
communities, and two inventories of plant abundances were obtained for each site.
Plant traits
Eleven seed and morphological plant traits were selected from the Traits Of Plants In
Canada database (TOPIC, Aubin et al. 2012), and supplemented by a literature review of
scientific articles and local floras (Flora of North America Editorial Committee 1993, Marie-
Victorin et al. 1995). Seed dispersal and seedling establishment strategies were represented by
four seed traits, while competition, resource use and tolerance to disturbance strategies of adult
plants were characterized by six morphological traits (Weiher et al. 1999, Cornelissen et al.
2003, Table 1).
9
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
1718
Explanatory variables
Two types of between-sites geographic distances were measured based on the coordinates
of each site to perform spatial analyses: overland distances (calculated with a minimum spanning
tree) and distance along watercourses (measured on ArcGIS). Seventeen environmental variables
related to tree planting, agriculture, hydrology, soils and landscape structure were also measured
at each site (Table 2). These variables were used to determine the environmental determinants
(either spatially structured or not) of riparian plant communities: their effects were thus removed
prior to performing spatial analyses. This procedure allowed us to investigate the effect of pure
spatial autocorrelation on the composition of riparian plant communities.
Statistical analysis
To partial out the effect of environmental variables on Hellinger-transformed species
abundances, a forward selection was first conducted to retain significant variables.
Environmental variables significantly related to species abundances were next used in a
Redundancy Analysis (RDA). Residuals were extracted and used as response variables in the
spatial analyses described below, as they corresponded to the environment-independent part of
the plant communities.
Three spatial models were then used on the extracted RDA residuals to assess the
importance of spatial processes on species composition: (1) a bidirectional overland model, (2) a
bidirectional watercourse model and (3) a unidirectional upstream-downstream watercourse
model (Fig. 1). The model representing overland bidirectional spatial processes was constructed
using Moran’s Eigenvector Maps (MEM, Dray et al. 2006). A matrix of overland between-site
distances (calculated by minimum spanning tree from XY site coordinates) was analyzed using a
10
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
1920
Principal Coordinates Analysis (PCoA) with resulting eigenvectors corresponding to spatial
components. The second model, representing bidirectional spatial processes along watercourses
was built using the same procedure, based instead on watercourse distances. The third model,
representing unidirectional spatial processes between sites along watercourse in the direction of
water flow (from upstream to downstream), was constructed using Asymmetric Eigenvector
Maps (AEM, Blanchet et al. 2008). A node-by-edge matrix coding unidirectional between-site
connections along watercourse was analyzed by means of a PCA with resulting eigenvectors
corresponding to spatial components.
For each of the three spatial models, we calculated 30 spatial submodels representing
linear, concave-down or concave-up spatial relationships with three different types of geographic
weighting functions, i.e.,
f 1α=1−(
d ij
max (d ij ))
α
, f 2α= 1
(d ij)α and f 3
α=(d ij
max ( dij ))
α
where d ij corresponds to the distances between the sites i and j and α, ranked from 1 to 10 (Dray
et al. 2006, Blanchet et al. 2011). For each submodel, the spatial components representing a
positive and significant spatial autocorrelation were determined using the Moran’s I statistics and
forward selection. These components were then used as explanatory variables in separate RDA
on detrended species abundances for MEM models and on non-detrended species abundances for
AEM models in order to identify the best weighting function (i.e., the best submodel) for each of
the three spatial models. Only the weighting function that resulted in the highest adjusted R2
among the 30 submodels tested was conserved for each spatial model: functions f 13 in field edges
and f 21 in riverbanks for the bidirectional overland model, function f 2
1 in both field edges and
riverbanks for the bidirectional watercourse model, and functions f 11 in field edges and f 1
2 in
11
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
2122
riverbanks for the unidirectional upstream‒downstream watercourse model. In both types of
plant communities (field edge and riverbank), we thus selected the weighting function with the
best explanatory power for each of the three spatial models (Fig. 1).
The three final spatial models were compared based on the adjusted R2 produced by the
RDA, and submitted to variation partitioning to identify the best geographic pathway (overland
vs. watercourse bidirectional processes) and direction (bidirectional vs. unidirectional
watercourse processes) for explaining the spatial structure of non detrended species abundances
in field edge and riverbank riparian plant communities separately.
Ecological mechanisms underlying spatial patterns that drive riparian plant communities
were then identified by the use of plant traits. For this, we selected the best spatial model based
on the above-mentioned adjusted R2 comparison and variation partitioning, and extracted species
scores along the significant RDA axes representing spatial gradients (i.e., linear combinations of
spatial components). These scores were used in a regression tree with the 11 plant traits as
explanatory variables in order to determine the traits responsible for species response to spatial
gradients (Kleyer et al. 2012, Fig. 1). To maximize representativeness, only species with more
than three informed trait values were used in this analysis; these included 95 species for field
edge communities and 116 for riverbank communities.
All analyses were performed on R version 3.0.3 (R Core Team 2014) using packages
vegan (Oksanen et al. 2013), spacemaker (Dray 2013), AEM (Blanchet et al. 2014) and packfor
(Dray et al. 2013).
RESULTS
Environmental control of riparian plant communities
12
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
2324
Field edge and riverbank communities were composed of 124 and 157 species,
respectively, for a total of 167 species inventoried in the 565 1-m² plots sampled within the 51
studied riparian zones (Appendix S2). Four selected environmental variables significantly
explained 18% of the species composition (adjusted R2; F = 2.86, P = 0.005) in the field edges:
elevation above river, soil gravel content, crop type in the adjacent field and tree canopy cover
(Table 3). In riverbank communities, five selected environmental variables significantly
explained 22% of species composition (adjusted R2; F = 2.96, P = 0.005): tree canopy cover, soil
gravel content, crop type in the adjacent field, river width and riverbank width (Table 3). The
effect of the environmental variables (either non-spatially or spatially structured) on plant
community composition was partialled out to conserve only the variability in species
composition that was independent from environmental variables and thereby investigate pure
spatial autocorrelation among these plant communities.
Spatial processes structuring riparian plant communities
The composition of riparian communities was spatially structured predominantly by
watercourse processes as opposed to overland ones and unidirectionally rather than
bidirectionally. Upstream‒downstream processes along watercourses indeed contributed to the
highest explained variation of the residuals of plant composition in both field edges (adjusted R2
= 11.1% which makes a total of 27% of variance explained, if combined with the environmental
variables) and riverbanks (adjusted R2 = 16.5% which makes a total of 34% of variance
explained, if combined with the environmental variables, Table 4), once the effect of
environmental variables was partialled out as explained above. This was especially true for
riverbank communities (closer to the river channel), for which the adjusted R2 of the
13
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
2526
unidirectional watercourse model was 1.2 to 1.5 times higher than the adjusted R2 of the two
bidirectional models. The variation explained by the bidirectional overland and watercourse
models was, however, still quite high in both field edge and riverbank communities (on the order
of 10%). However, variation partitioning indicated that a large part of this variation (7.0% and
10.9% for field edge and riverbank communities, respectively) was shared with the
unidirectional watercourse model (see Appendix S3 for the complete figures). Furthermore, the
variation in non-detrended plant communities solely explained by the unidirectional watercourse
model amounted to 6.7% in field edges and 4.1% in riverbanks and was higher than the variation
in plant communities solely explained by the two bidirectional models.
In the unidirectional model along watercourses, three spatial gradients (i.e., linear
combinations of spatial components) significantly influenced plant composition in field edge
communities (RDA axis 1 to 3; proportion explained = 8.6%, 4.5% and 3.9% respectively) and
five in riverbank communities (RDA axis 1 to 5; proportion explained = 8.9%, 5.7%, 4.8%, 4.4%
and 3.7% respectively). Considering the higher adjusted R2 provided by the unidirectional model
in both field edge and riverbank communities, the significant spatial gradients of this model were
selected for the subsequent analysis, which sought to identify the ecological mechanisms
underlying autogenic spatial patterns found in riparian communities based on plant traits.
Plant traits related to spatial processes
Among the 11 traits tested, community assembly processes under upstream‒downstream
watercourse spatial gradients were predominantly related to seed traits (Fig. 2). In field edge and
riverbank communities, the plant traits selected by regression trees explained 22% and 24% of
14
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
2728
species scores along significant upstream‒downstream spatial gradients and sorted species into
three and six distinct groups respectively.
In field edge communities, seed length predominantly explained the autogenic spatial
structure of plant composition from upstream to downstream along watercourse by separating
short-seeded species from long-seeded, which were negatively correlated with the two first
spatial gradients (seed length threshold = 5.3 mm, RDA axis 1, RDA Axis 2, Fig. 2a). For short-
seeded species, the start of the seed release period secondly influenced the response to
unidirectional upstream‒downstream gradients as species with seeds dispersed in autumn formed
a distinct group from species with seeds dispersed in spring or summer. These latter species with
earlier seed dispersal were then separated according to their seed weight (seed weight threshold =
0.12 mg). Finally, maximum height classified species in two distinct subgroups for species with
light seeds (weight inferior to 0.12 mg) dispersed in spring or summer (maximum height
threshold = 1.3 m) as well as for species dispersing seeds in autumn (maximum height threshold
= 1.4 m).
In riverbanks, species composition was also mostly explained by seed length, which first
separated species with short seeds from those with long seeds (seed length threshold = 6.8 mm),
the latter being negatively correlated with the first unidirectional upstream‒downstream spatial
gradient (RDA Axis1, Fig. 2b). Short-seeded species were secondly sorted into two subgroups
according to their maximum height, with species taller than 2.1 m forming a distinct subgroup.
The seed dispersal vector further subdivided species smaller than 2.1 m, as anemochorous
species responded differently to unidirectional upstream‒downstream spatial gradients than
species with other dispersal vectors. Seed weight (threshold = 0.07 mg) and seed length
(threshold = 3.3 mm) then contributed to species response for anemochorous species while
15
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
2930
species with hydrochorous, myrmechorous, zoochorous and unassisted dispersal were only
classified according to their seed length (threshold = 2.7 mm).
DISCUSSION
Spatial processes structuring riparian plant communities in agricultural landscapes
Our study highlights the prevalence of upstream‒downstream gradients along watercourses
over bidirectional gradients along watercourses or overland. Even for tree-planted sites scattered
in agricultural landscapes, water flow remains the major force contributing to the autogenic
spatial structure riparian communities at the watershed scale. Our study thus extends the well-
known structuring role of rivers (Hupp and Osterkamp 1996, Poff et al. 1997, Tabacchi et al.
1998) to recently restored sites that occur within heavily transformed landscapes. This underlines
the need for upscaling the evaluation of riparian restoration projects, which have mainly been
conducted locally to date (González et al. 2015b), and supports theoretical frameworks that
provide a landscape-scale perspective on riparian communities, as being structured around
hierarchical spatial scales (Gregory et al. 1991, Wu and Loucks 1995, Thorp et al. 2006). Using
similar spatial models, macrophyte, diatom, macroinvertebrate and fish communities have also
been found to be strongly structured by longitudinal gradients along rivers in the direction of
water flow (Landeiro et al. 2011, Liu et al. 2013, Padial et al. 2014). This study therefore extends
the importance of longitudinal gradients previously identified in such aquatic communities to the
riparian ecotone. Moreover, these results together provide quantitative data to support the
importance of hydrological connectivity, a key spatial process for biotic communities of river
ecosystems that was originally intended to be applied at watershed level, but was only
investigated at the scale of river reach (Ward and Stanford 1995, Ward et al. 1999, Pringle 2003).
16
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
3132
Surprisingly, overland gradients contributed to a large part of the spatial autogenic
structure of riparian plant communities, even if most of their explained variation was also shared
by watercourse gradients. This contribution of overland gradients might be attributed to land-use
changes due to agricultural intensification, which are known to impact the spatial connectivity of
riparian zones (Allan 2004, Tetzlaff et al. 2007). While the three environmental variables
representing landscape composition did not affect significantly the composition of riparian plant
communities here, the loss of structural heterogeneity in the studied agricultural landscapes
could still have promoted overland spatial processes. The habitat fragmentation and landscape
simplification that accompany agricultural intensification indeed affect the dispersal of plant
species differently, depending on their means of dispersion (Vellend et al. 2007, Ozinga et al.
2009). Anemochory highly favors species dispersal despite patch isolation and strongly drives
the diversity of plant communities in fragmented areas (Herlin and Fry 2000, Grashof-Bokdam
2009, Ozinga et al. 2009). In addition to this large-scale dispersal through seed rain (Gabriel et
al. 2005, Roschewitz et al. 2005, Buisson et al. 2006), small-scale dispersal from nearby patches
due to neighborhood effects also influences plant communities within agricultural open
landscapes (Kleyer 1999, Roschewitz et al. 2005). In the agricultural watersheds studied here,
the contribution of overland spatial processes to the composition of riparian communities is thus
likely related to a combination of seed rain and neighborhood effect fostering the immigration of
plant species from agricultural fields to riparian zones. Further investigation of the contribution
of overland spatial processes in the context of agricultural landscapes having different
intensification levels is needed, however, in order to better identify the underlying ecological
mechanisms.
17
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
3334
Seed traits drive species responses to upstream-downstream gradients
Riparian plant species responded to upstream‒downstream spatial gradients along
watercourses according to their seed traits. Seed length was the main trait explaining the affinity
of plant species to upstream‒downstream spatial gradients, especially in riverbank communities
closer to the river channel. In our study, long-seeded species were negatively correlated to
upstream‒downstream spatial gradients along watercourses, meaning that water flow had less
influence on the distribution of long-seeded species than on short-seeded ones. Besides seed
length, seed weight, seed dispersal vector and the start of the seed release period secondly
contributed to species sorting. While seed weight is considered a better functional marker (sensu
Garnier et al. 2004) than seed length (Westoby 1998), both are associated with several ecological
mechanisms operating at different plant developmental stages, from seed dispersal to
establishment of seedlings (Weiher et al. 1999, Moles and Westoby 2004, 2006). In general,
long, large and heavy seeds are produced in lower number and tend to travel shorter distances
than smaller seeds (Thompson et al. 1993, Westoby 1998, Leishman et al. 2000), while, in the
case of hydrochory, seed buoyancy, also conditions dispersal distance but remains poorly
informed in the literature (Merritt and Wohl, 2002, Nilsson et al. 2010). The importance of seed
traits over morphological traits in explaining the response of species to spatial gradients along
watercourse from upstream to downstream points to hydrochory as the main mechanism
responsible for the watershed-scale spatial structure of riparian communities. The spatial
structure of riparian communities at the reach scale has been related to hydrochory by several
authors (Nilsson et al. 1991, Boedeltje et al. 2003, Gurnell et al. 2008), but according to our
results, the key influence of this process can also be detected at the watershed-scale.
Furthermore, the distribution of a plant species in a riparian habitat was previously attributed to
18
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
3536
several traits: long seed buoyancy fostering the hydrochorous ability of plant species (Merritt and
Wohl 2002, Boedeltje et al. 2003, 2004), small and compact size that can facilitate burial in
sediment and persistence over a long period of time (Thompson et al. 1993), or large seeds which
produce seedlings that establish more successfully (Xiong et al. 2011). Our study showed that the
spatial structure of riparian plant communities from upstream to downstream was predominantly
related to seed traits, indirectly pointing out the importance of long distance hydrochorous
dispersal for species assembly.
Role of morphological traits in upstream-downstream species distribution
Although hydrochory played a predominant role, additional processes related to
morphological traits also helped to explain community composition along upstream‒downstream
gradients. Maximum plant height indeed contributed to species sorting in both field edge and
riverbank communities. While morphological traits are known to influence competition ability
and resource-use (Pérez-Harguindeguy et al. 2013), their contribution to the spatial structure of
riparian communities in the direction of water flow could be mainly related to tolerance to flow
disturbances. Such adaptation is an important feature of riparian plant strategies (Lytle and Poff
2004, Naiman et al. 2005, Catford and Jansson 2014). For example, tall stems can decrease the
impact of water flow by reducing biomass loss, while the higher resistance of small and compact
species to mechanical damage can explain their positive correlation with upstream‒downstream
spatial gradients observed here in field edges (Karrenberg et al. 2002, Boeger and Poulson 2003,
Puijalon and Bornette 2004). The predominant role of seed traits in structuring riparian
communities spatially and the secondary role of morphological traits suggest the existence of a
hierarchy within the adaptive strategies of plant species to riparian habitats. Alternatively, the
19
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
3738
higher influence of seed vs. morphological traits in structuring riparian plant communities could
be due to the spatial scale considered in this study. Seed traits might be of primary importance
for watershed-scale processes such as dispersal (as shown here) while morphological traits
should rather influence species assembly at the local scale (not assessed in this study) by
conditioning competition ability, resource-use strategies or resistance to disturbance (e.g., plant
height could be related to flood tolerance as higher plants can escape from inundation easier than
shorter plants).
Integrating spatial processes in riparian community restoration
This study demonstrates that longitudinal processes merit greater consideration in
management and restoration strategies. Restoring the longitudinal hydrological connectivity of
rivers through renaturalization of flow regimes (Poff et al. 1997, Poff and Zimmerman 2010),
should be among the first alternatives to be considered by restoration projects, especially in small
rivers running through agricultural landscapes where implementing this approach is
socioeconomically and technically more realistic than in other industrialized areas or larger
rivers. This approach fosters the recovery of the longitudinal gradients that structure river
ecosystems by removing dams, implementing environmental flows and reallocating water
resources, as initial necessary steps for the re-establishment of riparian communities. To
accelerate and optimize restoration efforts, such an approach would preferentially locate active
introductions of plant species in upstream sites, as downstream sites might be naturally colonized
via hydrochorous dispersal, especially in the case of small-seeded species. Restoration of riparian
zone should be implemented from upstream to downstream along river networks (Dietrich et al.
2014). As previously pointed out by Holl et al. (2003), there is indeed an urgent need to expand
20
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
3940
riparian zone restoration to the scale of the watershed. While legacies of past agricultural land-
use, long-lasting degradations and uninvestigated environmental factors could have contributed
to the unexplained variation in plant community composition, our statistical framework provides
an innovative perspective to take into account spatial processes in future restoration projects.
Considering large-scale processes in decision-making and land-use planning via spatial models is
a promising approach for developing efficient management strategies for riparian plant
communities, or, more holistically, entire river ecosystems. Spatial modeling of ecosystem
recovery is a new tool to scale up ecological restoration in a landscape perspective.
ACKNOWLEDGMENTS
We would thank local restoration practitioners, stakeholders and decision-makers from
Organisme de Bassin Versant de la Côte-du-Sud, Club conseil Bélair-Morency, Ministère de
l’Agriculture, des Pêcheries et de l’Alimention du Québec (respectively F. Lajoie, L. Beaulieu
and A. Goudreau) for their help in sampling design; field assistants (A. Rablat, P. Israël-Morin
and M. Vaillancourt) for data collection; P. Legendre from Université de Montréal for statistical
advice; K. Grislis for English revision; all the farmers who welcomed us onto their lands to
conduct this study; and Christer Nilsson and two anonymous reviewers for their help in
improving our manuscript. This project was funded by a research grant from the Ministère de
l’Agriculture, des Pêcheries et de l’Alimention du Québec received by A.V. and M.P., and by a
NSERC discovery grant to M.P. (RGPIN-2014-05663).
LITERATURE CITED
21
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
4142
Allan, J. D. 2004. Landscapes and riverscapes: the influence of land use on stream ecosystems.
Annual Review of Ecology, Evolution and Systematics 35:257–284.
Aubin, I., C. Messier, S. Gachet, K. Lawrence, D. McKenney, A. Arseneault, W. Bell, L. De
Grandpré, B. Shipley, J.-P. Ricard, and A.D. Munson. 2012. TOPIC: traits of plants in
Canada. Natural Resources Canada, Canadian Forest Service, Sault Ste. Marie, Ontario,
Canada. http://topic.nrcan.gc.ca/ (consulted on 2016-02-22).
Battaglia, L. L., D. W. Pritchett, and P. R. Minchin. 2008. Evaluating dispersal limitation in
passive bottomland forest restoration. Restoration Ecology 16:417–424.
Bayley, P. B. 1991. The flood pulse advantage and the restoration of river-floodplain systems.
Regulated Rivers: Research and Management 2:75–86.
Bendix, J., and C. R. Hupp. 2000. Hydrological and geomorphological impacts on riparian plant
communities. Hydrological Processes 14:2977–2990.
Blanchet, F. G., P. Legendre, and D. Borcard. 2008. Modelling directional spatial processes in
ecological data. Ecological Modelling 215:352–336.
Blanchet, F. G., P. Legendre, and O. Gauthier. 2014. AEM: tools to construct Asymmetric
Eigenvector Maps (AEM) spatial variables. R package version 0.5-2/r119. http://R-
Forge.R-project.org/projects/sedar/ (consulted on 2016-02-22).
Blanchet, F. G., P. Legendre, R. Maranger, D. Monti, and P. Pepin. 2011. Modelling the effect of
directional spatial ecological processes at different scales. Oecologia 166:357–368.
Boedeltje, G., J. P. Bakker, R. M. Bekker, J. M. van Groenendael, and M. Soesbergen. 2003.
Plant dispersal in a lowland stream in relation to occurrence and three specific life-history
traits of the species in the species pool. Journal of Ecology 91:855–866.
22
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
4344
Boedeltje, G., J. P. Bakker, A. Ten Brinke, J. M. van Groenendael, and M. Soesbergen. 2004.
Dispersal phenology of hydrochorous plants in relation to discharge, seed release time and
buoyancy of seeds: the flood pulse concept supported. Journal of Ecology 32:786–796.
Boeger, M. R. T., and M. E. Poulson. 2003. Morphological adaptations and photosynthetic rates
of amphibious Veronica anagallis-aquatica L. (Scrophulariaceae) under different flow
regimes. Aquatic Botany 75:123–135.
Bornette, G., E. Tabacchi, C. Hupp, S. Puijalon, and J. C. Rostan. 2008. A model of plant
strategies in fluvial hydrosystems. Freshwater Biology 53:1692–1705.
Bossuyt B., M. Hermy, and J. Deckers. 1999. Migration of herbaceous plant species across
ancient‒recent forest ecotones in central Belgium. Journal of Ecology 87:628–638.
Buisson, E., T. Dutoit, F. Torre, C, Romermann, and P. Poschlod. 2006. The implications of seed
rain and seed bank patterns for plant succession at the edges of abandoned fields in
Mediterranean landscapes. Agriculture, Ecosystems and Environment 115:6–14.
Cabezas, A., E. González, B. Gallardo, M. García, M. González, and F. A. Comín. 2008. Effects
of hydrological connectivity on the substrate and understory structure of riparian wetlands
in the Middle Ebro River (NE Spain): implications for restoration and management.
Aquatic Sciences 70:361–376.
Catford, J. A., B. J. Downes, C. J. Gippel, and P. A. Vesk. 2011. Flow regulation reduces native
plant cover and facilitates exotic invasion in riparian wetlands. Journal of Applied Ecology
48:432–442.
Catford, J. A., and R. Jansson. 2014. Drowned, buried and carried away: effect of plant traits on
the distribution of native and alien species in riparian ecosystems. New Phytologist
204:19–36.
23
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
4546
Cornelissen, J. H. C., S. Lavorel, E. Garnier, S. Diaz, N. Buchmann, D. E. Gurvich, P. B. Reich,
H. ter Steege, H. D. Morgan, M. G. A. van der Heijden, J. G. Pausas, and H. Poorter. 2003.
A handbook of protocols for standardized and easy measurement of plant functional traits
worldwide. Australian Journal of Botany 51:335–380.
Dietrich, A. L., L. Lind, C. Nilsson & R. Jansson. 2014. The use of phytometers for evaluating
restoration effects on riparian soil fertility. Journal of Environmental Quality 43:1916‒
1925.
Dray, S. 2013. SpacemakeR: spatial modelling. R package version 0.0-5/r113. http://R-Forge.R-
project.org/projects/sedar/ (consulted on 2016-02-22).
Dray, S., P. Legendre, and F. G. Blanchet. 2013. Packfor: forward selection with permutation
(Canoco p.46) R package version 0.0-8/r109. http://R-Forge.R-project.org/projects/sedar/
(consulted on 2016-02-22).
Dray, S., P. Legendre, and P. R. Peres-Neto. 2006. Spatial modelling: a comprehensive
framework for principal coordinate analysis of neighbor matrices (PCNM). Ecological
Modelling 196:483–493.
Dray, S., R. Pelissier, P. Couteron, M.-J. Fortin, P. Legendre, P. R. Peres-Neto, E. Bellier, R.
Bivand, F. G. Blanchet, M. de Caceres, A.-B. Dufour, E. Heegaard, T. Jombart, F. Munoz,
J. Oksanen, J. Thioulouse, and H. H. Wagner. 2012. Community ecology in the age of
multivariable multiscale spatial analysis. Ecological Monographs 82:257–275.
Flora of North America Editorial Committee. 1993. Flora of North America North of Mexico.
19+ vols. New York and Oxford. http://floranorthamerica.org/ (consulted on 2016-02-22).
Gabriel, D., C. Thies, and T. Tscharntke. 2005. Local diversity of arable weeds increases with
landscape connectivity. Perspectives in Plant Ecology, Evolution and Systematics 7:85–93.
24
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
4748
Garnier, E., J. Cortez, G. Billès, M.-L. Navas, C. Roumet, M. Debussche, G. Laurent, A.
Blanchard, D. Aubry, A. Bellmann, C. Neill, and J.-P. Toussaint. 2004. Plant functional
markers capture ecosystem properties during secondary succession. Ecology 85:2630–
2637.
Gibert, J., M.-J. Dole-Olivier, P. Marmonier, and P. Vervier. 1990. Surface water‒groundwater
ecotones. Pages 199–225 in R.J. Naiman and H. Decamps, editors. The ecology and
management of aquatic‒terrestrial ecotones. Parthenon Publishers, Carnforth, UK.
González, E., A. A. Sher, E. Tabacchi, A. Masip, and M. Poulin. 2015b. Restoration of riparian
vegetation: a global review of implementation and evaluation approaches in the
international, peer-reviewed literature. Journal of Environmental Management 158:85–94.
González, E., B. Bourgeois, A. Masip, and A. A. Sher. 2015a. Trade-offs in seed dispersal
strategies across riparian trees: the how matters as much as the when. River Research and
Applications.
Gouvernement du Québec. 1987. Loi sur la qualité de l’environnement: politique des rives, du
littoral et des plaines inondables. Gouvernement du Québec, Québec, Canada.
http://www2.publicationsduquebec.gouv.qc.ca/ (consulted on 2016-02-22).
Grashof-Bokdam, C. 2009. Forest species in an agricultural landscape in the Netherlands: effects
of habitat fragmentation. Journal of Vegetation Science 8:21–28.
Gregory, S. V., F. J. Swanson, W. A. McKee, and K. W. Cummins. 1991. An ecosystem
perspective of riparian zones. BioScience 41:540–551.
Gurnell, A., K. Thompson, J. Goodson, and H. Moggridge. 2008. Propagule deposition along
river margins: linking hydrology and ecology. Journal of Ecology 96:553–565.
25
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
4950
Harris, C. J., M. R. Leishman, K. Fryirs, and G. Kyle. 2012. How does restoration of native
canopy affect understory vegetation composition? Evidence from riparian communities of
the Hunter Valley Austrialia. Restoration Ecology 20:584–592.
Herlin, I. L. S., and G. L. A. Fry. 2000. Dispersal of woody plants in forest edges and hedgerows
in a southern Swedish agricultural area: the role of site and landscape structure. Landscape
Ecology 15:229–242.
Holl, K. D., E. E. Crone, and C. B. Schultz. 2003. Landscape restoration: moving from
generalities to methodologies. BioScience 53:491–502.
Hupp, C. R., and W. R. Osterkamp. 1996. Riparian vegetation and fluvial geomorphic processes.
Geomorphology 14:277–295.
Jansson, R., C. Nilsson, and B. Renöfält. 2000. Fragmentation of riparian floras in rivers with
multiple dams. Ecology 81:899–903.
Junk, W. J., P. B. Bayley, and R. E. Sparks. 1989. The flood-pulse concept in river-floodplain
systems. Pages 110‒127 in D. P. Dodge, editor. Proceedings of the International Large
Rivers Symposium (LARS). Canadian Special Publication of Fisheries and Aquatic
Sciences 106.
Karrenberg, S., P. J. Edwards, and J. Kollmann. 2002. The life history of Salicaceae living in the
active zone of floodplains. Freshwater Biology 47:733–748.
Kleyer, M. 1999. Distribution of plant functional types along gradients of disturbance intensity
and resource supply in an agricultural landscape. Journal of Vegetation Science 10:697–
708.
26
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
5152
Kleyer, M., S. Dray, F. de Bello, J. Leps, R. J. Pakeman, B. Strauss, W. Thuiller, and S. Lavorel.
2012. Assessing species response and community functional responses to environmental
gradients: which multivariate methods? Journal of Vegetation Science 23:805–821.
Landeiro, V. L., W. E. Magnusson, A. S. Melo, H. M. V. Espirito-Santo, and L. M. Buini. 2011.
Spatial eigenfunction analyses in stream networks: do watercourse and overland distances
produce different results? Freshwater Biology 56:1184–1192.
Leishman, M. R., I. J. Wright, A. T. Moles, and M. Westoby. 2000. The evolutionary ecology of
seed size. Pages 31–57 in M. Fenner, editor. Seeds: the ecology of regeneration in plant
communities, 2nd edition. CAB International, Wallingford, UK.
Legendre, P. 1993. Spatial autocorrelation: trouble or new paradigm? Ecology 74:1659–1673.
Liu, J., J. Soininen, B.-P. Han, and S. A. J. Declerck. 2013. Effects of connectivity, dispersal
directionality and functional traits on the metacommunity structure of river benthic
diatoms. Journal of Biogeography 40:2238–2248.
Lytle, D. A., and N. L. Poff. 2004. Adaptation to natural flow regimes. Trends in Ecology and
Evolution 19:94–100.
Marie-Victorin, F. E .C., L. Brouillet, and I. Goulet. 1995 Flore laurentienne, 3rd edition. Gaëtan
Morin éditeur, Montréal.
McClain, C. D., K. D. Holl, and D. M. Wood. 2011. Successional models as guides for
restoration of riparian forest understory. Restoration Ecology 19:280–289.
Merritt, D. M., and E. E. Wohl. 2002. Processes governing hydrochory along rivers: hydraulics,
hydrology, and dispersal phenology. Ecological Applications 12:1071–1087.
Moles, A., and M. Westoby. 2004. Seedling survival and seed size: a synthesis of the literature.
Journal of Ecology 92:372–383.
27
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
5354
Moles, A., and M. Westoby. 2006. Seed size and plant strategy across the whole life cycle. Oikos
113: 91–115.
Naiman, R. J., H. Décamps, and M. Pollock. 1993. The role of riparian corridors in maintaining
regional biodiversity. Ecological Applications 3:209–212.
Naiman, R. J., H. Décamps, M. E. McClain, and G. E. Likens. 2005. Riparia: ecology,
conservation, and management of streamside communities. Elsevier Academic Press,
London, 448 pp.
Nilsson, C., M. Gardfjell, and G. Grelsson. 1991. Importance of hydrochory in structuring plant
communities along rivers. Canadian Journal of Botany 69:2631–2633.
Nilsson, C., R. Jansson, and U. Zinko. 1997. Long-term responses of river-margin vegetation to
water-level regulation. Science 276:798–800.
Nilsson, C., C. A. Reidy, M. Dynesius, and C. Revenga. 2005. Fragmentation and flow
regulation of the world’s large river systems. Science 308:405–408.
Nilsson, C., R. L. Brown, R. Jansson, and D. M. Merritt. 2010. The role of hydrochory in
structuring riparian and wetland vegetation. Biological Reviews 85:837–858.
OBV Côte-du-Sud/GIRB. 2011. Plan directeur de l’eau du bassin versant de la rivière Boyer.
Organisme des bassins versants (OBV) de la Côte-du-Sud et Groupe d’intervention pour la
restauration de la Boyer (GIRB), Québec.
Oksanen, J., F. G. Blanchet, R. Kindt, P. Legendre, P. R. Minchin, R. B. O’Hara, G. L. Simpson,
P. Solymos, M. H. H. Stevens, and H. Wagner.2013. Vegan: community ecology package.
R package version 2.0-10. http://CRAN.R-project.org/package=vegan (consulted on 2016-
02-22).
28
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
5556
Ozinga, W. A., C. Romermann, R. M. Bekker, A. Prinzing, W. L. M. Tamis, J. H. J. Schaminee,
S. M. Hennekens, K. Thompson, P. Poschlod, M. Kleyer, J. P. Bakker, and J. M. van
Groenendael. 2009. Dispersal failure contributes to plant losses in NW Europe. Ecology
Letters 12:66–74.
Palmer, M. A., R. F. Ambrose, and N. L. Poff. 1997. Ecological theory and community
restoration ecology. Restoration Ecology 5:291–300.
Padial, A. A., F. Ceschin, S. A. J. Declerck, L. De Meester, C. C. Bonecker, F. A. Lansac-Toha,
L. Rodrigues, L. C. Rodrigues, S. Train, L. F. M. Velho, and L. M. Bini. 2014. Dispersal
ability determines the role of environmental, spatial and temporal drivers of
metacommunity structure. PlosOne 9: e111227.
Pérez-Harguindeguy, N., S. Díaz, E. Garnier, S. Lavorel, H. Poorter, P. Jaureguiberry, M. S.
Bret-Harte, W. K. Cornwell, J. M. Craine. D. E. Gurvich, C. Urcelay, E. J. Veneklaas, P.
B. Reich, L. Poorter, I. J. Wright, P. Ray, L. Enrico, J. G. Pausas, A. C. de Vos, N.
Buchmann, G. Funes, F. Quetier, J. G. Hodgson, K. Thompson, H. D. Morgan, H. ter
Steeger, M. G. A. van der Heijden, L. Sack, B. Blonder, P. Poschold, M. V. Vaieretti, G.
Conti, A. C. Staver, S. Aquino, and J. H. C. Cornelissen. 2013. New handbook for
standardised measurement of plant functional traits worldwide. Australian Journal of
Botany 61:167–234.
Petts, G. E., and Amoros C. 1996. Fluvial hydrosystems. Chapman & Hall, London, UK. 322 pp.
Poff, N. L., J. D. Allan, M. B. Bain, J. R. Karr, K. L. Prestegaard, B. D. Richter, R. E. Sparks,
and J. C. Stromberg. 1997. The natural flow regime: a paradigm for river conservation and
restoration. BioScience 47:769–784.
29
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
5758
Poff, N. L., and J. K. H. Zimmerman. 2010. Ecological responses to altered flow regimes: a
literature review to inform science and management of environmental flows. Freshwater
Biology 55:194–205.
Postel, S. L., G. C. Daily, and P. R. Ehrlich. 1996. Human appropriation of renewable fresh
water. Science 271:785–788.
Pringle, C. 2003. What is hydrologic connectivity and why is it ecologically important?
Hydrological Processes 17:2685–2689.
Puijalon, S., and G. Bornette. 2004. Morphological variation of two taxonomically distant plant
species along a natural flow velocity gradient. New Phytologist 163:651–660.
R Core Team. 2014. R: a language and environment for statistical computing. R Foundation for
Statistical Computing, Vienna, Austria. http://www.R-project.org/ (consulted on 2016-02-
22).
Richardson, D. M., P. M. Holmes, K. J. Esler, S. M. Galatowitsch, J. C. Stromberg, S. P.
Kirkman, P. Pysek, and R. J. Hobbs. 2007. Riparian vegetation: degradation, alien plant
invasions, and restoration prospects. Diversity and Distributions 13:126–139.
Richter, B. D., R. Mathews, D. L. Harrison, and R. Wigington. 2003. Ecologically sustainable
water management: managing river flows for ecological integrity. Ecological Applications
13:206–224.
Roschewitz, I., D. Gabriel, T. Tscharntke, and C. Thies. 2005. The effects of landscape
complexity on arable weed species diversity in organic and conventional farming. Journal
of Applied Ecology 42:873–882.
30
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
5960
Stanford, J. A., and J. V. Ward. 1993. An ecosystem perspective of alluvial rivers: connectivity
and the hyporheic corridor. Journal of the North American Benthological Society 12:48–
60.
Stella, J. C., J. J. Battles, B. K. Orr, J. R. McBride. 2006. Synchrony of seed dispersal, hydrology
and locate climate in a semi-arid river reach on California. Ecosystems 9:1200–1214.
Tabacchi, E., D. L. Correll, R. Hauer, G. Pinay, A.-M. Planty-Tabacchi,, and R. C. Wissmar.
1998. Development, maintenance and role of riparian vegetation in the river landscape.
Freshwater Biology 40:497–516.
Tetzlaff, D., C. Soulsby, P. J. Bacon, A. F. Youngson, C. Gibbins, and I. A. Malcolm. 2007.
Connectivity between landscapes and riverscapes: a unifying theme in integrating
hydrology and ecology in catchment science? Hydrological Processes 21:1385–1389.
Thompson, K., S. R. Brand, J. G. Hodgson. 1993. Seed size and shape predict persistence in soil.
Functional Ecology 7:236–241.
Thorp, J. H., M. C. Thoms, and M. D. Delong. 2006. The riverine ecosystem synthesis:
biocomplexity in river networks across space and time. River Research and Applications
22:123–147.
Tockner, K., M. S. Lorang, and J. A. Stanford. 2010. River flood plains are model ecosystems to
test general hydrogeomorphic and ecological concepts. River Research and Applications
26:76–86.
Tockner, K., F. Malard, and J. V. Ward. 2000. An extension of the flood pulse concept.
Hydrological Processes 14:2861–2883.
Townsend, C. R. 1989. The patch dynamics concept of stream community ecology. Journal of
the North American Benthological Society 8:36–50.
31
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
6162
Vannote, R. L., G. W. Minshall, K. W. Cummins, J. R. Sedell, and C. E. Cushing. 1980. The
river continuum concept. Canadian Journal of Fisheries and Aquatic Sciences 37:130–137.
Vellend, M., K. Verheyen, K. M. Flinn, H. Jacquemyn, A. Kolb, H. van Calster, G. Peterken, B.
J. Graae, J. Bellemare, O. Honnay, J. Brunet, M. Wulf, F. Gerhardt, and M. Hermy. 2007.
Homogenization of forest plant communities and weakening of species‒environment
relationships via agricultural land use. Journal of Ecology 95:565–573.
Ward, J. V. 1989. The four-dimensional nature of lotic ecosystems. Journal of the North
American Benthological Society 8:2–8.
Ward, J. V. 1998. Riverine landscapes: biodiversity patterns, disturbance regimes, and aquatic
conservation. Biological Conservation 83:269–278.
Ward, J. V., and J. A. Stanford. 1995. Ecological connectivity in alluvial river ecosystems and its
disruption by flow regulation. Regulated Rivers: Research and Management 11:105–119.
Ward, J. V., K. Tockner, and F. Schiemer. 1999. Biodiversity of floodplain river ecosystems:
ecotones and connectivity. Regulated Rivers: Research and Management 15:125–139.
Weiher, E., A. van der Werf, K. Thompson, M. Roderick, E. Garnier, and O. Eriksson. 1999.
Challenging Theophrastus: a common core list of plant traits for functional ecology.
Journal of Vegetation Science 10:609–620.
Westoby, M. 1998. A leaf-height-seed (LHS) plant ecology strategy scheme. Plant and Soil
199:213–227.
Wu, J., and O. L. Loucks. 1995. From balance of nature to hierarchical patch dynamics: a
paradigm shift in ecology. The Quaterly Review of Biology 70:439–466.
32
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
6364
Xiong, S., C. Nilsson, M. E. Johansson, and R. Jansson. 2001. Responses of riparian plants to
accumulation of silt and plant litter: the importance of plant traits. Journal of Vegetation
Science 12:481–490.
33
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
6566
TABLE 1. Description of the traits selected to characterize the 167 plant species inventoried in the
51 sampled riparian zones.
Trait name Trait valueMissing
value (%)
Regeneration and dispersion
Seed length numerical (mm) 7
Seed weight numerical (mg.seed-1) 11
Seed production abundant (> 1000 seeds.ind-1.year-1);
semi-abundant (20-1000 seeds.ind-1.year-
1); scarce (1-20 seeds.ind-1.year-1)
0
Seed dispersal vector anemochore; autochore; barochore;
hydrochore; myrmechore; zoochore
0
Seed release period: start spring; summer; autumn 0
Seed release period:
duration
numerical (seasons) 0
Morphology and adult plant
strategy
Life cycle annual; bi-annual; perennial 0
Maximum height numerical (m) 0
Morphological type fern; grass and sedge; herb; horsetail; 0
34
747
748
6768
vine
Raunkiaer life form chamaephyte; geophyte;
hemicryptophyte; helophyte; micro- and
nano-phanerophyte; therophyte
0
Vegetative propagation absence; presence 0
35
749
750
751
752
753
754
755
756
757
758
759
760
761
6970
TABLE 2. Description of the 17 environmental variables measured in the 51 sampled riparian
zones.
Variable Unit Methodology
Tree planting
Time elapsed since tree
planting
Years Discussions with stakeholders
and landowners
Tree canopy cover % Stereoscopic measurements
taken at 1 m height within each
plot, above understory species
(WINSCANOPY system, Régent
Instruments Inc., Québec,
Canada)
Agriculture
Crop type in the adjacent
field during sampling
Hay, corn,
soybean, wheat
Visual assessment
Hydrology
Elevation above the river
water level
m Height between the plot and
river water level (calculated for
each plot from the slope and the
distance to the river)
36
762
763
7172
River discharge Low, average,
strong
Visual assessment and
discussions with stakeholders
River sinuosity Ordinal categories
from 0 to 3 (0:
rectilinear river
channel)
Visual assessment
Riverbank erosion Ordinal categories
from 0 to 3 (0:
rectilinear river
channel)
"
River width m Measured with a tape
Riverbank width m "
Fiel edge width m "
Soil
Proportion of stones (250-
600 mm)
% Visual assessement of soil
samples (of ca15 000 cm3)
collected with an auger in the 0-
15 cm soil horizon at each plot
Proportion of cobbles (75-
250 mm)
% "
Proportion of gravels (2- % "
377374
75 mm)
Proportion of fine particles
(<2 mm)
% "
Landscape structure
Proportion of forests % Measured in a 500 m-wide
buffer zone around each site
from governemental
georeferenced databases* using
ArcGIS
Proportion of grasslands % "
Proportion of croplands % "
* Base de Données Topographiques du Québec from the Ministère des Ressources Naturelles et de la
Faune and Base de Données des Cultures Assurées from la Financière Agricole du Québec
38
764
765
766
767
768
769
770
771
7576
TABLE 3. Environmental variables significantly influencing understory species composition in
riparian plant communities of field edges and riverbanks, obtained by forward selection (adjusted
R2environmental variables = 18% in field edges and adjusted R2
environmental variables = 22% in riverbanks).
df F Pr > P
Field edge
Elevation above river 1 3.23 0.01
Soil gravel content 1 3.13 0.02
Crop type in the adjacent
field
3 3.00 0.01
Tree canopy cover 1 1.87 0.03
Riverbank
Tree canopy cover 1 8.07 0.01
Soil gravel content 1 2.30 0.02
Crop type in the adjacent
field
3 2.25 0.01
River width 1 1.88 0.01
Riverbank width 1 1.70 0.01
39
772
773
774
775
776
7778
TABLE 4. Proportion of variance (adjusted R2) in plant composition explained by the three spatial
models for field edge and riverbank riparian plant communities (MEM: Moran’s Eigenvector
Maps; AEM: Asymmetric Eigenvector Maps) once the effects of significant environmental
variables were partialled out (see Table 3; adjusted R2environmental variables = 18% in field edges and
adjusted R2environmental variables = 22% in riverbanks).
Adjusted R² (%)
Bidirectional overland
(MEM)
Bidirectional
watercourse (MEM)
Unidirectional
watercourse (AEM)
Field edge 9.94 10.14 11.09
Riverbank 11.16 13.23 16.46
40
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
7980
FIG. 1. Diagram of the analytical steps performed with the five datasets to address the two
research objectives of this study.
41
794
795
796
797
798
799
800
801
802
8182
FIG. 2. Traits-based classification of plant species response to upstream‒downstream spatial
process obtained by a regression tree for (a) field edge and (b) riverbank riparian communities.
Species scores along significant axes of the RDA modelling unidirectional upstream-downstream
watercourse spatial gradient (AEM model) were used as dependent variables and a selection of
functional traits (Table 1) as explanatory variables.
42
803
804
805
806
807
808
8384
43
809
8586