Urban Forestry & Urban Greening 84 (2023) 127925

Contents lists available at ScienceDirect

Urban Forestry & Urban Greening

journal homepage: www.elsevier.com/locate/ufug

Using plant community uniqueness and floristic quality assessment in

management decision-making in an urban park setting
Audrey Paquette a, b, Stéphanie Pellerin b, c, d, Monique Poulin a, b, *
a
Department of phytology, FSAA., 2425 rue de l’Agriculture, Université Laval, Québec, QC, G1V 0A6, Canada
b
Quebec Centre for Biodiversity Science, Department of Biology, McGill University, Stewart Biology Building, 1205 Dr. Penfield Avenue, Montreal, QC H3A 1B1, Canada
c
Institut de recherche en biologie végétale, Département de sciences biologiques, Université de Montréal, 4101 Sherbrooke E., Montréal, QC H1X 2B2, Canada
d
Jardin botanique de Montréal, 4101 Sherbrooke E., Montréal, QC H1X 2B2, Canada

A R T I C L E I N F O A B S T R A C T

Handling Editor: Cecil Konijnendijk van den With the densification of cities, it is imperative to identify urban ecosystems that should be protected or restored.
Bosch We aimed to determine the conservation and restoration needs in a large urban park (1,35 km2) in Quebec City
(Canada), based on site history, current species richness, floristic uniqueness, and floristic quality assessment of
Keywords: its diverse ecosystems: forest, swamp, wooded peatland, open peatland, and marsh. We evaluated the cover of all
Conservation
vascular species in 70 plots (400 m2) and assessed 18 environmental variables. We found that forest and swamp
Local contribution to beta diversity (LCBD)
plots were the richest while peatland plots were the poorest, with marsh plots showing intermediate values.
Mean coefficient of conservatism (Mean C)
Peatland Ecological uniqueness (LCBD) was not correlated with richness (ρ = 0.17; p > 0.05), with marsh and peatland
Restoration plots showing the highest and lowest uniqueness, respectively. With a regression tree, we identified canopy
Richness openness as the most influential variable explaining plot uniqueness across all ecosystems, especially in the
Tree encroachment peatland, indicating that future recreational development should be avoided in open ecosystems. By plotting
ecological uniqueness (LCBD) with tolerance to disturbance (Mean C) values, we identified areas that could
benefit from conservation or restoration, and areas that could sustain future development for recreational use.
For each area, floristic composition, site characteristics, and past land-use history were investigated further to
identify appropriate actions. The open peatland was identified as the main conservation priority, but actions will
be needed to limit rapid tree encroachment. Three marshes were identified as areas that would floristically
benefit from restoration actions. Still, since they also act as natural retention basins hosting species adapted to
the soil conditions, we suggested monitoring the expansion of exotic and invasive species. Approaches developed
and lessons learned from this project will serve as guidelines for municipalities aiming to implement a restoration
and management plan in urban parks.

1. Introduction and ruderal species than undisturbed natural areas (Dearborn and Kark,
2010; Nielson et al., 2013; Lepczyk et al., 2017).
More than half of the world’s population lives in urban areas and To achieve this two-fold mission, management strategies in UGS
future population growth is expected to occur mainly in cities (United must reduce the environmental impacts of recreational activities within
Nations, 2018). While the increase in urban population leads to both their boundaries. In that respect, infrastructures devoted to facilitating
expansion and densification of built areas, there is growing recognition and enhancing human experiences (e.g., trails, access roads, parking
of the importance of urban green spaces (UGS; e.g., forests, wetlands, lots) are well recognized to fragment the area (Ballantyne et al., 2014;
parks, gardens) for improving human well-being and protecting biodi­ Williams-Mounsey et al., 2021); to affect the biogeochemical cycles of
versity (Dearborn and Kark, 2010; Sandifer et al., 2015; Lepczyk et al., ecosystems, nutrient availability, and hydraulic conductivity (Chen
2017). Yet, reconciling both the social and ecological dimensions of the et al., 2014; Lin et al., 2014); and, in turn, to impact biodiversity (Zedler
mission of UGS remains a challenge, especially since they are often and Leach, 1998; Zielinska et al., 2013). It is also important to under­
small, isolated, and fragmented areas with a higher abundance of exotic stand how the transformation of the surrounding landscape may affect

* Correspondence to: Département de phytologie, Faculté des sciences de l’agriculture et de, l’alimentation, Pavillon Paul-Comtois, local 33072425 rue de
l’Agriculture, Québec, Québec G1V 0A6, Canada.
E-mail address: [email protected] (M. Poulin).

https://doi.org/10.1016/j.ufug.2023.127925
Received 25 October 2022; Received in revised form 23 February 2023; Accepted 4 April 2023
Available online 6 April 2023
1618-8667/© 2023 Elsevier GmbH. All rights reserved.
A. Paquette et al. Urban Forestry & Urban Greening 84 (2023) 127925

plant communities of UGS especially since there is often a time lag be­ photographs to explore the potential effect of past disturbances on the
tween the cause and the effect of disturbances (Lira et al., 2019). Yet, a present-day floristic communities, 2) to identify the biodiversity hot­
detailed portrait of human impacts on the biodiversity of UGS is lacking, spots (high floristic species richness) and areas with a high plant com­
which hampers decisions about how and where to intervene to protect munity uniqueness and those with high floristic quality, and 3) to
local biodiversity (Feltynowski et al., 2018). examine the variables associated with unicity and their link to the park’s
The biodiversity in UGS can be protected using different manage­ history. In addition, we explored the potential of using FQA values in
ment strategies, such as delimiting strict conservation zones, restoring conjunction with uniqueness values of plant communities to identify
degraded ecosystems, and improving infrastructure for promoting rec­ conservation and restoration needs. We predicted that ecological
reational activities in localized areas (Maxwell et al., 2020). Several uniqueness would be higher for vegetation communities less affected by
criteria can be used to identify conservation and restoration priorities past land use changes in their surroundings and mostly associated with
within UGS, among them species richness, ecological uniqueness, and specialist species and high diversity, thus indicating sites for
floristic quality assessment. Ecological uniqueness is a criterion recently conservation.
proposed to identify sites with a unique species composition that
contribute greatly to regional diversity (Legendre and De Caceres, 2. Material and methods
2013). Many researchers have since highlighted the potential of this
approach for assessing the conservation value of sites within forests and 2.1. Study area
wetlands (e.g., Heino and Grönroos, 2017; Tan et al., 2019; Dubois et al.,
2020). Since high ecological uniqueness can be associated with both The study was conducted at the Base de plein air de Sainte-Foy
high and low conservation values (Legendre and De Caceres, 2013), (BPASF; 46◦ 47’ 13.9914", − 71◦ 19’ 57.108" W; Fig. 1), a municipal
species composition requires further investigation (Dubois et al., 2020). park of 1,35 km2 in Quebec City (ca. 540 000 inhabitants). The BPASF is
Floristic quality assessment (FQA) is another criterion that has been located in a proximal suburb of 1 095 people/km2 and is surrounded by
used to assess the conservation value of sites. FQA reflects the quality of two highways and a marshalling yard. The BPASF lies on a thick sand
plant communities and is calculated using the coefficient of conserva­ and gravel layer from the deglaciation period, above a layer of silty clay,
tism (C-value) of a species, which indicates its sensitivity to disturbance a remnant of the Champlain Sea invasion (Filion et al., 2009). The
and habitat specificity (Spyreas, 2019). Most of the research on FQA impermeable clay layer explains the high level of water retention and
index performance suggests that FQA is correlated to disturbance the presence of wetlands in the park. The BPASF is characterized by
gradient (e.g., Bourdaghs et al., 2006; Bried et al., 2013; Kutcher and several ecosystem types, including forests and wetlands (swamps,
Forrester, 2018), and can thus help pinpoint areas that could benefit peatlands, and marshes), which represent 12 % and 56 % of the area,
from conservation, or inversely, degraded assemblages in need of respectively (Appendix 1). Most of the remaining area consisted of lawn
restoration measures. Since different indexes like species richness, regularly mowed for recreational purposes, lakes or human in­
ecological uniqueness and FQA can provide complementary informa­ frastructures. The climate of the area is humid continental, with a mean
tion, a multicriteria approach should be used to select conservation annual temperature of 4˚C and mean annual precipitation of about 99
areas, restoration needs, and best management practices (Fleishman mm, 26 % of which falls as snow (Environment Canada, 2020).
et al., 2006; Dubois et al., 2020). Yet, there is limited research on how to Before the site was acquired by Quebec City in 1968, it was char­
combine and apply various conservation indexes to guide UGS man­ acterized mainly by wooded areas, whether they were mesic forests,
agement decisions. wooded peatlands or swamps (37 %), and by open peatlands (17 %), but
This study aimed to identify the conservation and restoration needs it was also used for agriculture (14 %) and gravel extraction (17 %). The
in a large urban park (1,35 km2) in Quebec City (Canada), based on the rest of the park (15 %) was devoted to lawns or infrastructure such as
species richness, floristic uniqueness, and FQA of diverse ecosystems buildings, roads, and parking lots. The abandonment of the gravel
(forests, swamps, wooded peatlands, open peatlands, and marshes). Our quarry in the late 1950s created two oligotrophic lakes separated by a
specific objectives were 1) to assess the land use changes in the park and single road. Today, these lakes are supplied mainly from the ground­
its surrounding landscape over the past decades through historical aerial water table in spring, whereas this flow is reversed during dry periods,

Fig. 1. Location of the urban park named Base de plein-air de Sainte-Foy (BPASF), in Québec (Canada), the 70 sampled plots (square and rectangular) and the
drainage network (white lines). The BPASF limits are indicated by the red line. Forest stands and open wetlands on which the sampling was based are identified in
green and blue.

2
A. Paquette et al. Urban Forestry & Urban Greening 84 (2023) 127925

causing the lakes to recharge the aquifer (Québec City, unpubl. data). years (Lavoie et al., 2022). That study ruled out climatic conditions as an
Before 1948, a drainage network was built in the wet wooded part of the explanation for the rapid forestation in the peatland, leaving anthro­
BPASF (Fig. 1), most likely for forestry-related objectives, but it seems to pogenic causes to explain the shift. Today, the open peatland represents
have been left unmanaged ever since. While there is little water flowing about 5 % of its original area and around 2 % of the park (Fig. 2).
in the drainage channels that are oriented northwest-southeast, water
flows out of the park from the main one, which is oriented northeast-
southwest (Québec City, unpubl. data). Since the city acquired the 2.2. Evolution of land uses
site, some infrastructures, like buildings, parking lots, hiking trails, and
a beach, have been developed to provide recreational activities. To assess the change in soil occupancy over the years in the BPASF
The peatland (0.5 km2) was initiated 6 300 years BP ago and has and its surroundings, we analyzed historic aerial photographs. More
accumulated more than 2 m of organic matter since (Larouche, 1979). precisely, we compared supervised parametric (maximum likelihood)
This ecosystem is threatened by tree encroachment (Blanchard et al., per-pixel classifications of georeferenced aerial photographs from 1948,
2018), a process that started around 1970 (Lavoie et al., 2022). At that 1965, 2008, 2014, and 2019 for both the park and a one-kilometer
time, a marshalling yard was built and a drainage channel dug on the buffer area using the supervised Image classification tools of ArcGIS
peatland, which altered the hydrology and could explain tree Pro (version 2.7.2). The land was classified as either open, wooded, or
encroachment (Blanchard et al., 2018). This tree encroachment started impervious surface areas. We removed the area covered by the artificial
with the already present tamarack (Larix laricina), followed by coloni­ lakes from the analysis. The per-pixel classification uses the spectrum
zation by red maple (Acer rubrum), which became dominant over the information from training samples and applies the classification algo­
rithm to individual pixels (Lu and Weng, 2007). The BPASF and the

Fig. 2. Aerial view of the urban park named Base de plein-air de Sainte-Foy (BPASF) in (a) 1948 (photograph: Québec City, 1948) and in (b) 2020 (photograph:
Google Earth). The open bog ecosystems are delineated in yellow in each photograph.

3
A. Paquette et al. Urban Forestry & Urban Greening 84 (2023) 127925

buffer area were classified separately because pre-tests showed that evaluated based on the georeferenced aerial photographs. To capture
some contrasting habitats of these two areas that appeared dark on the the change in soil occupancy over the years, the state of the plots in 2019
aerial photographs were misclassified when using a single classification. was compared to the state in 1948. From these observations, each plot
In addition, the straightforward per-pixel classification created a salt was categorized as stable (no change) or assigned to one of the seven
and pepper effect in the classified pixels (Lu and Weng, 2007) in the ecological transition states: open peatland to wooded peatland, agri­
buffer area of the two black and white photographs (1948 and 1965), for cultural field to marsh, agricultural field to forest, quarry to forest, open
which we used object-based classification instead (Vogels et al., 2017). area to forest, forest to marsh, and peatland to marsh.
Object-based classification applies the classification algorithm on the To create a fine scale disturbance index, all disturbances of human
merged pixels (objects) obtained through an image segmentation (Lu origin (agriculture, mining activities, roads, trails, ditches, etc.) inside or
and Weng, 2007). We used a minimum of 50 training samples for each of near the plots (using buffer zones of 50 m and 100 m) were manually
the two classification types and for each of the five years assessed. For all delineated in each historical aerial photograph (1948, 1965, 2008,
resulting images, errors were manually corrected, and accuracy evalu­ 2014, and 2019) with ArcGIS. Then, the percentage of disturbed areas
ated based on an error matrix. We built this matrix using on-ground for each aerial photograph were added for each plot at both scales (50
truth points; that is, 100 random stratified points for the BPASF and and 100 m).
200 for its buffer area. The accuracy of our classifications ranged be­
tween 0.89 and 0.97 (Appendix 2). 2.5. Floristic quality assessment

2.3. Vascular vegetation sampling A Floristic Quality Assessment (FQA) was carried out for each plot
using the mean coefficient of conservatism (Mean C; Spyreas, 2019). We
For the vegetation sampling, the BPASF was divided into 17 zones used the Mean C as FQA metric since it is more robust in the presence of
based on forest stand boundaries and identified open wetlands (Québec different types of ecosystems compared to the well-known Floristic
City, unpubl. data). A forest stand represents a homogeneous commu­ Quality Index (Bourdaghs et al., 2006; Kutcher and Forrester, 2018).
nity of codominant tree species, without considering the type of sub­ Mean C is the mean of the coefficient of conservatism (C-values) of all
strate (mineral or organic). A total of 70 plots (400 m2) were sampled species present in the sampling site (Swink and Wilhelm, 1979). A
across all zones (Fig. 1), with 1–7 plots per zone according to its C-value is assigned a priori to each species based on its habitat speci­
respective area (Fig. 1). Sampled plots were positioned randomly within ficity and tolerance to disturbance (Spyreas, 2019). C-values range from
each zone, with a minimum distance of 10 m between plots or from the 0 to 10, where species exclusively found in undisturbed natural eco­
zone boundaries. We mostly used square plots, but 14 were rectangular systems will have a high C-value, while generalist species that can
(same sampling area) to capture only a homogeneous portion of the tolerate disturbance will have a low C-value (Swink and Wilhelm, 1994).
plant community and avoid ecotones (Fig. 1). Since there are no C-values available for the province of Québec, we
Vegetation sampling was carried out in July and August 2019. In averaged C-values from southern Ontario (Oldham et al., 1995), a
each plot, the cover of all vascular plant species (herbs, forbs, vines, neighboring province, and the two nearest regions in Northeastern
lianas, shrubs, tree seedlings and saplings) was visually assessed using 7 United States (Northern Appalachian and Atlantic Maritime Highlands,
classes: <1 %, 1–5 %, 6–10 %, 11–25 %, 26–50 %, 51–75 %, 76–100 %. Eastern Great Lakes and Hudson Lowlands; Faber-Langendoen et al.,
Plant nomenclature follows Database of Vascular Plants of Canada 2019). Since the three databases did not have C-values for Sphagnum
(VASCAN, Brouillet et al., 2019). Species were classified into three species, we averaged the C-values from Ohio (Andreas et al., 2004) and
exclusive groups according to their origin and behavior: native, exotic, New Jersey (Walz et al., 2018). The Sphagnum species C-values were
or exotic invasive following VASCAN and Lavoie (2019). Species affinity validated based on distribution information available for Québec
for wetlands was assigned using wetland indicator status categories for (Ayotte and Rochefort, 2019). Traditionally, in FQA, exotic species are
the Northcentral and Northeast wetland region found in the PLANTS either excluded or automatically attributed a C-value of 0 (e.g., Spyreas
database (USDA and NRCS, 2021). Species classified as "obligate" or et al., 2012; Bried et al., 2013). Numerous scholars have voiced concern
"facultative wetland" were considered wetland species in further about this practice, arguing that C-values should be attributed based on
analyses. species fidelity and not origin (e.g., Cohen et al., 2004; Matthews et al.,
2015; Kutcher and Forester, 2018). It follows that exotic species should
2.4. Local environmental conditions and disturbance have a low C-value, but not automatically a C-value of 0. In this study,
we assigned a C-value of 0 to exotic species that are considered invasive
In each plot, we evaluated microtopography (flat, low, medium, or and harmful for biodiversity, according to Lavoie (2019), but for the
high soil surface variation), drainage index (1 for well-drained to 6 for other exotic species, we assigned C-values based on their level of inva­
poorly drained sites), type of mineral soil (13 categories based on soil siveness following the southern Ontario classification (Appendix 3;
texture), forest stand structure (even or uneven-aged), and forest stand Havinga, 2004).
successional stage (open, stand initiation, stem exclusion, understory
reinitiation, or steady-state). All trees and snags with a diameter ≥10 cm 2.6. Data analysis
at breast height (DBH) were counted. Mean canopy openness was
calculated based on four readings per plot from a Lemmon’s model C Prior to statistical analyses, each plot was assigned to one of the
spherical densiometer. Canopy dominance (open, deciduous, conifer, following ecosystems, based on soil and vegetation indicators evaluated
mixed) was noted. We measured the thickness of the organic matter during the field campaign (Bazoge et al., 2014): forest, swamp, wooded
horizon at the center of each plot. In 2020, a 10-cm long soil core was peatland, open peatland, and marsh (Appendix 1). For the analysis
collected in each plot to calculate soil conductivity and pH. In the lab­ comparing the ecosystems, the open peatland and the wooded peatland
oratory, the soil samples were oven-dried for 48 h at 37 ◦ C, then crushed plots were grouped.
(mortar) and sieved (2 mm mesh). Soil pH and electrical conductivity All statistical analyses were performed using R (version 4.0.3; R Core
were determined in a 5:1 deionized water: soil suspension using a Hanna Team, 2020).
Instruments 98130 Waterproof pH/conductivity meter.
The altitude of each plot was obtained from the Quebec City 2017 2.6.1. Diversity and unicity
digital terrain model (Québec City, unpubl. data). The distance in meters We calculated species richness as the number of distinct vascular
between each plot and the nearest road was calculated in ArcGIS. The species (including trees) in a specific plot. We also calculated the
presence or absence of past agricultural activities for each plot was Simpson’s evenness index (D divided by species richness) in each plot

4
A. Paquette et al. Urban Forestry & Urban Greening 84 (2023) 127925

using the following formula for Simpson’s reciprocal index: D = best tree was selected based on parsimony, low relative error (high
∑ 1
, where n is the total cover of a particular species and N is predictive power), and high cross-validated relative error (high
n(n− 1)/(N(N− 1))
explanatory power). Regression trees were calculated with the function
the total cover of all species. The Simpson’s evenness index (D/species
and package ’mvpart’ (Therneau et al., 2014).
richness) defines the probability that two randomly selected individuals
will be of the same species (see Simpson, 1949). It produces values from
2.6.2. Comparison of diversity indexes between ecosystem types
0 to 1, where low values indicate that a few species are dominating while
To test whether there was a significant difference between ecosystem
high values indicate that a more even number of individuals belongs to
types for each diversity index (species richness, evenness, and LCBD), we
each species present. The uniqueness of the plant assemblage of each
used Kruskal-Wallis tests (one-way ANOVA on ranks; Gibbons, 1985)
plot and each type of ecosystem (open and wooded peatlands grouped
since the conditions of normality and equal variance were not met.
together due to the low number of open sites) were estimated using the
When a difference was detected, Dunn’s Kruskal-Wallis multiple com­
Local Contribution to Beta Diversity (LCBD) values. LCBD values
parison tests with Bonferroni correction were used to identify the
represent the relative contribution of a given site to the beta diversity of
significantly different ecosystems. Kruskal-Wallis tests were carried out
a group of sites and range from 0 to 1 (Legendre and De Caceres, 2013).
with the function ’Kruskal.test’ and ’dunnTest’ from the ’FSA’ package
High LCBD values indicate a unique site in terms of floristic composi­
(Ogle et al., 2020). Conditions of applications were assessed visually
tion. The calculation of LCBD values was carried out with the ’beta.div’
with a Q-Q plot. Normality was verified with a Shapiro-Wilk test using
function in the ’adespatial’ package (Dray et al., 2020) using log-chord
the ’shapiro.test’ function from the ’stats’ package (R Core Team, 2020).
transformed species abundance data. LCBD indexes were tested for
Residual heteroskedasticity was assessed with a Breusch-Pagan test
significance using 9 999 independent random permutations (Legendre
using the ’bptest’ function from the ’lmtest’ package (Zeileis and Hot­
and De Caceres, 2013) and p-values were corrected using Holm’s
horn, 2002). A one-way ANOVA was used to test for differences in Mean
method (Dray et al., 2020). Spearman rank correlation was used to test
C values between the ecosystems. Tukey’s Honest Significant Difference
the correlation between species richness and LCBD values.
multiple comparison test was used to identify ecosystems that signifi­
Regression tree analyses were used to identify the most influential
cantly differed in Mean C values. ANOVA and Tukey post-hoc tests were
variables (Table 1). for grouping the sites based on their LCBD value,
performed with the function ’aov’ and ’TukeyHSD’ from the ’stats’
across all plots as well as for each of the four ecosystem types considered
package (R Core Team, 2020).
individually. Regression trees are a robust compromise between linear
models and nonparametric approaches that deal with nonlinear re­
2.6.3. Identification of conservation and restoration needs
lationships and high order interaction (De’ath and Fabricus, 2000). The
To help identify plots in need of conservation or restoration mea­
sures, a scatter plot showing the sampled plots’ Mean C and LCBD values
Table 1 was used. Two thresholds were added to the scatter plot, one being the
Description of all explanatory variables used for predicting LCBD values with Mean C value of 3.5 which is a threshold commonly used to determine
regressions tree analyses. whether a floristic community is of good floristic quality (Herman et al.,
Explanatory variables Units or categories 2001; Rocchio, 2007; U.S. Fish and Wildlife Service, 2019) and the other
being the mean of all LCBD plot values which can help visually identify
Site characteristics
Microtopography Flat, low, medium, or high
plots with an important uniqueness for the study area as suggested by
Drainage 1 (well-drained) to 6 (poorly drained) Legendre and De Caceres (2013). More precisely, plots with a low Mean
Altitude m C value combined with high uniqueness (rare communities dominated
Ecological transition Open peatland to wooded peatland, agricultural by exotic or ruderal species) were identified as candidates for restora­
field to marsh, agricultural field to forest, quarry
tion, whereas those with high values for both parameters (rare com­
to forest, open area to forest, forest to marsh,
peatland to marsh, or no change munities composed with specialist or sensitive species) were considered
Forest stand characteristics candidates for conservation measures. Plots with low values for both
Canopy openness % indicators could be selected for further recreational use to enhance
Canopy dominance Open, deciduous, conifer, mixed
visitors’ experience in the park.
Structure Even or uneven-aged
Successional stage Open, stand initiation, stem exclusion,
understory reinitiation, or steady-state 3. Results
Number of trees with DBH Total per plot
≥ 10 cm 3.1. Evolution of land use
Number of snags with DBH Total per plot
≥ 10 cm
Soil characteristics Since 1948, land use surrounding the BPASF has changed from
Organic matter thickness cm agricultural to urbanized, with the landscape now dominated mainly by
Type of mineral soil 13 categories based on soil texture impervious surfaces (Figs. 2 and 3), as shown by a 62 % decrease in open
Soil pH
areas and a 56 % increase in impervious surfaces. In the BPASF itself,
Soil electroconductivity mS/cm
Disturbance index
tree encroachment occurred at the expense of open areas (Fig. 3), both in
Distance to nearest road m the abandoned agricultural fields and the peatland (Fig. 3). Wooded
Past agricultural activities Presence or absence areas increased by 38 %, while open areas decreased by 35 % (Fig. 3).
Disturbance index (50 m) % The surface area of the open peatland decreased from 0.5 km2 in
Disturbance index (100 m) %
1948–0.02 km2 in 2020, a contraction of 95 % (Fig. 2). Wooded areas in
Floristic composition
Mean C Mean of the coefficient of conservatism (C- the buffer zone and impervious surfaces in the BPASF itself remained
values) of all species present in a plot stable over time (Fig. 3).
Facultative wetland species %
cover
3.2. Floristic hotspots
Obligates wetland species %
cover
Obligates and facultative % A total of 300 species (vascular and Sphagnum spp.) were identified
wetland species cover in the BPASF (Appendix 4). The average richness per plot was 32 ± 1
Exotic species cover % species, with the lowest richness in a marsh plot (12 species) and the
Exotic invasive species cover %
highest in a swamp plot (61 species). Species richness differed

5
A. Paquette et al. Urban Forestry & Urban Greening 84 (2023) 127925

Fig. 3. Land use classification maps

bases on aerial photographs taken in (a)
1948, (b) 1965, (c) 2008, (d) 2014 and
(e) 2019 and (f) percentage area classi­
fied as open, wooded, or impervious
surfaces in the BPASF (solid lines) and
the 1-km buffer area (dashed lines) (see
Appendix 2 for unclassified photo­
graphs). In the land use classification
sub- figures, the BPASF is delineated
along with a buffer area of 1 km. Open
areas are mainly agricultural fields, but
although some are open peatland (see
Fig. 2). The artificial lakes were
removed from the analysis (white
areas).

significantly between ecosystems (Kruskal-Wallis χ2 = 24.3, df = 3, ecosystems (Kruskal-Wallis χ2 = 20.6, df = 3, p < 0.001; Fig. 4c). Marsh
p < 0.0001), with forest and swamp plots having significantly more plots had significantly higher LCBD values than swamps and peatlands
species than peatland plots (Fig. 4a). Swamp plots had the highest while forest plots had significantly higher LCBD values than those in
variability in species richness (19–61 species). The swamp plots peatlands (Fig. 4c). The highest LCBD value was in the open peatland
harboring the highest number of species were located close to the and the lowest in the wooded one. The highest LCBD values in the
northwest limits of the BPASF (Fig. 4b). The evenness values varied peatland were found in four plots located in the largest open peatland
significantly between ecosystems (Kruskal-Wallis χ2 = 18.8, df = 3, area (Figs. 2 and 4d). The floristic composition of two plots in the open
p < 0.001), but overall values were low for all plots (x‾ = 0.22 ± 0.01). peatland (plots 37 and 38) and three plots in marshes (plots 15, 55 and
Plot Simpson’s evenness was 1.5 times higher for peatlands than for 57) were identified as significantly unique in the BPASF (9 999 per­
forests and swamps (p < 0.01), with marshes showing intermediate mutations, Holm’s correction, p < 0.05, Fig. 4d). There was no signifi­
values (0.11–0.45; Appendix 5 and 6). cant correlation between species richness and LCBD values (ρ = 0.17;
p > 0.05).

3.3. Unique plots and ecosystems

There was a significant difference in LCBD values between

6
A. Paquette et al. Urban Forestry & Urban Greening 84 (2023) 127925

Fig. 4. Differences in (a) species richness and (c) Local Contribution to Beta Diversity (LCBD) between ecosystems. The peatland ecosystem includes open and
wooded peatland plots. Shown here are mean richness (black squares), median (line), 25–75 % quartiles (boxes), ranges (whiskers), and outliers (black circles).
Different letters indicate a significant difference between ecosystems (p ≤ 0.05; Dunn’s Kruskal-Wallis multiple comparison tests with Bonferroni correction). Sig­
nificant differences are indicated with asterisks (significance codes: p ≤ 0.05*, p ≤ 0.01**, p ≤ 0.001***). Bubble map of (b) species richness (radius = species
richness / 3) and (d) uniqueness (LCBD) (radius = LCBD values*650) at each plot. Plots with significant ecological uniqueness values are indicated with asterisks
(significance codes: p ≤ 0.05*, p ≤ 0.01**). The plot number of plots with ecological uniqueness values significantly different from others is indicated on the bubbles
for further reference.

3.4. Cover of wetland, exotic and rare species had a very low cover of wetland obligate species (0.5 %) while about
half of the plant cover was composed of facultative wetland species.
The cover of wetland species was high in all three wetland ecosys­ Exotic species were mainly found in the forest and marsh plots (mean
tems (mean species cover ranged from 75.7 % to 92.4 %), with the marsh species cover 7.7 % and 15.3 %, respectively), where very few were
plots having the highest wetland species cover (Table 2). The forest plots invasive. No exotic species were identified in the open peatland, but

Table 2
Mean species cover (% ± standard error) in terms of wetland status and origin for all the plots in each ecosystem. For the peatland ecosystem, mean species cover was
calculated for three subgroups: open peatland plots, open peatland plots with a significant LCBD value (Fig. 6), and wooded peatland plots. For the marsh ecosystem,
mean species cover was calculated for one subgroup: the plots with a significant LCBD value (Fig. 6).
Wetland status Origin

Type of ecosystem Obligate Facultative Total Exotic Invasive Total

Forest 0.5 ± 0.1 49.0 ± 5.4 49.5 ± 5.4 7.5 ± 2.8 0.2 ± 0.1 7.7 ± 2.8
Swamp 5.1 ± 2.3 72.4 ± 3.0 77.5 ± 3.0 1.0 ± 0.5 0.5 ± 0.3 1.5 ± 0.6
Peatland 6.0 ± 3.0 69.7 ± 3.1 75.7 ± 2.9 0.1 ± 0.0 0.0 ± 0.0 0.1 ± 0.0
Open 31.4 ± 10.8 60.8 ± 11.3 92.2 ± 4.1 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0
Significant LCBD 44.4 ± 22.5 55.4 ± 22.5 99.8 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0
Wooded 0.2 ± 0.1 71.8 ± 2.8 72.0 ± 2.8 0.1 ± 0.1 0.0 ± 0.0 0.1 ± 0.1
Marsh 28.7 ± 7.6 63.7 ± 8.0 92.4 ± 2.7 11.8 ± 6.2 3.5 ± 1.3 15.3 ± 5.7
Significant LCBD 38.9 ± 10.5 51.4 ± 12.7 90.3 ± 2.6 17.4 ± 14.5 5.8 ± 2.4 23.2 ± 12.2

7
A. Paquette et al. Urban Forestry & Urban Greening 84 (2023) 127925

some were present in the wooded peatland. Peatland (either open or < 21.3 %) (Fig. 5e). Conversely, in the peatland, uniqueness can be
wooded) was the only ecosystem with no invasive species (Table 2). explained by an open canopy (≥ 58.4 %) and forest stand characteristics
(Fig. 5d). Peatland communities with the highest uniqueness were found
in two open peatland plots with no trees present and corresponded to the
3.5. Influential variables in plots unicity two plots with significant LCBD values (Figs. 4d and 5d). In the marsh,
the three plots with the highest uniqueness had a soil electro­
The regression tree analysis conducted on all plots showed that an conductivity ≥139.5 mS/cm and corresponded to the plots with signif­
open canopy (openness ≥ 25.7 %) combined with the presence of icant LCBD values (Figs. 4d and 5c).
wetland species (cover > 16.8 %) led to unique plots (Fig. 5a). Canopy
openness was overall the most discriminating variable explaining plot
uniqueness (Fig. 5a). Sites with the lowest uniqueness were mainly 3.6. Identification of conservation and restoration needs
wooded peatland and swamp plots, while the highest uniqueness was
mainly found in the open peatland and marsh plots. All the open peatland plots were above the thresholds for habitat
Among the forest plots, those with the highest uniqueness had a quality (high Mean C values) and mean LCBD, suggesting a potential
floristic composition with a high tolerance to disturbance (Mean C < need for conservation measures, with the two open peatland plots with
3.8) (Fig. 5b). The four swamp plots with the highest uniqueness had a significant LCBD values being the conservation priorities (Fig. 6). The
soil pH ≥ 5.2 (Fig. 5e), while the swamp plots with a low uniqueness wooded peatland plots with a coniferous canopy were close to the mean
were associated with a low soil pH (< 5.2) and a dense canopy (openness LCBD threshold (plots 68, 70, 71), while the wooded peatland plots with

Fig. 5. Regression trees for predicting LCBD values for (a) all plots (Error: 0.268; CV: 0.493), (b) forest plots (Error: 0.186; CV: 1.78), (c) marsh plots (Error: 0.125;
CV: 1.03), (d) peatland (open and wooded) plots (Error: 0.0301; CV: 0.296), (e) swamp plots (Error: 0.209; CV: 0.851) from the set of explanatory variables (Table 1).
Each leaf shows the mean LCBD value and the number of plots in the node. When there is more than one split, branch lengths are proportional to variable importance
(reduction in error due to the split). In each tree, the rightmost branch has the highest mean LCBD value.

8
A. Paquette et al. Urban Forestry & Urban Greening 84 (2023) 127925

Fig. 6. Uniqueness (LCBD) vs Mean C (tolerance to disturbance) values for all plots (n = 70). Plot numbers are indicated on circles. Significant LCBD plot indices are
indicated with asterisks (significance codes: p ≤ 0.05 *, p ≤ 0.01 **).

a deciduous canopy were below the mean LCBD threshold (Fig. 6). The continuous metric, sites not statistically identified as significantly
three marsh plots showing ecological uniqueness values that differ from unique could nevertheless be of interest biologically. By plotting
plots of other ecosystems had a tolerance to disturbance (Mean C values) uniqueness and Mean C values, we also identified open peatland as the
at or below the habitat quality threshold (Fig. 6). These three plots ecosystem most in need of conservation efforts, which should be pro­
should therefore be prioritized for restoration actions. Most of the tected from any potential impacts and public activities should be avoi­
BPASF plots were in the area in which no development for recreational ded. The three marshes were identified as needing restoration. Contrary
use is recommended (Fig. 6) because floristic composition integrity to our expectation, sites with a high ecological unicity (high LCBD) were
would be at risk in the presence of new recreational development (low not those with high species richness, as richness hotspots were found in
tolerance to disturbance), but these plots would not be a priority when forests and swamps. These findings highlight the importance of preser­
targeting conservation of BPASF diversity (low LCBD). In terms of ving all the types of ecosystems to maintain the floristic diversity of the
identifying locations for future development for recreational use, two park. Overall, our results show clearly that plant community unicity
peatland plots (wooded with deciduous canopy) stood out, due to their combined with a few other criteria (e.g., species richness and floristic
low ecological uniqueness (below the mean LCBD) and high tolerance to quality assessment) can be used to assess management priorities for
disturbance (Mean C below the habitat quality threshold; plots 51 and areas where both conservation and recreational activities are
52; Fig. 6). These two plots would be the least affected by an increase in mandatory.
outdoor activities.

4. Discussion 4.1. Open peatland, an ecosystem to conserve

This study aimed to identify the conservation and restoration needs At the BPASF, we identified open peatland as the ecosystem to be
of diverse ecosystems in a large urban park based on species richness, conserved, but its integrity is threatened by tree encroachment. Indeed,
uniqueness of plant communities (LCBD), and floristic quality assess­ our aerial photograph analysis showed rapid tree encroachment on the
ment (Mean C). Our results indicate that unique plant communities open peatland, with the non-wooded sector having decreased by 95 % in
(significant LCBD) were found in the open peatland and marshes. As surface area since 1948. Tree encroachment is increasingly being
predicted, in the open peatland, unique communities were dominated by observed in open wetlands, where heliophilous and waterlogging-
native wetland species (> 99 % of the cover). On the other hand, about tolerant open peatland specialist species are often replaced with shade
one-fifth of the cover of unique plant communities in marshes was and drought- tolerant terrestrial species (Favreau et al., 2019; Langdon
composed of exotic species. Our analyses of historical photographs et al., 2020; Pellerin et al., 2021). Tree encroachment is primarily
indicate that the unique sites in the open peatland have not undergone in associated with human disturbances like drainage, nutrient deposition,
situ land use change over the years, while those in marshes originated and land use change (Berendse et al., 2001; Pellerin and Lavoie, 2003;
from past agricultural fields. We identified canopy openness as the most Pasquet et al., 2015; Beauregard et al., 2020), which leave wetlands that
influential variable explaining plot uniqueness across all ecosystems, lie isolated in an anthropogenic matrix, like the BPASF peatland,
and especially in the peatland, indicating that new recreational devel­ particularly at risk.
opment should be avoided in open areas. Because uniqueness is a In the BPASF, tree encroachment started in the southwestern portion
of the site, at the ecotone with the swamp and wooded peatland. The

9
A. Paquette et al. Urban Forestry & Urban Greening 84 (2023) 127925

drainage network and forestry roads built before 1948 (the year of our disturbances, recreational use and development in areas with higher
first aerial photograph) in the wooded area there likely triggered, or at canopy openness and higher soil wetness. On the other hand, the
least facilitated, the phenomenon. The increase in human disturbances species-rich forest plots are in areas that have sustained various in situ
(e.g., expanding railways, ditch digging, and building recreational trails) land use changes and disturbances over the years. This land-use legacy
and impervious surfaces in and around the park in the 1960 s and 1970 s as well as the linear shape and reduced core area of forests at the BPASF
coincides with the intensification of tree encroachment and must have created a suitable environment for the colonization of exotic, invasive,
played a crucial role. Such disturbances often induce soil compaction, and ruderal species, thus reducing the conservation value of these plots.
lower regional water levels, and/or increase nutrient availability, Thus, forests may not be the next priority after open peatlands for
thereby stimulating tree growth and reducing the competitive advan­ conservation efforts, but swamps would be because of their high rich­
tage of open peatland species (Ohlson et al., 2001; Heijmans et al., 2013; ness (especially in wetland species) and absence of past disturbances. As
Williams-Mounsey et al., 2021). a matter of fact, swamps encompassed more species of high conservation
If forestation of the BPASF peatland continues, a decrease in regional values (Mean C) than forests (Fig. 6), which should foster willpower for
floristic diversity is expected, as the flora of the open area will homog­ their protection within the park. These results indicate that biodiversity
enize with the surrounding wooded area (Pasquet et al., 2015; Favreau hotspots should be analyzed in conjunction with other criteria and
et al., 2019; Pinceloup et al., 2020). Our results show that once trees combined with a good knowledge of local history, to properly assess
become dominant in a peatland, the uniqueness of the plot decreases ecosystem value and guide management actions in the park.
along with the cover of wetland species, while exotic species establish.
The uniqueness also decreases once the canopy becomes dominated by 5. Future applications
deciduous species like Acer rubrum, which is associated with a longer
period of time elapsed since initial encroachment (Lavoie et al., 2022). Although the results obtained in this case study are specific to the
These results are consistent with previous research carried out in the BPASF, other parks can apply the methodology developed here to help
BPASF that compared the peatland plant communities between 1978 establish priorities and facilitate management decisions for one or more
and 2015 and found that forest species replaced wetland ones (Blan­ types of ecosystems. By using various floristic community measures, we
chard et al., 2018). Once trees invade, they promote conditions favor­ were able to identify areas that could benefit from conservation. Con­
able to their growth (Heijmans et al., 2013) and the ecosystem is servation priorities include areas with unique floristic composition
unlikely to return to its previous unforested state naturally. Canopy (LCBD) due to the presence of rare and specialist species, at-risk floristic
openness should thus be maintained at the site, whether considered as a communities due to a low tolerance to disturbances, and high species
consequence or a cause of ecological uniqueness. Tree cutting could be richness. In contrast, areas with a high floristic uniqueness due to the
undertaken to reduce water loss from tree evapotranspiration and pre­ presence of exotic and invasive species could be targeted for restoration
cipitation interception and, in turn, induces a rise in water table level actions unless drivers that favored these species cannot be modified, as is
(Jutras et al., 2006). The opening of the canopy and the increased soil the case for the BPASF marshes that are continuously receiving enriched
humidity can limit tree encroachment, since these conditions favor road runoff water. The scatter plot of the Mean C and LCBD values is a
Sphagnum species and other typical open bog species over trees, and tool that can help land managers make decisions by allowing them to
reduce seed rain pressure from adjacent wooded areas (Fenton and visualise where various sites are located on the spectrum of “restoration
Bergeron, 2006; Langdon et al., 2020). Such interventions would need to vs. conservation vs. recreational development”. This tool can be updated
be recurrent (Kamocki et al., 2017). with new information easily, which is especially important in an urban
setting where the species composition of an area can change rapidly due
4.2. Is restoration of the marshes the best solution? to disturbances. In addition, the regression tree can help identify plot
characteristics that best predict its uniqueness and thus inform managers
According to our results, three marsh plots should be prioritized for on operation options. Our methodology also showed that taking into
restoration efforts. These plots are in topographic depressions that account land-use legacy is essential to better plan management prac­
receive parking and road surface water runoff. Such runoff is known to tices, as past land-uses can still affect current plant communities.
be charged with salts (Kelly et al., 2012), nutrients (Carey et al., 2013) Finally, the use of uniqueness, in combination with other indices, is
and heavy metals (Gill et al., 2014), which could explain the higher soil particularly interesting as it is not region and taxon specific and can be
electroconductivity of the three plots (Fig. 5c) and the presence of spe­ easily computed from lists of species already available (Dubois et al.
cies highly tolerant to disturbances. In fact, these marshes could act as 2020).
natural retention basins, reducing the charge of sediments, contami­
nants, and bacteria in water through sedimentation, ionic exchange, Funding
agglomeration, and biological uptake (Robotham et al., 2021). To allow
indigenous and sensitive species to thrive, the runoff would need to be This research was supported by the MITACS Accelerate program
intercepted and restoration efforts would need to improve the soil and (grant to A.P.), the Ouranos consortium (project 554030) via the Quebec
water conditions (Wenger et al., 2009). In this context, restoration government 2013–2020 action plan on climate change, the watershed
measures may not be worthwhile unless the abundance of exotic and organization of the study region (OBV de la Capitale), and Quebec City.
invasive species expands and threatens the neighboring ecosystems.
CRediT authorship contribution statement
4.3. Biodiversity hotspots
Audrey Paquette: Conceptualization, Methodology, Field work,
As in previous studies, our results showed that plots with a high Formal analysis, Writing- Reviewing and Editing Stéphanie Pellerin:
uniqueness differed from plots with high species richness (Heino and Supervision, Writing- Reviewing and Editing. Monique Poulin: Funding
Grönroos, 2017; Landeiro et al., 2018; Dubois et al., 2020). Therefore, acquisition, Supervision, Writing- Reviewing and Editing.
evaluating species richness can provide complementary information
regarding the conservation value of a site and should be considered in Declaration of Competing Interest
UGS management. In the BPASF, we found the highest number of species
in swamps and forests. The swamp plots that were species-rich had The authors declare that they have no known competing financial
higher habitat heterogeneity, soil wetness, and canopy openness than interests or personal relationships that could have appeared to influence
other swamp plots. Their conservation can be achieved by limiting the work reported in this paper.

10
A. Paquette et al. Urban Forestry & Urban Greening 84 (2023) 127925

Acknowledgements Stéphane Daigle for statistical advice, and Karen Grislis for English
revision. We also thank the field and laboratory assistants.
We thank Denis Bastien for his help in identifying plant specimens,

Appendix 1

Wetlands and forest at the study site (Base de plein air de Sainte-Foy).

Appendix 2

11
A. Paquette et al. Urban Forestry & Urban Greening 84 (2023) 127925

.
Land use analysis: unclassified aerial photographs, source of aerial photographs and analysis details. Unclassified aerial photographs used for the
land use analysis.
Percentage area classified as open, wooded, and impervious surfaces and user accuracy assessment of the classification in the BPASF and the 1-km
buffer area based on aerial photographs taken in 1948, 1965, 2008, 2014 and 2019.

Year Area Type of classification Land cover (%) User accuracy

Open Wooded Impervious

1948 BPASF Per-pixel 52.4 40.2 7.4 0,92

Buffer Object-based 77.8 17.3 4.9 0,97
1965 BPASF Per-pixel 42.6 47.1 10.4 0,93
Buffer Object-based 68.6 15.3 16.2 0,95
(continued on next page)

12
A. Paquette et al. Urban Forestry & Urban Greening 84 (2023) 127925

(continued )
Year Area Type of classification Land cover (%) User accuracy

Open Wooded Impervious

2008 BPASF Per-pixel 20.2 76.4 3.4 0,94

Buffer Per-pixel 28.0 22.4 49.6 0,90
2014 BPASF Per-pixel 15.7 81.2 3.1 0,95
Buffer Per-pixel 20.9 23.5 55.6 0,91
2019 BPASF Per-pixel 17.7 78.0 4.3 0,94
Buffer Per-pixel 15.4 23.9 60.7 0,89

Source of orthophotographs used for the classification of the BPASF and 1-km buffer as open, wooded, and impervious surfaces.

Year Source of orthophotographs

1948 Quebec City. Orthophotographies noir et blanc de la ville de Québec - 1948 (50 cm). [orthophotograph]. 1:25 000. Roll A17115. Géoindex. 1948.
1965 Quebec City. Orthophotographies noir et blanc de la Ville de Québec - 1963–65 (25 cm). [orthophotograph]. 1:15 480. Géoindex. 1965.
2008 Communauté métropolitaine de Québec (CMQ). Orthophotographies couleur de la Communauté métropolitaine de Québec - 2008 (30 cm). [orthophotograph]. Géoindex.
2008.
2014 Quebec City. Orthophotographies couleur de la ville de Québec - 2014 (10 cm). [orthophotograph]. Géoindex. 2014.
2019 Quebec City. Orthophotographies couleur de la ville de Québec - 2019 (10 cm). [orthophotograph]. Géoindex. 2019.

Appendix 3

Coefficient of conservatism (CC) assigned to exotic species according to the level of invasiveness classification developed by Havinga (2004).

C-value Level of invasiveness

1 Highly invasive species with the potential to dominate a site or a niche

2 Moderately invasive species that do not pose a serious threat to natural areas
3 Potentially exotic invasive species to monitor

Appendix 4

Species list, origin, wetland status, and coefficient of conservatism (C-values). Species origin (indigenous or exotic) follows VASCAN classification.
Exotic invasive species are harmful for biodiversity according to Lavoie (2019). Species affinity for wetlands was assigned using wetland indicator
status categories found for the Northcentral and Northeast wetland region in the PLANTS database (USDA and NRCS, 2021). Species classified as
"obligate" or "facultative wetland" were considered wetland species. Species C-values is the average of values from southern Ontario (Oldham et al.
1995) and Northeastern United States (Northern Appalachian and Atlantic Maritime Highlands and Eastern Great Lakes and Hudson Lowlands;
Faber-Langendoen et al., 2019). Sphagnum species C-value is the average of the C-values found in Ohio (Andreas et al. 2004) and New Jersey (Walz
et al. 2018) databases for Sphagnum species.

Name Origin Wetland status C-value

Abies balsamea Indigenous Facultative 4.7

Acer negundo Exotic Facultative 1.0
Acer pensylvanicum Indigenous Non-indicator 6.0
Acer rubrum Indigenous Facultative 3.3
Acer saccharum Indigenous Non-indicator 5.0
Acer saccharinum Indigenous Obligate 5.7
Acer spicatum Indigenous Non-indicator 6.3
Achillea millefolium Exotic Non-indicator 3.0
Agrimonia gryposepala Indigenous Non-indicator 3.3
Agrostis stolonifera Exotic Facultative 3.0
Alisma triviale Indigenous Obligate 3.7
Alnus incana Indigenous Facultative 5.0
Amelanchier arborea Indigenous Non-indicator 5.0
Amelanchier bartramiana Indigenous Facultative 6.7
Amelanchier laevis Indigenous Non-indicator 4.0
Andromeda polifolia Indigenous Obligate 9.3
Antennaria howellii Indigenous Non-indicator 2.7
Apocynum androsaemifolium Indigenous Not indicator 2.3
Aralia nudicaulis Indigenous Non-indicator 4.7
Arctium lappa Exotic Not indicator 3.0
Arctium minus Exotic Not indicator 3.0
Arisaema triphyllum Indigenous Facultative 5.0
(continued on next page)

13
A. Paquette et al. Urban Forestry & Urban Greening 84 (2023) 127925

(continued )
Name Origin Wetland status C-value

Aronia melanocarpa Indigenous Facultative 5.7

Asclepias syriaca Indigenous Non-indicator 1.6
Athyrium filix-femina Indigenous Non-indicator 4.3
Berbersi linnaeus Exotic Non-indicator 2.0
Betula alleghaniensis Indigenous Facultative 5.7
Betula papyrifera Indigenous Non-indicator 3.3
Betula populifolia Indigenous Facultative 3.7
Bidens frondosa Indigenous Facultative 2.3
Brachyelytrum aristosum Indigenous Non-indicator 6.7
Calamagrostis canadensis Indigenous Facultative 4.3
Calystegia sepium Indigenous Non-indicator 2.0
Campanula rapunculoides Exotic Non-indicator 2.0
Carex aquatilis Indigenous Obligate 7.3
Carex arctata Indigenous Obligate 5.3
Carex baileyi Indigenous Facultative 6.0
Carex brunnescens Indigenous Facultative 6.0
Carex communis Indigenous Non-indicator 5.7
Carex conoidea Indigenous Facultative 5.0
Carex crawfordii Indigenous Facultative 4.7
Carex crinita Indigenous Facultative 4.3
Carex debilis Indigenous Facultative 4.0
Carex deweyana Indigenous Non-indicator 5.3
Carex hirta Exotic Non-indicator 3.0
Carex intumescens Indigenous Facultative 5.7
Carex laxiflora Indigenous Non-indicator 5.7
Carex lupulina Indigenous Obligate 6.0
Carex oligosperma Indigenous Obligate 9.0
Carex pallescens Indigenous Facultative 3.0
Carex retrorsa Indigenous Obligate 5.3
Carex scoparia Indigenous Facultative 3.2
Carex trisperma Indigenous Obligate 7.3
Centaurea nigra Exotic Non-indicator 3.0
Cerastium arvense Indigenous Facultative 4.0
Ceratophyllum demersum Indigenous Obligate 3.7
Chamaenerion angustifolium Indigenous Facultative 3.0
Chamaedaphne calyculata Indigenous Obligate 7.3
Chelone glabra Indigenous Obligate 6.3
Chimaphila umbellata Indigenous Non-indicator 7.0
Cicuta bulbifera Indigenous Obligate 5.3
Cicuta maculata Indigenous Obligate 5.0
Cinna latifolia Indigenous Facultative 6.0
Circaea canadensis Indigenous Non-indicator 3.0
Cirsium muticum Indigenous Obligate 7.3
Clematis virginiana Indigenous Non-indicator 3.0
Clintonia borealis Indigenous Facultative 6.3
Convallaria majalis Exotic Non-indicator 2.0
Coptis trifolia Indigenous Facultative 6.7
Cornus alternifolia Indigenous Non-indicator 6.0
Cornus canadensis Indigenous Facultative 6.0
Corylus cornuta Indigenous Non-indicator 5.0
Cornus sericea Indigenous Facultative 3.7
Cuscuta gronovii Indigenous Facultative 3.7
Cypripedium acaule Indigenous Facultative 5.7
Daucus carota Exotic Non-indicator 3.0
Dendrolycopodium obscurum Indigenous Non-indicator 6.0
Dennstaedtia punctilobula Indigenous Non-indicator 4.0
Diervilla lonicera Indigenous Non-indicator 5.0
Dirca palustris Indigenous Facultative 7.3
Doellingeria umbellata Indigenous Facultative 4.0
Drosera rotundifolia Indigenous Obligate 6.0
Dryopteris cristata Indigenous Facultative 7.0
Dryopteris spinulosa Indigenous Facultative 5.7
Eleocharis acicularis Indigenous Obligate 4.3
Eleocharis aestuum Indigenous Obligate 8.0
Eleocharis palustris Indigenous Obligate 5.0
Elymus repens Exotic invasive Facultative 0.0
Epilobium ciliatum Indigenous Facultative 3.0
Epilobium coloratum Indigenous Obligate 2.7
Epipactis helleborine Exotic Non-indicator 3.0
Epilobium palustre Indigenous Obligate 7.3
Equisetum arvense Indigenous Facultative 1.3
Equisetum fluviatile Indigenous Obligate 5.3
Equisetum sylvaticum Indigenous Facultative 6.0
Erigeron strigosus Indigenous Facultative 1.8
Eriophorum virginicum Indigenous Obligate 8.3
Erythronium americanum Indigenous Non-indicator 4.7
(continued on next page)

14
A. Paquette et al. Urban Forestry & Urban Greening 84 (2023) 127925

(continued )
Name Origin Wetland status C-value

Euthamia graminifolia Indigenous Facultative 2.2

Eutrochium maculatum Indigenous Facultative 3.3
Fallopia cilinodis Indigenous Non-indicator 3.3
Fagus grandifolia Indigenous Non-indicator 5.3
Fraxinus americana Indigenous Non-indicator 3.3
Fraxinus nigra Indigenous Facultative 7.0
Fragaria virginiana Indigenous Non-indicator 2.2
Galium labradoricum Indigenous Obligate 8.3
Galium palustre Indigenous Facultative 4.3
Galeopsis tetrahit Exotic Non-indicator 3.0
Gaultheria procumbens Indigenous Non-indicator 5.0
Geum aleppicum Indigenous Facultative 3.0
Geum canadense Indigenous Facultative 3.7
Geum laciniatum Indigenous Facultative 4.0
Geum macrophyllum Indigenous Facultative 7.7
Geum urbanum Exotic Non-indicator 3.0
Glyceria canadensis Indigenous Obligate 5.7
Glyceria grandis Indigenous Obligate 4.3
Glyceria striata Indigenous Obligate 4.0
Gymnocarpium dryopteris Indigenous Non-indicator 6.7
Hieracium vulgatum Indigenous Non-indicator 0.0
Huperzia lucidula Indigenous Facultative 6.3
Hypericum fraseri Indigenous Obligate 7.7
Hypericum perforatum Exotic Non-indicator 2.0
Ilex mucronata Indigenous Facultative 7.0
Ilex verticillata Indigenous Facultative 5.3
Impatiens capensis Indigenous Facultative 3.3
Iris versicolor Indigenous Obligate 4.3
Juncus bufonius Indigenous Facultative 2.2
Juncus effusus Indigenous Facultative 2.7
Kalmia angustifolia Indigenous Facultative 6.7
Kalmia polifolia Indigenous Obligate 9.3
Lactuca biennis Indigenous Facultative 3.7
Lactuca serriola Exotic Facultative 3.0
Larix laricina Indigenous Facultative 5.7
Lathyrus pratensis Exotic Facultative 3.0
Leersia oryzoides Indigenous Obligate 4.7
Lemna minor Indigenous Obligate 2.3
Leucanthemum vulgare Exotic Non-indicator 3.0
Lonicera canadensis Indigenous Non-indicator 6.3
Lonicera hirsuta Indigenous Facultative 7.7
Lonicera morrowii Exotic Non-indicator 1.0
Lonicera tatarica Exotic Non-indicator 1.0
Lotus corniculatus Exotic Facultative 1.0
Lycopus americanus Indigenous Obligate 4.3
Lycopus uniflorus Indigenous Obligate 4.7
Lysimachia borealis Indigenous Facultative 5.3
Lysimachia ciliata Indigenous Facultative 4.0
Lysimachia terrestris Indigenous Obligate 5.3
Lythrum salicaria Exotic invasive Facultative 0.0
Maianthemum canadense Indigenous Non-indicator 4.0
Maianthemum racemosum Indigenous Non-indicator 5.0
Maianthemum trifolium Indigenous Obligate 8.7
Medeola virginiana Indigenous Non-indicator 6.0
Melilotus albus Exotic Facultative 1.0
Mentha canadensis Indigenous Facultative 3.0
Mentha piperita Exotic Obligate 2.0
Mimulus ringens Indigenous Obligate 5.7
Mitella nuda Indigenous Facultative 7.3
Mitchella repens Indigenous Non-indicator 6.0
Monotropa uniflora Indigenous Non-indicator 6.3
Muhlenbergia frondosa Indigenous Facultative 4.0
Myosotis laxa Indigenous Facultative 4.7
Myrica gale Indigenous Obligate 6.7
Nabalus altissimus Indigenous Non-indicator 5.7
Oclemena acuminata Indigenous Non-indicator 6.3
Onoclea sensibilis Indigenous Facultative 2.7
Orthilia secunda Indigenous Facultative 6.3
Osmundastrum cinnamomeum Indigenous Facultative 6.0
ClaytOsmunda claytoniana Indigenous Facultative 5.0
Osmunda regalis Indigenous Facultative 6.0
Oxalis montana Indigenous Non-indicator 7.3
Oxalis stricta Exotic Non-indicator 3.0
Panicum philadelphicum Indigenous Facultative 5.7
ParaThelypteris noveboracensis Indigenous Facultative 5.0
Parthenocissus quinquefolia Exotic Non-indicator 3.0
(continued on next page)

15
A. Paquette et al. Urban Forestry & Urban Greening 84 (2023) 127925

(continued )
Name Origin Wetland status C-value

Persicaria maculosa Exotic Facultative 3.0

Persicaria pensylvanica Indigenous Facultative 2.3
Persicaria sagittata Indigenous Obligate 4.3
Phalaris arundinacea Exotic invasive Facultative 0.0
Phegopteris connectilis Indigenous Non-indicator 6.3
Phleum pratense Exotic Facultative 3.0
Phragmites australis Exotic invasive Facultative 0.0
Picea glauca Indigenous Non-indicator 6.0
Picea mariana Indigenous Facultative 7.3
Picea rubens Indigenous Non-indicator 5.0
Pilosella aurantiaca Exotic Non-indicator 3.0
Pilosella caespitosa Exotic Non-indicator 3.0
Pinus strobus Indigenous Facultative 3.3
Plantago major Exotic Non-indicator 3.0
Poa compressa Exotic Facultative 3.0
Poa palustris Indigenous Facultative 3.7
Polygonatum pubescens Indigenous Non-indicator 6.0
Populus balsamifera Indigenous Facultative 3.0
Populus deltoides Indigenous Facultative 3.3
Populus grandidentata Indigenous Non-indicator 4.0
Populus tremuloides Indigenous Non-indicator 2.3
Potentilla norvegica Indigenous Facultative 1.6
Prunus nigra Indigenous Non-indicator 2.0
Prunus pensylvanica Indigenous Non-indicator 2.3
Prunus virginiana Indigenous Non-indicator 2.7
Pteridium aquilinum Indigenous Non-indicator 2.3
Pyrola elliptica Indigenous Not indicator 5.7
Quercus alba Indigenous Non-indicator 5.7
Quercus macrocarpa Indigenous Non-indicator 5.7
Quercus rubra Indigenous Non-indicator 5.0
Ranunculus abortivus Indigenous Facultative 4.7
Ranunculus acris Exotic Facultative 3.0
Ranunculus recurvatus Indigenous Facultative 5.3
Ranunculus repens Exotic Facultative 2.0
Reynoutria japonica Exotic invasive Facultative 0.0
Rhododendron canadense Indigenous Facultative 4.7
Rhododendron groenlandicum Indigenous Obligate 5.7
Ribes glandulosum Indigenous Facultative 6.3
Ribes lacustre Indigenous Facultative 7.0
Ribes triste Indigenous Obligate 7.0
Rorippa palustris Indigenous Facultative 3.3
Rubus allegheniensis Indigenous Non-indicator 2.3
Rubus idaeus Indigenous Non-indicator 1.3
Rubus pensilvanicus Indigenous Non-indicator 4.0
Rubus pubescens Indigenous Facultative 6.0
Rumex britannica Indigenous Obligate 6.3
Sagittaria latifolia Indigenous Obligate 3.7
Salix alba Exotic Facultative 2.0
Salix discolor Indigenous Facultative 3.0
Salix eriocephala Indigenous Facultative 3.3
Salix fragilis Exotic Facultative 2.0
Salix interior Indigenous Facultative 4.3
Salix lucida Indigenous Facultative 5.0
Salix pellita Indigenous Obligate 6.0
Salix pyrifolia Indigenous Facultative 8.0
Salix triandra Exotic Facultative 3.0
Sambucus canadensis Indigenous Facultative 5.0
Sambucus racemosa Indigenous Non-indicator 5.3
Schizachne purpurascens Indigenous Non-indicator 7.3
Scirpus atrocinctus Indigenous Obligate 4.3
Scirpus cyperinus Indigenous Obligate 3.7
Scirpus microcarpus Indigenous Obligate 4.7
Scutellaria lateriflora Indigenous Obligate 6.0
Silene latifolia Exotic Non-indicator 3.0
Sium suave Indigenous Obligate 5.0
Solidago altissima Indigenous Non-indicator 2.2
Solidago canadensis Indigenous Facultative 2.2
Solanum dulcamara Exotic Facultative 2.0
Solidago flexicaulis Indigenous Non-indicator 6.0
Solidago rugosa Indigenous Facultative 2.7
Sorbus americana Indigenous Facultative 7.7
Sorbus aucuparia Exotic Non-indicator 2.0
Sphagnum angustifolium Indigenous Facultative 6.0
Sphagnum capillifolium Indigenous Facultative 5.0
Sphagnum centrale Indigenous Facultative 8.0
Sphagnum fallax Indigenous Facultative 5.5
(continued on next page)

16
A. Paquette et al. Urban Forestry & Urban Greening 84 (2023) 127925

(continued )
Name Origin Wetland status C-value

Sphagnum fimbriatum Indigenous Facultative 4.0

Sphagnum girgensohnii Indigenous Facultative 5.5
Sphagnum magellanicum Indigenous Facultative 7.0
Sphagnum quinquefarium Indigenous Facultative 7.0
Sphagnum russowii Indigenous Facultative 5.5
Sphagnum squarrosum Indigenous Facultative 6.0
Spiraea alba Indigenous Facultative 3.3
Stachys palustris Exotic Obligate 3.0
Stellaria graminea Exotic Non-indicator 3.0
Streptopus amplexifolium Indigenous Facultative 9.0
Streptopus lanceolatus Indigenous Non-indicator 7.3
Symphoricarpos albus Indigenous Non-indicator 8.0
Symphyotrichum cordifolium Indigenous Non-indicator 4.0
Symplocarpus foetidus Indigenous Obligate 6.3
Symphyotrichum lanceolatum Indigenous Facultative 2.3
Symphyotrichum lateriflorum Indigenous Facultative 2.3
Symphyotrichum puniceum Indigenous Facultative 4.7
Tanacetum vulgare Exotic Facultative 2.0
Taraxacum officinale Exotic invasive Non-indicator 0.0
Taxus canadensis Indigenous Non-indicator 7.3
Thalictrum pubescens Indigenous Facultative 4.0
Thelypteris palustris Indigenous Obligate 4.3
Tiarella cordifolia Indigenous Non-indicator 6.0
Trillium erectum Indigenous Non-indicator 6.3
Trifolium hybridum Exotic Not indicator 3.0
Trifolium pratense Exotic Non-indicator 2.0
Trifolium repens Exotic Non-indicator 2.0
Trillidium undulatum Indigenous Non-indicator 7.7
Tsuga canadensis Indigenous Non-indicator 6.0
Tussilago farfara Exotic Non-indicator 2.0
Typha angustifolia Exotic Obligate 3.0
Typha latifolia Indigenous Obligate 2.3
Ulmus americana Indigenous Facultative 3.3
Uvularia grandiflora Indigenous Non-indicator 7.0
Vaccinium angustifolium Indigenous Non-indicator 4.3
Vaccinium myrtilloides Indigenous Facultative 6.0
Vaccinium oxycoccos Indigenous Obligate 8.7
Valeriana officinalis Exotic Non-indicator 3.0
Veronica arvensis Exotic Non-indicator 3.0
Veronica officinalis Exotic Non-indicator 3.0
Veronica serpyllifolia Indigenous Facultative 2.7
Veratrum viride Indigenous Facultative 5.0
Viburnum edule Indigenous Facultative 8.0
Viburnum lantanoides Indigenous Non-indicator 7.3
Viburnum cassinoides Indigenous Facultative 6.0
Viburnum opulus Indigenous Facultative 1.6
Viburnum rafinesqueanum Indigenous Non-indicator 7.7
Vicia cracca Exotic Non-indicator 1.0
Vicia sepium Exotic Non-indicator 3.0
Vitis riparia Indigenous Facultative 2.0

Appendix 5

Plot diversity and evenness indices values and wetland, exotic and exotic invasives species cover (%).

Plot Ecosystem Species Evenness LCBD Mean Species cover (%)

number richness C
Obligate Facultative Obligate and facultative Exotic Exotic
wetland wetland wetland invasive

1 Peatland 19 0.38 0.010 5.4 0.0 67.4 67.4 0.1 0.0

2 Swamp 33 0.20 0.009 4.9 0.2 57.5 57.6 0.2 0.0
3 Peatland 28 0.23 0.011 4.8 0.0 77.2 77.2 0.0 0.0
4 Peatland 28 0.20 0.010 5.1 0.2 76.6 76.8 0.2 0.0
5 Peatland 21 0.29 0.010 5.8 0.6 63.4 64.1 0.1 0.0
6 Swamp 32 0.19 0.013 4.1 0.1 82.7 82.8 2.7 0.0
7 Swamp 45 0.19 0.012 4.6 1.7 84.1 85.8 0.1 0.0
8 Peatland 22 0.21 0.010 6.9 0.0 58.4 58.4 0.2 0.0
9 Peatland 21 0.23 0.009 5.7 0.2 84.6 84.8 0.0 0.0
10 Peatland 18 0.28 0.010 5.7 0.0 73.7 73.7 0.0 0.0
11 Peatland 27 0.11 0.011 5.2 0.0 44.3 44.3 0.0 0.0
13 Swamp 34 0.12 0.018 3.7 1.1 40.4 41.5 1.8 1.8
(continued on next page)

17
A. Paquette et al. Urban Forestry & Urban Greening 84 (2023) 127925

(continued )
Plot Ecosystem Species Evenness LCBD Mean Species cover (%)
number richness C
Obligate Facultative Obligate and facultative Exotic Exotic
wetland wetland wetland invasive

14 Swamp 20 0.25 0.023 3.9 29.4 61.8 91.2 1.0 8.3

15 Marsh 12 0.25 0.024 3.1 51.8 38.2 90.0 46.3 1.3
16 Forest 39 0.23 0.014 4.6 0.6 31.8 32.5 0.0 0.0
17 Swamp 36 0.23 0.019 4.9 1.9 58.9 60.8 0.6 0.0
18 Forest 36 0.16 0.012 4.7 0.0 58.2 58.2 0.3 0.0
19 Swamp 27 0.16 0.013 4.8 0.4 90.9 91.3 0.0 0.0
20 Forest 39 0.16 0.015 4.4 0.9 64.9 65.8 0.9 0.0
21 Marais 45 0.26 0.019 3.5 4.0 76.9 80.9 8.2 1.0
23 Forest 41 0.25 0.017 3.7 0.0 39.0 39.0 14.0 0.3
25 Peatland 26 0.29 0.009 5.7 0.0 67.5 67.5 0.0 0.0
26 Peatland 21 0.33 0.010 5.4 0.1 72.4 72.5 0.0 0.0
27 Swamp 46 0.14 0.016 4.5 0.2 80.0 80.2 0.8 0.0
28 Swamp 53 0.17 0.013 4.5 0.3 74.4 74.6 0.4 0.0
Plot Ecosystem Species Evenness LCBD Mean Species cover (%)
number richness index C
Obligate Facultative Obligate and facultative Exotic Exotic
wetland wetland wetland invasive

29 Swamp 42 0.21 0.014 5.1 5.7 55.5 61.2 0.0 0.0

30 Forest 41 0.15 0.015 5.0 0.9 61.2 62.0 6.9 0.0
31 Swamp 19 0.28 0.010 5.7 0.0 84.7 84.7 0.0 0.0
32 Swamp 26 0.19 0.009 5.2 0.2 90.5 90.7 0.0 0.0
33 Swamp 25 0.24 0.010 5.3 0.2 91.2 91.4 0.0 0.0
34 Swamp 21 0.25 0.011 5.1 0.0 91.4 91.4 0.0 0.1
35 Marsh 24 0.08 0.018 3.3 3.7 95.5 99.2 2.8 2.0
36 Marsh 31 0.24 0.021 4.2 35.9 63.5 99.5 0.0 4.3
37 Peatland 13 0.32 0.025 7.0 66.9 32.9 99.8 0.0 0.0
38 Peatland 17 0.30 0.023 6.8 21.9 77.9 99.8 0.0 0.0
39 Peatland 28 0.19 0.009 5.4 0.2 85.4 85.6 0.0 0.0
40 Peatland 25 0.23 0.011 5.5 0.3 82.5 82.8 0.0 0.0
41 Swamp 29 0.24 0.010 5.1 0.1 80.6 80.8 0.0 0.0
42 Swamp 41 0.18 0.011 4.6 0.0 71.4 71.4 0.9 0.0
43 Swamp 19 0.27 0.010 5.2 0.2 91.8 92.0 0.0 0.0
44 Peatland 28 0.26 0.012 4.5 0.1 45.6 45.7 0.4 0.0
45 Peatland 27 0.22 0.011 5.5 0.4 62.6 63.0 0.0 0.0
46 Swamp 28 0.23 0.010 6.5 1.1 57.8 58.9 0.0 0.0
47 Swamp 35 0.16 0.011 5.6 0.3 63.2 63.5 0.0 0.0
48 Peatland 14 0.41 0.010 5.3 0.1 67.4 67.5 0.0 0.0
49 Swamp 38 0.18 0.012 5.5 0.5 78.5 79.0 0.2 0.0
50 Swamp 55 0.13 0.011 4.8 15.2 68.5 83.7 0.0 0.0
51 Peatland 38 0.30 0.010 3.4 0.0 71.5 71.5 0.0 0.0
52 Peatland 39 0.25 0.009 3.0 0.0 86.8 86.8 0.0 0.0
53 Forest 40 0.17 0.014 3.8 0.2 58.4 58.6 2.0 0.0
Plot Ecosystem Species Evenness LCBD Mean Species cover (%)
number richness index C
Obligates Facultative Obligates and facultative Exotic Exotic
wetland wetland wetland invasive

54 Swamp 61 0.20 0.016 4.3 3.4 53.0 56.4 11.3 0.2

55 Marsh 35 0.26 0.024 3.5 46.8 39.2 86.0 3.9 6.8
56 Forest 55 0.20 0.022 3.2 0.2 25.1 25.3 31.4 0.0
57 Marsh 33 0.13 0.024 3.4 18.1 76.7 94.8 2.1 9.3
58 Forest 40 0.19 0.019 3.0 1.0 26.7 27.7 10.6 0.3
61 Forest 48 0.20 0.020 3.7 0.5 49.3 49.8 7.8 0.0
62 Peatland 24 0.45 0.022 6.7 37.4 49.1 86.5 0.0 0.0
63 Peatland 15 0.42 0.022 5.8 30.5 48.6 79.1 0.0 0.0
64 Peatland 47 0.15 0.010 4.3 0.9 79.7 80.5 0.7 0.0
65 Forest 42 0.18 0.020 3.3 0.1 42.4 42.6 7.2 0.1
66 Swamp 61 0.13 0.016 4.6 48.5 49.2 97.8 0.7 1.3
67 Swamp 57 0.17 0.013 5.1 15.5 71.7 87.2 0.6 0.0
68 Peatland 24 0.30 0.014 6.5 0.2 64.8 64.9 0.0 0.0
69 Peatland 31 0.19 0.009 4.8 0.0 62.7 62.7 0.9 0.0
70 Peatland 29 0.17 0.015 5.6 1.0 98.2 99.2 0.2 0.0
71 Peatland 22 0.35 0.016 6.2 0.0 86.3 86.3 0.0 0.0
72 Forest 42 0.14 0.014 4.0 1.4 82.0 83.4 1.8 1.0
73 Swamp 35 0.18 0.013 3.8 0.5 81.1 81.5 4.0 0.8
75 Peatland 17 0.31 0.020 5.2 0.4 95.4 95.7 0.0 0.0
76 Marsh 40 0.18 0.018 4.5 40.7 55.9 96.6 19.1 0.0

18
A. Paquette et al. Urban Forestry & Urban Greening 84 (2023) 127925

Appendix 6 Simpson’s evenness across the 4 types of ecosystems

.
Differences in Simpson’s evenness between ecosystems. Shown here are mean richness (black circles), median (line), 25–75 % quartiles (boxes)
and ranges (whiskers). Different letters indicate a significant difference between ecosystems (p ≤ 0.05; Dunn’s Kruskal-Wallis multiple comparison
tests with Bonferroni correction).

References De’ath, G., Fabricius, K.E., 2000. Classification and regression trees: a powerful yet
simple technique for ecological data analysis. Ecology 81, 3178–3192. https://doi.
org/10.1890/0012-9658(2000)081[3178:CARTAP]2.0.CO;2.
Andreas, B.K., Mack, J.J., McCormac, J.S., 2004. Floristic Quality Assessment Index
Dray, S., Bauman, D., Blanchet, G., Borcard, D., Clappe, S., Guenard, G., Jombart, T.,
(FQAI) for vascular plants and mosses for the State of Ohio. Ohio Environmental
Larocque, G., Legendre, P., Madi, N., Wagner, H.H., 2020. adespatial: Multivariate
Protection Agency, Division of Surface Water, Wetland Ecology Group.
Multiscale Spatial Analysis. R package version 0.3–8. https://CRAN.R-project.org/
Ayotte, G., Rochefort, L., 2019. Les sphaignes de l’Est du Canada: Biologie – Anatomie –
package=adespatial.
Morphologie – Techniques de conservation d’herbier et de préparations
Dubois, R., Proulx, R., Pellerin, S., 2020. Ecological uniqueness of plant communities as a
microscopiques. Editions JFD, Québec.
conservation criterion in lake-edge wetlands. Biol. Conserv. 243, 108491 https://
Ballantyne, M., Gudes, O., Pickering, C.M., 2014. Recreational trails are an important
doi.org/10.1016/j.biocon.2020.108491.
cause of fragmentation in endangered urban forests: a case-study from Australia.
Faber-Langendoen, D., Cameron, D., Gilman, A.V., Metzler, K.J., Ring, R.M., Sneddon, L.,
Landsc. Urban Plan. 130, 112–124. https://doi.org/10.1016/j.
2019. Development of an ecoregional floristic quality assessment method for the
landurbplan.2014.07.004.
Northeastern United States. Northeast. Nat. 26, 593–608. https://doi.org/10.1656/
Bazoge, A., Lachance, D., Villeneuve, C., 2014. Identification et délimitation des milieux
045.026.0312.
humides du Québec méridional. Ministère de l’Environnement et de la Lutte contre
Favreau, M., Pellerin, S., Poulin, M., 2019. Tree encroachment induces biotic
les changements climatiques, Québec.
differentiation in Sphagnum-dominated bogs. Wetlands 39, 841–852. https://doi.
Beauregard, P., Lavoie, M., Pellerin, S., 2020. Recent gray birch (Betula populifolia)
org/10.1007/s13157-018-1122-6.
encroachment in temperate peatlands of eastern North America. Wetlands 40,
Feltynowski, M., Kronenberg, J., Bergier, T., Kabisch, N., Laszkiewicz, E., Strohbach, M.
351–364. https://doi.org/10.1007/s13157-019-01186-3.
W., 2018. Challenges of urban green space management in the face of using
Berendse, F., Van, Breemen, N., Rydin, H., Buttler, A., Heijmans, M., Hoosbeek, M.R.,
inadequate data. Urban For. Urban Green. 31, 56–66. https://doi.org/10.1016/j.
Lee, J.A., Mitchell, E., Saarinen, T., Vasander, H., Wallen, B., 2001. Raised
ufug.2017.12.003.
atmospheric CO2 levels and increased N deposition cause shifts in plant species
Fenton, N.J., Bergeron, Y., 2006. Facilitative succession in a boreal bryophyte
composition and production in Sphagnum bogs. Glob. Change Biol. 7, 591–598.
community driven by changes in available moisture and light. J. Veg. Sci. 17, 65–76.
https://doi.org/10.1046/j.1365-2486.2001.00433.x.
https://doi.org/10.1111/j.1654-1103.2006.tb02424.x.
Blanchard, F., Pellerin, S., Poulin, M., 2018. La tourbière de la base de plein air de Sainte-
Filion, L., Lavoie, M., Querrec, L., 2009. The natural environment of the Québec City
Foy: quatre décennies de changements floristiques. Nat. Can. 142 (3), 22–30.
region during the Holocene. Post-Mediev. Archaeol. 43, 13–29. https://doi.org/
https://doi.org/10.7202/1050995ar.
10.1179/007942309x12457508843847.
Bourdaghs, M., Johnston, C.A., Regal, R.R., 2006. Properties and performance of the
Fleishman, E., Noss, R., Noon, B., 2006. Utility and limitations of species richness metrics
floristic quality index in great lakes costal. Wetl Wetl 26, 718–735. https://doi.org/
for conservation planning. Ecol. Indic. 6, 543–553. https://doi.org/10.1016/j.
10.1672/0277-5212(2006)26[718:PAPOTF]2.0.CO;2.
ecolind.2005.07.005.
Bried, J.T., Jog, S.K., Matthews, J.W., 2013. Floristic quality assessment signals human
Gibbons, J.D., 1985. Nonparametric Methods for Quantitative Analysis. American
disturbance over natural variability in a wetland system. Ecol. Indic. 34, 260–267.
Sciences Press.
https://doi.org/10.1016/j.ecolind.2013.05.012.
Gill, L.W., Ring, P., Higgins, N.M.P., Johnston, P.M., 2014. Accumulation of heavy metals
Brouillet, L., Coursol, F., Meades, S.J., Favreau, M., Anions, M., Bélisle, P., Desmet P.,
in a constructed wetland treating road runoff. Ecol. Eng. 70, 133–139. https://doi.
2019. VASCAN, the Database of Vascular Plants of Canada. http://data.canadensys.
org/10.1016/j.ecoleng.2014.03.056.
net/vascan/.
Environment Canada, 2020. Canadian Climate Normals 1981–2010 Station Data –
Carey, R.O., Hochmuth, G.J., Martinez, C.J., Boyer, T.H., Dukes, M.D., Toor, G.S.,
QUEBEC/JEAN LESAGE INTL A. https://climate.weather.gc.ca/climate_normals.
Cisar, J.L., 2013. Evaluating nutrient impacts in urban watersheds: Challenges and
Havinga, D., 2004. Sustaining Biodiversity: A Strategic Plan for Managing Invasive Plants
research opportunities. Environ. Pollut. 173, 138–149. https://doi.org/10.1016/j.
in Southern Ontario. Ontario Invasive Plants Working Group. http://ufora.ca/wp-
envpol.2012.10.004.
content/uploads/2011/03/sustain.pdf.
Chen, Y., Day, S.D., Wick, A.F., McGuire, K.J., 2014. Influence of urban land
Heijmans, M., van der Knaap, Y.A.M., Holmgren, M., Limpens, J., 2013. Persistent versus
development and subsequent soil rehabilitation on soil aggregates, carbon, and
transient tree encroachment of temperate peat bogs: effects of climate warming and
hydraulic conductivity. Sci. Total Environ. 494, 329–336. https://doi.org/10.1016/
drought events. Glob. Change Biol. 19, 2240–2250. https://doi.org/10.1111/
j.scitotenv.2014.06.099.
gcb.12202.
Cohen, M.J., Carstenn, S., Lane, C.R., 2004. Floristic quality indices for biotic assessment
Heino, J., Grönroos, M., 2017. Exploring species and site contributions to beta diversity
of depressional marsh condition in Florida. Ecol. Appl. 14, 784–794. https://doi.org/
in stream insect assemblages. Oecologia 183, 151–160. https://doi.org/10.1007/
10.1890/02-5378.
s00442-016-3754-7.
Dearborn, D.C., Kark, S., 2010. Motivations for conserving urban biodiversity. Conserv.
Herman, K.D., Masters, L.A., Penskar, M.R., Reznicek, A.A., Wilhem, G.S., Brodovich, W.
Biol. 24, 432–440. https://doi.org/10.1111/j.1523-1739.2009.01328.x.
W., Gardiner, K.P. 2001. Floristic quality assessment with wetland categories and
examples of computer applications for the state of Michigan. Michigan Department

19
A. Paquette et al. Urban Forestry & Urban Greening 84 (2023) 127925

of Natural Resources Wildlife Division Natural Heritage Program. https:// Pellerin, S., Lavoie, M., Talbot, J., 2021. Rapid broadleave encroachment in a temperate
richardchinn.com/pdffiles/FQA_text.pdf. bog induces species richness increase and compositional turnover. Écoscience 28,
Jutras, S., Plamondon, A.P., Hökkä, H., Bégin, J., 2006. Water table changes following 283–300. https://doi.org/10.1080/11956860.2021.1907976.
precommercial thinning on post-harvest drained wetlands. For. Ecol. Manag. 235, Pinceloup, N., Poulin, M., Brice, M.H., Pellerin, S., 2020. Vegetation changes in
252–259. https://doi.org/10.1016/j.foreco.2006.08.335. temperate ombrotrophic peatlands over a 35-year period. PLoS One 15, e0229146.
Kamocki, A.K., Kołos, A., Banaszuk, P., 2017. Can we effectively stop the expansion of https://doi.org/10.1371/journal.pone.0229146.
trees on wetlands? Results of a birch removal experiment. Wetl. Ecol. Manag. 25, R Core Team, 2020. R: A Language and Environment for Statistical Computing. R
359–367. https://doi.org/10.1007/s11273-016-9522-y. Foundation for Statistical Computing, Vienna, Austria (URL). 〈https://www.R-proje
Kelly, W.R., Panno, S.V., Hackley, K.C., 2012. Impacts of road salt runoff on water ct.org/〉.
quality of the Chicago, Illinois, region. Environ. Eng. Geosci. 18, 65–81. https://doi. Robotham, J., Old, G., Rameshwaran, R., Sear, D., Gasca-Tucker, D., Bishop, J., Old, J.,
org/10.2113/gseegeosci.18.1.65. McKnnight, D., 2021. Sediment and nutrient retention in ponds on an agricultural
Kutcher, T.E., Forrester, G.E., 2018. Evaluating how variants of floristic quality stream: evaluating effectiveness for diffuse pollution mitigation. Water 13, 1640.
assessment indicate wetland condition. J. Environ. Manag. 217, 231–239. https:// https://doi.org/10.3390/w13121640.
doi.org/10.1016/j.jenvman.2018.03.093. Rocchio, J. 2007. Floristic quality assessment indices for Colorado plant communities.
Landeiro, V.L., Bini, L.M., Franz, B., Heino, J., Siqueira, T., 2018. Poor and low-lying Prepared for Colorado Department of Natural Resources Division of Wildlife,
sites are more ecologically unique in a hyperdiverse Amazon region: evidence from Wetlands Program. https://mountainscholar.org/bitstream/handle/10217/46964/
multiple taxonomic groups. Divers. Distrib. 24, 966–977. https://doi.org/10.1111/ Floristic_Qual_Assess_Indices_Colo_2007.pdf.
ddi.12734. Sandifer, P.A., Sutton-Grier, A.E., Ward, B.P., 2015. Exploring connections among
Langdon, S.F., Dovciak, M., Leopold, D.J., 2020. Tree encroachment varies by plant nature, biodiversity, ecosystem services, and human health and well-being:
community in a large boreal peatland complex in the boreal-temperate ecotone of Opportunities to enhance health and biodiversity conservation. Ecosystem Serv. 12,
Northeastern USA. Wetlands 40, 2499–2511. https://doi.org/10.1007/s13157-020- 1–15. https://doi.org/10.1016/j.ecoser.2014.12.007.
01319-z. Simpson, E.H., 1949. Measurement of diversity. Nature 163, 688. https://doi.org/
Larouche, A., 1979. Histoire postglaciaire comparée de la végétation à Sainte-Foy et au 10.1038/163688a0.
mont des Éboulements, Québec, par l’analyse macrofossile et l’analyse pollinique. Spyreas, G. 2019. Floristic quality assessment: a critique, a defense and a primer.
M.Sc. Thesis, Université Laval, Québec. Ecosphere 10, e02825. https://doi.org/e02825.10.1002/ecs2.2825.
Lavoie, C., 2019. 50 plantes envahissantes: protéger la nature et l’agriculture. Les Spyreas, G., Meiners, S.J., Matthews, J.W., Molano-Flores, B., 2012. Successional trends
publications du Québec, Québec. in floristic quality. J. Appl. Ecol. 49, 339–348. https://doi.org/10.1111/j.1365-
Lavoie, M., Grondin, P., Caron, C., Larouche, A.C., Poirier, V., 2022. Rapid mixed forest 2664.2011.02100.x.
encroachment and the expansion of red maple (Acer rubrum) in a peatland: a Swink, F., Wilhelm, G., 1979. Plants of the Chicago Region, revised and expanded edition
dendrochronological and palaeoecological approach. Botany. with keys. The Morton Arboretum, Lisle, IL, USA.
Legendre, P., De Caceres, M., 2013. Beta diversity as the variance of community data: Swink, F., Wilhelm, G., 1994. Plants of the Chicago region. Indiana Academy of Science.
dissimilarity coefficients and partitioning. Ecol. Lett. 16, 951–963. https://doi.org/ Indianapolis, Indiana.
10.1111/ele.12141. Tan, L., Fan, C., Zhang, C., Zhao, X., 2019. Understanding and protecting forest
Lepczyk, C.A., Aronson, M.F.J., Evans, K.L., Goddard, M.A., Lerman, S.B., MacIvor, J.S., biodiversity in relation to species and local contributions to beta diversity. Eur. J.
2017. Biodiversity in the city: fundamental questions for understanding the ecology For. Res. 138, 1005–1013. https://doi.org/10.1007/s10342-019-01220-3.
of urban green spaces for biodiversity conservation. BioScience 67, 799–807. Therneau, T.M., Atkinson, B., Ripley, B., Oksanen, J., De’ath, G., 2014. mvpart:
https://doi.org/10.1093/biosci/bix079. Multivariate partitioning. R package version 1.6–2. https://CRAN.R-project.org/
Lin, T., Gibson, V., Cui, S., Yu, C.P., Chen, S., Ye, Z., Zhu, Y.G., 2014. Managing urban package=mvpart.
nutrient biogeochemistry for sustainable urbanization. Environ. Pollut. 192, United Nations, 2018. World urbanization prospects: the 2018 revision, key facts. 〈https
244–250. https://doi.org/10.1016/j.envpol.2014.03.038. ://population.un.org/wup/Publications/Files/WUP2018-KeyFacts.pdf〉.
Lira, P.K., de Souza Leite, M., Metzger, J.P., 2019. Temporal lag in ecological responses U.S. Fish, Wildlife Service, 2019. Floristic Quality Assessment. Midwest Region
to landscape change: where are we now? Curr. Landsc. Ecol. Rep. 4, 70–82. https:// Endangered Species. https://www.fws.gov/midwest/endangered/section7/
doi.org/10.1007/s40823-019-. s7process/plants/FQA.html.
Lu, D., Weng, Q., 2007. A survey of image classification methods and techniques for USDA and NCRS, 2021. The PLANTS Database. National Plant Data Team, Greensboro,
improving classification performance. Int. J. Remote Sens. 28, 823–870. https://doi. NC, USA. http://plants.usda.gov.
org/10.1080/01431160600746456. Vogels, M.F.A., de Jong, S.M., Sterk, G., Addink, E.A., 2017. Agricultural cropland
Matthews, J.W., Spyreas, G., Long, C.M., 2015. A null model test of Floristic Quality mapping using black-and-white aerial photography, object-based image analysis and
Assessment: are plant species’ coefficients of conservatism valid? Ecol. Indic. 52, random forests. Int. J. Appl. Earth Obs. Geoinf. 54, 114–123. https://doi.org/
1–7. https://doi.org/10.1016/j.ecolind.2014.11.017. 10.1016/j.jag.2016.09.003.
Maxwell, S.L., Cazalis, V., Dudley, N., Hoffmann, M., Rodrigues, A.S.L., Stolton, S., Walz, K.S., Kelly, L., Anderson, K., Bowman, K., Andreas, B., Andrus, R., Schuette, S.,
Visconti, P., Woodley, S., Kingston, N., Lewis, E., Maron, M., Strassburg, B.B.N., Schumacher, W., Robinson, S., O’Brien, T., Karlin, E., Hafstad, J., 2018. Universal
Wenger, A., Jonas, H.D., Venter, O., Watson, J.E.M., 2020. Area-based conservation Floristic Quality Assessment Index for Vascular Plants and Mosses of New Jersey:
in the twenty-first century. Nature 586, 217–227. https://doi.org/10.1038/s41586- Coefficient of Conservancy (CoC) Values for Species and Genera (Updated November
020-2773-z. 2019). New Jersey Department of Environmental Protection, New Jersey Forest
Nielsen, A.B., van den Bosch, M., Maruthaveeran, S., van den Bosch, C.K., 2013. Species Service, Office of Natural Lands Management, Trenton, NJ.
richness in urban parks and its drivers: a review of empirical evidence. Urban Wenger, S.J., Roy, A.H., Jackson, C.R., Bernhardt, E.S., Carter, T.L., Filoso, S., Gibson, C.
Ecosyst. 17, 305–327. https://doi.org/10.1007/s11252-013-0316-1. A., Hession, W.C., Kaushal, S.S., Martí, E., Meyer, J.L., Palmer, M.A., Paul, M.J.,
Ogle, D.H., Wheeler, P., Dinno, A., 2020. FSA: Fisheries Stock Analysis. R package Purcell, A.H., Ramírez, A., Rosemond, A.D., Schofield, K.A., Sudduth, E.B., Walsh, C.
version 0.8.31, https://github.com/droglenc/FSA. J., 2009. Twenty-six key research questions in urban stream ecology: an assessment
Ohlson, M., Økland, R.H., Nordbakken, J.F., Dahlberg, B., 2001. Fatal interactions of the state of the science. J. North Am. Benthol. Soc. 28, 1080–1098. https://doi.
between Scots pine and Sphagnum mosses in bog ecosystems. Oikos 94, 425–432. org/10.1899/08-186.1.
https://doi.org/10.1034/j.1600-0706.2001.940305.x. Williams-Mounsey, J., Grayson, R., Crowle, A., Holden, J., 2021. A review of the effects
Oldham, M.J., Bakowsky, W.D., Sutherland, D.A., 1995. Floristic Quality Assessment of vehicular access roads on peatland ecohydrological processes. Earth-Sci. Rev. 214,
System. Ontario. Natural Heritage Information Center Report. Ministry of Natural 103528 https://doi.org/10.1016/j.earscirev.2021.103528.
Resources, Peterborough, Ontario, Canada. Zedler, J.B., Leach, M.K., 1998. Managing urban wetlands for multiple use: research,
Pasquet, S., Pellerin, S., Poulin, M., 2015. Three decades of vegetation changes in restoration, and recreation. Urban Ecosystems 2, 189–204. https://doi.org/10.1023/
peatlands isolated in an agricultural landscape. Appl. Veg. Sci. 18, 220–229. https:// A:1009528505009.
doi.org/10.1111/avsc.12142. Zeileis, A., Hothorn, T., 2002. Diagnostic checking in regression relationships. R. N. 2,
Pellerin, S., Lavoie, C., 2003. Reconstructing the recent dynamics of mires using a 7–10. 〈https://CRAN.R-project.org/doc/Rnews/〉.
multitechnique approach. J. Ecol. 91, 1008–1021. Zielinska, K.M., Misztal, M., Zielinska, A., Zywiec, M., 2013. Influence of ditches on plant
species diversity in the managed forests of Central Poland. Balt. For. 19, 270–279.

20