Content uploaded by Jean-Christophe Clement
Author content
All content in this area was uploaded by Jean-Christophe Clement on Jun 09, 2015
Content may be subject to copyright.
The Cultural Dimensions of Freshwater Wetland Assessments:
Lessons Learned from the Application of US Rapid Assessment
Methods in France
Ste
´phanie Gaucherand
1
•Euge
´nie Schwoertzig
2,3
•Jean-Christophe Clement
3
•
Brad Johnson
4
•Fabien Que
´tier
5
Received: 30 October 2014 / Accepted: 30 March 2015 / Published online: 7 April 2015
Springer Science+Business Media New York 2015
Abstract Given the recent strengthening of wetland
restoration and protection policies in France, there is need
to develop rapid assessment methods that provide a cost-
effective way to assess losses and gains of wetland func-
tions. Such methods have been developed in the US and we
tested six of them on a selection of contrasting wetlands in
the Ise
`re watershed. We found that while the methods
could discriminate sites, they did not always give consis-
tent rankings, thereby revealing the different assumptions
they explicitly or implicitly incorporate. The US assess-
ment methods commonly use notions of ‘‘old-growth’’ or
‘‘pristine’’ to define the benchmark conditions against
which to assess wetlands. Any reference-based assessment
developed in the US would need adaptation to work in the
French context. This could be quite straightforward for the
evaluation of hydrologic variables as scoring appears to be
consistent with the best professional judgment of hydro-
logic condition made by a panel of French local experts.
Approaches to rating vegetation condition and landscape
context, however, would require substantial reworking to
reflect a novel view of reference standard. Reference
standard in the European context must include acknowl-
edgement that many of the best condition and biologically
important wetland types in France are the product of in-
tensive, centuries-long management (mowing, grazing,
etc.). They must also explicitly incorporate the recent trend
in ecological assessment to focus particularly on the wet-
land’s role in landscape-level connectivity. These context-
specific, socio-cultural dimensions must be acknowledged
and adjusted for when adapting or developing wetland
assessment methods in new cultural contexts.
Keywords Wetlands Mitigation France Rapid
assessment methods Best professional judgment
Introduction
Wetlands provide numerous ecosystem services such as
habitat provision for wildlife, mitigation of diffuse pollu-
tants and water quality protection (Clement et al. 2002;
Sabater et al. 2003), water and sediment flows regulation
(Pinay et al. 2002;MEA2005). The conservation and
restoration of wetlands are increasingly justified to sustain
the provision of these services (Maltby and Acreman
2011). Their delivery relies on specific wetland functions,
which in turn depend on the wetland’s ecological condition
and on their surrounding watersheds. Wetland condition,
which is commonly assessed in relation to pristine refer-
ence sites, indicates the range of function or services pro-
vided. Condition assessments are widely adopted in
wetland management (Sutula et al. 2006; Fennessy et al.
2007; Stein et al. 2009b; McLaughlin and Cohen 2013). It
is the case, in particular, for the implementation of envi-
ronmental impact mitigation rules (Robertson 2009). There
is an abundant literature on wetland impact mitigation in
&Ste
´phanie Gaucherand
stephanie.gaucherand@irstea.fr
1
IRSTEA, Unite
´de Recherche sur les Ecosyste
`mes
Montagnards, BP 76, 38402 St-Martin d’He
`res Cedex, France
2
Laboratoire Image, Ville, Environnement, UMR 7362 du
CNRS, Universite
´de Strasbourg, 3 rue de l’Argonne,
67083 Strasbourg, France
3
Laboratoire d’Ecologie Alpine CNRS UMR 5553, Universite
´
Grenoble Alpes, BP 53, 38041 Grenoble Cedex 09, France
4
Department of Biology, Colorado State University,
Fort Collins, CO 80523-1878, USA
5
Biotope, 22 Boulevard Foch, BP 58, 34140 Me
`ze, France
123
Environmental Management (2015) 56:245–259
DOI 10.1007/s00267-015-0487-z
the US, where ‘‘not net loss of wetland function’’ has been
set as a policy goal since 1990s (Hough and Robertson
2009) and numerous assessment methods have been de-
veloped in this context (Fennessy et al. 2007).
No net loss goals have gained interest in Europe in re-
cent years under the impulse of government commitments
under the Convention for Biological Diversity (Bull et al.
2013). The European Union, for example, is preparing a no
net loss Initiative targeting ecosystems and ecosystem
services (Tucker et al. 2014). In France, although the
mitigation hierarchy of avoidance, reduction, and com-
pensation of impacts to the environment has been in force
since 1976, it is only very recently that a goal of achieving
no net loss has been formulated in government guidance
(Ministe
`re de l’Ecologie, du Developpement Durable et de
l’Energie-MEDDE 2012,2013; Que
´tier et al. 2014). Offset
requirements to address residual impacts on wetlands were
laid out in the River Basin Management Plans established
under the European Water Framework Directive (2000/60/
EC) and known in France as Sche´mas Directeur
d’Ame´nagement et de Gestion des Eaux (SDAGE). In
2009, these SDAGEs were reviewed and updated (Que
´tier
et al. 2014) and area-based ratio, where 1.5–2 ha restored
for 1 ha destroyed by development have been adopted in
most river basins. For example, the SDAGE of the Rho
ˆne
river basin requires either the construction or the restora-
tion of twice the destroyed area, in the same catchment
area, or measures that ensure equivalent function and
biodiversity within the catchment (Comite
´de bassin
Rho
ˆne-Me
´diterrane
´e2009). This second approach not only
allows for loss–gain calculations and explicit no net loss
goals, but also calls for assessment methods targeting
wetland functions (Que
´tier and Lavorel 2011).
The legal definition of a wetland in France, established
through Ministerial Order DEVO0813942A in 2008, is
very broad and any habitat with a hydromorphic soil is
considered a wetland, including recently drained agricul-
tural soils. As a result, considerable offsetting requirements
have been imposed on development projects in lowlands
and river valleys. However, no shared methodological
framework for the assessment of losses and gains when
designing and sizing offsets has yet been developed in
France, there is on-going work in this direction (MEDDE
2014). Currently, wetland functions are assessed on a case-
by-case basis, typically through best professional judgment
by trained and experienced professionals hired by devel-
opers. While some guidance is provided on how to pro-
ceed, there are no compulsory methods yet. This situation
is forcing developers and regulators to innovate approaches
on an ad hoc basis (Que
´tier et al. 2014).
One key innovation is the development of wetland
assessment methods that are robust, easily applied, af-
fordable, and which provide sufficient discrimination to
guide management or regulatory decision making. The
recent strengthening of wetland mitigation policies in
France, including a pilot ‘‘banking’’ scheme, calls for im-
proved methods to assess the suitability of proposed
restoration approaches, determines the restoration size re-
quired to meet policy imperatives, and evaluates project
success. Development of robust wetland assessment
methods is a first step toward empowering wetland
mitigation programs. Because of the long-standing work on
rapid wetland assessment in the US, we tested whether US
rapid assessment methods could be ‘‘imported’’ wholesale
into France, or how the methods would have to be modified
to work in the novel context.
Fennessy et al. (2007) analyzed 40 wetland assessment
methodologies developed in the US and concluded that six
of them met the study’s criteria related to sufficiency in
meeting the objectives of the US Clean Water Act pro-
grams, namely, (1) they are rapid taking less than half a day
for two people to carry out an evaluation; (2) they gather
information in the field; and (3) they are repeatable. These
types of methods are called rapid assessment methodolo-
gies (RAMs). RAMs occupy the second level in the three-
tiered wetland assessment hierarchy developed by the US
Environmental Protection Agency (EPA) (EPA 2006), re-
siding between remote landscape assessment (Level 1) and
intensive site-level quantitative studies (Level 3).
We tested six methods (Table 1) on 13 depressional and
riverine wetlands across a gradient of degradation in the
Ise
`re watershed of the Rho
ˆne-Alpes region. All the selected
methods assessed wetland hydrology, biological structure
(mainly through an assessment of the vegetation structure),
and landscape setting, but they differed in their indicators
and how these were combined. Fennessy et al. (2007) dis-
tinguish two types of indicators: first, ‘‘essential indicators’’
with broad, general applicability and second, ‘‘regionally
refined indicators,’’ highlighting the importance of adapting
some indicators to local conditions. The methods also dif-
fered in their choice of reference conditions. The latter can
be ‘‘culturally unaltered,’’ which implies a state that existed
prior to human management activities such as grazing,
agriculture, fire suppression, land development, water re-
source management, and flood control. An alternative ap-
proach to defining reference conditions is termed ‘‘best
attainable conditions.’’ It refers to the highest possible state
that may exist given permanent or semi-permanent con-
straints on the landscape, such as major dams, urban centers,
or flood control facilities (Sutula et al. 2006).
We focused our analysis on three aspects of the RAMs:
(1) their discriminatory power, i.e., the ability to discern
poor versus good condition, (2) the consistency between
RAMs, i.e., the degree of agreement between the different
RAMs on which wetlands were in the best and in the
poorest condition. Where inconsistencies were identified
246 Environmental Management (2015) 56:245–259
123
we endeavored to identify the root cause, and (3) the
consistency with best professional judgment of wetland
condition made by panels of local wetland experts. This
comparison was used to gage the congruence between
evaluation scores and established best-management prac-
tices and local priorities.
Materials and Methods
Selection of Rapid Assessment Methods
Six rapid assessment methods were used (Table 1). We
selected five of the six methods identified by Fennessy
et al. (2007), dismissing the Massachusetts Coastal Zone
Management Rapid Habitat Assessment (Hicks and Car-
lisle 1998) because it was too specific to coastal wetlands.
Instead we added the California Rapid Assessment Method
(CWMG 2012) which has seen wide use and for which
abundant documentation was available (Hanson et al. 2008;
Stein et al. 2009a).
All methods provided an identification key based on the
Hydrogeomorphic (HGM) Classification (Brinson 1993)to
help users determine the type of wetland being assessed. In
all selected methods, the assessment was organized around
the vegetation, the hydrology, and the landscape setting,
except MWAM which does not directly evaluate the
landscape context. The indicators used by the selected
methods described the structure of wetland components
(vegetation, geomorphology, etc.) and/or the disturbance
factors (dams, levees, drains, management, etc.). However,
the methods differed in the choice of indicators because
they were designed in different biogeographic regions,
posing different mitigation issues and in response to a
variety of policies.
A fundamental characteristic that differentiates RAMs is
whether they employ ‘‘relative’’ or ‘‘absolute’’ evaluation
criteria. Most RAMs can be described as being the relative
type; that is, ratings are based on the evaluation of func-
tioning in comparison to a reference standard. In use,
evaluators gage the departure of the observed condition of
a variable from what is expected based on the normal range
Table 1 List of the six rapid assessment methods used for this study
Name and reference Version used Reference condition Scoring Time/expertise
Rapid Assessment Method
Washington (WAFAM)
(Hruby 2012)
2006 Best attainable condition
(25 reference sites)
Vegetation
Hydrology
Landscape
0.5 days
No specific expertise required
Uniform Mitigation Assessment
Method Florida (UMAM 2004)
2004 Culturally unaltered
reference
Overall score
Vegetation
Hydrology
Landscape
1 day
Expertise required
Montana Wetland assessment Method
(MWAM)
(Berglund and McEldowney 2008)
2008 Best attainable condition Overall score
Vegetation
Hydrology
0.5 days
Expertise required
Ohio Rapid Assessment Method
(ORAM) (Mack 2001)
2001 Culturally unaltered
reference
Overall score
Vegetation
Hydrology
Landscape
0.5 days
Expertise required
Delaware Rapid Assessment
Procedure (DERAP)
(Jacobs 2010)
v. 6.0 (2010) Best attainable condition
(calibrated with a
statistical model)
Overall score
Vegetation
Hydrology
Landscape
0.5 days
No specific expertise required
California Rapid Assessment Method
(CRAM)
(CWMG 2012)
v. 5.0.2 (2008) Best attainable condition Overall score
Vegetation
Hydrology
Landscape
0.5 days
No specific expertise required
Most methods assess the vegetation, the hydrology, and the landscape setting of the wetlands, and give an overall score. If one of these four
components is not assessed by a method, the missing component does not appear in the ‘‘scoring’’ column
Environmental Management (2015) 56:245–259 247
123
of variation exhibited by like-kinded reference standard
wetlands. The rates or capacities of functions are merely
implied based on the wetland’s functional type (Johnson
et al. 2013).
Fewer RAMs consider absolute functioning of a discrete
number of functions in terms of capacity or the rate at which
they are performed (e.g., cubic meters of surface water re-
tained). To varying degrees, RAMs that rate variables in
absolute terms step away from the assessment of pure con-
dition and interject elements of societal value into the
evaluation outcome. For example, the MWAM attributes a
higher score to wetlands sheltering rare species regardless of
wetland type or ecological condition ,and ORAM explicitly
awards bonus points to certain wetland types, such as bogs,
fens, old growth forests, etc. This is an important means for
widely agreed upon conservation objectives to be interjected
into the mitigation process. Likewise, WAFAM uses abso-
lute measures of functions such as reducing water velocities
and trapping sediment that describes an implied societal
value (‘‘removal of metals and toxic organic compounds’’).
DERAP is specific in that it focuses on stressors rather than
measurable proxies of the target functions: the presence and
intensity of stressors (habitat, hydrologic, and buffer stres-
sors) lower the final score.
ORAM and UMAM use ‘‘culturally unaltered’’ refer-
ence standards, whereas the other 4 mainly refer to the
‘‘best attainable condition,’’ though this may vary from one
indicator to the other. CRAM and UMAM focus on wild-
life habitat functions. CRAM was designed to assess con-
dition based on the capacity of a wetland to support
characteristic native flora and fauna. This means that hy-
drology and physical structure are assessed based on their
contribution to supporting plant and animal habitat rather
than on the ability of the wetland to provide services such
as flood attenuation or water quality improvement (Stein
et al. 2009b). Also, this method is not reference-based but
scaled to a theoretical optimum condition. Like CRAM,
UMAM focuses on the ability of the wetland to support
wild and aquatic life but in this case the scoring is done
according to best professional judgment (BPJ).
Selection of Study Sites and Field Data Collection
The RAMs are all based on the theory that as wetland eco-
logical condition degrades along a disturbance gradient
there is a corresponding impairment of functioning (Sutula
et al. 2006; Fennessy et al. 2007). The sampling design
aimed to cover gradients in vegetation structure, hydro-
logical connectivity, and landscape context. Study sites were
selected by cross-referencing information provided by the
regional wetland inventory of 2006 (DREAL 2006), pub-
licly available maps (e.g., through the www.geoportail.fr
web portal), field visits, and expert recommendation.
Thirteen sites were selected in the Romanche and Isere flood
plains (Fig. 1). They encompassed the vegetation types (wet
meadows, reed beds, and riverine forests), susceptibility to
regular flooding, and landscape context that is more or less
intensively cultivated or urbanized (Table 2; Fig. 2). Some
sites were managed by local land trusts, and others were
planted forests. Through a panel, local experts were asked to
rank the sites on the basis of their quality and their impor-
tance in relation to local conservation priorities (Table 2).
The six methods were implemented by a single
evaluator during spring 2011 after a trial period and ex-
tensive discussions on the interpretation of guidance
manuals for each method, with involvement of the local
expert panel. Scores for each method and study site are
provided in Appendix.
Data Analysis
All scores were normalized by calculating, for each score
and method, a percentage of the maximum total achievable
score using the method. All analysis was done using Sta-
tistica (StatSoft Inc. 2004).
Discriminatory power is a measure of the ability of the
method to differentiate between wetlands in poor versus
good condition. For each method, we compared the mini-
mum, average, and maximum scores obtained across the 13
wetlands, and distinguished overall score and scores for
vegetation, hydrology, and landscape context. WAFAM
does not attribute an overall score so this method was ex-
cluded from that comparison. A large spread of the scores
across the broad range of site conditions sampled is inter-
preted as being indicative of high discriminatory power.
Consistency between methods was measured by com-
paring scores obtained for each wetland using different
methods. A Principal Component Analysis (PCA) of
assessment scores was used to measure and provide graphic
illustration of the correlations between scores across meth-
ods and wetlands. This analysis was carried out on overall
scores and on the scores obtained for vegetation, hydrology,
and landscape context. Strong correlations across wetlands
were interpreted as indicating consistency between methods.
The contribution of each component (vegetation, hy-
drology, and landscape context) to the overall score was
explored for each RAM by running single and stepwise
regression analyses. Correlations between the different
components of the overall score were also measured.
Finally, we compared the scores obtained using each
method with the ratings given by local experts in order to
gage the relevance of rapid assessment methods in the local
context. The relationship was tested for hydrology, land-
scape, and overall score by regressing the scores given by
each rapid assessment method against the BPJ scores (no
BPJ rating was done on vegetation).
248 Environmental Management (2015) 56:245–259
123
Results
Discriminatory Power
Normalized on a scale from 0 to 100, the overall scores
assigned by different methods range from 24 (MWAM) to
41.6 (CRAM) for the most degraded sites, and 86 (ORAM)
to 98 (DERAP) for the best sites. The mean score (all sites
together) is 66 ±18 SD. Figure 3shows a spread of scores
across the study sites for each method, which obviously
discriminate different sites. Drained meadows received the
lower scores with all methods and this result was confirmed
in the PCA analysis on overall scores (Fig. 4).
Considering vegetation, hydrology, and landscape con-
text separately, the discriminatory power of different
methods varied. This was most evident in the case of
CRAM’s scoring of the landscape context. This method
had little ability to discriminate between the variety of
settings sampled, and scores were high compared to all
other methods clustering within a narrow range between 66
and 93.
Consistency Between RAMs
The RAMs were projected on the first two axis of the four
PCAs (Figs. 4b, 5,6,7b). The arrows indicate the direction
of the highest scores given by the RAMs in the factor-
plane. Axis 1 of the PCA on the vegetation, hydrology, or
overall scores explains over 70 % of the variance (Figs. 4,
5,6), whereas axis 1 of the PCA on the landscape scores
only explains 47 % of the variance (Fig. 7). When
projected on axis 1, all RAMs were located on the same
side of the axis (left side, Figs. 4b, 5,6,7b). Axis 1 of all
PCAs thus discriminates sites that received, overall, higher
scores with all RAMs (left side of axis 1) against sites that
received, overall, lower scores with all RAMs (right side of
axis 1). This shows that the information provided by the
RAMs on hydrology, vegetation, and the overall condition
of the wetlands was very redundant (correlated). This is
less clear for landscape assessment. Axis 2 of the PCAs
explains a much lower part of the variance (around 15 %,
for all PCAs except for the one on landscape scores, where
axis 2 explains up to 30 % of the variance). This axis
discriminates the least correlated RAMs, i.e., methods that
gave the highest scores to different wetlands.
Indeed, all methods agreed on the sites to be considered
‘‘in poor condition’’ (Fig. 4a, sites 2, 13 and 6). The best
sites, however, appeared to be different from one method to
another. Meadows (sites 1; 2; 5; 6; 13) were generally
discriminated against, receiving all the lowest scores (right
side of the first PCA axis) and forests (sites 3; 4; 7; 10; 11;
12) mostly occupied the left side of the PCA (average to
high scores). The second axis of the PCA separates sites
that received high scores with different methods: only
forested wetlands received the highest scores with CRAM,
whereas MWAM gave high scores to a variety of wetland
types, including a meadow (site 5). The PCA on vegetation
scores confirmed the trend of all methods to discriminate
against meadows (Fig. 5). It also showed that two methods,
CRAM and WAFAM, clearly favored forested wetlands.
The overall scores obtained using different methods
were significantly correlated (Table 3), except between
Site
code Long. Lat.
1 879562 2038628
2 879670 2038647
3 887221 2057248
4 886897 2057258
5 848014 2036430
6 847301 2035999
7 831115 2015860
8 831157 2015907
9 830794 2012742
10 830543 2010879
11 892847 2010714
12 893283 2011349
13 889733 2017772
Fig. 1 Code and location of the 13 study sites (latitude and longitude are given in the French geodetic system RGF93, Lambert 93 projection)
Environmental Management (2015) 56:245–259 249
123
CRAM and MWAM (which was consistent with the result
of the PCA on overall scores, Fig. 4). Similar results were
obtained for the vegetation scores: only CRAM and
MWAM were inconsistent. For Hydrology, only DERAP
gave significantly different scores; all other methods are
well correlated. The PCA on hydrology scores reiterated
this result (Fig. 6).
In contrast, scores for Landscape Context ranged widely
among RAMs: only the scores given by CRAM and
ORAM were well correlated (r
2
=0.67, P\0.001). The
scores from UMAM and WAFAM were also correlated,
but the relationship was weaker (r
2
=0.48, P=0.09). The
scores obtained from DERAP were not consistent with any
other methods, and MWAM did not give a score for
landscape context. This is consistent with the result of the
PCA on landscape scores (Fig. 6). While the methods do
not agree on which sites should receive the best or the
worst scores, the least degraded landscapes (wetlands sur-
rounded by forests and meadows) are grouped on the left
side of the first axis (sites 7; 11; 12; 13).
Contributions of Vegetation, Hydrology,
and Landscape Context to Overall Scores
Simple regressions between overall scores and scores for
vegetation, hydrology, or landscape context showed that
for all methods, vegetation and hydrology explained most
of the variance of the overall score (Table 4; the results for
WAFAM were not presented because this method had no
overall score). The exception was DERAP, where hy-
drology had no significant relationship with the overall
score. Only DERAP and UMAM show a significant rela-
tionship between landscape scores and the overall scores.
Results from the stepwise regressions showed that, for all
methods except UMAM, vegetation alone explained 87 % or
more of the variance of the overall score (Table 4). In the
case of UMAM, hydrology explained 92 % of the variance of
the overall score. As a consequence, adding another com-
ponent of the overall score explained only slightly better its
variance. Indeed, scores given to hydrology and vegetation
are correlated in all methods except DERAP (Table 5).
Table 2 Site description
Site
code
Name Type AA
(ha)
Vegetation type (Corine biotope Code) Maintenance
activities
BPJ
rank
1 Montfort Marsh DEP 6.5 Eutrophic wet meadow with Juncus
subnodulosus colonized by Phragmites and
Solidago (37.2 354.2)
Mowing (once
every 2 years)
VII
2 Montfort Meadow DEP 0.7 Wet meadow with Filipendula ulmaria, invaded
by Solidago (37.1)
Grazing VI
3 Chapareillan Riparian Forest—Ise
`re
side
RIV 2.8 Willow stands, dominated by Salix alba (44.13)– IV
4 Chapareillan Riparian Forest—
Chartreuse side
DEP 1.7 Willow & Poplar stands with Salix Alba,
Populus alba,P. nigra and P. tremula (44.1)
–IV
5 Moı
¨les Wet Meadow DEP 6.2 Eutrophic wet meadow with Juncus articulatus
and patrimonial species such as Orchis
palustris and Senecio paludosus (37.21)
Mowing (once/
year)
V
6 Moı
¨les Meadow DEP 2.2 Mesophile pastures with poaceae and semi-
ruderal species (Plantago major, Lampsane
communis…)(38.11)
Grazing III
7 Loyes Riparian Forest RIV 0.4 Willow stands (44.1)–VI
8 Loyes Reed bed RIV 0.8 Reed bed (53.1)–V
9 Creux Reed bed DEP 2.9 Reed bed (53.1)–V
10 Co
ˆte Chaude Riparian Forest RIV 2.3 Ash and alder stands (44.3) – III
11 Buclet Riparian Forest—Left bank RIV 1.3 Willow stands (44.1) – VII
12 Buclet Riparian Forest—Right bank DEP 13 Ash and alder stands (44.3)–I
13 Bourg d’Oisans Meadow DEP 0.2 Oligotrophic wet meadow with Molinia caerulea
(37.31)
Mowing or
grazing
II
The type of wetland is either depressional (DEP) or riverine (RIV). To describe the vegetation we used the Corine biotope code: a European
typology for natural habitats (European Communities 1991). AA is for ‘‘Assessment Area.’’ The best professional judgment rank (BPJ) was
based on a general classification provided by a panel of local experts, from I (lowest quality) to VII (best quality). Different sites can have the
same rank
250 Environmental Management (2015) 56:245–259
123
Axis 1 – Vegetaon
Axis 2 – Hydrology
Axis 3 – Landscape context
Reed Beds
8; 9
Wet Meadows
1; 2; 5; 6; 13
Forested Wetlands
3; 4; 7; 10; 11; 12
Water supply remains
mostly natural
(no drains or development
on the wetland )
3; 7; 8; 9; 10; 11
Water supply
moderately affected
by small development
(drains, water diversion,
water retention…) but the
wetland is still supplied by
surface or ground water
1; 5; 12; 13
Water supply strongly
affected by development
(deep drains, water diversion,
reservoirs…). Hydrological
connectivity with the local
surface or ground water flow
system drastically reduced
2; 4; 6
Mostly “natural”
(meadows, forests… )
7; 11; 12; 13
Moderately degraded
(fallow lands, tree
plantations, parks,
residences…)
3; 4; 5; 6; 8
Heavily degraded
(Industries, urban areas,
intensive agriculture…)
1; 2; 9; 10
Fig. 2 The study sites have
been selected among three types
of habitats and along two
gradients of degradation:
Landscape and Hydrology. The
numbers are the sample site
identification codes (cf. Fig. 1)
Fig. 3 Mean score attributed by
different RAMs to vegetation,
hydrology, landscape, and
overall score. The square
represents the mean score;
whiskers represent the minimum
and maximum scores
Environmental Management (2015) 56:245–259 251
123
Consistency Between Scores and Local Ranking
of Wetlands
A panel of local experts was consulted to position 13
wetlands on a gradient of hydrological and landscape
conditions (Fig. 1), and to rank them from ‘‘best quality’’
to ‘‘poorest condition’’ (Table 2). The PCA on hydrology
scores showed that the gradient identified by local experts
was consistent with the gradient detected using the tested
methods (Fig. 6). Indeed, all the sites considered in good
hydrological condition by local experts also have higher
RAM scores (left side of the first axis of the PCA), and all
Fig. 4 PCA on overall scores
obtained for 13 sites using five
different rapid assessment
methods (RAMs). aProjection
of the wetlands on the first two
axis of the PCA, dots are
forested wetlands, diamonds are
wet meadows, and triangles are
reed beds. bProjection of the
RAMs on first two axis of the
PCA
Fig. 5 PCA on vegetation
scores obtained for 13 sites
using 6 different rapid
assessment methods (RAMs).
aProjection of the wetlands on
the first two axis of the PCA,
dots are forested wetlands,
diamonds are wet meadows,
triangles are reed beds.
bProjection of the RAMs on
first two axis of the PCA
Fig. 6 PCA on hydrology scores obtained for 13 sites using 6
different rapid assessment methods (RAMs). aProjection of the
wetlands on the first two axis of the PCA. The marker’s shape
indicates the BPJ classification of the site regarding hydrology: stars
are wetlands with mostly natural water supply, squares are wetlands
where the water supply was moderately affected by small develop-
ment, and diamonds are wetlands where the water supply was heavily
affected by development. bProjection of the RAMs on first two axis
of the PCA
252 Environmental Management (2015) 56:245–259
123
Fig. 7 PCA on landscape scores obtained for 13 sites using 5
different rapid assessment methods (RAMs). aProjection of the
wetlands on the first two axis of the PCA. The marker’s shape
indicates the BPJ classification of the site regarding the landscape
context: stars for wetlands surrounded by mostly ‘‘natural’’
ecosystems (forests, meadows), squares are wetland surrounded by
moderately disturbed landscape (fallow lands, tree plantation,
residences…), and diamonds are wetland surrounded by urban areas,
industries, or intensive agriculture. bProjection of the RAMs on first
two axis of the PCA
Table 3 Correlation matrix of the scores obtained using different methods
CRAM ORAM UMAM DERAP MWAM
OSHLVOSHLVOSHLVOSHLVOSHLV
ORAM * ** * *
UMAM * ** ns * * * ns *
DERAP * ns ns * * ns ns * * ns ns **
MWAM ns ** – ns * *–** **–***ns–**
WAFAM – *ns ** – *– * – ** * * – ns ns * – * – *
Correlation was measured on Overall score (OS), Hydrology (H), Landscape (L), and Vegetation (V). * when P\0.05, ** when P\0.01, ns
when correlation is not significant. MWAM does not measure Landscape. WAFAM gives no Overall score
Table 4 Results of the forward stepwise regression analysis done to analyze the effect of the different components evaluated (vegetation,
hydrology, and landscape) on the overall score of the wetlands for each RAM
Predictive variable Step Multiple R
stepwise
Multiple Rsquare stepwise
(simple regression)
Rsquare
change
FP
CRAM Veg 1 0.87 0.76 (0.76**) 0.76 34.25 0.0001
CRAM Hydro 2 0.94 0.89 (0.64**) 0.13 11.98 0.0061
CRAM Landscape 3 0.97 0.94 (0.17
ns
) 0.05 7.99 0.0198
ORAM Veg 1 0.86 0.74 (0.74**) 0.74 31.68 0.0002
ORAM Landscape 2 0.96 0.93 (0.14
ns
) 0.19 26.23 0.0004
ORAM Hydro 3 0.98 0.97 (0.64**) 0.04 11.25 0.0085
UMAM Hydro 1 0.96 0.92 (0.92**) 0.92 126.02 0.0000
UMAM Landscape 2 0.98 0.96 (0.56**) 0.04 10.58 0.0087
UMAM Veg 3 1.00 1.00 (0.63**) 0.04 2910 0.0000
DERAP Veg 1 0.87 0.76 (0.76**) 0.76 34.57 0.0001
DERAP Hydro 2 0.96 0.91 (0.11
ns
) 0.16 18.26 0.0016
DERAP Landscape 3 0.98 0.96 (0.38*) 0.04 9.26 0.0139
MWAM Veg 1 0.96 0.93 (0.93**) 0.93 138.70 0.0000
MWAM Hydro 2 0.98 0.95 (0.47*) 0.03 5.87 0.0359
The results of simple regressions (r
2
) between the components and the overall score are also indicated between brackets, * Level of Pvalues
(* P\0.05, ** P\0.01)
Environmental Management (2015) 56:245–259 253
123
the sites considered in poor hydrological condition by local
expert had lower RAM scores. The regression showed that
BPJ rating and RAM Hydrology scores were correlated for
all methods except DERAP (Table 6).
For the Landscape Context, only one method, WAFAM,
showed a significant correlation with BPJ ratings (Table 6)
and that relationship was not strong (r
2
=0.25). The only
congruence between RAM and BPJ Landscape Context
ratings was that they tended to agree on the least disturbed
landscapes (sites 7; 11; 12; 13; Fig. 1).
Finally, and importantly, there was no significant corre-
lation between the overall scores provided by RAMs and the
overall condition ratings provided by local experts (Table 6).
This lack of correlation is because expert opinion favored
meadow sites over forested ones. Indeed, focusing on the 3
lowest overall scores, only ORAM and UMAM included a
forest among them (site 4), whereas 2 forests (site s 10 and 12)
and 1 meadow (site 13) are considered as low quality wet-
lands by local experts. At the other end of the gradient, two
meadows (sites 1; 2) and one forest (site 11) were considered
as the highest quality by local experts, when only MWAM
included a meadow in the best scores (site 5). One site in
particular (site 2) was classified as the most degraded by all
rapid assessment methods yet received one of the highest
ranks from local experts as 2nd best wetland. These results
reveal a real challenge in ‘‘importing’’ these US methods to
the French context, which we discuss below.
Discussion
The methods tested in this study were developed in specific
geographical areas, in response to local socio-economic,
ecological, and regulatory circumstances in the US. They
were not developed for the type of wetlands sampled in this
study. Moreover, although the reproducibility of the scor-
ing was assessed through repeated multi-assessor trials
(unpublished), the methods were tested by a single
evaluator who was not formally trained by an agreed in-
stitution in the use of these methods (we used the manual
provided by each method). Despite these limits, applying
six US rapid assessment methods in a novel situation, in
France, provided a means of exposing the underlying as-
sumptions and workings of each. It has generated several
results which we discuss below.
Selected Rapid Assessment Methods were Consistent
in Their Overall Ranking
The different RAMs enabled study sites to be ranked.
Meadows disconnected from the natural water flow system
had the lowest scores with all methods. Highest scores,
however, depended on the method used: CRAM clearly
favored forested habitats (the highest overall scores were
attributed to 3 forested wetlands), while MWAM gave high
overall scores to very different vegetation types. In fact,
MWAM was the only method that gave a high overall
score to a meadow (site 5). A closer look at the way
vegetation was scored by these different methods provided
insights on these differences. In both CRAM and MWAM,
the number of plant layers was one of the indicators used to
score vegetation. However, in CRAM (for riverine and for
depressional wetlands), a higher number of plant layers led
automatically to a higher score, whereas in MWAM the
number of plant layers was considered relative to the po-
tential of the wetland. Nevertheless, wetlands where the
number of plant layers was artificially maintained low
through management practices, such as tree cutting or
mowing to prevent shrub encroachment or spontaneous
afforestation, received a lower score from both methods.
Table 5 Correlation between scores of the RAM components (veg:
vegetation structure, hydro: hydrology and landscape)
Method Scores significantly correlated rr
2
P
CRAM Veg–hydro 0.58 0.33 0.038
ORAM Veg–hydro 0.63 0.39 0.021
UMAM Veg–hydro
Hydro–landscape
0.77
0.61
0.60
0.37
0.002
0.026
DERAP Veg–landscape 0.69 0.48 0.009
MWAM Veg–hydro 0.57 0.32 0.043
WAFAM Veg–hydro
Veg–landscape
0.65
0.70
0.42
0.49
0.016
0.007
Only significant correlations (P\0.05) are shown
Table 6 Correlation between
scores given by RAMs and BPJ
rating
Method Hydrology Landscape Overall score
rr
2
Prr
2
PRr
2
P
CRAM 0.72 0.52 0.0057 0.19 0.11 0.1412 0.01 -0.08 0.7434
ORAM 0.77 0.60 0.0020 0.13 0.05 0.2311 0.08 -0.00 0.3456
UMAM 0.87 0.75 0.0001 0.15 0.08 0.1873 0.06 -0.02 0.4065
DERAP 0.02 0.00 0.9576 0.03 -0.06 0.5600 0.04 -0.05 0.5022
MWAM 0.69 0.47 0.0095 – – – 0.05 -0.03 0.4524
WAFAM 0.81 0.66 0.0007 0.31 0.25 0.0490 –– –
Bold is used for significant correlations (P\0.05)
254 Environmental Management (2015) 56:245–259
123
Hydrology scores were very consistent across RAMs,
except DERAP, which was based on the identification of
stressors and how these impaired the functioning of the
wetland. The same stressor could have little or large impact
depending on the type of wetland. The lack of correlation
between DERAP scores and other RAMs for hydrology was
mainly due to the attribution of a low score to sites 9 and 3
using DERAP, while other methods considered these sites to
be in average or good condition. Site 9 was difficult to
classify as depressional or riverine due to limited available
information. Because of the way the score was calculated, a
misclassification could have an important impact on the final
score. If site 9 had been classified riverine instead of de-
pressional, its score with DERAP would have been sub-
stantially higher and more consistent with other methods.
Misclassification problems are a key issue in developing and
using rapid assessment methods and our application to
wetlands in the Ise
`re watershed clearly illustrated its poten-
tial effects on site condition scores.
The scores obtained for landscape context were quite
inconsistent across methods, with significant correlations
only existing between CRAM and ORAM and UMAM and
WAFAM scores. The disparities in scores could be ex-
plained by the way landscape context was considered in the
methods. Assessment of wetland buffer condition is im-
portant to all methods and, in fact, DERAP, CRAM, and
ORAM limit landscape assessment to consideration of the
surrounding buffer area. On the other hand, WAFAM and
UMAM also consider connections and corridors connect-
ing the assessed wetland to other similar habitats. Another
distinction among methods is that CRAM and ORAM
evaluate the ability of the buffer to protect the wetland
from human activity, whereas DERAP lists the stressors
(road, development, etc.) that could affect the buffer as
well as the wetland. Although the width of buffer consid-
ered differs between CRAM (250 m) and ORAM (50 m),
the ratings were generally in concurrence. DERAP, with a
100-m wide buffer, gave very different results from the
other methods, probably because of different indicators
(‘‘stressors’’) used to describe the buffer. Here again, the
choice of indicators was a key determinant of how wet-
lands were assessed, ranked, and therefore valued in de-
velopment and mitigation decisions.
Combining Vegetation, Hydrology, and Landscape
Context into an Integrated Assessment
Most of the tested RAMs generated consistent scores for
vegetation and hydrology, but not for the landscape con-
text. Regressions between the overall scores and the scores
obtained for vegetation, hydrology, and the landscape
context showed that vegetation and hydrology explained
most of the variance in overall scores. The role that
landscape factors actually played in these methods was
called into question, and this would certainly have to be
considered in detail since land-use planning regulations
increasingly consider landscape-level connectivity in
France (under French law 2009-967 of 03 August 2009).
Stepwise regression results showed that scores obtained for
either vegetation or hydrology were enough to explain
most of the overall score variance. Because changes in
hydrology lead to changes in vegetation, a correlation be-
tween these scores is inevitable (they were especially
strong with UMAM). However, methods should be con-
structed in a way that each component of the overall score
brings in as much unrelated information, to be as com-
prehensive as possible and to limit redundancies.
Another key finding from this study is that most of the
evaluated methods favor sites with vertical vegetation
structure, particularly forest vegetation. Moreover, all
methods discriminated against wetlands where the number
of plant layers was artificially maintained low through
management practices. This finding reveals fundamental
differences between Europe and the US in terms of both
landscape ecology and socio-cultural views of wetlands. In
the US, RAM scores are usually somehow tied to a notion
of ‘‘natural’’ functioning, with most references alluding to
passive maintenance of ecosystem processes in a way that
would mirror pre-European settlement conditions. The
methodological preference for multiple canopy layers and
vertical structure is a reflection of the foundation of US
references. Use of a naturalistic-type reference is prob-
lematic in France, however, because many highly desirable
and valuable wetland habitat types are sustained solely
through active maintenance. This is particularly the case
with meadows that are regularly mowed to avoid invasion
by trees, for fodder or, in a conservation context, to favor
animal and plant species associated with herbaceous
vegetation. If assessment methods assign higher scores to
‘‘pristine’’ or ‘‘old growth’’ vegetation types, then most
maintained meadows are automatically downgraded.
The case of site 2 (Montfort Meadow) illustrates the
conundrum of using naturalistic references in the assess-
ment of wetlands in France. At this site, invasive plants
dominated the vegetation, it has been drained and separated
from the river by levees, and it is set in an unfavorable
landscape context. For all these reasons, the site was
ranked as degraded by all the RAMs tested. Local experts,
however, classified it as the second best site. They valued
this wetland for its cultural heritage value: it was a rare
reminder of past land-use systems from the ninetieth and
early twentieth century called ‘‘chantournes.’’ Moreover, in
spite of invasive species, the site was included in a larger
group of wetlands that sheltered protected butterfly species
(Coenonympha oedippus,Maculinea teleius and Lycaena
dispar; Seigne-Martin et al. 2007). The conjunction of a
Environmental Management (2015) 56:245–259 255
123
protected species and cultural value ‘‘boosted’’ the score of
the site, locally, despite it being dysfunctional. Thus, we
conclude that wetland assessments developed for use in
France, and likely other European settings, need to incor-
porate elements societal value in the site scores in addition
to evaluation of ecological integrity or natural functioning.
Applicability of Rapid Assessment Methods
to French Wetland Management
The case of the Montfort Meadow (site 2) raises the question
of how to consider managed semi-natural ecosystems in
assessing wetland condition. The answer to this question is
crucial in intensively used landscapes where few wetlands
are unmanaged, as this is the case in France where historical
practices are often maintained on ‘‘protected’’ ecosystems in
order to keep historical landscapes and habitats that would
otherwise change (Prins 1998; Fischer and Wipf 2002;
Muller 2002; Middleton et al. 2006). Using ‘‘pristine’’ or
‘‘old growth’’ reference conditions is consistent with fre-
quent references to wildness in North America that find less
echo in Europe or France (Arnould and Glon 2006; Carrie
`re
and Bidaud 2012; Nash 2014). Given the longer land-use
history in France, and the different objectives with regard to
which functions and which land covers are considered of
higher value, the development of wetland assessment
methods in France (or Europe) should include careful
statements of intent and objectives. It appears that the vari-
ables in the set of assessment methods evaluated could be
reconfigured to design an appropriate set of metrics that
score the field observations to reflect preferences. Not only
scientists and managers but also citizens should be consulted
to fully comprehend which ecological functions and
ecosystem services are valued by the collective populace.
Reference conditions against which to assess wetlands are
key elements in the design of assessment methods. For the
reasons discussed previously, the ‘‘best attainable conditions
approach’’ should be preferred to make up for the lack of
pristine wetlands (Smith et al. 1995; Brooks et al. 2013).
Another is the reference against which losses and gains are
assessed to design and size mitigation and offsets and the set
of indicators or proxies used. The capacity of wetland offsets
to achieve no net loss has been questioned repeatedly (e.g.,
Matthews and Endress 2008; Hossler et al. 2011;
McLaughlin and Cohen 2013; Strain et al. 2014; Yepsen
et al. 2014). Our capacity to restore their complex structure
and function has also been questioned (Mitsch and Wilson
1996; Moreno-Mateos et al. 2012). All of these topics should
be addressed in the development of a rapid assessment
method for wetland functions in France (Que
´tier and Lavorel
2011; Lavoux et al. 2013; MEDDE. 2014).
Interestingly, none of the rapid assessment methods used
in this study focused on the importance of the wetland for
the surrounding landscape’s value (rather than the oppo-
site), except in the more recent (2012) version of WAFAM.
Yet, when a wetland is filled or damaged because of a new
development, the consequence of this loss for surrounding
similar habitat, and even the possible loss of connectivity
between natural habitats at larger spatial scales, should be
taken into account (Bruggeman et al. 2005, Bruggeman
et al. 2009; Hartig and Drechsler 2009). Also, the size of
the buffer and the indicators used varied a lot among
methods, leading to different assessments of a same wet-
land landscape. A possible explanation for this situation
could be historic and cultural. Historic because land-use
policy and regulations in the US have their origin in local
(state or municipal) government. Local regulation (and
assessment) of wetlands can still be more stringent than
state and national level regulations. Cultural because there
is a strong support in the US for protection of the rights of
individual landowners, preventing precise delineation of a
legal boundary of an individual wetland. These issues
probably influenced the details of many US wetland
assessment methodologies. This should be kept in mind
when developing landscape assessment indicators for a
French (or European) method.
In conclusion, this study has shown that American rapid
assessment methods cannot be directly transferred to the
French context. Some adaptation will be necessary. This
would be quite straightforward for assessments of hydrology
as these specific indicators are widely shared and scoring
appeared to be consistent with current best professional
judgment (revealed here by a panel of local experts). Vege-
tation and landscape context, however, would require sub-
stantial reworking to reflect conservation priorities tagged to
historical reference conditions that involve intensive man-
agement (mowing, grazing, etc.) and a recent focus away
from site-based approaches to landscape-level connectivity.
Acknowledgments The authors are very grateful to the local ex-
perts consulted as part of this study: Jean-Christophe Cle
´ment,
Christian Gay, Jacky Girel, Olivier Manneville, and Roger Marciau.
We also wish to thank the reviewers for their very helpful comments,
some of them now included in the text. This study was funded by the
Conseil Ge
´ne
´ral de l’Ise
`re through the Po
ˆle de
´partemental de
Recherches sur la biodiversite
´en Ise
`re. Subsequent analysis was
partly funded by the European Union’s Seventh Framework Pro-
gramme (FP7/2007-2013) under grant agreement 308393
‘‘OPERAs.’’ We also acknowledge funding by the Mission Biodi-
versite
´of the Caisse des De
´pots et Consignations, the Cluster
Recherche Rho
ˆne Alpes, and the PIR IngECOTech ‘‘Inge
´nierie et
equivalence.’’ LECA and IRSTEA Grenoble are part of Labex
OSUG@2020 (ANR10 LABX56).
Appendix
See Tables 7and 8.
256 Environmental Management (2015) 56:245–259
123
Table 7 Overall scores
Site code Site Category CRAM ORAM UMAM DERAP MWAM
1 Montfort Marsh DEP 55.8 66.0 63.3 64.6 75.0
2 Montfort Meadow DEP 41.6 32.0 29.3 29.3 24.0
3 Chapareillan Riparian Forest—Ise
`re side RIV 80.2 74.5 80.0 58.2 71.2
4 Chapareillan Riparian Forest—Chartreuse side DEP 63.8 41.0 58.8 90.2 78.3
5 Moı
¨les Wet Meadow DEP 56.8 69.0 73.3 68.6 88.3
6 Moı
¨les Meadow DEP 51.7 31.0 55.0 39.0 46.0
7 Loyes Riparian Forest RIV 70.3 68.0 86.3 70.3 85.0
8 Loyes Reed bed RIV/Flat 59.4 67.0 91.3 81.7 88.6
9 Creux Reed bed DEP 54.5 56.0 82.9 79.1 76.2
10 Co
ˆte Chaude Riparian Forest RIV 90.5 69.5 78.3 79.1 60.0
11 Buclet Riparian Forest—Left bank RIV 89.7 86.5 86.9 98.4 83.8
12 Buclet Riparian Forest—Right bank DEP 59.5 61.0 54.7 69.9 74.0
13 Bourg d’Oisans Meadow DEP 48.3 52.0 59.0 47.6 53.0
Table 8 Hydrology, Landscape, and vegetation scores
Site code Site CRAM ORAM UMAM DERAP MWAM WAFAM
Vegetation
1 Montfort Marsh 61.1 60.0 100.0 87.9 96.7 34.6
2 Montfort Meadow 30.5 0.0 30.0 29.3 23.3 15.4
3 Chapareillan Riparian Forest—Ise
`re side 80.5 65.0 90.0 90.9 90.0 65.4
4 Chapareillan Riparian Forest—Chartreuse side 86.1 40.0 86.0 100.0 93.3 61.5
5 Moı
¨les Wet Meadow 59.7 85.0 85.0 76.3 96.7 53.8
6 Moı
¨les Meadow 44.4 15.0 70.0 46.6 63.3 30.8
7 Loyes Riparian Forest 91.6 90.0 91.0 90.9 96.7 76.9
8 Loyes Reed bed 44.5 85.0 95.0 87.7 100.0 42.3
9 Creux Reed bed 41.6 45.0 81.0 94.0 80.0 38.5
10 Co
ˆte Chaude Riparian Forest 91.6 70.0 75.0 70.9 66.7 69.2
11 Buclet Riparian Forest—Left bank 91.6 85.0 92.0 100.0 100.0 80.8
12 Buclet Riparian Forest—Right bank 66.7 40.0 68.0 82.8 83.3 50.0
13 Bourg d’Oisans Meadow 33.3 15.0 57.0 58.6 66.7 30.8
Hydro
1 Montfort Marsh 52.8 46.7 70.0 76.7 10.0 40.0
2 Montfort Meadow 33.3 20.0 24.0 83.3 0.0 20.0
3 Chapareillan Riparian Forest—Ise
`re side 75.0 75.0 92.0 57.5 40.0 69.6
4 Chapareillan Riparian Forest—Chartreuse side 41.6 26.7 53.0 86.7 40.0 35.6
5 Moı
¨les Wet Meadow 84.5 56.7 90.0 93.3 70.0 51.1
6 Moı
¨les Meadow 41.6 16.7 50.0 76.7 0.0 17.8
7 Loyes Riparian Forest 66.7 40.0 85.0 71.2 50.0 76.1
8 Loyes Reed bed 75.0 50.0 96.0 100.0 80.0 73.3
9 Creux Reed bed 58.3 53.3 90.0 56.7 70.0 65.2
10 Co
ˆte Chaude Riparian Forest 100.0 95.0 92.0 100.0 60.0 56.5
11 Buclet Riparian Forest—Left bank 100.0 81.7 92.0 100.0 95.0 60.9
12 Buclet Riparian Forest—Right bank 33.3 26.7 48.0 66.7 0.0 30.4
13 Bourg d’Oisans Meadow 33.3 46.7 52.0 76.7 0.0 4.4
Environmental Management (2015) 56:245–259 257
123
References
Arnould P, Glon E (2006) Wilderness, usages et perceptions de la
nature en Ame
´rique du Nord. Annales de ge
´ographie 115:
227–238
Berglund J, McEldowney R (2008) MDT Montana wetland assess-
ment method. Prepared for: Montana Department of Transporta-
tion Post, Buckley, Schuh & Jernigan, Montana
Brinson MM (1993) A hydrogeomorphic classification for wetlands.
Wetlands Research Program TR-WRP-DE-4. US Army Water-
ways Exp. Station, Vicksburg, MS
Brooks RP, Brinson MM, Wardrop DH, Bishop JA (2013) Hydro-
geomorphic (HGM) classification, inventory, and reference
wetlands. In: Brooks RP, Wardrop DH (eds) Mid-atlantic
freshwater wetlands: advances in wetlands science, management,
policy, and practice. Springer, New York, pp 39–59
Bruggeman D, Jones M, Lupi F, Scribner K (2005) Landscape
equivalency analysis: methodology for estimating spatially
explicit biodiversity credits. Environ Manage 36:518–534
Bruggeman D, Jones M, Scribner K, Lupi F (2009) Relating tradable
credits for biodiversity to sustainability criteria in a dynamic
landscape. Landsc Ecol 24:775–790
Bull JW, Suttle KB, Gordon A, Singh NJ, Milner-Gulland E (2013)
Biodiversity offsets in theory and practice. Oryx 47:369–380
Carrie
`re SM, Bidaud C (2012) En que
ˆte de naturalite
´: Repre
´sentations
scientifiques de la nature et conservation de la biodiversite
´. In:
Ramiarantsoa HR, Blanc-Pamard C, Pinton F (eds) Ge
´opolitique
et environnement: les lec¸ ons de l’expe
´rience malgache. IRD
Marseille, France, pp 43–71
Clement JC, Pinay G, Marmonier P (2002) Seasonal dynamics of
denitrification along topohydrosequences in three different
riparian wetlands. J Environ Qual 31:1025–1037
Comite
´de Bassin Rho
ˆne-Me
´diterrane
´e (2009) Programme de
mesures. Le SDAGE 2010-2015 du bassin Rho
ˆne-Me
´diterrane
´e.
Vers le bon e
´tat des milieux aquatiques, directive cadre
europe
´enne sur l’eau. Agence de l’eau Rho
ˆne-Me
´diterrane
´e
DR-A, Office national de l’eau et des milieux aquatiques,
De
´le
´gation re
´gionale Rho
ˆne-Alpes, Bassin Rho
ˆne-Me
´diterrane
´e
CWMG (California Wetlands Monitoring Workgroup) (2012)
California Rapid Assessment Method (CRAM) for Wetlands
and Riparian Areas. Version 5:2
DREAL (2006) Inventaires de
´partementaux des zones humides de
Rho
ˆnes-Alpes. http://www.zoneshumides-rhonealpes.fr/index.
php/linventaire
EPA (2006) Application of elements of a state water monitoring and
assessment program for wetlands. U.S. Environmental Protection
Agency. Wetlands Division OoW, Oceans and Watersheds),
Washington. http://www.epa.gov/owow/wetlands/monitor/
European Communities (1991) CORINE Biotopes: the design,
compilation and use of an inventory of sites of major importance
for nature conservation in the European Community. Report and
Manual. Office for Official Publications of the European
Communities, Luxembourg
Fennessy MS, Jacobs AD, Kentula ME (2007) An evaluation of rapid
methods for assessing the ecological condition of wetlands.
Wetlands 27:543–560
Fischer M, Wipf S (2002) Effect of low-intensity grazing on the
species-rich vegetation of traditionally mown subalpine mead-
ows. Biol Conserv 104:1–11
Hanson AR, Swanson L, Ewing D, Grabas G, Meyer S, Ross L,
Watmough M, Kirkby J (2008) Aperc¸u des me
´thodes d’e
´valuation
des fonctions e
´cologiques des terres humides. Se
´rie de Rapports
techniques no 497. Service canadien de la faune, Sackville, Canada
Hartig F, Drechsler M (2009) Smart spatial incentives for market-
based conservation. Biol Conserv 142:779–788
Hicks A, Carlisle B (1998) Rapid habitat assessment of wetlands,
macro-invertebrate survey version: brief description and method-
ology. Massachusetts Coastal Zone Management Wetland
Assessment Program, Amherst
Hossler K, Bouchard V, Fennessy MS, Frey SD, Anemaet E, Herbert
E (2011) No-net-loss not met for nutrient function in freshwater
marshes: recommendations for wetland mitigation policies.
Ecosphere 2:art82
Hough P, Robertson M (2009) Mitigation under Section 404 of the
Clean Water Act: where it comes from, what it means. Wetl Ecol
Manage 17:15–33
Hruby T (2012) Calculating credits and debits for compensatory
mitigation in wetlands of Western Washington, Final Report,
March 2012. Washington State Department of Ecology Publication
Jacobs AD (2010) Delaware Rapid Assessment Procedure Version
6.0. Delaware Department of Natural Resources and Environ-
mental Control. Dover, Delaware
Table 8 continued
Site code Site CRAM ORAM UMAM DERAP WAFAM
Landscape
1 Montfort Marsh 84.3 78.6 20.0 0.0 15.4
2 Montfort Meadow 77.7 64.3 33.0 0.0 15.4
3 Chapareillan Riparian Forest—Ise
`re side 90.2 57.1 58.0 33.3 58.3
4 Chapareillan Riparian Forest—Chartreuse side 77.7 35.7 37.0 66.7 61.5
5 Moı
¨les Wet Meadow 66.2 35.7 45.0 16.7 23.1
6 Moı
¨les Meadow 70.8 21.4 45.0 0.0 46.2
7 Loyes Riparian Forest 85.3 50.0 83.0 33.3 83.3
8 Loyes Reed bed 80.8 50.0 83.0 11.1 53.8
9 Creux Reed bed 80.6 57.1 77.0 79.2 50.0
10 Co
ˆte Chaude Riparian Forest 82.9 50.0 66.0 0.0 58.3
11 Buclet Riparian Forest—Left bank 92.9 100.0 75.0 50.0 58.3
12 Buclet Riparian Forest—Right bank 88.1 85.7 48.0 44.2 58.3
13 Bourg d’Oisans Meadow 93.3 100.0 68.0 0.0 46.2
258 Environmental Management (2015) 56:245–259
123
Johnson BJ, Beardsley M, Doran J (2013) The Functional Assessment
of Colorado Wetlands (FACWet) Methodology—Version 3.0.
Colorado Department of Transportation. https://www.codot.gov/
programs/environmental/resources/environmental-cards/wetlands/
01-0010-11.pdf
Lavoux T, Barrey G, Perret B, Rathouis P (2013) Evaluation du Plan
national d’action pour les zones humides 2010-2013 (PNZH).
Rapport no 008343-01 au Conseil Ge
´ne
´ral de l’Environnement et
du De
´veloppement Durable (CGEDD), Ministe
`re en charge de
l’environnement. Paris, France
Mack JJ (2001) Ohio rapid assessment method for wetlands, manual
for using version 5.0. Ohio EPA Technical Bulletin Wetland/
2001-1-1. Ohio Environmental Protection Agency, Division of
Surface Water, 401 Wetland Ecology Unit, Columbus, Ohio
Maltby E, Acreman MC (2011) Ecosystem services of wetlands:
pathfinder for a new paradigm. Hydrol Sci J 56:1341–1359
Matthews JW, Endress AG (2008) Performance criteria, compliance
success, and vegetation development in compensatory mitigation
wetlands. Environ Manage 41:130–141
McLaughlin DL, Cohen MJ (2013) Realizing ecosystem services:
wetland hydrologic function along a gradient of ecosystem
condition. Ecol Appl 23:1619–1631
MEA (2005) Millennium ecosystem assessment, ecosystems and
human well-being: synthesis. (ed. Press I) Washington
MEDDE (2012) Doctrine relative a la se
´quence e
´viter, re
´duire et
compenser les impacts sur le milieu naturel. Ministe
`re de
l’Ecologie, du De
´veloppement Durable et de l’Energie. Paris,
France
MEDDE (2013) Lignes directrices nationales sur la se
´quence e
´viter,
re
´duire et compenser les impacts sur les milieux naturels.
Ministe
`re de l’Ecologie, du De
´veloppement Durable et de
l’Energie. Paris, France
MEDDE (2014) 3e
`me plan national d’action en faveur des milieux
humides (2014-2018). Ministe
`re en charge de l’environnement,
Paris
Middleton BA, Holsten B, van Diggelen R (2006) Biodiversity
management of fens and fen meadows by grazing, cutting and
burning. Appl Veg Sci 9:307–316
Mitsch WJ, Wilson RF (1996) Improving the success of wetland
creation and restoration with know-how, time, and self-design.
Ecol Appl 6:77–83
Moreno-Mateos D, Power ME, Comı
´n FA, Yockteng R (2012)
Structural and functional loss in restored wetland ecosystems.
PLoS Biol. doi:10.1371/journal.pbio.1001247
Muller S (2002) Diversity of management practices required to ensure
conservation of rare and locally threatened plant species in
grasslands: a case study at a regional scale (Lorraine, France).
Biodivers Conserv 11(1173):1184
Nash RF (2014) Wilderness and the American mind. Yale University
Press, New Haven
Pinay G, Clement JC, Naiman RJ (2002) Basic principles and
ecological consequences of changing water regimes on nitrogen
cycling in fluvial systems. Environ Manage 30:481–491
Prins HH (1998) Origins and development of grassland communities
in northwestern Europe. In: WallisdeVries MF, Bakker JP, Van
Wieren SE (eds) Grazing and conservation management.
Springer, Berlin, pp 55–105
Que
´tier F, Lavorel S (2011) Assessing ecological equivalence in
biodiversity offset schemes: key issues and solutions. Biol
Conserv 144:2991–2999
Que
´tier F, Regnery B, Levrel H (2014) No net loss of biodiversity or
paper offsets? A critical review of the French no net loss policy.
Environ Sci Policy 38:120–131
Robertson M (2009) The work of wetland credit markets: two cases in
entrepreneurial wetland banking. Wetl Ecol Manage 17:35–51
Sabater S, Butturini A, Clement JC, Burt T, Dowrick D, Hefting M,
Maitre V, Pinay G, Postolache C, Rzepecki M, Sabater F (2003)
Nitrogen removal by riparian buffers along a European climatic
gradient: patterns and factors of variation. Ecosystems 6:20–30
Seigne-Martin A, Bernard M, Marciau R (2007) Plan de pre
´servation
et d’interpre
´tation du Marais de Montfort 2007–2011 (Commune
de Crolles). AVENIR - CREN, Grenoble
Smith RD, Ammann A, Bartoldus C, Brinson MM (1995) An
approach for assessing wetland functions using hydrogeomor-
phic classification, reference wetlands, and functional indices.
Wetlands Research Program Technical Report WRP-DE-9. U.S.
Army Engineers Waterways Experiment Station, Vicksburg, MS
StatSoft Inc (2004) STATISTICA (data analysis software system),
version 7. www.statsoft.com
Stein ED, Brinson M, Rains MC, Kleindl W, Hauer FR (2009a)
Wetland assessment debate: wetland assessment alphabet soup:
how to choose (or not choose) the right assessment method. Soc
Wetl Sci Bull 26:20–24
Stein ED, Fetscher AE, Clark RP, Wiskind A, Grenier JL, Sutula M,
Collins JN, Grosso C (2009b) Validation of a wetland rapid
assessment method: use of EPA’s level 1-2-3 framework for
method testing and refinement. Wetlands 29:648–665
Strain GF, Turk PJ, Anderson JT (2014) Functional equivalency of
created and natural wetlands: diet composition of red-spotted
newts (Notophthalmus viridescens viridescens). Wetlands Ecol
Manage. doi:10.1007/s11273-014-9362-6
Sutula MA, Stein ED, Collins JN, Fetscher AE, Clark R (2006) A
practical guide for the development of a wetland assessment
method: the californian experience. JAWRA 42:157–175
Tucker G, Allen B, Conway M, Dickie I, Hart K, Rayment M, Schulp
C, van Teeffelen A (2014) Policy options for an EU No Net Loss
initiative. Report to the European Commission. Institute for
European Environmental Policy, London
UMAM (2004) Uniform Wetland Assessment Method (UMAM) rule.
Chapter 62-345, Florida Statutes (F.S.) Florida Department of
Environmental Protection
Yepsen M, Baldwin AH, Whigham DF, McFarland E, LaForgia M,
Lang M (2014) Agricultural wetland restorations on the USA
Atlantic Coastal Plain achieve diverse native wetland plant
communities but differ from natural wetlands. Agric Ecosyst
Environ 197:11–20
Environmental Management (2015) 56:245–259 259
123
