A new diatom-based multimetric index to assess lake ecological status

Abstract

Eutrophication impairs lake ecosystems at a global scale. In this context, as benthic microalgae are well-established warnings for a large range of stressors, particularly nutrient enrichment, the Water Framework Directive required the development of diatom-based methods to monitor lake eutrophication. Here, we present the diatom-based index we developed for French lakes, named IBDL (Indice Biologique Diatomées en Lacs). Data were collected in 93 lakes from 2015 to 2020. A challenge arose from the discontinuous pressure gradient of our dataset, especially the low number of nutrient-impacted lakes. To analyze the data we opted for the so-called “Threshold Indicator Taxa ANalysis” method, which makes it possible to determine a list of “alert taxa.” We obtained a multimetric index based on specific pressure gradients (Kjeldahl nitrogen, suspended matter, biological oxygen demand, and total phosphorous). Considering the European intercalibration process, the very good correlation between IBDL and the common metric (R² from 0.52 to 0.87 according to the lake alkalinity type) makes us very confident in our ability to match future IBDL quality thresholds with European standards. The IBDL proved at last to be particularly relevant as it has a twofold interest: an excellent relationship with total phosphorus (R² from 0.63 to 0.83 according to the lake alkalinity type) and a possible application to any lake metatype. Its complementarity with macrophyte-based indices moreover justifies the use of at least two primary producer components for lake ecological status classification.

Supplementary Information
The online version contains supplementary material available at 10.1007/s10661-023-11855-w.

Full text

Vol.: (0123456789)
1 3
https://doi.org/10.1007/s10661-023-11855-w
RESEARCH
A new diatom‑based multimetric index toassess lake
ecological status
J.Tison‑Rosebery· S.Boutry· V.Bertrin·
T.Leboucher· S.Morin
Received: 16 May 2023 / Accepted: 6 September 2023
© The Author(s) 2023
Abstract Eutrophication impairs lake ecosystems
at a global scale. In this context, as benthic microal-
gae are well-established warnings for a large range of
stressors, particularly nutrient enrichment, the Water
Framework Directive required the development of
diatom-based methods to monitor lake eutrophica-
tion. Here, we present the diatom-based index we
developed for French lakes, named IBDL (Indice
Biologique Diatomées en Lacs). Data were collected
in 93 lakes from 2015 to 2020. A challenge arose
from the discontinuous pressure gradient of our data-
set, especially the low number of nutrient-impacted
lakes. To analyze the data we opted for the so-called
“Threshold Indicator Taxa ANalysis” method, which
makes it possible to determine a list of “alert taxa.”
We obtained a multimetric index based on specific
pressure gradients (Kjeldahl nitrogen, suspended
matter, biological oxygen demand, and total phos-
phorous). Considering the European intercalibration
process, the very good correlation between IBDL and
the common metric (R2 from 0.52 to 0.87 according
to the lake alkalinity type) makes us very confident
in our ability to match future IBDL quality thresholds
with European standards. The IBDL proved at last to
be particularly relevant as it has a twofold interest: an
excellent relationship with total phosphorus (R2 from
0.63 to 0.83 according to the lake alkalinity type) and
a possible application to any lake metatype. Its com-
plementarity with macrophyte-based indices moreo-
ver justifies the use of at least two primary producer
components for lake ecological status classification.
Keywords Ecological assessment· Lakes·
Phytobenthos· Water framework directive
Introduction
Eutrophication is one of the most frequent conse-
quences of human pressure on lake ecosystems at
a global scale (Stenger-Kovács et al., 2007). Pri-
mary producers are directly impacted since they
are the base of the aquatic food web (Brauer etal.,
2012). As the ability of species to compete differs
according to nutrient availability, nutrient enrich-
ment results in significant changes in community
structure and function (Birk etal.,2012). For this
reason, scientists and policymakers developed indi-
ces based on primary producer attributes to moni-
tor eutrophication (Stevenson, 2014). In the early
Supplementary Information The online version
contains supplementary material available at https:// doi.
org/ 10. 1007/ s10661- 023- 11855-w.
J.Tison-Rosebery(*)· S.Boutry· V.Bertrin· S.Morin
INRAE, UR EABX, 33612Cestas, France
e-mail: juliette.rosebery@inrae.fr
J.Tison-Rosebery· S.Boutry· V.Bertrin· S.Morin
Pôle R&D ECLA, LeBourget-du-Lac, France
T.Leboucher
Université de Lorraine, CNRS, LIEC, 57000Metz, France
/ Published online: 13 September 2023
Environ Monit Assess (2023) 195:1202
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Environ Monit Assess (2023) 195:1202
1 3
Vol:. (1234567890)
2000s, the Water Framework Directive (European
Union, 2000) required all EU member states to
implement bioassessment methods based, among
other aspects, on the biological quality of “macro-
phytes and phytobenthos” to assess lake ecological
status. This led to the development of numerous
methods at the European level.
Poikane etal. (2016) reviewed this panel of meth-
ods and observed that countries generally developed
separate assessment tools for macrophytes and phy-
tobenthos, and that most of them considered diatoms,
which are unicellular microalgae, to be a good proxy
for phytobenthos. Diatoms are indeed early and well-
established warnings for a large range of stressors,
particularly nutrient enrichment (Stevenson, 2014).
As a first step, indices originally dedicated to rivers
were applied to lakes by the majority of member states
(Kelly etal., 2014b), considering that many processes
influencing diatom assemblages were comparable
between lakeshores and shallow rivers (Cantonati &
Lowe, 2014).
In some rare cases, diatom-based indices were
developed specifically for lakes, based on species
composition and abundance as for rivers (Bennion
etal., 2014; Poikane etal., 2016). Diatoms from mud
and silts were generally not considered, as they would
respond to pore-water chemistry rather than water
quality. The recommended sampling substrate varied
according to authors, from macrophytes to cobbles or
even artificial substrates when no natural substrates
are found in all water bodies (King etal., 2006).
To harmonize the different national approaches, a
European intercalibration exercise was performed,
involving eleven member states (Kelly etal., 2014b).
France participated in this exercise with the Biological
Diatom Index (BDI, Coste etal., 2009), routinely used
to assess river ecological status. Although previous
results tended to suggest there was a good correlation
between BDI and the environmental pressure gradi-
ents, at least in shallow lakes (Cellamare etal., 2012),
this intercalibration exercise revealed a poor correla-
tion between BDI values and total phosphorous across
France (Kelly etal., 2014b). This was explained by the
absence of many lake taxa from the list of key species
used to calculate the BDI, resulting in an overall poor
relevance of the final status assessment.
The aim of the present study was, therefore, to develop
a new diatom-based index for lakes in metropolitan
France: the IBDL (Indice Biologique Diatomées en Lac:
Diatom Biological Index for Lakes). To collect the nec-
essary data, we proposed a method (Morin etal., 2010)
consistent with a potential subsequent combination of this
index with the existing French macrophyte index IBML
(Indice Biologique Macrophytique en Lac: Macrophyte
Biological Index for Lakes, Boutry etal., 2015). We detail
here how diatom data were sampled and analyzed and
how we developed the IBDL. Finally, we discuss the rel-
evance of this new index, comparing the results obtained
with index scores based on macrophytes, and assessing its
ability to reveal environmental gradients.
Materials andmethods
Data collection
Samples were collected from 93 French lakes during
the summer period, between 2015 and 2020 as part
of national assessment surveys, according to Morin
etal. (2010) (Fig.1 and S1 Table1). The lakes were
classified into three metatypes based on alkalinity,
according to the European intercalibration exercise
previously performed (Kelly et al., 2014b): low
alkalinity (LA, alkalinity ≤ 0.2 meq.l−1), medium
alkalinity (MA, 0.2 meq.l−1 < alkalinity < 1 meq.
l−1), and high alkalinity (HA, alkalinity ≥ 1meq.l−1).
Diatoms were collected from both mineral substrates
and lakeshore macrophyte surfaces in observation
units (OUs), whose number and location varied
according to the lake surface area and the riparian
zone types. Such units are defined in the French
macrophyte sampling protocol for lakes NF T90-328
(AFNOR, 2022).
Biological data
Samples from hard mineral substrates were taken from at
least five boulders or cobbles selected at random for each
OU. The total surface area sampled was equivalent to 100
cm2, as defined in the NF T90-354 standard (AFNOR,
2016). Selected substrates had to be submerged within
the euphotic zone at a maximum depth of 0.5m.
Samples performed on macrophytes were taken
from helophytes (mainly Phragmites australis (Cav.)
Trin. ex Steud.). Green stem segments submerged for
at least 4 to 6weeks were collected from a minimum of
5 macrophytes chosen at random. These stem segments
had to be located at a maximum depth of 0.2m.
1202 Page 2 of 12
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Environ Monit Assess (2023) 195:1202
1 3
Vol.: (0123456789)
Diatoms were sampled from both substrates accord-
ing to the NF T90-354 protocol, in line with the Euro-
pean standards (EN 13946; CEN, 2003). Cells were
identified at 1000× magnification by examining per-
manent slides of cleaned diatom frustules (400 valves
per slide) using, among others, Krammer and Lange-
Bertalot (1986–1991) and Lange-Bertalot (1995–2015,
2000–2013). Taxonomic homogenization was per-
formed with Omnidia 6 software (Lecointe etal., 1993).
All OUs from a single lake were sampled within a
maximum of 21days. Diatom counts had to include
at least 350 cells per slide, with more than 50% of the
diatom cells determined at the species level, to com-
ply with the NF T90-354 requirements.
Physico‑chemical data
Parameter values were determined in summer in
the euphotic layer at the deepest point of each
lake, according to European standards. Data were
obtained from national surveillance monitoring pro-
grams. Water quality analysis was not systematically
Fig. 1 Study sites, number
of surveys per site, and lake
alkalinity classes (LA, low
alkalinity; MA, medium
alkalinity; HA, high alkalin-
ity) (Kelly etal., 2014a)
Table 1 Physico-chemical data available for analysis
Variable % of
missing
values
Mean sd Median p25 p75 Maximum
Kjeldahl nitrogen (NKJ, mg.l−1) 0.292 0.661 0.959 0.25 0.25 0.7 6.9
Ammonium (NH4, mg.l−1) 0.292 0.09 0.35 0.015 0.01 0.06 3.3
Biological oxygen demand (BOD5, mg.l−1) 0.584 2.157 2.615 1.3 0.9 1.8 12
Conductivity (Cond, µs.cm2) 0.309 230.108 124.368 243.5 158 297 815
Nitrates (NO3, mg.l−1) 0.292 1.113 1.222 0.6 0.25 1.4 6.07
Nitrites (NO2, mg.l−1) 0.292 0.011 0.025 0.005 0.005 0.01 0.3
Orthophosphates (PO4, mg.l−1) 0.292 0.015 0.026 0.005 0.005 0.01 0.22
Oxygen (O2, mg.l−1) 0.333 8.938 1.654 8.7 8.1 9.665 14.74
Oxygen saturation (% O2) 0.333 110.203 20.91 108 101 117.65 187
Suspended particles (SP, mg.l−1) 0.292 7.979 18.145 2.8 1.6 5 153
Total phosphorous (Pt, mg.l−1) 0.292 0.027 0.067 0.005 0.005 0.015 0.51
Page 3 of 12 1202
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Environ Monit Assess (2023) 195:1202
1 3
Vol:. (1234567890)
performed each year: in a few cases (10% of the sam-
ples), the most recent physicochemical data available
were collected the year after or before the diatom
samples. The following parameters were recorded:
biological oxygen demand (BOD5, mg.l−1), oxygen
(O2, mg.l−1), oxygen saturation (% O2), conductiv-
ity (Cond, µs.cm2), Kjeldahl nitrogen (NKJ, mg.l−1),
ammonium (NH4, mg.l−1), nitrates (NO3, mg.l−1),
nitrites (NO2, mg.l−1), orthophosphates (PO4, mg.l−1),
total phosphorous (Pt, mg.l−1), and suspended parti-
cles (SP, mg.l−1).
Data analysis and index settlement
All analyses were performed with R version 4.1.2
(2021–11-01) (R Core Team, 2021) (Platform: x86_64-pc-
linux-gnu (64-bit), Running under: Ubuntu 22.04.1 LTS).
Considering that the final dataset revealed a dis-
continuous trophic gradient, we opted for the so-
called Threshold Indicator Taxa ANalysis method
(TITAN2 package, Baker etal., 2020), which, based
on bootstrapping and permutations, makes it possible
to determine a list of “alert taxa.” The presence and/or
increasing abundance of alert taxa reveals the exist-
ence of anthropogenic pressures. TITAN replaces the
community‐level response along a composite gra-
dient with taxon‐specific responses toward single-
environmental variables (Dufrêne & Legendre, 1997).
Negative and positive responses are distinguished,
and cumulative decreasing or increasing responses in
the community are tracked. This method is particu-
larly suitable for setting up multimetric indices.
A three-step procedure was necessary to build
our biological diatom index for lakes (IBDL): iden-
tification of alert taxa, choice of relevant metrics,
and aggregation of these metrics to obtain the final
index score.
Identification ofalert taxa
For the next part of the analysis, we set an occurrence
threshold ≥ 3 for taxa to be included in the index cal-
culation (the so-called index taxa).
TITAN combines change-point analysis (nCPA; King
& Richardson, 2003) and indicator species analysis (Ind-
Val, Dufrêne et al., 1997). Basically, the change-point
analysis compares within-group vs. between-group dis-
similarity to detect shifts in community structure along
the environmental variable considered (for further details
concerning this method, see Baker and King (2010)).
Indicator species analysis then identifies the strength of
association between any particular taxon and this sample
grouping. At the end of the process, two IndVal scores
are calculated for a single taxon in a two-group classifica-
tion. The algorithm finally classifies taxa into three differ-
ent categories: Z+ taxa, showing a significant increase in
abundance along the increasing environmental gradient;
Z− taxa, showing a significant decrease along this gradi-
ent; and indifferent taxa, with no significant trend.
Alert taxa were defined as Z+ or Z− taxa whose
shift thresholds were greater or lesser than the com-
munity shift threshold.
Building metrics andselecting therelevant ones
For each environmental variable, a metric was calcu-
lated at the OU scale according to (1)
where Alerttaxa is the number of alert taxa and
Indextaxa is the number of index taxa in the sample.
The metric value is bounded between 0 and 1. The
lowest value (0) corresponds to a species list entirely
composed of alert taxa (determined for the environ-
mental variable considered).
To build our index, we then selected the most rele-
vant metrics, i.e., those with the best relationship with
the environmental parameter considered. We used
Pearson’s correlation coefficients to measure this
statistical association and only kept metrics show-
ing a Pearson’s coefficient over 0.6. Metrics should
significantly increase with impairment, significantly
decrease with impairment, or show no particular pat-
tern. We obtained the response patterns of the differ-
ent metrics by transforming raw values into normal-
ized deviations (standardized effect size: SES, Gotelli
& McCabe, 2002; Mondy etal., 2012) (2). SES val-
ues made it possible to obtain a single response pat-
tern for a metric whatever the lake metatype and sub-
strate type considered.
where MetricM is the observed value of the metric,
and Mgroup and sdgroup are the mean and standard
(1)
MetricM=1−
(
Alerttaxa
Index
axa )
(2)
SES
M=(
Metric
M
−M
group
sd
group )
1202 Page 4 of 12
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Environ Monit Assess (2023) 195:1202
1 3
Vol.: (0123456789)
deviation, respectively, of the metric value for a given
group of samples (i.e., substrate type × lake alkalinity
metatype) (values of Mgroup and sdgroup are given in
Table1 S2).
The next step consisted of the normalization of
SES values (SESnorM) to make comparable metric
variation ranges (3):
where SESM is the observed value of SES for a given
metric, Min its minimum value, and Max its maxi-
mum value in the whole dataset (values of Min and
Max are given in Table2 S2).
We further transformed metric values from nor-
malized SES into the ecological quality ratio (EQR)
(4), i.e., the ratio between the observed value of a
metric (SESnorM) and its expected value under ref-
erence conditions, for any lake metatype and any
substrate (SESnorMref, values given in Table 3 S2)..
National reference conditions were set based on lakes
characterized by very low or negligible anthropic
pressure. This selection was checked according to
the land use criteria applied during the initial lake
intercalibration exercise (Kelly et al., 2014a). Lakes
were deemed to be in reference condition if show-
ing < 0.4% artificial land use and < 20% agriculture
within the catchment area.
Finally, for each metric, we performed a Wilcoxon
test to detect the potential influence of substrate type
on the EQR values obtained at the OU scale.
Aggregating metric values toobtain thefinal IBDL score
The final index score was obtained at the OU scale by
averaging the selected metric values, expressed in EQR.
For a score calculated for both mineral and macro-
phyte substrates, the lowest value was considered the
final score.
Each OU belongs to one of the four riparian
zone types, as required in the NF T90-328 standard
(AFNOR, 2022). These types were defined from
the vegetation composition and/or anthropogenic
alterations of the lakeshore. The percentage of each
(3)
SESnor
M=
(
SESM−Min
)
(Max −Min)
(4)
EQR
=
(
SESnorM
SESnor
Mref )
riparian zone type was estimated insitu, on the whole
lake perimeter, during the sampling surveys. The
final index score for the whole lake was derived from
a weighted average of the ScoreOU (5), taking into
account the percentage of the lake perimeter each OU
represented in terms of riparian zone type (Pctype).
Finally, the resulting IBDL scores varied between 0
(worst water quality) and 1. Relationships between IBDL
scores and the different environmental variables consid-
ered were tested a posteriori with simple linear regres-
sions (R “mass” package, Venables & Ripley, 2002).
Comparing IBDL and IBML scores
We compared IBDL and IBML scores, based, respec-
tively, on diatom and macrophyte communities to
evaluate their complementarity or redundancy. IBML
scores were computed with the online application
https:// seee. eaufr ance. fr/ api/ indic ateurs/ IBML/1. 0.1
and the “httr” package (Wickham, 2022).
We built a multiple linear regression model
(“mass” package) to test which index correlated best
with Pt values: IBML, IBDL, or a combination of
both (mean value).
Preparing intercalibration
Considering a future intercalibration exercise, we
analyzed the relationships between IBDL scores and
Pt for each lake metatype. A good correlation of the
candidate metric with Pt constitutes a key criterion
for considering the index ready for integration into
the intercalibration process (Kelly etal., 2014b).
We also plotted IBDL against CM scores (intercal-
ibration common metric, i.e., the trophic index devel-
oped by Rott etal., 1998) to check their compliance.
The CM was calculated with Omnidia 6 software.
Results
Our data revealed discontinuous pressure gradients
(Table 1), with a clear lack of impacted conditions
and an over-representation of lakes characterized by
low eutrophication levels.
(5)
IBDL
=
∑4
type=
1
(
ScoreOU ∗Pctype
)
Page 5 of 12 1202
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Environ Monit Assess (2023) 195:1202
1 3
Vol:. (1234567890)
sd, standard deviation; p25, 25th percentile; p75,
75th percentile.
Biotic and abiotic data were obtained for 958 sam-
ples. Considering the data validation criteria, 99%
of the samples were included in the analysis. Sixty-
eight, 202, and 402 OUs were, respectively, sampled
on LA, MA, and HA lakes (unknown alkalinity type
for 8 lakes). Table2 S1 specifies the substrates sam-
pled for each alkalinity type. Data from both substrate
types were available for 552 OUs. Seven hundred
eighty taxa were recorded, 8% of which were identi-
fied to the genus level. One hundred and twenty-one
alert taxa were determined out of 590 index taxa (S3).
We obtained the following Pearson test values for
the different metrics at the OU scale: R = −0.715 for
the metric based on the parameter NKJ, R = −0.754
for BOD5, R = −0.688 for Pt, R = −0.666 for SP,
R = −0.553 for PO4, R = −0.329 for conductiv-
ity, R = −0.174 for O2, R = −0.265 for NO2, and
R = −0.204 for %O2. Considering the selection rule
proposed (|R|> 0.6), only the metrics based on NKJ,
BOD5, Pt, and SP were considered to build the IBDL.
Metric values (in EQR) calculated from the lists
of taxa sampled on mineral substrates and macro-
phytes for a single OU did not differ significantly
(p-value = 0.65).
IBDL scores at the lake level were calculated from
the selected metrics following the aggregation rules
proposed. The scores obtained were distributed as
given in Fig.2. IBDL could not be calculated for 20%
of the samples due to incomplete floristic data.
The relationships between IBDL scores and
the different environmental variables considered
were very good (Fig.3) in both high-alkalinity and
medium-alkalinity lakes. IBDL scores showed high
correlations with these variables, particularly Pt,
in both high alkalinity (R2 = 0.63, p = 1.8e−15) and
medium alkalinity lakes (R2 = 0.83, p = 8.3e−11). Note
that data from low alkalinity lakes were too scarce to
perform such correlations.
IBDL scores were also strongly associated with CM
scores (R2 = 0.52 and p = 2.2e−16 for high-alkalinity
lakes; R2 = 0.87 and p = 1.8 e−7 for medium-alkalinity
lakes) (Fig.4).
IBDL scores showed a better correlation with Pt
(AIC = −171.44) than did IBML (AIC = −129.25)
or a combination of both indices (AIC = −169.44).
Nevertheless, IBDL tended to be generally less strin-
gent than IBML (in 18 out of 22 samples), especially
for scores higher than 0.8 (clearly dominant here).
Figure 5 presents the difference between IBDL and
IBML scores according to IBDL scores.
Discussion
As required by the WFD, we developed a diatom
index for the assessment of the ecological status of
French lakes. We obtained very good correlations
between IBDL and key environmental variables. One
major challenge arose from the discontinuous pres-
sure gradient of our dataset, especially the low avail-
able number of nutrient-impacted lakes.
The scarcity of impacted lakes in the datasets used
to build diatom indices is not rare and has already
been pointed out by some authors (Bennion et al.,
2014). This lack makes it impossible to capture the
entire trophic gradient or to build reliable species’
ecological profiles. However, the majority of existing
indices are calculated as an abundance-weighted aver-
age of the ecological profiles of every taxon from a
sample, according to the Zelinka and Marvan formula
(Zelinka & Marvan, 1961). This method is far from
optimal for datasets showing discontinuous or very
specific environmental conditions (Carayon et al.,
2020). In such cases, the identification of alert taxa
seems more appropriate than considering diatom com-
munities as a whole. This has made the TITAN algo-
rithm increasingly popular for detecting specific taxa
providing reliable signals of a specific stress (Carayon
etal., 2020; Costas etal., 2018; Gieswein etal., 2019;
Gonzalez-Paz etal., 2020; Khamis etal., 2014).
Using this method, we built a multimetric index
based on different pressure gradients (NKJ, SP,
BOD5, and Pt). Although the strong influence of
nutrients and organic matter on diatom community
composition is well established (Jüttner etal.,2010;
Stevenson etal., 2013), diatom-based metrics rarely
take into account suspended particles for water qual-
ity assessment (but see Larras etal., 2017). Diatoms
are indeed directly impaired by turbidity, reducing
light availability for photosynthesis. Multimetric indi-
ces thus offer simple tools to summarize the effect
of multi-pressure gradients on communities (Riato
et al., 2018), and can be considered more effec-
tive for assessing biological conditions than a single
metric (Stevenson et al., 2013). However, despite
1202 Page 6 of 12
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Environ Monit Assess (2023) 195:1202
1 3
Vol.: (0123456789)
their increasing use, multimetric indices suffer from
the subjectivity that can arise from metric selection
(Reavie etal., 2008). Here, we attempted to avoid this
pitfall by proposing a method of selecting metrics
based on the robustness of their response to environ-
mental gradients.
IBDL appears less stringent than IBML when
assessing lakes’ ecological status. Literature com-
paring results from different indices in lakes, though
scarce, tends to agree with this overestimation of water
quality by diatom-based methods (Kolada etal., 2016).
Phytobenthos has long been paid less attention than
macrophytes for the assessment of lake ecological sta-
tus. It is true that recent diatom-based metrics barely
detected newly impacted lakes that would not have
been detected by macrophyte metrics. Bennion etal.
(2014) showed, for example, that their index (LTDI)
performed well for lakes with good ecological status,
but diatoms and other methods agreed less for lakes
of lower status. This was particularly the case in the
presence of morphological alterations, for which dia-
toms are poor indicators. A possible general explana-
tion for the lower stringency of diatom-based indices
in lakes is the high abundance of species complexes
like Achnanthidium minutissimum or Gomphonema
parvulum. Such complexes merge taxa that are mor-
phologically close but with different ecological pref-
erences. Due to the existence of different taxa within
the A. minutissimum complex, many authors consider
it an indicator of good water quality (Almeida etal.,
2014), whereas others consider it tolerant toward toxic
contaminants (micropollutants) and hydrologic dis-
turbances (Cantonati etal.,2014; Lainé etal., 2014).
Considering the generally high abundance of A.
minutissimum in samples, this tends to blur the overall
pressure-response relationship between index scores
and environmental variables (Potapova & Hamilton,
2007). TITAN provides a means to avoid this pitfall,
as such complexes are not selected as alert taxa, given
that their abundance dynamics do not show clear
Fig. 2 Distribution of the IBDL scores obtained (p25, 25th percentile; p50, median value; p75, 75th percentile)
Page 7 of 12 1202
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Environ Monit Assess (2023) 195:1202
1 3
Vol:. (1234567890)
response patterns to environmental gradients. Indeed,
A. minutissimum, although highly abundant in our
dataset (22% of total species abundances), was not
considered an alert taxon.
The fact remains that IBDL tends to be less strin-
gent than IBML, despite better relationships with Pt.
In consequence, we have to explain why we think that
the use of diatom-based indices to assess lake ecolog-
ical status is justified.
First, the discrepancy between macrophyte and
diatom responses relies mainly on the differences
between their integration periods, given that indices
provide information on ecological conditions over
the time an assemblage develops. Lavoie etal. (2009)
showed the integration period of diatom-based indi-
ces to be about 2–5 weeks for nutrients, whereas
macrophytes react on yearly time scales (Kelly etal.,
2016). As diatoms catch nutrients directly from the
water column (Wetzel, 2001),they also may be more
directly sensitive to rapid changes in trophic status
than macrophytes (Vermaat et al., 2022). The rapid
response of phytobenthos should justify its routine
use (Schneider etal.,2019), in particular, for lakes in
non-equilibrium states (Kelly etal., 2016).
Second, diatom-based indices are essential where
hydrologic pressures in littoral areas prevent the devel-
opment of macrophytes, and in lake typologies where
macrophyte communities are naturally species poor or
even absent (Schneider etal.,2019). Thus, while mac-
rophyte-based indices cannot be calculated in all lakes,
this is not true for diatom-based indices. Moreover, our
results show that, with IBDL, water quality managers
can directly compare ecological status assessments
from different lakes even if the substrate sampled is
different. Many studies highlighted that allelopathic
relationships between macrophytes and epiphytic dia-
toms may be responsible for specific associations
between macrophytes and diatom species and, thus,
Fig. 3 Relationships between IBDL and the environmental variables considered (MA, medium-alkalinity lakes; HA, high-alkalinity
lakes; BOD5, biological oxygen demand; NKJ, Kjeldahl nitrogen; Pt, total phosphorous; SP, suspended particles)
1202 Page 8 of 12
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Environ Monit Assess (2023) 195:1202
1 3
Vol.: (0123456789)
may contribute to the organization of particular assem-
bly patterns (Hinojosa-Garro etal., 2010). In any case,
in terms of ecological preferences, and consequently
in terms of IBDL scores, our results did not show any
significant differences between communities sampled
on mineral substrates or macrophytes at the OU level,
corroborating previous results obtained by Kitner and
Poulíčková (2003) and Bennion et al. (2014). Other
studies even support the use of epiphytic diatoms as
biological indicators for lakes irrespective of the domi-
nant macrophyte species sampled (Cejudo-Figueiras
etal., 2010). The key point is to avoid senescent mate-
rial or recently grown shoots that would potentially
induce a colonization stage effect (King etal., 2006).
The next challenge was to check the consist-
ency of the resulting classification of lakes based
on IBDL to the harmonized definition of good eco-
logical status established in the completed intercali-
bration exercise (Kelly etal., 2014b). The first step
consisted in testing the correlation between IBDL
scores and total phosphorus in our dataset. Only
HA and MA typologies were considered here but,
in any case, the last intercalibration exercise could
not be performed for LA lakes. We obtained very
good correlations that are clearly an improvement
compared to the non-significant relationship pre-
viously obtained between BDI (diatom index used
for the assessment of rivers) and Pt, and even better
than the pressure-impact relationships observed at a
pan-European scale (R2 between national methods
and Pt ranged from 0.32 to 0.66 max., Kelly etal.,
2014b). The second step consisted in testing the
correlation between IBDL scores and the intercali-
bration common metric (CM) scores, in EQR. Here,
the correlations demonstrated a very good agree-
ment between IBDL and CM scores in both medium
(R2 = 0.87) and high alkalinity (R2 = 0.82) lakes.
We are, therefore, confident in our ability to match
IBDL ecological status thresholds with those vali-
dated at the European level.
Fig. 4 Relationships between IBDL and the common metric (CM) in medium alkalinity (MA) and high alkalinity (HA) lakes
Page 9 of 12 1202
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Environ Monit Assess (2023) 195:1202
1 3
Vol:. (1234567890)
Conclusion
The new diatom index proposed here meets the
requirements of the WFD and makes it possible to
assess lakes’ ecological status in metropolitan France.
The IBDL has proved to be particularly relevant as
it has a twofold interest: an excellent relationship
with total phosphorus and an application in any lake
metatype. Its complementarity with IBML justifies
the use of at least two primary producer components
for ecological status classification (Kelly etal., 2016).
Acknowledgements We thank all Water Agencies for data
sharing and all Regional Departments for Environment for data
collection. We also thank the two reviewers for their helpful
comments on this work.
Author contribution All authors participated in designing
the study and developing aims and research questions. S.B.
designed methodology, extracted data and made the analyses,
supported by T.L. concerning pretreatments before intercali-
bration. J.T.R. led the writing of the manuscript supported by
S.B., S.M., and V.B. All authors contributed critically to the
drafts, contributed to the final version of the manuscript, and
gave final approval for publication.
Funding The research leading to these results received fund-
ing from the French Biodiversity Agency (OFB, pôle ECLA).
Availability of data and code The data that support the findings of
this study are openly available athttps:// doi. org/ 10. 57745/ PDKBGB.
Declarations
Competing interests The authors declare no competing interests.
Open Access This article is licensed under a Creative Com-
mons Attribution 4.0 International License, which permits
use, sharing, adaptation, distribution and reproduction in any
medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Crea-
tive Commons licence, and indicate if changes were made. The
images or other third party material in this article are included
in the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not
Fig. 5 Difference between IBDL and IBML scores according to IBDL scores
1202 Page 10 of 12
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Environ Monit Assess (2023) 195:1202
1 3
Vol.: (0123456789)
included in the article’s Creative Commons licence and your
intended use is not permitted by statutory regulation or exceeds
the permitted use, you will need to obtain permission directly
from the copyright holder. To view a copy of this licence, visit
http://creativecommons.org/licenses/by/4.0/.
References
AFNOR. (2016). Qualité de l’eau – Échantillonnage, traitement et
analyse de diatomées benthiques en cours d’eau et canaux.
Association française de normalisation, Norme NF 90 T-354.
AFNOR. (2022). Qualité de l’eau - Echantillonnage des com-
munautés de macrophytes en plans d’eau. Norme NF
T90–328.
Almeida, S. F., Elias, C., Ferreira, J., Tornés, E., Puccinelli, C.,
Delmas, F., & Sabater, S. (2014). Water quality assess-
ment of rivers using diatom metrics across Mediterranean
Europe: A methods intercalibration exercise. Science of
the Total Environment, 476, 768–776.
Baker, M. E., & King, R. S. (2010). A new method for detecting and
interpreting biodiversity and ecological community thresholds.
Methods in Ecology and Evolution, 1(1), 25–37.
Baker, E., King, R. S., & Kahle, D. (2020). TITAN2: Thresh-
old indicator taxa analysis. R package version 2.4.1.
Retrieved May, 15, 2023, from https:// CRAN.R- proje ct.
org/ packa ge= TITAN2
Bennion, H., Kelly, M. G., Juggins, S., Yallop, M. L., Burgess,
A., Jamieson, J., & Krokowski, J. (2014). Assessment
of ecological status in UK lakes using benthic diatoms.
Freshwater Science, 33(2), 639–654.
Birk, S., Bonne, W., Borja, A., Brucet, S., Courrat, A., Poikane,
S., & Hering, D. (2012). Three hundred ways to assess
Europe’s surface waters: An almost complete overview of
biological methods to implement the Water Framework
Directive. Ecological Indicators, 18, 31–41.
Boutry, S., Bertrin, V., & Dutartre, A. (2015). Indice
Biologique Macrophytique en Lac (IBML): notice de cal-
cul. Irstea, pp.25. ⟨hal-02602320⟩.
Brauer, V. S., Stomp, M., & Huisman, J. (2012). The nutrient-
load hypothesis: Patterns of resource limitation and com-
munity structure driven by competition for nutrients and
light. The American Naturalist, 179(6), 721–740.
Cantonati, M., & Lowe, R. L. (2014). Lake benthic algae:
Toward an understanding of their ecology. Freshwater
Science, 33(2), 475–486.
Carayon, D., Eulin-Garrigue, A., Vigouroux, R., & Delmas,
F. (2020). A new multimetric index for the evaluation
of water ecological quality of French Guiana streams
based on benthic diatoms. Ecological Indicators, 113,
106248.
Cejudo-Figueiras, C., Alvarez-Blanco, I., Bécares, E., & Blanco, S.
(2010). Epiphytic diatoms and water quality in shallow lakes:
The neutral substrate hypothesis revisited. Marine and Fresh‑
water Research, 61(12), 1457–1467.
Cellamare, M., Morin, S., Coste, M., & Haury, J. (2012). Eco-
logical assessment of French Atlantic lakes based on phy-
toplankton, phytobenthos and macrophytes. Environmen‑
tal Monitoring and Assessment, 184(8), 4685–4708.
CEN (Comité Européen de Normalisation). (2003). Water qual-
ity - Guidance standard for the routine sampling and pre-
treatment of benthic diatoms from rivers. EN 13946:2003.
Comité Européen de Normalisation, Geneva, Switzerland.
Costas, N., Pardo, I., Méndez-Fernández, L., Martínez-Madrid,
M., & Rodríguez, P. (2018). Sensitivity of macroinverte-
brate indicator taxa to metal gradients in mining areas in
Northern Spain. Ecological Indicators, 93, 207–218.
Coste, M., Boutry, S., Tison-Rosebery, J., & Delmas, F. (2009).
Improvements of the biological diatom index (BDI):
Description and efficiency of the new version (BDI-
2006). Ecological Indicators, 9(4), 621–650.
Dufrêne, M., & Legendre, P. (1997). Species assemblages and
indicator species: The need for a flexible asymmetrical
approach. Ecological Monographs, 67(3), 345–366.
European Union. (2000). Directive 2000/60/EC of the european
parliament and of the council of 23rd october 2000 establish-
ing a framework for community action in the field of water
policy. Official journal of european communities, european
commission, brussels (2000) (22 December, L 327/1).
Gieswein, A., Hering, D., & Lorenz, A. W. (2019). Development
and validation of a macroinvertebrate-based biomonitor-
ing tool to assess fine sediment impact in small mountain
streams. Science of the Total Environment, 652, 1290–1301.
Gonzalez-Paz, L., Delgado, C., & Pardo, I. (2020). Under-
standing divergences between ecological status classifica-
tion systems based on diatoms. Science of the Total Envi‑
ronment, 734, 139418.
Gotelli, N. J., & McCabe, D. J. (2002). Species co-occurrence:
A meta-analysis of J.M. Diamond’s assembly rules model.
Ecology, 83(8), 2091–2096.
Hinojosa-Garro, D., Mason, C. F., & Underwood, G. J. (2010).
Influence of macrophyte spatial architecture on periphyton and
macroinvertebrate community structure in shallow water bodies
under contrasting land management. Fundamental and Applied
Limnology, 177(1), 19–28.
Kelly, M., Urbanic, G., Acs, E., Bennion, H., Bertrin, V., Burgess,
A., Denys, L., Gottschalk, S., Kahlert, M., Karjalainen, S.,
Kennedy, B., Kosi, G., Marchetto, A., Morin, S., Picinska-
Fałtynowicz, J., Poikane, S., Rosebery, J., Schoenfelder, I.,
Schoenfelder, J., & Varbiro, G. (2014a). Comparing aspira-
tions: Intercalibration of ecological status concepts across
European lakes for littoral diatoms. Hydrobiologia, 734,
125–141.
Kelly, M., Acs, E., Bertrin, V., Bennion, H., Borics, G., Bur-
gess, A., Denys, L., Ecke, F., Kahlert, M., Karjalainen,
S., Kennedy, B., Marchetto, A., Morin, S., Picinska Fal-
tynowicz, J., Phillips, G., Schönfelder, I., Schönfelder, J.,
Urbanic, G., Van Dam, H., & Zalewski, T. (2014b). Water
Framework Directive Intercalibration Technical Report:
Lake Phytobenthos ecological assessment methods Publi-
cations Office of the European Union, 125.
Kelly, M. G., Birk, S., Willby, N. J., Denys, L., Drakare, S.,
Kahlert, M., Karjalainen, S. M., Marchetto, A., Pitt, J.-A.,
Urbanic, G., & Poikane, S. (2016). Redundancy in the
ecological assessment of lakes: Are phytoplankton, macro-
phytes and phytobenthos all necessary? Science of the Total
Environment, 568, 594–602.
Khamis, K., Hannah, D. M., Brown, L. E., Tiberti, R., & Mil-
ner, A. M. (2014). The use of invertebrates as indicators of
Page 11 of 12 1202
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Environ Monit Assess (2023) 195:1202
1 3
Vol:. (1234567890)
environmental change in alpine rivers and lakes. Science of
the Total Environment, 493, 1242–1254.
King, L., Clarke, G., Bennion, H., Kelly, M., & Yallop, M.
(2006). Recommendations for sampling littoral diatoms in
lakes for ecological status assessments. Journal of Applied
Phycology, 18, 15–25.
King, R. S., & Richardson, C. J. (2003). Integrating bioassess-
ment and ecological risk assessment: An approach to
developing numerical water-quality criteria. Environmen‑
tal Management, 31(6), 795–809.
Kitner, M., & Poulícková, A. (2003). Littoral diatoms as indi-
cators for the eutrophication of shallow lakes. Hydrobio‑
logia, 506(1), 519–524.
Kolada, A., Pasztaleniec, A., Bielczyńska, A., & Soszka, H. (2016).
Phytoplankton, macrophytes and benthic diatoms in lake clas-
sification: Consistent, congruent, redundant? Lessons learnt
from WFD-compliant monitoring in Poland. Limnologica, 59,
44–52.
Krammer, K., & Lange-Bertalot, H. (1986–1991). Bacillario‑
phyceae. Suswasserflora von Mitteleuropa. Gustav Fisher
Verlag.
Jüttner, I., Chimonides, P. J., & Ormerod, S. J. (2010). Using
diatoms as quality indicators for a newly-formed urban
lake and its catchment. Environmental Monitoring and
Assessment, 162, 47–65.
Lainé, M., Morin, S., & Tison-Rosebery, J. (2014). A multi-
compartment approach -diatoms, macrophytes, benthic
macroinvertebrates and fish- to assess the impact of toxic
industrial releases on a small French river. PLoS ONE,
9(7), e102358.
Lange-Bertalot, H. (1995–2015). Iconographia Diatomologica.
Annotated Diatom Micrographs. Koeltz Scientific Books.
Lange-Bertalot, H. (2000-2013). Diatoms of Europe ‑ Diatoms
of the European inland waters and comparable habitats.
Konigstein: Koeltz Scientific Books.
Lavoie, I., Hamilton, P. B., Wang, Y. K., Dillon, P. J., & Campeau,
S. (2009). A comparison of stream bioassessment in Québec
(Canada) using six European and North American diatom-
based indices. Nova Hedwigia, 135, 37–56.
Larras, F., Coulaud, R., Gautreau, E., Billoir, E., Rosebery, J., &
Usseglio-Polatera, P. (2017). Assessing anthropogenic pres-
sures on streams: A random forest approach based on benthic
diatom communities. Science of the Total Environment, 586,
1101–1112.
Lecointe, C., Coste, M., & Prygiel, J. (1993). “Omnidia”: Soft-
ware for taxonomy, calculation of diatom indices and inven-
tories management. Hydrobiologia, 269, 509–513.
Mondy, C. P., Villeneuve, B., Archaimbault, V., & Usseglio-
Polatera, P. (2012). A new macroinvertebrate-based mul-
timetric index (I2M2) to evaluate ecological quality of
French wadeable streams fulfilling the WFD demands: A
taxonomical and trait approach. Ecological Indicators, 18,
452–467.
Morin, S., Valade, D., Tison-Rosebery, J., Bertrin, V., Cellamare,
M., & Dutartre, A. (2010). Utilisation du phytobenthos
pour la bioindication en plans d’eau : Etat de l’art des
méthodes disponibles et test de métriques sur les plans
d’eau aquitains. Rapport scientifique, Irstea.
Poikane, S., Kelly, M., & Cantonati, M. (2016). Benthic algal
assessment of ecological status in European lakes and
rivers: Challenges and opportunities. Science of the Total
Environment, 568, 603–613.
Potapova, M., & Hamilton, P. B. (2007). Morphological and
ecological variation within the Achnanthidium minutis‑
simum (Bacillariophyceae) species complex. Journal of
Phycology, 45, 561–575.
R Core Team. (2021). R: A language and environment for sta-
tistical computing. Vienna, Austria: R Foundation for Sta-
tistical Computing. Retrieved May, 15, 2023, from http://
www.R- proje ct. org/
Reavie, E. D., Kireta, A. R., Kingston, J. C., Sgro, G. V., Danz,
N. P., Axler, R. P., & Hollenhorst, T. P. (2008). Compari-
son of simple and multimetric diatom-based indices for
great lakes coastline disturbance. Journal of Phycology,
44(3), 787–802.
Riato, L., Leira, M., Della Bella, V., & Oberholster, P. J.
(2018). Development of a diatom-based multimetric index
for acid mine drainage impacted depressional wetlands.
Science of the Total Environment, 612, 214–222.
Rott, E., Pipp, E., Pfister, P., van Dam, H., Ortler, K., Binder, N.,
& Pall, K. (1998). Indikationslisten fur Aufwuchsalgen. In
M. Scheffer (Ed.), Ecology of shallow lakes. Chapman and
Hall.
Schneider, S. C., Hjermann, D. O., & Edvardsen, H. (2019).
Do benthic algae provide important information over and
above that provided by macrophytes and phytoplankton in
lake status assessment?–Results from a case study in Nor-
way. Limnologica, 76, 28–40.
Stenger-Kovács, C., Buczko, K., Hajnal, E., & Padisák, J. (2007).
Epiphytic, littoral diatoms as bioindicators of shallow lake
trophic status: Trophic Diatom Index for Lakes (TDIL)
developed in Hungary. Hydrobiologia, 589(1), 141–154.
Stevenson, J. (2014). Ecological assessments with algae: A
review and synthesis. Journal of Phycology, 50, 437–461.
Stevenson, R. J., Zalack, J. T., & Wolin, J. (2013). A mul-
timetric index of lake diatom condition based on sur-
face-sediment assemblages. Freshwater Science, 32(3),
1005–1025.
Venables, W. N., & Ripley, B. D. (2002). Modern applied sta‑
tistics with S (4th ed.). Springer.
Vermaat, J. E., Biberdžić, V., Braho, V., Gjoreska, B. B., Cara,
M., Dana, Z., & Schneider, S. C. (2022). Relating envi-
ronmental pressures to littoral biological water quality
indicators in Western Balkan lakes: Can we fill the largest
gaps? Science of the Total Environment, 804, 150–160.
Wetzel, R. G. (2001). Protists: Key ecosystem regulators. Bio‑
Science, 51(12), 997.
Wickham, H. (2022). Tools for working with URLs and HTTP.
R package version 1.4.4. Retrieved May, 15, 2023, from
https:// CRAN.R- proje ct. org/ packa ge= httr
Zelinka, M., & Marvan, P. (1961). Zur Priizisierung derbiol-
ogischen Klassification der Reinheit fliessender Gewàsser.
Archiv Für Hydrobiology, 57, 389–407.
Publisher’s Note Springer Nature remains neutral with regard
to jurisdictional claims in published maps and institutional
affiliations.
1202 Page 12 of 12
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”),
for small-scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are
maintained. By accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use
(“Terms”). For these purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or
a personal subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or
a personal subscription (to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the
Creative Commons license used will apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data
internally within ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking,
analysis and reporting. We will not otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of
companies unless we have your permission as detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that
Users may not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to
circumvent access control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil
liability, or is otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by
Springer Nature in writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer
Nature journal content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates
revenue, royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain.
Springer Nature journal content cannot be used for inter-library loans and librarians may not upload Springer Nature journal
content on a large scale into their, or any other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any
information or content on this website and may remove it or features or functionality at our sole discretion, at any time with or
without notice. Springer Nature may revoke this licence to you at any time and remove access to any copies of the Springer Nature
journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express
or implied with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or
warranties imposed by law, including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be
licensed from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other
manner not expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com