Aquatic Botany 113 (2014) 8389

Contents lists available at ScienceDirect

Aquatic Botany
j our nal homepage: www. el sevi er . com/ l ocat e/ aquabot
Flooding tolerance and horizontal expansion of wetland plants:
Facilitation by oating mats?
Bert Hidding
a,
, Judith M. Sarneel
a,b
, Elisabeth S. Bakker
a
a
Department of Aquatic Ecology, Netherlands Institute of Ecology (NIOO-KNAW), PO Box 50, 6700 AB Wageningen, The Netherlands
b
Landscape Ecology Group, Department of Ecology and Environmental Sciences, Ume University, SE 901 87 Ume, Sweden
a r t i c l e i n f o
Article history:
Received 24 April 2013
Received in revised form5 November 2013
Accepted 7 November 2013
Available online 6 December 2013
Keywords:
Water level uctuations
Facilitation
Floating mats
Stratiotes aloides
Riparian vegetation
Hydrosere succession
a b s t r a c t
Water level uctuations (WLF) can be important disturbances promoting the diversity of riparian plant
communities, but are currently absent from many managed aquatic ecosystems. A lack of WLF is thought
to reduce plant diversity and hamper hydrosere succession. However, a positive impact of WLF on plant
diversity may crucially depend on nutrient availability and the presence of a potential ecosystem engi-
neer, the oating plant Stratiotes aloides, that may provide structural support to riparian plants. We tested
the interactive effects of 40cmooding, presence of S. aloides and sediment nutrient availability (N and
P) on growth and horizontal expansion of eight wetland plant species in a 10 week experiment. Seven
out of eight species showed a signicant elongation response to ooding. Compared to stagnant water
levels, ooding in combination with high nutrient availability decreased horizontal expansion in two
short species and increased it in two tall species, whereas ooding decreased horizontal expansion in
two other short species under both nutrient levels. Inthis 10 week experiment, we observed no effect of S.
aloides on the measured plant parameters. This experiment shows short-term negative effects of ooding
on most of the short species. On the long-term, we hypothesize that improvements in water quality and
seedling recruitment due to drawdown may result in net positive effects of WLF in the riparian zone,
but as the species that were rare in the eld happened to be short, care should be taken to maintain rare
species when allowing more WLF.
2013 Elsevier B.V. All rights reserved.
1. Introduction
Disturbances are thought to potentially enhance species diver-
sity in communities because at a moderate level of disturbance
species with different responses to such disturbances may coexist
(Connell, 1978; Roxburgh et al., 2004). Recruitment of subordinate
species can be facilitated by disturbances (Bakker and Olff, 2003;
Hidding et al., 2010b) while dominants can persist. Disturbances
may be generated biotically, for instance by disease, bioturbation
or herbivory, or they may be generated abiotically, through re,
drought or ooding. In (semi-) aquatic habitats water level uctu-
ations (WLF) may formsuch type of disturbance and therefore the
diversity of wetland vegetation may be enhanced through intra-
annual variation in water levels (Riis and Hawes, 2002; Van Geest
et al., 2005a; Jansson et al., 2005). Lack of water level uctuations
may under some circumstances promote hypoxia of the water col-
umn and of the sediment (Bunch et al., 2010). Water drawdown

Corresponding author. Present address: Leibniz Institute of Freshwater Ecology

and Inland Fisheries, Mggelseedamm301, 12587 Berlin, Germany.
Tel.: +49 030 641 81 5.
E-mail addresses: abhidding@gmail.com, hidding@igb-berlin.de (B. Hidding).
may provide oxygen to the sediment and promote the establish-
ment of submerged aquatic vegetation (Van Geest et al., 2007). In
addition, the disturbance dynamics related to WLF may favor coex-
istence betweencompeting wetlandspecies (Bonis et al., 1995; Van
Eck et al., 2004) and create windows of opportunity for the recruit-
ment fromseeds (Casanova and Brock, 2000). The natural variation
in water level, which often exhibits a seasonal pattern with low
water levels in summer and high water level in winter may pro-
mote lake-ward expansion of fringing vegetation, as established
frompaleorecords (Korhola, 1992).
However, WLF may have negative consequences for wetland
plant diversity as well (Zohary and Ostrovsky, 2011). The ip-side
of positive drawdown effects is the possibility of ooding in spring
or summer. Such scenario can be common with excessive precipi-
tation in spring and/or when lakes are inuenced by water levels in
rivers (Van Geest et al., 2005b). Under such a scenario, plants may
experience CO
2
and O
2
shortage (e.g. Mommer and Visser, 2005;
Perata et al., 2011) and a changed light climate, in particular when
the ooding water is turbid. Plants growing in frequently ooded
areas have evolved mechanisms of ooding mitigation such as
shoot elongation or quiescence (Voesenek et al., 2006), which ulti-
mately affects their distributionalong ooding gradients (Mommer
et al., 2007) and subsequently community composition.
0304-3770/$ see front matter 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.aquabot.2013.11.003
84 B. Hidding et al. / Aquatic Botany 113 (2014) 8389
In many managed aquatic systems, including those in the
Netherlands, water levels are xed at a target level and strictly
controlled by sluice drainage, such that WLF type disturbances are
absent (Coops andHosper, 2002). Together witheutrophication, the
lowfrequency and intensity of WLF has been hypothesized to be a
cause of the lack of expansion of riparian vegetation and reduced
biodiversity (Lamers et al., 2002; Sarneel and Soons, 2012; Van der
Putten, 1997). In particular, oating mat vegetation has become
rare, although it was once a common component of peat lakes in
the Netherlands (Lamers et al., 2002; Sarneel et al., 2011). This
highly valued habitat type, considered to be an important stage in
hydrosere succession(successionof aquatic habitats towardterres-
trial systems, sensuTutin(1941)) has nowfrequentlybeenreplaced
by less diverse, tall reed and Thypha vegetation.
Stratiotes aloides L., suspected to be a vital component in the
formation of oating mats in peat lakes (Sarneel et al., 2011) has
particularly declined (Smolders et al., 2003). The plant has been
associatedwithhighdiversityof wetlandplants (Sugier et al., 2010),
and may be an important constituent of oating mats because the
sturdy leaves in S. aloides mats may facilitate horizontal expan-
sion of marsh plants by providing physical support. Facilitation
has been shown to be important under adverse conditions in salt
marshes (Bertness and Hacker, 1994), and more recently also in
submerged macrophytes (Le Bagousse-Pinguet et al., 2012). Sim-
ilarly, for emergent wetland plants facilitation by oating plants
may be an important factor in the expansion of helophytes, as was
shown in Lake Victoria (Azza et al., 2006). In this sense, oating
plants may be ecosystemengineers (sensu Jones et al., 1994). In the
Netherlands, a lack of dense S. aloides beds has been hypothesized
to contribute to decreased horizontal expansion of wetland plants
inpeatlands (Lamers et al., 2002; Beltmanet al., 2008; Sarneel et al.,
2011). Facilitation through physical support from oating plant
species suchas S. aloides may mitigate the potential negative effects
of ooding.
Here, we experimentallytest the resistance of bothtall andshort
emergent wetland species from fen systems to ooding, with one
treatment consisting of submergence and subsequent drawdown
and a control treatment with a stable water level but waterlogged
sediments. This treatment was performed at two nutrient levels
(high and low). We also test whether the presence of S. aloides may
mitigate negative effects of ooding and may facilitate horizontal
expansion of these plants. To this end we performed an experi-
ment using 12 experimental ponds and measured horizontal, and
vertical growth and biomass development during 10 weeks. We
hypothesized (1) that tall plant species (Typha, Phragmites) would
be better at resisting water level uctuations than short plants, as
tall plants may continue to photosynthesize effectively andprovide
the root systemwith oxygen at high water levels. We hypothesized
(2) nutrient availability to positively affect biomass mitigating the
negative effects of the ooding period while stimulating growth
during drawdown. We hypothesized that (3) the presence of S.
aloides would mitigate negative ooding effects on horizontally
expanding plants as they facilitate their expansion by offering
physical support. We measured the plant response to the treat-
ments in total biomass, aboveground biomass, root shoot ratio and
horizontal and vertical expansions.
2. Methods
2.1. Study system
Peat lakes in the Netherlands are mostly the result of peat exca-
vations that date back to the middle ages (Lamers et al., 2002). Peat
lake vegetation is characterized by extensive reed elds and the
submerged vegetation, if at all present, is dominated by Potamoge-
ton species. The drier areas that represent later successional stages,
are characterized by alder forest. In such systems, oating mat veg-
etation used to be rather common until the 1970s (Beltman et al.,
2008). However, nowadays these vegetation types have become
rare. In attempting to meet Natura 2000 and Water Framework
Directive requirements, water managers in the Netherlands aimat
reinstalling exible water levels in a range of wetlands in order to
promote natural disturbance regimes.
2.2. Field survey
To obtain an idea about the abundance of the focal plant species
that potentially colonize oating mats, ten wetlands in the west
of the Netherlands that were candidate for WLF type management
were surveyed (Appendix A). To this end, depending on wetland
size, three to six transects with a surface area of 200m
2
were
established in mid-August to mid-September 2010. Length was
50m along the bank and width 4m across the slope from dry to
wet conditions. In ditches however, that have narrower fringes,
transects were 100m long and 2m wide across the slope. At each
transect, percent cover of our focal species (Table 1) was estimated
per species following the Dutch Water Framework Directive stan-
dards (Bijkerket al., 2010) whichis derivedfromtheBraunBlanquet
method (Barkman et al., 1964) (Table 1).
2.3. Experimental design
We tested the effect of ooding, nutrient availability and
of S. aloides presence on plant performance parameters of a
set of eight wetland species that occur in oating mats. To
this end, we set up a large full factorial mesocosm experiment
using twelve experimental ponds (5m5m) at the Loenderveen
experimental pond facility owned by the water board Waternet
(52

12

41

N, 5

18

E). Water was taken froma nearby lake that

contained dephosphatized water (water properties after introduc-
tion in the ponds: NNH
4
+
=0.09mg/L, NNO
3

=0.03mg/L and
PPO
4
3
=0.01mg/L). Alkalinity was 3.5meq and the pH varied
between 8.0 and 8.9. Such pH is high but not uncommon in
our focal peat lakes (Smolders et al., 2012). Based on these val-
ues we estimated CO
2
concentrations to uctuate between 0.01
and 0.09mM (Stumm and Morgan, 1996). Hence, underwater
Table 1
Percentage presence of the eight focal species in ten wetland systems in the west of the Netherlands in areas, transects and mean cover respectively.
Latin name Code name Height (cm)
b
Perc. areas Perc. transects Mean cover if present Max. cover
Calla palustris L. CalPal 15 10 1 1 1
Menyanthes trifoliata L. MenTri
a
20 20 3 <1 1
Comarumpalustre (L.) ComPal
a
30 30 4 <1 15
Berula erecta (Huds.) Coville BerEre 48 90 33 <1 515
Cicuta virosa L. CicVir 83 70 19 1 1520
Ranunculus lingua L. RanLin 85 40 10 <1 1525
Typha angustifolia L. TypAng 150 80 34 1 5075
Phragmites australis (Cav.) Steud. PhrAus 250 100 73 515 75100
a
Red list status in the Netherlands according to Van der Meijden, 2005.
b
Data fromthe LEDA trait base. Median value was taken if more than one record was present.
B. Hidding et al. / Aquatic Botany 113 (2014) 8389 85
photosynthesis of the focal plants may have been signicant at
the lower pH range (Maberly and Madsen, 1998). Adjacent ponds
were paired resulting in six pairs, with one randomly chosen pond
receiving a xed water level and the other a ooding treatment.
Each of the twelve ponds was divided in two using chicken wire
fence. One side was randomly selected to receive 75 S. aloides
rosettes that were collected in a pond in Amsterdam(52

23

10

N,
4

56

59

E) between May 23 and May 29, 2012. 75 Rosettes (cover-

ing approximately 3540% of the pond surface) provided S. aloides
on all sides of the experimental pots without obstructing growth
and light interception above the pots. Inside the S. aloides treat-
ments eight plant species (obtained from commercial growers,
Table 1) were planted in square pots (18cm18cm18cm) and
were lled with 5 L sand between May 10 and 16, 2012. All plants
had been grown from vegetative cuttings (Berula erecta, Cicuta
virosa, Menyanthes trifoliata, Calla palustris, Comarum palustre) or
originated from seedlings of at least a year old (Phragmites aus-
tralis, Typha angustifolia, Ranunculus lingua). Pots were assigned to
a nutrient treatment and one out of two harvest dates resulting
in four pots per species per pond side. Sand was used as a sub-
strate rather than peat or clay in order to conveniently manipulate
nutrient supplybymeans of slowreleasefertilizer addition. Organic
substrates would have inhibited or supplemented release of N and
P making control of nutrient availability difcult. Low nutrient
treatments received 1.7g slow release osmocote (52gm
2
, 5.8%
NO
3

, 8.2% NH
4
+
, 14% P
2
O
5

, 14% K
2
O) whereas high nutrient pots
received 4 times as much (6.7g). The high nutrient treatment cor-
responded with optimal growth conditions found for submerged
macrophytes (Potamogeton sp.) using slowrelease fertilizers (Elger
et al., 2007; Hidding et al., 2010a; Bakker et al., 2013). Although no
direct measurements of pore water nutrient concentrations were
made, the highest nutrient level yields a cumulative release of
0.29gl
1
PO4
2

, 0.16gl
1
NO
3

and 0.14gl
1
NH
4
+
in the inter-
stitial water after ve weeks, assuming a porosity of 0.5 and 65%
cumulative release as established for sandy substrates under eld
capacity by Prasad and Woods (1971). Ammonium is potentially
toxic to wetlandplants, but the givenconcentrations wouldat most
be marginally toxic (Clarke and Baldwin, 2002) and would only be
reached under the conservative assumption of absence of uptake
by plant roots, microbes and sediment leakage. Pots were placed
in ponds randomly, under the condition that high nutrient and low
nutrient pots were alternated to achieve a checkerboard pattern
in terms of nutrient levels, so that each pond side had 32 pots. S.
aloides rosettes and focal plants were left to acclimatize until June
11th, until when the water level was kept at the level of the pot
rims. On this date, week 0, the largest horizontal expansion of the
plants and the largest height were recorded. Also in week 0, the
water level was raised for the rst time in the ooding treatment.
With 10cm increments each week, the water level was increased
up to 40cm above pot rim level in week 3 at July 2nd. In week 4,
the +40cm was maintained and height and width of plants was
again recorded in week 5 (July 17). Of all plants 50% was harvested
on July 17 by randomly selecting one pot of each treatment com-
bination. Plant parts were rinsed and separated in above-ground
and below-ground parts. Plant fragments were dried at 70

C for
72h and thereafter weighed to the nearest 0.01g. Also starting
in week 5 the water level was again allowed to drop with 10cm
decrements each week until the rim level was reached in week
8. In week 9 the water level was kept constant and in week 10,
height and width of plants was measured and the nal harvest
was done (August 21). Estimates of S. aloides surface cover were
made on July 3rd and 31st by counting rosettes and measuring the
diameter of 5 haphazard S. aloides rosettes. A repeated measures
ANOVA with water level treatment as an explanatory variable was
applied to check whether there were treatment related differences
in cover.
2.4. Statistical analyses
The effect of S. aloides, water level treatment and nutrients was
tested on aboveground biomass, total biomass, shoot to root ratio,
height and width of plants. Mixed effects models were used to ana-
lyze the split-plot design of the experiment using the nlme library
(Pinheiro et al., 2011) in R. Randomlevels in the model were pond
pairs (n=6), ponds andS. aloides treatment. Inmodels onheight and
width measures, also a random factor was included for pot iden-
tity, whereas in the biomass data, pots were treatment replications
and were hence not included as random variable. Fixed factors in
the models were species, S. aloides treatment, water level, nutri-
ent level, harvest time (measurement date for width and height
measures) and interactions between these factors. Normal distri-
bution of model residuals was assessed visually using QQ plots. If
the residuals deviated strongly fromnormality, response data were
transformed to their natural logarithm or square-root. Variance
homogeneity was veried using a Levenes test. When variance
turned out heterogeneous at the level of species, heterogeneous
variance structures were tted using the weights parameter of the
model (Zuur et al., 2009). ANOVAusingF-statistics was thenapplied
to the selected models and higher order interactions were omitted
when insignicant at =0.05. If a variable and its interactions had
no signicant effect, only the maineffect was retainedinthe model.
Afterwards, a multiple comparison procedure was followed using
likelihood ratio tests (Crawley, 2007). To this end we tted a full
model including the highest interaction order. The highest order of
interaction was redened as a newfactor where each combination
of factor levels became a separate factor level. Contrasts were cal-
culated for the effects within species within time units, since we
were primarily interested in between-treatment differences over
time. Multiple comparisons were only performed for signicant
interactions (Crawley, 2007).
3. Results
3.1. Field survey
Calla palustris was the rarest species in the ten surveyed wet-
lands, with presence only in one wetland, in one transect and
with a very low abundance (Table 1). M. trifoliata was found in
two wetlands, three transects and also at very low abundance. C.
palustre was found in three wetlands and in four transects. Both
M. trifoliata and C. palustre have a red list status. T. angustifolia,
P. australis and B. erecta were found in the majority of sites. B.
erecta, although common, was never dominant with the highest
percent cover reaching 15%. T. angustifolia and P. australis on the
other hand were locally dominant, reaching cover percentages of
5075% and 75100% respectively. There was a positive relation-
ship between abundance and attainable plant height (based on the
LEDA traitbase, Kleyer et al., 2008): in a regression with attain-
able height as predictor value and percentage of transects with
focal species occurrences as a response, taller species were more
common (F
1,6
=30.75, p=0.001, R
2
=0.84).
3.2. Pond experiment
The water remained highly transparent throughout the experi-
ment withonlylowgrowingbenthic algae at pot sediment surfaces.
After ve weeks in which the water level increased, seven out
of eight species showed shoot elongation in response to ooding,
except Berula erecta (Fig. 1). Instead, B. erecta was the only species
to respond to increased nutrient levels by increasing plant height
during ooding. After 10 weeks, when the water level was back to
its initial height in the ooding treatment, C. palustris, M. trifoliata
86 B. Hidding et al. / Aquatic Botany 113 (2014) 8389
0
20
40
60
80
a) CalPal
WL WL
b) MenTri
WL WL
c) ComPal
WL
d) BerEre
nut WL
Jun Jul Aug
0
50
100
150
e) CicVir
WL
Jun Jul Aug
stable
stable +
flooding
flooding +
f) RanLin
WL
Jun Jul Aug
g) TypAng
WL
Jun Jul Aug
h) PhrAus
WL WL
M
a
x
i
m
u
m

h
e
i
g
h
t

(
c
m
)
Fig. 1. Effects of ooding and nutrients on the maximumheight of plant species in centimeters in a pond experiment. Error bars indicate standard errors of the mean (n=6).
The model on plant height only had signicant three-way interactions, therefore post hoc procedures only tested for the signicance of the main treatments within species
within time unit. WL denotes a signicant water level effect (ooding vs. stable), and NUT denotes a signicant nutrient effect.
Table 2
Analysis of variance on the effects of a exible water level, Stratiotes aloides presence, and nutrient levels on growth parameters of eight plant species in a mesocosm
experiment. Data were log or square-root transformed and a heterogeneous error structure was tted to the effect of species to account for non-homogeneity of variances.
Interactions involving Stratiotes were removed as these effects were never signicant.
Maximumplant height (

) Maximumplant width (log) Aboveground biomass (

) Total biomass (

)
df F p df F p df F p df F p
Time (T) 2,989 13.07 <0.001
***
2,975 4.57 0.011
*
1,674 3.02 0.083 1,674 3.45 0.064
Water level (Wl) 1,721 0.13 0.724 1,721 0.01 0.919 1,5 0.63 0.465 1,5 0.69 0.444
Nutrients (Nut) 1,721 4.56 0.033
*
1,721 7.60 0.006
**
1,674 5.80 0.016
*
1,674 5.99 0.015
*
Species (Sp) 7,721 75.68 <0.001
***
7,721 8.17 <0.001
***
7,674 16.43 <0.001
***
7,674 16.58 <0.001
***
Stratiotes (Str) 1,721 0.53 0.466 1,721 0.00 0.958 1,11 0.04 0.843 1,11 0.00 0.950
T Wl 2,989 13.22 <0.001
***
2,975 10.18 <0.001
***
1,674 7.57 0.006
**
1,674 8.82 0.003
**
T Nut 2,989 1.90 0.150 2,975 3.72 0.025
*
1,674 1.28 0.259 1,674 1.90 0.169
T Sp 14,989 29.26 <0.001
***
14,975 5.32 <0.001
***
7,674 3.50 0.001
**
7,674 4.05 <0.001
***
Wl Nut 1,721 0.09 0.761 1,721 0.10 0.749 1,674 1.12 0.289 1,674 1.78 0.183
Wl Sp 7,721 0.88 0.520 7,721 2.43 0.018
*
7,674 1.45 0.182 7,674 2.35 0.022
*
Nut Sp 7,721 2.98 0.004
**
7,721 2.82 0.007
**
7,674 1.57 0.141 7,674 2.86 0.006
**
T Wl Nut 2,989 0.59 0.556 2,975 1.77 0.171 1,674 4.50 0.034
*
1,674 2.63 0.105
T Wl Sp 14,989 10.46 <0.001
***
14,975 2.99 <0.001
***
7,674 2.45 0.017
*
7,674 2.09 0.042
*
T Nut Sp 14,989 2.38 0.003
**
14,975 3.33 <0.001
***
7,674 1.80 0.084 7,674 0.56 0.787
Wl Nut Sp 7,721 1.68 0.111 7,721 1.72 0.102 7,674 2.84 0.006
**
7,674 2.40 0.020
*
T Wl Nut Sp 14,975 2.39 0.003
**
Stars indicate signicant main effects and interactions.
*
p<0.05.
**
p<0.01.
***
p<0.001.
and B. erecta were higher in the stable water level compared to the
plants in the ooding treatment. No signicant differences were
found for the other species and between nutrient and Stratiotes
treatments. Themaximumplant height responseover timetowater
level andnutrient treatments variedwithspecies, hence there were
signicant three-way interactions Time species water level and
time species nutrients whereas four-way interactions were not
signicant (Fig. 1andTable 2), meaning that the plant response was
dependent upon the interaction between water level and nutri-
ent availability. There was a contrast in the response to the water
level treatment between on the one hand the creeping growth
forms like C. palustris, C. palustre, M. trifoliata and B. erecta and on
the other hand the taller species R. lingua, C. virosa T. angustifolia,
and P. australis. The rst group showed inhibition of horizon-
tal expansion under ooding, for 2 species when grown under
high nutrient availability (Fig. 2). In the second group, there was
either no evidence of horizontal expansion or a positive effect:
under a combination of high nutrient levels and ooding C. virosa
and P. australis exhibited most horizontal expansion (Fig. 2). The
width of the plants did not differ signicantly between S. aloides
treatments. Aboveground biomass and total biomass of plants
showed similar patterns. Threeway-interactions were signicant
B. Hidding et al. / Aquatic Botany 113 (2014) 8389 87
0
20
40
60
80
a) CalPal
a
b
ab
a
ab
a
b
c
b) MenTri
a
a
b
b
a
b
c
c
c) ComPal
a
ab
bc
c
a
a
ab
b
d) BerEre
ab
b
c
a
a
c
b
ab
Jun Jul Aug
0
20
40
60
80
e) CicVir
a
b
ab
ab
a
a
a
b
Jun Jul Aug
f) RanLin
NS
ab
ab
a
b
Jun Jul Aug
stable
stable +
flooding
flooding +
g) TypAng
NS NS
Jun Jul Aug
h) PhrAus
NS
a
a
a
b
M
a
x
i
m
u
m

h
o
r
i
z
o
n
t
a
l

e
x
p
a
n
s
i
o
n

(
c
m
)
Fig. 2. Effects of ooding (WL) and nutrients on maximumwidth of plant species in centimeters in a pond experiment. Error bars indicate standard errors of the mean (n=6).
Because in the maximumwidth model, the four-way interaction was signicant, signicant differences are indicated using letters, juxtaposed in the same fashion as the data
points. Different letters indicate signicantly different plant width among treatments at a sampling date.
WL nut
0
10
20
30
40
a) CalPal WL b) MenTri
stable
stable +
flooding
flooding +
c) ComPal WL nut d) BerEre
Jul Aug
0
10
20
30
40
e) CicVir
Jul Aug
f) RanLin
Jul Aug
0
50
100
150
g) TypAng nut
Jul Aug
0
10
20
30
40
h) PhrAus
M
e
a
n

t
o
t
a
l

b
i
o
m
a
s
s

(
g

d
w
)
Fig. 3. Effects of ooding and nutrients on the mean total biomass of 8 plant species in a pond experiment. Error bars indicate standard errors (n=6). If an effect was signicant
in the post hoc test, here nutrients or ooding within species within time, it is indicated by WL (water level) or nut (nutrient level).
with oodingnutrients species and time oodingspecies
(Fig. 3). At the rst sampling time in July, there was no signicant
effect of any of the treatments on total plant biomass. However,
ve weeks later, C. palustris, M. trifoliata and B. erecta had a sig-
nicantly lower biomass in the ooding treatment compared to
the stable treatment. B. erecta, C. palustris and P. australis exhib-
ited a positive effect of nutrient availability on biomass. S. aloides
did not affect biomass in any of the species. In the analysis on
rootshoot ratios a signicant effect was found of the interaction
water level nutrients species (Appendix B, Table B1). Apost hoc
88 B. Hidding et al. / Aquatic Botany 113 (2014) 8389
analysis was impossible because the full model would not t due to
convergence issues. Hence, treatment effects could not be detected
at the species level.
S. aloides cover declined signicantly over time (F
1,16
=6.22,
p=0.024), from40%4 (SE) to 32%+7 for ooded ponds and from
34%4 to 19%3 for stable water levels. S. aloides tended to per-
formbetter in the ooded ponds, because the treatment effect was
marginally signicant (F
1,16
=3.77, p=0.070). This trend is in line
with eld observation showing that S. aloides prefers deeper water
(Sarneel et al., 2011; Veen et al., 2013). Whereas S. aloides cover
declined since the start of the experiment, rosettes of the species
were still present on all sides of the pots within the S. aloides treat-
ment. The S. aloides-treatment did not have any signicant effect
on any of the measured parameters, not in interaction terms, nor
in main effect terms.
4. Discussion
In a 10-week mesocosm experiment, we studied the effect of
ooding, nutrient availability and the presence of a potential struc-
tural support providing ecosystem engineer on wetland plants
associated with oating mat formation. Our focus was on vertical
and horizontal expansion as well as on plant biomass. Flooding and
subsequent drawdown inhibited the horizontal expansion of a set
of key emergent macrophytes: C. palustris, M. trifoliata, C. palustre
and B. erecta. Of these species, M. trifoliata and C. palustre are vul-
nerable species in fen systems and C. palustris is relatively rare. The
same effect is to a lesser extent observable in the total biomass
of these species (except C. palustre). This indicates that if permis-
sive water level management allows for ooding events in spring
or early summer, negative effects on target oating mat species
may be observed. Plant species that already dominate the fringes
of fen systems such as T. angustifolia and P. australis did not appear
negatively affected by ooding. R. lingua also remained relatively
unaffected under ooding. So, if these short term effects can be
extrapolated to community composition, ooding may favor tall
species over creeping growth forms in wetland vegetation. The
short statured species risk being completely submerged, and thus
experience all the related stress effects. A tall growth form may
help plants overcome ooding events without having to depend on
theinefciencyof submergedphotosynthesis (Mommer andVisser,
2005), gas transport and on shoot elongation.
We foundnoeffect of thepresenceof S. aloides onanyof theplant
parameters. So on the short term S. aloides rosettes do not appear
to mitigate the impact of ooding, not on focal plant biomass, and
neither on horizontal expansion. Hence, if there is a role for struc-
tural support to horizontal expansion of marsh plants by S. aloides
as suggested by Sarneel et al. (2011), then this may play a role on
larger time scales, requiring long-termexperiments. Also, the effect
of S. aloides offering structural support may depend on the density
of the Stratiotes rosettes and the density in our experiment may
have been too low, as in the eld Stratiotes may sometimes cover
100% of the water surface (up to 35 rosettes per m
2
; JS, personal
observation). Facilitation among species can be strongly density
dependent (Bruno et al., 2003; Le Bagousse-Pinguet et al., 2012).
It is currently unknown whether facilitation is important for hor-
izontal expansion and whether this process is density-dependent
for S. aloides. The observed decline in S. aloides cover during the
experiment may have been due to the lowavailability of nutrients
fromthe water column, whereas S. aloides grows better at high(er)
nutrient availability (Veen et al., 2013). In our experiment S. aloides
tended to be more abundant in the ooding treatment. In line with
this, eld experiments have shown that S. aloides plants survive
and growbetter in relatively deep water (Sarneel et al., 2011; Veen
et al., 2013).
Our study shows that water level uctuations in the form of
ooding may inhibit the colonization of open water by some of
the studied marsh plants, provided that no oating mat has yet
formed. Subsequent drawdown did not mitigate the stress impact
of ooding. Our results are valid in the absence of oating mat for-
mation, which is the prevailing condition in most Dutch peat lakes
(Lamers et al., 2002). We are aware that in our study certain nec-
essary experimental conditions may not be typical for peat lakes,
such as the sandy sediment and the use of slow release fertilizers.
However, ooding may hamper growth irrespective of sediment
type (Mommer and Visser, 2005). In addition, the results are in
line with a study comparing lakes in Finland that showed that
increased spring ooding may decrease hydrosere succession in
boreal lakes (PartanenandLuoto, 2006). InanAustralianstudy WLF
were shown to mitigate the effects of nutrient addition: water level
uctuations hampered growth of a set of emergent macrophytes
from river systems under nutrient rich conditions (Deegan et al.,
2012). The current experiment only considered ooding effects
on the short term. On the long term WLF may have a positive
impact on riparian communities because exible water levels may
improve aquatic chemistry. Fixed water levels require the inow
of allochthonous river water, which may notably boost sulphate
concentrations. These inturnmay leadtothe formationof toxic sul-
de and in addition cause internal eutrophication (Smolders et al.,
2003). Together with the increase in seedling establishment, long
term effects of WLF on biodiversity, oating mat formation, and
eventually hydrosere succession may very well be positive.
Our study shows that in order to promote the horizontal expan-
sion of desired vulnerable marsh species and hence an increase
of biodiversity of marsh vegetation, restoring exible water level
management may possibly pose a risk to the rare species with
creeping growth forms, whereas it may promote common species
if ooding events indeed occur in the growth season. This com-
monness relates to the height of the plants: rare species were
short, abundant species were tall. This interacted with nutrient
availability: two of thetall species showedhigher horizontal expan-
sion under higher nutrient availability whereas two short plant
species showed highest horizontal expansion under stable water
levels. Many of the Dutch fen systems contain nutrient-rich sed-
iment (Lamers et al., 2002), where sediment nutrient availability
is already negatively related with the occurrence of red-list ripar-
ian plant species (Geurts et al., 2009), because under high fertility
tall species replace creeping species. Such processes may thus be
further enhanced by ooding in the growing season. We suspect
that the timing and magnitude of drawdown and ooding in rela-
tion to plant growth is important, as they represent frequency
and intensity of the dominant disturbance regime. Moreover, sea-
sonal timing interacts with plant phenology and controls whether
seedling establishment is positively affected or that growing plants
are damaged. Hence, the balance between positive and negative
effects of water level uctuations may depend strongly on the real-
ized water level uctuations and their timing under a more natural
water regime.
Acknowledgements
We thank Waternet for access to the experimental site; we
are indebted to Waternet, Natuurmonumenten, Staatsbosbeheer,
Landschap Noord-Holland and the Amsterdam-North city counsel
for access to the eld sites. Nico Bouwman, Koos Swart, Jose van
Paassen, Casper van Leeuwen, Michaela Brehm, Dennis Waasdorp,
Monica Estebaranz-Bustamante and Edwin van den Berg are very
much acknowledged for practical assistance. We thank the anony-
mous reviewers for valuable comments on an earlier version of this
manuscript. This is publication 5556 of the NIOO-KNAW.
B. Hidding et al. / Aquatic Botany 113 (2014) 8389 89
Appendix A. Supplementary data
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.aquabot.
2013.11.003.
References
Azza, N., Denny, P., van de Koppel, J., Kansiime, F., 2006. Floating mats: their occur-
rence and inuence on shoreline distribution of emergent vegetation. Freshw.
Biol. 51, 12861297.
Bakker, E.S., Dobrescu, I., Straile, D., Holmgren, M., 2013. Testing the stress gra-
dient hypothesis in herbivore communities: facilitation peaks at intermediate
nutrient levels. Ecology 94, 17761784.
Bakker, E.S., Olff, H., 2003. Impact of different-sized herbivores on recruitment
opportunities for subordinate herbs in grasslands. J. Veg. Sci. 14, 465474.
Barkman, J.J., Doing, H., Segal, S., 1964. Kritische Bemerkungen und Vorschlge zur
quantitativen Vegetationsanalyse. Acta Bot. Neerlandica 13, 394419.
Beltman, B., Weijs, W., Sarneel, J.M., 2008. Werken de KRW-en Natura 2000-criteria
voor sloten en veenplassen? H
2
O 41, 2527.
Bertness, M.D., Hacker, S.D., 1994. Physical stress and positive associations among
marsh plants. Am. Nat. 144, 363372.
Bijkerk, R., Beers, M., Bonhof, G., Brans, B., Buskens, R., Coops, H., Dam, H., van, Fock-
ens, K., Kampen, J., van Maanen, B., van, Mertens, A., Moeleker, M., Nieuwenhuis,
R., Pilon, J., Pot, R., Spier, J., Swarte, M., van Tongeren, O., Torenbeek, R., Vermaat,
J., Wagenvoort, A., Wilhelm, M., de Wit, M., 2010. Handboek Hydrobiologie (No.
2010-28). Stichting Toegepast Onderzoek Waterbeheer, Amersfoort.
Bonis, A., Lepart, J., Grillas, P., 1995. Seed bank dynamics and coexistence of annual
macrophytes in a temporary and variable habitat. Oikos 74, 8192.
Bruno, J.F., Stachowicz, J.J., Bertness, M.D., 2003. Inclusion of facilitation into ecolog-
ical theory. Trends Ecol. Evol. 18, 119125.
Bunch, A.J., Allen, M.S., Gwinn, D.C., 2010. Spatial and temporal hypoxia dynamics
in dense emergent macrophytes in a Florida lake. Wetlands 30, 429435.
Casanova, M.T., Brock, M.A., 2000. How do depth, duration and frequency of ood-
ing inuence the establishment of wetland plant communities? Plant Ecol. 147,
237250.
Clarke, E., Baldwin, A.H., 2002. Responses of wetland plants to ammonia and water
level. Ecol. Eng. 18, 257264.
Connell, J.H., 1978. Diversity in tropical rain forests and coral reefs. Science 199,
13021310.
Coops, H., Hosper, S.H., 2002. Water-level management as a tool for the restoration
of shallowlakes in the Netherlands. Lake Reserv. Manage. 18, 293298.
Crawley, M.J., 2007. The R Book. John Wiley & Sons, Ltd., Chichester, West Sussex,
UK.
Deegan, B.M., White, S.D., Ganf, G.G., 2012. Nutrients and water level uctuations:
a study of three aquatic plants. River Res. Appl. 28, 359368.
Elger, A., de Boer, T., Hanley, M.E., 2007. Invertebrate herbivory during the
regeneration phase: experiments with a freshwater angiosperm. J. Ecol. 95,
106114.
Geurts, J.J.M., Sarneel, J.M., Willers, B.J.C., Roelofs, J.G.M., Verhoeven, J.T.A., Lamers,
L.P.M., 2009. Interacting effects of sulphate pollution, sulphide toxicity and
eutrophication on vegetation development in fens: a mesocosm experiment.
Environ. Pollut. 157, 20722081.
Hidding, B., Brederveld, R.J., Nolet, B.A., 2010a. How a bottom-dweller beats the
canopy: inhibition of an aquatic weed (Potamogeton pectinatus) by macroalgae
(Chara spp.). Freshw. Biol. 55, 17581768.
Hidding, B., Nolet, B.A., de Boer, T., de Vries, P.P., Klaassen, M., 2010b. Above- and
below-ground vertebrate herbivory may each favour a different subordinate
species in an aquatic plant community. Oecologia 162, 199208.
Jansson, R., Zinko, U., Merritt, D.M., Nilsson, C., 2005. Hydrochory increases riparian
plant species richness: a comparison between a free-owing and a regulated
river. J. Ecol. 93, 10941103.
Jones, C.G., Lawton, J.H., Shachak, M., 1994. Organisms as ecosystemengineers. Oikos
69, 373386.
Kleyer, M., Bekker, R.M., Knevel, I.C., Bakker, J.P., Thompson, K., Sonnenschein, M.,
Poschlod, P., Van Groenendael, J.M., Klime s, L., Klime sov, J., Klotz, S., Rusch,
G.M., Hermy, M., Adriaens, D., Boedeltje, G., Bossuyt, B., Dannemann, A., Endels,
P., Gtzenberger, L., Hodgson, J.G., Jackel, A.-K., Khn, I., Kunzmann, D., Ozinga,
W.A., Rmermann, C., Stadler, M., Schlegelmilch, J., Steendam, H.J., Tackenberg,
O., Wilmann, B., Cornelissen, J.H.C., Eriksson, O., Garnier, E., Peco, B., 2008. The
LEDATraitbase: a database of life-history traits of the Northwest European ora.
J. Ecol. 96, 12661274.
Korhola, A., 1992. The early holocene hydrosere in a small acid hill-top basin studied
using crustacean sedimentary remains. J. Paleolimnol. 7, 122.
Lamers, L.P.M., Smolders, A.J.P., Roelofs, J.G.M., 2002. The restoration of fens in the
Netherlands. Hydrobiologia 478, 107130.
Le Bagousse-Pinguet, Y., Liancourt, P., Gross, N., Straile, D., 2012. Indirect facilitation
promotes macrophytesurvival andgrowthinfreshwater ecosystems threatened
by eutrophication. J. Ecol. 100, 530538.
Maberly, S.C., Madsen, T.V., 1998. Afnity for CO
2
in relation to the ability of fresh-
water macrophytes to use HCO
3

. Funct. Ecol. 12, 99106.

Mommer, L., Visser, E.J.W., 2005. Underwater photosynthesis in ooded terrestrial
plants: a matter of leaf plasticity. Ann. Bot. 96, 581589.
Mommer, L., Wolters-Arts, M., Andersen, C., Visser, E.J.W., Pedersen, O., 2007.
Submergence-induced leaf acclimation in terrestrial species varying in ooding
tolerance. New. Phytol. 176, 337345.
Partanen, S., Luoto, M., 2006. Environmental determinants of littoral paludication
in boreal lakes. Limnol. Ecol. Manage. Inland Waters 36, 98109.
Perata, P., Armstrong, W., Voesenek, L.A.C.J., 2011. Plants and ooding stress. New
Phytol. 190, 269273.
Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D., 2011. R development core team. NLME:
Linear Nonlinear Mixed Effects Models.
Prasad, M., Woods, M.J., 1971. Release characteristics of nitrogen fertilizers in peat
and sand. J. Agric. Food Chem. 19, 9698.
Riis, T., Hawes, I., 2002. Relationships between water level uctuations and veg-
etation diversity in shallow water of New Zealand lakes. Aquat. Bot. 74,
133148.
Roxburgh, S.H., Shea, K., Wilson, J.B., 2004. The intermediate disturbance hypoth-
esis: patch dynamics and mechanisms of species coexistence. Ecology 85,
359371.
Sarneel, J.M., Soons, M.B., 2012. Post-dispersal probability of germination and estab-
lishment on the shorelines of slow-owing or stagnant water bodies. J. Veg. Sci.
23, 517525.
Sarneel, J.M., Soons, M.B., Geurts, J.J.M., Beltman, B., Verhoeven, J.T.A., 2011. Multiple
effects of land-use changes impede the colonization of open water in fen ponds.
J. Veg. Sci. 22, 551563.
Smolders, A.J.P., Lamers, L.P.M., Hartog, C., den Roelofs, J.G.M., 2003. Mechanisms
involved in the decline of Stratiotes aloides L. in The Netherlands: sulphate as a
key variable. Hydrobiologia 506509, 603610.
Smolders, A.J.P., Loermans, J., Lamers, L.P.M., 2012. Effecten van exibel peilbeheer
op bodemprocessen en waterkwaliteit (Report no. 2012.51) Flexibel peil: van
denken naar doen. B-WARE Research Centre, Nijmegen.
Stumm, W., Morgan, J.J., 1996. Aquatic chemistry: chemical equilibria and rates in
natural waters, environmental science. John Wiley & Sons, Inc., NewYork.
Sugier, P., Lorens, B., Chmiel, S., Turczy nski, M., 2010. The inuence of Ceratophyl-
lum demersum L. and Stratiotes aloides L. on richness and diversity of aquatic
vegetation in the lakes of mid-eastern Poland. Hydrobiologia 656, 4353.
Tutin, T.G., 1941. The hydrosere and current concepts of the climax. J. Ecol. 29, 268.
Vander Meijden, R., 2005. Heukels oravanNederland, 23rded. Wolters-Noordhoff,
Groningen.
Van der Putten, W.H., 1997. Die-back of Phragmites australis in European wetlands:
an overview of the European Research Programme on Reed Die-back and Pro-
gression (19931994). Aquat. Bot. 59, 263275.
Van Eck, W.H.J.M., Van De Steeg, H.M., Blom, C.W.P.M., De Kroon, H., 2004. Is
tolerance to summer ooding correlated with distribution patterns in river
oodplains? A comparative study of 20 terrestrial grassland species. Oikos 107,
393405.
Van Geest, G.J., Coops, H., Roijackers, R.M.M., Buijse, A.D., Scheffer, M., 2005a. Suc-
cession of aquatic vegetation driven by reduced water-level uctuations in
oodplain lakes. J. Appl. Ecol. 42, 251260.
Van Geest, G.J., Coops, H., Scheffer, M., van Nes, E.H., 2007. Long transients near the
ghost of a stable state in eutrophic shallow lakes with uctuating water levels.
Ecosystems 10, 3646.
Van Geest, G.J., Wolters, H., Roozen, F.C.J.M., Coops, H., Roijackers, R.M.M., Buijse,
A.D., Scheffer, M., 2005b. Water-level uctuations affect macrophyte richness
in oodplain lakes. Hydrobiologia 539, 239248.
Veen, G.F., Ciska Sarneel, J.M., Ravensbergen, L., Huig, N., van Paassen, J., Rip, W.,
Bakker, E.S., 2013. Aquatic grazers reduce the establishment and growth of
riparian plants along an environmental gradient. Freshw. Biol. 58, 17941803.
Voesenek, L.A.C.J., Colmer, T.D., Pierik, R., Millenaar, F.F., Peeters, A.J.M., 2006. How
plants cope with complete submergence. NewPhytol. 170, 213226.
Zohary, T., Ostrovsky, I., 2011. Do water level uctuations matter? Inland Waters 1,
4759.
Zuur, A.F., Ieno, E.N., Walker, N., Saveliev, A.A., Smith, G.M., 2009. Mixed Effects
Models and Extensions in Ecology with R. Springer, NewYork, NY.