Content uploaded by Érika Maria Neif
Author content
All content in this area was uploaded by Érika Maria Neif on Oct 03, 2017
Content may be subject to copyright.
ORIGINAL ARTICLE
Responses of benthic algal communities and their traits to
experimental changes in fine sediments, nutrients and flow
Erika M. Neif
1
|
Daniel Graeber
2
|
Liliana Rodrigues
1
|
Simon Rosenhøj-Leth
2
|
Tinna M. Jensen
2
|
Peter Wiberg-Larsen
2
|
Frank Landkildehus
2
|
Tenna Riis
2
|
Annette Baattrup-Pedersen
2
1
Programa de P
os-graduacß
~
ao em Ecologia
de Ambientes Aqu
aticos Continentais –
PEA –Nup
elia, Universidade Estadual de
Maring
a, Maring
a, PR, Brazil
2
Department of Bioscience, Aarhus
University, Aarhus, Denmark
Correspondence
Daniel Graeber, Department of Bioscience
Aarhus University, Silkeborg, Denmark.
Email: dgr@bios.au.dk
Funding information
Seventh Framework Programme, Grant/
Award Number: 603378; Coordenacß
~
ao de
Aperfeicßoamento de Pessoal de N
ıvel
Superior
Summary
1. Lowland stream ecosystems are subjected to multiple anthropogenic stressors,
usually nutrient enrichment in combination with sedimentation of fine particles
and low flow periods in summer. Here, we investigated the temporal develop-
ment of the benthic algae community in response to these three stressors and
linkages to the trait characteristics of the community to explore the mechanisms
responsible for stress-induced community changes.
2. We investigated the response of benthic algae species composition, traits (life
forms, cell size categories), biovolume and chlorophyll a(Chl-a) concentration to
low flow in combination with nutrient enrichment and fine sedimentation in
twelve large outdoor stream flumes (12 m long) resembling small streams in size
and habitat characteristics. The experiment consisted of two phases: a normal-
flow phase followed by a low-flow phase (90% current velocity reduction), each
spanning 4 weeks. We applied a eutrophication scenario (mean increases of
1.14–5.48 mg N/L and 0.01–0.06 mg P/L in the flumes for dissolved inorganic
nitrogen and phosphate respectively) throughout the experiment. Under low
flow, we supplemented this with a fine sedimentation scenario (>90% stream
bed cover). We took samples once in the normal-flow phase and every week
during the low-flow phase.
3. We observed strong responses in the benthic algae community to sudden
changes in low flow and fine sedimentation, mediating rapid species turnover
with a decreased algal biovolume and increased abundance of large, motile spe-
cies. However, we did not observe any pronounced responses to nutrient enrich-
ment. In contrast to the observations for other variables, we found a continuous
increase in Chl-aconcentration during low flow. This was likely due to continu-
ous fine sedimentation during this phase, reducing light availability which proba-
bly resulted in an increase of cell-level Chl-aconcentration in response to light
limitation and lower rates of light-induced Chl-adegradation.
4. The rapid response of the benthic algal community to the applied stressors sug-
gests that even short periods of major stressor exposure may significantly affect
benthic algae in lowland systems. We suggest that short-term stress events may
have cascading effects on several important ecosystem processes given the
importance of benthic algae for the productivity of these systems.
Accepted: 22 May 2017
DOI: 10.1111/fwb.12965
Freshwater Biology. 2017;1–12. wileyonlinelibrary.com/journal/fwb ©2017 John Wiley & Sons Ltd
|
1
KEYWORDS
biological traits, fine sediment, low-flow event, nutrients, periphyton
1
|
INTRODUCTION
Human activities strongly affect natural stream ecosystems through
changes in a number of key environmental factors (Hering et al.,
2015). Concurrent changes in these factors are defined as multiple
stresses if the range of normal variability is exceeded and if the fac-
tors affect the structural and functional characteristics of the ecosys-
tems (Auerbach, 1981). Multiple stressors can have additive,
synergistic, or antagonistic effects relative to a single stressor, which
can be dependent on the stressor intensity and identity (Allan et al.,
2013). This necessitates the evaluation of stressor combinations to
quantify the total ecosystem impact (e.g. Matthaei, Piggott, & Town-
send, 2010; Piggott, Townsend, & Matthaei, 2015).
In the coastal climates of northern Europe, climate change-
mediated alterations in flow regimes in lowland streams are pro-
nounced and periods with critically low flows are becoming more
and more common during summer (Arnell, 1999). Furthermore,
worldwide stream ecosystems are affected by high water abstrac-
tion, high loads of fine sediments and elevated nutrient concentra-
tions due to agricultural activities (Graeber, Pusch, Lorenz, &
Brauns, 2013; Hille et al., 2014; Piggott, Salis, Lear, Townsend, &
Matthaei, 2015; Townsend, Uhlmann, & Matthaei, 2008; Wagen-
hoff, Lange, Townsend, & Matthaei, 2013). Therefore, multiple
stress in lowland agricultural streams is a widespread phenomenon
that strongly affects biological stream communities, including ben-
thic algae.
Benthic algae are highly successful primary producers in streams
and constitute an important component of the structure and func-
tion of these ecosystems (Finlay, Maberly, & Cooper, 1997). Benthic
algae are a part of biofilms that develop on all types of surfaces (e.g.
sand, rocks, aquatic vegetation). Biofilms themselves consist of
water, polysaccharide excretions and other organic and inorganic
materials (Sutherland, 2001) and contribute to stream ecosystem
functions by assimilating and transforming nutrients such as ammo-
nium and nitrate (Baker, de Guzman, & Ostermiller, 2009; Levi et al.,
2015), thereby altering the nutrient availability for other biota in
adjacent and downstream ecosystems.
Development of the benthic algae community is controlled by a
complex array of factors and interactions (Biggs & Thomsen, 1995),
with irradiance, nutrient availability, physical disturbance and grazing
being the most important (Biggs, 1996). Extreme low flow, nutrient
availability and fine sediment deposition strongly impact these fac-
tors, and are stressors often occurring in human-impacted lowland
streams (Dewson, James, & Death, 2007; Townsend et al., 2008).
Clear responses of the biomass of benthic algae to low flow,
nutrient availability and fine sediment deposition have been reported
in the literature. Algal biomass may increase with enhanced nutrient
availability in the water column and reduced flow disturbance which
occurs, for instance, at low-flow conditions (Biggs, 1996; Ferreira &
Chauvet, 2011; Piggott, Lange, Townsend, & Matthaei, 2012). Fur-
thermore, algal cell density has been shown to increase with increas-
ing nutrient availability as well as with enhanced fine sediment
deposition (Piggott et al., 2012).
Stressors occurring in agricultural streams affect not only the
biomass but also the species composition of benthic algae. Benthic
algae taxa vary in size, growth form and strength of adherence to
the substrate (Passy, 2007), and the prevailing taxa reflect environ-
mental characteristics, in particular the level of disturbance and
resource supply (Biggs, Stevenson, & Lowe, 1998; McCormick &
O’Dell, 1996; Passy & Larson, 2011; Piggott, Salis et al., 2015). Fur-
thermore, trait composition of benthic algae reacts strongly in terms
of growth form to combinations of fine sediments and nutrient
enrichment (Piggott et al., 2012; Piggott, Salis et al., 2015). For
example, Piggott et al. (2012) found that high sediment deposition
reduced the number of adnate and prostrate species as they became
covered by sediment and likely suffered from lack of light in the
photosynthesis process. Concurrently, the number of medium and
large species, as well as motile species increased with enhanced sed-
iment deposition, reflecting the ability of larger motile species to
stay on the sediment surface, thus ensuring light availability for pri-
mary production (Piggott et al., 2012; Piggott, Salis et al., 2012;
2015). Furthermore, motile diatoms and diatoms with high profiles,
as well as small and medium algal species were shown to profit from
nutrient enrichment (Passy, 2007; Piggott, Salis et al., 2015). Finally,
firmly adhered species are more resistant to natural disturbances
(e.g. high current velocities and herbivory) compared with those that
are loosely attached but likely do not profit from low current veloci-
ties with low disturbances (Biggs et al., 1998; Passy, 2007; Schneck
& Melo, 2012). Altogether, if water velocity reduction co-occurs with
fine sedimentation and nutrient enrichment in streams, the species
composition probably changes towards lower abundances of smaller,
firmly adhered adnate/prostrate species and higher abundances of
larger motile species.
Knowledge on the interactions of low flow, fine sediment and
nutrient availability on algal biomass or other algal quantity and com-
position variables is scarce since no study tried to investigate the
combined effects of these stressors in an experiment. This is impor-
tant as low flow, fine sedimentation and nutrient availability often
occur together in human-impacted lowland streams and benthic algal
responses can only be fully understood if the combined effects of
these stressors are investigated. Furthermore, we expect changes in
biofilm biomass and community composition to occur within days to
2
|
NEIF ET AL.
weeks after the stressors are induced. For example, we think it likely
that rapid deposition of sediment on a benthic biofilm affects the
primary production of the benthic algae and thereby algal biomass
within a few days due to increased mortality. In contrast, changes in
species composition may occur over weeks rather than days because
it is related to the benthic algae life cycle and growth rates. How-
ever, these assumptions remain to be tested since, to our knowl-
edge, no literature exists on the temporal development of the
benthic algae response to the combined effects of low flow, fine
sedimentation, and nutrient enrichment.
In this study, the responses of the benthic algal community to
low flow, fine sediment load and eutrophication were investigated.
The experiment was conducted in twelve replicated large outdoor
flumes with a size comparable to small lowland streams. We
combined two nutrient levels during a normal-flow phase and con-
tinued the same nutrient levels and two additional fine sedimenta-
tion scenarios during a low-flow phase. Each of the two phases
lasted 4 weeks and samples were taken at the end of the normal-
flow phase and in each week of the low-flow phase. We hypothe-
sised (1) that the adaptation of the benthic algae species and trait
composition to the applied stressors would occur gradually from
the start of the low-flow phase and continue throughout the
experiment, (2) that deposition of fine sediment would mediate a
fast reduction in the biomass of the community, resulting in lower
chlorophyll a (Chl-a) concentrations and biovolumes and a shift
from adnate/prostrate to motile algae species, and (3) that high
amounts of nutrients at low flow would enhance the biovolume
of species with a more erect growth, allowing them to circumvent
reduced light availability. Therefore, the high-nutrient treatment
should partly mitigate the effects of fine sediment deposition on
biomass.
2
|
MATERIAL AND METHODS
2.1
|
Stream flume design
We conducted the experiment in twelve outdoor flumes during sum-
mer 2014 in Denmark (56°40N, 9°310E) with different pre-experi-
mental and experimental phases. The flumes consisted of rectangular
12 m long, 60 cm wide and 30 cm deep channels in which various
types of inorganic sediments were introduced to create a run-riffle
sequence (please see Appendix S1 for details on the flumes and
habitats).
The stream flumes were fed with unfiltered water from a nearby
source stream. Each of the flumes was fed with stream water from
the source stream by a pump through 1,000 L plastic feeder tanks
by water pipes. The water output of the stream feeder pump was
controlled by a frequency regulator (VLT Aqua Drive FC202 7.5 kW,
Danfoss) and individual flow regulators installed at the pipes before
the feeder tanks for fine adjustment of the flow. An additional recy-
cling pump was installed at the end of each of the flumes to increase
the flow when needed (please see Appendices S1 and S2 for further
details).
2.2
|
Experimental design
The flumes were started with c. 5 L/s on 26 June 2014 for 2 weeks
to allow benthic algae colonisation and organic matter accumulation
before benthic macroinvertebrates were introduced on the 7 July
2014. For each flume, the benthic macroinvertebrates were taken
from the source stream by standardised kick sampling along a reach
of c. 500 m to construct a realistic stream environment (more details
on the kick sampling can be found in Appendix S1). This macroinver-
tebrate colonisation approach has been shown to result in a
macroinvertebrate composition similar to the one of the source
stream (Graeber et al., 2017).
On 27 July 2014, we started the normal-flow experiment phase
(32 days) with nutrient enrichment in six randomly chosen flumes
(NP treatment). Hence, six flumes were not treated with additional
nutrients. Nutrients were added with a 12-channel peristaltic pump
(BVP-Process with a 12-channel CA pump head, Ismatec, Wertheim,
Germany) from a central 600 L tank, which was refilled when empty.
We used fertilizer (SweDane NPK 21-3-10 and GrowHow NS 24-6,
DLG, Copenhagen, Denmark) to enrich with dissolved inorganic
nitrogen and phosphate to obtain a fivefold increase for dissolved
inorganic nitrogen (on average 5.5 mg N/L in the nutrient enriched
flumes compared to a background concentration of 1.1 mg N/L) and
a sevenfold increase for phosphate (on average 0.064 mg P/L com-
pared to 0.009 mg P/L) throughout the experiment (please see
Appendix S1 for further statistics on the nutrient concentrations in
the different treatments).
Thus, in the six flumes with nutrient enrichment, a high-nutrient
environment was simulated (typical for streams in catchments with
intensive agriculture), and in the six flumes without nutrient enrich-
ment, a moderate-nutrient environment was simulated (typical for
streams in catchments with extensive agriculture). The flow regime
was continued as before, hence stream water was fed into the chan-
nels and the recycling pumps were running to reach a sufficient dis-
charge. Nutrient addition was adjusted for the effects of water
recycling, so that the flumes with nutrient addition experienced a
constant nutrient concentration.
The low-flow phase was started on 28 August 2014 and lasted
for 28 days. Here, the flow was reduced to simulate a flow event
typical of lowland streams in Central European intensive agricultural
landscapes in summer (Graeber, Gelbrecht, Pusch, Anlanger, & von
Schiller, 2012; Graeber et al., 2015; Hille et al., 2014). The discharge
was lowered from, on average, 5.2 L/s during the normal-flow phase
to 1 L/s during the low-flow phase, rendering a current velocity
reduction of 83%–93% in the run habitat (further statistics on the
hydraulics in the different treatments are available in Appendices S1
and S2). The lowering of discharge was achieved by switching off
the recycling pumps and fine-adjusting the discharge with the fre-
quency regulator of the pump feeding stream water into the system.
Due to the architecture of the system, the lowering of discharge
was near-instantaneous. The nutrient-enrichment treatments of the
normal-flow phase were continued in the low-flow phase to keep
stable eutrophic and mesotrophic conditions.
NEIF ET AL.
|
3
At the start of the low-flow phase, fine sediment was added manu-
ally in six randomly chosen flumes, three without nutrient enrichment
(FS treatment) and three with nutrient enrichment (NP +FS treat-
ment). Hence, for the low-flow phase, the combination of fine sedi-
ment and nutrient treatments resulted in low flow combined with NP
alone, FS alone and the combination the NP and FS (NP +FS) treat-
ments for three channels each, plus three channels with low-flow but
without any secondary stressor. The fine sediment was pumped from
the source stream into several carboys and manually introduced into
the flumes until >90% fine sediment cover was reached. During the
whole low-flow phase, the sediment cover in the fine sediment treat-
ment was >95%. Furthermore, we observed a continuous increase in
the coverage of fine sediments during the low-flow phase in the non-
fine sediment treatments as fine sediment transported with stream
water settled within the stream flumes due to the low current veloci-
ties. However, the fine sediment cover was always significantly lower
in the non-fine sediment treatments than in the fine-sediment treat-
ments (see Appendix S1 for further details on fine sediment coverage).
Water temperature was measured every ten minutes in each of
the flumes with a temperature logger (HOBO Pendant UA-001,
Onset, USA) and was, on average, 12.1°C with a minimum of 4.7°C
and a maximum of 18.8°C. Average water temperature did not
change significantly during the low-flow phase.
2.3
|
Benthic algae
Quadratic stones (tiles, side length c. 9 cm, height c. 2.5 cm, with a
roughly textured surface) were used to investigate algal growth.
Three days after the start of the normal-flow phase, 20 stones were
introduced in two run habitats per each flume. To minimise distur-
bance of the benthic habitat by later removal of stones, we added
twice as many stones as we sampled. They were positioned along
the main flow path of each run habitat and slow flowing water or
eddies were avoided.
The density and composition of the benthic algae were analysed in
all flumes for the last week of the normal-flow phase and during the
4 weeks of the low-flow phase. We sampled two stones from two dif-
ferent run habitats on each of the five sampling occasions. Biofilm was
scrubbed carefully from the upper side of two stones from two differ-
ent run habitats for each flume and sampled, using a soft toothbrush
and stream water and combined into one sample. Scrubbed-off stones
were discarded and not re-used. The exact length and width of the
stones were measured. The sample from the stones was divided into
two subsamples for analysis of species composition and Chl-a, the lat-
ter using ethanol extraction of filter residues (Whatman glass microfi-
ber filters, GF/C, 47 mm, GE Healthcare, Brondby, Denmark)
according to Jespersen and Christoffersen (1987).
For later analysis, the benthic algae material (50 ml) was fixed
and preserved in a solution of acetic lugol (5%). Counting was con-
ducted in the laboratory in random fields (mean of 150 fields was
calculated for each sample), using an inverted microscope according
to Uterm€
ohl (1958). Based on this, the algal density was calculated
as the number of counted individuals (a unicellular organism, a
colony, a filament, or coenobium) divided by the area of the 150
fields. Algae were identified at 4009magnification to the lowest
practical taxonomic level. Identification of the species was carried
out using literature sources (Croasdale & Flint, 1986; Dillard, 1990,
1991; F€
orster, 1982; Kom
arek & Anagnostidis, 1999, 2005; Krammer
& Lange-Bertalot, 1986, 1988, 1991; Prescott, Croasdale, Vinyard, &
Bicudo, 1982). Very concentrated samples were diluted to facilitate
visualisation and counting of individuals. If samples were diluted, the
dilution factor was included in the calculation of the algal densities.
The benthic algae biovolume was calculated by multiplying the
density of each taxon by its respective volume. The cell volume was
calculated from geometric models, according to the species-specific
shape of the cells (Hillebrand et al. 1999; Sun & Liu 2003). In detail,
for an individual (a unicellular organism, a colony, a filament, or coeno-
bium), the length and number of all cells was measured and the appro-
priate geometric model was applied to calculate the cell volume.
Subsequently the sums of all cells of an individual were summed up.
This was done for 50 individuals of each taxon within each sample, or
for all individuals of taxa with 50 individuals or less.
The life forms of the algae were categorised into one of five
types based on literature (Guiry & Guiry, 2017; Passy, 2007; Passy
& Larson, 2011; Schneck, Schwarzbold, & Melo, 2011) and personal
observation: (1) adnate or prostrate (Achnanthidium,Cocconeis,Char-
acium,Coleochaete and Chamaesiphon), (2) erect or with mucilage
stalks or tubes (Cymbella,Eunotia,Fragilaria,Gomphonema,Ulnaria
and Chroococcus), (3) motile (Cymatopleura,Gyrosigma,Navicula,Nitz-
chia,Pinnularia,Pandorina,Mallomonas,Cryptomonas,Euglena,Phacus
and Trachelomonas), (4) metaphyton (Closterium,Cosmarium, Scene-
desmus, Dictyosphaerium, Pediastrum and Tetraedron) and (5) filamen-
tous (Planktolyngbya, Pseudoanabaena and Geitlerinema). We are
aware that our life-form trait classification contains growth-form
types (prostrate, erect/with mucilage stalks, metaphyton, filamen-
tous), modes of attachments (adnate) and the ability to move actively
(motile). However, our classification encompasses the typical life-
form types of benthic algae (Schneck & Melo, 2012), and should give
a realistic representation of their trait composition. Furthermore,
these were used to ensure comparability with earlier studies on mul-
tiple stress (e.g. Piggott et al., 2012; Piggott, Salis et al., 2015).
Moreover, the algae were categorised into three size classes depend-
ing on the length of algae cells: small: <30 lm, medium: 30–90 lm
and large: >90 lm (Piggott, Salis et al., 2015). Following to Piggott,
Salis et al. (2015), we considered unicellular organisms, colonies, fila-
ments and coenobium as individuals, and used the average length of
the longest axis of the individuals to define the size classes. Length
was used instead of biovolume for the size classes to maximise the
comparability of the size classes to earlier multiple-stress studies
(e.g. Piggott et al., 2012; Piggott, Salis et al., 2015).
2.4
|
Data analysis
Biovolumes were used for all tests on species and trait data. The
biovolumes were ln (x+1) transformed since the data followed a
log-normal distribution.
4
|
NEIF ET AL.
The responses of benthic algae species composition, life form
composition and size composition to the low flow, FS, NP and
FS +NP treatments were analysed by applying principal response
curves (PRC; please see Van den Brink & Ter Braak, 1999 for details
on the calculation of the PRC). The PRC model was based on the
first axis of a principal coordinate analysis using Bray–Curtis similar-
ity (Hille et al., 2014) to generate site and species scores (capscale
function, vegan package; Oksanen et al., 2015). The PRC consists of
treatment scores and species weights. The treatment scores can be
interpreted as principal response of the community to a treatment
(Van den Brink & Ter Braak, 1999). Furthermore, the species weights
allow determining the taxon-specific reaction, since higher species
weights indicate stronger responses of the species to the treatment
patterns in the PRC (Van den Brink & Ter Braak, 1999). Furthermore,
taxa with near zero species weights either show no response or a
response that is unrelated to the patterns represented in the PRC.
Moreover, the direction of the species weights determines the direc-
tion of the species response to the treatments (Van den Brink & Ter
Braak, 1999).
Within the PRCs, the six channels without nutrient addition dur-
ing the last week of the normal-flow phase were used as control.
The developments of the benthic algae species and trait composition
for the FS, NP and FS +NP treatments and the treatment without
any secondary stressor (hence neither FS nor NP) were plotted
against this control. For benthic algae species composition, species
with a score >0.1 were used in the PRC plots to reveal only the spe-
cies strongly affected by the treatments. An ANOVA-like permuta-
tion test was used to assess the significance of the PRC model
(model df =1, residual df =62, 9,999 permutations, anova.cca func-
tion, vegan package; Oksanen et al., 2015). Based on this test, all
PRCs were significant and thus significantly represented the effects
of the treatments on the benthic algae species composition, life form
type composition and size composition (F>41.2, p<.001).
A permutational multivariate ANOVA (PERMANOVA) (9,999 per-
mutations, adonis function, vegan package; Oksanen et al., 2015) was
used to assess the effects of the NP treatment on species, life form
type and size composition for the last week of the normal-flow phase.
PERMANOVAs were also used to assess the effects of the FS and NP
treatments and the interaction of these stressors on species, life form
type and size composition during the 4 weeks of the low-flow phase
(model df =3, residual df =44). Weeks were used as strata for the
permutations to take the repeated nature of the sampling into account
(PERMANOVA model: adonis(log10(species +1) ~nutrient *fs,
strata =week, distance =“bray”, permutations =9,999). For treat-
ments with significant effects according to the PERMANOVAs, it was
investigated which taxa and trait categories could be used as indica-
tors for these treatments. For this, an indicator value analysis was
used (multipatt function within the indicspecies package for R;
C
aceres & Legendre, 2009).
Furthermore, PERMANOVAs were used to assess the effect of
low flow on species, life form type and size composition. In detail, a
one-way PERMANOVA with sampling week as the main factor and
the flumes as strata was conducted using the last week of the
normal-flow phase and the first week of the low-flow phase for
each species, life form type and size composition (PERMANOVA
model: adonis(log10(species +1) ~flow.phase, method =“bray”,
strata =flumes, permutations =9,999), model df =1, residual
df =10). The channels without FS treatment were investigated sepa-
rately from the channels with FS treatment to separate the effect of
fine sedimentation from the pure low-flow phase effect. For each of
the dominant algal species (species constituting >1% of total algal
biovolume in the entire experiment), it was checked if either the FS
treatment, the NP treatment or an interaction between these
affected biovolumes by using linear mixed-effects models (lme func-
tion, nlme package; Pinheiro, Bates, Debroy, & Deepayan, 2015).
Flume identity was used as a random intercept in the linear
mixed-effects models. Biovolumes of the dominant algal species
were ln (x+1)-transformed to reach normal distribution and
variance homogeneity of the residuals (model: lme(log(y)~fine.sed *
nutrients *week, random =~1|channel, weights =varIdent
(form =~1|channel), model df =7, residual df =40).
The changes in benthic algae quantity were assessed by analys-
ing Chl-aconcentrations and total algal biovolumes. We used a one-
way analysis of variance (ANOVA, model df =1, residual df =10,
aov function, R core team 2015) to investigate whether the NP
treatment influenced either Chl-aor total biovolume in the last week
of the normal-flow phase (model: aov(log(y)~nutrients)). For the
low-flow phase, linear mixed-effects models were applied to assess
if FS or NP treatment, sampling week or an interaction between
these influenced Chl-aand total algal biovolume. The same model
formulation and data transformation as for the biovolumes of the
dominant algal species was applied to the linear mixed-effects mod-
els of Chl-aand total biovolume. Finally, paired permutation tests
based on binomial distribution (non-parametric alternative to a
paired ttest, 99,999 iterations, model df =1, residual df =10) were
used to investigate whether Chl-aconcentration and biovolume
changed in the first or last week of the low-flow phase relative to
the last week of the normal-flow phase for the channels without fine
sediment addition during the low-flow phase.
3
|
RESULTS
A total of 45 benthic algae taxa were recorded (see Table S1 for
taxon list). Of these, seven taxa dominated the community during
the experiment (Table 1).
3.1
|
Temporal development of multiple stressor
effects during the low-flow phase
After 1 week of low flow, a rapid change relative to that in the last
week of the normal-flow phase was observed in the PRCs regarding
both species composition, life form type and size composition. This
effect was enhanced for the flumes with FS treatments, as indicated
in the PRCs by the higher deviance of the FS treatment curves than
the non-FS treatment curves from the control (zero line) for the
NEIF ET AL.
|
5
weeks of the low-flow phase (Figures 1 and 2). This observation was
supported by the statistical comparison between the last week of
the normal-flow phase and the first week of the low-flow phase
where a significant change in the community of benthic algae spe-
cies occurred, both alone and in combination with the FS treatment
(PERMANOVA, F>2.7, p=.03, R²> .21). Furthermore, for the FS
treatment, the change from normal to low flow had a significant
effect on benthic algae life form type and size composition (PERMA-
NOVA, F>11.8, p=.03, R²> .54), as opposed to the treatments
without FS (PERMANOVA, F<3.7, p>.07, R²< .28).
After the rapid reaction recorded in the first week of the low-
flow phase, the species composition remained stable in the treat-
ments without FS (Figure 1). In contrast, species composition contin-
ued to change in the FS treatments and was most pronounced after
3 weeks of low flow (Figure 1). Furthermore, the life form type com-
position remained stable throughout the low-flow phase (Figure 2a),
whereas also size composition continued to change, with the largest
difference to the control after 2 weeks of low flow, except for the
NP treatment (Figure 2b).
The two variables applied to assess the quantity of the algae
community, Chl-aconcentration and algal biovolume, responded dif-
ferently to the low flow. Chl-aremained constant after 1 week of
low flow (paired permutation test, p=.09, Figure 3a), whereas bio-
volume decreased significantly after 1 week of low flow relative to
the last week of normal flow (paired permutation test, p=.03, Fig-
ure 3b). Additionally, we observed a significant effect of the sam-
pling week on the Chl-aconcentration (linear mixed-effects model,
t=4.61, p<.001) with a significantly higher Chl-aconcentrations
after 4 weeks of low flow than during normal flow (paired permuta-
tion test, p<.001, Figure 3a). In contrast, we did not observe any
significant effect of the sampling week on the biovolume of benthic
algae, which remained low during the low-flow phase (linear mixed-
effects model, t=1.12, p=.27). However, at the end of the experi-
ment, from the third to the fourth week of low flow, the biovolume
rapidly increased and was statistically similar to that observed during
the normal-flow phase (paired permutation test, p=.18, Figure 3b).
3.2
|
Effects of fine sediment during the low-flow
phase
Four of the dominant algae taxa were negatively affected by the
FS treatment (Table 1), all being within the small size class. Fur-
thermore, two motile taxa were positively affected by the FS treat-
ment (Table 1). Here, the effect of the FS treatments was stronger
than the effect of the NP treatment under low-flow conditions
(Table 2).
When looking at the traits of all taxa, those with motile life form
and large cell size (>90 lm) responded positively to the FS treat-
ments during the low-flow phase (Figure 2, Table 2). In contrast, the
small taxa (<30 lm) with adnate/prostrate form, erect/mucilage
stalks or metaphyton life forms showed a weak negative response
(Figure 2, Table 2). The medium-sized (30–90 lm) taxa showed no
response (Figure 2b). Based on the R²of the PERMANOVAs
(Table 2), the responses of benthic algae life form type and size
composition to the FS treatments were stronger than the reaction of
taxon composition to the FS treatments.
In terms of algal quantity, we found contrasting effects of the FS
treatment during the low-flow phase: No effect of the FS treatment
on Chl-awas found (linear mixed-effects model, tvalue =2.2,
p=.06) (Figure 3a), however, a significant negative effect of the FS
treatments (linear mixed-effects model, tvalue =2.6, p=.03) on
benthic algae biovolume was observed (Figure 3b).
3.3
|
Effects of nutrient enrichment
Neither taxon composition, life form type nor size composition were
affected by the NP treatment under normal flow (Figures 1, 2; PER-
MANOVA, F<0.27, R²< .23, p>.09). However, significant
TABLE 1 Dominant benthic algae taxa during the experiment, trait characteristics and reaction to stressors during the low-flow phase. Each
taxon comprises >1% of the total algal biovolume detected. The effects of the nutrient enrichment (NP) and fine sedimentation (FS)
treatments, as well as their interaction during the low-flow phase, were assessed by a linear mixed-effects model using channel identity as
random intercept
Species Life form type
a
Size class
b
Sum of
biovolume
c
Relative
contribution (%)
Treatment effect
during low flow
d
Achnanthidium minutissimum Adnate/Prostrate Small 9.1E +07 54.3 FS**
cf. Coleochaete sp. Adnate/Prostrate Small 2.7E +07 16.2 none
Cocconeis placentula Adnate/Prostrate Small 1.6E +07 9.7 FS**
Gomphonema parvulum Erect/mucilage stalks Medium 1.2E +07 7.2 FS**
Navicula cryptocephala Motile Medium 9.2E +06 5.5 +FS***
Chamaesiphon sp. Adnate/Prostrate Small 5.7E +06 3.4 FS**
Pinnularia lundii Motile Large 4.0E +06 2.4 +FS***
a
Passy, 2007, Passy and Larson 2011 and Schneck et al. 2011.
b
Piggott, Salis et al. (2015).
c
Sum of all samples (n=60) in experiment, mm
3
/cm
2
.
d
Results of linear mixed-effects model, +/indicates direction of effect, ***p<.001, **p=.001–.01.
6
|
NEIF ET AL.
responses in taxon composition to the NP treatment were recorded
during the low-flow phase, but with a low R²of .04, implying only a
minor effect (Figure 1; Table 2). Finally, there was no effect of the
NP treatment or the interaction of the NP and the FS treatment for
the life-form trait composition or size composition during the low-
flow phase (PERMANOVA, Table 2, Figure 2).
NF 1.wk LF 2.wk LF 3.wk LF 4.wk LF
−1.5 −1.0 −0.5 0.0 0.5 1.0 1.5
PRC score
●●●●
Chamaesiphon sp.
Pseudoanabaena sp.
Cocconeis placentula
Gomphonema sp.
Gomphonema parvulum
Cymbellla cistula
Achnanthidium minutissimum
Fragilaria crotonensis
Planktolyngbya sp.
Cryptomonas sp.
Mallomonas sp.
Gomphonema pumilum
Geitlerinema sp.
Trachelomonas hispida
Cosmarium regnellii
Scenedesmus denticulatus
Eunotia minor
Nitzschia lineares
Pediastrum tetras
cf. Coleochaete sp.
Phacus sp.
Scenedesmus armatus
Euglena sp.
Pinnularia mesolepta
Gyrosigma acuminatum
Fragilaria capucina
Nitzschia palea
Navicula cryptocephala
Pinnularia lundii
●
Control
No sec. stress
FS
NP
FS+NP
FIGURE 1 Principal response curves
(PRCs) for benthic algae taxon
composition. Symbols represent the mean
PRC scores for each sampling date and
treatment (n=3). Control =channels
without NP during normal-flow phase
(n=6); No sec. stress =no secondary
stressor (FS or NP); FS =fine sediment;
NP =Nutrients; FS +NP =fine
sediment +nutrients, NF =normal-flow
phase, LF =low-flow phase. For clarity,
only the species with a weight less than
0.1 or larger than +0.1 are displayed (Van
den Brink & Ter Braak, 1998)
−0.5 0.0 0.5 1.0
●
●
●●
Filamentous
Erect/mucilage stalks
Adnate/Prostrate
Metaphyton
Motile
●
Control
No sec. stress
FS
NP
FS+NP
(a)
NF 1.wk LF 2.wk LF 3.wk LF 4.wk LF
−0.5 0.0 0.5 1.0
PRC score
●
●
●●
Small
Medium
Large
(b)
FIGURE 2 Principal response curves
(PRCs) for life forms (a) and size classes (b)
of benthic algae. Symbols represent the
mean PRC scores for each sampling date
and treatment (n=3). Control =channels
without NP during normal-flow phase
(n=6); No sec. stress =no secondary
stressor (FS or NP); FS =fine sediment;
NP =Nutrients; FS +NP =fine
sediment +nutrients, NF =normal-flow
phase, LF =low-flow phase
NEIF ET AL.
|
7
Similarly, only minor effects of the NP treatment on algae quan-
tity were recorded. Chl-aresponded positively to the NP treatment
in the last week of the normal-flow phase (ANOVA, F=8.1, p=.02,
Figure 3a), but during the low-flow phase no effect of the NP treat-
ment on Chl-acould be traced (linear mixed-effects model,
t=0.29, p=.77). Algal biovolume did not react to the NP treat-
ment in either the normal-flow phase (ANOVA, F=0.11, p=.75) or
the low-flow phase (linear mixed-effects model, t=0.11, p=.91).
4
|
DISCUSSION
We hypothesised that a gradual response of benthic algae to multi-
ple stressors would occur during the low-flow phase, but already
after the first week of low flow we observed a significant reduction
in benthic algal biovolume, strong effects on the dominant algae taxa
and shifts from small to large taxa and from prostrate/adnate to
motile taxa. The effects were strongest in combination with the slug
deposition of fine sediments. Our findings support those of previous
studies showing similar effects but, at least in our flume systems,
demonstrate that changes can occur at a much faster rate than pre-
viously documented (Magbanua et al., 2013; Piggott et al., 2012;
Piggott, Salis et al., 2015). Therefore, even a short, low-flow phase
will likely affect the abundance and composition of benthic algae
communities in small lowland streams, especially when co-occurring
with deposition of fine sediments.
For the interpretation of the results on the flow change, the
near-instantaneous change of the discharge from normal to low flow
within our experiment needs to be considered. A slower change in
discharge might have led to a slower change in algal compositional
patterns, but based on the fast response of the benthic algae to the
abrupt change in flow, we assume that benthic algae community
would have reacted with only a short delay. For the flumes with fine
sediment (FS) treatment, a large algal mortality due to the acute
major disturbance by the fine-sediment slug addition is implied by
low algal biovolume at the start of the low-flow event, followed by
increases in algal biovolume. Therefore, we think it is likely that the
major shifts in species composition and trait composition due to the
FS treatment are a combined effect of community-level adaption
and re-colonisation from the stream water. In contrast, the slower
fine sedimentation in the flumes without FS treatment (see figure in
Appendix S1 for details on fine sedimentation development) treat-
ment also shifted the community composition of the algal species,
but significantly less strong than with FS treatment.
Following the initial fast response, the algal life form composition
remained largely stable, while the responses of Chl-a, biovolume,
taxon composition and size composition continued to change. Con-
sequently, evaluating the responses of benthic algae to multiple or
single stressor effects at only one point in time seems insufficient.
To avoid misinterpretation of multiple stressor effects, we strongly
recommend observing them repeatedly in time within future studies.
As described above, the effect of low flow on the benthic algae
taxon or trait composition was exacerbated when fine sediment was
added to the flumes. Furthermore, six of the seven dominant algal
taxa responded significantly to the fine sediment treatment. The
strong effects observed in the FS treatments support our second
hypothesis. In addition, despite different taxa in our study compared
to those of earlier studies from New Zealand (Magbanua et al.,
2013; Piggott et al., 2012; Piggott, Salis et al., 2015), it was the
same traits that responded to fine sedimentation. Likely due to their
low profile, the taxa with adnate/prostrate life form had difficulties
in resisting fine sedimentation, whereas motile taxa profited from
fine sedimentation, likely due to their ability to move across and
between fine substratum particles to reach the light (Passy, 2007).
The consistency in the observed response of the trait characteristics
of the benthic algae community to fine sedimentation across studies
suggests that benthic algae traits provide an excellent tool to predict
effects of fine sedimentation. This is also theoretically supported by
the habitat templet theory (Southwood, 1977, 1988), which suggests
that species with the same suitable traits are always selected by the
same habitat characteristics, independent of the species in the com-
munity but with the prerequisite that species with suitable traits are
available to colonise the habitat in question. Consequently, traits
rather than species composition are likely a better predictor for the
environmental condition (Poff, 1997) as recently shown also for
macrophytes, fish and macroinvertebrates in European lowland
streams (G€
othe et al., 2016). Furthermore, a change in trait composi-
tion due to changes in environmental condition may have further
effects on the ecological functions of the benthic algae community
Chl−a (mg/m²)
(a)
Week
Biovolume
(mm³ cells/cm²) x 1,000
NF 1.wk LF 2.wk LF 3.wk LF 4.wk LF
02468 12
0 4000 8000
No sec. stress
FS
NP
FS+NP
(b)
FIGURE 3 Benthic algae Chl-a(a) and total biovolume (b).
Symbols represent means for each sampling date and treatment
(n=3). No sec. stress =no secondary stressor (FS or NP); FS =fine
sediment; NP =nutrients; FS +NP =fine sediment +nutrients,
NF =normal-flow phase, LF =low-flow phase
8
|
NEIF ET AL.
within streams, e.g. algal phosphorus uptake, carbon fixation and vul-
nerability to herbivory (Steinman, Mulholland, & Hill, 1992).
In contrast to the total biovolume of the benthic algal commu-
nity, Chl-awas not affected by fine sedimentation or nutrient enrich-
ment during the low-flow phase. In fact, the Chl-aconcentration
continued to increase over time in all treatments, whereas the algal
biovolume remained stable. For Chl-a, this opposes our second
hypothesis in which we expected a strong negative effect of fine
sedimentation. We believe this effect to be a consequence of inten-
sified light limitation in the benthic zone, caused by the continued
deposition of fine sediments on the stream bottom during low flow,
which likely acted in two ways: (1) By increasing the concentration
of Chl-awithin their cells, the benthic algae likely compensated for
the reduced light availability (Baulch, Turner, Findlay, Vinebrooke, &
Donahue, 2009; Falkowski & LaRoche, 1991; Hill & Dimick, 2002).
(2) Light has, furthermore, been shown to be an important factor for
Chl-adegradation (Carpenter, Elser, & Elser, 1986). Hence, fine sedi-
mentation may also have reduced Chl-adegradation indirectly by
reducing the availability of light, resulting in a build-up of Chl-ain
the benthic biofilms. However, since we lack measurements of Chl-a
degradation products, we cannot further disentangle the responsible
mechanisms. Alternative explanations, such as water temperature
differences between the flow phases and species turnover in the
low-flow phase, cannot explain this effect because: (1) the water
temperature regime was stable during both the normal-flow phase
and the low-flow phase, and (2) taxon composition did not exhibit a
similar gradual change over time, indicating that the change in Chl-a
was most likely not related to species turnover (Felip & Catalan,
2000).
In our third hypothesis, we proposed a mitigating effect of nutri-
ent enrichment on the adverse effects of fine sediment. However,
we did not find strong support for our hypothesis in the community
statistics of species or trait composition, as no interactions were
observed between the NP and FS treatment in the low-flow phase.
In fact, our results imply that the addition of nutrients does not
change the benthic algae response to fine sedimentation or low flow.
Earlier studies have clearly shown nutrient enrichment to affect ben-
thic algae trait composition (e.g. Passy, 2007; Piggott et al., 2012),
and one reason for the different result obtained in our study could
be the low nutrient enrichment in our experiment (approximately
fivefold for dissolved inorganic nitrogen and sevenfold for phos-
phate). In comparison, earlier multiple stress studies used an approxi-
mate inorganic nitrogen and phosphate increase of up to 10-fold
and 45-fold, respectively, and found interacting effects of nutrient
enrichment on the effects of fine sediment and low flow (Piggott,
Salis et al., 2015; Wagenhoff et al., 2013). Furthermore, Passy
(2007) demonstrated that the response of diatom traits largely
occurred between 0 and 2 mg N/L nitrate, above which the
response was negligible and which was lower than the dissolved
inorganic nitrogen concentration (on average 5.49 mg N/L) in our
nutrient-enriched flumes. This notion is, moreover, supported by the
subsidy-stress reaction of the motile guild to nutrient addition,
where motile species benefited from low nutrient addition levels, but
not from high nutrient addition levels (Wagenhoff et al., 2013). The
absence of a mitigating impact of nutrient enrichment on the effect
of fine sedimentation during the low-flow phase can furthermore
reflect that the delivery of nutrients from the water column to the
benthic algae was reduced due to thicker boundary layers under low
flow. We changed the average current velocity close to the stream
bottom from 9.9 cm/s during the normal-flow phase to 0.5 cm/s
during the low-flow phase (see Appendix S1 for further details). This
is exactly the level of change of current velocity for which a limita-
tion of phosphorus and nitrogen transport to the benthic algae due
to boundary layer effects was reported (Borchardt, 1996).
The different reactions of benthic algal traits to nutrient enrich-
ment between studies raise a more general question of how far the
level of one stressor modulates the response of other stressors. It
has recently been discussed how different synergistic and antagonis-
tic multiple stressor effects can be defined and categorised relative
to a control which itself has a certain stress level (Piggott, Townsend
et al., 2015), and to this we can add that also the initial stress level
modulates the kind and direction of multiple stressor responses. We
suggest that this modulating effect likely depends on the type of
stressor.
Here, we propose a first outline of how we understand the
“dose–response relationship”between the strength of the benthic
algae trait response and stressor event intensity for typical stressors
in lowland streams (Figure 4). First, we propose that nutrient
TABLE 2 Results of the permutational multivariate analyses of variance (PERMANOVA) for the effects of nutrient enrichment with nitrogen
and phosphorus (NP) and fine sediment addition (FS) on benthic algae taxon composition, life form composition and size composition during
the low-flow phase. The taxa names and trait categories in parentheses are indicators (p<.05, indicator value analysis) for the significant
treatments. Fragilaria capucina appeared as indicator species in the FS +NP treatment and was hence attributed as indicator species to both
treatments
Treatment Species composition Life form type composition Size composition
FS F=16.8, R²= .26***
(Euglena sp., Pinnularia lundii, Navicula cryptocephala, F. capucina)
F=39.9, R²= .46***
(Motile taxa)
F=33.8, R²= .41***
(Large taxa)
NP F=2.5, R²= .04*
(Gomphonema parvulum, Gomphonema sp., F. capucina)
F=1.8, R²= 0.02
n.s.
F=2.6, R²= .03
n.s.
Interaction term F=1.0, R²= .02
n.s.
F=0.1, R²= .002
n.s.
F=1.5, R²= .02
n.s.
***p<.001, *p=.01–.05, n.s. =not significant.
NEIF ET AL.
|
9
enrichment events affect benthic algae trait composition up to only
a certain level. As already explained above, this is supported for
nitrate by our study as well as the results obtained in other studies
(Passy, 2007; Piggott et al., 2012; Wagenhoff et al., 2013). Thus,
when we added more nutrients to the relatively high background
levels of our control, only minor effects appeared. Hence, other
stressor effects (e.g. fine sedimentation) on trait responses should be
modulated only to a low degree by changes in nutrient concentra-
tions, if these changes in nutrient concentration occur at a high level
(e.g. above c. 2 mg N/L nitrate for diatom traits, Passy, 2007) (Fig-
ure 4). We suggest that the same principle applies to increasing tem-
peratures; thus, in a study by Piggott, Salis et al. (2015) a small
temperature rise was observed to affect some algal traits, whereas a
further increase of temperature only had negligible effects on ben-
thic algae traits (Piggott, Salis et al., 2015). This implies that a ben-
thic algal community already affected by high water temperatures is
likely to be less affected by a further water temperature increase
hence effects of other stressors become more important. However,
at extreme water temperatures, we again expect strong effects on
benthic algae trait composition, as conditions at such extreme condi-
tions should select few tolerant benthic algal species. Furthermore,
very high nutrient concentrations can inhibit benthic algae species
growth and photosynthesis (Azov & Goldman, 1982; Hoellein, Tank,
Kelly, & Rosi-Marshall, 2010; Tank & Dodds, 2003), also likely select-
ing algae tolerant to such conditions. Several studies have shown
strong effects of fine sediment on benthic algae traits (our study;
Piggott et al., 2012; Piggott, Salis et al., 2015), intermediate levels (c.
80% fine sediment cover) creating substantial changes but with only
modest effects of an additional rise (c. 90%; Piggott, Salis et al.,
2015). Hence, in systems with high fine sediment stress levels, the
benthic algal community is not likely to respond significantly to a
further addition. For current velocity, we propose an exponential
relationship between current velocity-related stress and the effect
on benthic algae traits (Figure 4) at both high-flow and low-flow
stress. Thus, high-flow events are expected to result in a strong
selection of pioneer communities dominated by small, low-profile
species with fast colonisation and high growth rates (Biggs et al.,
1998; Passy, 2007). In our experiment, we also found strong effects
of low flow on the composition of benthic algae traits, with large
and motile species showing the strongest responses. At the levels
between the two extremes, benthic algae trait composition should
be less strongly affected by changes in flow velocity (Passy, 2007).
Therefore, changes in other stressors might overrule the effects of
minor changes in current velocity, its impact increasing to dominance
at more severe flow events. Therefore, in future multiple stress
experimental studies on benthic algae or other biological elements,
stress background levels and different levels of stress should be con-
sidered for the design of the experiment, formulation of the
hypotheses and analyses of the results.
ACKNOWLEDGMENTS
This study was supported by the MARS project (Managing Aquatic
ecosystems and water Resources under multiple Stress) funded
under the 7th EU Framework Programme, Theme 6 (Environment
including Climate Change), Contract No.: 603378 (http://www.mars-
project.edu).
Erika Maria Neif would like to thank the Coordenacß
~
ao
de Aperfeicßoamento de Pessoal de N
ıvel Superior (CAPES) for pro-
viding her PhD scholarship. We thank Anne Mette Poulsen for valu-
able editorial comments. We furthermore thank Sandra Hille and
Stefan Lorenz, who helped to start the experiments, Marlene Venø
Skjærbæk and Dorte Nedergaard, who conducted the nutrient mea-
surements in the laboratory and Anne Mette Poulsen for linguistic
assistance.
REFERENCES
Allan, J. D., McIntyre, P. B., Smith, S. D. P., Halpern, B. S., Boyer, G. L.,
Buchsbaum, A. ... Steinman, A. D. (2013). Joint analysis of stressors
and ecosystem services to enhance restoration effectiveness. Pro-
ceedings of the National Academy of Sciences of the United States of
America,110, 372–377.
Arnell, N. W. (1999). The effect of climate change on hydrological
regimes in Europe: A continental perspective. Global Environmental
Change,9,5–23.
Auerbach, S. I. (1981). Ecosystem response to stress: A review of con-
cepts and approaches. In G. W. Barrett, & R. Rosenberg (Eds.), Stress
effects on natural ecosystems (pp. 29–41). New York: John Wiley &
Sons Ltd.
Azov, Y., & Goldman, J. C. (1982). Free ammonia inhibition of algal pho-
tosynthesis in intensive cultures. Applied and Environmental Microbiol-
ogy,43, 735–739.
Baker, M. A., de Guzman, G., & Ostermiller, J. D. (2009). Differences in
nitrate uptake among benthic algal assemblages in a mountain stream.
Journal of the North American Benthological Society,28,24–33.
Baulch, H. M., Turner, M. A., Findlay, D. L., Vinebrooke, R. D., & Don-
ahue, W. F. (2009). Benthic algal biomass —measurement and errors.
Canadian Journal of Fisheries and Aquatic Sciences,66, 1989–2001.
Biggs, B. J. (1996). Patterns in benthic algae in streams. In J. Stevenson,
M. Bothwell, & R. Lowe (Eds.), Algal ecology (pp. 31–56). New York,
NY: Academic Press.
Biggs, B. J. F., Stevenson, R. J., & Lowe, R. L. (1998). A habitat matrix
conceptual model for stream periphyton. Archiv f€
ur Hydrobiologie,
143,25–56.
FIGURE 4 Conceptual diagram of the relationship between the
intensity of stressor events for typical stressors of lowland streams
and the strength of the reaction of benthic algae traits. The
hypothesised effect of changes in flow velocity is assumed for low-
flow and high-flow stress
10
|
NEIF ET AL.
Biggs, B. J. F., & Thomsen, H. A. (1995). Disturbance of stream periphyton
by perturbations in shear stress: Time to structural failure and differ-
ences in Community Resistance1. Journal of Phycology,31, 233–241.
Borchardt, M. A. (1996) Nutrients. In: R. J. Stevenson, M. L. Bothwell &
R. L. Lowe (Eds.), Algal ecology: Freshwater benthic ecosystem (pp.
183–227). San Diego, USA: Academic press.
C
aceres, M. D., & Legendre, P. (2009). Associations between species and
groups of sites: Indices and statistical inference. Ecology,90, 3566–
3574.
Carpenter, S. R., Elser, M. M., & Elser, J. J. (1986). Chlorophyll produc-
tion, degradation, and sedimentation: Implications for paleolimnology.
Limnology and Oceanography,31, 112–124.
Croasdale, H., & Flint, E. A. (1986). Flora of New Zealand: Freshwater
algae, Chlorophyta, Desmids, with ecological comments on their habitats.
Wellington: V. R. Ward, Government Printer.
Dewson, Z. S., James, A. B. W., & Death, R. G. (2007). A review of the
consequences of decreased flow for instream habitat and macroin-
vertebrates. Journal of The North American Benthological Society,26,
401–415.
Dillard, G. E. (1990). Freshwater algae of the southeastern United States.
Part 3. Chrolophyceae: Zygnematales: Mesotaeniaceae and Desmidia-
ceae. J. Cramer, Berlin, Stuttgart.
Dillard, G. E. (1991). Freshwater algae of the southeastern United States.
Part 4. Chrolophyceae: Zygnematales: Desmidiaceae. J. Cramer, Ber-
lin, Stuttgart.
Falkowski, P. G., & LaRoche, J. (1991). Acclimation to spectral irradiance
in Algae. Journal of Phycology,27,8–14.
Felip, M., & Catalan, J. (2000). The relationship between phytoplankton
biovolume and chlorophyll in a deep oligotrophic lake: Decoupling in
their spatial and temporal maxima. Journal of Plankton Research,22,
91–106.
Ferreira, V., & Chauvet, E. (2011). Synergistic effects of water tempera-
ture and dissolved nutrients on litter decomposition and associated
fungi. Global Change Biology,17, 551–564.
Finlay, B. J., Maberly, S. C., & Cooper, J. I. (1997). Microbial diversity and
ecosystem function. Oikos,80, 209–213.
F€
orster, K., & Huber-Pestalozzi, G. (1982). Das Phytoplankton des
S€
ußwassers: Systematik und Biologie. Teil 8: H€
alfte 1. Conjugatophy-
ceae: Zygnematales und Desmidiales (excl. Zygnemataceae). Schwei-
zerbart’sche Verlagsbuchhandlung, Stuttgart.
G€
othe, E., Baattrup-Pedersen, A., Wiberg-Larsen, P., Graeber, D., Kris-
tensen, E. A., & Friberg, N. (2016). Environmental versus spatial con-
trol of taxonomic and trait composition of stream biota in a
European lowland region. Freshwater Biology,62, 397–413.
Graeber, D., Gelbrecht, J., Pusch, M. T., Anlanger, C., & von Schiller, D.
(2012). Agriculture has changed the amount and composition of dis-
solved organic matter in Central European headwater streams.
Science of the Total Environment,438, 435–446.
Graeber, D., Goyenola, G., Meerhoff, M., Zwirnmann, E., Ovesen, N. B.,
Glendell, M., et al. (2015). Interacting effects of climate and agricul-
ture on fluvial DOM in temperate and subtropical catchments.
Hydrology and Earth System Sciences,19, 2377–2394.
Graeber, D., Jensen, T. M., Rasmussen, J. J., Riis, T., Wiberg-Larsen, P., &
Baattrup-Pedersen, A. (2017). Multiple stress response of lowland
stream benthic macroinvertebrates depends on habitat type. Science
of the Total Environment,600, 1517–1523.
Graeber, D., Pusch, M., Lorenz, S., & Brauns, M. (2013). Cascading effects
of flow reduction on the benthic invertebrate community in a low-
land river. Hydrobiologia,717, 147–159.
Guiry, M. D., & Guiry, G. M. (2017). AlgaeBase. URL: http://www.
algaebase.org. Last accessed: 11 February 2017.
Hering, D., Carvalho, L., Argillier, C., Beklioglu, M., Borja, A., Cardoso, A.
C., Duel, H., et al. (2015). Managing aquatic ecosystems and water
resources under multiple stress —An introduction to the MARS pro-
ject. Science of the Total Environment,503–504,10–21.
Hillebrand, H., D€
urselen, C.-D., Kirschtel, D., Pollingher, U. & Zohary, T.
(1999). Biovolume Calculation for Pelagic and Benthic Microalgae.
Journal of Phycology,35, 403–424.
Hill, W. R., & Dimick, S. M. (2002). Effects of riparian leaf dynamics on
periphyton photosynthesis and light utilisation efficiency. Freshwater
Biology,47, 1245–1256.
Hille, S., Kristensen, E. A., Graeber, D., Riis, T., Jrgensen, N. K., & Baat-
trup-Pedersen, A. (2014). Fast reaction of macroinvertebrate commu-
nities to stagnation and drought in streams with contrasting nutrient
availability. Freshwater Science,33, 847–859.
Hoellein, T. J., Tank, J. L., Kelly, J. J., & Rosi-Marshall, E. J. (2010). Seasonal
variation in nutrient limitation of microbial biofilms colonizing organic
and inorganic substrata in streams. Hydrobiologia,649, 331–345.
Jespersen, A.-M., & Christoffersen, K. (1987). Measurements of chloro-
phyll-a from phytoplankton using ethanol as extraction solvent. Archiv
f€
ur Hydrobiologie,109, 445–454.
Kom
arek, J., & Anagnostidis, K. (1999). In H. Ettl, J. Gerloff, H. Heynig, &
D. Mollenhauer (Eds.), Cyanoprokaryota 1 - Chroococcales. Jena: Gus-
tav Ficher Verlag.
Kom
arek, J., & Anagnostidis, K. (2005). In H. Ettl, J. Gerloff, H. Heynig, &
D. Mollenhauer (Eds.), Cyanoprokaryota 2 - Oscillatoriales. Heidelberg:
Elsevier Spektrum Akademischer Verlag.
Krammer, K., & Lange-Bertalot, H. (1986). In H. Ettl, J. Gerloff, H. Heynig,
& D. Mollenhauer (Eds.), Bacillariophyceae: Naviculaceae. Stuttgart:
Gustav Ficher Verlag.
Krammer, K., & Lange-Bertalot, H. (1988). In H. Ettl, J. Gerloff, H. Heynig,
& D. Mollenhauer (Eds.), Bacillariophyceae: Bacillariaceae, Epithemi-
aceae, Surirellaceae. Stuttgart: Gustav Ficher Verlag.
Krammer, K., & Lange-Bertalot, H. (1991). In H. Ettl, J. Gerloff, H.
Heynig, & D. Mollenhauer (Eds.), Bacillariophyceae: Centrales,
Fragilariaceae, Eunotiaceae. Stuttgart: Gustav Ficher Verlag.
Levi, P. S., Riis, T., Alnøe, A. B., Peipoch, M., Maetzke, K., Bruus, C., et al.
(2015). Macrophyte complexity controls nutrient uptake in lowland
streams. Ecosystems,18, 914–931.
Magbanua, F. S., Townsend, C. R., Hageman, K. J., Lange, K., Lear, G.,
Lewis, G. D., & Matthaei, C. D. (2013). Understanding the combined
influence of fine sediment and glyphosate herbicide on stream peri-
phyton communities. Water Research,47, 5110–5120.
Matthaei, C. D., Piggott, J. J., & Townsend, C. R. (2010). Multiple stres-
sors in agricultural streams: Interactions among sediment addition,
nutrient enrichment and water abstraction. Journal Of Applied Ecology,
47, 639–649.
McCormick, P. V., & O’Dell, M. B. (1996). Quantifying periphyton
responses to phosphorus in the florida everglades: A Synoptic-Experi-
mental approach. Journal of the North American Benthological Society,
15, 450–468.
Oksanen, J., Blanchet, F. G., Kindt, R., Legendre, P., Minchin, P. R., & O’Hara,
R. B., Simpson, G. L., et al. (2015). vegan: Community Ecology Package.
Passy, S. I. (2007). Diatom ecological guilds display distinct and pre-
dictable behavior along nutrient and disturbance gradients in running
waters. Aquatic Botany,86, 171–178.
Passy, S. I., & Larson, C. A. (2011). Succession in stream biofilms is an
environmentally driven gradient of stress tolerance. Microbial Ecology,
62, 414–424.
Piggott, J. J., Lange, K., Townsend, C. R., & Matthaei, C. D. (2012). Multi-
ple stressors in agricultural streams: A mesocosm study of interac-
tions among raised water temperature, sediment addition and
nutrient enrichment. PLoS ONE,7, e49873.
Piggott, J. J., Salis, R. K., Lear, G., Townsend, C. R., & Matthaei, C. D.
(2015). Climate warming and agricultural stressors interact to deter-
mine stream periphyton community composition. Global Change Biol-
ogy,21, 206–222.
Piggott, J. J., Townsend, C. R., & Matthaei, C. D. (2015). Reconceptualiz-
ing synergism and antagonism among multiple stressors. Ecology and
Evolution,5, 1538–1547.
NEIF ET AL.
|
11
Pinheiro, J., Bates, D., Debroy, S., & Deepayan, S.; Core Team R. eds.
(2015). nlme: Linear and nonlinear mixed effects models. http://cran.
r-project.org/web/packages/nlme/.
Poff, N. L. (1997). Landscape filters and species traits: Towards mecha-
nistic understanding and prediction in stream ecology. Journal of the
North American Benthological Society,16, 391–409.
Prescott, G. W., Croasdale, H. T., Vinyard, W. C., & Bicudo, C. E. M.
(1982). A Synopsis of North American Desmids. Part II. Desmidia-
ceae: Placodermae. University Nebraska Press, Lincoln.
R Development Core Team. (2016). R: A Language and Environment for
Statistical Computing. R Foundation for Statistical Computing, Vienna,
Austria. Retrieved from https://www.R-project.org/.
Schneck, F., & Melo, A. S. (2012). Hydrological disturbance overrides the
effect of substratum roughness on the resistance and resilience of
stream benthic algae. Freshwater Biology,57, 1678–1688.
Schneck, F., Schwarzbold, A., & Melo, A. S. (2011). Substrate roughness
affects stream benthic algal diversity, assemblage composition, and
nestedness. Journal of the North American Benthological Society,30,
1049–1056.
Southwood, T. R. E. (1977). Habitat, the templet for ecological strategies?
Journal of Animal Ecology,46, 337–365.
Southwood, T. R. E. (1988). Tactics, strategies and templets. Oikos,52,3–
18.
Steinman, A. D., Mulholland, P. J., & Hill, W. R. (1992). Functional
responses associated with growth form in stream algae. Journal of the
North American Benthological Society,11, 229–243.
Sun, J. & Liu, D. (2003). Geometric models for calculating cell biovolume
and surface area for phytoplankton. Journal of Plankton Research,25,
1331–1346.
Sutherland, I. W. (2001). The biofilm matrix –an immobilized but
dynamic microbial environment. Trends in Microbiology,9, 222–227.
Tank, J. L., & Dodds, W. K. (2003). Nutrient limitation of epilithic and
epixylic biofilms in ten North American streams. Freshwater Biology,
48, 1031–1049.
Townsend, C. R., Uhlmann, S. S., & Matthaei, C. D. (2008). Individual and
combined responses of stream ecosystems to multiple stressors. Jour-
nal of Applied Ecology,45, 1810–1819.
Uterm€
ohl, H. (1958). Zur Vervollkommnung der quantitativen Phyto-
plankton-Methodik. Mitteilungen Internationale Vereinigung f€
ur Theo-
retische und Angewandte Limnologie,9,1–38.
Van den Brink, P. J., & Ter Braak, C. J. F. T. (1998). Multivariate analysis
of stress in experimental ecosystems by Principal Response Curves
and similarity analysis. Aquatic Ecology,32, 163–178.
Van den Brink, P. J., & Ter Braak, C. J. F. (1999). Principal response curves:
Analysis of time-dependent multivariate responses of biological com-
munity to stress. Environmental Toxicology and Chemistry,18, 138–148.
Wagenhoff, A., Lange, K., Townsend, C. R., & Matthaei, C. D. (2013). Pat-
terns of benthic algae and cyanobacteria along twin-stressor gradi-
ents of nutrients and fine sediment: A stream mesocosm experiment.
Freshwater Biology,58, 1849–1863.
SUPPORTING INFORMATION
Additional Supporting Information may be found online in the sup-
porting information tab for this article.
How to cite this article: Neif
EM, Graeber D, Rodrigues L,
et al. Responses of benthic algal communities and their traits
to experimental changes in fine sediments, nutrients and
flow. Freshwater Biol. 2017;00:1–12. https://doi.org/10.1111/
fwb.12965
12
|
NEIF ET AL.
