Comparison of metabolic rates among macrophyte and nonmacrophyte habitats in streams
Little is known about the relative contribution of different stream habitats to reach-scale metabolism. We measured in situ metabolism in sand, gravel, stone, and macrophyte habitats to compare metabolic rates among these habitat types and to compare habitat-weighted measurements with reach-scale measurements. We used open-bottom chambers in sand, gravel, and macrophyte habitats and closed-bottom chambers in stones, and we estimated reach-scale metabolism from 2-station O2 budgets. Macrophyte habitats had a significantly higher gross primary production (GPP) and community respiration (CR) than stone, gravel, and sand habitats. A large part of this difference was associated with epiphytic biofilm: 28% of net ecosystem production (NEP), 20% of CR, and 24% of GPP. Macrophyte habitats contributed proportionally more to reach-scale metabolism than did the other habitat types. Forty-one percent of reach-scale NEP, 60% of ecosystem respiration (ER), and 50% of GPP were associated with this habitat type even though only 14% of the reach was covered by macro-phytes. We found significant linear relationships between GPP and CR and the amount of autotrophic biomass in the streams. The rates reported in 11 literature studies fit into our observed relationships, showing the generality of our findings. The rates we obtained expand the range of reported metabolic values in relation to auto-trophic biomass for both low and high biomass. The importance of macrophyte habitat can be ascribed to the macrophytes themselves, the associated epiphytic biofilm, and the fine organic material trapped in the dense stands. We conclude that besides having an effect on the structural elements in streams, macrophytes contribute significantly to stream ecosystem functions.
Comparison of metabolic rates among macrophyte
and nonmacrophyte habitats in streams
Anette Baisner Alnoee
, Tenna Riis
, and Annette Baattrup-Pedersen
Department of Bioscience, Aarhus University, Ole Worms Allé 1, DK-8000 Aarhus, Denmark
Department of Bioscience, Aarhus University, Vejlsøvej 25, DK-8600 Silkeborg, Denmark
Abstract: Little is known about the relative contribution of diﬀerent stream habitats to reach-scale metabolism.
We measured in situ metabolism in sand, gravel, stone, and macrophyte habitats to compare metabolic rates
among these habitat types and to compare habitat-weighted measurements with reach-scale measurements. We
used open-bottom chambers in sand, gravel, and macrophyte habitats and closed-bottom chambers in stones,
and we estimated reach-scale metabolism from 2-station O
budgets. Macrophyte habitats had a signiﬁcantly
higher gross primary production (GPP) and community respiration (CR) than stone, gravel, and sand habitats.
A large part of this diﬀerence was associated with epiphytic bioﬁlm: 28% of net ecosystem production (NEP),
20% of CR, and 24% of GPP. Macrophyte habitats contributed proportionally more to reach-scale metabolism
than did the other habitat types. Forty-one percent of reach-scale NEP, 60% of ecosystem respiration (ER), and
50% of GPP were associated with this habitat type even though only 14% of the reach was covered by macro-
phytes. We found signiﬁcant linear relationships between GPP and CR and the amount of autotrophic biomass
in the streams. The rates reported in 11 literature studies ﬁt into our observed relationships, showing the gen-
erality of our ﬁndings. The rates we obtained expand the range of reported metabolic values in relation to auto-
trophic biomass for both low and high biomass. The importance of macrophyte habitat can be ascribed to the
macrophytes themselves, the associated epiphytic bioﬁlm, and the ﬁne organic material trapped in the dense
stands. We conclude that besides having an eﬀect on the structural elements in streams, macrophytes contribute
signiﬁcantly to stream ecosystem functions.
Key words: benthic metabolism, autotrophic biomass, Ranunculus aquatilis, epiphytic metabolism, stream
metabolism, primary production
Ecosystem production and respiration are functions that
are related directly to energy ﬂows within the ecosystem.
Measurements of metabolism provide information about
these functions in the form of gross primary production
(GPP), which expresses the total C ﬁxed, and ecosystem
respiration (ER), which expresses the total amount of or-
ganic matter respired in the ecosystem. In general, if ER >
GPP, the stream is heterotrophic, whereas if GPP > ER,
the stream is autotrophic.
Comparisons of GPP and ER over spatial and temporal
scales provide knowledge on how physicochemical and bi-
otic variables aﬀect energy ﬂow. Structural properties, in
terms of physical and chemical conditions controlled by
hydrological and geomorphological conditions, and the pres-
ence and abundance of autotrophic and heterotrophic or-
ganisms aﬀect stream metabolism directly and indirectly
because they may act as stressors or facilitators on the me-
tabolism of other organisms. Stream canopy cover con-
trols stream metabolism via the combined eﬀects of light
availability and allochthonous supply of detritus (Bernot
et al. 2010, Clapcott and Barmuta 2010). Furthermore,
across biomes, reach-scale GPP is related to light availabil-
ity as a descriptor of autotrophic biomass (i.e., periphyton),
whereas ER is related to water P concentration and the
extent of depositional areas for organic matter within the
reach (Mulholland et al. 2001). The eﬀect of autotrophic
biomass on reach-scale GPP and ER was demonstrated in
a Swiss stream where GPP and ER decreased up to 70%
after 90% of the macrophytes was removed (Kaenel et al.
2000). In addition to their eﬀect on reach-scale metabo-
lism, macrophytes can be large 3-dimensional structures
capable of changing current velocity and sediment com-
position (Sand-Jensen 1998, Kleeberg et al. 2010) and add-
ing to the physical complexity of streams.
Studies on habitat-speciﬁc metabolism in streams gen-
erally are based on in-stream or at-stream measurements
of changes in dissolved O
(DO) in chambers. In-stream
measurements done in open-bottom chambers embed-
DOI: 10.1086/687842. Received 12 August 2015; Accepted 22 March 2016; Published online 12 July 2016.
Freshwater Science. 2016. 35(3):834–844. © 2016 by The Society for Freshwater Science.
ded in the stream bed include hyporheic metabolic pro-
cesses (Fellows et al. 2006), whereas closed-bottom cham-
bers placed in the stream (e.g., Reid et al. 2006, Aristegi
et al. 2010, Acuña et al. 2011) include only the metabo-
lism of the components inside the chamber. At-stream
chambers are placed at the stream bank but with contin-
uous renewal of stream water in the chambers through-
out the experimental period (Prahl et al. 1991).
Eﬀects of macrophytes on stream metabolism have
been assessed by measuring reach-scale metabolism from
1- or 2-station measurements (e.g., Marzolf et al. 1994,
Young et al. 1998) or in the laboratory (Van et al. 1976,
Binzer et al. 2006). However, measurements at the reach-
scale give no information on how diﬀerent types of habi-
tat contribute to reach-scale metabolism. Habitat-speciﬁc
metabolism in streams has been studied in sand, gravel/
cobble, and stone habitats (e.g., Bott et al. 1985, Fellows
et al. 2001, Aristegi et al. 2010), whereas fewer investiga-
tors have studied habitat-speciﬁc metabolism in macro-
phyte habitats (but see Prahl et al. 1991, Acuña et al. 2011,
Leggieri et al. 2013) despite their dominant role in many
lowland streams. Koopmans and Berg (2015) reported an
eddy covariance technique for studying stream metabolism
at the habitat scale, but this technique has not yet been
tested in macrophyte beds.
Measuring habitat metabolism in chambers has several
disadvantages. First, when in-stream or at-stream closed
chambers are used, the hyporheic zone, which may con-
tribute signiﬁcantly to metabolism (Fellows et al. 2001), is
excluded. Second, supersaturation of CO
ent limitation can develop in the chambers because they
are closed systems. This problem can be avoided by limit-
ing the incubation time. Third, when measuring habitat
metabolism in chambers, potential hotspots in the reach
can be overlooked. For instance, leaf packs, amphibious veg-
etation, or macrophytes, which could contribute signiﬁ-
cantly to the reach-scale metabolism, might cover only a
small area of the reach. Macrophytes can be potential hot-
spots because they have high autotrophic biomass, large
deposits of ﬁne organic sediments in the beds (Sand-
Jensen 1998, Kleeberg et al. 2010), roots that respire, and
surface area that acts as substrate for epiphytic bioﬁlm.
We measured metabolism in stream habitats dominated
by stone, gravel, sand, and macrophytes to estimate the
proportional contribution of these typical lowland-stream
habitats to reach-scale metabolism. We used open-bottom
chambers except when measuring metabolism on stone
habitats, where we used closed-bottom chambers. The spe-
ciﬁc aims of our study were to: 1) measure in situ meta-
bolic rates in stream habitats dominated by stone, gravel,
sand, and macrophytes, 2) quantify the contribution of hab-
itat metabolism to reach-scale metabolism, 3) estimate the
contribution of epiphytic bioﬁlm to macrophyte habitat
metabolism, 4) identify predictors of habitat metabolism,
and 5) compare habitat metabolism among a range of
stream types reported in the literature to evaluate more
generally the role of macrophyte habitats as potential met-
abolic hotspots. We hypothesized that macrophyte habi-
tats would have higher metabolic rates/stream area be-
cause of higher standing biomass and contributions from
epiphytic bioﬁlm. Furthermore, metabolism and biomass
are highly correlated, so we expected that autotrophic
biomass would be one of the main predictors of habitat
Study sites and stream characteristics
We conducted our study at 2 temperate Danish low-
land streams, Linå and Skader. The study reach at Linå is
-order and meandering with high occurrence of sub-
merged macrophytes, whereas the study reach at Skader
-order with limited occurrence of submerged mac-
rophytes. We placed 100 evenly distributed transects per-
pendicular to the stream bank in each 500-m-long ex-
perimental reach. At each transect, we measured stream
width, depth at 5 equidistant points along the transect,
and cover of benthic substrates and macrophyte species.
The substrates were stone, gravel, sand, and mud. We cal-
culated cover (%) of each substrate type and macrophytes
at the reach scale as relative frequency at all examined points
and estimated areal cover (m
) by scaling relative cover to the
whole reach area.
We estimated average reach-scale water velocity in 2
representative 100-m reaches within the 500-m experi-
mental reach by adding a conservative tracer (NaCl) to
the stream and measuring conductivity 100 m down-
stream as the pulse passed (Webster and Valett 2006).
We calculated discharge based on the average water ve-
locity and stream proﬁle area of the 100-m reach. We used
temperature and light loggers (HOBO Pendant
Bourne, Massachusetts) to measure temperature at the mid-
dle of the reach every 10 min. We used a geographic infor-
mation system (GIS; ArcGIS, version 10.0; Environmental
Research Systems Institute, Redlands, California) to estimate
land use in a 50-m-wide buﬀer along all up-stream reaches
(GIS layers; ‘AIS_landuse’and ‘Catchment’from the national
catchment database; Aarhus University). We categorized land
uses as agriculture, forest, fresh water, open, urban areas, and
We collected water samples twice (once before and
once after the metabolism measurements) during a 1.5-mo
period at base ﬂow. We ﬁltered some samples with glass-
ﬁber ﬁlters (Whatman
; Buckinghamshire, UK) and stored
all samples on ice until returning to the laboratory, where
they were frozen. We analyzed ﬁltered samples for inor-
ganic nutrient concentrations (NO
, soluble reactive
P[SRP])withaﬂow-injection analyzer (Lachat Instruments,
Loveland, Colorado). We analyzed unﬁltered samples for
total N (TN) with a TOC-VCPH (Shimadzu, Kyoto, Japan)
Volume 35 September 2016 | 835
and total P (TP) with ammonium molybdate following a
persulfate digestion. We measured water alkalinity by end-
point titration with 0.05M HCl.
We measured habitat metabolism at each experimen-
tal reach in habitat types covering >10% of the stream bed.
We conducted measurements only on sunny days. At Linå,
we conducted all habitat measurements within 16 d in late
July–early August. At Skader, we conducted all habitat
measurements within 6 d at the end of August. We con-
ducted 39 habitat measurements, distributed among 4 hab-
itat types: stone, gravel, sand, and macrophytes. At Linå,
we measured habitat metabolism on stone (n= 8), gravel
(n= 4), sand (n= 10), and macrophytes (n= 5). At
Skader, we measured habitat metabolism only on stone
(n=4)andsand(n= 8) because macrophyte and gravel
habitats either were not present or had limited occurrence.
We conducted the macrophyte habitat measurements
at Linå in Ranunculus aquatilis L. beds because this was
the predominant species present. Contributors to metab-
olism in the macrophyte habitat consisted of the macro-
phytes themselves, the epiphytic bioﬁlm, and O
fused from the water into the upper sediment beneath
the macrophytes because of low O
concentrations in the
sediment. The deep sediment under the sand, gravel, and
macrophytes was ﬁne sand, which enclosed the cylinder
tightly. Therefore, we expected that any exchange of wa-
ter between chamber and hyporheic water would be in-
signiﬁcant. We also measured metabolism on artiﬁcial
plants colonized by epiphytes at Linå (n= 6).
We measured habitat metabolism in sand and gravel
in hemispheric Plexiglas
domes (diameter = 29 cm, total
height = 25 cm, maximum volume = 13.7 L) mounted
with a circulation pump and an O
probe (ProOdo; Yellow
Springs Instruments, Yellow Springs, Ohio) that logged
DO (mg O
saturation (DO%), and temperature (°C)
every 10 min (Fellows et al. 2006, 2009). We embedded the
dome ∼10 cm into the substrate to prevent or reduce wa-
ter exchange through the hyporheic zone. To calculate the
exact water volume inside the dome we removed the O
probe and inserted a ruler. We measured the height from
the streambed to the top of the dome after each habitat
measurement and before moving the dome to another
habitat. We measured metabolism in stone habitats in a
closed dome because it was impossible to embed the dome
into the stone substrate. We collected stones correspond-
ing to the area of the bottom of the dome and placed
them on a Plexiglas bottom before closing the dome and
sealing it with sticky tack to prevent exchange of water or
gasses between the dome and the surrounding waters.
We left the domes totally submerged in the stream hab-
itats for 24 to 30 h to calculate the metabolism during a
whole day and to avoid nutrient limitation caused by cham-
ber eﬀects. The domes were surrounded by stream water,
so the temperature did not diﬀer between the habitat and
We collected substrate samples from all sand and gravel
habitats to estimate autotrophic biomass and organic mat-
ter content. In sand habitats, we collected 3 samples with
a tipless syringe (area [A] = 6.6 cm
), whereas in gravel hab-
itats, we collected 3 samples with a pipe (A = 22.7 cm
In stone habitats, we collected samples for measurement
of autotrophic biomass by scraping bioﬁlm from 3 stones
(A = 6.6 cm
/sample). We extracted chlorophyll a(chl a;
g chl a/m
) in 90% ethanol and measured concentrations
with a spectrophotometer (UV-1700, UV-visible spectro-
photometer; Shimadzu, Suzhou, China). We calculated or-
ganic matter in the sediment from the ash-free dry mass
(AFDM) measured after dried samples were weighed, com-
busted at 550°C for 24 h, and reweighed. We then calcu-
lated AFDM as % organic matter in the sediment.
In macrophyte habitat, we measured metabolism in Plex-
iglas cylinders (A = 102.1 cm
, maximum volume = 4.96 L).
We placed the cylinder on top of the macrophyte bed,
pushed it down through the macrophyte bed, and embed-
ded it ∼10 cm into the sediment. This method enabled us
to measure metabolism in a natural macrophyte bed and
to include the upper sediment and the eﬀect of shading
from the surrounding macrophyte bed. We mounted the
cylinders with a circulation pump and an O
them with a Plexiglas lid at the water surface, and sealed
them with sticky tack to prevent gas exchange with the
atmosphere. We left the cylinders in place for ≥2hduring
midday and measured temperature and DO every 10 min
with ProOdo probes. After the 2-h light incubation, we
placed a dark cylinder around the Plexiglas cylinder so we
could measure respiration rates in the dark. We waited 1 h
after covering the cylinders so the autotrophs could adjust
temperature and DO every 10 min for ≥2 h. Next, we mea-
sured the distance from the stream bed to the water level
in the cylinder to calculate the exact water volume. We col-
lected all macrophytes in the cylinder, estimated the surface
area of the macrophytes with the software AnalyzingDigital
Images (version 2.0; MVH Image, Amherst, Massachusetts),
and estimated their biomass after drying at 60°C for ≥48 h.
The surface area ranged from 24.3 to 434.4 cm
We measured epiphytic bioﬁlm metabolism with bio-
ﬁlm established on artiﬁcial plants as a way to separate epi-
phytic bioﬁlm metabolism from macrophyte metabolism
in the macrophyte habitat. In total, we conducted 6 epi-
phytic bioﬁlm metabolism measurements at Linå. We
glued the artiﬁcial plants to a Plexiglas plate and placed
the plates in the stream for ≥4 wk before measurements
to allow the bioﬁlm to become established on the plants.
We used 3 types of artiﬁcial plant: 1 with strap-like leaves
resembling Sparganium spp. and the other 2 with dissected
leaves resembling the submerged species Ranunculus spp.
836 | Habitat metabolism in streams A. B. Alnoee et al.
and Myriophyllum spp. All of these species are common
in many temperate lowland streams. After ≥4 wk, we at-
tached a cylinder similar to the one used for measurement
of macrophyte metabolism to the plate with the artiﬁcial
plant and measured metabolism as described above for
macrophytes. At the end of the incubation, we recorded
water depth in the cylinder to calculate the exact water
volume and brought the artiﬁcial plants to the laboratory
where we gently scrubbed the bioﬁlm from the artiﬁcial
plant and measured chl aafter ﬁltering the slurry onto
GF/C glass-ﬁber ﬁlters and extracting in ethanol. We
scaled chl ato g/m
plant by estimating the surface area
of the artiﬁcial plants with the software AnalyzingDigital-
Images. The surface area of the 3 types of artiﬁcial plants
ranged from 154.4 to 429.5 cm
. We calculated epiphytic
bioﬁlm production as g O
plant surface d
increase or decrease in DO under light or dark conditions,
respectively, and the area of the artiﬁcial plants.
At Linå, we recorded surface irradiance every 10 min
during light and dark metabolism measurements with a
temperature and light logger (HOBO, UA-002-64), and
at Skader we recorded surface irradiance every 5 min with
a light logger (LI 190SA; Li-COR, Lincoln, Nebraska). We
calculated average photosynthetically active radiation (PAR)
per measurement of dome/cylinder for the period when
DO was being measured in the habitat.
We calculated GPP in the domes and cylinders based
on 2 to 3 h of DO production measurements under light
conditions to ensure that it reﬂected the maximum pro-
duction rate. Estimates of CR of the organisms in the hab-
itat were based on 2 to 3 h of measurements under dark
conditions. Only CR measurements with DO concentra-
tions >3 mg O
/L were included to prevent inclusion of
hypoxia eﬀects during the measurements. Furthermore,
to ensure that any change in metabolism was not caused
by short-term eﬀects of cloud shading, DO had to be lin-
early related to time (r
> 0.85). We multiplied metabolic
values (g O
) by water volume (L) in the dome/
cylinder and divided by dome/cylinder bottom area (m
to obtain g O
. Next, we calculated metabolic met-
rics (CR and NEP) for the biotic community in the domes.
We calculated NEP as the rate of increase in DO in the
dome/cylinder and scaled it to daily NEP rates by multi-
plying by the number of daily light hours. We calculated
CR from the rate of decrease in DO during dark or night
conditions and multiplied measurements by 24 h to obtain
daily CR rate. The sum of NEP and CR for the photic
period was GPP, and we calculated the ratio of produc-
tion/respiration (P/R) for each habitat measurement as
GPP/CR. Instead of using P/R = 1 as the point at which
the stream shifts between autotrophy and heterotrophy,
we used the 0.63 ×GPP line presented by Hall and Beau-
lieu (2013). They compared the fraction from diﬀerent
streams worldwide and found that autotrophic ER consti-
tutes 63% of GPP, whereas the rest is from the heterotro-
phic organisms. Hereafter, we refer to this line as the
We scaled habitat metabolism to estimate habitat-
weighted reach-scale net ecosystem production (NEP
ecosystem respiration (ER
), and gross primary pro-
) by multiplying each habitat cover (%)
by the respective habitat metabolism (g O
summed the contribution from all habitats to obtain
habitat-weighted reach-scale metabolism.
To quantify the metabolic contribution from the plank-
tonic algae in the domes and cylinders, we incubated
stream water in 3 light and 3 dark 200-mL glass bottles
for 3 h in the middle of the water column in the stream.
We measured temperature and DO before and after in-
cubation. We measured P and R as the rate of change in
concentration from the start to end of the ex-
periment (as described above for the domes/cylinders).
We measured reach-scale metabolism with the
upstream–downstream 2-station DO-change technique de-
scribed by Odum (1956) and modiﬁed by Marzolf et al.
(1994, 1998) and Young et al. (1998). We used YSI 6600
V2-2 Multiparameter Water Quality Probes (Yellow Springs
Instruments, Yellow Springs, Ohio) to measure DO (mg/L)
and temperature (°C) every 10 min for 24 h during the
habitat measurements. We calibrated the probes at 100%
saturation with calibration caps before deploying them in
the stream and corrected for drift in DO by intercalibrat-
ing the probes in the stream before and after measurements
for ≥½ h. We assumed the drift was linear over time and
adjusted for diﬀerences among probes accordingly. At Linå,
we measured surface irradiance with a temperature and
light logger (HOBO, UA-002-64) every 10 min during me-
tabolism measurement, and at Skader, we measured surface
irradiance every 5 min with a LI 190SA meter. We calcu-
lated average PAR per day of metabolism measurement.
Reach-scale metabolic values obtained from open-
water measurements are given as net ecosystem production
) and ecosystem respiration (ER
). The reach-scale
metabolic metrics, NEP
,diﬀer from habitat-
because they are based on
changes in DO in open-water measurements across the
whole reach rather than on measurements scaled up from
as the sum
of the change in DO adjusted for temperature and reaera-
tion according to Bott (2006) from 1 h after sunrise to 1 h
before sunset (according to timeanddate.com). We calcu-
as the nighttime change in DO concentration
caused by DO consumed, DO deﬁcit, and reaeration from
0000 to 0300 h, assuming constant rates throughout the
day, and multiplied by 24 to get the daily rate. We calcu-
Volume 35 September 2016 | 837
as the sum of DO produced in the photic
period (sum of NEP
during the photic period) plus
DO consumed during the photic period calculated from
nighttime respiration (ER
from night multiplied by pho-
tic hours). The reaches were too long to allow measure-
ment of metabolism along the whole reach, so we separated
them into 2 parts and averaged the metabolism per stream
At Skader, we measured reaeration with propane re-
lease and salt as a conservative tracer (Marzolf et al. 1994).
At Linå, propane measurement was not possible because
the stream ﬂooded immediately after metabolism measure-
ments were made and before the planned propane release,
so we estimated a reaeration coeﬃcient, k
, from an em-
pirical model, the Surface Renewal Model (SRM) (Owens
et al. 1964, Owens 1974) cited by Hauer and Lamberti
(2006). We compared the measured values from the pro-
pane release in Skader with calculated values from the
SRM, and the variability between the 2 methods was low
(trace method: k= 0.014/min and 0.007/min for the 2
reaches; SRM model: k= 0.010/min and 0.007/min), so we
decided to use this model for Linå because the streams are
physically comparable. We also tried other empirical meth-
ods from the literature (Aristegi et al. 2009), but none of
these showed a reaeration coeﬃcient as similar to the one
calculated from the tracer. We could not use night time
regression because the change in O
during night was not
>1 mg/L during winter (Thyssen et al. (1987).
We conducted a literature search on stream habitat
metabolism to compare our habitat metabolism with other
measurements. Measurements had been obtained by dif-
ferent methods, but all the studies included were con-
ducted during summer. We obtained 73 comparable mea-
surements from 11 studies on stream habitat metabolism
and autotrophic biomass from the literature.
To compare the studies, we converted the metabolism
(g C m
) estimates by Minshall et al. (1992), Fellows
et al. (2006), and Reid et al. (2006) to g O
dividing by 0.375 g C/g O
(following Lampert 1984 and
Bender et al. 1987). We did not include studies in which
metabolic rates were measured on an hourly basis (g O
) because day length was not given, so we could
not convert hourly rates to g O
except for Biggs
et al. (1999) and Fellows et al. (2001), for which we multi-
plied respiration rates by 24 h to obtain daily rates.
We used 1-way analysis of variance (ANOVA) and a
multiple range test (Least Signiﬁcant Diﬀerence [LSD])
to test for diﬀerences in metabolic rates among habitats.
Before applying the ANOVA, we tested for homogeneity
of variance in all data. To identify which variables con-
trolled habitat metabolism, we applied a General Linear
Model (GLM). Variables included in the GLM were au-
totrophic biomass, stream (Linå, Skader), habitat type,
water quality, % cover of agricultural land use in a 50-m-
wide buﬀer zone up-stream of the study reach, maximum
temperature inside the dome/cylinder, water velocity in
the stream reach, water depth and transect width where
the dome/cylinder was placed, sediment organic matter (%)
inside the dome, and average PAR measured during com-
munity production in the domes. We used comparison of
regression lines to compare the slopes of lines. When nec-
essary, data were log(x)-transformed to obtain homosce-
Physical and chemical characteristics
Nutrient concentrations were 2 to 5×higher in Skader
than in Linå except for NH
, for which concentrations
were similar in the 2 streams (Table 1). Average temper-
ature was comparable in the 2 streams. Linå had twice
the discharge of Skader, and both catchments were dom-
inated by agricultural land use (Table 1).
NEP, CR, and GPP estimated from change in DO in
the dome/cylinder diﬀered among the 4 habitat types and
between Linå and Skader and were positively correlated
Table 1. Mean (± SD, n= 2) physicochemical measurements
for Linå and Skader. Temperature is mean (± SD) from the
middle of June to the middle of October, and catchment land
use is % total catchment area. * indicates signiﬁcant diﬀerences
between streams (p< 0.05).
Alkalinity (meq/L) 1.59 2.74
(mg/L) 0.05 0.04
(mg/L) 2.37 5.60
Total N (mg/L) 2.62 5.37
(mg/L) 0.01 0.05
Total P (mg/L) 0.08 0.14
Temperature (°C) 13.93 ± 1.18 13.86 ± 2.35
/s) 0.17 0.08
Stream width (m) 2.87 ± 0.67 3.25 ± 1.09*
Stream depth (m) 0.34 ± 0.22* 0.30 ± 0.14
% agriculture 50 66
% forest 7 11
% open nature 2 10
% urban 3 6
% freshwater areas 38 5
838 | Habitat metabolism in streams A. B. Alnoee et al.
with the autotrophic biomass in the habitat (GLM, r
0.77 for NEP, r
phyte habitats had signiﬁcantly higher CR and GPP than
stone, gravel, or sand habitats, and sand habitats had
higher CR and GPP than stone habitats (Table 3). NEP did
not vary among habitat types. All habitats had P/R > 0.63.
Macrophyte habitats had the lowest ratio, but only P/R for
gravel was signiﬁcantly higher than for macrophytes (Ta-
ble 3). Epiphytic bioﬁlm contributed 28% to total macro-
phyte habitat NEP, 20% to macrophyte habitat CR, and
24% to macrophyte habitat GPP on a plant-area basis
(Table 3), but P/R did not diﬀer signiﬁcantly between epi-
phytes and macrophytes. Planktonic algae did not contrib-
ute to habitat metabolism because DO did not diﬀer be-
tween bottles incubated under light and dark conditions
(data not shown).
In general, GPP and CR were higher at Linå than at
Skader across all habitats, mainly because of the higher
activity in sand habitats and the high GPP and CR in mac-
rophyte habitat, which was present only in Linå (Fig. 1A,
B). CR and GPP and autotroph biomass were positively
related across all habitats at Skader and Linå (CR: r
0.81, p< 0.0001; GPP: r
= 0.78, p< 0.0001), and the pos-
itive relationship was supported more generally when mea-
surements from the literature were considered together
with our measurements (Fig. 1A, B). The data obtained
from measurements in Linå and Skader included the full
range of biomasses reported in the literature. The 4 high-
est biomass measurements in our study originated from
macrophyte habitats (circle in Fig. 1A, B) and the lowest
biomass measurements originated from stone habitats.
Habitat-weighted reach-scale metabolism
At both streams, NEP
was ∼3×higher than NEP
was 3 to 8×higher than ER
(Table 4). ER
and GPP were higher at Linå than Skader regardless of the
Habitat cover at Linå was 14% macrophytes (R. aqua-
tilis), 41% sand, 22% gravel, and 23% stone, whereas the
habitat cover at Skader was 68% sand, 8% gravel, and 24%
Table 2. Results of a General Linear Model for the eﬀects
of habitat variables on net ecosystem production (NEP;
), community respiration (CR; g O
gross primary production (GPP; g O
) for all habitat
measurements. Only variables with signiﬁcant eﬀects are
included. Habitat includes stone, gravel, sand, and macrophyte.
Biomass = autotrophic biomass in the diﬀerent habitat types
and is directly related to metabolism.
NEP CR GPP
Habitat 18.72 <0.0001 47.28 <0.0001 41.15 <0.0001
Stream 48.16 <0.0001 39.35 <0.0001 73.04 <0.0001
Log(biomass) 7.54 0.0097 14.70 0.0005 13.93 0.0007
Table 3. Mean (± SD) net ecosystem production (NEP; g O
), community respiration (CR; g O
), gross primary
production (GPP; g O
), and production/respiration (P/R) for the diﬀerent habitat types (n= 12 for stone, n= 4 for gravel,
n= 18 for sand, n= 5 for macrophytes) and areal NEP (g O
), CR (g O
), GPP (g O
and P/R for macrophytes and epiphytic bioﬁlm (n= 5 for macrophytes and n= 6 for epiphytic bioﬁlm) for measurements from
both streams. The chamber type used for each measurement is indicated, and closed/open refers to whether the chamber included
a bottom (closed) or not (open). Macrophyte habitats are beds of Ranunculus aquatilis. Values with the same letters are not
signiﬁcantly diﬀerent between habitat types or R. aquatilis and epiphytic bioﬁlm ( p< 0.05). Epiphytic bioﬁlm is included in the R.
Habitat or autotroph type
Stone Gravel Sand Macrophyte Epiphytic bioﬁlm
Dome, closed Dome, open Dome, open Cylinder, open Cylinder, closed
NEP (g O
± 0.33 0.24
± 0.27 0.35
± 0.53 −1.54
CR (g O
± 0.44 0.71
± 0.31 1.39
± 0.78 11.85
GPP (g O
± 0.45 0.94
± 0.36 1.75
± 1.05 10.31
± 0.38 1.43
± 0.45 1.27
± 0.41 0.89
NEP (g O
± 2.60 0.85 ± 0.39
CR (g O
± 3.97 1.24
GPP (g O
± 4.17 1.39
± 0.15 1.17
Volume 35 September 2016 | 839
stone (Fig. 2). At Linå, habitat-weighted whole-reach me-
tabolism showed that the 14% of the reach covered by mac-
rophytes contributed disproportionally more to metabo-
lism than other habitats. Macrophyte habitat contributed
41, 60, and 50% to NEP
, and GPP
tively, whereas sand habitat in Linå contributed 43, 28,
and 36%, respectively. At Skader, the 68% of the reach
covered by sand contributed 78, 77, and 79% to NEP
, and GPP
, respectively, whereas the 24% cov-
ered by stone contributed 22, 23, and 21%, respectively.
Our measurements of habitat metabolism were within
the range of other habitat metabolism measurements re-
ported in the literature (Fig. 3). The distribution of dif-
ferent types of habitats was highly uneven among the 73
measurements reported in the literature. Most studies
were conducted in gravel/cobble habitat (32 studies) or
sand habitat (15 studies), and only 8 studies were con-
ducted in stone habitat, 5 in macrophyte habitat, 1 in epi-
pelon, and 3 in ﬂoating macroalgae. For 8 of the mea-
surements reported in the literature, we used only values
for respiration and biomass because we were unable to
convert GPP rates into daily rates. GPP varied between
0.0 and 35.3 g O
and CR varied between 0.07
and 17.2 g O
(Table S1), but most habitats had
GPP and CR between 0.5 and 4 g O
. The highest
values of GPP and CR in the literature (>15 g O
were for ﬂoating macroalgae (Acuña et al. 2011) and in a
polluted stream aﬀected by intensive managed tree plan-
tations (Aristegi et al. 2010). The lowest CR values were
measured by Biggs et al. (1999) on unstable substrate in
streams with high water velocities. In general, stone, gravel,
and sand habitats varied within the same range, and no
habitat-speciﬁc patterns were found in their metabolism.
The regressions for CR vs GPP were signiﬁcant with
our data from Linå and our data pooled with the meta-
bolic rates from the literature, despite the diﬀerence in auto-
trophic organisms among habitats (Fig. 3). The slopes of
both regressions diﬀered from the 0.63GPP line (All: F=
24.04, p< 0.001; Linå: F= 8.66, p= 0.004). When we
considered the data by habitat type, the slopes of the
regressions for CR vs GPP diﬀered from the 0.63GPP line
in all habitats (stone: F= 129.09, p<0.001; gravel: F=
3.99, p= 0.049; sand: F= 49.59, p= < 0.001; macrophytes:
F= 7.59, p= 0.008; Fig. 4A–D).
Metabolic rates were much higher in macrophyte than
other habitats in streams, and the epiphytic bioﬁlm con-
tributed signiﬁcantly to this metabolism. GPP and CR were
6to17×higher in R. aquatilis habitats than in stone, gravel,
and sand habitats, with microalgae being the dominant pri-
mary producer. CR and GPP measurements in macrophyte
Table 4. Reach-scale net ecosystem production (NEP), ecosys-
tem respiration (ER), gross primary production (GPP), and
GPP/ER estimated based on habitat-weighted metabolism
) and the 2-station method (NEP
). Data on 2-station metabolism are averaged
for 2 reaches in Linå and Skader. Numbers are means and do
not necessarily sum up.
Linå Skader Linå Skader
) 1.58 0.44 −2.52 −1.04
) 2.76 0.69 8.72 5.48
) 2.88 0.67 2.57 1.71
1.04 0.97 0.45 0.36
Figure 1. Relationship between community respiration (CR)
(A) and gross primary production (GPP) (B) and autotrophic
biomass from the literature (Table S1) and this study. Data
from Linå and Skader are reported for each habitat type. Reg. =
840 | Habitat metabolism in streams A. B. Alnoee et al.
habitats in our study (CR: 10.9–19.9 g O
7.19–17.1 g O
) and metabolic rates in cobble hab-
itats dominated by ﬁlamentous green algae in Spanish
streams (maximum CR = 17 g O
GPP = 35 g O
; Aristegi et al. 2010) were among
the highest values reported in the literature.
The epiphytic bioﬁlm contributed, on average, 28, 20,
and 24% to NEP, CR, and GPP, respectively. The high
metabolic contribution from the bioﬁlm emphasizes the
importance of macrophytes as substrate for bioﬁlm and,
therefore, as a hotspot for microbial metabolism (Pomaz-
kina et al. 2012, Tunca et al. 2014). The distribution of
epiphytic bioﬁlm diﬀers within macrophyte beds, and the
diﬀerences must be considered when scaling metabolic
rates to the whole-ecosystem level. For example, bioﬁlm
can grow thick mats on macrophyte surfaces in summer,
but its biomass is likely to vary within macrophyte beds
because light may be limiting in the deeper areas, and
diﬀering water velocities within the bed can aﬀect CO
and nutrient uptake (Biggs et al. 2005). Our results were
based on the assumption that bioﬁlm was evenly distrib-
uted throughout the macrophyte beds. The beds were
only 0.2 m deep, so light probably was not limited at that
We also assumed that metabolic rates of bioﬁlm would
be similar on natural and artiﬁcial plants. However, natural
and artiﬁcial plants could have aﬀected biomass of the
epiphytic bioﬁlm diﬀerently. For example, artiﬁcial plants
might aﬀect water velocity in a way that diﬀers from nat-
ural plants if they diﬀer physically from natural plants (e.g.,
in surface area). Moreover, natural and artiﬁcial plants
might diﬀer chemically because some natural macrophytes
release allelopathic chemicals that can inhibit the growth
of other organisms. However, we expected these eﬀects to
be minimal and that bioﬁlm would grow on almost every-
thing. We also expected that once established, bioﬁlm
would grow in all 3 dimensions, regardless of the substrate
being natural or artiﬁcial plants. However, more studies
are needed to predict more precisely how bioﬁlm may af-
fect the habitat metabolism within macrophyte beds and
whether plant surface area or allelopathic activity are key
factors for colonization.
All 4 habitats had P/R > 0.63, but macrophyte habitats
had the lowest P/R (0.89). We expected NEP to be much
higher on an areal basis in macrophyte than in nonmacro-
phyte habitat because of the high autotrophic biomass of
the macrophyte habitats. However, CR rates were higher
than GPP rates in macrophyte habitat. High CR probably
reﬂects a combination of several factors. First, water ve-
locity is lower in a macrophyte bed than in other stream
habitats. Therefore, deposition of ﬁne substrate and or-
ganic matter is high in macrophyte beds and leads to higher
consumption because of high breakdown rates (Sand-
Jensen 1998, Jones et al. 2011). Second, macrophytes must
sustain large amounts of nonphotosynthetic tissue and trans-
port ions over long distances. These processes require res-
piration by the plant.
Overall, CR was linked to GPP in the 4 habitat types, es-
pecially in gravel habitat. In all 4 habitat types, the slope
of the relationship diﬀered from the 0.63GPP line, but the
slope was higher only in macrophyte habitat. This result
indicates that the high CR in macrophyte habitats arose
Figure 2. Percent substrate cover and % contribution of sand,
gravel, stone, macrophyte, and other habitats to reach-scale net
ecosystem production (NEP
), ecosystem respiration (ER
and gross primary production (GPP
) for Linå (L) and Skader (S).
Figure 3. Relationship between community respiration (CR)
and gross primary production (GPP) from Linå, Skader, and
data from the literature (Table S1). Dotted line indicates the
0.63GPP relationship. The regression lines show the overall
trends for all values (Linå, Skader, and literature data) and for
Linå alone. Reg. = regression.
Volume 35 September 2016 | 841
from the macrophytes themselves and from the sediment.
The slope was lowest in stone habitats. This result could
have been because of underestimation of CR in the closed
chambers where the contribution from the surface and in-
terstitial sediments was not estimated.
Habitat-weighted estimation of reach-scale metabolism
based on both habitat measurements and the reach-scale
method showed that ER and GPP were higher at Linå,
which had 14% cover of R. aquatilis, than at unvegetated
Skader. At Linå, the 14% of the streambed covered by
R. aquatilis contributed disproportionately to reach-scale
(60%), and GPP
that lend support to the suggestion that the diﬀerence in
reach-scale metabolism between the 2 study streams was
caused mainly by macrophyte cover. Nutrient concentra-
tion did not diﬀer between the streams, and organic mat-
ter, light, and temperature were not among the best pre-
dictive variables in the GLM. Macrophyte beds may be
hotspots of high metabolic activity by the macrophytes,
their epiphytic bioﬁlm, and the heterotrophic organisms
in the upper layer of ﬁne sediment in the macrophyte bed.
Madsen et al. (1988) showed that reach-scale GPP and ER
were reduced signiﬁcantly after removal of macrophyte bio-
mass via weed cutting, even though only half of the bio-
mass was removed. Furthermore, in a manipulative study
of macrophyte cover during growing season, O’Brien et al.
(2014) found a positive relationship between GPP and mac-
rophyte cover in 3 New Zealand lowland streams.
Estimates of metabolic rates diﬀered between 2-station
measurements and habitat-weighted measurements. ER
was 3×higher than ER
in Linå and 8×higher in Skader.
GPP values were more-or-less similar between methods at
Linå, but GPP
was 2.5×higher than GPP
The higher GPP
and higher ER
at Skader may reﬂect our inability to ﬁt the large
stones in the reach into the chambers. Thus, we had to use
smaller stones to measure metabolic rates in stone habitats.
because big stones are more stable. Thus, GPP and ER may
have been underestimated for ER
tively. Furthermore, the small stones used were from a
riﬄe where biomass on stones could have been reduced
because stones were continuously rotated and moved, pro-
cesses that reduce the biomass. Moreover, a lower ER
also might reﬂect that we did not measure the
contribution from the upper sediments in the stone cham-
bers and, therefore, underestimated ER
. In addition,
the lower habitat-weighted rates potentially may be ascribed
to the fact that some habitats, such as leaf litter and other
organic detritus that might contribute signiﬁcantly to reach-
scale metabolism, were missing from the habitat-weighted
Figure 4. Relationship between community respiration (CR)
and gross primary production (GPP) from stone (A), gravel (B),
sand (C), and macrophyte (D) habitat from this study and data
from the literature (Table S1). Dotted line indicates 0.63GPP
842 | Habitat metabolism in streams A. B. Alnoee et al.
metabolism (Hedin 1990, Fuss and Smock 1996, Houser
et al. 2005). Amphibious plants growing at the margins of
the streams also may have contributed to ER because they
are rooted in the stream bottom. However, they contribute
only negligibly, if at all, to in-stream DO production be-
cause the gas exchange occurs primarily over the leaves
above the water surface.
Our results showed that macrophyte habitats had a
signiﬁcant and disproportionately higher contribution to
stream metabolism than stone, gravel, and sand habitats
in lowland streams. The high GPP in the macrophyte hab-
itats was caused by the plant itself and its epiphytic bio-
ﬁlm, which constituted 24% of the GPP. The high CR in
the macrophyte habitats was caused by macrophytes, epi-
phytic bioﬁlm (20%), and microbial heterotrophic pro-
cesses in the upper sediment deposited in the macro-
phyte beds. Deposition and mineralization of ﬁne organic
matter in the stream reduce the transport of organic mat-
ter to downstream recipients, such as lakes and coastal
waters. The higher metabolic rates in macrophyte than
in other habitats could be explained largely by the higher
autotrophic biomass/streambed area, a conclusion sup-
ported by a comparison of metabolic metrics from a wide
range of stream types. Our study conﬁrms that besides
having an eﬀect on the structural elements in streams,
macrophytes contribute signiﬁcantly to stream ecosystem
function through metabolism.
The authors acknowledge the Carlsberg Foundation (TR),
the Danish Natural Science Research Council (TR), the Euro-
pean Union 7
Framework Project REFRESH under contract
244121 (ABP) and ‘Managing aquatic ecosystems and water re-
sources under multiple stress’(MARS) under contract 603378
(ABP) for ﬁnancial support. We thank Søren E. Larsen for help
with the statistical analyses and Søren B. Alnøe for technical
support and data collection. Furthermore, we thank the editors
and 2 anonymous referees for comments and improvements on
Acuña, V., C. Vilches, and A. Giorgi. 2011. As productive and
slow as a stream can be—the metabolism of a Pampean
stream. Journal of the North American Benthological Society
Aristegi, L., O. Izagirre, and A. Elosegi. 2009. Comparison of
several methods to calculate reaeration in streams, and their
eﬀects on estimation of metabolism. Hydrobiologia 635:113–
Aristegi, L., O. Izagirre, and A. Elosegi. 2010. Metabolism of
Basque streams measured with incubation chambers. Lim-
Bender, M., K. Grande, K. Johnson, J. Marra, P. J. L. Williams,
J. Sieburth, M. Pilson, C. Langdon, G. Hitchcock, J.
Orchardo, C. Blunt, P. Donaghay, and K. Heinemann. 1987.
A comparison of four methods for determining planktonic
community production. Limnology and Oceanography 32:
Bernot, M. J., D. J. Sobota, R. O. Hall, P. J. Mulholland, W. K.
Dodds, J. R. Webster, J. L. Tank, L. R. Ashkenas, L. W.
Cooper, C. N. Dahm, S. V. Gregory, N. B. Grimm, S. K.
Hamilton, S. L. Johnson, W. H. McDowell, J. L. Meyer, B.
Peterson, G. C. Poole, H. M. Valett, C. Arango, J. J. Beaulieu,
A. J. Burgin, C. Crenshaw, A. M. Helton, L. Johnson, J.
Merriam, B. R. Niederlehner, J. M. O’Brien, J. D. Potter, R. W.
Sheibley, S. M. Thomas, and K. Y. M. Wilson. 2010. Inter-
regional comparison of land-use eﬀects on stream metabo-
lism. Freshwater Biology 55:1874–1890.
Binzer, T., K. Sand-Jensen, and A. Middelboe. 2006. Community
photosynthesis of aquatic macrophytes. Limnology and Ocean-
Biggs, B. J. F., V. I. Nikora, and T. H. Snelder. 2005. Linking
scales of ﬂow variability to lotic ecosystem structure and
function. River Research and Applications 21:283–298.
Biggs, B. J. F., R. A. Smith, and M. J. Duncan. 1999. Velocity
and sediment disturbance of periphyton in headwater streams:
biomass and metabolism. Journal of the North American
Benthological Society 18:222–241.
Bott, T. L. 2006. Primary productivity and community respiration.
Pages 663–690 in F. R. Hauer and G. A. Lamberti (editors).
Methods in stream ecology. Academic Press, Amsterdam, The
Bott, T. L., J. T. Brock, C. S. Dunn, R. J. Naiman, R. W. Ovink,
and R. C. Petersen. 1985. Benthic community metabolism in
four temperate stream systems: an inter-biome comparison
and evaluation of the river continuum concept. Hydrobio-
Clapcott, J. E., and L. A. Barmuta. 2010. Forest clearance in-
creases metabolism and organic matter processes in small
headwater streams. Journal of the North American Benthol-
ogical Society 29:546–561.
Fellows, C. S., S. E. Bunn, F. Sheldon, and N. J. Beard. 2009.
Benthic metabolism in two turbid dryland rivers. Freshwater
Fellows, C. S., J. E. Clapcott, J. W. Udy, S. E. Bunn, B. D. Harch,
M. J Smith, and P. M. Davies. 2006. Benthic metabolism as
an indicator of stream ecosystem health. Hydrobiologia 572:
Fellows, C. S., H. M. Valett, and C. N. Dahm. 2001. Whole-
stream metabolism in two montane streams: contribution of
the hyporheic zone. Limnology and Oceanography 46:523–
Fuss, C. L., and L. A. Smock. 1996. Spatial and temporal varia-
tion of microbial respiration rates in a blackwater stream.
Freshwater Biology 36:339–349.
Hall, R. O., and J. J. Beaulieu. 2013. Estimating autotrophic
respiration in streams using daily metabolism data. Freshwa-
ter Science 32:507–516.
Volume 35 September 2016 | 843
Hauer, F. R., and G. A. Lamberti. 2006. Methods in stream
edition. Academic Press, Amsterdam, The Neth-
Hedin, L. O. 1990. Factors controlling sediment community res-
piration in woodland stream ecosystems. Oikos 57:94–105.
Houser, J. N., P. J. Mulholland, and K. O. Maloney. 2005.
Catchment disturbance and stream metabolism: patterns in
ecosystem respiration and gross primary production along a
gradient of upland soil and vegetation disturbance. Journal
of the North American Benthological Society 24:538–552.
Jones, J. I., A. L. Collins, P. S. Naden, and D. A. Sear. 2011. The
relationship between ﬁne sediment and macrophytes in rivers.
River Research and Applications 28:1006–1018.
Kaenel, B. R., H. Buehrer, and U. Uehlinger. 2000. Eﬀects of
aquatic plant management on stream metabolism and oxy-
gen balance in streams. Freshwater Biology 45:85–95.
Kleeberg, A., J. Köhler, T. Sukhodolova, and A. Sukhodolov.
2010. Eﬀects of aquatic macrophytes on organic matter de-
position, resuspension and phosphorus entrainment in a
lowland river. Freshwater Biology 55:326–345.
Koopmans, D. J., and P. Berg. 2015. Stream oxygen ﬂux and me-
tabolism determined with the open water and aquatic eddy co-
variance techniques. Limnology and Oceanography 60:1344–
Lampert, W. 1984. The measurement of respiration. Page 413–
468 in J. A. Downing and F. H. Rigler (editors). A manual
on methods for the assessment of secondary productivity in
fresh waters. 2
edition. Blackwell Scientiﬁc, Oxford, UK.
Leggieri, L., C. Feijoó, A. Giorgi, N. Ferreiro, and V. Acuña.
2013. Seasonal weather eﬀects on hydrology drive the metabo-
lism of non-forest lowland streams. Hydrobiologia 716:47–58.
Madsen, J. D., M. S. Adams, and P. Ruﬃer. 1988. Harvest as a
control of sago pondweed (Potamogeton pectinatus L.) in
Badﬁsh Creek, Wisconsin: frequency, eﬃciency and its im-
pact on stream community oxygen metabolism. Journal of
Aquatic Plant Management 26:20–25.
Marzolf, E. R., P. J. Mulholland, and A. D. Steinman. 1994.
Improvements to the diurnal upstream–downstream dissolved
oxygen change technique for determining whole-stream me-
tabolism in small streams. Canadian Journal of Fisheries and
Aquatic Sciences 51:1591–1599.
Marzolf, E., P. Mulholland, and A. D. Steinman. 1998. Reply:
Improvements to the diurnal upstream–downstream dissolved
oxygen change technique for determining whole-stream me-
tabolism in small streams. Canadian Journal of Fisheries and
Aquatic Sciences 55:1786–1787.
Minshall, G. W., R. C. Petersen, T. L. Bott, C. E. Cushing, K. W.
Cummins, R. L. Vannote, and J. R. Sedell. 1992. Stream eco-
system dynamics of the Salmon River, Idaho: an 8
system. Journal of the North American Benthological Society
Mulholland, P. J. , C. S. Fellows, J. L. Tank, N. B. Grimm, J. R.
Webster, S. K. Hamilton, E. Martí, L. Ashkenas, W. B.
Bowden, W. K. Dodds, W. H. McDowell, M. J. Paul, and B. J.
Peterson. 2001. Inter-biome comparison of factors control-
ling stream metabolism. Freshwater Biology 46:1503–1517.
O’Brien, J. M., J. L. Lessard, D. Plew, S. E. Graham, and A. R.
McIntosh. 2014. Aquatic macrophytes alter metabolism and
nutrient cycling in lowland streams. Ecosystem 17:405–417.
Odum, H. T. 1956. Primary production in ﬂowing waters. Lim-
nology and Oceanography 1:102–117.
Owens, M. 1974. Measurements on non-isolated natural com-
munities in running waters. Pages 111–119 in R. A. Vollen-
weider (editor). A manual on methods for measuring primary
production in aquatic environments. Blackwell Scientiﬁc, Ox-
Owens, M., R. W. Edwards, and J. W. Gibbs. 1964. Some reaer-
ation studies in streams. Limnology and Oceanography 18:
Prahl, C., E. Jeppesen, K. Sand-Jensen, and T. Moth-Iversen.
1991. A continuous-ﬂow system for measuring in vitro oxy-
gen and nitrogen metabolism in separate stream communi-
ties. Freshwater Biology 26:495–506.
Pomazkina, G., L. Kravtsova, and E. Sorokovikova. 2012. Struc-
ture of epiphyton communities on Lake Baikal submerged
macrophytes. Limnological Review 12:19–27.
Reid, M. A., M. C. Thoms, and F. J. Dyer. 2006. Eﬀects of
spatial and temporal variation in hydraulic conditions on me-
tabolism in cobble bioﬁlm communities in an Australian up-
land stream. Journal of the North American Benthological
Rier, S. T., and D. K. King. 1996. Eﬀects of inorganic sedimen-
tation and riparian clearing on benthic community metabo-
lism in an agriculturally-disturbed stream. Hydrobiologia
Sand-Jensen, K. 1998. Inﬂuence of submerged macrophytes on
sediment composition and near-bed ﬂow in lowland streams.
Freshwater Biology 39:663–679.
Thyssen, N., M. Erlandsen, E. Jeppesen, and C. Ursin. 1987. Re-
aeration of oxygen in shallow, macrophyte rich streams. I. De-
termination of the reaeration coeﬃcient. Internationale Revue
der gesamten Hydrobiologie und Hydrographie 72:405–429.
Tunca, H., T. Ongun Sevíndík, D. N. Bal, and S. Arabaci. 2014.
Community structure of epiphytic algae on three diﬀerent
macrophytes at Acarlar ﬂoodplain forest (northern Turkey).
Chinese Journal of Oceanology and Limnology 32:845–857.
Van, T. K., W. T Haller, and G. Bowes. 1976. Comparison of
the photosynthetic characteristics of three submersed aquatic
plants. Plant Physiology 58:761–768.
Webster, J. R., and H. M. Valett. 2006. Solute dynamics. Pages 169–
185 in F. R. Hauer and G. A. Lamberti (editors). Methods in
stream ecology. Academic Press, Amsterdam, The Netherlands.
Young, R. G., A. D Huryn, E. R. Marzolf, P. J. Mulholland, and
A. D. Steinman. 1998. Comment: Improvements to the diur-
nal upstream–downstream dissolved oxygen change tech-
nique for determining whole-stream metabolism in small
streams. Canadian Journal of Fisheries and Aquatic Sciences
844 | Habitat metabolism in streams A. B. Alnoee et al.