ArticlePDF Available

Invertebrate grazing and epilithon assemblages control benthic nitrogen fixation in an N-limited river network

Authors:
  • Saint Catherine University, St. Paul, MN

Abstract and Figures

The effects of top-down (e.g., herbivory) and bottom-up (e.g., nutrient supply) processes on primary producers are often interdependent. In stream ecosystems, interactions between herbivorous grazers and physical factors, such as light and temperature, can alter the abundance and taxonomic composition of epilithic nitrogen (N) fixers. To examine how grazing and physical factors mediate the source of N to stream ecosystems, we conducted an in-situ grazer exclusion experiment by removing crawling invertebrate grazers from epilithon-covered rocks in 3 streams with varying drainage areas, representing a gradient of temperature and light levels, within a northern California river network. After ∼1 mo of grazer exclusion, we measured epilithon biomass and composition, N2 fixation, and ammonium (NH4), nitrate (NO3), and phosphate (PO4) uptake rates. Epilithic biomass, N2-fixation rates, and N and phosphorus uptake rates differed among the 3 streams, and rates were highest in the largest, open-canopy stream. Increases in N2-fixation rates with stream size were due to higher nitrogenase activity per unit of biomass as well as higher absolute biomass of N2 fixers. The presence of grazers interacted with physical factors to control nutrient fluxes. Ecologically-significant grazer removal effects only occurred in the largest stream, where grazing increased the influx of atmospheric N2 to the benthic biofilm. N2-fixation rates increased with grazing while NH4 uptake rates decreased by a similar proportion, shifting the predominant N source from the assimilation of dissolved N to atmospheric N2 via fixation by cyanobacteria. By altering the balance between N2 fixation and water column N uptake, grazers can mediate the flux of N to stream ecosystems.
Content may be subject to copyright.
Invertebrate grazing and epilithon assemblages control
benthic nitrogen xation in an N-limited river network
Brooke L. Weigel
1,3
, Jill R. Welter
2,4
, and Paula C. Furey
2,5
1
Committee on Evolutionary Biology, University of Chicago, Culver Hall 402, 1025 East 57
th
Street, Chicago, Illinois 60637 USA
2
Department of Biology, St Catherine University, 2004 Randolph Avenue, St Paul, Minnesota 55105 USA
Abstract: The effects of top-down (e.g., herbivory) and bottom-up (e.g., nutrient supply) processes on primary pro-
ducers are often interdependent. In stream ecosystems, interactions between herbivorous grazers and physical fac-
tors, such as light and temperature, can alter the abundance and taxonomic composition of epilithic nitrogen (N)
xers. To examine how grazing and physical factors mediate the source of N to stream ecosystems, we conducted an
in-situ grazer exclusion experiment by removing crawling invertebrate grazers from epilithon-covered rocks in
3 streams with varying drainage areas, representing a gradient of temperature and light levels, within a northern Cal-
ifornia river network. After 1 mo of grazer exclusion, we measured epilithon biomass and composition, N
2
xation,
and ammonium (NH
4
), nitrate (NO
3
), and phosphate (PO
4
) uptake rates. Epilithic biomass, N
2
-xation rates, and N
and phosphorus uptake rates differed among the 3 streams, and rates were highest in the largest, open-canopy
stream. Increases in N
2
-xation rates with stream size were due to higher nitrogenase activity per unit of biomass
as well as higher absolute biomass of N
2
xers. The presence of grazers interacted with physical factors to control
nutrient uxes. Ecologically-signicant grazer removal effects only occurred in the largest stream, where grazing
increased the inux of atmospheric N
2
to the benthic biolm. N
2
-xation rates increased with grazing while
NH
4
uptake rates decreased by a similar proportion, shifting the predominant N source from the assimilation of
dissolved N to atmospheric N
2
via xation by cyanobacteria. By altering the balance between N
2
xation and water
column N uptake, grazers can mediate the ux of N to stream ecosystems.
Key words: nitrogen xation, herbivory, grazers, cyanobacteria, nutrient uxes, epilithon, Rhopalodiaceae,
Epithemia, consumerresource interactions, drainage area
In aquatic ecosystems, primary production is controlled by
interactions between bottom-up factors, such as the avail-
ability of nutrients and light (Warren et al. 2017), as well as
top-down consumption by grazers (Power et al. 1985, Hill
and Knight 1988, Hillebrand 2002, Warren et al. 2017).
Rather than acting in isolation, these bottom-up and top-
down effects are often interdependent. For example, when
conditions are favorable for nitrogen (N
2
)xation (i.e., high
light and temperature, N limitation), primary producers
that x atmospheric N
2
into ammonia (NH
3
) can provide
a vital source of N to stream ecosystems (Grimm and Pe-
trone 1997, Marcarelli et al. 2008, Kunza and Hall 2014,
Williamson et al. 2016). Grazers can also directly inuence
the supply of N to primary producers by altering the bio-
mass, relative abundance, and species composition of N
2
-
xing algae and cyanobacteria (Feminella and Hawkins
1995, Chan et al. 2006, Arango et al. 2009). Despite the
global prevalence of N-poor freshwater and marine ecosys-
tems (Elser et al. 2007), few studies have examined how
grazers may mediate the ux of N to stream ecosystems
by altering benthic N
2
xation and dissolved nutrient up-
take rates.
Physical factors that change with stream drainage area,
including light, temperature, and hydrology, mediate pri-
mary production (Fig. 1A;Vannote et al. 1980). These abi-
otic factors can also exert a strong inuence on N
2
-xation
rates, as these rates increase with temperature, light avail-
ability, and decreasing dissolved inorganic nitrogen (DIN)
concentrations (Fig. 1B; Grimm and Petrone 1997, Kunza
and Hall 2014, Williamson et al. 2016). The increase in ben-
thic N
2
-xation rates with stream temperature is the re-
sult of temperature dependence of the nitrogenase enzyme,
which is responsible for N
2
xation (Welter et al. 2015).
N
2
xation is also energetically expensive and may be limited
E-mail addresses:
3
brookeweigel@uchicago.edu;
4
jrwelter@stkate.edu;
5
pcfurey@stkate.edu
DOI: 10.1086/710023. Received 29 October 2018; Accepted 19 January 2020; Published online 10 July 2020.
Freshwater Science. 2020. 39(3):000000. © 2020 by The Society for Freshwater Science. 000
This content downloaded from 073.075.149.002 on July 13, 2020 17:42:56 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
by the availability of xed carbon, even for autotrophic N
2
xers including photosynthetic cyanobacteria (Rabouille
et al. 2006). As a result, greater light availability leads to higher
N
2
-xation rates (Lewis and Levine 1984). Light availability
in forested catchments typically increases as stream chan-
nels widen and tree canopy cover decreases (Vannote et al.
1980, Warren et al. 2017). Therefore, larger drainage areas
may support higher rates of N
2
xation. In open-canopy des-
ert streams, where temperature and light availability are
high, N
2
-xation rates can be substantial but still exhibit var-
iation with light availability and time of day (Grimm and
Petrone 1997). Light and temperature also interact with wa-
ter column N availability to control rates of N
2
xation. High
DIN concentrations can suppress N
2
xation (Marcarelli
and Wurtsbaugh 2007, Kunza and Hall 2014), and variability
in DIN concentrations within catchment ecosystems can fur-
ther inuence N
2
xation and watershed-scale nutrient cycling.
In addition to the physical factors that change with
stream size (Finlay et al. 2011), grazer assemblages (Holo-
muzki et al. 2013) and the effects of grazing on primary pro-
duction can vary withstream size (Fig. 1A, C; Feminella et al.
1989, McNeely and Power 2007). Grazers can alter biolm
physical structure by consuming detritus and dislodging
sediment, which can increase light and nutrient availability
to understory epilithon (Lamberti and Resh 1983, Power
1990, Liess and Kahlert 2009). This increase in light avail-
ability can increase N
2
-xation rates (Grimm and Petrone
1997, Rabouille et al. 2006). However, this effect varies with
stream canopy cover. In streams with dense forest canopy
cover, primary production is often light limited (Hill and
Knight 1988, McNeely and Power 2007), and grazers may
not substantially alter primary producer biomass or have the
potential to increase light availability. Temperature can also
alter grazer feeding preferences, which in turn inuences
epilithon biomass and taxonomic structure. Gordon et al.
(2018) found that snails and blackylarvaebecomemore
selective feeders in warmer streams, favoring high-prole
epilithic diatoms and motile diatoms. In N-poor streams,
N
2
-xing taxa may become more abundant as higher
temperatures and light availability release energetic con-
straints, but an increase in selective consumption by graz-
ers may alter this response. The outcome is dependent on
the interaction between the physical abiotic environment,
which determines the potential for N
2
xation, and grazing
preferences for different epilithon taxa, including N
2
xers
(Fig. 1AD).
There are also direct feedbacks between grazers, epili-
thon assemblage structure, and nutrient cycling in streams
(Fig. 1D). Low water column DIN favors N
2
-xing taxa
(Grimm and Petrone 1997, Marcarelli et al. 2008), but nutri-
ent regeneration by grazers can reduce the competitive ad-
vantage of N
2
-xing taxa by excreting DIN (Elser and Urabe
1999, Gettel et al. 2007). Selective consumption by grazers
can also shape the taxonomic composition of epilithon
(Feminella and Hawkins 1995). Diverse grazer taxa includ-
ing chironomids (Furey et al. 2012), mayies, and caddis-
ies (Peterson et al. 1998), algivorous minnows (Power et al.
1988), and snails (Steinman et al. 1991) preferentially con-
sume diatoms over cyanobacteria. By altering the relative
proportion of N
2
-xing and non-N
2
-xing taxa, selective
grazing could have strong impacts on N
2
-xation rates
(Hambright et al. 2007). In addition, Kunza and Hall (2014)
found that rates of DIN uptake and N
2
xation were inversely
related in stream epilithon assemblages, thus, any selective
grazing that changes the relative proportion of N
2
-xing
and non-N
2
-xing taxa may also inuence DIN uptake
rates. In this way, grazers could shift the balance of primary
N inputs between DIN uptake and atmospheric N
2
xation,
with potential consequences for the retention of limiting
nutrients and downstream export of N.
This study investigated how grazing, light availability,
and temperature inuence epilithon assemblage structure
and biomass, N
2
xation, and water column N and phos-
phorus (P) uptake rates. We conducted an experimental
grazer removal within the South Fork Eel River and 2 of its
smaller tributaries to assess 1) how grazers affect epilithon
assemblages, nutrient uptake, and N
2
-xation rates, and
2) whether the effects of grazing differ among streams of
varying size, drainage area, and light availability. Our study
provides insight into how grazers and physical drivers may
interact to affect ecosystem-scale nutrient cycling in N-
limited river networks.
Figure 1. Physical drivers, mediated by stream size, interact
with community dynamics to inuence biogeochemical cycling
in river networks. Physical drivers can interact directly with
community dynamics (A), or with biogeochemical cycling (B).
Grazers can interact with epilithon assemblages to alter pro-
ductivity or assemblage composition (C), which can in turn al-
ter biogeochemical cycling (D).
000 | Grazing and epilithon control N
2
xation B. L. Weigel et al.
This content downloaded from 073.075.149.002 on July 13, 2020 17:42:56 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
METHODS
Study site
We examined grazer-mediated shifts in N uxes to epi-
lithon in 3 streams with varying drainage areas, each with
different light and temperature regimes. To experimentally
investigate how interactions among stream physical attri-
butes and grazer activity inuence epilithon assemblages,
nutrient uptake, and N
2
-xation rates, we manipulated
grazers in rife areas within the South Fork Eel River and
2 of its tributaries at the Angelo Coast Range Reserve
(lat 397440N, long 1237390W) in Mendocino County, Cali-
fornia, USA (Fig. 2). Thisriver experiences a Mediterranean
climate with rainy winters followed by dry and sunny sum-
mers characterized by low ow and high biological activity
(Power et al. 2008, 2013). In the Eel River basin, ambient
N availability is low relative to P, and N limits primary pro-
duction (Hill and Knight 1988, Power 1991, Schade et al. 2011).
Study sites included 3 streams of different drainage areas:
Fox Creek, a small, shaded tributary of the Eel River (drain-
age area 2 km
2
); Elder Creek, a larger tributary (17 km
2
);
and the South Fork Eel River, a highly-productive stream
(120 km
2
) (Fig. 2). The 2 larger streams support high levels
of autotrophic production, whereas production in the small-
est stream may be light limited (Hill and Knight 1988). We
did not have replication of streams with similar drainage
areas, thus we were unable to directly test for the variation
attributable to drainage area alone.
In this river network, autotrophic epilithon contains di-
atoms (Bacillariophyta), lamentous green algae including
Cladophora glomerata (L.) Kütz. 1843, and both N
2
-xing
and non-N
2
-xing cyanobacteria. Within the diatoms,a sub-
set that belong to the Rhopalodiaceae (Epithemia Kütz. 1844
and Rhopalodia O.Müll. 1895) (Furey et al. 2012) contain
endosymbiotic cyanobacteria capable of N
2
xation, hereaf-
ter referred to as N
2
-xing diatoms. Dominant grazers in-
clude stone-cased caddisies (Dicosmoecus gilvipes Hagen
1875, Glossosoma Curtis 1834, Neophylax McLachlan 1871),
mayy larvae (Heptageniidae and Baetidae), and midges
(Chironomidae) (McNeely and Power 2007).
Experimental grazer exclusion
We excluded benthic grazers by elevating epilithon-
covered rocks above the streambed on articial platforms
(42 33 14 cm high), which made rocks on the platform
inaccessible to non-swimming and low-drifting grazers
(Feminella et al. 1989). We lined platform edges with petro-
leum jelly to further reduce rock accessibility to the crawling
stone-cased caddisyD. gilvipes (McNeely and Power 2007).
We placed submerged platforms in an area representative
of local water velocity and randomly selected 7 epilithon-
covered rocks from the surrounding stream rifetobe
cleared of invertebrate grazers and placed on the surface
of the platforms (hereafter referred to as the ungrazed treat-
ment). In previous experiments, elevated platforms effec-
tively excluded crawling grazers but less effectively reduced
swimming grazers like mayies (Feminella et al. 1989). To
minimize the presence of mayies, an observer with a snor-
kel mask removed visible grazers from the rocks with soft
forceps at least once every 3 d. Intermittent grazing by
swimming grazers likely occurred between removals, but
the experimental platforms prevented access to large crawl-
ing grazers such as D. gilvipes, which have a large effect on
algal biomass in the Eel River (Wootton et al. 1996). Grazer
removal treatments lasted 24 d (7 July31 July 2010) in the
largest stream, 33 d (July 9Aug 11) in the intermediate
stream, and 30 d (July 14Aug 13) in the smallest stream.
At the end of 4 wk, we randomly selected 7 control rocks
from within the same rife area to represent ambient graz-
ing levels (hereafter referred to as the grazed treatment).
We added only 2 grazer exclusion platforms in each stream
(Fig. S1). To avoid pseudo-replication, we did all statistical
analyses using mean values obtained from rocks selected
from each of the 2 experimental platforms in each stream
(3 rocks on platform 1 and 4 rocks on platform 2). Therefore,
the replication within each stream was n52 for the un-
grazed treatment and n57 for the grazed treatment.
Nutrient uptake assay
We conducted nutrient uptake assays for 4 h during
maximum daylight. We incubated individual grazed and
Figure 2. Location of study sites (black dots) in the South
Fork Eel River watershed within the Angelo Coast Range Re-
serve in Mendocino County, California, USA. This map
was adapted from McNeely and Power (2007).
Volume 39 September 2020 | 000
This content downloaded from 073.075.149.002 on July 13, 2020 17:42:56 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
ungrazed rocks in clear recirculating chambers (circula-
tion provided by a magnetic stir bar) in a ow-through
system located on the adjacent streambed, which main-
tained ambient stream temperature and consistent light
availability across chambers. We measured nutrient up-
take by adding a combined nutrient cocktail of ammo-
nium(NH
4
),nitrate (NO
3
), and phosphate (PO
4
) with tar-
get concentrations of 20 lg/L for each nutrient, and we
sampledthe chamber water at 0, 1, 2, and 4 h. We ltered
water samples in the eld with 0.7-lmglassber lters
(GF/F; Whatman
®
, Maidstone, United Kingdom) and ana-
lyzed NH
4
-N and PO
4
-P concentrations immediately fol-
lowing sample collection. We analyzed NH
4
-N via the uo-
rimetric method (Holmes et al. 1998, Taylor et al. 2007)
and PO
4
-P concentrations with molybdate colorimetric anal-
ysis (Murphy and Riley 1962). Prior to analysis, we kept
NO
3
-N samples frozen on a Lachat Quikchem 8000 auto-
analyzer (Lachat Instruments, Loveland, Colorado). We cal-
culated areal nutrient uptake rates (lgnutrientm
22
h
21
)with
the following equation:
Nutrientuptake lg nutrient m22h21

5BN0VðÞ
A
(Eq. 1),
where Bis the slope of the regression between the natural
log-transformed nutrient concentration and time for each
individual chamber, N
0
is the initial nutrient concentration
(lg/L) in each chamber, Vis volume (L), and Ais rock sur-
face area (m
2
). We calculated biomass-specic nutrient up-
take rates (lgnutrientmg
21
ash-free dry mass [AFDM] h
21
)
by dividing uptake rates by the AFDM associated with each
rock rather than the surface area. We completed nutrient
uptake assays the day prior to N
2
-xation assays, and we
placed rocks back in the stream in open-top ow-through
buckets overnight between assays.
N
2
-xation assay
We measured in-situ N
2
xation with an acetylene re-
duction assay (Flett et al. 1976, Capone 1993) by introduc-
ing acetylene to gas-tight septum-equipped chambers (2 L,
containing the epilithon-covered rock and lled with stream
water) after reacting calcium carbide with water. We shook
chambers for 5 min to ensure homogeneous mixing of the
acetylene. We collected gas samples at 0 and 2 h and ana-
lyzed them within 12 h on a gas chromatograph (model
8610; SRI Instruments, Torrance, California) equipped with
aame ionization detector (HayeSep
®
T, 80/100 mesh). We
converted rates of ethylene production to areal N
2
-xation
rates (lgN
2
xed m
22
h
21
) using a ratio of 31 moles of
acetylene reduced:N
2
xed and dividing byrock surface area
(Capone 1993). For biomass-specicN
2
-xation rates, we
divided by AFDM (lgN
2
xed mg
21
AFDM h
21
).
Epilithic biomass and assemblage composition
We analyzed homogenized subsamples of rock epilithon
for AFDM, algal taxonomic composition, and cell biovol-
ume. To calculate AFDM, we used mass loss on ignition
(4 h at 5507C after drying at 607C for 72 h). We preserved
additional subsamples in 3% formaldehyde for algal cell
counts. For estimates of taxonomic composition, we counted
a minimum of 10 elds of view or 200 cells (or counting
units) with a Palmer Cell (#1803-B30; Wildlife Supply Com-
pany, Yulee, Florida) on a Labophot Microscope (Nikon,
Tokyo, Japan) at 400magnication. We determined epili-
thon cell biovolume by counting the number of overlapping
intersections on a 10 10 Whipple grid for each taxon in
the eld of view. For enumeration and biovolume determi-
nation, we partitioned epilithon assemblages into distinct
algal or cyanobacteria taxa and functional groups to sepa-
rate N
2
-xing taxa from non-N
2
-xing taxa and to group
less common taxa together (Table S1). We identied dia-
toms to the genus or species level according to Diatoms of
North America (https://diatoms.org/) and Krammer and
Lange-Bertalot (1991).
Statistical analysis
We used 2-way analysis of variance (ANOVA) tests to
examine the effects of stream and grazing treatment, as well
as their interaction, on AFDM, total epilithon cell biovol-
ume, and both areal (m
2
) and biomass-specic (mg AFDM)
N
2
xation and nutrient (DIN, NH
4
,NO
3
, and PO
4
) uptake
rates. Despite the variation in sample sizes between the un-
grazed treatment and the control (grazed) treatment in each
stream, all tests met the assumption of homogeneity of vari-
ance for ANOVA using Bartletts test (grazed vs ungrazed
treatment: areal N
2
xation, Bartletts K-squared 52.5, p5
0.11; biomass-specicN
2
xation, Bartletts K-squared 51.2,
p50.28; NH
4
uptake, Bartletts K-squared 50.82, p5
0.36; NO
3
uptake, Bartletts K-squared 52.5, p50.11; PO
4
uptake, Bartletts K-squared 50.07, p50.80; DIN uptake,
Bartletts K-squared 50.12, p50.73). For signicant
ANOVA outcomes, we conducted Tukeyshonestsigni-
cant difference (HSD) pairwise tests with a post-correction
experiment-wide error rate of 0.05. We completed ANOVA
tests and graphical visualizations in R (version 3.4.4; R Proj-
ect for Statistical Computing, Vienna, Austria).
For epilithon assemblage-level comparisons among streams
and between grazer treatments, we generated BrayCurtis
similarity matrices from square root-transformed abundance
and biovolume matrices and visualized them with non-metric
multidimensional plots. We used a 2-way permutational mul-
tivariate analysis of variance (PERMANOVA) with stream
and grazer treatment as xed factors to test for signicant
differences in epilithon assemblage structure among the
3 streams and between treatments. The PERMANOVA test
statistic, pseudo-F, is a multivariate analog to the ANOVA
F-ratio, and it compares the variability within groups to
000 | Grazing and epilithon control N
2
xation B. L. Weigel et al.
This content downloaded from 073.075.149.002 on July 13, 2020 17:42:56 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
the variability among groups (Anderson 2001). We com-
pleted all multivariate analyses in PRIMER (version 6.1.11;
PRIMER-e, Auckland, New Zealand).
RESULTS
Physical stream characteristics
During the experimental grazer removal, mean stream
temperature (7C) was lowest in the smallest tributary (15.7 ±
0.9), higher in the larger tributary (16.3 ±1.4), and highest
in the South Fork Eel River (19.1 ±1.1) (Table 1). Incuba-
tion temperatures during the N
2
-xation assays were higher
than the mean stream temperature at each site, likely re-
ecting the inuence of solar heating on the recirculating
water inside the chambers. However, the trend of increas-
ing temperature with stream size remained the same (Ta-
ble 1). Light intensity during the in situ N
2
-xation assay
conducted in the largest stream was 10 and 23higher
than light intensity for the streams of intermediate and
small size, respectively. The light intensity that the interme-
diate stream received was 2higher than the smallest stream
(Table 1). DIN concentrations were generally higher in the 2
larger streams compared to the smallest stream (Table 1).
N
2
xation
Areal N
2
-xation rates varied greatly among the 3
streams (2-way ANOVA, p54.33 10
29
; Table 2, Fig. 3).
N
2
-xation rates were 3 and 8higher in the largest
stream (859 ±115 lgNm
22
h
21
)thanintheintermediate
(312 ±33.4) and small (105 ±7) streams (Table 2). N
2
-xation
rates did not differ between the intermediate and small streams
(Table 2). Biomass-specicN
2
-xation rates also varied
among the 3 streams (2-way ANOVA, p50.03; Table 2),
with signicantly higher rates in the 2 larger streams than
in the smallest stream (Table 2).
Grazer removal did not affect areal N
2
-xation rates (2-
way ANOVA, p50.18; Table 2). However, the interaction
between stream and grazer removal explained some of the
variation in areal N
2
-xation rates (2-way ANOVA, p5
0.047; Table 2). Pairwise tests for this interaction term re-
vealed that grazer removal affected N
2
-xation rates only
in the largest stream, where the grazed treatment exhibited
a higher areal N
2
-xation rate (Tukeys HSD pairwise com-
parison, p50.06). The magnitude of this response was
large; grazing enhanced areal N
2
-xation rates by 73% rela-
tive to the ungrazed treatment (grazed treatment mean 5
1088 ±119 lgNm
22
h
21
; ungrazed treatment mean 5
629 ±159) (Fig. 3). Finally, neither grazer nor stream*grazer
effects altered biomass-specicN
2
xation (Table 2).
Nutrient uptake
Areal uptake rates of all inorganic nutrients were signif-
icantly different among the 3 streams (2-way ANOVA, p5
0.001 for NH
4
,p50.008 for NO
3
,p50.002 for PO
4
; Ta-
ble 2). Areal uptake rates of DIN, NH
4
, and PO
4
were
2.5higher in the largest stream compared to the
2 smaller streams, and these differences were signicant
(Table 2, Fig. 4). The areal uptake rate of NO
3
was 15
higher in the 2 larger streams compared to the smallest
stream (Table 2, Fig. 4), while the biomass-specic uptake
rate of NO
3
was 5higher in the stream of intermediate
size compared to the larger and smaller streams (Table 2).
Biomass-specic uptake rates of DIN, NH
4
, and PO
4
did
not vary among streams (Table 2).
The interaction between stream and grazer removal ex-
plained a substantial amount of the variation in biomass-
specicNH
4
uptake (2-way ANOVA, p50.02) but not
areal NH
4
uptake (2-way ANOVA, p50.22) (Table 2). In
the largest stream, biomass-specicNH
4
uptake was 4
lower in the grazed treatment (TukeysHSDpairwisecom-
parison, p50.06). Similarly, areal NH
4
uptake was 1.6
lower in the grazed treatment of the largest stream. Grazed
areal NH
4
-uptake rates averaged 915 ±224 lgNm
22
h
21
,
Table 1. Summary of physical stream characteristics including drainage area and mean (±standard deviation) temperature, light in-
tensity, and nutrient concentrations. Temperatures are means of 1) in-stream temperatures from HOBO data loggers (Onset
®
,
Bourne, Massachusetts) installed at each site and recorded for the duration of the experimental grazer exclusion, and 2) temperatures
recorded with a handheld YSI (Yellow Springs Instruments, Yellow Springs, Ohio) temperature probe from chambers during
nitrogen (N
2
)-xation assays. Light intensity was measured from HOBO data loggers installed adjacent to the chambers during N
2
-
xation assays. Nutrient concentrations represent means of the ambient stream water nutrient concentrations during the nutrient
uptake assays. NH
4
-N 5ammonium, NO
3
-N 5nitrate, PO
4
-P 5phosphate.
Characteristic Fox Creek Elder Creek South Fork Eel River
Drainage area (km
2
) 2.2 16.9 120
Stream temperature (7C) 15.7 ±0.9 16.3 ±1.4 19.1 ±1.1
Temperature (7C) during N
2
-xation assays 15.7 ±0.1 17.7 ±0.3 21.4 ±0.4
Light intensity (lux) during N
2
-xation assays 9910 ±8226 21,325 ±47,219 232,197 ±39,001
NO
3
-N (lg/L) 5.47 ±3.23 17.63 ±11.01 14.02 ±5.75
NH
4
-N (lg/L) 2.12 ±0.83 1.55 ±0.80 14.42 ±6.35
PO
4
-P (lg/L) 21.61 ±7.26 24.78 ±2.55 15.61 ±1.14
Volume 39 September 2020 | 000
This content downloaded from 073.075.149.002 on July 13, 2020 17:42:56 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
while ungrazed uptake rates averaged 1495 ±157 (Fig. 4).
This difference (580 lgNm
22
h
21
) was similar in magni-
tude to the increase in N
2
-xation rate in the grazed treat-
ment (459 lgNm
22
h
21
) from the largest stream. NH
4
-
uptake rates responded to grazer removal, but grazers or
stream*grazer interactions did not affect areal or biomass-
specic uptake rates of DIN, NO
3
,andPO
4
(Table 2).
AFDM and total algal biovolume
AFDM varied among the 3 streams (2-way ANOVA, p5
0.004; Table 2, Fig. 5A). The largest stream contained 2.6
more epilithic AFDM than the 2 smaller streams (Fig. 5A).
While there were no grazer effects or stream*grazer interac-
tions (2-way ANOVA, p50.74 for grazers, p50.13 for
stream*grazer interactions; Table 2), the presence of grazers
notably doubled AFDM relative to the ungrazed treatment
in the largest stream (Fig. 5A). Epilithon cell biovolume dis-
played the same trends as AFDM (Fig. 5B). Biovolume was
2higher in the largest stream relative to the 2 smaller
streams, and there were no grazer effects or stream*grazer
interactions (Table 2). Epilithon cell biovolume was also
twice as high in the presence of grazers in the largest stream
(Fig. 5B).
Epilithon assemblage structure and composition
The taxonomic composition of epilithon, based on cell
biovolume or abundances of 26 algal or cyanobacteria
taxa and functional groups, varied among the 3 streams
(PERMANOVA, p50.001), with the factor of stream
explaining 22.9 and 39.5% of the total variation in
biovolume-based and abundance-based epilithon assem-
blages, respectively (Table 3, Fig. 6). Diatoms dominated
epilithon assemblages in all 3 streams, including the non-
N
2
-xing diatom Cocconeis Ehrenb. 1837 spp., comprising
8 to 38% of epilithon cell biovolume, and the N
2
-xing di-
atom Epithemia spp., which constituted 10 to 25% of epi-
lithon cellbiovolume (Fig.7A, Table S2). The largest stream
had the highest absolute biovolume of diatoms, including
Table 2. Statistical results for 2-way analysis of variance tests (residual degrees of freedom [df ] 521) examining the effects of grazing
(df 51), stream (df 52), and stream*grazing (df 52) on ash-free dry mass (AFDM), total epilithon cell biovolume, and both areal
(m
2
) and biomass-specic (mg AFDM) nitrogen (N
2
)xation, dissolved inorganic nitrogen (DIN-N), ammonium (NH
4
-N), nitrate
(NO
3
-N), and phosphate (PO
4
-P) uptake rates. Signicant Tukeys honest signicant difference pairwise tests for streams are indi-
cated in the rightmost column, where streams are listed by drainage area (2, 17, 120 km
2
) and differences among streams are indicated
with equality signs: <(less than), >(greater than), 5(equal to). Signicant tests are bolded, and NS 5not signicant.
Response variable Grazing effect Stream*grazing effect Stream effect Pairwise stream effect
Areal N
2
xation F51.95 F53.56 F555.2 2 517 <120
p50.17 p50.047 p54.33 10
29
N
2
xation per AFDM F50.60 F50.81 F54.18 2 <17 5120
p50.44 p50.46 p50.03
ADFM F50.12 F52.25 F57.37 2 517 <120
p50.74 p50.13 p50.004
Epilithon cell biovolume F50.13 F51.83 F511.43 2 517 <120
p50.73 p50.18 p50.0004
Areal DIN-N uptake F50.39 F50.09 F54.91 2 517 <120
p50.54 p50.91 p50.02
Areal NH
4
-N uptake F51.36 F51.63 F59.62 2 517 <120
p50.25 p50.22 p50.001
Areal NO
3
-N uptake F50.12 F50.22 F56.23 2 <17 5120
p50.73 p50.81 p50.008
Areal PO
4
-P uptake F51.10 F50.60 F58.35 2 517 <120
p50.31 p50.56 p50.002
DIN-N uptake per AFDM F50.08 F52.17 F52.80 NS
p50.78 p50.14 p50.08
NH
4
-N uptake per AFDM F50.05 F54.72 F53.13 NS
p50.83 p50.02 p50.06
NO
3
-N uptake per AFDM F50.64 F51.37 F59.21 2 5120 <17
p50.43 p50.28 p50.001
PO
4
-P uptake per AFDM F50.01 F52.71 F51.46 NS
p50.92 p50.09 p50.26
000 | Grazing and epilithon control N
2
xation B. L. Weigel et al.
This content downloaded from 073.075.149.002 on July 13, 2020 17:42:56 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Epithemia spp. (Fig. 7B). Green algal biovolume was rela-
tively low in the smallest and largest streams (04% of epi-
lithon cell biovolume). However, green algal biovolume
was 8higher in the stream of intermediate size (Fig. 7A,
B). The absolute and relative cell biovolume of cyanobac-
teria was low in all 3 streams. Non-heterocystous cyano-
bacteria constituted <5% of epilithon assemblages, and the
biovolume of N
2
-xing cyanobacteria varied across streams
and treatments, comprising 1 to 18% of epilithon biovolume
(Fig. 7A, B, Table S2).
Overall, grazing explained 3.4 to 3.8% of the total varia-
tion in epilithon assemblage composition, whereas stream*
grazing interactions explained 6.2 to 7.4% of the variation
(Table 3, Fig. 6). Epilithon assemblage composition did
not differ signicantly with grazing or stream*grazing inter-
actions (PERMANOVA tests based on abundance; grazing:
p50.16; stream*grazing: p50.66; PERMANOVA tests
based on biovolume, grazing: p50.09, stream*grazing:
p50.64; Table 3), but there were notable shifts in the rel-
ative and absolute biovolume of important functional
groups of epilithic taxa between grazed and ungrazed rocks
(Fig. 7A, B). In the largest stream, the absolute cell bio-
volumes of both N
2
-xing cyanobacteria (including Rivularia
(C.Agardh) Born. et Flah.1886 spp. and other heterocyte-
bearing cyanobacteria) and non-heterocystous cyanobacteria
were higher in the grazed treatment (Fig. 7B). The bio-
volumes of N
2
-xing and non-heterocystous cyanobacteria
were 12and 6higher, respectively, in the grazed treat-
ment of the largest stream (Fig. 7B, Table S2). In addition,
the total cell biovolume of N
2
-xing Epithemia spp. was lower
in the grazed treatment of the largest stream (Fig. 7B, Table S2),
mainly due to a reduction in the biovolume of large E. turgida
cells. The relative biovolume of Cocconeis spp. was also 4.5
lower in the grazed treatment of the largest stream compared
to the ungrazed treatment (Fig. 7A, Table S2).
DISCUSSION
In this study, we tested whether interactions between
grazers and benthic epilithon assemblages are modied
by physical stream characteristics that change with steam
size, such as light availability and temperature, with resulting
impacts on N cycling. Our results demonstrate that epilithic
biomass, N
2
-xation rates, and epilithon nutrient uptake
rates increased with temperature and light availability in
the Eel River network. N
2
-xation rates peaked in the largest
open-canopy stream as a result of higher biomass-specic
N
2
-xation rates and higher total biomass and biovolume
of N
2
-xing taxa. We further investigated whether grazers
have the potential to inuence the inux of N to stream eco-
systems by shifting the community composition, biomass,
or activity of epilithic N
2
-xing assemblages. This important
Figure 3. Mean areal N
2
-xation rates between grazed and
ungrazed treatments in each stream. Stream sites are in order
of increasing drainage area: Fox Creek (2 km
2
), Elder Creek
(17 km
2
) and the South Fork Eel River (120 km
2
). Error bars in-
dicate the standard error of the mean. Different letters indicate
signicant differences among the 3 streams, and asterisks (*) in-
dicate a signicant within-stream grazer effect (see Table 2).
Figure 4. Mean areal dissolved inorganic nitrogen (DIN),
ammonium (NH
4
), nitrate (NO
3
), and phosphate (PO
4
) uptake
rates between grazed and ungrazed treatments in each steam.
Stream sites are in order of increasing drainage area: Fox Creek
(2 km
2
), Elder Creek (17 km
2
), and the South Fork Eel River
(120 km
2
). Error bars indicate the standard error of the mean.
Different letters indicate signicant differences among the
3 streams (see Table 2).
Volume 39 September 2020 | 000
This content downloaded from 073.075.149.002 on July 13, 2020 17:42:56 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
interaction had yet to be examined in the context of variable
stream sizes, and we found that grazers only inuenced N
2
-
xation and NH
4
-uptake rates in the largest stream. Ob-
served increases in N
2
-xation rates due to grazing in the
largest stream were most likely explained by a taxonomic
shift in epilithon assemblages toward a greater proportion
of N
2
-xing cyanobacteria. Here, we discuss how physical
stream characteristics and grazer effects together inuenced
the composition and activity of N
2
-xing assemblages, ulti-
mately shifting the balance between N
2
xation and water
column DIN uptake, with potential downstream impacts
on N cycling.
Interactive effects of physical stream variables
and autotrophic assemblages on nutrient cycling
In the South Fork Eel River, stream size and associated
physical characteristics inuenced nutrient uptake rates, ar-
eal N
2
-xation rates, and epilithon assemblage structure and
composition. In our study, we found increased N
2
xation
associated with stream size, and it is likely that differences
in the physical variables among the 3 streams, including tem-
perature and light availability, drove these among-site differ-
ences. Previous experiments have demonstrated increased
N
2
-xation rates by stream epilithon at elevated tempera-
tures (Marcarelli and Wurtsbaugh 2006, Williamson et al.
2016), and the nitrogenase enzyme responsible for N
2
xation is temperature dependent (Welter et al. 2015). In-
creased light availability with stream size can also result in
higher rates of N
2
xation (Grimm and Petrone 1997,
Rabouille et al. 2006). Although we did not directly test the
effects of temperature or light on N
2
xation, we did nd that
N
2
-xation rates, light availability, and temperature all in-
creased with stream size.
It is important to note that simultaneous changes in
light, temperature, and nutrient availability can generate
complex nonlinear responses in biological processes, in-
cluding N
2
xation (Marcarelli and Wurtsbaugh 2006, Kunza
and Hall 2014, Hiatt et al. 2017), which are typically mediated
through changes in species composition or biological
activity. The largest stream exhibited the highest absolute
biovolume of diatoms that house N
2
-xing cyanobacterial
endosymbionts, including Epithemia spp., which likely
contributed to the observed increase in areal N
2
xation.
Further, we observed higher biomass-specicN
2
-xation
rates in the 2 larger streams, suggesting that nitrogenase
enzyme activity was stimulated by higher temperature and
light availability (Grimm and Petrone 1997, Welter et al.
2015), in addition to the accumulation of N
2
-xer biomass.
A previous study in the South Fork Eel River found that
DIN concentrations increased markedly at a stream size
of 100 km
2
during late summer, when N
2
-xing taxa were
abundant (Finlay et al. 2011). This nonlinear increase in N
availability may be indicative of a threshold response that
occurred when conditions were appropriate to support
high rates of primary production, N
2
-xation activity, and
consequentially, N regeneration and release.
In addition to N
2
xation, areal N and P demand were
also highest in the largest stream, likely as a result of higher
epilithon biomass and cell biovolume. Areal uptake rates of
DIN, NH
4
,NO
3
, and PO
4
were 2.5higher in the largest
stream, demonstrating a higher demand for N and P in
the larger open-canopy stream. As we expected, epilithon
biomass (AFDM) and cell biovolume also increased with
stream size. Both were twice as high in the largest stream
compared to the 2 smaller streams, likely contributing to
the increased demand for inorganic nutrients. Rates of nu-
trient cycling per unit of epilithic biomass reect the ef-
ciency or cell-specicactivityofnutrientexchange.
Biomass-specic nutrient uptake rates can decrease with
increasing epilithic biomass because of shading or reduced
nutrient transport into thicker epilithon matrices (Steven-
son and Glover 1993). While biomass-specicNO
3
uptake
Figure 5. Mean ash-free dry mass (AFDM) (A), and total
epilithon cell biovolume (B) between grazed and ungrazed
treatments in each stream. Stream sites are in order of increas-
ing drainage area: Fox Creek (2 km
2
), Elder Creek (17 km
2
),
and the South Fork Eel River (120 km
2
). Error bars indicate the
standard error of the mean. Different letters indicate signicant
differences among the 3 stream sites (see Table 2).
000 | Grazing and epilithon control N
2
xation B. L. Weigel et al.
This content downloaded from 073.075.149.002 on July 13, 2020 17:42:56 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
was higher in the stream of intermediate size compared to
both larger and smaller streams, we found no difference in
biomass-specic uptake rates of DIN, NH
4
, and PO
4
among
the 3 streams. Thus, increased areal N and P demand in the
largest stream was likely caused by the increase in epilithon
biomass alone rather than a change in epilithon uptake
efciency.
Finally, we found that physical stream variables (e.g.,
light and temperature) associated with drainage area al-
tered the effects of grazing on autotrophic assemblages. In
this study, we found signicant grazer effects only in the larg-
est stream, where grazers altered epilithic biomass, assem-
blage composition, and N
2
-xation and NH
4
-uptake rates.
Light limitation in some tributaries in this river basin
(Power et al. 2013), particularly those with watershed areas
<10 km
2
(Finlay et al. 2011), may limit primary production
more so than grazers, which was likely the case in our
smallest stream. Other studies have found similar patterns.
For instance, Feminella et al. (1989) found that grazers had
no effect on epilithon biomass in a densely-shaded 2
nd
-order
stream. Research within our study basin (McNeely and Power
2007) demonstrated that grazers in watersheds >1km
2
ex-
hibited variable effects on epilithon, reducing biomass only
during periods of peak productivity. It is also possible that
temporal shifts in herbivore communities, driven by the
timing of aquatic insect emergence (Uno 2016), could pro-
duce intermittent effects of grazing on epilithon assem-
blages across the growing season. This is most certainly
the case in the South Fork Eel River drainage basin, where
scouring oods and water turbidity alter both primary
Table 3. Statistical results for permutational multivariate analysis of variance (residual degrees of freedom [df ] 521) and pairwise
tests examining the effects of stream (df 52), grazing (df 51), and stream*grazing (df 52) on epilithon assemblage structure based
on abundance and biovolume matrices. The proportion of variation explained refers to the relative importance of each factor (grazing,
stream*grazing, stream) in explaining the variation in epilithon assemblages across all 3 streams. Signicant tests are bolded.
Epilithon assemblages (abundance) Epilithon assemblages (biovolume)
Factor
Pseudo-F
(tfor pairwise) p-value
Proportion of
variation explained
Pseudo-F
(tfor pairwise) p-value
Proportion of
variation explained
Grazing 1.55 0.16 3.8% 1.66 0.09 3.4%
Stream*grazing 0.73 0.66 7.4% 0.82 0.64 6.2%
Stream 5.59 0.001 39.5% 3.16 0.001 22.9%
Pairwise tests
217 km
2
t51.80 0.004 t51.60 0.02
2120 km
2
t52.32 0.002 t51.83 0.01
17120 km
2
t52.85 0.001 t51.90 0.01
Figure 6. Non-metric multidimensional scaling ordinations, based on BrayCurtis similarities of square-root transformed epilithon
abundance matrices from 26 algal species and functional groups (Table S1). Symbols indicate ungrazed (closed) and grazed (open)
epilithon assemblages from the largest stream (square), intermediate stream (circle), and smallest stream (triangle). To account for
pseudo-replication, we did permutational analysis of variance tests with the mean assemblages (*) from each of the 2 ungrazed plat-
forms in each stream.
Volume 39 September 2020 | 000
This content downloaded from 073.075.149.002 on July 13, 2020 17:42:56 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
producer and grazer assemblages in the winter and spring
(Wootton et al. 1996, Power et al. 2013), and winter oods
control primary producer biomass, most likely by reducing
spring grazer densities (Sculley et al. 2017). Despite the pos-
sibility of temporal differences in grazer effects, we found
that grazers impacted epilithon assemblages in the largest
stream during peak summertime productivity, with impor-
tant consequences for nutrient cycling.
Cascading effects of grazing on autotrophic
assemblages and nitrogen cycling
Most studies report that grazing reduces epilithic bio-
mass (Feminella and Hawkins 1995, Hillebrand 2002),
but herbivory can increase biomass-specic productivity
(Abe et al. 2007, McNeely and Power 2007) as well as total
autotrophic biomass (McCormick 1990, Power 1990) through
indirect mechanisms. Here, we found that grazers increased
epilithon biomass in the largest stream. Grazing can increase
light availability to epilithon by removing senescent algae and
detritus, alleviating light limitation for photosynthesis and N
2
xation (Grimm and Petrone 1997, Arango et al. 2009).
Grazers can also stimulate the growth of epilithon through
nutrient regeneration (Sterner 1986, Urabe et al. 2002, Knoll
et al. 2009). However, it is unlikely that grazer-mediated nu-
trient release led to the observed increase in biomass in our
study, as grazer excretion can increase dissolved N availabil-
ity (McCormick 1990) and higher DIN concentrations are
negatively correlated with N
2
-xation rates (Kunza and Hall
2014). As a result, we would expect to see lower, not higher,
N
2
xation in the presence of grazers. In N-limited algal turf
communities on coral reefs, grazing by Diadema urchins
enhanced chlorophyll-specic rates of N
2
xation, likely
by increasing light availability and water ow to N
2
xers
(Williams and Carpenter 1997). Here, grazing did not alter
biomass-specicN
2
xation, but AFDM and epilithic
biovolume was twice as high on grazed rocks in the largest
stream, suggesting that a grazer-induced increase in the to-
tal biomass of N
2
xers was likely responsible for the ob-
served increase in areal N
2
-xation rates.
A number of studies have observed an increase in the
abundance of cyanobacteria (Steinman et al. 1991, Rose-
mond et al. 1993, Flecker 1996, Peterson et al. 1998) with
Figure 7. Epilithon assemblage composition in each stream, separated by grazing treatment, for the mean relative biovolume of
epilithon taxa (see Table S1 for the species identied under each epilithon taxa category) (A), and the absolute biovolume of epilithon
functional groups (B), highlighting the total biovolume of nitrogen (N
2
)-xing and non-N
2
-xing algae and cyanobacteria. In
panel (B), the diatoms (non-N
2
-xing) functional group encompasses all of the non-N
2
-xing diatoms listed in panel A. Stream sites
are in order of increasing drainage area: Fox Creek (2 km
2
), Elder Creek (17 km
2
), and the South Fork Eel River (120 km
2
). In panel B,
error bars indicate the standard error of the mean. Note that the y-axis in panel B was scaled for visibility. The grazed epilithon cell
biovolume is 3higher than the ungrazed epilithon cell biovolume.
000 | Grazing and epilithon control N
2
xation B. L. Weigel et al.
This content downloaded from 073.075.149.002 on July 13, 2020 17:42:56 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
grazing, and Arango et al. (2009) found that grazing in-
creased the abundance of diatoms with N
2
-xing endosym-
bionts. Here, we observed compositional changes to epilithon
assemblages with grazing that altered the abundances of
N
2
-xing taxa in the largest stream. The total cell bio-
volume of N
2
-xing cyanobacteria (including Rivularia
spp. and other heterocyte-bearing cyanobacteria) was 12
higher on grazed rocks in the South Fork Eel River. In ad-
dition, the relative abundance of cyanobacteria without he-
terocytes, some which have recently been shown to be
capable of N
2
xation (Berrendero et al. 2016), was also 6
higher in the grazed treatment. The 73% increase in N
2
-
xation rates observed in the grazed treatment was there-
fore likely the result of shifts in the taxonomic composition
of epilithon toward a higher abundance of N
2
-xing cya-
nobacteria. It is interesting, however, that the grazed treat-
ment exhibited a lower relative biovolume of diatoms with
N
2
-xing endosymbionts. This result was explained by a re-
duction in the biovolume of large Epithemia turgida (Ehrenb.)
Kütz. 1844 cells, which are known to be palatable to grazers
(Marks and Power 2001, Power et al. 2009, Furey et al. 2012).
Free-living cyanobacteria, in contrast, are less palatable be-
cause of a number of defense mechanisms including the pro-
duction of toxins, thick mucilage, and rapidly-regenerating
basal trichomes (Power et al. 1988), all of which facilitate
their resistance to grazing and increase the potential for high
rates of N
2
xation.
Here, we found that grazing stimulated N
2
xation in the
largest stream and lowered DIN uptake by epilithon, shift-
ing the predominant source of N away from the dissolved
pool of available N in stream water to atmospheric N
2
.
The observed increase in areal N
2
xation with grazing in
the largest stream was well matched by a decrease in areal
NH
4
-uptake rates (459 and 580 lgNm
22
h
21
, respectively),
suggesting that biological N demand remained relatively
constant despite changes in the taxonomic composition of
the epilithon. Overall, a greater proportion of N demand
was met by atmospheric N
2
xation when grazers were pre-
sent. By increasing the proportion of N assimilated as atmo-
spheric N
2
, grazing also increased total N loading in the
stream, with important potential consequences for down-
stream export and the availability of N in receiving systems.
Further, increases in N
2
-xation rates due to grazing in the
largest stream were most likely explained by a taxonomic
shift in epilithon assemblages toward a greater proportion
of N
2
-xing cyanobacteria. This increase in N
2
xation rep-
resents an inux of new N to the ecosystem that has the po-
tential to enhance stream productivity (Karlson et al. 2015).
However, the proliferation of less palatable cyanobacteria
with grazing may instead divert N in epilithic biomass away
from the stream food web, resulting in increased down-
stream N export. Thus, both stream size (likely as a result
of higher light and temperature conditions) and community
dynamics (herbivory) have the potential to inuence the com-
position and activity of N
2
-xing assemblages in streams,
with important implications for total N availability and its
path through ecosystems.
We note that while grazing decreased epilithic NH
4
up-
take, total DIN uptake (measured as NH
4
-N 1NO
3
-N up-
take) did not change with grazing. This lack of change may
be because of a simultaneous decrease in NH
4
uptake and
increase in NO
3
removal in the presence of grazers (Fig. 4).
It is important to note that microbial reductive processes,
including denitrication, could have contributed to NO
3
removal, while nitrication could have articially increased
our estimates of NH
4
uptake and lowered rates of NO
3
up-
take. While the dataset precludes calculation of these mi-
crobial transformation rates, the proportion of NH
4
removal
from nitrication is likely low relative to assimilatory uptake
(Arango et al. 2008). This is especially true in streams with
relatively low DIN availability (Bernhardt et al. 2002), as ni-
trifying microbes are poor competitors for NH
4
compared
to photoautotrophs (Smith et al. 2014). Future studies inves-
tigating the effects of grazers on autotrophic algal assem-
blages as well as microbial communities will provide a richer
understanding of grazer-mediated nitrogen transformations
in stream ecosystems.
ACKNOWLEDGEMENTS
Author contributions: BLW, JRW, and PCF conceived of the
project, designed methodology, and collected data in the eld.
BLW analyzed the data and led the manuscript writing. JRW and
PCF contributed to manuscript preparation and the writing pro-
cess. All authors gave nal approval for publication.
We thank Mary E. Power for conceptual guidance, advice, and
knowledge of the Eel River basin that helped shape this project.
We are grateful to the University of California Natural Reserve
System as well as Peter Steel and his family for providing the An-
gelo Coast Range Reserve as a protected research reserve. Many
thanks to Grace M. Wilkinson, Angela J. Rosendahl, and Jessica
Cormier for assistance in the eld. Thank you to Mike Limm
for designing the ow-through recirculating chamber system that
we used in this study. Thanks to Charles E. Umbanhowar for pro-
viding microscopy facilities and project guidance, John D. Schade
for helpful analytical advice, and the St Olaf College Biology De-
partment for supporting this research. J. T. Wootton, C. A. Pster,
and M. E. Power provided helpful feedback and comments on this
manuscript. This research was supported by the National Science
Foundation grants NSF-DEB 0950016 and NSF-DBI 0923234,
awarded to JRW. JRW and PCF also received support from the
National Science Foundations National Center for Earth Surface
Dynamics, awarded to project PI EFoufoula-Georgiou, and allo-
cated by co-PI Mary E. Power. The authors have no conicts of
interest to declare.
LITERATURE CITED
Abe, S. I., K. Uchida, T. Nagumo, and J. Tanaka. 2007. Alterations
in the biomass-specic productivity of periphyton assemblages
mediated by sh grazing. Freshwater Biology 52:14861493.
Volume 39 September 2020 | 000
This content downloaded from 073.075.149.002 on July 13, 2020 17:42:56 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Anderson, M. J. 2001. A new method for non-parametric multi-
variate analysis of variance. Austral Ecology 26:3246.
Arango, C. P., L. A. Riley, J. L. Tank, and R. O. Hall. 2009. Her-
bivory by an invasive snail increases nitrogen xation in a
nitrogen-limited stream. Canadian Journal of Fisheries and
Aquatic Sciences 66:13091317.
Arango, C. P., J. L. Tank, L. T. Johnson, and S. K. Hamilton. 2008.
Assimilatory uptake rather than nitrication and denitrica-
tion determines nitrogen removal patterns in streams of vary-
ing land use. Limnology and Oceanography 53:25582572.
Bernhardt, E. S., R. O. Hall Jr, and G. E. Likens. 2002. Whole-
system estimates of nitrication and nitrate uptake in streams
of the Hubbard Brook Experimental Forest. Ecosystems 5:
419430.
Berrendero, E., E. F. Valiente, E. Perona, C. L. Gómez, V. Loza,
M. Á. Muñoz-Martín, and P. Mateo. 2016. Nitrogen xation
in a non-heterocystous cyanobacterial mat from a mountain
river. Scientic Reports 6:30920.
Capone, D. G. 1993. Determination of nitrogenase activity in aquatic
samples using the acetylene reduction procedure. Pages 621
631 in P. F. Kemp, B. F. Sherr, and J. J. Cole (editors). Handbook
of methods in aquatic microbial ecology. Lewis Publishers, Boca
Raton, Florida.
Chan, F., R. L. Marino, R. W. Howarth, and M. L. Pace. 2006. Eco-
logical constraints on planktonic nitrogen xation in saline es-
tuaries. II. Grazing controls on cyanobacterial population dy-
namics. Marine Ecology Progress Series 309:4153.
Elser, J. J., M. E. S. Bracken, E. E. Cleland, D. S. Gruner, W. S.
Harpole, H. Hillebrand, J. T. Ngai, E. W. Seabloom, J. B.
Shurin, and J. E. Smith. 2007. Global analysis of nitrogen and
phosphorus limitation of primary producers in freshwater,
marine and terrestrial ecosystems. Ecology Letters 10:1135
1142.
Elser, J. J., and J. Urabe. 1999. The stoichiometry of consumer-
driven nutrient recycling: Theory, observations, and conse-
quences. Ecology 80:735751.
Feminella, J. W., and C. P. Hawkins. 1995. Interactions between
stream herbivores and periphyton: A quantitative analysis of
past experiments. Journal of the North American Benthological
Society 14:445509.
Feminella, J. W., M. E. Power, and V. H. Resh. 1989. Periphyton
responses to invertebrate grazing and riparian canopy in three
northern California coastal streams. Freshwater Biology 22:445
457.
Finlay, J. C., J. M. Hood, M. P. Limm, M. E. Power, J. D. Schade, and
J. R. Welter. 2011. Light-mediated thresholds in stream-water
nutrient composition in a river network. Ecology 92:140150.
Flecker, A. S. 1996. Ecosystem engineering by a dominant detri-
tivore in a diverse tropical stream. Ecology 77:18451854.
Flett, R. J., R. D. Hamilton, and N. E. R. Campbell. 1976. Aquatic
acetylene-reduction techniques: Solutions to several problems.
Canadian Journal of Microbiology 22:4351.
Furey, P. C., R. L. Lowe, M. E. Power, and A. M. Campbell-Craven.
2012. Midges, Cladophora, and epiphytes: Shifting interactions
through succession. Freshwater Science 31:93107.
Gettel, G. M., A. E. Giblin, and R. W. Howarth. 2007. The effects
of grazing by the snail, Lymnaea elodes, on benthic N
2
xation
and primary production in oligotrophic, arctic lakes. Limnol-
ogy and Oceanography 52:23982409.
Gordon, T. A. C., J. Neto-Cerejeira, P. C. Furey, and E. J.
OGorman. 2018. Changes in feeding selectivity of freshwater
invertebrates across a natural thermal gradient. Current Zool-
ogy 64:231242.
Grimm, N. B., and K. C. Petrone. 1997. Nitrogen xation in a des-
ert stream ecosystem. Biogeochemistry 37:3361.
Hambright, K. D., N. G. Hairston, W. R. Schaffner, and R. W.
Howarth. 2007. Grazer control of nitrogen xation: Phyto-
plankton taxonomic composition and ecosystem functioning.
Fundamental and Applied Limnology/Archiv für Hydrobio-
logie 170:103124.
Hiatt, D. L., C. J. Robbins, J. A. Back, P. K. Kostka, R. D. Doyle,
C. M. Walker, M. C. Rains, D. F. Whigham, and R. S. King.
2017. Catchment-scale alder cover controls nitrogen xation
in boreal headwater streams. Freshwater Science 36:523532.
Hill, W. R., and A. W. Knight. 1988. Nutrient and light limitation
of algae in two Northern California streams. Journal of Phy-
cology 24:125132.
Hillebrand, H. 2002. Top-down versus bottom-up control of
autotrophic biomassA meta-analysis on experiments with
periphyton. Journal of the North American Benthological So-
ciety 21:349369.
Holmes, R. M., J. W. McClelland, D. M. Sigman, B. Fry, and B. J.
Peterson. 1998. Measuring
15
NNH
4
1
in marine, estuarine and
fresh waters: An adaptation of the ammonia diffusion method
for samples with low ammonium concentrations. Marine
Chemistry 60:235243.
Holomuzki, J. R., P. C. Furey, R. L. Lowe, and M. E. Power.
2013. Microdistributional variability of larval caddisies in
Mediterranean-climate streams in northern California. West-
ern North American Naturalist 73:261269.
Karlson, A. M. L., J. Duberg, N. H. Motwani, H. Hogfors,
I. Klawonn, H. Ploug, J. B. Svedén, A. Garbaras, B. Sundelin,
S. Hajdu, U. Larsson, R. Elmgren, and E. Gorokhova. 2015.
Nitrogen xation by cyanobacteria stimulates production in
Baltic food webs. AMBIO 44:413426.
Knoll, L. B., P. B. McIntyre, M. J. Vanni, and A. S. Flecker. 2009.
Feedbacks of consumer nutrient recycling on producer bio-
mass and stoichiometry: Separating direct and indirect effects.
Oikos 118:17321742.
Krammer, K., and H. Lange-Bertalot. 1991. Bacillariophyceae 4.
Teil: Achnanthaceae, Kritische Ergänzungen zu Navicula (Line-
olatae) und Gomphonema, Gesamtliteraturverzeichnis Teil 1
4. In H. Ettl, G. Gärtner, J. Gerloff, H. Heynig, and D. Mollen-
hauer (editors). Süßwasserora von Mitteleuropa, Band 2/4.
Gustav Fischer Verlag, Stuttgart, Germany.
Kunza, L. A., and R. O. Hall. 2014. Nitrogen xation can exceed
inorganic nitrogen uptake uxes in oligotrophic streams. Bio-
geochemistry 121:537549.
Lamberti, G. A., and V. H. Resh. 1983. Stream periphyton and in-
sect herbivores: An experimental study of grazing by a cad-
disy population. Ecology 64:11241135.
Lewis Jr, W. M., and S. N. Levine. 1984. The light response of ni-
trogen xation in Lake Valencia, Venezuela. Limnology and
Oceanography 29:894900.
Liess, A., and M. Kahlert. 2009. Gastropod grazers affect periph-
yton nutrient stoichiometry by changing benthic algal taxon-
omy and through differential nutrient uptake. Journal of the
North American Benthological Society 28:283293.
000 | Grazing and epilithon control N
2
xation B. L. Weigel et al.
This content downloaded from 073.075.149.002 on July 13, 2020 17:42:56 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Marcarelli, A. M., M. A. Baker, and W. A. Wurtsbaugh. 2008. Is
in-stream N
2
xation an important N source for benthic com-
munities and stream ecosystems? Journal of the North Amer-
ican Benthological Society 27:186211.
Marcarelli, A. M., and W. A. Wurtsbaugh. 2006. Temperature
and nutrient supply interact to control nitrogen xation in ol-
igotrophic streams: An experimental examination. Limnology
and Oceanography 51:22782289.
Marcarelli, A. M., and W. A. Wurtsbaugh. 2007. Effects of up-
stream lakes and nutrient limitation on periphytic biomass and
nitrogen xation in oligotrophic, subalpine streams. Freshwater
Biology 52:22112225.
Marks, J. C., and M. E. Power. 2001. Nutrient induced changes in
the species composition of epiphytes on Cladophora glomerata
Kütz. (Chlorophyta). Hydrobiologia 450:187196.
McCormick, P. V. 1990. Direct and indirect effects of consumers
on benthic algae in isolated pools of an ephemeral stream. Ca-
nadian Journal of Fisheries and Aquatic Sciences 47:20572065.
McNeely, C., and M. E. Power. 2007. Spatial variation in caddisy
grazing regimes within a northern California watershed. Ecol-
ogy 88:26092619.
Murphy, J., and J. P. Riley. 1962. A modied single solution method
for the determination of phosphate in natural waters. Analytica
Chimica Acta 27:3136.
Peterson, C. G., K. A. Vormittag, and H. M. Valett. 1998. Inges-
tion and digestion of epilithic algae by larval insects in a heav-
ily grazed montane stream. Freshwater Biology 40:607623.
Power, M., R. Lowe, P. Furey, J. Welter, M. Limm, J. Finlay, C.
Bode, S. Chang, M. Goodrich, and J. Sculley. 2009. Algal mats
and insect emergence in rivers under Mediterranean climates:
Towards photogrammetric surveillance. Freshwater Biology
54:21012115.
Power, M. E. 1990. Resource enhancement by indirect effects
of grazers: Armored catsh, algae, and sediment. Ecology
71:897904.
Power, M. E. 1991. Shifts in the effects of tuft-weaving midges on
lamentous algae. American Midland Naturalist 125:275285.
Power, M. E., J. R. Holomuzki, and R. L. Lowe. 2013. Food webs in
Mediterranean rivers. Hydrobiologia 719:119136.
Power, M. E., W. J. Matthews, and A. J. Stewart. 1985. Grazing
minnows, piscivorous bass, and stream algae: Dynamics of a
strong interaction. Ecology 66:14481456.
Power, M. E., M. S. Parker, and W. E. Dietrich. 2008. Seasonal re-
assembly of a river food web: Floods, droughts, and impacts of
sh. Ecological Monographs 78:263282.
Power, M. E., A. J. Stewart, and W. J. Matthews. 1988. Grazer
control of algae in an Ozark mountain stream: Effects of short-
term exclusion. Ecology 69:18941898.
Rabouille, S., M. Staal, L. J. Stal, and K. Soetaert. 2006. Modeling
the dynamic regulation of nitrogen xation in the cyanobacte-
rium Trichodesmium sp. Applied and Environmental Micro-
biology 72:32173227.
Rosemond, A. D., P. J. Mulholland, and J. W. Elwood. 1993. Top-
down and bottom-up control of stream periphyton: Effects of
nutrients and herbivores. Ecology 74:12641280.
Schade, J. D., K. MacNeill, S. A. Thomas, F. C. McNeely, J. R.
Welter, and J. Hood. 2011. The stoichiometry of nitrogen
and phosphorus spiraling in heterotrophic and autotrophic
streams. Freshwater Biology 56:424436.
Sculley, J. B., R. L. Lowe, C. A. Nittrouer, T. M. Drexler, and M. E.
Power. 2017. Eighty years of food-web response to interannual
variation in discharge recorded in river diatom frustules from
an ocean sediment core. Proceedings of the National Academy
of Sciences 114:1015510159.
Smith, J. M., F. P. Chavez, and C. A. Francis. 2014. Ammonium
uptake by phytoplankton regulates nitrication in the sunlit
ocean. PLoS ONE 9:e108173.
Steinman, A. D., P. J. Mulholland, and D. B. Kirschtel. 1991. In-
teractive effects of nutrient reduction and herbivory on bio-
mass, taxonomic structure, and P uptake in lotic periphyton
communities. Canadian Journal of Fisheries and Aquatic Sci-
ences 48:19511959.
Sterner, R. W. 1986. Herbivoresdirect and indirect effects on al-
gal populations. Science 231:605607.
Stevenson, R., and R. Glover. 1993. Effects of algal density and
current on ion transport through periphyton communities.
Limnology and Oceanography 38:12761281.
Taylor, B. W., C. F. Keep, R. O. Hall Jr, B. J. Koch, L. M. Tronstad,
A. S. Flecker, and A. J. Ulseth. 2007. Improving the uoromet-
ric ammonium method: Matrix effects, background uores-
cence, and standard additions. Journal of the North American
Benthological Society 26:167177.
Uno, H. 2016. Stream thermal heterogeneity prolongs aquatic-
terrestrial subsidy and enhances riparian spider growth. Ecol-
ogy 97:25472553.
Urabe, J., J. J. Elser, M. Kyle, T. Yoshida, T. Sekino, and Z. Kawa-
bata. 2002. Herbivorous animals can mitigate unfavourable
ratios of energy and material supplies by enhancing nutrient
recycling. Ecology Letters 5:177185.
Vannote, R. L., G. W. Minshall, K. W. Cummins, J. R. Sedell, and
C. E. Cushing. 1980. The river continuum concept. Canadian
Journal of Fisheries and Aquatic Sciences 37:130137.
Warren, D. R., S. M. Collins, E. M. Purvis, M. J. Kaylor, and H. A.
Bechtold. 2017. Spatial variability in light yields colimitation
of primary production by both light and nutrients in a forested
stream ecosystem. Ecosystems 20:198210.
Welter, J. R., J. P. Benstead, W. F. Cross, J. M. Hood, A. D. Huryn,
P. W. Johnson, and T. J. Williamson. 2015. Does N
2
xation
amplify the temperature dependence of ecosystem metabo-
lism? Ecology 96:603610.
Williams, S. L., and R. C. Carpenter. 1997. Grazing effects on
nitrogen xation in coral reef algal turfs. Marine Biology
130:223231.
Williamson, T. J., W. F. Cross, J. P. Benstead, G. M. Gíslason,
J. M. Hood, A. D. Huryn, P. W. Johnson, and J. R. Welter.
2016. Warming alters coupled carbon and nutrient cycles
in experimental streams. Global Change Biology 22:2152
2164.
Wootton, J. T., M. S. Parker, and M. E. Power. 1996. Effects of dis-
turbance on river food webs. Science 273:15581561.
Volume 39 September 2020 | 000
This content downloaded from 073.075.149.002 on July 13, 2020 17:42:56 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
... The watershed maintains low inorganic nutrient levels throughout the river network (Supporting Information Table S1; Finlay et al. 2011). During summer months, N 2 -fixation increases with stream size, which enhances autotrophic production and DIN availability at downstream sites (Power et al. 2008;Weigel et al. 2020). Finally, previous research has extensively described patterns in metabolism, whole stream nutrient uptake, and biotic structure throughout the watershed (McNeely et al. 2006;Schade et al. 2011), providing a strong basis for analyses of organic matter dynamics across sites (Table 1, Supporting Information Table S1, Fig. S1). ...
... The longitudinal increase in benthic POM-associated N uptake we observed, in conjunction with similar, previously observed increases in N 2 -fixation and nitrogen availability (Power et al. 2008;Finlay et al. 2011;Weigel et al. 2020), indicate that ecosystem N demand (but not N limitation) increased longitudinally in these watersheds. The ultimate driver of these longitudinal patterns is likely light and the associated increase in primary production and algal biomass (Supporting Information Fig. S1), supporting previous work indicating that light availability controls primary production and nutrient cycling at the ecosystem level and in biomass compartments (Finlay 2011;Finlay et al. 2011;Peipoch et al. 2016;Tank et al. 2018;Kaylor et al. 2019). ...
Article
Full-text available
Longitudinal changes in the structure and function of river ecosystems have long been recognized, yet our understanding of how such patterns shape elemental cycles remains limited. In particular, while benthic fine particulate organic matter (POM, 0.7–1000 μm) may control many stream nutrient cycles, less is known about longitudinal patterns or controls of benthic POM‐associated nutrient uptake. We conducted a survey of benthic POM‐associated respiration and nutrient uptake as well as microbial biomass (bacteria and algae) and benthic POM composition in four size classes (0.7–53 μm, 53–106 μm, 106–250 μm, and 250–1000 μm) in six streams in the forested South Fork Eel River watershed (California), encompassing a longitudinal gradient in light availability and primary production. Benthic POM at downstream sites was composed of smaller particles with lower organic matter content that were richer in nitrogen and autotrophic material. Areal respiration and nutrient uptake rates increased 11‐ to 67‐fold with stream size. While microbial activity rates did not increase with stream size, benthic POM‐associated microbial biomass increased 20‐fold with stream size, and closely tracked a 15‐fold increase in light availability, and primary production. Thus, microbial biomass, not activity, determined longitudinal patterns in benthic POM‐associated areal nutrient uptake and respiration rates. We attribute longitudinal patterns in microbial biomass to increases in light availability and primary production. Our findings help clarify the role of local (primary production) and upstream processes in shaping ecosystem structure and function.
... Stream type arthropod prey in larger open-canopy reaches where epilithon densities have been shown to be greater compared to smaller streams (Weigel et al., 2020). ...
Article
Stream intermittency is predicted to increase where water withdrawals and climate warming are increasing. In regions coupled with high fish diversity, understanding how intermittency influences fish trophic ecology is critical for informing ecosystem function. This study compared fish diets across seasons in perennial and intermittent streams to estimate the immediate and cumulative effects of stream drying on fish foraging patterns. We used gut content analysis to compare the diets of small-bodied, secondary consumer fishes, including two minnow and three darter species found in the lower Flint River Basin of southwestern Georgia, during both the summer (before stream dry-down) and fall (post flow resumption) seasons. Fish communities in perennial streams had greater diet richness compared to fishes in intermittent streams for both seasons. Darter diets were characterised by rheophilic aquatic insects in perennial streams and by benthic crustaceans (copepods, cladocerans and isopods) and predatory aquatic insects in intermittent streams. Minnow diets were typified by freshwater sponges, eggs and organic detritus in intermittent streams and by terrestrial insects and diatoms in perennial streams. Fishes in intermittent streams consumed significantly more benthic crustaceans in the fall (37% increase in proportional volume) compared to preflow cessation conditions in the summer, suggesting these organisms play an important, yet relatively unrecognised role in supporting fish communities in southeastern streams. Our findings enhance our understanding of how stream intermittency influences the trophic dynamics of secondary consumer fishes in an agricultural watershed increasingly affected by water scarcity.
Article
Full-text available
Environmental warming places physiological constraints on organisms, which may be mitigated by their feeding behaviour. Theory predicts that consumers should increase their feeding selectivity for more energetically valuable resources in warmer environments to offset the disproportionate increase in metabolic demand relative to ingestion rate. This may also result in a change in feeding strategy or a shift towards a more specialist diet. This study used a natural warming experiment to investigate temperature effects on the feeding selectivity of three freshwater invertebrate grazers: the snail Radix balthica, the blackfly larva Simulium aureum, and the midge fly larva Eukiefferiella minor. Chesson’s Selectivity Index was used to compare the proportional abundance of diatom species in the guts of each invertebrate species with corresponding rock biofilms sampled from streams of different temperature. The snails became more selective in warmer streams, choosing high profile epilithic diatoms over other guilds and feeding on a lower diversity of diatom species. The blackfly larvae appeared to switch from active collector gathering of sessile high profile diatoms to more passive filter feeding of motile diatoms in warmer streams. No changes in selectivity were observed for the midge fly larvae, whose diet was representative of resource availability in the environment. These results suggest that key primary consumers in freshwater streams, which constitute a major portion of invertebrate biomass, can change their feeding behaviour in warmer waters in a range of different ways. These patterns could potentially lead to fundamental changes in the flow of energy through freshwater food webs.
Article
Full-text available
Little is known about the importance of food-web processes as controls of river primary production due to the paucity of both long-term studies and of depositional environments which would allow retrospective fossil analysis. To investigate how freshwater algal production in the Eel River, northern California, varied over eight decades, we quantified siliceous shells (frustules) of freshwater diatoms from a well-dated undisturbed sediment core in a nearshore marine environment. Abundances of freshwater diatom frustules exported to Eel Canyon sediment from 1988 to 2001 were positively correlated with annual biomass of Cladophora surveyed over these years in upper portions of the Eel basin. Over 28 years of contemporary field research, peak algal biomass was generally higher in summers following bankfull, bed-scouring winter floods. Field surveys and experiments suggested that bed-mobilizing floods scour away overwintering grazers, releasing algae from spring and early summer grazing. During wet years, growth conditions for algae could also be enhanced by increased nutrient loading from the watershed, or by sustained summer base flows. Total annual rainfall and frustule densities in laminae over a longer 83-year record were weakly and negatively correlated, however, suggesting that positive effects of floods on annual algal production were primarily mediated by "top-down" (consumer release) rather than "bottom-up" (growth promoting) controls.
Article
Full-text available
Colimitation of primary production is increasingly recognized as a dominant process across aquatic and terrestrial ecosystems. In streams, both nutrient availability and light availability have been shown to independently limit primary production, but colimitation by both light and nutrients is rarely considered. We used a series of nutrient-diffusing substrates (NDS) bioassays deployed across a range of light availability conditions in a single-study stream over two summers to determine the light level at which the limiting factor for benthic periphyton accrual transitioned from light to nutrients. Stream periphyton accrual was nutrient-limited in high-light patches, and light-limited in low-light patches, with the transition from being predominantly light-limited to being predominantly nutrient-limited occurring when daily light fluxes exceeded 3.5 mol m−2 day−1. We quantified light at each NDS bioassay location and at 5 m intervals throughout our two adjacent 160 m study reaches—one in structurally complex old-growth riparian forest and one bordered by more uniform second-growth forest. Although both reaches were colimited overall, the resource (light or nutrients) dominating limitation differed between the two riparian forest age/structure conditions. In the old-growth section, about three quarters of the reach was predominantly nutrient-limited, whereas in the second-growth reach only about a quarter of the streambed was nutrient-limited. In this stream, colimitation of benthic periphyton accrual by light and nutrients at the reach scale was an emergent property of the ecosystem that manifested as a result of high heterogeneity in riparian forest structure.
Article
Full-text available
Emerging aquatic insects from streams are important food sources for riparian predators, yet their availability is seasonally limited. Spatial heterogeneity in stream water temperature was found to spatially desynchronize the emergence timing of aquatic insects, and prolong their flight period, potentially enhancing consumer growth. While a mayfly Ephemerella maculata emergence lasted for 12-22 days in local sites along a river, mayflies emerged 19 days earlier from warmer than cooler sites. Therefore, the overall emergence of E. maculata from the river lasted for 37 days, and adult swarms were observed over that same period in an adjacent reproductive habitat. A feeding experiment with the riparian spider Tetragnatha versicolor showed that a prolonged subsidy, as would occur in a heterogeneous river, led to higher juvenile growth than a synchronous pulsed subsidy of equal total biomass, as would typify a more homogeneous river. Since larger female adult spiders produce more eggs, spiders that received prolonged subsidy as juveniles should achieve higher fecundity. Restoring spatial heterogeneity in streams may benefit not only stream communities but also riparian predators. This article is protected by copyright. All rights reserved.
Article
Full-text available
In situ nitrogen fixation was investigated in a cyanobacterial mat growing on the bed of rocks of the Muga River, Spain. The filamentous non-heterocystous cyanobacterium Schizothrix dominated the mat, showing nitrogenase activity in the light at similar rates to those found in nearby heterocystous Rivularia colonies. N2 fixation in the light was significantly increased by an inhibitor of PSII and oxygen evolution, DCMU (3-[3,4-dichlorophenyl]-1,1-dimethylurea), and anaerobic conditions. However, no nitrogenase activity was found in the dark. Addition of fructose as a respiratory substrate induced nitrogenase activity in samples incubated under aerobic conditions in the dark but not in anaerobic conditions. Microelectrode oxygen profiles showed internal microaerobic microzones where nitrogen fixation might concentrate. Analyses of the 16S rRNA gene revealed only the presence of sequences belonging to filamentous non-heterocystous cyanobacteria. nifH gene diversity showed that the major phylotypes also belonged to this group. One of the three strains isolated from the Schizothrix mat was capable of fixing N2 and growing in the absence of combined N. This was consistent with the nifH gene analysis. These results suggest a relevant contribution of non-heterocystous cyanobacteria to nitrogen fixation in these mats.
Article
Alder (Alnus spp.) is a woody plant with bacterial symbionts that fix atmospheric N2 into bioavailable N. We studied 12 North American boreal headwater streams spanning a steep gradient of catchment alder cover (0-27%) to test the hypothesis that increasing inputs of inorganic N associated with increasing alder cover would reduce or eliminate in-stream benthic N2 fixation. We measured N2 fixation rates, chlorophyll a, and ash-free dry mass (AFDM) of periphyton in early (May) and late (August) summer 2011. Dissolved inorganic N (DIN) concentrations, composed almost entirely of NO3/NO2-N, ranged from below detection limits to nearly 2 mg/L and were strongly predicted by catchment alder cover in both months. Higher N2 fixation rates were observed in August than in May. N2 fixation rates declined sharply when alder cover exceeded ~2% of catchment cover, corresponding to 20 to 40 lg/L DIN. This pattern also was evident among 3 streams with contrasting catchment alder cover sampled approximately every 2 wk during the study. The stream with no catchment alder cover exhibited a steady increase in N2 fixation rates over the summer, whereas the streams with low and high alder in their catchments had very low N2 fixation rates that did not vary over time. The influence of alder cover on periphyton biomass was not clear. The strong regulation of alder cover on in-stream N availability leads us to suggest that alder presence may be an important terrestrial regulator of stream N cycling.
Article
Although much effort has been devoted to quantifying how warming alters carbon cycling across diverse ecosystems, less is known about how these changes are linked to the cycling of bioavailable nitrogen and phosphorus. In freshwater ecosystems, benthic biofilms (i.e. thin films of algae, bacteria, fungi and detrital matter) act as biogeochemical hotspots by controlling important fluxes of energy and material. Understanding how biofilms respond to warming is thus critical for predicting responses of coupled elemental cycles in freshwater systems. We developed biofilm communities in experimental streamside channels along a gradient of mean water temperatures (7.5-23.6°C), while closely maintaining natural diel and seasonal temperature variation and a common water and propagule source. Both structural (i.e. biomass, stoichiometry, assemblage structure) and functional (i.e. metabolism, N2 -fixation, nutrient uptake) attributes of biofilms were measured on multiple dates to link changes in carbon flow explicitly to the dynamics of nitrogen and phosphorus. Temperature had strong positive effects on biofilm biomass (2.8- to 24-fold variation) and net ecosystem productivity (44- to 317-fold variation), despite extremely low concentrations of limiting dissolved nitrogen. Temperature had surprisingly minimal effects on biofilm stoichiometry: carbon:nitrogen (C:N) ratios were temperature-invariant, while carbon:phosphorus (C:P) ratios declined slightly with increasing temperature. Biofilm communities were dominated by cyanobacteria at all temperatures (>91% of total biovolume) and N2 -fixation rates increased up to 120-fold between the coldest and warmest treatments. Although ammonium-N uptake increased with temperature (2.8- to 6.8-fold variation), the much higher N2 -fixation rates supplied the majority of N to the ecosystem at higher temperatures. Our results demonstrate that temperature can alter how carbon is cycled and coupled to nitrogen and phosphorus. The uncoupling of C fixation from dissolved inorganic nitrogen supply produced large unexpected changes in biofilm development, elemental cycling, and likely downstream exports of nutrients and organic matter. This article is protected by copyright. All rights reserved.