Comparisons of stable isotope (C, H, N) signatures for revealing organic matter sources and trophic relationships in headwater streams

Abstract

We compared the efficacy of stable carbon, hydrogen, and nitrogen isotope ratios in identifying the resources used by insect consumers in headwater streams of southern California. We also compared gut contents with consumer stable isotope ratios and mixing model estimates of resource contributions to predator diet. Stable hydrogen isotope ratios (as δ²H) of algivores were well separated from ratios for detritivores, whereas relationships between stable carbon (as δ¹³C) and nitrogen (as δ¹⁵N) ratios of consumers and their expected diets were weaker and more ambiguous. δ²H values of primary consumers more strongly reflected the proportions of their gut contents consisting of algae than δ¹³C values. Τhe proportions of algivorous prey in predator gut contents increased with mixing model estimates of algivore contributions to predator diet using δ²H but not δ¹³C values. Our findings support the use of hydrogen isotope ratios in food web studies of streams in southern California and their potential use in assessing the effects of anthropogenic and natural disturbance on basal resource contributions to food webs that might not otherwise be identified using carbon isotope ratios.

Full text

ARTICLE
Comparisons of stable isotope (C, H, N) signatures for revealing
organic matter sources and trophic relationships in headwater
streams
Henry M. Page, Scott D. Cooper, Sheila W. Wiseman, Danuta Bennett, Kristie Klose, Steven Sadro,
Craig Nelson, and Thomas Even
Abstract: We compared the efficacy of stable carbon, hydrogen, and nitrogen isotope ratios in identifying the resources used by
insect consumers in headwater streams of southern California. We also compared gut contents with consumer stable isotope
ratios and mixing model estimates of resource contributions to predator diet. Stable hydrogen isotope ratios (as ␦
2
H) of algivores
were well separated from ratios for detritivores, whereas relationships between stable carbon (as ␦
13
C) and nitrogen (as ␦
15
N)
ratios of consumers and their expected diets were weaker and more ambiguous. ␦
2
H values of primary consumers more strongly
reflected the proportions of their gut contents consisting of algae than ␦
13
C values. ⌻he proportions of algivorous prey in
predator gut contents increased with mixing model estimates of algivore contributions to predator diet using ␦
2
H but not ␦
13
C
values. Our findings support the use of hydrogen isotope ratios in food web studies of streams in southern California and their
potential use in assessing the effects of anthropogenic and natural disturbance on basal resource contributions to food webs that
might not otherwise be identified using carbon isotope ratios.
Résumé : Nous avons comparé l’efficacité des isotopes stables de carbone, d’hydrogène et d’azote pour cerner les ressources
utilisées par des consommateurs d’insectes dans des cours d’eau d’amont du sud de la Californie. Nous avons également comparé
les contenus stomacaux aux rapports d’isotopes stables de consommateurs et a
`des estimations obtenues de modèles de mélange
des contributions de ressources au régime alimentaire de prédateurs. Les rapports d’isotopes stables d’hydrogène (␦
2
H)
d’alguivores étaient bien distincts des rapports de détritivores, alors que les relations entre les rapports d’isotopes stables de
carbone (␦
13
C) et d’azote (␦
15
N) de consommateurs et leurs régimes alimentaires attendus étaient plus faibles et plus ambiguës.
Les valeurs de ␦
2
H de consommateurs primaires reflétaient plus fortement les proportions d’algues de leurs contenus stomacaux
que les valeurs de ␦
13
C. Les proportions de proies alguivores dans les contenus stomacaux de prédateurs augmentaient parallèle-
ment aux estimations des modèles de mélange de la contribution d’alguivores au régime alimentaire de prédateurs obtenues en
utilisant les valeurs de ␦
2
H, mais non les valeurs de ␦
13
C. Nos constatations appuient l’utilisation des rapports d’isotopes
d’hydrogène dans les études des réseaux trophiques dans des cours d’eau du sud de la Californie et leur utilité potentielle pour
évaluer les effets de perturbations humaines et naturelles sur les contributions de ressources de base aux réseaux trophiques que
les rapports d’isotopes de carbone pourraient ne pas faire ressortir. [Traduit par la Rédaction]
Introduction
Terrestrial organic matter and stream algae are recognized as
the two most important sources of primary production fueling
stream food webs (Cummins 1974;Finlay 2001;Doucett et al. 2007;
Caraco et al. 2010). Although the general importance of terrestrial
plant and stream algal production to stream communities is well
known, quantifying the relative contributions of these basal re-
sources to food webs is challenging (e.g., Cummins and Klug 1979;
France 1996;Doucett et al. 1996). Consumer nutrition inferred
from gut contents is invaluable in describing recent diets (Cross
et al. 2013;Rosi-Marshall et al. 2016); however, the digestibility and
assimilation of different foods may vary, and consumer diets can
change over time, even within the same taxon, making inferences
about consumer nutritional support from only gut content anal-
yses problematic (Chapman and Demory 1963;Moore 1977;Mihuc
and Minshall 1995;Lancaster et al. 2005;Jardine et al. 2014).
Stable isotope analysis can provide additional and complemen-
tary insights into the contributions of terrestrial plant and stream
algal production to stream food webs. Generally, isotope ratios of
carbon are used to identify organic carbon sources (Finlay 2001;
McNeely et al. 2007;Ishikawa et al. 2012), whereas isotope ratios
of nitrogen are used to assist in this evaluation and to identify
trophic levels or positions (Vander Zanden and Rasmussen 1999;
Post 2002;Anderson and Cabana 2007). Stable carbon isotope ra-
tios provide a time-integrated signature of dietary support pro-
vided by different carbon sources and possess an advantage over
nitrogen isotopes in that trophic discrimination, the enrichment
of the consumer relative to the resource in the heavy isotope (
13
C
Received 19 July 2016. Accepted 13 February 2017.
H.M. Page, S.W. Wiseman, D. Bennett, and K. Klose. Marine Science Institute, University of California, Santa Barbara, CA 93106-6150, USA.
S.D. Cooper. Marine Science Institute, University of California, Santa Barbara, CA 93106-6150, USA; Department of Ecology, Evolution, and Marine
Biology, University of California, Santa Barbara, CA 93106, USA.
S. Sadro. Department of Environmental Science and Policy, University of California, Davis, CA 95616, USA.
C. Nelson. Marine Science Institute, University of California, Santa Barbara, CA 93106-6150, USA; Department of Oceanography, University of Hawai‘i at
Maˉnoa, Honolulu, HI 96822, USA.
T. Even. Department of Ecology, Evolution, and Marine Biology, University of California, Santa Barbara, CA 93106, USA.
Corresponding author: Henry M. Page (email: page@lifesci.ucsb.edu).
Copyright remains with the author(s) or their institution(s). Permission for reuse (free in most cases) can be obtained from RightsLink.
2110
Can. J. Fish. Aquat. Sci. 74: 2110–2121 (2017) dx.doi.org/10.1139/cjfas-2016-0322 Published at www.nrcresearchpress.com/cjfas on 21 February 2017.
Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by Santa Barbara (UCSB) on 12/31/17
For personal use only.

or
15
N), is smaller for ␦
13
C values and, therefore, a better direct
indicator linking prey and consumers across trophic levels
(Peterson and Fry 1987;Post 2002). Mixing models using isotope
data can quantify the relative contributions of different resources
to consumer nutrition (Parnell et al. 2010;Layman et al. 2012).
The use of carbon isotope ratios to identify basal resource con-
tributions to aquatic food webs may be limited in some cases by
insufficient separation in carbon isotope end-member values
(typically <1.0‰–2.0‰, Jardine et al. 2009;Skinner et al. 2016) for
terrestrial detritus and algal basal resources. Insufficient separation
in end-member values makes descriptions of food web structure
using isotope biplots or quantitative mixing models challenging
(France 1996;Doucett et al. 1996,2007;Finlay 2001;McNeely et al.
2006;Ishikawa et al. 2012). Although the carbon isotope ratios of
leaf detritus, the principal allochthonous basal resource in stream
food webs, vary over a relatively narrow range (Finlay 2001), the
carbon isotope ratios of autochthonous algal biomass can vary
greatly (reviewed in Ishikawa et al. 2012). Variation in the carbon
isotope ratios of algae has been related to variation in the avail-
ability and isotopic signatures of aqueous CO
2
and dissolved inor-
ganic carbon and to factors that affect the fractionation of carbon
isotopes (
13
C/
12
C) during algal uptake, including algal productivity
and water current velocity (Doucett et al. 1996;Finlay et al. 1999;
Finlay 2004;Ishikawa et al. 2012).
Stable isotope ratios of hydrogen (as ␦
2
H) may provide a com-
plementary or alternative tracer for use in food web studies (Estep
and Dabrowski 1980;Macko et al. 1983). Recent investigations
have suggested that isotope ratios of hydrogen provide more con-
sistent and clear end-member separation of terrestrial and algal
basal resources than isotope ratios of carbon in stream, river, and
lake environments (Doucett et al. 2007;Finlay et al. 2010;Cole
et al. 2011;Karlsson et al. 2012). Distinct isotopic separation be-
tween terrestrial and algal sources is achieved largely because of
the fractionation of hydrogen isotopes during evapotranspiration
of water from the leaves of terrestrial plants, causing isotopic
enrichment of the remaining leaf water, which is assimilated into
plant biomass during photosynthesis (Smith and Epstein 1970).
Although the use of hydrogen isotope ratios to trace allochtho-
nous and autochthonous sources of production in aquatic food
webs appears promising (Cole et al. 2006,2011;Doucett et al. 2007;
Finlay et al. 2010;Wilkinson et al. 2013,2015;Vander Zanden et al.
2016), the effectiveness of these ratios as a tracer may vary with
study system. For example, Jardine et al. (2009) reported less over-
lap of basal resource end-member values with carbon than with
hydrogen isotope ratios in streams in eastern Canada. In addition,
there are concerns about the effects of dietary water, as well as
trophic and tissue discrimination, on the hydrogen isotope values
of consumers (Macko et al. 1983;Solomon et al. 2009;Peters et al.
2012;Soto et al. 2013;Wilkinson et al. 2015;Vander Zanden et al.
2016). Relatively few studies have compared the use of stable carbon
and hydrogen isotope ratios as tracers in streams or, importantly,
examined the assumption that hydrogen isotope signatures reflect
consumer diets.
Preliminary data suggested that carbon isotope ratios could not
resolve sources of terrestrial versus aquatic production support-
ing food webs in the small headwater streams of southern Cali-
fornia (subwatershed areas <13 km
2
), necessitating the evaluation
of an alternative tracer for use in these systems. The streams of
southern California are located in a semiarid Mediterranean cli-
mate and are subject to natural (e.g., floods, drought) and anthro-
pogenic (e.g., fire, contaminants, exotic species, dams, diversions)
disturbances that affect the structure of stream communities and
food webs (Cooper et al. 2013,2015;Verkaik et al. 2013).
In this study, we provide new information on the stable carbon,
hydrogen, and nitrogen isotope ratios of primary and secondary
insect consumers in the streams of semiarid southern California.
We explore the hypothesis that stable isotope ratios of hydrogen
provide more distinct separation in isotopic end-member values
between terrestrial detritus and stream algae-based resources
than carbon or nitrogen isotope ratios, and thus are more useful
in identifying the relative contributions of these resources to the
food webs of southern California streams. Because of potential
overlap in the carbon isotope signatures of terrestrial detritus and
stream algae, we also hypothesized that hydrogen isotope ratios
would provide better agreement with recent diets, as determined
from gut content analyses, than carbon isotope ratios for both
primary consumers and predators. Because the isotope ratios of
consumer tissues are assumed to reflect those of their diets inte-
grated over time, tissue values may not necessarily reflect the
isotopic signatures of recently consumed food items. Neverthe-
less, a concordance between isotope ratios and diets assessed by
gut content analysis can provide ancillary support that isotope
ratios reflect dietary sources (Peterson 1999). Although gut con-
tent data have been compared with data on stable isotope ratios of
carbon and nitrogen (e.g., Mihuc and Toetz 1994;Lancaster et al.
2005;McNeely et al. 2007), no studies, to our knowledge, have
explored these relationships for hydrogen isotope ratios.
Materials and methods
Study streams
We measured the stable carbon, nitrogen, and hydrogen iso-
tope ratios of basal resources and a diverse array of invertebrate
consumers in nine streams. The study streams drain independent
subwatersheds, ranging in area from ⬃3to13km
2
, on the south
side of the Santa Ynez Mountains (Santa Barbara County, Califor-
nia, USA), a steep, coastal mountain chain separated from the
Pacific Ocean by a narrow coastal plain. The Santa Ynez Moun-
tains have a Mediterranean climate, with mean annual rainfall
ranging from 45 cm·year
−1
at sea level to 100 cm·year
−1
in the
mountains, with nearly all rain falling between November and
March. The drainage basins for our sites largely occurred above
developed areas (<4% urban and <6% agricultural). Wetted chan-
nel width at base flow averaged ⬃2to5minourstudy streams
and early summer water velocity ranged from 0.02 to 0.10 m·s
−1
in
pools and from 0.06 to 0.36 m·s
−1
in riffles. NO
3−
-N concentrations
at the time that samples were collected for isotope analysis
ranged from 5 to 184 ␮mol·L
−1
across streams, with the highest
value from Mission Creek.
Riparian canopy cover at our study sites ranged from 27% to
98%. Riparian vegetation was dominated by white alder (Alnus
rhombifolia), California laurel (Umbellularia californica), coast live
oak (Quercus agrifolia), willow (Salix spp.), and California sycamore
(Platanus racemosa). The riparian vegetation bordering two of our
study streams in larger (⬃10 km
2
) subwatersheds (Mission and San
Antonio creeks) had burned approximately 1 year prior to our
sampling, whereas upland, but not riparian vegetation, burned in
the catchment of another study site (Rattlesnake Creek). Our
stream sites represented a subset of sites sampled in a larger study
that explored the effects of wildfire on stream community struc-
ture and food webs (see Cooper et al. 2015 for a more detailed
description).
Sampling design: basal resources and consumers
We collected samples of basal resources (conditioned leaves,
algae, fine particulate organic matter) and invertebrates from the
major functional feeding groups for stable carbon, nitrogen, and
hydrogen isotope analysis, as well as invertebrates for gut content
analyses, from pool and riffle habitats separately in nine streams
in June 2010, approximately 2 months after the last rains of the
wet season.
Conditioned leaves of A. rhombifolia,U. californica,Q. agrifolia,
Salix spp., and (or) P. racemosa were collected from stream bottoms
with Surber samplers at each site, and a subset was subjected to
stable isotope analysis. Macroalgae (primarily Cladophora spp.)
were collected by hand from hard substrata (cobbles, boulders,
Page et al. 2111
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by Santa Barbara (UCSB) on 12/31/17
For personal use only.

bedrock). Benthic microalgae, although sampled, were not used
in our analysis because samples contained substantial amounts
of detritus based on microscopic examination. Previous studies
showed that microalgae were more enriched in deuterium (
2
H)
than Cladophora spp. (Finlay et al. 2010) and “filamentous algae”
(Doucett et al. 2007), but all algae were well separated isotopically
from stream leaf litter. Fine particulate organic matter (FPOM)
was sampled by coring soft substrata (silt, sand, gravel, pebbles) to
a depth of 0.8 cm with the barrel of a syringe sampler (Davies and
Gee 1993). After agitation, elutriation, and mixing, suspended
FPOM in subsamples was filtered through GF/C glass microfiber
filters. Because FPOM samples from riffles did not contain suffi-
cient quantities of organic matter for hydrogen isotope analysis,
we used FPOM isotope values from pools to represent FPOM values
for both pool and riffle habitats. Macroinvertebrates for isotope
analysis were collected at each stream site using Surber samplers
(250 ␮m mesh) and aquatic D-nets (1 mm mesh). For all basal
resources and invertebrates, samples were taken from five pools
and five riffles at each site, samples were combined by habitat
(pools versus riffles) and thoroughly mixed, and representative
subsamples were taken from these amalgamated samples.
All samples taken for stable isotope and gut content analyses
were stored frozen at −20 °C. The guts of individuals of common
primary and secondary consumer taxa, with the exception of chi-
ronomids, were removed and preserved in vials containing either
formalin (primary consumers) or ethanol (predators). Invertebrate
specimens without guts or entire chironomids were used for sta-
ble isotope analysis. Specimens of each taxon were not always
represented in samples from each stream.
Sample preparation and isotope analysis
In the laboratory, subsamples of leaves and macroalgae were
thawed and rinsed in deionized water to remove adhering mate-
rial. FPOM was removed wet from GF/C filters. Samples of leaves,
macroalgae, FPOM, and insects were dried in new glass scintilla-
tion vials without caps at 60 °C and ground (or gently crushed for
some small insect specimens) to a fine powder using a mortar and
pestle. The ground material was divided in half, and one portion
was analyzed for hydrogen isotopes while the other half was an-
alyzed for carbon and nitrogen isotopes. The number of insect
specimens processed varied by taxon depending on availability
and individual size; 10 to 100 individuals of each taxon from pools
versus riffles at each site were combined to obtain sufficient ma-
terial for analysis. Lipids were not removed from samples prior to
analysis, similar to earlier studies (Doucett et al. 2007;Finlay et al.
2010;Cole et al. 2011). The wide separation in hydrogen isotope
ratios between detritivores and detritus, on the one hand, and
algae and algivores, on the other hand, suggested that variation in
lipid content would not be an important driver of hydrogen iso-
tope patterns in our study (see also Wilkinson et al. 2015). This
assumption was further supported by the congruence between
gut contents and isotope mixing model estimates of resource con-
tributions to predator diets (see Results). However, because of
smaller differences in carbon isotope ratios between resources
and predators, we applied the correction for lipids recommended
by Skinner et al. (2016) based on C:N ratios (generally 4.0–4.5) to
data used in mixing model analyses (see below).
Isotopic analysis of ground samples or subsamples (typically
⬃0.5 mg for hydrogen, ⬃1 mg for carbon and nitrogen) was con-
ducted by the Facility for Isotope Ratio Mass Spectrometry (http://
ccb.ucr.edu/firms.html) at the University of California, Riverside,
using a thermochemical elemental analyzer interfaced to a Thermo-
Finnigan Delta V Advantage isotope ratio mass spectrometer (Thermo
Fisher Scientific Corp., Bremen, Germany). Hydrogen isotope ra-
tios are expressed as ␦
2
H. The ␦
2
H of non-exchangeable hydrogen
in all samples was measured using the comparative equilibration
method (Wassenaar and Hobson 2003;Kelly et al. 2009) with the
isotopic values normalized to the Vienna Standard Mean Ocean
Water – Standard Light Antarctic Precipitation scale. The follow-
ing standards were equilibrated with the samples and included in
each analytical run: Caribou Hoof Standard (␦
2
H = −197‰), Kudu
Horn Standard (␦
2
H = −54‰), and Spectrum Chemical keratin
(␦
2
H = −121‰) (Flockhart et al. 2013). The natural abundances of
carbon and nitrogen isotopes are expressed relative to the Pee Dee
Belemnite standard for carbon and atmospheric N
2
for nitrogen.
Variation in isotope values between replicate portions of the same
ground sample averaged 3.4‰, 0.1‰, and 0.2‰ for ␦
2
H(n= 10),
␦
13
C(n= 15), and ␦
15
N(n= 15), respectively. Dividing these preci-
sion estimates by the differences in mean stable isotope ratios
between algivores and detritivores in Table 1 produces values of
3.5%, 3.0%, and 8.7% for hydrogen, carbon, and nitrogen signa-
tures in pools and 4.0%, 2.9%, and 7.7% for hydrogen, carbon, and
nitrogen signatures in riffles. Thus, although the absolute preci-
sion value was highest for hydrogen isotope ratios, relative preci-
sion for hydrogen isotopes was comparable to or better than the
Table 1. Mean (±1 SE) ␦
13
C, ␦
15
N, and ␦
2
H values of basal resources, algivores, and detritivores in
Fig. 1 computed across study streams.
Habitat Category Taxon ␦
13
C (‰) ␦
15
N (‰) ␦
2
H (‰)
Pool Algivores Baetis −31.8±1.0 (9) 2.3±0.9 (9) −209±8 (9)
Callibaetis −32.6±1.2 (4) 3.7±2.5 (4) −228±10 (4)
Eubrianax −30.9±1.3 (5) 1.9±1.7 (5) −192±7 (5)
Centroptilum −30.8±1.2 (5) 3.2±1.8 (5) −216±11 (7)
Mean −31.5 2.8 −211
Detritivores Paraleptophlebia −29.1±0.7 (7) 0.9±0.2 (7) −122±10 (7)
Lepidostoma −27.9±0.3 (8) −0.4±0.6 (8) −105±5 (8)
Mean −28.5 0.5 −114
Macroalgae Cladophora −34.3±2.3 (6) 1.2±1.8 (6) −217±9 (6)
Conditioned leaves Various −29.0±0.3 (9) −2.1±0.4 (9) −110±4 (9)
FPOM −26.4±0.7 (9) 1.3±0.5 (9) −95±4 (5)
Riffle Algivores Baetis −31.7±1.2 (7) 3.2±1.3 (7) −190±9 (8)
Eubrianax −32.8±2.5 (3) 2.5±3.1 (3) −187±2 (3)
Mean −32.3 2.9 −189
Detritivores Paraleptophlebia −29.4±1.2 (4) 1.0±0.6 (4) −115±6 (5)
Lepidostoma −27.7±0.3 (3) −0.4±0.3 (3) −92±1 (3)
Mean −28.6 0.3 −104
Macroalgae Cladophora −34.6±1.4 (6) 0.3±1.1 (6) −219±8 (6)
Note: FPOM, fine particulate organic matter. Sample size (number of streams for each isotope value) is shown in
parentheses.
2112 Can. J. Fish. Aquat. Sci. Vol. 74, 2017
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by Santa Barbara (UCSB) on 12/31/17
For personal use only.

relative precision estimates for carbon and nitrogen isotope ratios
given the measured differences in mean stable isotope ratios be-
tween algivores and detritivores in Table 1.
Gut content analyses
We examined the gut contents of primary consumers and preda-
tors from our study sites to explore relationships between consumer
isotope values and recently consumed food items. Gut contents of
selected primary consumer taxa (Lepidostoma,Paraleptophlebia,
Baetis,Callibaetis,Centroptilum,Diphetor,Leucrocuta,Eubrianax,Simulium)
processed for isotope analysis (see above) were homogenized and
a subsample was removed to a microscope slide for the identifi-
cation, measurement, and enumeration of food items under a
compound microscope at 400× magnification. Counts were con-
verted to biovolumes using known geometric formulae for algal
taxa and fungi (Hillebrand et al. 1999). In addition, the slide area
occupied by leaf fragments or amorphous detritus was measured
and this was converted to a biovolume by multiplying the area by
the measured mean thickness (18.8 ␮m) of detrital particles.
We also examined the gut contents of predatory invertebrate
taxa from pools and riffles. The gut contents of individuals of
predatory taxa were placed in a small dish and examined under a
dissecting microscope. Any prey body parts (mandibles, claws,
tarsi, labra, heads) were removed and placed in a drop of 100%
glycerol on a microscope slide. Body parts were identified to the
lowest practical taxonomic level (usually genus) under a com-
pound microscope by comparing them with photographs and il-
lustrations of body parts taken from known taxa. The number of
each taxon per gut sample was calculated as the rounded-up inte-
ger of the number of claws/6, tarsi/6, mandibles/2, heads, oper-
cula, or labra (whichever was greatest). Because left and right
mayfly mandibles were distinguishable, the number of mayfly
nymphs per gut was determined by the number of left mandibles,
right mandibles, or other body parts (as calculated above), which-
ever was greatest.
Statistical analysis
We used general linear models to test for significant separation
in stable isotope ratios (expressed as ␦
2
H, ␦
13
C, or ␦
15
N) among
categories of resources and primary consumers (conditioned leaves,
FPOM, Cladophora, detritivores, algivores). The detritivore cate-
gory included two common taxa presumed to rely primarily on
terrestrial detritus and for which gut content data were available,
Lepidostoma (shredder caddisflies) and Paraleptophlebia (collector
mayflies). Many taxa in the subfamilies Orthocladiinae and Chi-
ronominae (family Chironomidae) are also considered detritivo-
rous (Armitage et al. 1997), but the enriched nitrogen isotope
ratios of chironomids collected in one creek (Mission) with ele-
vated nitrate levels (see Results) suggested that this group could
be using appreciable amounts of algae at some times or places. As
a consequence, the isotope results for the Orthocladiinae and
Chironominae are presented in Fig. 1, but these chironomids were
not considered resource specialists on leaves or algae and were
not included in the specialist primary consumer and predator
mixing model analyses (see below). The algivore category in-
cluded common taxa for which gut content data were available, i.e.,
mayflies of the Baetidae (Baetis,Callibaetis,Centroptilum) and larvae
of the beetle genus Eubrianax. Streams were treated as replicates
in analyses, with pools and riffles analyzed separately because
differences in current velocity and other factors between these
habitats could influence stable isotope ratios (Finlay et al. 1999,
2010).
Preliminary review of the isotope data suggested that values
varied with the subwatershed area of each stream site, similar to
Finlay et al.’s (2010) results. Therefore, we included subwatershed
area as a covariate in the analysis. When the linear model showed
no significant interaction effects of resource or consumer cate-
gory with subwatershed area, we used the Šidák test to identify
significant differences in isotope values between categories (e.g.,
leaves versus algae, detritivores versus algivores) for both pools
and riffles at a mean subwatershed area of ⬃8km
2
. The Šidák test
adjusts pvalues in multiple pairwise comparisons to reduce type I
error (Day and Quinn 1989). If the interaction between category
and subwatershed was significant, we evaluated differences be-
tween resource and primary consumer categories at two covariate
values (5 and 10 km
2
basin areas, representing “small” and “large”
subwatersheds, respectively).
We tested for an effect of habitat (pool versus riffle) on the isotope
ratios of algivores (Baetis,Eubrianax) and detritivores (Lepidostoma,
Paraleptophlebia) that were collected in both habitats using a
paired ttest (pairing habitats for each taxon by stream). We used
regression analyses to explore relationships between stable iso-
tope ratios and gut contents. Proportion data were arcsine square-
root transformed prior to analysis, producing results similar to
those obtained with logit transformations. Statistical analyses
were performed using SigmaPlot 13 and SPSS 22.
Comparison of predator stable isotope ratios with predator
gut contents
To estimate the proportional contribution of algivorous prey to
predator diets using stable hydrogen and carbon isotope ratios,
we used the Stable Isotope Analysis in R (SIAR), a commonly ap-
plied statistical package that computes dietary contributions of
resources to a consumer using a Bayesian framework (Parnell
et al. 2010). The SIAR model includes variability in source, con-
sumer, and trophic enrichment in iterative model fittings to gen-
erate probability estimates of source proportions in consumer
diets. We used the “siarsolo” routine in SIAR because we did not
have within-stream estimates of variation in isotope ratio or di-
etary composition values for predators (one value per prey taxon
or group per stream).
For the algivore end-member in the SIAR model, we used mean
␦
2
Hor␦
13
C values (and standard deviation) for individual algivo-
rous taxa (Baetis,Callibaetis,Centroptilum,Eubrianax) as replicates
from each stream. Based on the high proportion of algae in the gut
contents of these taxa (see Results), the isotope values of these
taxa also served as proxies for the isotope values of microalgae (as
in Vander Zanden and Rasmussen 2001;Finlay 2001,2004;Post
2002), which could not be determined directly because periphy-
ton samples contained considerable amounts of terrestrial detri-
tus that could not be removed by silicon centrifugation or other
methods prior to isotope analysis. For the detritivore end-
member, we used mean ␦
2
H and ␦
13
C values of Lepidostoma and
Paraleptophlebia, which had isotope values similar to those of con-
ditioned leaves and gut contents consisting almost entirely of
detrital particles (see Table 1;Fig. 1).
For carbon isotopes, we used a trophic enrichment factor (TEF)
of 0.4‰ ± 1.3‰ from prey to predator (Post 2002), whereas the lack
of consensus regarding trophic enrichment for hydrogen isotopes
(Vander Zanden et al. 2016) precluded the inclusion of a TEF for
hydrogen isotopes. Proportional algivore contributions to preda-
tor diets were modeled separately for hydrogen and carbon iso-
tope ratios and compared with proportions of algivore prey in
predator gut contents. For ␦
2
H, solutions of the SIAR model for
algivore contributions to predator diet were highly correlated
with the solutions of a single-isotope two-source mixing model
(Fry 2006)(R
2
= 0.95, p< 0.001, SIAR algivore proportion = 0.75 ×
(two-source algivore proportion) + 0.12), but the outputs of SIAR
and single-isotope mixing models were not related when ␦
13
C
values were used, because of greater overlap in prey end-member
values for ␦
13
C than for ␦
2
H (discussed below).
We also explored how incorporating environmental water into
the SIAR model affected our conclusions by varying the stable
hydrogen isotope ratios of stream water (␦
2
H
H2O
) and the propor-
tion of consumer tissue consisting of stream water hydrogen (
␻
)
(Solomon et al. 2009;Wilkinson et al. 2015). ␦
2
H
H2O
values for our
Page et al. 2113
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by Santa Barbara (UCSB) on 12/31/17
For personal use only.

study streams were not available. Given the proximity and similar
elevations of our study streams (Klose et al. 2015), we expected
that ␦
2
H
H20
values would be similar among streams, as supported
by estimates of precipitation ␦
2
H
H2O
based on the georeferenced
locations of study sites and the online tool Waterisotopes.org
(Bowen and Revenaugh 2003;Bowen et al. 2005). Estimated an-
nual precipitation ␦
2
H
H20
values were similar across our study
area (−53‰ to −48‰, xˉ = −49‰) and comparable to ␦
2
H
H20
values
reported for dry season flows measured in eight streams with
undeveloped watersheds in the Santa Monica Mountains, ap-
proximately 100 km south (Hibbs et al. 2012).
Some information is available on possible
␻
values for shred-
ders (Lepidostoma) (0.12) and scrapers (0.06) from Finlay et al. (2010)
in Wilkinson et al. (2015). Modifying the method of Wilkinson
et al. (2015), we adjusted the detritivore and algivore ␦
2
H end-
member values for each stream prior to mixing model calcula-
tions by adjusting for the contribution of environmental water
(
␻
1
) as follows: ␦
2
H
adjusted end-member
=[␦
2
H
end-member
−(
␻
1
×
Fig. 1. ␦
2
H, ␦
13
C, and ␦
15
N values of basal resources and specialist algivore and detritivore taxa from pool (left) and riffle (right) habitats of the study
streams. Streams are arranged, left to right, from smallest to largest subwatershed area (in parentheses, km
2
). Mean values for detritivores and
algivores from pools and riffles computed across streams are shown as dashed lines (from Table 1). Leaf and fine particulate organic matter (FPOM)
values are from pools only.
2114 Can. J. Fish. Aquat. Sci. Vol. 74, 2017
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by Santa Barbara (UCSB) on 12/31/17
For personal use only.

␦
2
H
H2O
)]/(1 −
␻
1
). Because environmental water contributions
change with trophic level, we also adjusted ␦
2
H values for each
predator (␦
2
H
predator
) as follows (modified from Wilkinson et al.
2015): ␦
2
H
adjusted predator
=[␦
2
H
predator
−(
␻
2
×␦
2
H
H2O
)]/(1 −
␻
2
),
where
␻
2
=1−(1−
␻
1
)
2
. We compared unadjusted mixing model
results with results adjusted for three values of environmental
water ␦
2
H (−45‰, −50‰, −55‰) reported by Hibbs et al. (2012) for
dry season flows and suggested from precipitation values at Wateri-
sotopes.
org, and for two levels of
␻
1
, i.e., 0.1, representative of shredders
and scrapers, and 0.2, suggested as a default value when
␻
is
unknown (Wilkinson et al. 2015). Proportional contributions of
algae-based resources to predator diet were then calculated using
the SIAR mixing model with algivore and detritivore end-member
and predator hydrogen and carbon stable isotope ratios.
Results
Isotope values of basal resources and specialist primary
consumers
Across all streams, ␦
2
H values of conditioned leaves, FPOM, and
detritivores were enriched, on average, by ⬃100‰ in pools and
⬃80‰ in riffles compared with values for Cladophora and algivores
(Table 1;Figs. 1a,1b). ␦
2
H values of algivores, but not of Cladophora
(p> 0.8), declined significantly with subwatershed area in pools
and riffles (pools: p< 0.002, R
2
= 0.76; riffles: p= 0.015, R
2
= 0.60,
n= 9). ␦
2
H values of detritivores were not correlated with sub-
watershed area (p> 0.5).
In both pools and riffles, ␦
2
H values of detritivores were signifi-
cantly enriched relative to values for algivores in accordance with
differences in the ␦
2
H values of their expected food sources (pools:
resource-consumer category, p< 0.001, F
[4,28]
= 16.3, category ×
subwatershed area, p= 0.001, F
[4,28]
= 5.9; Fig. 1a; riffles: category,
p< 0.001, F
[4,26]
= 10.4, category × subwatershed area, p= 0.13,
F
[4,26]
= 2.0; Šidák post hoc test results in Table 2;Fig. 1b). ␦
2
H
values of detritivores were not significantly different from those
for conditioned leaves and FPOM (Table 2). ␦
2
H values of algivores
were not significantly different from values for Cladophora in
pools, but they were enriched relative to values for these algae in
riffles (Table 2).
In both pools and riffles, ␦
13
C values also differed among basal
resources, detritivores, and algivores (pools: resource-consumer
category, p= 0.001, F
[4,32]
= 6.6, category × subwatershed area,
p= 0.06, F
[4,32]
= 2.5; Table 2;Fig. 1c; riffles: category, p= 0.05,
F
[4,28]
= 2.7, category × subwatershed area, p= 0.65, F
[4,28]
= 0.6;
Table 2;Fig. 1d). In contrast with the results for ␦
2
H, there was no
significant difference in ␦
13
C values between algivores and detri-
tivores or leaves in pools or riffles (Table 2;Fig. 1c). In both pools
and riffles, there were no significant differences in ␦
15
N values
among resource-consumer categories, and there were no cate-
gory × subwatershed area interaction effects (pools: category,
p= 0.36, F
[4,31]
= 1.1, category × subwatershed, p= 0.51, F
[4,31]
= 0.8;
Table 2;Fig. 1e; riffles: category, p= 0.70, F
[4,27]
= 0.5, category ×
subwatershed, p= 0.51, F
[4,27]
= 0.8; Table 2;Fig. 1f). Algivores,
Orthocladiinae, and Chironominae in both pools and riffles were
enriched in
15
N in Mission Creek relative to other creeks, which
was associated with the appreciable
15
N enrichment of algae at
this site (Figs. 1e,1f).
There was no difference between pool and riffle habitats in the
isotope values (␦
2
H, ␦
13
C, or ␦
15
N) of detritivores (Lepidostoma,
Paraleptophlebia) or algivores (Baetis,Eubrianax)(pvalues > 0.1, n=9,
paired ttests).
Relationship between isotope ratios and gut contents of
primary consumers
␦
2
H values of primary consumers from both pools and riffles
were negatively correlated with the proportions of gut contents
consisting of algae (Figs. 2a,2b). As expected, the gut contents of
Lepidostoma (shredders) and Paraleptophlebia (collectors) consisted
almost entirely of detrital particles (pools: 99.9% ± 0.1%, n=9;
riffles: 95.5% ± 2.5%, n=4,xˉ ± 1 SE), consistent with their enriched
␦
2
H values (Figs. 2a,2b). Algae constituted a much higher propor-
tion of the gut contents (pools: 70% ± 3%, n= 11; riffles: 40% ± 3%,
n= 10) of algivores (Baetis,Callibaetis,Centroptilum,Eubrianax), con-
sistent with their more depleted ␦
2
H values. ␦
13
C values of pri-
mary consumers from pools also were negatively correlated with
Table 2. Significance (pvalue) of Šidák multiple pairwise tests comparing the stable isotope signatures (␦
2
H, ␦
13
C, ␦
15
N) of
basal resources and specialist primary consumers from pool and riffle habitats.
Pools Riffles
Cladophora Algivores Detritivores FPOM Cladophora Algivores Detritivores FPOM
␦
2
H*␦
2
H
Cladophora Cladophora
Algivores 0.1, 1 Algivores 0.006
†
Detritivores 0.001 0.001 Detritivores 0.001 0.001
FPOM 0.001 0.001 1, 0.2 FPOM 0.001 0.001 0.5
Leaves 0.001 0.001 0.5, 0.8 0.6, 0.9 Leaves 0.001 0.001 1 0.5
␦
13
C*␦
13
C
Cladophora Cladophora
Algivores 0.1 Algivores 0.7
Detritivores 0.001 0.1
†
Detritivores 0.02 0.4
†
FPOM 0.001 0.001 0.3 FPOM 0.001 0.001 0.2
Leaves 0.001 0.2
†
1 0.3 Leaves 0.001 0.06
†
1 0.3
␦
15
N␦
15
N
Cladophora Cladophora
Algivores 0.9 Algivores 0.6
Detritivores 1
†
1
†
Detritivores 0.4
†
1
†
FPOM 1
†
0.8
†
1 FPOM 1
†
1
†
0.8
Leaves 0.6
†
1
†
0.6 0.8 Leaves 0.5
†
0.005 0.005
†
0.05
†
Note: FPOM, fine particulate organic matter.
*Category × subwatershed interaction effect significant (p< 0.05) in general linear model resulting in Šidák tests evaluated at subwatershed
areas of 5 and 10 km
2
, with pvalues provided for both subwatershed areas only if different. Significant differences (p< 0.05) identified in
pairwise tests are in boldface.
†
Result from pairwise test not consistent with known trophic relationships.
Page et al. 2115
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by Santa Barbara (UCSB) on 12/31/17
For personal use only.

the proportions of their gut contents consisting of algae, but the
relationship was weaker than that for ␦
2
H(Fig. 2c) and not signif-
icant for consumers from riffles (Fig. 2d).
Trophic support of predators: comparison of isotope ratios
␦
2
H values of common predator taxa declined with subwater-
shed area (pools: p< 0.001, R
2
= 0.80; riffles: p= 0.04, R
2
= 0.62, n=9;
Figs. 3a,3b). This analysis excluded the beetle Stictotarsus sp.,
which appeared to be a specialist predator on algivores based on
consistently depleted ␦
2
H values. In the smallest streams, ␦
2
H
values of predators in pools were generally similar to values for
detritivores and well separated from algivore values, suggesting
primary use of detritivorous prey in these streams (Fig. 3a). How-
ever, the dramatically lower values of damselflies of the genus
Archilestes in two larger streams with elevated nitrate concentra-
tions, and where the riparian vegetation had burned in a wildfire
(Mission and San Antonio creeks), suggested greater use of algivo-
rous prey at those sites (Fig. 3a; see also Cooper et al. 2015). These
patterns were less evident for predators sampled from riffles. No-
tably, stoneflies of the genera Isoperla and Calineura tended to have
depleted ␦
2
H values in the four smaller streams where they oc-
curred, suggesting a greater use of algivorous prey by these stone-
fly predators than by other predators at those sites (Fig. 3b).
There were no relationships between ␦
13
Cor␦
15
N values of
predator taxa from pools or riffles and subwatershed area
(pvalues > 0.1; Figs. 3c–3f). There was inconsistent separation
among streams between detritivores and algivores in ␦
13
C (range
of separation, +0.5‰ to −7.7‰) and ␦
15
N (+0.1‰ to +7.2‰) values,
although most predator ␦
13
C values were similar to detritivore
values in the three smallest streams (Figs. 3c–3f). Predators were
more enriched in
15
N than detritivores (Figs. 3e,3fversus 1e,1f),
even in small streams, reflecting the trophic enrichment from
prey to predator expected with N isotopes (Vander Zanden and
Rasmussen 2001;Post 2002). However, ␦
15
N values of predators in
Mission and San Antonio creeks also were enriched relative to
those in other streams (Figs. 3e,3f), suggesting the consumption of
algivores feeding on microalgae enriched in
15
N at these sites.
Trophic support of predators: mixing model estimates
versus gut contents
There was no difference between pools and riffles in the relation-
ship between either ␦
2
Hor␦
13
C values of predators and the propor-
tions of algivorous prey in their guts (all habitat and habitat ×
proportion of gut contents consisting of algivores effects were not
significant, p> 0.2), so data from pools and riffles were analyzed
and presented together (Figs. 4a,4c). ␦
2
H values of predators were
negatively correlated with the proportions of algivorous prey in
their guts (p< 0.001; Fig. 4a) and there was a positive correlation
between the percent algae-based contributions to predator diet,
as estimated using the SIAR model, and the proportions of
predator gut contents consisting of algivorous prey (p< 0.001;
Fig. 4b).
Including the possible effects of environmental water on prey
end-member and predator ␦
2
H values prior to mixing model cal-
culations increased estimates of algae-based prey resource use
relative to uncorrected values by 6% to 16%, depending on the
␻
and ␦
2
H
H2O
values used (Table 3). Increasing environmental water
content (
␻
) from 0.1 to 0.2 had a greater effect on mixing model
Fig. 2. Relationships between (a–b)␦
2
H and (c–d)␦
13
C tissue values and the proportion of algae in guts (algal biovolumes/algal + detrital biovolumes,
arcsine square-root transformed) for known detritivores (triangles: Lepidostoma,Paraleptophlebia), algivores (circles: Baetis,Callibaetis,Eubrianax,
Centroptilum), and other primary consumer taxa with variable diets (squares: Leucrocuta,Diphetor,Simulium) from pool (left) and riffle (right) habitats.
Linear regression equations (for significant relationships), coefficients of determination (R
2
), associated pvalues, and sample sizes are also shown.
2116 Can. J. Fish. Aquat. Sci. Vol. 74, 2017
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by Santa Barbara (UCSB) on 12/31/17
For personal use only.

estimates than decreasing ␦
2
H
H2O
values from –45‰ to –55‰
(Table 3). Despite the possible effects of environmental water on
mixing model results, the positive relationship between model
estimates of algae-based contributions to predator tissue sup-
port and the proportions of algivores in gut contents remained
highly significant across all combinations of
␻
and ␦
2
H
H2O
val-
ues (pvalues < 0.001, shown only for
␻
=0.2,␦
2
H
H2O
= −55‰ in
Fig. 4b).
Fig. 3. ␦
2
H, ␦
13
C, and ␦
15
N values of predator taxa from pool (left) and riffle (right) habitats of the study streams. Streams are arranged as in
Fig. 1. Values for detritivores (open bars) and algivores (solid bars) are shown for each stream where data were available.
Page et al. 2117
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by Santa Barbara (UCSB) on 12/31/17
For personal use only.

There was a weak relationship between predator ␦
13
C values
and the proportions of their gut contents composed of algivorous
prey (p= 0.001, R
2
= 0.19; Fig. 4c). However, there was no relation-
ship between SIAR model estimates of algivore-based carbon con-
tributions to predator diets and the proportions of algivorous
prey in predator guts (p= 0.5; Fig. 4d).
Discussion
Stable hydrogen isotope ratios (␦
2
H) more consistently and
clearly distinguished algae and algivores from conditioned terres-
trial leaves, FPOM, and detritivores in our headwater streams than
did carbon isotope ratios (␦
13
C), with a mean ␦
2
H separation
of >80‰ between algal and detrital resources and consumers
across habitats and streams. The correspondence between the ␦
2
H
values of specialist primary consumers and their expected diets in
our small streams is congruent with comparisons of ␦
2
H values
between consumers and their assumed food sources in streams in
northern California (Finlay et al. 2010). In that study, mean differ-
ences in ␦
2
H values between algivorous invertebrates and algae,
and between detritivorous invertebrates and terrestrial detritus,
were much smaller (<5‰) than the differences between these two
feeding groups and food sources (>60‰).
There also was remarkably good agreement between our results
and those of Finlay et al. (2010) for the mean ␦
2
H values of condi-
tioned leaves (−110‰ for our data versus −109‰ for Finlay et al.
2010), detritivores (pools, −108‰ versus −105‰), Cladophora (rif-
fles, −219‰ versus −225‰), and algivores (riffles, −189‰ versus
Table 3. Percentage difference in algae-
based contributions to predator diet
among Stable Isotope Analysis in R mixing
model estimates using measured algivore,
detritivore, and predator end-member
values compared with values adjusted for
environmental water using the different
␻
and ␦
2
H
H2O
values listed below (xˉ ± 1 SE,
n= 47 predator values).
␻
␦
2
H
H2O
(‰) 0.10 0.20
−45 7.3±0.8 16.1±1.6
−50 6.8±0.8 15.2±1.6
−55 6.3±0.8 9.2±1.4
Note:
␻
, proportion of consumer tissue
consisting of stream water hydrogen.
Fig. 4. Relationships between the proportions of predator gut contents composed of algivorous prey (algivorous prey/algivorous + other prey,
arcsine square-root transformed) and (a)␦
2
H values of predator taxa; (b) SIAR mixing model estimates (±1 SD) of algivore contributions to
predator nutrition based on ␦
2
H values of predator, algivore, and detritivore tissues for each stream (Figs. 1,3) (open points and solid line
represent unadjusted values; solid points and dashed line represent an example with values adjusted for environmental water (Table 3),
setting
␻
1
= 0.2, ␦
2
H
H20
= −55‰); (c)␦
13
C values of predator taxa; and (d) SIAR mixing model estimates of algivore contributions to predator
nutrition based on ␦
13
C values of predator, algivore, and detritivore tissues for each stream. Regression lines were fit to data from pools and
riffles combined. Regression equations, coefficients of determination, associated pvalues, and sample sizes are also shown for significant
relationships.
2118 Can. J. Fish. Aquat. Sci. Vol. 74, 2017
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by Santa Barbara (UCSB) on 12/31/17
For personal use only.

−175‰), despite differences in the riparian vegetation, consumer
taxa, and hydrological and other conditions between our study
sites. The close correspondence and wide separation in ␦
2
H values
between specialist consumers and their food resources in our
study and that of Finlay et al. (2010) suggest that the effects of
trophic and tissue fractionation, and ␦
2
H values of environmental
water, on body tissue ␦
2
H values (Solomon et al. 2009;Wilkinson
et al. 2015) were similar between our sites and those of Finlay et al.
(2010).
In contrast, Jardine et al. (2009), working in streams in New
Brunswick, Canada, reported a smaller overall mean difference
and more variation in the separation of ␦
2
H values in stream
“biofilm” versus leaves (alder) (mean difference±1SDof39.2‰ ±
31.2‰, all 31 sites, from Table 1, range from +15.9‰ to −109‰)
than found in our study and that of Finlay et al. (2010). Further-
more, the ␦
2
H values for leaves and biofilm overlapped at more
sites than did ␦
13
C values, precluding the application of mixing
models using ␦
2
H data because of insufficient ␦
2
H end-member
separation (as based on criteria of nonoverlapping standard devi-
ations around mean values). They suggested that local factors,
such as the chemistry and ␦
2
H values of the water body, and
perhaps consumer lipid content, affected the relative effective-
ness of carbon and hydrogen isotope ratios in tracing food web
pathways at their sites.
Although the separation in ␦
2
H values between algivore and
detritivore taxa was distinct in our study streams, ␦
2
H values of
algivores declined with increasing subwatershed area and with
the loss of canopy cover owing to wildfire in two of the larger
subwatersheds (Mission, San Antonio, Cooper et al. 2015). Finlay
et al. (2010) also reported decreasing ␦
2
H values with increasing
catchment area for filamentous green algae, diatoms, and herbiv-
orous caddisflies in northern California streams. One possible ex-
planation for this pattern is that the microalgae consumed by
algivores were more enriched in ␦
2
H in the smaller than larger
streams. Finlay et al. (2010) proposed that such enrichment could
arise through greater fractionation of hydrogen isotopes during
photosynthesis or higher algal lipid concentrations, which are
depleted in deuterium (
2
H), in large than in small streams.
However, in our study, ␦
2
H values of the filamentous macroal-
gae Cladophora did not decrease with subwatershed area, which
is inconsistent with this explanation.
Another possibility is that algivores consumed more detritus in
small than in large streams, congruent with McNeely et al.’s (2006)
results showing increasing algal biomass, and increasing propor-
tions of algae in the guts of some taxa, with increases in watershed
area. This possibility is also supported by studies showing flexibil-
ity in primary consumer diets, even within functional groups,
including scrapers or grazers, generally considered to be largely
algivorous, along stream or river continua (Rosi-Marshall et al.
2016;Collins et al. 2016). Until more information is available, re-
searchers using algivore ␦
2
H values as proxies for the microalgal
end-member in mixing models need to consider that the ␦
2
H val-
ues of microalgae may vary among streams and that algivores may
derive increased nutritional support from terrestrial detritus (ver-
sus algae) in small streams. Although some shredder or collector
taxa have been reported to consume large amounts of algae in
other systems (Rosi-Marshall et al. 2016), the detritivore taxa rep-
resenting the terrestrial end-member in our study were distinctly
detritivorous across streams, as evidenced by enriched ␦
2
H values
that were widely separated from algivore values and by gut con-
tents consisting almost entirely of detrital particles.
In contrast with ␦
2
H, ␦
13
C values of basal resources, algivores,
and detritivores were not consistently well separated in all
streams. A lack of consistent separation in ␦
13
C values among
basal resources (algae, coarse or fine particulate organic matter)
has limited the utility of carbon isotopes in tracing the sources
and fates of organic matter in freshwater ecosystems and has been
attributed to variation in the ␦
13
C values of algae rather than
terrestrial detritus (France 1996;Doucett et al. 1996,2007;Finlay
2001;McNeely et al. 2006;Ishikawa et al. 2012). ␦
13
C values of
terrestrial detritus vary over a relatively narrow range (e.g., con-
ditioned leaves: −29.0‰ ± 0.3‰, our study, Table 1; −28.2‰ ± 0.2‰
(±1 SE) in Finlay 2001), whereas ␦
13
C values of algae and algivores
can vary widely with factors that affect the fractionation of carbon
isotopes during the uptake of CO
2
and dissolved inorganic carbon
by algae, such as algal productivity and aqueous CO
2
availability
(Finlay 2001).
Using nitrogen isotopes to trace trophic pathways also requires
sufficient and reliable separation in resource end-member values.
Nitrogen isotope ratios can be highly variable in time and space
owing to variation in the isotopic signature and availability of
dissolved inorganic N (DIN) assimilated by producers, changes in
the ␦
15
N values of detritus during decomposition by microbes, and
variation in producer ␦
15
N values resulting from anthropogenic
DIN inputs (Peterson 1999;Post 2002;Moore et al. 2014). In this
study, stable nitrogen isotope values were not consistently or
clearly different between different resource and primary con-
sumer categories. As a consequence, ␦
15
N values were not useful
in identifying basal resource contributions to consumers in our
study streams.
Predators in Mission and San Antonio creeks, however, were
enriched in
15
N relative to predators in other streams, suggesting
the consumption of algivores feeding on microalgae with en-
riched ␦
15
N values, engendered by increased light and nutrient
levels following a wildfire the year before (Cooper et al. 2015).
Interestingly, there also was some
15
N enrichment of the detriti-
vores, Lepidostoma and Paraleptophlebia (⬃+2‰), in these two
streams. ␦
15
N values of conditioned leaves were not enriched,
suggesting that these detritivores were deriving some nutri-
tional support from microbial cells that had incorporated DIN
(Hamilton et al. 2004). Variation in the ␦
15
N values of predators
across streams emphasizes the importance of calculating incre-
ments in
15
N values from prey to predators for each stream
independently to assess the trophic level of individual taxa
(Vander Zanden and Rasmussen 1999;Post 2002) and also illus-
trates the potential use of ␦
15
N to trace nutrient enrichment
effects through the food web (McClelland et al. 1997;Moore
et al. 2014).
There was better agreement (higher R
2
values) of ␦
2
H than ␦
13
C
values with recent consumer diets. One likely explanation for this
finding is the more consistent and greater separation of ␦
2
H than
␦
13
C values between algal and terrestrial basal resources across
streams. The better congruence of ␦
2
H than ␦
13
C signatures with
recent diets also could occur because of differences in isotope
turnover times, such as hydrogen isotopes turning over more
rapidly than carbon isotopes, as has been found for some verte-
brate tissues (Wolf et al. 2012), thus better reflecting recently con-
sumed food items. The concordance between consumer ␦
2
H
values and gut contents supports conclusions regarding the rela-
tive contributions of algae and terrestrial detritus to primary con-
sumer diets and nutrition, but the time frame over which ␦
2
H
values integrate past consumption is unknown (e.g., Jardine et al.
2005;McNeely et al. 2007;Li and Dudgeon 2008).
The efficacy of hydrogen isotope ratios in delineating food webs
in our streams was further supported by the significant correla-
tion between algivore contributions to predator nutrition, esti-
mated using a commonly applied Bayesian mixing model, and the
proportions of algivorous prey in predator gut contents, a rela-
tionship that was unaffected by adjusting measured resource and
predator ␦
2
H values for the possible effects of environmental wa-
ter. The variable and smaller separation of carbon isotope ratios
between algivores and detritivores resulted in a lack of agreement
between estimates of algivore contributions to predator diet us-
ing the mixing model and predator gut contents, and there was no
pattern of carbon isotope ratios with subwatershed area to indi-
cate better separation of end-members in larger catchments
Page et al. 2119
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by Santa Barbara (UCSB) on 12/31/17
For personal use only.

(Finlay et al. 2010). As a consequence, carbon isotope signatures
were not a consistent indicator of resource use by consumers in
our streams and were not reliable as a complementary tracer. Our
results indicate that individual streams need to be evaluated a
priori to determine whether sufficient separation exists in re-
source end-member values before specific stable isotopes are used
to delineate food web structure.
In conclusion, although the importance of allochthonous ter-
restrial organic matter subsidies and autochthonous algal produc-
tion to the food webs of streams is widely recognized, the ability
of stable carbon isotope ratios to quantify the relative contribu-
tions of these resources to higher trophic levels has been limited
in many cases. Our findings support the use of hydrogen isotope
ratios to trace the sources and fates of organic matter in the food
webs of small headwater streams of southern California, and
probably other streams in arid and semiarid climates and (or)
during drought (Doucett et al. 2007). Our results also support the
use of hydrogen isotope ratios in delineating food web structure
along longitudinal river continua (Finlay et al. 2010;Collins et al.
2016) and in evaluating food web responses to human activities
and natural disturbances that alter riparian shading, riparian in-
puts, and algal production, such as land use changes (logging,
livestock grazing, agricultural and urban development), wildfires,
and floods (Kiffney 2008;Wootton 2012;Cooper et al. 2013,2015),
producing more consistent and clearer results than those based
on carbon isotope ratios.
Acknowledgements
We thank Kristen Mollura for assistance in the laboratory and
Kyle Emery for discussion and comments on the manuscript. We
also thank Robert Sherwood of the Santa Barbara Botanic Garden,
Ralph Philbrick of the San Jose Trout Club, Terri Bowman of El
Capitan Canyon, the Land Trust for Santa Barbara County, and Jeff
Brinkman for assistance with obtaining access to study sites. This
research was supported by funding from the US National Science
Foundation (awards OCE-0620276, OCE-1232779, and DEB-0952599)
and the University of California Santa Barbara Academic Senate.
References
Anderson, C., and Cabana, G. 2007. Estimating the trophic position of aquatic
consumers in river food webs using stable nitrogen isotopes. J. N. Am. Benthol.
Soc. 26: 273–285. doi:10.1899/0887-3593(2007)26[273:ETTPOA]2.0.CO;2.
Armitage, P., Cranston, P.S., and Pinder, L.C.V. 1997. The Chironomidae. The
biology and ecology of non-biting midges. Springer Science, Dordrecht, the
Netherlands.
Bowen, G.J., and Revenaugh, J. 2003. Interpolating the isotopic composition of
modern meteoric precipitation. Wat. Res. Res., 39: 1299. doi:10.1029/
2003WR002086.
Bowen, G.J., Wassenaar, L.I., and Hobson, K.A. 2005. Global application of stable
hydrogen and oxygen isotopes to wildlife forensics. Oecologia, 143: 337–348.
doi:10.1007/s00442-004-1813-y.
Caraco, N., Bauer, J.E., Cole, J.J., Petsch, S., and Raymond, P. 2010. Millennial-aged
organic carbon subsidies to a modern river food web. Ecology, 91: 2385–2393.
doi:10.1890/09-0330.1. PMID:20836460.
Chapman, D.W., and Demory, R.L. 1963. Seasonal changes in the food ingested
by aquatic insect nymphs in two Oregon streams. Ecology, 44: 140–146.
Cole, J.J., Carpenter, S.R., Pace, M.L., Van de Bogert, M.C., Kitchell, J.L., and
Hodgson, J.R. 2006. Differential support of lake food webs by three types of
terrestrial organic carbon. Ecol. Lett. 9: 558– 568. doi:10.1111/j.1461-0248.2006.
00898.x. PMID:16643301.
Cole, J.J., Carpenter, S.R., Kitchell, J., Pace, M.L., Solomon, C.T., and Weidel, B.
2011. Strong evidence for terrestrial support of zooplankton in small lakes
based on stable isotopes of carbon, nitrogen, and hydrogen. Proc. Natl. Acad.
Sci. U.S.A. 108: 1975–1980. doi:10.1073/pnas.1012807108. PMID:21245299.
Collins, S.M., Kohler, T.J., Thomas, S.A., Fetzer, W.W., and Flecker, A.S. 2016. The
importance of terrestrial subsidies in stream food webs varies along a stream
size gradient. Oikos, 125: 674–685. doi:10.1111/oik.02713.
Cooper, S.D., Lake, P.S., Sabater, S., Melack, J.M., and Sabo, J.L. 2013. The effects
of land use changes on streams and rivers in Mediterranean climates. Hydro-
biologia, 719: 383–425. doi:10.1007/s10750-012-1333-4.
Cooper, S.D., Page, H.M., Wiseman, S., Klose, K., Bennett, D., Even, T., Sadro, S.,
Nelson, C.E., and Dudley, T.L. 2015. Physicochemical and biological responses
of streams to wildfire severity in riparian zones. Freshw. Biol. 60: 2600–2619.
doi:10.1111/fwb.12523.
Cross, W.F., Baxter, C.V., Rosi-Marshall, E.J., Hall, R.O., Jr., Kennedy, T.A.,
Donner, K.C., Wellard Kelly, H.A., Seegert, S.E.Z., Behn, K.E., and Yard, M.D.
2013. Food-web dynamics in a large river discontinuum. Ecol. Monogr. 83:
311–337. doi:10.1890/12-1727.1.
Cummins, K.W. 1974. Structure and function of stream ecosystems. Bioscience,
24: 631–641. doi:10.2307/1296676.
Cummins, K.W., and Klug, M.J. 1979. Feeding ecology of stream invertebrates.
Annu. Rev. Ecol. Syst. 10: 147–172. doi:10.1146/annurev.es.10.110179.001051.
Davies, A.L., and Gee, J.H.R. 1993. A simple periphyton sampler for algal biomass
estimates in streams. Freshw. Biol. 30: 47–51. doi:10.1111/j.1365-2427.1993.
tb00787.x.
Day, R.W., and Quinn, G.P. 1989. Comparisons of treatments after an analysis of
variance in ecology. Ecol. Monogr. 59: 433–463. doi:10.2307/1943075.
Doucett, R.R., Barton, D.R., Guiguer, K., Power, G., and Drimmie, R.J. 1996. Com-
ment: Critical examination of stable isotope analysis as a means for tracing
carbon pathways in stream ecosystems. Can. J. Fish. Aquat. Sci. 53(8): 1913–
1915. doi:10.1139/f96-114.
Doucett, R.R., Marks, J.C., Blinn, D.W., Caron, M., and Hungate, B.A. 2007. Mea-
suring terrestrial subsidies to aquatic food webs using stable isotopes of
hydrogen. Ecology, 88: 1587–1592. doi:10.1890/06-1184. PMID:17601150.
Estep, M., and Dabrowski, H. 1980. Tracing food webs with stable hydrogen
isotopes. Science, 209: 1537–1538. doi:10.1126/science.209.4464.1537. PMID:
17745967.
Finlay, J.C. 2001. Stable-carbon isotope ratios of river biota: Implications for
energy flow in lotic food webs. Ecology, 82: 1052–1064. doi:10.2307/2679902.
Finlay, J.C. 2004. Patterns and controls of lotic algal stable carbon isotope ratios.
Limnol. Oceanogr. 49: 850–861. doi:10.4319/lo.2004.49.3.0850.
Finlay, J.C., Power, M.E., and Cabana, G. 1999. Effects of water velocity on algal
carbon isotope ratios: Implications for river food web studies. Limnol. Ocean-
ogr. 44: 1198–1203. doi:10.4319/lo.1999.44.5.1198.
Finlay, J.C., Doucett, R.R., and McNeely, C. 2010. Tracing energy flow in stream
food webs using stable isotopes of hydrogen. Freshw. Biol. 55: 941–951. doi:
10.1111/j.1365-2427.2009.02327.x.
Flockhart, D.T.T., Wassenaar, L.I., Martin, T.G., Hobson, K.A., Wunder, M.B., and
Norris, D.R. 2013. Tracking multi-generational colonization of the breeding
grounds by monarch butterflies in eastern North America. Proc. R. Soc. B Biol.
Sci. 280: 20131087. doi:10.1098/rspb.2013.1087.
France, R.L. 1996. Carbon-13 conundrums: Limitations and cautions in the use of
stable isotope analysis in stream ecotonal research. Can. J. Fish. Aquat. Sci.
53(8): 1916–1919. doi:10.1139/f96-121.
Fry, B. 2006. Stable isotope ecology. Springer, New York.
Hamilton, S.K., Tank, J.L., Raikow, D.F., Siler, E.R., Dorn, N.J., and Leonard, N.E.
2004. The role of instream vs allochthonous N in stream food webs: Modeling
the results of an isotope addition experiment. J. N. Am. Benthol. Soc. 23:
429–448. doi:10.1899/0887-3593(2004)023<0429:TROIVA>2.0.CO;2.
Hibbs, B.J., Hu, W., and Ridgeway, R. 2012. Origin of stream flows at the
wildlands-urban interface, Santa Monica Mountains, California, U.S.A. Envi-
ron. Eng. Geosci. 18: 51–64. doi:10.2113/gseegeosci.18.1.51.
Hillebrand, H., Durselen, C.-D., Kirschtel, D., Pollingher, U., and Zohary, T. 1999.
Biovolume calculation for pelagic and benthic microalgae. J. Phycol. 35: 403–
424. doi:10.1046/j.1529-8817.1999.3520403.x.
Ishikawa, N.F., Doi, H., and Finlay, J.C. 2012. Global meta-analysis for controlling
factors on carbon stable isotope ratios of lotic periphyton. Oecologia, 170:
541–549. doi:10.1007/s00442-012-2308-x. PMID:22466861.
Jardine, T.D., Curry, R.A., Heard, K.S., and Cunjak, R.A. 2005. High fidelity: Iso-
topic relationship between stream invertebrates and their gut contents. J. N.
Am. Benthol. Soc. 24: 290–299. doi:10.1899/04-092.1.
Jardine, T.D., Kidd, K.A., and Cunjak, R.A. 2009. An evaluation of deuterium as a
food source tracer in temperate streams of eastern Canada. J. N. Am. Benthol.
Soc. 28: 885–893. doi:10.1899/09-046.1.
Jardine, T.D., Hadwen, W.L., Hamilton, S.K., Hladyz, S., Mitrovic, S.M., Kidd, K.A.,
Tsoi, W.Y., Spears, M., Westhorpe, D.P., Fry, V.M., Sheldon, F., and Bunn, S.E.
2014. Understanding and overcoming baseline isotopic variability in running
waters. Riv. Res. Appl. 30: 155–165. doi:10.1002/rra.2630.
Karlsson, J., Berggren, M., Ask, J., Byström, P., Jonsson, A., Laudon, J., and
Jansson, M. 2012. Terrestrial organic matter support of lake food webs: Evi-
dence from lake metabolism and stable hydrogen isotopes of consumers.
Limnol. Oceanogr. 57: 1042–1048. doi:10.4319/lo.2012.57.4.1042.
Kelly, J.F., Bridge, E.S., Fudickar, A.M., and Wassenaar, L.I. 2009. A test of com-
parative equilibration for determining non-exchangeable stable hydrogen
isotope values in complex organic materials. Rapid Commun. Mass Spec-
trom. 23: 2316–2320. doi:10.1002/rcm.4150.
Kiffney, P.M. 2008. Response of lotic producer and consumer trophic levels to
gradients of resource supply and predation pressure. Oikos, 117: 1428–1440.
doi:10.1111/j.0030-1299.2008.16584.x.
Klose, K., Cooper, S.D., and Bennett, D. 2015. Effects of wildfire on stream algal
abundance, community structure, and nutrient limitation. Freshw. Sci. 34:
1494–1509. doi:10.1086/683431.
Lancaster, J., Bradley, D.C., Hogan, A., and Waldron, S. 2005. Intraguild om-
nivory in predatory stream insects. J. Anim. Ecol. 74: 619–629. doi:10.1111/j.
1365-2656.2005.00957.x.
Layman, C.A., Araujo, M.S., Boucek, R., Hammerschlag-Peyer, C.M., Harrison, E.,
Jud, Z.R., Matich, P., Rosenblatt, A.E., Vaudo, J.J., Yeager, L.A., Post, D.M., and
2120 Can. J. Fish. Aquat. Sci. Vol. 74, 2017
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by Santa Barbara (UCSB) on 12/31/17
For personal use only.

Bearhop, S. 2012. Applying stable isotopes to examine foodweb structure: an
overview of analytical tools. Biol. Rev. 87: 545–562. doi:10.1111/j.1469-185X.
2011.00208.x. PMID:22051097.
Li, A.O.Y., and Dudgeon, D. 2008. Food resources of shredders and other benthic
macroinvertebrates in relation to shading conditions in tropical Hong Kong
streams. Freshw. Biol. 53: 2011–2025. doi:10.1111/j.1365-2427.2008.02022.x.
Macko, S.A., Estep, M.L.F., and Lee, W.Y. 1983. Stable hydrogen isotope analysis
of foodwebs on laboratory and field populations of marine amphipods. J. Exp.
Mar. Biol. Ecol. 72: 243–249. doi:10.1016/0022-0981(83)90109-0.
McClelland, J.W., Valiela, I., and Michener, R.H. 1997. Nitrogen-stable isotope
signatures in estuarine food webs: a record of increasing urbanization on
isotope signatures in coastal watersheds. Limnol. Oceanogr. 42: 930–937.
doi:10.4319/lo.1997.42.5.0930.
McNeely, C., Clinton, S.M., and Erbe, J.M. 2006. Landscape variation in C sources
of scraping primary consumers in streams. J. N. Am. Benthol. Soc. 25: 787–
799. doi:10.1899/0887-3593(2006)025[0787:LVICSO]2.0.CO;2.
McNeely, C., Finlay, J.C., and Power, M.E. 2007. Grazer traits, competition, and
carbon sources to a headwater-stream food web. Ecology, 88: 391–401. doi:
10.1890/0012-9658(2007)88[391:GTCACS]2.0.CO;2. PMID:17479757.
Mihuc, T., and Toetz, D. 1994. Determination of diets of alpine aquatic insects
using stable isotopes and gut analyses. Am. Midl. Nat., 131: 146–155. doi:10.
2307/2426617.
Mihuc, T.B., and Minshall, G.W. 1995. Trophic generalists vs. trophic specialists:
Implications for food web dynamics in post-fire streams. Ecology, 76: 2361–
2372. doi:10.2307/2265813.
Moore, J.W. 1977. Some factors effecting algal consumption in subarctic
Ephemeroptera, Plecoptera and Simuliidae. Oecologia, 27: 261–273. doi:10.
1007/BF00347471.
Moore, J.W., Lambert, T.D., Heady, W.N., Honig, S.E., Osterback, A.-M.,
Phillis, C.C., Quiros, A.L., Retford, N.A., and Herbst, D.B. 2014. Anthropogenic
land-use signals propagate through stream food webs in a California, USA,
watershed. Limnologica, 46: 124–130. doi:10.1016/j.limno.2014.01.005.
Parnell, A., Inger, R., Bearhop, S., and Jackson, A.L. 2010. Source partitioning
using stable isotopes: coping with too much variation. PLoS ONE, 5: e9672.
doi:10.1371/journal.pone.0009672. PMID:20300637.
Peters, J.M., Wolf, N., Stricker, C.A., Collier, T.R., and Martínez del Rio, C. 2012.
Effects of trophic level and metamorphosis on discrimination of hydrogen
isotopes in a plant–herbivore system. PloS ONE, 7: e32744. doi:10.1371/journal.
pone.0032744. PMID:22470423.
Peterson, B.J. 1999. Stable isotopes as tracers of organic matter input and trans-
fer in benthic food webs: a review. Act. Oecol. 20: 479– 487. doi:10.1016/S1146-
609X(99)00120-4.
Peterson, B.J., and Fry, B. 1987. Isotopes in ecosystem studies. Annu. Rev. Ecol.
Syst. 18: 293–320. doi:10.1146/annurev.es.18.110187.001453.
Post, D.M. 2002. Using stable isotopes to estimate trophic position: models,
methods, and assumptions. Ecology, 83: 703–718. doi:10.1890/0012-9658(2002)
083[0703:USITET]2.0.CO;2.
Rosi-Marshall, E.J., Vallis, K.L., Baxter, C.V., and Davis, J.M. 2016. Retesting a
prediction of the River Continuum Concept: autochthonous versus alloch-
thonous resources in the diets of invertebrates. Freshw. Sci. 35: 534–543.
doi:10.1086/686302.
Skinner, M.M., Martin, A.A., and Moore, B.C. 2016. Is lipid correction necessary in
the stable isotope analysis of fish tissues? Rapid Commun. Mass Spectrom.
30: 881–889. doi:10.1002/rcm.7480. PMID:26969930.
Smith, B.N., and Epstein, S. 1970. Biogeochemistry of the stable isotopes of
hydrogen and carbon in salt marsh biota. Plant Physiol. 46: 738–742. doi:10.
1104/pp.46.5.738. PMID:16657539.
Solomon, C.T., Cole, J.J., Doucett, R.R., Pace, M.L., Preston, N.D., Smith, L.E., and
Weidel, B.C. 2009. The influence of environmental water on the hydrogen
stable isotope ratio in aquatic consumers. Oecologia, 161: 313–324. doi:10.
1007/s00442-009-1370-5. PMID:19471971.
Soto, D.X., Wassenaar, L.I., and Hobson, K.A. 2013. Stable hydrogen and oxygen
isotopes in aquatic food webs are tracers of diet and provenance. Funct. Ecol.
27: 535–543. doi:10.1111/1365-2435.12054.
Vander Zanden, H.B, Soto, D.X., Bowen, G.J., and Hobson, K.A. 2016. Expanding
the isotopic toolbox: applications of hydrogen and oxygen stable isotope
ratios to food web studies. Front. Ecol. Evol. 4: 1–19. doi:10.3389/fevo.2016.
00020.
Vander Zanden, M.J.V., and Rasmussen, J.B. 1999. Primary consumer ␦
13
C and
␦
l5
N and the trophic position of aquatic consumers. Ecology, 80: 1395–1404.
doi:10.1890/0012-9658(1999)080[1395:PCCANA]2.0.CO;2.
Vander Zanden, M.J., and Rasmussen, J.B. 2001. Variation in ␦
l5
N and ␦
13
C trophic
fractionation: Implications for aquatic food web studies. Limnol. Oceanogr.
46: 2061–2066. doi:10.4319/lo.2001.46.8.2061.
Verkaik, I., Rieradevall, M., Cooper, S.D., Melack, J.M., Dudley, T.L., and Prat, N.
2013. Fire as a disturbance in mediterranean climate streams. Hydrobiologia,
719: 353–382. doi:10.1007/s10750-013-1463-3.
Wassenaar, L.I., and Hobson, K.A. 2003. Comparative equilibration and online
technique for determination of non-exchangeable hydrogen of keratins for
use in animal migration studies. Isotopes Environ. Health Stud. 39: 211–217.
doi:10.1080/1025601031000096781. PMID:14521282.
Wilkinson, G.M., Carpenter, S.R., Cole, J.J., and Pace, M.L. 2013. Terrestrial sup-
port of pelagic consumers: patterns and variability revealed by a multi-lake
study. Freshw. Biol. 58: 2037–2049. doi:10.1111/fwb.12189.
Wilkinson, G.M., Cole, J.J., and Pace, M.L. 2015. Deuterium as a food source
tracer: sensitivity to environmental water, lipid content, and hydrogen ex-
change. Limnol. Oceanogr. Methods, 13: 213–223. doi:10.1002/lom3.10019.
Wolf, N., Newsome, S.D., Fogel, M.L., and del Rio, C.M. 2012. An experimental
exploration of the incorporation of hydrogen isotopes from dietary sources
into avian tissues. J. Exp. Biol. 215: 1915–1922. doi:10.1242/jeb.065219. PMID:
22573770.
Wootton, J.T. 2012. River food web response to large-scale riparian zone manip-
ulations. PLoS ONE, 7: e51839. doi:10.1371/journal.pone.0051839. PMID:
23284786.
Page et al. 2121
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by Santa Barbara (UCSB) on 12/31/17
For personal use only.