Effects of a diatom ecosystem engineer (Didymosphenia geminata) on stream food webs: implications for native fishes

Abstract

Stream habitat changes affecting primary consumers often indirectly impact secondary consumers such as fishes. Blooms of the benthic algae Didymosphenia geminata (Didymo) are known to affect stream macroinvertebrates, but the potential indirect trophic impacts on fish consumers are poorly understood. In streams of the Kootenai River basin, we quantified the diet, condition, and growth rate of species of trout, char, and sculpin. In 2018, macroinvertebrate taxa composition was different between a stream with Didymo and a stream without, but trout diets, energy demand, and growth rates were similar. Trout abundance was higher in the stream with Didymo, but the amount of drifting invertebrates was higher in the stream without. In 2019, we surveyed 28 streams with a gradient of coverage. Didymo abundance was correlated only with the percentage of aquatic invertebrates in trout diets and was not related to diets of char or sculpin or condition of any species. Thus, we found no evidence for a trophic link between Didymo blooms and the condition or growth of trout, char, or sculpin in mountainous headwater streams.

Full text

ARTICLE
Effects of a diatom ecosystem engineer (Didymosphenia geminata)
on stream food webs: implications for native fishes
Niall G. Clancy, Janice Brahney, James Dunnigan, and Phaedra Budy
Abstract: Stream habitat changes affecting primary consumers often indirectly impact secondary consumers such as fishes.
Blooms of the benthic algae Didymosphenia geminata (Didymo) are known to affect stream macroinvertebrates, but the potential
indirect trophic impacts on fish consumers are poorly understood. In streams of the Kootenai River basin, we quantified the
diet, condition, and growth rate of species of trout, char, and sculpin. In 2018, macroinvertebrate taxa composition was different
between a stream with Didymo and a stream without, but trout diets, energy demand, and growth rates were similar. Trout abun-
dance was higher in the stream with Didymo, but the amount of drifting invertebrates was higher in the stream without. In 2019,
we surveyed 28 streams with a gradient of coverage. Didymo abundance was correlated only with the percentage of aquatic inverte-
brates in trout diets and was not related to diets of char or sculpin or condition of any species. Thus, we found no evidence for a
trophic link between Didymo blooms and the condition or growth of trout, char, or sculpin in mountainous headwater streams.
Résumé : Les modifications des habitats de cours d’eau qui ont une incidence sur les consommateurs primaires ont souvent
deseffetsindirectssurlesconsommateurssecondairescommelespoissons.S’il est établi que les proliférations de l’algue
benthique Didymosphenia geminata (didymo) ont une incidence sur les macroinvertébrés lotiques, leurs impacts trophiques
indirects potentiels sur les poissons consommateurs ne sont pas bien compris. Nous avons quantifiélerégimealimentaire,
l’embonpoint et le taux de croissance de truites, d’ombles et de chabots dans des cours d’eau du bassin de la rivière Koote-
nai. En 2018, un cours d’eauoùl’algue didymo était présente et un autre dont elle était absente présentaient des composi-
tions taxonomiques de macroinvertébrés différentes, alors que les régimes alimentaires, la demande énergétique et les
taux de croissance des truites y étaient semblables. L’abondance de truites était plus grande dans le cours d’eau contenant
des algues didymo, mais la quantité d’invertébrés à la dérive était plus grande dans le cours d’eau exempt de l’algue. En
2019, nous avons inventorié 28 cours d’eau définissant un gradient de couverture par l’algue didymo. L’abondance des
algues était seulement corrélée au pourcentage d’invertébrésaquatiquesdanslerégimealimentairedestruitesetn’était
pas corrélée au régime alimentaire des ombles ou des chabots, ni à l’embonpoint d’aucune espèce. Ainsi, nous n’avons rel-
evéaucunepreuved’un lien entre les proliférations d’algues didymo et l’embonpoint ou la croissance des truites, ombles
ou chabots dans des cours d’eau supérieurs de montagne. [Traduit par la Rédaction]
Introduction
Trophic structure determines the availability of food resources
to organisms within a food web. The growth of these organisms
depends not only on the amount of energy in a system, but also
on the proportion of that energy that can be consumed and digested
(Horton 1961;Huryn 1996;Bellmore et al. 2013). In human-
modified landscapes, activities such as logging, mineral extrac-
tion, and river impoundment have altered in-stream habitats and
riparian areas (Hand et al. 2018), often resulting in a lack of habi-
tat complexity and nutrient availability, and thus affecting the
amount of energy available as food for aquatic organisms (Meredith
et al. 2014;Minshall et al. 2014). Such change can alter stream mac-
roinvertebrate assemblages and impact consumers of both larval
and adult life stages of aquatic insects (Power et al. 1996;Nakano
et al. 1999;Baxter et al. 2005). Fo r many fish species, aquatic mac-
roinvertebrates are a primary source of energy (Behnke 2010).
Understanding how specific environmental changes alter the
flow of in-stream energy to fish can thus be of great importance to
conservation and management efforts (Cross et al. 2011;Bellmore
et al. 2012;Scholl et al. 2019).
Organisms that alter physical and biological habitats upon which
other organisms rely are ubiquitous in many environments and have
long been a topic of biological study (Darwin 1881). In fresh waters, a
wide variety of animal ecosystem engineers such as dam-building
beavers, nest-digging fishes, pebble-moving crayfish, sediment-
disturbing tadpoles, and net-spinning caddisflies alter aquatic
habitat through the manipulation of their environment (Moore
2006;Albertson and Daniels 2016;Tumolo et al. 2019). Other spe-
cies, such as corals, act as agents of biogeomorphic change through
the growth of their own physical structures —altering or creating
habitats due to their own architectures (Jones et al. 1994). Corre-
sponding ecosystem engineers in streams are almost all autotrophs
(but see Beckett et al. 1996 and Hopper et al. 2019)andincludetrees
that form large woody debris (LWD) jams, macrophytes that shelter
juvenile fishes, and algae that harbor invertebrates.
One such autotrophic ecosystem engineer is the diatomaceous
algae Didymosphenia geminata (hereinafter, Didymo). Overgrowths
Received 10 April 2020. Accepted 28 September 2020.
N.G. Clancy* and J. Brahney. Department of Watershed Sciences, Utah State University, 5210 Old Main Hill, Logan, UT 84322, USA.
J. Dunnigan. Montana Fish, Wildlife & Parks, 385 Fish Hatchery Road, Libby, MT 59923, USA.
P. Budy. US Geological Survey, Utah Cooperative Fish and Wildlife Research Unit & Department of Watershed Sciences, Utah State University, 5205 Old
Main Hill, Logan, UT 84322, USA.
Corresponding author: Niall G. Clancy (email: niall.clancy@mt.gov).
*Present address: Montana Fish, Wildlife & Parks, 490 N. Meridian Road, Kalispell, MT 59901, USA.
Copyright remains with the author(s) or their institution(s). Permission for reuse (free in most cases) can be obtained from copyright.com.
Can. J. Fish. Aquat. Sci. 00: 1–11 (0000) dx.doi.org/10.1139/cjfas-2020-0121 Published at www.nrcresearchpress.com/cjfas on 20 January 2021.
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
1
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 63.153.124.239 on 01/19/21
For personal use only.

(colloquially, blooms) of this North American native are charac-
terized by the production of a long polysaccharide stalk from
individual diatoms, which can cover large areas of the substrate.
Prodigious growth of this bifurcating stalk differentiates Didymo
from most other diatoms, and the interwoven aggregate of stalks
from a multitude of cells produces a thick mat that contains
other algae, detritus, and macroinvertebrates —sometimes cov-
ering entire streambeds (Gretz 2008). While phosphorus limita-
tion is thought to play a role in Didymo bloom formation, the
precise causes of blooms remain a current topic of investigation
(Taylor and Bothwell 2014). In recent years, increasing reports of
severe Didymo blooms have led to major concern about its conse-
quences for freshwater organisms (Bickel and Closs 2008;James
and Chipps 2016;Jellyman and Harding 2016).
At high Didymo coverage, stream invertebrate assemblages origi-
nally dominated by Ephemeroptera, Plecoptera, and Trichoptera
(EPT taxa) typically shift towards dominance by Chironomidae, Oli-
gochaeta, Nematoda, or Cladocera taxa generally associated with
reduced habitat quality in trout streams (Kilroy et al. 2009;Gillis and
Chalifour 2010;Larned and Kilroy 2014). There has been widespread
concern about the consequences of Didymo blooms for trout (Gillis
and Chalifour 2010;James et al. 2010a;Jellyman and Harding 2016)
because EPT taxa are often assumed to be a primary food source for
salmonid species (Behnke 2010). However, to date, it is unclear
whether Didymo blooms have had any substantial negative or posi-
tive impacts on salmonids via changes to the macroinvertebrate
food base. In several New Zealand rivers, blooms were correlated
with lower trout abundances, dietary percent EPT, and stomach full-
ness (Jellyman and Harding 2016). In contrast, production of Atlantic
salmon (Salmo salar) in Icelandic and Norwegian rivers has remained
high despite the presence of severe Didymo blooms (Jonsson et al.
2008;Lindstrøm and Skulberg 2008), and escapement and juvenile
production of Pacific salmon and steelhead (Oncorhynchus spp.) in
Vancouver Island streams either increased or did not change in rela-
tion to blooms (Bothwell et al. 2008). In four South Dakota streams,
the condition and forage of large brown trout (Salmo trutta)wasnot
correlated with Didymo blooms, while body condition of juveniles
washigher(James and Chipps 2016). However, Didymo growth in
the study was also correlated with drought, making causal inference
difficult (James et al. 2010b). As such, no study has successfully exam-
ined the mechanistic links among Didymo blooms, macroinverte-
brates, and fishes necessary to make causal inference. Further, no
studies have addressed the potential effects of blooms on inland
native trout populations or on nongame species.
To better understand the trophic consequences of Didymo
blooms, we assessed the relationship among blooms, fish diet,
condition, and growth over two summers in the mountainous
Kootenai (Kootenay in Canada) basin of British Columbia, Idaho,
and Montana, much of which falls within the globally rare inland
temperate rainforest biome (Dellasala et al. 2011). We employed a
multifaceted research approach in which we examined potential
Didymo bloom impacts on fish via two methods: (i) a detailed,
mechanistic study comparing food webs in a stream with Didymo
blooms against one without and (ii) an observational study survey-
ing fish diet and condition across 28 streams representing a gradi-
ent of Didymo bloom severity. We also refine a novel method for
estimating food consumption by fish that may serve as a viable al-
ternative to traditional bioenergetics, in some cases. The results
of this study increase ecological understanding of the consequen-
ces of Didymo blooms and will help determine whether treat-
ment of Didymo blooms is a necessary strategy to benefitfishes.
Methods
Study location
Twice-monthly through the summer of 2018, we sampled two
streams located in the Cabinet Mountains of northwestern Mon-
tana, Bear Creek and nearby Ramsey Creek (Fig. 1). Both creeks
have similar physical characteristics (Table 1), but Bear Creek
contains obvious Didymo blooms while Ramsey Creek does not.
The two streams thus offer an opportunity to examine potential
effects of blooms on biotic communities in a paired, reference-
impact framework.
Over the course of the summers of 2018 and 2019, we examined
131 locations on 103 individual streams for the presence of Didymo
blooms throughout the Kootenai River basin (Clancy 2020). In 2019,
we surveyed fishes in 28 of those streams (Fig. 1), representing large
differences in bloom coverage: 0%–80% (Table A1).
Didymo reference-impact study —2018
We selected a 300 m long reach for study in both Bear and Ram-
sey Creeks with similar habitat conditions. The fish assemblages
of both were predominantly composed of Columbia River red-
band trout (Oncorhynchus mykiss gairdneri)andbulltrout(Salvelinus
confluentus). Ramsey Creek also contained Columbia slimy sculpin
(Uranidea cognata syn. Cottus cognatus) at relatively low abun-
dance. We measured five habitat variables to ensure Bear and
Ramsey creeks were suitable for comparison: mean substrate
size (sensu Wolman 1954), channel wetted width, mesohabitat
composition (percent cascade, riffle, and pool), water tempera-
ture (30 min recording interval, Onset HOBO data loggers), and
water chemistry (soluble reactive phosphorus, nitrate, and sul-
fate; Lachat 8500 QuikChem FIA and IC). Every 2 weeks (early
June –early September), we systematically estimated percentage
of substrate covered by blooming Didymo using a 19 L bucket
with a clear bottom, making five evenly spaced estimates along
lateral transects, each 20 paces apart from reach top to bottom.
We then combined twice-monthly estimates to form monthly
Didymo bloom coverage estimates.
Food-web structure was determined by macroinvertebrate and
fish sampling concurrent with Didymo coverage estimation. In
conjunction with Didymo bloom measurements, we collected
drifting macroinvertebrates by placing two separate 25.4 cm 
45.72 cm drift nets in the stream for 30 min at the middle-to-end
of the reach and pooling the combined samples in 70% ethanol.
Sampling occurred between the hours of 10:00 and 17:00. The day
following each Didymo and macroinvertebrate sampling event,
we collected fishes through single-pass backpack electroshock-
ing (LR-24 Backpack Shocker Smith-Root, Vancouver, Washing-
ton). We completed multiple passes during the final sampling
event (September) to maximize summer-long recapture. Each
fish was anesthetized with clove oil, weighed, measured, and
marked by clipping a small section of the caudal fin. We gastri-
cally lavaged individuals larger than 100 mm to collect diets and,
if captured during June or July, implanted a uniquely coded,
12 mm passive integrated transponder (PIT) tag (Model HDX12,
Biomark, Boise, Idaho). Gut evacuation was assumed to be mini-
mal due to cold temperatures, and processing was generally less
than an hour after capture. Using the average percent growth
between individuals measured in June and July, we back-calculated
June weights for individuals tagged in July. This represented 57%
of redband trout in Bear Creek and 68% in Ramsey Creek. We also
observed no evidence of a tag effect on growth of PIT-tagged fish
(Clancy 2020).
Because a shift to a macroinvertebrate assemblage of smaller
and more abundant individuals may favor juvenile fishes (James
and Chipps 2016), we identified large and small size classes of red-
band (cutoff at 105 mm) and bull trout (cutoff at 130 mm) using
monthly length–frequency histograms (Clancy 2020). We then
calculated size-specific abundances using Lincoln–Petersen
mark–recapture estimation in which the final sampling date was
the recapture event and all previous sampling events a single
marking event (Lincoln 1930). We determined this approach to
be reasonable because movement of PIT-tagged fishes between
the abutting upper and lower halves of Bear Creek was negligible,
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
2 Can. J. Fish. Aquat. Sci. Vol. 00, 0000
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 63.153.124.239 on 01/19/21
For personal use only.

thus meeting the population closure assumption of Lincoln–Petersen
estimation (Clancy 2020).
We identified and measured drift and diet macroinvertebrates
to the taxonomic level of family and used published length-
to-mass conversions to estimate biomass and caloric content
(Clancy 2020). We compared taxon-specificproportionsof
drifting macroinvertebrates in Bear and Ramsey creeks by cal-
culating the monthly percent similarity (Schoener 1970):
ð1ÞPercent Similarity ¼100 0:5X
n
i¼1




BiRi



0
@1
A
where B
i
is the percentage of aquatic invertebrates of taxa iin
Bear Creek, and R
i
is the percent of invertebrates of taxa iin
Ramsey Creek. Using the same equation, we compared trout
diets with the availability of invertebrates in the drift (both
aquatic and terrestrial) to determine whether feeding behavior
was different between the two streams. Then, we also compared
trout diets between the two streams using percent energetic
content for each diet taxa. To evaluate how likely observed dif-
ferences between groups were (drift versus drift, diet versus
drift, and diet versus diet), we used Pearson’s
x
2
tests. We fur-
ther report monthly and summer-long gut fullness and relative
number and energetic content of invertebrates in the drift
between the two streams.
By pairing individual caloric demand with trout diet composi-
tion, we created energy-flow food webs. To calculate individual
caloric demand, we used a novel and simple bioenergetics equa-
tion that combines aspects of the Benke–Wallace trophic basis
of production method for calculating energetic demand (Benke
and Wallace 1980) and traditional fish bioenergetics (Hanson
et al. 1997) while accounting for fish thermal growth optima. We
developed this novel thermal optima bioenergetics equation
(TOBE) because traditional fish bioenergetics models require
extensive species-specific parameterization (Chipps and Wahl
2008) and are thus not available for many species. By creating the
TOBE, where the only species-specific parameter is thermal
Fig. 1. Location of study streams (red dots) within the Kootenai River basin (left) and the upper Libby Creek subbasin (right). Inset A
shows the location of the Kootenai basin within the larger Columbia River watershed. Map created using QGIS (2019) and shapefiles from
the National Hydrography Dataset (USGS 2019). [Colour online.]
Table 1. Habitat measurements and nutrient
concentrations in Bear Creek (Didymo) and
Ramsey Creek (no Didymo) during summer 2018
(June–September).
Bear Ck.
(Didymo)
Ramsey Ck.
(No Didymo)
Temp. (°C) 6SD 9.7962.32 9.7962.40
Mesohabitat
Cascade 76% 83%
Riffle 16% 10%
Pool 8% 7%
Substrate size (mm) 26.7 23.2
Wetted width (m) 7.24 7.17
Nutrients (lg·L
–1
)6SD
SRP 1.99560.368 1.53060.409
Nitrate 74.5 25
Sulfate 1235 930
N:P ratio 82.5 36.0
Note: The N:P ratio is the molar ratio of nitrate to
soluble reactive phosphorus (SRP).
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
Clancy et al. 3
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 63.153.124.239 on 01/19/21
For personal use only.

growth optima, and comparing consumption estimates of red-
band trout and bull trout with those generated using traditional
bioenergetics models (Deslauriers et al. 2017), we were able to
determine whether the TOBE is adequate for use with species for
which traditional bioenergetics models are not available. Inputs
to the traditional fish bioenergetics models were summer stream
temperatures (taken at 30 min intervals), start and end weights
of fish, and the output was summer-long energetic consumption.
We used species-specific bioenergetics models for redband (rain-
bow) trout (Railsback and Rose 1999)andbulltrout(Mesa et al.
2013) and substituted a model for prickly sculpin (Cottus asper)for
Columbia slimy sculpin (Moss 2001).
The Benke–Wallace method, upon which the TOBE is based,
was originally developed for use with benthic macroinverte-
brates and does not account for differential allocation of energy
by organism size and water temperature, factors known to
strongly influence fish growth (Brown et al. 2004). Thus, we used
two different numbers for the proportion of total assimilated
energy allocated to growth (net production efficiency) in large
versus small fish as suggested by Bellmore et al. (2013).Wethen
applied a temperature-specific factor to adjust for the effects of
suboptimum temperatures. Thus, consumption in kilocalories
was calculated as follows:
ð2ÞConsumption ¼X
n
i¼1
DietProportioniGrowth EnergyDensity
ðÞ
TempFactor TissueAllocation Digestiblei0:2Digestiblei
ðÞ½
where DietProportion
i
is the average proportion by kilocalories
of food type iin the diet, Growth is the average accrued body
mass in grams of the fish over the summer (June–September),
EnergyDensity is the energy density (kcal·g
–1
)ofthefish, and Tis-
sueAllocation is the theoretical maximum proportion of assimi-
lated energy allocated to fish tissue growth (net production
efficiency), which was set as 0.22 for large size class trout and 0.5
for small size class trout and sculpin. Digestible
i
is the estimated
digestible proportion of food type i. We used Digestible
i
for each
food type from Hanson et al. (1997) and subtracted a value of
0.2Digestible
i
to account for specific dynamic action (Hanson
et al. 1997 ). Thus, Digestible
i
–0.2Digestible
i
is the assimilation ef-
ficiency of food type i. TempFactor is the temperature correction
factor calculated according to the following equation:
ð3ÞTempFactor ¼e0:2StreamTempOptimTemp
ðÞ½
4
fg
where StreamTemp is the average stream temperature for the
measurement interval over the growth period, and OptimTemp
is the thermal optimum for each species. This equation is an
approximation of a fish’s thermal optimum curve (based on equa-
tions in Bear et al. 2007) that asymptotes at an energy-allocation-to-
tissue value of zero (Clancy 2020). We derived thermal optimum
values from previous field and laboratory studies: 13.1°C for red-
band trout (Bear et al. 2007), 12.0 °C for bull trout (Dunham et al.
2003), and 12.1 °C for Columbia slimy sculpin (Wehrly et al. 2003).
To derive total estimated consumption by each species, we
multiplied estimated summer TOBE consumption values by cal-
culated fish abundances in each stream. Then, we multiplied the
proportion of energy of each prey item in the average diet of each
fish species by the reach-level consumption estimates. Thus, we
obtained estimates of total energy flow from all prey to fish pred-
ators and compared results for Bear and Ramsey creeks.
Multistream Didymo survey —2019
In a 30.5 m reach of each selected stream, we estimated Didymo
coverage using the same method as in 2018. We also recorded six
other habitat variables: wetted width (n= 5), canopy density (n=5
using a densitometer; Strickler 1959), dominant vegetation type,
substrate type (Cummins 1962), Rosgen channel type (Rosgen
1994), number of LWD items (sensu Kershner et al. 2004), and
mean August stream temperature. From reach top to bottom, we
measured wetted width and canopy density, while we qualita-
tively assessed vegetation, substrate, and channel type. Tempera-
ture loggers were placed in three streams: Bear, Outlet, and Trail
creeks. We estimated mean August temperatures of all other
streams by adding the time-specific difference of each stream’s
temperature (taken with a handheld thermometer) to the mean
August value of one of the three temperature loggers (Bear Creek
for streams flowing into the Kootenai River below the Fisher
River confluence, Outlet Creek for those above the Fisher conflu-
ence, and Trail Creek for Fisher River tributaries).
In the same reach, we collected fishes through two-pass (one
upstream, one downstream) backpack electroshocking. We anes-
thetized, weighed, and measured all fishes and then released leu-
ciscids and catostomids. Using an in-field assessment in which
we gastrically lavaged fish, we assessed the diets of salmonids
and cottids by spreading the diet contents in a 30 cm 15 cm
white pan and recording the number of individuals of each inver-
tebrate taxa. We identified insects to Order except for Simuliidae
and Chironomidae, which we identified to Family. Other inverte-
brates we identified to Class or Phylum, and vertebrates to the
lowest practical taxonomic level (usually species).
We generated two response metrics of fish condition (Fulton’s
K(Heincke 1908;Ricker 1975) and residual analysis of observed
versus predicted weights (Fechhelm et al. 1995)) and four metrics
of diet composition (%Diptera, %EPT, %Aquatics, and gut fullness
(number of diet items/fish length]) for each fish. Using weighted,
univariate logistic (%Diptera, %EPT, %Aquatics) and linear regres-
sions (gut fullness and fish condition) in which fish sample size
was the relative weight of each stream in the regression, we ana-
lyzed each response metric compared with Didymo coverage and
the other six habitat variables. We removed four streams (Koka-
nee, Coffee, Mobbs, and Solo Joe creeks) from regressions due to
low sample size or substantially different substrate type. We
grouped fish by genus due to otherwise small sample size if com-
pared only within species (char (Salvelinus)andsculpin(Uranidea))
or substantial hybridization in the basin that made some field
IDs difficult (trout, Oncorhynchus spp.). For each comparison of a
habitat variable with a diet metric, we calculated an R
2
value (or
Nagelkerke’spseudo-R
2
for logistic regression; Nagelkerke 1991)
and pvalue and considered variables with an R
2
greater than 0.2
and a pvalue less than 0.2 to be a nonspurious correlation.
Results
Didymo versus reference stream study —2018
Differences in all four physical habitat variables were small
between Bear Creek (Didymo) and Ramsey Creek (no Didymo),
giving us confidence the two were suitable for comparison (Table
1). Nutrient concentrations differed between the two (Table 1),
likely playing a role in Didymo bloom presence (Capito 2020).
Didymo bloom severity in Bear Creek increased from 10.9% cover-
age in June to 22.6% coverage in August before falling to 18.9% in
September (Fig. 2). The June to August change in Didymo cover-
age was significant (p<0.01), but the decline from August to Sep-
tember was not (p=0.21).
Composition of drifting aquatic insects (EPT larvae and Diptera
larvae and pupae) between the two streams became less similar
as Didymo coverage increased throughout the summer and more
similar when Didymo coverage decreased in September (June–
September: 84.4%, 78.5%, 67.2%, and 92.0% similar; Fig. 2). From
June through August, percent EPT in the drift was higher in Ram-
sey Creek (no Didymo) and percent larval and pupal Diptera was
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
4 Can. J. Fish. Aquat. Sci. Vol. 00, 0000
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 63.153.124.239 on 01/19/21
For personal use only.

higher in Bear Creek (Didymo; Fig. 2). Percentages were similar
in September. Both total drifting invertebrates and total energy
of drifting invertebrates similarly diverged later in the summer
with the streams having similar numbers in June, Ramsey
Creek having higher numbers in July and August, and Bear
Creek having higher numbers in September. The summer-long
amount of total energy of drifting invertebrates was 2.2 times
higher in Ramsey Creek due to greater total biomass in the
drift.
Reach abundance estimates for redband and bull trout were
higher in Bear Creek (Didymo; Table 2). Relative growth (size-
dependent growth per unit of time) of redband trout varied by
size class. We estimated summer relative growth of small trout
(<105 mm) to be 0.0292 g of growth per gram of starting weight
per day (g·g
–1
·day
–1
) in Bear Creek and only 0.0033 g·g
–1
·day
–1
in
Ramsey Creek, but this difference was likely driven by a very
small June sample size (three in Bear Creek and one in Ramsey
Creek). Relative growth of large size class redband trout (>105 mm)
Fig. 2. Monthly percentage of stream substrate covered by Didymo in Bear Creek, 2018 (top). Pie charts show proportions of aquatic life-
stage insect taxa in the drift in Bear and Ramsey creeks and indicate higher proportion of Diptera larvae and pupae in Bear Creek
(Didymo). [Colour online.]
Table 2. Population (reach) abundance, growth, and consumption estimates for each fish
species and size class in Bear and Ramsey creeks during summer 2018.
Stream
Population
abundance
estimate
Individual
growth
(g·g
–1
·day
–1
)
TOBE
estimated
consumption
(kcal)
FB4
estimated
consumption
(kcal)
Population
consumption
(kcal)
Redband trout
Small Bear 132 0.0292 33.3 20.9 4 385.6
Ramsey 91 0.0033 31.9 18.8 2 902.9
Large Bear 196 0.0027 74.9 75.3 14 643.0
Ramsey 81 0.0029 53.6 54.2 4 347.0
Bull trout
Small Bear 60 0.0136 38.9 27.1 2 334.0
Ramsey 2 NA NA NA NA
Large Bear 45 0.0011 46.0 24.8 2 083.8
Ramsey 3 NA NA NA NA
Slimy sculpin
Ramsey 20 0.0030 5.0 8.8 100.0
Note: Consumption estimates generated using traditional bioenergetics with Fish Bioenergetics 4.0 (FB4;
Deslauriers et al. 2017) are shown for comparison with the novel thermal optimum bioenergetics equation
(TOBE) estimates. Population-level consumption estimates are the product of the population abundance
estimate and the TOBE consumption estimate.
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
Clancy et al. 5
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 63.153.124.239 on 01/19/21
For personal use only.

was similar between the two streams: 0.002760.0004 g·g
–1
·day
–1
in
Bear Creek and 0.0029 60.0007 g·g
–1
·day
–1
in Ramsey Creek (mean 6
standard error; Table 2).
Redband trout diets (including both terrestrial and aquatic
insects of all life stages) were 40.7% similar to the drift in Bear
Creek (
x
2
test: p<0.01) and 40.1% similar to the drift in Ramsey
Creek (
x
2
test: p<0.01). By energetic content, redband trout diets
were 81.2% similar between Bear and Ramsey creeks for the
whole summer (
x
2
test: p= 0.84): 55.6% similar in June, 77.5% simi-
lar in July, 99.7% similar in August, and 75.0% similar in Septem-
ber. Monthly gut fullness was not significantly different between
the two streams. Diets of small individual redband trout in Bear
Creek (Didymo) had more EPT (78.6% 68.4%) than large individu-
als (46.4% 63.0%), while gut fullness and %Diptera were similar.
Individual consumption estimates for large redband trout
were 39% higher in Bear Creek (Didymo), while small size class
estimates were similar between the two streams (Table 2). We
estimated reach-level energetic demand by all redband trout at
19 029 kcal in Bear Creek and 7250 kcal in Ramsey Creek (Table
2). Consumption estimates using the TOBE were similar to those
estimated using traditional, species-specificbioenergeticsmod-
els (Table 2).
The primary sources of energy (>5% of demand) for redband trout
in Bear Creek (Didymo) were Ephemeroptera (38.0% of energy
intake), Hymenoptera (15.1%), Trichoptera (14.4%), Plecoptera (9.5%),
and Diptera (7.6%) (Fig. 3; also refer to online Supplementary Table
S1
1
). Primary energy sources for Ramsey Creek (no Didymo) redband
trout were Ephemeroptera (45.8%), Hymenoptera (15.7%), Diptera
(9.8%), Trichoptera (9.0%), and Plecoptera (6.3%) (Fig. 3). Primary sour-
ces of energy for bull trout in Bear Creek were Ephemeroptera
(48.0%), Trichoptera (13.1%), Nematoda (7.2%), Plecoptera (6.3%), and
Hymenoptera (5.1%) (Fig. 3). We collected only four bull trout diets
during the entire summer in Ramsey Creek, and we did not consider
this sufficient to calculate average diet compositions. Columbia
slimy sculpin were only captured in Ramsey Creek and thus had no
Didymo comparison.
Multistream Didymo survey —2019
Between-site variation in Fulton’sKwas too low to assess possi-
ble explanatory variables (coefficients of variation (CV) ≤0.1;
Fig. 3. Energy-flow food web for fishes in Bear Creek (Didymo) and Ramsey Creek (no Didymo) indicating that similar energy sources
sustained fish growth in both streams. Line thickness represents proportion of total energy demand by the given fish species met by each
invertebrate taxon. Only taxa representing at least 5% of energy demand are shown. Left to right: Hymenoptera, Diptera, Ephemeroptera,
Plecoptera, Trichoptera, and Nematoda. [Colour online.]
1
Supplementary data are available with the article at https://doi.org/10.1139/cjfas-2020-0121.
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
6 Can. J. Fish. Aquat. Sci. Vol. 00, 0000
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 63.153.124.239 on 01/19/21
For personal use only.

Supplementary Table S2
1
). Between-site variation in fish relative
condition, calculated as a fish’s observed weight compared with
its predicted weight, was similarly low for trout and sculpin (CV
of 0.12 and 0.04, respectively) and moderately low for char (CV =
0.28). Despite slightly more variation in char relative condition
between sites, there was no relationship between condition and
Didymo coverage (R
2
=0.03,p= 0.46).
For all diet metrics across all three fish taxa, percent Didymo
cover was only correlated with percentage of aquatic life-stage
invertebrates in Oncorhynchus diets (Fig. 4). Canopy cover, LWD, ri-
parian vegetation type, and stream temperature were also corre-
lated with percent aquatic invertebrates in Oncorhynchus diets,
with LWD having the highest pseudo-R
2
(Supplementary Table
S3
1
). In fact, few fish diet metrics were correlated with habitat
variables (Supplementary Table S3
1
).
Discussion
During the summers of 2018 and 2019, we examined the
response of trout, char, and sculpin species to Didymo blooms at
reach and regional scales. Similar to other studies, Didymo in
Bear Creek was positively correlated with an increase in the
proportion of larval Diptera and a decrease in the proportion of
EPT taxa (Marshall 2007;James et al. 2010a;Anderson et al. 2014).
Despite changes to their macroinvertebrate food base, redband
trout diets and growth rates with Didymo were similar to those
in the no-Didymo reference stream. In fact, diets were most simi-
lar in August when Didymo coverage was at its peak. While a no-
Didymo comparison was not available for bull trout because so
few were captured in Ramsey Creek, bull trout diets from Bear
Creek also contained relatively few Diptera. Corroborating these
results, we similarly did not observe major differences in the diets
or condition of trout, char, or sculpin species across a gradient of
Didymo bloom coverages in 2019. This held true even in Outlet
Creek, British Columbia, where Didymo blooms covered over 80% of
the streambed, but rainbow trout diets remained similar to those in
streams with little to no Didymo.
Stream-resident trout are considered generalist invertivores
(Behnke 1992), but strong selection by redband trout in Bear
Creek (Didymo) and Ramsey Creek (no Didymo) for the same taxa
indicates this subspecies may show a strong preference for may-
flies (Ephemeroptera; Fig. 3). Despite this strong selection, the
similarity in trout growth rates between the two streams in 2018
Fig. 4. Correlations of Didymo coverage to each fish taxon’s diet and condition response metrics from 2019. Each dot represents the
average value for fish in a single stream.
r
2
is Nagelkerke’spseudo-R
2
value. Asterisks (***) indicate p≤0.05. Results indicate few
relationships between Didymo coverage and fish diet metrics.
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
Clancy et al. 7
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 63.153.124.239 on 01/19/21
For personal use only.

and condition across streams in 2019 may indicate summer trout
growth in Kootenai River headwaters is more limited by factors in-
dependent of forage such as nontrophic interspecific competition.
Consumption estimates using the novel TOBE were similar to
those generated using traditional bioenergetics models, and the
small discrepancies between bull trout estimates from the two
approaches were probably due to the fact we used a thermal opti-
mum va lue of 12.0 °C ( Dunham et al. 2003) that was likely more
appropriate for resident Kootenai basin bull trout than the 16.0 °C
optimum (Mesa et al. 2013) used by Fish Bioenergetics 4.0
(Deslauriers et al. 2017). While further refinement of TOBE (espe-
cially of the size-specific tissue allocation) could certainly make
estimates more accurate, the relative similarities between the
two estimates demonstrate the potential utility of TOBE as a
means to generate consumption estimates when bioenergetics
models are not available or otherwise appropriate (Donner 2011).
Although not the impetus of our study, we observed interest-
ing differences in percentage of aquatic invertebrates in trout
diets in streams with differing riparian vegetation. We observed
higher proportions of riparian invertebrates in trout diets in
alder-dominated (Alnus spp.) streams than in pine-dominated
streams (largely lodgepole pine, Pinus contorta), a pattern similar
to that observed in juvenile salmon (Oncorhynchus kisutch)in
Alaska coastal temperate rainforests (Picea spp.; Allan et al.
2003). However, unlike the Alaska juvenile salmon, trout in our
inland temperate rainforest streams dominated by cedar (Thuja
plicata)andhemlock(Tsuga spp.) had similar aquatic–terrestrial
ratios to alder-dominated streams.
Our 2018 study was limited to two streams, and while Didymo
was associated with differences in the macroinvertebrate com-
munity, it is possible differences would exist independent of
Didymo. Although not collected for this study, analysis of
benthic invertebrate samples would likely be more sensitive to
variation in Didymo coverage because they are less affected by
potential drift from upstream sources. Further, we examined the
impacts of Didymo blooms only from June to September during
both years, a time when terrestrial invertebrate inputs, and trout
reliance upon them, are high (Nakano and Murakami 2001). It is
possible some negative or positive consequence of Didymo can
only be observed by studying fishes across seasons (e.g., winter).
In fact, some studies have reported severe Didymo blooms during
winter months (e.g., Kolmakov et al. 2008), and we observed
severe blooms in the Lardeau River during April of 2018, prior to
the freshet. Trout growth in headwater streams is higher in summer
months, but foraging (Thurow 1997)andgrowth(Al-Chokhachy
et al. 2019) still occur over winter. Further, our study did not
examine the potential effects of Didymo on fish spawning sites or
early life-stage survival (Bickel and Closs 2008). We therefore sug-
gest future work examine the relationship of Didymo blooms to
fish diet and growth in winter, as well as egg and larval fish sur-
vival. Nonetheless, we demonstrate at both reach and regional
scales that Didymo blooms do not seem to affect fish diets, condi-
tion, or growth in mountainous headwaters during summer.
Implications for management
Because Didymo blooms are visually obvious and macroinver-
tebrates assemblages are altered, it has been routinely hypothe-
sized Didymo negatively impacts fisheries (Bickel and Closs 2008;
Beville et al. 2012;Klauda and Hanna 2016). Authors of previous
studies have suggested nutrient amendments (James et al. 2015;
Coyle 2016) and dam releases (Cullis et al. 2015)asviablemeansto
manage nuisance Didymo blooms. Indeed, both methods show
promise for reduction of blooms at local scales. The impetus for
this bloom reduction may be independent of concern for fishes,
including aesthetics, fouling of infrastructure, or to prevent hy-
poxia. However, we did not observe any large impacts of Didymo
blooms on the diet, condition, or growth of trout in Kootenai ba-
sin headwaters. This overall result is similar to those for brown
trout in a South Dakota stream (James and Chipps 2016). There-
fore, clearly identifying objectives of Didymo control may be
useful. At present, it is not clear that objectives to increase
growth or condition of fish would be met by reducing Didymo
blooms in headwaters. Where Didymo control is implemented,
additional monitoring of invertebrates, fish diets, and growth
will lead to greater understanding of specific impacts. While less
often an objective for major stream restoration efforts, consider-
ing potential impacts of Didymo blooms to imperiled inverte-
brates (especially sedentary freshwater mussels that may not be
able to avoid blooms) may identify conservation opportunities of
Didymo control.
Acknowledgements
This work was supported by the British Columbia Ministry of
the Environment; the US Forest Service; Montana Fish, Wildlife &
Parks (in-kind); the US Geological Survey –Utah Cooperative Fish
and Wildlife Research Unit (in-kind); and the Utah State University
School of Graduate Studies, Department of Watershed Sciences,
and Ecology Center. We thank Chuck Hawkins for helpful reviews
of previous versions of this manuscript. Jon McFarland, Ryan West,
Chris Clancy, and Marshall Wolf provided valuable assistance in
the field. Thank you to Jay DeShazer, Ryan Sylvester, Jared
Lampton, Jordan Frye, Brian Stephens, Monty Benner, and Mike
Hensler for help at the Libby Field Office and to Jeff Burrows, Greg
Andrusak, Joe Thorley, Murray Pearson, and Jeff Curtis in British
Columbia. Gary Thiede at Utah State University (USU) and Brett
Roper and Mike Young of the US Forest Service provided additional
support. The National Aquatic Monitoring Center in Logan, Utah,
provided generous help with identification of fish diet samples.
Thanks also to the associate editor and two anonymous reviewers,
whose comments greatly improved the manuscript. Any use of
trade, firm, or product names is for descriptive purposes only and
does not imply endorsement by the United States Government.
This study was performed under the auspices of the USU IACUC
protocol No. 10006, The University of British Columbia animal
care protocol No. A19-0171, Montana scientific collector’s permits
07-2018 and 16-2019, and British Columbia fish collection permit
CB19-512335.
References
Albertson, L.K., and Daniels, M.D. 2016. Effects of invasive crayfish on fine
sediment accumulation, gravel movement, and macroinvertebrate com-
munities. Freshw. Sci. 35(2): 644–653. doi:10.1086/685860.
Al-Chokhachy, R., Kovach, R.P., Sepulveda, A., Strait, J., Shepard, B.B., and
Muhlfeld, C.C. 2019. Compensatory growth offsets poor condition in
native trout populations. Freshw. Biol. 64(11): 1–11. doi:10.1111/fwb.13400.
Allan, J.D., Wipfli,M.S.,Caouette,J.P.,Prussian,A.,andRodgers,J.2003.
Influence of streamside vegetation on inputs of terrestrial invertebrates
to salmonid food webs. Can. J. Fish. Aquat. Sci. 60(3): 309–320. doi:10.1139/f03-
019.
Anderson, I.J., Saiki, M.K., Sellheim, K., and Merz, J.E. 2014. Differences
in benthic macroinvertebrate assemblages associated with a bloom of
Didymosphenia geminata in the Lower American River, California. South-
west. Nat. 59(3): 389–395. doi:10.1894/FRG-05.1.
Baxter, C.V., Fausch, K.D., and Saunders, W.C. 2005. Tangled webs: recipro-
cal flows of invertebrate prey link streams and riparian zones. Freshw.
Biol. 50(2): 201–220. doi:10.1111/j.1365-2427.2004.01328.x.
Bear,E.A.,McMahon,T.E.,andZale,A.V.2007.Comparativethermal
requirements of westslope cutthroat trout and rainbow trout: implica-
tions for species interactions and development of thermal protection
standards. Trans. Am. Fish. Soc. 136(4): 1113–1121. doi:10.1577/T06-072.1.
Beckett, D.C., Green, B.W., Thomas, S.A., and Miller, A.C. 1996. Epizoic inver-
tebrate communities on upper Mississippi River unionid bivalves. Am.
Midl. Nat. 135(1): 102–114. doi:10.2307/2426876.
Behnke, R.J. 1992. Native trout of Western North America. AFS Monograph
6. American Fisheries Society, Bethesda, Md.
Behnke, R.J. 2010. Trout and salmon of North America. 2nd ed. Simon and
Schuster, New York.
Bellmore,J.R.,Baxter,C.V.,Ray,A.M.,Denny,L.,Tardy,K.,andGalloway,E.
2012. Assessing the potential for salmon recovery via floodplain restora-
tion: A multitrophic level comparison of dredge-mined to reference
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
8 Can. J. Fish. Aquat. Sci. Vol. 00, 0000
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 63.153.124.239 on 01/19/21
For personal use only.

segments. Environ. Manage. 49(3): 734–750. doi:10.1007/s00267-012-9813-
x.PMID:22323109.
Bellmore, J.R., Baxter, C.V., Martens, K., and Connolly, P.J. 2013. The flood-
plain food web mosaic: a study of its importance to salmon and steel-
head with implications for their recovery. Ecol. Appl. 23(1): 189–207.
doi:10.1890/12-0806.1.PMID:23495646.
Benke, A.C., and Wallace, J.B. 1980. Trophic basis of production among net-
spinning caddisflies in a southern Appalachian stream. Ecology, 61(1):
108–118. doi:10.2307/1937161.
Beville, S.T., Kerr, G.N., and Hughey, K.F.D. 2012. Valuing impacts of the
invasive alga Didymosphen ia geminat a on recreational angling. Ecol. Econ.
82:1–10. doi:10.1016/j.ecolecon.2012.08.004.
Bickel, T.O., and Closs, G.P. 2008. Impact of Didymosphenia geminata on hypo-
rheic conditions in trout redds: reason for concern? Mar. Freshw. Res. 59(11):
1028–1033. doi:10.1071/MF08011.
Bothwell, M., Holtby, B., Lynch, D.R., Wright, H., and Pellet, K. 2008. Did
blooms of Didymosphenia geminata affect runs of anadromous salmonids
on Vancouver Island? In Proceedings of the 2007 International Workshop
on Didymosphenia geminata.Edited by M.L. Bothwell and S.A. Spaulding.
Canadian Technical Report of Fisheries and Aquatic Sciences 2795, Nanaimo,
B.C. pp. 50–53.
Brown, J.H., Gillooly, J.F., Allen, A.P., Savage, V.M., and West, G.B. 2004. To-
ward a metabolic theory of ecology. Ecology, 85(7): 1771–1789. doi:10.1890/
03-9000.
Capito, L. 2020. Environmental controls on Didymosphenia geminata bloom
formation. M.S. thesis, Utah State University.
Chipps, S.R., and Wahl, D.H. 2008. Bioenergetics modeling in the 21st cen-
tury: reviewing new insights and revisiting old constraints. Trans. Am.
Fish. Soc. 137(1): 298–313. doi:10.1577/T05-236.1.
Clancy, N.G. 2020. Do Didymosphenia geminata blooms affect fishes in the
Kootenai River basin? M.S. thesis, Utah State University. doi:10.13140/
RG.2.2.20745.80488.
Coyle, M.K. 2016. Didymosphenia geminata in the Kootenai River in Libby,
Montana: nuisance mat characteristics and management strategies for
suppression. M.S. thesis, University of Idaho.
Cross,W.F.,Baxter,C.V.,Donner,K.C.,Rosi-Marshall,E.J.,Kennedy,T.A.,
Hall, R.O., et al. 2011. Ecosystem ecology meets adaptive management:
food web response to a controlled flood on the Colorado River, Glen Can-
yon. Ecol. Appl. 21(6): 2016–2033. doi:10.1890/10-1719.1.
Cullis, J.D.S., Mcknight, D.M., and Spaulding, S.A. 2015. Hydrodynamic con-
trol of benthic mats of Didymosphenia geminata at the reach scale. Can. J.
Fish.Aquat.Sci.72:902–914. doi:10.1139/cjfas-2014-0314.
Cummins, K.W. 1962. An evaluation of some techniques for the collection
and analysis of benthic samples with special emphasis on lotic waters.
Am. Midl. Nat. 67(2): 477–504. doi:10.2307/2422722.
Darwin, C. 1881. The formation of vegetable mould through the action of
worms. John Murray, London.
Dellasala, D.A., Alaback, P., Craighead, L., Goward, T., Paquet, P., and Spri-
bille, T. 2011. Temperate and boreal rainforests of inland northwestern
North America. In Temperate and boreal rainforests of the world: Ecol-
ogy and Conservation. pp. 82–110.
Deslauriers, D., Chipps, S.R., Breck, J.E., Rice, J.A., and Madenjian, C.P. 2017.
Fish bioenergetics 4.0: an R-based modeling application. Fisheries, 42(11):
586–596. doi:10.1080/03632415.2017.1377558.
Donner, K.C. 2011. Secondary production rates, consumption rates, and
trophic basis of production of fishes in the Colorado River, Grand Can-
yon, AZ: an assessment of potential competition for food. M.S. thesis,
Idaho State University.
Dunham, J., Rieman, B., and Chandler, G. 2003. Influences of temperature
and environmental variables on the distribution of bull trout within
streams at the southern margin of its range. N. Am. J. Fish. Manage. 23(3):
894–904. doi:10.1577/M02-028.
Fechhelm, R.G., Griffiths, W.B., Wilson, W.J., Gallaway, B.J., and Bryan, J.D.
1995. Intra- and interseasonal changes in the relative condition and prox-
imate body composition of Broad Whitefish from the Prudhoe Bay region
of Alaska. Trans. Am. Fish. Soc. 124:508–519. doi:10.1577/1548-8659(1995)
124<0508:IAICIT>2.3.CO;2.
Gillis, C.A., and Chalifour, M. 2010. Changes in the macrobenthic community
structure following the introduction of the invasive algae Didymosphenia
geminata in the Matapedia River (Québec, Canada). Hydrobiologia, 647(1):
63–70. doi:10.1007/s10750-009-9832-7.
Gretz, M.R. 2008. The stalks of didymo. In Proceedings of the 2007 Interna-
tional Workshop on Didymosphenia geminata.Edited by M.L. Bothwell and
S.A. Spaulding. Canadian Technical Report of Fisheries and Aquatic Sci-
ences 2795. pp. 21.
Hand, B.K., Flint, C.G., Frissell, C.A., Muhlfeld, C.C., Devlin, S.P., Kennedy, B.
P., et a l. 2 01 8. A so ci al–ecological perspective for riverscape management
in the Columbia River basin. Front. Ecol. Environ. 16:23–33. doi:10.1002/
fee.1752.
Hanson, P.C., Johnson, T., Schindler, D.E., and Kitchell, J.F. 1997. Fish bioen-
ergetics 3.0. In Technical Report WISCU-T-97-001. Madison, Wisc.
Heincke, F. 1908. Bericht €
uber die untersuchungen der biologischen anstalt
auf helgoland zur naturgeschichte der nutzfische. Die Beteiligung Deutsch-
lands an der Int. Meeresforsch. (4/5). pp. 67–155.
Hopper, G.W., Dubose, T.P., Gido, K.B., and Vaughn, C.C. 2019. Freshwater
mussels alter fish distributions through habitat modifications at fine spa-
tial scales. Freshw. Sci. 38(4): 702–712. doi:10.1086/705666.
Horton, P.A. 1961. The bionomics of brown trout in a Dartmoor stream.
J. Anim. Ecol. 30(2): 311–338. doi:10.2307/2301.
Huryn, A.D. 1996. An appraisal of the Allen paradox in a New Zealand trout
stream. Limnol. Oceanogr. 41(2): 243–252. doi:10.4319/lo.1996.41.2.0243.
James, D.A., and Chipps, S.R. 2016. Influence of Didymosphenia geminata
blooms on prey composition and associated diet and growth of Brown
Trout. Trans. Am. Fish. Soc. 145:195–205. doi:10.1080/00028487.2015.1111255.
James, D.A., Ranney, S.H., Chipps, S.R., and Spindler, B.D. 2010a. Inverte-
brate composition and abundance associated with Didymosphenia geminata
in a montane stream. J. Freshw. Ecol. 25(2): 235–241. doi:10.1080/02705060.
2010.9665073.
James, D.A., Wilhite, J.W., and Chipps, S.R. 2010b.Influence of drought con-
ditions on Brown Trout biomass and size structure in the Black Hills,
SouthDakota.N.Am.J.Fish.Manage.30:791–798. doi:10.1577/M09-199.1.
James,D.A.,Bothwell,M.L.,Chipps,S.R.,andCarreiro,J.2015.Useofphos-
phorus to reduce blooms of the benthic diatom Didymosphenia geminata in
an oligotrophic stream. Freshw. Sci. 34(4): 1272–1281. doi:10.1086/683038.
Jellyman, P.G., and Harding, J.S. 2016. Disentangling the stream community
impacts of Didymosphenia geminata: how are higher trophic levels
affected? Biol. Invasions, 18(12): 3419–3435. doi:10.1007/s10530-016-1233-z.
Jones, C.G., Lawton, J.H., and Shachak, M. 1994. Organisms as ecosystem
engineers. In Ecosystem management. Edited by F.B. Samson and F.L.
Knopf. Springer, New York. pp. 130–147.
Jonsson, I.R., Jonsson, G.S., Olafsson, J.S., Einarsson, S.M., and Antonsson, T.
2008. Occurrence and colonization pattern of Didymosphenia geminata in
Icelandic streams. In Proceedings of the 2007 International Workshop on
Didymosphenia geminata.Edited by M.L. Bothwell and S.A. Spaulding. Cana-
dian Technical Report of Fisheries and Aquatic Sciences 2795. pp. 41–44.
Kershner, J.L., Coles-Ritchie, M., Henderson, R.C., Archer, E.K., Cowley, E.R.,
Kratz, K., et al. 2004. Guide to effective monitoring of aquatic and ripar-
ian resources. USDA Forest Service General Technical Report RMRS-GTR-
121. doi:10.2737/RMRS-GTR-121.
Kilroy, C., Larned, S.T., and Biggs, B.J.F. 2009. The non-indigenous diatom
Didymosphenia geminata alters benthic communities in New Zealand riv-
ers. Freshw. Biol. 54(9): 1990–2002. doi:10.1111/j.1365-2427.2009.02247.x.
Klauda, R.J., and Hanna, K.V. 2016. Didymosphenia geminata infestation in
Maryland: reactions and responses by the Maryland Department of Natu-
ral Resources, 2008–2014. Maryland Department of Natural Resources
report.
Kolmakov, V.I., Anishchenko, O.V., Ivanova, E.A., Gladyshev, M.I., and
Sushchik, N.N. 2008. Estimation of periphytic microalgae gross primary
production with DCMU-fluorescence method in Yenisei River (Siberia,
Russia). J. Appl. Phycol. 20:289–297. doi:10.1007/s10811-007-9246-8.
Larned, S.T., and Kilroy, C. 2014. Effects of Didymosphenia geminata removal
on river macroinvertebrate communities. J. Freshw. Ecol. 29(3): 345–362.
doi:10.1080/02705060.2014.898595.
Lincoln, F.C. 1930. Calculating waterfowl abundance on the basis of banding
returns. US Department of Agriculture Circular No. 118. Washington, D.C.
Lindstrøm, E. and Skulberg, O.M. 2008. Didymosphenia geminata —anative
diatom species of Norwegian rivers coexisting with Atlantic salmon. In
Proceedings of the 2007 International Workshop on Didymosphenia
geminata.Edited by M.L. Bothwell and S.A. Spaulding. Canadian Technical
Report of Fisheries and Aquatic Sciences 2795. pp. 35–40.
Marshall, B.D. 2007. Effects of Libby Dam, habitat, and an invasive diatom,
Didymosphenia geminata, on benthic macroinvertebrate assemblages of the
Kootenai River, Montana. EcoAnalysts, Inc., Bozeman, Mont.
Meredith, C., Roper, B., and Archer, E. 2014. Reductions in instream wood in
streamsnearroadsintheinteriorColumbiaRiverbasin.N.Am.J.Fish.
Manage. 34:493–506. doi:10.1080/02755947.2014.882451.
Mesa, M.G., Weiland, L.K., Christiansen, H.E., Sauter, S.T., and Beauchamp, D.A.
2013. Development and evaluation of a bioenergetics model for Bull Trout.
Trans. Am. Fish. Soc. 142(1): 41–49. doi:10.1080/00028487.2012.720628.
Minshall, G.W., Shafii, B., Price, W.J., Holderman, C., Anders, P.J., Lester, G.,
and Barrett, P. 2014. Effects of nutrient replacement on benthic macroin-
vertebrates in an ultraoligotrophic reach of the Kootenai River, 2003–2010.
Freshw. Sci. 33(4): 1009–1023. doi:10.1086/677900.
Moore, J.W. 2006. Animal ecosystem engineers in streams. Bioscience, 56(3):
237–246. [doi:10.1641/0006-3568(2006)056[0237:AEEIS]2.0.CO;2]
Moss, J.H.H. 2001. Development and application of a bioenergetics model
for Lake Washington prickly sculpin (Cottus asper). M.S. thesis, University
of Washington.
Nagelkerke, N. 1991. A note on a general definition of the coefficient of
determination. Biometrika, 78(3): 691–692. doi:10.1093/biomet/78.3.691.
Nakano, S., and Murakami, M. 2001. Reciprocal subsidies: dynamic interde-
pendence between terrestrial and aquatic food webs. Proc. Natl. Acad.
Sci. U.S.A. 98(1): 166–170. doi:10.1073/pnas.98.1.166.
Nakano, S., Fausch, K.D., and Kitano, S. 1999. Flexible niche partitioning via a for-
aging mode shift: a proposed mechanism for coexistence in stream-dwelling
charrs. J. Anim. Ecol. 68:1079–1092. doi:10.1046/j.1365-2656.1999.00355.x.
Power, M.E., Dietrich, W.E., and Finlay, J.C. 1996. Dams and downstream
aquatic biodiversity: potential food web consequences of hydrologic and geo-
morphic change. Environ. Manage. 20(6): 887–895. doi:10.1007/BF01205969.
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
Clancy et al. 9
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 63.153.124.239 on 01/19/21
For personal use only.

QGIS. 2019. QGIS Geographic Information System. Open Source Geospatial
Foundation Project. Available from http://qgis.org.
Railsback, S.F., and Rose, K.A. 1999. Bioenergetics modeling of stream trout
growth: temperature and food consumption effects. Trans. Am. Fish. Soc.
128(2): 241–256. doi:10.1577/1548-8659(1999)128<0241:bmostg>2.0.co;2.
Ricker, W.E. 1975. Computation and interpretation of biological statistics of
fish populations. Bulletin 191 of the Fisheries Research Board of Canada.
Rosgen, D.L. 1994. A classification of natural rivers. Catena, 22(3): 169–199.
doi:10.1016/0341-8162(94)90001-9.
Schoener, T.W. 1970. Nonsynchronous spatial overlap of lizards in patchy
habitats. Ecology, 51(3): 408–418. doi:10.2307/1935376.
Scholl, E.A., Dutton, A.J., Cross, W.F., and Guy, C.S. 2019. Density of pallid
sturgeon and food web dynamics in the Missouri River: Inferences
regarding carrying capacity and density-dependent response of pallid
sturgeon to the contemporary stocking protocol. Montana State Univer-
sity, Bozeman, Mont.
Strickler, G.S. 1959. Use of the densiometer to estimate density of forest can-
opy on permanent sample plots. US Department of Agriculture Research
Note, Portland, Oreg.
Taylor, B.W., and Bothwell, M.L. 2014. The origin of invasive microorgan-
isms matters for science, policy, and management: the case of Didymos-
phenia geminata. Bioscience, 64(6): 531–538. doi:10.1093/biosci/biu060.
Thurow, R.F. 1997. Habitat utilization and diel behavior of juvenile bull
trout (Salvelinus confluentus) at the onset of winter. Ecol. Freshw. Fish, 6:1–
7. do i : 10.1111/j.1600-0633.1997.tb00136.x.
Tumolo, B.B., Albertson, L.K., Cross, W.F., Daniels, M.D., and Sklar, L.S. 2019.
Occupied and abandoned structures from ecosystem engineering differ-
entially facilitate stream community colonization. Ecosphere, 10(5): 1–15.
doi:10.1002/ecs2.2734.
USGS. 2019. National hydrography dataset plus high resolution. US Geologi-
cal Survey. Available from https://www.usgs.gov/core-science-systems/ngp/
national-hydrography/access-national-hydrography-products.
Wehrly, K.E., Wiley, M.J., and Seelbach, P.W. 2003. Classifying regional vari-
ation in thermal regime based on stream fish community patterns.
Trans. Am. Fish. Soc. 132(1): 18–38. doi:10.1577/1548-8659(2003)132<0018:
crvitr>2.0.co;2.
Wolman, M.G. 1954. A method of sampling coarse river-bed material. Trans.
Am. Geophys. Union, 35(6): 951–956. doi:10.1029/TR035i006p00951.
Appendix Table A1 appears on the following page.
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
10 Can. J. Fish. Aquat. Sci. Vol. 00, 0000
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 63.153.124.239 on 01/19/21
For personal use only.

Appendix A
Table A1. Steams surveyed in 2019.
Stream
Latitude
(°N)
Longitude
(°W)
Subbasin (state
or prov.)
% Didymo
coverage
%
overstory
cover
Wetted
width
(m)
Dominant
vegetation
Rosgen
channel
type
No. of
large
woody
debris
Aug.
stream
temp.
(°C)
Fish spp.
observed
Blacktail
Cr.
48.95124 115.54154 Yaak R. (Mont.) 1.7 82.6 3.4 Pine B 12 10.1 RB
Boulder
Cr.
48.82052 115.29097 Koocanusa
(Mont.)
33.3 80.2 4.7 Cedar A 6 12.5 WCT
Bear Cr. 48.16886 115.58853 Kootenai R.
(Mont.)
30.6 88.7 8.3 Cedar B 19 11.9 BULL, RB
Big Cherry
Cr.
48.20577 115.59134 Kootenai R.
(Mont.)
29.0 74.5 5.9 Cedar A 75 12.5 BULL, RB,
RBCT, WCT,
SLCOT
Burnt Cr. 48.72936 115.87002 Yaak R. (Mont.) 31.6 64.5 7.3 Cedar A 12 14.4 EB, LN DC, MWF,
RB, SLCOT
Coffee Cr. 49.69663 116.91781 Kootenay L. (B.C.) 2.0 —14.3 Cedar A —12.9 BULL, WCT
Davis Cr. 50.14236 116.95520 Kootenay L. (B.C.) 20.5 60.6 9.8 Cedar A 5 11.8 BULL, MWF, RB,
SLCOT
E. Fork
Pipe Cr.
48.61675 115.61885 Kootenai R.
(Mont.)
24.6 83.7 3.2 Alder B 12 10.9 EB, RB, RBCT,
SL COT
E. Fork
Yaak R.
48.94885 115.53378 Yaak R. (Mont.) 52.6 31.5 7.6 Pine B 4 11.4 RB
Granite Cr. 48.29544 115.62011 Kootenai R.
(Mont.)
21.3 68.1 10.3 Cedar B 7 11.6 BULL, EB, RB,
SLCOT
Hope Cr. 50.45751 117.19081 Lardeau R. (B.C.) 16.6 67.3 4.8 Cedar A 16 12.7 BULL, MWF, RB
Kokanee
Cr.
49.60506 117.12635 Kootenay L. (B.C.) 0.0 32.0 15.9 Alder A —13.9 RB
Lake Cr. 48.44899 115.87918 Kootenai R.
(Mont.)
27.0 6.3 19.0 Alder B 0 —RB, TCOT
Leigh Cr. 48.22127 115.60603 Kootenai R.
(Mont.)
3.2 88.0 4.6 Cedar A 20 9.9 EB, RBCT
Lizard Cr. 49.48972 115.10481 Elk R. (B.C.) 15.4 26.0 6.8 Pine B 5 11.5 EB, RBCT
Lockhart
Cr.
49.50892 116.78628 Kootenay L. (B.C.) 40.6 80.4 5.0 Cedar A —11.5 BULL, LNDC,
MWF, RB
Mobbs Cr. 50.50673 117.27104 Lardeau R. (B.C.) 48.3 28.1 3.2 Alder Side
channel
8 10.6 BULL, RB, SLCOT
Lost Ledge
Cr.
49.90834 116.90733 Kootenay L. (B.C.) 5.7 69.2 4.9 Cedar A 9 12.8 RB
N. Fork 17
Mile Cr.
48.66022 115.76750 Yaak R. (Mont.) 17.7 66.0 3.8 Cedar A 15 11.9 EB, RB, SLCOT,
WCT
Outlet Cr. 50.16812 115.46405 White R. (B.C.) 80.5 64.0 6.3 Pine B 8 18.2 RB
Parmenter
Cr.
48.37814 115.62908 Kootenai R.
(Mont.)
0.9 83.2 7.7 Cedar B 40 10.7 EB, RB, SLCOT
Pinkham
Cr.
48.82799 115.24295 Koocanusa
(Mont.)
39.4 53.4 4.5 Alder A —11.3 EB, RB
Solo Joe
Cr.
48.92425 115.53963 Yaak R. (Mont.) 0.0 81.1 2.8 Pine A 21 11.7 RB
Trail Cr. 48.03825 115.46030 Fisher R. (Mont.) 13.0 63.7 4.9 Pine B 10 13.9 EB, TCOT, WCT
W. Fisher
Cr.
48.04253 115.47351 Fisher R. (Mont.) 1.3 26.9 5.6 Pine B 2 11.1 BULL, TCOT,
WCT
Wolf Cr. 48.23393 115.28529 Fisher R. (Mont.) 0.0 32.0 9.8 Alder B 3 20.2 LNDC, LSSU,
MWF, RSSH,
TCOT, WCT
Woodbury
Cr.
49.80625 117.02835 Kootenay L. (B.C.) 0.2 37.9 10.1 Cedar B —8.6 BULL, WCT
Weasel Cr. 48.94904 114.73400 Wigwam R.
(Mont.)
52.5 58.0 3.5 Pine B 5 15.3 RBCT
Note: Species codes: EB, brook trout; BULL, bull trout; LNDC, longnose dace; LSSU, largescale sucker; MWF, mountain whitefish; RB, rainbow trout; RBCT,
rainbow–cutthroat hybrid; RSSH, redside shiner; SLCOT, slimy sculpin; TCOT, torrent sculpin; WCT, westslope cutthroat trout.
Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
Clancy et al. 11
Published by NRC Research Press
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by 63.153.124.239 on 01/19/21
For personal use only.