Volume 70

2013

An NRC Research Press Journal

Un journal deNRC Research Press

www.nrcresearchpress.com

Canadian Journal of

Fisheries andAquatic Sciences

Journal canadien des

sciences halieutiqueset aquatiques

ARTICLE

Carcass analog addition enhances juvenile Atlantic salmon (Salmo salar)growth and conditionMargaret Q. Guyette, Cynthia S. Loftin, and Joseph Zydlewski

Abstract: Our study used historic marine-derived nutrient (MDN) delivery timing to simulate potential effects of restoredconnectivity on juvenile Atlantic salmon (ATS; Salmo salar) growth and condition. Four headwater streamswere stockedwith ATSyoung of the year (YOY) and received carcass analog additions (0.10 kg·m–2 wetted area) in treatment reaches tomatch the timingof sea lamprey (Petromyzon marinus) spawning. Individual ATSmass was 33%–48% greater and standard length was 9%–15% greaterin treatment reaches relative to control reaches for 4 months following nutrient additions. Percent total lipids in YOY ATS weretwice as great in treatment reaches 1 month following carcass analog additions and remained elevated in treatment fish for2moremonths. Absolute growth rates, based on otolithmicrostructure analysis, correlatedwithwater temperature fluctuationsin all reaches and were elevated by an average of 0.07 mm·day–1 in treatment reaches for 1 month following carcass analogadditions. Simulated sea lampreyMDNs increased juvenile ATS growth, which, via potential increases in overwinter survival anddecreases in smolt age, may contribute to population persistence and ecosystem productivity.

Résumé : Notre étude fait appel a des données historiques sur le moment de l’apport de nutriants de source marine (MDN) poursimuler les effets potentiels du rétablissement de la connectivité sur la croissance et l’embonpoint des saumons atlantiques (ATS;Salmo salar) juvéniles. Quatre cours d’eau d’amont ont été empoissonnés avec des jeunes de l’année (YOY) d’ATS et ont fait l’objetd’ajouts d’analogues de carcasses (0,10 kg·m–2 de surface mouillée) dans des tronçons traités pour simuler le moment du frai deslamproies (Petromyzon marinus). La masse d’ATS individuels était de 33 % a 48 % plus importante et leur longueur standard, de 9 %a 15 % plus grande dans les tronçons traités que dans les tronçons témoins pour les 4 mois suivant l’ajout de nutriants. Lepourcentage de lipides totaux dans les YOY d’ATS était deux fois plus élevé dans les tronçons traités 1 mois après l’ajoutd’analogues de carcasses et est demeuré élevé dans les poissons de tronçons traités pendant 2 autres mois. Il y avait unecorrélation entre les taux de croissance absolue, tirés de l’analyse de microstructures d’otolites, et les fluctuations de latempérature de l’eau dans tous les tronçons, et ces taux étaient de 0,07 mm·jour–1 en moyenne plus élevés dans les tronçonstraités pendant le mois suivant les ajouts d’analogues de carcasses. Les MDN de lamproies simulés se sont traduits par unecroissance accrue des ATS juvéniles ce qui, par l’entremise d’augmentations potentielles de la survie hivernale et de réductionsde l’âge des saumoneaux, peut contribuer a accroître la persistance des populations et la productivité des écosystèmes. [Traduitpar la Rédaction]

IntroductionAtlantic salmon (ATS; Salmo salar) populations have declined

precipitously across much of their historic range largely becausepassage between marine and freshwater habitats has been re-duced or blocked, decreasing connectivity (Moring 2005). Thisconnectivity loss similarly has reduced or eliminated populationsof other co-evolved diadromous fishes (Limburg and Waldman2009). Growth opportunities are great in the highly productivemarine environment (McCormick et al. 1998), and the majority ofanadromous fish biomass is gained in the ocean (Flecker et al.2010; Thorpe 1984). As a result, anadromous fish act as vectors ofmarine-derived nutrients (MDNs) into freshwater systems duringtheir spawning runs, when metabolic waste, unsuccessful eggs,and carcasses may be deposited within the spawning grounds.Many streams inMaine are oligotrophic (Danielson 2010), and thisin part may be due to the absence of anadromous fish runs. Re-storing anadromous fish populations may increase productivity,creating a positive feedback loop (Chaloner et al. 2002; Wipfliet al. 1998) as productivity increases the success of juvenile Atlan-tic salmon and other anadromous fish.

The diversity of anadromous species' life histories and spawn-ing event timing influences the magnitude of MDNs enteringfreshwater environments and the subsequent potential effects ofthese MDNs on the ecosystems (Flecker et al. 2010). In westernNorth American ecosystems, MDNs imported by Pacific salmon(Oncorhynchus spp.) have resulted in increases in periphyton bio-mass (Johnston et al. 2004); taxa-specific increases and decreasesin macroinvertebrate abundance, biomass, and secondary pro-duction (Chaloner et al. 2004; Lessard and Merritt 2006; Lessardet al. 2009); site-specific variability in response by juvenile sal-monids (Lang et al. 2006); increases in growth and survival ofjuvenile salmonids (Stockner andMacIsaac 1996); and increases inadult salmonid recruitment (Michael 1995). The role of ATS asnutrient vectors between marine and freshwater environmentshas been studied in Europe (Jonsson and Jonsson 2003; Lyle andElliott 1998); however, the effects ofMDNs on easternNorth Amer-ican ecosystems have received less study (Garman and Macko1998; Nislow and Kynard 2009; Walters et al. 2009).

There are clear differences between western and eastern NorthAmerican anadromous fish assemblages and life histories suchthat extrapolating Pacific MDN research results to the northwest

Received 15 November 2012. Accepted 29 March 2013.

Paper handled by Associate Editor Bror Jonsson.

M.Q. Guyette. Department of Wildlife Ecology, 5755 Nutting Hall, University of Maine, Orono, ME 04469, USA.C.S. Loftin and J. Zydlewski. Department of Wildlife Ecology, 5755 Nutting Hall, University of Maine, Orono, ME 04469, USA; US Geological Survey, Maine Cooperative Fish and WildlifeResearch Unit, 5755 Nutting Hall, Orono, ME 04469, USA.

Corresponding author: Margaret Q. Guyette (e-mail: margaret.guyette@gmail.com).

860

Can. J. Fish. Aquat. Sci. 70: 860–870 (2013) dx.doi.org/10.1139/cjfas-2012-0496 Published at www.nrcresearchpress.com/cjfas on 11 April 2013.

Can

. J. F

ish.

Aqu

at. S

ci. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y C

SP S

taff

on

06/1

1/13

For

pers

onal

use

onl

y.

Atlantic may be inappropriate. In contrast with Pacific salmon,ATS are iteroparous, and nutrient delivery occurs primarily asmetabolic waste products and unsuccessful eggs. In addition, ATSspawning occurs in the fall, when temperature and productivityare reduced in lotic systems (Huryn andWallace 2000; Morin et al.1999). Unlike ATS, sea lamprey (Petromyzon marinus) spawn in thespring, are semelparous, and nutrient delivery includes carcasses.Spawning inmuch of the same habitat as ATS, sea lampreymay bea major vector for MDNs into ATS nursery streams.

Improving juvenile ATS condition is a management objectivethroughout its range, because growth and survival of juvenile ATSaffect smolt production and persistence of ATS populations(McCormick et al. 1998). MDNs may indirectly affect juvenile ATSby increasing productivity of the freshwater community and theprey base for fish (Bilby et al. 1996; Kohler et al. 2008; Wipfli et al.1999), as well as through direct ingestion of eggs and carcass ma-terial (Bilby et al. 1998). The timing of the sea lamprey spawningevent may be particularly advantageous for juvenile ATS, becauseit occurs concurrently with ATS emergence and the advent offeeding (Saunders et al. 2006). Sea lamprey spawning events likelywill increase ATS growth and survival more than nutrient loadingin the fall from ATS spawning events because the magnitude ofnutrient influx is greater and fish growth potential increases withincreasing temperature (Brett 1979).

Restoring nutrients to freshwater environments that have lostthe annual nutrient contribution from spawning salmon is a con-servation goal in western North American lotic systems. Artificialdelivery of nutrients as inorganic fertilizers (Peterson et al. 1993;

Slaney et al. 2003; Wipfli et al. 2010) and fish carcasses (Giannicoand Hinch 2007; Wipfli et al. 2010) has been used experimentallyto assess effects of elevated nutrient concentrations on recov-ery of aquatic systems, and nutrient supplementation has beenadopted as a restoration strategy in systems with depleted ana-dromous fish runs (Griswold et al. 2003; Oregon WatershedEnhancement Board 2006). While carcass analogs do not simu-late all aspects of spawning run contributions and effects tofreshwater ecosystems, they can be used as a potential nutrientsource to simulate MDN delivery (Kohler et al. 2008; Martinet al. 2010; Wipfli et al. 2004). Carcass analogs added to Pacificsalmon streams have increased productivity analogous to thatfrom carcass additions (Martin et al. 2010; Wipfli et al. 2004).

We determined effects of carcass analog application timed tomatch sea lamprey spawning on juvenile ATS. We hypothesizedthat carcass analog application in sea lamprey spawning habitatin headwater streams would (i) increase juvenile ATS size andgrowth rates and (ii) increase lipids in the first year of ATS growth.As connectivity is restored in the region, the diverse life historiesof anadromous species in these systems likely will benefit theirrecovery. Our approach explores the potential effects of one ana-dromous species on the juvenile life stages of another, whichmayhave far-reaching consequences for restoration efforts.

Materials and methodsStudy area

We conducted this study in four headwater tributaries of Kings-bury Stream (Kingsbury Plantation township, Piscataquis County,

Fig. 1. Location of study sites within Penobscot River watershed (outlined), Maine, USA. Four study streams (A, B, C, and D) are withinKingsbury Plantation township and are tributaries of Kingsbury Stream, which flows west to east from Kingsbury Pond to the PiscataquisRiver (not shown). Each study stream contains a control reach, a buffer, and a treatment reach.

AB

C

D

Control

(300m)

Treatment

(300m)

Buffer

(100-500m)

water

flow

Kingsbury Stream

km

Guyette et al. 861

Published by NRC Research Press

Can

. J. F

ish.

Aqu

at. S

ci. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y C

SP S

taff

on

06/1

1/13

For

pers

onal

use

onl

y.

Maine, USA) in the Penobscot River basin (Fig. 1). Small dams onthe river's tributaries inhibited fish passage beginning in the late1700s, and the first large main stem dam was built in the 1820s(MDMR and MDIFW 2007). The basin contains thousands of cul-verts and at least 116 dams, 24 of which are federally licensed forhydropower (Fay et al. 2006). These barriers greatly limit anadro-mous species' access to spawning habitat. The region primarily ismixed deciduous forest with a long logging history. Brook trout(Salvelinus fontinalis) dominate the resident fish populations inthese streams; eastern blacknose dace (Rhinichthys atratulus) arefound in low densities.

We established an upstream control reach (300 m) and a down-stream treatment reach (300 m) in each stream, with a 100–500 mbuffer between these reaches (Fig. 1). The maximum distanceyoung-of-the-year (YOY) ATSmove upstream 4months after stock-ing in Maine is �500 m in systems with few physical barriers(Mike Loughlin, Maine Department of Marine Resources, MDMR,USA, personal communication). The buffer length in each streamwas determined by the stream gradient and height of physicalbarriers between control and treatment reaches. Stream A had a100 m buffer that contained two >40 cm tall falls that created abarrier for upstream YOY ATS movement. Stream B had a longer,500 m buffer; there were no physical barriers >30 cm tall thatwould limit YOY ATS movement upstream. Both Streams C and Dhad 400–500 m buffers with multiple physical barriers >30 cm.

Physical characteristics of reaches within and between streamswere similar (Table 1). Gradient of each reach was measured witha laser transit (Rugby 100, Leica Geosystems, Switzerland). Sub-strate characterization included visual estimates of percent coverof silt, sand, gravel, cobble, boulder, leaf litter, and moss within0.25 m2 survey plots sampled in the center of riffles at 25 mintervals. Water temperature was recorded at 5 min intervals forthe duration of the study with HOBO loggers (UA-001-64 PendantTemp, Onset Computer Corporation, Bourne, Massachusetts,USA) submerged and anchored to the substrate in flowingwater inshaded sections of each reach. Canopy cover was measured withhemispherical photography. Photographs were taken at 25 m in-tervals with a Canon Powershot G3 (Canon U.S.A. Inc., Lake Suc-cess, Massachusetts, USA) with a fisheye lens placed on a platformon a tripod in the center of the stream 1mabove thewater surface.Percent openness was measured with Winphot 5.0 (ter Steege1996). The study streams were oligotrophic prior to nutrient ma-nipulations (total dissolved nitrogen (TDN): 0.10–0.32mg·L–1; totaldissolved phosphorus (TDP): 4–28 �g·L–1), and conductivity waslow (10–35 �S·cm–1).

Experimental design

Atlantic salmon marking and stockingFirst filial generation (F1) hatchery-raised ATS otoliths were

marked thermally before hatching in January and February 2010as two separate groups at the US Fish and Wildlife Service Craig

Brook National Fish Hatchery, East Orland, Maine, USA. Onegroup received a series of four marks in a chilled recirculatingsystem by exposing the eggs to alternating cycles of 36 h with a4 °C increase in temperature and 36 h at ambient temperature(1.7 °C). The second group received a series of two marks, a 72 hambient period, and a series of two more marks. The procedureused for marking otoliths in 2009 was not effective, and thermalmarks on the 2009 year class ATS were indistinguishable.

Salmon fry were stocked throughout the streams from 100 mabove to 12.5 m below each control and treatment reach duringconditions specified by the MDMR, Bureau of Sea-Run Fisheriesand Habitat, Bangor, Maine, USA. Control and treatment sectionswere stocked with fish from different thermally marked groups.Fry density in the research streams was increased to 400 fry per100 m2 wetted area, which is four times the typical stocking den-sity in Maine (Fay et al. 2006), to maximize the potential numberof surviving fry. Unfed fry were stocked on 18 May 2009, and fedfry were stocked on 14 May 2010. The developmental index (Kane1988), which is 0 at fertilization and 100 at first feeding, was�93 in2009 and �123 in 2010. Fry were distributed evenly every 12.5 mthroughout the control and treatment reaches.

Carcass analog additionCarcass analogs were applied to treatment reaches (Fig. 1) at a

density of 0.1 kg analog·m–2 on 14 July and 31 October 2009 and1 July 2010 (Fig. 2). July delivery matched sea lamprey spawningtiming, and October delivery matched ATS spawning timing. Thesame application density was used for all dates to facilitate com-parisons. Carcass analogs were placed in Vexar 6.35 mm meshbags, anchored to bags filledwith rocks collected from the stream-bed, randomly distributed throughout each treatment reach, andleft in place until analog pellets had dispersed. Eachmesh bagwasfilled with �800 g of carcass analog material. Carcass analog ma-terial, produced by BioOregon, Inc. (Warrenton, Oregon, USA),primarily was derived from fall run Chinook salmon (Oncorhynchustschawytcha) of hatchery origin from the Pacific Northwest; thepellets were approximately 10% nitrogen (N), 2% phosphorus (P),13% lipids, and 56% crude protein (Pearsons et al. 2007).

Field samplingWater samples were collected (Fig. 2) in flowing water in the

center of the stream at three locations within each control andtreatment reach in 2010. Duplicate samples were collected with a60 mL syringe and filtered through 0.40 �m polycarbonate mem-brane filters (GE Osmonics, Minnetonka, Minnesota, USA) with aMillipore Swinnex filter holder (Millipore Corporation, Billerica,Massachusetts, USA) into clean amber glass jars. One sample wasacidifiedwith 50% sulfuric acid and held at 4 °C prior to N analysis.The other sample was frozen (–20 °C) or held at 4 °C for less than48 h prior to P analysis.

Table 1. Physical characteristics in control (C) and treatment (T) reaches in the four study streams.

Stream A Stream B Stream C Stream D

C T C T C T C T

Mean bankfull width (m) 2.9 (2.2–3.8) 3.5 (2.3–4.7) 5.4 (4.8–6.6) 5.4 (3.8–6.9) 3.9 (3.4–4.7) 3.7 (2.6–5.0) 3.8 (2.5–5.9) 5.3 (3.6–7.5)Gradient (%) 3.8 (2.7–5.5) 3.2 (1.5–6.2) 2.9 (1.9–3.5) 2.2 (1.0–4.3) 4.6 (1.5–8.1) 3.7 (1.8–5.2) 4.4 (3.4–5.0) 3.1 (1.6–5.2)Gravel–cobble (%) 68 (10–100) 57 (10–100) 59 (10–100) 73 (10–100) 49 (0–100) 53 (0–100) 58 (10–100) 63 (5–100)Water temperature (°C) 13.4 (4.0–20.7) 13.9 (4.0–22.1) 14.0 (3.3–22.4) 14.3 (3.2–22.3) 13.0 (3.9–19.4) 13.8 (3.6–21.5) 13.1 (3.8–20.9) 12.7 (3.8–19.9)Elevation (m) 334 346 306 308Riparian overstory vegetation� Fagr Abba Abba

Fagr Fagr Bepa Fagr Fagr Fagr Bepa FramTsca Tsca Tsca Bepa Pima Fagr

Canopy openness (%) 33 (25–43) 27 (21–35) 35 (31–40) 36 (33–40) 34 (29–38) 29 (23–35) 30 (24–37) 29 (24–34)

Note: Values are reported as means with ranges in parentheses (minimum to maximum).�Dominant overstory vegetation: Fagus grandifolia (Fagr, American beech), Abies balsamea (Abba, balsam fir), Fraxinus americanus (Fram, white ash), Tsuga canadensis

(Tsca, eastern hemlock), Betula papyrifera (Bepa, white birch), Picea mariana (Pima, black spruce).

862 Can. J. Fish. Aquat. Sci. Vol. 70, 2013

Published by NRC Research Press

Can

. J. F

ish.

Aqu

at. S

ci. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y C

SP S

taff

on

06/1

1/13

For

pers

onal

use

onl

y.

Juvenile ATS were collected (Fig. 2) by backpack electrofishingconducted at 100 m increments from downstream sections to up-stream sections in each control and treatment reach. Samplingcontinued until the target sample size (eight to nine YOY and fourage 1 per 100 m section) was reached, which usually occurredwithin the first 25 m of the section. Atlantic salmon were eutha-nized on dry ice individually and stored at –20 °C prior to prepa-ration for stable isotope analysis (Guyette 2012) or –80 °C prior topreparation for lipid analysis.

Laboratory analysis

Water nutrient analysisTDN was measured with the automated cadmium reduction

method after persulfate digestion (APHA 1999; Delia et al. 1977).TDP was measured with the ascorbic acid method after persulfateoxidation (APHA 1999). Water analyses were performed at theSawyer Environmental Chemistry Research Laboratory (Univer-sity of Maine, Orono, Maine, USA). Detection limits for these anal-yses were determined by the laboratory (0.1 mg TDN·L–1 and 1.0 �gTDP·L–1).

Sample preparationWemeasured the blotted mass of each thawed ATS to the near-

est 0.1 mg. Each fish was photographed, and the standard lengthwas measured to the nearest 0.01 mmwith ImageJ (National Insti-tutes of Health, Bethesda, Maryland, USA). Sagittal otoliths weredissected from all sampled fish with fine forceps and a dissectingmicroscope under cross-polarized light (EMZ-13TR, Meiji TechnoAmerica, Santa Clara, California, USA). Otoliths were mounted onmicroscope slides with clear nail polish.

Otolith microstructure analysisPrepared otoliths were sanded withmicrofinishing and lapping

film to level the core of the otolith. Otoliths were viewed under100× and 200× magnification with a compound microscope todistinguish thermal marks (Fig. 3a). All fish with thermal marksindicating movement between control and treatment reaches orwith indistinguishable thermal marks were excluded from thestudy. Otoliths were photographed under 80× magnification witha compound microscope linked to a high-resolution camera andtiled together with Image-Pro Insight version 8.0 software (MediaCybernetics, Bethesda, Maryland, USA). Increments were countedand measured with the line profile tool in Image-Pro Insight inone of each pair of otoliths along a transect starting in the nucleusand forming a 45 degree angle from the posterior axis in theventral region (Johnston et al. 2005; Fig. 3b).

ComputationsThe standard length of ATS at each measurable otolith incre-

ment was back-calculated with the time-varying growth (TVG)method (Sirois et al. 1998):

(1) Lt � L0 � �i�1

t

[Wi � R(Wi � W)](Lc � L0)(Oc � O0)�1

where L is the length at age t (Lt), capture (Lc), and the biologicalintercept (L0), and O is the otolith radius at capture (Oc) and thebiological intercept (O0; Fig. 3c).Wi is the increment width at age i,andW is themean increment width. No differences were detectedbetween using values ofW based on shorter averages (1 or 2weeks)or longer averages (the entire range; paired t test: t2060 = 1.03, p =0.30); therefore, a global mean W was used for each individualfish. The growth effect (R) was calculated as described in Siroiset al. (1998) with a linear regression using all measured individualfish:

(2) S � b � RG

(3) S � (Lc � L0)(Oc � O0)�1

(4) G � (Lc � L0)t�1

where S is the estimated slope of the fish size – otolith size rela-tionship, b is the estimated y intercept, and G is the linear growthrate in length for each fish.

The TVG method requires continuous increment measure-ments from the biological intercept to the point of back-calculation. The biological intercept for salmonids usually isidentified as the day of hatching (Campana 1990; Meekan et al.1998); however, increments between hatching and stocking onthe otoliths in this study were indistinct, reflecting consistentambient daily temperature and light in the hatchery. Daily incre-ments were distinct following stocking, and we used a sample offish (n = 72) collected on the day of stocking to establish the bio-logical intercept (mean ± 1 SD: L0 = 26.6 ± 0.13 mm; O0 = 218.2 ±10.61 �m). Stocking checks were verified on a minimum of fivefish per July and August sampling period by counting incrementsfrom the stocking check to the outer edge. The assumptions ofdaily increment formation and proportionality between somaticgrowth and otolith growth for this back-calculation techniquehave been validated for juvenile Atlantic salmon (Meekan et al.1998).

Back-calculated lengths at each increment were used to calcu-late absolute growth rate, in �m·day–1:

(5) AGR � Lt � Lt�1

where the time increment measured is always 1 day.

Lipid analysisWemeasured total lipids of 11–14 individual YOY ATS from con-

trol and treatment sections of Stream C on each sample date in2010. We chose to analyze fish from a single stream because oftime and budget limitations. After otolith extraction, we mea-sured ATS wet mass, dried the whole fish at 60 °C for 24 h, mea-sured dry mass to the nearest 0.1 mg, and ground the dried fishwith a mortar and pestle into a fine powder. Whole body totallipids were extracted following the Folch method (Folch et al.

Fig. 2. Timing of carcass analog additions (black vertical bars) and fish and water sampling (grey vertical bars) in experimental streams.

Fish

14 July 1 Aug.200917

Fish

Week

Water

9 31215

2010 1 July

4

1 Sept.

0 611- 2 3

1 Aug.

1 Sept.

1 Oct.

15 167 81 Oct.

10 111 Nov.

14

Guyette et al. 863

Published by NRC Research Press

Can

. J. F

ish.

Aqu

at. S

ci. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y C

SP S

taff

on

06/1

1/13

For

pers

onal

use

onl

y.

1957) with 0.5 g aliquots of dried whole fish with sample repeatson 10% of the fish. The mass of each lipid sample was measured tothe nearest 0.01 mg, and measurements of total lipids were ex-pressed as percentages of the original wet mass of the whole fish.

Statistical analysisResponse variables for analyses were dissolved nutrient concen-

trations, fish mass, standard length, total percent lipid, and abso-lute growth rate (AGR). We assessed treatment and temporaleffects on ATS mass, length, and AGR with Wilcoxon rank sumtests because assumptions of normality and homogeneity of vari-ance were not met. We tested for breakpoints in the AGR patternover time for both control and treatment datasets with segmentedlinear regression (Muggeo 2003, 2008).We assessed treatment and

temporal effects on natural logarithm-transformed ATS total per-cent lipid with a two-way ANOVA, followed by post hoc pairwisecomparisons with Tukey's honestly significant differences. Sam-ples were pooled across all streams, and statistical significancewas determined with � of 0.05. All analyses were performed withthe statistical package R, version 2.15.0 (R Development CoreTeam 2012), with the R packages “MASS” (Venables and Ripley2002), “car” (Fox and Weisberg 2011), “fBasics”, and “segmented”(Muggeo 2003, 2008).

ResultsNutrient concentrations were similar (TDP: F[1,70] = 0.024, p =

0.88; TDN: F[1,60] = 0.006, p = 0.94) in the control and treatment

Fig. 3. (a) Young-of-the-year Atlantic salmon otoliths collected on stocking day (14 May 2010) under 100× magnification. Otoliths (i) and (ii)display the two thermal mark patterns used in this study. (b) The measurement transect starts in the center of the nucleus of the otolith andforms a 45 degree angle from the posterior axis in the ventral region. (c) Measurements indicated on the transect are the otolith radius atcapture, Oc, and otolith radius at the biological intercept, O0.

200 µm200 µm

O0

Oc

(a)

(b)

(c)

(i) (ii)

864 Can. J. Fish. Aquat. Sci. Vol. 70, 2013

Published by NRC Research Press

Can

. J. F

ish.

Aqu

at. S

ci. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y C

SP S

taff

on

06/1

1/13

For

pers

onal

use

onl

y.

reaches prior to analog addition (Fig. 4). Nutrient concentrationswere 405% (P) and 82% (N) greater in treatment reaches relative tocontrol reaches 2 days after carcass analog addition (TDP: F[1,22] =170, p < 0.001; TDN: F[1,20] = 30.0, p < 0.001). There was a significantinteraction between treatment and date for TDP samples col-lected until early August (F[8,198] = 3.04, p = 0.003). TDP was 131%–151% greater in treatment reaches on all sample dates until5 August (p < 0.001). There were significant treatment (F[1,191] = 164,p < 0.001) and date (F[8,191] = 2.12, p = 0.04) effects for TDN samplescollected until early August, and TDN was 61%–135% greater until19 July (p < 0.05), no significant differences were detected on 22July, and TDN was 85% greater on 29 July (p = 0.02). No significantdifferences between treatment and control reaches were detectedfor TDP or TDN after 5 August.

Thermal marks on ATS otoliths stocked in 2009 were not dis-tinct, and analyses of YOY ATS in 2009 and age 1 ATS in 2010included all sampled fish. Thermal marks on YOY ATS otoliths

stocked in 2010 were clear (Fig. 3a) and were used to exclude fishmoving between control and treatment sections, as well as fishwith indistinguishable thermal marks or unreadable otoliths(6.2% of the fish sampled from control reaches and 15.5% of thefish sampled from treatment reaches).

A significant difference before carcass analog addition in onestream (Stream C) in 2009 resulted in significantly larger treat-ment fishwhen control fish and treatment fishwere pooled acrossall streams (mass: p = 0.020, length: p = 0.027). There were nosignificant differences inmass and standard length before carcassanalog addition when Stream C was excluded in 2009. TreatmentYOY ATS were 39%–48% larger in mass and 11%–12% larger inlength in 2009 than control fish (n = 99–200, p < 0.001; Figs. 5a, 5c).Mass and standard length of YOY ATS were not significantly dif-ferent between control and treatment reaches before carcass an-alog addition in 2010. Treatment YOY ATS were 33%–48% larger inmass and 9%–15% larger in length in 2010 than control fish (n =81–196, p < 0.001; Figs. 5b, 5d). Treatment age 1 ATS were 6% largerin length in early May 2010, before 2010 YOY stocking (n = 67–71,p = 0.015; Table 2) and 21% and 19% larger inmass in July (n = 29–30,p = 0.038) and September (n = 25–28, p = 0.053), respectively(Table 2).

The estimated magnitude of the growth effect (R) was 0.369(Fig. 6). Absolute growth rates in 2010 YOY ATS were greater by anaverage of 0.07 mm·day–1 in treatment fish within 1 week aftercarcass analog addition (Fig. 7). Breakpoints in the pattern of AGRchange over time were not different between control and treat-ment fish; however, the slopes of the segmented lines were differ-ent, particularly in the week following carcass analog addition(Fig. 7). Changes in both control and treatment AGR correlatedwith changes in water temperature (Fig. 7).

There were significant treatment (F[1,106] = 72.4, p < 0.001) andtemporal (F[4,106] = 7.3, p < 0.001) effects on 2010 YOY ATS percenttotal lipid by wet mass, and there was a significant interactionbetween treatment and time (F[4,106] = 14.3, p < 0.001), reflectingchanges in the magnitude of the treatment effect over time. Posthoc pairwise comparisons revealed significant differences in per-cent total lipid between control and treatment fish after carcassanalog addition in July (p < 0.001), August (p < 0.001), and Septem-ber (p = 0.026; Fig. 8).

DiscussionMarine-derived nutrients may play a role in the recovery of

anadromous fish populations. In western North America, re-search suggests MDNs in freshwater ecosystems can increasestream productivity, potentially enhancing juvenile anadromoussalmonid growth and survival (Kiernan et al. 2010; Wipfli et al.1999); however, few studies experimentally examine effects ofMDNs on juvenile anadromous fish (Giannico and Hinch 2007;Lang et al. 2006; Michael 1995). Nutrient additions as inorganicP or ATS carcasses can increase juvenile ATS density and biomass(Weng et al. 2001; Williams et al. 2009); however, these studiesused inorganic nutrient sources or did not incorporate naturaltiming of MDN delivery. Our study simulates potential MDN ef-fects after restored freshwater–marine system connectivity byconsidering the timing of MDN contribution by co-evolved ana-dromous fish also found in northeasternNorthAmerican streams.The sea lamprey spawning event simulated with carcass analogadditions in our four study streams was followed by increasedYOY ATS mass, standard length, percent total lipids, and absolutegrowth rates during the first growth season, suggesting the poten-tial value of this nutrient source to comparable stream systems innortheastern North America.

YOY ATS size increased after carcass analog additions, and bylate October YOY ATS were 41% larger in 2009 and 34% larger in2010 by mass in treatment reaches than in control reaches, whichis consistent with results from studies of Pacific salmon and resi-

Fig. 4. (a) Total dissolved nitrogen and (b) total dissolved phos-phorus in stream water during the 2010 sampling season. Datapoints shown are means ± 1 SE, pooled across all sample streams.The vertical dashed line indicates the 1 July carcass analog addition,and shaded regions indicate significant differences between controland treatment reaches.

Guyette et al. 865

Published by NRC Research Press

Can

. J. F

ish.

Aqu

at. S

ci. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y C

SP S

taff

on

06/1

1/13

For

pers

onal

use

onl

y.

dent salmonids (Bilby et al. 1998; Wipfli et al. 2003). Greater bodysize increases overwinter survival for juvenile ATS (Huusko et al.2007; Johnston et al. 2005) and has been associated with fastergrowth rates during winter (Murphy et al. 2006). Fish are able toconsume larger prey as body size increases (Brett 1979), in effectenhancing diet quality and quantity and creating a positive feed-back loop. The physiological decision to smolt strongly dependson growth (Thorpe et al. 1998). Larger juveniles have a higherprobability of smolting at age, whereas smaller fish delay theirmigration for at least 1 additional year (Heggenes and Metcalfe1991; Nicieza et al. 1994; Thorpe 1977; Utrilla and Lobon-Cervia1999). These data are consistent with the view that the parr–smolttransformation is a physiological threshold of development(Thorpe et al. 1998). Fall size thresholds for smolting in Atlanticsalmon have been reported as 115–125 mm from passive inte-grated transponder (PIT) telemetry studies in Maine (Horton et al.2009). A similar threshold of 112 mm is reported for hatchery-stocked parr in the Penobscot River drainage (J. Zydlewski, unpub-lished data). Thus an early 10% to 15% increase in growth observedin this study is not likely to result in age 1+ smolts, but this growthadvantage will likely increase the probability that these fish willmigrate as age 2+ rather than age 3+ smolts if this growth advan-tage continues into the second year. The data for age 1+ fish fromthis study are suggestive but underscore the need for more assess-

Fig. 5. Young-of-the-year Atlantic salmon mass (a, b) and standard length (c, d) in 2009 (a, c) and 2010 (b, d). Carcass analog addition (verticaldashed line) occurred on 14 July 2009 and 1 July 2010. Data points are means ± 1 SE, pooled across all sample streams.

Table 2. Sample sizes (n), means (±1 SD) of mass and standard length, and results of statistical comparisons (p) for age 1 Atlantic salmon in 2010.

n Mass (g) Standard Length (mm)

Control Treatment Control Treatment p� Control Treatment p�

11 May 67 72 3.0±1.08 3.5±1.37 0.064 57.5±6.88 61.2±8.44 0.01522 June 31 34 4.9±1.59 5.9±2.26 0.087 71.6±7.37 75.2±10.46 0.14526 July 29 30 5.7±2.09 6.9±2.42 0.038 75.7±8.79 79.9±10.79 0.09623 August 22 31 5.6±1.57 5.2±2.23 0.212 77.5±6.56 73.5±11.42 0.13327 September 25 28 6.0±1.30 7.1±2.08 0.053 80.0±5.54 83.4±8.31 0.12123 October 27 34 6.1±1.65 7.0±2.85 0.302 79.2±6.47 82.3±13.14 0.276

�Reported p values reflect results from Wilcoxon rank sum tests. Values in bold font indicate statisitically significant data (p < 0.05).

Fig. 6. The relationship between the estimated slope (S) of the fishsize – otolith size relationship (S = (Lc – L0)(Oc – O0)–1) and the lineargrowth rate in length (G) for each fish (G = (Lc – L0)t–1) with data fromall measured fish (n = 443, F[1,441] = 1042, p < 0.001). Regression isbased on pooled control and treatment fish.

866 Can. J. Fish. Aquat. Sci. Vol. 70, 2013

Published by NRC Research Press

Can

. J. F

ish.

Aqu

at. S

ci. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y C

SP S

taff

on

06/1

1/13

For

pers

onal

use

onl

y.

ment of long-term advantages conferred by augmented earlygrowth. The larger body sizes in treatment YOY ATS documentedin this studymay improve long-term success for these populationsby enabling earlier smoltification in subsequent growth years.

Few measurable differences were detected in age 1 ATS, andsome inconsistencies were detected in YOY ATS in 2009 in onestream. Although the buffer between treatment and controlreaches potentially limited movement between the reaches, wecannot be certain that age 1 ATS did not move between treatmentand control reaches. We were unable to distinguish thermalmarks for this year class of stocked fish, so that age 1 ATS movingbetween treatment and control reaches would be indistinguish-able. Alternatively, limited differences between control and treat-ment age 1 ATSmay indicate size-selectivemortality. The YOY ATSin Stream C in 2009 may have included ATS stocked by MDMR inKingsbury Stream in 2008 that may have moved into the treat-ment reach of this stream. The treatment reach of Stream Cis <500 m from Kingsbury Stream, while all other treatmentreaches are ≥1 km from Kingsbury Stream. Movement of age 1 fishfrom Kingsbury Stream into Stream C would account for the sig-nificant size difference detected between treatment and controlreaches in that stream in 2009.

Ameasureable increase in fish size indicates faster growth ratesin treatment fish, and absolute growth rates in 2010 YOY ATSweregreater in fish from treatment reaches relative to control reaches.

We used otolith microstructure analysis to explore daily changesin juvenile salmonid growth. While otolith microstructure analy-sis requires lethal sampling, mark–recapture methods describechanges based only on the interval of time between captures.Other studies have shown elevated growth rates in coho salmon(Oncorhynchus kisutch) following carcass additions based on percentchange in mass or length or instantaneous growth rates in mark–recapture studies (Giannico and Hinch 2007; Wipfli et al. 2003).Others found no change in age 1+ juvenile salmonid growth ratesafter carcass additions based on specific growth rate measure-ments in mark–recapture designs (Cram et al. 2011; Harvey andWilzbach 2010).

Size differences in YOY ATS within the first month and growthrate divergence within the first week after nutrient addition indi-cate direct ingestion of the carcass analog material or a rapidincrease inmacroinvertebrate prey availability. The primarymac-roinvertebrate prey item for YOY ATS in these systems is Chirono-midae (Guyette 2012; Keeley and Grant 1995; Martinussen et al.2011). We found a significant increase in Chironomidae densityand biomass 1 week after nutrient addition in one study stream;however, no change was detected in until 4 weeks after additionin a second study stream (Guyette 2012). The growth rate diver-gence occurred in the first week in both streams, and these resultssuggest that carcass analog material ingestion during this firstweek led to increased growth rates in YOY ATS. Atlantic salmonconsumed nutrients delivered by the carcass analog, as evidencedby changes in tissue N and carbon (C) stable isotope values in fishcollected from treatment reaches; up to 41% of C and 25% of N insampled fish muscle tissue was of analog origin (Guyette 2012).

Percent total lipids in YOY ATS increased after carcass analogaddition. These levels remained elevated for 2 months followingthe nutrient addition and then began to decline, correspondingwith changing surface water TDN and TDP concentrations. Ourresults are consistent with responses to salmon carcass additionsin Pacific salmon streams in Washington state; juvenile cohosalmon and steelhead (Oncorhynchus mykiss) condition factor re-mained elevated while carcasses were present (Bilby et al. 1998).Treatment of Alaskan streams with carcasses and carcass analogsincreased percent total lipid by �225% in resident YOY cohosalmon and �40%–100% in resident cutthroat trout (Oncorhynchusclarkii) relative to fish in control mesocosms or natural streams(Heintz et al. 2004; Wipfli et al. 2004), comparable to the �125%increase 1month after analog addition in this study. Ingested foodenergy is allocated to metabolism, growth, or excretion, andgreater energy stores during a growth period implies that excess

Fig. 8. Percent total lipid by wet mass of Atlantic salmon in StreamC in 2010. The vertical dashed line indicates the date the carcassanalog addition occurred (1 July 2010). Data points shown aremeans ± Tukey's confidence intervals.

Fig. 7. (a) Absolute growth rate means (±1 SE) for young-of-the-yearAtlantic salmon sampled on 23 October 2010 and (b) average dailywater temperature within the stream reaches. Carcass analogaddition (vertical dashed line) occurred on 1 July 2010. Significantdifferences between control and treatment absolute growth ratebased on Wilcoxon rank sum tests are highlighted in gray. Data arepooled across all sample streams.

Guyette et al. 867

Published by NRC Research Press

Can

. J. F

ish.

Aqu

at. S

ci. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y C

SP S

taff

on

06/1

1/13

For

pers

onal

use

onl

y.

resources are available (Adams 1999). When energy is allocated toreserves, the probability of overwintering survival increases(Adams 1999; Henderson et al. 1988; Thompson et al. 1991). Theincrease in percent total lipids during the summer months, cou-pled with the greater body size, suggests that the sea lampreyspawning could increase the likelihood of success for juvenileATS.

A key advantage of the timing of sea lamprey spawning is itscoincidence with increasing water temperatures, which increasesgrowth opportunities for juvenile ATS. In western North America,steelhead spawn in the spring before water temperatures havereached their elevated summer levels, and most other Pacificsalmon species spawn in the fall, when water temperatures aredeclining. Pink salmon (Oncorhynchus gorbuscha) may spawn in thesummer; however, timing ranges from August through Novem-ber, and runs typically do not occur during peak summer temper-atures (Heard 1991). Faster growth in YOY ATS was correlated withincreasing water temperature in both control and treatmentreaches. This increased temperature and the nutrient additiontimed tomatch sea lamprey spawning resulted in a greater rate ofAGR change in treatment reaches for 7–10 days following thecarcass analog addition, demonstrating a causal relationship be-tween the presence of the carcass analog and elevated growthrates rather than simply a response to changing temperatures.

The positive effects of analog addition on ATS in this studysuggest restoration of anadromous fish may benefit ATS. Thereare, however, several issues involved in restored access for migra-tory fish. The experimental design did not incorporate other fac-tors present in natural spawning events. For example, sea lampreyspawning does not occur on a single day, and carcasses would bedeposited in the streams over an extended period (Nislow andKynard 2009). The magnitude of the nutrient delivery from a nat-ural spawning event may differ from the loading rate used in thisstudy; historical sea lamprey population sizes are unknown. Ad-ditionally, the N:P ratio of sea lamprey carcasses in other Mainestreamswith natural lamprey spawning is four times greater thanthat of the carcass analog we used in this study (Hogg 2012). Car-cass or analog supplementation that reflects the ecological stoi-chiometry of the spawning fish may affect the proportions ofN and P entering the system and subsequently the nutrient limi-tation status of the ecosystem. Nutrient amendments with the N:Pratio of sea lamprey may have a greater effect on productivity ifthe system is N-limited; however, a larger nutrient addition maybe required to attain the same effects in a P-limited system. Sub-strate restructuring during Pacific salmon redd building de-creases aquatic invertebrate densities (Honea and Gara 2009;Moore et al. 2007); similar effects from sea lamprey redd buildingmay reduce the positive effects of MDNs on juvenile ATS by reduc-ing the prey base. Examining incorporation ofMDNs into the foodwebs of a variety of stream, river, and lake ecosystems with ahistory of anadromous species at a variety of nutrient densitieswill increase our understanding of the potential effects of re-stored habitat and connectivity on juvenile anadromous fishgrowth, survival, and persistence.

AcknowledgementsFunding was provided by the National Oceanic and Atmo-

spheric Administration, the US Geological Survey, and the Depart-ment of Wildlife Ecology at the University of Maine, Orono,Maine, USA. The authors thank the landowners for their generos-ity in allowing access to their land. The manuscript was improvedwith a review provided by Keith Nislow and two anonymous re-viewers. The research was performed under University of Maineapproved IACUC Protocol No. 2008-07-01. Mention of trade namesand commercial parts does not constitute endorsement or recom-mendation for use by the US Government.

ReferencesAdams, S.M. 1999. Ecological role of lipids in the health and success of fish

populations. In Lipids in freshwater ecosystems. Edited by M.T. Arts and B.C.Wainman. Springer, New York. pp. 132–160.

APHA. 1999. Standard methods for the examination of water and wastewater.20th ed. American Public Health Association, Washington, D.C.

Bilby, R.E., Fransen, B.R., and Bisson, P.A. 1996. Incorporation of nitrogen andcarbon from spawning coho salmon into the trophic systemof small streams:evidence from stable isotopes. Can. J. Fish. Aquat. Sci. 53(1): 164–173. doi:10.1139/f95-159.

Bilby, R.E., Fransen, B.R., Bisson, P.A., andWalter, J.K. 1998. Response of juvenilecoho salmon (Oncorhynchus kisutch) and steelhead (Oncorhynchus mykiss) to theaddition of salmon carcasses to two streams in southwestern Washington,U.S.A. Can. J. Fish. Aquat. Sci. 55(8): 1909–1918. doi:10.1139/f98-094.

Brett, J.R. 1979. Environmental factors and growth. In Fish physiology: bioener-getics and growth. Edited by W.S. Hoar, D.J. Randall, and J.R. Brett. AcademicPress, London. pp. 599–675.

Campana, S.E. 1990. How reliable are growth back-calculations based on oto-liths? Can. J. Fish. Aquat. Sci. 47(11): 2219–2227. doi:10.1139/f90-246.

Chaloner, D.T., Martin, K.M.,Wipfli, M.S., Ostrom, P.H., and Lamberti, G.A. 2002.Marine carbon and nitrogen in southeastern Alaska stream food webs: evi-dence from artificial and natural streams. Can. J. Fish. Aquat. Sci. 59(8):1257–1265. doi:10.1139/f02-084.

Chaloner, D.T., Lamberti, G.A., Merritt, R.W., Mitchell, N.L., Ostrom, P.H., andWipfli, M.S. 2004. Variation in responses to spawning Pacific salmon amongthree south-eastern Alaska streams. Freshw. Biol. 49(5): 587–599. doi:10.1111/j.1365-2427.2004.01213.x.

Cram, J.M., Kiffney, P.M., Klett, R., and Edmonds, R.L. 2011. Do fall additions ofsalmon carcasses benefit foodwebs in experimental streams? Hydrobiologia,675(1): 197–209. doi:10.1007/s10750-011-0819-9.

Danielson, T.J. 2010. Assessing the biological condition of Maine streams andrivers using benthic algal communities. Doctoral dissertation, Ecology andEnvironmental Science, University of Maine, Orono, Maine.

Delia, C.F., Steudler, P.A., and Corwin, N. 1977. Determination of total nitrogenin aqueous samples using persulfate digestion. Limnol. Oceanogr. 22(4): 760–764. doi:10.4319/lo.1977.22.4.0760.

Fay, C., Bartron, M., Craig, S., Hecht, A., Pruden, J., Saunders, R., Sheehan, T., andTrial, J. 2006. Status review for anadromous Atlantic salmon (Salmo salar) inthe United States. Report to theNationalMarine Fisheries Service andUS Fishand Wildlife Service.

Flecker, A.S., McIntyre, P.B., Moore, J.W., Anderson, J.T., Taylor, B.W., and Hall,R.O., Jr. 2010. Migratory fishes as material and process subsidies in riverineecosystems. In Community ecology of stream fishes: concepts, approaches,and techniques. Edited by K.B. Gido and D.A. Jackson. American FisheriesSociety, Symposium 73, Bethesda, Md. pp. 559–592.

Folch, J., Lees,M., and Stanley, G.H.S. 1957. A simplemethod for the isolation andpurification of total lipides from animal tissues. J. Biol. Chem. 226(1): 497–509.

Fox, J., and Weisberg, S. 2011. An R companion to applied regression, 2nd ed.Sage, Thousand Oaks, Calif.

Garman, G.C., and Macko, S.A. 1998. Contribution of marine-derived organicmatter to an Atlantic coast, freshwater, tidal stream by anadromous clupeidfishes. J. N. Am. Benthol. Soc. 17(3): 277–285. doi:10.2307/1468331.

Giannico, G.R., and Hinch, S.G. 2007. Juvenile coho salmon (Oncorhynchus kisutch)responses to salmon carcasses and in-stream wood manipulations duringwinter and spring. Can. J. Fish. Aquat. Sci. 64(2): 324–335. doi:10.1139/f07-011.

Griswold, R.G., Taki, D., and Stockner, J.G. 2003. Redfish lake sockeye salmon:nutrient supplementation as ameans of restoration. InNutrients in salmonidecosystems: sustaining production and biodiversity. Edited by J.G. Stockner.pp. 197–211.

Guyette, M.Q. 2012. Responses of Atlantic salmon stream communities tomarine-derived nutrients. Doctoral dissertation, Wildlife Ecology, The Uni-versity of Maine, Orono, Maine.

Harvey, B.C., and Wilzbach, M.A. 2010. Carcass addition does not enhance juve-nile salmonid biomass, growth, or retention in six northwestern Californiastreams. N. Am. J. Fish. Manage. 30(6): 1445–1451. doi:10.1577/M10-022.1.

Heard, W.R. 1991. Life history of pink salmon (Oncorhynchus gorbuscha). In Pacificsalmon life histories. Edited by C. Groot and L. Margolis. University of BritishColumbia Press, Vancouver. pp. 121–229.

Heggenes, J., and Metcalfe, N.B. 1991. Bimodal size distributions in wild juvenileAtlantic salmon populations and their relationship with age at smolt migra-tion. J. Fish Biol. 39(6): 905–907. doi:10.1111/j.1095-8649.1991.tb04421.x.

Heintz, R.A., Nelson, B.D., Hudson, J., Larsen,M., Holland, L., andWipfli,M. 2004.Marine subsidies in freshwater: effects of salmon carcasses on lipid class andfatty acid composition of juvenile coho salmon. Trans. Am. Fish. Soc. 133(3):559–567. doi:10.1577/T03-035.1.

Henderson, P.A., Holmes, R.H.A., and Bamber, R.N. 1988. Size-selective overwin-tering mortality in the sand smelt, Atherina boyeri Risso, and its role in popu-lation regulation. J. Fish Biol. 33(2): 221–233. doi:10.1111/j.1095-8649.1988.tb05465.x.

Hogg, R. 2012. Fish community response to a small stream dam removal in acoastal Maine tributary. Master's thesis, Wildlife Ecology, The University ofMaine, Orono, Maine.

868 Can. J. Fish. Aquat. Sci. Vol. 70, 2013

Published by NRC Research Press

Can

. J. F

ish.

Aqu

at. S

ci. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y C

SP S

taff

on

06/1

1/13

For

pers

onal

use

onl

y.

Honea, J.M., and Gara, R.I. 2009. Macroinvertebrate community dynamics:strong negative response to salmon redd construction and weak response tosalmon-derived nutrient uptake. J. N. Am. Benthol. Soc. 28(1): 207–219. doi:10.1899/08-030.1.

Horton, G.E., Letcher, B.H., Bailey, M.M., and Kinnison, M.T. 2009. Atlanticsalmon (Salmo salar) smolt production: the relative importance of survivaland body growth. Can. J. Fish. Aquat. Sci. 66(3): 471–483. doi:10.1139/F09-005.

Huryn, A.D., and Wallace, J.B. 2000. Life history and production of stream in-sects. Annu. Rev. Entomol. 45: 83–110. doi:10.1146/annurev.ento.45.1.83.

Huusko, A., Greenberg, L., Stickler, M., Linnansaari, T., Nykänen, M.,Vehanen, T., Koljonen, S., Louhi, P., and Alfredsen, K. 2007. Life in the icelane: the winter ecology of stream salmonids. River Res. Appl. 23(5): 469–491.doi:10.1002/rra.999.

Johnston, N.T., MacIsaac, E.A., Tschaplinski, P.J., and Hall, K.J. 2004. Effects ofthe abundance of spawning sockeye salmon (Oncorhynchus nerka) on nutrientsand algal biomass in forested streams. Can. J. Fish. Aquat. Sci. 61(3): 384–403.doi:10.1139/f03-172.

Johnston, P., Bergeron, N.E., and Dodson, J.J. 2005. Assessment of winter size-selective mortality of young-of-the-year Atlantic salmon (Salmo salar) usingotolith microstructure analysis. Ecol. Freshw. Fish, 14(2): 168–176.

Jonsson, B., and Jonsson, N. 2003. Migratory Atlantic salmon as vectors for thetransfer of energy and nutrients between freshwater and marine environ-ments. Freshw. Biol. 48(1): 21–27. doi:10.1046/j.1365-2427.2003.00964.x.

Kane, T.R. 1988. Relationship of temperature and time of initial feeding of At-lantic salmon. Progr. Fish-Cult. 50: 93–97. doi:10.1577/1548-8640(1988)050<0093:ROTATO>2.3.CO;2.

Keeley, E.R., and Grant, J.W.A. 1995. Allometric and environmental correlates ofterritory size in juvenile Atlantic salmon (Salmo salar). Can. J. Fish. Aquat. Sci.52(1): 186–196. doi:10.1139/f95-019.

Kiernan, J.D., Harvey, B.N., and Johnson, M.L. 2010. Direct versus indirect path-ways of salmon-derived nutrient incorporation in experimental lotic foodwebs. Can. J. Fish. Aquat. Sci. 67(12): 1909–1924. doi:10.1139/F10-110.

Kohler, A.E., Rugenski, A., and Taki, D. 2008. Stream food web response to asalmon carcass analogue addition in two central Idaho, U.S.A. streams.Freshw. Biol. 53(3): 446–460. doi:10.1111/j.1365-2427.2007.01909.x.

Lang, D.W., Reeves, G.H., Hall, J.D., and Wipfli, M.S. 2006. The influence offall-spawning coho salmon (Oncorhynchus kisutch) on growth and productionof juvenile coho salmon rearing in beaver ponds on the Copper River Delta,Alaska. Can. J. Fish. Aquat. Sci. 63(4): 917–930. doi:10.1139/f05-268.

Lessard, J.L., and Merritt, R.W. 2006. Influence of marine-derived nutrients fromspawning salmon on aquatic insect communities in southeast Alaskanstreams. Oikos, 113(2): 334–343. doi:10.1111/j.2006.0030-1299.14512.x.

Lessard, J.L., Merritt, R.W., and Berg, M.B. 2009. Investigating the effect ofmarine-derived nutrients from spawning salmon on macroinvertebrate sec-ondary production in southeast Alaskan streams. J. N. Am. Benthol. Soc.28(3): 683–693. doi:10.1899/08-141.1.

Limburg, K.E., and Waldman, J.R. 2009. Dramatic declines in north Atlanticdiadromous fishes. Bioscience, 59(11): 955–965. doi:10.1525/bio.2009.59.11.7.

Lyle, A.A., and Elliott, J.M. 1998. Migratory salmonids as vectors of carbon, ni-trogen and phosphorus between marine and freshwater environments innorth-east England. Sci. Total Environ. 210(1–6): 457–468. doi:10.1016/S0048-9697(98)00031-X.

Martin, A.E.,Wipfli, M.S., and Spangler, R.E. 2010. Aquatic community responsesto salmon carcass analog and wood bundle additions in restored floodplainhabitats in an Alaskan stream. Trans. Am. Fish. Soc. 139(6): 1828–1845. doi:10.1577/T09-096.1.

Martinussen, P.A., Robertsen, G., and Einum, S. 2011. Density-dependent dietcomposition of juvenile Atlantic salmon (Salmo salar). Ecol. Freshw. Fish,20(3): 384–392.

McCormick, S.D., Hansen, L.P., Quinn, T.P., and Saunders, R.L. 1998. Movement,migration, and smolting of Atlantic salmon (Salmo salar). Can. J. Fish. Aquat.Sci. 55: 77–92. doi:10.1139/d98-011.

MDMR and MDIFW. 2007. Strategic plan for the restoration of diadromous andresident fishes to the Penobscot River. Maine Department of Marine Re-sources and Maine Department of Inland Fisheries and Wildlife.

Meekan, M.G., Dodson, J.J., Good, S.P., and Ryan, D.A.J. 1998. Otolith and fish sizerelationships, measurement error, and size-selective mortality during theearly life of Atlantic salmon (Salmo salar). Can. J. Fish. Aquat. Sci. 55(7): 1663–1673. doi:10.1139/f98-047.

Michael, J.H. 1995. Enhancement effects of spawning pink salmon on streamrearing juvenile coho salmon: managing one resource to benefit another.Northw. Sci. 69(3): 228–233.

Moore, J.W., Schindler, D.E., Carter, J.L., Fox, J., Griffiths, J., and Holtgrieve, G.W.2007. Biotic control of stream fluxes: spawning salmon drive nutrient andmatter export. Ecology, 88(5): 1278–1291. doi:10.1890/06-0782.

Morin, A., Lamoureux, W., and Busnarda, J. 1999. Empirical models predictingprimary productivity from chlorophyll a and water temperature for streamperiphyton and lake and ocean phytoplankton. J. N. Am. Benthol. Soc. 18(3):299–307. doi:10.2307/1468446.

Moring, J. 2005. Recent trends in anadromous fishes. In The decline of fisheriesresources in New England: evaluating the impact of overfishing, contamina-tion, and habitat degradation. Edited by R. Buchsbaum, J. Pederson, and W.E.Robinson. MIT Sea Grant College Program, Cambridge, Mass. Publ. No. 05–5.

Muggeo, V.M.R. 2003. Estimating regression models with unknown break-points. Statist. Med. 22: 3055–3071. doi:10.1002/sim.1545.

Muggeo, V.M.R. 2008. segmented: an R package to fit regression models withbroken-line relationships [online]. In R News, 8/1, 20–25. Available fromhttp://cran.r-project.org/doc/Rnews/.

Murphy, M.H., Connerton, M.J., and Stewart, D.J. 2006. Evaluation of winterseverity on growth of young-of-the-year Atlantic salmon. Trans. Am. Fish.Soc. 135(2): 420–430. doi:10.1577/T04-109.1.

Nicieza, A.G., Reyes-Gavilán, F.G., and Braña, F. 1994. Differentiation in juvenilegrowth and bimodality patterns between northern and southern popula-tions of Atlantic salmon (Salmo salar L.). Can. J. Zool. 72(9): 1603–1610. doi:10.1139/z94-213.

Nislow, K.H., and Kynard, B.E. 2009. The role of anadromous sea lamprey innutrient and material transport between marine and freshwater environ-ments. In Challenges for diadromous fishes in a dynamic global environ-ment. Edited by A.J. Haro, K.L. Smith, R.A. Rulifson, C.M. Moffitt, R.J. Klauda,M.J. Dadswell, R.A. Cunjak, J.E. Cooper, K.L. Beal, and T.S. Avery. AmericanFisheries Society, Symposium 69, Bethesda, Md. pp. 485–494.

Oregon Watershed Enhancement Board. 2006. Oregon plan biennial report2005- 2007, Salem, Oregon.

Pearsons, T.N., Roley, D.D., and Johnson, C.L. 2007. Development of a carcassanalog for nutrient restoration in streams. Fisheries, 32(3): 114–124. doi:10.1577/1548-8446(2007)32[114:DOACAF]2.0.CO;2.

Peterson, B.J., Deegan, L., Helfrich, J., Hobbie, J.E., Hullar, M., Moller, B.,Ford, T.E., Hershey, A., Hiltner, A., Kipphut, G., Lock, M.A., Fiebig, D.M.,McKinley, V., Miller, M.C., Vestal, J.R., Ventullo, R., and Volk, G. 1993. Biolog-ical responses of a Tundra river to fertilization. Ecology, 74(3): 653–672.doi:10.2307/1940794.

R Development Core Team. 2012. R: a language and environment for statisticalcomputing. R Foundation for Statistical Computing, Vienna, Austria.

Saunders, R., Hachey, M.A., and Fay, C.W. 2006. Maine's diadromous fish com-munity: past, present, and implications for Atlantic salmon recovery. Fisher-ies, 31(11): 537–547. doi:10.1577/1548-8446(2006)31[537:MDFC]2.0.CO;2.

Sirois, P., Lecomte, F., and Dodson, J.J. 1998. An otolith-based back-calculationmethod to account for time-varying growth rate in rainbow smelt (Osmerusmordax) larvae. Can. J. Fish. Aquat. Sci. 55(12): 2662–2671. doi:10.1139/f98-170.

Slaney, P.A., Ward, B.R., and Wightman, J.C. 2003. Experimental nutrient addi-tion to the Keogh River and application to the Salmon River in coastal BritishColumbia. In Nutrients in salmonid ecosystems: sustaining production andbiodiversity. Edited by J.G. Stockner. American Fisheries Society Symposium34, Bethesda, Md. pp. 111–126.

Stockner, J.G., and MacIsaac, E.A. 1996. British Columbia lake enrichment pro-gramme: two decades of habitat enhancement for sockeye salmon. Regul.Rivers Res. Manage. 12(4-5): 547–561. doi:10.1002/(SICI)1099-1646(199607)12:4/5<547::AID-RRR407>3.0.CO;2-M.

ter Steege, H. 1996.Winphot version 5.0 (copyright 12/2/1996). Tropenbos GuyanaProgramme, Tropenbos, Guyana.

Thompson, J.M., Bergersen, E.P., Carlson, C.A., and Kaeding, L.R. 1991. Role ofsize, condition, and lipid content in the overwinter survival of age-0 Coloradosquawfish. Trans. Am. Fish. Soc. 120(3): 346–353. doi:10.1577/1548-8659(1991)120<0346:ROSCAL>2.3.CO;2.

Thorpe, J.E. 1977. Bimodal distribution of length of juvenile Atlantic salmon(Salmo salar L.) under artificial rearing conditions. J. Fish Biol. 11(2): 175–184.doi:10.1111/j.1095-8649.1977.tb04111.x.

Thorpe, J.E. 1984. Downstream movements of juvenile salmonids: a forwardspeculative view. In Mechanisms of fish migration. Edited by J.D. McCleave,G.P. Arnold, J.J. Dodson, and W.H. Neill. Plenum Press, New York. pp.387–396.

Thorpe, J.E., Mangel, M., Metcalfe, N.B., and Huntingford, F.A. 1998. Modellingthe proximate basis of salmonid life-history variation, with application toAtlantic salmon, Salmo salar L. Evol. Ecol. 12(5): 581–599. doi:10.1023/A:1022351814644.

Utrilla, C.G., and Lobon-Cervia, J. 1999. Life-history patterns in a southern pop-ulation of Atlantic salmon. J. Fish Biol. 55(1): 68–83. doi:10.1111/j.1095-8649.1999.tb00657.x.

Venables, W.N., and Ripley, B.D. 2002. Modern applied statistics with S. 4th ed.Springer, New York.

Walters, A.W., Barnes, R.T., and Post, D.M. 2009. Anadromous alewives (Alosapseudoharengus) contribute marine-derived nutrients to coastal stream foodwebs. Can. J. Fish. Aquat. Sci. 66(3): 439–448. doi:10.1139/F09-008.

Weng, Z., Mookerji, N., and Mazumder, A. 2001. Nutrient-dependent recovery ofAtlantic salmon streams from a catastrophic flood. Can. J. Fish. Aquat. Sci.58(8): 1672–1682. doi:10.1139/f01-107.

Williams, K.L., Griffiths, S.W., Nislow, K.H., McKelvey, S., and Armstrong, J.D.2009. Response of juvenile Atlantic salmon, Salmo salar, to the introduction ofsalmon carcasses in upland streams. Fish. Manage. Ecol. 16(4): 290–297. doi:10.1111/j.1365-2400.2009.00673.x.

Wipfli, M.S., Hudson, J., and Caouette, J. 1998. Influence of salmon carcasses onstream productivity: response of biofilm and benthic macroinvertebrates insoutheastern Alaska, U.S.A. Can. J. Fish. Aquat. Sci. 55(6): 1503–1511. doi:10.1139/f98-031.

Wipfli, M.S., Hudson, J.P., Chaloner, D.T., and Caouette, J R. 1999. Influence of

Guyette et al. 869

Published by NRC Research Press

Can

. J. F

ish.

Aqu

at. S

ci. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y C

SP S

taff

on

06/1

1/13

For

pers

onal

use

onl

y.

salmon spawner densities on stream productivity in southeast Alaska. Can. J.Fish. Aquat. Sci. 56(9): 1600–1611. doi:10.1139/f99-087.

Wipfli, M.S., Hudson, J.P., Caouette, J.P., and Chaloner, D.T. 2003. Marine subsi-dies in freshwater ecosystems: salmon carcasses increase the growth rates ofstream-resident salmonids. Trans. Am. Fish. Soc. 132(2): 371–381. doi:10.1577/1548-8659(2003)132<0371:MSIFES>2.0.CO;2.

Wipfli, M.S., Hudson, J.P., and Caouette, J.P. 2004. Restoring productivity of

salmon-based food webs: contrasting effects of salmon carcass and salmoncarcass analog additions on stream-resident salmonids. Trans. Am. Fish. Soc.133(6): 1440–1454. doi:10.1577/T03-112.1.

Wipfli, M.S., Hudson, J.P., Caouette, J.P., Mitchell, N.L., Lessard, J.L., Heintz, R.A.,and Chaloner, D.T. 2010. Salmon carcasses increase streamproductivitymorethan inorganic fertilizer pellets: a test on multiple trophic levels in stream-side experimental channels. Trans.Am. Fish. Soc. 139(3): 824–839. doi:10.1577/T09-114.1.

870 Can. J. Fish. Aquat. Sci. Vol. 70, 2013

Published by NRC Research Press

Can

. J. F

ish.

Aqu

at. S

ci. D

ownl

oade

d fr

om w

ww

.nrc

rese

arch

pres

s.co

m b

y C

SP S

taff

on

06/1

1/13

For

pers

onal

use

onl

y.