Ecology, 94(2), 2013, pp. 356–367� 2013 by the Ecological Society of America

Limitation and facilitation of one of the world’s most invasive fish:an intercontinental comparison

PHAEDRA BUDY,1,9 GARY P. THIEDE,2 JAVIER LOBON-CERVIA,3 GUSTAVO GONZALEZ FERNANDEZ,4 PETER MCHUGH,5

ANGUS MCINTOSH,6 LIEF ASBJØRN VØLLESTAD,7 ELOY BECARES,8 AND PHILLIP JELLYMAN6

1U.S. Geological Survey, Utah Cooperative Fish and Wildlife Research Unit, Department of Watershed Sciences,Utah State University, Logan, Utah 84322-5210 USA

2Department of Watershed Sciences, Utah State University, Logan, Utah 84322-5210 USA3National Museum of Natural Sciences (CSIC), C/ Jose Gutierrez Abascal 2, Madrid 28006 Spain

4Icthios Environmental, Pablo Ruiz Picasso 38, Leon 24009 Spain5Oregon Department of Fish and Wildlife, 17330 Southeast Evelyn Street, Clackamas, Oregon 97015 USA6School of Biological Sciences, University of Canterbury, Private Bag 4800, Christchurch, New Zealand

7Center for Ecological and Evolutionary Synthesis, Department of Biology, University of Oslo, 0316 Oslo, Norway8Department of Ecology, Genetics, and Microbiology, Faculty of Biology, University of Leon, Leon 24071 Spain

Abstract. Purposeful species introductions offer opportunities to inform our understand-ing of both invasion success and conservation hurdles. We evaluated factors determining theenergetic limitations of brown trout (Salmo trutta) in both their native and introduced ranges.Our focus was on brown trout because they are nearly globally distributed, considered one ofthe world’s worst invaders, yet imperiled in much of their native habitat. We synthesized andcompared data describing temperature regime, diet, growth, and maximum body size acrossmultiple spatial and temporal scales, from country (both exotic and native habitats) and majorgeographic area (MGA) to rivers and years within MGA. Using these data as inputs, we nextused bioenergetic efficiency (BioEff ), a relative scalar representing a realized percentage ofmaximum possible consumption (0–100%) as our primary response variable and a multi-scale,nested, mixed statistical model (GLIMMIX) to evaluate variation among and within spatialscales and as a function of density and elevation. MGA and year (the residual) explained thegreatest proportion of variance in BioEff. Temperature varied widely among MGA and was astrong driver of variation in BioEff. We observed surprisingly little variation in the diet ofbrown trout, except the overwhelming influence of the switch to piscivory observed only inexotic MGA. We observed only a weak signal of density-dependent effects on BioEff;however, BioEff remained ,50% at densities .2.5 fish/m2. The trajectory of BioEff across thelife span of the fish elucidated the substantial variation in performance among MGAs; themaximum body size attained by brown trout was consistently below 400 mm in native habitatbut reached ;600 mm outside their native range, where brown trout grew rapidly, feeding inpart on naive prey fishes. The integrative, physiological approach, in combination with theintercontinental and comparative nature of our study, allowed us to overcome challengesassociated with context-dependent variation in determining invasion success. Overall ourresults indicate ‘‘growth plasticity across the life span’’ was important for facilitating invasion,and should be added to lists of factors characterizing successful invaders.

Key words: bioenergetics; brown trout; conservation; diet; distribution; exotic; fish growth; invasion;life history; metabolism; piscivory; Salmo trutta.

INTRODUCTION

The hierarchy of factors determining the natural

distribution of animals continues to fascinate ecologists;

however, globalization and a diversity of contemporary

anthropogenic activities have altered these filters at

multiple scales and in ways difficult to predict (e.g.,

Dukes and Mooney 1999, Brook et al. 2008). Nonethe-

less these global changes make the fundamental

questions about factors that determine the distribution

of an animal important, both in trying to conserve and

restore native biota, as well as to understand, minimize,

and prevent the spread or distribution of invasive exotic

species (Kolar and Lodge 2001).

While most invasions fail, successful invasions often

have devastating effects on plant and animal communi-

ties and result in significant economic costs in some cases

(e.g., Pimentel et al. 2000). As such, successful invasions

have generally been the focus of research on invasive

species (Perrings 2002, Panov et al. 2011). Consequently,

we have made considerable advances in identifying the

suite of characteristics that facilitate the invasion process

for fishes, focusing both on characteristics of the

Manuscript received 17 April 2012; revised 18 July 2012;accepted 17 September 2012. Corresponding Editor: D. E.Schindler.

9 E-mail: phaedra.budy@usu.edu

356

successful invader (e.g., Marchetti et al. 2004, Valiente et

al. 2010), and characteristics of the successfully invaded

habitat (Shea and Chesson 2002, Westley and Fleming

2011).

Despite these advances in understanding invasion

success, there are still some important information gaps

in our understanding of the invasion process (reviewed

in Mack et al. 2000). Much invasion ecology has been

limited to large-scale occupancy (e.g., presence–absence

evaluations), and ecological niche modeling studies,

theoretical modeling, or small-scale species interaction

studies (Richardson 2011). While occupancy and niche

modeling are often feasible from a data perspective and

broadly applicable, they are rarely mechanistic and can

fail to detect important local patterns and interspecific

interactions (e.g., Bruno et al. 2005, Stohlgren et al.

2006). Theoretical models of the invasion process are

stimulating and critical, but serve only to provide

theories that must be tested and revised based on

empirical data and natural field studies (e.g., Shea and

Chesson 2002). Last, while small-scale species interac-

tion studies may provide mechanistic insights about the

invasion process (e.g., McHugh and Budy 2005), the

results may be difficult to apply at larger regional and

global scales relevant to invasion ecology (Dukes and

Mooney 1999, Korsu et al. 2010; but see Carey and

Wahl 2010).

There remains a need for global comparisons of

invader performance in their native habitat vs. their

exotic habitat using mechanistic approaches such as

functional relationships (e.g., Bollache et al. 2008) or

metrics of performance that are clear indicators of

organism fitness (e.g., growth, survival; Frank et al.

2011). Developing a better mechanistic understanding of

invader performance may be most feasible for ecto-

therms such as fishes because temperature and food are

the two primary drivers of fish growth and are

reasonable reflections of their ultimate fitness. The net

metabolic efficiency that results from the interaction of

food availability and temperature presents a potential

fitness-based metric, similar to ‘‘local determinism’’

(Ricklefs 2004). High growth and consumption rates,

for example, are often associated with high fitness in fish

and are thus expected in the case of a successful invader.

Thus metabolic efficiency, as estimated using bioener-

getics, diet, and temperature, can aid in understanding

differences in performance between an organism in their

native vs. exotic habitat.

Our overall goal was to understand the factors that

determine the energetic limitations of brown trout both

within and across their natural, or native, and exotic

distributions. Bioenergetic efficiency (BioEff ) is a scalar

that adjusts theoretical estimates of maximum consump-

tion (gram per gram per day) to fish growth observed in

the field. Bioenergetic efficiency has been used as a

integrative measure of overall fish performance to

answer a variety of different ecological questions across

a variety of different systems (e.g., Johnson et al. 2008,

Armstrong and Schindler 2011, Budy et al. 2011). We

focused on brown trout because they are: (1) nearly

globally distributed and considered in the top 30

(McIntosh et al. 2011) worst invaders worldwide

(Appendix A); (2) economically valuable sport fish and

relatively well documented; (3) growing in distribution

in their exotic habitat, yet in many areas of their native

habitat, they are imperiled or declining (e.g., Spain,

Almodovar and Nicola [2004]; France, Keith and

Allardi [1996]); and (4) representative of many of their

invasive and imperiled trout relatives. As such, our

results can be broadly applied (e.g., IUCN 2000, Clout

and Williams 2009).

First, we synthesized and compared data describing

temperature regime, diet, growth, and maximum body

size among multiple spatial and temporal scales, from

country or continent (both exotic and native habitats),

to major geographic area (MGA) within a country.

Next, as an integrated response metric capturing the

interaction between diet, temperature, and allometry, we

estimated bioenergetic efficiency (see Methods: Statisti-

cal analysis) among MGAs, as river within a MGA, and

then among years and age classes within a river. Last, we

used bioenergetic efficiency to explore the effect of

density dependence and variation in local environmental

conditions on the bioenergetic efficiency of brown trout.

We use the integrated metric of bioenergetic efficiency to

better understand the factors facilitating the global

invasion and limiting the conservation and recovery of

brown trout within their native range.

METHODS

Our study was focused on fluvial (riverine) brown

trout, a trout indigenous to Europe, North Africa, and

western Asia but now introduced and present globally

(McIntosh et al. 2011). More detailed information

describing (1) the general biology and life history of

brown trout is available in Klemetsen et al. (2003) and

Jonsson and Jonsson (2011), and (2) their contemporary

distribution and role as widespread invader worldwide

in McIntosh et al. (2011) and Appendix B.

Among and within several countries, we acquired

best-available field measurements of stream tempera-

ture, diet composition, and fish growth, and used these

to predict what we term ‘‘bioenergetic efficiency’’

(BioEff ) of brown trout in their native and exotic range.

The approach we used to quantify BioEff was based on

the balanced energy budget of a poikilothermic organ-

ism (Hartman and Kitchell 2008). In combination with

observations of temperature, diet, and growth, we use

BioEff as our primary response variable. The BioEff is a

scalar (elsewhere termed ‘‘p value,’’ or proportion of

maximum consumption) that adjusts theoretical esti-

mates of maximum consumption (gram per gram per

day) to fish growth observed in the field. The maximum

estimates of consumption are based on laboratory

relationships between consumption and temperature

and allometry, which are then scaled to observed growth

February 2013 357ENERGETIC LIMITATION OF GLOBAL INVADERS

and field observations of stream temperature, diet

composition, and prey energy content (i.e., realized

maximum possible consumption rate). We refer to the p

value herein as BioEff to avoid confusion with statistical

P values and because the term BioEff better captures the

physiological relationship of this variable to fish

performance overall.

Here we express BioEff as a percentage with a

theoretical range of 0% to 100%. A BioEff near 100%indicates fish are feeding near their maximum possible

consumption rate, given the environmental temperature,

diet, and their body size, while BioEff near 0% indicates

fish are feeding at a rate far below their theoretical

maximum consumption rate and performing poorly.

Bioenergetic efficiencies are relative and integrate across

the entire energy budget of fishes (i.e., assess how

consumed energy is allocated to metabolism, wastes, and

growth), thereby enabling comparison among popula-

tions or streams. High BioEff values often indicate rapid

growth rates were observed, but not always. If food

availability is high relative to the temperature regime,

BioEff will also be high in the absence of other limiting

factors. Conversely, a fish can perform well (i.e.,

relatively high BioEff ) in a low food environment if

temperatures are less than optimal. The relative

strengths and weaknesses of bioenergetics as a tool are

well documented and discussed elsewhere (e.g., reviewed

in Hartman and Kitchell 2008), and the accuracy of the

brown trout model specifically has been fully evaluated

elsewhere (Whitledge et al. 2010).

Data set and bioenergetics analyses

The data we could use were restricted by the

availability of the three primary inputs for our analysis:

daily stream temperature, fish diet composition, and fish

growth (Appendix B). We a priori restricted our

bioenergetics analysis of brown trout to entirely fluvial

populations to avoid changes in BioEff associated with

movement into a different habitat type (i.e., lake, ocean)

and changes in prey availability and type. In addition, we

chose to model only streams characterized as being in

‘‘good’’ to nearly pristine physical habitat condition and

as having no other, or very low densities of, exotic fishes.

Thus selected streams should experience little thermal or

hydrologic alteration and have a prey base generally

native to, and representative of, that area.We did include

one tailwater fishery (Arkansas; U.S. South), as in the

United States brown trout appear to prosper in these

altered, but not necessarily degraded systems (Hudy

1990). We made these restrictions based on data

availability, and given the availability of data, to try to

isolate the relative performance based exclusively on

native vs. exotic habitat. Nonetheless, the streams and

populations of brown trout we modeled included an

extremely wide range of biological and physical charac-

teristics including elevation, water temperature, hydro-

logic regime, productivity, prey diversity, and predator

density (see Appendix A for a complete summary of

model runs and details of bioenergetic inputs and

simulations). All populations represented in this analysis

experience substantial angling pressure. In sum, we

modeled 305 different brown trout ‘‘histories,’’ where

history typically represented the average growth rate and

diet composition for each age class within each stream.

Statistical analysis

Using BioEff as our primary response variable, we

used a multi-scale, nested, mixed modeling approach

with both random and fixed effects to assess the relative

contribution of different sources of variation in BioEff

and to test the effects and interactions among additional

abiotic and biotic factors. Although comprehensive, our

international data set is asymmetric. We did not have

records for all combinations of age classes and diet,

consequently, we separately analyzed two subsets of the

full data set, to allow for full factorial analyses.

In the first analysis (Model I), we used a general linear

mixed model to partition the variance and to assess

differences in BioEff values associated with density of

brown trout, elevation, and all age classes and for which

sample size was greater than n¼ 3 (‘‘all age classes’’; n¼258). Brown trout density (on the logarithmic scale) and

elevation were incorporated as continuous-scale fixed-

effects factors with linear relationships to BioEff. Age

class (with seven levels) was a categorical fixed-effects

factor. Total variance was partitioned into variance

among countries, variance among rivers within coun-

tries, and variance among years within rivers (residual

error).

In the second analysis (Model II), we used the same

statistical model, but also including whether fish were

identified in stomach contents and its interaction with

age class as fixed-effects factors. Data analyzed by

Model II were limited to brown trout age classes that are

able to consume fish (no age-0 fish) and for which sample

size was greater than n ¼ 3 (‘‘age classes with the

potential to have fish in the diet’’; n ¼ 222). We used

Model I as the basis for most comparisons as it was the

most inclusive and relied on Model II only for evaluating

the effect of piscivory and the associated interaction

between age class and piscivory. Where appropriate,

pairwise comparisons among age class means were

adjusted for family-wise Type I error rate using a step-

down simulation-based approach (Westfall et al. 1999).

Computations were made using the GLIMMIX proce-

dure in SAS/STAT Software for Windows version 9.2

using a normal distribution and an identity link.

Assumptions of normality, homogeneity of variance,

and linearity were assessed using graphical examinations

of residuals. Given the asymmetry and variability in the

data, we generally assessed significance at a ¼ 0.10 but

provide all raw P values in the results.

RESULTS

Across the summer season when most fish growth

occurs, temperature varied widely among MGAs (Fig.

PHAEDRA BUDY ET AL.358 Ecology, Vol. 94, No. 2

1), and was an inherently strong driver of variation in

BioEff (i.e., as a primary model input). In Denmark and

Norway (i.e., native habitat), for example, mean daily

temperatures were nearly always below the optimal

temperature range for brown trout growth (13–178C;

e.g., Ojanguren et al. [2001]). In New Zealand (i.e.,

exotic habitat), mean daily temperature remained far

below the optimal temperature for growth. In contrast,

in both MGAs of Spain (i.e., native habitat), mean daily

temperatures were nearly always within the optimal

range. In the United States mean daily temperature

regimes varied among MGA, remaining below optimal

in the U.S. West and U.S. South and within optimal

most of the season in the U.S. Midwest.

In contrast to stream temperature, we observed

surprisingly little variation in the diet of brown trout

FIG. 1. Average temperatures during the summer growth period of brown trout (Salmo trutta) for each major geographic area.Vertical bars represent maximum and minimum temperatures where available. As a consequence of small range, bars are hidden bysymbols for Denmark. Shaded bands indicate the optimal range for brown trout growth, 13–178C. ‘‘Summer’’ growth period (June–October) was adjusted for the New Zealand summer growth period in the southern hemisphere (December–April). ‘‘C-L’’ standsfor Castilla-Leon.

February 2013 359ENERGETIC LIMITATION OF GLOBAL INVADERS

excluding the overwhelming influence of the switch to

piscivory in some MGA (Fig. 2). Taxonomic diversity of

invertebrate prey varied little among MGA (Fig. 2a).

The diversity of vertebrate (fish) prey potentially

available varied from a low of two species in Spain to

a high of six species in the U.S. Midwest; however,

potential forage fish were present in all MGAs. Diets of

brown trout in native habitats (i.e., Spain, Norway,

Denmark) were dominated by invertebrates; diets of

brown trout in exotic regions (i.e., United States, New

Zealand) were also dominated by aquatic invertebrates,

but we also observed substantial variation in the

proportion of the diet that included fish (Fig. 2b).

Notably, fish were present in the diets only in exotic

habitats, where they made up, on average, 22% of the

diet (mass) among age classes.

Bioenergetic efficiency varied among MGAs as a

function of differences in the interaction between diet

and temperature (Fig. 3a). Mean BioEff by MGA was

similar and lowest in the U.S. South and Denmark and

greatest in the U.S. West and Spain Castilla-Leon (C-L).

Notably the second and third highest observations of

mean BioEff in our data set occurred in exotic habitat in

the United States, and BioEff was most variable in the

U.S. South and in New Zealand (also exotic habitat).

In Model I (Table 1; ‘‘all age classes; no fish diet’’),

while ‘‘year’’ (residual) and ‘‘MGA’’ explained the

greatest variation, a substantial portion was also

explained by ‘‘river’’ (within MGA). Effects of all

individual age classes less than age 4 were highly

significant (P , 0.04), whereas the effects of age classes

age 4 and above were not significant, likely due to high

variability and smaller sample size. The relationship

between BioEff and ‘‘brown trout density’’ (fish/m2; all

age classes combined) demonstrated a weak, negative

slope (log scale; Table 1; P ¼ 0.10). Bioenergetic

efficiency ranged widely across brown trout densities

ranging 0.1–1.5 fish/m2, but remained ,50% at densities

.2.5 fish/m2. We observed no relationship between

BioEff and ‘‘elevation’’ (Table 1; Model I; P . 0.3). In

Model II (Table 1; age classes with the potential to have

fish in the diet), MGA explained a greater portion of the

variance overall, with less and near equal variation

explained by river and accounted for by year. While

‘‘fish in the diet’’ alone was only significant at P , 0.10,

fixed effects tests of ‘‘age class’’ and the interaction

between fish in the diet (yes or no) and age class were

highly significant (P , 0.001).

Thus when we compared BioEff across all brown

trout that did or did not include fish in the diet, the

mean BioEff was only slightly higher for those that

demonstrated piscivory, but the variation was quite

large (Fig. 3b; Table 1, Model II; P . 0.08). This

variation was not surprising and likely driven by natural

diet variation among individual fish and size/age

influences on diet (Vinson and Budy 2011). However,

although variability in BioEff was high among all age

classes, the highest individual estimates of BioEff were

with piscivory, and there was a strong and significant

interaction between the effect of fish in the diet and age

class (Fig. 2c; Table 1, Model II; P , 0.001). This

interactive effect resulted in increasing BioEff of

piscivores after they reach age 5 (see also Fig. 3c), a

switch that occurred only outside their native habitat.

The maximum body size attained by brown trout

varied by several magnitudes among MGAs (Fig. 4a).

Maximum body sizes were lowest in Denmark, Norway,

and Spain-Asturias (i.e., native habitats) and were

consistently ,300 mm; fish generally lived only up to

age 3. Even in Norway, where fish lived up to age 6,

maximum body size was ,300 mm. Brown trout

reached slightly greater maximum body sizes, ;400

mm, in Spain C-L and U.S. Midwest. In contrast, in

U.S. West, New Zealand, and U.S. South (i.e., exotic

habitats), brown trout reached maximum body sizes

;600 mm, corresponding to approximately age-5 and

older fish (for those cases where true ages were known).

Following the strong metabolic effect of allometry at

earlier developmental stages, BioEff consistently de-

FIG. 2. (a) Prey diversity (number of taxa) of aquaticmacroinvertebrates (‘‘aqua’’ by order) and ‘‘fish’’ (by species)found in streams among different major geographic areas(MGA). (b) Brown trout diets percentage diet composition bymass among different major geographic areas (MGA): ‘‘aqua’’refers to aquatic macroinvertebrates, ‘‘terr’’ are terrestrialinvertebrates, and ‘‘fish’’ is fish. Note break in y-axis scale.

PHAEDRA BUDY ET AL.360 Ecology, Vol. 94, No. 2

creased as a function of increasing age (and size) until

fish reached age 4 (Fig. 4b, c). However, at age 4 and

above, both the mean BioEff and variability in BioEff

began to increase. Thus BioEff was extremely variable

for the largest age and size classes (.age 4) primarily

due to whether or not individual brown trout consumed

fish.

The trajectory of BioEff across the life history of an

average brown trout explained, in part, both the

substantial variation in maximum body size achieved

among MGAs and the increase in the mean and

variability in BioEff at age 4 and older (Fig. 4c). In

Spain-Asturias and Norway (i.e., native habitats), for

example, the BioEff of brown trout was high at age 0,

then dropped substantially at age 1 and remained low

from that point on. In contrast, in Denmark and Spain

C-L (i.e., native habitats), BioEff varied little among age

classes, was always low in Denmark, and always high in

Spain C-L; fluvial fish did not reach sizes .400 mm in

either case. In Spain C-L, BioEff was relatively high and

decreased only slightly by age 3, yet fluvial fish still did

not reach sizes .400 mm. Most interesting, however,

was the U.S. West, where brown trout demonstrated the

same pattern of decreasing BioEff with age class up to

age 3, BioEff remaining stable and low for 1–2 years,

then increasing substantially between age 4 and age 5,

FIG. 3. (a) Bioenergetic efficiency averaged among the different major geographic areas. Error bars indicate 2 SE. Black squaresare native habitats, and gray circles are exotic habitats. (b) Box plots of bioenergetic efficiency as a function of inclusion of fish inthe diet (‘‘piscivory’’) or ‘‘no fish’’ in the diet for age-3 and older brown trout. The box represents the 25th through 75th percentiles,the solid line in the box represents the median, the dashed line represents the mean, error bars represent the 10th and 90thpercentiles, and the dots represent the 5th and 95th percentiles. (c) Least-square means of bioenergetic efficiency among five ageclasses for brown trout that are piscivorous (gray circles, exotic habitats) and brown trout that do not consume fish (black squares,native habitat), all major geographic areas combined. Error bars represent 95% confidence intervals.

February 2013 361ENERGETIC LIMITATION OF GLOBAL INVADERS

and diets included .35% fish (by mass). Note,

unfortunately this life-history level of evaluation was

not available for all MGAs.

DISCUSSION

We assembled a comprehensive, international data set

to assess factors that limit and facilitate a fish that is

simultaneously socioeconomically and ecologically im-

portant and imperiled in its native range, yet also one of

the world’s most successful invaders. Although our data

set does not include all the variation in performance of

brown trout within or among our selected countries and

MGAs, the streams and populations of brown trout we

evaluated comprise a wide range of variability in

biological and physical characteristics. Despite this

variability, we observed an intriguing and generally

consistent difference in the performance of fluvial brown

trout in most countries of their native habitat compared

to their exotic habitat. While prey fishes were present in

all MGAs, brown trout switched to piscivory at older

age classes only in their exotic habitats, frequently

exhibiting high BioEff. Further, exotic brown trout grew

to maximum body sizes .400 mm, body sizes and ages

not observed in their native habitats.

We observed the greatest amount of variation in

BioEff among MGAs, a somewhat coarse and arbitrary

categorization from a biological standpoint, but one

that has strong links to temperature and likely

productivity (or food availability) at this large spatial

scale (e.g., Stohlgren 2002). Notably, MGA also acts as

a surrogate category for native vs. exotic, while still

allowing analysis at a slightly finer scale of resolution

(e.g., among rivers within MGAs) when the data

permitted. The overall pattern of BioEff among MGAs

appears to be driven first, by the interaction between

food availability and temperature, and second, by

whether or not brown trout switch to a diet including

piscivory. For example, in the relatively cold native

habitat of Denmark, BioEff was consistently low.

Metabolically, water temperatures that remain consis-

tently below those optimal for consumption limit the

ability of fishes to capitalize on abundant resources

TABLE 1. Model I and Model II summary statistics for brown trout (Salmo trutta) age classes intheir native and introduced ranges.

Parameters and effects Estimate SE df t F P

Model I: all age classes

Covariance

MGA 0.005 0.004River 0.003 0.001Residual (year) 0.005 0.001

Fixed effects

Brown trout density log scale �0.010 0.006 6 �1.63 0.104Elevation 0.001 0.001 207 0.98 0.330Age class, age 0 0.240 0.053 207 4.56 ,0.001Age class, age 1 0.169 0.053 207 3.22 0.002Age class, age 2 0.129 0.053 207 2.45 0.015Age class, age 3 0.112 0.054 207 2.09 0.038Age class, age 4 0.027 0.056 207 0.48 0.629Age class, age 5 �0.019 0.063 207 �0.31 0.758Age class, age 6 0 � � � � � � � � � � � �

Type III tests off fixed effects

Brown trout density log scale 1, 207 2.67 0.104Elevation 1, 207 0.95 0.330Age class 6, 207 25.72 ,0.001

Model II: age classes with the potential tohave fish in the diet

Covariances

MGA 0.005 0.004River 0.003 0.001Residual (year) 0.003 0.001

Type III tests of fixed effects

Brown trout density log scale 1, 165 4.33 0.039Elevation 1, 165 3.22 0.075Age class 4, 165 9.19 ,0.001Fish in diet 1, 165 3.147 0.077Fish in diet 3 age class 4, 165 11.12 ,0.001

Notes: Model I summary statistics include covariance parameter estimates, parameter estimatesfor fixed effects, and Type III tests of fixed effects, from multi-scale, mixed model analysis using allage classes and with sample size greater than n¼ 3 (n¼ 253). Model II summary statistics includecovariance parameter estimates and Type III tests of fixed effects, from multi-scale, mixed-modelanalysis using only data for brown trout age classes with potential to include fish in their diet (noage-0 fish) and with sample size greater than n ¼ 3 (n ¼ 222).

PHAEDRA BUDY ET AL.362 Ecology, Vol. 94, No. 2

(Hanson et al. 1997). However, many streams in these

northern regions are not only relatively cold on average,

but also low in productivity (Vøllestad et al. 2002) such

that the interaction between temperature and low food

availability co-limit brown trout growth and perfor-

mance. In contrast, in the MGA of Spain C-L,

temperatures are also within the optimal range and

BioEff is relatively high (maximum of 71%), indicating

that food is more available.

In the remaining regions of predominantly exotic

habitat for brown trout, the pattern of variation in

BioEff among MGAs is more complex. For example, in

U.S. West streams, water temperatures were nearly

always far below the optimal temperature for consump-

tion and growth, yet despite these low temperatures,

BioEff was relatively high on average. There are several

co-related factors that can explain this pattern. First, the

scale of our analysis may be too coarse to detect the

effects of some important local physical habitat charac-

teristics. For example, published optimal temperatures

for consumption and growth may be imprecise and do

not account for local thermal adaptation (Jensen et al.

FIG. 4. (a) Maximum body size attained by brown trout in different major geographic areas, MGA. Black squares are nativehabitat, and gray circles are exotic habitat. Downward bars indicate range of maximum body size among streams within a majorgeographic area. (b) Box plots of bioenergetic efficiency among different age classes of brown trout in all MGAs combined. Thebox represents the 25th through 75th percentiles, the solid line in the box represents median, the dashed line represents the mean,error bars represent the 10th and 90th percentiles, and the dots represent the 5th and 95th percentiles. (c) Mean bioenergeticefficiency among age classes for those MGAs where available.

February 2013 363ENERGETIC LIMITATION OF GLOBAL INVADERS

2008). Similarly, the hydrologic regime (Poff et al. 1997)

drives other physical factors we did not consider that

can affect the growth and consumption of stream fishes,

including stream velocity and microhabitat features

(e.g., Hill and Grossman 1993). Nonetheless, the

relatively high BioEff in the U.S. West results, in part,

from the ability to switch to a diet including piscivory, a

switch that occurs only in the exotic habitat of brown

trout. All else equal in terms of foraging (e.g., handling,

pursuit), a diet that includes some portion of fish is

much higher in energy content (Werner and Mittlebach

1981, Hart 1993, Sih and Christensen 2001). In both the

United States and New Zealand, the native sculpin and

galaxiids, respectively, represent potential novel prey

fish that have not coevolved with exotic brown trout,

and both make up a substantial portion of the diet

(McIntosh et al. 2010, Richardson 2011).

In addition to greater mean BioEff at size or age, an

ability to switch to a diet including piscivory also allows

an individual fish to grow to a much larger body size

than possible on a diet of invertebrates alone (Juanes et

al. 2002; but see Hayes et al. 2000). This size advantage

can then lead to even greater rates of piscivory as size-

based gape limits are overcome, which then increases

predation pressure on native prey over a longer time

period of the brown trout’s lifetime. In the United

States, for example, BioEff actually increases with older

age classes, and fluvial brown trout in the United States

grow to greater than 550 mm and ages exceeding age 6.

While stream productivity certainly plays a role in this

pattern, this is a fluvial life-history pattern facilitated by

switching to a diet including piscivory (Juanes et al.

2002). In contrast, in their native range, fluvial brown

trout rarely exceed 300 mm or ages of age 3, before they

senesce (Lobon-Cervia 2008) or must switch to an

adfluvial or anadromous life history to achieve larger

biomass (e.g., Valiente et al. 2010). The incidence of

piscivory increases rapidly around 150–200 mm in

brown trout, and once trout exceed 300 mm, they are

rarely gape limited (McIntosh 2000). Greater growth

rates resulting from piscivory or predation on other

energetically rich prey (e.g., Hayes et al. 2000) can also

lead to superior competitive abilities with conspecifics

(McHugh and Budy 2005), thus potentially exacerbating

the community-level effects (i.e., competition and

predation) of exotic fish invaders.

In direct contrast to the pattern observed in their

exotic habitat, in native regions brown trout appear to

coexist with potential native prey fish (e.g., sculpin,

Cottus spp.; a major dietary source in their exotic

habitat), and cottids may even be superior competitors

for food and space when sympatric with brown trout

(Hesthagen and Heggenes 2003, Holmen et al. 2003).

The life-history pattern of growth and associated high

rates of piscivory of exotic brown trout on naive natives

has been shown to have contributed to the decline of

native fishes in Chile (Young et al. 2010), Australia

(Jackson et al. 2004), and New Zealand (McIntosh et al.

2010) and has been implicated in the United States (see

Results) and elsewhere (McIntosh et al. 2011). The niche

for a large, fluvial trout species was historically occupied

by different native trout (e.g., fluvial cutthroat trout,

Oncorhynchus clarkii utah) in the exotic range of brown

trout (U.S. West). However in this MGA, invasive

brown trout have essentially created a new niche

characterized by novel traits associated with diet,

growth, and consumption rates across the lifetime

(Moyle et al. 1986, Simberloff 1995, Richardson 2011).

Although these complex effects of life history, diet,

and thermal environment can be highly intertwined and

difficult to separate, BioEff is a sensitive and integrated

response variable for evaluating overall fish perfor-

mance. Nonetheless, this metric describes only the

performance of survivors and is often only possible to

accurately quantify for juvenile and older fish. Herein,

we observed only a weak signal of density-dependent

effects on BioEff, a result that is perhaps unsurprising as

we modeled the BioEff only of average survivors. In the

rich body of research on this topic, density-dependent

population regulation typically occurs at very early life

stages and is manifested as cohort recruitment strength

(e.g., Lobon-Cervia 2005). In addition, density-depen-

dent population regulation may be determined by both

intra- and interspecific interactions (Holmen et al. 2003).

Nonetheless, interannual variability in BioEff was high,

a pattern that can be driven both by annual variation in

climate (e.g., Lobon-Cervia and Rincon 1998) and by

internal population regulation (e.g., Milner et al. 2003,

Vøllestad and Olsen 2008). And notably, the majority of

extremely high densities of brown trout observed

occurred primarily in countries where they are exotic,

suggesting brown trout may also be able to reach and

sustain greater densities in exotic habitat relative to

within their native distribution (Shea and Chesson 2002,

Benjamin and Baxter 2010). Assuming an ecologically

similar native species was formerly present in high

densities, this finding may parallel that of Stohlgren et

al. (2006) that the density of exotic plants, birds, and

fishes can be accurately predicted based on positive

correlations with historically high densities of natives.

Our approach can be extended to help prioritize

actions aimed at minimizing the impact of exotic brown

trout on native species and ecosystems in their exotic

range and conversely conserving native brown trout.

Given that a BioEff value closer to 100% indicates

performance closer to the theoretical maximum growth

and consumption rate for a given temperature regime,

we might expect BioEff to be high in good quality,

native habitat. Deviations from this expectation could

suggest the potential limiting effect of another factor

(i.e., other than temperature and food availability).

However, in a review of 639 published estimates of

BioEff covering a wide variety of trophic niches, median

BioEff was 43%, and ,5% of fish populations exhibited

BioEff .80% (Armstrong and Schindler 2011). The

authors suggest this pattern occurs when daily foraging

PHAEDRA BUDY ET AL.364 Ecology, Vol. 94, No. 2

opportunities are extremely heterogeneous, such that

episodes of gorging and fasting are common. Nonethe-

less, when fishes consistently demonstrate extremely

high, or extremely low, values of BioEff in exotic

habitat, this could indicate their likely success as

invaders or point to potential limiting factors. For

example, our results demonstrate that in their native

habitat, fish performance (BioEff ) is relatively low,

which suggests that fish performance is also influenced

by other anthropogenic impacts. Many native popula-

tions of brown trout in Europe have experienced

dramatic reductions in range and abundance (e.g.,

FAO 2002) with overharvest and genetic introgression

with hatchery fish both thought to represent potential

limiting factors (Almodovar and Nicola 2004, Almodo-

var et al. 2006). In addition, the BioEff of exotic brown

trout was generally .43% (in several cases far greater),

the median value observed in our meta-analysis. This

observation highlights an area worthy of greater

investigation across invasive species—do exotic fishes

consistently demonstrate greater than average or ex-

pected BioEff?

Furthermore, the driving role of temperature in

determining all physiological rates of ectotherms makes

our bioenergetics approach well suited for predicting

and understanding the effects of climate change (Dukes

and Mooney 1999, Lockwood et al. 2007). The likely

effects of climate change for fishes include not only

commonly considered changes in range size (e.g., Hari et

al. 2006, Wenger et al. 2011), but may also include

energetic changes (positive or negative) in some areas of

their native range. Within this context, our approach has

potential for addressing contemporary questions of both

invasion success and native fish conservation now and

under a future climate (see Results; Fleishman et al.

2011). However, we note that while temperature and fish

size data were widely available and part of some large

regional data sets, diet composition was the most

common missing input, eliminating potential sites from

our analysis. Our ability to complete these large-scale,

energetic analyses for fishes might be strengthened by

trophic and diet information consistently collected both

in more places and at finer spatial and temporal scales.

The intercontinental and comparative nature of our

study eliminates, to some degree, the challenges of

context-dependent variation in determining invasion

success. Like others (Fitzpatrick et al. 2007), we warn

that models predicting invasion success based on the

performance of exotic species within their native range

or based on ‘‘guilt by association’’ (i.e., taxonomy) may

underestimate their performance in exotic habitat. Last,

we suggest that growth plasticity across the life span, a

potential life-history trait that could be expressed by

many fishes, should be added to growing lists of factors

characterizing successful invaders worldwide.

Brown trout, one of the world’s most notorious

invaders, but also a species of conservation concern

appears to possess a genetic deck of cards from which

any number of successful strategies for growth and life-

history expression can be drawn given biota present and

the environment. Although there are limitations to thistype of large-scale, energetic analyses, our understand-

ing of invasion at the global scale could be improved by

similar applications to other organisms where large data

sets permit.

ACKNOWLEDGMENTS

We thank Jon Flinders, Douglas Dieterman, and JohnHayes for graciously sharing data sets and interpretation. Thiswork was funded and supported by a National ScienceFoundation ADVANCE Program grant to Phaedra Budy, theUtah Division of Wildlife Resources, Project XII, SportFisheries Research, Grant Number F-47-R, Segment 20, theU.S. Geological Survey–Utah Cooperative Fish and WildlifeResearch Unit (in-kind), and the Department of WatershedSciences and Ecology Center at Utah State University. Wethank Susan Durham for statistical guidance, Christy Meredithfor developing the map, and Brian Bailey for formatting andeditorial assistance. Two anonymous reviewers substantiallyimproved the clarity and rigor of previous drafts of themanuscript. Mention of brand names in this paper does notimply endorsement by the U.S. Government.

LITERATURE CITED

Almodovar, A., and G. G. Nicola. 2004. Angling impact onconservation of Spanish stream-dwelling brown trout Salmotrutta. Fisheries Management and Ecology 11:173–182.

Almodovar, A., G. G. Nicola, and B. Elvira. 2006. Spatialvariation in brown trout production: the role of environ-mental factors. Transactions of the American FisheriesSociety 135:1348–1360.

Armstrong, J. B., and D. E. Schindler. 2011. Excess digestivecapacity in predators reflects a life of feast and famine.Nature 476:84–88.

Benjamin, J. R., and C. V. Baxter. 2010. Do nonnativesalmonines exhibit greater density and production than thenatives they replace? A comparison of nonnative brook troutwith native cutthroat trout. Transactions of the AmericanFisheries Society 139:641–651.

Bollache, L., J. T. A. Dick, K. D. Farnsworth, and W. I.Montgomery. 2008. Comparison of the functional responsesof invasive and native amphipods. Biology Letters 4:166–169.

Brook, B. W., N. S. Sodhi, and C. J. A. Bradshaw. 2008.Synergies among extinction drivers under global change.Trends in Ecology and Evolution 23:453–460.

Bruno, J. F., J. D. Fridley, K. D. Bromberg, and M. D.Bertness. 2005. Insights into biotic interactions from studiesof species invasions. Pages 13–40 in D. F. Sax, J. J.Stachowicz, and S. D. Gaines, editors. Species invasions:insights into ecology, evolution, and biogeography. Sinauer,Sunderland, Massachusetts, USA.

Budy, P., M. E. Baker, and S. K. Dahle. 2011. Predicting fishgrowth potential and identifying potential water qualityissues: a spatially explicit bioenergetics approach. Environ-mental Management 48:691–709.

Carey, M. P., and D. H. Wahl. 2010. Native fish diversity altersthe effects of an invasive species on food webs. Ecology91:2965–2974.

Clout, M. N., and P. A. Williams. 2009. Invasive speciesmanagement: a handbook of principles and techniques.Oxford University Press, New York, New York, USA.

Dukes, J. S., and H. A. Mooney. 1999. Does global changeincrease the success of biological invaders? Trends in Ecologyand Evolution 14:135–139.

FAO (Food and Agriculture Organization of the UnitedNations). 2002. Report of the symposium on inland fisheriesmanagement and the aquatic environment. European Inland

February 2013 365ENERGETIC LIMITATION OF GLOBAL INVADERS

Fisheries Advisory Commission, FAO Fisheries Report 681,Supplement. FAO, Rome, Italy.

Fitzpatrick, M. C., J. F. Weltzin, N. J. Sanders, and R. R.Dunn. 2007. The biogeography of prediction error: why doesthe introduced range of the fire ant over-predict its nativerange? Global Ecology and Biogeography 16:24–33.

Fleishman, E., et al. 2011. Top 40 priorities for science toinform US conservation and management policy. BioScience61:290–300.

Frank, B. M., J. J. Piccolo, and P. V. Baret. 2011. A review ofecological models for brown trout: towards a new demoge-netic model. Ecology of Freshwater Fish 20:167–198.

Hanson, P. C., T. B. Johnson, D. E. Schindler, and J. F.Kitchell. 1997. Fish bioenergetics 3.0. University of Wiscon-sin, Sea Grant Institute, Technical Report WISCU-T-97-001.Madison, Wisconsin, USA.

Hari, R. E., D. M. Livingstone, R. Siber, P. Burkhardt-Holm,and H. Guttinger. 2006. Consequences of climatic change forwater temperature and brown trout populations in Alpinerivers and streams. Global Change Biology 12:10–26.

Hart, P. J. B. 1993. Teleosts foraging: facts and theories. Pages253–284 in T. J. Pitcher, editor. Behaviour of teleost fishes.Chapman and Hall, London, UK.

Hartman, K. J., and J. F. Kitchell. 2008. Bioenergeticsmodeling: progress since the 1992 symposium. Transactionsof the American Fisheries Society 137:216–223.

Hayes, J. W., J. D. Stark, and K. A. Shearer. 2000.Development and test of a whole-lifetime foraging andbioenergetics growth model for drift feeding brown trout.Transactions of the American Fisheries Society 129:315–332.

Hesthagen, T., and J. Heggenes. 2003. Competitive habitatdisplacement of brown trout by Siberian sculpin: the role ofsize and density. Journal of Fish Biology 62:222–236.

Hill, J., and G. D. Grossman. 1993. An energetic model ofmicrohabitat use for rainbow trout and rosyside dace.Ecology 74:685–698.

Holmen, J., E. M. Olsen, and L. A. Vøllestad. 2003.Interspecific competition between stream-dwelling browntrout and Alpine bullhead. Journal of Fish Biology62:1312–1325.

Hudy, M. 1990. Brown trout population structures in WhiteRiver tailwaters currently managed under no special regula-tions. Pages 94–97 in J. C. Borawa, editor. Brown troutworkshop: biology and management. American FisheriesSociety, Special Symposium, Bethesda, Maryland, USA.

IUCN (International Union for Conservation of Nature). 2000.IUCN guidelines for the prevention of biodiversity losscaused by alien invasive species. Invasive Species SpecialistGroup, IUCN, Gland, Switzerland.

Jackson, J. E., T. A. Raadik, M. Lintermans, and M. Hammer.2004. Alien salmonids in Australia: impediments to effectivemanagement, and future directions. New Zealand Journal ofMarine and Freshwater Research 38:447–455.

Jensen, L. F., M. M. Hansen, C. Pretoldi, G. Holdensgaard,K. L. D. Mensberg, and V. Loeschcke. 2008. Localadaptation in brown trout early life-history traits: implica-tions for climate change adaptability. Proceedings of theRoyal Society B 275:2859–2868.

Johnson, B. M., P. J. Martinez, J. A. Hawkins, and K. R.Bestgen. 2008. Ranking predatory threats by nonnative fishesin the Yampa River, Colorado, via bioenergetics modeling.North American Journal of Fisheries Management 28:1941–1953.

Jonsson, B., and N. Jonsson. 2011. Ecology of Atlantic salmonand brown trout. Pages 1–21 in Habitat as a template for lifehistories. Fish and fisheries series 33. Springer, London, UK.

Juanes, F., J. A. Buckel, and F. S. Scharf. 2002. Feedingecology of piscivorous fishes. Pages 267–283 in P. J. B. Hartand J. D. Reynolds, editors. Handbook of fish biology andfisheries. Volume 1, Fish biology. Blackwell Publishing,Malden, Massachusetts, USA.

Keith, P., and J. Allardi. 1996. Endangered freshwater fish: thesituation in France. Pages 35–54 in A. Kirchhofer and D.Hefti, editors. Conservation of endangered freshwater fish inEurope. Birkhauser Verlag, Basel, Switzerland.

Klemetsen, A., P. A. Amundsen, J. B. Dempson, B. Jonsson, N.Jonsson, M. F. O’Connell, and E. Mortensen. 2003. Atlanticsalmon (Salmo salar L.), brown trout (Salmo trutta L.) andArctic char (Salvelinus alpinus L.): a review of aspects of theirlife histories. Ecology of Freshwater Fish 12:1–59.

Kolar, C. S., and D. M. Lodge. 2001. Progress in invasionbiology: predicting invaders. Trends in Ecology and Evolu-tion 16:199–204.

Korsu, K., A. Huusko, and T. Muotka. 2010. Invasion of northEuropean streams by brook trout: hostile takeover or pre-adapted habitat niche segregation? Biological Invasions12:1363–1375.

Lobon-Cervia, J. 2005. Spatial and temporal variation in theinfluence of density dependence on growth of stream-livingbrown trout (Salmo trutta). Canadian Journal of Fisheriesand Aquatic Sciences 62:1231–1242.

Lobon-Cervia, J. 2008. Why, when and how do fish populationsdecline, collapse and recover? The example of brown trout(Salmo trutta) in Rio Chaballos northwestern Spain.Freshwater Biology 54:1149–1162.

Lobon-Cervia, J., and P. A. Rincon. 1998. Field assessment ofthe influence of temperature on growth rate in a brown troutpopulation. Transactions of the American Fisheries Society127:718–728.

Lockwood, J. L., M. F. Hoopes, and M. P. Marchetti. 2007.Invasion ecology. Blackwell, Malden, Massachusetts, USA.

Mack, R. N., D. Simberloff, W. M. Lonsdale, H. Evans, M.Clout, and F. A. Bazzaz. 2000. Biotic invasions: causes,epidemiology, global consequences, and control. EcologicalApplications 10:689–710.

Marchetti, M. P., P. B. Moyle, and R. Levine. 2004. Alien fishesin California watersheds: characteristics of successful andfailed invaders. Ecological Applications 14:587–596.

McHugh, P., and P. Budy. 2005. An experimental evaluation ofcompetitive and thermal effects on brown trout (Salmotrutta) and Bonneville cutthroat trout (Oncorhynchus clarkiiutah) performance along an altitudinal gradient. CanadianJournal of Fisheries and Aquatic Sciences 62:2784–2795.

McIntosh, A. R. 2000. Habitat- and size-related variations inexotic trout impacts on native galaxiid fishes in New Zealandstreams. Canadian Journal of Fisheries and Aquatic Sciences57:2140–2151.

McIntosh, A. R., P. A. McHugh, and P. Budy. 2011. Browntrout (Salmo trutta). Pages 285–296 in R. A. Francis, editor.A handbook of global freshwater invasive species. Earthscan,London, UK.

McIntosh, A. R., P. A. McHugh, N. R. Dunn, J. M. Goodman,S. W. Howard, P. G. Jellyman, L. K. O’Brien, P. Nystromand J. D. Woodford. 2010. The impact of trout on galaxiidfishes in New Zealand. New Zealand Journal of Ecology34:195–206.

Milner, N. J., J. M. Elliott, J. D. Armstrong, R. Gardiner, J. S.Welton, and M. Ladle. 2003. The natural control of salmonand trout populations in streams. Fisheries Research 62:111–125.

Moyle, P. B., H. W. Li, and B. A. Barton. 1986. TheFrankenstein effect: impact of introduced fishes on nativefishes in North America. Pages 415–426 in R. H. Stoud,editor. Fish culture in fisheries management. AmericanFisheries Society, Bethesda, Maryland, USA.

Ojanguren, A. F., F. G. Reyes-Gavilan, and F. Brana. 2001.Thermal sensitivity of growth, food intake and activity ofjuvenile brown trout. Journal of Thermal Biology 26:165–170.

Panov, V. E., S. Gollasch, and F. Lucy. 2011. An importantpart of the developing European information and early

PHAEDRA BUDY ET AL.366 Ecology, Vol. 94, No. 2

warning system on invasive alien species. Aquatic Invasions6:1–5.

Perrings, C. 2002. Biological invasions in aquatic systems: theeconomic problem. Bulletin of Marine Science 70:541–552.

Pimentel, D., L. Lach, R. Zuniga, and D. Morrison. 2000.Environmental and economic costs associated with non-indigenous species in the United States. BioScience 50:53–65.

Poff, N. L., D. Allan, M. B. Bain, J. R. Karr, K. L. Prestegard,B. D. Richter, R. E. Sparks, and J. C. Stromberg. 1997. Thenatural flow regime. BioScience 47:769–784.

Richardson, D. M. 2011. Fifty years of invasion ecology: thelegacy of Charles Elton. Wiley-Blackwell, Oxford, UK.

Ricklefs, R. E. 2004. A comprehensive framework for globalpatterns in biodiversity. Ecology Letters 7:1–15.

Shea, K., and P. Chesson. 2002. Community ecology theory asa framework for biological invasions. Trends in Ecology andEvolution 17:170–176.

Sih, A., and B. Christensen. 2001. Optimal diet theory: whendoes it work, and when and why does it fail? AnimalBehaviour 61:379–390.

Simberloff, D. 1995. Why do introduced species appear todevastate islands more than mainland areas? Pacific Science49:87–97.

Stohlgren, T. J. 2002. Beyond theories of plant invasions:lessons from the natural landscape. Comments in TheoreticalBiology 7:355–379.

Stohlgren, T. J., D. Barnett, C. Flather, P. Fuller, B. Peterjohn,J. Kartesz, and L. L. Master. 2006. Species richness andpatterns of invasion in plants, birds, and fishes in the UnitedStates. Biological Invasions 8:427–447.

Valiente, A. G., F. Juanes, P. Nunez, and E. Garcia-Vazquez.2010. Brown trout (Salmo trutta) invasiveness: plasticity inlife-history is more important than genetic variability.Biological Invasions 12:451–462.

Vinson, M., and P. Budy. 2011. Sources of variability andcomparability between salmonids stomach contents andisotopic analyses: study design lessons and recommendations.

Canadian Journal of Fisheries and Aquatic Sciences 68:137–151.

Vøllestad, L. A., and E. M. Olsen. 2008. Non-additive effects ofdensity-dependent and density-independent factors on browntrout vital rates. Oikos 117:1752–1760.

Vøllestad, L. A., E. M. Olsen, and T. Forseth. 2002. Growth-rate variation in brown trout in small neighbouring streams:evidence for density-dependence? Journal of Fish Biology61:1513–1527.

Wenger, S. J., D. J. Isaak, C. H. Luce, H. M. Veville, K. D.Fausch, J. B. Dunham, D. C. Dauwalter, M. K. Young,M. M. Elsner, B. E. Rieman, A. F. Hamlet, and J. E.Williams. 2011. Flow regime, temperature, and bioticinteractions drive differential declines of trout species underclimate change. Proceedings of the National Academy ofSciences USA 34:14175–14180.

Werner, E. E., and G. G. Mittlebach. 1981. Optimal foraging:field tests of diet choice and habitat switching. AmericanZoologist 4:813–829.

Westfall, P. H., R. D. Tobias, R. D. Wolfinger, and Y.Hochberg. 1999. Multiple comparisons and multiple testsusing the SASt System. SAS Institute, Cary, North Carolina,USA.

Westley, P. A. H., and I. A. Fleming. 2011. Landscape factorsthat shape a slow and persistent aquatic invasion: browntrout in Newfoundland 1883–2010. Diversity and Distribu-tions 17:566–579.

Whitledge, G. W., P. G. Bajer, and R. S. Hayward. 2010.Laboratory evaluation of two bioenergetics models or browntrout. Transactions of the American Fisheries Society139:929–936.

Young, K. A., J. B. Dunham, J. F. Sephenson, A. Terreau,A. F. Thailly, G. Gajardo, and C. Garcia de Leaniz. 2010. Atrial of two trouts: comparing the impacts of rainbow andbrown trout on a native galaxiid. Animal Conservation13:399–410.

SUPPLEMENTAL MATERIAL

Appendix A

A table describing the countries, major geographic areas (MGAs), streams, age classes, and time series of data for which wemodeled brown trout bioenergetics, along with the range of densities and the number of individual model simulations for eachMGA and the details of bioenergetic simulations (Ecological Archives E094-030-A1).

Appendix B

World map showing native and introduced ranges of brown trout (Salmo trutta), and identifying major geographic areas(MGAs) (Ecological Archives E094-030-A2).

February 2013 367ENERGETIC LIMITATION OF GLOBAL INVADERS