food habits and fish prey size selection of a newly colonizing population of river otters (lontra...

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors,nonprofit publishers, academic institutions, research libraries, and research funders in the common goal ofmaximizing access to critical research.

Food Habits and Fish Prey Size Selection of a NewlyColonizing Population of River Otters (Lontracanadensis) in Eastern North DakotaAuthor(s): Cory R. Stearns and Thomas L. SerfassSource: The American Midland Naturalist, 165(1):169-184. 2011.Published By: University of Notre DameDOI: http://dx.doi.org/10.1674/0003-0031-165.1.169URL: http://www.bioone.org/doi/full/10.1674/0003-0031-165.1.169

BioOne (www.bioone.org) is a nonprofit, online aggregation of coreresearch in the biological, ecological, and environmental sciences. BioOneprovides a sustainable online platform for over 170 journals and bookspublished by nonprofit societies, associations, museums, institutions, andpresses.

Your use of this PDF, the BioOne Web site, and all posted and associatedcontent indicates your acceptance of BioOne’s Terms of Use, available atwww.bioone.org/page/terms_of_use.

Usage of BioOne content is strictly limited to personal, educational, andnon-commercial use. Commercial inquiries or rights and permissionsrequests should be directed to the individual publisher as copyright holder.

http://dx.doi.org/10.1674/0003-0031-165.1.169

http://www.bioone.org/doi/full/10.1674/0003-0031-165.1.169

http://www.bioone.org

http://www.bioone.org/page/terms_of_use

Food Habits and Fish Prey Size Selection of a Newly ColonizingPopulation of River Otters (Lontra canadensis) in Eastern

North Dakota

CORY R. STEARNS AND THOMAS L. SERFASS1

Department of Biology and Natural Resources, Frostburg State University, Frostburg, Maryland 21532

ABSTRACT.—The food habits of river otters (Lontra canadensis) on three rivers in the RedRiver of the North drainage of eastern North Dakota were evaluated using an analysis of 569scats collected between 4 Oct. 2006 and 26 Nov. 2007. Fish and crayfish were the primary preyitems, occurring in 83.0% and 51.1% of scats, respectively. Other prey included insects(26.7%), birds (7.9%), amphibians (6.7%), mammals (6.0%) and freshwater mussels (0.2%).Fish of Cyprinidae (carp and minnows) were the most prominent fish in the diet, occurringin 64.7% of scats. Other relatively common fish in the diet included Ictaluridae (catfish,17.4% frequency of occurrence), Catostomidae (suckers, 13.0%), and Centrarchidae(sunfish, 11.2%). The diet of river otters changed seasonally, including a decline in thefrequency of fish in the summer diet, and a corresponding increase in the occurrence ofcrayfish. Consumed fish ranged from 3.5 to 71.0 cm total length, with a mean of 20.7 cm (SE 6

0.5, n 5 658). Fish 10.1–20.0 cm were the most frequently consumed size class (36.5% relativefrequency), with the majority of other consumed fish being #10.0 cm (24.6%), 20.1–30.0 cm(14.1%), 30.1–40.0 cm (14.0%), or 40.1–50.0 cm (8.2%). The size of consumed fish changedseasonally, with spring having the largest mean prey size.

INTRODUCTION

Historically, the nearctic river otter (Lontra canadensis) occurred on most rivers in NorthDakota, and was relatively common into the 1890s (Bailey, 1926; Adams, 1961). River ottersstill occurred in the 1920s along the major rivers and some lakes but had become rare by the1960s, and were considered extirpated soon after (Bailey, 1926; Adams, 1961). However, inrecent years reports of river otters have increased, with most coming from the Red River ofthe North (hereafter referred to as Red River) drainage, and Lake Sakakawea in theMissouri river drainage (Hagen et al., 2005).

River otters are opportunistic aquatic predators. Although the diet is diverse, most dietaryanalyses have shown fish to be the primary prey (e.g., Greer, 1955; Melquist and Hornocker,1983; Serfass et al., 1990). River otters are presumed to select fish in proportion to theirabundance and in inverse proportion to swimming speed and agility (Ryder, 1955).Therefore, the most abundant and slowest swimming fishes tend to be taken most often.Catostomidae (suckers), Centrarchidae (sunfish and bass), Cyprinidae (carp and minnows)and Ictaluridae (catfish) are usually among the most frequently occurring fish familiesdetected in river otter diet studies (e.g., Wilson, 1954; Greer, 1955; Hamilton, 1961; Griess,1987; Serfass et al., 1990; Noordhuis, 2002; Giordano, 2005). When available, crayfish areusually the second most important prey item and in a few studies have occurred mostfrequently in the diet (Grenfell, 1974; Griess, 1987; Noordhuis, 2002). Other organismsconsumed by river otters include amphibians, insects and other invertebrates,birds, mammals and reptiles (Ryder, 1955; Melquist and Hornocker, 1983; Serfass et al.,1990).

1 Corresponding author present address: Department of Biology and Natural Resources, FrostburgState University, Frostburg, Maryland 21532; Telephone: (301) 687-4171; e-mail: [email protected].

Am. Midl. Nat. 165:169–184

169

Despite many previous food studies on river otters, rarely have studies been conducted toassess the size of their fish prey. Previous studies have made general inferences about preysize, indicating that fish prey ranges from 2–80 cm and that most fish consumed areprobably 10–30 cm in length (Lagler and Ostenson, 1942; Greer, 1955; Ryder, 1955;Hamilton, 1961; Toweill, 1974; Melquist and Hornocker, 1983; Stenson et al., 1984; Griess,1987; Tumlison and Karnes, 1987; Noordhuis, 2002; Giordano, 2005). However, thesestudies typically did not indicate the methods used in their assessments, or establishpredictive relationships between anatomical structures (that are recoverable from thedigestive tracts or scats, such as bones and scales) and fish length. Also, inferences have beenlimited to one or a few species (occasionally only a few individuals) and only providedinformation on the size range (maximum and minimum) or common prey sizes. This studywas part of a larger project to determine the distribution of river otters in the Red Riverdrainage of eastern North Dakota and was the first research conducted on the reestablishingpopulation. The objectives of this aspect of the study were to: (1) assess the food habits ofriver otters in recolonized areas, (2) determine the size of fish preyed on by river otters and(3) evaluate seasonal variation in the diet and prey size.

STUDY AREA

The Red River forms at the convergence of the Bois de Sioux River and the Ottertail Riverat Wahpeton, North Dakota and Brackenridge, Minnesota (46u15.849N, 96u35.929W). Theriver flows north forming the boundary between North Dakota and Minnesota for nearly640 km before entering Manitoba, Canada (Koel and Peterka, 1998). The landscape of theRed River drainage has low relief, and mostly occurs within the former lake bed of LakeAgassiz (Stoner et al., 1993). The majority of the Red River valley (about 80%) is cropland,but pasture also is present (Stoner et al., 1993). Forested regions mostly are confined tonarrow riparian strips (Stoner et al., 1993). Riparian areas consist of strips of grass or trees,but in some areas agricultural fields extend to the river banks (Stoner et al., 1993). TheNorth Dakota tributaries of the Red River are very similar in physical structure-meanderingand typically have low gradients and high turbidity (Copes and Tubb, 1966; Stoner et al.,1993). In spring 2007, primary study areas were established on the Red River in GrandForks, ND, and East Grand Forks, MN, Forest River in Ardoch township (Walsh County) andTurtle River in Turtle River township (Grand Forks County) (Fig. 1). Although the RedRiver study site was located in an urban area, both cities have a protected riparian greenwayproviding abundant riparian vegetation. The study areas on the Forest and Turtle riversoccurred within agricultural landscapes. There was a narrow forested riparian strip in theTurtle River study area. But, there was little woody riparian vegetation in the Forest Riverstudy area because agricultural fields extend to its banks. The western border of the ForestRiver study area was Ardoch National Wildlife Refuge, which provided river otters access tolacustrine and wetland habitats. In addition to the primary study areas, latrines on theTongue River were monitored after their discovery in May 2007 (Fig. 1). The Tongue issmaller than the other rivers in the study and is a tributary of the Pembina River (which is atributary of the Red River). The latrines on the Tongue River were located in the townshipsof Akra and Bathgate (Cavalier County).

METHODS

The diet of river otters was assessed by analyzing scats collected between 4 Oct. 2006 and26 Nov. 2007. Initially, scats were collected during sign surveys (surveys along river banks todetect scats, tracks or other sign) or when checking latrine sites. Beginning in spring 2007,

170 THE AMERICAN MIDLAND NATURALIST 165(1)

the primary study areas were surveyed on foot at least monthly but were usually checkedweekly or biweekly. Scats were collected on the Tongue River during sporadic checks oflatrine sites.

Scats were collected in individual plastic bags, which subsequently were labeled withidentifying information (i.e., date, river and site) and frozen until analysis. In preparationfor analysis, scats were washed by soaking over night in soapy water and then rinsed througha 0.125 mm mesh sieve to eliminate small organic material and other debris. After drying,food particles were separated to facilitate identification. Fish remains were identified tospecies or family using Daniels’ (1996) scale identification key, and reference collections ofscales and other bony structures. We used Spiers (1973) to identify mammalian prey, andother remains (e.g., amphibian bones) were identified using reference collections.

The diet was assessed using frequency of occurrence, determined by tabulating thenumber of scats the prey occurred in and dividing by the total number of scats. Seasonalvariation in the diet was evaluated by assigning scats to a season depending on theircollection date. Scats collected from 1 Mar.–31 May were defined as spring, from 1 Jun.–31Aug. as summer, 1 Sep.–30 Nov. as fall, and 1 Dec.–28 Feb. as winter. Food habits of riverotters on the Tongue River were assessed separately because the vast majority of scats werecollected in late spring and summer, and this river differed physically (much smaller at thecollection site) than the other rivers in the study.

The size of fish consumed by river otters was estimated using body-scale relationships(relating fish total length to scale size) established from samples collected throughout theNorth Dakota tributaries of the Red River, Jun.–Nov. 2007. The samples were collected incollaboration with researchers from South Dakota State University and were obtainedthrough electrofishing and the use of seines and clover leaf traps. The size of fish prey wasestimated by inserting the measurements of scales sorted from river otter scats into species-specific or group models. Lateral line scales were preferred for size estimation (because they

FIG. 1.—Map of the primary (ovals) and secondary (rectangle) study areas in an analysis of the foodhabits of a reestablishing population of river otters in eastern North Dakota, Oct. 2006–Nov. 2007

2011 STEARNS AND SERFASS: RIVER OTTER FOOD HABITS AND PREY SIZE 171

TA

BL

E1.

—Se

aso

nal

com

par

iso

no

fth

efo

od

hab

its

of

rive

ro

tter

so

nth

eF

ore

st,

Red

,an

dT

urt

leR

iver

so

fea

ster

nN

ort

hD

ako

ta,

by

the

nu

mb

ero

fo

ccu

rren

ces

(No

.)an

dfr

equ

ency

of

occ

urr

ence

(Fo

O,

exp

ress

edas

ap

erce

nta

ge)

insc

ats

coll

ecte

dO

ct.

2006

–No

v.20

07.

Scat

sco

llec

ted

1M

ar.–

31M

ayw

ere

con

sid

ered

assp

rin

g,1

Jun

.–30

Au

g.as

sum

mer

,1

Sep

t.–3

0N

ov.

asfa

llan

d1

Dec

.–28

Feb

.as

win

ter.

Bo

ldin

dic

ates

asi

gnif

ican

td

iffe

ren

ce(P

,0.

05) Pre

yit

em

Spri

ng

(n5

133)

Sum

mer

(n5

150)

Fal

l(n

526

4)W

inte

r(n

522

)T

ota

l(n

556

9)

x2

PN

o.

Fo

ON

o.

Fo

ON

o.

Fo

ON

o.

Fo

ON

o.

Fo

O

Cra

yfis

h48

36.1

9362

.014

253

.88

36.4

291

51.1

21.8

,0.0

01

Inse

cts

3627

.128

18.7

8431

.84

18.2

152

26.7

9.3

0.0

26

Fre

shw

ater

mu

ssel

s1

0.8

00.

00

0.0

00.

01

0.2

——

Am

ph

ibia

ns

64.

527

18.0

41.

51

4.5

386.

743

.3,

0.0

01

Bir

ds

2015

.013

8.7

124.

50

0.0

457.

915

.40.0

02

Mam

mal

s8

6.0

42.

721

8.0

14.

534

6.0

4.8

0.18

4F

ish

111

83.5

101

67.3

238

90.2

2210

047

283

.040

.1,

0.0

01

Cat

ost

om

idae

(su

cker

s)15

11.3

53.

350

18.9

418

.274

13.0

21.5

,0.0

01

Cat

osto

mu

sco

mm

erso

ni

(wh

ite

suck

er)

118.

33

2.0

186.

81

4.5

335.

8M

oxos

tom

asp

p.

(red

ho

rses

)2

1.5

10.

724

9.1

29.

129

5.1

Icti

obu

ssp

p.

or

Car

piod

escy

prin

us

10.

81

0.7

51.

91

4.5

81.

4C

entr

arch

idae

(su

nfi

sh)

2518

.83

2.0

2911

.07

31.8

6411

.229

.8,

0.0

01

Pom

oxis

spp

.(c

rap

pie

)3

2.3

00.

07

2.7

29.

112

2.1

Am

blop

lite

sru

best

ris

or

Lep

omis

spp

.18

13.5

32.

014

5.3

627

.341

7.2

Un

kno

wn

cen

trar

chid

s6

4.5

00.

011

4.2

14.

518

3.2

Cyp

rin

idae

(car

pan

dm

inn

ow

s)71

53.4

8355

.320

075

.814

63.6

368

64.7

27.4

,0.0

01


Pre

yit

em

Spri

ng

(n5

133)

Sum

mer

(n5

150)

Fal

l(n

526

4)W

inte

r(n

522

)T

ota

l(n

556

9)

x2

PN

o.

Fo

ON

o.

Fo

ON

o.

Fo

ON

o.

Fo

ON

o.

Fo

O

Cyp

rin

us

carp

io(c

arp

)61

45.9

5738

.018

570

.113

59.1

316

55.5

Rhi

nic

hthy

ssp

p.

(dac

e)1

0.8

00.

01

0.4

00.

02

0.4

Lar

gen

on

-car

pcy

pri

nid

11

0.8

10.

76

2.3

00.

08

1.4

Smal

lcy

pri

nid

20

0.0

138.

726

9.8

14.

540

7.0

Eso

cid

aeE

sox

luci

us

(no

rth

ern

pik

e)11

8.3

32.

011

4.2

313

.628

4.9

9.8

0.0

20

Hio

do

nti

dae

(mo

on

eyes

)5

3.8

21.

36

2.3

00.

013

2.3

1.8

0.40

43

Icta

luri

dae

(cat

fish

)14

10.5

10.

784

31.8

00.

099

17.4

76.4

,0.0

01

Per

cid

ae(p

erch

es)

86.

03

2.0

166.

11

4.5

284.

93.

80.

282

Eth

eost

oma

spp

.o

rP

erci

na

spp

.(d

arte

rs)

53.

80

0.0

103.

81

4.5

162.

8P

erca

flav

esce

ns

or

San

der

spp

.4

3.0

32.

06

2.3

00.

013

2.3

Mo

ron

idae

/Sc

iaen

idae

Mor

one

chry

sops

(wh

ite

bas

s)o

rA

plod

inot

us

gru

nn

ien

s(f

resh

wat

erd

rum

)19

14.3

10.

75

1.9

29.

127

4.7

38.0

,0.0

01

1In

clu

des

no

n-c

arp

cyp

rin

ids

wit

hsc

ale

len

gth

s$

2.50

mm

2In

clu

des

cyp

rin

ids

wit

hsc

ale

len

gth

s,

2.50

mm

,ex

clu

din

gC

ypri

nu

sca

rpio

and

Rhi

nic

hthy

ssp

p.

3C

om

par

iso

nex

clu

ded

win

ter

TA

BL

E1.

—C

on

tin

ued


reduce variation in the body-scale regressions; see Appendix A for the regression modelsused, and Stearns 2008 for methods and analysis) and always were used when present in ascat. When a lateral line scale could not be located within a scat, a representative (what wesubjectively determined to be of average shape and size) non-lateral line scale of the species(or group) was used in models constructed with non-lateral line scales (see Stearns 2008 forregression models used). If scales of distinctly different sizes (but the same species) werepresent within a scat, multiple scales were taken and used for size estimation.

Fish size estimates were obtained for each of the scales, and 95% prediction intervals wereconstructed around the estimates. Doing so allowed multiple individuals of the same speciesto be identified in a single scat when the prediction intervals did not overlap. When multipleindividuals were detected within a scat each individual was included in prey size estimates.Fish prey size was evaluated by categorizing into six total length classes: #10.0 cm, 10.1–20.0 cm, 20.1–30.0 cm, 30.1–40.0 cm, 40.1–50.0 cm and .50.0 cm. Ten cm was chosen as theupper limit of the smallest size class because it is approximately the maximum length ofsmall cyprinid species (minnows). The size of fish prey is expressed using relative frequency(calculated as the number of fish in a size category divided by the total number of fish withestimated lengths). Due to their lack of scales the size of ictalurid prey was not analyzed.Furthermore, because of inadequate scale samples to construct body-scale regressions thesize of mooneyes (Hiodon spp.), quillback (Carpiodes cyprinus) and Ictiobus spp. was notestimated, but these species comprised a very small portion of the diet (Table 1).

Seasonal variation in the diet was assessed by comparing the occurrence of each prey itembetween seasons using chi-square (x2) analyses. Additionally, mean prey sizes werecompared among seasons and study areas through 1-way ANOVA. Chi-square analyses alsowere used to assess the seasonal variation in the relative frequency of prey size categories. Allstatistical analyses were conducted using Minitab (Minitab Version 14, Minitab Inc., StateCollege, Pennsylvania).

RESULTS

A total of 569 scats were collected and analyzed from our main study areas-142 scats fromthe Red River, 245 from the Forest River and 182 from the Turtle River. Additionally, 89scats were collected from the Tongue River. Due to seasonal variation in river otter scentmarking there was considerable variation in the number of scats collected among seasons,ranging from 22 in winter to 264 in fall (Table 1).

Fish and crayfish were the primary foods of river otters, occurring in 83.0% and 51.1% ofscats, respectively (Table 1). Insects also were relatively common in the diet, occurring in26.7% of scats. Other prey included birds (7.9%), amphibians (6.7%), mammals (6.0%) andfreshwater mussels (0.2%). Cyprinids were the most common fishes consumed (64.7%), andcarp (Cyprinus carpio) (55.5%) was the most frequently consumed species (Table 1). Otherfishes that were relatively common in the diet included ictalurids (17.4%), catostomids(13.0%) and centrarchids (11.2%). Other fish prey included: Percidae (4.9%), northernpike (Esox lucius, 4.9%), white bass (Morone chrysops) or freshwater drum (Aplodinotusgrunniens, 4.7%) and mooneyes (2.3%).

The diet of river otters changed seasonally (Table 1, Fig. 2). Fish were the most importantprey item throughout the year but became less important in summer (x2 5 40.10, 3 df, P ,

0.001) (Table 1). The frequency of occurrence of most fish families varied seasonally, withthe relatively common families in the diet (i.e., cyprinids, ictalurids and catostomids)tending to occur most often in fall or winter, and families consumed less often occurring athigher frequencies in spring (Table 1). In contrast to fish, crayfish increased in importance


in summer (x2 5 21.80, 3 df, P , 0.001) when they became the prey item with the highestfrequency of occurrence on the Turtle River (Fig. 2). Seasonal differences also weredetected in the occurrence of insects (x2 5 9.30, 3 df, P 5 0.026), birds (x2 5 15.39, 3 df, P5 0.002) and amphibians (x2 5 43.31, 3 df, P , 0.001) (Table 1).

The size of fish prey ranged from 3.5 (a carp) to 71.0 cm (a northern pike), with a meanof 20.7 cm (SE 6 0.5, n 5 658). Northern pike x̄ 5 36.7 cm, SE 6 2.2, n 5 34) and darters(Etheostoma spp. or Percina spp.) (x̄ 5 6.1 cm, SE 6 0.3, n 5 15) were on average the largestand smallest fish prey, respectively (Table 2). The size category with the highest relativefrequency (36.5%) in the diet was 10.1–20.0 cm total length. In order of relative frequencythe remaining fish were #10.0 cm (24.6%), 20.1–30.0 cm (14.1%), 30.1–40.0 cm (14.0%),40.1–50.0 cm (8.4%) and .50.0 cm (2.4%). Carp and catostomids were represented in allsize classes, whereas centrarchids, non-carp cyprinids and percids occurred predominatelyin the smaller size classes (Fig. 3).

The size of fish consumed by river otters changed seasonally, with mean prey size (F3,654 5

27.87, P , 0.001) and all size categories varying across seasons (Tables 2 and 3). Spring (x̄ 5

28.1 cm, SE 6 1.0, n 5 171) had the highest mean prey size and was followed by winter (x̄ 5

21.0 cm, SE 6 1.8, n 5 41), summer (x̄ 5 18.7 cm, SE 6 1.6, n 5 89) and fall (x̄ 5 17.6 cm, SE

6 0.6, n 5 357) (Table 2). In summer, most fish prey (46.1%) were represented in thesmallest size category (#10.0 cm) (Table 3). With increasing frequency of fish in the diet infall and winter, there was a shift towards fish 10.1–20.0 cm in length. In spring, fish 30.1–40.0 cm reached their seasonal maximum in relative frequency and were the most frequentsize in the diet (26.3%). Similarly, fish 20.1–30.0 cm, 40.1–50.0 cm, and .50.0 cm peaked inrelative frequency in spring.

FIG. 2.—Seasonal variation in the frequency of occurrence of fish and crayfish in the diet of riverotters on the Forest and Turtle rivers of eastern North Dakota. Scats collected 1 Mar.–31 May wereconsidered as spring, 1 Jun.–30 Aug. as summer, 1 Sep.–30 Nov. as fall and 1 Dec.–28 Feb. as winter


TA

BL

E2.

—Se

aso

nal

com

par

iso

no

fm

ean

tota

lle

ngt

h(c

m)

of

fish

con

sum

edb

yri

ver

ott

ers

on

the

Fo

rest

,R

ed,

and

Tu

rtle

rive

rso

fea

ster

nN

ort

hD

ako

ta,

Oct

.20

06–N

ov.

2007

.Sc

ats

coll

ecte

d1

Mar

.–31

May

wer

eco

nsi

der

edas

spri

ng,

1Ju

n.–

30A

ug.

assu

mm

er,

1Se

p.–

30N

ov.

asfa

llan

d1

Dec

.–28

Feb

.as

win

ter.

Bo

ldin

dic

ates

asi

gnif

ican

td

iffe

ren

ce(P

,0.

05)

Spri

ng

Sum

mer

Fal

lW

inte

rT

ota

l

FP

Mea

nSE

nM

ean

SEn

Mea

nSE

nM

ean

SEn

Mea

nSE

n

Cat

ost

om

idae

(su

cker

s)38

.73.

713

44.0

10.2

420

.21.

744

16.2

1.8

325

.31.

964

11.6

,0.0

01

Cat

osto

mu

sco

mm

erso

ni

(wh

ite

suck

er)

42.6

3.1

1152

.08.

93

22.2

3.1

2012

.7—

130

.92.

935

9.4

,0.0

01

Mox

osto

ma

spp

.(r

edh

ors

es)

17.3

2.4

219

.9—

118

.51.

724

18.0

0.3

218

.51.

429

0.0

0.99

4

Cen

trar

chid

ae(s

un

fish

)14

.80.

835

12.8

1.7

312

.11.

131

13.7

1.6

1013

.50.

679

1.3

0.28

5

Pom

oxis

spp

.(c

rap

pie

)15

.76.

63

——

014

.73.

17

19.1

1.8

215

.72.

312

0.2

0.82

0A

mbl

opli

tes

rube

stri

so

rL

epom

issp

p.

14.8

0.9

2612

.81.

73

10.7

1.6

1312

.51.

97

13.3

0.7

492.

10.

112

Un

kno

wn

cen

trar

chid

s14

.01.

66

——

012

.11.

711

11.8

—1

12.7

1.2

180.

30.

749

Cyp

rin

idae

(car

pan

dm

inn

ow

s)

Cyp

rin

us

carp

io(c

arp

)32

.21.

379

19.0

1.8

6218

.20.

821

123

.52.

421

21.6

0.7

373

27.4

,0.0

01

Oth

ercy

pri

nid

s22

.211

.52

8.4

0.6

138.

50.

534

6.2

—1

8.9

0.6

509.

8,

0.0

01

Eso

cid

ae

Eso

xlu

ciu

s(n

ort

her

np

ike)

41.0

2.3

1331

.014

.63

33.4

3.6

1540

.14.

13

36.7

2.2

341.

10.

369

Mo

ron

idae

/Sc

iaen

idae

Mor

one

chry

sops

(wh

ite

bas

s)o

rA

plod

inot

us

gru

nn

ien

s(f

resh

wat

erd

rum

)27

.41.

621

24.2

—1

37.1

3.3

624

.19.

42

29.0

1.6

302.

80.

061

Per

cid

ae(p

erch

es)

11.2

1.7

814

.41.

83

9.7

1.4

166.

9—

110

.51.

028

0.9

0.44

3

Eth

eost

oma

spp

.o

rP

erci

na

spp

.(d

arte

rs)

6.8

0.8

4—

—0

5.7

0.1

106.

9—

16.

10.

315

2.8

0.10

2P

erca

flav

esce

ns

or

San

der

spp

.15

.50.

34

14.4

1.8

316

.31.

06

——

015

.60.

613

0.7

0.53

0

To

tal

28.1

1.0

171

18.7

1.6

8917

.60.

635

721

.01.

841

20.7

0.5

658

27.9

,0.0

01


Although the diet was similar among study areas, some prey items occurred morefrequently on one river as compared to the others. For example, birds were more frequentlyconsumed on the Forest River (15.9%, compared with #2.2% on other rivers), and crayfishwere eaten most often on the Turtle River (79.7%, compared with 34.7% and 42.7% on theForest and Red Rivers, respectively) where they were consumed more frequently than fish(71.4%). The diet on the Tongue River was dominated by crayfish (96.6%) with fishoccurring relatively rarely (29.2%). Catfish were the most common fish consumed (9.0%)on the Tongue River, whereas carp (which were not documented in the diet on the TongueRiver) dominated the diet in the primary study areas.

DISCUSSION

The general food habits of river otters documented in this study are similar to thosereported elsewhere, with fish and crayfish being the primary prey items (e.g., Greer, 1955;Melquist and Hornocker, 1983; Serfass et al., 1990). For instance, in the primary study areasthe occurrence of fish ranged from 67.3% in summer to 100% in winter, values comparableto many previous studies at similar latitudes (Lagler and Ostenson, 1942; Greer, 1955;Ryder, 1955; Knudsen and Hale, 1968; Route and Peterson, 1988; Reid et al., 1994).Nonetheless, there were some results that differed from what would be expected based onprevious studies. Fish, for example, occurred in only 29.2% of scats on the Tongue River,with only Grenfell (1974) reporting a lower frequency of occurrence. Also, birds wererelatively frequent in the diet on the Forest River (15.9%), in comparison to 7.9% for theentire primary study area, and most previous studies where occurrence typically was ,5%

(e.g., Wilson, 1954; Hamilton, 1961; Melquist and Hornocker, 1983; Route and Peterson,1988; Serfass et al., 1990; Reid et al., 1994; Noordhuis, 2002).

FIG. 3.—The number of fish observed in the diet of river otters Oct. 2006–Nov. 2007 on the Forest,Red, and Turtle rivers of eastern North Dakota, by total length category (cm)


TA

BL

E3.

—Se

aso

nal

com

par

iso

no

fto

tal

len

gth

(cm

)ca

tego

ries

of

fish

con

sum

edb

yri

ver

ott

ers

on

the

Fo

rest

,R

edan

dT

urt

leR

iver

so

fea

ster

nN

ort

hD

ako

tab

yn

um

ber

of

fish

det

ecti

on

s(N

o.)

and

rela

tive

freq

uen

cy(R

F,e

xpre

ssed

asa

%).

Scat

sco

llec

ted

1M

ar.–

31M

ayw

ere

con

sid

ered

assp

rin

g,1

Jun

.–30

Au

g.as

sum

mer

,1

Sep

.–30

No

v.as

fall

and

1D

ec.–

28F

eb.

asw

inte

r.B

old

ind

icat

esa

sign

ific

ant

dif

fere

nce

(P,

0.05

)

To

tal

Len

gth

(cm

)

Spri

ng

(n5

171)

Sum

mer

(n5

89)

Fal

l(n

535

7)W

inte

r(n

541

)T

ota

l(n

565

8)

x2

PN

o.

RF

No

.R

FN

o.

RF

No

.R

FN

o.

RF

#10

127.

041

46.1

104

29.1

512

.216

224

.657

.9,

0.0

01

10.1

–20

4224

.620

22.5

159

44.5

1946

.324

036

.529

.7,

0.0

01

20.1

–30

4124

.06

6.7

3910

.97

17.1

9314

.121

.0,

0.0

01

30.1

–40

4526

.311

12.4

287.

88

19.5

9214

.034

.1,

0.0

01

40.1

–50

2313

.57

7.9

236.

42

4.9

558.

48.

20.0

43

$50

84.

74

4.5

41.

10

0.0

162.

48.

80.0

31


River otters have been reported to capture prey in proportion to the prey’s availability andinversely with the prey’s swimming ability (Ryder, 1955). Accordingly, the most importantfish in the diet (i.e., cyprinids, ictalurids, catostomids and centrarchids) are relatively slowswimmers and were the most numerous fishes (in the same order of importance) insampling we conducted in the Forest and Turtle river study areas (Stearns, 2008). Similarly,these fishes are typically reported as the most frequently occurring fishes in the diet of riverotters in other studies (e.g., Wilson, 1954; Greer, 1955; Hamilton, 1961; Griess, 1987; Serfasset al., 1990; Noordhuis, 2002; Giordano, 2005). However, cyprinids have been reported asthe most frequently occurring fish family in relatively few studies (i.e., Wilson, 1954;Hamilton, 1961; Griess, 1987).

Seasonal variation in the diet of river otters likely reflects changes in the availabilityand vulnerability of prey. This and previous studies have shown fish decreasing inoccurrence in summer with a corresponding increase in the occurrence of crayfish(Tumlison and Karnes, 1987; Serfass et al., 1990; Noordhuis, 2002; Giordano, 2005).Changes in the catchability of fish and crayfish may contribute to this transition in diet.For instance, fish are probably more difficult to capture in summer because fishswimming speeds increase as water temperature increases (Erlinge, 1968; Wardle, 1980).Also, and perhaps a greater contributing factor, crayfish activity and overall availabilityincreases in warmer water temperatures, potentially making them more vulnerable topredation in summer (Flint, 1977). This shift in the vulnerability of prey likelycontributed to the predominance of crayfish (and relative scarcity of fish) in the diet onthe Tongue River (because the majority of scats were collected in summer) and on theTurtle River in summer (Fig. 2).

The breeding season of potential prey likely influences their vulnerability. For example,fish become concentrated during spawning and, therefore, are probably more vulnerable topredation. In this study, despite total fish consumption being higher in fall, fish that wereconsumed relatively rarely (e.g., centrarchids, northern pike and percids) occurred moreoften in spring, which corresponds with their spawning period (Lee et al., 1980; Koel andPeterka, 2003; Werner, 2004). More commonly consumed fishes (i.e., carp, ictalurids andcatostomids) also spawn at this time, but because they are the most abundant fishes in thestudy area they were consumed relatively frequently throughout the year (Lee et al., 1980;Koel and Peterka, 2003; Werner, 2004). Similarly, birds and amphibians were consumedmost frequently in spring and summer, which corresponds to seasonally high abundanceand with higher activity levels associated with the breeding season. The higher frequency ofoccurrence of birds on the Forest River in comparison to the other study areas and previousstudies (only Grenfell (1974) and Gilbert and Nancekivell (1982) have reported higheroccurrences) is likely the result of a high abundance of breeding waterfowl in the area(Stearns, 2008).

The fish prey of river otters ranged from 3.5 to 71.0 cm total length with most being#30.0 cm, which is similar to the few previous reports of river otter prey size (e.g., Lagler andOstenson, 1942; Greer, 1955; Ryder, 1955; Hamilton, 1961; Melquist and Hornocker, 1983;Griess, 1987; Giordano, 2005). The only other study to thoroughly examine prey size(Giordano, 2005) reported similar relative frequencies as this study, including fish 10.1–20.0 cm as the most frequent prey size. Also, fish 30.1–40.0 cm, 40.1–50 cm, and .50.0 cmoccurred in similar proportions in both studies. However, Giordano (2005) reported amuch lower frequency of fish #10.0 cm (10.3%, compared with 24.4% in this study) and ahigher frequency of fish 20.1–30.0 cm.


Many factors likely influence the likelihood that a fish of a particular length will beconsumed. In particular, the abundance and species composition of the size class may playan important role. For instance, more abundant size classes of a fish species would suggestgreater encounter rates with river otters. Also, if a size class is dominated by slow swimmingspecies it would be expected to be more vulnerable to predation (Ryder, 1955). Otherfactors that could influence the likelihood of a fish being taken include its nutritive value,swimming speed and detectability. Large individual fish are a more concentrated source ofcalories than smaller individuals of the same species, therefore, making it more profitableto catch the larger fish, if catchability is similar. However, fish of different sizes likely dohave differential catchability because large fish swim at faster speeds than smallerindividuals of the same species (Rowe-Rowe, 1977; Videler, 1993). Detectability also likelyplays a role in determining the size of the fish that a river otter will pursue while foraging.Because of their size, small fish are less readily detectable and, therefore, may be less likelyto be pursued. Furthermore, there are probably more hiding spaces available to small fishseeking refuge from predation, possibly making small fish less available to river otters thanwhat their abundance would suggest. Accordingly, Adrian and Delibes (1987) reportedEuropean otters (Lutra lutra) capturing small fish less than expected based on theirabundance, and Erlinge (1968) noted captive European otters taking fish #10.0 cm withdifficulty. Although fish in the #10.0 cm size category occurred second most frequently inour study (24.6%, Table 3), this size class was represented much more frequently (61.3% offish collected) in fish sampling we conducted in the Forest and Turtle River study areas (seeStearns, 2008).

There are limitations to using scales to estimate the size of fish prey. Knowing the actualnumber of fish of a particular species a scat contains is difficult because of wide predictionintervals around fish size estimates. Therefore, remains of multiple individuals occurring ina scat generally could only be differentiated for species that attain large sizes (e.g., carp andnorthern pike). Consistently underestimating the number of small fish within a scat (i.e., notdetecting all fish present) would cause an overall overestimation of mean prey size andconversely overestimating the number of small fish (i.e., detecting more fish than the actualamount) would lead to an underestimation of mean prey size. Therefore, because a scatcould contain more individuals of smaller lengths, we suspect that the relative frequency ofsmaller size classes are underestimated, leading to an overestimation of the importance oflarger size classes and mean prey size.

Previous studies on the diet of river otters and European otters have shown that prey sizeis largest in spring or winter (Erlinge, 1968; Carss et al., 1990; Kozena et al., 1992; Dolloff,1993; Giordano, 2005). Similarly, in our study, river otters consumed larger prey in springthan in other seasons. The increased vulnerability of large fish during spawning is apotential contributing factor to the observed increase in river otter prey size. Swimmingspeed also may be an influence on prey selection. Large fish may be more easily capturedand handled during periods of colder water because of lowered activity levels (i.e., theirmetabolism is lowered and swimming speed reduced) (Wardle, 1980).

Acknowledgments.—We would like to thank the North Dakota Game and Fish Department forproviding funding (through a State Wildlife Grant) and assistance throughout the project. Thankyou to South Dakota State University researchers Lucas Borgstrom and Cari-Ann Hayer for allowingus to join them during their fish surveys so that we could obtain scale samples. Thanks to TurtleRiver State Park (Arvilla, ND) for providing housing. Finally, thank you to our field and lab crew,including Zach Olson, Brenda Curry, Sarah Negium, Lou Allard, Melissa Smith, Steven Loughery,Thomas Baden and Melissa Brannon.


LITERATURE CITED

ADAMS, A. W. 1961. Furbearers of North Dakota. North Dakota Game and Fish Department, Bismarck,North Dakota.

ADRIAN, M. I. AND M. DELIBES. 1987. Food habits of the otter (Lutra lutra) in two habitats of the DonanaNational Park, SW Spain. J. of Zool. (Lond.), 212:399–406.

BAILEY, V. 1926. A biological survey of North Dakota. USDA Bureau of Biological Survey. N. Am. Fauna.No. 49.

CARSS, D. N., H. KRUUK AND J. W. H. CONROY. 1990. Predation on adult Atlantic salmon, Salmo salar L., byotters, Lutra lutra L., within the River Dee system, Aberdeenshire, Scotland. J. of Fish Biol.,37:935–944.

COPES, F. A. AND R. A. TUBB. 1966. Fishes of the Red River tributaries in North Dakota. Contrib. Inst. Ecol.Stud. No. 1.

DANIELS, R. A. 1996. Guide to the identification of scales of inland fishes of northeastern North America.New York State Mus. Bull. No. 488. The University of New York, Albany, New York. 97 p.

DOLLOFF, C. A. 1993. Predation by river otter (Lutra canadensis) on juvenile coho salmon (Oncorhynchuskisutch) and dolly varden (Salvelinus malma) in southeast Alaska. Can. J. Fish. Aqua. Sci.,50:312–315.

ERLINGE, S. 1968. Food studies on captive otters Lutra lutra. Oikos, 19:259–270.FLINT, R. W. 1977. Seasonal activity, migration, and distribution of the crayfish, Pacifastacus leniusculus, in

Lake Tahoe. Am. Midl. Nat., 97:280–292.GILBERT, F. F. AND E. G. NANCEKIVELL. 1982. Food habits of the mink (Mustela vison) and otter (Lutra

canadensis) in northeastern Alberta. Can. J. of Zool., 60:1282–1288.GIORDANO, A. J. 2005. Feeding ecology of a reintroduced river otter (Lontra canadensis) population in

northcentral Pennsylvania. Thesis, Frostburg State University, Frostburg, Maryland.GREER, K. R. 1955. Yearly food habits of the river otter in the Thompson Lakes Region, northwest

Montana, as indicated by scat analysis. Am. Midl. Nat., 54:299–313.GRENFELL, W. E., JR. 1974. Food habits of the river otter in Suisin Marsh, central California. Thesis,

California State University, Sacramento, California.GRIESS, J. M. 1987. River otter reintroduction in Great Smokey Mountains National Park. Thesis,

University of Tennessee, Knoxville, Tennessee.HAGEN, S. K., P. T. ISAKSON AND S. R. DYKE. 2005. North Dakota comprehensive wildlife conservation

strategy. North Dakota Game and Fish Department, Bismarck, North Dakota.HAMILTON, W. J., JR. 1961. Late fall, winter, and early spring foods of 141 otters from New York. New York

Fish Game J., 8:106–109.KNUDSEN, G. J. AND J. B. HALE. 1968. Food habits of otters in the Great Lakes region. J. Wldlf. Manag.,

32:89–93.KOEL, T. M. AND J. J. PETERKA. 1998. Stream fishes of the Red River of the North basin, United States: a

comprehensive review. Can. Field Nat., 112:631–646.——— AND ———. 2003. Stream fish communities and environmental correlates in the Red River of the

North, Minnesota and North Dakota. Envir. Biol. Fish., 67:137–155.KOZENA, I., P. URBAN AND I. STOURACOVA. 1992. The diet of the otter (Lutra lutra Linn.) in the Polana

protected landscape region. Folia Zool., 41:107–122.LAGLER, K. F. AND B. T. OSTENSON. 1942. Early spring food of the otter in Michigan. J. Wldlf. Manag.,

6:244–254.LEE, D. S., C. R. GILBERT, C. H. HOCUTT, R. E. JENKINS, D. E. MCALLISTER AND J. R. STAUFFER, JR. 1980. Atlas of

North American freshwater fishes. North Carolina State Museum of Natural History, Raleigh,North Carolina. 854 p.

MELQUIST, W. E. AND M. G. HORNOCKER. 1983. Ecology of river otters in west central Idaho. Wldlf. Monog.,83:1–60.

NOORDHUIS, R. 2002. The river otter (Lontra canadensis) in Clarke County (Georgia, USA): survey, foodhabits, and environmental factors. IUCN Otter Spec. Group Bull., 19:75–86.

REID, D. G., T. E. CODE, A. C. H. REID AND S. M. HERRERO. 1994. Food habits of the river otter in a borealecosystem. Can. J. of Zool., 72:1306–1313.


ROUTE, W. T. AND R. O. PETERSON. 1988. Distribution and abundance of river otters in Voyageurs NationalPark, Minnesota. National Park Service, Omaha, Nebraska.

ROWE-ROWE, D. T. 1977. Variations in the predatory behavior of the clawless otter. Lammergeyer, 23:22–27.RYDER, R. A. 1955. Fish predation by the otter in Michigan. J. of Wldlf. Manag., 19:497–498.SERFASS, T. L., L. M. RYMON AND R. P. BROOKS. 1990. Feeding relationships of river otters in northeastern

Pennsylvania. Trans. Northeast Sect. Wldlf. Soc., 47:43–53.SPIERS, J. K. 1973. A microscopic key to the hairs of Virginia land mammals. Thesis, Virginia Polytechnic

Institute and State University, Blacksburg, Virginia.STEARNS, C. R. 2008. Food habits of river otters (Lontra canadensis) in the Red River of the North drainage

of eastern North Dakota. Thesis, Frostburg State University, Frostburg, Maryland.STENSON, G. B., G. A. BADGERO AND H. D. FISHER. 1984. Food habits of the river otter Lutra canadensis in the

marine environment of British Columbia. Can. J. of Zool., 62:88–91.STONER, J. D., D. L. LORENZ, G. J. WICHE AND R. M. GOLDSTEIN. 1993. Red River of the North basin,

Minnesota, North Dakota, and South Dakota. Water Res. Bull., 29:575–615.TOWEILL, D. E. 1974. Winter food habits of river otters in western Oregon. J. Wldlf. Manag., 38:107–111.TUMLISON, R. AND M. KARNES. 1987. Seasonal changes in food habits of river otters in southwestern

Arkansas beaver swamps. Mammalia, 51:225–231.VIDELER, J. J. 1993. Fish swimming. Chapman and Hall Publishing, New York, New York. 260 p.WARDLE, C. S. 1980. Effects of temperature on the maximum swimming speed of fishes, p. 519–531. In:

M. A. Ali (ed.). Environmental Physiology of Fishes. Plenum Press, New York, New York.WERNER, R. G. 2004. Freshwater fishes of the northeastern United States. Syracuse University Press,

Syracuse, New York. 335 p.WILSON, K. 1954. The role of mink and otter as muskrat predators in northwest North Carolina. J. Wldlf.

Manag., 18:199–207.

SUBMITTED 16 OCTOBER 2009 ACCEPTED 22 JULY 2010

APPENDIX A

The body-lateral line scale relationships used in estimating the size of fish consumed byriver otters (Lontra canadensis) on the Forest, Red, and Turtle rivers of eastern North Dakota.Scale samples were obtained throughout the Red River tributaries of North Dakota, Jun.–Oct. 2007. Some models represent individual species, whereas others include multiplespecies that could not reliably be differentiated by their scales. The relationships presentedwere constructed using the scale measurement that resulted in the best model for eachspecies (or group) of six measurements that were considered. Refer to Stearns (2008) forthe methods of establishing the relationships, and body-scale relationships of non-lateralline scales.


food habits and fish prey size selection of a newly colonizing population of river otters (lontra...

Documents