Dangerous prey and daring predators: a review

Biol. Rev. (2013), 88, pp. 550–563. 550doi: 10.1111/brv.12014

Dangerous prey and daring predators:a review

Shomen Mukherjee∗,† and Michael R. HeithausDepartment of Biological Sciences, Florida International University, 3000 NE 151ST, North Miami, FL 33154, USA

ABSTRACT

How foragers balance risks during foraging is a central focus of optimal foraging studies. While diverse theoretical andempirical work has revealed how foragers should and do manage food and safety from predators, little attention hasbeen given to the risks posed by dangerous prey. This is a potentially important oversight because risk of injury cangive rise to foraging costs similar to those arising from the risk of predation, and with similar consequences. Here, wesynthesize the literature on how foragers manage risks associated with dangerous prey and adapt previous theory tomake the first steps towards a framework for future studies. Though rarely documented, it appears that in some systemspredators are frequently injured while hunting and risk of injury can be an important foraging cost. Fitness costs offoraging injuries, which can be fatal, likely vary widely but have rarely been studied and should be the subject of futureresearch. Like other types of risk-taking behaviour, it appears that there is individual variation in the willingness totake risks, which can be driven by social factors, experience and foraging abilities, or differences in body condition.Because of ongoing modifications to natural communities, including changes in prey availability and relative abundanceas well as the introduction of potentially dangerous prey to numerous ecosystems, understanding the prevalence andconsequences of hunting dangerous prey should be a priority for behavioural ecologists.

Key words: foraging behaviour, predator injury, predator-prey interaction, prey avoidance, risk of injury.

CONTENTS

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 550II. What makes prey dangerous? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551

III. Frequency of injuries from dangerous prey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551IV. Ethology of reducing the risk of injury . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 555V. Costs of hunting dangerous prey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556

VI. A framework for investigating foraging on dangerous prey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556(1) Interpreting the foraging costs of hunting dangerous prey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556(2) Is there intraspecific variation in willingness to take risks? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 558

VII. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561VIII. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561

IX. References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561

I. INTRODUCTION

Studies of predator–prey interactions continue to be oneof the most fascinating and important aspects of ecologicalresearch. The intense focus on this topic can be attributedto the central role of foraging in the lives of predatorsand their prey, and the importance of predation in driving

* Address for correspondence (Tel: +27 714390560; E-mail: [email protected])† Present address: School of Life Sciences (Biological & Conservation Sciences), University of KwaZulu-Natal, Westville Campus, PO

Box 54001, Durban 4000, Republic of South Africa.

population, community, and evolutionary dynamics. Morerecently, behavioral ecologists have begun to investigatethe population and ecosystem consequences of predators inmodifying the behaviour of their prey [see Lima & Dill (1990),Lima (1998) and Brown & Kotler (2004) for reviews], whichcan have profound consequences for prey populations andthe dynamics of their communities [see reviews by Werner

Biological Reviews 88 (2013) 550–563 © 2013 The Authors. Biological Reviews © 2013 Cambridge Philosophical Society

Daring predators 551

& Peacor (2003), Schmitz, Krivan & Ovadia (2004), Preisser,Bolnick & Benard (2005), Heithaus et al. (2008) and Creel &Christanson (2008)].

Predator–prey studies, especially of prey choice bypredators, are becoming increasingly important due toanthropogenic modification of ecosystems. For example,changes in energetic demands instigated by climate changecould modify predator foraging needs and decisions, ascould the introduction of exotic prey species or reductions innaturally important prey. Predators have more at stake whilehunting than simply the risk of missing a meal if unsuccessful.Some potential prey may harm or even kill their predator, orthe habitat in which particular prey are found may pose aninjury or mortality risk to a predator. While the potential forpoisonous prey to harm predators has been widely consideredin studies of diet choice, as has foraging under the risk ofpredation, there has been less attention focused on foragingbehaviours and decisions of predators hunting other types ofdangerous prey.

The ability of predators to recognize dangerous prey mayvary widely depending on the novelty of such prey (e.g. exoticprey). Regardless, the decisions of when to attack or avoidsuch prey and the potential costs of doing so could be animportant aspect of ecological dynamics. Here, we reviewour current understanding of interactions between dangerousprey and their predators, and suggest a framework for futurestudies of such situations.

II. WHAT MAKES PREY DANGEROUS?

Prey exhibit a wide array of secondary defences that maybe physical, chemical, or behavioural (Edmunds, 1974;Ruxton, Speed & Kelly, 2004, Table 1). Weapons anddefensive structures include hooves, spines, flukes, horns,tusks, shell, teeth, and noxious chemicals among others.While some structures are permanent body parts, others areinduced in response to increased predation risk. For example,Daphnia lumholtzi has a rounded head when planktivorousfish are rare (e.g. during the winter months) or absent, buthave a sharp helmet and long extended tail spine, whichreduces predation, when predators are present (Greene,1967; Agrawal, 2001). Although some structures used indefence by prey have likely evolved for other purposes suchas intraspecific conflicts (e.g. sexual selection etc.; Edmunds,1974), they still may serve an important anti-predator roleand pose substantial risks to predators.

Certain prey use chemicals to deter or injure predators(i.e. for more than modifying their palatability). Forexample African bombardier beetles (Stenaptinus insignis) sprayhot quinonoid liquid towards attacking predators (Eisner& Aneshansley, 1999). In deep-sea habitats, consumingbioluminescent prey can increase the chances of a predatorbeing attacked by their own predators (Haddock, Moline &Case, 2010). Prey behaviour, such as kicking and biting whenattacked or counter-attacking predators, also may influencetheir danger to predators. For example, female white-tailed

deer (Odocoileus virginianus) will face and attack coyotes todeter them (Garner & Morrison, 1980) and honey bees maysmother and kill predatory wasps that try to attack their hive(Ono et al., 1995).

The overall level of danger posed by a particular preytype may extend beyond its own attributes. Indeed, habitatcharacteristics may influence injury and mortality risk topredators as could the particular tactics used to hunt prey,but these sources of danger to predators have been largelyoverlooked with respect to injury risk. For example preysuch as Nubian ibex (Capra ibex; Kotler, Gross & Mitchell,1994) and elk (Cervus elaphus; Mao et al., 2005) take shelter onsteep cliff faces to take advantage of its inaccessibility to theirpredators. Similarly, a fast pursuit in an undulating terraincould put a predator at risk of a fall causing fracture or limbdislocation. Risk of injury could be one of the reasons whypredators such as cougars (Dickson, Jenness & Beier, 2004)prefer to hunt in less undulating terrain. Physical threats inhabitats can include the presence of dense thickets or thornyvegetation, that can inflict debilitating injuries, particularlyto eyes (e.g. in owls, see Holt & Layne, 2008) or limbs.The presence of a predator’s own predator in the habitat isanother component of the risks of hunting particular prey.The latter situation has been the subject of considerableresearch (e.g. Lima & Dill, 1990; Donadio & Buskirk, 2006;Frid, Burns & Baker, 2009) and is not considered furtherhere.

III. FREQUENCY OF INJURIES FROMDANGEROUS PREY

There is little information on how frequently predatorsare injured by their prey (Table 1). This can be partiallyattributed to the difficulty in measuring prey-inflicted injurieswithout close inspection. Therefore, studies of injuries topredators generally are limited to studies of hard parts(e.g. teeth, bones) and therefore surely underestimate actualinjury rates to predators. To our knowledge, no studies haveestimated mortality rates from foraging-related injuries.

Injuries have been relatively well studied in raptors andcarnivores, and can be fairly common (Table 1). Dislocationof joints (Ackermann & Redig, 1997), and broken toes, talons,flight feathers, and injured eyes are widely documentedin raptors. A study in Canada found that 5.9% of 1120American kestrels had hunting-related injuries (Murza,Bortolotti & Dawson, 2000). In another study 14% of 98individuals sampled from three species of raptors in northernArkansas, USA, had injuries (Bedrosian & St. Pierre, 2007).Inspection of museum specimens of accipter hawks have alsorevealed that hunting injuries may be relatively common(18.6% of 339 individuals) (Roth, Jones & French, 2002).

Carnivores also appear to be injured frequently by theirprey. High rates of fractured canines were recorded formany species of carnivores (see table 3 in Van Valkenburg,2009). These included 5.4% of lions (Panthera leo), 9.2%of tigers (Panthera tigris), 9.8% of leopards (Panthera pardus),


552 Shomen Mukherjee and Michael R. Heithaus

Tab

le1.

Are

pres

enta

tive

listo

fqua

ntita

tive

stud

ies

and

anec

dota

lobs

erva

tions

cont

aini

ngev

iden

ceof

inju

ry,a

ndri

sk-r

educ

ing

beha

viou

rs,o

fpre

dato

rshu

ntin

gda

nger

ous

prey

Pred

ator

taxa

Prey

Pred

ator

inju

ry

Ris

k-re

duci

ngbe

havi

our

ofpr

edat

or

Typ

eof

evid

ence

(qua

ntita

tive

stud

yor

anec

dota

levi

denc

e)R

efer

ence

s

Inve

rteb

rate

sSe

ast

ar(L

epta

ster

ias

hexa

ctis

)Sn

ail(

Am

phis

saco

lum

bian

a)Im

mob

iliza

tion

ofse

ast

arar

mdu

eto

bite

inju

ryon

radi

alne

rve

Mos

tlyav

oida

nce.

Sea

star

drap

esits

elft

ight

ly,a

ndm

anoe

uvre

sits

arm

sso

that

the

snai

lcan

not

exte

ndits

prob

osci

s

Qua

ntita

tive

Bra

ithw

aite

etal

.(20

10)

Snai

l(Sin

istr

oful

gur

sini

stru

m)

Biv

alve

(Mer

cena

ria

mer

cena

ria)

Dam

age

thei

row

nsh

ell

Snai

lssh

ould

dela

yat

tack

ing

larg

ebi

valv

esun

tilth

eyre

ach

the

shel

l-thi

cken

ing

stag

e.

Qua

ntita

tive

Die

tl(2

003)

Cru

stac

ean

(Can

cer

mag

iste

)C

lam

(Pro

toth

aca

stam

ina)

Cla

wda

mag

ePr

efer

red

clam

sof

the

smal

lest

size

clas

sQ

uant

itativ

eJu

anes

&H

artw

ick

(199

0)A

nts

Afr

ican

bom

bard

ier

beet

le(S

tena

ptin

usin

sign

is)

Che

mic

alir

rita

nt(p

-ben

zoqu

inon

es)

Avo

id/r

elea

sepr

ey∗

Qua

ntita

tive

Eis

ner

&A

nesh

ansle

y(1

999)

Ant

lion

(Myr

mel

eon

caro

linu

s)A

nt(C

ampo

notu

sflo

rida

nus)

Che

mic

alir

rita

nt(fo

rmic

acid

)Fe

eds

onan

tsw

ithou

tru

ptur

ing

the

acid

sac

Qua

ntita

tive

Eis

ner

etal

.(19

93)

Spid

ers(

Ara

neus

trifol

ium

and

Arg

iope

trifas

ciat

a)B

eean

dw

asp

Stin

g∗A

void

phys

ical

cont

acta

ndus

ew

ebto

subd

ueda

nger

ous

prey

Qua

ntita

tive

Oliv

e(1

980)

Hor

net(

Ves

pam

anda

rina

)B

ee(A

pis

cera

na)

Let

hals

ting

Hor

nets

atta

ckm

ostly

duri

ngau

tum

n,w

hen

they

need

extr

apr

otei

nto

rear

new

quee

nsan

dm

ales

Qua

ntita

tive

Ono

etal

.(19

95)

Afr

ican

pone

rine

ant

(Pac

hyco

ndyl

apa

chyd

erm

a)

Cen

tiped

eB

ite∗

Cen

tiped

eshe

ldfr

oman

teri

orpa

rtof

the

body

,an

dst

ung

onve

ntra

lfac

ew

here

neur

alch

ain

pass

es

Qua

ntita

tive

Dej

ean

&L

acha

ud(2

011)

Zel

uslo

ngip

esC

ater

pilla

r(S

podo

pter

afr

ugip

erda

)

Bite

∗A

void

sat

tack

ing

med

ium

and

larg

eca

terp

illar

sQ

uant

itativ

eC

ongi

,Fri

tus

&Fi

lho

(200

2)

Fish

esG

reat

whi

tesh

ark

(Car

char

odon

carc

hari

s)C

alifo

rnia

sea

lion

Bite

∗B

itean

dre

leas

epr

eyA

necd

otal

Tri

cas

&M

cCos

ker

(198

5)B

lueg

illsu

nfish

(Lep

omis

mac

roch

irus

)D

aphn

ia(D

aphn

ialu

mho

ltzi

)D

aphn

iasp

ine

lodg

edon

the

top

offis

hm

outh

Avo

idan

cebe

havi

our.

Ori

enta

teda

phni

asu

chth

atth

ean

teri

orsp

ine

ente

rsth

ebu

ccal

cavi

tyfir

st

Qua

ntita

tive

Swaf

fer

&O

’Bri

en(1

996)



Tab

le1.

(con

t.)

Pred

ator

taxa

Prey

Pred

ator

inju

ry

Ris

k-re

duci

ngbe

havi

our

ofpr

edat

or

Typ

eof

evid

ence

(qua

ntita

tive

stud

yor

anec

dota

levi

denc

e)R

efer

ence

s

Rep

tiles

Kom

odo

mon

itor

(Var

anus

kom

odoe

nsis

)W

ater

buffa

loes

Phys

ical

trau

ma∗

Qui

ckat

tack

onbu

ffalo

leg,

tote

arte

ndon

sA

necd

otal

Auf

fenb

erg

(198

1)

Bur

ton’

sle

gles

sliz

ard

(Lia

lis

burt

onis

)W

ater

skin

ks(E

ulam

prus

heat

wol

ei)

Bite

∗L

arge

prey

alw

ays

atta

cked

onhe

ad,o

rne

ck,t

opr

even

tthe

mfr

ombi

ting.

Pred

ator

can

retr

acti

tsey

es.

Qua

ntita

tive

Wal

l&Sh

ine

(200

7)

Nor

ther

nPa

cific

ratt

lesn

akes

(Cro

talu

sor

egan

us)a

ndM

alay

pit-

vipe

rs(C

allo

sela

sma

rhod

osto

ma)

Mic

eB

ite∗

Rel

ease

larg

em

ice

afte

ra

pred

ator

yst

rike

buto

ften

reta

insm

alle

rin

divi

dual

sin

thei

rja

ws

Qua

ntita

tive

Kar

dong

(198

6)an

dB

arr,

Wie

burg

&K

ardo

ng(1

988)

Ret

icul

ate

pyth

on(P

ytho

nre

ticu

latu

s)Po

rcup

ine

(Hys

trix

brac

hyur

a)Ph

ysic

alin

jury

from

quill

s∗A

void

ance

∗A

necd

otal

Shin

eet

al.(

1998

)

Bir

dsPr

edat

ory

bird

s(e

.g.

kitt

iwak

e–

Ris

satr

idac

tyla

,her

ring

gull

–L

arus

arge

ntat

us)

Fulm

ars

(Gen

us–

Ful

mar

us,D

aption

,T

hala

ssoi

ca,

Pag

odro

ma,

Mac

rone

ctes

Pred

ator

isdr

ench

edw

ithoi

l,w

hich

mat

tsth

eir

feat

hers

and

dest

roys

insu

latio

n

Unk

now

nA

necd

otal

War

ham

(197

7)

Rap

tors

(Fal

cope

regr

inus

,O

tus

asio

,F

alco

colu

mba

rius

)

Unk

now

nL

uxat

ion

ofth

eel

bow

Unk

now

nA

necd

otal

Ack

erm

ann

&R

edig

(199

7)

Rap

tors

(But

eoja

mai

cens

is,F

alco

spar

veri

us,A

ccip

iter

coop

erii

)

Unk

now

nM

issi

ngta

lons

,m

issi

ngto

es,

win

gfr

actu

re,i

ris

dam

age

Unk

now

nA

necd

otal

Bed

rosi

an&

St.P

ierr

e(2

007)

Bar

red

owl(

Str

ixva

ria)

Rod

ent

Bite

/dea

th∗

Unk

now

nA

necd

otal

Gib

son

etal

.(19

98)

Lon

g-ea

red

owl(

Asi

oot

us)

Unk

now

nE

yein

jury

(from

hunt

s)U

nkno

wn

Ane

cdot

alH

olt&

Lay

ne(2

008)

Am

eric

anke

stre

l(F

alco

spar

veri

us)

Rod

ent(

Mus

mus

culu

s)B

ite∗

Bir

dsin

good

cond

ition

avoi

ded

dang

erou

spr

eyQ

uant

itativ

eM

urza

etal

.(20

00)

Buz

zard

(But

eobu

teo

ovul

pinu

s)R

oden

tB

ite∗

Bir

dsin

good

cond

ition

avoi

ded

dang

erou

spr

eyQ

uant

itativ

ePe

arlm

an&

Tsu

rim

(200

8)R

apto

r(A

ccip

iter

stri

atus

,A

.co

oper

ii,A

.ge

ntilis

)U

nkno

wn

Frac

ture

inpe

ctor

algi

rdle

Unk

now

nA

necd

otal

Rot

het

al.(

2002

)



Tab

le1.

(con

t.)

Pred

ator

taxa

Prey

Pred

ator

inju

ry

Ris

k-re

duci

ngbe

havi

our

ofpr

edat

or

Typ

eof

evid

ence

(qua

ntita

tive

stud

yor

anec

dota

levi

denc

e)R

efer

ence

s

Mam

mal

sK

iller

wha

le(O

rcin

usor

ca)

Stin

gray

(Das

yatis

sp.)

Dea

thA

void

ance

∗A

necd

otal

Dui

gnan

etal

.(20

00)

Chi

mpa

nzee

(Pan

trog

lody

tes)

Prim

ates

(Col

obus

and

Cer

copi

thec

ussp

ecie

s)Po

tent

ialf

orph

ysic

alin

jury

from

bite

∗

Gro

uphu

ntin

gQ

uant

itativ

eB

oesc

h(1

994)

and

Stan

ford

etal

.(19

94)

Tig

er(P

anth

era

tigr

is)

Porc

upin

eFa

cial

inju

ryfr

omqu

illA

void

ance

∗A

necd

otal

Cor

bett

(194

6,19

57)

Wild

dogs

(Lyc

aon

pict

us)

Ung

ulat

esD

eep

cuts

,bro

ken

teet

h,in

jure

dlim

bs

Gro

uphu

ntin

g,re

stra

inin

ghe

ad(e

.g.

whi

lehu

ntin

gho

gs)

Ane

cdot

alC

reel

&C

reel

(200

2)

Coy

ote

(Can

isla

tran

s)W

hite

taile

dde

er(O

doco

ileu

svi

rgin

ianu

s)Ph

ysic

alin

jury

∗A

void

ance

∗A

necd

otal

Gar

ner

&M

orri

son

(198

0)A

fric

anlio

n(P

anth

era

leo)

Ele

phan

t(L

oxod

onta

afri

cana

)Po

tent

ialf

orph

ysic

alin

jury

orev

ende

ath

∗

Gro

uphu

ntQ

uant

itativ

eL

over

idge

etal

.(20

06)

Wol

ves

(Can

islu

pus)

Elk

(Cer

vus

elap

us)

Pote

ntia

lfor

phys

ical

inju

ryor

even

deat

h∗

Lar

ger

body

size

ofw

olve

spr

ovid

edpr

edat

ory

adva

ntag

e

Qua

ntita

tive

Mac

Nul

tyet

al.(

2009

)

Mar

supi

als

(Das

yuru

svi

vern

nus,

Das

yuru

sha

lluc

atus

,D

asyu

roid

esby

rnei

and

Pha

scog

ale

tapo

ataf

a)an

ddo

mes

ticca

t(F

elis

catu

s)

Rod

ent

Bite

∗A

ttac

kfr

omba

ckor

flank

Qua

ntita

tive

Pelli

s&

Offi

cer

(198

7)

Mar

tin(M

arte

sam

eric

ana)

,ly

nx(L

ynx

cana

dens

is),

fishe

r(M

arte

spe

nnan

t)an

dre

dfo

x(V

ulpe

svu

lpes

)

Porc

upin

eQ

uills

embe

dded

inst

omac

hlin

ing

Avo

idan

ce∗

Ane

cdot

alQ

uick

(195

3)

Dir

ew

olf(

Can

isdi

rus),

coyo

te(C

anis

latr

ans),

sabr

e-to

othe

dca

t(S

milod

onfa

talis)

and

Am

eric

anlio

n(P

anth

era

atro

x)

56he

rbiv

ore

spec

ies

Frac

ture

dto

oth†

Unk

now

nQ

uant

itativ

eV

anV

alke

nbur

g&

Her

tel(

1993

)

Car

nivo

res

(36

spec

ies)

Not

men

tione

dFr

actu

red

toot

h†U

nkno

wn

Qua

ntita

tive

Van

Val

kenb

urg

(200

9)

∗ Dat

ano

tspe

cific

ally

prov

ided

inst

udy.

Est

imat

esar

eba

sed

onna

tura

lhis

tory

ofpr

edat

oran

dpr

ey.

† Too

thfr

actu

res

also

occu

rdu

eto

food

habi

ts(c

hew

ing

bone

s).



9.6% of spotted hyenas (Crocuta crocuta), 9.8% of grey wolves(Canis lupus), 17.3% of stoats (Mustela erminea) and 12% ofweasels (Mustela frenata). Since carnivores drive their caninesinto moving and struggling prey, the observed breakagerates are likely to be due to injuries sustained duringhunting (Van Valkenburg & Hertel, 1993). While these dataprovide compelling evidence that injuries during hunting arecommon in carnivores, they are certainly underestimates. Nodata are available for the proportion of individuals injuredor killed by prey through injuries to soft tissues or other partsof the body. Such injuries, however, are known to occur. Forexample, Creel & Creel (2002) found that African wild dogsmay incur deep cuts, broken teeth and injured limbs.

IV. ETHOLOGY OF REDUCING THE RISK OFINJURY

At one extreme, predators can simply avoid huntingdangerous prey even though they are capable of attackingand killing them. For example, a meta-analysis of leoparddiet studies suggests that they prey upon larger sized prey,such as Cape buffalo (Syncerus caffer), plains zebra (Equusquagga) and giraffes (Giraffa camelopardalis), less frequently thanexpected because they are too dangerous to hunt (Haywardet al., 2006). Some seas stars (Leptasterias hexactis) avoid snail(Amphissa columbiana) prey because they defend themselvesby biting into the radial nerve of the sea star, which canimmobilize an arm for several days (Braithwaite, Stone &Bingham, 2010). Orb-web spiders tend to grab and bitenon-venomous prey that gets caught in their web, but fordangerous prey (such as wasps and bees) they avoid physicalcontact and instead use webbing to wrap and subdue them(Olive, 1980, Fig. 1).

Predators can minimize their risk of prey-inflicted injuriesbehaviourally by minimizing contact or handling time (seeFig. 1; Table 1). By swiftly attacking and injuring dangerousprey, and then leaving it to die or become less active,predators can reduce their contact time with dangerousprey. For example Komodo dragons (Varanus komodoensis)target the legs of large and dangerous prey (e.g. waterbuffaloes). A sudden attack on the prey’s leg reduces thechances of their own injury, but leaves the prey crippled withtorn tendons and an infected wound, making them morevulnerable to future attacks (Auffenberg, 1981). Similarly,great white sharks (Carcharodon carcharias) appear to use a bite-and-spit (release) hunting tactic for dangerous Californiasea lions (Zalophus californianus), which results in the sea lionsbleeding to death (or going into shock) after an attack. Sharksdo not use this tactic for the less dangerous elephant seal(Mirounga angustirostris, Tricas & McCosker, 1985) and do notappear to use it when hunting smaller sea lions in otherlocations (Klimley, 1994; Klimley, Pyle & Anderson, 1996;Martin et al., 2005). Ant lions (Myrmeleon carolinus) are able tokill formic-acid-spraying ants (Camponotus floridanus) withoutinducing the ants to spray, and while feeding on the deadants they suck out the ant’s body contents without puncturing

0

10

20

30

40

50

60

Orthoptera Lepidoptera Diptera Orthoptera Lepidoptera Diptera

Num

ber

of a

ttac

ks

bite attacks wrap attacks

Araneus Argiope

Fig. 1. Differences in attack strategies upon safe and dangerousprey in orb web spiders. Spiders of the genera Araneus andArgiope are willing to risk direct contact (bite and inject venom)to capture large but safe prey (Lepidoptera and Diptera). Butthey avoid direct contact and instead use their web to wrap,subdue and hunt dangerous prey (Orthopera). Data from Olive(1980).

the ant’s formic acid sac (Eisner, Baldwin & Conner, 1993).Being more careful (i.e. increasing hunting or handling time)also can reduce the risk of injury. Grasshopper mice takesignificantly longer to subdue scorpions with neurotoxins(Centruroids spp.) compared to those species that do not havethese toxins (Vaejovis spp.; Rowe & Rowe, 2006).

The direction and position of an attack also plays animportant role in reducing the chances of injury. Forexample terrestrial predators such as domestic cats (Feliscatus) likely reduce their chances of injury by avoidinga frontal attack; instead they attack their prey from theback or the flanks (Pellis & Officer, 1987). Similar attackbehaviour has been observed in Asiatic lions (Leo leo persica)hunting water buffaloes in the Gir forests of western India(S. Mukherjee, personal communication). African wild dogs(Lycaon pictus) are able to successfully tackle armed prey suchas warthogs (Phacochoerus aethiopicus, which have dangeroustusks) only if they manage to restrain their head since itreduces their chances of injury from the warthog’s tusks(Creel & Creel, 2002). Australian limbless lizards (Lialisburtonis), lacking venom and constriction capabilities, useprey-size-specific hunting tactics. They almost always attacklarge lizard prey on the head or in the neck region therebyreducing their chances of being bitten (Wall & Shine, 2007).The predator lizards always wait for their larger lizard preyto be incapacitated before eating them, but consume smallerlizard prey immediately, even while it is still struggling (Wall& Shine, 2007). Finally, African ponerine ants (Pachycondylapachyderma) grab small and less dangerous prey (termites) bytheir thorax, but catch more dangerous prey (centipedes) bythe anterior part of their body and often sting them (Dejean& Lachaud, 2011).

Group hunting can reduce the per capita risk of an individualbeing injured when hunting large and dangerous prey. Forexample, in African wild dogs both hunting success andthe size of prey killed increases with pack size (Creel &Creel, 1995, 2002). Interestingly, however, the proportion



of the total attacks (i.e. frequency) that are on dangerousprey (e.g. adults of impala, Aepyceros melampus, wildebeest,Connochaetes taurinus, warthog, zebra) does not change withpack size (Creel & Creel, 2002). In African lions, althoughsociality is suggested to have evolved to reduce the probabilityof infanticide (Packer, Scheel & Pusey, 1990) and increasedterritory-holding potential (McComb, Packer & Pusey, 1994;Grinnell, Packer & Pusey, 1995), group living also facilitatescapturing large or dangerous prey (see Packer & Ruttam,1988). Scheel (1993) found that prides of five or morelionesses were more likely to attack prey such as buffaloes,topi (Damaliscus korrigum) or kongoni (Alcelaphus buselaphus)compared to smaller prides. Individual lions also prefer tojoin group hunts only when hunting large and dangerousprey such as zebra and Cape buffalo (Scheel & Packer,1991). Like mammals, social insects may also minimizerisk and maximize hunting success by working together.For example, African ponerine ants (Pachycondyla pachyderma)hunt small and less dangerous prey (termites) solitarily, butseveral workers cooperate when catching dangerous preysuch as Scolopendra centipedes (Dejean & Lachaud, 2011).Hence, in order to reduce the risk of injury, predators mayprefer to hunt in larger groups than would be optimal formaximizing energy intake rates alone. Although data to testthis hypothesis specifically are not currently available, therich literature on the advantages of group hunting amongpredators (e.g. Packer et al., 1990; Boesch, 1994; Creel &Creel, 1995) and group formation for defence among prey(e.g. Berger, 1979) suggests that such dynamics are likely.

V. COSTS OF HUNTING DANGEROUS PREY

Understanding the fitness consequences of huntingdangerous prey is critical for integrating these costs andbenefits into our understanding of ecological dynamics.Unfortunately, few studies have quantified such costs. Inpart, this is due to the wide variation in the costs of evensingle encounters with dangerous prey. Avoiding dangerousprey entirely only carries the cost of a lost resource of amissed successful hunt. When predators pursue dangerousprey, however, the costs can be extreme. Killer whales(Orcinus orca) have died from injuries inflicted from the spinesof stingrays (Duignan et al., 2000) and rays also have killeddolphins foraging on other prey in habitats with high raydensities (Walsh et al., 1988, McLellan, Thayer & Pabst,1996). Similarly, live rodent prey can inflict lethal internalinjuries to owls (Gibson, Gibson & Bardelmeier, 1998) andlarge carnivores in Africa can be killed or badly injuredby their ungulate prey (Creel & Creel, 2002). An injuredpredator can also become more vulnerable to its competitors.For instance hyenas may kill lions injured during hunts(Schaller, 1972).

Sublethal impacts also can be extreme. Predators attackinga porcupine can end up with quills embedded in their limbs,body, or face, which will negatively impact future foraging(Quick, 1953). While trying to open their bivalve prey,

some snails (Sinistrofulgur sinistrum) may damage their ownshell and become more vulnerable to their own predators(Dietl, 2003). Some of the most obvious sublethal costsof hunting dangerous prey are reduced foraging efficiencythrough increased handling time, energetic and lost foragingtime costs of recovering from injuries and decreased abilityto capture particular prey types. For example, in Dungenesscrabs (Cancer magister), claw damage (chela breakage andclaw-tooth wear) due to fatigue failure (injury due to repeatedforce cycles) is one of the most important factors determiningboth crab foraging efficiency and the size of the prey crabschoose (Juanes & Hartwick, 1990). Crabs with damaged clawsbecome less efficient in opening their bivalve prey (Protothacastamina, Juanes & Hartwick, 1990, Fig. 2). Furthermore, longperiods of starvation because of an injury can lead to changesin foraging behaviour and willingness to incur risks thatcan increase the probability of accruing additional injuriesduring foraging. Murza et al. (2000) found that handicapped(broken flight feather or missing talons) American kestrels(Falco sparverius) are less willing to risk injury or spend energyin hunting dangerous prey.

VI. A FRAMEWORK FOR INVESTIGATINGFORAGING ON DANGEROUS PREY

Under what conditions are predators more or less likelyto attack dangerous prey? Although few studies havefocused on this specific question, the diverse literatureon optimal diet theory and foraging under the risk ofpredation offer important insights. Although the diets ofpredators foraging on mobile prey may be determinedprimarily by the effectiveness of anti-predator behaviour (Sih& Christensen, 2001), profitability remains an importantconcept when investigating predator foraging choices. Inmost cases, profitability refers to the quotient of the netenergetic gain from consuming the resource divided byhandling time (the time required to pursue, capture andconsume prey). Profitability is used as a proxy of fitness basedon the assumption that maximizing net energy intake ratealso maximizes fitness (Stephens & Krebs, 1986). Obviously,when predators hunt dangerous prey, this formulation ofprofitability is inadequate since attacking a particular preyitem can incur fitness costs as outlined above (e.g. reducedforaging ability, early death). Indeed, the absence of suchforaging costs from foraging models may explain why prey-choice data from the field do not match the predictions ofsimple foraging models (Creel & Creel, 2002). Hence, thereis a need to consider the risk of injury in foraging models forpredators hunting behaviourally responsive and dangerousprey.

(1) Interpreting the foraging costs of huntingdangerous prey

Although a framework for understanding the costs of huntingdangerous prey might seem to derive easily from that of



1.5 2.0 2.5 3.0 3.5 4.0 4.5

Clam length (cm)

0

200

400

600

800

1000C

lam

bre

akin

g tim

e (s

)

worn

old

new

1.5 2.0 2.5 3.0 3.5 4.0 4.5

Clam length (cm)

0

200

400

600

800

1000

Cla

m e

atin

g tim

e (s

)

Fig. 2. Simulated regression equations from table 1 of Juanes& Hartwick (1990) show variations in clam breaking and eatingtime (the two components of handling time) of Dungeness crabswith ‘new’ (freshly moulted), ‘old’, and ‘worn’ (claw teeth fileddown) claws. Comparison of regression slopes (using equality-of-slope test) for both ‘old’ and ‘worn’ crabs to ‘new’ crabs showedsignificant differences in slope for eating time between ‘old’and ‘new’ crabs (P = 0.001). However for other comparisons,where slopes were homogeneous, analysis of covariance wasused to compare adjusted means between the groups. Thisshowed significant differences in eating time between worn–new(P=0.0001; higher in worn crabs), and also in breaking timebetween old–new (P = 0.003; higher in old) and worn–newcrabs (P = 0.006; higher in worn) over the range of clamstested. High breaking and eating times in ‘worn’ crabs can beconsidered similar to foraging costs of risk of injury to a predator.Since larger clams required more energy to open, the crabs alsohad an increased chance of claw muscle fatigue, hence higherrisk of injury.

hunting unpalatable or chemically defended prey, the costsare more likely to be akin to those of foraging under the riskof predation. This stems primarily from the uncertainty inthe consequences of hunting dangerous prey. The costs(e.g. sickness) of consuming chemically defended prey,as long as they are recognizable, are likely to be fairlypredictable. Hunting dangerous prey, however, may leadto a wide range of costs – from none at all to death –that may occur with different probabilities that foragersmay be able to assess. This is similar to making trade-offs

between foraging opportunities and reducing predationrisk.

Models addressing the foraging costs of predation risk canbe adapted to investigate the costs of foraging while huntingdangerous prey. Prey encounter rates and handling timeare critical factors influencing foraging decisions (Holling,1959; Schoener, 1971). A predator’s harvest rate (f ) can becalculated using Holling’s (1959) disc equation:

f = Ta × N

1 + (Ta × Tb × N )(1)

where, T a = prey encounter rate, T h = prey handlingtime and N = total number of prey. Dangerous prey canaffect the harvest rate in two ways. If predators take moretime to subdue and handle dangerous prey, T h will behigher (e.g. Juanes & Hartwick, 1990, Fig. 2). If predatorscontinue to encounter prey at a constant rate (T a), foragingon dangerous prey will result in lower harvest rates for thepredator (Fig. 3A). Therefore, a shift from safe to dangerousprey could result in considerably reduced energy intake ratesfor predators. Such may be the case for native predators facedwith increasing populations of dangerous invasive prey. Therelationship between energetic gains possible from attackingdangerous prey and increased handing time, discounted forthe fitness consequences of injury and probability of injurywhile foraging on that prey, should be critical to determiningthe probability that predators will attack a dangerous preyspecies.

Changing encounter rates would have a similar effect,but may operate through different mechanisms. Attackingdangerous prey and sustaining injury would increase therecovery time for predators, thereby reducing T a. For apredator hunting dangerous prey, if T h is held constant,even a large magnitude of change (e.g. eightfold) in T adoes not increase the harvest rate of a predator (Fig. 3B).Hence the more dangerous the prey, the lower is the harvestrate. Regardless of the mechanism involved, the differencebetween the harvest rate curves of dangerous and safe preytypes indicates the predator’s risk from its prey (Berger-Talet al., 2009). The greater the difference in risk between preytypes, the greater the difference in these curves, and thehigher the risk to the predator.

Data on predator hunting behaviour and success,specifically the relationship between the total time spenthunting a given prey and the number of attacks/attempts(i.e. physical contact with prey) required to successfully huntit, may provide insights into the perceived risk of injuryfrom prey when direct observations of prey-inflicted injuriesare not available. The least dangerous prey should requirerelatively short hunting times to capture and, in general,fewer attempts should be required for a successful hunt (areaI of Fig. 4). For African wild dogs this category of prey wouldinclude piglets, fawns and ungulate yearlings (Creel & Creel,2002). A second category of prey is one requiring moreattempts for a successful capture, but is quickly subdued(area II, Fig. 4). Finding a prey to fit this category is likelyto be biologically challenging since number of attempts



0 10 20 30 40 50 60

Prey density

0

2

4

6

8

10

12

14

16

18

20

22

24 (A)

(B)

Har

vest

rat

e, f

(#

kille

d/da

y)

Safe prey (Th=0.032)

Dangerous prey (Th=0.256)

(Th=0.064)

(Th=0.128)

Ta=1.38 (fixed)

0 10 20 30 40 50 60

Prey density

0

2

4

6

8

10

12

14

16

18

20

22

24

Har

vest

rat

e, f

(#

kille

d/da

y)

(Ta=1.38)

(Ta=0.173)

(Ta=0.69)

(Ta=0.345)

Th = 0.256 (fixed)

Fig. 3. Simulated harvest rate curves [using Holling’s (1959)disc equation] for a predator, showing the effect of varying prey(A) handling time (T h), and (B) encounter rate (T a).

and hunting time are generally positively correlated. Agood example of such a prey would be Thompson’s gazelle(Eudorcas thomsonii), which are good at escaping, but whencaught by a cheetah (Acinonyx jubatus) are killed swiftly. Preythat require many attempts before a hunt is successful aswell as relatively long times to capture should generally bethe most dangerous prey for a predator (area III in Fig. 4).Many attempts are needed because the predator has tocatch and release its prey several times in order to avoidphysical injury. It also helps tire the prey so that the predatorcan easily subdue it at a later stage. Buffaloes, which havebeen observed to injure and even kill lions (Mangani, 1962;Mitchell, Shenton & Uys, 1965; Makacha & Schaller, 1969)fit such a description. A fourth category of prey includesspecies which require long times to capture and subdue, butfew attempts are generally successful (area IV in Fig. 4). Thiscould be either because the prey has to be stalked for along time before an attack, or because habitat characteristicsfavour the prey (e.g. hunting in a difficult terrain). Forexample, a snow leopard hunting blue sheep (Pseudois nayaur)in the steep slopes of the Himalayas has to reach a fairly closedistance (i.e. long stalking time) before attempting to attack,

Num

ber

of a

ttack

s pe

r su

cces

sful

hun

t

Time spent during an individual hunt

(II) (III)

(IV)(I)

Fig. 4. The relationship between the total time spent huntinga given prey and the number of attacks/attempts required tosuccessfully hunt can provide insights into a predator’s risk ofinjury. Prey that fall in area I are the easiest (few attempts andshort hunting time) to catch (least dangerous). Prey in area IIare tricky (more attempts) to catch, but biologically it maybedifficult to find such a prey since the number of attempts bya predator is generally positively correlated with hunting time.Prey in area III are difficult to catch (more attempts and longhunting time) and likely to be the most dangerous. Prey that fallin area IV require a long time to stalk, hence long hunting time,but are captured swiftly.

but once they reach this position, are likely to be successful.While number of hunting attempts per successful hunt, andtotal time spent hunting will typically covary positively (seeScheel, 1993), factors such as aggressiveness of the prey(i.e. prey defence) and the predator’s energetic state andpersonality (both discussed later in Section IV.2) will alsoinfluence this relationship.

(2) Is there intraspecific variation in willingness totake risks?

Theory based on food versus predation-risk trade-offs canprovide a basis for developing insights into when predatorsshould risk injury. For example, an approach similar to thatof Lima & Dill (1990) can be used to quantify risk of injuryP(Injury) for a predator, as follows:

P(Injury

) = 1 –exp (–αiT ) (2)

where α is the rate of encounter between a predator anda dangerous prey, i is the probability of injury given anencounter, and T is the time spent vulnerable during anencounter. We can think of α, i and T as the basiccomponents of risk of injury, and these are assessable bya predator. The rate of encounter will depend on variousfactors such as prey density, search tactics, habitat structure,etc. The probability of injury, i, given an encounter, isdependent on a set of conditional probabilities in thesituation in which an encounter occurs, the predatorattacks, and potentially gets injured (see Fig. 5), where i



Encounter situation(for a predator)

Detects prey No encounter

1-p p

Avoids prey(prey too dangerous)

Attacks prey

a 1-a

Captures prey Prey escapes

1-c c

Predator gets injured Successful capture

s1-r -s

Predator releases prey(prey too dangerous)

r

Fig. 5. Schematic representation of a predator hunting adangerous prey. The symbols adjacent to the arrows representthe conditional probability of the predator following a certainpathway. Only a portion of all encounters lead to potentialinjury to the predator.

can be defined as:

i = (1–p

)(1–a) (1–c) (1– r –s) (3)

where 1-p is the probability of detecting a prey, 1-a is theprobability that the predator decides to attack the prey, 1-cis the probability that it captures the prey, s is the probabilitythat the capture is successful, r is the probability that itreleases the prey after capture since it is dangerous, and1-r-s is the probability that the predator gets injured. Mostof these sub components (see Fig. 5) can be assessed by thepredator. For example, after sighting a prey, the predatorcan assess the level of danger posed by that potential preyitem and may choose to avoid it or to attack. If it doesdecide to attack it, the predator may choose to releaseits prey depending on its assessment of the probability ofinjury and the potential severity of any possible injury.Time, T , in the context of risk of injury, can be thoughtof as time spent actually pursuing the dangerous prey,and this again can be managed by the predator. Thus,predators would be expected to manage their risk of injuryat least to some extent at multiple points in a predator–preyencounter.

Brown’s (1988, 1992) model can be interpreted to suggestthat a predator should stop hunting a dangerous prey whenits harvest rate or reward (f ) equals sum of its metabolic cost(c), missed opportunity cost (MOC ) and cost of risk of injury(RI ; Berger-Tal et al., 2009).

f = c + MOC + RI . (4)

MOC represents the alternatives that the predator misses,e.g. hunting less dangerous prey, or simply resting. Thecost of risk of injury (RI ) has units of energy per unit timeor resources per unit time. The currency of risk of injury,γ , can be converted into the currency of f by multiplyingthe risk of injury by the marginal rate of substitution ofenergy (MRS) for safety. The MRS depends upon the fitnessformulation, and is the ratio of survivor’s fitness (F ) to themarginal fitness value of energy (MVE, ∂F/∂e; Brown, 1992;Brown & Kotler, 2004). Therefore, following Brown (1992),the energetic cost of risk of injury (RI ) to a predator is:

RI = γ F(

∂F∂e

) (5)

This equation suggests that as the value of acquiring energyincreases for a predator, its overall cost of RI is reduced,and hence the predator should be willing to take greaterrisks (Berger-Tal et al., 2009). This is consistent with state-dependent foraging theory which suggests that individualsthat are in a poor state (e.g. close to starvation) are morelikely to take greater risks while foraging (McNamara &Houston, 1987) because their MVE is greater and costs ofpredation are lower (Charnov, 1976; Brown, 1988). Diversefield studies have provided support for state-dependentforaging under predation risk (e.g. Godin & Smith, 1988;Heithaus et al., 2007; Berger-Tal et al., 2010). Berger-Talet al. (2009) extended this to injury costs and tested whetherRI is a true foraging cost to a predator (red fox, Vulpesvulpes). They found that foxes exploited safe patches moreintensively, by foraging for a longer time and also removingmore food, compared to risky patches. They also found thathungrier foxes allocated more time to foraging from riskierpatches. Other studies support state-dependent foraging ondangerous prey. For example, lions in Hawange NationalPark, Zimbabwe, are more likely to hunt elephant calves(which are strongly protected by their herd) during the drymonths and drought years when the energetic state of lions islow (Loveridge et al., 2006). These decisions, however, mayalso be influenced by the relative abundance of alternativeprey affecting the predator’s MOC.

State-dependent decisions about attacking dangerousprey, however, may not always occur in a consistent manner.For instance, foragers in poor condition may be more likelyto be injured by dangerous prey (high RI ) and hence avoidthem. Murza et al. (2000) found that injured male Americankestrels (Falco sparverius) were more reluctant to attack largeand potentially dangerous rodent prey compared to non-handicapped birds. Perhaps injured birds have a highercost of missed opportunity, hence preferring to rest andrecover and thus are reluctant to attack. Predators may alsobe willing to take more risk when their food resources areabundant, since the costs of experiencing a negative payoffcan be recuperated rapidly. For example, in chimpanzees(Pan troglodytes), both hunting rate and the probability ofhunting upon encountering red colobus monkeys (Procolobusspp.) are positively correlated with seasonal consumption of



ripe drupe fruits (Gilby & Wrangham, 2007). These fruitsare a preferred food and they are associated with elevatedreproductive performance by females (Gilby & Wrangham,2007).

In some cases injury to a predator can lead to inter-specific conflict and further costs. For example, large cats inAsia and Africa may resort to killing humans (less profitablebut easier prey) after suffering an injury. Corbett (1946,1957) found that some man-eating leopards (Panthera pardus)and tigers, which killed several hundred villagers in northernIndia, suffered from a missing (or broken) tooth or claw, whileothers had injuries from porcupine quills or gunshot wounds.

It is likely that predators of different age, sex, andreproductive classes vary in their willingness to attackdangerous prey. Clark’s (1994) asset protection principlesuggests that the larger the reproductive asset of an individual,the more inclined it is to protect it. Following this, one mightexpect pregnant females to take less risk for two main reasons.They might be less adept at avoiding injury (physically lessagile, e.g. Magnhagen, 1991), but they also have more tolose (i.e. greater immediate loss of fitness) if they are injuredor killed than a non-pregnant female or a male. Hence,following Equations (4) and (5), since F is high for a pregnantfemale, it also has a high cost of RI .

Reproductive value can also depend on the age of theindividual. For example, adult redshanks (Tringa totanus)minimize their risk of predation at the cost of reducedenergetic intake rate, while juveniles maximize their intakerate at the cost of higher predation risk (Cresswell, 1994).Variations in anti-predator behaviours between sexes and ageclasses have also been reported. Male and yearling yellow-bellied marmots (Marmota flaviventris) reduce their foragingbehaviour under risk of predation more than females andadult marmots (Lea & Blumstein, 2011). Such decisionsalmost certainly will apply with regard to the probability ofan individual hunting dangerous prey.

Social rank might also affect willingness to attackdangerous prey since it is related to their fitness. For example,in spotted hyenas (Crocuta crocuta) even though low- andhigh-ranked females have similar hunting success, it is thelow-ranked females who hunt more often (Holekamp et al.,1997), and hence are more prone to risk of injury fromtheir prey. Direct fitness benefits of risk-taking behaviourhave been documented in primates. For example, differentspecies of Colobus and Cercopithecus monkeys are hunted bychimpanzees (Boesch, 1994; Stanford et al., 1994), which areone of the few primates that hunt for meat, even thoughthese prey can potentially be dangerous (can fight back andbite, Gomes & Boesch, 2009). After a successful hunt, femalechimpanzees often beg for meat from males, and the malesare more likely to share the meat with these females thanwith other males (Gomes & Boesch, 2009). Females not onlybenefit energetically, but also by reducing the potential riskof being injured while hunting (Boesch, 1994; Gomes &Boesch, 2009). Successful males gain direct fitness advantageby sharing their meat since female chimpanzees copulatemore frequently with males that share meat, thereby directly

increasing the propensity of the males to father offspring(Gomes & Boesch, 2009).

Recent studies have shown important roles of personalitytraits on the foraging decisions of animals [see Wilson et al.(1994), Sih, Bell & Johnson (2004) and Reale et al. (2007)for reviews], with some individuals being inherently moreaggressive and bold than others (Riechert & Hedrick, 1993;Maupin & Riechert, 2001). It is possible that boldnesscould be related to propensity to attack risky prey, and thiscould be referred to as the predator’s ‘daringness’ (Brown &Kotler, 2004; Berger-Tal et al., 2009). Under standardizedconditions, the smaller the difference in hunting behaviour(e.g. harvest rate) of a predator between dangerous and safeprey, the more daring (i.e. willingness to risk injury) is thepredator (Berger-Tal et al., 2009). The relationship betweena predator’s body size and that of its prey will also affect theprobability of injury; therefore, we might expect larger indi-viduals of a population to be more likely to hunt dangerousprey. In Yellowstone National Park, USA, larger wolves arebetter at strength-related tasks (grappling and subduing elk)compared with smaller individuals (MacNulty et al., 2009).

Do predators take risk of injury into account when decidingwhere to hunt? Although previous studies have consideredhow physical features of the environment affect prey-captureprobabilities (Cresswell & Quinn, 2004; Hebblewhite,Merrill & McDonald, 2005; Hopcraft, Sinclair & Packer,2005; Heithaus et al., 2009; Wirsing, Cameron & Heithaus,2010), how physical danger posed by microhabitat or habitattype affects hunting decision and success of predators is poorlyunderstood. Indeed, predator avoidance of certain habitattypes (i.e. prey refugia) is largely studied in the context of howeasily predators are detected by their prey (i.e. prey–predatorencounter rate). What is largely overlooked is the extent towhich predators avoid certain areas to minimize their risk ofinjury. For example wolves avoid steep slopes (see Kauffmanet al., 2007) most likely because it helps reduce their chancesof injury from falling, and hence these areas likely act asrefuges for their prey. Resource selection functions (Boyce &McDonald, 1999; Manly et al., 2002) not only form anefficient framework for quantifying the spatial probabilityof predator–prey encounter rates and kills in ecologicallandscapes (Hebblewhite et al., 2005), but also could provideinsights into a predator’s spatial probability of risk of injurywhile hunting. Decisions made by predators hunting indangerous landscapes are likely to mirror those made byanimals foraging under the risk of predation, but with lowerfitness costs.

The selection of a more dangerous prey (or a daring preda-tor) can be driven by escalation or coevolution (Vermeij,1994; Brodie & Brodie, 1999). While predators accumulateadaptations (including behaviours) that increase their hunt-ing success, prey adapt to reduce their risk of being killed. Ina game between daring predators and dangerous prey, thereis the intriguing possibility that there would be selection forprey adaptations (either physical or behavioural) that wouldincrease risk to predators and, therefore, require even moredaring predators. For example, could daring in predators



select for more aggressive prey that are willing to confrontpredators? Aggression in prey has several advantages beyondpotentially reducing predation risk including enhancedresource (food, mate, offspring, etc.) defence. Previousstudies have found that there is higher correlation betweenaggressiveness and boldness (via correlative selection) inpopulations facing strong predation pressure (Bell & Sih,2007). In certain prey species this may translate to increasedaggression towards the predator, thereby increasing thepredators RI. The prey’s aggression may feedback andinfluence the predator’s behaviour. The predator may eitheravoid the prey, or show increased boldness by attackingthe prey while risking injury. Hence risk-taking by daringpredators may ultimately be related to ‘daring’ behaviour inthe prey themselves. This hypothesis remains to be tested.

VII. CONCLUSIONS

(1) Predators that hunt dangerous prey represent a specialclass of prey–predator interactions that have not yet beenwell integrated into our understanding of foraging behaviourand predator–prey games theory. It is likely that mostforagers pay a risk of injury foraging cost, with these costsbeing greater for predators than herbivores.

(2) Understanding the foraging costs of hunting dangerousprey and the frequency and costs of prey-inflicted injurieslikely will provide important insights into the dynamics ofsome communities.

(3) Just as anti-predator behaviours of prey help stabilizepredator–prey dynamics in fear-driven systems (Brown,Laundre & Gurung, 1999), the foraging decisions andbehavioural responses of predators to minimize risk of injurymay play an important role in at least some predator–preyforaging systems.

(4) Both theoretical and empirical studies are still neededto answer basic questions about the frequency and severityof injuries to predators hunting across a range of ecologicalconditions, the fitness costs of these injuries, and the tacticsemployed to balance energy gain and costs of injury.While these costs and risk-reducing behaviours amongpredators seem to have been investigated to some extent ininvertebrates (Table 1) we need similar quantitative studiesin other groups.

(5) Although it will be a challenge to quantify costs of riskof injury in the wild, the collection of such data especiallyin situations where there are changes in prey availability(that might affect willingness to take risks) or invasions ofpotentially dangerous prey, should be an important avenueof future research for behavioural ecologists.

VIII. ACKNOWLEDGEMENTS

Funding was provided by the Florida InternationalUniversity College of Arts and Sciences, National ScienceFoundation (grants OCE0745606, DEB-9910514). Thanks

to Alison Cooper, Burt Kotler, Keren Embar, Oded Berger-Tal and an anonymous reviewer for their valuable comments.

IX. REFERENCES

Ackermann, J. & Redig, P. (1997). Surgical repair of elbow luxation in raptors.Journal of Avian Medicine and Surgery 11, 247–254.

Agrawal, A. A. (2001). Phenotypic plasticity in the interactions and evolution ofspecies. Science 294, 321–326.

Auffenberg, W. (1981). Behavioral Ecology of the Komodo Monitor. University PressFlorida, Gainesville.

Barr, A. D., Wieburg, S. A. & Kardong, K. V. (1988). The predatory strikebehavior of the mamushi (Agkistrodon b. blomhoffii) and the Malay Pitviper (Calloselasma

rhodostoma). Japanese Journal of Herpetology 12, 135–138.Bedrosian, B. E. & St. Pierre, A. M. (2007). Frequency of injuries in three raptor

species wintering in northern Arkansas. Wilson Journal of Ornithology 119, 296–298.Bell, A. & Sih, A. (2007). Exposure to predation generates personality in three spined

sticklebacks (Gasterosteus aculeatus). Ecology Letters 10, 828–834.Berger-Tal, O., Mukherjee, S., Kotler, B. P. & Brown, J. S. (2009). Look

before you leap: is risk of injury a foraging cost? Behavioral Ecology and Sociobiology 63,1821–1827.

Berger-Tal, O., Mukherjee, S., Kotler, B. P. & Brown, J. S. (2010). Complexstate-dependent game between owls and gerbils. Ecology Letters 13, 302–310.

Berger, J. (1979). Predator harassment as a defense strategy in ungulates. American

Midland Naturalist 102, 197–199.Boesch, C. (1994). Cooperative hunting in wild chimpanzees. Animal Behaviour 48,

653–667.Boyce, M. S. & McDonald, L. L. (1999). Relating populations to habitats using

resource selection functions. Trends in Ecology & Evolution 14, 268–272.Braithwaite, L. F., Stone, B. & Bingham, B. L. (2010). Defensive behaviors of

gastropod Amphissa columbiana. Journal of Shellfish Research 29, 217–222.Brodie, E. D. III & Brodie, E. D. Jr. (1999). Predator–prey arms races: asymmetrical

selection on predators and prey may be reduced when prey are dangerous. Bioscience

49, 557–568.Brown, J. S. (1988). Patch use as an indicator of habitat preference, predation risk,

and competition. Behavioral Ecology and Sociobiology 22, 37–47.Brown, J. S. (1992). Patch use under predation risk. I. Models and predictions. Annales

Zoologici Fennici 29, 301–309.Brown, J. S. & Kotler, B. P. (2004). Hazardous duty pay and the foraging cost of

predation. Ecology Letters 7, 999–1014.Brown, J. S., Laundre, J. W. & Gurung, M. (1999). The ecology of fear: optimal

foraging, game theory, and trophic interactions. Journal of Mammalogy 80, 385–399.Charnov, E. L. (1976). Optimal foraging, the marginal value theorem. Theoretical

Population Biology 9 (129), 136.Clark, C. W. (1994). Antipredator behavior and the asset-protection principle.

Behavioral Ecology 5, 159–170.Congi, R., Fritus, A. V. L. & Filho, B. F. A. (2002). Influence of prey size on

predation success by Zelus longipes L. (Het., Reduviidae). Journal of Applied Entomology

126, 74–78.Corbett, J. (1946). Man-Eaters of Kumaon. Oxford University Press, London.Corbett, J. (1957). Man-Eaters of India. Oxford University Press, New York.Creel, S. & Christanson, D. (2008). Relationships between direct predation and

risk effects. Trends in Ecology & Evolution 23, 194–201.Creel, S. & Creel, N. M. (1995). Communal hunting and pack size in African wild

dogs, Lycaon pictus. Animal Behaviour 50, 1325–1339.Creel, S. & Creel, N. M. (2002). The African Wild dog: Behavior, Ecology and Conservation.

Princeton University Press, Princerton.Cresswell, W. (1994). Age-dependent choice of redshank (Tringa totanus) feeding

location: profitability or risk? The Journal of Animal Ecology 63, 589–600.Cresswell, W. & Quinn, J. L. (2004). Faced with a choice, sparrowhawks more

often attack the more vulnerable prey group. Oikos 104, 71–76.Dejean, A. & Lachaud, J. (2011). The hunting behavior of the African ponerine ant

Pachycondyla pachyderma. Behavioral Processes 86, 169–173.Dickson, B. G., Jenness, J. S. & Beier, P. (2004). Influence of vegetation, topography,

and roads on cougar movement in Southern California. The Journal of Wildlife

Management 69, 264–276.Dietl, G. P. (2003). Interaction strength between a predator and dangerous prey:

Sinistrofulgur predation on Mercenaria. Journal of Experimental Marine Biology and Ecology

289, 287–301.Donadio, E. & Buskirk, S. W. (2006). Diet, morphology and interspecific killing in

carnivore. The American Naturalist 167, 524–536.Duignan, P. J., Hunter, J. E. B., Visser, I. N., Jones, G. W. & Nutman, A. (2000).

Stingray spines: a potential cause of killer whale mortality in New Zealand. Aquatic

Mammals 26, 143–147.



Edmunds, M. (1974). Defense in Animals: A Survey of Anti-Predator Defenses.Longman, Harlow.

Eisner, T. & Aneshansley, D. J. (1999). Spray aiming in the bombardier beetle:photographic evidence. Proceedings of the National Academy of Sciences 96, 9705–9709.

Eisner, T., Baldwin, I. T. & Conner, J. (1993). Circumvention of prey defense by apredator: ant lion vs ant. Proceedings of the National Academy of Sciences 90, 6716–6720.

Frid, A., Burns, J. & Baker, G. G. (2009). Predicting synergistic effects of resourcesand predators on foraging decisions by juvenile steller sea lions. Oecologia 158,775–786.

Garner, G. W. & Morrison, J. A. (1980). Observations of interspecific behaviorbetween predators and white-tailed deer in southwestern Oklahoma. Journal of

Mammalogy 61, 126–130.Gibson, M. J., Gibson, D. C. & Bardelmeier, D. G. (1998). Prey conquers

predator: a case study – barred owl dies from rodent-inflicted wound. Journal of

Wildlife Rehabilitation 21, 19–21.Gilby, I. C. & Wrangham, R. W. (2007). Risk-prone hunting by chimpanzees (Pan

troglodytes schweinfurthii) increases during periods of high diet quality. Behavioral Ecology

and Sociobiology 61, 1771–1779.Godin, J. J. & Smith, S. A. (1988). A fitness cost of foraging in the guppy. Nature 333,

69–71.Gomes, C. M. & Boesch, C. (2009). Wild chimpanzees exchange meat for sex on a

long- term basis. PLoS ONE 4 (4), e5116 (doi: 10.1371/journal.pone.0005116).Greene, J. (1967). The distribution and variation of Daphnia Zumholtzi (Crustacea:

Cladocera) in relation to fish predation in Lake Albert, East Africa. Journal of Zoology

151, 181–197.Grinnell, J., Packer, C. & Pusey, A. E. (1995). Cooperation in male lions: kinship,

reciprocity or mutualism? Animal Behaviour 49, 95–105.Haddock, S. H. D., Moline, M. A. & Case, J. F. (2010). Bioluminescence in the sea.

Annual Review of Marine Science 2, 443–493.Hayward, M. W., Henschel, P., O’Brien, J., Hofmeyr, M., Balme, G. &

Kerley, G. I. H. (2006). Prey preferences of the leopard (Pantherapardus). Journal of

Zoology 270, 298–313.Hebblewhite, M., Merrill, E. H. & McDonald, T. L. (2005). Spatial

decomposition of predation risk using resource selection functions: an examplein a wolf-elk predator–prey system. Oikos 111, 101–111.

Heithaus, M. R., Frid, A., Wirsing, A. J., Dill, L. M., Fourqurean, J.,Burkholder, D., Thompson, J. & Bejder, L. (2007). State-dependent risk-takingby green sea turtles mediates top-down effects of tiger shark intimidation in a marineecosystem. The Journal of Animal Ecology 76, 837–844.

Heithaus, M. R., Frid, A., Wirsing, A. J. & Worm, B. (2008). Predicting ecologicalconsequences of marine top predator declines. Trends in Ecology & Evolution 23,202–210.

Heithaus, M. R., Wirsing, A. J., Burkholder, D., Thomson, J. & Dill, L. M.(2009). Towards a predictive framework for predator risk effects: the interaction oflandscape features and prey escape tactics. The Journal of Animal Ecology 78, 556–562.

Holekamp, K. E., Smale, L., Breg, R. & Cooper, S. M. (1997). Hunting rates andhunting success in the spotted hyena (Crocuta crocuta). Journal of Zoology 242, 1–15.

Holling, C. S. (1959). Some characteristics of simple types of predation andparasitism. The Canadian Entomologist 91, 385–398.

Holt, D. W. & Layne, E. A. (2008). Eye injuries in long-eared owls (Asiootus):prevalence and survival. Journal of Raptor Research 42, 243–247.

Hopcraft, J. G. C., Sinclair, A. R. E. & Packer, C. (2005). Planning for success:Serengeti lions seek prey accessibility rather than abundance. The Journal of Animal

Ecology 74, 559–566.Juanes, F. & Hartwick, E. B. (1990). Prey selection in dungeness crabs: the effect of

claw damage. Ecology 71, 744–758.Kardong, K. V. (1986). Predatory strike behavior of the rattlesnake, Crotalus viridis

oreganus. Journal of Comparative Psychology 100, 304–313.Kauffman, M. J., Varley, N., Smith, D. W., Stahler, D. R., MacNulty, D.

R. & Boyce, M. S. (2007). Landscape heterogeneity shapes predation in a newlyrestored predator–prey system. Ecology Letters 10, 690–700.

Klimley, A. P. (1994). The predatory behavior of the white shark. American Scientist

52, 122–133.Klimley, A. P., Pyle, P. & Anderson, S. D. (1996). The behavior of white sharks and

their pinniped prey during predatory attacks. In Great White Sharks: The Biologyof Carcharodon Carcharias (eds A. P. Klimley and D. G. Ainley), pp. 175–191.Academic Press, New York.

Kotler, B. P., Gross, J. E. & Mitchell, W. A. (1994). Applying patch use to assessaspects of foraging behavior in Nubian ibex. The Journal of Wildlife Management 58,299–307.

Lea, A. J. & Blumstein, L. (2011). Age and sex influence marmot antipredatorbehavior during periods of heightened risk. Behavioral Ecology and Sociobiology 65,1525–1533.

Lima, S. L. (1998). Stress and decision making under the risk of predation: recentdevelopments from behavioral, reproductive, and ecological perspectives. Advances

in the Study of Behavior 27, 215–290.Lima, S. L. & Dill, L. M. (1990). Behavioral decisions made under the risk of

predation: a review and prospectus. Canadian Journal of Zoology 68, 619–640.

Loveridge, A. J., Hunt, J. E., Murindagomo, F. & Macdonald, D. W. (2006).Influence of drought on predation of elephant (Loxodonta africana) calves by lions(Panthera leo) in an African wooded savannah. Journal of Zoology 270, 523–530.

MacNulty, D. R., Smith, D. W., Mech, L. D. & Eberly, L. E. (2009). Body sizeand predatory performance in wolves: is bigger better? The Journal of Animal Ecology

78, 532–539.Magnhagen, C. (1991). Predation risk as a cost of reproduction. Trends in Ecology &

Evolution 6, 183–186.Makacha, S. & Schaller, G. B. (1969). Observations on lions in the Lake Manyara

National Park, Tanzania. East African Wildlife Journal 7, 99–103.Mangani, B. (1962). Buffalo kills lion. African Wildlife 16, 27.Manly, B. F. J., McDonald, L. L., Thomas, D. L., McDonald, T. L. & Erickson,

W. P. (2002). Resource Selection by Animals: Statistical Analysis and Design for Field Studies.Second Edition. Kluwer, Massachusetts.

Mao, J. S., Boyce, M. S., Smith, D. W., Singer, F. J., Vales, D. J., Vore, J. M. &Merrill, E. H. (2005). Habitat selection by elk before and after wolf reintroductionin Yellowstone National Park. The Journal of Wildlife Management 69, 1691–1707.

Martin, R. A., Hammerschlag, N., Collier, R. S. & Fallows, C. (2005).Predatory behaviour of white shakrs (Carcharodon carcharias) at Seal Island, SouthAfrica. Journal of the Marine Biological Association of the United Kingdom 85, 1121–1135.

Maupin, J. L. & Riechert, S. E. (2001). Superfluous killing in spiders: a consequenceof adaptation to food-limited environments? Behavioral Ecology 12, 569–576.

McComb, K., Packer, C. & Pusey, A. (1994). Roaring and numerical assessment incontests between groups of female lions, Panthera leo. Animal Behaviour 47, 379–387.

McLellan, W. A., Thayer, V. G. & Pabst, D. A. (1996). Stingray spine mortalityin a bottle nose dolphin, Tursiops truncates, from North Carolina waters. The Journal

of the Elisha Mitchell Scientific Society 112, 98–101.McNamara, J. M. & Houston, A. I. (1987). Starvation and predation as factors

limiting population size. Ecology 68, 1515–1519.Mitchell, B. L., Shenton, J. B. & Uys, J. C. M. (1965). Predation on large mammals

in the Kafue National Park, Zambia. Zoologica Africana 1, 297–318.Murza, G. L., Bortolotti, G. R. & Dawson, R. D. (2000). Handicapped American

kestrels: needy or prudent foragers. Journal of Raptor Research 34, 137–142.Olive, C. W. (1980). Foraging specialization in orb-weaving spiders. Ecology 61,

1133–1144.Ono, M., Igarashu, T., Ohno, E. & Sasaki, M. (1995). Unusual thermal defense

by a honeybee against mass attack by hornets. Nature 377, 334–336.Packer, C. & Ruttam, L. (1988). The evolution of cooperative hunting. The American

Naturalist 132, 159–198.Packer, C., Scheel, D. & Pusey, A. E. (1990). Why lions form groups: food is not

enough? The American Naturalist 136, 1–19.Pearlman, Y. & Tsurim, I. (2008). Daring, risk assessment and body condition

interactions in steppe buzzards Buteobuteo vulpinus. Journal of Avian Biology 39,226–228.

Pellis, S. M. & Officer, R. C. E. (1987). An analysis of some predatory behaviorpatterns in four species of carnivorous marsupials (Dasyuridae), with comparativenotes on the Eutherian cat Felis catus. Ethology 75, 177–196.

Preisser, E., Bolnick, D. I. & Benard, M. F. (2005). Scared to death? The effects ofintimidation and consumption in predator–prey interactions. Ecology 86, 501–509.

Quick, H. F. (1953). The occurrence of porcupine quills in carnivorous mammals.Journal of Mammalogy 34, 256–259.

Reale, D., Reader, S. M., Sol, D., McDougall, P. T. & Dingemanse, N. J. (2007).Integrating animal temperament within ecology and evolution. Biological Reviews 82,291–318.

Riechert, S. E. & Hedrick, A. V. (1993). A test for correlations among fitness-linked behavioural traits in the spider Agelenopsisaperta (Aranear, Agelenidae). Animal

Behaviour 46, 669–675.Roth, A. J., Jones, G. S. & French, T. W. (2002). Incidence of naturally-healed

fractures in the pectoral bones of North American accipiters. Journal of Raptor Research

36, 229–230.Rowe, A. H. & Rowe, M. P. (2006). Risk assessment by grasshopper mice (Onychomys

spp.) feeding on neurotoxic prey (Centruroides spp.). Animal Behaviour 71, 725–734.Ruxton, G. D., Speed, M. P. & Kelly, D. J. (2004). What, if anything, is the adaptive

function of countershading? Animal Behaviour 68, 445–451.Schaller, G. B. (1972). The Serengeti Lion: A Study of Predator-Prey Relations.

University of Chicago Press, Chicago.Scheel, D. (1993). Profitability, encounter rates, and prey choice of African lions.

Behavioral Ecology 4, 90–97.Scheel, D. & Packer, C. (1991). Group hunting behaviour of lions: a search for

cooperation. Animal Behaviour 4, 697–709.Schmitz, O., Krivan, V. & Ovadia, O. (2004). Trophic cascades: the primacy of

trait-mediated indirect interactions. Ecology Letters 7, 153–163.Schoener, T. W. (1971). Theory of feeding strategies. Annual Review of Ecology and

Systematics 2, 369–404.Shine, R., Harlow, P. S., Keogh, J. S. & Boeadi (1998). The influence of body

size on the food habits of a giant tropical snake, Python reticularis. Functional Ecology 12,248–258.



Sih, A., Bell, A. & Johnson, J. C. (2004). Behavioral syndromes: an ecological andevolutionary overview. Trends in Ecology & Evolution 19, 372–378.

Sih, A. & Christensen, B. (2001). Optimal diet theory: when does it work, and whenand why does it fail? Animal Behaviour 61, 379–390.

Stanford, C. B., Wallis, J., Mpongo, E. & Goodall, J. (1994). Hunting decisionsin wild chimpanzees. Behaviour 131, 1–18.

Stephens, D. W. & Krebs, J. R. (1986). Foraging Theory. Princeton University Press,Princeton.

Swaffer, S. M. & O’Brien, W. J. (1996). Spines of Daphnia lumholtzi create feedingdifficulties for juvenile bluegill sunfish (Lepomis macrochirus). Journal of Plankton Research

18, 1055–1061.Tricas, T. C. & McCosker, J. E. (1985). Predatory behavior of the white shark

(Carcharodon carcharias), with notes on its biology. Proceedings of the California Academy of

Sciences 43, 221–238.Van Valkenburg, B. (2009). Costs of carnivory: tooth fracture in Pleistocene and

recent carnivorans. Biological Journal of the Linnean Society 96, 68–81.Van Valkenburg, B. & Hertel, F. (1993). Tough times at La Brea: tooth breakage

in large carnivores of the late Pleistocene. Science 261, 456–459.

Vermeij, G. J. (1994). The evolutionary interaction among species: selection, escalationand coevolution. Annual Reviews of Ecology and Systematics 25, 219–236.

Wall, M. & Shine, R. (2007). Dangerous food: lacking venom and constriction, howdo snake-like lizards (Lialisburtonis, Pygopodidae) subdue their lizard prey? Biological

Journal of the Linnean Society 91, 719–727.Walsh, M. T., Beusse, D., Bossart, G. D., Young, W. G., Odell, D. K. &

Patton, G. W. (1988). Ray encounters as a mortality factor in the Atlanticbottlenose dolphins (Tursiops truncatus). Marine Mammal Science 4, 154–162.

Warham, J. (1977). The incidence, function and ecological significance of petrelstomach oils. Proceedings of the New Zealand Ecological Society 24, 84–93.

Werner, E. E. & Peacor, S. D. (2003). A review of trait-mediated indirect interactionsin ecological communities. Ecology 84, 1083–1100.

Wilson, D. S., Clark, A. B., Coleman, K. & Dearstyne, T. (1994). Shyness andboldness in humans and other animals. Trends in Ecology & Evolution 9, 442–446.

Wirsing, A., Cameron, K. & Heithaus, M. R. (2010). Spatial responses to predatorsvary with prey escape mode. Animal Behavior 79, 531–537.

(Received 1 September 2011; revised 17 November 2012; accepted 11 December 2012; published online 21 January 2013)


Documents

Dangerous prey and daring predators: a review