energetics and evasion dynamics of large predators and ... · gray wolves (canis lupus), for...

Submitted 5 June 2017Accepted 26 July 2017Published 17 August 2017

Corresponding authorCaleb M Bryce cbryceucsceducalebmbrycegmailcom

Academic editorDonald Kramer

Additional Information andDeclarations can be found onpage 15

DOI 107717peerj3701

Copyright2017 Bryce et al

Distributed underCreative Commons CC-BY 40

OPEN ACCESS

Energetics and evasion dynamics of largepredators and prey pumas vs houndsCaleb M Bryce12 Christopher C Wilmers3 and Terrie M Williams1

1Department of Ecology amp Evolutionary Biology University of California Santa Cruz CAUnited States of America

2Botswana Predator Conservation Trust Maun Botswana3Center for Integrated Spatial Research Environmental Studies Department University of CaliforniaSanta Cruz CA United States of America

ABSTRACTQuantification of fine-scale movement performance and energetics of hunting bylarge carnivores is critical for understanding the physiological underpinnings of trophicinteractions This is particularly challenging for wide-ranging terrestrial canid and felidpredators which can each affect ecosystem structure through distinct hunting modesTo compare free-ranging pursuit and escape performance from group-hunting andsolitary predators in unprecedented detail we calibrated and deployed accelerometer-GPS collars during predator-prey chase sequences using packs of hound dogs (Canislupus familiaris 26 kg n = 4ndash5 per chase) pursuing simultaneously instrumentedsolitary pumas (Puma concolor 60 kg n= 2) We then reconstructed chase paths speedand turning angle profiles and energy demands for hounds and pumas to examineperformance andphysiological constraints associatedwith cursorial and cryptic huntingmodes respectively Interaction dynamics revealed how pumas successfully utilizedterrain (eg fleeing up steep wooded hillsides) as well as evasive maneuvers (egjumping into trees running in figure-8 patterns) to increase their escape distance fromthe overall faster hounds (avg 23times faster) These adaptive strategies were essential toevasion in light of the mean 16times higher mass-specific energetic costs of the chase forpumas compared to hounds (mean 076 vs 129 kJ kgminus1 minminus1 respectively) On aninstantaneous basis escapes were more costly for pumas requiring exercise at ge90of predicted VO2MAX and consuming as much energy per minute as approximately5 min of active hunting Our results demonstrate the marked investment of energyfor evasion by a large solitary carnivore and the advantage of dynamic maneuvers topostpone being overtaken by group-hunting canids

Subjects Animal Behavior Conservation Biology Ecology Environmental Sciences ZoologyKeywords Large carnivore Physiology Movement ecology Tradeoffs Performance Adaptivestrategies Accelerometer GPS telemetry Hunting modes Energetics

INTRODUCTIONHunting modes in sympatric large carnivores have evolved and diversified with membersof the families Felidae and Canidae exhibiting nearly opposite prey detection and capturetechniques characterized by cryptic ambushing or cursorial pursuit respectively (Table1) Gray wolves (Canis lupus) for example often hunt cooperatively in packs (Mech1970 Mech Smith amp MacNulty 2015) and rely on endurance pursuit (Snow 1985 Poole

How to cite this article Bryce et al (2017) Energetics and evasion dynamics of large predators and prey pumas vs hounds PeerJ5e3701 DOI 107717peerj3701

Table 1 Comparison of hunting mode divergence observed in large felids and canids Selected references for each topic (superscripts) are pro-vided below

Family Felidae Family Canidae

Eg puma leopard jaguar Eg gray wolf hound dingo

Hunting modea Cryptic stalking amp pouncing lsquolsquoSurprise ampsubduersquorsquo

Cursorial pursuit lsquolsquoCharge amp chasersquorsquo

Hunting socialityb Solitary Often grouppackRelative prey selectivity and timingc Low (opportunistic) Prior to attack High (selective) Often during pursuitInteraction with amp risk imposed by preyd Short Lower risk of injurydeath Prolonged Higher risk of injurydeathKill site attributese Sufficient structural cover for concealment

during stalking and brief pursuitRelatively open terrain that facilitatesprolonged pursuit

Scale of habitat features impacting huntsuccessf

Small-scale habitat features Large-scale landscape heterogeneity

Relative activity and energetic demand ofhuntrsquos attack phase

High intensity short duration Low intensity long duration

NotesaHornocker 1970 Koehler amp Hornocker 1991 Ruth amp Murphy 2009a Seidensticker et al 1973 Young amp Goldman 1946 Poole amp Erickson 2011 Snow 1985Mech amp Korb 1978Mech amp Cluff 2011

bGittleman 1989 Hornocker amp Negri 2009Mech Smith amp MacNulty 2015Mech 1970cHusseman et al 2003Wilmers Post amp Hastings 2007 Kunkel et al 1999 Okarma et al 1997Mech Smith amp MacNulty 2015 Peterson amp Ciucci 2003 but see Karanth amp Sun-quist 1995 Krumm et al 2010

dHornocker amp Negri 2009Mech Smith amp MacNulty 2015Mech amp Boitani 2003eAlexander Logan amp Paquet 2006 Hebblewhite amp Merrill 2008 Husseman et al 2003 Ruth et al 2011 Schmidt amp Kuijper 2015 and references thereinfHebblewhite Merrill amp McDonald 2005 Kauffman et al 2007 Laundreacute amp Hernaacutendez 2003 Podgoacuterski et al 2008 Schmidt amp Kuijper 2015

amp Erickson 2011) rather than speed or agility to test and ultimately outperform morevulnerable prey (Peterson amp Ciucci 2003 Mech Smith amp MacNulty 2015) In contrastpumas (Puma concolor) exhibit an opportunistic (ie less selective Husseman et al2003 Wilmers Post amp Hastings 2007) solitary and cryptic hunting mode by which theystealthily ambush and overpower prey (Hornocker amp Negri 2009 Ruth amp Murphy 2009a)through matching pounce force to prey size (Williams et al 2014) Such divergence inlocomotion sociality prey selectivity and even preferred terrain while hunting reducesexploitative and interspecific competition (Husseman et al 2003 Elbroch et al 2015)through spatiotemporal niche partitioning and has cascading ecosystem-wide effects(Rosenzweig 1966 Linnell amp Strand 2000 Donadio amp Buskirk 2006 Elbroch et al 2015)Less is known however of how the fine-scale movement performance and metabolicdemands associated with these distinct predatory hunting modes interact to affectpredation success

During a predation event an animalrsquos ability to readily adjust its speed (Howland 1974Domenici 2001) acceleration (Combes et al 2012 Wilson et al 2013b) and turn capacity(Howland 1974 Maresh et al 2004 Wilson et al 2013a) becomes critical for survivalDespite its relative brevity the attack phase of the hunt may be the most energeticallyexpensive stage of prey acquisition particularly for ambush predators (Williams et al2014) Given the two-dimensional confines of the terrestrial environment both predators

Bryce et al (2017) PeerJ DOI 107717peerj3701 223

and prey have restricted behavioral options during this critical phase and are primarilyleft with modulating their speed (Elliott Cowan amp Holling 1977) andor maneuverability(Howland 1974) in order to hunt successfully or survive respectively (reviewed in Wilsonet al 2015) Furthermore these constraints may result in an lsquolsquoarms racersquorsquo evolutionaryescalation of matched specialized morphologies and behavioral strategies that promotecapture ability or evasion capacity in species that have co-evolved (Brodie amp Brodie 1999Cortez 2011) although the strength of selective forces acting on predators vs prey maydiffer (ie the lsquolsquolife-dinner principlersquorsquo Dawkins amp Krebs 1979)

The impacts of locomotor performance and energetics in altering chase outcomes haslong been recognized with the majority of our understanding of these interactions comingfrom studies of animals maneuvering in aerial (Warrick 1998 Hedenstroumlm amp Roseacuten 2001Combes et al 2012) or aquatic (Domenici amp Blake 1997 Domenici 2001) environmentsConsiderably less attention has been given to describing these complex dynamics interrestrial species particularly large carnivores and their prey (Wilmers et al 2015) Thisis likely because our ability to describe such interactions is substantially impaired by thewide-ranging often cryptic behaviors of these mammals (Gese 2001 Williams et al 2014Wang Allen amp Wilmers 2015) Recently however advancements and miniaturization ofbiologging sensor technology now enable scientists to concurrently measure previouslyunavailable metrics including the fine-scale behavior physiological performance andenergetics of wild animals (Kays et al 2015 Wilmers et al 2015) In addition these noveltools have the capacity to quantify chase dynamics and identify features of the landscapeand the animals themselves that determine whether or not prey evade capture (Wilson etal 2013a)

Here using simultaneously instrumented pumas and scent hounds (Canis lupusfamiliaris) we examined the performance and energetic tradeoffs of divergent terrestrialhunting modes in real time Packs of trained hounds pursued solitary pumas in need ofrecapture for a separate monitoring study and afforded a comprehensive look at houndgroup hunting cohesion and its effect on puma maximal performance escape tactics inrugged terrain Given their local adaptation and stalk-and-pounce hunting mode wepredicted that pumas would exhibit greater acceleration top speed and turning abilityin rugged terrain relative to hounds but could only sustain this peak performance over ashort distance and duration Furthermore we predicted that the cursorial hounds wouldcompensate for poorer sprinting performance by coursing continually over long distancesat slower speeds with greater energetic efficiency relative to pumas Due to the scarcity ofstudies investigating detailed chase performance parameters and their associated metaboliccosts in terrestrial mammals our goal was to assess how terrain and evolved differencesin morphology physiology and behavioral strategies among large felids and canids affectchase dynamics and outcomes

MATERIALS amp METHODSCollar amp energetic calibrationsWe used a laboratory-to-field approach in which the locomotor biomechanics andenergetics of scent hounds (n= 7 242 plusmn 09 kg mean plusmn SE) and captive pumas


(n= 3 657 plusmn 44 kg) instrumented with accelerometer-GPS collars were measured in anenclosure and laboratory environment prior to deployment on free-ranging conspecificsin the wild (Bryce amp Williams 2017 Wang Allen amp Wilmers 2015 Williams et al 2014Wilmers et al 2015) Hounds wore a 16 Hz accelerometer (TDR10-X Wildlife ComputersRedmond WA USA) affixed to a GPS collar (Astro Garmin Ltd Switzerland) capableof taking a GPS satellite fix every 3 s (total collar mass = 328 g) and pumas wore anintegrated accelerometer-GPS collar (GPS Plus Vectronics Aerospace Germany totalcollar mass = 480 g) that sampled acceleration continuously at 32 Hz and took GPSfixes every 6 s in the field during hound-assisted puma recaptures For both collar typestri-axial accelerometer orientation was such that the X- Y - and Z - axes were parallel tothe transverse anterior-posterior and the dorsal-ventral planes of the animal respectivelyFor collar calibration captive pumas (Williams et al 2014 Wang Allen amp Wilmers 2015)and hounds (Bryce amp Williams 2017) were filmed (Sony HDR-CX290B 1080 HD 60 p)moving across a range of natural speeds (rest to 2 m sminus1 and 47 m sminus1 respectively) whileon a treadmill enclosed by a metabolic chamber Collar-derived accelerometer signatureswere then correlated to gait-specific locomotor costs by simultaneously measuring oxygenconsumption (VO2) and overall dynamic body acceleration (ODBA Qasem et al 2012Wilson et al 2006) of the animals during steady-state resting and treadmill running(Williams et al 2014 Bryce amp Williams 2017) Because both speed and metabolic rateare linked to the dynamic component of an animalrsquos body acceleration (Gleiss Wilson ampShepard 2011 Bidder Qasem ampWilson 2012 Qasem et al 2012 Bidder et al 2012) weused ODBA to translate sensor output from the collars into the speed turning maneuversand energetics of free-ranging individuals

FieldworkAn estimated population of 50ndash100 pumas resides in our 1700 km2 study area in the SantaCruz Mountains of California (371000primeN 122300primeW) The climate is Mediterraneanand elevation ranges from sea level to 1155 m with rugged forested canyons characterizingmuch of the preferred puma habitat Human development (ranging from low-density tourban) is surrounded by native vegetation comprised of redwood and Douglas fir oakwoodland coastal scrubland and grassland communities As a result pumas and nativemesopredators (ie coyotes foxes and bobcats) in the region exhibit spatial and temporalpartitioning of activities that varies with human use (Wang Smith amp Wilmers in pressWilmers et al 2013 Smith Wang amp Wilmers 2015 Smith Wang amp Wilmers 2016 WangAllen amp Wilmers 2015)

Previous work validated the use of accelerometer-GPS collars for describingspatiotemporally explicit puma energetics (Wang Smith amp Wilmers in press Williamset al 2014) and behaviors (Wang Allen amp Wilmers 2015) in the field Here we separatelyrecaptured two adult male pumas (36 M and 26 M 597 plusmn 07 kg) in autumn 2015using packs of 5 and 4 hounds (n= 8 243 plusmn 08 kgWilmers et al 2013) respectively Wetook advantage of this routine capture technique to simultaneously record and quantify thedetailed chase-escape dynamics and associated energetic costs for hounds and pumas In thefield we filmed (Sony HDR-CX290B 1080 HD 60 p) the hound collar being manually


shaken prior to and immediately following deployment for subsequent accelerometerand GPS clock synchronization to Greenwich Mean Time (GMT) Similarly we filmedthe screen (including the clock) of the handheld UHF terminal (Vectronic AerospaceGermany) while uploading the rapid GPS fix schedule to later synchronize the exact timeof the puma collar schedule upload to that of the hound collar clock

Puma chases occurred during daylight hours between 0900 and 1500 local time aperiod that typically corresponds with inactivity for these nocturnal hunters (Fig S1) Eachpuma initially escaped into terrain that precluded darting As a result we re-chased eachpuma after several hours and thus measured a total of four puma pursuits by hounds Allhounds were released simultaneously for each recapture and although only one hound ineach chase wore the combined accelerometer-GPS collar all hounds wore identical GPStracking collars to enable an analysis of hound pack hunting dynamics After escapingto a tree suitable for darting pumas were tranquilized with Telazol at a concentration of100 mgmL measured and re-collared while we collected the previous collar for chasereconstruction and analysis

Ethics statementThis study was conducted in strict accordance with animal ethics including captureand handling as approved by the California Department of Fish and Wildlife(Scientific Collection Permit SC-11968) and the UC Santa Cruz Animal Care and UseCommittee (IACUC Protocol Wilmc1101) All human interventions including captureadministration of immobilizing drugs radio collaring monitoring were done to minimizenegativeadverse impacts on the welfare of the study species

AnalysesFrom each chase we quantified the speed turning energy expenditure and elevation profilerun by each instrumented animal Instantaneous energetics and cost of transport (COTthe energy expended per meter) of pumas and hounds were determined by correlating 2 ssmoothed ODBA (Wilson et al 2006 Shepard et al 2008) to laboratory-derived rates ofoxygen consumption We then used Eqn (5) from Williams et al (2014) to compare theCOT of 60 kg pumas during typical 2-hour pre-kill active hunting activity (ie searchingand stalking) to that of the brief high intensity escape bouts during hound-assistedrecapture To assess the extent of anaerobic exercise for each species we comparedaccelerometer-derived estimates of VO2 during chases to published values of VO2MAX forlions (approx 52 ml O2 kgminus1 minminus1 Taylor et al 1980 Williams et al 2014) and dogs(approx 160 ml O2 kgminus1 minminus1 Seeherman et al 1981Weibel et al 1983) of similar mass

Overground pursuit and escape speeds were quantified by GPS-derived means for allanimals with accelerometer-derived speeds also computed for both pumas and focalhounds instrumented with combined sensors The proportion of time spent not movingwithin each chase was calculated for each species based on the number of 2-secondwindowswhereODBAlt05 gWedownsampled all houndGPSdata to fixes taken every 6 s to accountfor differences in GPS sample rate during chases and permit direct comparisons of houndand puma spatial datasets The precise start and end of pursuits and escapes for hounds


and pumas respectively were determined by post-hoc comparison of GMT-synchronizedvideo recordings obtained in the field and from each collarrsquos raw accelerometer output Thebeginning of each escape was readily apparent from puma accelerometer records as eachanimal had been resting prior to hound release Tag synchronization and data visualizationwas performed in Igor Pro (Wavemetrics Lake Oswego OR USA) Statistical analyses andfigures were produced using JMP Pro13 (SAS Institute Inc Cary NC USA) programR (v 311 R Core Team 2014) and Matlab (Mathworks Inc Natick MA USA) Studyresults are expressed as the mean plusmn SE (α= 005 a priori)

RESULTSCaptive calibrationsFor both hounds and pumas mass-specific metabolic rate increased linearly as a functionof ODBA as described previously for a variety of other terrestrial quadruped species (egBrown et al 2013 Halsey et al 2009 Wilmers et al 2015) according to

VO2 HOUND= 2287 middotODBA+639 (r2= 086n= 83plt 0001) (1)

VO2 PUMA= 5842 middotODBA+352 (r2= 097n= 9plt 0001) (2)

respectively where VO2 is in ml O2 kgminus1 minminus1 and ODBA is in g Similarly speed wasstrongly predicted fromODBA (Bidder Qasem ampWilson 2012 Bidder et al 2012) for bothspecies according to

SpeedHOUND= 256 middotODBAminus032 (r2= 082n= 83plt 0001) (3)

SpeedPUMA= 532 middotODBAminus042 (r2= 085n= 9plt 0001) (4)

where speed is inm sminus1 andODBA is in g Equations (2) and (4) as well as additional pumacollar calibration data are available from Williams et al (2014) and Wang et al (2015)

Chase reconstructionsThe duration distance average speed elevation change and number of hounds involvedin each recapture are summarized for hounds and pumas in Table 2 We presentindividual chase tracks and parameters (Figs 1 S2ndashS4) as well as a Google Earth Pro(earthgooglecom) visualization of chase 4 generated from synchronized puma and houndcollar data (Video S1) In general mean chase distance was three times farther for hounds(1020 plusmn 249 m) than pumas (335 plusmn 63 m t (6)=minus266 p= 0037 Table 2) becausewe released hounds from a distance great enough to not startle pumas prior to release Inthis way we measured the complete and varied escape maneuvers of the puma in responseto the approaching hounds As a result of these longer pursuit distances hound chaseduration (0859 plusmn 0305 mmss) was longer than the associated escape time in pumas(0348 plusmn 0116 mmss) Compared to the initial escape each pumarsquos second escape wasshorter in both distance (247 plusmn 69 vs 423 plusmn 60 m) and duration (0153 plusmn 0056 vs 0544plusmn 0112 mmss Table 2)

As predicted tortuosity (total distance traveled divided by straight-line distance fromstart to end point of run) did not differ significantly between hounds and pumas when


A

C B D E

F

Figure 1 Chase 1 pursuit (red lines= hounds) and escape (blue line= puma) paths (A) including theelevation profile for Brandy a GPS-accelerometer collar equipped hound Insets display ODBA (g B)speed (m sminus1 C) and estimated mass-specific metabolic demand (VO2 in ml O2 kgminus1 minminus1 D) For (B)(C) and (D) mean values are presented as dashed horizontal lines and solid horizontal lines in (D) de-pict VO2MAX for each species Tortuosity plots (proportion of turns in each compass direction (E) andthe elevation profile for the accelerometer-GPS-equipped hound (F) are also presented Map data ccopy 2016Google

Table 2 Summary of pursuit and escape parameters from hounds and pumas respectivelyMeasurement units are enclosed in parentheses Av-erage speed (m sminus1) is GPS-rather than accelerometer-derived and across-chase averages (plusmnSE) are presented

Hound Puma

Hounds(n)

Distance(m)

Duration(mmss)

Avgspeed(m sminus1)

ElevGainLoss (m)

Distance(m)

Duration(mmss)

Avgspeed(m sminus1)

Elevgainloss(m)

Chase 1 5 1270 0737 278 228minus161 482 0656 116 121minus84Chase 2 5 1400 1513 153 306minus157 178 0248 106 70minus32Chase 3 4 1120 1208 154 99minus339 363 0415 133 88minus139Chase 4 4 291 0059 493 80minus75 316 0057 554 89minus110Average 1020 0839 27 178 (54) 3348 0344 227 92 (11)

(250) (0305) (08) minus183 (56) (628) (0115) (109) minus91 (23)

all individuals and chases were grouped (t (22) 104 p= 031 Table 3) because thescent hounds roughly followed each pumarsquos escape path Overground distance traveledaveraged 23 to 29 times farther than straight-line distance indicating extent of turningmaneuvers while running through rugged terrain To prolong the time until capturedpumas employed several adaptive strategies that compensated for physiological constraintsand being outnumbered Evasive maneuvers such as temporarily jumping into treesrunning hairpin turns or figure-of-8 patterns and fleeing up steep wooded hillsides wereall used repeatedly to increase escape distance and postpone being overtaken (Table S1)For example when the hounds were within 35 m of puma 36 M (chase 1) the puma ran


Table 3 Average (plusmnSE) chase tortuosity and speed performance during hound-assisted puma recaptures Average and maximum speeds (m sminus1)are presented for both GPS and accelerometer-derived estimates Sample sizes and measurement units are enclosed in parentheses and results fromWelch two-sample t -tests comparing hound and puma data are included

GPS speed (m sminus1) Accel speed (m sminus1)

Chase Species (animals) Path tortuosity Avg Max Avg Max

1 Hounds (n= 5) 201plusmn 008 233plusmn 009 853plusmn 029 307plusmn 007 52Puma 36 M 222 093plusmn 021 527 369plusmn 027 145

t =minus65 t = 23plt 001 p= 002


t =minus64 t = 31plt 001 plt 001


t =minus87 t =minus22plt 001 p= 003


t =minus14 t = 121p= 016 plt 001

Avg Hound 232plusmn 024 17plusmn 003 70plusmn 038 389plusmn 018 597plusmn 028Puma 292plusmn 052 074plusmn 009 36plusmn 063 294plusmn 004 139plusmn 073

t = 104 t =minus104 t =minus39 t = 52 t = 103p= 031 plt 001 plt 001 plt 001 plt 001

NotesDenotes significant relationship (ple 005) GPS speeds are inherently averaged over 6 s whereas the accelerometer speeds are near instantaneous (see methods)

a figure-of-8 pattern and briefly jumped into a tree As a result hound-puma separationdistance increased by nearly 15 m and the pumarsquos capture was delayed by an additional5 min (Fig S5)

Overall average chase speed as measured by chase distance divided by chase durationwas similar between species (27plusmn 08 m sminus1 and 23 plusmn 109 m sminus1 for hounds and pumasrespectively t (6)=minus031 p= 077 Table 2) However GPS-derived average speeds fromall hounds (including those not outfitted with accelerometers) and pumas suggested thatacross chases hound pursuit speed was twice that of the escaping pumas (17 vs 074 m sminus1

for hounds and pumas respectively t (1870)=minus104 plt 001 Table 3 Fig 2) sincepumas spent larger proportions of each encounter stationary (avg 31 vs 15 stationaryfor pumas and hounds respectively t (6)= 137 p= 022) Using accelerometer-derivedspeed estimates (Eqs (3) and (4)) to resolve running dynamics in finer temporal resolutionwe note that pumas briefly hit peak speeds in excess of 14m sminus1 more than twice the topspeed of the pursuing hounds (52ndash63m sminus1 Table 3) Puma escapes were characterized bysequential high-speed evasive maneuvers interspersed with slow low acceleration periods(Fig 3) In contrast hound pursuit speeds were more constant over the course of eachchase (Fig 1C and Figs S2ndashS4C)


00

01

02

03

04

05

06

07

08

Dens

ity

0 1 2 3 4 5 6 7 8 9Speed (ms)

Figure 2 GPS-derived pursuit and escape speeds for hounds (red) and pumas (blue) respectively dur-ing all chases The mean (plusmnSE) speeds in ms for hounds (17plusmn 003) and pumas (074plusmn 009) are de-picted as dashed vertical lines

Immediately after release all hounds concurrently worked to detect the nearby pumarsquosscent and give chase Average hound chase speed did not differ across individuals for bothpursuits of puma 36 M (chase 1 252 plusmn 007 m sminus1 F4711 071 p= 059 chase 2 174 plusmn003 m sminus1 F4 1653 195 p= 01) but Hound 4 (Crocket) was significantly slower thanthe three other hounds during both pursuits of puma 26 M (Hound 4 137 plusmn 006 m sminus1others 171 plusmn 004 m sminus1 t988 minus44 plt 0001) This was probably a result of Hound4rsquos age (11) over twice as old as the other hounds (average age of 5) involved in 26 Mrsquosrecapture Hound group cohesion (the spacing of individual members in proximity to themoving group centroid measured every 3 s) varied across chases (Fig 4) likely due tointeracting effects including pack composition individual characteristics (eg experienceage sex) topographic complexity and puma scent freshness Tighter spatial clusteringwas observed between the five members of the hound pack pursuing puma 36 M (Figs4A and 4B) than that of the 4-member pack that chased puma 26 M (Figs 4C and 4D)Across chases the maximum path deviation of individual hounds from the centroid of themoving group averaged 131 (plusmn28) meters No single hound was ever beyond 55 m of thetrue path of the puma although the average maximum deviation was 191 plusmn117 m

Energetic demandsAcross chases the metabolic rates (kJ minminus1) of pumas during escape (765 plusmn 151) werenearly four times higher than those of the pursuing hounds (201 plusmn 151 t (6)= 265p= 0038 Fig 5) On a mass-specific basis metabolic rates (kJ kg minminus1) were still gt16timesgreater in pumas relative to hounds (129 plusmn 027 vs 076 plusmn 027 kJ kg minminus1) Similarly


$amp

A B

D C $amp $amp

B A

D amp$amp

C amp$amp

A B

C D

Figure 3 Escape acceleration signatures of adult male pumas 36 (A B) and 26 (C D) Acceleration (g)is scaled to the same range for comparison Chase distance is in m and chase duration is in mmss Colorscorrespond to pumasrsquo accelerometer-GPS collar orientation in the X (transverse sway black) Y (anterior-posterior surge blue) and Z (dorsal-ventral heave red) planes

COT (J kgminus1 mminus1) was gt2times as high for pumas (117 plusmn 14) than hounds (55 plusmn 14t (6)= 305 p= 0023 Fig 5)

Hounds remained below their gas exchange threshold (ie VO2MAX) for the durationof pursuits with peak hound VO2 estimates during the highest-intensity chase (chase 4)of 60 ml O2 kgminus1 minminus1 or just 40 of VO2MAX (Fig 6) In contrast pumas routinelyexceeded VO2MAX during escapes with an average of 525 plusmn 16 (range 32ndash100) ofeach escape requiring energy from anaerobic metabolism

Exercise effort was comparatively larger for pumas compared to hounds (Figs 1D 6and Figs S2ndashS4D) and on average one minute spent escaping cost pumas 464 times (plusmn13)as much energy as a typical minute spent actively hunting In other words the averagepuma escape duration of 0348 (plusmn0116) was metabolically equivalent to about 18 min ofroutine active hunting

DISCUSSIONIn quantifying the fine-scale pursuit and evasion dynamics of two large carnivores includingfree-ranging cryptic pumas we present evidence for morphological and physiologicalconstraints imposed by specialization towards divergent hunting modes Although thehighly cursorial endurance-adapted canids (here scent hounds) exhibited relatively poor


Figure 4 Hound pursuit paths (AndashD) and 2D-spatial histograms (EndashH) of pack cohesion over thecourse of each chase with black arrows indicating direction of chase Every 3 s the group centroidthroughout each pursuit path is marked as a red plus (+) In the spatial histogram insets the relativeposition of each hound relative to the group centroid is scaled by color with warm colors representingclose group cohesion and cool colors depicting more distant spacing

turning ability and maximum speed compared with pumas these animals maintainedlower metabolic rates (Fig 5B) and transport costs (Fig 5D) than their felid quarry Canidshave a higher index of aerobic athleticism (Gillooly Gomez amp Mavrodiev 2017 Taylor etal 1987 Weibel et al 1983) thanks to relatively larger hearts (Williams et al 2015) andgreater lung volumes (Kreeger 2003Murphy amp Ruth 2009) compared with felids of similarsize Like other endurance-adapted species (eg humans horses) during high-intensitymuscular exercise canids likely exhibit higher critical speedpower and lower energy storage(Wprime) than felids (Jones et al 2010 Poole et al 2016) Canid skeletal specializations includelsquobox-likersquo elbow joints and limbs locked in a more prone position (Figueirido et al 2015)enabling wolves for example to travel for several kilometers at 56ndash64 km hrminus1 (Mech1970Mech Smith amp MacNulty 2015) pursue prey over distances in excess of 20 km (Mechamp Korb 1978 Mech Smith amp MacNulty 2015) and cover 76 km in 12 h (Mech amp Cluff2011) As with other social canids hounds worked together effectively as a pack (Fig 4) todetect and maintain each pumarsquos scent while giving chase through steep terrain and densebrush understory Our results indicate that group-hunting hounds exhibit fissionndashfusion


Figure 5 Energetic costs of pursuit and evasion for hounds (white) and pumas (grey) respectivelysummarized across four chases Total metabolic cost (kJ A) metabolic rate (kJ minminus1 B) mass-specificmetabolic rate (kJ kgminus1 minminus1 C) and cost of transport (COT J kgminus1 mminus1 D) are shown Asterisks () de-note significant (ple 005) differences between species

spatial dynamics while giving chase Furthermore they suggest that pack size as well as theage sex and experience level of individuals can influence these complex dynamics Theseand many other factors have been shown to impact group composition and kill success ingroup-hunting predators (reviewed in Gittleman 1989)

In contrast as solitary highly adapted stalk-and-pounce predators pumas rely heavilyon an element of surprise coupled with a short pursuit (le10 m Laundreacute amp Hernaacutendez2003) before making contact with and subduing prey (Murphy amp Ruth 2009) Comparedto canids in felids pouncing and grappling with prey are aided by wider elbow joints(Figueirido et al 2015) greater spinal flexibility (Spoor amp Badoux 1988 Ruben 2010) andother limb and pelvic adaptations (Taylor 1989)We documented the extreme performancecapabilities of this ambush-hunting mode (Table 3 and Fig 3 also seeWilliams et al 2014)as well as the physiological limitations for stamina exacted during the flight response(Figs 5 and 6) For example although brief the maximum puma escape G-forcemeasurements (Fig 3) in excess of plusmn5 g are similar to those experienced during anOlympic luge race or under maximum braking force in a Formula 1 racecar (Gforcesnet2010) Such peak performance capacity is energetically expensive (Figs 1D 6 S2DndashS4D)


Figure 6 Estimated meanmetabolic rate (VO2 ml O2 kgminus1 minminus1) expressed as a percentage ofVO2MAX for hounds (red) and pumas (blue) during each chase

and each pumasrsquo second escape was much shorter in distance and duration In escape 4 forexample puma 26 Mrsquos brief burst of speed (Tables 2 and 3 Fig S4) required short-termenergy stores well beyond those met aerobically (ie above the gas-exchange thresholdFig 6) Our recapture results provide field-based empirical support for the locomotorramifications of these hunting mode differences between large canids and felids

Describing species-specific energetic costs and movement ecology can elucidatepopulation-level consequences of anthropogenic disturbance and environmental change(Stephens amp Krebs 1986 Gorman et al 1998 Wikelski amp Cooke 2006 Somero 2011Seebacher amp Franklin 2012 Cooke et al 2013 Humphries amp McCann 2014 Tomlinsonet al 2014 Laundreacute 2014 Scantlebury et al 2014 Wong amp Candolin 2014) Cumulativecosts associated with exposure to disturbance can tip the energy balance for large carnivoresand potentially lead to demographic changes that reverberate through the ecologicalcommunity (Ripple et al 2014 New et al 2014 King et al 2015)

A recent terrestrial predatorndashprey pursuit model developed by Wilson et al (2015)suggested that during predation interactions the larger animal would be absolutely fasterbut have inferior turning ability Using our GPS-accelerometer datasets we found greatermaximum speeds and turning performance in pumas (weighing over twice the mass ofeach hound) but slower average speeds Differences between our findings and predictionsof the Wilson et al model can be explained in part by recognizing our studyrsquos violationof several underlying model assumptions For instance our recaptures did not occurbetween a solitary predator pursuing solitary prey on flat and homogenous terrain norwere the predator(s) and prey morphologically similar In addition some differences may


be explained by anatomical and physiological specialization in canids and felids as wellas local adaptations to rugged topography enhancing momentary escape performancein pumas

We recognize that our opportunistic hound-assisted puma recaptures constitute semi-natural interactions but nevertheless their analysis serves as an important first stepin understanding the complexities and tradeoffs of locomotor performance vs energyexpenditure in large felids and canids (Hubel et al 2016) For example recaptures alsoenabled us to record the maximal or near-maximal performance capacity of a wild felidpredator shedding light on hunting adaptations for and limits to managing speedmaneuverability and energy demand during prey capture Our approach also providesa framework for quantifying natural competitive or predatory interactions and theiroutcomes in the future Although dogs can adversely affect free-ranging carnivore behavior(eg Vanak et al 2013 Wierzbowska et al 2016) hound-elicited escapes by pumas mayreflect direct interspecific interactions between wolf packs and solitary pumas in sympatriclandscapes While some evidence suggests that pumas are capable of killing subadult (Ruthamp Murphy 2009b) and adult wolves (Schmidt amp Gunson 1985) the pack hunting strategyof wolves generally makes them dominant competitors against solitary pumas during directconflicts (Husseman et al 2003 Kortello Hurd amp Murray 2007 Ruth amp Murphy 2009bRuth et al 2011 Bartnick et al 2013)

Where they coexist wolves and pumas often exhibit temporal as well as spatial nichepartitioning with pumas often utilizing edge habitat (Laundreacute amp Hernaacutendez 2003) orrugged terrain (eg steep slopes boulders) dominated by vegetative cover for concealmentwhen hunting (Logan amp Irwin 1985 Laing amp Lindzey 1991 Williams McCarthy amp Picton1995 Jalkotzy Ross amp Wierzchowski 2002 Husseman et al 2003) Wolves in comparisontend to prefer valley bottoms and open country for hunting (Husseman et al 2003Alexander Logan amp Paquet 2006 Atwood Gese amp Kunkel 2007 Kortello Hurd amp Murray2007) In addition to being critical for hunting cover (Kleiman amp Eisenberg 1973) pumasand other solitary felids rely on structural complexity and vegetative cover as escape terrainduring direct intra- and interspecific conflict (Duke 2001 Ruth 2004 Dickson Jenness ampBeier 2005 Kortello Hurd amp Murray 2007) Furthermore trees may serve as primary andimmediate refuge from wolves and other threats as pumas do not readily utilize trees forother purposes such as arboreal prey caching and consumption observed in other felids(eg lynx leopards Balme et al 2017Vander Waal 1990)We observed this phenomenonin the field each puma escape was characterized by agile high-performance maneuveringthat terminatedwith jumping into a tree immediately prior to being overtaken by the houndpack Maintaining adequate vegetative cover therefore provides a dual concealment-safetybenefit to pumas indicating the importance of protecting complex habitat in addition toadequate prey to ensure the long-term persistence of these cryptic predators (Beier 2009Burdett et al 2010 Laundreacute 2014 Williams et al 2014 Wilmers et al 2013) especiallywhere they co-occur with wolves (Bartnick et al 2013 Kortello Hurd amp Murray 2007Ruth amp Murphy 2009a)


ACKNOWLEDGEMENTSWe thank the California Department of Fish and Game P Houghtaling D TichenorT Collinsworth and several undergraduate assistants for their significant field supportduring puma recaptures We also thank JA Estes O Dewhirst L McHuron B Nickel andP Raimondi for their thoughtful discussions and contributions to the analyses

ADDITIONAL INFORMATION AND DECLARATIONS

FundingSupport was provided by the National Science Foundation (DBI-0963022 DBI-1255913and GK-12 DGE-0947923) and the Gordon and Betty Moore Foundation Additionalsupport came to CMB from the UCSC Science Internship Program (SIP) Mazamas theARCS Foundation and the Ecology and Evolutionary Biology (EEB) Department at UCSanta Cruz The funders had no role in study design data collection and analysis decisionto publish or preparation of the manuscript

Grant DisclosuresThe following grant information was disclosed by the authorsNational Science Foundation DBI-0963022 DBI-1255913 GK-12 DGE-0947923Gordon and Betty Moore FoundationUCSC Science Internship Program (SIP)MazamasARCS FoundationEcology and Evolutionary Biology (EEB)

Competing InterestsThe authors declare there are no competing interests

Author Contributionsbull Caleb M Bryce conceived and designed the experiments performed the experimentsanalyzed the data contributed reagentsmaterialsanalysis tools wrote the paperprepared figures andor tables reviewed drafts of the paperbull Christopher C Wilmers conceived and designed the experiments contributedreagentsmaterialsanalysis tools reviewed drafts of the paperbull Terrie M Williams conceived and designed the experiments reviewed drafts of thepaper

Animal EthicsThe following information was supplied relating to ethical approvals (ie approving bodyand any reference numbers)

This study was conducted in strict accordance with animal ethics including capture andhandling as approved by the the UC Santa Cruz Animal Care and Use Committee


Field Study PermissionsThe following information was supplied relating to field study approvals (ie approvingbody and any reference numbers)

This study was conducted in strict accordance with animal ethics including captureand handling as approved by the California Department of Fish and Wildlife (Permit NoSC-11968)

Data AvailabilityThe following information was supplied regarding data availability

The raw data is provided in the Supplemental Files

Supplemental InformationSupplemental information for this article can be found online at httpdxdoiorg107717peerj3701supplemental-information

REFERENCESAlexander SM Logan TB Paquet PC 2006 Spatio-temporal co-occurrence of

cougars (Felis concolor) wolves (Canis lupus) and their prey during winter acomparison of two analytical methods Journal of Biogeography 332001ndash2012DOI 101111j1365-2699200601564x

Atwood TC Gese EM Kunkel KE 2007 Comparative patterns of predation by cougarsand recolonizing wolves in Montanarsquos Madison Range The Journal of WildlifeManagement 711098ndash1106 DOI 1021932006-102

Balme GA Miller JRB Pitman RT Hunter LTB 2017 Caching reduces klep-toparasitism in a solitary large felid Journal of Animal Ecology 86634ndash644DOI 1011111365-265612654

Bartnick TD Deelen TR Van Quigley HB Craighead D 2013 Variation in cougar(Puma concolor) predation habits during wolf (Canis lupus) recovery in thesouthern Greater Yellowstone Ecosystem Canadian Journal of Zoology 9382ndash93DOI 101139cjz-2012-0147

Beier P 2009 A focal species for conservation planning In Hornocker MG Negri Seds Puma ecology and conservation Chicago University of Chicago Press 177ndash189

Bidder OR Qasem LAWilson RP 2012 On higher ground how well can dy-namic body acceleration determine speed in variable terrain PLOS ONE 71ndash7DOI 101371journalpone0050556

Bidder OR SoresinaM Shepard ELC Halsey LG Quintana F Goacutemez-Laich AWilson RP 2012 The need for speed testing acceleration for estimating an-imal travel rates in terrestrial dead-reckoning systems Zoology 11558ndash64DOI 101016jzool201109003

Brodie EDI Brodie EDJ 1999 Predator-prey arms races asymmetrical selection onpredators and prey may be reduced when prey are dangerous BioScience 49557ndash568DOI 1023071313476


Brown DD Kays RWikelski MWilson R Klimley A 2013 Observing the unwatch-able through acceleration logging of animal behavior Animal Biotelemetry 120DOI 1011862050-3385-1-20

Bryce CMWilliams TM 2017 Comparative locomotor costs of domestic dogs revealenergetic economy of wolf-like breeds Journal of Experimental Biology 220312ndash321DOI 101242jeb144188

Burdett CL Crooks KR Theobald DMWilson KR Boydston EE Lyren LM FisherRN Vickers TWMorrison SA BoyceWM 2010 Interfacing models of wildlifehabitat and human development to predict the future distribution of puma habitatEcosphere 11ndash21 DOI 101890ES10-000051

Combes SA Rundle DE Iwasaki JM Crall JD 2012 Linking biomechanics and ecologythrough predatorndashprey interactions flight performance of dragonflies and their preyJournal of Experimental Biology 215903ndash913 DOI 101242jeb059394

Cooke SJ Hinch SG Lucas MC Lutcavage M 2013 Biotelemetry and biologging InZale AV Parrish DL Sutton TM eds Fisheries techniques Bethesda AmericanFisheries Society 1ndash64

Cortez MH 2011 Comparing the qualitatively different effects rapidly evolving andrapidly induced defences have on predatorndashprey interactions Ecology Letters14202ndash209 DOI 101111j1461-0248201001572x

Dawkins R Krebs JR 1979 Arms races between and within species Proceedings of theRoyal Society B Biological Sciences 205489ndash511 DOI 101098rspb19790081

Dickson BG Jenness JS Beier P 2005 Influence of vegetation topography and roadson cougar movement in southern California Journal of Wildlife Management69264ndash276 DOI 1021930022-541X(2005)069lt0264IOVTARgt20CO2

Domenici P 2001 The scaling of locomotor performance in predatorndashprey encountersfrom fish to killer whales Comparative Biochemistry and Physiology Part A A Molecu-lar and Integrative Physiology 131169ndash182 DOI 101016S1095-6433(01)00465-2

Domenici P Blake RW 1997 The kinematics and performance of fish fast-startswimming The Journal of Experimental Biology 2001165ndash1178

Donadio E Buskirk SW 2006 Diet morphology and interspecific killing in carnivoraThe American Naturalist 167524ndash536 DOI 101086501033

Duke D 2001Wildlife use of corridors in the central Canadian Rockies multivariate use ofhabitat characteristics and trends in corridor use Edmonton University of Alberta

Elbroch LM Lendrum PE Newby J Quigley H Thompson DJ 2015 Recolonizingwolves influence the realized niche of resident cougars Zoological Studies 5441DOI 101186s40555-015-0122-y

Elliott JP Cowan IM Holling CS 1977 Prey capture by the African lion CanadianJournal of Zoology 551811ndash1828 DOI 101139z77-235

Figueirido B Martiacuten-Serra A Tseng ZJ Janis CM 2015Habitat changes and changingpredatory habits in North American fossil canids Nature Communications 6Article8976 DOI 101038ncomms8976

Gese EM 2001 Monitoring of terrestrial carnivore populations In Gittleman JL FunkSM MacDonald DW Wayne RK eds Carnivore conservation London CambridgeUniversity Press 372ndash396


Gforcesnet 2010 Some typical examples of G forces Available at httpwwwgforcesnet(accessed on 01 January 2016)

Gillooly JF Gomez JP Mavrodiev EV 2017 A broad-scale comparison of aerobicactivity levels in vertebrates endotherms versus ectotherms Proceedings of the RoyalSociety B Biological Sciences 28420162328 DOI 101098rspb20162328

Gittleman JL 1989 Carnivore group living comparative trends In Gittleman JL edCarnivore behavior ecology and evolution New York Cornell University Press183ndash207

Gleiss ACWilson RP Shepard ELC 2011Making overall dynamic body accelerationwork on the theory of acceleration as a proxy for energy expenditureMethods inEcology and Evolution 223ndash33 DOI 101111j2041-210X201000057x

GormanML Mills MG Raath JP Speakman JR 1998 High hunting costs makeAfrican wild dogs vulnerable to kleptoparasitism by hyaenas Nature 852479ndash481DOI 10103835131

Halsey LG Shepard ELC Quintana F Gomez Laich A Green JAWilson RP 2009The relationship between oxygen consumption and body acceleration in a range ofspecies Comparative Biochemistry and Physiology Part A Molecular amp IntegrativePhysiology 152197ndash202 DOI 101016jcbpa200809021

Hebblewhite M Merrill E 2008Modelling wildlife-human relationships for socialspecies with mixed-effects resource selection models Journal of Applied Ecology45834ndash844 DOI 101111jl365-2664200801466x

Hebblewhite M Merrill EH McDonald TL 2005 Spatial decomposition of predationrisk using resource selection functions an example in a wolf-elk predatorndashpreysystem Oikos 111101ndash111 DOI 101111j0030-1299200513858x

Hedenstroumlm A RoseacutenM 2001 Predator versus prey on aerial hunting and escapestrategies in birds Behavioral Ecology 12150ndash156 DOI 101093beheco122150

Hornocker MG 1970 An analysis of mountain lion predation upon mule deer and elkin the Idaho Primitive Area InWildlife monographs Vol 21 Bethesda The WildlifeSociety 3ndash39

Hornocker M Negri S (eds) 2009 Cougar ecology and conservation Chicago Universityof Chicago Press

Howland HC 1974 Optimal strategies for predator avoidance the relative impor-tance of speed and manoeuvrability Journal of Theoretical Biology 47333ndash350DOI 1010160022-5193(74)90202-1

Hubel TY Myatt JP Jordan NR Dewhirst OP McNutt JWWilson AM 2016 Energycost and return for hunting in African wild dogs and cheetahs Nature Communica-tions 7Article 11034 DOI 101038ncomms11034

Humphries MMMcCann KS 2014Metabolic ecology The Journal of Animal Ecology837ndash19 DOI 1011111365-265612124

Husseman JS Murray DL Power G Mack CWenger CR Quigley H 2003 Assessingdifferential prey selection patterns between two sympatric large carnivores OIKOS101591ndash601 DOI 101034j1600-0706200312230x


Jalkotzy MG Ross PI Wierzchowski J 2002 Regional scale cougar habitat modeling inSouthwestern Alberta Canada In Proceedings of the Sixth Mountain Lion WorkshopAustin Texas Parks and Wildlife Department

Jones AM Vanhatalo A Burnley M Morton RH Poole DC 2010 Critical powerimplications for determination of VO2MAX and exercise toleranceMedicine ampScience in Sports amp Exercise 421876ndash1890 DOI 101249MSS0b013e3181d9cf7f

Karanth KU Sunquist ME 1995 Prey selection by tiger leopard and dhole in tropicalforests Journal of Animal Ecology 64439ndash450

KauffmanMJ Varley N Smith DW Stahler DR MacNulty DR Boyce MS 2007Landscape heterogeneity shapes predation in a newly restored predatorndashprey systemEcology Letters 10690ndash700 DOI 101111j1461-0248200701059x

Kays R Crofoot MC JetzWWikelski M 2015 Terrestrial animal tracking as an eye onlife and planet Science 348 aaa2478-1ndashaaa2478-9 DOI 101126scienceaaa2478

King SL Schick RS Donovan C Booth CG BurgmanM Thomas L Harwood J 2015An interim framework for assessing the population consequences of disturbanceMethods in Ecology and Evolution 61150ndash1158 DOI 1011112041-210X12411

Kleiman DG Eisenberg JF 1973 Comparisons of canid and felid social systems from anevolutionary perspective Animal Behaviour 21637ndash659DOI 101016S0003-3472(73)80088-0

Koehler GM Hornocker MG 1991 Seasonal resource use among mountain lionsbobcats and coyotes Journal of Mammalogy 72391ndash396

Kortello AD Hurd TE Murray DL 2007 Interactions between cougars (Puma con-color) and gray wolves (Canis lupus) in Banff National Park Alberta Ecoscience14214ndash222 DOI 1029801195-6860(2007)14[214IBCPCA]20CO2

Kreeger TJ 2003 The internal wolf physiology pathology amp pharmacology In MechLD Boitani L edsWolves behavior ecology amp conservation Chicago University ofChicago Press 192ndash217

KrummCE Conner MM Thompson Hobbs N Hunter DOMiller MW 2010Moun-tain lions prey selectively on prion-infected mule deer Biology Letters 6209ndash211

Kunkel KE Ruth TK Pletscher DH Hornocker MG 1999Winter prey selection bywolves and cougars in and near Glacier National Park Montana Journal of WildlifeManagement 63901ndash910

Laing SP Lindzey FP 1991 Cougar habitat selection in south-central utah In BraunCE edMountion lion-human interaction symposium and workshop DenverColorado Division of Wildlife 86ndash94

Laundreacute JW 2014How large predators manage the cost of hunting Science 34633ndash34DOI 101126science1260170

Laundreacute JW Hernaacutendez L 2003Winter hunting habitat of pumas (Puma concolor) innorthwestern Utah and southern Idaho USAWildlife Biology 9123ndash129

Linnell JDC Strand O 2000 Interference interactions co-existence and con-servation of mammalian carnivores Diversity and Distributions 6169ndash176DOI 101046j1472-4642200000069x

Logan KA Irwin LL 1985Mountain lion habitats in the Big Horn MountainsWyomingWildlife Society Bulletin 13257ndash262 DOI 1023073782489


Maresh JL Fish FE Nowacek DP Nowacek SMWells RS 2004 High performanceturning capabilities during foraging by bottlenose dolphins (Tursiops Truncatus)Marine Mammal Science 20498ndash509 DOI 101111j1748-76922004tb01175x

Mech LD 1970 The wolf the behavior and ecology of an endangered species New YorkNatural History Press

Mech LD Boitani L (eds) 2003Wolves behavior ecology amp conservation ChicagoUniversity of Chicago Press

Mech LD Cluff HD 2011Movements of wolves at the northern extreme of thespeciesrsquo range including during four months of darkness PLOS ONE 62ndash6DOI 101371journalpone0025328

Mech LD KorbM 1978 An unusually long pursuit of a deer by a wolf Journal ofMammalogy 59860ndash861 DOI 1023071380155

Mech LD Smith DWMacNulty DR 2015 Introduction the wolf as a killing machineInWolves on the hunt the behavior of wolves hunting wild prey Chicago Universityof Chicago Press

Murphy K Ruth TK 2009 Diet and prey selection of a perfect predator In HornockerM Negri S eds Cougar Ecology amp Conservation Chicago University of ChicagoPress 118ndash137

New L Clark J Costa D Fleishman E Hindell M Klanjšček T Lusseau D Kraus SMcMahon C Robinson P Schick R Schwarz L Simmons S Thomas L TyackP Harwood J 2014 Using short-term measures of behaviour to estimate long-term fitness of southern elephant sealsMarine Ecology Progress Series 49699ndash108DOI 103354meps10547

Okarma H JedrzejewskiW Schmidt K Kowalczyk R Jedrzejewska B 1997 Predationof Eurasian lynx on roe deer and red deer in Białowieza Primeval Forest PolandActa Theriologica 42203ndash224

Peterson RO Ciucci P 2003 The wolf as a carnivore In Mech LD Boitani L edsWolves behavior ecology amp conservation Chicago University of Chicago Press104ndash130

Podgoacuterski T Schmidt K Kowalczyk R Gulczyntildeska A 2008Microhabitat selectionby Eurasian lynx and its implications for species conservation Acta Theriologica5397ndash110 DOI 101007BF03194243

Poole DC Burnley M Rossiter HB Jones AM 2016 Critical power an importantfatigue threshold in exercise physiologyMedicine amp Science in Sports amp Exercise48(11)2320ndash2334

Poole DC Erickson HH 2011Highly athletic terrestrial mammals horses and dogsComprehensive Physiology 11ndash37 DOI 101002cphyc091001

Qasem L Cardew AWilson A Griffiths I Halsey LG Shepard ELC Gleiss ACWilson R 2012 Tri-axial dynamic acceleration as a proxy for animal energyexpenditure should we be summing values or calculating the vector PLOS ONE 7DOI 101371journalpone0031187

R Core Team 2014 R a language and environment for statistical computing Availableat httpwwwr-projectorg


RippleWJ Estes JA Beschta RLWilmers CC Ritchie EG Hebblewhite M Berger JElmhagen B Letnic M NelsonMP Schmitz OJ Smith DWWallach ADWirsingAJ 2014 Status and ecological effects of the worldrsquos largest carnivores Science343151ndash163 DOI 101126science1241484

Rosenzweig ML 1966 Community structure in sympatric carnivora Journal of Mam-malogy 47602ndash612 DOI 1023071377891

Ruben JA 2010 The role of the vertebral column during jumping in quadrupedalmammals PhD thesis Oregon State University

Ruth TS 2004 Patterns of resource use among cougars and wolves in northwesternMontana and southeastern British Columbia PhD thesis University of IdahoMoscow Idaho

Ruth TK HaroldsonMAMurphy KM Buotte PC Hornocker MG Quigley HB 2011Cougar survival and sourcendashsink structure on Greater Yellowstonersquos NorthernRange Journal of Wildlife Management 751381ndash1398 DOI 101002jwmg190

Ruth T Murphy K 2009a Cougar-prey relationships In Hornocker M Negri S edsCougar ecology amp conservation Chicago University of Chicago Press 138ndash162

Ruth T Murphy K 2009b Competition with other carnivores for prey In HornockerM Negri S eds Cougar ecology amp conservation Chicago University of ChicagoPress 163ndash172

Scantlebury DMMills MGLWilson RPWilson JWMills MEJ Durant SM BennettNC Bradford P Marks NJ Speakman JR 2014 Flexible energetics of cheetahhunting strategies provide resistance against kleptoparasitism Science 34679ndash81DOI 101126science1256424

Schmidt KP Gunson JR 1985 Evaluation of wolf-ungulate predation near NordeggAlberta Second year progress report (1984ndash1985) Alberta Energy and NaturalResources Fish and Wildlife Division Alberta

Schmidt K Kuijper DPJ 2015 A lsquolsquodeath traprsquorsquo in the landscape of fearMammalResearch 60275ndash285 DOI 101007s13364-015-0229-x

Seebacher F Franklin CE 2012 Determining environmental causes of biologicaleffects The need for a mechanistic physiological dimension in conservation biologyPhilosophical Transactions of the Royal Society of London Series B Biological Sciences3671607ndash1614 DOI 101098rstb20120036

Seeherman HJ Taylor CR Maloiy G Armstrong RB 1981 Design of the mammalianrespiratory system II measuring maximum aerobic capacity Respiration Physiology4411ndash23 DOI 1010160034-5687(81)90074-8

Seidensticker J Hornocker MWilesWMessick J 1973 Mountain lion social organi-zation in the Idaho Primitive Area InWildlife monographs Vol 35 Bethesda TheWildlife Society 60

Shepard EWilson R Halsey L Quintana F Goacutemez Laich A Gleiss A Liebsch NMyers A Norman B 2008 Derivation of body motion via appropriate smoothingof acceleration data Aquatic Biology 4235ndash241 DOI 103354ab00104

Smith JAWang YWilmers CC 2015 Top carnivores increase their kill rates on preyas a response to human-induced fear Proceedings of the Royal Society B BiologicalSciences 2821ndash6


Smith JAWang YWilmers CC 2016 Spatial characteristics of residential developmentshift large carnivore prey habits The Journal of Wildlife Management 801040ndash1048DOI 101002jwmg21098

SnowDH 1985 The horse and dog elite athletesmdashwhy and how Proceedings of theNutrition Society 44267ndash272 DOI 101079PNS19850046

Somero GN 2011 Comparative physiology a lsquolsquocrystal ballrsquorsquo for predicting consequencesof global change American Journal of Physiology Regulatory Integrative and Compar-ative Physiology 301R1ndashR14 DOI 101152ajpregu007192010

Spoor CF Badoux DM 1988 Descriptive and functional myology of the back andhindlimb of the striped hyena (Hyaena hyaena L 1758) Anatomischer Anzeiger167313ndash321

Stephens DW Krebs JR 1986 Foraging theory Princeton NJ Princeton UniversityPress

Taylor ME 1989 Locomotor adaptations by carnivores In Gittleman JL ed Carnivorebehavior ecology and evolution Ithaca Cornell University Press 382ndash409

Taylor CR Karas RHWeibel ER Hoppeler H 1987 Adaptive variation in the mam-malian respiratory system in relation to energetic demand II Reaching the limits tooxygen flow Respiration Physiology 697ndash26 DOI 1010160034-5687(87)90098-3

Taylor R Maloiy GWeibel E Al E 1980 Design of the mammalian respiratory systemIII Scaling maximum aerobic capacity to body mass wild and domestic mammalsRespiration Physiology 4425ndash37

Tomlinson S Arnall SG Munn A Bradshaw SD Maloney SK Dixon KW DidhamRK 2014 Applications and implications of ecological energetics Trends in Ecologyamp Evolution 291ndash11 DOI 101016jtree201403003

Vanak AT Dickman CR Silva-Rodriguez EA Butler JRA Ritchie EG 2013 Top-dogsand under-dogs competition between dogs and sympatric carnivores In GompperME ed Free-ranging dogs and wildlife conservation Oxford Oxford University Press

VanderWaal SB 1990 Food hoarding in animals Chicago University of Chicago PressWang Y AllenMLWilmers CC 2015Mesopredator spatial and temporal responses to

large predators and human development in the Santa Cruz Mountains of CaliforniaBiological Conservation 19023ndash33 DOI 101016jbiocon201505007

Wang Y Nickel B Rutishauser M Bryce CMWilliams TM Elkaim GWilmersCC 2015Movement resting and attack behaviors of wild pumas are re-vealed by tri-axial accelerometer measurementsMovement Ecology 31ndash12DOI 101186s40462-015-0030-0

Wang Y Smith JAWilmers CC Residential development alters behavior movementand energetics in a top carnivore PLOS ONE In Press

Warrick DR 1998 The turning- and linear-maneuvering performance of birds the costof efficiency for coursing insectivores Canadian Journal of Zoology 761063ndash1079DOI 101139z98-044

Weibel ER Taylor CR Neil JJO Leith DE Gehr P Hoppeler H Langman V BaudinetteRV 1983Maximal oxygen consumption and pulmonary diffusing capacity a directcomparison of physiological and morphometric measurements in canids RespirationPhysiology 54173ndash188 DOI 1010160034-5687(83)90055-5


Wierzbowska IA HędrzakM Popczyk B Okarma H Crooks KR 2016 Preda-tion of wildlife by free-ranging domestic dogs in Polish hunting grounds andpotential competition with the grey wolf Biological Conservation 2011ndash9DOI 101016jbiocon201606016

Wikelski M Cooke SJ 2006 Conservation physiology Trends in Ecology amp Evolution2138ndash46 DOI 101016jtree200510018

Williams TM Bengtson P Steller DL Croll DA Davis RW 2015 The healthy heartlessons from naturersquos elite athletes Physiology 30349ndash357DOI 101152physiol000172015

Williams JS McCarthy JJ Picton HD 1995 Cougar habitat use and food habits on theMontana Rocky Mountain front Intermountain Journal of Science 116ndash28

Williams TMWolfe L Davis T Kendall T Richter BWang Y Bryce C Elkaim GHWilmers CC 2014 Instantaneous energetics of puma kills reveal advantage of felidsneak attacks Science 34681ndash85 DOI 101126science1254885

Wilmers CC Nickel B Bryce CM Smith JAWheat RE Yovovich V 2015 The goldenage of bio-logging how animal-borne sensors are advancing the frontiers of ecologyEcology 961741ndash1753 DOI 10189014-14011

Wilmers CC Post E Hastings A 2007 The anatomy of predatorndashprey dynamics in achanging climate The Journal of Animal Ecology 761037ndash1044DOI 101111j1365-2656200701289x

Wilmers CCWang Y Nickel B Houghtaling P Shakeri Y AllenML Kermish-WellsJ Yovovich VWilliams T 2013 Scale dependent behavioral responses to humandevelopment by a large predator the puma PLOS ONE 8DOI 101371journalpone0060590

Wilson AM Lowe JC Roskilly K Hudson PE Golabek KA McNutt JW 2013aLocomotion dynamics of hunting in wild cheetahs Nature 498185ndash189DOI 101038nature12295

Wilson JWMills MGLWilson RP Peters G Mills MEJ John R Durant SM BennettNC Marks NJ ScantleburyM Speakman JR 2013b Cheetahs Acinonyx jubatusbalance turn capacity with pace when chasing prey Biology Letters 92ndash6

Wilson RP Griffiths IWMills MG Carbone CWilson JW Scantlebury DM 2015Mass enhances speed but diminishes turn capacity in terrestrial pursuit predatorsLife 41ndash18 DOI 107554eLife06487

Wilson RPWhite CR Quintana F Halsey LG Liebsch N Martin GR Butler PJ2006Moving towards acceleration for estimates of activity-specific metabolic ratein free-living animals the case of the cormorant The Journal of Animal Ecology751081ndash1090 DOI 101111j1365-2656200601127x

Wong BBM Candolin U 2014 Behavioral responses to changing environmentsBehavioral Ecology 1ndash9 DOI 101249MSS0000000000000939

Young SP Goldman EA 1946 The puma mysterious American cat Washington DCAmerican Wildlife Institute


Table 1 Comparison of hunting mode divergence observed in large felids and canids Selected references for each topic (superscripts) are pro-vided below

Family Felidae Family Canidae

Eg puma leopard jaguar Eg gray wolf hound dingo

Hunting modea Cryptic stalking amp pouncing lsquolsquoSurprise ampsubduersquorsquo

Cursorial pursuit lsquolsquoCharge amp chasersquorsquo

Hunting socialityb Solitary Often grouppackRelative prey selectivity and timingc Low (opportunistic) Prior to attack High (selective) Often during pursuitInteraction with amp risk imposed by preyd Short Lower risk of injurydeath Prolonged Higher risk of injurydeathKill site attributese Sufficient structural cover for concealment

during stalking and brief pursuitRelatively open terrain that facilitatesprolonged pursuit

Scale of habitat features impacting huntsuccessf

Small-scale habitat features Large-scale landscape heterogeneity

Relative activity and energetic demand ofhuntrsquos attack phase

High intensity short duration Low intensity long duration

NotesaHornocker 1970 Koehler amp Hornocker 1991 Ruth amp Murphy 2009a Seidensticker et al 1973 Young amp Goldman 1946 Poole amp Erickson 2011 Snow 1985Mech amp Korb 1978Mech amp Cluff 2011

bGittleman 1989 Hornocker amp Negri 2009Mech Smith amp MacNulty 2015Mech 1970cHusseman et al 2003Wilmers Post amp Hastings 2007 Kunkel et al 1999 Okarma et al 1997Mech Smith amp MacNulty 2015 Peterson amp Ciucci 2003 but see Karanth amp Sun-quist 1995 Krumm et al 2010

dHornocker amp Negri 2009Mech Smith amp MacNulty 2015Mech amp Boitani 2003eAlexander Logan amp Paquet 2006 Hebblewhite amp Merrill 2008 Husseman et al 2003 Ruth et al 2011 Schmidt amp Kuijper 2015 and references thereinfHebblewhite Merrill amp McDonald 2005 Kauffman et al 2007 Laundreacute amp Hernaacutendez 2003 Podgoacuterski et al 2008 Schmidt amp Kuijper 2015

amp Erickson 2011) rather than speed or agility to test and ultimately outperform morevulnerable prey (Peterson amp Ciucci 2003 Mech Smith amp MacNulty 2015) In contrastpumas (Puma concolor) exhibit an opportunistic (ie less selective Husseman et al2003 Wilmers Post amp Hastings 2007) solitary and cryptic hunting mode by which theystealthily ambush and overpower prey (Hornocker amp Negri 2009 Ruth amp Murphy 2009a)through matching pounce force to prey size (Williams et al 2014) Such divergence inlocomotion sociality prey selectivity and even preferred terrain while hunting reducesexploitative and interspecific competition (Husseman et al 2003 Elbroch et al 2015)through spatiotemporal niche partitioning and has cascading ecosystem-wide effects(Rosenzweig 1966 Linnell amp Strand 2000 Donadio amp Buskirk 2006 Elbroch et al 2015)Less is known however of how the fine-scale movement performance and metabolicdemands associated with these distinct predatory hunting modes interact to affectpredation success

During a predation event an animalrsquos ability to readily adjust its speed (Howland 1974Domenici 2001) acceleration (Combes et al 2012 Wilson et al 2013b) and turn capacity(Howland 1974 Maresh et al 2004 Wilson et al 2013a) becomes critical for survivalDespite its relative brevity the attack phase of the hunt may be the most energeticallyexpensive stage of prey acquisition particularly for ambush predators (Williams et al2014) Given the two-dimensional confines of the terrestrial environment both predators




























A

C B D E

F



Hound Puma

Hounds(n)

Distance(m)

Duration(mmss)

Avgspeed(m sminus1)

ElevGainLoss (m)

Distance(m)

Duration(mmss)

Avgspeed(m sminus1)

Elevgainloss(m)


(250) (0305) (08) minus183 (56) (628) (0115) (109) minus91 (23)







t =minus65 t = 23plt 001 p= 002






t =minus14 t = 121p= 016 plt 001








00

01

02

03

04

05

06

07

08

Dens

ity

0 1 2 3 4 5 6 7 8 9Speed (ms)





$amp

A B

D C $amp $amp

B A

D amp$amp

C amp$amp

A B

C D




























































































































































































A

C B D E

F



Hound Puma

Hounds(n)

Distance(m)

Duration(mmss)

Avgspeed(m sminus1)

ElevGainLoss (m)

Distance(m)

Duration(mmss)

Avgspeed(m sminus1)

Elevgainloss(m)


(250) (0305) (08) minus183 (56) (628) (0115) (109) minus91 (23)







t =minus65 t = 23plt 001 p= 002






t =minus14 t = 121p= 016 plt 001








00

01

02

03

04

05

06

07

08

Dens

ity

0 1 2 3 4 5 6 7 8 9Speed (ms)





$amp

A B

D C $amp $amp

B A

D amp$amp

C amp$amp

A B

C D






































































































































































t =minus65 t = 23plt 001 p= 002






t =minus14 t = 121p= 016 plt 001








00

01

02

03

04

05

06

07

08

Dens

ity

0 1 2 3 4 5 6 7 8 9Speed (ms)





$amp

A B

D C $amp $amp

B A

D amp$amp

C amp$amp

A B

C D


































































































































































00

01

02

03

04

05

06

07

08

Dens

ity

0 1 2 3 4 5 6 7 8 9Speed (ms)





$amp

A B

D C $amp $amp

B A

D amp$amp

C amp$amp

A B

C D


































































































































































$amp

A B

D C $amp $amp

B A

D amp$amp

C amp$amp

A B

C D


































































































































































energetics and evasion dynamics of large predators and ... · gray wolves (canis lupus), for...

Documents