Using Radiotelemetry to Monitor Cardiac Response of Free-Living Tule Greater White-Fronted Geese (Anser albifrons elgasi) to Human Disturbance

2004·4.2 MB·English

by  Joshua T Ackerman

Wilson Bulletin, 116(2), 2004, pp. 146—151
USING RADIOTELEMETRY TO MONITOR CARDIAC RESPONSE
OF FREE-LIVING TULE GREATER WHITE-FRONTED GEESE
(ANSER ALBIFRONS ELGASI) TO HUMAN DISTURBANCE
JOSHUA T. ACKERMAN,'-’ JOHN Y. TAKEKAWA,^ KAMMIE L. KRUSE,“
DENNIS L. ORTHMEYER,’^ JULIE L. YEE,’ CRAIG R. ELY," DAVID H. WARD,"
KAREN S. BOLLINGER,"* AND DANIEL M. MULCAHY"
—
ABSTRACT. We monitored the heart rates offree-living Tule Greater White-fronted Geese {Anseralhifrons
elgasi) during human disturbances on their wintering range in the Sacramento Valley ofCaliforniaduring 1997.
We used implanted radio transmitters to record the heart ratesofgeese as an observerexperimentallyapproached
them at a constant walking speed. On average, geese flushed when observers were 47 m (range: 25-100 m)
away. Change point regression was used to identify the point in time when heart rate abruptly increased prior
to flushing and when heart rate began to level off in flight after flushing. Heart rates ofgeese increased as the
observerapproached them during five ofsix disturbance trials, from 1 14.1 ± 6.6 beats/min duringtheobserver’s
initial approach to 154.8 ± 7.4 beats/minjust priorto flushing at the first change point. On average, goose heart
rates began to increase most rapidly 5 sec priorto taking flight, and continued to increase rapidly for4 sec after
flushing until reaching flight speed. Heart rate was 456.2 ± 8.4 beats/min at the second change point, which
occurred immediately after flushing, and 448.3 ± 9.5 beats/min 1 min later during flight. Although goose heart
rates increased as an observer approached, the largest physiological change occurred during a 9-sec period
(range: 1.0-15.7 sec) immediately before and afterflushing, when heart rates nearly tripled. Received28October
2003, accepted 12 April 2004.
Wildlife may incur energetic costs when to human activities (Richardson and Miller
disturbed by humans; therefore, spatial buffer 1997, Rodgers and Smith 1997, Ward et al.
zones are commonly used by managers to sep- 1999, Swarthout and Steidl 2001, Blumstein
arate wildlife from human activities (Knight et al. 2003). However, wildlife may have more
and Temple 1995, Camp et al. 1997, Richard- subtle physiological responses to disturbances,
son and Miller 1997, Rodgers and Smith such as increased heart rate, that also have en-
1997). Buffer zones often are based on the ergetic consequences (MacArthur et al. 1982,
distance at which a species flushes in response Gabrielsen and Smith 1995, Andersen et al.
1996, Weisenberger et al. 1996). Heart rate is
a good indicator of an animal’s energy expen-
sea'rUc.hS.CenGteeorl,ogDiacvailsSFuirevledyS,taWteiosnt,erOnneEcSohlioeglidcsalAvRee.-, diture (Wooley and Owen 1977, 1978; Woak-
Univ. ofCalifornia, Davis, CA 95616, USA. es and Butler 1983; Nolet et al. 1992) and
2U.S. Geological Survey, Western Ecological Re- depends not only on an animal’s physical ac-
search Center, San Francisco Bay Estuary Field Sta- tivity, but also on the level of external stim-
tion, 505 Azuar Dr., Vallejo, CA 94592, USA. ulation caused by disturbances (Harms et al.
U.S. Geological Survey, Western Ecological Re-
search Center, Dixon Field Station, 6924Tremont Rd., 1997, Ely et al. 1999). Therefore, heart rate,
Dixon, CA 95620, USA. rather than flushing distance, may be a better
U.S. Geological Survey, Western Ecological Re- indicator of an animal’s stress response to hu-
search Center, 3020 State Univ. Dr. East, Modoc Hall, man disturbance (Fernandez-Juricic et al.
Rm. 3006, Sacramento, CA 95819, USA. 2001
1011U.ES..TGuedoolrogRidc.a,lAnScurhvoerya,ge,AlAasKka99S5c0i3e,ncUeSACe.nter, In )t.his paper, we provide the first radiote-
^Current address: U.S. Fish and Wildlife Service, lemetry-based data on the cardiac response of
P.O. Box 25486, DEC, Denver, CO 80225, USA. free-living waterfowl to human disturbance.
^Currentaddress: CaliforniaWaterfowl Association, We used implanted radio transmitters to mon-
4630 Northgate Blvd., Suite 150, Sacramento, CA itor the heart rates of Tule Greater White-
95834, USA. fronted Geese {Anser alhifrons elgasi, here-
^Current address: U.S. Fish and Wildlife Service,
1 1“5^0C0orArmeesrpiocndainngHoalultyhoDrr;.,e-Lmaauirle:l, MD 20708, USA. taafltleyr dTiuslteurGbeeedseea)chasraadnioo-bmsaerrkveerd beixrpderbiymeanp--
[email protected] proaching it at a constant walking speed. We
146
Ackerman et al. • GOOSE HEART RATES AND DISTURBANCE 147
report the variation in heart rate throughout simple, interrupted sutures of 0-vicryl placed
the disturbance event from roosting or feeding through the mesh so that the antenna was sit-
to flushing and flight, and use change point uated on the dorsal side ofthe abdominal cav-
analysis to identify points in time when heart ity, thus increasing transmission range. The
rate abruptly changed. incision was closed in two layers using 0-vi-
cryl in a simple continuous pattern. We re-
METHODS
leased the radio-marked geese 2 days after
Tule Geese are larger (>0.5 kg) than the their capture.
other subspecies of Greater White-fronted We examined the variation in heart rate
Geese {A. a. frontalis) wintering in California when Tule Geese were disturbed by approach-
(Ely and Dzubin 1994). We captured flightless ing a radio-marked goose on foot and simul-
Tule Geese by driving them into net corrals. taneously recording both its heart rate and the
We implanted radio transmitters in 28 adult time since we initiated the disturbance. During
female geese on 6 and 8 July 1997 at Kahiltna daylight hours, we located radio-marked in-
River Valley, Alaska (see Ely et al. 1999), as dividuals that were roosting or feeding in ref-
part of an aircraft disturbance study of Tule uge ponds or in harvested rice fields that were
Geese in southcentral Alaska. Thirteen of the protected from hunting. After locating a radio-
28 transmitters were programmed to be active marked goose in a flock, one observer ap-
during 15 days the following winter. To min- proached on foot (about 0.53 m/sec) while si-
imize potential confounding effects of captur- multaneously recording goose behavior (i.e.,
ing and surgically implanting radio transmit- on ground, flush, or flying) into a data audio
ters in Tule Geese, we collected heart rate data tape recorder (DAT) that added a digital time-
5.5 months after the radio-marking procedure stamp. At the same time, heart rate signals
when geese were on their wintering grounds were being received by a truck-mounted Yagi
in the Sacramento Valley of California. We antenna and ATS receiver (Advanced Telem-
conducted disturbance tria—ls during the last 2 etry Systems, Isanti, Minnesota), and recorded
weeks of December 1997 immediately after together with the time on a data logger (Ad-
closure of the hunting—season for dark geese vanced Telemetry Systems, Isanti, Minnesota)
(including Tule Geese) within the Sacramen- that was connected directly to the receiver. We
to and Delevan National Wildlife Refuges and synchronized the observations recorded on the
the surrounding agricultural (rice) fields. The DAT with the heart rate data by documenting
main types of human disturbance in this area precise start times on each recording device.
during our study included waterfowl hunting, In each trial, the observer continued to ap-
farming, vehicular traffic, and birdwatching. proach the radio-marked goose until it flushed,
Transmitters (model HRT-150, Telonics then immediately recorded the precise time of
Inc., Mesa, Arizona) were covered with a flushing and estimated flushing distance by
braided surgical mesh (Marlex Mesh, Davol, pacing the distance between the observer and
Charlotte, North Carolina) and implanted into the site where the bird flushed. We continued
the right abdomen using standard surgical pro- monitoring heart rates of radio-marked birds
cedures (except that the transmitters had in flight for 60 sec after they flushed from the
coiled, internal antennas rather than extended, disturbance. We then estimated the starting
percutaneous antennas; Korschgen et al. distance from the observer to the bird by re-
1996). The radio packages could transmit pacing the distance between the site of Hush-
heart rates at a maximum rate of 600 beats/ ing and the starting point of the obscr\ation.
min. Leads ofthe 44 g transmitter were placed Heart rates recorded on the tlala logger
in a base-apex configuration and anchored us- were transcribed using an A'fS DCX' 11 tlata
ing 2-vicryl (Ethicon, .Somerville, New Jer- logger (Advanced rclemetry .Systems. Isanti.
sey). The caudal lead was sutured to the dorsal Minnesota) aiul a lajitop ctnii|'uitcr with a
side of the caudal end of the sternum and the ProC’omm macro iirogram tlcvelopcd by Af.S.
anterior lead was brought through the incision which calculatetl the radio-marked bird's heart
and passed subcutaneously to the left clavicle, rate every 0.2 see. We then ealeulaled the a\-
where it was suturetl just lateral to the ster- erage heart rate during eaeh seeond. I or eaeh
num. Lkich transmitter was .secured with two trial, we used ehange point analysis (I’ROC'
148 THE WILSON BULLETIN • Vol. 116, No. 2, June 2004
NLIN, SAS Institute, Inc. 1999; also called creased from the beginning of the disturbance
segmented regression, Draper and Smith trial to the first change point as an observer
1998) to fit a three-segment regression and approached the radio-marked goose (Trial 1:
identify the two points in time that flank the m = 2.54 ± 0.56, = 4.54, P < 0.001; Trial
period when heart rate changed most rapidly. 2: m = 0.28 ± 0.0^ - 3.51, P < 0.001;
We examined whether heart rate increased as Trial 3: m = 0.10 ± 0.06, G = 1-77, P =
an observer approached radio-marked geese 0.08; Trial 4: m = -1.50 ± 1.2)44, o, = -1.21,
from the beginning of the trial until the hrst P = 0.23; Trial 5: m = 0.40 ± 0.13, =
change point (just before flushing) by testing 3.07, P = 0.003; Trial 6: m = 0.56 ± 0.12,
the slope (m) ofthis segment ofthe regression /,5o = 4.65, P < 0.001; Fig. 1). The average
for significant difference from the null hy- distance from observers to radio-marked
pothesis of no slope (a slope of zero). We cal- geese at the start ofdisturbance trials was 102
culated the weighted average ± SE heart rate m (range: 46-189 m). Heart rates increased
of radio-marked birds among trials in 5-sec from an average of 114.1 ± 6.6 beats/min
time periods using PROC MEANS (SAS In- (range means; 86.6-164.8 beats/min) during
stitute, Inc. 1999). Each measurement was in- the first time period (0-5 sec) when observers
versely weighted by the sample size for its began their approach to 154.8 ± 7.4 beats/min
corresponding time period, so that each trial (range means; 116.3—199.7 beats/min) during
would have equal total weights ofone. We did the second time period just prior to flushing
this for (1) the first 5 sec during our initial at the first change point. Average flushing dis-
approach, (2) the 5 sec immediately before the tance was 47 m (range; 25-100 m). During
first change point in heart rate (before flush- the third time period, immediately after flush-
ing), (3) the 5 sec immediately after the sec- ing at the second change point, heart rates av-
ond change point in heart rate (after flushing), eraged 456.2 ± 8.4 beats/min (range means;
and (4) during the 55- to 60-sec time period 417.1-478.0 beats/min); heart rates averaged
following flushing when the radio-marked 448.3 ± 9.5 beats/min (range means; 379.0-
goose was in flight. Statistical significance 518.8 beats/min) during the 55- to 60-sec time
was set at a = 0.10. period after flushing while in flight. In gen-
eral, heart rates increased most rapidly at 5
RESULTS
sec (range; 1.2—9.8 sec) prior to flushing and
We successfully documented the cardiac re- continued increasing rapidly for an additional
sponses of Tule Geese to an approaching ob- 4 sec (range; 0.2—8.7 sec) after flushing until
server during six disturbance trials, represent- reaching flight levels (Fig. 1).
ing four different birds. Four trials were con-
DISCUSSION
ducted at Delevan National Wildlife Refuge
and two trials were conducted nearby within The population of Tule Geese is one of the
harvested rice fields. Four additional trials on smallest goose populations in North America
different birds were excluded due to equip- (about 8,000); the winter range, centered at
ment malfunction (two trials), outside inter- Delevan National Wildlife Refuge in the Sac-
ference by approaching hunters (one trial), or ramento Valley of California (Callaghan and
logistical limitations during approach of a ra- Green 1993, Ely and Dzubin 1994, Green
dio-marked goose (one trial). The number of 1996) is also small. This concentrated winter-
radio-marked birds available for study also ing distribution potentially subjects the entire
was limited because we captured and radio- population of Tule Geese to negative effects
marked Tule Geese 5.5 months before we col- of human disturbance caused by farming,
lected heart rate data in order to minimize any birdwatching, and hunting. In five of six dis-
effects of handling and transmitter implanta- turbance trials, we found that the heart rates
tion. By the time radio-marked geese arrived of radio-marked Tule Geese increased as an
on their wintering grounds, a few radio trans- observer approached them. Heart rates in-
mitters had failed and some radio-marked creased by 36% from the time the trial began
birds could not be located following fall mi- to the abrupt increase in heart rate that oc-
gration. curred 5 sec before a bird flushed. However,
During five of six trials, heart rates in- the largest increase in heart rate occurred dur-
Ackerman et al • GOOSE HEART RATES AND DISTURBANCE 149
(beats/min)
rate
Heart
600
500
400
300
200
100
0
0 40 80 120 160 200 240 0 30 60 90 120 150 180
Time since initial approach (sec)
FIG. I. Heart rates (beats/min) of rule CJreater White-fronted Geese {A/ner alhifmns d^asi) in rela(ii>n to
time elapsed since an observer began approacliing ratlio-marketl birds. F'ach panel represents a single trial
conducted in the .Sacramento Valley of C’alifbrnia, December 1997. Trial numbers one aiul six. and three anti
four, were each conducted on the same two intlivitluals, respectively, but on tlifferent tlays. l-ach data pt)int is
the average heart rate of an individual birti tluring a I-sec periotl. The \ertical broken lines itulicate the time
when the radio-marketl bird flushed; tlata to the left of the broken line rc|Mesent birtls on the ground, whereas
data to the right ofthe broken line represent birds in (light. Approximate human-to-bird ilistances whenobser\ers
began their approaches for trials I b were 46. I2.T 1X9. 6.^. 99, and 91 m, res|iecti\ely. ,\|'>pn>ximate flushing
distances for trials I 6 were 25. 5>(). 100. .'>0. 30. and 25 m, respectively. Ihe solid iineam lines were (itteil
using change point analysis.
150 THE WILSON BULLETIN • Vol. 116, No. 2, June 2004
ing a 9-sec (range: 1.0-15.7 sec) period im- er et al. 1978, Wooley and Owen 1978, Mac-
mediately before and after a bird flushed, Arthur et al. 1982, Weisenberger et al. 1996,
when heart rates nearly tripled (Fig. 1). Using Harms et al. 1997, Perry et al. 2002). There-
our average approach speed of 0.53 m/sec, 5 fore, brief disturbances that do not elicit a
sec represents only 3 m more than the flushing flushing response by wildlife probably result
distance, when their stress response increased in minimal energetic costs. We suggest that
appreciably. Therefore, flushing behavior is a flushing distance is a reasonable basis for es-
reasonable indicator for acute changes in the tablishing buffer zones when disturbances are
physiological state of Tule Geese. likely to be short-term (e.g., automobile orair-
Heart rates, however, were slightly elevated craft noise), but that alert distances should be
for the majority of the time prior to flushing, estimated to create buffer zones when distur-
and self-maintenance activities were disrupted bances are likely to be long-term (e.g., bird-
by our disturbance. For example, Tule Geese watchers, hikers).
were alert to our presence and discontinued
ACKNOWLEDGMENTS
foraging for the entire duration of our ap-
proach. Greater White-fronted Geese (both A. We are grateful to R. Taylor for field assistance; J.
a. elgasi and A. a. frontalis) typically spend Sarvis, J. Hupp, M. Anthony, T Rothe, and D. Mather
30% of their time in alert behaviors at forag- for help capturing geese; R Tuomi and T. Bowser for
ing sites in the Sacramento Valley during win- implanting transmitters; and D. Mather and C. Mul-
ter (Ely 1992, Hobbs 1999), and heart rates cahy for assistance with the surgeries. G. Mensik and
the staffof Sacramento National Wildlife Refuge pro-
tend to be higher during alert behaviors than vided facilities and assistance. We also thank A. Fowl-
when they are walking, feeding, preening, er, J. Eadie, K. Phillips, A. Blackmer, D. Blumstein,
resting, or standing (Ely et al. 1999). If, as we and two anonymous reviewers for helpful comments
found, alert behavior typically is associated on earlier versions of the manuscript. This research
with an elevated heart rate, then alert activity was reviewed and approved under the Animal Care
could be energetically expensive, particularly and Use Committee of the Western Ecological Re-
if human disturbances are prolonged prior to tsheearUc.hS.CeGnteeorl,ogUi.cSa.lGSeuorlvoegyicaanldStuhreveLyegaancdyfAuinrdeFdorbcye
flushing. Therefore, alert distance may be a Fund.
more conservative basis for establishing wild-
life buffer zones than flushing distance, es- LITERATURE CITED
pecially when disturbances are persistent (Fer-
Andersen, R., J. D. C. Linnell, and R. Langvatn.
nandez-Juricic et al. 2001, Swarthout and 1996. Short term behavioural and physiological
Steidl 2001). response of moose Alces alces to military distur-
Other studies of disturbance effects on wa- bance in Norway. Biological Conservation 77:
terfowl have revealed acute changes in heart 169-176.
rate in response to stressful, short-term stim- Blumstein, D. T, L. L. Anthony, R. Harcourt, and
uli. For example, Ely et al. (1999) found that G. Ross. 2003. Testing a key assumption ofwild-
the heart rates of Tule Geese showed the fol- life bufferzones: is flight initiation distance aspe-
cies-specific trait? Biological Conservation 110:
lowing increases: 80-140 beats/min during 97-100.
non-strenuous activities, 180 beats/min when Callaghan, D. A. and A. J. Green. 1993. Waterfowl
the birds were alert due to natural stimuli, and at risk, 1993. Wildfowl 44:149-169.
>400 beats/min during antagonistic social in- Camp, R. J., D. T Sinton, and R. L. Knight. 1997.
teractions with conspecifics. Wooley and Viewsheds: a complementary management ap-
Owen (1978) observed brief increases in the proach to buffer zones. Wildlife Society Bulletin
25:612-615.
heart rates of Black Ducks {Anas rnbripes) in Draper, N. R. and H. Smith. 1998. Appliedregression
response to approaching herons, hawks, and analysis, 3rd ed. John Wiley and Sons, New York.
low-flying planes. Harms et al. (1997) found Ely, C. R. 1992. Time allocation by Greater White-
that heart rates ofcaptive Black Ducks briefly fronted Geese: influence ofdiet, energy reserves,
increased in response to simulated aircraft and predation. Condor 94:857-870.
Ely, C. R. and A. X. Dzubin. 1994. Greater White-
noise. However, in most studies examining the
fronted Goose {Anser alhifrons). The Birds of
cardiac response of wildlife to stressful stim- North America, no. 131.
uli, heart rates quickly returned to normal lev- Ely, C. R., D. H. Ward, and K. S. Bollinger. 1999.
els after the stimulus was removed (Kanwish- Behavioral correlates of heart rates of free-living
Ackerman et al. • GOOSE HEART RATES AND DISTURBANCE 151
Greater White-fronted Geese. Condor 101:390- Nolet, B. a., P. j. Butler, D. Masman, and A. J.
395. WoAKES. 1992. Estimation ofdaily energy expen-
Eernandez-Juricic, E., M. D. Jimenez, and E. Lucas. diture from heart rate and doubly labeled waterin
2001. Alert distance as an alternative measure of exercising geese. Physiological Zoology65:1 188-
bird tolerance to human disturbance: implications 1216.
for park design. Environmental Conservation 28: Perry, E. A., D. J. Boness, and S. J. Insley. 2002.
263-269. Effects ofsonic boomson breedinggray sealsand
Gabrielsen, G. W. and E. N. Smith. 1995. Physiolog- harbor seals on Sable Island, Canada. Journal of
ical responses of wildlife to disturbance. Pages the Acoustical Society of America 1 1 1:599-609.
95-107 in Wildlife and recreationists: coexistence Richardson, C. T. and C. K. Miller. 1997. Recom-
through management and research (R. L. Knight mendationsforprotectingraptors fromhumandis-
and K. J. Gutzwiller, Eds.). Island Press, Wash- turbance: a review. Wildlife Society Bulletin 25:
634-638.
ington, D.C.
Green, A. J. 1996. Analyses of globally threatened Rodgers,J. A.,Jr. and H. T. Smith. 1997. Bufferzone
Anatidae in relation tothreats, distribution, migra- distances to protect foraging and loahng water-
tion patterns, and habitat use. Conservation Biol- birds from human disturbance in Florida. Wildlife
ogy 10:1435-1445. Society Bulletin 25:139-145.
Harm1s9,97C..AA.,teWc.hnJi.quFelefmoirngd,orsaanldsMu.bcuKt.anSetoousskopifm.- SASSIAnSstiItnusttiet,uteI,ncI.nc.19C9a9r.y,SANoSrtOhnlCianroelDionca,. Ovenrl.in8e.
plantation of heart rate biotelemetry transmitters at <http://v8doc.sas.com/sashtml/>.
in Black Ducks: application in an aircraft noise SwARTHOUT, E. C. H. AND R. J. Steidl. 2001. Flush
response study. Condor 99:231-237. responses ofMexican Spotted Owls to recreation-
Hobbhsa,biJt.atH.use199o9f. tFhalel Taunlde wGirnetaetrerdisWthriitbeut-ifornonatnedd Wardi,sts.D.JoHu.r,naRl.oAf.WSitldelhinf,eWM.anP.agEerimceknston6,5:a3n1d2-D3.1V7..
Derksen. 1999. Response of fall-staging Brant
Goose (Anser alhifrons gambelli) in the Sacra- and Canada Geese to aircraft overflights in south-
mento Valley, California. M.Sc. thesis, California western Alaska. Journal of Wildlife Management
State University, Sacramento.
63:373-381.
Kanwisher, j. W., T. C. Williams, J. M. Teal, and K. Weisenberger, M. E., P. R. Krausman, M. C. Wal-
O. Lawson, Jr. 1978. Radiotelemetry of heart lace, D. W. De Young, and O. E. Maughan.
rates from free-ranging gulls. Auk 95:288-293. 1996. Effects of simulated jet aircraft noise on
Knight, R. L. and S. A. Temple. 1995. Wildlife and heart rate and behavior of desert ungulates. Jour-
recreationists: coexistence through management. nal of Wildlife Management 60:52-61.
Pages 327-333 in Wildlife and recreationists: co- WoAKES, A. J. AND P. J. BuTLER. 1983. Swimming and
existence through management and research (R. diving in Tufted Ducks,Aythyafnligiikn with par-
L. Knightand K. J. Gutzwiller, Eds.). IslandPress, ticular reference to heart rate and gas exchange.
Washington, D.C. Journal of Experimental Biology 107:31 1-329.
Kor.schgen, C. E., K. P. Kenow, A. Gendron-Fitz- WooLEY, J. B., Jr., and R. B. Owen, Jr. 1977. Met-
PATRiCK, W. L. Green, and F. J. Dein. 1996. Im- abolic rates and heart rate-metabolism relation-
planting intra-abdominal radio transmitters with ships in the Black Duck (Anas nihripes). Com-
external whip antennas in ducks. Journal ofWild- parative Biochemistry and Physiology: part A 57:
life Management 60:132-137. 363-367.
MacArthur, R. a., V. Gei.st, and R. H. John.ston. WooLiiY, J. B., Jr., and R. B. Owi;n, Jr. 1978. Energy
1982. Cardiac and behavioral responses ofmoun- costs of activity and daily energy expenditure in
tain sheep to human disturbance. Journal ofWild- the Black Duck. Journal ol Wildlile Management
life Management 46:351-358. 42:739-745.