J. Raptor Res.

21(1):14-20
¸ 1987 The Raptor Research Foundation, Inc.

THE EFFECT OF VEGETATIVE COVER ON FORAGING

STRATEGIES, HUNTING SUCCESS AND NESTING
DISTRIBUTION OF AMERICAN KESTRELS IN
CENTRAL MISSOURI

BRIAN R. TOLAND

ABSTRACT.--The huntingmethodsusedby the AmericanKestrel(Falcosparverius) in relationto seven

habitattypeswerestudiedin BooneCounty,Missouri,September 1981 throughAugust1984.Kestrels
spentan averageof 75% of eachday huntingincluding63% perch-hunting,7% hover-hunting,3.5%
changingperchsites,and 1.5% in horizontalpursuitflight. Of 6359 kestrelforagingsitesobserved, use
of disturbedgrasslandswas greaterthan expected(61%), useof croplandsand woodlotswas lessthan
expected (3.5 and4.0%),anduseof oldfields,undisturbed
grasslands,andplowedfieldswasin proportion
to availability.Kestrelscuedon human-related
disturbances
in managed grasslandhabitat.There was
no sexbiasin useof habitattypesby kestrelsduringany season.Kestrelswere successful
in 69.5% of
their captureattemptsand maleswere more successful than females.Invertebrateswere capturedmost
easily(82%), thenrodents(66%) and birds(33%). Hunting success declinedwith increasing vegetation
height. Hunting efficiency(estimateddaily energyexpendituresfrom time budgetsand multiplesof
standardmetabolicrate) was highestduring perch-hunting.Time spentperch-huntingby kestrelsde-
creasedwith increasingvegetationheight. In capture/costratios, kestrelsforagedmost efficientlyin
disturbedgrasslands, and leastefficientlyin old fieldsand croplands.Of 56 kestrelhomeranges,95%
were in disturbedgrasslandhabitat (which comprised18% of the studyarea). These data suggestthat
the absenceof suitableplant coverfor kestrelforagingmay effectivelylimit the distributionof American
Kestrels in central Missouri.

Sincethe American Kestrel (Falcosparverius)is Croplands(wheat, corn, soybeans,etc.) made up 49%

a relativelycommonand conspicuous raptor in Mis- of the area. Lesser amounts of milo, oats, and tobaccowere
found in the study area. Three categoriesof croplands
souri(Toland 1984), inhabitingopenareasand for- were recognized:1) plowed fields, including light wheat
aging along roadsides,farms and other habitatseas- stubbleand newly plantedwinter wheat, 0-13 cm high
ily accessibleto observers,it is a good subjectfor comprising25% of the studyarea; 2) crops,mainly wheat,
studyof raptor foraginghabitats.My objectives were corn,andsoybeans 60-183 cmhigh,with an averageground
coverof 90%, comprising14% of the studyarea;3) heavy,
to describethe huntingmethodsusedby wild Amer-
tall stubble30-60 cm high, with an averageground cover
ican Kestrelsthroughan analysisof the relationship of about 30%, comprising10% of the area. Woodlots of
betweenvegetationheight and densityto foraging 0.40-16.0 ha comprised15% of the study area and had
siteselection,huntingstrategy,huntingsuccess, and 75% ground cover. Important speciesincluded white oak
nestingdistributionof kestrelsin central Missouri. (Quercusalba), black oak (Quercusrubra), hickory (Carya
spp.),Americanelm (Ulmusamericana),Americansyca-
more(Plantanusoccidentalis), blacklocust(Robiniapseudo-
STUDY AREA AND METHODS
acacia), honey locust (Gleditsiatriacanthos),persimmon
The study area comprised175 km2 in Boone County, (Diosprosvirginiana), and Eastern red cedar (Juniperus
Missouriandwascomposed mainlyof farmland,woodlots, virginiana).Old fields comprised5% of the study area.
old fields, and meadows,and was interlacedby gravel Vegetationranged 90-254 cm in height with 95% ground
roads.The areawasdividedinto sevenhabitattypesbased cover.
on vegetation height, percent ground cover, and compo- Grasslandscomprised31% of the area and were sub-
sition. Aerial photographsby the U.S. GeologicalSurvey divided into two categories:1) idle, undisturbedpastures
(1981), and ground reconnaissance were used in measur- and meadows60-91 cm high, with an averageground
ing percentagesof different habitats. Vegetation height coverof 90%, comprising11% of the study area; and 2)
wasmeasuredto the nearest2.5 cm and estimatedvisually disturbedgrasslands(mowed, and grazed pasturesand
for comparisonuntil no differencebetweenthe two meth- harvestedhay fields) 5-25 cm high where ground cover
ods resulted.This was done every two wk and habitat was 90%, comprising20% of the studyarea.
availability was reassessedto providemonthly as well as Actual habitat use by kestrelswas comparedwith rel-
annual means.I calculatedmean habitat availability for ative habitat occurrence. American Kestrels were observed
the kestrelnestingperiod(March-August). Percentground in the field for three yearsfrom September1981 through
coverwas estimatedby sightingthrough an occulartube August 1984. During the first six mo, 36 kestrelswere in
(four cm diameter/10 cm length;Weller 1956) pointedat the studyarea. During the next six mo (March-August
the ground at arm's length. In each habitat type means 1982) I observed13 territorial pairs. BetweenSeptember
from 20 readings were taken. 1982 and February 1983 an averageof 48 birds was in

14
 SPRING 1987 AMERICAN KESTRELSIN MISSOURI 15

Table 1. Relative habitat usea by American Kestrelsin central Missouri.

%
% HABITAT
HABITAT USE
GROUND AVAIL-
HABITAT TYPE HEIGHT (CM) COVER ABLE No. %

Plowed fields,light stubble,newly planted crops 0-13 10 25 681 11.0

(11) (295) (9.0)
D•sturbed grasslandand fields (mowed,hayed,grazed) 5-25 90 20 3892 61.0
(18) (2230) (68.0)
Heavy, tall stubblefields 30-60 30 10 340 5.5
(7) (131) (4.0)
Idle, undisturbedpasturesand meadows 60-91 95 11 660 10.5
(13) (393) (12.0)
Crops (corn,wheat, soybeans,
milo, oats) 60-183 90 14 228 3.5
(31) (98) (3.0)
Old fields,overgrownpasturesand meadows 90-254 95 5 291 4.5
(5) (66) (2.0)
Woodlots > 300 75 15 267 4.0
(15) (66) (2.0)
6359 100.0
Total (3280) (100.0)
During nestingseasonin parentheses.

the 175 km2 studyarea. During the periodMarch-August alert posture.Observationsof kestrelswere madeduring
1984, 50 territorial birds (25 pairs) were present.Kestrels all daylight hr includingas many full days as possible.
were capturedwith bal-chatri traps (Berger and Mueller Otherwise,half-day observationswere alternatedin morn-
1959) or noose-harnessed House Sparrows(Passerdomes- ings and afternoons.
twus;Toland 19854), and then marked with paintedU.S. Birdswerelocatedby drivingsecondary roads.A census
Fish and Wildlife Servicebandsand coloredplasticleg routeof about120 km was drivenat an averagespeedof
bandsto facilitate individual recognition. 40 km/hr. An averagelateral distanceof 800 m eachside
I used a 30x spotting scopeand 9x binoculars to ob- of the routewas effectivelycoveredby two observers. The
serve each kestrel. Each bird was observed a minimum of areawas intersected by a networkof roadsevery1.6 km,
20 min for a combinedtotal of 1810 hr. I recordedtype thusthe routeallowedcompletesurveillance of the study
of activity (perch-hunting,hovering,soaring,preening, area.At leastonecensus/wkwas completedduringthree
etc.), changesof perch, and duration (sec)of all flights. yr. Censuseswere conductedbetween 1000 and 1400 H
Type or types of vegetationwithin 50 m of a hunting on dayswith conditionsof goodvisibility and low wind
kestrel,huntingmethod,capturesuccess, typeof prey (in- velocity.All kestrelsiteswereplottedoncovermapsduring
vertebrate, small mammal, or bird), and the sex of each eachof the three nestingseasons.
foragingkestrelwere recorded.When unableto seeprey I estimateddaily energy expendituresusing observed
that kestrelsunsuccessfully attemptedto captureon the time budgetsand multiples of standard(basal) metabolic
ground, I distinguishedbetweeninvertebratesand small rate (SMR) (Koplin et al. 1980; King 1984) to determine
mammals by differencesin strike characteristics. Wild differencesin daily energybudgets(DEB) dueto different
American Kestrelsattackground-basedinvertebratesfrom huntingmethodsin vegetationof variousheights.A mul-
a buoyant,parachutingstrikeand oftenhop or run after tiple of 1.7 x SMR has been usedfor restingmetabolic
an insect if the initial pounceis unsuccessful.When at- rate (RMR) (Wolf and Hainsworth 1971). This is the
tackinga smallmammal,however,kestrelsemploya dive rate of diurnal inactive metabolism and includes SMR as
or stoopwithout breaking their momentum until the last well as heat liberatedin thermoregulationand digestion
moment(pers.observ.).Thesedifferences in huntingtech- (Gessaman 1973). Therefore, I used 1.7 as the value of
niqueswerereinforcedby the performances of 12 falconry the energeticcostof inactiveperchingand loafing. I used
kestrelswhich I trained to hunt known prey types or 1.0asan indexto theenergetic costof nocturnalinactivity
"bagged"quarry during four yr (Toland, unpubl. data). (Gessaman 1973).I estimated costof preening,stretching,
All hunting attemptswith undeterminedoutcomeswere eating,caching,and other maintenanceactivityas 2.0 x
excludedfromanalysis.I distinguished betweenstill-hunt- SMR. I useda multiple of 3.5 x SMR for still-hunting
ing froma perchandotherperchingactivityby observation from a perch(Wakeley 1978b). I used8.0 as an indexto
of associated behaviorssuchas head-bobbing, sleekplum- thecostofchanging perching spots(Tucker1971;Wakeley
age with frequent plumage rousals(shaking), and erect, 1978b). Energy consumption during fast forward flight
16 BgI^N R. TOL^ND VOL. 21, No. 1

usedin pursuithasbeenestimatedat 12.5 x SMR (Wake- turbancesby farm workers,machineryand livestock

ley 1978b; Rudolph 1982). Therefore, I used 12.5 as an would increasemovementand thus vulnerability of
index to the cost of swift horizontal chases as well as
smallmammals.Shrubb(1980) reportedsimilar be-
hovering--kestrelscompensate for the lift lost by lack of
forward velocityby fanning the tail and utilizing wind haviorby the EurasianKestrel(Falcotinnunculus),
and surfaceupdrafts (Tucker 1968; Rudolph 1982). I which made62% of their kills in uncultivatedgrass-
convertedprey capturerates to estimatesof the number lands,roadsides,and field edgescomprising24% of
of captures/unitcostby dividingeachrateby itsrespective a studyarea in England. Shrubbalsoreportedthat
energy:costindex (Wakeley 1978b). A comparisonwas
made betweenthesecapture:costratios and amount of Eurasian Kestrelsavoidedcereal cropsduring the
ume kestrelsusedeach hunting methodand each type of nestingseason. His opinionwasthatthecombination
habitat. of height,density,and evenness of cerealcropsin-
hibited successfulsearching,and the stiff, dense,
RESULTS AND DISCUSSION
spikeynature of the plantsmadeprey capturedif-
ficult. A preferencefor haylandsand pastureswith
During the average10 hr period of daylight, goodinterspersion and avoidanceof large tractsof
American Kestrels spent an average75% of their croplandhasalsobeenreportedfor the Ferruginous
time hunting and 25% loafing,eating and caching, Hawk (Buteoregalis)(Wakeley 1978a, 1978b; Gil-
and preeningand stretching.Averagedaily activity mer and Stewart 1983), the Swainson'sHawk (B.
of kestrelsincluded63% still-hunting, 3.5% direc- swainsoni)(Bechard 1982), the Red-tailed Hawk (B.
tionalchangeof perch,7.0%hoveringand 1.5%hor- jamaicensis),and the Rough-leggedHawk (B. la-
izontal pursuit or tail-chasing. gopus)(Baker and Brooks 1981). Craigheadand
Foraging Site Selection.I observed foragingkes- Craighead(1956) reportedhigherbuteodensitiesin
trels 2131 timesduringthe first yr, 2363 the second habitatswith shortervegetationand sparserground
yr, and 1865 the third yr. The distributionof sites covereventhoughvole populationswere lower.
selectedwere not significantlydifferent amongthe When choosing huntingsites,kestrelsin my study
threeyrs, sodata were combined(Table 1). area were quick to cue on recentlyharvestedcrop
Kestrelshuntedoverdisturbedgrasslands61% of and hay fieldsas well as other human-relateddis-
the time or three times the frequencywith which turbancessuchas plowing and mowing.These dis-
this habitat occurredin the studyarea (x2 = 84.05, turbances result in sudden decreases of cover and
P < 0.01, df -- 6; Table 1). By the samemeasure, increasesin rodentvulnerability.Kestrelsresponded
kestrelssignificantlyunder-utilizedcropsandwood- soconsistently to thesedisturbances that I was able
lots (x2 = 61.6, P < 0.01, df = 5; Table 1). Undis- to predicttheir foragingsiteson the basisof farming
turbedgrasslands, old fieldsand plowedfieldswere activities. Kestrels also cue on other human-caused
usedin proportionto availability (x2 = 1.67, P > disturbances, suchas irrigation in California (Ru-
0.05, df = 2). dolph 1982) and controlledfiresin Florida (Small-
During the nestingseason(March-August) hab- wood et al. 1982). Kestrelsin central Missouri com-
itat availabilitychangedsubstantially,croplandsin- monly hunted in and around herds of livestock,
creasingfrom 14% to 31%. Kestrelsexhibitedeven apparentlyfindingvoles(Microtusspp.)highly con-
strongerpreferencefor disturbedgrasslandduring spicuouswhen flushedby foragingsheep,cattleor
this season,conducting 68% of their foragingin this horses.Usually kestrelshovered4-10 m abovesheep
habitat (Table 1). During the nestingseason,kes- but sometimes flew quicklyover,around,andunder
trels hunted in disturbedgrasslandmore than ex- grazingcattle.Shrubb(1980, 1982)foundthat Eur-
pectedandin cropsandwoodssignificantlylessthan asianKestrelsin Englandalsorespondedto habitat
expected(X2 = 177.8, P < 0.01, df -- 6). Kestreluse disturbances. Bechard(1982) foundnestingSwain-
of available old fields, undisturbedgrasslands,and son'sHawks avoidedcropland before harvest and
plowed fieldsdid not deviatesignificantlyfrom ex- concentrated on pasturesand edgewith lesscover,
pectedvalues(x2 = 3.52, P > 0.05, df = 2). althoughgoodconcentrations of rodentswerepresent
Kestrels were probably attracted to disturbed in wheat fields. However, when harvestreducedcov-
grassland,since 1) low vegetationin pasturesand er, fieldswere extensivelyhuntedby the hawks.
fieldsaffordedgoodvisibilityof small mammals,2) Differential use of winter habitatsby both sexes
shorter,flexible grasseswould give little resistance has beenreportedfor kestrelsin Texas, California,
to the strikeof the light-weightkestrel,and 3) dis- Arizona, Mexico (Koplin 1973; Mills 1976), and
SPRING 1987 AMERICAN KESTRELS IN MIssoum 17

Georgia (Stinsonet al. 1981), but was not found in founda similarproportionof vertebrateprey (70%)
Kentucky(Sferra 1984) nor in my area (X2 = 10.77, in kestrel diets in California. However, most studies
P > 0.05, df = 6). reporthigherpercentages of invertebratesthan ver-
Hunting Success.Kestrelswere successful in 988 tebrates,includingJenkins (1970; 39% success, 33%
of 1414 (69.5%) captureattemptsduring the three vertebrates)in Costa Rica, Sparrowe (1972; 33%
yr. Hunting success was higher during the nesting success, 21% vertebrates)in Michigan, Cruz (1976;
period (March-June) than in winter (November- 42% success,39% vertebrates) in Puerto Rico, Col-
February)thoughthe differencewasnot statistically lopy (1979; 55% success, 6% vertebrates)in Cali-
significant(X2 = 3.72, P > 0.05, df = 1). Higher fornia and Rudolph (1982; 57% success, 5% verte-
hunting successduring nesting probably reflects brates) in California.
greaterpreyabundance and/or vulnerabilityduring Height and densityof vegetationin kestrelranges
the spring.Males were moresuccessful hunters(72%) had a considerableeffectupon their hunting success
than females(67%) year-round(X2= 3.98, P < 0.05, (Table 2). With the exceptionof plowedfieldswhere
df = 1), but there was no differencein hunting kestrelshunted mostly for insectsand earthworms,
successof males and femalesduring winter (X2 = hunting success declinedsignificantlywith increas-
0.014, P > 0.05, df = 1). Betterhuntingsuccess by ing vegetationheight (X2 = 182.14, P < 0.01, df =
malesmay be an adaptationby which malesprovide 6). The greatestnumber of hunting attempts(705)
food for both femalesand nestlingsduring much of and captures(83%) were made in managedor dis-
the nestingperiod (Cade 1982; Toland 1986), or turbed grassland(5-25 cm high), while only 41%
greaterprey abundanceand vulnerabilityduring the of 79 attemptswere successful in cropsand wood-
nestingcyclewhen malesare primary foragersmay land.
affect their higher successrates. Hunting Efficiency. Kestrelsuse three distinct
Kestrelsin my studyarea had an 82% success rate hunting methodswhich vary in efficiencyand en-
in capturinginvertebrates,66% success in capturing ergeticcost(Rudolph1982). Kestrelsstill-huntfrom
rodentsand 33% success in capturingbirds (X2 --- an elevatedperch 70-97% of the time (Cruz 1976;
127.08, P < 0.01, df = 2) (Toland 1983, 1986). Balgooyen1976; Cade 1982; Rudolph 1982). Kes-
During the nestingseason,the overall capture rate trels hunt from a hover 2-20% of the time (Bal-
increasedto 74% of 580 attemptseven though the gooyen 1976; Rudolph 1982), and in swift, hori-
percentageof vertebrate prey increasedto 81.5% zontal flight <5% of the time. Kestrelsin my study
(Toland 1983). Kestrelsbecamemore aggressiveand still-huntedfrom a perch88% of the time and from
rapaciousduring the nesting seasonwhen it was a hover 10% of the time, while swift, horizontal
probablymore energyefficientto capturelarger ver- flights to includetail-chasingand contour-hugging
tebrate prey than invertebrate prey when raising were used 2% of the time (Table 2). Vegetation
broods(Cade1982). To investigate thisphenomenon height apparently influencedthe hunting strategy
I offeredhandicappedEuropeanStarlings(Sturnus usedby kestrels.Time spentstill-hunting declined
vulgaris)to kestrelsduring both nestingand non- with increasing height of vegetation while time
nestingseasons. Although attractedto within a few spenthoveringsignificantlyincreased(X2 = 50.74,
m of 16 starlingsoffered,kestrelsdid not attempt to P < 0.05, df = 6; Table 2). Sincehunting methods
kill them during the non-nestingperiod. However, differ in energeticcostsand kestrels use them in
during the nesting season,kestrelskilled 12 of 16 proportionsvarying with the vegetationat foraging
starlings(X2= 21.87, P < 0.01, df = 1). sites,vegetativestructure probably influencedthe
The overallhuntingsuccess of kestrelsin my study ability of kestrelsto maximize energygain.
washigherthan previouslyreportedelsewhere,per- Kestrels were successfulin 76% of hunting at-
hapsdueto a high densityof Prairie Voles(Microtus temptsfrom perches(Table 3), which was signifi-
ochrogaster). Volesweresoabundantthat theycould cantlyhigherthan success from hover-hunting(52%)
be seenfrequently in all habitats. Interviews with (X2 = 55.15, P < 0.001, df = 1) or horizontalflights
farmersin my studyarea supportedthis qualitative (45%) (X2 = 48.2, P < 0.001, df = 1). I calculated
assessment. The high hunting success rate of kestrels capture/costvaluesof 22.0, 4.7, and 3.7 for still-
in central Missouri is even more significantwhen hunting, hovering,and horizontalpursuit, respec-
consideringthat vertebrates (mainly voles) com- tively. The averageuseof thesehunting methodsby
prised67% of the prey captured.Balgooyen(1976) kestrelswas roughly proportionalto their respective
18 BRIAN R. TOLAND VOL. 21, NO. I

Table 2. Hunting strategiesand success

of AmericanKestrelsforagingin sevenhabitat typesin centralMissouri.

% HUNTING STRATEGIES USED HUNTING SUCCESS

% PERCH-
GROUND HUNT- HOVER- FLAPPING CAPTURES/
HABITAT TYPE HEIGHT (CM) COVER ING ING FLIGHT ATTEMPTS % SUCCESS

Plowed fields, light stubble,

newly plantedcrops 0-13 10 92 7 1 67/91 74
Disturbed grasslandand
fields (mowed, hayed,
grazed) 5-25 90 91 8 1 583/705 83
Heavy, tall stubblefields 30-60 30 81 18 1 79/122 65
Idle, undisturbed pastures
and meadows 60-91 95 72 27 1 92/175 53
Crops (corn, wheat, beans,
milo, oats) 60-183 90 68 30 2 28/67 42
Old fields, overgrown
pasturesand meadows 90-254 95 68 30 2 35/106 33
Woodlots > 300 75 92 6 2 4/12 33
Means -- -- 88 10 2 -- --
Totals ..... 888/1278 69.5

capture/costratios. Thus, kestrelsused the most theycanforagemostefficiently.Of 56 nestingseason

efficientmethod(perch-hunting)mostoften, as pre-
viouslyreportedby Sparrowe(1972), Collopy(1979)
and Shrubb (1982). Perch-huntingwas used91% of
the time in disturbedgrasslandswhere kestrelsfor-
aged 61% of the time with a successrate of 83%.
Kestrelsperch-huntedonly 68% of the time in crop-
land and old fieldsand huntedonly four and 6% of
the time, respectively,in thesetwo habitat types;
capturesuccess rateswere 42 and 33%, respectively.
I estimated an average daily energy budget of
about 60 kcal, althoughthis value couldvary with
daily temp or season,and bodyweightof the birds
(Koplin et al. 1980). During the nesting season,
when adultsin centralMissouri are usually feeding
five nestlings(Toland 1985b), it becomesincreas-
ingly obviouswhy kestrelshunt in habitat where
home ranges,95% (all but three) were in disturbed
grasslandhabitat. Thus, mostof the nestingkestrels
in my study area were concentratedin 18% of the
availablehabitat. As vegetationincreasesin height,
detectionand captureof prey becomemore difficult.
Becausethe prey animal is vulnerableto predation
only for brief moments,kestrelsforaging in these
habitats must depend on hunting methodswhich
afford closeproximity to prey. These methodsin-
cludehoveringand horizontalflight during contour-
huggingor tail-chasing,andare energeticallyat least
four times more costlythan perch-hunting,which
may explain why kestrelsspendso much time for-
aging in habitat where they mostlystill-hunt from
perches.
The importanceof habitat physiographyis com-

Table 3. Successof American Kestrel hunting strategies1981-83.

STILL-HUNT HOVER TAIL-CHASE

CAPTURES/ CAPTURES/ CAPTURES/

SEX ATTEMPTS % ATTEMPTS ATTEMPTS

Males 478/611 78 8o/141 57 34/69 49

Females 333/451 74 48/103 15/39

Total 811/1062 76 128/244 49/108
SPRING 1987 AMERICAN KESTRELS IN MISSOURI 19

pounded during the nesting season when adult ED. Ecologicalenergeticsof homeotherms.Monogr.
American Kestrelsmust provisionfive or six nest- Ser. 20, Utah State Univ. Press, Logan.
lingswhosedaily energyrequirementsexceedadults GILMER, D. S. AND R. E. STEWART. 1983. Ferruginous
(Cade 1982). This critical time period demandsthat Hawk populationsand habitat use in North Dakota.
J. Wildl. Manage. 47:146-157.
kestrelsforageas efficientlyas possible,and could
JENKINS,R.E. 1970. Food habitsof wintering Sparrow
explain why 95% of the nesting pairs had home Hawks in Costa Rica. Wilson Bull. 82:97-98.
rangescomposed of disturbedgrasslands.The scar- KING,J.R. 1974. Seasonalallocationof time and energy
city of suitableplant covereffectivelylimits the dis- resourcesin birds. Pages4-70. In R. A. Paynter, Jr.,
tribution of American Kestrels in central Missouri ED. Avian energetics.Nuttall Ornithol. Club, Cam-
and may explain declinesof severalspeciesof hawks bridge, MA.
in areas where expansivemonoculturefarms pre- KOPLIN,J. R. 1973. Differential habitat use by sexes
dominate. of AmericanKestrelswintering in northernCalifornia.
Raptor Res. 7:39-42.
ACKNOWLEDGMENTS • M. W. COLLOPY,A. R. BAMANN AND H. LEV-
ENSON. 1980. Energetics of two wintering raptors
William H. Elder, School of Forestry, Fisheries and Auk 97:795-806.
Wildlife, University of Missouri, Columbia, provided
MILLS, E.G. 1976. American Kestrel sex ratios and
guidancethroughoutthisstudy.NancyThompson-Toland
contributedfinancial, moral and field support. I appre- habitat separation.Auk 93:740-748.
ciated field assistancefrom David Scarbrough,Toney RUDOLPH,S.G. 1982. Foraging strategiesof American
Chiles, and Tim Haithcoat. Thomas S. Baskett, Curtice Kestrelsduring breeding.Ecology63:1268-1276.
Griffin and James R. Koplin providedcriticism of an SFERRA,N.J. 1984. Habitat selectionby the American
earlier draft of the manuscript.Helmut C. Mueller, Clay- Kestrel(Falcosparverius) and Red-tailedHawk (Buteo
ton m. White, Peter Bloom, and an anonymousreferee jamaicensis)wintering in Madison County, Kentucky.
made editorial commentsthat improvedthe paper. This Raptor Res. 18:148-150.
studywas fundedby the Natural History Sectionof the SHRUBB,M. 1980. Farming influenceson the food and
Missouri Departmentof Conservation.
hunting of kestrels.Bird Study29:109-115.
1982. The hunting behaviorof somefarmland
LITERATURE CITED
kestrels.Bird Study 29:121-128.
BAKER,J. A. AND R. J. BROOKS. 1981. Distribution SMALLWOOD,J. A., M. WOODREY,M. J. SMALLWOOD
patternsof raptors in relation to densityof meadow ANDM. A. KETTLER. 1982. Foragingby Cattle Egrets
voles. Condor 83:42-47. and American Kestrelsat a fire's edge.J. Field Ornithol.
B^LGOOYEN, R. G. 1976. Behavior and ecologyof the 53:171-172.
American Kestrel (Falcosparverius)in the Sierra Ne- SPARROWE, R. D. 1972. Prey-catchingbehavior in the
vada of California. Univ. Calif. Publ. Zool. 103:1-87. Sparrow-hawk.J. Wildl. Manage. 36:297-308.
BECHARD,M. J. 1982. Effect of vegetative cover on STINSON,C. H., D. L. CRAWFORDAND J. LAUTHNER
foraging site selectionby Swainson'sHawk. Condor 1981. Sex differences in winter habitat of American
84:153-159. Kestrelsin Georgia.J. Field Ornithol.53:29-35.
BERGER, D. D. AND H. C. MUELLER. 1959. The bal- TOLAND,B. R. 1983. The ecologyand biology of the
chatri: a trap for the birdsof prey. Bird-Band.30:19- American Kestrel in central Missouri. Unpubl. M.S.
27. Thesis, University of Missouri, Columbia, MO.
CADE, T.J. 1982. The falconsof the world. Cornell 1984. Missouri's eagles,hawks, falcons, and
Univ. Press, Ithaca, NY. vultures. Missouri Cons. 45:16-31.
COLLOPY,M. W. 1973. Predatory efficiency of the 1985a. A trapping technique for trap-wary
American Kestrel wintering in northwesternCalifor- American Kestrels. North Am. Bird Band. 10:11.
nia. RaptorRes. 7:25-31. 1985b. Double brooding by American Kestrels
1979. Behavioral and predatory dynamicsof in central Missouri. Condor 87:434-436.
American Kestrels wintering in the Arcata Bottoms. 1986. Hunting success
of someMissouri rap-
Abst. Thesis, Raptor Res. 13:32. tors. Wilson Bull. 98:116-125.
CRAIGHEAD,J. J. AND F. C. CRAIGHEAD,JR. 1956. TUCKER,V.A. 1968. Respiratory exchangeand evap-
Hawks, owls and wildlife. StackpoleCo., Harrisburg, orativewater lossin the flying Budgerigar.J. Exp. Biol
PA. 48:67-87.
CRUZ,A. 1976. Food and foragingecologyof the Amer- 1971. Flight energeticsin birds. Am. Zool. 11'
ican Kestrel in Jamaica. Condor78:409-423. 115-124.

GESSAMAN, J.A. 1973. Methodsof estimatingthe energy WAKELEY, J.S. 1978a. Factorsaffectingthe useof hunt-
costof free existence.Pages3-31. In J. A. Gessaman, ing sitesby FerruginousHawks. Condor80:316-326.
20 BRIAN R. TOLAND VOL. 21, NO. 1

1978b. Hunting methodsand factorsaffecting Department of Forestry, Fisheries and Wildlife, 112
their use by FerruginousHawks. Condor80:327-333. StephensHall, University of Missouri, Columbia,
WELTER, M. W. 1956. A simple field candler for wa- MO 65211. Present address: Florida Game and Fresh
terfowl eggs.J. Wildl. Manage. 20:111-113. Water Fish Commission,P.O. Box 1840, Vero Beach,
WOLF, L. L. AND F. R. HAINSWORTH. 1971. Time and FL 32961.
energy budgetsof territorial hummingbirds.Ecology
52:980-988. Received10 July 1986; Accepted12 January 1987

Research/TeachingAssistantship Wanted. SeriousraptorstudentseekingM.Sc. research/teaching assistantship

to
beginFall 1987. Interestedin virtually any aspectof raptor ecologyor behavior,especiallyraptor-preyecology,habitat
requirements,and populationmodeling.Willing to consideralmostany locality,but prefer westernU.S. or Mexico.
Have B.S. in Wildlife Science,publishedresearch,and a variety of experience.GRE scores,transcripts,recommen-
dations,etc., available.Please contact: Bryan Kimsey, P.O. Box 278, Anahuac, TX 77514, (409) 267-6527.

Newly-AppointedVice Presidentof the Societyfor the Preservationof Birds of Prey. The ReverendEdward
D. McGinnis of Elizabeth City, North Carolina, was named Vice Presidentof the Societyfor the Preservationof
Birds of Prey on 29 December 1986. Reverend McGinnis was born in Durham, North Carolina, and receivedhis
B.A. degreein religionand philosophyfrom Elon Collegein 1969.He can be contactedat P.O. Box 2448, Elizabeth
City, NC 27909.