J. Field Ornithol., 58(3):372-382

AN INEXPENSIVE CAMERA SETUP FOR THE STUDY OF AT ARTIFICIAL NESTS

JAROSLAVPICMAN Departmentof Biology Universityof Ottawa 30 Somerset E. Ottawa, Ontario KIN 6N5, Canada

Abstract.--Nest predationis an important factor reducingreproductive success of many .However, few data are availableon the relativerole of differentpredators. This paper describesan automaticcamera setup that can be usedto studypredation on artificial nests. Becausethe setup is relatively inexpensive,it shouldallow fieldworkersto conductlarge- scaleinvestigations of nest predation and determineits importanceas a selectiveforce in the evolutionof avian reproductivestrategies. MONTAJEDE CAMARASFOTOGRJ, FICAS EN NIDOS ARTIFICIALES, PARA ESTUDIOS DE DEPREDACI•)N DE HUEVOS Sinopsis.--La depredaci6nen nidoses un factor de importanciaen la reducci6ndel •xito reproductivode muchasaves. Sin embargo,hay pocostrabajos que documentenel rol de diferentesdepredadores. Este trabajo describeel montajede una cfimaraautomfitica que puedeser utilizada para estudiarla depredaci6nen nidos artificiales.Debido a que el montaje cs de bajo costo,esto debe permitirle a los investigadores,11evar a cabo amplios estudiossobre el efectode depredadoresy susimplicaciones en la evoluci6nde estrategias reproductivasde aves.

Predation is the most important causeof nestingmortality in many birds (e.g., Murphy 1983, Nilsson 1984, Pageet al. 1985, Ricklefs1969) and has played an important role in the evolutionof avian reproductive strategies(e.g., Crook 1964, Lack 1968). Very little work has beendone, however,to identifyindividual nest predators and determinetheir impact on prey.This is presumablybecause predation events are rarely witnessed and there has not been any inexpensivetechnique available that would allow fieldworkersto effectivelystudy this phenomenonon a larger scale. In 1985 I began a large scalestudy of intensity,type, and temporal and spatial patterns of predation on artificial nestsin different habitats. To record the predation at artificial nests, I designedan inexpensive automatic camera setup, describedhere.

DESCRIPTION OF THE CAMERA SETUP To avoiddifficulties associated with the useof battery-operatedflashes and to keep the price of the setup minimal, I selectedan inexpensive camera,the KodakInstamatic X-15F (approximately$17 Cdn). Although the camera'soptics and smallerfilm format(126) resultin onlya moderate quality picture, the advantagesof this camera(flip flashesand low cost) justify its use. The setupconsists of four main components:(1) a nest supportwith trigger mechanism,(2) the Kodak Instamatic cameraX-15F attachedto the support, (3) a clock placed on the support near the nest, and (4)

372 Vol.5a, •o. • CameraSetup for Predation Studies [37 3

FIGURE1. The main componentsof the camerasetup; nest support with triggermechanism (a), camerawith support(b), clock(c), and coversprotecting camera, trigger and clock. The brake cable(d) is connectingthe triggermechanism on nestsupport and the camera shutterlever. Details of individual componentsare shownin Figures 2-4. protectivecovers (Fig. 1). In the following descriptionthe capital letters refer to parts shown on Figures 2-4. The nestsupport and the triggermechanism (Fig. 2).--To the top of a 120 cm long stake(40 x 16 ram; A) I attacheda basket(support for the artificial nest),which I madefrom vinyl-coatedfencing wire (G, H). The attachmentwas achievedby pushingthe two endsof the basketthrough the pre-drilledholes (the samediameter as the wires) in the stakeand bendingthe endsof the wires toward the stake(H). The bottomend of the stakeA is pointedso that it couldbe pushedinto the soil.The trigger mechanismconsists of three parts. An L-shapedwire (B) made of piano wire (diameter1.75 mm) with a large loop (diameter20 ram) at oneend for eggsupport, a small hook at the other end that holdspart C in place when the camerais set up, and a small loop (diameter2 mm) approxi- mately 30 mm from the hookend. The part B is attachedto the support stake A through the small loop with a screw (length 10 mm), which functionsas a hinge.Part C is a short(about 35 mm) straightwire (piano 374] J. Picman J.Field Ornithol. Summer 1987

FIGURE2. Nestsupport, parts of thetrigger mechanism and movement of triggerparts followingdepression of the trigger loop which supports an egg. wire, diameter0.75 mm) with a smallloop (diameter2 mm) at oneend. This part is attachedthrough its loop to the stakeA with a smallscrew (length10 mm), whichalso functions as a hinge,allowing it to rotate freelyaround the screw. The thirdpart is an approximately120 cm long bicyclebrake cable (D). The nestend of the cableis placedin a notch (110 mm long,6 mm wide,4 mm deep;E) madein the stakeA. To hold the cable,a 50 mm longmending plate (F) is attachedto the stake.The vol.58, No. • CameraSetup for Predation Studies [375

116

J • •Jl M

FIGURE5. Details of the camerasupport and the attachmentof the brake cable to the camera shutter lever, the top view of the camera base (J) indicatingdimensions and positionof the notchfor the flip flash,and the detail of the cameraend of the brake cable(O) with the compressionspring (M) installed. 376] J. Picman J.Field Ornithol. Summer 1987 brake line at the nest end of the cableis cut off and a small loop is made by twisting the steel wires together.The wire C is then placedthrough this loop.A screwW (length 15 mm) is screwedabout two thirds of the way into the stakeA abovethe top of the mendingplate F (Fig. 2). The function of this screw is to restrict the movement of wire C, therefore preventingit from escapingfrom the brake line loop. The movementof the trigger parts B, C, and D, following the depressionof the loop end of the part B in which the egg is placed,is shownin Figure 2. The camerasupport (Fig. 3) .--The camerasupport consists of two main components,a brake cableattachment and a camerasupport. The camera end of the brake cable is not altered with the exceptionof the addition of the compressionspring (M). The springis constantlyexerting pressure on the cylindricalend of the brake line, pulling it out of the cable(this end of the brake line setsoff the camera).This spring createsconstant downwardpressure on the trigger bar C (Fig. 2) at the nest end of the brake cable. I found it necessaryto glue a small (5 mm diameter and length) plasticcylinder to the shutterlever of the camera.This cylinder housesthe end of the brake line, therebypreventing it from slippingaway from the shutter lever when the trigger is depressed. The woodencamera base (J) is attachedwith two screws(length 40 mm) to the stake P (40 x 16 mm, but 20 cm longer than the supportA in Fig. 2, so that the picture is taken from slightly above). The joint betweenthe two parts (J and P) is further strengthenedby gluing these partstogether with Epoxyglue. The baseJ hasa notchfor the flip flash (seeFig. 5 for dimensionsand positionof this notch on the base). The camera, in an upside down position, is placed on a camera base and attached with several elastic bands. The camera end of the brake cable (O) is then placedin the notch 4 mm deep in the camerabase (Fig. 3) and positionedby tighteningthe wing nut (S) on bolt R (length 60 mm, diameter4 mm), which pressesthe washer (Q) againstthe brake cable. When the brake cable is in the right position,the metal cylinder at the end of the brake line must be pressingagainst the shutter lever of the camera. The clockattachment ('Fig. 4).--To recordthe time of the predation event,I useda battery-operatedclock (Brown AG; Type 4750/AB 3) to which I added an AM/PM indicator (X). This indicator is a thin (1 mm), light disk (Y) cut out of plasticgardening labels, with a small hole (diameterI mm) drilled in its centerand is turned by the passageof the hour hand. The length of the hour hand determinesthe positionand the size of the disk (i.e., the centerof the AM/PM indicator must be located abovethe end of the hour hand). If the hour hand is too long and does not leave enoughspace for the disk, it has to be shortenedslightly. To oneside of the disk I gluedfour dividers(cut out of the gardeninglabels) subdividingthe disk area into four equal fields(Z in Fig. 4). The height of these plastic dividers is determinedby the distanceof the hour and minute handsfrom the faceof the clock(i.e., the dividersshould be slightly higher than the distanceof the hour hand from the face of the clock,but Vol. 58, No. 3 CameraSetup for PredationStudies [377

--y

FIGU•.E 4. The clock attachment and the detail of the AM/PM indicator. lower than the distanceof the minute hand from the clock).Two opposite 90ø fields are painted black and the other two white. The disk is then hinged through the hole in the center of the disk to the top part of the clockwith a small nail (diameter 1 mm, length 10 mm, but any excess length that would interfere with the mechanisminside the clockmust be cut off). The AM/PM indicatorhas to be positionedas closeas possible 378] J. Picman J.Field Ornithol. Summer 1987 to the face of the clockbut must be allowed to rotate freely around the nail. The nail is then glued in positionfrom inside the clockwith Instant Krazy Glue ©. Becauseonly the hour hand canrotate the disk,the indicator is rotated every 12 hours by 90ø and the position of the white and black fieldscan be usedto determineAM/PM from the picturesof predation events.The clockis positionedabout 30-40 cm asidefrom the nest. This placementis less likely to result in predatorsstanding in front of the clock. The clock T is attachedwith two elastic bands to the clock support baseU which is hingedthrough the stakeV to the nestsupport A. Hinging the clockattachment to the nestsupport allows collapsingthe clockto the nestsupport stake, making transportationof the setupless awkward. The screwW (length 20 mm), which is screwedabout two thirds of the way into the stakeA, fixes the supportingstake V in a selectedposition. I havetested different types of clocksand founddigital clocksunsuitable. In addition, I recommend the use of a clock with a black face and white handsand numbersbecause these are more legibleon picturesthan black numbersand hands on a white background. Protectivecovers.--The trigger mechanism,the camera and the clock must be protectedfrom moisture.I used a vinyl ground sheet (sold for camping)from which I made small protectivecovers for the camera,clock and the trigger parts on the nest support stake (see Fig. 1). I attached these coversto the equipment with elastic bands. The vinyl coversare durableand providedadequate protection for 50 setupsfor approximately four months. In addition,to protectthe cameralenses from moisture,I glued 30 mm long shieldsto the camera objectives.These shieldswere cut from the plasticcovers sold for the metal pipessupporting shower curtains. A hole was cut in the vinyl cameracovers so that they would fit tightly around the protectivelens shield. To better camouflagethe camera setup, I painted the exposedparts, including the vinyl coversand the entire brake cablesbrown and green.

USE OF THE CAMERA SETUP Nests of most passetinesare usually well concealedin supporting vegetation.Therefore, I placed the camera setupsin densevegetation (denseshrubs and small treesallow the bestconcealment of the support stakes,the cableconnecting them, and the camera).The cameraand nest supportsare placedabout 100 cm apart (the minimum focusingdistance of the camerais approximately90 cm) and are drivenabout 20 cm into the ground. Ideally the camera should face to the north to prevent sun flares. The artificial nest is placedinside the supportingbasket and at- tachedby bendingthe endsof the wires aroundits edge.An eggis placed on the trigger loop and the trigger set up. This requiresthat the camera endof the brakecable be removedfrom the camerasupport and the brake line pushedinside the brake cableby compressingthe spring.The vinyl protectivecover is placed over the trigger parts and attachedwith an vol.58, No. • CameraSetup for Predation Studies [379 elasticband near the top of the stake (Fig. la). The camera is then inverted and fastenedto the camera support with elasticbands and a protectivecover is placedover it and attachedwith one elasticband. The flip flash is pushedinto its slot in the camera.The end of the brake cable is installed by pressingthe end of the line against the shutter lever and, simultaneously,tightening the wing nut. The position of the camera is checked(i.e., both the nest and the clock have to be in the camera's view) and the surroundingvegetation arranged to better camouflagethe setup. To set up a camera and a nest in a new location may take 10-20 min, dependingon the complexityof vegetation.The checkof a setupwhere predationoccurred which requiresplacement of a new eggon the trigger loop, winding the camera, and setting up the trigger mechanismtakes about 5 min. I recommendusing artificial nestsbecause they are more durable. I make thesenests by pressingtwo layers of dry grassinto a mold, with severalstrips of gluepoured over the first layer. The studyof eggpredation alsorequires a large number of .I have usedBlue-breasted (Coturnix chinensis)and Common quail (C. coturnix)eggs which are inexpensive and readily available.However, this techniquecould be adaptedto study predationon any eggs,regardless of their size, althoughit may require somechanges in choiceof materialsfor nest supportand trigger parts. The nest height could be altered by changingthe length of the nest support(A). The camerasupport (P) shouldbe adjustedaccordingly (the camerashould be positionedapproximately 20 cm higher than the nest). However, the study of predation on eggsof ground-nestingbirds would require one additional modification.Because in this casethe nest support would haveto be short(about 40 cm), the clocksupport must be separate from it. Finally, I recommendthe use of the Kodachrome64 slide film which resultsin the bestquality pictures. Problemsencountered and their solutions.--There are three possible causesof camera failure. First, the trigger is sometimesstiff and small predatorssuch as micewill not depressit. This problemcan be eliminated by moving the loop of the brake line (D, Fig. 2) closerto the hinge of the trigger bar (C, Fig. 2). In addition, becausethe trigger parts, made of piano wire, will corrodewhen exposedto moisture, oiling them on a regular basiswill slow down this processand reducefriction betweenthe moving parts. Second,the picture may not be taken becausethe predator has lifted an egg from the trigger loop without exerting the necessarypressure to set off the trigger. This is more likely to happen when the trigger loop (B, Fig. 2) is small relative to the egg placedon it. Therefore, making the loop larger sothat the eggwill fit in it moretightly will reducechances of this failure. Third, the cameramight fail to take a picture becauseof cablefriction or a weak spring. These problems can be correctedwith oiling, and adjustingthe cablemechanism. 380] J. Picman J.Field Ornithol. Summer 1987

Accordingto my experienceapproximately 10-20% of predation cases result in camera failures (inexperiencemay, however, result in higher failure rates). Therefore, I recommendthat the trigger parts be carefully examined each time the trigger is set up. Oiling the moving parts should be doneregularly (i.e., onceevery 2 wks) and more frequently in rainy weather. I have usedflip flashesat all times, regardlessof the light conditions. For this reason, pictures taken on sunny days and in the open habitats were slightly overexposed.In spite of overexposure,however, the iden- tification of predators on the slides was relatively easy. However, to improvequality of picturesin sucha situationone coulduse flip-flashes only when light conditionsare poor (but this would also increasethe number of visits to the setups). Flip flashesrarely fail. However, if one flash fails, all the remaining flasheswill also fail. Thus, to ensure that the flip flash is operating properly, the number of the flash usedshould be recordedfollowing the predation event. In 1985 I tested 10 setupsand in 1986 I used a total of 60 setups during a studyof predationon passerinenests in differenthabitats. During this studyin eachof the four habitatsstudied (marsh, meadow, scrubland, and coniferousforest) I distributed10 nestsin 10 m intervalsalong one or two transects.I obtained a total of about 900 pictures of predation events.These picturesrevealed a number of birds (e.g., Blue Jays, Cy- anocittacristata, Fig. 5; crows, Corvusbrachyrhynchos; Marsh Wrens, Cistothoruspalustris,_' House Wrens, Troglodytesaedon ; Catbirds,Dumetella carolinensis)and mammals (e.g., red squirrels,Tamiasciurus hudsonicus; flying squirrels,Glaucomys sabrinus; raccoons, Procyon lotor, Fig. 5; striped skunks,Mephitis mephitis;longtail weasels,Mustela frenata; deer mice, Peromyscusmaniculatus), suggesting that the cameraswill sample and identify a wide variety of egg predators.Predation was extremely high in the coniferousforest where usually all eggs would be removed by predators within 24 hours, intermediate in the meadow and scrubland areas (on averageabout 40% of eggsdepredated per day) and relatively low in the marsh (about 20% of eggsdepredated per day). Predation rates on experimentalnests varied, however,with the time in a season.Results on different predation rates, however, suggestthat the time between consecutive camera and nest checks should be shortened in habitats with high predationto allow samplingof all predatorsoperating at different times of day. Perhapsthe most important shortcoming of thistechnique results from the use of artificial nests. Predation on artificial nests could differ from predationon real nestsbecause: (1) the artificialnests and their location differ from real nests;(2) there are no birds defendingthe experimental nestsor coveringthe experimentalclutches; (3) the presenceof cameras nearnests might distract some predators; and (4) the scentleft by human observersmay affectchances of predation.Although some of thesediffi- cultiescould be solvedby comparingpredation rates on real and artificial xoL •8, •o. • CameraSetup for PredationStudies [381

r \

FIGURE5. Samplepredation events as recordedby the camerasetup described above: top, raccoon;bottom, Blue Jay (Cyanocittacristata). nestsand examiningthe appearanceof depredatednests, results on pre- dationfrom the camerastudy must be interpretedwith caution.Despite theseproblems, however, I believethat the cameratechnique should prove very useful for studiesexamining nest mortality factorsof individual speciesas well as passerinecommunities in different habitats. In conclusion,the automaticcamera setup described is an efficienttool for studyingpredation on nests.Because each setup is relatively inexpensive(between $40-50 Cdn), it is possibleto build manyof these setupsand usethem for large-scalestudies of predation. 382] j. Picman j. FieldOrnithol. Summer 1987

ACKNOWLEDGMENTS

D. Beedelland J. McAllister participatedin constructionof camerasetups in 1986 and, alongwith B. Jobinand A. Quintin, assistedwith the fieldwork.The NationalCapital Commissionkindly allowedme to conductmy fieldworkon their property.I wouldlike to thank J.-C. Belles-Isles,M. Leonard,D. Moore, and C. Schaefffor commentson an earlier draft of this paper.E. H. Burtt, Jr., P. Frederick,and an anonymousreviewer helped to improvethis manuscriptthrough many constructivecomments. This work was supported by a NSERC operatinggrant to J.P.

LITERATURE CITED

CROOK,J. H. 1964. The evolutionof socialorganization and visual communicationin the weaverbirds (Ploceinae). Behaviour Suppl. 10. LACK,D. 1968. Ecologicaladaptations for breedingin birds. Methuen, London. MURPHY,M. T. 1983. Nest successand nestinghabits of EasternKingbirds and other flycatchers.Condor 85:208-219. NILSSON,S. G. 1984. The evolutionof nest-siteselection among hole-nesting birds: the importanceof nestpredation and competition.Ornis Scand.15:167-175. PAGE, G. W., L. E. STENZEL,AND C. A. RIBIC. 1985. Nest site selection and clutch predationin the Snowy Plover.Auk 102:347-353. RICKLEFS,R.E. 1969. An analysisof nestingmortality in birds.Smithson. Contrib. Zool. 9:1-48. Received31 Oct. 1986; accepted14 Mar. 1987.