The Wilson Journal of Ornithology 120(3):505–512, 2008

COMMON POORWILL ACTIVITY AND CALLING BEHAVIOR IN RELATION TO MOONLIGHT AND PREDATION

CHRISTOPHER P. WOODS1,2 AND R. MARK BRIGHAM1,3

ABSTRACT.—We investigated the influence of lunar and environmental factors on behavior of Common Poorwills (Phalaenoptilus nuttallii) in southern Arizona under a diverse set of natural and artificial conditions. Radio-marked poorwills were most active shortly after sunset during the new moon. Movements declined as evening progressed. Activity remained high for several hours after sunset when the moon was full. Poorwills were heard calling from March through October, but most calling occurred between early May and September. Only ambient light was correlated with number of poorwills heard calling. More poorwills responded to play- backs of conspecifics when the moon was full than when it was new. Poorwills did not change their response to conspecifics during full moon when playback of poorwill calls followed playback of Great Horned Owl (Bubo virginianus) calls but, during the new moon, fewer birds responded following the owl call. Poorwill behavior is strongly influenced by lunar conditions; their ability to detect and evade predators is important when calling advertises their location. Received 22 May 2006. Accepted 1 September 2006.

The behavioral influence of sunlight on are reduced during those periods (Cooper birds is evident, but the influence of moon- 1981, Mills 1986, Brigham and Barclay 1992, light on diurnal and nocturnal birds is less ap- Jetz et al. 2003). parent. For example, many nocturnal birds The Caprimulgidae is a circumglobal fam- call most actively at dusk in contrast to the ily of visually acute, crepuscular or nocturnal well known singing or calling at dawn by di- insectivores for which some aspects of lunar urnal birds (Brauner 1952, Cooper 1981, Ga- influence have been evaluated. The most com- ney 1990). Frequencies of calling, as well as monly studied aspect of caprimulgid biology overall activity patterns, are less well known in the context of lunar illumination is the pos- during true night, when moonlight is the pri- sible synchrony of breeding with lunar cycle mary source of illumination. Some animals (Jackson 1985, Mills 1986, Brigham and Bar- that are eaten by owls limit activity during clay 1992, Perrins and Crick 1996). The hy- periods when the moon is full (Price et al. pothesis that light levels constrain foraging 1984, Brown et al. 1988, Kotler et al. 1994, activity has also been proposed (Csada et al. Brigham et al. 1999, Beier 2005, Lang et al. 1992, Bayne and Brigham 1995, Jetz et al. 2006). Bright moonlight coincides with the 2003). Common Poorwills (Phalaenoptilus highest rates of nocturnal predation on Black- nuttallii; hereafter poorwill) are the smallest vented Shearwaters (Puffinus opisthomelas) North American caprimulgid (45–50 g), and and Cassin’s Auklets (Ptychoramphus aleuti- are one of the least understood of all North cus) by Western Gulls (Larus occidentalis) American birds, principally because of their (Nelson 1989, Keitt et al. 2004). Predation by cryptic coloration and nocturnal habits (Csada Great Black-backed Gulls (Larus marinus)on and Brigham 1992). We studied Common Manx Shearwaters (Puffinus puffinus) is also Poorwills under a diverse set of natural and strongly influenced by lunar condition (Brooke artificial conditions to assess the influence of 1990). The behavioral effects of different lu- environmental and lunar factors on their be- nar conditions are not universal, however, and havior. Specifically, we compared the likeli- other nocturnal animals increase activity when hood of movement by radio-marked birds be- the moon is full and the sky is relatively tween the full and new moons, and measured bright, presumably because visual constraints the rate of vocalization under a range of en- vironmental, lunar, and seasonal situations. 1 Department of Biology, University of Regina, Re- We also used playbacks of poorwill and Great gina, SK S4S 0A2, Canada. Horned Owl (Bubo virginianus) vocalizations 2 Current address: 11181 West Seneca Drive, Boise, ID 83709, USA. to examine whether the proximity to a poten- 3 Corresponding author; tial predator influences calling responses by e-mail: [email protected] poorwills. 505 506 THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 120, No. 3, September 2008

METHODS (Ͼ95% of the moon’s face illuminated; 82 bird-nights from 6 males, 5 females, and 2 We collected data during 1996–1999 at Ͻ three sites within 100 km of Tucson, Arizona, unknown gender juveniles) or no moon ( 5% of the moon’s face illuminated; 92 bird-nights USA: the National Audubon Society’s Apple- from 9 males, 4 females, and 2 unknown gen- ton-Whittell Research Ranch (31Њ 36Ј N, 110Њ der juveniles). We only used data from birds 30Ј W), central portions of the Buenos Aires that were euthermic (i.e., not torpid based on National Wildlife Refuge (31Њ 41Ј N, 111Њ 26Ј body temperature above 30Њ C) and for which W), and the east side of the Tortolita Moun- there was either movement or inactivity in at tains (32Њ 32Ј N, 111Њ 00Ј W). All three sites least one of the 5-min periods. We inferred are within the Sonoran desert ecosystem and that movement had occurred when the stan- share environmental and ecological attributes, dard deviation of the intensity of three signals including an elevation range of 1,000 to 1,550 recorded by the Lotek receiver was Ͼ10. This m, hot days and cool nights, limited rainfall value was derived from comparisons of vari- that typically occurs during the summer mon- ation in signal strength when movement or in- soon, and an abundance of cacti. Poorwills oc- activity by birds could be observed directly curred year-round at the Tortolita and Buenos (cf. Sutter et al. 1996, Brigham et al. 1999). Aires sites; some over-wintered at the Re- We used Chi-square tests to evaluate whether search Ranch site, although most were absent there were differences in the number of times during the coldest months. that movement occurred during each 5-min Poorwills were captured at night either us- period, for the two different lunar conditions. ing mist nets in conjunction with playbacks of Environmental Influences on Vocaliza- territorial calls or with a spotlight and long- tion.—We examined the influence of climatic, handled net (Swenson and Swenson 1977, lunar, and seasonal variables on poorwill vo- Jackson 1984, Brigham 1992). A few birds calization by conducting night-time counts of were captured or recaptured during daylight calling poorwills at established points in the by flushing them into mist nets placed near three study sites. Counts were made on 89 sur- roosts. Captured birds were banded with a vey nights between June 1996 and January USGS metal band and fitted with a tempera- 1999, although effort varied seasonally and ture-sensitive radio transmitter (2 models that between sites. Most surveys, especially during differed in mass and range were used: Model spring and summer, were at 7 to 10 day in- PD-2T, 2.8 g and 1–4 km range, and Model tervals, but logistics meant that five were sep- BD-2GT, 1.7 g and 0.5–2 km range; Holohil arated from a prior survey by only 2 to 4 days Systems Inc., Carp, ON, Canada). Transmit- and 10 were separated by 2 weeks to 10 ters were affixed using an elastic harness months. Two counts were done on 55 of 89 slipped over the wings (Brigham 1992, Hill et survey nights with 1–3 hrs separating each al. 1999). count making for 144 total counts. Poorwills Likelihood of Activity.—Radio-marked were counted at the Research Ranch and poorwills were monitored remotely using a Buenos Aires sites at nine points: three, each Lotek SRX 400 Telemetry Receiver (Lotek separated by 1 km, along each of three narrow Engineering Inc., Newmarket, ON, Canada) to and lightly-used dirt roads. Counts at the Tor- assess activity patterns by quantifying varia- tolita Mountains site consisted of either five tion in signal strength. Activity can be in- individual points or two sets of three points; ferred because signal strength of successive in both cases points were separated by 1 km radio pulses varies widely when radio-marked or more (logistics at Tortolita required two animals are active due to rapid changes in ori- separate routes). entation of the transmitting antenna (Sutter et Calling poorwills were counted at each al. 1996, Brigham et al. 1999). point for 3 min during which time CPW, who We evaluated the relationship between time conducted all counts, remained in darkness of night, lunar illumination, and movement by and silent (playbacks of poorwill calls were poorwills using telemetry data from 5-min pe- not used). The number of calling birds was riods exactly 1, 3, and 5 hrs past sunset on assigned based on differing direction and dis- nights for which there was either a full moon tance of calls and temporal overlap in calling Woods and Brigham • POORWILL ACTIVITY IN RELATION TO MOONLIGHT 507

(i.e., 2 or more birds calling at the same time). imum number of birds calling, not the number Counts were started at least 1 hr after sunset of birds specifically. We used two separate and completed in 1–2 hrs. No counts were backward stepwise multiple regressions to as- conducted during civil twilight (when the sun sess the importance of environmental vari- is 0Њ to 6Њ below the horizon), even though ables. First, for all 92 counts with light, tem- poorwills called frequently during this time, perature, and percent cloud cover as variables, so that solar light did not complicate our in- and second, for the smaller sample of 43 terpretation of the influence of lunar illumi- counts for which data on relative humidity and nation. No counts were conducted when it wind speed were also available. rained. Influence of Owls on Vocalization.—We as- We recorded environmental variables twice sessed the influence of owls on calling behav- (Tortolita Mountains) or three (Research ior by poorwills by measuring the response to Ranch and Buenos Aires) times during each playbacks of poorwill calls both with and survey, and averaged them to establish overall without the implied presence of a Great count conditions. We measured ambient light Horned Owl (these owls are common noctur- levels, temperature, relative humidity, average nal predators and were heard in all 3 study wind speed, and percent cloud cover. Wind areas). We surveyed for calling poorwills (in- speed and/or relative humidity were not mea- dependently from the counts described) using sured during 49 counts. A Beseler PM2L Col- two playback treatments to discern the effect or Analyzer (Charles Beseler Co., Linden, NJ, of potential owl presence on poorwill calling USA) was used to measure ambient light lev- behavior under differing light levels. Play- els (Hecker and Brigham 1999). This device backs were conducted at 1.6-km intervals on generates a unit-less measure of light intensity dirt road systems at the Research Ranch and to facilitate setting exposure times for film Tortolita Mountains sites, as well as on lightly printing. It works at light intensities at which traveled dirt roads on the east side of the Rin- many light meters are ineffective. We used a con Mountains, 50 km east of Tucson (32Њ Cole-Parmer thermistor thermometer (Model 07Ј N, 110Њ 28Ј W), on desert flats near Black 8402-00; Cole-Parmer Instrument Co., Vernon Mountain, 60 km north of Tucson (32Њ 49Ј N, Hills, IL, USA) to measure ambient tempera- 110Њ 57Ј W), and on the west side of the Sil- ture and a General Eastern thermo-hygrometer verbell Mountains, 55 km west of Tucson (Model 880, General Eastern Instruments, (32Њ 26Ј N, 111Њ 30Ј W). At each stop, we Woburn, MA, USA) to measure relative hu- waited in silence and darkness for 30 sec and midity. We measured wind speed (Ϯ 3 km/hr) then played a treatment tape. The test was over a 3–4 min period with a Kestrel wind conducted between 6 June and 11 July 1998 meter (NK Electronics, Chester, PA, USA), on calm or nearly calm nights with little or no which averaged measurements taken at 1-sec cloud cover, beginning at least 90 min after intervals. Percent cloud cover was visually es- sunset. The test usually took several hours and timated. was not repeated on the same evening. Three Most birds called during the warmer routes were surveyed twice, once during the months and we included for our analysis the full moon and again during no moon, and two 92 counts between 5 May and 3 September, were surveyed three times each, twice during during which time 80% of all calling oc- a full moon (separated temporally by 1 month) curred. Forty-seven of the 92 counts included and once when there was no moon. in our analysis were at the Research Ranch, We randomly assigned a treatment to each 42 at Buenos Aires, and 3 at Tortolita Moun- stop prior to surveys, but with the condition tains. We assumed there was no site specific that each set of two stops included each treat- lunar effect. The maximum number of birds ment. Counts were based on response to two heard at each point along each route within playbacks treatments: poorwill only (15 sec any year was used as an estimate of year- and poorwill, 60 sec silence, 15 sec poorwill) and site-specific bird maximums. The number of poorwill, Great Horned Owl, poorwill (15 sec calls detected during any count were stan- poorwill, 60 sec silence, 15 sec Great Horned dardized against the maximums and our anal- Owl, 15 sec poorwill). The treatment variation ysis was based on the percentage of the max- thus occurred in the second call sequence. Re- 508 THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 120, No. 3, September 2008 cordings used in treatments were taken from a standard collection of bird songs (Peterson Field Guides: Western Bird Songs, Houghton Mifflin Co., Boston, MA, USA). Complete 90–120 sec treatment tapes were created so that switching of tapes during counts was un- necessary. We used a portable cassette player, set at 3/4 maximum volume and judged that poorwills within ϳ300 m responded. We measured the number of responding birds and calling intensity during the 60-sec interval following playback of the initial call and again during the 60-sec interval following FIG. 1. Likelihood of activity by Common Poor- playback of the treatment call(s). The number wills during 5-min periods 1, 3, and 5 hrs after sunset of birds calling was based on the direction and in southern Arizona. There were between 97 and 135 distance of calls, and overlap in calling. To sampling periods within each time period, and activity characterize intensity, we counted the total differs within each (␹2 tests, all P Ͻ 0.05). number of calls, regardless of the number of birds calling, and assigned a categorical value based on six categories: (0) no calls, (1) 1–5 61% (3 hrs), and 69% (5 hrs) of each. The calls, (2) 6–10 calls, (3) 11–20 calls, (4) 21– likelihood of activity decreased over time dur- 30 calls, and (5) 31ϩ calls. Categorical values ing the new moon with movements occurring were used to minimize potential errors in in 86, 39, and 25% of the three periods, re- counting when calling was particularly in- spectively (Fig. 1). Birds were more likely to tense. We completed 12 surveys with a total be active 3 and 5 hrs past sunset when the ␹2 ϭ of 168 individual stops, 84 for each treatment. moon was full versus when it was new ( 1 ϭ ␹2 ϭ Ͻ Stops at which no poorwills were heard (29 5.1, P 0.023 and 1 19.1, P 0.001, in total), were eliminated from our analysis. respectively). The situation was reversed, We initially compared the full versus new however, 1 hr past sunset, when activity was moon results prior to the treatment playback more common during the new than the full ␹2 ϭ ϭ (i.e., the numerical and categorical response moon ( 1 5.1, P 0.024). to the first poorwill call only). A subsequent Environmental Influences on Vocaliza- analysis to assess treatment effects was based tions.—Poorwills in southern Arizona called on the per point change in number of birds or throughout an extended period from March calls from the 60-sec period after the first through October (Fig. 2). Most calling oc- playback to the 60-sec period after the second curred during the summer months. The central playback. Preliminary analysis indicated that 50 and 80% of all calls were recorded between data for both survey treatments were not nor- 27 May and 29 July, and 5 May and 9 Sep- mally distributed (Shapiro-Wilks’ W with as- tember, respectively. Ambient light levels sig- sociated P Ͻ 0.01); therefore, we used a non- nificantly predicted poorwill calling, whether parametric Mann-Whitney U-test for all com- based on the data from 92 counts for which parisons. only light, temperature, and cloud cover were ϭ 2 ϭ Ͻ analyzed (F1,89 92.2, R 0.509, P 0.001, RESULTS ␤ ϭ Light 0.71; Fig. 3), or based on the data Likelihood of Activity.—We used data from from 43 counts that also included relative hu- ϭ 2 ϭ 135 activity records (59 full moon, 76 new midity and wind speed (F1,41 49.1, R Ͻ ␤ ϭ moon) for the period 1 hr past sunset, 112 (56 0.545, P 0.001, Light 0.74). No other en- full moon, 56 new moon) for 3 hrs past sunset, vironmental variable measured was signifi- and 97 (48 full moon, 49 new moon) for 5 hrs cantly related to the number of calling birds, past sunset to examine poorwill movements. regardless of the data set used. The likelihood that a bird was active during Influence of Owls on Vocalizations.—The full moon was similar in the three time peri- effect of moonlight on calling was also evi- ods with movement occurring in 70% (1 hr), dent from playback surveys whose purpose Woods and Brigham • POORWILL ACTIVITY IN RELATION TO MOONLIGHT 509

FIG. 2. Seasonality of calling by Common Poor- FIG. 3. Relationship between ambient light at wills in southern Arizona from point count surveys night and calling by Common Poorwills during spring without playback of a poorwill call. Solid lines encom- and summer in southern Arizona without playback of pass the central 50% of all calls, and stippled lines the a poorwill call. Light levels are unit-less measurements central 80%. from a photographic light meter, with ‘‘10’’ represen- tative of no moon and little light, and ‘‘50’’ of a full moon and bright conditions. was to evaluate the effect of owl calls on poorwill vocalizing. Poorwills responded in significantly greater numbers and more in- responding poorwills following the poorwill tensely during full moons versus new moons only treatment during the new moon increased after playback of the first poorwill call. The 59% (from 0.56 to 0.89 birds) after the second number of birds responding during the new call compared to after the first call, but de- moon averaged only 38% of the number call- creased 11% (from 0.63 to 0.56 birds) after ing during the full moon (mean of 0.59 birds the Great Horned Owl then poorwill treatment ϭ ϭ ϭ at each point vs. 1.55 birds; Z54,85 5.66, P (Z27,27 2.42, P 0.016). Categorically, the Ͻ 0.001). The categorical number of calls de- number of calls increased 35% (from 1.59 to tected during the new moon averaged 52% of 2.15) after the poorwill only treatment, but de- the number heard during full moon (1.57 vs. creased 4% (from 1.56 to 1.48) following the ϭ Ͻ 3.02; Z54,85 4.51, P 0.001). Great Horned Owl then poorwill treatment ϭ ϭ There was no significant difference in the (Z27,27 2.19, P 0.029). change in calling response following playback of the second treatment call(s) during the full DISCUSSION moon either in number of birds responding or The intensity of night-time lunar illumina- intensity of calling. The number of responding tion varies by more than a hundred-fold de- poorwills increased 16% (from 1.64 to 1.90 pending on phase (Austin et al. 1976). Con- birds) per point following the poorwill only sequently, it is not surprising that lunar phase treatment during the full moon from the pe- should influence behavior of nocturnal ani- riod after the first call to the period after the mals, especially those that orient visually, al- second. Calling increased 14% (from 1.47 to though the direction of this influence may not 1.67 birds) after the Great Horned Owl then be obvious. In the absence of extenuating fac- ϭ Ͼ poorwill treatment (Z42,43 0.14, P 0.1). tors, visually-oriented nocturnal predators like The number of calls increased 21% (from 3.05 poorwills should concentrate foraging activity to 3.69) after the poorwill only treatment, and during bright moonlight but vocalize more 11% (from 3.00 to 3.33) following the Great frequently when there is little moonlight, ϭ Horned Owl then poorwill treatment (Z42,43 when foraging should be least efficient 0.85, P Ͼ 0.1). In contrast, there was a sig- (Brigham and Barclay 1992, Csada et al. nificant difference in the change in response 1992, Bayne and Brigham 1995). Chuck- by poorwills when the moon was new, both will’s-widows (Caprimulgus carolinensis) and in the number of birds responding and the cat- Whip-poor-wills (C. vociferus) call with great- egorical intensity of calling. The number of er frequency during moonlit conditions, how- 510 THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 120, No. 3, September 2008 ever, and lunar illumination is the only envi- al. 1992). These results suggest a trade-off ex- ronmental factor correlated with calling dur- ists between frequent calling and efficient for- ing the breeding season (Cooper 1981, Mills aging, assuming that calling reduces foraging 1986, Ganey 1990). Poorwills in our study vo- efficiency. The extent to which vocalizing calized most actively during the full moon and may restrict foraging by caprimulgids is un- lunar illumination was the only environmental certain, but males deplete fat reserves during variable we measured with which the number the breeding season whereas females, which of birds heard calling was correlated. In ad- probably call infrequently, do not (Csada and dition, poorwills tended to be more active Brigham 1992, Thomas et al. 1996). Energetic (which we interpret as foraging) during full shortfalls resulting from forgoing feeding moon, although the number of flying insect when prey are most visible are probably min- prey depended principally on temperature and imized since full moons are above the horizon was independent of ambient light (Woods for the most time. Consequently, more forag- 2002). Consequently, increased activity by ing time is available when the moon is rela- poorwills during the brightest periods was not tively full. During new moon, activity was related to increased prey density, and vice ver- highest shortly after sunset and diminished as sa. evening progressed, whereas during full Jetz et al. (2003) found a strong relationship moon, activity levels were consistent through- between lunar cycle and timing of foraging out late evening (cf. Brauner 1952). These activity by Standard-winged (Macrodipteryx contrasting patterns suggest that during new longipennis) and Long-tailed (Caprimulgus moons, poorwills must forage with greater in- climacurus) nightjars in equatorial West Af- tensity at and just past twilight, since low light rica. They noted that foraging was concen- levels will reduce foraging efficiency and may trated during the crepuscular period during all increase predation risk later in the night. phases of the lunar cycle but increased signif- Why are poorwills more vocal during pe- icantly around new moon. They also detected riods of high lunar illumination? In addition a small but significant increase in abundance to the offsetting increase in time available for of some larger insects during full moon peri- foraging, the ease with which males can move ods which correlated with increased foraging within and defend their territories is likely en- during periods of the night with lunar light. In hanced by relatively bright conditions. Our re- cooler more seasonal subtropical environ- sults suggest the ability to detect and evade ments, temperature as well as light levels predators is also important when calling ad- strongly affects activity patterns of Freckled vertises location. This explanation is support- Nightjars (Caprimulgus tristigma;R.A.M. ed by the higher intensity of calling in re- Ashdown and A. E. McKechnie, unpubl. sponse to conspecific calls following owl data). Unfortunately, calling behavior was not playbacks during the full moon, but an appar- evaluated in either of these studies. ent reluctance to respond under similar cir- Peaks in calling occur for some nocturnal cumstances during the new moon. Other rel- birds throughout the breeding season, and are atively small nocturnal birds that are espe- often attributed to different breeding stages cially vocal when the moon is bright also have (e.g., Clark and Anderson 1997). We found no acute vision and rely heavily on vision to for- seasonal peaks in calling, possibly because of age (e.g., Whip-poor-will). We surmise that the cyclic influence of lunar phase on calling visual acuity influences timing or extent to frequency. Poorwills are known to call persis- which smaller nocturnal animals vocalize at tently (Bent 1940, Gabrielson and Jewett night. For nocturnal animals that may less eas- 1940, Bailey and Niedrach 1965, this study), ily detect and out-maneuver predators, the in- and we heard birds call in every month but creased risk of predation during the full moon January. These observations support the hy- selects for less activity. Gerbils (Gerbillus pothesis that individuals are inclined to call spp.) and other desert rodents, as well as Aus- frequently where they occur as year-round res- tralian Owlet-Nightjars (Aegotheles cristatus), idents, regardless of breeding stage, but in lo- adjust foraging patterns to avoid bright moon- cations where they are migratory, calling light and are all preyed upon by owls (Price tends to cease in late summer (Kalcounis et et al. 1984, Brown et al. 1988, Kotler et al. Woods and Brigham • POORWILL ACTIVITY IN RELATION TO MOONLIGHT 511

1994, Brigham et al. 1999). Notably, one male LITERATURE CITED radio-marked poorwill was killed by an owl (Otus sp.) in our study. The bird was taken AUSTIN,R.H.,B.F.PHILLIPS, AND D. J. WEBB. 1976. A method for calculating moonlight illuminance under relatively dim conditions when the at the earth’s surface. Journal of Applied Ecology moon was in its first quarter and setting. 13:741–748. Whether specific predators and or predator BAILEY,A.M.AND R. J. NIEDRACH. 1965. Birds of foraging strategies are the major selective Colorado. Denver Museum of Natural History, pressures which have shaped the differences Denver, Colorado, USA. in response to lunar conditions between ca- BAYNE,E.M.AND R. M. BRIGHAM. 1995. Prey selec- tion and foraging constraints in Common Poor- primulgids and some rodents, remains to be wills (Phalaenoptilus nuttallii: Aves: Caprimul- understood. gidae). Journal of Zoology 235:1–8. Future research is needed with an emphasis BEIER, P. 2005. Effects of artificial night lighting on on explaining the role of lunar illumination on terrestrial mammals. Pages 19–42 in Ecological behavior of other birds that are active at night, consequences of artificial night lighting (C. Rich including those that are generally considered and T. Longcore, Editors). Island Press, Washing- ton, D.C., USA. diurnal (e.g., Johnson et al. 2003). The lunar BENT, A. C. 1940. Life histories of North American cycle is relatively short in comparison with cuckoos, goatsuckers, hummingbirds, and their al- the overall breeding season for most birds, lies. U.S. National Museum Bulletin 176. and consideration of this variable could pro- BRAUNER, J. 1952. Reactions of poor-wills to light and vide insight into foraging strategies and tim- temperature. Condor 54:152–159. ing of breeding. Knowledge of the lunar phase BRIGHAM, R. M. 1992. Daily torpor in a free-ranging and its effect on both foraging by and preda- goatsucker, the Common Poorwill (Phalaenopti- lus nuttallii). Physiological Zoology 65:457–472. tion risk on nocturnal birds may also be an BRIGHAM,R.M.AND R. M. R. BARCLAY. 1992. Lunar important consideration for surveying noctur- influence on foraging and nesting activity of Com- nal species (Ganey 1990, Downs 1998). Ad- mon Poorwills (Phalaenoptilus nuttallii). Auk ditionally, our study sites were relatively dis- 109:315–320. tant from large urban centers and the light pol- BRIGHAM, R. M., R. C. A. GUTSELL,R.S.WIACEK, AND lution associated with them. Light pollution F. G EISER. 1999. Foraging behaviour in relation to could relieve some dependence on moonlight the lunar cycle by Australian Owlet-Nightjars Ae- gotheles cristatus. Emu 99:253–261. for efficient foraging by predators but, as a BROOKE, M. 1990. The Manx Shearwater. T. & A. D. consequence, enhance predation pressure on Poyser, London, United Kingdom. their prey (Buchanan 2005, Frank 2005, Lloyd BROWN, J. S., B. P. KOTLER,R.J.SMITH, AND W. O. 2005). Perhaps more importantly, the loss of WIRTZ II. 1988. The effects of owl predation on cyclic variation in nocturnal illumination may the foraging behavior of heteromyid rodents. Oec- disrupt behaviors that evolved in association ologia 76:408–415. with regular fluctuation between bright and BUCHANAN, B. W. 2005. Observed and potential effects of artificial night lighting on anuran amphibians. dark periods (Rich and Longcore 2005). Ad- Pages 192–220 in Ecological consequences of ar- ditional research into the impact of light pol- tificial night lighting (C. Rich and T. Longcore, lution on the behavior of nocturnal animals is Editors). Island Press, Washington, D.C., USA. essential, considering the extent to which it CLARK,K.A.AND S. H. ANDERSON. 1997. Temporal, has altered the night sky in many developed climatic and lunar factors affecting owl vocaliza- regions. tions of western Wyoming. Journal of Raptor Re- search 31:358–363. ACKNOWLEDGMENTS COOPER, R. J. 1981. Relative abundance of Georgia caprimulgids based on call-counts. Wilson Bulle- P. A. Bradshaw, R. K. Sechler, and B. A. Woods tin 93:363–371. assisted in trapping and radio-marking poorwills. The CSADA,R.D.AND R. M. BRIGHAM. 1992. Common National Audubon Society and W. V. Branan provided Poorwill (Phalaenoptilus nuttallii). The birds of generous access to the Appleton-Whittell Research North America. Number 32. Ranch and logistic support. Comments by C. E. Braun CSADA, R. D., R. M. BRIGHAM, AND B. R. PITTENDRIGH. and two anonymous reviewers greatly improved the 1992. Prey selection in relation to insect avail- manuscript. Funding for this research was provided in ability by the Common Poorwill (Phalaenoptilus part by awards to CPW from the Faculty of Graduate nuttallii). Canadian Journal of Zoology 70:1299– Studies and Research at the University of Regina and 1303. an NSERC research grant to RMB. DOWNS, K. D. 1998. Common Snipe surveys, habitat 512 THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 120, No. 3, September 2008

and evaluation of the Breeding Bird Survey. The- of predatory risk and resource renewal on the tim- sis. University of Wyoming, Laramie, USA. ing of foraging activity in a gerbil community. FRANK, K. D. 2005. Effects of artificial night lighting Oecologia 100:391–396. on moths. Pages 305–344 in Ecological conse- LANG, A. B., E. K. V. KALKO,H.RO¨ MER,C.BOCK- quences of artificial night lighting (C. Rich and T. HOLDT, AND D. K. N. DECHMANN. 2006. Activity Longcore, Editors). Island Press, Washington, levels of bats and katydids in relation to the lunar D.C., USA. cycle. Oecologia 146:659–666. GABRIELSON,I.N.AND S. G. JEWETT. 1940. Birds of LLOYD, J. E. 2005. Stray light, fireflies, and fireflyers. Oregon. Oregon State Monographs 2. Oregon Pages 345–364 in Ecological consequences of ar- State College, Corvallis, USA. tificial night lighting (C. Rich and T. Longcore, GANEY, J. L. 1990. Calling behavior of Spotted Owls Editors). Island Press, Washington, D.C., USA. in northern Arizona. Condor 92:485–490. MILLS, A. M. 1986. The influence of moonlight on the HECKER,K.R.AND R. M. BRIGHAM. 1999. Does moon- behavior of goatsuckers (Caprimulgidae). Auk light change vertical stratification of activity by 103:370–378. forest dwelling insectivorous bats? Journal of NELSON, D. A. 1989. Gull predation on Cassin’s Auklet Mammalogy 80:1196–1201. varies with lunar cycle. Auk 106:495–497. ERRINS AND RICK HILL, I. F., B. H. CRESSWELL, AND R. E. KENWARD. P ,C.M. H. Q. P. C . 1996. Influence of 1999. Field-testing the suitability of a new back- lunar cycle on laying dates of European Nightjars pack harness for radio-tagging passerines. Journal (Caprimulgus europaeus). Auk 113:705–708. of Avian Biology 30:135–142. PRICE, M. J., N. W. WASER, AND T. A . B ASS. 1984. Effects of moonlight on microhabitat use by desert JACKSON, H. D. 1984. Finding and trapping nightjars. rodents. Journal of Mammalogy 65:211–219. Bokmakierie 36:86–89. RICH,C.AND T. L ONGCORE (EDITORS). 2005. Ecological JACKSON, H. D. 1985. Aspects of the breeding biology consequences of artificial night lighting. Island of the Fierynecked Nightjar. Ostrich 56:263–276. Press, Washington, D.C., USA. JETZ, W., J. STEFFEN, AND K. E. LINSENMAIR. 2003. Ef- SUTTER, G. C., D. J. SAWATZKY,D.M.COOPER, AND fects of light and prey availability on nocturnal, R. M. BRIGHAM. 1996. Renesting intervals in lunar and seasonal activity of tropical nightjars. Sprague’s Pipit, Anthus spragueii. Canadian Field- Oikos 103:627–639. Naturalist 110:694–697. JOHNSON, M., J. P. BECKMANN, AND L. W. ORING. 2003. SWENSON,J.E.AND S. SWENSON. 1977. Nightlighting Diurnal and nocturnal behavior of breeding Amer- as a method for capturing Common Nighthawks ican Avocets. Wilson Bulletin 115:176–185. and other caprimulgids. Bird Banding 48:279– KALCOUNIS, M. C., R. D. CSADA, AND R. M. BRIGHAM. 280. 1992. The status and distribution of the Common THOMAS, D. W., R. M. BRIGHAM, AND H. LAPIERRE. Poorwill in the Cypress Hills, Saskatchewan. Blue 1996. Field metabolic rates and body mass chang- Jay 50:38–44. es in Common Poorwills (Phalaenoptilus nuttallii: KEITT, B. S., B. R. TERSHY, AND D. A. CROLL. 2004. Caprimulgidae). Ecoscience 3:70–74. Nocturnal behavior reduces predation pressure on WOODS, C. P. 2002. Ecological aspects of torpor use Black-vented Shearwaters Puffinus opisthomelas. and inactivity during winter by Common Poor- Marine Ornithology 32:173–178. wills. Dissertation. University of Regina, Regina, KOTLER, B. P., Y. AYAL, AND A. SUBACH. 1994. Effects Saskatchewan, Canada.