Bull. SouthernCaliforniaAcad. Sci. 109(2),2010,pp.23-36 ©SouthernCaliforniaAcademyofSciences,2010 Notes on the Breeding Biology ofthe White-Throated Swift in Southern California Charles T. Collins Department ofBiological Sciences, California State University, Long Beach, CA 90840, [email protected] — Abstract. Reproductive activities of White-throated Swifts (Aeronautes saxatalis) were examined from 1997 to 2006 at two nest sites in a human-made structure in southern California. The start of egg-laying was from 27 April to 30 May and hatching ofthe first chick ranged from 21 May to 23 June; all chicks had fledged by late July, 42^3 days after hatching. The annual molt ofadults began in early June and broadly overlapped with the chick-rearing period. Year to year adult survival was minimally 73.9% and nesting pairs showed strong mate and nest site fidelity; pairs reused nests up to five consecutive years. The composition of the arthropod food (insects and spiders) brought to nestlings was different in the periods 1997-2000 and 2001-2004 but prey size was similar in both periods. The onset ofbreeding was more varied from year to year than the start ofthe primary feather molt suggesting differing environmental stimuli for these important components ofthe annual cycle. The White-throated Swift {Aeronautes saxatalis) is a widespread and familiar element ofthe southern California avifauna. Its staccato chattering vocalization often announces its presence, searching for aerial prey high overhead, long before visual contact is made. In recent decades, this swift has utilized human-made structures as buildings, bridges and highway overpasses for nesting and roosting (Collins and Johnson 1982; Ryan and Collins 2000). In so doingit hasexpandedits local range fartherfromits more traditional rocky cliffsites. Recent studies have documented the molt and weight pattern ofdesert- inhabitingWhite-throated Swifts (Marin 2003) and the diet and behavior ofcoastal birds (Rudalevige et al. 2003; Ryan and Collins 2003a, b, c). Still there remains a paucity of information about the reproductive biology ofthe species (Ryan and Collins 2000). This is inlarge part due to the inaccessibility oftheirnestsin natural cliffsites; rock-quarrying techniqueshadto beemployedbyearlycollectorsto obtainthefirstnestandeggs (Hanna 1917; Bradbury 1918). The utilization ofhuman-made sites by these swifts holds promise of better access to breeding sites and the acquisition of more information on their reproductive activities. In Europe, breeding colonies of the Common Swift {Apus apus) (Lack 1956; Gory 1987; Kaiser 1992), Pallid Swift {A. pallidus)(Boano and Cucco 1989) and Alpine Swift (Tachymarptis melbd) (Arn-Willi 1960), situated in human-made structures have provided the necessary access fordetailed study. Until a similarcolony of White-throated Swifts is established, data on their reproductive activities will be characterized, as in this study, by regrettably small sample sizes. Methods The data presented here are based on a 9-year study ofWhite-throated Swifts nesting in a human-made highway overpass in Glendale, Los Angeles County, California (34° 11.16' N. 118° 13.01' W; elevation 361 m) between 1997 and 2006. The land-cover type 23 24 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES within 15 km ofthe study location is largely urban with some areas ofchaparral, coastal scrub and montane hardwood forest (Davis et al. 1998). The swifts entered the nest sites through 12.3 cm diameter drainage holes located on the underside ofthe overpass. There were eight such holes in the overpass ofwhich 5-7 appeared to be utilized each year by nesting swifts. This was determined by observation ofadults entering/leaving holes, fecal material below the hole orhearingvocalizingnestlings within. Only a single pairofswifts utilized any given opening. Two ofthe holes were accessible from inside making detailed observations possible each year. This program ofstudy included capture and banding of both adults and nestlings with numbered aluminum bands issued by the U. S. Geological Survey (Bird Banding Laboratory), aswell as documentation ofchick growth anddiet as indicated by analysis of food boluses brought to nestlings; adult body weight and molt were also studied. Food boluses were stored in 70% ethyl alcohol for later examination. Prey sizes (head to abdomen length omitting antennae and caudal appendages) were mm measured to the nearest 0.1 using an ocular micrometer. Adults and nestlings were weighed with a Pesola spring balance to the nearest 0.1 g and flattened wing lengths measuredwithastoppedwingruleto thenearest 1.0mm. Eggweightswereobtainedwith a portable electronic balance to 0.01 g. The molt of adult primary feathers was scored using a numerical system where each feather is given a value from to 5 (Newton 1966). In swifts, primary molt is usually symmetrical so only the primaries of the right wing were scored in this study. An unmoltedfeatherwas scored as 0, adropped ormissingfeatheras 1, anemergingnewpin feather < Vi regrownas 2, anewfeather lA to 3/4 regrownandpartiallyemergedfromthe sheathingas 3, anearlyfully (> 3/4) regrown featherwith onlysome sheathingat thebase as 4, and a fully regrown new feather as 5. A maximum score for replacement ofthe 10 primarieswouldthusbe 50. Amoltscoreof20mightbeachievedbyhavingprimaries 1-3 new(15), thefourthprimary 3/4 regrown (3) and thefifthprimary beingpresent asa short new pin feather (2) and primaries 6-10 old and unmolted (0). Results and Discussion Adult Body Weight Nine adult White-throated Swifts of unknown sex were captured and weighed on 19 occasions in June-July 1997-2006. The mean body weight was 32.5 g (SE = 0.54). This is similar to other data for this species (Table 1) including birds from both desert (Marin 2003) and coastal (Collins, in, Dunning 1993) areas of southern California. Individuals weighed on 2-5 occasions in different years varied in weight from -4.6 to +8.5 g (mean 2.64, SD = 2.53) orabout 8.2% ofthemeanforacoastal population(Collins, in Dunning 1993). The greatest annual difference in body weight was from an individual which weighed 29.8 g on 13 July 2005 and 38.3 g on 15 June 2006. Seasonal variation in body weight of up to 28.7% between the highest and lowest values were reported for a desert Tabic 1. Body weights ofadult White-throated Swifts in southern California. Mean weight (g) SE n Range Source 31.7 0.23 89 26.5-37.2 Marin 2003 32.1 0.55 20 27.8-36.0 Collins, in Dunning 1993 30.5 19 29.0-33.5 Bartholomew et al. 1957 32.5 0.54 19 29.0-38.5 this study WHITE-THROATED SWIFT BIOLOGY 25 Fig. 1. Nest, eggs, and developing chicks of White-throated Swifts: a - eggs show spotting by ectoparasiteexcrement, b-8-9dayoldchickswithdevelopingsemiplumefeathers seenas subcutaneous dots,c-23-24dayoldchickswithgreysemiplumespartiallyovergrownbyincomingcontourfeathers,d- 38^-0 day oldchick with sheathing still present at baseofsome dorsal contourfeathers. population of White-throated Swifts (Marin 2003). This is less than the variation recorded for some other species ofswifts (Gladwin andNau 1964; Naik and Naik 1966). Nests andnestparasites All ofthe White-throated Swift nests observed at this location were shallow saucers of plant material and feathers glued together with saliva (Figure la). A strip ofcellophane was incorporated into one nest. All the material was presumably gathered on the wing; construction of the nests was not observed. Two nests measured 8X8 cm, and 9.4 X 9.4 cm rim to rim and were 2.5-2.8 cm deep. New nests used for the first time contained less material and were often nearly torn apart by the normal movements of older nestlings. Nests that survived and were re-used in subsequent years (up to 5 successive years) became more substantial with the addition of new material each year. At the mouth ofone ofthe other ventilation holes a dead young swift was observed tangled in 26 SOUTHERN CALIFORNIAACADEMY OF SCIENCES string, presumably brought in by the adults as nesting material. A similar incident was reported previously at another location (Collins and Johnson 1982). These observations are in general agreement with earlier descriptions of White-throated Swift nests (Ryan and Collins 2000 and references therein). Parasites Infestations of mites (Acarina) and bedbugs (Cimicidae) were noted in varying intensities in different years. They were observed in substantial numbers on nests, eggs, and chicks in some years and nearly absent in others. These have been previously identified as Mimenstathia aeronautou (mites) and Snyexenoderus comosus (bedbugs) (Peterson et al 1980; Usinger 1996). Samples ofthe bedbugs collected in 2006 were tested at the USDA National Wildlife Research Center and all were negative for exposure to West Nile Virus (WNV; P. Oesterle, pers.com.). Samples of the feather louse Dennyus bruneri, endemic to White-throated Swifts (Carriker 1954), were collected from both adults and late stage nestlings. Eggs and Clutch Size The eggs ofthe White-throated swift, like those ofother swift species (Chantler 2000), were uniformly creamy white without gloss although they sometimes became spotted by excrement ofectoparasitic cimicids (Figure la). They showed little taper and have been described as varying from elongate ovate to cylindrical ovate (Bent 1940). In previous studiestheyhaverangedinsizefrom 18.6-22.6 X 12.9-14.7mm(RyanandCollins2000). Twelve partially incubated eggs in this study weighed, on average, 1.95 g (SE = 0.04, = range 1.7-2.2 g). Fresh egg weight of27 eggs from southern California was 2.01 g (SE 0.04) or 6.3% ofadult weight (M. Marin, in Ryan and Collins 2000). Hatching success was 82% (14 of 17 eggs). Unhatched eggs remained in the nest for up to two weeks until they were eventually crushed or accidentally ejected. Clutch size forfournests initially foundwitheggswas4, 4, 4and 5 eggs. Brood sizesin thirteen nests containing chicks when initially found ranged from 2 to 4, averaging 3.38 (SD =0.87). These values are in close agreement with previously recorded clutch sizes of 3-6 but normally 4^5 (Hanna 1917; Bradbury 1918; Bent 1940). Chick Development At hatching, White-throated Swift chicks were pink-skinned, devoid of any natal down, with eyes closed and a prominentegg-tooth. Theireyeswere partially open byday 8-10 post hatching (Figure lb) and fully open by day 15. The egg-tooth was still observable on day 8-9 but gone by day 18-19. The first feather coat was a covering of moderatelylonglightgreydown-likesemiplumes. Thesewereobservableassubcutaneous dark dots by day 7-8 (Figure lb) and erupted through the skin by day 11-13. The semiplumes were fully developed by day 20-23 but were covered over by the emerging contour feathers by day 28. The tracts of the body contour feathers were observable under the skin by day 11, and emerged through the skin by day 15. These pin feathers began to erupt through their sheaths by day 22-24 at which time they began to appear through the semiplume coveringofthechicks (Figure lc). Thecontour feathers appeared fully developed except for some remaining sheaths at their base by day 32-33; this sheathing was entirely gone by day 40 (Figure Id). The primary flight feathers emerged through the skin on day 8-11 and started to erupt through the pin feather sheaths byday 15 18. They had emerged from the sheaths for halftheirlength byday 23. The outermost WHITE-THROATED SWIFT BIOLOGY 27 140 F 120 E __ £ 100 » +« r O) .i!*** c 80 !:** _l 60 re 1o- .* -— c 40 (0 ' Oc) 20 > 10 20 30 40 Age (days) 50 Fig.2. Growthof(a)wing,tailand(b)bodyweightofWhite-throatedSwiftsinsouthernCalifornia,a. winglength shownbyblackdiamonds; taillengthbyblack squaresb. Dailymeanbodyweight shownby black squares. Clear symbols indicate daily maximum (diamonds) and minimum (triangles) weights. three primaries had some sheathing at their bases as late as day 39 but this was gone by day 42. Tail feathers emerged through the skin at day 9-10 and began erupting through the sheaths at day 17-18. The growth of the wing and tail was rapid and nearly linear for most of the growth period (Figure 2a). Byday42 thewinglengthsoftwochickscloselyapproximated that of adults captured in this study (131-143 mm, mean =138 mm); three other chicks at ages mm of40-41 days had only reached wing lengths of 105-117 or 76-85% of adult wing mm length. Tail feathers reached full adult length of 54-58 by day 35-37 (Figure 2a). 28 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES The pattern of plumage development in White-throated Swifts closely approximated that documented for the Common Swift (Lack and Lack 1951, 1954). The absence of natal downs in swifts has also been previously documented (Collins 1978). White-throated Swift chicks increased in body weight from about 2 grams at hatching to as much as 46 grams prior to fledging. From day 2 to day 17 chick growth was nearly linearandincreasingat arate of2.0 grams perday (Figure 2b). Thereafter, thepatternof growth is highly variable with some individuals reaching a peak weight in excess of40 g. by day 24 which is >123% ofadult mean weight (Table 1) followed by a decline to 30- 35 g prior to fledging at day M-AZ. This pattern of weight recession has been noted previously in both swifts and swallows (Ricklefs 1968). Weight recession has been attributed to maturation of the integument and feathers accompanied by decrease in water content (Ricklefs 1967, 1968). In this study however, some individuals did not reachpeakweight until much later (> day 32) and three chicks reached theirpeakweight of43.9^-5.5 g at days 40^11 immediately prior to fledging. Whathas been termed "abnormal growth" has been documented in a numberofavian species (Ricklefs 1976) including swifts (Lack and Lack 1951). It has been largely attributedto a reduction in fooddelivered to theyoung, particularlyearlyinthebreeding seasonwhennormal food suppliesare lowornotavailable(Lack andLack 1951; Ricklefs 1976). Such abnormal growth is often manifested in lower peak weights achieved by the chicks as well as a delay in the time to reach the peak weight. In addition, the rate of growth ofthewingand tail feathersmayalso be slowed (Ricklefs 1976). In this study, the abnormal growth ofsome White-throated Swift chicks, is illustrated here (Figure 2b) by lower peak weights, delay in the time to reach peak weight and a slower than typical growth ofthe wing and tail. This is most likely attributed to observer disturbance and, particularly, the capture or recapture of adults returning to the nest with food for the chicks. This in turn may have resulted in a temporarily decreased food delivery to the young and the observed abnormal growth. The more natural growth pattern, shown by some White-throated Swift chicks, was characterized by early acquisition ofpeak weight followed by moderate weight recession and fledging at or near adult weight. Asynchrony among and between broods Thepattern ofegglayingandincubationwasnotobservedinthis study. Elsewhere,egg laying takes place over 4-6 days, incubation does not start until the clutch is complete and hatching is asynchronous (Ryan and Collins 2000 and references therein). In the broods observed in this study, the chicks appeared to be ofunequal ages suggesting that hatchingtook place over2-3 daysinbroodsof4. Thesizedifferencesweremoreapparent when the chicks were older (>14 days old) and may have been accentuated by nestling competition for food. However, no chicks were observed to suffer severe growth impairment or starvation attributable to such intra-brood competition. Annual synchrony among broods was generally high with the median age ofthe chicks in the two nests under observation each year usually differing by only one to five days. However, in 1999, the inter-brood difference in age was 9 days and in 2004 it was an exceptional 13 days. The overall annual difference in age of the broods was 4.8 days (range 1 13 days, SD = 4.2). By backdating from the age of the oldest chick in the study broods and, assuming a mean incubation period of 24 days (H. Richardson, unpublished, in Ryan and Collins 2000). it was possible to calculate the date ofhatching ofthe first chick and the date of laving ofthe first egg. The date oflaying ofthe first egg ranged from 27 April (in 2005) to WHITE-THROATED SWIFT BIOLOGY 29 Fig. 3. Adult White-throated Swift carrying a bolus ofnestling food resulting in a distension ofthe floor ofthe mouth. Photo by Kevin Smith; reproduced by permission. 30 May (in 2000) with a median date of 13 May; hatching ofthe first chick ranged from 21 May to 23 June (median 6 June). A pair, which had remained together for two consecutive years, had laying dates of22 May (in 2005) and 8 June (in 2006). Food Swifts ofthe subfamilies Chaeturinae and Apodinae, including White-throated Swifts, bringfood to nestlings in the form ofa bolus ofaerial arthropods gummed togetherwith saliva. The bolus, consisting mostly ofinsects and some spiders, is carried in the mouth which results in a visible distension of the floor of the buccal cavity when adults are carrying food to nestlings (Figure 3). The first depiction ofthis in White-throated Swifts is a sketch by Louis Agassiz Fuertes made in the Chisos Mountains, Texas, in 1901. This sketch (Image # 215804, American Museum of Natural History Library) has been reproduced by Castro and Burke (2007). All prey items are caught on the wing at unknown distances from the nesting location. Each bolus can contain a wide array of prey taxa which range in number from 4-5 larger items to over 200 smaller ones (Rudalevige et al. 2003). During the study period nine such boluses were collected from adults returning to the nests to feed nestlings. Five boluses (1-5) collected between 1997 and2000wereanalyzedpreviously(Rudalevigeetal. 2003); analysis ofanadditionalfour boluses(6-9)collectedbetween2001 and2004isincluded here. Onebolusmeasured 12 X mm 15 and weighed 0.6 g. Inthe initialanalysis (boluses 1-5) 547 preyitemswere identifiedincluding spiders and seven orders ofinsects distributed among 50 families or superfamilies (Rudalevige et al. 2003). Hemiptera made up 41.9% ofthe prey items while Diptera (26.4%) were the most 30 SOUTHERN CALIFORNIAACADEMYOF SCIENCES Table 2. Annual variation in prey types captured by White-throated Swifts from the same nesting location in southern California. 1997-2000a (Boluses 1-5) 2001-2004 (Boluses 6-9) Hymenoptera 36 (6.58%) 58 (15.34%) Hemiptera 229(41.86%) 104(27.51%) Diptera 120(21.94%) 206(54.49%) Coleoptera 7(1.28%) 6(1.58%) Psocoptera 147 (27.24%) -0- Lepidoptera 1 (0.18%) -0- Thysanura 1 (0.18%) 1 (0.76%) Araneae 4 (0.73%) 3 (0.79%) Total preyitems 547 378 Prey Items/bolus 4-231 44-126 Mean prey items/bolus 109.4 94.50 SD 103.19 35.37 afrom Rudalevigeet al. 2003 diverse with 22 families represented. The four additional boluses (boluses 6-9) contained 378 individuals distributed among five orders of insects and three spiders; family level identifications were not attempted. A comparison ofthese two data sets (Table 2) shows substantial variation in prey types captured by swifts from the same nesting location in different years. Coleoptera, Thysanura, and Araneae were captured in about the same frequency in boluses 6-9 as in boluses 1-5 (Table 2). Hymenoptera and Diptera were more than twice as frequent in boluses 6-9 while Hemiptera made up a substantially greater percent of the prey items in boluses 1-5 (41.9%) than in boluses 6-9 (27.5%). Psocoptera which made up 27.2% ofthe prey items in boluses 1-5 were entirely absent from boluses 6-9. Similar year-to-year variation in the taxonomic composition of swift prey items has been noted in comparisons ofboluses obtained from Common Swifts in England (Lack and Owen 1955). Prey diversity also varied in a comparison of prey taken in geographically separated populations ofthe Alpine Swifts in Europe and South Africa (Collins et al. 2009). There was also substantial bolus to bolus variation in prey type captured duringa single breeding season by the Plain Swift (Apusunicolor) in the Canary Islands (Garcia-del-Rey et al. 2010). The size ofpreyitemstakenbyWhite-throated Swiftswasmuchlessvariablethanprey mm taxabetweenboluses 1-5 andboluses6-9 (Figure 4). Themeanpreysizewas 3.9 and 3.7 mm respectively (Table 3). These means are not significantly different (T = 1.94, p > .05). The largest prey item identified in boluses 1-5 was 16.7 mmbut only9.9 mmin boluses 6-9. The modal size ofprey itemswas 3 mm in both samplingperiods (Figure 4). The larger prey items (> 12 mm) in boluses 1-5 were honeybees {Apis mellifera) which made up three of the four prey items in one bolus. Both drone and worker honeybees have been recorded in the diets of other swifts (Lacey 1910; Hess 1927; Bartels 1931; Morse and Laigo 1969; Collins et al. 2009). The taxonomicdiversityofpreyitemstaken byWhite-throatedSwiftsindifferentyears at the same location, as well as the bolus-to-bolus variation suggests that these swifts are opportunistic foragers which readily exploit localized or seasonal abundances of prey within a preferred size range. This is in agreement with data from other swift species (Lack and Owen 1955; Thirumurthi and Krisna Doss 1981; Bull and Beckwith 1993; WHITE-THROATED SWIFT BIOLOGY 31 180 160 140 120 100 80 60 == 40 FHB iu.- 20 z 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Preysize (mm) Fig. 4. Sizes of food items delivered to chicks by White-throated Swifts at the same location in southern California in 1997-2000 (black bars) and 2001-2004 (shaded bars). Tarburton 1986, 1993). The mean prey size taken by White-throated Swifts in both sampling periods is similar to the mean prey size taken by six similar sized swifts in the genera Apus and Tachymarptis where there is a significant positive correlation between prey size and adult body weight (Collins et al. 2009). Survival rate, mate andnest sitefidelity White-throated Swifts occupiedmost orall oftheavailable sitesatthe studylocationin all years of this study. Continued site utilization was also noted at other locations in southern California (Bailey 1907; Skinner 1933; Collins 1973). Elsewhere they have been observed atthe samenest sites overmuchlongertime intervals (Dobkinet al. 1986). Such traditional nest-site use suggests a high survival rate ofadult breeders. No predation on adult White-throated Swifts was documented in this study. One brood offour chicks 9- 11 days old, two chicks 30-31 days old in a brood offour, and one chick 35 days old in a brood ofthree disappeared but no predator was identified. The small number ofadults banded in this studyprecludes a detailed survival analysis. In addition, not all breeding adults were captured, or recaptured, each year due to the probable disruption of chick provisioning and possible desertion. The available data indicates that 73.9% ofthe adults were known to be alive from one year to the next. This Table 3. Annual variation in prey body sizes (mm) taken by White-throated Swifts from the same nesting location in different years in southern California. Sample Years ofcollection Number ofprey items Mean prey size (mm) SE Range Boluses 1-5 1997-2000 547 3.9 0.09 0.7-16.7 Boluses 6-9 2001-2004 378 3.7 0.08 1.4-9.9 32 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Nest 1 Nest2 1997 1998 1999 c B ^^\.^^"^?? D ^ 2000 B /J> c 77 ?? / 2001 D* / ?? 77 77 2002 C E 77 F | 2003 77 F^ 77 G 2004 77 i'G 77 H 2005 ?? rG i 77 I 2006 ' B 1rG Ci h f CorrespondingU.S.G.S. Bandnumbers: A=752-08627 F=752-08661 B=752-08628 G=752-08671 C=752-08629 H=752-08684 D-752-08636 1=752-08691 E=752-08660 Fig. 5. Nest site and mate site fidelity ofbanded White-throated Swifts in southern California. No datawerecollected in 1998. Sexes were not determined. No attempt was made tocaptureall adultseach year to avoid disturbance and possible desertion. is a minimal value as it does not include the probability that others were alive but not recaptured or had emigrated to other nest sites at this or a more distant location One adult banded in July 1997 was still present and breeding in June 2006. The age of first breeding in White-throated Swifts is not known. Ifit is assumed to be when they are one year old then this individual would have been 10 years old when last recaptured. Similarly, another individual was eight years old when last recaptured in 2006. Nest site fidelity was variable in this study (Figure 5). One individual moved from nest hole one in 1997 to nest hole two in 1999 and was found back at nest hole one in 2000. Another individual was at nest hole one in 1999, nest hole two in 2000 and back at nest hole one in 2002. These two individuals were together at nest hole one in 2006 and were probably there together from 2002 to 2006 but were not recaptured during this interval.