Philornis Ectoparasitism of Pearly-Eyed Thrashers. Ii

PHILORNIS ECTOPARASITISM OF PEARLY-EYED THRASHERS. II. EFFECTS ON ADULTS AND REPRODUCTION

WAYNE J. ARENDT U.S.D.A.Forest Service, Institute of TropicalForestry, Southern Forest Expen'ment Station, P.O. Box AQ, RœoPiedras, Puerto Rico 00928 USA

ABSTRACT.--Arain-forest population of the Pearly-eyedThrasher (Margarops fuscatus) suffered heavy ectoparasitismfrom the larvae of a tropical fly recently identified as Philornis deceptivus.The impact of parasiticattack was of lessconsequence to adult thrashersthan to nestlings.Nestling mortality was high and was causedmostly by philomid ectoparasitism, other factorsbeing negligible(3%). The extent of larval infestationsfluctuated seasonally and was related to monthly rainfall. Infestationdid not have to be heavy to causedebilitation and death. Larval infestationsites varied with the nestling'sontogeny, especially pterylae development.Overall 4-yr fledging successwas 53.3%,fairly high for a tropical passefine. However, observationsof heavily parasitized nestlings that died during fledging attempts and the deathsof postfledgingjuvenile thrasherssuggest that first-yearmortality may be as high as 80%in nestlingsthat sufferedheavy larval infestations(357 of 448 nestlings).Juve- niles facekeen intraspecificcompetition after leaving the nest.A combinationof high nestling and juvenile mortalityand little opportunityfor juvenilesto enter the breedingpopula- tion couldgreatly reduce this host'snumbers. There are signs,however, that the Pearly-eyed Thrasheris adjustingto heavy philomid ectoparasitismby lengtheningthe breedingseason and reproducingwhile fly populationsare seasonallylow (December-March)and, thus,may remain in abundancein the rain forest.Received 31 August1983, accepted 10 June1984.

MYIASlS "is the infestation of live human and Thrasher(Margarops fuscatus), a more abundant vertebrate animals with dipterous larvae, cavity-nesting specieswith a similar ecology. which, at leastfor a certain period, feed on the Pearly-eyed Thrasher nestlingsalso are infest- host'sdead or living tissue ..." (Zumpt 1965). ed by anddie fromphilornid larvae. The Pearly- In the Old World, 5 genera of bird-infesting eyed Thrasher is restricted to the West Indies diptera from 3 major families (Neottiophilidae, and hasa patchydistribution, ranging from St. Calliphoridae, and Muscidae) are reported to Vincent in the Lesser Antilles to the Bahamas infest adults and nestlingsor inhabit nestsof (Bond 1979). This rapaciouspasserine, well at least 9 avian orders, 28 families, and some known for its predatory habits (Rolle 1965a,b; 70 speciesof birds (Owen 1954, Zumpt 1965, pers. obs.),is omnivorous,feeding on a variety Pont 1974, Bakkal 1980). In the New World, of fruits, invertebrates, and vertebrates. In the species.ofthe common Europeanbird parasite Luquillo Mountains it is monogamousand a Protocalliphoraspp. (Calliphoridae;Plath 1919a, secondarycavity-nester, often nestingin crev- b) and a unique genus Philornis(Muscidae; ices and tree hollows. Dodge 1955, 1963, 1968;Pont 1972) have been Becauseseveral Philornisspecies are polytyp- reported. Some 30 Philornisspecies have been ic (Dodge and Aitken 1968), the taxonomyof collected and described from birds south of the this group has alwaysbeen extremelydifficult United Statesto the Neotropics(Pont 1972).On (reviewed by Aldrich 1923). In Puerto Rico two Trinidad alone 10 speciesof Philornisassociated Philornisspecies have been reported: P. obscura with 29 avian speciesfrom 17 families are rec- (Van der Wulp) from Ponceand P. pici (Mac- ognized (Dodge and Aitken 1968). quart) from Mayaguez (Wolcott 1948). Recent- In Puerto Rico the endangeredPuerto Rican ly, adult specimensthat emergedfrom puparia Parrot (Amazonavittata), a cavity-nesting bird, that I collected from thrasher nests in the Lu- is parasitizedby larval philornids, which have quillo Mountains were identified as P. decepti- causedthe death of parrot nestlings.As part of vusDodge and Aitken, 1968,by M. S. Couri, of a large-scalerestoration program to save the the National Museum of Rio de Janeiro, Brazil. parrot, study was begun on the Pearly-eyed Philornisdeceptivus previously had been record-

281 The Auk 102:281-292. April 1985 282 W^¾•E J. ARENDT [Auk, Vol. 102

ed only on Trinidad, its type locality (Dodge Over a 4-yr period (1979-1982),I took data on 272 and Aitken 1968, Pont 1972). nests, 681 nestlings, and 105 adults of the Pearly- Although many host recordsinvolving phi- eyed Thrasher.I usedartificial nest boxesdesigned lornid flies have been cited (Pont 1972), few as part of the Puerto Rican Parrot restorationpro- studies have concentrated on the deleterious gram (Snyder1977, Snyder and Taapken1977). Boxes were installed at 0.1-kin intervals between kms 12 effectsthat infesting fly larvae can have on an and 17 alongHighway 191,3-50 m from the roadway avian host over extended periods of time. With within the rain forest. Additional nest boxes were the exceptionof recent studiesby Smith (1968, added until the number of available boxes increased 1980), Bakkal (1980), Hector (1982), and Win- from 26 the first year to 48 the fourth year. From terstein and Raitt (1983), most of the literature January1979 to August 1982nest boxeswere visited on avian ectoparasitisminvolving Old and New throughout the breeding seasonon an average of World dipterans has been limited to the tax- every two days, but often on a daily basis during onomyof the adult fliesand to partial descrip- criticalperiods (laying, hatching, and fledging).Dur- tion of their life history stages(Nielsen 1911; ing the nonbreedingseason (September-November) boxes were checked once a week or once every two Gilbert 1919; Dodge 1968; Ledger 1969; Pont weeks for signsof nesting,depending on the prox- 1972, 1974). Some reports are anecdotal ac- imity of breeding activity. After eggs hatched, I countsof larval impactson host species(Plath checkedthe nestlingsfor fly larvae at every nest vis- 1919a, b; Hindwood 1930). it, notingnumbers, positions, and infestationperiods This study was designed to better under- of the larvae. I trapped larvae in funnels below the stand the host-parasite relationship and to nest cup and allowed them to pupate in small saw- quantify the direct effects that dipteran ecto- dust and woodchipvials coveredwith screens.I also parasitismcan have on an avian host and its collectedpuparia that adheredto nest twigs. During reproduction.Specific objectives were to deter- the 1979and 1980breeding seasonsI removed larvae mine 1) the generalprevalence (percentage of from some nestlingsas soon as they were detected in order to determine thrasher fledging successin host individuals infested;terminology of Mar- the absenceof ectoparasitism. golis et al. 1982)and intensity (mean number To determineif ectoparasitismwas dependent upon of larvae/infested host) of larval infestations, siblinghatch order, I randomlychose 90 parasitized 2) seasonalchanges and how they affect the nestlings(3/nest), separatedthem by hatch order, host-parasiterelationship, 3) at which of the and performed a one-way analysisof variance test host's anatomical sites larvae are most often (c•= 0.05; Shedecorand Cochran 1980) on the mean found, 4) the overall effects on host reproduc- number of infesting larvae/nestling (from the total tive success,and 5) possibleresponses by the number of infesting larvae found during the entire host to counter Philornisectoparasitism. nestling period). Bartlett'stest revealed heteroge- neousvariance among classes,so I useda square-root transformation to stabilize the variance. STUDY •REA AND METHODS I testedthe significanceof the differencebetween This studytook placewithin the 11,200-haLuqui- two proportions(Sokal and Rohlf 1969)to determine llo ]LxperimentalForest located in easternPuerto Rico if adult female thrashersexperienced significantly (18ø19'N,65ø45'W) at an elevation of ca. 800 m within morefly larvaethan adult malesand to determineif subtropicallower montaneforest (Holdridge 1947, site specificityof the infesting larvae on thrashers 1967;Ewel and Whitmore 1973). A detailed descrip- differed significantlybetween adults and nestlings. tion of this forest can be found in Odum and Pigeon To testfor a significantcorrelation between site spec- (1970). ificity of the larvaeon nestlingthrashers (dorsal lo- Annual rainfall varies from 3,000 mm in the foot- cationson young nestlingsvs. ventral siteson older hills to more than 5,000 mm on the highest peaks young), and to test for possibletemporal variation (Odum et al. 1970). Rainfall within the main study (earlyvs. late season)in larvalsite specificity, I used site averaged4,660 mm from 1935 to 1943 (Wads- a nonparametricstest for correlation(phi coefficient; worth 1948).A more recent12-yr summary of NOAA Conover 1971). data (1970-1981) from the Pico del Este weather sta- Adult and nestling thrashers were individually tion located about 5 km from the study site shows a marked with color bands and a metal USFWS band mean rainfall of 4,117 ram, which indicates that rain- for field identification. Outside of the breeding sea- fall patternsremain similar over long periods.Pho- son adult thrasherswere capturedin mist nets locat- toperiod varies little, with 12-14 h of daylight ed in their territories. During the breeding season throughoutthe year. For a 9-yr period, Wadsworth adultswere capturedby the use of pull-string trap- (1948)showed that the temperaturewithin the study doors mounted on the nest boxes. Observations were site averaged21.1øC (range 11.1-32.2øC). madefrom burlapblinds with fiberglassroofs placed April 1985] PhilornisEctoparasitism onThrashers 283

TABLE1. Monthly prevalenceand mean intensity of TABLE2. Monthly prevalence and mean intensity of Philornisdeceptivus larvae on 105 adult Pearly-eyed Philornisdeceptivus larvae on 448 nestling Pearly- Thrashers (1979-1982). eyed Thrashers(1979-1982).

Percent Intensity Number Month prevalence(n) a (range) of nest- Number January 0 (7) 0 lings Preva- of February 0 (6) 0 sam- lence para- Intensity March 15.0 (20) 3.3 (2-6) Month pied (%) sites (range) April 21.1 (19) 1.7 (1-2) May 25.9 (27) 4.0 (1-6) January 3 100 69 23.0 (19-26) June 64.3 (14) 3.5 (2-7) February 25 80.0 432 20.6 (0-49) July 91.7 (12) 3.2 (1-6) March 43 81.4 1,086 25.3 (0-73) April 97 96.9 4,215 40.9 (0-220) number of parasitized adult thrashers. May 115 100 4,720 45.4 (6-131) June 95 100 6,294 61.7 (3-169) July 70 100 2,946 42.1 (5-135) 5-20 m from the nest boxes.Four nestlingswere fit- ted with radio transmittersin a study of postfledging mortality. Three of these birds hatched in the same nest box, 1 young from the first brood (March) and were parasitized, while only 6 of 45 males 2 from the second(May). The fourth individual was (13.3%) were infested (P < 0.0001; Sokal and selected from a distant nest box at the end of June. Rohlf 1969). Females constituted 80% (28 of 34 Each radio package consistedof a model 1220-LD transmitter with elastic harnesses. Each transmitter individuals) of adult thrashersfound to be par- weighed 5.5 g, or ca. 5% of the nestling'sbody mass. asitized by philornid flies. A 3-element Yagi null-peak system antenna and Ninety-six percent of 448 nestling Pearly- model TRX 24 receiverwere used to locateradio sig- eyed Thrashersexamined from 1979 to 1982 nals. Radiotelemetry equipment was supplied by were infested with philornid larvae (Table 2). Wildlife Materials, Carbondale, Illinois. Prevalenceof larval infestation over the reproductive seasonranged from 80 to 100%.With RESULTS few exceptions,all nestlings were parasitized eachmonth. Overall mean intensitywas 37 lar- Generalpatterns of infestation.--Examinationof vae/nestling (range 0-220). However, mean in- 105 adult Pearly-eyedThrashers during the re- tensity increasedfrom 20.6 larvae/nestling in productive period (January-July)in 1979-1982 Februaryto 61.7 larvae/young in June.This in- showed that 31% were infested with larvae of creasecorresponded with an increasein the P. deceptivus(Table 1). A mean intensity of 3.1 number of nestling thrashersavailable for phi- larvae/bird (range 1-7) was observed.Because lornid infestation and the total number of fly samplesizes were small, between-yearcompar- larvae found on nestlings(range 69 in January isions were not made. However, it was appar- to 6,294 in June). Wide monthly ranges in the ent that prevalence of infestation (percentage number of larvae/nestling show the variability of host individuals infested) did change over of larval infestations at different nest boxes each the reproductiveperiod. No adult thrasherwas month. found infested during January or February, Initial infestation by philornid larvae on while prevalenceof infestation increasedfrom nestling thrashersvaried greatly over the re- 15.0% in March to 91.7% in July. Mean inten- productive period. Throughout the first two- sity of infestation (mean number of larvae/in- thirds of the breeding season(February-May) festedhost) ranged from 3.2 to 4.0 for all months adult flies did not parasitize a nest until the except April, when 1.7 larvae/bird were ob- oldestnestling was about one week old (Table served. The presenceof large cutaneouslesions 3). Although the oldestnestling was not always (ca. 3-4 mm in diameter),left by vacatinglar- the first sibling parasitized,I basedthe timing vae about to pupate, indicated that successive of initial parasitismon the vulnerability of the larval infestations occurred. Incubating and first available host (first-hatchednestling). Lar- brooding female thrashers were infested more vae remained in the hosts an average of 5.1 frequently than males, which do not incubate days(SD = 1.07, range 3-9, n = 500). Although or brood. Of 60 captured females, 28 (46.7%) successive infestations were common, larvae 284 WAYNEJ.ARENDT [Auk,Vol. 102

TABLE3. Mean age (days)of the oldestPearly-eyed Thrasher nestling in a brood at initial infestation 160 by Philornisdeceptivus larvae (1979-1982).

Month Mean age _+SD n Rangea February 6.90 + 2.49 23 5-14 :00 ' '""il.' March 7.00 _+ 4.02 30 2-14 ,,oE.,o .. ...' : &t...... - April 7.26 _+4.35 50 1-17 ._1 May 7.07 + 3.32 50 3-14 June 4.32 + 1.87 50 0-8 2 4 6 8 10 12 14 16 18 20 22 July 2.41 _+ 1.17 50 0-4 Age (days) hatch day, i.e. lessthan 24 h old. Fig. 1. Numbers of Philornisdeceptivus larvae in- festingPearly-eyed Thrasher young during the nestling period.Outlying points,especially those at the tended to complete development during the beginning and end of the nestling period, represent middle portion of the nestlings'development late-seasonyoung. period (Fig. 1). Late-season nestlings that hatched in June and July, however, were heavily parasitizedwithin the first two days along the tracts,lying in clustersof 10 or more. after hatching.Late-season young that lived to Many larvae also were found on the wings of fledge often harbored 50 or more larvae during the nestlingsthroughout development (Tables their last day in the nest (Fig. 1). 4, 5). Common sites included the rapidly de- No significant difference was found in the veloping manus and ulnas, future sites of the total number of larvae/young and hatch order fast-growingbrachial feather papillae (Table 5). of the nestlings within individual nest boxes. Interestingly, in the tarsi, whose growth was All siblings within a brood received about the greatly affected by philornid ectoparasitism samenumbers of infesting larvae (t = 30.8, 32.0, (Arendt 1983),larvae aggregatedin the region and 32.8 larvae/nestling, respectiveto hatch of the tibiotarsusand at the tibiotarsal joints order; F2,•47= 0.04, P > 0.250). throughout the nestling period (Table 5). Site specificity.--Althoughphilornid larvae The distribution of infesting larvae varied were observedon all body surfacesof the birds, with the nestling'sontogeny (Table 6). There significantdifferences in somespecific sites of was a positive correlation [c•= 0.05, phi coef- infestation were found between adult and ficient= 0.4, critical level (P-value)< 0.001] nestling thrashers(Table 4). Almost 60% of the showingthat significantlymore larvae infested larvae found on adults were located on the dorsalareas in young nestlings(0-9 daysold) ventral surfaceof the wings, with 30% in the and significantlymore larvae infestedventral patagial membranes,17% in the axillaries, and areasin older nestlings(10-19 daysold) during the remaining larvae in brachial sites more or the first half (early season)of the reproductive less evenly distributed among the humerus, period. A positive correlation also was found ulna, and manus. A large larva (15 mm long x for nestlings hatched during the second half 5 mm wide) was found at the baseof a recently replaced primary pinfeather on one adult TABLE4. Site specificityof Philornisdeceptivus larvae thrasher. The second most common site of in- on adult and nestling Pearly-eyedThrashers (1979- festation on adult birds was the head region. 1982). Most larvae were found at the base of the bill, Larvae at lores,and above the eye orbits.One larva was Site of each site (%) found developingin a lower eyelid, forcingthe larval lid to closepartially. Larvaealso were found in infesta- Adults Nestlings the ventral tracts, humeral tracts, and inter- tion (n = 34) (n = 432) scapularregions of the capital tract. Wing 58.8 19.2 P < 0.0001 Of the larvaeinfesting nestling thrashers, 54% Head 14.7 17.2 NS were found on the trunk, with 38% on ventral Trunk surfacesin associationwith feather papillae at Ventral 11.8 30.4 P < 0.02 Dorsal 8.8 14.1 NS the base of the two ventral tracts (Tables 4, 5). It was not uncommon to find 50 or more larvae Legs 5.9 19.1 P < 0.05 April1985] PhilornisEctoparasitism onThrashers 285

TABLE5. Comparisonof relative abundanceof Philornisdeceptivus larvae at specificanatomical sites on 205 nestling Pearly-eyedThrashers (1982).

Early season Late season (9 Feb. to 15 May) (16 May to 27 July) Larvae Larvae

Site of infestation (n)a n % Site of infestation (n)a n % Ventral tracts (93) 1,690 34.2 Ventral tracts (82) 1,555 23.6 Tibiotarsus (91) 582 11.8 Tibiotarsus (97) 864 13.1 Ulna (84) 374 7.6 Crown, orbits (83) 580 8.8 Manus (83) 312 6.3 Ulna (84) 448 6.8 Crown, orbits (66) 283 5.7 Manus (74) 404 6.1 Tibiotarsal joint (70) 220 4.5 Tibiotarsal joint (72) 373 5.7 Humeral tract (66) 210 4.3 Humeral tract (80) 315 4.8 Throat and underneck (61) 182 3.7 Throat and underneck (71) 288 4.4 Dorsal region (62) 157 3.2 Dorsalregion (87) 288 4.4 Flanges(76) 148 3.0 Pelvic region (77) 227 3.4 Cloaca (50) 146 3.0 Humerus (76) 222 3.4 Humerus (66) 136 2.8 Patagialmembrane (69) 198 3.0 Pelvic region (66) 135 2.7 Flanges(88) 191 2.9 Patagial membrane(54) 123 2.5 Foot pads (65) 121 1.8 Baseof culmen (37) 59 1.2 Upper neck (55) 118 1.8 Upper neck (34) 54 1.1 Cloaca(32) 114 1.7 Interscapularregion (33) 51 1.0 Interscapularregion (50) 113 1.7 Oil gland (20) 37 0.7 Oil gland (55) 91 1.4 Foot pads (25) 36 0.7 Baseof culmen (47) 85 1.4 Total 4,935 100.0 6,595 100.0 number of parasitizednestling thrashers.

(late season)of the reproductive period, but it lings in general (Fig. 1). The numbers of in- was less obvious (phi coefficient= 0.28) be- festinglarvae increased during the middle por- causeof the presenceof more larvaein the ven- tion of nestling development and declined tral areasearly in the nestlingstage (Table 6). rapidly as the normal fledging period ap- However, the critical level remained less than proached (Fig. 3). Among nestlings that suc- 0.001. cumbed to ectoparasitism,first- and second- Nestlingdeath and debilitation.--Nestling mor- hatchedsiblings withstood significantly greater tality was high and was causedalmost exclu- numbers(Duncan's New Multiple Range Test) sively by Philornis ectoparasitism. Of 448 thrasher nestlingsinfested with fly larvae in naturally parasitizednests over the 4-yr study TABLE 6. Distribution of infestation sites of Philornis period, 209 young died as a result of ectopara- deceptivuslarvae on 205 nestling Pearly-eyed sitism(Fig. 2). Other causesof mortality were Thrashersduring developmentand throughoutthe reproductiveseason (1982). a incidental(3%). Fifteen nestlings died from such causesas disease,death during hatching, rat Early season Late season and unknown predation, windstorms, and (n = 88) b (n = 117)• being trampled by siblings. During the 1979 Age (days) Age (days) and 1980breeding seasons 73 of 74 (98.6%)un- Site 0-9 10-19 0-9 10-19 parasitized nestlings successfullyfledged at nests in which all larvae were removed from Dorsal 785 495 1,383 479 nestlings.In contrast,during the 1981and 1982 Ventral 767 2,888 1,951 2,782 •bc = 0.402, •bc = 0.282, breedingseasons nestling mortality at natural- P < 0.0001 P < 0.0001 ly parasitizednests reached 55.8% in 1981and 41.6% in 1982. • Presented as total numbers of larvae from each anatomicalzone. Early season is 9 Februaryto 15 May; In nestlingsthat succumbedto philomid ec- late seasonis 16 May to 27 July. toparasitism,the pattern of larval infestation • n = number of parasitizednestling thrashers. was similar to that found in parasitized nest- cPhi coefficient of correlation (see Methods). 286 W^¾NEJ. ARENDT [Auk, Vol. 102

450

65 = 400 ... 6O

c 55 •z 350 -- 50 o

300 Z 45 ; 4o E• 2_50 Z

20O o 2 4 6 8 I0 12 1416 182022 24 • 20

15 Age (days) I0 Fig. 2. Survivorshipof 448 Pearly-eyedThrasher 5 young during the nestling period (1979-1982). Not all nestlings were parasitized by Philornisdeceptivus 0 larvae, nor did all infested nestlingsdie. 0 2 4 6 8 I0 12 14 16 18 2022 24

Age (days) of infesting larvae (œ= 65.1 and 59.6 larvae, Fig. 3. Daily numbersof Philornisdeceptivus larvae respectively;F2,87 = 4.13, P < 0.025) and sur- on Pearly-eyedThrasher young that died from ec- vived longer (œ= 12.9 and 12.1 days, respec- toparasitismduring the nestlingperiod. Vertical lines tively;F2,87 = 3.12,P < 0.05)than third-hatched are _+1 SE, horizontal lines are mean numbers, and nestlings(• = 41.0 larvae and 9.1 days). closedrectangles are 95%confidence intervals; width The extent of larval infestations on thrashers of rectangleswas chosen arbitrarily. fluctuatedseasonally, increasing each month (Fig. 4), and apparentlywas related to monthly rainfall. Increased rainfall was associated with succumbedto philomid ectoparasitism,the av- highermortality (Fig. 5). Nestlingmortality was erage age at death was 11.1 days (SD = 5.19; higher in the wetter 1979 (50.0%) and 1981 range 2-31) from Februaryto June.It dropped (55.8%)breeding seasons than in the drier 1980 to 7.4 daysfor July of eachyear, when fly pop- (35.6%) and 1982 (41.6%) seasons.Similarly, ulations had greatly increased. Nestlings nestling mortality from Februaryto May aver- hatchedin late June and July were often par- aged 60% in the wetter years,as comparedto asitized within a day after hatching, suffered only 20%in the drier years.Post-May mortality higher larval infestations,and died within a increasedin all yearsas hostand parasitepop- week. ulations increased. If larval infestation occurred at or just after The age at which thrasher nestlings suc- hatching, even light larval loads causeddeath cumbed to philomid ectoparasitismwas de- up to the fifth day (Table 7). Heavy larval in- pendentupon their initial age and size at the festationssuffered by nestlingsin Juneand July onset of larval infestation, the total numbers of raised the average number of larvae at death infesting larvae,and the season.I arbitrarily greatly,as shown by the wide rangesin Table7. designatedlarval infestationas light (1-30 lar- A single or a few infesting larvae causedde- vae/individual), moderate (31-60), and heavy bilitation when they occurredin sensitiveareas (>60). In recentlyhatched nestlings as few as of the nestling'sbody. In a nestling infested 5 larvae caused death. Therefore, lethal infes- with 3 larvae, scar tissue remaining after the tations (no matter how few infesting larvae) evacuationof single larvae from the left and were consideredheavy. Of thosenestlings that right auditory meatus caused them to close April 1985] PhilornisEctoparasitism onThrashers 287

1979 1980

160 160

140

120 12C

I00

80 140801- _•_

40 4O

•0 20

I I I I I I Feb Mor Apr May June July Feb Mar Apr May June July

1981 1982 • 160t 160 Z½ 140l- 14C •. 120F 120 o I001- I00

60 6C

Z 40 4C

20 2C

Feb Mor Apr Moy June July Feb Mot Apr Moy June July

Month Month

Fig. 4. Numbers of Philornisdeceptivus larvae on Pearly-eyedThrasher nestlings each month in the Lu- quillo Mountains,Puerto Rico (1979-1982).Vertical lines are ranges(minimum and maximumnumbers of infestinglarvae/nest), horizontal lines are means,and height of eachrectangle is + 1 SD; width of rectangles was chosen arbitrarily. completely.Even if the developing larvae did An averageof 1.47 nestlingsfledged/nesting not destroyauditory tissuesduring develop- attempt, and exceptfor 1.16 young in 1981, an ment, lossof hearing was certainsimply by the average of 1.50 nestlings fledged/nesting at- blockageof the external orifices.Similarly, in tempt eachyear. With the exceptionof a small otheryoung, larvae evacuating from orbitalsites samplein 1979, when 5 of 8 femaleslost all and even eyelidsleft eyescompletely covered broods attempted to ectoparasitism,slightly with scar tissue. Clusters of larvae found on the more than 20% (22.2, 24.0, and 22.2%) of the ventral surfacesof the throat and neck physi- nestingfemales lost all broodsattempted in one cally blockedthe passageof food and caused seasonto philornid larvae. The percentageof audible interference with normal respiration. females that succeededin fledging at least 1 Somenestlings with infesting larvae in the re- young/nestingattempt eachyear varied great- gion of the esophagusstarved to death. ly (25.0, 44.0, 12.0, and 30.6%, respectively). Hostfiedging success.--Although philomid ec- However, a much higher fledging successwas toparasitismwas high, with an averageof 95.8% observedin the two drier years(1980 and 1982). of all nests and 96.4% of all nestlings being Juvenilethrasher mortality.--To study the ef- parasitizedeach year, the average4-yr fledging fectsof philornid ectoparasitismon postfledg- successwas 53.3%(range 44.2-64.4%;Table 8). ing thrashers,I radiotagged4 individuals.The 288 w^¾•eJ. ARENDT [Auk,Vol. 102

1979 :' ;', •eo 1980 ; 80 ._>, Mortality ': '" /I I00 • I00 ...... Rainfall 90 80 70 50 ?- 60 40 n- o, 6050 . ,: ' ""'-...... "'.. 40 5{3 ?- 40 . .."•' 30 4C 30 E • 30 ' ""., 20 3(3 20 • 20 20 Z I0 ' I I I I I 1I0 I0 Dec Jan Feb Mar Apr May June July Dec Jan Feb Mar Apr May June July

Month M onth

80 1981 I00 Mortality I00, Martalily ...... Rainfall 80 6O \ , 7o 5O ? 50 80 il 1982 4O 50 50 \ .... '.... / '...... 40 40 30 46•0• "..'•..../." '; ",.. .,::' , 30 3o 20

I0 io

Dec Feb Mar Apr May June July Dec Jan Feb Mar Apr May June Jul,

Month Month

Fig. 5. Relationshipbetween mortality of Pearly-eyedThrasher young and rainfall each month in the Luquillo Mountains, Puerto Rico (1979-1982).

first juvenile fledged in March and was 1 of 7 watched it from a blind on the day that it unparasitized March nestlings. I was able to fledged. It weakly fluttered to the understory follow this bird successfullyfor 73 daysbefore but managedto hop up to safety. Both parents the transmitter failed. This thrasher had emi- fed it. I never saw it fly, nor did it move more grated about 2 km from the nest area and ap- than 10 m from the nest box tree. It died within peared healthy at the last sighting. Unfortu-. a few daysand I retrieved the transmitterabout nately,the 3 remainingfledglings were heavily 10 m from the tree. If I had not observed the parasitized before leaving the nest, and all 3 March and May fledglingsflying normallywith died after fledging. Two May nestlingsfledged their transmitter packs,I would have suspected from the samenest box as the March fledgling. that the transmitter weighed too much for the Each suffered more than 60 infesting larvae as latter 3 nestlings,inhibited normal activity, and nestlings,but all infesting larvae had evacuat- contributed to their deaths. ed by the time of fledging and both chicks weighed as much as the first radio-taggedin- DISCUSSION dividual (97 g). I watched them make short flights of 5-10 m in the canopy surrounding The reproductive cycle of Philornisdeceptivus the next box. However, one sibling died 9 days is highly adapted to that of the Pearly-eyed after fledging, while the other died 15 days af- Thrasher, Ninety-six percent of all nestling ter leaving the nest. They had remained within thrashers examined over a 4-yr period were 10 m and 30 m, respectively,of the nest box. parasitized by philornid larvae. Nestling The fourth radio-taggedthrasher left the nest thrasherswere more vulnerable to philornid in June. It had received more than 90 larvae as ectoparasitismthan adults. Although the per- a nestling, and all but 4 larvae had evacuated. centageof adult thrashersparasitized increased This nestling weighed 95 g at fledging. I to more than 90%by the end of the thrasher's April1985] PhilornisEctoparasitism onThrashers 289

TABLE7. Mean number of Philornisdeceptivus larvae However, most larvae were shown to complete infesting 81 Pearly-eyedThrasher nestlingsthat developmentwithin the middle portionsof the died early in the nestling stage(1979-1982). thrasher's nestling stage. This decreasesthe Age at Mean possibilityof killing young hostsprematurely, death number Number of Larval while increasingthe chancesof larval survival. (days) of larvae nestlings ranges Once a nestling has fledged, the chancesfor a 2 5.3 5 5-6 successfullarval evacuation and pupation are 3 9.8 6 6-16 greatly reduced. 4 18.8 16 7-52 Larvalinfestation sites.--On adult Pearly-eyed 5 21.0 14 5-46 6 23.4 12 9-58 Thrashers,all infesting larvae were found in 7 30.8 15 7-66 areasinaccessible, for the most part, to a preen- 8 28.4 13 8-63 ing bird. Almost 15%of the larvae were found in regions of the head. Most larvae, however, were found in the patagial membranes,whose fleshy surfacesmay facilitate larval entry. In- breeding season,when fly populationswere terestingly,larvae found in brachial areaswere high, numbersof larvae/adultdid not increase. always on ventral sides of the wings. Adult In contrast,the percentageof nestling thrash- philornids must avoid detection and possible ers parasitizedand numbers of infesting lar- capture by adult birds. vae/nestling were high throughout the repro- In thrashernestlings, the appearanceof phi- ductive season. Larval numbers/nestling lornid larvae was associatedwith the young's increased until late in the season, when fewer ontogeny, especiallypterylae development.The thrasherswere breeding and more nestlings general seq•aenceof larval infestationswas as were dying at a younger age from the effects follows. Until midseason (May), larvae ap- of increasedparasitism. pearedin dorsoanteriorregions of young nest- For most of the thrasher's reproductive sea- lings; the most common sites included the son, philornid larvae did not appear on nest- crown and nape, orbits, auditory meatus, lings until they were aboutone week old. Most flanges,and the spinal column from the inter- thrashernestling growth occurswithin the first scapularto the pelvic regions.Young nestlings week or so of life (Arendt 1983). It is during were more or less immobile, with heads and this period of development that nestlings are gapingmouths more accessibleto adult philor- most susceptibleto the effectsof philomid ec- nids. Older nestlings were more active and toparasitism.Therefore, it is advantageousfor readily moved about in the nest. This probably the parasiteto delay infestationuntil the host allowed for easier accessto ventral areas by is better able to withstand the effectsof para- adult philornids. The ventral tract, one of the sitism.With a larval stageof about 5 days for last major feather tractsto develop, appearedat the parasiteand a nestling period of about 21 about two weeks after hatching. At that time days for the host, successivelarval infestations larvae began to appearalong the ventrolateral are possible and were commonly observed. feather tracts from the vent to their apex on

TABLE8. Pearly-eyedThrasher fledging successin responseto heavy ectoparasitismby Philornisdeceptivus larvae.

1979 1980 1981 1982

n (%) n (%) n (%) n (%)

Females 7 14 27 38 Nestings attempted• 8 18 63 78 Nests parasitized 8 (100) 16 (88.9) 61 (96.8) 75 (96.2) Nestlingshatched 24 45 165 214 Nestlings parasitized 22 (91.7) 43 (95.6) 162 (98.2) 205 (95.8) Nestlingsfledged 12 (50.0) 29 (64.4) 73 (44.2) 125 (58.4) Nestlings fledged/nest 1.52 1.61 1.16 1.6 aAt least 1 egg hatched/nest. 290 WAYNEJ.ARENDT [Auk,Vol. 102

the upper breast.Other regionsof the nestling to heavylarval infestationsfalling to the ground used less frequently were the dorsal and ven- (two immediately falling prey to mongooses) tral basesof the somewhat"stubby" tail feath- when they fledged (jumped) from their nest ers (in early developmentjust prior to fledg- boxesshow that the 4-yr overall fledging suc- ing), the oil gland, and the sphincter muscle cessrate of 53.3% is questionablyhigh. If all aroundthe cloaca.The chin and throat regions nestlingsthat died and/or received more than occasionallywere parasitized in young nest- 30 larvae during the 1979-1982 seasonsare lings, but larvae appeared in those areasmore considered lost to ectoparasitism, the 4-yr frequentlywhen other ventral siteswere par- fledging successwould drop to 20.3% (357 of asitized. When host and parasitepopulations 448 nestlingsmay have perished). were higher (Juneand July),site specificitydis- Juvenile thrashers face keen intraspecific appearedand larvae appearedanywhere that competition after leaving the nest. Play-back spacewas still available. experiments using various thrasher vocaliza- Ectoparasitism'seffects on the host.--In birds, tions showedthat residentadults remain paired mortality causedby nest predation forms a lat- from year to year and that both sexesguard the itudinal gradient ranging from often negligi- nest box and territory throughout the year ble in the Arctic to as high as 86% in the Neo- (Arendt unpubl. data). Becausethrasher den- tropics (Ricklefs 1969). Most continental nest sities are high and adequate nest cavities are predation is causedby vertebrate predators,in- scarce(Snyder and Taapken 1977), less energy cluding a variety of ground and arboreal mam- may be expended maintaining year-round ter- mals and reptiles. With few exceptions (i.e. ritories and nest sites than would be spent Smith 1968, 1980; Bakkal 1980; Hector 1982; trying to obtain new ones prior to each repro- Winterstein and Raitt 1983), dipteran ectopar- ductive bout. Within their territories resident asiteshave not been considereda major factor adults have been observed guarding food affecting bird populations.Ricklefs (1969) stat- sources(e.g. fruits of the sierra palm, Miconia ed that "The effectsof other bird parasites,such spp., and Margravia)from other thrashersand as fly larvae... are poorly known, but it is not suchspecies as the Black-throatedBlue Warbler likely that these constitutea major mortality (Dendroica caerulescens),Scaly-naped Pigeon factor for many species.It has been shown, (Columbasquamosa), Puerto Rican Tanager (Ne- however, that botflies exert a significant influ- sospingusspeculiferus), Stripe-headed Tanager ence on the nesting successof oropend[o]lasin (Spindaliszena), and Puerto Rican Bullfinch Panama (Smith 1968)." On oceanic islands, na- (Loxigillaportoricensis). The radio-tagged juve- tive vertebrate predatorsare often absent and nile that emigrated some 2 km from its birth- nest predation may be greatly reduced. Under placewas attackedvigorously as it moved from such circumstances it is not unreasonable to as- territory to territory and when it approached sume that avian ectoparasitismmay serve as a fruiting trees guarded by resident pairs. It was regulating factor or as an additional reproduc- forced to remain in areas of few visible fruits. tive stresson insular bird populations. If interloping juveniles suffer heavy parasite In the Luquillo Mountains, prevalence of loadsas nestlings,their chancesof survival may philornid ectoparasitismwas low in adults,and be greatly reduced in such a competitive en- no adult is known to have died as a direct re- vironment. This also was suggestedby the suit of larval infestation. The average 4-yr re- postfledgingdeaths of the other three radio- turn rate was over 80% for adults of both sexes. taggedjuveniles. They died even though they Extended observations (sunrise to sunset on were still receiving additional food from both consecutivedays) did not reveal any overt be- adults. havioral changesin either sexwhen harboring Longevity in adult thrashersfurther lowers philornid larvae. However, infesting larvae the probability of successfor juvenile recruits significantlyaffected body massand the growth into the breeding population. Observationsof and development of certain long bones and marked thrashers,including an 8-yr-old (min- feathers in thrasher nestlings (Arendt 1983). imum) female, yielded an 80-90% yearly return Survival of young Pearly-eyed Thrashers, rate of nesting adults and showed that few both as nestlingsand as postfledgingjuveniles, young birds enter the breeding population.Of was reducedby philornid ectoparasitism.Thir- 452 young that fledgedbetween 1979and 1982 ty-two observationsof nestlingswith moderate from naturally parasitizednests and from nests April 1985] PhilornisEctoparasitism onThrashers 291 in which larvae were removedfrom nestlings, lingsof the Pearly-eyedThrasher (Margarops fus- only 7 individuals (ca. 2%) have subsequently catus)in the Luquillo Mountains, Puerto Rico. been recorded replacing older nesters in my Unpublished M.A. thesis. Columbia, Univ. Mis- souri. sampled population. The longevity and sus- tained productivity of adults,while hindering BAKKAL,S.N. 1980. O gibeli ptentsovvorobinnykh ptits ot paraziticheskikhmykh. Vestn. Leningr. juvenile recruitment, is helping to maintain Univ. Biol. 9: 106-108. thrasher numbers in the face of high juvenile BOND, J. 1979. Birds of the West Indies, fourth ed. mortality. London, Collins. Snyder and Taapken (1977) summarizedthe CONOVER,W.J. 1971. Practicalnonparametric sta- status of the Pearly-eyed Thrasher in Puerto tistics.New York, Wiley. Rico.Thrashers were unknown in the Luquillo DODGE, H. R. 1955. New muscid flies from Florida Mountains before the 1950's. Their numbers and the WestIndies (Diptera: Muscidae). Fla. Ent. greatly increasedfor the first 20 yr or so and 38: 147-151. then beganto level off, possiblyin responseto 1963. A new Philorniswith coprophagous habitat saturation. Thrashers have remained larva,and somerelated species (Diptera: Musci- dae). J. Kans. Entomol. Soc. 36: 239-247. abundant in recentyears and have shown signs 1968. Some new and little-known species of adjustingto philornid ectoparasitism.While of Philornis(Diptera: Muscidae).J. Kans. Ento- lowland populationsbreed mainly from April tool. Soc. 41: 155-164. to July, a period of heavy rainfall and increased ß & t. H. G. AITKEN. 1968. Philornis flies from fruit and insect crops, highland populations Trinidad. J. Kans. Entomol. Soc. 41: 134-154. have pushed back the initiation of the repro- EWEL,J. J., & J. L. WHITMORE.1973. The ecological ductivecycle, possibly to increasefledging suc- life zones of Puerto Rico and the U.S. Virgin cessin the seasonalabsence of heavy philornid Islands.USDA Forest Service Research Paper ITF- 18. ectoparasitism.In the sampledrain-forest pop- GILBERT,P.A. 1919. A dipterousparasite of nestling ulation older femalesbegan nesting in late No- birds. Emu 19: 48-49. vember and early December,and many were HECTOR,D.P. 1982. Botfly (Diptera, Muscidae)par- able to lay 3 and 4 clutches/year. Becauseof asitismof nestlings of Aplomado Falcons.Con- this, one might expect this prolific passerine dor 84: 443-444. always to remain in abundancein the Luquillo HINDWOOD,K.A. 1930. A sub-cutaneousavian par- Mountains, despite continuing ill effectsfrom asite. Emu 30: 131-137. philornid ectoparasitism. HOLDRIDGE,L.R. 1947. Determinationof world plant formationsfrom simpleclimatic data. Science 105: 367-368. ACKNOWLEDGMENTS 1967. Life zoneecology. San Jose, Costa Rica, The field assistanceand continuedencouragement Tropical ScienceCenter. from my wife, Angela, is gratefully acknowledged.I LEDGER,J. A. 1969. Notes on the tropicalnest fly. Bokmakerie 21: 28-30. sincerelythank my major advisor, Dr. John Faaborg, and my thesis committee members, Drs. Susan B. MARGOLIS,L., G. W. ESCH,J. C. HOLMES,A.M. KURIS, Chaplin, Terry Finger, and Leslie Uhazy, for their & G. A. SCHAD.1982. The useof ecologicalterms continued support and encouragementthroughout in parasitology(report of an ad hoc committee my studiesat the University of Missouri-Columbia. of the Amer. Soc.of Parasitologists).J. Parasitol. 68: 131-133. I thank S. Chaplin, J. Faaborg,T. Finger, P. Price, N. G. Smith, L. Uhazy, and two anonymousreviewers NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRA- for helpful commentson the manuscript.L. Uhazy TION. 1970-1981. Climatologicaldata of Puerto offered much assistanceand insight into host-para- Rico and the Virgin Islands. Asheville, North siterelations. Mrs. MarciaKelly-Nelson kindly helped Carolina, NOAA Environ. Data and Information with the illustrations. The study was supported by Serv. Natl. Climatic Center. the U.S. Forest Service. NIELSEN,J. C. 1911. Mydaeaanomala Jaenn., a parasite of South-American birds. Medd. Nat. For. Kobenhaven 63: 195-208. LITERATURE CITED ODUM,H. T., & R. F. PIGEON(Eds.). 1970. A tropical ALDRICH,J. M. 1923. The genus Philornis--abird- rain forest:a studyof irradiationand ecologyat infesting group of Anthomyiidae. Ann. Ento- E1 Verde, Puerto Rico. Washington,D.C., U.S. tool. Soc. Amer. 16: 304-309. Atomic Energy Commission. ARENDT,W.J. 1983. The effectsof dipteranectopar- OWEN,D. F. 1954. Protocalliphorain birdsß nests. asitismon the growth and developmentof nest- Brit. Birds 47: 236-243. 292 WAYNEJ. ARENDT [Auk,Vol. 102

PLATH,O.E. 1919a. Parasitismon nestling birds by $NEDECOR,G. W., & W. G. COCHRAN. 1980. Statistical fly larvae. Condor 21: 30-38. methods. Amesß Iowa State Univ. Press. 1919b. The prevalence of Phormia azurea SNYDER,N. F.R. 1977. Puerto Rican Parrots and nest Fallen (larva parasiticon nestling birds) in the site scarcity. Pp. 47-53 in Endangered birds, Puget Sound region and data on two unde- management techniques for preserving threat- scribed flies of similar habit. Ann. Entomol. Soc. ened species(S. A. TempleßEd.). MadisonßUniv. Amer. 12: 373-381. Wisconsin Press. PONT,A.C. 1972. A catalogueof the Diptera of the ß& J. D. TAAP•CEN.1977. Puerto Rican Parrots Americas south of the United States: Family and nestpredation by Pearly-eyedThrashers. Pp. Muscidae, Genus Neomusca.Pp. 55-57. Museo 113-120 in Endangered birdsßmanagement tech- Zoologia Universidade de Sio Paulo, Brazil. niques for preserving threatened species(S. A. 1974. A revision of the Genus Passeromyia TempleßEd.). Madison, Univ. Wisconsin Press. Rodhain and Villeneuve (Diptera: Muscidae). $OKAL,R. R., & F. J. ROHLF. 1969. Biometry. The Bull. Brit. Mus. Nat. Hist. (Entomol.) 30: 339-372. principlesand practicesof statisticsin biological RICKLEESßR. E. 1969. An analysisof nestingmortal- research. San Franciscoß W. H. Freeman and Co. ity in birds. SmithsonianContrib. Biol. 9. WADSWORTH,F.H. 1948. The climate of the Luqui- ROLLE,F.J. 1965a. An observationof heavy preda- 11oMountains and its significanceto the people tion by Pearly-eyed Thrasher. Wilson Bull. 77: of Puerto Rico. Caribb. For. 9: 321-344. 296-297. WINTERSTEIN,$. R., & R. J. RAITT. 1983. Nestling --. 1965b. Destruction of Red-legged Thrush growth and developmentand the breeding bi- nest by Pearly-eyedThrasher. Auk 82: 643. ology of the BeecheyJay. Wilson Bull. 95: 256- SMITH,N. G. 1968. The advantageof being parasit- 268. ized. Nature 219: 690-694. WOLCOTT,G.N. 1948. The insects of Puerto Rico. J. 1980. Some evolutionaryßecological, and Agr. Univ. Puerto Rico 22. behavioral correlatesof communal nesting by ZUMPT,F. 1965. Myiasis in man and animals in the birds with wasps or bees. Acta 17th Congr. In- Old World. Londonß Butterworth. tern. Ornithol. 2: 1199-1205.

JUST PUBLISHED

ORNITHOLOGICAL MONOGRAPHS NO. 35

EcogeographicalVariation in Size and Proportionsof Song Sparrows(Melospiza meloclia),by John W. Aldrich. x + 134 pp., 9 tablesß34 text figuresß5 appendices.1984. Price $10.50 ($8.50) to A.O.U. members).All orders must be prepaid and include $1.00 handling charge. Order from Frank R. MooreßAssistant to the Treasurer,A.O.U., Department of BiologyßUniversity of Southern Mississippi,Southern Station Box 5018, Hattiesburg,Mississippi 39046 USA.