The use of supplementary foods by Australian magpies Gymnorhina tibicen: Implications for wildlife feeding in suburban environments.

Transcription

1 The use of supplementry foods by Austrlin mgpies Gymnorhin tibicen: Implictions for wildlife feeding in suburbn environments. Author O'Lery, Rebecc, Jones, Drryl Published 2006 Journl Title Austrl Ecology DOI Copyright Sttement 2006 Wiley-Blckwell Publishing. This is the uthor-mnuscript version of the pper. Reproduced in ccordnce with the copyright policy of the publisher. The definitive version is vilble t Downloded from Link to published version Griffith Reserch Online

2 1 The use of supplementry foods by Austrlin mgpies Gymnorhin tibicen: Implictions for wildlife feeding in suburbn environments REBECCA O LEARY AND DARRYL N. JONES Suburbn Wildlife Reserch Group, Austrlin School of Environmentl Studies, Griffith University, Nthn, Queenslnd 4111, Austrli (Emil: D.Jones@griffith.edu.u) Dte: 29 April 2005 Word Count: 6198 Key words: Mgpie, Gymnorhin tibicen, wildlife feeding, supplementry feeding, urbn ecology Correspondence: Dr Drryl N. Jones, Austrlin School of Environmentl Studies, Griffith University, Nthn, Queenslnd 4111, Austrli

3 2 Abstrct The provision of foods to wild birds is n extremely common prctice mong householders through out the Western world. Nonetheless, concerns over potentil impcts of the prctice re currently being rised, including the possibility tht some species my become relint on humn-provided food. We compred the forging nd breeding ecology of pirs of fed nd unfed Austrlin mgpies Gymnorhin tibicen living suburbn environments in Brisbne, Queenslnd, Austrli. Detiled behviourl observtions of both forging nd chick provisioning were mde for mles nd femles of both groups throughout the breeding seson. Nturl foods dominted the diets of both fed nd unfed mgpies, mking up 76% nd 92% of ll items consumed respectively. During the morning, fed mgpies obtined fewer food items during ground forging thn did unfed birds, pprently becuse they visited suburbn feeding sttions more often during the erly prt of the dy. At other times, the mount of food items obtined during forging ws similr for both fed nd unfed birds. Mgpies utilising suburbn feeding sttions strted ll breeding ctivities significntly erlier thn unfed mgpies, except during the fledgling phse. Both fed nd unfed mgpie prents provisioned their chicks predominntly with nturl food. Mgpies were not relint or dependent on supplementry food provided by wildlife feeders t ny time during the breeding seson. Although mny mgpies did utilise suburbn feeding sttions extensively, they continued to forge for nd provision their chicks with nturl food.

4 3 INTRODUCTION Attrcting wildlife, especilly birds, to suburbn houseyrds by the provision of foods ( wildlife feeding ) is n extremely common prctice throughout the Western world. Numerous studies indicte tht 25% to 75% of householders surveyed in suburbn res regulrly provided food specificlly for visiting wildlife, with the highest rtes occurring during the winter in temperte regions (Cowie & Hinsley 1987, 1988; Brittinghm & Temple 1988). Approximtely million Americns feed birds during the winter (Brittinghm & Temple 1992; Wilson 2001), distributing round 450 million kilogrms of seed every yer (Brittinghm 1991). Throughout Europe, bird feeding is lso ubiquitous, especilly in the United Kingdom (Cnnon 1999, 2000). In Austrli, wildlife feeding is similrly populr with recent studies finding prticiption rtes of % of households in severl lrge cities (Woodll 1995; Thoms 2000; McLees 2001; Rollinson et l. 2003). There re, however, some importnt differences in the prctice of bird feeding in Austrli compred to the Northern Hemisphere. Austrlin bird feeders, for exmple, ttrct reltively few smller grnivorous psserines (Jones 2003); more likely re lrger insectivorous or crnivorous species such s the Austrlin mgpie Gymnorhin tibicen nd the lughing kookburr Dcelo noveguinee (McLees 2001; Rollinson et l. 2003). These species re typiclly ttrcted by mets, lthough bred nd food scrps re lso widely used (Rollinson et l. 2003). Furthermore, Austrlin wildlife feeders tend to provide food yer-round (Thoms 2000), rther thn primrily during winter, s is the cse the Northern Hemisphere (Cowie & Hinsley 1988; Brittinghm 1991). The feeding of wildlife is openly promoted throughout the Northern Hemisphere s being of positive benefit to the recipient nimls, s mens of incresing community interest in nture, nd s hving n importnt role in conserving wildlife, especilly in urbn res (e.g. see Cnnon 1999). Prominent orgnistions such s the British Trust for Ornithology (BTO), the Royl Society for the Protection of Birds (RSPB) nd Cornell Lbortory of Ornithology (CLO), mong others, ctively encourge the feeding of birds, nd provide detils on how to ttrct birds nd minimise ssocited problems (Dunn & Tessgli-Hynes 1990; Grubb & Cimprich 1990; RSPB 2002; BTO 2003).

5 4 Tht the benefits of feeding, for both the people involved nd the birds themselves, outweigh ny possible negtive effects, is widely ssumed, lthough remrkbly little specific reserch hs been conducted on these issues nywhere in the world (Brittinghm 1991). Severl recent ssessments (Brittinghm 1991; Anderson et l. 1997; Green & Higginbottom 2000; Orms 2002) of the possible negtive nd positive impcts of wildlife feeding in generl were significntly constrined by the derth of relible informtion. Nonetheless, numerous studies hve demonstrted benefits of supplementry feeding in enhncing popultions of severl endngered species (e.g. Wilbur et l. 1974), the winter survivl rte nd helth of certin nonmigrtory species (Grubb & Cimprich 1990). Other benefits include enhnced physicl condition nd reproduction rtes, greter resilience to disese, prsites nd predtion (Brittinghm 1991). Despite the lmost universl ppel nd promotion of wildlife feeding, the potentil for negtive impcts re lso beginning to be ddressed (Brittinghm 1991, 1992; Green & Higginbottom 2000; Orms 2002). Concern over, for exmple, the spred of disese, dietry deficiencies, incresing bundnce of certin species, nd potentilly disdvntgeous behviourl chnges, hs prompted some to question whether feeding should be regrded so fvourbly (Deis 1982; Anderson et l. 1997). This ttitude is especilly pronounced in Austrli. Mny stte-bsed wildlife gencies, hving promoted feeding in the pst, now ctively discourge the prctice of wildlife feeding (Skir & Smith 1991; Hunter 2001), lthough the context of interest is usully tht of reserves nd ntionl prks. Concern over the impcts of feeding wildlife in suburbn settings is rpidly growing, however. Cnnon (1999: 288), for exmple, quoted n Austrlin corespondent s stting: Generly, the more conservtionminded nd knowledgeble individuls in Austrli do not feed in their grden. Nonetheless, wildlife feeding, in Austrli, s elsewhere, remins very common prctice (Rollinson et l. 2003; Howrd & Jones 2004). Prominent mong the numerous potentil impcts of wildlife feeding is tht of dependency (see Brittinghm 1991; Anderson et l. 1997; Green & Higginbottom 2002; Orms 2002). This suggestion proposes tht nimls tht become used to receiving predictble nd regulr supply of food my become entirely relint upon this food source with possible short-term nd long-term implictions (Brittinghm & Temple

6 5 1992). In the short-term, birds my be temporrily ffected when feeding sttion is moved or discontinued. Over longer periods these birds my be unble to fend for themselves, potentilly resulting in strvtion, s well s ssocited issues such s diminished bility to find pproprite food, or relince on unsuitble or nutritionlly poor foods (Anderson et l. 1997). The impct of these dependency-relted ctivities on chick welfre nd survivl is of prticulr concern (Brittinghm 1991). The potentil for dependency is cknowledged by both proponents of wildlife feeding (see e.g. Dunn & Tessgli-Hynes 1990; Grubb & Cimprich 1990; BTO 2003) nd opponents (Green & Higginbottom 2002; Orms 2002), yet unequivocl findings remin virtully nonexistent (but see Brittinghm & Temple 1988, 1992). Generlly, the continuing debte is lrgely bsed on inference, opinion nd necdote (Deis 1982; Nttrss 2001; Jones & Howrd 2001). To investigte this issue, it is impertive tht birds being fed by humns re compred with birds in the sme environment not being fed. It would be prticulrly instructive to investigte the food provisioning behviour of fed birds during the high food-demnd period prior to the fledging of chicks. Do such birds continue to obtin nturl food sources or do they rely entirely on esily obtined supplementry foods? The Austrlin mgpie is n idel cndidte for such study. Not only is it bundnt in suburbn environments throughout Austrli (Jones 2002), it is mong the most common species to utilise feeding sttions (McLees 2001; Rollinson et l. 2003). Suburbn wildlife feeders, while common, re often distributed ptchily throughout the suburbs, providing nturl experiment in which to investigte the rection of mgpies to the presence of different tretments of supplementry foods. In the present study we compred the forging nd chick provisioning behviour of mgpies resident in the suburbs of Brisbne, Austrli. Two groups of mgpies were studied: those using feeding sttions ( fed ) nd those tht did not ( unfed ). As l pirs were ctively engged in breeding, we predicted tht fed mgpies would obtin significntly less nturl food for themselves nd especilly for their chicks thn would unfed birds. For the purposes of this study, we regrded mgpies s being dependnt upon supplementry foods if they obtined most of their food from feeding sttions even during low food-demnd times.

7 6 METHODS Study re nd nimls The study ws conducted during the 2001 mgpie s breeding seson (June to November) in suburbn res of Brisbne ( E, S), in south-est Queenslnd, Austrli. The climte is typiclly sub-tropicl without pronounced sesonlity. During the study the rinfll recorded June-August (minimum 8.4 mm, mximum 27.6 mm) ws less thn the long-term verge (57 mm), September nd October were similr (34 mm nd 92 mm), nd November (192 mm) received higher thn the verge rinfll (98 mm) (Queenslnd Bureu of Meteorology 2003). The suburbn environment consisted lrgely of mtrix of lots with single or two-storey privte houses, interspersed with smller res of ntive woodlnd nd prklnd. Approximtely 45-78% of the re consisted of well-mnged lwns (the primrily forging substrte of the species) in houseyrds, sports fields, school grounds nd prks (Jones 2002). We selected twenty pirs of mgpies with territories comprised predominntly of prks nd sports fields; pirs with territories bsed minly in suburbn streets nd house yrds were voided due to the difficulties of observing birds behind fences nd other structures. All birds hd previously been individully colour-ringed (Rollinson 2003) nd sexes were esily distinguished through plumge differences (Jones 2002). Pirs were ctegorised into one of two groups: those regulrly utilising suburbn feeding sttions ( fed ) nd those tht were not fed by humns ( unfed ). Pirs were confirmed s belonging to either group through informtion obtined by observtion nd by questionnire survey of ll households ner or djcent to known focl mgpie territories (see Rollinson et l. 2003). Birds utilising one or more feeding sttions were designted s fed (totl of 12 pirs). Unfed (totl of eight pirs) birds lso visited some houseyrd lwns during their norml forging ctivities but did not encounter or use feeding sttions. Some unfed mgpies did consume food wste left unintentionlly in prks nd picnic res but such forging ws irregulr nd opportunistic. Informtion obtined seprtely on the prctices of the wildlife feeders operting in the study re (O Lery 2002; Rolinson et l. 2003) indicted tht fed pirs hd

8 7 ccess to one to four feeding sttions situted in bckyrds, vrndhs or ptios offering mets nd food scrps (minly beef mince, susge, stek, hm, bcon, bred nd cheese. Approximtely 78% of the supplementry food provided consisted of met with the reminder including t lest some bred. Wildlife feeders provided supplementry food on regulr bsis, vrying between one to ten times per dy (men 1.9 times per dy)(o Lery 2002). Behviourl observtions Ech mgpie within ll pirs ws observed t lest weekly for totl of 6-18 hours throughout the breeding seson using binoculrs (8x40 mm) nd zoom spotting scope (45x56 mm to 20x25 mm). Smpling ws divided into three sessions per dy: morning (dwn to 10:00); middy (10:00 to 14:00); nd fternoon (14:00 to dusk). Ech observtion of n individul lsted 15 minutes, during which ll forging or chickfeeding ctions were recorded per minute. Observtion visits to pirs were rndomised for time of dy nd loction. Forging behviour Ground forging ctivities ws only recorded when the birds were on lwns or short grss, the preferred forging substrte for mgpies in the suburbn environment (O Lery & Jones 2002). During the observtion sesion, forging events were recorded for the minute in which they occurred; food items nd mount were lso identified where possible nd counted. For different food types (nturl or supplementry food), the forging mount per minute ws scertined by counting the number of items consumed for prticulr type of food item within minute. Hed tossing nd swllowing (Veltmn & Hickson 1989) distinguish the successful cpture of prey, nd most food items could be determined before being swllowed. Some smll food items could not be identified nd were ctegorised s unknown. Forging events only included food obtined from the lwns nd not from feeding sttions.

9 8 Chick provisioning All pirs (both fed nd unfed) were ssessed weekly for breeding ctivity from June to November 2001 nd chick-provisioning events were recorded during ech session with the sex of the provisioner, mount of food nd food source (nturl or supplementry food) being recorded per minute. The breeding seson ws divided into the following phses: nest building; incubtion; htchlings (chicks not visible from the ground but being fed); medium chicks (pproximtely 7-14 dys since htching dte); lrge chicks (chicks frequently visible from ground); nd fledglings (chicks permnently wy from nest for mximum of seven dys). The strting dte of ech breeding phse for ech pir ws determined s the dte on which either the mle or femle ws first observed engged in the prticulr phse. For ech pir, 15 minute observtionl periods of both mle nd femle were combined per dy. During the chick phses, both prents were usully observed forging for food for the chicks. Food intended for the chicks ws held in the prent s bek nd once sufficient food hd been obtined the prent bird returned to the nest to feed the chicks. The item nd mount of ech food item collected for the chicks ws recorded while the prent bird ws forging prior to the prent provisioning the chicks. If the prent bird ws not observed forging before provisioning the chicks it ws difficult to determine the quntity nd type of food being provisioned. When this occurred, the mount nd type of food ws estimted if the bill ws seen before the food ws provisioned to the chicks; otherwise it ws recorded s unknown. Al food provisioned to the chicks ws recorded, including food obtined from feeding sttions. Sttisticl nlysis Anlyses were performed using the SAS sttisticl nlysis pckge (SAS System for Windows, SAS Institute, Cry, NC). Dt lcking normlity were logrithmiclly trnsformed, due to the fctor effects being multiplictive rther thn dditive (Zr 1999). Mens nd stndrd errors re, however, presented s untrnsformed. The level of significnce used ws 5% unless otherwise stted. The test of proportions (Zr 1999) ws used to compre the proportion of six ctegories of foods consumed by the dult fed versus unfed mgpies. These ctegories consisted on four nturl food types: (1) insect lrve/worms (LrvWm); (2) flying

10 9 insects (including ll Coleopter, Lepidopter nd Hymenopter other thn non-flying nts)(flyins); (3) ll other insects (minly nts, Blttodee, Phsmodee, Arnee, Orthopter nd Diplopod)(OthIns); nd (4) other nturl (ll other living items including plnt mteril)(othnt); nd two humn-derived food types: (5) met items (Met); nd (6) non-met (primrily non-met items such s bred)(nmet). For ech ctegory, test of proportions ws used to ssess whether there ws n equl chnce of fed nd unfed mgpies consuming prticulr food ctegory. Similrly, test of proportions ws performed on the proportion of nturl nd supplementry food provisioned by fed nd unfed mgpies to the chicks. Breeding dtes were converted into numerls, strting t 1 for the first dy of June nd ending t 153 for 31 st October. Student s t-test ws used to test between fed nd unfed mgpies for ech breeding event dte. A fctoril nlysis of vrince ws performed on the mount of food forged by the dults nd the mount of food provisioned to the chicks. Outliers identified during explortory dt nlysis were exmined to ssess whether or not they could be explined. If n outlier could not be explined it ws removed to void possible violtions of normlity nd vrince equlity ssumptions (Zr 1999). As the dt (forging mount nd mount of food provisioned) hd unequl smple sizes, the nlysis of vrince ws crried out using the generl liner models lest squres method (Zr 1999). Anlyses of the dt were conducted using split-plot model, with mgpie pirs being split into mles nd femles. A t-test using the lest-squres mens ws used to test for differences between the mens when n effect ws identified s significnt in the nlysis of vrince. RESULTS Forging Both fed nd unfed mgpies forged on wide vriety of foods in suburbn res of Brisbne. Nturl food clerly dominted the diets of both groups of mgpies; pproximtely 76% nd 92% of totl items consumed by fed nd unfed birds respectively were nturlly occurring invertebrtes. There were no differences in the

11 10 proportions of ech of nturl food types consumed by fed nd unfed mgpies (Fig. 1). Fed mgpies, however, consumed significntly lrger proportion of both met (test of proportions: z = 40.06, d.f. = 1, P < 0.001) nd non-met items (z = 6.85, d.f. = 1, P < 0.01) thn unfed birds. FIG 1 The men number of nturl food items ingested per minute by forging dults ws compred for feedtype (fed or unfed), sex nd time of dy (Tble 1, Fig. 2). Fed mgpies consumed similr numbers of items while forging throughout the dy. In contrst, unfed mgpies fed t significntly higher rte during the morning, with the feeding rte declining to similr rte of fed birds by middy (Tble 1, Fig.2). Thus, n obvious effect of mgpies utilising feeding sttions ws to substntilly reduce their rte of forging during the morning; both fed nd unfed birds forged t similr rtes t ll other times. TABLE 1 FIG 2 Reproductive ctivities Fed mgpies initited ech phse of breeding significntly erlier thn unfed mgpies, except during the fledgling phse (Fig. 3). The initition of nest building (the first discernible event denoting the commencement of breeding) for fed mgpies ws on verge dys erlier thn tht of unfed birds (Student s t-test: t = 2.47, d.f. = 16, P = 0.031), clutch initition ws dys erlier (t = 2.39, d.f. = 16, P = 0.030) nd dte of first htch ws dys erlier (t = 2.45, d.f. = 15, P = 0.034). Although fed birds first dy of fledging ws 8.17 dys erlier thn unfed birds, this ws not significnt. FIG 3 A cler mjority of the types of food provided to chicks by both fed nd unfed mgpies ws nturl, lthough the proportions were significntly different: 0.68 of the items provided by fed mgpies were nturl compred to 0.97 for unfed mgpies (test of proportions: z = 49.6, d.f. = 1, P = 0.001). This is further emphsised by compring the totl mounts of nturl versus supplementry food provided to chicks by their fed nd unfed prents (Fig. 4).

12 11 FIG 4 The totl mount of food provisioned to chicks ws lso compred for feedtype (fed or unfed), sex of provisioner nd food source (nturl or supplementry) (Tble 2). There ws significnt difference mong pirs within feedtype, lthough the overll feedtype effect on the men totl mount food provisioned to the chicks ws not significnt. The interction between feedtype nd food source ws significnt (Tble 2), however, with both fed nd unfed mgpies clerly preferring to provision their chicks with nturl food (Fig. 4). TABLE 2 DISCUSSION The possibility tht nimls utilising suburbn wildlife feeding sttions my become dependent on this source of supplementry food remins prominent concern for both opponents nd dvoctes of this extremely populr prctice (Deis 1982; Anderson et l. 1997). The present study ws designed to ssess one spect of this complex behviourl phenomenon by investigting the extent to which breeding Austrlin mgpies utilised supplementry rther thn nturlly occurring foods when forging for themselves nd when provisioning their chicks. We predicted tht mgpies with continul ccess to feeding sttions would feed from nd provision chicks primrily with supplementry foods rther thn nturl foods. In contrst, unfed mgpies, without this redy source of food would obtin ll food items, for themselves nd their chicks, vi ground forging. Forging Contrry to our predictions, we found tht the diets of suburbn mgpies, both fed nd unfed, were dominted by nturl foods with fed mgpies consuming similr proportion of ech nturl food ctegory s unfed mgpies. An erlier study from the sme loction (O Lery & Jones 2002), conducted during prolonged drought when ground-derived invertebrtes were presumbly difficult to obtin, found similr result, with pproximtely 82% of the diet of mgpies being nturl rther supplementry. This strongly suggests tht despite there being n bundnce of supplementry food

13 12 vilble (Rollinson et l. 2003), fed mgpies did not rely on this food for their dily energy requirements. Other ground-forging birds in suburbn environments, such s the common strling Sturnus vulgris, lso forge on brod rnge of nturl foods, even where feeding sttions were redily vilble (Mennechez & Clergeu 2001). The observed diet of unfed mgpies ws different to tht of fed mgpies with respect to incidentl supplementry food obtined while forging on the ground wy from feeding sttions. Unfed mgpies were rrely observed consuming food scrps, wheres 24% of the diet of fed mgpies consisted of food wstes nd other discrded items. O Lery & Jones (2002) lso found tht mgpies regulrly obtined food scrps, with pproximtely 18% of ll items consumed consisting of this food. This difference between fed nd unfed birds my hve been due to greter vilbility of incidentl foods in the territories of fed mgpies, lthough this ws not investigted t the time. Certinly, ll of the territories surveyed in this study contined picnic grounds or public prks where humn foods were often discrded. In generl ll birds need to replenish their energy levels erly in the morning fter the enforced fst of night roosting; for species relint on protein-rich foods such s insects feeding rtes re typiclly fstest during the erly morning (Sutherlnd 1996). Observtions of forging ptterns (Fig. 2) provided some evidence tht fed mgpies were ble to stisfy their erly-morning nutritionl requirements more rpidly thn unfed birds. During the morning, fed mgpies consumed significntly less food per minute thn unfed mgpies, suggesting tht fed mgpies obtined most of their morning food from feeding sttions. (This suggestion is corroborted by n independent study of visits by mgpies to feeding sttions (Rollinson 2003).) At ll other times of the dy, fed mgpies consumed similr mount of food per minute s unfed mgpies. Reproductive ctivities Timing of breeding Mgpies utilising feeding sttions commenced nest building lmost two weeks erlier thn unfed birds resident in the sme environment. Similrly, in seprte study of mgpies breeding in the sme region, Rollinson & Jones (2002) found suburbn mgpies breeding on verge dys erlier thn birds in nerby rurl res,

14 13 difference ttributed primrily to the greter food vilbility of the suburbn environment. In ddition to the regulr supply of supplementry food t the mny feeding sttions (Rollinson et l. 2003), suburbn lwns typiclly contin greter bundnce of invertebrtes thn grsslnds in rurl res nd bushlnd (Rollinson 2003). An dvncement in lying dte by two to 25 dys (men 8.84 dys) is consistent finding of experimentl supplementry feeding studies of wide vriety of species (Boutin 1990; Soler & Soler 1996), results which clerly demonstrte the fundmentl influence of food vilbility on reproduction in birds (Mrtin 1987). These experimentl results re consistent with numerous studies (Dhont et l. 1984; Bol & Mnnn 1999) compring reproductive prmeters of urbn versus rurl or wildlife species; urbn popultions invribly commenced breeding erlier, finding ttributed to the greter vilbility of supplementry foods. Although fed mgpies strted their breeding much erlier thn unfed birds, this difference ws sttisticlly insignificnt by the fledgling phse. This result my hve resulted from smll smple sizes in the fledgling phse, s mny pirs of both fed (bout 40%) nd unfed (bout 50%) mgpies did not successfully fledge their brood. It is lso possible tht dietry difference between fed nd unfed mgpies my hve cused the chicks of fed mgpie to grow more slowly nd therefore fledge t similr dte s tht of unfed mgpie chicks (see below). Chick provisioning Although regulr nd bundnt supply of supplementry food (minly intentionl) ws vilble, fed mgpies continued to provision their chicks primrily with nturl food. Unexpectedly, the proportion of nturl food provisioned to the chicks ws similr for both fed nd unfed mgpies. Furthermore, the rte t which nturl food ws provisioned per minute ws significntly greter thn supplementry food for both groups. These results were the opposite to tht predicted; given the ese with which food could be gthered from feeding sttions, nd its generlly high-protein levels, we expected tht fed mgpies would feed their chicks minly with supplementry food. Those opposed to wildlife feeding often clim tht birds utilising supplementl food bndon nturl forging, nd come to rely entirely on food provided by humns for themselves nd their chicks, with rnge of negtive implictions likely (Anderson et

15 14 l. 1997; Orms 2002). Our results demonstrte tht, for dult Austrlin mgpies t lest, these concerns re unfounded. Nonetheless, we re lso wre of numerous situtions where lrge numbers of juvenile un-pired mgpies pper to remin close to prticulr suburbn feeding sttions throughout the dy nd for extended periods of time (D. Jones, unpublished dt, 2003). Although the forging behviour of these birds hs yet to be investigted, some degree of relince upon nthropogenic foods could be expected. Other studies on the diets of chicks rised in suburbn res were ll conducted in the Northern Hemisphere. In southern Wles, Cowie & Hinsley (1988) found tht pproximtely 15% of the diets of blue tits Prus ceruleus nd gret tits Prus mjor chicks consisted of penuts, bred nd ft obtined from suburbn feeding sttions. A similr result ws found mong common strlings, with 11-20% of items consumed by nestlings consisting of supplementry foods (fruits nd humn refuse) in Frnce (Mennechez & Clergeu 2001), nd 35% in Wrsw, Russi (Gromdzk & Lunik 1978). In the present study, bout one third of the food provided to the chicks of fed mgpies consisted of supplementry food, compred to only 2% for unfed birds. In none of these studies did the proportion of supplementl food in the diet of chicks exceed more thn bout 40%. In their investigtion of suburbn strlings in Frnce, Mennechez & Clergeu (2001) found tht the proportion of supplementry food provided to chicks peked during the lter stges of chick development, when nestling feeding demnds were t mximum. These uthors suggest tht strling prents ctively voided supplementry foods for s long s possible, scrificing food qulity for more redily obtined foods only when stisfying chick hunger ws most difficult. The supplementry foods vilble to these birds consisted minly of humn food wstes such s bred nd met scrps, food generlly regrded s being of low energetic content or poor nutritionl vlue (Pierotti & Annett 2001). The humn-provided foods used by suburbn Austrlin mgpies, however, were more often high-qulity, protein-rich mets, lthough poorer qulity foods were lso provided (Rollinson et l. 2003). Given tht virtully nothing is currently known of the nutritionl qulity of commonly used nthropogenic foods we strongly dvocte this issue s priority issue for future reserch.

16 15 This observtionl study found tht the suburbn mgpies studied were not dependent on supplementry food (both intentionl nd incidentl) s both fed nd unfed mgpies forged minly on wide vriety of nturl items nd provisioned their chicks primrily with nturl foods. Mgpies my be regrded s opportunists in the suburbn environment s they utilise wide vriety of both nturl nd supplementry foods. However, further study is needed to determine whether the commonest types of supplementry foods tht the mgpies re consuming hve detrimentl effect on dults, nd in prticulr, the helth nd survivl of chicks. Given the populrity of wildlife feeding throughout the world, nd the mny benefits to humn prticipnts, it is unlikely tht ssertions of potentil negtive impcts re likely to curb public enthusism for the prctice, especilly when little fctul informtion exists. Nonetheless, the considerble literture on experimentl supplementry feeding studies (Boutin 1990) indictes clerly tht the ddition of food resources lmost lwys influences mny prmeters of the popultion ecology of the species involved. Although suburbn wildlife feeding represents supplementry feeding experiment on truly globl scle, it hs received remrkbly little ttention by reserchers. As interest in the opportunities for ecologicl reserch in urbn res grows (Collins et l. 2000), we suggest tht further studies into the phenomenon of suburbn wildlife feeding my be prticulrly importnt. ACKNOWLEDGEMENTS We thnk Robyn Chithm for the use of vehicle nd Jnet Chseling nd Rchel King for their help with sttisticl nlyses. Comments by John White (Dekin University) nd John Wright (University of Trondheim) gretly improved n erlier version of the mnuscript. This study ws conducted with the pprovl of the Griffith University Animl Ethics Committee nd the Queenslnd Prks nd Wildlife Service. REFERENCES Anderson, T., Bitossi, K., Berhmmer, L., Chill, D., Denning, D., Lnhm, S., Low, C., Muller, A. & Pinwell, S. (1997) Wildlife feeding: Is it problem? School of

17 16 Nturl nd Rurl Systems Mngement, University of Queenslnd, Gtton College, Queenslnd, Austrli. Bol, C.W. & Mnnn, R.W. (1999) Comprtive breeding ecology of Cooper's hwks in urbn nd exurbn res of southestern Arizon. Journl of Wildlife Mngement 63, Boutin, S. (1990) Food supplementtion experiments with terrestril vertebrtes: Ptterns, problems nd the future. Cndin Journl of Zoology 68, Brittinghm, M.C. (1991). Effect of winter feeding on wild birds. Wildlife Conservtion in Metropolitn Environments (eds L.W. Adms & D.L. Leedy), pp Ntionl Institute for Urbn Wildlife, Columbi. Brittinghm, M.C. & Temple, S.A. (1988) Avin disese nd winter bird feeding. The Pssenger Pigeon, 50, Brittinghm, M.C. & Temple, S.A. (1992) Does winter bird feeding promote dependency? Journl of Field Ornithology, 63, British Trust for Ornithology (2003) Grden BirdWtch: Feeding grden birds. (ccessed 22 August 2003). Cnnon, A. (1999) The significnce of privte grdens for bird conservtion. Bird Conservtion Interntionl, 9, Cnnon, A. (2000) The Grden Birdwtch Hndbook. British Trust for Ornithology, The Nunnery, Thetford, UK. Collins, J.P., Kinzig, A., Grimm, N.B., Fgn, W.F., Hope, D., Wu, J. & Borer, E.T. (2000). A new urbn ecology. Americn Scientist 88, Cornell Lbortory of Ornithology. (2003) Project FeederWtch. ( Accessed 23 August Cowie, R.J. & Hinsley, S.A. (1987) Breeding success of blue tits nd gret tits in suburbn grdens. Arde, 75, Cowie, R.J. & Hinsley, S.A. (1988) Feeding ecology of gret tits (Prus mjor) nd blue tits (Prus ceruleus), breeding in suburbn grdens. Journl of Animl Ecology, 57, Deis, R. (1982) Is bird feeding no-no? Defenders, 57, Dhondt, A.A., Eyckermn, R., Moermns, R. & Huble, J. (1984) Hbitt nd lying dte of Gret nd Blue tits Prus mjor nd P. ceruleus. Ibis, 126,

18 17 Dunn, E.H. & Tessgli-Hymes, D.L. (1990). Birds t your Feeder. W.W. Norton, New York. Green, R.J. & Higginbottom, K. (2000) The effects of non-consumptive wildlife tourism on free-rnging wildlife: A review. Pcific Conservtion Biology, 6, Gromdzk, J. & Lunik, M. (1978) Food composition of the strling nestlings, Sturnus vulgris L., in Wrsw. Act Ornithologic, 16, Grubb, T.C. & Cimprich, D.A. (1990) Supplementry food improves the nutritionl condition of wintering woodlnd birds: evidence from ptilochronology. Ornis Scndinvic, 21, Howrd, P. & Jones, D.N. (2004) The mening of feeding: A qulittive study of prctices nd resons for feeding wildlife in urbn South-Est Queenslnd. Urbn Wildlife: More thn meets the eye. (eds S.K. Burgin & D. Lunney) pp Royl Zoologicl Society of N.S.W., Sydney. Hunter, J. (2001). Urbn ntics: To feed or not to feed? Lndscope 17(2): 54. Jones, D.N. (2002) Mgpie Alert: Lerning to Live with Wild Neighbour. University of New South Wles Press Ltd, Sydney, NSW. Jones, D. N. 2003). Contemporry urbn ecology: the view from the Antipodes. In: Ecosystems nd Sustinble Development IV, Vol. 2 (eds. E. Tiexxi, C. A. Brebbi & J. L. Usó) pp WIT Press; Southmpton. Jones, D.N. & Howrd, P. (2001) Feeding wildlife in urbn res: An indecent obsession? Wildlife Austrli 38(3), Mrtin, T.E. (1987) Food s limit on breeding birds: life-history perspective. Annul Review of Ecologicl Systems 18, McLees, B. (2001) Feeding wildlife: Right or wrong. BSc. (Hons) thesis. Dekin University, Austrli. Mennechez, G. & Clergeu, P. (2001) Settlement of breeding Europen strlings in urbn res: Importnce of lwns vs. nthropogenic wstes. Avin Ecology nd Conservtion in n Urbnizing World (eds J.M. Mrzluff, R. Bowmn & R. Donnelly), pp Kluwer Acdemic, Norwell, MA. Nttrss, R. (2001) To feed or not to feed. Wildlife Austrli, 38(3),

19 18 O'Lery, R.A. & Jones, D.N. (2002) Forging by suburbn Austrlin mgpies during dry conditions. Corell, 26, O'Lery, R.A. (2002) Implictions of Wildlife Feeding in the suburbn Environment: A Nturl Experiment involving Austrlin Mgpies. B.Sc. (Hon) thesis, Griffith University, Austrli. Orms, M.R. (2002) Feeding wildlife s tourism ttrction: A review of issues nd impcts. Tourism Mngement, 23, Pierotti, R. & Annett, C. (2001) The ecology of Western gulls in hbitts vrying in degree of urbn influence. Avin Ecology nd Conservtion in n Urbnizing World (eds J.M. Mrzluff, R. Bowmn & R. Donnelly), pp Kluwer Acdemic, Norwell, MA. Queenslnd Bureu of Meteorology. (2003) ( Accessed 17 September Rollinson, D.J. (2003) Synnthropy in the Austrlin Mgpie: A Comprison of Urbn nd Rurl Popultions in Southestern Queenslnd, Austrli. PhD thesis, Griffith University, Austrli. Rollinson, D.J. & Jones, D.N. (2002) Vrition in breeding prmeters of the Austrlin mgpie Gymnorhin tibicen in suburbn nd rurl environments. Urbn Ecosystems, 6, Rollinson, D.J., O'Lery, R.A. & Jones, D.N. (2003) The prctice of wildlife feeding in suburbn Brisbne. Corell, 27, Royl Society for Protection of Birds. (2002) The Birds in Your Grden. RSPB, Sndy, Bedfordshire. Skir, I. & Smith, S. (1991) Feeding wildlife in ntionl prks. South Austrlin Regionl Seminr on Ntionl Prks nd Wildlife Mngement, Tsmni, Soler, M. & Soler, J.J. (1996) Effects of experimentl food provisioning on reproduction in the jckdw Corvus monedul, semi-colonil species. Ibis, 138, Sutherlnd, W.J. (1996) From Individul Behviour to Popultion Ecology. Oxford University Press, Oxford.

20 19 Thoms, L. (2000) Wildlife nd humns in suburbn setting. New pproches to the mngement of positive nd negtive interctions. PhD thesis. Griffith University, Brisbne. Veltmn, C.J. & Hickson, R.E. (1989) Predtion by Austrlin mgpies (Gymnorhin tibicen) on psture invertebrtes: Are non-territoril birds less successful? Austrlin Journl of Ecology 14, Wilbur, S.R., Crrier, W.D. & Bornemn, J.C. (1974) Supplementry feeding progrm for Cliforni condors. Journl of Wildlife Mngement, 38, Wilson, W.H. (2001) The effects of supplementry feeding on wintering blck-cpped chickdees (Poecile tricpill) in centrl Mine: Popultion nd individul responses. Wilson Bulletin 113, Woodll. P.R. (1995) Results of the QOS grden bird survey, , with prticulr reference to South-est Queenslnd. Sunbird 25, Zr, J.H. (1999) Biosttisticl Anlysis, 4 th edn. Prentice Hll, New Jersey.

21 20 Tble 1. Anlysis of vrince (lest squres generl liner model) of forging rte for nturl food (number of items per minute) for feedtype (fed nd nonfed mgpies), sex nd time of dy (morning, middy nd fternoon). Pirs within feedtype were used s the error for feedtype. *Significnt vrible or interction. Source d.f. Men Squre F-Vlue P-Vlue Feedtype * Pirs:Feedtype Sex Feedtype X Sex Time of dy Feedtype X Time of dy * Sex X Time of dy Feedtype X Sex XTime of dy Residul

22 21 Tble 2. Anlysis of vrince (lest squres generl liner model) of the totl number of items of food provisioned per 30 min for feedtype (fed nd nonfed mgpies), sex of provisioner nd food source (nturl nd supplementry). Pirs within feedtype were used s the error for feedtype. *Significnt vrible or interction. Source d.f. Men Squre F-Vlue P-Vlue Feedtype Pirs:Feedtype * Food Source * Feedtype X Food Source * Sex Feedtype X Sex Food Source X Sex Feedtype X Food Source X Sex Residul

23 22 Figure Legends Fig. 1. Totl proportion of nturl nd supplementry food ctegories ingested by fed (solid brs) nd unfed mgpies (open brs) (LrvWm = insect lrve/worms, FlyIns = flying insects, OthIns = ll other insects, OthNt = other nturl food, Met = ll mets, NMet = supplementry non-met items.)(totl number of items detected: 1736 for fed, 2409 for unfed.) Proportions with the sme letter not significntly different. Fig. 2. Men ( SE) (lest squres) number of items ingested per minute during forging for fed ( ) nd unfed ( ) mgpies. Mens with the sme letter not significntly different. Fig. 3. Men ( SE) dtes (1 = 1 st June) for strt of phses of breeding for fed ( ) nd unfed ( ) mgpies. Mens with the sme letter not significntly different. Fig. 4. Lest squres Men ( s.e.) number of items of nturl nd supplementl foods provisioned per 30 minutes to chicks by fed ( ) nd nonfed ( ) mgpies. Mens with the sme letter not significntly different.

24 b LrvWm FlyIns OthIns OthNt Met NMet b Ctegory of Food

25 b Morning Middy Afternoon Time of Dy

26 b b b Build Clutch Htch Fledge Stge of Breeding

27 b Nturl c d Supplementl Food Sourc