(Charadrius vociferous): testing models of avian parental defense

Transcription

1 Behav Ecl Scibil (1990) 26: Behaviral Eclgy and Scibilgy Springer-Verlag 1990 The effects f nesting stage, sex, and type f predatr n parental defense by killdeer (Charadrius vciferus): testing mdels f avian parental defense Dianne H. Bruntn* Schl f Natural Resurces, University f Michigan, Ann Arbr, Michigan, MI , USA Received July 19, 1988 / Accepted September 16, 1989 Summary. Tw mdels predicting the tempral patterns f parental investment in ffspring defense ver the nesting cycle were tested. The first is based n ffspring age, the ther n the vulnerability f ffspring t predatin. Bth mdels make very similar predictins fr altricial species after eggs have hatched, i.e., increases in intensity f parental defense until fledging. Fr preccial species, hwever, the pst-hatching predictins f each mdel are different: the ffspring age mdel predicts a cntinued increase in defense intensity, while the vulnerability mdel predicts a decline. I examined the tempral patterns f parental defense f a preccial shrebird, the killdeer (Charadrius vciferus), and determined which mdel was supprted. Killdeer respnses t human and natural predatrs were bserved. Killdeer were less willing t leave the nest, respnded mst intensely, and displayed clsest t a ptential predatr arund hatching. Defense intensity increased frm early t late incubatin as predicted by the ffspring age mdel. Hwever, after hatching killdeer parental defense declined fr bth males and females, thus supprting the vulnerability mdel fr this stage. Males and females respnded significantly differently t all types f predatrs. Males tk greater risks, remained n the nest lnger, defended ffspring mre intensely, and displayed clser t the predatr than females at the apprach f a ptential predatr. Respnses t natural predatrs depended n the type f predatr and the apprach made by the predatr; a greater range f defense behavir was used fr predatrs appraching n the grund cmpared t aerial predatrs. In general, killdeer respnses t humans were mre intense and less variable than their re- * Present address: Department f Bilgy, Yale University, OML, P.O. Bx 6666, New Haven, CT , USA spnses t natural predatrs. This was mst likely because the human intruder apprached nests and chicks mre directly and clser than natural predatrs. Intrductin Parental defense and its assciated risks are an imprtant cmpnent f parental investment (Trivets 1972). Predatin is a pwerful selective pressure and in species with parental care, parents ften enhance the survival f their ffspring by actively defending them frm predatrs (Gttfried 1979; Greig-Smith 1980; Gchfeld 1984). Defense behavir is assumed t be perfrmed at sme risk t the defending bird (Simmns 1955; Armstrng 1956; Kruuk 1964; Barash 1975; Anderssn et al. 1980; Greig-Smith 1980; Waiters 1982; Weatherhead 1982; Bruntn 1986). Althugh distractin displays, in particular injury-feigning, appear t place the perfrming bird at a greater risk f predatin, the actual risks are difficult t assess and may be small. Few bservatins f mrtality f displaying adults have been made (Jurdain 1936; Myers 1978; Bruntn 1986). The alertness f the displaying bird is cnsidered evidence f the small prbability f mrtality (Gchfeld 1984), i.e., an alert bird is unlikely t be captured by a ptential predatr. At the very least, defending parents are expending time and energy that cannt be expended n smatic activities. Thus, selectin shuld result in a level f defense that is a cmprmise between the benefit t the current brd and the survival f the parent and future brds (Trivers 1972). The hyptheses develped within the cst/benefit framewrk f parental investment have been

2 182 Incubatin Hatching FI.edging Incubatin Hatchingl Fledging Incubatin Hatching F[edging Nesting Stage Fig. 1. Graphical representatin f mdels f parental defense. Mdel 1 predicts that levels f parental defense crrelate with ffspring age (a). Mdel 2 predicts levels f parental defense crrelate with ffspring vulnerability and, thus, predicts different relatinships fr altricial species (b), where vulnerability cntinues t increase after hatching until fledging, and fr preccial species (e), where vulnerability peaks at hatching recently reviewed by Mntgmerie and Weatherhead (1988). Tw f these hyptheses make predictins cncerning the tempral changes in the intensity f defense behavir ver the nesting cycle. The first hypthesis is based n ffspring age (Barash 1975; Anderssn et al. 1980). This hypthesis predicts an increase in nest defense thrugh the breeding seasn as the difference between the reprductive value f ffspring and parent decreases (Fig. 1 a; Anderssn et al. 1980; Mntgmerie and Weatherhead 1988). Offspring value increases expnentially as ffspring vercme initial high juvenile mrtality and near independence (Curi et al. 1984). The secnd hypthesis is based n the vulnerability f ffspring t predatin (Skutch 1949; Harvey and Greenwd 1978). This hypthesis predicts that the level f nest defense crrelates with the vulnerability f ffspring t predatin (Fig. 1 b, c). The mre vulnerable ffspring are, due t their stage f develpment r physical lcatin, the greater the intensity f parental defense. Studies f parental defense have cncentrated n altricial species and have revealed increases in the intensity f nest defense by parents as ffspring increase in age (Erpin 1968; Barash 1975; Curi 1975; Weatherhead 1979; Greig-Smith 1980; Biermann and Rbertsn 1981; Merritt 1984; Shields 1984). Althugh the ffspring age hypthesis and the vulnerability hypthesis make different predictins fr altricial species during the egg stage, after hatching bth hyptheses generate the predictin that defense shuld increase and peak just prir t fledging (Fig. l a, b). Barash (1975) predicted a majr difference in the tempral pattern f parental defense between altricial and preccial spe- cies based n vulnerability t predatin. Chick vulnerability tends t increase until fledging (r when chicks can mve ut f the nest) fr altricial species because f increasing cnspicuusness f the nest (Skutch 1949; Barash 1975; Harvey and Greenwd 1978; East 1981; Gchfeld 1984; Anderssn et al. 1980). In cntrast, chick vulnerability peaks at hatching and decreases thereafter fr preccial species because f increasing chick mbility (Armstrng 1956; Barash 1975; Fig. lc). Furthermre, parental defense fr preccial species is independent f brd size after hatching, as defense is unshared parental investment and predatrs usually nly take a single ffspring (Lazarus and Inglis 1986). Hence, if ffspring age has the greater impact n intensity f defense fr preccial species, then defense shuld increase thrughut the nesting cycle and peak just prir t fledging (Fig. 1 a), but if chick vulnerability t predatin is mre imprtant, then defense shuld be cnstant during the egg stage, peak at hatching, and thereafter decrease (Fig. 1 c). Variatin in levels f defense may als differ with the sex f the parent (Burger 1981 ; Ashkenazie and Safriel 1979; Greig-Smith 1980; Mntevecchi and Prter 1980; Piertti 1981 ; Mundahl 1982; Regelmann and Curi 1983). These differences are usually cnsistent with thse predicted by sexual selectin and parental investment thery (but see Hwe 1982). If there are differences between the sexes in levels f defense, then these must be cntrlled fr in any study examining tempral changes in defense. Finally, parental defense behavir will be influenced by the "risk" psed by a ptential predatr (Kruuk 1964; Lemmetyinen 1971; Curi 1975; Greig-Smith 1980; Pattersn et al. 1980; Buitrn 1983; Gchfeld 1984). The species f predatr, its apprach, and the distance the predatr is frm the ffspring are all imprtant factrs. Individuals shuld respnd differently t predatrs that are capable f taking bth adults and ffspring cmpared t predatrs that can nly take ffspring. In this study I examine hw the intensity f

3 183 parental defense by killdeer (Charadrius vciferus) varies with nesting stage, sex f the defending parent, and the type f predatr. I test which f tw mdels is supprted by cmparing parental defense behavir f a preccial species ver all nesting stages. I als examine the effects f the sex f the defending parent, type f predatr, and predatr apprach, n defense behavir. Killdeer are cmmn Nrth American shrebirds. They are grund nesting birds that suffer heavy nest and hatchling mrtality (Nl 1980; Nl and Lambert 1984; Bruntn 1987). Studies f killdeer time-budgets (Leningtn 1980; Bruntn 1988a) and energy expenditures (Bruntn 1988b) supprt the general hypthesis that males and females maximize fitness in different ways. The parental rles f the sexes, hwever, were nt thse generally predicted fr a mngamus species; males allcated mre time and energy than females t parental activities (Phillips 1972; Mundahl 1982; Bruntn 1988a, b). Chicks are preccial, leave the nest within 24 h f hatching, and are nt fed by their parents. Killdeer parental care includes incubating and guarding eggs, brding, and leading and defending chicks (Phillips 1972; Leningtn 1980; Mundahl 1982). Methds Study areas During 1984 and studied 87 nesting attempts by 41 pairs f killdeer (see Bruntn 1987, 1988a fr descriptins f the study sites). Nests were lcated by bserving the behavir f adults. All defense behavir bservatins were made n individuals identifiable by dyed plumage, clr bands r unique plumage characteristics. Observatins were made frm tw 5 m twers and frm a car. Nests were checked every 2-5 days until nests failed r chicks hatched. After hatching, pairs with brds were mnitred at least weekly until chicks fledged. Definitins f parental defense Parental defense behavir f killdeer was studied by bserving their respnses t cntrlled appraches by human intruders and appraches by natural predatrs. Amng Charadriidae, distractin displays are a cmmn defense behavir, and apparently are quite effective at leading ptential predatrs away frm the nest and chicks (Larsn 1960; Bengtsn 1970; Hbbs 1972). Gchfeld (1984) gives the mst thrugh review f shrebird defense behavir, and I use his terminlgy. Killdeer respnd t ptential predatrs in a variety f ways, including bth cryptic behavir and distractin displays (Phillips 1972; Mundahl 1982). These respnses were ranked in terms f intensity: quiet departure (0), standing alert r calling (1), cruched r upright run (2), false-brding (3), lw, medium and high intensity injury-feigning (4, 5, 6), and threat display (7). As injury-feigning invlved a cntinuum f intensities, the divisin f this behavir int lw, medium, and high was determined based n the fllwing descriptins : lw intensity partial extensin f bth wings < 30 frm the bdy, legs flexed, and partial r n fanning f the tail; medium intensity asymmetrical extensin f the wings frm the bdy, with ne wing extended > 30 frm the bdy, legs flexed t fully cruched, and partial t full fanning f the tail; high intensity almst symmetrical extensin f the wings with bth wings > 30 t~rm the bdy, legs fully cruched with breast pressed t the substrate, and tail fully fanned. Threat displays (see als "ungulate display" Gchfeld 1984) invlve the killdeer running twards the predatr while hlding the wings symmetrically ut frm the bdy. All defense behavir, with the exceptin f quiet departure and standing alert, were cnsidered "distractin displays". T assess differences between nesting stages, I divided the nesting cycle int six stages: pre-laying (x= 10 days), pair frmatin and arrival n the breeding grunds; egg-laying (x= 6 days), first egg laying t clutch cmpletin; incubatin (x= 24 days), divided int early incubatin, 1 t 12 days after clutch cmpletin, and late incubatin, 12 days after clutch cmpletin until hatching; and chick dependence (apprximately 34 days), divided int brding, 1 t 15 days, and pst-brding, 15 days until fledging. Male and female defense behavir during early nesting attempts, initiated prir t 10 June, and late nesting attempts, initiated after 10 June, were als cmpared. Respnses t human intruders Parental defense by killdeer was measured by recrding three aspects f defense behavir t a human intruder at the nest r chicks: (1)"respnse distance", the distance between the intruder and nest, r chicks, at which I first perceived a change in behavir f the incubating, r tending parent, (initial respnses t an appraching intruder usually differed frm respnses given t an intruder at the nest), (2) "respnse intensity", as ranked abve; fr each sample the mst risky behavir was recrded, and (3) minimum "distractin display distance", the minimum distance between an intruder and a displaying bird. Distractin display distances were ranked: greater than 33 m (1), 16 t 32 m (2), 8 t 15 m (3), 4 t 7 m (4), 2 t 3 m (5), and less than 2 m (6). I made a ttal f 626 appraches t 87 nests. There was n crrelatin between the frequency f human visits t nests and the prbability f predatin (Bruntn 1987). The sexes f all birds were knwn frm bservatins f cpulatin behavir. T avid habituatin t the human intruder, nest r chick appraches were made n mre than nce every tw days (mean interval duratin ver entire breeding attempt was 6 days). T reduce the influence f envirnmental variables, nest appraches were made between 0700 and 1000 hurs r 1700 and 2000 hurs, but nt when it was raining. After bserving the pair in rder t determine the identity f the incubating r tending parent and the lcatin f its mate, I apprached the nest r chicks at a slw walk, beginning apprximately 200 m frm the nest r clsesl chick. Observatins were made using a prtable tape recrder. Upn reaching the nest r vicinity f the chicks, I remained standing fr 3 min while cntinuusly recrding the behavir and distance f the fcal animal and, when pssible, the behavir and lcatin f the ther parent. Samples sizes f males and females were as equal as pssible in rder t avid biases due t the different amunts f time the sexes spend incubating, brding, and tending yung (Bruntn 1988 a). Risk-taking Regardless f the actual cst f distractin displays, it is likely that the clser an adult is t a ptential predatr, and the

4 184 mre intense its display, the greater the risk relative t ther behavir. T cmpare risk taking by the sexes and ver different nesting stages, I used a risk index derived frm Windt and Curi (1986); risk=~r(t,/dr), where R is the rank value f each defence behavir r, tr is time spent in hehavir r, and dr is the mean distance frm the predatr during tr. A risk value was calculated fr each 3-rain sample perid. Lw values fr R mean that nly a lw risk is taken. The majr assumptins f the risk index were that risk f injury r death increases (1) the clser the defending parent is t the ptential predatr and (2) with increasing intensity f respnse behavir (Barash 1975; Anderssn et al. 1980). Respnses t natural predatrs Apprximately 1000 appraches by natural predatrs were bserved during mre than 2000 h f behaviral bservatins made thrughut all nesting stages. A ptential predatr was cnsidered t apprach a nest when it gt t within 33 m f the nest r chicks (clear bservatins past this distance were difficult because f grund cver r tpgraphy). Respnse intensity and the minimum display distance f bth parents were recrded when pssible. Respnse distance was difficult t assess as predatrs were ften sptted by killdeer pairs befre I nticed their apprach. Appraches by avian predatrs culd be aerial, grund, r a cmbinatin. The type f predatr, directin f apprach, duratin, r time f day culd nt be cntrlled; therefre, the sample sizes were unequal and killdeer respnses likely t be quite variable. Statistical prcedures Data were analyzed n an Apple Macintsh using the sftware packages Multiplan, Excel, Statfast, and Systat. Differences in frequency distributins were tested using the Chi-squared test f independence (Z 2 tests). The nn-parametric Wilcxn Rank Sum test (WRS test) was used n all categrical data. A significance level f c~ < 0.05 (2-tailed) was used fr all statistical tests. Results Sex, nesting stage, and grund predatrs (humans) Respnse distance varied frm 4-7 m t >33 m fr all nesting stages, and the nly significant differences between males and females were bserved during incubatin and pst-brding (Fig. 2). When significant differences existed between the sexes, it was always female respnse distance that was greater; females mved frm the vicinity f the nest r chicks sner than males. Females shwed n variatin in the mst frequent respnse distance during all stages; it was always > 33 m, whereas males shwed cnsiderable variatin (Fig. 2). The mst frequent male respnse distance decreased frm >33m during egg-laying t m during the remaining stages. Killdeer respnses t human intruders during early and late nesting attempts were nt significantly different ~ 2 tests, P > 0.23). Fr this reasn, z > ~ 100 EGG-LAYING NS, ~i Oi,, :~., ] EARLY INCUBATION P< ~ ~ 4i ~ 20. ~ 0 ~ ~l'~k--.,~ ~ I I 60, LATE INCUBATION P< !~ , 0 I I I'~-1, BROODING 80, , I :~ POST-BROOD ING 80, 60 40, 20 0 ~,r-~ < RESPONSE NS p< >33 DISTANCE (m) Fig. 2. Histgrams cmparing male (pen bars) and female (striped bars) respnse distance t a human intruder at the nest r chicks. Respnse distance is the distance between the incubating r tending bird and the intruder at which a change in behavir is first bserved. Chi-squared tests were used t determine significant differences. Samples sizes (males, females) fr each nesting stage were: egg-laying (31, 27), early incubatin (58, 55), late incubatin (78, 66), brding (97, 71), and pstbrding (89, 54) bservatins f appraches t early and late nests were cmbined fr all subsequent cmparisns. Althugh respnses t human intruders varied frm quiet departure t threat (Fig. 3), males and females respnded significantly differently during all nesting stages (Fig. 3, X 2 tests, P<0.0005), and the mst frequent respnse was mre intense fr males than females. Males and females differed in their respnses accrding t nesting stage, with the mst intense respnses ccurring during late incubatin (Fig. 3). Respnses were mderate during egg-laying; females departed the nest site quietly and did nt attempt t distract a human at the nest site, whereas males respnded by standing nearby and calling. The intensity f respnse fr bth sexes increased nce incubatin started. The mst frequent male respnse during early incubatin was lw intensity injury-feigning cmpared t the mst frequent female respnse f false brding. As incubatin

5 EGG-LAYING ~ p< t '7. 20 [~ 0, ~ ' : : : ~ ~ H 60 EARLY INCOBATI~N ><~ ~ F~ ~ 2< ~ O ===, ~.... r'-k--~. =:~ : : ~ LATE INCUBATION < m 40 ~ ~ r'k-'l, ;,. : : 60 0 BR00~ING 40 P< : POST-BROODING.00 z i-~,~ > N ~n s3 p 60 IEARLY INCUBATION ~ P<0.0005, I,ATE ' NCUB ' ION " " ' ~ ~<0 03 ~--.~. ~..~, ~. I BROOD'~NG ' ' I P< " ~ ~ ' ~, ~ POS~-BROOD~N~ ~< ~ ~, :.,,---,r"~.,.r-] : : TYPE OF RESPONSE Fig. 3. Histgrams cmparing the frequency f varius killdeer respnses t human intruders. Chi-squared tests were used t determine significant differences between male (pen bars) and female (striped bars). The intensities f defense behavirs are : 0 = quiet departure, 1 =standing alert r calling, 2=cruched r upright run, 3 = false brding, 4 = lw intensity injury-feigning, 5 = medium intensity injury-feigning, 6 = high intensity injury-feigning, 7=threat display. Sample sizes are shwn in Fig. 2 =.,R,FI < >33 DISPLAY DISTANCE Fig. 4. Histgrams cmparing the distractin display distance f males (pen bars) and females (striped bars) t human intruders at the nest r chicks ver each nesting stage. Distributins were cmpared using Chi-squared tests. Samples sizes (males, females) fr each nesting stage were: early incubatin (57, 53), late incubatin (78, 66), brding (97, 67), and pstbrding (84, 38) (m) prgressed, mre intense respnses were bserved. Respnses remained intense during brding, with the mst frequent male respnse being medium intensity injury-feigning and the mst frequent female respnse being lw intensity injury-feigning. During pst-brding, the intensity f respnse decreased fr bth sexes. Specifically cmparing injury-feigning respnses (lw, medium, and high) fr all stages, I fund that males were mre likely t injury-feign than females; 18.2% cmpared t 55.8% f human appraches (n=626, ~2 test, P<0.0001). Males were als mre likely t injury-feign when they were incubating r tending yung; 62.8% (182/290) cmpared t 28.3% (75/265) fr females (~ 2 test, P < 0.001). Furthermre, nn-incubating r nn-tending males were mre likely t injuryfeign (52.5%, 167/318) than nn-incubating r nn-tending females (25.3%, 39/154, X 2 test, P< 0.001). The minimum distractin display distances var- ied with the sex f the displaying bird and the nesting stage (Fig. 4). Males displayed significantly clser than females during every nesting stage (Fig. 4). The mst frequent minimum distractin display distances were lwer fr males than females during all stages except brding, when the mst frequent distance was 2-3 m. Fr bth sexes, display distances decreased frm early t late incubatin, were lwest during brding, and increased slightly during pst-brding (Fig. 4). During 287/452 appraches where bth parents were present, warning calls were given by nn-incubating r nn-tending males and females prir t a respnse frm the incubating r tending p~/rent. Risk-taking fr bth sexes increased frm prelaying t hatching, peaked at hatching, and decreased after hatching (Fig. 5). Risk-taking values were significantly greater fr males than females during early incubatin (WRS test, P<0.03), late incubatin (WRS test, P < 0.009), and pst-brding (WRS test, P<0.04). Male and female risk-

6 186 (A) MALES 20. b., 10. I ~4 O3 H 0::; 5 m -. ~ ~ " "..... l--lm" -- 1 ~ rm --:'g i" "--~ :. "i" )-=- {..~ ".. - II. "'~ j I~:;.1~; :1" ",.,].':. " -" m. m.:- ".: "- " ":=:i=77 %=:", (B) FEMALES 20 ~ 15 ~-I ~ 1 I M H 5 -).:.-..:. "~i.. ": ~". ~.. m... ) "m ~'-m- m ~m ~ -ram ~.m -. ~. "_.,.-_.-~'..,. EGG-LAYING EARLY LATE BROODING POST-BROODING PRE-LAYING INCUBATION CHICK DEPENDENCE NESTING STAGE Fig. 5. Risk-taking by (a) male and (b) female killdeer during a nesting attempt (fr explanatin f risk index see "Methds"). Sample sizes (males, females) fr each f nesting stage were: pre-laying (22, 16), egg-laying (13, 16), early incubatin (41, 40), late incubatin (44, 42), brding (68, 64), pst-brding (89, 76) taking was nt significantly different during egglaying (WRS test, P=0.65) and brding (WRS test, P = 0.37). Natural predatrs Althugh 53.4% f eggs were lst, I bserved the depredatin f nly five nests: tw by herring gulls (Larus argentatus), tw by cmmn crws (Crvus crvus), and ne by a raccn (Prcyn ltr). Crws and skunks (Mephitis mephitis) were the mst cmmn avian and mammalian predatrs at Hughtn Lake (Table 1). During 1984 and 1985 I bserved 1017 appraches by ptential predatrs. Respnses t natural predatrs by killdeer were bserved thrughut all nesting stages. One adult mrtality was bserved; the injury-feigning parent was captured by a fx a shrt distance frm the nest (Bruntn 1986). In general, the level f respnse t predatrs was greater during incubatin than during chick dependence (Fig. 6). The ne exceptin was male respnses t aerial predatrs, where respnses were mre intense during chick dependence. Males and females respnded in different ways t grund and aerial predatrs (Fig. 6). The mst frequent respnse f males was mre intense than that f females t grund predatrs during incubatin (medium intensity injury-feigning cmpared t quiet departure, Z 2 test, P < 0.01) and chick dependence (standing calling cmpared t standing quietly, Z 2 test, P<0.0001) and t aerial predatrs during chick dependence (standing calling cmpared t standing quietly, Z 2 test, P < 0.000l). During incubatin, males and females did nt respnd significantly differently (Z 2 test, P--0.35); quiet departure was the mst frequent respnse. Killdeer respnded differently t grund predatrs cmpared t aerial predatrs. Killdeer behaved with a greater range f respnses t grund predatrs cmpared t aerial predatrs, using injury-feigning behavir mre ften twards grund predatrs (Fig. 6). Neither sex respnded t an aerial predatr by false brding. Male respnses t grund predatrs varied frm sit-tight t threat, while respnses t aerial predatrs varied frm sit-

7 187 Table 1. The bserved number f appraches (<33 m) t killdeer nests r chicks by avian and mammalian ptential predatrs during incubatin (I) and chick dependence (CD), and the mde f each apprach Predatr Avian Nrthern harrier Circus cyaneus Red-tailed hawk Bute jamaicensis American kestrel Falc sparverius Herring gull Larus argentatus Cmmn crw CO?'VUS crvbis Cmmn grackle Quisealus quiscalus Mde f apprach Grund Aerial I CD I CD Ttal Mammalian Cat Fells dmesticus Raccn Prcyn ltr Wdchuck Marmta mnax Striped skunk Mephitis mephitis Red fx 3 2 Vulpes fulva Muskrat 6 1 Ondatra zibethica Ttal (a) INCUBATION 9 males P< % OBSERVATIONS 3O 3O 6O, ~ ~ SIT TIGHT Il111~1111~ II]]l]l]~ DEPARTURE QUIET ~1111~111 Ill,,ii,,,ill, NN STAND CALLING [] CROUCHED ~ RUN [~ FALSE ~ BROODING [] ~N~UR - [] FEIGNING 1 [ [] ~ I] (b) CHICK DEPENDENCE t? females P< males STAND I { lemales P< ~ QUIET Il11] ]#111# NS QUIET DEPARTURE IIIIIIIIIINIIIIINIIIIIIIII[II CALLINGSTAND ~ I1~ CROUCHED RUN ~ ~] FALSE BROODING ~ INJURY- FEIGNING 1 ~] ] I z rl [] a [] Fig. 6. Cmparisn between grund (pen bars) and aerial (striped bars) predatrs and the type f respnse exhibited by males and females during (a) incubatin and (b) chick dependence. Respnses t grund and aerial predatrs were cmpared using Chi-squared tests. Male sample sizes (grund, aerial) were: incubatin (132, 112) and chick dependence (58, 65). Female sample sizes (grund, aerial) were: incubatin (123, 87) and chick dependence (45, 29) tight t lw intensity injury-feigning. Female respnses t grund predatrs varied frm quiet departure t high intensity injury-feigning, while respnses t aerial predatrs included nly sit-tight and quiet departure. I bserved 177 distractin displays by males and 115 by females directed twards grund predatrs. Distractin display distances decreased frm early t late incubatin, were lwest during brding, and increased during pst-brding. Male and female distractin display distances were significantly different during all nesting stages with males displaying clser t ptential predatrs Of z tests, P<0.001). Predatr respnses t killdeer displays varied cnsiderably. Killdeer displaying at lw intensities were basically ignred by ptential predatrs, althugh it was difficult t tell if predatrs were ac- tively searching fr nests. Mammalian predatrs essentially ignred killdeer false-brding displays (18/22), while avian predatrs n the grund tended t fllw false-brding killdeer (23/31). All injury-feigning had a significant effect n mammalian predatr behavir; predatrs fllwed injury-feigning killdeer in 76/146 bservatins. Hwever, avian predatrs n the grund ignred injury-feigning killdeer 121/132 bservatins. Mde f predatr apprach: grund versus aerial Gulls and crws were the mst cmmn avian predatrs and are f apprximately equal size. I bserved 329 grund appraches and 239 aerial appraches by gulls and crws. There was a significant difference in the respnses f killdeer t

8 188 (a) INCUBATION %OBSERVATIONS i i i males ~] SIT TIGHT females P = O i ~i~i~:i~] ~ P = O. 009 I~!~ (b) CHICK DEPENDENCE QUIET DEPARTURE [iiii~ ~ ii;~i~i~!ii!~ii:!t~i ii~ i~c~.i STAND CALLING ~ g C.OUC.E~ ~ am ~ ~s~ U BR~DING ~ ~uav- ~ FEI~ING 1 C3 2 D I'] 3 I O males [--] STAND fen~ales P< ~ QUIET i!iiii~i~i::i::i[ NS QUIET DEPARTURE li)ii ii i::i:: ~ ii ilil i~i~ CALUNG STAND ~ ~ CROUCHED [~ RUN B FALSE BROODING ~ ' INJURY- [] i I FE~GN~NG~ I I 2 B a Fig. ~. Cmparisn between grund (pen bars) and aerial (dtted bars) appraches by gulls and crws, and the type f respnse exhibited by males and females during (a) incubatin and (~) chick dependence. Respnses t grund versus aerial appraches were cmpared using Chi-squared tests. Male sample sizes (~und, aerial) were: incubatin (117, 95) and chick dependence (76~ 45). Female sample sizes (grund, aerial) were: incubatin (]01, 70) and chick dependence (41, 24) grund versus aerial appraches by avian predatrs (Fig. 7). Respnses t grund appraches by gulls and crws were nt significantly different frm thse given t mammalian predatrs (incubatin: males X2=0.94, females Z2=1.76, P>0.05; chick dependence: males X z = 2.67, females Z 2 = 3.92, P>0.05). As with respnses t mammalian appraches, females respnded less intensely than males, and respnses t grund appraches by crws and gulls varied mre. Discussin Evidence is presented that suggests there are cnsistent differences between the sexes - with nesting stage and with the species f predatr and its mde f apprach. The varius measures f parental defense examined in this study, i.e., level f respnse, distractin display distance, and respnse distance, shw that killdeer parental defense crrelates with the ability f ffspring t gain frm defense. The mst intense respnses, medium and high intensity injury-feigning, were given by bth males and females just after hatching. Distances between displaying birds and ptential predatrs were shrtest fr bth sexes arund hatching and decreased after hatching as the chicks became mre independent. Similar results have been seen in ther preccial species (Simmns 1955; Gramza 1967). Hwever, the verall pattern f risk-taking by killdeer supprts bth defense mdels at different stages; during early nesting stages increasing levels f defense are cnsistent with increasing ffspring age (Fig. 1 b), while later stages shw decreasing levels f defense cnsistent with decreasing chick vulnerability (Fig. I c). Variatin in the frm, fi'equency, and level f defense between shrebird species has been bserved; fr example, species such as greater glden plvers Pluvialis apricaria (Ratcliffe 1977), lapwings, Vanellus spp, (Walters 1982), and Dunlins, Calidris alpina (Bent 1929) depart the nest early, whereas Australian Dtterels, Pelthyas australis (Maclean 1973), snipes, Gallinag spp, and wdcck, Sclpax spp, (Bent 1929) remain n the nest until the predatr is very clse. This study has shwn that cnsiderable variatin als exists within a species. Sexual differences in parental defense Male and female killdeer perfrmed all defense behavirs with the exceptin f the threat display. Thus, the bserved frequencies f the varius behavirs may represent differences in male and female strategies f defense behavir. Thrughut all reprductive stages, female killdeer respnses t human intruders were less intense than male respnses. Females used cryptic behavir such as false brding and quiet departure mre ften than males. Killdeer respnse distance differed fr males and females. Females tended t respnd early t human intruders regardless f the reprductive stage. Earlier female respnse may be an indicatin f female reluctance t place herself at risk by remaining n the nest. Alternatively, this difference may have been due t the tendency fr nn-incubating males t give alarm calls mre ften than nn-incubating females (315/452 cmpared t

9 /452). If males warn incubating r tending females mre ften, then females may be able t depart frm the nest sner. The cntributin f the secnd parent needs further investigatin. Due t the difficulty f sexing shrebirds in the field, and f fllwing birds after chicks have hatched, quantitative studies examining sexual differences in parental defense behavir are rare (Brwn 1962; Gchfeld 1984). Males perfrm mre defense than females in the majrity f cases where differences ccur (Gchfeld 1984). The apparent greater rle f males in parental activities may be related t high nest lsses amng grund nesters and thus t the imprtance f females being in gd cnditin t prduce replacement clutches (Mundahl 1982; Walters 1982; Bruntn 1988b). Furthermre, the cmmn ccurrence f female desertin amng shrebirds als supprts the idea that males play the predminant rle in nest defense (Gibsn 1971; Hussell and Page 1976; Ashkenazie and Safriel 1979; Maxsn and Oring 1980; Walters 1982; Hwe 1982; Kmeda 1983). Natural predatrs Despite variatin in predatr frm and in number f appraches by different species, it was pssible t make cmparisns between grund (mammalian and avian) and aerial (avian) predatrs. The mre intense defense behavir such as injury-feigning was mre ften directed twards grund predatrs. Respnses t aerial predatrs usually invlved sitting tight n the nest r quietly departing frm the vicinity f the nest. Furthermre, intense respnses were given t grund predatrs regardless f whether they were avian r mammalian even thugh avian predatrs tended t ignre injuryfeigning killdeer. The effectiveness with which shrebirds distract ptential predatrs has been nted by many authrs (Vgt 1938; Larsn 1960; Bengtsn 1970; Gibsn 1971; Hbbs 1972; Jnes 1979). Killdeer successfully distracted ptential predatrs frm eggs and chicks, during 1012 f 1017 bserved appraches. The mst bvius difference between killdeer respnses t human intruders and natural predatrs was the level f respnse. This is mst likely because f the differences in apprach and hw clse the ptential predatr was t the nest. The average clsest distance that natural predatrs came t nests r chicks was 21 m (n= 1017). The human intruder always apprached t within i m f the nest r chicks, clearly evking a mre intense respnse. Cnclusins The imprtance f cntrlling fr sex, type f predatr, and mde f predatr apprach has been underestimated when relating tempral patterns f parental defense t theretical mdels. The sex f the defending bird is especially imprtant; male killdeer defend mre intensely than females. Fr bth sexes defense peaks when ffspring are ld enugh t represent a large investment, but are still able t benefit frm cntinued investment by the parent. Many factrs ptentially influence whether the benefits t an individual f cntinued parental investment utweigh the csts. These include the genetic relatedness f the ffspring t a parent, the alternatives available, such as additinal mating pprtunities, future breeding r ther ffspring, and the ability f the ffspring t gain frm cntinued care.! suggest that patterns f parental defense by killdeer prvide further supprt fr the idea that an individual's decisin t cntinue investing in an ffspring des nt depend upn hw much has already been invested, rather, the level f defense crrelates mst strngly with the vulnerability f the ffspring t predatin. Acknwledgements. Financial supprt fr this study was prvided by the University f Michigan, a Rackham Predctral Fellwship, Rackham Dissertatin Grant, and Sigma Xi Grants-in-aid f Research. The Hughtn Lake Waste-Water Treatment Authrity and Rscmmn Cunty, Michigan, gave permissin t wrk n their land; I thank t]hem fr cperating with my research. The Department f Natural Resurces f the State f Michigan prvided ldging at ]Hughtn Lake and sme equipment. Assistance with data analysis by G. Fwler is gratefully acknwledged. Fr reviews f varius drafts f this wrk I thank R. Alexander, S. Beissinger, E. Curi, G. Fwler, B. Lw, D. Rabe, B. Stutchbury, and P. Weatherhead. References Anderssn M, Wiklund CG, Rundgren H (1980) Parental defense f ffspring: a mdel and an example. Anim Behav 28 : Armstrng EA (1956) Distractin display and the human predatr. Ibis 98 : Ashkenazie S, Safriel UN (1979) Time-energy budget f the Semipalmated Sandpiper (Calidris pusilla) at Barrw, Alaska. Eclgy 60: Barash DP (1975) Evlutinary aspects f parental behavir: distractin display f the alpine accentr. Wilsn Bull 87: Bengtsn SA (1970) Breeding behaviur f the Purple Sandpiper Calidris maritima in West Spitsbergen. Ornis Scand 1:17-25 Bent AC (1929) Life histries f Nrth American shrebirds. Part 2, US Natl Mus Bull 142 Biermann GC, Rbertsn RJ (1981) An increase in parental

10 190 investment during the breeding seasn. Anim Behav 29 : Brwn RG (1962) The aggressive and distractin display behavir f the Western Sandpiper Ereunetes mauri. Ibis 104:1-12 Bruntn DH (1986) Fatal antipredatr behavir by Killdeer. Wilsn Bull 98(4) : Bruntn DH (1987) Reprductive effrt f male and female Killdeer (Charadrius vciferus). PhD dissertatin University Michigan, Ann Arbr, Michigan Bruntn DH (1988a) Sexual differences in time budgets f Killdeer during the breeding seasn. Anim Behav 36: Bruntn DH (1988b) Energy expenditure in reprductive effrt: the reprductive strategies f male and female Killdeer (Charadrius vciferus). Auk 105: Buitrn D (1983) Variability in the respnses f Black-billed Magpies t natural predatrs. Behaviur 78: Burger J (1981) Sexual differences in parental activities f breeding Black Skimmers. Am Nat 117 : Curi E (1975) The functinal rganizatin f antipredatr behaviur in the Pied Flycatcher: a study f avian visual perceptin. Anita Behav 23:1 115 Curi E, Regelmann K, Zimmerman U (1984) The defence f first and secnd brds by Great Tit (Parus majr) parents: a test f predictive scibilgy. Z Tierpsychl 66 : East M (1981) Alarm calling and parental investment in the Rbin, Erithacus rubecula. Ibis 123(2): Erpin MJ (1968) Nest-related activities f Black-billed Magpies. Cndr 70: Gibsn F (1971) The breeding bilgy f the American Avcet Recurvirstra americana in central Oregn. Cndr 73 : Gchfeld M (1984) Antipredatr behavir: aggressive and distractin displays f shrebirds. In: Burger J, Olla BL (eds) Shrebirds: breeding behavir and ppulatins. Plenum Publ. Crp., New Yrk, pp Gttfried BM (1979) Anti-predatr aggressin in birds nesting in ld field habitats: an experimental analysis. Cndr 81 : Gramza A (1967) Respnses f brding Nighthawks t a disturbance stimulus. Auk 84: 7~86 Greig-Smith PW (1980) Parental investment in nest defense by Stnechats (Saxicla trquata). Anim Behav 28: Harvey PH, Greenwd PJ (1978) Antipredatr defence by Stnechats (Saxicla trquata). Anim Behav 28: Hbbs JN (1972) Breeding f Red-capped Dtterel at Fletcher's Lake Daretn, NSW. Emu 72: Hwe MA (1982) Scial rganizatin in a nesting ppulatin f eastern Willets (Catptrphrus semipalmatus). Cndr 84: Hussell D, Page GW (1976) Observatins n the breeding bilgy f the Black-bellied Plvers n Devn Island, N.W.T., Canada. Wilsn Bull 88: Jnes A (1979) Ntes n the behaviur f the Variable Oystercatcher. Ntrnis 26 : Jurdain FC (1936) The s-called injury feigning in birds. Olgists Rec 16:25-37 Kmeda S (1983) Nest attendance f parent birds in the Painted Snipe. Auk 100:48-55 Kruuk H (1964) Predatrs and anti-predatr behavir f the Black-headed Gull (Larus ridibundus L.). Behav Suppl 11 : Larsn S (1960) On the influence f the Arctic fx (Alpex lagpus) n the distributin f Arctic birds. Oiks 11 : Lazarus J, Inglis IR (1986) Shared and unshared parental investment, parent-ffspring cnflict and brd size. Anim Behav 34: Lemmetyinen R (1971) Nest defence behaviur f Cmmn and Arctic Terns and its effects n the success achieved by predatrs. Ornis Fenn 48:13-24 Leningtn S (1980) Bi-parental care in Killdeer: an adaptive hypthesis. Wilsn Bull 92(1): 8-20 Maclean GL (1973) A review f the bilgy f the Australian Desert Waders, Stilta and Pelthyas. Emu 73 : Maxsn S J, Oring LW (1980) Breeding seasn time and energy budgets f the plyandrus Sptted Sandpiper. Behaviur 74: Merritt PG (1984) Observer recgnitin by the Nrthern Mckingbird. J Field Ornithl 55:25~253 Mntevecchi W, Prter J (1980) Parental care and related behavir befre fledging in Nrthern Gannets (Mrus bassanus) with reference t ther marine birds. In: Burger J, Olla B, Winn H (eds) Behavir f marine animals. Vl. VI. Marine birds. Plenum Press, New Yrk, pp Mntgmerie RD, Weatherhead PJ (1988) Risks and rewards f nest defence by parental birds. Q Rev Bil 63: Mundahl JT (1982) Rle specializatin in the parental and territrial behavir f the Killdeer. Wilsn Bull 94(4): Myers JP (1978) One deleterius effect f mbbing in the Suthern Lapwing (Vanellus chilensis). Auk 95:419 Nl E (1980) Factrs affecting the nesting success f the Killdeer (Charadrius vciferus) n Lng Pint, Ontari. M.S. thesis, University Guelph, Guelph, Ontari Nl E, Lambert A (1984) Cmparisn f Killdeer, Charadrius veiferus, breeding in Mainland and peninsula sites in Suthern Ontari. Can Field Nat 98 : 7-11 Pattersn TL, Petrinvich L, James DK (1980) Reprductive value and apprpriateness f respnse t predatrs by White-crwned Sparrws. Behav Ecl Scibil 7: Phillips RE (1972) Sexual and agnistic behavir in the Killdeer (Charadrius vciferus). Anim Behav 20 : 1-9 Piertti R (1981) Male and female parental rles in the Western Gull under different envirnmental cnditins. Auk 98 : Ratcliffe DA (1977) Observatins n the breeding f the Glden Plver in Great Britain. Bird Study 23: Regelmann K, Curi E (1983) Determinants f brd defence in the Great Tit Parus majr L. Behav Ecl Scibil 13: Shields WM (1984) Barn Swallw mbbing: self-defence, cllateral kin defence, grup defence, r parental care? Anim Behav 32: Simmns KEL (1955) The nature f the predatr reactins f waders twards humans, with special reference t the rle f the aggressive-, escape- and brding-drives. Behaviur 8: Skutch AF (1949) D trpical birds rear as many yung as they can nurish? Ibis 91: Trivets RL (1972) Parental investment and sexual selectin. In: Campbell B (ed) Sexual selectin and the descent f man, Aldine Press, Chicag, pp Vgt W (1938) Preliminary ntes n the behavir and eclgy f the eastern Willet. Prc Linn Sc NY 49 : 8-42 Walters JR (1982) Parental Behavir in Lapwings (Charadriidae) and its relatinships with clutch sizes and mating systems. Evlutin 36: Weatherhead PJ (1979) D Savannah Sparrws cmmit the Cncrde Fallacy? Behav Ecl Scibil 5: Weatherhead PJ (1982) Risk-taking by red-winged blackbirds and the Cncrde Fallacy. Z Tierpsychl 60: