Does large size make daughters of the blue-footed booby more expensive than sons?

Ecology 1999, 68, Does large size make daughters of the blue-footed booby more expensive than sons? ROXANA TORRES and HUGH DRUMMOND Instituto de EcologõÂa, Universidad Nacional AutoÂnoma de MeÂxico, A.P. 70±275, MeÂxico DF, 04510, MeÂxico Summary 1. In species where size dimorphism emerges during parental care, individuals of the larger sex are presumably more expensive to produce than individuals of the smaller sex. 2. We evaluated whether female o spring of the blue-footed booby, because of their larger size, require greater parental feeding expenditure than males. 3. A eld experiment compared growth rates of sons and daughters when hand-fed the same amount of food, and a descriptive eld study compared the parental feeding rates to edgling sons and daughters. 4. Females maintained greater rates of mass and ulna increase than males when fed the same amount of food. 5. First-hatched chicks had higher probability to be fed than second-hatched chicks. However, male and female edglings were fed at similar rates by their parents. 6. No evidence that daughters of the blue-footed booby receive a greater feeding expenditure than sons was found. Faster motor development of males in this facultatively siblicidal species might explain why they grow more slowly than females with the same food budget. Key-words: feeding investment, hand-fed, parental care, size dimorphism, Sula nebouxii. Ecology (1999) 68, Introduction Correspondence: Dr Roxana Torres, Instituto de Ecologõ a, Universidad Nacional Auto noma de Me xico, A.P. 70±275, Me xico D.F. 04510, Me xico. Tel.: (52±5) 622 90 07. Fax: (52±5) 616 19 76. E-mail: rtorres@miranda. ecologia.unam.mx In animal species where the sexes di er in size, it is generally assumed that members of the larger sex are more expensive to produce than members of the smaller sex (Clutton-Brock, Albon & Guinness 1981). Individuals of the larger sex have higher growth and metabolic rates (Fiala 1981; Charnov 1982; Teather 1987; Teather & Weatherhead 1988; Anderson et al. 1993b), and presumably require more food to fuel growth than the smaller sex (review in Anderson et al. 1993b; but see Newton 1978; Richter 1983; Stamps 1990). Greater parental expenditure on o spring of the larger sex has been reported in several species of mammals and birds in which males are larger than females (review in Clutton-Brock 1991). For example, in mammals, greater milk intake by male o spring occurs in the red deer (Cervus elaphus, Clutton-Brock et al. 1981), the African elephant (Loxodonta africana, Lee & Moss 1986) the American bison (Bison bison, Wol 1988) and the California sea lion (Zalophus californianus, Ono & Boness 1996), although similar maternal investment in male and female o spring has been observed in the southern elephant seal (Mirounga leonina, McCann, Fedak & Harwood 1989) the grey seal (Halichoerus grypus, Smiseth & Lorentsen 1995) and the Antarctic fur seal (Arctocephalus gazella, Lunn & Arnould 1997). In birds, greater energy demands and food consumption by the larger sex have been reported in species where male nestlings are larger than females (Fiala 1981; Fiala & Congdon 1983; Slagsvold, Rùskaft & Engen 1986; Teather 1987, 1992; Teather & Weatherhead 1988). For example, in red-winged blackbirds (Angelaius phoeniceus), where males are roughly 30% heavier than females, total energy assimilated between hatching and edging was 27% greater for males than for females (Fiala & Congdon 1983). Field observations in two di erent study sites, suggest that male nestlings were

1134 Size dimorphism in the blue-footed booby fed 29% more often than females (Teather 1992), and that parental feeding rates increased with the proportion of males in the brood (Yasukawa et al. 1990). When females are the larger sex, parental expenditure on daughters and sons has not been compared in mammals (Clutton-Brock 1991), and the data for birds are scarce and inconclusive. Observations of a small sample of sparrowhawks (Accipiter nisus) in captivity suggest that female nestlings, which grow to be roughly 73% heavier than males, may consume more food than males when food is provided ad libitum (Frumkin 1988), although in the eld both sexes appeared to consume similar amounts (Newton 1978; Newton & Marquiss 1979). In the golden eagle (Aquila chrysaetos), female nestlings grew to be roughly 30% heavier than males, but they did not consume signi cantly more food than males over the 10-week rearing period, either in the eld or in captivity (Collopy 1986). However, females in the eld consumed on average 39% more food than males during the last 5 weeks of the rearing period and captive females were capable of consuming more food per meal than were males (Collopy 1986). Di erences in the relative costs of producing sons compared to daughters have implications for the population sex ratio. According to Fisher's (1930) theory, total investment by parents in o spring of each sex should be equal at the end of parental care. If individuals of one sex are more costly, then equal investment implies producing more o spring of the less expensive sex and a sex ratio biased towards the cheaper sex is expected at independence. Females of the blue-footed booby (Sula nebouxii) are larger and about 27% heavier than males at 79 days of age, when growth rates of size and mass reach an asymptote (Drummond et al. 1991; Guerra & Drummond 1995). Producing females may thus represent a greater cost to the parents. Although male and female nestlings do not di er in size (ulna and culmen lengths) and mass at hatching (Torres & Drummond 1999), females grow faster and reach a higher asymptotic size and mass (Drummond et al. 1991). The blue-footed booby raises from one to three chicks that compete aggressively for food in the nest (Drummond, Gonzalez & Osorno 1986). Both parents incubate the clutch and feed their nestlings by regurgitating sardines and anchovies into their mouths (Nelson 1978; Drummond et al. 1986). Parents predigest food before transferring it to very young chicks and transfer whole shes to older chicks (Guerra & Drummond 1995). As chicks age, meals increase in size and decrease in frequency: diurnal meals decreased from approximately 10 to 2 meals per day between ages 10 and 60 days (Guerra & Drummond 1995). Parental feeding continues after chicks have completed plumage development (mean edgling age = 86 0 days for males and 90 5 days for females; Drummond et al. 1991). Chicks remain near the nest and are attended continuously by at least one parent during the rst 8±9 weeks of the chick's life (Nelson 1978). Thereafter, chicks frequently wander during the day and return to nest sites to be fed by their parents. However, there is no data on the duration of the parental feeding period after edging (i.e. the time of edglings' independence). The blue-footed booby colony at Isla Isabel, Mexico, has on average 825 nests per reproductive season (SD, 2 266 data from 10 years; H. Drummond, unpublished). This colony produces a edging sex ratio biased towards males. In 1989, the sample of 751 chicks was male biased at hatching (56%) and at edging (56%; Torres & Drummond 1999). Fledging sex ratios during the ve subsequent years were male-biased varying from 56 to 70%, except for one year when similar numbers of males and females were produced (Torres & Drummond 1999). Greater production of sons agrees with Fisherian expectations under the assumption that daughters, because of their larger size are more costly to rear. To assess whether the observed male-biased sex ratio does indeed re ect equal overall expenditure on each sex, we evaluated whether daughters cost more to rear than sons. To do this, we evaluated whether daughters require a greater feeding investment than sons. We compared the growth rates of male and female chicks experimentally fed the same amount of food in the natal nest during the period of maximum daily growth rates. Two feeding treatments were established, one where chicks received a food amount similar to natural broods, and one where chicks received a food amount smaller than natural broods. If daughters normally sustain their faster growth by consuming more food than sons, we predicted that daughters and sons would have similar growth rates when fed the same amount of food. Also, investment in daughters and sons may di er toward the end of the period of parental care, when size dimorphism is established. To detect this, we compared natural feeding rates to sons and daughters during the last weeks before the end of parental care. If daughters require a greater investment than sons to reach independence, we predicted that female edglings should be provisioned more frequently or for a longer period than male edglings. Methods Field work was conducted from March to July 1994 in the blue-footed booby colony on Isla Isabel, o the Paci c coast of Mexico. All nests in a long-term

1135 R. Torres & H. Drummond study area were marked and surveyed every second day, which allowed us to estimate the age of chicks (22 days). Nestlings were marked with a coloured wire 6-g leg band at hatching and a numbered plastic leg band after age 7 days. Survival of nestlings was recorded until edging, when edglings were tted with numbered steel bands. Chicks were sexed by the length of their ulna at edging, when this structure has reached an asymptotic non-overlapping bimodal size (Drummond et al. 1991; Torres & Drummond 1997). This is a reliable method of sexing blue-footed boobies. In a random sample of 16 edglings, sex determination by ulna length and later gonadal examination agreed in all cases (Drummond et al. 1991). Additionally, the sex of 207 edglings determined by ulna length was veri ed 4 years later by sexing on the basis of the highly distinct adult vocalizations. Sex determination by these two methods agreed in all cases except 1 (Torres & Drummond 1997). FEEDING EXPERIMENT To exclude variance due to sibling interactions, only chicks from single-chick broods were used in the feeding experiment. In the colony at Isla Isabel single-chick broods are common: the proportions of one-, two- and three-chick broods are on average 55, 38 and 6%, respectively (data from 10 reproductive seasons; H. Drummond, unpublished). Subjects Seventy-eight 22-day-old singletons were randomly assigned to three treatment groups: 27 underfed chicks (hereafter, the underfed group), 24 abundantly fed chicks (the fed group), and 27 chicks fed by their parents (the control group). Chicks were sorted according to their hatching date, in blocks of 7 days, and then randomly assigned to each treatment group. Since male and female chicks cannot be distinguished when they are young, we were blind with respect to the sex of the subjects during the feeding experiment. The mean absolute mass of male and female chicks at the beginning of the experiment were similar among the three treatment groups. However, overall, females were signi cantly heavier than males (Table 1). In contrast, mean absolute ulna length was not signi cantly di erent among nestlings of the three treatment groups or between the sexes (Table 1). In each group females were heavier than males by 12% in the underfed group, 6% in the fed group, and 10% in the control group. Female ulna lengths were 2% larger than males in the underfed and control groups, and 0 7% smaller than males in the fed group. Treatment To control the amount of food received, we handfed chicks at their nests and placed an adhesive cloth tape (Micropore) around their necks at the base of the cranium to prevent parental feeding between hand-feeding events. This method of controlling food delivery to chicks has been used for up to 3 days without any signs of permanent harm (Drummond & Garcõ a-chavelas 1989; NunÄ ez-de la Mora, Drummond & Wing eld 1996; Rodrõ guez- Girone s, Drummond & Kacelnik 1996). Chicks from the underfed and fed groups were fed during nine consecutive days, from ages 22±30 days. This age interval includes the maximum daily growth rates for ulna length, and mass of female and male chicks from natural nests (Torres 1996). Nestlings were fed beside their home nests by opening their beaks and putting fragments of fresh local sh (skipjack, Katsuwonus pelamis, and bonita, Sarda sarda) into their mouths. Hand-fed chicks Table 1. Mean mass (g) and ulna length (mm) of male and female blue-footed booby chicks 1 day before they entered the feeding experiment. Sample sizes are 12 females and 11 males in the underfed group, 7 females and 9 males in the fed group, and 10 females and 16 males in the control group F Two-way ANOVA Females Males Treatment Sex Interaction Mass Underfed 653 08 2 22 70 582 91 2 23 70 0 78 7 48* 0 20 Fed 617 40 2 29 80 580 41 2 26 20 Control 618 08 2 24 90 560 51 2 19 70 Ulna Underfed 88 01 2 2 88 86 50 2 3 01 0 44 0 12 0 07 Fed 85 30 2 3 77 85 91 2 3 33 Control 85 40 2 3 16 83 61 2 2 50 *P < 0 05.

1136 Size dimorphism in the blue-footed booby accepted the new diet and feeding method. The amounts of food were based on the estimated baseline food intake of blue-footed booby chicks of similar ages (Guerra & Drummond 1995). The underfed and fed groups received 150 and 200 g (2 1) of sh per day, respectively, plus a 5% increment every second day. During the 9 days of treatment, chicks from the underfed group received a total of 1472 g, which is 21% less than baseline, while chicks from the fed group received a total of 1964 g or 6% more than baseline. Each chick's ration was delivered in three daily feeding visits, a frequency similar to parental deliveries to chicks of similar ages (mean = 3 4 feeds in a 12-h period, range = 1±12, on Isla Isabel, Mexico, Guerra & Drummond 1995; roughly 3 0 feeds in a 24-h period for two-chick broods, on Isla EspanÄ ola, Galapagos Islands, gure 7 in Anderson & Ricklefs 1992). Feeds were at roughly 06.00, 11.00 and 16.00 h, providing 40, 20 and 40%, respectively, of the daily ration. During the feeding experiment 13 chicks died: four in the underfed group, six in the fed group and three in the control group. From the chicks that died, six disappeared possibly victims of predation by feral cats, two died showing serious injuries that were probably caused by adult neighbours (Drummond et al. 1991), and another ve probably died because they were smaller and looked weaker than the rest of the experimental chicks from the start of the experiment. We could not exclude possible association between the experimental treatment and the death of these last ve chicks. However, they were all from the latest 26 broods included in the experiment, and from day 1 their masses were 11% lower and their ulnas were 9% shorter than those of the other experimental chicks. Hand-fed chicks were weighed on an Ohaus electronic balance (0±6000 2 1 g) before each feeding, and every morning the lengths of their left ulnas were measured (2 1 mm). To verify that parents resumed normal feeding, chicks were weighed and measured during three additional days after the last feeding (when tapes were removed at approximately 18.00 h). Chicks from the control group were simply weighed and measured on the same schedule as hand-fed chicks, but were not taped or fed by us. Hence, although chicks in this group were not a strict control, they illustrate normal growth of chicks subjected to our weighing schedule. Daily survival of hand-fed and control chicks was recorded until edging. We analysed growth by calculating independently the rate of increase of mass and ulna length of every chick during the 9-day experimental period. To control for di erences in chicks' initial sizes, the rate of increase was calculated as the di erence between the nal and initial measures divided by the initial measure. Rates of mass and ulna increase of male vs. female nestlings were compared using Two-way ANOVAs, with treatment (underfed, fed and control groups) and sex as the main factors. Duncans' multiple post hoc comparisons were made using a global signi cance level of 0 05. The growth rates of ulna and mass increase of hand-fed males vs. females were also compared by performing a Generalized Linear Model with an Identity Link function, and normal distribution of error, sex and treatment were included as factors in the model (Crawley 1993). This analysis simultaneously takes into account the initial size of the chicks and allometric e ects between mass and ulna length. PARENTAL FEEDING AFTER FLEDGING Parental feeds to edglings not involved in the experiment were recorded daily from 14.00 to 19.00 h, during the peak daylight feeding period (Torres 1996; see also Anderson & Ricklefs 1992). An observer walked slowly along a xed route through the study area, stopping for 30 min every 30 m, alternating travel direction every day to avoid any possible bias due to passing through the same sites at the same time. We recorded all feeds to edglings within an estimated 30-m radius of the observer, noting each edgling's band number (using 10 23 binoculars). Each chick's age, hatching sequence and sex determined by the length of its ulna were subsequently recovered from data in our project's routine measurements of all chicks. Feeding records began on 16 June, when roughly 20% of rst-hatched chicks from the marked nests in the study area were older than 90 days and continued until 15 July, just before the end of the reproductive season. This recording method produced a daily sample of the total feedings to marked edglings in the area surveyed. Assuming that male and female edglings are fed at similar times of day, samples allow us to compare feeding frequencies to the two sexes, but they do not yield an estimate of absolute daily frequencies. The statistical analysis was performed using a Generalized Linear Model with a Logit Link function and binomial distribution of error (Crawley 1993). The maximal model included sex and hatching sequence as factors and age of chicks as a continuous variable. Feeds to third-hatched chicks (31 to males and seven to females) were not included in the analysis because of small sample size. Results FEEDING EXPERIMENT Feeding treatments produced di erences in mass and ulna increase (mass, F 2,64 = 5 08, P = 0 009; ulna, F 2,64 = 9 01, P = 0 0004). Overall, the under-

1137 R. Torres & H. Drummond fed group (12 females and 11 males) showed lower rates of mass increase than the fed (seven females and nine males) and control groups (10 females and 16 males), while the rates of the fed and control groups were not signi cantly di erent (Duncans' multiple comparisons). The mean rates of mass increase during the 9-day experimental period (mean 2 SEM g) were 0 32 2 0 02 in the underfed group, 0 40 2 0 02 in the fed group, and 0 38 2 0 02 in the control group. Ulna growth rates of the underfed and fed groups did not di er signi cantly, and both of these groups grew more slowly than the control group (Duncans' multiple comparisons). The mean rates of ulna length increase (mm) were 0 33 2 0 01 in the underfed group, 0 36 2 0 01 in the fed group, and 0 40 2 0 01 in the control group. The prediction that under a regime of equal food intake female and male chicks should have similar growth rates was not supported. Female rates of mass increase were signi cantly greater than male rates (F 1,64 = 11 92, P = 0 001; Fig. 1a): 33, 25 and 19% greater than male rates in the underfed, fed and control groups, respectively. There was no signi cant interaction between treatment and sex (F 2,64 = 0 13, P = 0 87; Fig. 1a), suggesting that rates of mass increase of both sexes di ered similarly at di erent levels of provisioning. Similarly, female rates of ulna increase were greater than male rates in the three groups (F 1,64 = 5 95, P = 0 01; Fig. 1b); by 9% in the underfed group, 11% in the fed group and 13% in the control group. As for mass increase, there was no signi cant interaction between treatment and sex (F 1,64 = 0 15, P = 0 85; Fig. 1b). After the 9-day feeding treatment, hand-fed nestlings grew normally or recovered rapidly, and no noticeable harm was detected. Three days after tapes were removed, nestlings of the three groups did not di er in their mass or ulna lengths (mass, F 2,64 = 0 60, P = 0 51, ulna F 2,64 = 0 68, P = 0 50), suggesting normal parental feeding. Likewise, experimental chicks eventually achieved a size that was not signi cantly di erent from that of unmanipulated chicks (asymptotic ulna length for experimental and unmanipulated chicks, respectively, mean 2 SD; females 216 2 2 93 mm, n = 29 and Fig. 1. Mean rates (2 SEM) of (a) mass and (b) ulna increase during the 9-day experimental period of blue-footed booby male and female nestlings in the underfed (12 females and 11 males), fed (seven females and nine males) and control (10 females and 16 males) groups.

1138 Size dimorphism in the blue-footed booby 216 2 3 35 mm, n = 257, t = 1 00, P = 0 32; males 196 2 4 15 mm, n = 36, and 196 2 4 62 mm, n = 428, t = 0 08, P = 0 94). Survival to edging of experimental nestlings after treatment was high, only one female died at age 58 days. To further evaluate sex di erences in growth rates while controlling for possible allometric e ects, we compared the relationship between ulna and mass increase of hand-fed chicks (i.e. only the underfed and fed groups). Ulna and mass increases were positively correlated (F 1,36 = 9 98, P = 0 003), and chicks from the underfed group had on average 8% lower rates than chicks from the fed group (F 1,36 = 5 39, P = 0 02; Fig. 2). The di erences between the sexes were signi cant in both treatments: females had 30 and 24% higher rates of increase than males in the underfed and fed groups, respectively (F 1,36 = 5 97, P = 0 01; Fig. 2). Neither of the interactions were signi cant (treatment * sex: F 1,36 = 3 23, P = 0 08, ulna increase * treatment: F 1,36 = 0 44, P = 0 50, ulna increase * sex: F 1,36 = 0 01, P = 0 89, ulna increase * treatment * sex: F 1,36 = 3 36, P = 0 08). Thus, females maintained higher growth rates than males when fed the same amount of food. PARENTAL FEEDING AFTER FLEDGING The probability of feeding of rst- and secondhatched chicks varied with age of chick and hatching sequence in our sample of 186 male and 115 female edglings (age w 2 = 5 30, d.f. = 1, P = 0 02, sequence w 2 = 22 51, d.f. = 1, P < 0 0001; Fig. 3). First-hatched chicks had, on average, a 0 02 greater probability of being fed than second-hatched chicks. Disparities between rst- and second-hatched chicks did not change as chicks got older, though variations felt short of signi cance (interaction e ect, age * sequence w 2 = 3 18, d.f. = 1, P = 0 07; Fig. 3). The prediction of greater parental provisioning to daughters than sons after edging was not supported. Overall, the probability of feeding of male and female edglings did not di er (w 2 = 0 05, d.f. = 1, P = 0 82), whether age of chick (age * sex w 2 = 0 003, d.f. = 1, P = 0 95) and/or hatching sequence were considered (sex * sequence w 2 = 2 72, d.f. = 1, P = 0 09, sex * sequence * age w 2 = 0 07, d.f. = 1, P = 0 79). Mean probabilities to be fed were 0 06 for rst-hatched males, 0 06 for rsthatched females, 0 05 for second-hatched males and 0 03 for second-hatched females. Discussion FEEDING EXPERIMENT We found no evidence that singleton daughters, because of their larger size, require a greater feeding investment by parents than sons. When both sexes were fed the same amount of food, females maintained greater rates of mass and ulna increase than males. Also, there was no indication of any interaction between treatment and sex, showing that rates of mass and ulna increase of both sexes were similarly a ected by the amount of food delivered to a chick. Although 9 days represents only roughly 13% of the growth period (before asymptote of mass and Fig. 2. Rates of ulna and mass increase of blue-footed booby males and females of the underfed and fed groups. Regressions were calculated using a Generalized Linear Model with normal error distribution. The estimated parameters are from the minimal adequate model, i.e. all terms in the model are signi cant. Adjusted curves are dotted lines for males and solid lines for females (r = 0 63, d.f. = 36, P < 0 0001).

1139 R. Torres & H. Drummond Fig. 3. Probabilities of rst- and second-hatched edglings to be fed by parents. Regressions were calculated using a Generalized Linear Model with binomial error distribution. The estimated parameters are from the minimal adequate model (r = 0 34, d.f. = 190, P < 0 0001). The probability of feeding = exp (A + b * age)/1 exp (A + b * age). The estimated terms in the exponential equation were A chick1 = ± 3 63, A chick2 = ± 4 33 and b = 0 0094. ulna at approximately 70 days, Drummond et al. 1991), the duration of the feeding experiment was su cient to detect substantial e ects of di erent levels of food ingestion on chicks' growth rates. Mass and ulna growth rates varied with treatment and with the exception of ulna growth in the fed group, the magnitude of this variation was commensurate with the variation in food amount. For example, the underfed group received 21% less food than the estimated natural intake, and its rates of mass and ulna increase were 16 and 17% lower, respectively, that in the control group. The fed group received 6% more food than the estimated natural intake, and its rates of mass and ulna increase were 5% higher and 10% lower, respectively, than in the control group. We are unable to explain the di erent growth patterns of ulna length and body mass. We cannot discard the possibility that consumption of non-typical food species or frequent handling altered the growth of hand-fed chicks. However, hand-fed chicks that were receiving food amounts similar to the estimated natural baseline showed rates of mass increase similar to those of chicks fed by their parents. Likewise, a previous study of the blue-footed booby found that, although very frequent weighing (12±20 times per day) slightly depressed the growth of second-hatched chicks, overall the growth of handled broods was similar to that of control broods (Guerra & Drummond 1995). There is, as far as we know, no reason to suspect that our manipulations a ected the sexes di erentially and were responsible for superior growth of females. PARENTAL FEEDING AFTER FLEDGING Greater parental feeding expenditure on daughters than on sons was not apparent during the period of transition to independence. Throughout the observation period, rst-hatched chicks had a higher probability of feeding than second-hatched chicks. However, male and female edglings had similar probabilities to be fed, whether variations due to hatching sequence or age of chick were controlled. We did not continue observations until all the focal edglings dispersed, but the fact that none of the interaction terms were signi cant suggest that males and females may be fed for a similar period of time before independence. Of course, we cannot rule out the possibility that parents transfer a greater quantity of food to daughters than sons at each feeding. SEX RATIO On present evidence, the male-biased progeny sex ratio of the blue-footed booby is not an adaptive response to the relatively higher cost of provisioning daughters (Torres & Drummond 1999). However, biased sex ratios are also expected when the bene ts of raising daughters compared to sons di er. For example, local resource competition or enhancement between relatives may be an important selective

1140 Size dimorphism in the blue-footed booby force for adjusting the sex ratio when there are sex di erences in natal dispersal (Clark 1978; Gowaty 1993). Natal dispersal of female blue-footed boobies is greater than that of males (Osorio-Beristain & Drummond 1993), but we do not know whether parents are di erentially a ected by this. GROWTH, DEVELOPMENT AND SIBLICIDE We need to explain how one sex can grow to a greater size and mass than the other on the same input of food. It has been argued that avian size dimorphism may be related to di erent patterns of growth and development, re ecting di erent strategies of the sexes for using the same quota of resources (Richter 1983; Stamps 1990). However, there is little evidence that the smaller sex generally adopts potentially costly growth tactics, such as accelerated increase in size or plumage development. Growth analysis of dimorphic birds showed that in 13 out of 14 species the smaller sex actually had inferior rates of mass and tarsus increase, and in four out of ve species males and females developed their plumage at similar rates, although the smaller sex appeared to complete the process earlier (Richner 1991). According to Teather & Weatherhead (1994), di erences in growth and plumage development in dimorphic species appear to be an allometric e ect, rather than the result of distinctive inter-sexual strategies of resource allocation. Sexual di erences in resource allocation could result from accelerated neuromotor development of one sex and the increased activity accompanying it (Stamps 1990). In size dimorphic species where siblings compete for parental investment, more rapid motor development of the smaller sex might be favoured to compensate for its inferior size (Newton 1978, 1979; Stamps 1990; Drummond et al. 1991; Anderson et al. 1993a; Teather & Weatherhead 1994), particularly, or perhaps uniquely, if competition involves violent aggression a ecting the probability of survival (Drummond et al. 1991). Accordingly, in the sparrowhawk, a siblicidal species, more rapid motor development of males apparently allow them to compete more e ectively with their larger sisters (Newton 1978). In contrast, in the American kestrel (Falco sparverius) where sibling aggression is apparently absent, greater competitive superiority of females (the larger sex) over male siblings has been reported (Anderson et al. 1993a). In the intense sibling competition of the bluefooted booby, aggressive dominance by the elder chick is established shortly after hatching and even in broods where the younger chick is a female and eventually outgrows her elder brother, the dominance±subordinance relationship is seldom reversed (Drummond et al. 1991; Drummond & Osorno 1992). Aggressive dominance depends initially on relative size and maturity among nestlings, but subsequently it is social experience that in uence the tendency to dominate or submit (Drummond & Osorno 1992; Drummond & Canales 1998). The ability of males to sustain dominance over larger sisters in this siblicidal species could be enhanced by possibly faster neuromotor development of males, which in turn could explain their slower size and mass growth on the same food intake. Acknowledgements We thank the Universidad Nacional Auto noma de Me xico and PADEP (Student Grant to RT) for nancial support, the Secretarõ a de Marina (Armada de Me xico) for logistic support, and the Secretarõ a de Manejo de Recursos Naturales y Ecologõ a (SEMARNAP) for permission. In the eld, we were greatly helped by Miguel Rodrõ guez- Girone s, Cristina Rodrõ guez, Ma. Guadalupe Me ndez Ca rdenas and Francisco Jurado. Ce sar Domõ nguez, Constantino Macõ as Garcõ a, Jose Luis Osorno and Judy Stamps made useful comments on the manuscript. We especially want to thank the sherman from San Bla s and Boca de Camichõ n for providing the food for our experimental chicks and for their friendly support of the project in di erent ways. References Anderson, D.J., Budde, C., Apanius, V., Martinez Gomez, J.E., Bird, D. & Weathers, W. (1993a) Prey size in uences female competitive dominance in nestling American kestrels (Falco sparverius). Ecology, 74, 367± 376. Anderson, D.J., Reeve, J., Gomez, J.E.M., Weathers, W.W., Hutson, S., Cunningham, H.V. & Bird, D.M. (1993b) Sexual size dimorphism and food requirements of nestling birds. Canadian Journal of Zoology, 71, 2541±2545. Anderson, D.J. & Ricklefs, R. (1992) Brood size and food provisioning in masked and blue-footed boobies (Sula spp.). Ecology, 73, 1363±1374. Charnov, E.L. (1982) Sex Allocation. Princeton University Press, Princeton. Clark, A. (1978) Sex ratio and local resource competition in a prosimian primate. Science, 201, 163±165. Clutton-Brock, T.H. (1991) The Evolution of Parental Care. Princeton University Press, Princeton. Clutton-Brock, T.H., Albon, S.D. & Guinness, F.E. (1981) Parental investment in male and female o spring in polygynous mammals. Nature, 289, 487±489. Collopy, M.W. (1986) Food consumption and growth energetics of nestling golden eagles. Wilson Bulletin, 98, 445±458. Crawley, M.J. (1993) GLIM for Ecologists. Blackwell Scienti c Publications, Oxford. Drummond, H. & Canales, C. (1998) Dominance between booby nestlings involves winner and loser e ects. Animal Behavior, 55, 1669±1676. Drummond, H. & Garcõ a-chavelas, C. (1989) Food shortage in uences sibling aggression in the blue-footed booby. Animal Behaviour, 37, 806±819.

1141 R. Torres & H. Drummond Drummond, H., Gonzalez, E. & Osorno, J.L. (1986) Parent±o spring cooperation in the blue-footed booby (Sula nebouxii): social roles in infanticidal brood reduction. Behavioral Ecology and Sociobiology, 19, 365± 372. Drummond, H. & Osorno, J.L. (1992) Training siblings to be submissive: dominance between booby nestlings. Animal Behavior, 44, 881±893. Drummond, H., Osorno, J.L., Torres, R., Garcõ a Chavelas, C. & Merchant Larios, H. (1991) Sexual size dimorphism and sibling competition: implications for avian sex ratios. American Naturalist, 138, 623±641. Fiala, K.L. (1981) Reproductive cost and the sex ratio in red-winged blackbirds. Natural Selection and Social Behaviour (eds R. D. Alexander & D. W. Tinkle), pp. 198±214. Chiron Press, New York. Fiala, K. & Congdon, J. (1983) Energetic consequences of sexual size dimorphism in nestling red-winged blackbirds. Ecology, 64, 642±647. Fisher, R. (1930) The Genetical Theory of Natural Selection. Clarendon Press, Oxford. Frumkin, R. (1988) Egg quality, nestling development and dispersal in the sparrowhawk (Accipiter nisus). PhD dissertation, University of Oxford, Oxford. Gowaty, P. (1993) Di erential dispersal, local resource competition, and sex ratio variation in birds. American Naturalist, 141, 263±280. Guerra, M. & Drummond, H. (1995) Reversed sexual size dimorphism and parental care: minimal division of labor in the blue-footed booby. Behaviour, 132, 479± 496. Lee, P.C. & Moss, C. (1986) Early maternal investment in male and female African elephant calves. Behavioral Ecology and Sociobiology, 18, 353±361. Lunn, N.J. & Arnould, J.P. (1997) Maternal investment in Antarctic fur seals: evidence for equality in the sexes? Behavioral Ecology and Sociobiology, 40, 351±362. McCann, T.S., Fedak, M.A. & Hardwood, J. (1989) Parental investment in southern elephant seals Mirounga leonina. Behavioral Ecology and Sociobiology, 25, 81±87. Nelson, J.B. (1978) The Sulidae: Gannets and Boobies. Oxford University Press, London. Newton, I. (1978) Feeding and development of sparrowhawk nestlings. Journal of Zoology London, 184, 465±487. Newton, I. (1979) Population Ecology of Raptors. T. & A.D. Poyser Ltd, Berkhamsted. Newton, I. & Marquiss, M. (1979) Sex ratio among nestlings of the European sparrowhawk. American Naturalist, 113, 309±315. NunÄ ez-de la Mora, A., Drummond, H. & Wing eld, J.C. (1996) Hormonal correlates of dominance and starvation-induced aggression in chicks of the blue-footed booby. Ethology, 102, 748±761. Ono, K.A. & Boness, D.J. (1996) Sexual dimorphism in sea lion pups: di erential maternal investment, or sex speci c di erences in energy allocation? Behavioral Ecology and Sociobiology, 38, 31±41. Osorio-Beristain, M. & Drummond, H. (1993) Natal dispersal and deferred breeding in the blue-footed booby. Auk, 110, 234±239. Richner, H. (1991) The growth dynamics of sexually dimorphic birds and Fisher's sex ratio theory: does sex speci c growth contribute to balanced sex ratios? Functional Ecology, 5, 19±28. Richter, W. (1983) Balanced sex ratios in dimorphic altricial birds: the contribution of sex-speci c growth dynamics. American Naturalist, 121, 158±171. Rodrõ guez-gironeâ s, M.A., Drummond, H. & Kacelnik, A. (1996) E ect of food deprivation on dominance status in blue-footed booby (Sula nebouxii) broods. Behavioral Ecology, 7, 82±88. Slagsvold, T., Rùskaft, E. & Engen, S. (1986) Sex ratio, di erential cost of rearing young, and di erential mortality between the sexes during the period of parental care: Fisher's theory applied to birds. Ornis Scandinavica, 17, 117±125. Smiseth, T. & Lorentsen, S. (1995) Evidence of equal maternal investment in the sexes in the polygynous and sexually dimorphic grey seal (Halichoerus grypus). Behavioral Ecology and Sociobiology, 36, 145±150. Stamps, J.A. (1990) When should avian parents di erentially provision sons and daughters? American Naturalist, 135, 671±685. Teather, K. (1987) Intersexual di erences in food consumption by hand-reared great-tailed grackle (Quiscalus mexicanus) nestlings. Auk, 104, 635±639. Teather, K. (1992) An experimental study of competition for food between male and female nestlings of the redwinged blackbird. Behavioral Ecology and Sociobiology, 31, 81±87. Teather, K. & Weatherhead, P.J. (1988) Sex speci c energy requirements of great-tailed grackle (Quiscalus mexicanus) nestlings. Ecology, 57, 658±668. Teather, K. & Weatherhead, P.J. (1994) Allometry, adaptation, and growth and development of sexually dimorphic birds. Oikos, 71, 515±555. Torres, R. (1996) Proporcio n de sexos e inversioâ n parental diferencial en hijos e hijas en el Bobo de patas azules. PhD dissertation, Universidad Nacional Auto noma de Me xico, Me xico. Torres, R. & Drummond, H. (1997) Female-biased mortality in nestlings of a bird with size dimorphism. Journal of Animal Ecology, 66, 859±865. Torres, R. & Drummond, H. (1999) Variably male-biased sex ratios in a marine bird with females larger than males. Oecologia, 118, 16±22. Wol, J.O. (1988) Maternal investment and sex ratio adjustment in American bison calves. Behavioral Ecology and Sociobiology, 23, 127±133. Yasukawa, K., McClure, J.L., Boley, R.A. & Zanocco, J. (1990) Provisioning of nestlings by male and female red-winged blackbirds, Angelaius phoeniceus. Animal Behavior, 40, 153±166. Received 14 June 1998; revision received 2 February 1999