Behav Ecol Sociobiol (2005) 59: 115–123 DOI 10.1007/s00265-005-0016-9

ORIGINAL ARTICLE

Diana P´erez-Staples · Hugh Drummond

Tactics, effectiveness and avoidance of mate guarding in the blue-footed booby (Sula nebouxii)

Received: 15 October 2004 / Accepted: 22 May 2005 / Published online: 17 August 2005 C Springer-Verlag 2005


Abstract We examined the dynamics and avoidance of mate guarding, by males and females, in the blue-footed booby, in which the two social mates are usually simultaneously present on the territory but each of them is unmonitored by the other for one-quarter of its time. Both sexes were promiscuous and liable to switch mates. Cuckolded individuals did not increase their overall presence on the territory, but in response to the extra-pair (EP) courtships of their mates, both sexes doubled their rate of intra-pair (IP) courtship and sometimes showed aggression. The male or female’s presence depressed the social mate’s EP activity, but intra-pair courtship had no such effect, tending even to propitiate that EP activity. Similarly, when females responded to their social mates’ EP courtship with approach or aggression, disruption of EP activity was short-lived. Promiscuous females modified their diurnal pattern of attendance, as if attempting to sidestep monitoring by their mates, but cuckolded males matched the modification. Both sexes tended to perform their EP activities at a distance when their mates were present, possibly to evade monitoring or disruption by their mates. Male and female boobies cannot monitor their mates continuously, they do little to facultatively adjust their presence on territory to the risk of infidelity, and their immediate responses to overt infidelity have only the briefest impact; but the information they acquire while monitoring their mates may be critical Communicated by R. M. Gibson D. P´erez-Staples (
) · H. Drummond Instituto de Ecolog´ıa, Universidad Nacional Aut´onoma de M´exico, A.P. 70–275, 04510 D.F. M´exico, M´exico e-mail: [email protected] Tel.: +52-228-8421841 Fax: +52-228-8421800 Present address: D. P´erez-Staples Instituto de Ecolog´ıa A.C., Apartado Postal 63, Xalapa, Ver. M´exico

to constraining their mates’ infidelity and also to calibrating their own reproductive investment. Keywords Mate guarding . Sula nebouxii . Extra-pair activity Introduction In many socially monogamous avian species males guard their mates by remaining on territory (e.g. Westneat 1993) and closely following their social mate (e.g. Birkhead et al. 1992) prior to and during egg laying. However, not all species can guard mates in this way. Colonial birds, for example, depart from their territories and do not seem to follow their mates because they need to perform other activities such as foraging (Birkhead and Møller 1992). Nonetheless, in theory colonial birds can protect paternity by increasing the time spent with their mate, copulating frequently and showing aggression towards other individuals (Birkhead and Møller 1992), and in theory such guarding can be evaded by adopting countermeasures. This study examines the dynamics of such measures and countermeasures in the blue-footed booby, a long-lived colonial marine bird in which both sexes incubate and care for the brood (Nelson 1978; Guerra and Drummond 1995) and both sexes appear to guard their mates (Osorio-Beristain and Drummond 1998). Most studies of mate guarding have focused solely on males, who risk losing paternity if the mate’s extra-pair copulations (EPCs) result in fertilisations (Birkhead and Møller 1992), and it has been doubted that females of socially monogamous species guard their mates (Creighton 2000). However, even though they do not need to ensure their paternity, by mate guarding females could theoretically reduce the risk of quasi-parasitical egg-dumping by their social mate’s extra partner (Davies 2000) or strengthen the pair bond (Petrie 1992), especially if there is a risk of mate switching (Black 1996; Wagner 1991; Stamps 1998). This risk may be especially relevant if EPCs are a form of assessing future partners (Hatch 1987; Colwell and

116

Oring 1989; Slagsvold and Lifjeld 1997). And in principle, females with promiscuous partners may be exposed to more venereal diseases or parasites (Hamilton 1990; Petrie 1992), receive insufficient sperm from their social mates, or suffer from a reduction in paternal investment (Petrie 1992; Hunter et al. 1993). Thus, females may attempt to monopolise males by increasing attendance at the territory and by performing IP activity (Yasukawa and Searcy 1982; Summers 1989; Wagner 1992b; Petrie 1992; Hunter et al. 1993; Eens and Pinxten 1995). When one sex guards, the other may be selected to avoid being guarded. Forms of avoidance include performing EPCs outside the territory (Hatch 1987; Smith 1988; Wagner 1991; Kempenaers et al. 1992; Heg et al. 1993; Neurdorf et al. 1997; Sheldon 1994), moving away from the social mate as egg laying approaches, and having a prolonged receptive period (Davies 1985; Sheldon 1994; Lumpkin 1981; Hatch 1987). However, variation in the tactics, contexts and effectiveness of avoidance is poorly understood. Male budgerigars (Melopsittacus undulatus) restrict their extra-pair (EP) courtship to times when the social mate is unable to observe them (Baltz and Clark 1997), but female western bluebirds (Sialia mexicana) are no more likely to participate in EPCs when the resident male is visually occluded than when he is present (Dickinson 1997). Blue-footed boobies nest on open ground, mostly in dense neighbourhoods where neighbours and social mates are just a few meters apart and highly visible. Female boobies are 32% heavier than males (Castillo and Chavez-Pe´on 1983) and their willing co-operation is apparently necessary for copulation (Osorio-Beristain and Drummond 1998). As laying approaches, the male and female spend more and more time at their joint territory, until they are there for roughly half of the daylight period. At this time, the two social mates are jointly present during roughly 6 h per day and, importantly, each of them is alone on the territory for roughly 2 h per day (Osorio-Beristain and Drummond 1998). Females are probably fertile for 1–5 days before laying each egg, plus an unspecified previous period (Osorio-Beristain and Drummond 1998; Birkhead and Møller 1992). It has been shown that (1) roughly half of paired males and females perform EPCs; (2) females increase their EPCs during their fertile period; (3) male and female EPCs decline in the presence of the social mate, drastically in the case of males; (4) males that are temporarily removed during the female’s fertile period tend to destroy the first-laid egg (Osorio-Beristain and Drummond 1998, 2001); and (5) EP females sometimes dump eggs in the promiscuous male’s nest (Osorio-Beristain et al. submitted). In principle, guarding could be achieved by mere attendance (spending time at the territory) or by intra-pair (IP) courtship and copulation, and it is not clear which of these behaviours actually deters infidelity. Furthermore, although Osorio-Beristain and Drummond (1998) detected no aggressive responses to EP behaviour of social mates, we suspected that boobies may deter EP behaviour by approaching the interactants, and in the present study we therefore analysed approaches. In these events, we observed numerous incidents of not only

approach but also aggression, so we scored both of these behaviours and pooled them because they co-occured. Aggression was scored when the mate directed jabbing and/or yes-headshaking (Nelson 1978) at the rival. We aimed to confirm that both sexes guard, identify the particular behaviours that function as guarding and their particular effects on the social mate, and detect facultative flexibility in guarding. We also looked for evidence of facultative avoidance of guarding by both sexes. Three hypotheses were tested. (1) Individuals respond to EP activity of the social mate by increasing attendance, engaging in intra-pair activity and approaching the rival and mate. (2) Attendance, IP courtship and copulation, and approach suppress EP activity of the social mate. (3) Unfaithful individuals avoid guarding by modifying their attendance and performing EP activity away from the territory. The predictions of each hypothesis are presented in Results. We use the following terminology. An unfaithful individual courts with other individuals in addition to the social mate. A threatened individual is one whose social mate engages in EP courtship. A promiscuous individual copulates with another individual (the rival of the social mate) in addition to the social mate, while a cuckolded individual is one whose social mate is promiscuous. Methods Boobies were observed from January 25 to April 6, 1997, during the colony’s mating season on Isla Isabel, Mexico (21◦ 52 N, 105◦ 54 W). Most subjects in the colony had been previously banded with numbered metal bands. Five unbanded boobies were identified by natural markings on feathers or legs. Eighteen pairs consisting of two social mates each defending a common territory (patch of ground) were observed from three canvas blinds set in a lightly forested part of a dense nesting zone on the northeast edge of the island. Three to six pairs were observed from each blind at any time until a sample of 18 was accumulated. Individuals that courted or copulated with any focal individual were also observed. All banded or marked pairs within 15 m of the blind were observed, provided they did not have an egg, were actively courting, and could be easily observed from the blind. A numbered wooden peg was placed at each focal territory. Observations started daily as soon as light conditions allowed, at 06:00–06:30 h as the season progressed, and ended at 18:00 h when it became dark. At each blind two observers alternated 2-h shifts. Each pair was observed until the female laid an egg that was still intact (not broken) by the end of that day. The number of days each focal pair was observed (observation period) varied from 5 to 38 (12.9±7.5 days). Observers recorded the frequency and time of occurrence of all copulations, noting the identities of both participants and their location in relation to their own territory. They also recorded the identity of all focal individuals courting (see below) with either the social mate or another individual (the mate’s rival) during every 5-min interval (one-zero record). In addition, at the end of

117

each 5-min interval the presence of each focal individual and its rival within the territory or nearby, up to 15 m of their respective territories, was noted. The EP copulations of two males and one female were registered during preliminary observations before we began recording in 5-min intervals. Thus, these mates were considered promiscuous but the EP copulation events were not used for analysis (see Hypothesis 2, prediction a). Courtship occurred when two interacting birds performed skypointing or twiglifting (Nelson 1978; Stamps et al. 2003). Courtship could be simply received by the focal individual, or active; and if active, it was performed by the focal individual without a response from the receiver (non-reciprocal courtship), or performed with an immediate response from the receiver (reciprocal courtship). We recorded a copulation whenever a male stood on a female’s back and either (1) the two birds’ cloacas came into contact or, when these were out of sight, (2) the female rotated and elevated her tail, and the male simultaneously rotated and lowered his tail, and then stepped down to the ground (Osorio-Beristain and Drummond 1998). In the case of EP activity (courtship or copulation with a rival) we recorded: (1) the location of both interactants, (2) the presence or absence of the threatened mate in the observation area (within 15 m of any peg), and (3) the threatened mate’s immediate response (courtship, approach and/or aggression) if any, and its location. Courtship was scored when the threatened mate immediately performed skypointings or twiglifting. Approach was noted if the mate walked towards the individuals engaged in EP activity or entered the territory at any time during the 5-min interval while the focal individual and the rival were still together. Locations were identified as either in the focal individuals’ territory or at an estimated distance from the territory. To check and maintain inter-observer reliability, once per week each observer registered all behaviours alongside another observer for 60 min. During the first 30 min the two observers recorded behaviour jointly, discussing their criteria; during the next 30 min they recorded independently, then compared their records. The exercise was then repeated until there was 90% agreement on the frequency of each behavioural category. Observers were trained to estimate distances by comparing their estimations of distances between objects in the colony with subsequent measurements. Non-parametric tests were used when sample sizes did not allow normalization of data through transformations. For the total proportion of hours that individuals were present at the territory, we used a repeated measures ANOVA after performing an angular transformation on proportions (Zar 1999) which allowed us to test for an interaction between two independent variables. Graphs represent non-transformed data. Two-tailed statistical tests were used even though predictions were directional. Values given in the text are means ± SD. Some analyses were restricted only to prelaying days −1 to −5 so as to increase the sample size of pairs. Because not all individuals received or performed the same types of courtship, sample sizes for individuals engaged in EP courtship vary.

Results In the 18 focal pairs, all 18 males courted other females in addition to their social mate, while only 7 females actively courted another individual, and 11 focal males and 6 focal females were promiscuous. Promiscuous males performed an average of 6.11±4.62 EPCs, at a median of −5 days in relation to laying of their mates’ first eggs (range 0 to −22 days). Promiscuous females performed an average of 12.33±12.93 EPCs, at a median −4 days before egg laying (range 0 to −9 days). Two-thirds of these females copulated EP with only one male (two unpaired, one paired and one male of unknown status). In addition, another female copulated with two unpaired males, and another female copulated with four males (two unpaired, one paired and one male of unknown status). Four promiscuous males copulated EP with unpaired females, while six males copulated with paired females and one male copulated with a female of unknown status, none with more than a single female. Mate switches by males and females were observed (P´erezStaples et al. unpublished data). Hypothesis 1. Male and female blue-footed boobies respond to the EP activity of their mates by increasing attendance, performing IP courtship and approaching. Prediction a: Cuckolded individuals should spend more time at or near the territory than non-cuckolded individuals, as laying approaches. Comparing the average time spent daily by individuals at or near their territory over the 5-day period before egg laying, cuckolded males did not attend longer than non-cuckolded males (cuckolded, n=6; noncuckolded, n=12; Mann–Whitney test, U=30.0, P=0.62), and neither did cuckolded females (cuckolded, n=11; noncuckolded, n=7; U=36.0, P=0.86). Prediction b: Threatened mates should increase IP activity after EP activity has occurred. We compared the frequency of intervals with IP courtship by the threatened individual before versus after a partner’s EP courtship. Whenever an EP courtship occurred, we scored whether an IP courtship occurred during the remainder of that 5-min interval and also during the subsequent 5-min interval. The sum of these two one-zero scores was the “after” score, and the sum of the two scores before EP courtship was the “before” score. When there was a block of consecutive 5-min intervals with EP courtship, IP courtship occurring during these intervals was not taken into account. The whole period of observation for each focal pair was used, and the total one-zero scores across all these days were summed for each individual. Only reciprocal and non-reciprocal IP courtship performed by the threatened individual were taken into account. The average frequency of intervals with IP courtship by the threatened female approximately doubled after a male actively performed non-reciprocal or reciprocal EP courtship (Table 1). For threatened males IP courtship only increased after females engaged in reciprocal EP courtship (Table 1). The sample size for IP copulations occurring before versus after EP copulations was insufficient to test.

118 Table 1 Frequency of intervals with intra-pair courtships by threatened females and males responding to the EP courtships of the social mate

Values are average ± SD. P-values are Wilcoxon test, two-tailed, n=sample size after deleting ties

EP courtship Threatened females Simply received Non-reciprocal Reciprocal Threatened males Simply received Non-reciprocal Reciprocal

IP courtships Before

After

n

T

P

1.21±1.09 3.66±3.44 2.75±3.32

2.00±2.34 7.16±7.33 6.00±4.84

5 12 8

4.0 12.5 1.5

0.87 0.03 0.02

5.80±7.52 4.50±4.93 2.00±1.63

9.80±13.30 6.00±7.54 4.14±3.13

6 6 7

1.0 5.0 2.5

0.06 0.31 0.05

Prediction c: Threatened mates should respond with approach or aggression after their partner’s EP courtship has occurred, whether the unfaithful individual is within or outside the social pair’s territory. Alternatively, if approach/aggression occurs as a response to a territorial threat rather than an infidelity threat, then it should occur only when the unfaithful individual and the rival are within the social pair’s territory. Comparison of the proportion of 5-min intervals with responses (aggression and approach pooled) by the threatened mate to EP courtship when the unfaithful individual was inside the territory versus outside the territory were based on the whole period of observation for each focal pair. When males actively performed non-reciprocal EP courtship, there was no difference between the proportion of courtship events eliciting a response inside versus outside the territory (Table 2). However, when males actively performed reciprocal EP courtship, the social mate was approximately four times less likely to respond if the male was inside the territory than outside (Table 2). That is, females performed presumed guarding responses more frequently when their social mates were performing EP courtship outside the territory. Most of these approaches and aggressions resulted in the rival leaving the territory and the unfaithful mate ceasing its EP courtship temporarily. When females simply received or actively performed EP courtships there was no statistical significant difference in the responses by mates inside or outside the territory (Table 2). Hypothesis 2. Attendance, courtship, copulation, approach and aggression suppress EP activity of the social mate. Prediction a: Extra-pair activity should be less frequent when the threatened mate is present. We compared the rate Table 2 Average (±SD) proportion of responses (aggression and approach pooled) by threatened females and threatened males to EP courtship events of the social mate inside versus outside the territory

EP courtship Threatened females Simply received Non-reciprocal Reciprocal Threatened male Simply received Non-reciprocal Reciprocal

of EP activity that occurred when the threatened mate was present within 15 m of their respective territories or absent. We calculated the frequency of 5-min intervals with EP activity by each individual for the total intervals when both mates were present and also for the total intervals when the individual was present without its mate. The whole period of observation was analysed with the exception of day 0 (when females laid their first egg). We included in the analysis of each category of EP behaviour (simply received, non-reciprocal and reciprocal) only those individuals that ever performed that behaviour. Active performance of EP courtship by males occurred significantly more often when the mate was absent (Table 3), with more than a four-fold increase of nonreciprocal courtship and a doubling of reciprocal EP courtship in the mate’s absence. Females received approximately five times more EP courtship and performed more non-reciprocal EP courtships in the absence of the social mate (Table 3). Male and female EP copulations also appeared to increase more than four times when the mate was absent, but this difference was not significant (Table 3). Prediction b1 : EP activity should be less frequent when the threatened mate is engaging in IP activity than when it is merely present. Data did not support this prediction. Prediction b2 : The probability that an individual engages in EP courtship should increase as time passes after IP courtship (assuming that guarding effects decay over time). Only active courtships were used, for the whole period of observation. There was no clear pattern for intervals with EP courtships to increase in frequency as time advanced after the threatened mate had performed IP courtship (Fig. 1). Nevertheless, variation among successive 5-min intervals was significant (males: Friedman

Inside

Outside

n

T

P

– 0.42±0.42 0.21±0.31

– 0.43±0.42 0.79±0.31

13 10

27.5 5.0

1.00 0.02

0.36±0.38 – 0.59±0.45

0.41±0.39 – 0.27±0.39

7

11.0

1.0

7

15.0

0.31

Values are average ± SD. P-values are Wilcoxon test, two-tailed, n=sample size after deleting ties. (−) Refers to too few events recorded to be analysed

119 Table 3 Proportion of 5-min intervals with male and female EP activity in presence or absence of the threatened social mate

EP courtship

Present

Attendance of threatened female Simply received 0.005±0.007 Non-reciprocal 0.022±0.026 Reciprocal 0.025±0.033 Copulations 0.003±0.006 Attendance of threatened male Simply received 0.013±0.022 Non-reciprocal 0.000±0.000 Reciprocal 0.000±0.000 Copulations 0.001±0.002

Absent

n

T

P

0.012±0.015 0.106±0.106 0.054±0.059 0.013±0.012

7 18 14 9

3.5 5.0 9.0 9.0

0.08 <0.0001 <0.01 0.13

0.062±0.093 0.030±0.070 0.050±0.070 0.027±0.027

13 7 7 5

11.0 1.0 3.0 1.0

0.01 0.03 0.08 0.12

EP courtship divided into three categories: simply received, non-reciprocal and reciprocal courtship. Values are average ± SD. P-values are from Wilcoxon tests, two-tailed, n=sample size after deleting ties. Two promiscuous males and one promiscuous female were not included in this analysis because they performed EP copulations only during previous records when their attendance was not registered in 5-min intervals 0. 0.16

Males

0. 0.12

Probability of Ep courtship

0.08

0.04

0.00 IP

0.08

1

2

3

4

5

6

Females

0.06

0.04

0.02

0.00 IP

1

2

3

4

5

6

Intervals afer an IP courtship

Fig. 1 Probability of (a) males (n=12) and (b) females (n=5) performing active EP courtship following active IP courtship by the social mate. Showing the six 5-min intervals following the interval with IP courtship by the social mate (IP)

repeated measures test, Fr=20.95, n=12, P<0.01; females Fr=14.13, n=5, P=0.03). Surprisingly, EP courtship by both sexes increased in the interval immediately following IP courtship and then declined progressively over the next five intervals (Fig. 1), which could suggest that IP courtship by the threatened mate temporarily propitiates EP courtship. Prediction c: EP courtship should be less frequent after the threatened mate has made some immediate response (courtship, approach or aggression) to the unfaithful mate’s EP courtship, rather than when the threatened mate is merely present. We compared the proportion of 5-min intervals with EP courtship in the interval after the threat-

ened mate had made a response to the unfaithful mate’s EP courtship versus after the threatened mate had made no response and was only present. These proportions were calculated as follows. The frequency of 5-min intervals with EP courtship immediately after an interval with a response was divided by the total number of intervals where a response had occurred. This was compared to the frequency of intervals with EP courtship immediately after an interval with no response, divided by the total number of intervals where both social partners were present and no response occurred. The whole period of observation was used. Only active EP courtship performed when the social mate was present and the responses (courtship, approach or aggression) to this active EP courtship were taken into account. For males, there was a significant fivefold increase in EP courtship after the threatened mate had shown a response compared to when there was no response (males 0.32±0.30 EP courtships after a response versus 0.06±0.05 EP courtships after no response, Wilcoxon test, T=21.0, n=15, P=0.02). For females there was also a pattern for EP courtships to greatly increase after the mate had shown some sort of response, but this was not statistically significant (Wilcoxon test, T=21.0, n=7, P=0.29). Responses to EP copulations were too few to be analysed. Hypothesis 3. Promiscuous individuals evade guarding by modifying their attendance and performing EP activity away from the territory. Prediction a: Promiscuous individuals will overlap with the social mate during a lower proportion of their total attendance time than non-promiscuous individuals. Data did not support this prediction. Prediction b: The temporal pattern of attendance within or near the territory should be different for promiscuous and non-promiscuous individuals. During the 5 days prior to egg laying, the proportion of each hour that males spent within or near their territory did not differ between promiscuous and non-promiscuous individuals (Fig. 2) (repeated measures two-way ANOVA, promiscuity: F1,16 = 1.10, P=0.31; hours: F11,176 =4.07, P=0.00002; promiscuity × hours: F11,176 =0.62, P=0.80). Promiscuous females, however, appeared to stay in the territory less during the

120 Non-promiscuous males Promiscuous males

1.00 Proportion of hour (x, se)

Fig. 2 Proportion of each hour that promiscuous (n=11) and non-promiscuous males (n=7) spent within or near their territories during the 5 days before laying

0.80 0.60 0.40 0.20 0.00 6

7

8

9

10

11

12

13

14

15

16

17

Hour of day

vs. 0.54±0.60 m, n=9, respectively; T=38.0, P=0.07.) Males and females engaged in EP copulations farther from the territory when the threatened mate was present, but this difference was not significant (Mann–Whitney test, U=8.5, P=0.08 males, U=4.5, P=0.11 females). Prediction c2 : Individuals should perform a lower proportion of their EP activity within the territory (vs. outside the territory) when the threatened mate is present rather than absent. We compared the frequency of intervals with EP activity that occurred inside the focal pair’s territory in the presence of the social mate divided by the total frequency of intervals with EP activity anywhere in presence of the mate, to the frequency of intervals with EP activity that occurred inside the focal pair’s territory but in absence of the social mate divided by the total frequency of intervals with EP activity anywhere in absence of the mate. The whole period of observation was used. Both males and females performed a lower proportion of their active EP courtships inside the territory when the threatened mate was present, though this was significant only for males. Males performed 0.52±0.37 EP courtships in the presence of the mate versus 0.80±0.31 EP courtships in her absence, (Wilcoxon test, T=15.0, n=18 minus

mid-morning hours and more during the afternoon hours than non-promiscuous females (Fig. 3), and this was confirmed by a significant interaction between female promiscuity and hour of day (repeated measures two-way ANOVA, promiscuity: F1,16 =0.01, P=0.76; hours: F11,176 =2.88, P=0.002; promiscuity × hours: F11,176 =3.11, P=0.0008). Similarly, cuckolded males emphasized afternoon attendance more than non-cuckolded males (Fig. 4), as shown by the significant interaction (cuckoldedness: F1,16 =0.93, P=0.35, hour: F11,176 =4.04, P=0.00003; cuckoldness × hour: F11,176 =3.15, P=0.0006). If promiscuous females shifted attendance to the afternoon to partially escape the monitoring of their partners, then it seems that males thwarted the manoeuvre by making a similar shift. Prediction c1 : Individuals should perform EP activity farther from the territory when the threatened mate is present versus absent. The average distance at which active EP courtship took place was greater when the threatened mate was present than when absent, but this difference was significant only for males. (Males 1.42±1.53 m, n=16, vs. 0.80±1.01 m, n=18, respectively; Wilcoxon test, T=74.0, n=18 minus 5 ties, P<0.05; females 1.92±1.45 m, n=8,

Non-promiscuous females Promiscuous females

1.20 Proportion of hour (x, se)

Fig. 3 Proportion of each hour that promiscuous (n=6) and non-promiscuous females (n=12) spent within or near their territories during the 5 days before laying

1.00 0.80 0.60 0.40 0.20 0.00 6

7

8

9

10

11

12

Hour of day

13

14

15

16

17

121

Non-cuckolded males Cuckolded males

1.20 Proportion of hour (x, se)

Fig. 4 Proportion of each hour spent by cuckolded males (n=6) within or near their territories during the 5 days before laying

1.00 0.80 0.60 0.40 0.20 0.00 6

7

8

9

10

11

12

13

14

15

16

17

Hour of day

3 ties, P<0.05), for females (Wilcoxon test, T=10.0, n=10 minus 2 ties, P=0.32). Males also performed on average a significantly lower proportion of their EP copulations within the territory in the presence than in the absence of the threatened mate (0.52±0.37 vs. 0.85±0.24, respectively; Wilcoxon test, T=0.00, n=8, P<0.05). Females had a similar response but this difference was not significant (Wilcoxon test, T=0.00, n=6 minus 1 tie, P=0.06). Discussion Female infidelity and guarding by males Promiscuous females copulated with EP males cooperatively, actively seeking EPCs with males near their territories. In a previous study, 7 of 13 females copulated with extra males (Osorio-Beristain and Drummond 1998), and in our sample 6 of 18 females did so. Furthermore, all 13 females in the previous study and 7 in our sample actively courted EP males (see male infidelity below). This behaviour could suggest that while only about half of all pairbonded females engage in EP copulations, most of them could be assessing other males as potential EP partners, prospective social mates (Wagner 1992a) or potential hosts for dumped eggs (quasi-parasitism; Petrie 1986; Emlen and Wrege 1986; Griffith et al. 2004). Such assessing of prospective social mates is supported by observations showing that eight of 11 cases of mate switching involved the EP partner the switcher had been courting or copulating with (P´erez-Staples et al. unpublished data). Attendance by males clearly reduces the EP behaviour of their mates, but we found no sign that males facultatively moderate their attendance in response to their partners’ infidelity. When males were present, their unfaithful female mates received and performed non-reciprocal EP courtships five and 30 times less frequently, respectively, than when the males were absent (cf. Osorio-Beristain and Drummond 1998). But cuckolded males did not overlap

with their mates at the territory longer than non-cuckolded males. Males did respond to the social mate’s EP behaviour, however, with approach and aggression and by courting the mate. They were just as likely to respond with approach and aggression whether the unfaithful female was performing active EP courtship within or outside the pair’s own territory. Hence, approach and aggression towards intruding rivals probably function as guarding rather than simple territorial defence. By responding to the reciprocal EP courtships of their social mates by courting them, males might be expected to deter further infidelity, but our observations contradict this interpretation. Although they disrupted EP behaviour instantaneously, active courtship, copulation and even aggression by males did not depress the EP courtship and copulations of their mates over the next 5 min; if anything, during this time their mates courted the rival more. We therefore suggest that whereas (1) mere presence of males greatly deters their social mates’ EP behaviour, (2) approach/aggression and IP courtship/copulation are largely ineffective for deterring infidelity over the short term, and (3) brief enhancement of IP behaviour by EP behaviour, and enhancement of EP behaviour by IP behaviour, could be non-functional consequences of a general susceptibility to social facilitation. That is, the act of observing a social mate actively engage in courtship, or of actually courting with a social mate, may prime a booby of either sex to court with that individual or other individuals, as a purely proximate effect. Over a longer timescale, males’ immediate responses to EP behaviour (approach/aggression and IP courtship/copulation) could yield a sizeable benefit if they reduce the probability of the mate deserting and pairing up with the rival. That is, a plausible function of those immediate responses is that of shoring up a vulnerable pairbond (cf., Barrows 1995). Certainly, IP courtship is very frequent whenever pairbonded boobies are together (Osorio-Beristain and Drummond 1998), and seems likely to mediate reproductive commitment and coordination.

122

Male infidelity and guarding by females All 18 paired males courted extra females and 61% of them copulated with extra females. The function of the males’ behaviour could have been to initiate a change of social mate in the current season (pair switching) or the next season (divorce), to collude in egg dumping in the males’ own nests (quasi-parasitism; Petrie 1986; Emlen and Wrege 1986; Griffith et al. 2004) or simply to achieve fertilization of a clutch to be laid elsewhere. If the males’ EP behaviour functions to initiate a change of mate or collude in egg dumping, or if it depletes the sperm available for IP copulations, then their social mates stand to gain by guarding the males (Petrie 1992; Stamps 1998). Importantly, we have observed five cases of males switching to the EP partner (P´erez-Staples et al. unpublished data) and six cases of EP females dumping eggs in the nests of promiscuous males that copulated with them previously (Osorio-Beristain et al. unpublished data), showing that two of the putative costs to females of male infidelity are credible. Variation in the guarding behaviours of females and in the effects of such behaviours was highly similar to what we observed for male guarding behaviours. Cuckolded females did not attend the pair’s territory longer overall than non-cuckolded females. In 5 min after their social mates engaged in active courtship with a rival, females doubled their rate of IP courtship and also sometimes approached or attacked, and were twice as likely to do so when their rivals reciprocated their mates’ courtship or when the EP behaviour occurred outside the pair’s territory. And whereas female presence reduced reciprocal EP courtship of males by one half and non-reciprocal EP courtship of males by three-quarters; female IP courtship and copulations had no such effect (except instantaneously), even propitiating male EP courtship during the next 5 min; and female responses to infidelity (approach/aggression, IP courtship) were followed by (non-significantly) enhanced EP courtship by the male. Hence, for females too, it seems that mere presence serves to dampen the male mate’s EP activities, while for this function aggression and IP behaviour are largely ineffective over the short term; that responding to EP behaviour with IP courtship could reflect functionally unimportant social facilitation; and that aggression and increased IP courtship could function over the longer term to buttress the pairbond (cf. Wagner 1992b). Females may be especially vulnerable to abandonment when their social mate is courting with a rival outside the territory, so the pairbonding function might best explain their greater responsiveness in this context. However, we note that the decline in male EP behaviour induced by the presence of the social mate was considerably less than the effect of male presence on female EP behaviour, in the present study and in Osorio-Beristain and Drummond (1998). This could be because male infidelity imposes lower average costs than female infidelity and consequently elicits lesser punishment by their mates (and more minor selective consequences). At the proximate level, it is possible that males perform less EP behaviour when their mates are present, not because they become less

responsive to EP females but simply because much of the males’ time is taken up with IP behaviour, leaving them little opportunity for EP interactions. Avoidance of mate guarding For unfaithful males and females, it is the social mate’s presence that immediately reduces their level of EP activities. The social mate’s responses to infidelity (approach/aggression and IP courtship) only disrupt EP activities instantaneously, however these responses can possibly head off their longer term consequences such as mate switching, egg dumping and divorce. Therefore unfaithful boobies of both sexes are expected to sometimes sidestep their mates’ company and monitoring, and also to perform their EP activities at a distance where they are less readily disrupted. Both of these expectations were supported. First, our observations suggest that promiscuous females (but not males) may attempt to avoid guarding by increasing their mid-afternoon time there at the expense of their mid-morning time; and that cuckolded males negate this avoidance by also emphasizing afternoon attendance. (Of course, other interpretations of these coincident patterns are possible.) Second, in the presence of their social mates males were more likely to perform EP activity off the home territory and farther from the territory, and similar but nonsignificant distancing was evidenced by unfaithful females. Male and female blue-footed boobies guard their mates, and they do this principally by attending the territory, a behaviour that doubtless serves additional functions such as defending the territory against colony neighbours and maintaining/developing the pairbond. Because they have to go on long foraging trips, they cannot continuously monitor their mates on the territory, so the effectiveness of their guarding is constrained. Maybe it is for this reason that unfaithful females and males make little and no effort, respectively, to evade guarding. EP copulations can be reduced but not prevented. However, during the times when both mates are on territory, males (and probably females) carry out their infidelities at a distance where the risk of detection and disruption is probably attenuated and, importantly, both sexes engage in IP activities and cut back substantially on the frequency of their EP activities. In particular, females perform far fewer EP copulations when their mates are present (Osorio-Beristain and Drummond 1998), probably partly because males sometimes respond to a greater risk of cuckoldry by destroying their first-laid eggs (Osorio-Beristain and Drummond 2001). In the booby mating system, it may be critical to accompany one’s mate on the territory not only to minimize and obstruct the mate’s EP activities, and thereby instantly protect paternity (in the case of males), but also to protect the pairbond and acquire information on the mate’s EP activities. Such information could be used by a booby to calibrate its commitment to its mate and its investment in the current reproductive enterprise (cf. Gladstone 1979; Gowaty 1995). And monitoring could deter a mate from engaging in infidelities that might elicit reduction or termination in their partner’s

123

commitment and investment. Information may be the pivotal element in a system where physical prevention of the mate’s EP activities is seriously constrained, and we surely have much to learn about the sorts of information on social mate’s infidelities that boobies can process and how they use that information. Acknowledgements We thank the Mexican Navy for transportation and logistical support and the fishermen of San Blas and Boca de Camich´ın for their generous support and friendship. The Secretar´ıa de Desarrollo Urbano y Ecolog´ıa gave permission to work on Isla Isabel. We are very grateful to: B. Ayala, E. Carrillo, J. Contreras, R. Hern´andez, L. Lomas, A. L´opez, M. L´opez, A. Mancera, J. Meraz, J. Morales, C. Rodr´ıguez, I. Salvador, O. Salvatore, “el Tepo”, C. Wiley and G. Woolrich, for help with field work. This research was financed by a Kathleen S. Anderson Award and an E. Alexander Bergstrom Memorial Award given to D. P.-S. and a CONACyT Grant (4722-N9407) to H. D. M. Osorio-Beristain, C. Macias, J. Osorno, R. Macias, F. Ornelas and C. Cordero provided helpful comments. This article is based on D. P.-S.’s MSc thesis at the Instituto de Ecolog´ıa, UNAM

References Baltz AP, Clark AB (1997) Extra-pair courtship behaviour of male budgerigars and the effect of an audience. Anim Behav 53:1017– 1024 Barrows EM (1995) Animal behavior desk reference. CRC Press, Boca Raton, FL Black JM (1996) Introduction: pair bonds and partnerships. In: Black JM (ed) Partnerships in birds. Oxford University Press, Oxford, pp 3–20 Birkhead T, Møller A (1992) Sperm competition in birds: evolutionary causes and consequences. Academic Press, London Birkhead TR, Clarkson K, Reynolds MD, Koenig WD (1992) Copulation and mate guarding in the colonial yellow-billed magpie Pica nutalli and a comparison with the solitary black-billed magpie P. pica. Behaviour 121:110–130 Castillo A, Ch´avez Peon C (1983) Ecolog´ıa reproductiva e influencia del comportamiento en el control del n´umero de cr´ıas en el bobo de patas azules Sula nebouxii en la Isla Isabel, Nayarit. BSc Thesis, National Autonomous University of Mexico Creighton E (2000) Female mate guarding: no evidence in a socially monogamous species. Anim Behav 59:201–207 Colwell MA, Oring LW (1989) Extra-pair mating in the spotted sandpiper: a female mate acquisition tactic. Anim Behav 38:675–684 Davies NB (1985) Cooperation and conflict among dunnocks (Prunella modularis), in a variable mating system. Anim Behav 33:628–648 Davies NB (2000) Cheating on your own kind. In: Cuckoos, cowbirds and other cheats, Princeton University Press, Princeton, NJ Dickinson JL (1997) Male detention affects extra-pair copulation frequency and pair behaviour in western bluebirds. Anim Behav 53:561–571 Emlen ST, Wrege PH (1986) Forced copulations and intraspecific parasitism: two cost of social living in the white-fronted beeeater. Ethology 71:2–29 Eens M, Pinxten R (1995) Inter-sexual conflicts over copulations in the European starling: evidence for the female mate guarding hypothesis. Behav Ecol Sociobiol 36:71–81 Guerra M, Drummond H (1995) Reversed sexual dimorphism and parental care: minimal division of labour in the blue-footed booby. Behaviour 132:479–496

Gowaty PA (1995) Battles of the sexes and origins of monogamy. In: Black JM (ed) Partnerships in birds. Oxford University Press, Oxford, pp 1–17 Griffith SC, Lyon BE, Montgomerie R (2004) Quasi-parasitism in birds. Behav Ecol Sociobiol 56:191–200 Gladstone DE (1979) Promiscuity in monogamous colonial birds. Am Nat 114:545–557 Hamilton WD (1990) Mate choice near or far. Am Zool 30:341–352 Hatch SA (1987) Copulation and mate guarding in the northern fulmar. Auk 104:450–461 Heg D, Ens BJ, Burke T, Kruijt JP (1993) Why does the typically monogamous oystercatcher (Haematopus ostralegus) engage in extra-pair copulations? Behaviour 126:247–289 Hunter FM, Petrie M, Otonen M, Birkhead TR, Møller AP (1993) Why do females copulate repeatedly with one male? TREE 7:21–26 Kempenaers B, Verheyen GR, Van den Broeck M, Burke T, Van Broeckhoven C, Dhondt AA (1992) Extra-pair paternity results from female preference for high-quality males in the blue tit. Nature 357:494–496 Lumpkin S (1981) Avoidance of cuckoldry in birds: the role of the female. Anim Behav 29:303–304 Nelson JB (1978) The Sulidae: Gannets and boobies. Oxford University Press, Oxford Osorio-Beristain M, Drummond H (1998) Non-aggressive mate guarding by the blue-footed booby: a balance of female and male control. Behav Ecol Sociobiol 43:307–315 Osorio-Beristain M, Drummond H (2001) Male boobies expel eggs when paternity is in doubt. Behav Ecol 12:16–21 Petrie M (1986) Reproductive strategies of male and female moorhens (Gallinula chloropus) In: Rubenstein DI, Wrangham RW (eds) Ecological aspects of social evolution. Princeton University Press, Princeton, NJ, pp 43–63 Petrie M (1992) Copulation frequency in birds; why do females copulate more than once with the same male? Anim Behav 44:790–792 Sheldon BC (1994) Sperm competition in the chaffinch: the role of the female. Anim Behav 47:163–173 Slagsvold T, Lifjeld JT (1997) Incomplete female knowledge of male quality may explain variation in extra-pair paternity in birds. Behaviour 134:353–371 Smith SM (1988) Extra-pair copulations in black-capped chickadees: the role of the female. Behaviour 107:15–23 Stamps J (1998) Territoriality and the non-passive female. In: Gowaty PA (ed) Feminism and evolutionary biology: boundaries, intersections and frontiers. Chapman & Hall, New York, pp 249–319 Stamps J, Calder´on-de Anda M, P´erez C, Drummond H (2003) Collaborative tactics for nestsite selection by pairs of blue footed boobies, Sula nebouxii. Behaviour 139:1383–1412 Summers K (1989) Sexual selection and intra-female competition in the green poison-dart frog, Dendrobates auratus. Anim Behav 37:797–805 Wagner RH (1991) Pair-bond formation in the razorbill. Wilson Bull 103:682–685 Wagner RH (1992a) Behavioural and breeding-habitat related aspects of sperm competition in razorbills. Behaviour 123:1–26 Wagner RH (1992b) Mate guarding by monogamous female razorbills. Anim Behav 44:533–538 Westneat DF (1993) Polygyny and extra-par fertilizations in eastern red-winged blackbirds (Agelaius phoeniceus). Behav Ecol 4:49– 60 Yasukawa K, Searcy WA (1982) Aggression in female red-winged blackbirds: a strategy to ensure male parental investment. Behav Ecol Sociobiol 11:13–17 Zar JH (1999) Biostatistical analysis. Prentice Hall, Upper Saddle River, NJ