Negotiation of parental care when the stakes are high: experimental handicapping of one partner during incubation leads to short-term generosity


Correspondence author. E-mail:


1. Most game theoretical models of biparental care predict that a reduction in care by one partner should not be fully compensated by increased work of its mate but this may not be true for incubating birds because a reduction in care could cause the entire brood to fail.

2. I performed the first handicapping experiment of both males and females during incubation, by placing small lead weights on the tails of male and female northern flickers Colaptes auratus, a woodpecker in which males do most of the incubation.

3. Females responded to the acute stressor (handling and handicapping) by tending to abandon more readily than males and staying away from the nest longer in the first incubation bout. Among pairs that persisted, both males and females compensated fully for a handicapped partner, keeping the eggs covered nearly 100% of the time.

4. Partners did not retaliate by forcing their handicapped mate to sit on the eggs with a long incubation bout length subsequent to having a long bout length themselves. Instead, during the 24 h immediately after handicapping, males behaved generously by relieving handicapped females early.

5. Such generosity was probably not energetically sustainable as these male partners took on less incubation in the 72 h following handicapping compared to female partners of handicapped males. Males and females are probably generous in the short-term because of the high cost of nest failure during incubation but maintaining increased work loads in the longer term is probably limited by body condition and abandonment thresholds consistent with game theory models.


In species with biparental care, there will often be conflict between mates over the level of investment that each should provide to the offspring. Conflict occurs because each parent has limited resources and wants to maximize its own future reproduction by forcing its mate to invest more heavily in the current reproductive attempt (Lessells 1991; Parker, Royle & Hartley 2002). Such sexual conflict has been the focus of numerous mathematical models and empirical studies. The classic ‘sealed bid’ game theoretical model of Houston & Davies (1985) suggested that each parent benefits by exploiting the effort of the partner and so an evolutionarily stable (ESS) level of care occurs when each does not compensate fully for any reduced effort by the mate. Sealed bid models assume that an individual’s effort is fixed during its lifetime and so subsequent models incorporated negotiation in ‘behavioural time’, where each parent could adjust its effort in response to the partner in a series of steps (McNamara, Gasson & Houston 1999). Both types of models; however, predict that partners should not compensate fully when their mate reduces the level of parental care.

Such incomplete compensation has been the ruling paradigm for sexual conflict in parental care until recently. Dearborn (2001) suggested that ‘retaliation’ may occur in which an individual actually reduces its own contribution after observing a decrease in effort by its partner, similar to the iterated Prisoner’s Dilemma game. Jones, Ruxton & Monaghan (2002) emphasized that optimal parental behaviour may depend on shape of the fitness curve. In particular, when the curve resembles a step-function, as may be the case with incubation of eggs, full compensation may result (Fig. 1). Finally, Johnstone & Hinde (2006) incorporated uncertainty of information about the need of the brood in a model of negotiation and found that a whole range of responses (no compensation, incomplete or full compensation or matching) could be predicted depending on the circumstances.

Figure 1.

 Two ways that breeding success (offspring survival) could vary with total parental effort, redrawn after Jones et al. (2002). The solid line shows the decelerating fitness function assumed in most game theoretical models of biparental care. The dotted line shows a situation which could be more typical of incubation; the eggs must be attended at all times, below this level of care, they fail to hatch and breeding success plummets to zero.

Biparental care is common in birds and a dozen or more field experiments on sexual conflict in this taxon have ‘handicapped’ one parent to reduce its level of care in order to observe how the partner responds when provisioning young. Three main techniques have been used: clipping flight feathers (Slagsvold & Lifjeld 1988; Sanz, Kranenbarg & Tinbergen 2000), applying lead weights (Wright & Cuthill 1989) or increasing the level of testosterone in males (Hegner & Wingfield 1987) although the latter may also have undesired effects on aspects of behaviour outside of parental care. The responses to manipulation have been diverse (reviews in Sanz et al. 2000; Johnstone & Hinde 2006). Sometimes partners did not respond at all, suggesting a ‘sealed bid’ model (Schwagmeyer, Mock & Parker 2002). In other cases, mates of handicapped birds compensated partially, consistent with predictions, and rarely they compensated fully or matched the partner (Hinde 2006).

One gap in studies of sexual conflict and negotiation is that few experiments have focused on the incubation period although it may be a reproductive stage with conflict between mates involving energy demands or securing alternate mating attempts (Székely et al. 1996; Barta et al. 2002; Wiebe 2008). Mate removal experiments during incubation have been used to examine costs and benefits of one parent deserting the clutch (e.g. Persson & Öhrström 1989; Wiebe 2005) but few have examined the negotiation process concerning division of labour and the game theoretical model of Jones et al. (2002) has not been tested. A few studies have now experimentally reduced a male’s incubation effort through testosterone implants, finding that females either did not react or compensated only partially (De Ridder, Pinxten & Ens 2000; Alonso-Alverez 2001; McDonald, Buttemer & Astheimer 2001; Schwagmeyer, Schwabl & Mock 2005). Documenting full compensation during incubation as predicted by Jones et al. (2002) remains elusive but these previous testosterone experiments have used species in which the fitness function does not meet the assumptions of the ‘classic’ incubation curve (Fig. 1). As well, no study has reduced the contributions of both sexes during incubation to study the effects of gender roles.

A second gap in empirical field studies of sexual conflict is that few have addressed the process of negotiation between mates in ‘behavioural time’ to understand the multiple cues that partners may use to gain information about each other and the offspring (but see Hinde & Kilner 2007). In contrast to the nestling period, incubation provides a good opportunity to observe a discrete, alternating time series of decisions (incubation bouts) that can be quantified for both partners. The accuracy of information about the partner’s contribution may be higher during incubation than during nestling provisioning because it is easy to tell from egg temperature or direct observation how long the mate has been incubating (Schwagmeyer, Bartlett & Schwabl 2008). For species which cover the eggs almost constantly, the length of an incubation bout is determined by the ‘return-to-the-nest’ decisions of the mate which is currently foraging. Therefore, mates have perfect information about each other’s contributions and quantifying incubation bout lengths as in Dearborn (2001) and Schwagmeyer et al. (2008) may reveal how partners interpret, and react to, changes in their partner’s effort.

Here, my main goal was to test patterns of negotiation assuming a stepped fitness function (Fig. 1). This framework predicts that the mate of a handicapped partner will either simply abandon the clutch if it is below a certain body condition (causing failure of the entire reproductive attempt), or it will fully compensate for the reduced work of their partner (Jones et al. 2002). Unlike the situation of provisioning food, ‘overcompensation’ during incubation is not possible for species when eggs must be covered nearly 100% of the time. I observed the process of negotiation in ‘behavioural time’ by applying lead weights to one member of a pair and measuring subsequent bout lengths immediately post-treatment (short-term negotiation process) and how the proportion of incubation shifted between partners in the longer term as a result of such negotiation. I was particularly interested in whether the mate of a handicapped bird would retaliate in the subsequent incubation bout by staying away from the nest longer. Such retaliation between partners would be evident as a pattern of increasing bout lengths post-handicapping as each tries to make the other sit on the eggs for a longer period. A variety of predictions for bout lengths depending on the response of the partner and on whether incubation bouts are fixed, ‘sealed bids’ or are responsive to the handicapped bird are summarized in Table 1.

Table 1.   Predicted incubation bout and foraging recess lengths according to potential responses of the mate of the handicapped bird (HB) immediately following handicapping. Here it is assumed that the HB stays away from the nest for a longer than normal period in bout 1 immediately following the treatment
Time stepResponse of matePredicted bout length
  1. aMate of HB is sitting on the eggs.

  2. bHB is back on the eggs while its mate now forages.

  3. cIn the broad sense. The mate may forage longer because it actually needs more time to recoup lost energy stores OR because it is ‘punishing’ its mate for perceived negligence.

Bout 1aNo responseSame as control bout. Leaves nest before HB returns (‘sealed bid’ model)
Partial compensationLonger than control bout but still leaves before HB returns
Full compensationLonger than control bout and does not leave eggs before mate returns
Bout 2bNo responseAverage, mate returns to nest after average foraging bout (‘sealed bid’ model)
RetaliationcLonger than control bout: after being forced to incubate unusually long, mate stays away longer
GenerosityShorter than control bout: despite being forced to incubate unusually long, mate returns early

I studied the northern flicker Colaptes auratus, a common woodpecker, with partially reversed sex roles where males do all the nightime incubation and the sexes share incubation during the day. Eggs are attended nearly 100% of the time, incubation bouts vary between 2 and 4 h and bout-switching is accompanied by ritualized vocal and visual displays at the nest entrance (Wiebe 2008; Wiebe & Moore 2008). Females may be polyandrous and so have alternate reproductive opportunities including intraspecific brood parasitism (Wiebe & Kempenaers 2009). Hence, I also predicted that male flickers, rather than females, would be more willing to invest in the current reproductive attempt and would be less influenced by his own handicapping than females. Incubation in flickers seems to fit the pattern in Fig. 1 because all females and most males widowed naturally during incubation abandoned the attempt, suggesting that both parents are normally needed to maintain nearly constant incubation and successful hatching of any eggs (Wiebe 2005).

Materials and methods

Study area and species

Northern flickers have been studied since 1997 in British Columbia, Canada near Riske Creek (51°52′N, 122°21′W). The area covers about 100 km2 of grassland and wetlands interspersed with clumps of aspen trees (Populus tremuloides). Each year, I monitored the reproduction of between 120 and 160 breeding pairs from the time they arrived on territories after migration in late April until young fledged in late July. Clutch sizes range from 3 to 13 on the study area, incubation takes about 11 days and nestlings fledge at 25 days. Both sexes contribute to incubation, brooding and provisioning of nestlings at both monogamous and polyandrous nests (Wiebe & Elchuk 2003; Wiebe & Moore 2008).

Early in spring, potential cavity nests on active territories were checked with a mirror and flashlight until eggs were found and then a small, replaceable ‘door’ was cut into the tree trunk to give access to nest contents and to allow adults to be captured inside the cavity. Flickers are tolerant of such disturbance and trees with doors may be reused for many consecutive years (Fisher & Wiebe 2006a). Annually, over 95% of known breeding adults on the study area were trapped either by blocking the nest hole during incubation, or by pulling a net over the cavity entrance during brood rearing. Birds were weighed, measured, and aged up to 4 years according to molt patterns described by Pyle et al. (1997) and each received a unique colour band combination. For a multivariate measure of body size (Rising & Somers 1989), I used the score on the first axis of a Principle Components Analysis (PCA1) based on six measures: lengths of the wing, bill, tail, tarsus, 9th primary, and bill depth. Separate PCA analyses were done for each sex because of slight sexual size dimorphism. For an index of nutrient reserves or ‘body condition’ that controlled for structural body size, I used the residuals of a regression of PCA1 and body mass.

Handicapping and video recording

The handicapping experiments took place during 2003–2005 and used only monogamous pairs. Whether the male or the female would be handicapped at a given nest was determined by flipping a coin and the target bird was then captured early in the morning. Since males are always in the nest at night and females switch with them for their first incubation bout just after sunrise, the correct sex could be targeted for capture. Following Wright & Cuthill (1989), I glued small, lead fishing weights to tail feathers equivalent to 7–8% of the bird’s body mass which ranged from 130 to 150 g. The weights were distributed on at least four rectrices, and observed on the subsequent video tapes to ensure they did not fall off. Handicapped birds were released outside the cavity. Such handicapping is temporary as the weights drop off when the feathers are molted after the young fledge. Handicapped birds survived and returned to breed in the following year at a rate (20/42 = 47%) similar to the survival of control birds in the population (43%). For the experiment, the lead weights did not need to cause an immediate reduction on the body condition (nutrient stores) of the treated birds. To observe negotiation, it was only necessary that the treated individuals delayed their return to incubation. Mates of handicapped birds were not captured for banding until after data collection to reduce the effect of stress on their behaviour.

Because incubation bouts are long in flickers (2–4 h), conventional video tapes were too short to capture adequate observation intervals. Instead, I used an infra-red and motion-triggered video camera system (Trailmaster Inc., Lenexa, Kansas, USA) that filmed for several minutes when a bird entered or exited the cavity. Parents were allowed to acclimatize for a day to the presence of the tripod and camera placed about 3 m from the nest hole before data were collected. As in Schwagmeyer et al. (2002), I used a repeated measures design to assess how an individual’s incubation patterns changed relative to its previous behaviour and to take into account individual differences in baseline behaviour. Starting on day three or four of incubation, and prior to the lead-weighting, incubation at each nest was filmed for 2 days in order to obtain control (pre-treatment) data on bout lengths and daily contributions of each sex. I averaged the bout lengths and total daily contributions over these 2 days for the ‘pre-treatment’ effort for each male and female after determining there were no significant diel variation in bout length and no significant temporal variation according to the day of incubation (see Wiebe 2008).

Post-treatment video recording began immediately following lead application and usually was long enough to record four subsequent incubation switches before nightfall that first day. Video recording then continued for two subsequent days (i.e. to 72 h post-handicapping). Technical difficulties such as cows or bears knocking over cameras meant that not all nests were filmed for the entire 72 h, so sample sizes differed slightly for the time periods. Bout lengths and the daily proportion of incubation by females were continuous, normally distributed variables so these were analysed with parametric statistics using SPSS (2004). All statistical tests were two-tailed.


Immediate reactions of handicapped birds

I placed lead weights on 23 females and 19 males, hereafter handicapped males (HM) and females (HF), respectively. At these 42 experimental nests, the sexes tended to divide incubation between sunrise and sunset equally in the pre-treatment period, with males contributing an average proportion of 0·50 ± 0·05 SD of the incubation. Including their night time bout, the males performed an average of 67% of total incubation over the entire 24 h period. Prior to handicapping, the average bout lengths of males in the population (2·06 h) did not differ from females (1·91 h; t40 = 1·1, = 0·23) and neither was there a significant difference in bout length between members of a pair (paired t-test: t39 = 1·8, = 0·09). After handicapping, one male, a yearling, (5%) and four females (two yearlings and two older: 17%) abandoned incubation and the nest soon failed as the partner also then abandoned, usually the following day. The proportion of abandonments did not differ between sexes (X2 = 1·45, DF = 1, = 0·22) but the number of abandonments was low. The one male that abandoned had the lowest body condition index in the sample and the mean condition of females that abandoned averaged lower (−10·8) than females that continued (−3·2) but the difference was not quite significant again, perhaps, because of the small sample of abandonments and low power (t-test, t21 = 1·6, =0·12).

Both HM and HF left the vicinity of the nest when they were released after handicapping and so their partners took over incubation duties earlier than they would normally have done so (disturbance at the nest attracted the attention of the foraging mate which returned and resumed incubation within 5 min after the handicapped bird was released). As expected, HM and HF delayed their subsequent return-to-the-nest, forcing their mates to incubate for a longer bout than was typical for them (paired t-test of averaged male bout lengths in the control period vs. his first bout length after his mate received lead, t18 = 5·7, < 0·001; for females with HM: t17 = 2·5, = 0·02). However, HF delayed their return-to-the-nest longer (3·68 h ± 1·3 SD) than HM (2·7 h ± 1·0; t-test t35 = 2·45, = 0·02; Fig. 2). All mates remained incubating until their lead-weighted partner finally returned to relieve them.

Figure 2.

 Mean ± SE lengths of incubation bouts by northern flickers immediately following handicapping by lead weights applied either to the female (n = 19) or male (n = 18) of the pair. In bout 1 post-handicapping, the mate of the handicapped bird is on the nest, in bout 2 the handicapped bird is on the nest etc. Horizontal dashed line marks the average control bout length.

Reactions of partners ‘bouts 2–4’

After handicapping, there was enough time left that day to record four subsequent incubation bouts at most nests before nightfall. There was a significant difference in the pattern of these bout lengths between the HM and HF treatment groups (Repeated measures analysis, interaction term: F3 = 4·41, = 0·006, Fig. 2). Daytime bout lengths declined steadily but slowly at HM nests (post hoc repeated measures Bonferroni contrast, bout 1 vs. 4: F1 = 4·86, = 0·045) whereas there was a steeper decline in bout length during the first day in the HF nests (interaction term, bout 1 vs. 4, F1 = 6·68, = 0·015) and a more dramatic oscillation between long sitting bouts of the male and short bouts of the female.

After the handicapped birds finally returned to the nest after their first unusually long absence, neither their male nor female mates ‘retaliated’ by staying away a long time in the second bout. Instead, the mates of HF returned in a significantly shorter time than the previous bout (paired t-test between bouts 1 & 2: t19 = 4·39, < 0·001, Fig. 2). As a result, the HF were made to incubate for a similar bout length as their control bout length before handicapping (paired t-test between control averaged bout lengths and bout 2: t19 = 1·56, = 0·13). Female partners of HM returned with a bout that was also shorter, but not significantly so, than the time they had been made to incubate in the previous bout (bout 1 vs. bout 2: paired-t18 = 1·12, = 0·27). As a result, HM were made to incubate with a similar bout length as they had prior to handicapping (paired t-test between control male bout length and bout 2: t18 = 0·04, = 0·97).

After the next incubation switch (‘bout 3’Fig. 2), handicapped birds left for their second recess and HF showed greater lingering reluctance to incubate than HM as they stayed away from the nest for a slightly longer time than the HM (2·6 h vs. 2·1 h; t-test: t34 = 1·9, = 0·05). As a result, the male mates of HF were again forced to sit for a longer than normal time (paired t-test between control bout length of male and his bout three: t18 = 3·40, = 0·003). In contrast, female partners of HM incubated with a nearly normal bout length because the HM returned after a normal length of absence (paired t-test between control female bout length and bout three: t19 = 0·73, = 0·47).

In the fourth bout, after having been forced to sit for two unusually long periods by their handicapped mates, male partners of HF returned after only a very short time away (Fig. 2). As a result, HF were made to incubate for a significantly shorter time than their normal bout length in the control period (paired t-test between control bout length and bout 4: t17 = 2·2, = 0·04). A similar, but non-significant pattern occurred with female partners of HM: the females took a relatively short recess such that the incubation bout of their HM was slightly, but not significantly shorter than their normal bout length (paired t-test between control bout length of male and bout four: t15 = 1·85, = 0·09).

Longer term shifts in division of labour

Because the incubation period of flickers is short, it was not possible to monitor behaviour longer than 3 days post-handicapping but most of the shift in division of incubation duties seemed to occur already by the second day (24–48 h post-treatment = period 1, Fig. 3). The fraction of the daytime incubation done by males increased in HF nests and declined in HM nests over 3 days following handicapping (repeated measures, interaction term: F2,34 = 7·16, = 0·003; Fig. 3) but the magnitudes of the shifts were not large. HM decreased their daytime incubation from the control period to the third day post-treatment by 12% (repeated measures: F2,18 = 5·4, = 0·02) while HF decreased their incubation by 9% over the same period, but it was only marginally significant: F2,16 = 2·9, = 0·08). I compared the control bout lengths of individuals prior to handicapping to their bout lengths in period 2 (48–72 h post-treatment). In HM nests, male bout lengths decreased significantly suggesting their female partners were taking shorter recesses to relieve them from incubation earlier (paired t-test: t17 = 2·15, = 0·04) while average bouts of the female partners increased, but not significantly so (t17 = 2·15, = 0·48; Fig. 4). In HF nests, bout lengths of both males and females were on average longer than in the pre-treatment period (Fig. 4) but not significantly so (males: t18 = 1·35, = 0·20; females: t18 =0·67, = 0·51).

Figure 3.

 Mean ± SE proportion of incubation by flicker males during daylight hours at nests where either the female (n = 18) or male (n = 18) was handicapped by applying lead weights. Birds were filmed in all three time periods: pre-treatment (control), and days 1 and 2 following treatment.

Figure 4.

 Mean ± SE changes in average incubation bout lengths by females (= 18) and males (n = 18) at HF and HM nests over the long-term. Changes are shown relative to the pre-treatment (control) bout length for the individual so that positive changes are longer bouts and negative changes are shorter bouts in period 2, 48–72 h post-handicapping.


Compensation during incubation

The fitness function for incubation in flickers seems to match closely the curve in Jones et al. (2002; Fig. 1). The freezing temperatures on the study area together with a suite of predators and nest competitors that regularly threaten unattended cavity nests (Fisher & Wiebe 2006a) probably requires that one parent always incubates (Wiebe 2008). On the time scale of an incubation bout, this meant that a parent never left the nest before being relieved by its mate contrary to a few other species in which incubation effort has been studied. Indeed, during unusually long bouts, flickers sometimes called loudly from the nest hole apparently to encourage their partner to return, yet they did not leave the eggs unless the partner did return to switch with them (Wiebe 2008). With such a stepped fitness function, partners should either compensate fully or abandon the attempt. Consistent with this, among the pairs that persisted with incubation there was full compensation: whatever fraction of incubation the handicapped birds gave up, their mates accepted. In contrast, when the total amount of incubation is flexible and need not be 100%, partial compensation may occur as in house sparrows Passer domesticus (Schwagmeyer et al. 2005, 2008) and some Kentish plovers Charadrius alexandrinus (Kosztolányi, Cuthill & Székely 2009).

Full compensation by flickers meant that care was not based on sealed bids (Table 1). While sealed bid models may be simplistic, Schwagmeyer et al. (2002) suggested that they were consistent with the apparent lack of response to handicapping observed in many other studies of parental care during the nestling-feeding stage and may be more common than currently recognized (but see Johnstone & Hinde 2006; Hinde & Kilner 2007). However, a lack of response to the partner during incubation is risky, if the entire breeding attempt fails if the eggs are not constantly attended. Therefore, one would expect that monitoring the partner’s effort, negotiation, and flexible incubation bouts would evolve to cope with stochastic and acute stressors such as intruders or predators which may temporarily flush a parent off the nest. Flickers rarely forage farther than 1 km from the nest (Elchuk & Wiebe 2003) and loud vocalizations from the incubating partner at the nest probably mean that mates can communicate with each other readily.

Consistent with Jones et al. (2002), abandonment of the attempt by one partner forced the mate to do the same so that the entire clutch was lost. The decision to abandon by the handicapped bird seemed to be made immediately post-treatment and not after a series of negotiated incubation bouts with the partner. The fraction of birds that abandoned was small but there was a trend that more HF than HM abandoned and, in both sexes, the abandoning individuals tended to be the ones in poorest body condition. Even when HF did not abandon, they stayed away from the nest longer than HM during the bouts immediately following handicapping (Fig. 2). This reluctance to incubate was apparently a response to an ‘acute’ stressor rather than a chronic stressor of low body condition since the lead weights likely did not have time to influence nutrient stores in only the few hours following handicapping in a species that forages mainly on the ground for ants. The greater tendency of females to abandon immediately post-handicapping is consistent with the idea that they have greater alternate reproductive opportunities (Wiebe & Kempenaers 2009) and slightly greater annual survival than males (Fisher & Wiebe 2006b) linked to their lower parental effort during the breeding attempt (Wiebe & Elchuk 2003).

There are few comparative data on sexual differences in response to experimental handicapping during the incubation stage in other species. The relative responses of male and female Kentish plovers could not be compared directly (Kosztolányi et al. 2009). During the nestling period; however, male house sparrows showed more dramatic short-term impairment than females, similar to the situation in great tits Parus major, where handicapped females did not reduce provisioning rates but lost body condition (Sanz et al. 2000) suggesting that females had greater investment in the current attempt. However, other studies have not found any difference between responses of the sexes during the provisioning period (Wright & Cuthill 1989).

The process and cues of negotiation

There is no ESS according to Jones et al. (2002), but selfishness during biparental incubation is limited because the partner that is forced to work too hard will simply abandon as its body condition declines, to the detriment of both. Assessing the commitment of the partner to the current reproductive attempt (the ‘abandonment threshold’) is therefore critical, but little is known about how information is transferred between mates during incubation. Interestingly, the mates of both HM and HF did not retaliate in the second bout, but foraged for a ‘typical’ length of time with the result that the handicapped birds were made to incubate for a typical time in bout two. Schwagmeyer et al. (2008) found that subsequent bout lengths in house sparrows P. domesticus were uncorrelated but Dearborn (2001) found that, consistent with retaliation, a long incubation bout in great frigate birds (Fregata minor) was followed by another long bout. In flickers, any energetic costs of the unusually long incubation bout were absorbed by the unhandicapped partners. These costs happened to be greater for partners of HF, because HF avoided returning to the nest for a longer period.

Perhaps the most interesting pattern with the series of bout lengths was the difference in response of mates during in the second compared to the fourth bouts (Fig. 3). When the mates of handicapped birds were made to sit for an unusually longer bout for the second time after handicapping (i.e. the third bout), they responded by returning to the nest after a shorter than normal time away (males) or nearly so (females). This was remarkable because, despite the probable mounting energy costs of two reduced foraging sessions, the unhandicapped partners returned ‘voluntarily’ to relieve their partners early. This apparent generosity seems counter to principles of conflict and negotiation but suggests that the unusual behaviour of the handicapped birds for two consecutive incubation bouts was a reliable cue of the fragile motivation of the partner to continue with the breeding attempt. Therefore, in the short-term, and probably in response to an acute stressor, flickers of both sexes reacted generously to reduce the chance of abandonment by the partner. Handicapped birds seemed to easily win concessions from their mates during incubation but theory suggests that in the longer term, energy constraints on both sexes dampen individual selfishness (Jones et al. 2002). It is difficult to say how quickly the lead weights influenced foraging and body condition but average shares in incubation duties shifted over the subsequent 72 h and consistent with the probable greater cost to the body condition of HMs.

The shifting patterns of bout lengths over the 72 h also suggest different patterns of negotiation between HM and HF groups (Fig. 4). Cooperation is suggested in HM pairs, where the significantly shorter bouts of males result from their female partners continuing to relieve them ‘early’. Greater conflict is suggested in HF pairs where bout lengths of both sexes averaged longer, though not significantly so, than control bout lengths. Here, the greater daily contribution by males seemed to occur by HF taking longer recesses, forcing long bouts on their males. Male mates of HF seemed unwilling or unable to relieve their mates early, continuing to take average or slightly longer than average recesses themselves. This apparent lack of generosity by males 3 days after handicapping was in remarkable contrast to their apparent altruism and generosity the day handicapping occurred (Fig. 2).

Short-term and long-term stressors and a new model

There is little comparative data on behavioural negotiation between parents over time. In flickers, there was a fascinating difference in the way the sexes reacted to handicapping in the short- vs. long-term. Whereas HF avoided incubation more in the bouts immediately following capture (Fig. 2), after 72 h it was the bout lengths and daily contribution of the HM and not the HF which were significantly reduced (Fig. 4). Thus, females were more vulnerable to short-term stressors but more resilient to the probable longer term costs caused by lead weights to body condition. Interestingly, in the control population, the proportional contribution of male flickers to incubation was positively correlated with their body condition but it was not the case for females (Wiebe 2008). This suggests that males, with their greater overall share of incubation caused by their sole responsibility at night, are working closer to their energetic limit than females, and that the division of labour during incubation is closely tied to, and negotiated based on, the male’s energy reserves. In contrast, female great frigate birds performed most of the incubation and it was the bout lengths of females that were more responsive to body condition than bout lengths of males (Dearborn 2001).

The greater willingness of males to cope with acute stressors in the short-term but their lesser ability to cope with chronic energetic stress in the long-term suggests the sexes have different abandonment thresholds depending on the type of stressor. In Jones et al. (2002), a single abandonment threshold was based on an individual’s body condition. With flickers, the magnitude of the acute stressor treatment (handling times and lead weights) was the same for males and females, but females reacted more negatively, implying their abandonment threshold for acute stressors was at a higher level than the males (Fig. 5a). Flexible abandonment thresholds are suggested by the response of the male mates of HF which were very generous in the short-term but probably were operating temporarily in a zone of energy deficit in the bouts immediately following handicapping (Fig. 5b). However, males are actually less able to cope with such a deficit in the longer term because they are closer to the abandonment threshold based on body condition (even in a control situation they incubate more than females). Over time, therefore, males reduce their level of compensation until they once again can maintain body condition above the ‘long-term’ abandonment threshold.

Figure 5.

 The body condition (survival) curves for male and female partners are shown according to the proportion of incubation the male does (redrawn after Jones et al. 2002). (a) Shows the immediate reactions if either a male or female flicker is handicapped at a nest. Prior to handicapping, the male performs a greater proportion of incubation than the female so his survival (open circle) is slightly lower than the female’s (solid circle). After the acute stressor and before body condition is affected by the lead weights, the male or female perceives their future survival prospects at the current nest site to be lower (arrows). However, females have a higher abandonment threshold than males and so they abandon more readily or reduce their contribution more than males do. (b) (Bottom) shows longer term responses of partners of handicapped birds as they negotiate based on their body condition. Here the sexes share a common abandonment threshold based on their body condition. The circles show, what either the male mate of a handicapped female does (open circles) or what the female mate of a handicapped male does (filled circles) in response to the reduced effort by their handicapped partners. At first males compensate fully even to the point of energy deficit, but this point below the abandonment threshold is unsustainable. Thus, over time he will negotiate a share of incubation above the abandonment threshold after the acute stressor has passed (dashed arrow back above the threshold) leading to a workload division that is not much different from what it was previously. In comparison, if a male receives the lead weights, his female partner has a greater energy buffer (solid circle farther above the abandonment threshold) and so she is better able to absorb costs and therefore negotiation over time more easily leads to increased participation of the female in incubation.

In sum, flickers showed full compensation and short-term generosity when reacting to a decreased work effort by their mate during incubation. This is counter to many previous models where incomplete compensation is an ESS, but the high cost of complete nest failure may force partners into a ‘cruel bind’ and limit selfishness (Jones et al. 2002). The reactions of males and females differed in magnitude and changed in the short vs. longer term following handicapping. Future work should focus on individual variation in reaction to short vs. longer term stressors during incubation, and how individuals assess or communicate abandonment thresholds with their mate.


Helpful comments by Tore Slagsvold and an anonymous reviewer improved the paper. Thanks to numerous students over the years for helping to find and monitor flicker nests in the field. This study was funded by and NSERC discovery grant and the experiment was conducted with permission from the University of Saskatchewan Animal Care Committee (permit # 20010113).