Better together
Groenewoud, Frank
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from
it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2018
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Groenewoud, F. (2018). Better together: Cooperative breeding under environmental heterogeneity.
University of Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Chapter 7
Synthesis
In this thesis, I have addressed the environmental factors affecting the formation of groups, the benefits of helping by subordinates and group stability in cooperatively breeding Sey-chelles warblers and Neolamprologus pulcher. In the SeySey-chelles warbler, most studies had focused on the impact of differences in territory quality on delayed dispersal and the ben-efits of helping by subordinates, but temporal variation in food availability (chapter 3) and predation risk (chapters 6 and box A) had not yet been recognized as drivers of sociality in this species. In N. pulcher, experimental studies had shown changes in helping behavior and dispersal as a result of manipulated predation risk, but consequences of natural variation in predation risk between populations (chapter 2) and the implications for within-group conflict (chapter 5) were unknown. I will discuss these and other findings in relation to ecological drivers of cooperative breeding, further below.
DELAYED DISPERSAL
The evolution of cooperative breeding can be approached as a two-step process where group formation is a necessary, but not a sufficient, first step enabling individuals to jointly raise offspring. In many cases, these additional caregivers will be offspring from previous breeding attempts that have delayed dispersal, but other routes to group formation and cooperative breeding are also possible (see chapter 4; Riehl 2013). It has been suggested that help by unrelated subordinates is of a secondary nature – i.e. help by unrelated individuals only became possible after direct benefits became available for related subordinates that initially provided care for kin (Cockburn 2013). Several lines of evidence, such as phyloge-netic analyses (Cornwallis et al. 2010; Lukas & Clutton-Brock 2012) and the prevalence of family vs non-family groups (Riehl 2013), seem to point in the direction that this was in fact the evolutionary route to cooperative breeding for the majority of species. In this view, direct benefits of helping have played a lesser role during the initial evolution of family liv-ing, in the same way as they might do now in determining the benefits of delayed dispersal for subordinates (Drobniak et al. 2015; Griesser et al. 2017). I would add two exceptions to this scenario, for cases where unrelated subordinates either (i) provide and obtain direct (passive) grouping benefits, such as through group augmentation (Kokko et al. 2001; King-ma et al. 2014), or (ii) where helping behaviour is not strictly “altruistic”, such as territory defense or other behaviours, which are partly self-serving (e.g. Tanaka, Frommen & Kohda 2018). In these cases, it is easier to envision a situation where group formation by unrelated individuals, without evolved helping behaviours in place, is likely to occur. Nonetheless, in many cases, it seems that the factors that are currently maintaining group formation and cooperative breeding are not the same as those responsible for its initial evolution. Thus, it seems justified for researchers interested in the evolution of cooperative breeding, to
inves-tigate the factors leading to delayed dispersal, that are independent from the benefits that individuals can obtain by helping.
The evolution of delayed dispersal has been attributed mainly to combinations of habitat saturation and benefits the can be obtained by remaining in the natal territory (Emlen 1982, 1994; Stacey & Ligon 1991). Most studies investigating delayed dispersal in cooperative breeders have focused on stable environments, where habitat saturation creates a shortage of breeding opportunities, and where spatial variation in territory quality generates the conditions necessary for offspring to benefit by delaying dispersal and foregoing indepen-dent breeding (Emlen 1982; Koenig et al. 1992; Komdeur 1992). However, recent studies show that cooperative breeding in mammals (Lukas & Clutton-Brock 2017) and birds (Rubenstein & Lovette 2007) is more prevalent, rather than less, in areas with high temporal variability. The extent to which this pattern is due to changes in delayed dispersal and the propensi-ty to form family groups (i.e. “hard life” hypothesis; Koenig et al. 2011; Griesser et al. 2017), or to changes in the benefits of helping behaviour by subordinates (i.e. “bet-hedging” or “temporal variability” hypothesis; Rubenstein 2011; Shen et al. 2017), is unclear. In chapter 2, I show that in the Seychelles warbler, delayed dispersal and group formation is associ-ated with spatial and temporal variation in food availability. Offspring are more likely to disperse from their natal group at one-year of age when food availability – and therefore the conditions for dispersal and independent breeding – is favorable, and the benefits of remaining as a subordinate in the natal territory are perhaps less important. Increased lev-els of dispersal and breeding as a result of improved conditions have also been found in other cooperative breeders, such as azure winged magpies Cyanopica cyanus (Canario et al. 2004), acorn woodpeckers Melanerpes formicivorus ( Koenig & Walters 2011) and experimen-tally in sociable weavers Philetairus socius (Covas et al. 2004). However, the extent to which the relationship between decreased dispersal and adverse ecological conditions is due to increased costs of dispersal (and breeding alone), or increased benefits of natal philopat-ry and breeding together (i.e. cooperative breeding arises through mutualistic benefits) is mostly unknown. Chapter 3 and box A suggest that, in the Seychelles warbler, reduced dis-persal is both the consequence of increased costs of disdis-persal due to low food availability and the benefits that females can obtain by breeding together under high risk of predation. Predation risk has received very little attention as a driver of complex sociality and coopera-tive breeding. However, this view neglects potentially important effects of predation risk on the costs of dispersal, the survival benefits of group living and the benefits of cooperation for reproductive success, which could all play a role in delayed dispersal. In N. pulcher, large subordinates are sexually mature and can potentially disperse to vacant habitat, which is plentiful at colony borders, to breed. However, an earlier study has shown
experimental-133 SYNTHESIS
ly, that individuals do not always disperse to such areas and that this is due to predation risk acting as an ecological constraint preventing dispersal (Heg et al. 2004a). In chapter 1, I show the consequences of natural variation in predation risk between different popu-lations of N. pulcher on group composition and behaviour, which corroborates this earlier experimental study and further shows the importance of predation risk for the evolution of sociality in this system. Predation risk is also thought to play an important role in the Seychelles warbler, as egg predation by the endemic Seychelles fody Foudia sechellarum is the main cause of nest failure (Komdeur & Kats 1999). In chapter 6, I show experimentally that Seychelles warblers have evolved effective behavioural strategies to prevent egg preda-tion by fodies, and that particularly dominant males are responsive to changes in predapreda-tion risk. However, egg predation by fodies only occurs when nests are unprotected, and given that subordinate females participate in incubation, this too can be an effective strategy to reduce egg predation. In box A, I present a study into the effect of predation risk as a driver of delayed dispersal and cooperative breeding in the Seychelles warbler.
THE BENEFITS OF BREEDING TOGETHER
Limited dispersal and kin selection – i.e. selection on genes through its effects on others carrying the same gene (Hamilton 1963) – has been the predominant explanation of altru-istic helping behaviour in coope rative breeding species (West-Eberhard 1975; Cockburn 1998; Foster et al. 2006). One of the appealing characteristics of this scenario, is that such selection is the automatic consequence of limited dispersal (due to e.g. ecological con-straints) leading to kin neighbourhoods (but see Platt & Bever 2009). The importance of kin selection for the evolution of helping behaviour in cooperative breeders has been well documented, both between (e.g. Hughes et al. 2008; Cornwallis et al. 2010; Briga et al. 2012) and within (e.g. Emlen & Wrege 1989; Komdeur 1994b; Wright et al. 2010) species. Despite this observation, cooperation between unrelated individuals in common, and some argue that the direct benefits to subordinates in cooperative breeding species are sufficient to maintain cooperation and that indirect benefits have been overestimated (Clutton-Brock 2002; Riehl 2013). Direct benefits of cooperation can be obtained by various means such as when cooperation boosts group reproductive success and there are benefits to being in a larger group (i.e. group augmentation; Kokko et al. 2001), or when cooperation improves the reproductive success of all group members (i.e. mutualism; Clutton-Brock et al. 2001). However, environmental conditions can alter or even generate the potential of individuals to obtain direct and indirect fitness benefits through cooperative breeding, when group-liv-ing or cooperation functions to overcome some ecological or environmental obstacle (e.g. harsh abiotic conditions or high predation risk). Quantifying the costs and benefits of
co-operation in relationship to such conditions is thus important to explain the evolution and maintenance of cooperative breeding.
In chapter 3 and 4, I show that there are mutualistic benefits of helping and co-breeding in the Seychelles warbler. In chapter 3, I show that per capita reproduction does not decrease in groups with female subordinates compared to Seychelles warblers that breed in pairs, which means, at least for female subordinates, cooperative breeding is not necessarily cost-ly in terms of reproductive output. The exact reproductive benefits for subordinate females living in groups will depend on the level of reproductive sharing and within-group related-ness (b and r in Hamilton’s rule; Hamilton 1963). Furthermore, groups that contain female subordinates have lower fecundity variance, indicating more consistent reproduction, which can be an important aspect of fitness, especially under the conditions that are com-mon for many cooperative breeders (e.g. small, kin-structured populations). The extent to which fecundity variance (i.e. bet-hedging) contributes to fitness in the Seychelles warbler, or other cooperative breeding species, remains to be discovered (Rubenstein 2011; Koenig & Walters 2015). Estimating fitness in fluctuating environments is challenging (Sæther & Engen 2015), but recently developed statistical methods have been able to assess the (long term) genetic contribution of social traits in fluctuating environments (Engen et al. 2009; Sæther et al. 2016). However, our study shows that breeding together might offer fitness gains for females, irrespective of the constraints that are placed on dispersal and indepen-dent breeding. This conclusion is also supported by chapter 3, where I investigate the prox-imate and ultprox-imate factors associated with between-group dispersal by natal subordinates (i.e. subordinates joining a non-natal group as a subordinate) and show that mostly females are allowed to immigrate into another group. In the Seychelles warbler, eviction is thought to be a common phenomenon (e.g. Eikenaar et al. 2007; Kingma et al. 2016a; chapter 4 of this thesis) and groups are thus expected to only allow outsiders to join the group when such outsiders provide net benefits to the fitness of insiders (Shen et al. 2017). Since a previous study in the Seychelles warbler has shown survival costs of being in a larger group (Brouwer
et al. 2006), groups allowing unrelated immigrant females to join, might be doing so for
reproductive benefits.
Recent comparative studies have suggested that cooperative breeding is more common in regions with high environmental variability in temperature and rainfall (Jetz & Rubenstein 2011; Lukas & Clutton-Brock 2017). Such patterns have been explained primarily by posing that subordinates make reproduction possible under adverse conditions (i.e. “temporal variability” or “hard-life” hypothesis; Rubenstein & Lovette 2007; Koenig et al. 2011). How-ever, in chapter 3, I show that annual reproductive success is unaffected by temporal pat-terns in food availability, for pairs as well as groups with subordinates, but that groups with
135 SYNTHESIS
subordinate females do have higher reproductive success that those without. This result suggests that either food availability never drops under levels where it affects provisioning rates or that birds incur additional costs under low food availability. This would also mean that help by subordinates would be more important to reduce the costs of provisioning for dominant breeders when food availability is low (e.g. in terms of survival) than for repro-ductive success. However, an earlier study has shown that survival of dominant breeders actually declines in larger groups (Brouwer et al. 2006). This conflicting result might be explained by a preliminary analysis, which shows that subordinate females are more likely to help in years when food availability is high rather than low (Fig. 7.1), indicating that the costs of help might increase under low food conditions.
Female (R = 0.77*) Male (R = 0.09)
−0.6 −0.2 0.2 0.4 0.6 Log standardized food availability 0.0 0.2 0.4 0.6 0.8 1.0 Probability subo rdinate helps −0.4 0.0
FIGURE 7.1 The probability that a subordinate female (red) or male (blue) helps in relation to annual food availability. Solid and
dashed lines represent model predicted means ±95% CI, respectively, and points indicate the proportion of subordinates helping in each year.
In chapter 5, I show that in N. pulcher the benefits of having subordinates only become ap-parent under elevated risk of predation: territories are more likely to contain small fry or juveniles (indicative of successful reproduction) when they have more large or small sub-ordinates, but only in high predation risk populations. In low predation risk populations, reproductive success is higher overall, and independent of the number of subordinates. Previous studies have pointed out the importance of both large and small subordinates for the protection and survival of other group members (Balshine et al. 2001; Heg et al. 2004a; Groenewoud et al. 2016)
GROUP STABILITY
Conflict between individuals living in social groups is common, because individuals often compete for limited resources (e.g. food or reproduction) and fitness is rarely perfectly aligned. As such, conflict can be a repellant force in the evolution of sociality and diminish or even negate the benefits of group-living or cooperation (West et al. 2015). However, in-dividuals should invest less in competition when the benefits of cooperation (or the costs of competition) are higher. Understanding the conditions under which conflict is reduced can therefore give insights into the benefits of cooperation. In N. pulcher, conflict can occur on different levels of organisation: individuals within groups compete for limited resourc-es such as reproduction, group membership and status, and between groups, individuals compete mainly for space and shelters needed for breeding. However, individuals living un-der higher density (both larger groups and more neighbours) also benefit from increased protection (Jungwirth & Taborsky 2015; chapter 2 and 5). In chapter 5, I show that with-in-group conflict is reduced in groups living under high risk of predation: dominants are less likely to show aggression towards subordinates and subordinates are less likely to show aggression towards each other. However, aggression towards neighbouring groups does not change with increasing predation risk. This study supports and adds to the findings of chapter 2, which show that groups under high predation risk are more likely to contain large subordinates, either through delayed dispersal (Heg et al. 2004a) or through immi-gration (Bergmüller et al. 2005a; Zöttl et al. 2013). While the findings in chapter 3 were ini-tially attributed to increased costs of dispersal, reduced aggression might also function to incentivize large subordinates to stay in the group when dispersal decisions are influenced by such aggressive interactions. To our knowledge, this study is the first to show reduced social conflict in response to predation risk (but see Shen et al. 2012). The function of aggres-sive interactions between dominants and subordinates in N. pulcher has been attributed to at least two different functions, i.e. as punishment to subordinates to provide more help (Gaston 1978; Fischer et al. 2014; Quiñones et al. 2016) and as a way to suppress subordinate reproduction (Fitzpatrick et al. 2005; Heg & Hamilton 2008). If aggression functions mainly to control reproduction, reproduction by subordinates should increase under higher levels of predation risk, when the benefits of anti-predator defense by such subordinates is high-est. Interestingly, concession models of reproductive skew predict that dominants should be less inclined to concede reproduction to subordinates when ecological constraints are more severe (i.e. predation risk is higher), or that alternatively, that subordinates might try to claim a smaller share of reproduction, because the costs of expulsion from the group are higher (Reeve & Keller 1997; Johnstone 2000). To square this with our current data, this would have to mean that (i) the benefits of having larger subordinates in the group for
137 SYNTHESIS
dominants increase more with predation risk, than the extent to which predation risk is an ecological constraint to large subordinates, or that (ii) dominants are less likely to show ag-gression to large subordinates because they do not try to reproduce under high predation risk. Future studies interested in the function of aggression and the extent to which preda-tion risk is an ecological constraint to the dispersal of larger subordinates should include measures of reproductive skew, to test for such effects.
In the Seychelles warbler, groups can remain stable over multiple years, and aggressive in-teractions between group members are only seldom observed. Nevertheless, indirect evi-dence suggests that evictions are common and are mostly the result of the death and sub-sequent replacement of dominant breeding males (Eikenaar et al. 2007; Kingma et al. 2016a; Groenewoud et al. 2018). In chapter 3, I have shown that changes in annual food availabil-ity can also contribute to group stabilavailabil-ity: subordinates were more likely to disperse from their natal territory in years where food availability was high. Since reproductive success in those years (for pairs or for groups with subordinates) was not higher, this pattern is best explained by lower costs of dispersal and/or prospecting (Kingma et al. 2016b). This result adds on previous work by Komdeur (1992) who showed that subordinates were more likely to delay dispersal in high quality, rather than low quality territories. Additionally, in box A, I show that nest predation risk likely creates mutualistic benefits of communal breeding and is therefore an important driver of group formation in the Seychelles warbler. Such benefits are also supported by chapter 3 of this thesis, which show reduced reproductive variance for groups containing female subordinates, but no decrease in mean per capita reproduction. Theory predicts that given two individuals with exactly the same mean re-productive success, the one with the lowest rere-productive variance, will have the highest fitness (Starrfelt & Kokko 2012). Thus, females that stay together and breed communally may outperform those that breed in groups with only a dominant breeding male present, espe-cially under high risk of nest predation (see chapter 6).
CONCLUSIONS
In conclusion, I would like to make three important observations. First, I think that preda-tion risk has been undervalued as a driver of cooperative breeding and transipreda-tions to social complexity (chapter 2, 5 and box I). Obtaining accurate estimates of the risk of predation can be challenging, especially in avian systems, but simple proxies of predation risk (e.g. predator densities) can sometimes be sufficient. In addition to the well documented effects of predation risk of grouping (Krause & Ruxton 2002), predation risk can also affect the costs and benefits of delayed dispersal, the division of labour and cooperation in rearing
young. All of these are important aspects in the transition to complex sociality and play a crucial role in the evolution and maintenance of cooperative breeding (West et al. 2015). Second, while it has been suggested that cooperative breeding could function as a bet-hedg-ing strategy – partly because it occurs at higher frequencies in geographical regions with high temporal variation in environmental conditions (Rubenstein & Lovette 2007; Lukas & Clutton-Brock 2017) – I argue that his conclusion could be premature. Our data shows that while years of low food availability cause offspring to delay dispersal and remain as subor-dinates in the natal territory, cooperative breeding does not act as buffer to prevent low reproductive success in bad years (chapter 3). Thus, high temporal environmental variation could have facilitated the transition to cooperative breeding by leading to group formation and family living (Griesser et al. 2017), but not be responsible for the benefits of collective-ly rearing offspring. Third, where many studies have concluded that delayed dispersal and helping by subordinates is making the best of a bad job, I would argue that these conclu-sions may change if the ecological and environmental conditions under which breeding occurs are taken into account. Subordinate females in the Seychelles warbler females might have higher fitness by breeding together under increased risk of predation, but more de-tailed studies of individual reproductive success and the importance of reproductive vari-ance are necessary, to demonstrate this conclusively. Similarly, large groups are necessary for successful reproduction in N. pulcher, and the extent of reproductive sharing under high risk of predation, would therefore determine if individuals might do best by breeding co-operatively. Together, these results show that the costs and benefits of cooperative breeding can only be determined, if one considers the ecological and environmental conditions un-der which cooperative breeding occurs.
139 SYNTHESIS
