Per capita reproductive success decreases with group size in a communally breeding bird

doi:10.21203/rs.3.rs-5105936/v1

The benefits of cooperative breeding include anti-predator defense, access to resources, and inclusive fitness via kin-selection. Smooth-billed anis (Crotophaga ani) are communally breeding birds in which co-breeding females lay eggs in a shared nest. Within-group relatedness is low, so kin selection is not driving this system. Group size ranges widely, and larger groups often exhibit intense ovicide, suggesting it is costly to breed in large groups. Assuming there are tradeoffs between group size and reproductive success, we predicted that with increasing group size i) group reproductive success would increase, while ii) per-capita reproductive success would decrease, and iii) the probability of offspring surviving to a subsequent developmental stage would increase due to the presence of additional caregivers. Using data from 364 nests, we show that larger groups laid, incubated, and hatched more eggs and fledged more chicks, whereas per capita reproductive success decreased with increasing group size across all four measures. Group size did not affect the probability of offspring surviving to a subsequent developmental stage, and anis frequently breed in larger-than-optimal groups. We conclude that communal breeding in large groups is likely driven by ecological constraints, factors that increase long-term fitness of individuals, and conflicting selection pressures among individuals trying to join groups and those already in groups. Our findings highlight the complex selection pressures that likely influence communal breeding in non-kin groups with highly variable group structure.

Communal breeding

Crotophaga ani

seasonal reproductive success

group size

Cooperative breeding is often driven by inclusive fitness benefits, ecological constraints, or both. We evaluated short-term fitness benefits in smooth-billed anis, a facultative communal breeding bird with variable group sizes and low within-group relatedness. Previous work in this system showed that significant reproductive effort is wasted, particularly within large groups, because competing females destroy the eggs of co-breeders. Here, we show that as group size increases, per capita reproductive success decreases across all breeding stages. Despite this, smooth-billed anis typically breed in larger-than-optimal groups. To understand the factors maintaining communal breeding in non-kin groups experiencing short-term fitness costs, future work will examine how ecological constraints and conflicting selection pressures among group members affect short and long-term fitness in this species.

Cooperative breeding often benefits individual success through group living (Cockburn 1998; Russell et al. 2016). Group living can reduce individual predation risk (Miller 1922; Hamilton 1971; Pulliam 1973; Jeschke & Tollrian, 2007) and increase access to resources such as food (Heinsohn 1995), territory (Duca & Marini 2014; Kingma 2017; Lemoine et al. 2020), and mates (Double & Cockburn 2003; Groenewoud et al. 2018; Johnson & Pruett-Jones. 2018). Breeding groups may also form in response to ecological constraints (Emlen 1982; Hatchwell & Komdeur 2000) such as environmental uncertainty (e.g., in superb starlings, Lamprotornis superbus; Rubenstein 2016) or increased predation pressure (e.g., in greater anis, Crotophaga major; Riehl 2011). In terms of reproductive benefits, larger groups may be able to support larger broods than single pairs, since more provisioning adults can reduce individual parental load (Downing et al. 2021). In many cooperatively breeding species, groups are made up of relatives, and group members may thus gain indirect fitness benefits by helping kin (Hamilton 1963; Browning et al. 2012).

In cooperative breeding species where kin assist with rearing young, groups can include non-breeding helpers (i.e., helper-at-the-nest systems) and/or multiple breeders (i.e., co-breeders) (Vehrencamp & Quinn 2004). In joint-laying (communal) cooperative breeding systems, females lay eggs in a single shared nest, and within-group dynamics are diverse and complex, varying widely among species (Vehrencamp & Quinn 2004). For example, in acorn woodpeckers (Melanerpes formicivorus) non-breeding young from previous broods remain with their parents and help rear their siblings (Stacey 1979), while pūkeko (Porphyrio melanotus melanotus) retain young as both helpers and co-breeders in northern populations (Jamieson 1997). Kin selection can therefore drive both helping (Russel & Hatchwell 2001) and co-breeding (Baglione & Canestrari 2016), but co-breeders need not be kin to gain reproductive benefits from breeding in a group (Riehl 2011; Robertson et al, 2017).

Smooth-billed anis (Crotophaga ani) form breeding groups of socially monogamous pairs (Quinn & Startek-Foote 2020) with relatively low within-group relatedness (Robertson et al. 2017), so indirect fitness benefits are not likely driving cooperative breeding in this system. Groups consist of one to over twenty individuals, with group sizes of four to eight being most common. Groups typically break up after the breeding season and reform into different groups in subsequent years, and interannual site fidelity tends to be low among individuals (Quinn & Startek-Foote 2020). Multiple females lay eggs in a single shared nest, leading to competition as females presumably attempt to maximize their own contribution to the joint clutch by ejecting or burying other individuals’ eggs (Schmaltz et al. 2008; Grieves & Quinn 2018). Thus, joint laying females tend to lose some proportion of their eggs to ovicide by competing group members, and the likelihood of egg loss is higher in larger groups (Schmaltz et al. 2008). Despite competition via ovicide, most breeding groups eventually complete a joint-clutch, at which point parental care duties in subsequent stages of the reproductive cycle (i.e., incubation and provisioning) are shared among group members (Quinn & Startek-Foote 2020).

We evaluated the effect of group size on short-term reproductive success in smooth-billed anis, building on previous research by Schmaltz et al. (2008) showing that, per capita, females both lay and lose more eggs as group size increases. We explore related questions using a larger dataset (Schmaltz et al 2008: 5 breeding seasons, our study: 15) with a wider range of female group sizes (Schmaltz et al 2008: 1–5 females, our study: 1–10), and additional reproductive success metrics (Schmaltz et al. 2008: eggs laid and incubated, our study: eggs laid, incubated, and hatched, and chicks fledged).

We predicted that group reproductive success would increase with group size because of the greater number of adults available to reproduce and provide offspring care. However, we predicted that per capita reproductive success would decrease with group size due to increased competition (via ovicide) during egg laying. Specifically, we predicted that the total number of eggs laid, incubated, and hatched, and chicks fledged would increase with group size while the per capita number of eggs incubated, and hatched, and chicks fledged would decrease with group size. Consistent with Schmaltz et al. (2008), we predicted that the number of eggs laid per capita would increase with group size due to within-group competition during egg-laying. We further predicted that the probability of offspring surviving to successive stages of development would increase with group size, except between egg laying and incubation when competition is greatest (Schmaltz et al. 2008). Specifically, we predicted that the probability of surviving from incubation to hatching and from hatching to fledging would increase with group size.

Study site and data collection

We monitored smooth-billed anis at two U.S. Fish and Wildlife Service (USFWS) National Wildlife Refuges and surrounding property in southwestern Puerto Rico: Cabo Rojo (17°59′N, 67°10′W) and Laguna Cartagena (18°00′N, 67°10′W). Field observations were performed during the breeding season (September/October – December) in 1998–2004, 2006, 2007, 2011–2015, and 2021, for a total of 15 seasons. It was not possible to record data blind because our study involved focal animals in the field.

Each season, teams of 3–5 researchers surveyed these areas (Laguna Cartagena was surveyed over 7 seasons from 1998–2004; Cabo Rojo was surveyed discontinuously over 15 seasons from 1998–2021) to locate groups and monitor breeding activity. Observations typically occurred between 06:00–12:00 and 15:00–18:00 six days a week. Groups were observed opportunistically or targeted deliberately (e.g., to locate nests and count individuals arriving at or leaving roosts) and named according to their breeding territory. Territories were typically occupied each year, but the individuals occupying the territory (i.e., forming the breeding group) changed from year to year.

Group members roost together, so we determined breeding group size by repeatedly counting the number of adults leaving and/or entering roosts throughout the breeding season, supplemented with repeated observations of each group during which we counted the number of adults seen together, consistent with previously established methods in this system (Schmaltz et al. 2008; Robertson et al. 2017; Grieves et al. 2023). Anis are socially monogamous, so we assumed that groups were composed of half females and half males (Schmaltz et al. 2008). We assumed that all females within a group were breeding, as helping is typically restricted to juveniles in this species (Loflin 1983; Quinn & Startek-Foote 2020), and our analyses focused exclusively on groups of adults. To estimate female group size, we divided the total group size by half and rounded down, consistent with prior studies (Schmaltz et al. 2008; Schmaltz et al. 2016; Grieves & Quinn 2018).

We followed groups that showed evidence of breeding and searched for nests. Once found, we monitored nests every 1–4 days, recording the total number of eggs laid, documenting eggs that were ejected from the nest or buried within the nest, and recording the number and identity of eggs in the final incubated clutch. Eggs were numbered using a non-toxic permanent marker each nest visit in the presumed order that they were laid based on wear or their first appearance in the nest.

Ani eggs are initially white from a thin chalky layer of vaterite, a morph of calcium carbonate (Board and Perrott 1979; Portugal et al. 2017). Underneath, the base egg color is blue. The vaterite wears away as eggs are shifted and scratched on nest materials during the laying and incubation stages, and older eggs may be almost completely blue (Grieves & Quinn 2018). Thus, when more than one new egg was found in a nest at a time, we numbered the eggs according to their degree of wear, such that bluer eggs were marked as older.

We determined the date of clutch completion when two consecutive nest checks yielded no new eggs. Nest checks were then paused during incubation until approx. ten days from the date of clutch completion. Nest checks were resumed every 1–2 days thereafter to determine the first date of hatching. Subsequently, we checked nests every 1–3 days until the entire clutch was hatched and banded, or until the oldest chicks fled the nest upon our approach (presumably an anti-predator defense in this species; Quinn and Startek-Foote 2020). When chicks fled the nest at our arrival, we discontinued nest checks to reduce risk of premature fledging.

To ensure accurate monitoring, each chick had one or more toenails clipped for individual recognition when first discovered. Once large enough, chicks were fitted with a USGS metal band and 1–3 Darvic color bands for unique identification within the joint brood. We monitored groups to count fledglings when observed perched, flying, or foraging with adult group members away from the nest tree. We recorded the band combination of fledglings to cross-reference with nest monitoring and banding data to ensure accurate fledgling counts.

We attempted to follow all nests from egg laying through fledging, depending on the stage at which the nest was discovered. We recorded the number of eggs (laid, ejected, buried, and incubated), hatchlings, and fledglings for each group. Our initial dataset included 453 groups. After filtering to remove incomplete data, we retained 364 groups, of which 11% (40 groups) had multiple nests in a season. Of these 364 groups, we had egg laying data for 216, incubation data for 225, hatching data for 198, and fledging data for 176. The number of nests differs between breeding stages because some nests were found after egg laying, incubation, or hatching was completed, nests were abandoned or depredated prior to fledging, or we were unable to follow groups from laying through fledging because they did not complete the breeding attempt before researchers left the study sites.

If a breeding group had multiple nests within a single season, we pooled those nests to determine the total reproductive output for that group’s breeding season. As a check to be sure we were not distorting patterns with this approach, we repeated analyses (in supplementary materials) by collating the data in two additional ways: i) only including the first-discovered nest for each group, and ii) only including the first successful nest (i.e., nests that fledged at least one chick) for each group.

Statistical analyses

For all analyses, we used R version 4.4.1 (R Core Team 2022) run with RStudio (version 2024.4.2.764; RStudio Team 2022). We used generalized linear mixed models (GLMM) with the glmmTMB package (Brooks et al. 2017) to test for an effect of female group size (predictor variable) on the following reproductive success metrics (response variables): total eggs laid, eggs incubated, eggs hatched, and chicks fledged. We included the number of nesting attempts per season and the study site (USFWS refuge) as covariates and territory and year as random effects in all statistical models.

For the group and per capita reproductive success models, we used GLMMs fitted with a negative binomial distribution for analyses of eggs laid and eggs incubated due to overdispersion in the data, and Poisson distributions for analyses of eggs hatched and chicks fledged. To standardize measures of reproductive success in the per capita models, we included the natural log of female group size as an offset.

We used binomial models to test whether the probability of reproductive success changed with group size. The log odds of success \(\:\left(ln\frac{Pr\left(Success\right)}{Pr\left(Failure\right)}\right)\) was modeled with female group size as a fixed effect. We included the number of nesting attempts per group per season and the study site as covariates and territory and year as random effects in all statistical models. We analyzed the proportional number of eggs laid being incubated by dividing the number of incubated eggs by the number of eggs laid. Similarly, we analyzed proportional hatching success by dividing the number of hatched eggs by the number of incubated eggs, and proportional fledgling success by dividing the number of chicks that hatched by the number chicks that of fledged.

For all models, we adjusted the P-value for our predictor variable (female group size) using the sequential Bonferroni method to control the family wise error rate (Holm 1979).

In the group reproductive success models, the total number of eggs laid (\(\:{\beta\:}\) = 0.23,\(\:\:p\) < 0.001), incubated (\(\:{\beta\:}\) = 0.10, p < 0.001), hatched (\(\:{\beta\:}\) = 0.06,\(\:\:p\) = 0.01) and fledged (\(\:{\beta\:}\) = 0.08,\(\:\:p\) = 0.047) increased with group size (Table 1, Fig. 1a, c, e, g). The number of nesting attempts was also a significant predictor of the group number of eggs laid (β = 0.36, p < 0.001; Table 1) and incubated (β = 0.21, p < 0.001; Table 1).

Table 1 Results (in log scale) of generalized linear mixed models to test for an effect of female group size, the number of nesting attempts per season, and study site on the group and per capita number of eggs laid, incubated, and hatched, and chicks fledged in smooth-billed anis. P-values for female group size (our predictor variable) were adjusted using the sequential Bonferroni method (Holm 1979). Nest attempts refer to the number of nest attempts within a given breeding season for a given breeding group

		Group success				Per capita success
Model	Term	β	SE	Z	P		Β	SE	Z	P
Eggs laid	Fixed
	Intercept	1.43	0.15	9.32	<0.001		1.45	0.15	9.60	<0.001
	Female group size	0.23	0.03	7.83	<0.001		-0.09	0.03	-3.45	<0.001
	Nest attempts	0.36	0.08	4.63	<0.001		0.37	0.08	4.89	<0.001
	Study site	0.07	0.14	0.53	0.593		-0.03	0.14	-0.23	0.818
	Random
	Territory	0.15	–	–	–		0.16	–	–	–
	Year	0.21	–	–	–		0.24	–	–	–
Incubated	Fixed
	Intercept	0.98	0.22	4.46	<0.001		1.01	0.22	4.50	<0.001
	Female group size	0.13	0.04	3.50	<0.001		-0.19	0.04	-5.32	<0.001
	Nest attempts	0.32	0.11	3.07	<0.001		0.34	0.11	3.19	<0.001
	Study site	0.05	0.18	0.27	0.788		-0.08	0.18	-0.47	0.637
	Random
	Territory	0.00	–	–	–		0.00	–	–	–
	Year	0.43	–	–	–		0.48	–	–	–
Hatched	Fixed
	Intercept	1.32	0.12	11.12	<0.001		1.33	0.12	11.06	<0.001
	Female group size	0.06	0.02	2.79	0.011		-0.25	0.02	-11.04	<0.001
	Nest attempts	0.06	0.06	1.03	0.302		0.07	0.06	1.25	0.210
	Study site	0.01	0.13	0.12	0.908		-0.07	0.13	-0.56	0.577
	Random
	Territory	0.29	–	–	–		0.30	–	–	–
	Year	0.08	–	–	–		0.07	–	–	–
Fledged	Fixed
	Intercept	0.01	0.37	0.02	0.987		0.00	0.38	0.00	0.999
	Female group size	0.08	0.04	1.98	0.047		-0.24	0.04	-5.95	<0.001
	Nest attempts	0.10	0.13	0.73	0.465		0.12	0.13	0.93	0.351
	Study site	0.16	0.21	0.78	0.435		0.01	0.21	0.06	0.955
	Random
	Territory	0.47	–	–	–		0.46	–	–	–
	Year	0.96	–	–	–		1.00	–	–	–

Per capita, reproductive success decreased as group size increased across all four measures (eggs laid: β = -0.09, p < 0.001, incubated: β = -0.22, p < 0.001, hatched: β = -0.25, p < 0.001, and chicks fledged: β = -0.24, p < 0.001; Table 1, Fig. 1b, d, f, h). The number of nest attempts was also a significant predictor of the per capita eggs laid (β = 0.37, p < 0.001; Table 1) and incubated (β = 0.22, p < 0.001; Table 1).

The probabilities of eggs laid being incubated (β = -0.23,\(\:\:p\) <0.001; Table 2, Fig. 2a) and of incubated eggs hatching (β = -0.12,\(\:\:p\) = 0.033; Table 2, Fig. 2b) significantly decreased as group size increased. There was no significant effect of group size on the probability of chicks fledging (β = -0.08,\(\:\:p\) = 0.229; Table 2, Fig. 2c). The number of nesting attempts was not a significant factor for any probability models (Table 2).

Table 2 Binomial probability models (in log odds scale) used to test for an effect of female group size on the probability (Pr) of eggs laid being incubated, eggs incubated being hatched, and eggs hatched producing fledging chicks in smooth-billed anis. P-values for female group size (our predictor variable) were adjusted using the sequential Bonferroni method (Holm 1979). Nest attempts refer to the number of nest attempts within a given breeding season for a given breeding group

Model	Term	Β	SE	Z	P
Pr _{Incubation\|Laid}	Fixed
	Intercept	0.72	0.36	1.98	0.047
	Female group size	-0.23	0.04	-6.40	<0.001
	Nest attempts	-0.10	0.08	-1.17	0.241
	Study site	-0.16	0.38	-0.41	0.683
	Random
	Territory	1.24	–	–	–
	Year	1.09	–	–	–
Pr _{Hatching\|Incubated}	Fixed
	Intercept	1.10	0.35	3.13	<0.001
	Female group size	-0.12	0.05	-2.40	0.033
	Nest attempts	-0.20	0.12	-1.70	0.088
	Study site	0.10	0.44	0.23	0.814
	Random
	Territory	1.40	–	–	–
	Year	0.71	–	–	–
Pr _{Fledging\|Hatched}	Fixed
	Intercept	-1.11	0.63	-1.76	0.078
	Female group size	-0.08	0.07	-1.20	0.229
	Nest attempts	0.00	0.23	-0.01	0.988
	Study site	0.77	0.54	1.42	0.157
	Random
	Territory	1.47	–	–	–
	Year	1.57	–	–	–

To evaluate whether pooling data for the small subset of groups (40/364) with multiple nesting attempts within a breeding season affected our results, we conducted supplementary analyses considering only the first-discovered nest or the first successful nest (fledging at least one chick) for each of these groups and re-running the analyses with the full dataset (364 group-nests). Overall, our results remained consistent (Tables S2-S5), so we only note here cases in which a change in statistical significance was observed. When we included the first-successful nests only, the group-total number of chicks fledged still increased with group size, but the relationship was no longer statistically significant (β = 0.05, p = 0.105; Table S4). When considering first-discovered and first-successful nests only, the probability of an incubated egg hatching still decreased, but this was no longer statistically significant (first-discovered: β = -0.03, p = 0.615; Table S3; first-successful: β = -0.03, p = 0.570; Table S5).

We evaluated the effects of group size on annual reproductive success in smooth-billed anis. Our first prediction that group reproductive success would increase with group size was supported: larger groups laid, incubated, and hatched significantly more eggs and fledged more chicks. Consistent with our second prediction, per capita reproductive success decreased with group size across incubation, hatching, and fledging, but the per capita number of eggs laid also decreased (inconsistent with Schmaltz et al. 2008). Contrary to our second prediction, after being laid, the probability of an egg being incubated and hatched decreased with group size. We attribute this to ovicide, which occurs at a greater frequency in larger groups (Schmaltz et al. 2008). Finally, the probability of a hatchling surviving to fledge was not affected by group size. Overall, these results were robust to different subsetting methods, and we prefer to focus our discussion on the effects of female group size on total reproductive success within a season, rather than on first-discovered nests, which reflects researcher ability rather than necessarily representing the first nest of the season, or first-successful nests, both of which capture only a portion of the group’s reproductive output within a breeding season.

The number of nest attempts had a significant positive effect on group and per capita number of eggs laid and incubated, but not eggs hatched, or chicks fledged. The likely explanation for this result is simply that there is more opportunity for earlier nesting stages to be affected, given that most nest failures occur during the egg stages. In our dataset, we were able to assign nest fates for 47.6% (215/452) of nests. Nest failure occurred in 60.5% (130/215) of these nests, with abandonment by the group occurring in 43.8% of cases (57/130) and depredation in 56.2% (73/130) of cases. All but one abandonment occurred during egg stages, with one case of abandonment occurring after eggs had hatched, whereas depredation was equally likely during egg (36/73) and hatching (37/73) stages. Thus, when a nest failed during these stages and groups attempted additional nests, there would be more opportunity for additional eggs to be laid or incubated than for eggs to hatch or chicks to fledge as a function of group size.

Schmaltz et al. (2008) found that, per capita, females in large groups of 4 to 5 females produced more eggs than those in single female groups when competition was highest (i.e., when ovicide is most intense). We found that the number of eggs laid per capita decreased with increasing female group size. These seemingly contradictory results may be explained by the different range of group sizes evaluated in these two studies. Schmaltz et al. (2008) analyzed groups with of up to five females over five breeding seasons while we analyzed groups with up to 10 females over 15 seasons. Female great tits (Parus major) are phenotypically plastic regarding clutch size, reducing the number of eggs laid in response to both conspecific density (Nicolaus et al. 2013) and predation, a response that may increase individual survival (Juliard et al. 1997). Joint laying smooth-billed ani females may switch egg laying strategies depending on the number of co-breeders, with those in larger groups reducing their total egg output to avoid reproductive waste associated with increases in ovicide typically associated with larger groups (Schmaltz et al. 2008). In larger groups, the egg laying period of the whole group is often longer, potentially leading to extended hatching periods. In such situations, chicks from later laid eggs may be unable to compete with older more competitive nestlings (Schmaltz et al. 2008). If so, late laying females in large groups may cut their losses and lay fewer eggs, but this remains to be tested.

Despite the short-term reproductive costs we observed, there are potential benefits to co-breeding in large groups, such as increased adult survival. In superb starling individuals in larger groups have greater survival (Guindre-Parker & Rubenstein 2020). Larger groups may allocate more time to vigilance, possibly reducing the risk of predation and thereby increasing adult survival (Lott & Mastrup 1999; Sorato et al. 2012). Smooth-billed and greater anis both use functionally referential alarm calls that warn group members of aerial and terrestrial predators (Grieves et al. 2014; LaPergola et al. 2023), and during foraging individual smooth-billed anis often act as sentinels for the group (Hing et al. 2019). Nest predation is lower in greater anis breeding in larger groups (Riehl 2011), but it remains to be seen whether and how group size affects smooth-billed ani predation risk and adult survival.

Co-breeding in smooth-billed anis may also be driven by ecological constraints. In some species, larger groups may be able to monopolize larger, higher quality territories (Balshine et al. 2001; Duca & Marini 2014; Lemoine et al. 2020). If the habitat is saturated in our study site, with all suitable territories occupied, this could promote increased group sizes despite the short-term reproductive costs we observed. Koenig (1981) pointed out that, when group size is associated with anti-predator behaviour or resource defense, selection should favor an optimal group size. This optimization of group size is beneficial for both group and per capita reproductive success, particularly for groups with kinship among breeders who thereby gain indirect fitness benefits (Hatchwell 2010). Overall, it is still unclear whether habitat saturation and/or territory quality impacts communal breeding in smooth-billed anis, but these avenues are worthy of future study.

Increased group size may lead to individual benefits such as increased survival due to group roosting (Chappell et al. 2016), improved ability to defend territory, or anti-predator vigilance (Kokko et al. 2001; Shah and Rubenstein 2016), and associated behaviors that might be favored by reciprocity (Wright 2007). However, reciprocity in smooth-billed anis would be limited to within season interactions because group membership is unstable across breeding seasons (Schmaltz et al. 2008; Quinn and Startek-Foote 2020). Understanding the long-term (i.e., lifetime) fitness impacts on individuals based on group membership should provide insights into whether the potential benefits of breeding in large groups can offset the short-term reproductive costs of breeding in larger-than-optimal groups.

The formation of non-optimal group size may also be explained, at least partly, by differences in body condition. Joining a group may be advantageous for individuals in poorer conditions, but disadvantageous for those in good condition. For example, striped mice (Rhabdomys pumilio) and house mice (Mus musculus domesticus) switch reproductive strategies from cooperative breeding to solitary breeding if their condition allows them to rear pups singly (Hill et al. 2015; Ferrari et al. 2019) and they produce more pups per capita when breeding singly (Ferrari et al. 2019). In smooth-billed anis, individual group members attempt to prevent other birds from joining the group by engaging in chasing and fighting behaviors that can last for days, an activity that is likely energetically costly (Quinn & Startek-Foote 2020). Joining a group may lead to breeding opportunities for the joiner that would otherwise be unavailable, while decreasing per capita reproductive success within the group due to increased competition. If group members cannot exclude these would-be joiners, group sizes should exceed the theoretical optimal size (Sibley 1983; Seno 2006).

Based on our results, single pairs (Figs. 1g-h, 2c) may be optimal for per capita reproductive success, but if the energetic demands of keeping birds from joining the group are too great, it may be less costly to simply accept them. We therefore expect that individuals in better condition should be more likely to breed in smaller groups and successfully repel would-be joiners, whereas individuals in poorer condition should be more likely to breed in larger groups and be less likely to repel joiners. We also expect that low quality individuals should be more likely than high quality individuals to attempt to join established breeding groups.

In conclusion, individual smooth-billed anis incur short-term reproductive costs as group size increases. Despite this, smooth-billed anis commonly breed in large groups. Future work in this system will evaluate long-term fitness benefits in relation to group size, testing whether i) individuals exhibit higher survival in larger groups, ii) individuals in poorer condition are more likely to join groups, and iii) reproductive success of birds without territories that join established groups differs from pairs that initiate the group.

Supplementary Information The online version contains supplementary material available on Mendeley data: https://data.mendeley.com/v1/datasets/publish-confirmation/vp7bcyznj3/1

Acknowledgements: Fieldwork for this project was conducted on the traditional and contemporary lands of the Taíno and Jíbaro. We thank the many students and field assistants who have worked with anis over the past 25 years and who have provided support to our field teams: A. Cameron, A. Demko, A. Samuelson, B. Bravery, B.-M. Wadien, C. Lentz, E. Friesen, E. Nishikawa, E. Rutherford, H. Darrow, H. Jones, J. Eyster, J. Tabh, L. Rodriguez-Sanoguet, M. Barkley, M. Cruz, N. Roach, R. Land, S. Gregory, and, S. Hing. The Natural Science and Engineering Research Council (Canada) provided funding for this research (Discovery Grant to J.S.Q. RGPIN-2017-06609). We thank USFWS staff, especially F. Schaffner, A. Roman, F. Ramos, J. Padilla, O. Diaz and S. Silander; management and staff at Finca Altamira: A. Franqui and V. Sanchez; and B. Bolker, I. Dworkin, and J. Dushoff for statistical advice.

Funding: This work was supported by a Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Grant to J.S.Q. (RGPIN-2017-06609).

Data Availability: Analyses reported in this article can be reproduced using the data and code available on Mendeley data: doi: 10.17632/gv5wpbryn2.4.

Ethics approval: We followed all applicable international, national, and institutional guidelines for the care and use of animals in research. All birds were handled with permission from the USGS (banding permit 10529). All procedures were approved by McMaster University (Animal Use Protocol 00-05-19; 09-07-25 13-10-37 17-08-32 to JSQ).

Competing Interests: The authors declare that they have no competing interests.

Financial Interests: The authors have no relevant financial or non-financial interests to disclose.

Author Contributions

James S. Quinn conceived and designed the study. Data collection was performed by James S. Quinn, Leanne A. Grieves, Quinlan M. Mann, and Gregory Schmaltz. Data curation and analysis was performed by Quinlan M. Mann. Quinlan M. Mann, Leanne A. Grieves, and James S. Quinn drafted the manuscript. All authors contributed critically to manuscript drafts and gave final approval for publication.

Baglione V, Canestrari D (2016) Carrion crows: Family living and helping in a flexible social system. In: Dickinson JL, Koenig WD, (eds) Cooperative breeding in vertebrates: Studies of ecology, evolution, and behavior. Cambridge University Press, Cambridge, pp. 97-114.
Balshine S, Leach B, Neat F, Reid H, Taborsky M, Werner N (2001) Correlates of group size in a cooperatively breeding cichlid fish (Neolamprologus pulcher). Behav Ecol Sociobiol 50:134-140. https://doi.org/10.1007/s002650100343
Board RG, Perrott HR (1979) Vaterite, a constituent of the eggshells of nonparasitic cuckoos, Guira guira and Crotophaga ani. Calcif Tissue Int 29:63–69. https://doi.org/10.1007/BF02408056
Brooks ME, Kristensen Ka, van Bentham KJ, Magnusson A, Berg CW, Nielsen A, Skaug HJ, Maechler M, Bolker BM (2017) GlmmTMB balances speed and flexibility among packages for zero-inflated generalized linear mixed modeling. R J. 9:378-400. https://doi.org/10.32614/RJ-2017-066
Browning LE, Patrick SC, Rollins LA, Griffith SC, Russell AF (2012) Kin selection, not group augmentation, predicts helping in an obligate cooperatively breeding bird. Proc R Soc B. 279:3861-3869. https://doi.org/10.1098/rspb.2012.1080
Chappell MA, Buttemer WA, Russell AF (2016) Energetics of communal roosting in chestnut-crowned babblers: Implications for group dynamics and breeding phenology. J Exp Biol. 219:3321-3328. https://doi.org/10.1242/jeb.144972
Cockburn A (1998) Evolution of helping behavior in cooperative breeding birds. Annu Rev Ecol Syst 29:141-177. https://doi.org/10.1146/annurev.ecolsys.29.1.141
Double MC, Cockburn A (2003) Subordinate superb fairy-wrens (Malurus cyaneus) parasitize the reproductive success of attractive dominant males. Proc R Soc B 270:379-384. https://doi.org/10.1098/rspb.2002.2261
Downing PA, Griffin AS, Cornwallis CK (2021) Hard-working helpers contribute to long breeder lifespans in cooperative birds. Philos Trans R Soc Lond B Biol Sci 376:20190742. https://doi.org/10.1098/rstb.2019.0742
Duca C, Marini MÂ (2014) Territorial system and adult dispersal in a cooperative-breeding tanager. Auk. 131:32-40. https://doi.org/10.1642/AUK-13-005.1
Emlen ST (1982). The evolution of helping. I. An ecological constraints model. Am Nat 119:29-39. https://doi.org/10.1086/283888
Ferrari M, Lindholm AK, König B (2019) Fitness consequences of female alternative reproductive tactics in house mice (Mus musculus domesticus). Am Nat 193:106-124. http://dx.doi.org/0000-0003-4610-2806
Grieves L, Logue D, Quinn J (2014) Joint-nesting smooth-billed anis (Crotophaga ani) use a functionally referential alarm call system. Anim Behav 89:215–221. https://doi.org/10.1016/j.anbehav.2014.01.008
Grieves L, Quinn J (2018) Group size, but not manipulated whole-clutch egg color, contributes to ovicide in joint-nesting smooth-billed anis. Wilson J Ornithol 130:479-484 https://doi.org/10.1676/17-014.1
Grieves LA, Gloor GB, Quinn JS (2023) Symbiotic microbiota vary with breeding group membership in a highly social joint-nesting bird. Behav Ecol 34:653-661. https://doi.org/10.1093/beheco/arad034
Groenewoud F, Kingma SA, Hammers M, Dugdale HL, Burke T, Richardson DS, Komdeur J (2018) Subordinate females in the cooperatively breeding seychelles warbler obtain direct benefits by joining unrelated groups. J Anim Ecol 87:1251-1263. https://doi.org/10.1111/1365-2656.12849
Guindre-Parker S, Rubenstein DR (2020) Survival benefits of group living in a fluctuating environment. Am Nat 195:1027-1036. https://doi.org/10.1086/708496
Hamilton WD (1963) The evolution of altruistic behavior. Am Nat 97:354-356. https://doi.org/10.1086/497114
Hamilton WD (1971) Geometry for the selfish herd. J Theor Biol 31:295-311. https://doi.org/10.1016/0022-5193(71)90189-5
Hatchwell, BJ (2010). Cryptic kin selection: kin structure in vertebrate populations and opportunities for kin‐directed cooperation. Ethology 116:203-216. https://doi.org/10.1111/j.1439-0310.2009.01732.x
Hatchwell BJ, Komdeur J (2000) Ecological constraints, life history traits and the evolution of cooperative breeding. Anim Behav 59:1079-1086. https://doi.org/10.1006/anbe.2000.1394
Heinsohn R (1995) Hatching asynchrony and brood reduction in cooperatively breeding white-winged choughs Corcorax melanorhamphos. Emu 95:252-258. https://doi.org/10.1071/MU9950252
Hill DL, Pillay N, Schradin C (2015) Alternative reproductive tactics in female striped mice: Heavier females are more likely to breed solitarily than communally. J Anim Ecol 84:1497-1508. https://doi.org/10.1111/1365-2656.12431
Hing JS, Schmaltz G, Quinn JS (2019) Testing for sentinel coordination in smooth-billed anis (Crotophaga ani). J Ornithol 160:165-172. https://doi.org/10.1007/s10336-018-1584-6
Holm S (1979) A simple sequentially rejective multiple test procedure. Scand J Stat 6:65–70. http://www.jstor.org/stable/4615733
Jamieson, IG (1997) Testing reproductive skew models in a communally breeding bird, the pukeko, Porphyrio porphyrio. Proc R Soc B 264:335-340. https://doi.org/10.1098/rspb.1997.0048
Jeschke JM, Tollrian R (2007) Prey swarming: which predators become confused and why? Anim Behav 74:387-393. https://doi.org/10.1016/j.anbehav.2006.08.020
Johnson AE, Pruett-Jones S (2018) Reproductive promiscuity in the variegated fairy-wren: An alternative reproductive strategy in the absence of helpers? Anim Behav 139:171-180. https://doi.org/10.1016/j.anbehav.2018.03.015
Julliard R, McCleery RH, Clobert J, Perrins CM (1997) Phenotypic adjustment of clutch size due to nest predation in the great tit. Ecology 78:394-404. https://doi.org/10.1890/0012-9658(1997)078[0394:PAOCSD]2.0.CO;2
Kingma SA (2017). Direct benefits explain interspecific variation in helping behaviour among cooperatively breeding birds. Nat Commun 8:1094. https://doi.org/10.1038/s41467-017-01299-5
Koenig, WD (1981) Reproductive success, group size, and the evolution of cooperative breeding in the acorn woodpecker. Am Nat 11:7421-443. https://doi.org/10.1086/283726
Kokko, H, Johnstone, RA and TH, CB (2001) The evolution of cooperative breeding through group augmentation. Proc R Soc B 268:187-196. https://doi.org/10.1098/rspb.2000.1349
Lapergola JB, Savagian AG, Smith MG, Bennett BL, Strong MJ, Riehl C (2023) Referential signaling in a communally breeding bird. PNAS. 120:e2222008120. https://doi.org/10.1073/pnas.2222008120
Lemoine S, Boesch C, Preis A, Samuni L, Crockford C, Wittig RM (2020) Group dominance increases territory size and reduces neighbour pressure in wild chimpanzees. R Soc Open Sci 7:200577. https://doi.org/10.1098/rsos.200577
Loflin RK (1983) Communal behaviors of the smooth-billed ani (Crotophaga ani). Dissertation University of Miami.
Lott DF, Mastrup SN (1999) Facultative communal brood rearing in California quail. Condor 101:678-681. https://doi.org/10.2307/1370200
Miller RC (1922) The significance of the gregarious habit. Ecology 3:122-126. https://doi.org/10.2307/1929145
Nicolaus M, Brommer JE, Ubels R. Tinbergen JM, Dingemanse NJ (2013) Exploring patterns of variation in clutch size–density reaction norms in a wild passerine bird J Evol Biol 26:2031-2043. https://doi.org/10.1111/jeb.12210
Portugal SJ, Bowen J, Riehl C (2018) A rare mineral, vaterite, acts as a shock absorber in the eggshell of a communally nesting bird. Ibis. 160:173-178. https://doi.org/10.1111/ibi.12527
Pulliam HR (1973) On the advantages of flocking. J Theor Biol 38:419-422. http://doi.org/10.1016/0022-5193(73)90184-7
Quinn JS, Startek-Foote JM (2020) Smooth-billed ani (Crotophaga ani). Ithaca, NY, USA: Cornell Lab of Ornithology. https://doi.org/10.2173/BNA.SMBANI.02
R Core Team (2022) R: A language and environment for statistical computing. R Foundation for statistical computing, Vienna, Austria. URL https://www.R-project.org/
Riehl C (2011) Living with strangers: Direct benefits favor non-kin cooperation in a communally nesting bird. Proc R Soc B 278:1728-1735. https://doi.org/10.1098/rspb.2010.1752
Robertson J, Caldwell J, Grieves L, Samuelsen A, Schmaltz G, Quinn JS (2017) Male parental effort predicts reproductive contribution in the joint-nesting, smooth-billed ani (Crotophaga ani). J Ornithol 159:1-11. https://doi.org/10.1007/s10336-017-1522-z
RStudio Team (2022) Rstudio: Integrated development environment for R. RStudio, PBC Boston, MA. URL http://www.rstudio.com/
Rubenstein DR (2016) Superb starlings: Cooperation and conflict in an unpredictable environment. In: Dickinson JL, Koenig WD, (eds) Cooperative breeding in vertebrates: Studies of ecology, evolution, and behavior. Cambridge: Cambridge University Press. pp 181-196
Russell AF, Hatchwell BJ (2001) Experimental evidence for kin-biased helping in a cooperatively breeding vertebrate. Proc R Soc B 268:2169-2174. https://doi.org/10.1098/rspb.2001.1790
Russell, AF, Koenig WD, Dickinson JL (2016) Cooperative breeding in vertebrates: Studies of ecology, evolution, and behavior. Cambridge University Press, Cambridge
Schmaltz G, Quinn J, Lentz C (2008) Competition and waste in the communally breeding smooth-billed ani: Effects of group size on egg-laying behaviour. Anim Behav 76:153-162. https://doi.org/10.1016/j.anbehav.2007.12.018
Schmaltz G, Quinn JS, Schoech SJ (2016) Maternal corticosterone deposition in avian yolk: Influence of laying order and group size in a joint-nesting, cooperatively breeding species. Gen Comp Endocrin 232:145-150. https://doi.org/10.1016/j.ygcen.2016.04.022
Seno H (2006) Group size determined by fusion and fission a mathematical modelling with inclusive fitness. J Math Biol 52:70-92. https://doi.org/10.1007/s00285-005-0341-7
Shah SS, Rubenstein DR (2023) Group augmentation underlies the evolution of complex sociality in the face of environmental instability. PNAS 120:2212211120. https://doi.org/10.1073/pnas.2212211120
Sibly RM (1983) Optimal group size is unstable. Anim Behav 31:947-948. https://doi.org/10.1016/S0003-3472(83)80250-4
Sorato E, Gullett PR, Griffith SC, Russell AF (2012) Effects of predation risk on foraging behaviour and group size: Adaptations in a social cooperative species. Anim Behav 84:823-834. https://doi.org/10.1016/j.anbehav.2012.07.003
Stacey PB (1979) Kinship, promiscuity, and communal breeding in the acorn woodpecker. Behav Ecol Sociobiol 6:53-66. https://doi.org/10.1007/BF00293245
Vehrencamp SL, Quinn J (2004) Joint laying systems. In: Koenig WD, Dickinson JL, (eds) The ecology and evolution of cooperative breeding in birds. Cambridge University Press, Cambridge, pp 177-196
Wright, J (2007) Cooperation theory meets cooperative breeding: exposing some ugly truths about social prestige, reciprocity and group augmentation. Behav Processes 76:142-148. https://doi.org/10.1016/j.beproc.2007.01.017

Supplementalmaterials.docx

Per capita reproductive success decreases with group size in a communally breeding bird

Status:

Version 1

Abstract

Figures

Significance statement

Introduction

Materials and Methods

Study site and data collection

Statistical analyses

Results

Discussion

Declarations

References

Supplementary Files

Status:

Version 1