Alloparental care is central to human life history, which integrates exceptionally short interbirth intervals and large birth size with an extended period of juvenile dependency and increased longevity. Formal models, previous comparative research, and palaeoanthropological evidence suggest that humans evolved higher levels of cooperative childcare in response to increasingly harsh environments. Although this hypothesis remains difficult to test directly, the relative importance of alloparental care varies across human societies, providing an opportunity to assess how local social and ecological factors influence the expression of this behaviour. We therefore, investigated associations between alloparental infant care and socioecology across 141 non-industrialized societies. We predicted increased alloparental care in harsher environments, due to the fitness benefits of cooperation in response to shared ecological challenges. We also predicted that starvation would decrease alloparental care, due to prohibitive energetic costs. Using Bayesian phylogenetic multilevel models, we tested these predictions while accounting for potential confounds as well as for population history. Consistent with our hypotheses, we found increased alloparental infant care in regions characterized by both reduced climate predictability and relatively lower average temperatures and precipitation. We also observed reduced alloparental care under conditions of high starvation. These results provide evidence of plasticity in human alloparenting in response to ecological contexts, comparable to previously observed patterns across avian and mammalian cooperative breeders. This suggests convergent social evolutionary processes may underlie both inter- and intraspecific variation in alloparental care.

1. Introduction

Humans exhibit high degrees of cooperation—encompassing proactive food sharing and cooperative child care [1,2], teaching and cultural learning [3], collaborative foraging [4], and extensive division of labour [5,6]—in groups characterized by low average genetic relatedness [7] and low frequencies of non-reproductive individuals [8]. This propensity toward ‘hyper-cooperation’ [9] takes place within an unusual life history integrating short interbirth intervals and large birth size with a highly dependent juvenile period and a prolonged post-reproductive lifespan, as compared to other primates [10,11]. Collectively, these traits are hypothesized to have coevolved with humans' relatively large brains, skill-intensive foraging niche, and capacity for cumulative cultural evolution [11,12]. Understanding the selection pressures driving human hyper-cooperation and life history therefore remains a central explanatory task for evolutionary anthropology.

Allomaternal care, i.e. caretaking from individuals other than an offspring's mother, is crucial for human life history [12]. In extant human societies, allomaternal care promotes consistent energetic input for infants [13] and reduces demands on maternal time and energy allocation [14–17]. Palaeoanthropological and comparative evidence also suggest that enhanced allomaternal care may have underpinned the marked transitions in sociality, life history, and brain size of ancestral hominins [12]. This increase in allomaternal care involved the evolution of greater paternal care (i.e. help from fathers [6,18,19]), as well as greater alloparental care (i.e. help from individuals other than an offspring's parents [2,9,20]). Today, alloparental care is consistently observed across societies, with older siblings, extended family members, and non-kin providing food, knowledge, and protection to highly dependent offspring [2,11,21]. Such alloparenting reflects investments of time, energy, and resources into others' offspring and thus presents opportunity costs relative to direct investment in one's own reproduction [22]. What fitness benefits may have facilitated the evolution of this cooperative behaviour?

Palaeoanthropologists have attempted to explain the evolution of human cooperation and life history as a response to the environmental volatility of the Plio-Pleistocene period [23–25], which was characterized by high spatio-temporal heterogeneity in habitat quality and protracted intervals of climate instability and resource uncertainty [26,27]. Such harsh environments, characterized by unpredictable and extreme climates, likely affected the fitness benefits of collaboration and interdependency, as suggested by formal models demonstrating that environmental harshness can facilitate the evolution of various cooperative behaviours such as food sharing and alloparental care [28–32]. However, the role of environmental harshness as a selection pressure for human alloparenting remains difficult to test empirically with the palaeoanthropological record, notwithstanding suggestive shifts in hominin life history, brain size, and diet [23–25].

Nevertheless, empirical data from extant human societies and animal species across a broad range of ecological contexts provide important evidence to assess this environmental harshness hypothesis. For example, kin selection is a primary driver of alloparental care within many mammal and bird species [33,34], who often live in small family groups and defend territories in saturated habitats (e.g. [35]). However, alloparental care is also frequently observed among species inhabiting harsher environments [36–40]. This indicates that cooperative breeding can provide fitness benefits in less predictable and more extreme ecological contexts, where individuals face common challenges that prevent them from successfully surviving and reproducing alone. Such ecological pressures are expected to facilitate more extensive cooperation in larger groups of both kin and non-kin [28–32,41–43]. Furthermore, many species also exhibit plasticity in cooperative breeding behaviour across ecological contexts (e.g. [35,44,45]). This suggests that fluctuating environmental conditions can select for behavioural flexibility to optimize trade-offs between self-maintenance, reproductive effort, and offspring care [46].

In contrast to the literature on non-human animals, research on cooperative childcare in humans has given less attention to ecological variation, with prior work largely emphasizing kin-based fitness benefits through grandmaternal [14,47,48] and sibling care [48,49], as well as direct and indirect forms of paternal care [6,18,19,50]. While paternal and kin-directed care are crucial for explaining human life history, distal and non-kin also make significant contributions [2], and human alloparenting is known to be contingent on reciprocity among both close and distal kin [51,52]. Given the hypothesized role of harsh environments in the social evolution of ancestral hominins and other cooperatively breeding taxa, variation in shared ecological challenges—and thus the benefits of pooling risk through cooperation—may be an important but relatively understudied determinant of plasticity in human alloparental care across societies.

Here, we integrate behavioural-ecological and palaeoanthropological theory to test whether environmental harshness explains intraspecific variation in human alloparenting. In particular, we predicted that societies occupying harsher environments would exhibit higher rates of alloparental care, as the relative costs of this behaviour are expected to be outweighed by the increased benefits of cooperation for survival and reproduction. Conversely, we also predicted decreased rates of alloparental care in environments where the benefits of cooperation are insufficiently offset by its costs to self-maintenance and future reproduction, such as when societies experience conditions of severe starvation and resource pressure. To test these predictions, we examined 141 societies from the standard cross-cultural sample (SCCS) [53], a globally representative sample of non-industrialized societies, with measures of allomaternal and paternal infant care [54] (figure 1). We used multiple climatic and socioecological predictors to test whether alloparental care (i.e. allomaternal care independent of paternal care) associated with environmental harshness and starvation.

Figure 1. Global variation in allomaternal infant care. (a) Frequency of societies (N = 141 total) scored from 1 to 6 (verbal description on x-axis) on a measure of infant allomaternal care (SCCS v51), which was subsequently binned into either low (purple) or high (yellow) care. (b) Global distribution of study societies, colour-coded by low (purple) or high (yellow) allomaternal infant care. This measure covaries with paternal infant care (SCCS v53), which was subsequently controlled for to focus analyses on alloparental care. (Online version in colour.)

2. Material and methods

(a) Sample

All measures were retrieved from the Database of Places, Language, Culture and Environment (D-PLACE [55]). We examined 141 non-industrialized societies from the SCCS [53], for which both allomaternal and paternal infant care measures were available. Note that all societal variables have standardized indices within the SCCS and are referred to as ‘SCCS v’ with the appropriate index. We used ‘Non-maternal relationship, infancy’ (SCCS v51 [54]), a standardized six-point scale capturing the degree of infant-directed allomaternal care described in ethnographies for each society, with 1 indicating ‘Almost exclusively mother’ and 6 indicating ‘Mother minimal except for nursing’ (figure 1). Due to low counts of negligible allomaternal care (5 societies scored 1) or primary allomaternal care (13 societies scored 4 or above), we binned this scale into a more robust binary measure of low (1–2; N = 71) or high (3–6; N = 70) allomaternal infant care. We focused our analysis on infant care due to the particularly high energetic demands that mothers face during this period [56]. We did not analyse early childhood allomaternal care because of minimal variation (i.e. no societies had minor early childhood allomaternal care). This likely reflects indirect or low-cost forms of care typical of childhood (e.g. play groups [57]).

(b) Social and ecological predictors

(i) Environmental harshness

Despite subtle differences in measures and terminology, previous studies have consistently demonstrated associations between cooperative breeding and indicators of climate predictability and average temperature and precipitation. For example, Rubenstein & Lovette [36] found that cooperatively breeding starlings were more likely to occupy savannah habitats characterized by less predictable patterns of precipitation. Similarly, using local temperature range and the mean and variance of precipitation, Lin et al. [37] found that starlings engaged in higher rates of cooperative breeding in harsher environments. Jetz & Rubenstein [38] further demonstrated a positive association between cooperative breeding and precipitation variability across a broad range of avian species, while Cornwallis et al. [39] established an association between avian cooperative breeding and a phylogenetically controlled principal component capturing higher temperature mean and predictability, lower temperature variance, and higher precipitation mean and variance. Across mammals, Lukas & Clutton-Brock [40] found that cooperative breeders were more likely to inhabit environments with lower precipitation and temperature, as well as lower precipitation and temperature predictability, with low annual precipitation in particular identified as the key predictor of cooperative breeding in a multivariate analysis.

To maximize comparability with prior studies, we operationalized ‘harsh environments’ as less predictable climates with more extreme patterns of temperature and rainfall. We therefore extracted spatio-temporally localized means and variances of monthly temperature and precipitation for each society, as well as Colwell's [58] standardized indices of annual temperature and precipitation predictability, from the EcoClimate database hosted on D-PLACE [55,59]. Following previous work [39], we used phylogenetically controlled principal component analysis (pPCA) [60] to reduce multicollinearity and phylogenetic autocorrelation, and we subsequently applied a parallel analysis procedure to retain components that explained greater variance in our dataset than would be expected by chance. We thus selected two pPCs that were subjected to an orthogonal Quartimax rotation [61] to enhance biological interpretation. These two components explained 73% of the total variance, each distinctly capturing variance in temperature (pPC1; 43% of total variance) and precipitation (pPC2; 30% of total variance). See table 1 for component loadings. We examined both additive and nonlinear interaction effects among these components, as described below, because societies with both low climate predictability and extreme temperatures and rainfall should experience the harshest environments.

Table 1. Phylogenetic principal components of ecological measures. Loadings greater than 0.50 (i.e. component describes + 25% original measure variance) are bolded. Dashes indicate that the third principal component was independently derived due to missing data across societies.

Collapse
variable	temperature (pPC1)^a	precipitation (pPC2)^a	starvation (pPC3)
temperature monthly mean	0.92	−0.13	—
temperature monthly variance	−0.89	−0.15	—
temperature annual predictability	0.84	0.33	—
precipitation monthly mean	0.39	0.80	—
precipitation monthly variance	0.31	0.60	—
precipitation annual predictability	−0.06	0.80	—
endemic starvation	—	—	0.69
short-term starvation	—	—	0.75
seasonal starvation	—	—	0.88

^aComponents were rotated using the Quartimax criterion to enhance interpretability.

(ii) Starvation

To assess the influence of starvation on alloparenting, we used multiple indices of starvation (endemic, SCCS v1261; short-term, SCCS v1262; and seasonal starvation, SCCS v1263 [62]). In contrast to our climate measures, heterogeneous patterns of missing data were present for the starvation indicators (N = 83 societies with all observed, N = 87 endemic, N = 129 short-term, N = 128 seasonal). We therefore conducted a distinct pPCA using only observed starvation values to reduce collinearity among these measures without the need to artificially impute climate data. A single component (pPC3) explained 61% of variance in these starvation measures (table 1).

(iii) Additional predictors

Three measures were included in our analysis to account for variation due to subsistence practices. We included fixity of residence (SCCS v150 [63]) because it may influence alloparenting through the positive effects of sedentism on fertility [64], as well as food storage (SCCS v20 [65]), which can reduce the probability of daily food sharing [66] and may, therefore, lead to general reductions in the benefits of cooperation. To account for any clustering in our data otherwise not accounted for by the effects of fixity of residence and food storage, we also included primary subsistence mode (SCCS v820 [67]; see electronic supplementary material, table S1). Total pathogen stress (SCCS v1260 [68]) was included for comparison with previous work on ecological determinants of maternal and paternal care [50]. Finally, the biome [69] occupied by each society was included to capture any clustering in our data due to ecological factors that were not captured by our principal components (see electronic supplementary material, table S2).

While fathers are an important source of offspring care and energetic input across many human societies [6,19], we controlled for paternal infant care (SCCS v53 [54]) as a covariate rather than a separate outcome measure in our analysis. This ensured that our allomaternal care measure accurately reflected alloparenting, rather than being confounded by potentially distinct influences on paternal investment [6,19,50,70]. In addition, this facilitated more direct comparison with the broader literature on non-human cooperative breeders, which emphasizes the evolution of alloparenting (in contrast to the emphasis on allomaternal care within anthropology, e.g. [9,20]). A clear relationship was observed between the paternal and allomaternal infant care measures (electronic supplementary material, table S2), further supporting our decision to control for paternal care in order to obtain a more robust measure of alloparental care. See electronic supplementary material, appendix SI for further exploratory analyses of ecological effects on paternal care.

(iv) Phylogeny

To avoid the pitfalls of autocorrelation due to shared population history, we used a supertree of human populations based on genetic and linguistic data [71–73]. This supertree contains 388 source trees from 250 studies and a topological constraint based on linguistic classification [71,72]. The inference and time-calibration of the supertree for the SCCS populations is described in [73]. See electronic supplementary material, figure S1 for the subtree used in this study.

(c) Statistical analysis

A Bayesian multilevel phylogenetic regression model was fitted to the Bernoulli response of low or high allomaternal care, controlling for paternal care [54]. We included fixed effects for pPC1–pPC3 to assess our primary hypotheses, and further controlled for fixity of residence [63], food storage [65], and pathogen stress [68]. All continuous covariates were standardized to z-scores to enhance interpretation and facilitate effect size comparison, with pPCs centred on inferred ancestral mean states rather than average sample values. Random effects were estimated for primary mode of subsistence [67], biome [69], and phylogeny [73]. Additionally, to examine potential interactions between temperature (pPC1) and precipitation (pPC2), we also estimated a model with a linear interaction effect between these components (+ pPC1 * pPC2), as well as a second-order polynomial model with both interaction and quadratic terms (+ pPC1² + pPC2² + pPC1 * pPC2), which is commonly used to estimate nonlinear responses in evolutionary ecology (e.g. [74]). The inclusion of both interaction and quadratic parameters, also known as a response surface analysis [75], allowed for a nonlinear interaction between pPC1 and pPC2, such that their joint effects on alloparenting could differ as a function of their respective magnitudes. These interaction models, therefore, more directly tested our hypothesis that the greatest probability of high alloparental infant care would occur for low pPC1 and low pPC2 scores specifically, i.e. environments with both low climate predictability and relatively cooler and dryer climates, as compared to average or high scores indicating both higher climate predictability and relatively warmer and wetter climates. Random slopes for all linear and nonlinear parameters were estimated across primary subsistence modes for pPC1–pPC3, which accounted for potential differences in these effects due to unmeasured factors related to subsistence practices. Random effect correlations were fixed to 0 across subsistence modes because there were too few categories (N = 7) to estimate these parameters (see electronic supplementary material, table S1); similarly, random slopes were not included across biomes due to insufficient sampling (see electronic supplementary material, table S2). Regularizing priors were placed on fixed $β \sim Normal (0, 1)$ and random effects $σ \sim Exponential (3)$ . The ‘brms’ package for R statistical environment [76] was employed for all analyses.

By contrast to classical phylogenetic least-squares methods, our Bayesian multilevel modelling approach provided a number of benefits, namely allowing us to (i) properly account for the non-Gaussian, binary response variable and for hierarchical clustering in the data due to phylogeny, biome, and subsistence mode [77], (ii) impute missing starvation (pPC3) scores during model estimation with a fully conditional imputation model, which uses information from all variables to predict missing scores and thus avoids systematic biases due to non-random missingness [78], (iii) accurately estimate monotonic effects for ordinal-scale predictors (paternal infant care, fixity of residence) [79], (iv) and employ regularizing priors to produce more conservative estimates and reduce the risk of inferential error [78,80].

Bayesian models produce a probability distribution for each estimated parameter, known as a posterior distribution. Following suggestions for state-of-the-art practice [78,80,81], we used multiple quantitative measures to summarize these posterior distributions and draw scientific inferences from our models, rather than relying on null-hypothesis tests and dichotomous attributions of statistical significance. In particular, to interpret the strength and uncertainty of associations, we present the median regression coefficients (β, on the log odds scale), their median absolute deviation (MAD), a more robust measure of dispersion than the standard deviation, their 90% credible intervals (CI), and the proportion of the posterior distribution greater or smaller than 0 (p₊ or p₋). Note that in contrast to classical p-values, which indicate the probability of observing the data under a null hypothesis $p (data | H_{0})$ , the reported p₊ or p₋ directly estimate the probability in support of hypothesized positive or negative associations given the data $p (H_{1} | data)$ . In addition, we calculated standardized effect sizes (Cohen's d) for fixed effects to facilitate comparison within and across studies. We also used the Watanabe–Akaike information criterion (WAIC), a fully Bayesian extension of classical AIC accounting for posterior uncertainty [82], to compare models with main effects only and those with linear and nonlinear interactions. Posterior medians are denoted throughout by tilde accents (i.e. $\tilde{β}$ , $\tilde{d}$ ), while dispersion measures and CI are presented in brackets. Please see electronic supplementary material, appendix SI for further details on statistical models, effect size calculations, robustness checks, and results without imputation.

3. Results

We found a large negative main effect of the temperature component on alloparental infant care (pPC1; $\tilde{β} = - 2.50$ [MAD = 0.75], 90% CI [−3.78, −1.33], p₋ = 1.00, $\tilde{d} = - 1.38$ [0.41]), but also a small quadratic effect of temperature ( $\tilde{β} = - 0.41$ [0.26], 90% CI [−0.85, 0.03], p₋ = 0.94, $\tilde{d} = - 0.22$ [0.14]) and large interaction effect of temperature with the precipitation component (pPC2; $\tilde{β} = 1.62$ [0.65], 90% CI [0.58, 2.78], p₊ = 1.00, $\tilde{d} = 0.89$ [0.36]). Temperature, therefore, had a smaller, nonlinear effect in regions of high precipitation, but a large negative effect in regions of low precipitation (see figure 2a). In contrast, the precipitation component did not exhibit an independent main effect ( $\tilde{β} = 0.41$ [0.39], 90% CI [−0.21, 1.09], p₊ = 0.86, $\tilde{d} = 0.22$ [0.21]) or quadratic effect ( $\tilde{β} = 0.00$ [0.24], 90% CI [−0.39, 0.46], p₊ = 0.50, $\tilde{d} = 0.00$ [0.13]) on alloparental infant care. Consistent with our hypothesis, the nonlinear interaction between the temperature and precipitation components suggests that the probability of high alloparental infant care is greatest in environments characterized by both reduced climate predictability and relatively lower average temperatures and precipitation (figure 2a,b). In particular, our model predicts that societies with low temperature and precipitation scores (pPC1 = −1, pPC2 = −1) have a greater probability of expressing high alloparental care compared to those with either average scores (pPC1 = 0, pPC2 = 0; Δp = 0.61 [0.17], p₊ = 0.99), high scores (pPC1 = +1, pPC2 = +1; Δp = 0.71 [0.20], p₊ = 0.99), or low temperature and high precipitation scores (pPC1 = −1, pPC2 = +1; Δp = 0.43 [0.25], p₊ = 0.95). In addition to the climatic predictors, starvation (pPC3) also showed a moderately sized negative main effect on alloparental infant care ( $\tilde{β} = - 1.16$ [0.33], 90% CI [−1.73, −0.60], p₋ = 1.00, $\tilde{d} = - 0.64$ [0.18]; figure 2a,b). Societies with more starvation are, therefore, less likely to exhibit high infant alloparenting. The combined effects of these three ecological components on alloparental infant care are visualized in figure 2b. See electronic supplementary material, figure S2 for predictions across subsistence modes.

Figure 2. Predicted effects of ecology on alloparental infant care. (a) Predicted probability (y-axis) of observing high alloparental infant care as a function of variation in ecological variables (x-axes). The effect of temperature (pPC1) is shown for both low (−1) and high (+1) precipitation (pPC2), reflecting the non-linear interaction of these components. Raw data are also shown, colour-coded as either high (yellow) or low (purple) alloparental care. Raw data are separated by pPC2 < 0 and pPC2 > 0 for the low and high precipitation plots, respectively. Shaded bands indicate 10–90% posterior CI from the darkest to lightest bands, respectively, while the dark lines indicate posterior median values. (b) Posterior probability of high alloparental infant care (indicated by colour gradient) as a function of ecological variables—temperature (pPC1), precipitation (pPC2), and starvation (pPC3). Probability estimates are based on median posterior effect sizes, conditional on average covariate values, and minimum categories for monotonic effects, marginalized over random effects. Probability estimates from our model are therefore given by p = logistic(−1.41–2.50*pPC1 + 0.41*pPC2–1.16*pPC3–0.41*pPC1² + 0.00*pPC2² + 1.62*pPC1*pPC2), where −1.41 is the global intercept. (Online version in colour.)

These findings were robust to the exclusion of missing data imputation for pPC3, the exclusion of all non-essential fixed and random effects, as well as to binning of the response variable (see electronic supplementary material, appendix SI for further details and electronic supplementary material, table S3 for other covariate and random effect estimates). Model comparison provided stronger support for the nonlinear interaction model presented here as compared to a main effects only model (ΔWAIC = 10.01 [6.86]) or an additive interaction model without quadratic effects (ΔWAIC = 6.58 [5.10]). See electronic supplementary material, table S1 for comparison to models without random slopes across subsistence modes.

4. Discussion

Consistent with palaeoanthropological evidence [23–25], formal models [28–32], and comparative studies of other cooperative breeders [36–40], we found clear support for the hypothesis that environmental harshness promotes alloparental care across human societies, suggesting adaptive plasticity in this behaviour. In particular, we observed greater alloparental care among societies occupying harsh environments characterized by both relatively lower climate predictability and lower mean precipitation and temperature (table 1 and figure 2). The evolution of human alloparenting may, therefore, be promoted by the fitness benefits of cooperation in the face of shared ecological challenges. These findings are consistent with research demonstrating the importance of reciprocity in human alloparenting [51,52], as well as broader behavioural-ecological theory emphasizing the importance of collective action and reciprocity for maintaining various other cooperative behaviours and human institutions, such as interhousehold food sharing [66,83], leadership [84], and religious beliefs [85]. While previously supported by suggestive climate data and concurrent shifts in hominin life history, foraging practices, and brain size [12,18,23–25], our results provide more direct support for the claim that human alloparenting has evolved to buffer against the risks of volatile ecological contexts. In addition, while harsh environments increased alloparental care, starvation also had a strong negative effect (figure 2). This finding is consistent with previous research demonstrating that extreme resource stress is associated with increased infanticide [86]. Maternal care was also negatively associated with famine severity in previous analyses of the SCCS [50]. These patterns suggest that childcare can become exceedingly costly if subsistence practices and/or social structures do not sufficiently offset the risks of extreme malnourishment to the caretaker. Our results, therefore, provide evidence that human alloparenting is responsive to the effects of ecological conditions on both the costs and benefits of cooperation.

While our findings are broadly consistent with previous comparative research linking cooperative breeding to harsh environments [36–40], the specific climatic effects identified across studies have been notably heterogeneous. For example, cooperative breeding is more likely in regions of high rather than low annual temperatures and precipitation among birds [39], in contrast to the findings of the present study as well as previous work linking cooperative breeding to lower temperatures and precipitation across other mammals [40]. However, more variable and less predictable precipitation has been consistently linked to species differences in sociality and cooperative breeding across both birds and mammals [36–40,87]. Here, we found that precipitation did not exhibit an independent effect on alloparenting, but rather that lower temperature (pPC1) and precipitation (pPC2) scores jointly predicted the highest rates of alloparental care. Such apparent inconsistencies are not surprising, however, given the diverse life histories, foraging strategies, and biogeographic distributions of the species studied thus far. As noted by Lukas & Clutton-Brock [40], cooperatively breeding birds frequently occur in warmer equatorial environments [38], while cooperatively breeding mammals are often found in cooler latitudes. This mammalian pattern is consistent with our findings, where lower pPC1 scores also correlate with latitudinal distance from the equator (r = 0.87). Both temperature and precipitation are important predictors of global variation in biodiversity [88], and reduced biodiversity has been linked to reduced human population densities in less productive, high- and mid-latitude habitats [89]. This suggests that environments with both relatively lower and less predictable temperature and precipitation, as compared to equatorial habitats, may indeed be particularly harsh for non-industrialized human societies. For example, high alloparental infant care was observed among many of the indigenous societies who occupied deserts, xeric shrublands, and grasslands throughout Southwestern and Central North America, as well as among societies in the Mediterranean forests and woodlands of western South America and Europe, which are characterized by less predictable temperatures and precipitation. Notably, these were the only biomes that on average exhibited both negative pPC1 and negative pPC2 scores in our sample (electronic supplementary material, figure S3). Therefore, while harsh environments can be consistently characterized by extreme and variable climates, it is also expected that the specific effects of temperature and precipitation will differ across taxa and biogeographic distributions.

It is also important to emphasize that our analyses do not directly disentangle alternative causal hypotheses relating alloparenting and ecology. In particular, while we expect starvation to decrease rates of infant alloparenting, observed starvation may itself be a function of reduced alloparental care in response to additional socioecological factors. Similarly, the association between alloparenting and environmental harshness may arise from a plastic response or local adaptation to ecological conditions, as suggested by the environmental harshness hypothesis, or because alloparenting facilitates migration from benign to harsher environments, as argued for avian cooperative breeders [39]. Alternatively, cooperative breeders may be more resilient to environmental deterioration and thus better able to withstand increased ecological variability, as suggested by geographical hotspots of avian cooperative breeding in regions that underwent dramatic climate change throughout the Tertiary period [90]. These studies indicate that benign environments can promote the evolution of family living and cooperative breeding, at least among birds, followed by the subsequent maintenance of cooperative breeding in harsher environments. Nonetheless, the selection pressures on social group formation and alloparenting are expected to differ across these environmental gradients [31], as exemplified by contrasting associations between cooperative breeding and environmental harshness among starlings and hornbills [37]. Disentangling these distinct eco-evolutionary trajectories beyond the initial onset of cooperative breeding within a lineage, as well as their consequences for subsequent changes in group structure and variation in alloparental care, remains an important target for future research.

In conclusion, our study provides strong evidence that infant alloparenting is associated with environmental harshness across human societies, complementing previous work on intraspecific variation in avian alloparental care (e.g. [35,44–46]), as well as interspecific comparisons across both birds and mammals [36–40]. These findings collectively demonstrate that harsh environments (i.e. unpredictable and extreme climates) play a central role in the evolution of cooperative breeding. Therefore, our results indicate that the expression of human alloparental care is congruent with expectations of broader social evolutionary theory [31,91,92], despite the somewhat unique structure of human social systems [93]. Moreover, our findings provide indirect support for palaeoanthropological evidence emphasizing rapid ecological change as a key selection pressure in the evolution of hominin cooperation and life history [23–27].

Data accessibility

The dataset, R code, and phylogeny supporting all analyses reported in this article have been uploaded to a publicly available GitHub repository for open access. These files can be downloaded at https://github.com/Jordan-Scott-Martin/SCCS_alloparental-care.

Authors' contributions

J.S.M. devised and initiated the research project, developed the dataset, composed the manuscript, and conducted the statistical analyses with supervision from A.V.J. E.J.R. and P.D. made significant contributions to the theoretical development of the project, design and implementation of statistical models, and critical revision of the manuscript. All authors gave final approval for publication and agree to be held accountable for the work performed therein.

Competing interests

We declare we have no competing interests.

Funding

Emory University Robert W. Woodruff Fellowship to J.S.M. Czech Science Foundation (GACR) Grant 18-23889S to P.D.

Acknowledgements

We thank Kevin Delijani for assistance in data collection and organization, and Jessica Thompson, Abigail Page, Andreas Koenig, and Carel van Schaik for helpful feedback. We also benefitted from the constructive input of three anonymous reviewers and many colleagues during laboratory meetings. J.S.M. would like to dedicate this research to the memory of his beloved late grandmother, Sue Martin (3 June 1947–1 October 2019), who was the best alloparent he could have ever hoped for.

Footnotes

Electronic supplementary material is available online at https://doi.org/10.6084/m9.figshare.c.5089508.

References

1.
Jaeggi AV, Gurven M . 2013 Natural cooperators: food sharing in humans and other primates. Evol. Anthropol. 22, 186-195. (doi:10.1002/evan.21364) Crossref, PubMed, Web of Science, Google Scholar
2.
Kramer KL . 2010 Cooperative breeding and its significance to the demographic success of humans. Annu. Rev. Anthropol. 39, 417-436. (doi:10.1146/annurev.anthro.012809.105054) Crossref, Web of Science, Google Scholar
3.
Boyd R, Richerson PJ, Henrich J . 2011 The cultural niche: why social learning is essential for human adaptation. Proc. Natl Acad. Sci. USA 108, 10 918-10 925. (doi:10.1073/pnas.1100290108) Crossref, Web of Science, Google Scholar
4.
Tomasello M et al. 2012 Two key steps in the evolution of human cooperation: the interdependence hypothesis. Curr. Anthropol. 53, 673-692. (doi:10.1086/668207) Crossref, Web of Science, Google Scholar
5.
Hooper PL, Demps K, Gurven M, Gerkey D, Kaplan HS . 2015 Skills, division of labour and economies of scale among Amazonian hunters and South Indian honey collectors. Phil. Trans. R. Soc. B 370, 20150008. (doi:10.1098/rstb.2015.0008) Link, Web of Science, Google Scholar
6.
Gurven M, Hill K . 2009 Why do men hunt? A reevaluation of ‘man the hunter’ and the sexual division of labor. Curr. Anthropol. 50, 51-74. (doi:10.1086/595620) Crossref, PubMed, Web of Science, Google Scholar
7.
Hill KR et al. 2011 Co-residence patterns in hunter-gatherer societies show unique human social structure. Science 331, 1286-1289. (doi:10.1126/science.1199071) Crossref, PubMed, Web of Science, Google Scholar
8.
Ross CT, Hooper PL, Borgerhoff Mulder M . 2015 Data on the frequency of non-reproductive adults in a cross-cultural sample of small-scale human societies. bioRxiv 032318. (doi:10.1101/032318) Google Scholar
9.
Burkart JM et al. 2014 The evolutionary origin of human hyper-cooperation. Nat. Commun. 5, ncomms5747. (doi:10.1038/ncomms5747) Crossref, Web of Science, Google Scholar
10.
Miller IF, Churchill SE, Nunn CL . 2019 Speeding in the slow lane: phylogenetic comparative analyses reveal that not all human life history traits are exceptional. J. Hum. Evol. 130, 36-44. (doi:10.1016/j.jhevol.2018.12.007) Crossref, PubMed, Web of Science, Google Scholar
11.
Kaplan H, Hill K, Lancaster J, Hurtado AM . 2000 A theory of human life history evolution: diet, intelligence, and longevity. Evol. Anthropol. 9, 156-185. (doi:10.1002/1520-6505(2000)9:4<156::AID-EVAN5>3.0.CO;2-7) Crossref, Web of Science, Google Scholar
12.
Isler K, Van Schaik CP . 2012 How our ancestors broke through the gray ceiling: comparative evidence for cooperative breeding in early homo. Curr. Anthropol. 53, S453-S465. (doi:10.1086/667623) Crossref, Web of Science, Google Scholar
13.
Kaplan H, Gurven M, Hill K, Hurtado AM . 2005 The natural history of human food sharing and cooperation: a review and a new multi-individual approach to the negotiation of norms. In Moral sentiments and material interests: the foundations of cooperation in economic life, pp. 75-113. Cambridge, MA: MIT Press. Google Scholar
14.
Hewlett BS, Winn S . 2014 Allomaternal nursing in humans. Curr. Anthropol. 55, 200-229. (doi:10.1086/675657) Crossref, PubMed, Web of Science, Google Scholar
15.
Kramer KL, Veile A . 2018 Infant allocare in traditional societies. Physiol. Behav. 193, 117-126. (doi:10.1016/j.physbeh.2018.02.054) Crossref, PubMed, Web of Science, Google Scholar
16.
Meehan CL . 2009 Maternal time allocation in two cooperative childrearing societies. Hum. Nat. 20, 375-393. (doi:10.1007/s12110-009-9076-2) Crossref, Web of Science, Google Scholar
17.
Meehan CL, Quinlan R, Malcom CD . 2013 Cooperative breeding and maternal energy expenditure among Aka foragers. Am. J. Hum. Biol. 25, 42-57. (doi:10.1002/ajhb.22336) Crossref, PubMed, Web of Science, Google Scholar
18.
Isler K, Van Schaik CP . 2014 How humans evolved large brains: comparative evidence. Evol. Anthropol. 23, 65-75. (doi:10.1002/evan.21403) Crossref, PubMed, Web of Science, Google Scholar
19.
Quinlan RJ . 2008 Human pair-bonds: evolutionary functions, ecological variation, and adaptive development. Evol. Anthropol. 17, 227-238. (doi:10.1002/evan.20191) Crossref, Web of Science, Google Scholar
20.
Hrdy SB . 2009 Mothers and others. Cambridge, MA: Harvard University Press. Google Scholar
21.
Gurven M, Kaplan H, Gutierrez M . 2006 How long does it take to become a proficient hunter? Implications for the evolution of extended development and long life span. J. Hum. Evol. 51, 454-470. (doi:10.1016/j.jhevol.2006.05.003) Crossref, PubMed, Web of Science, Google Scholar
22.
Emmott EH, Page AE . 2019 Alloparenting. In Encyclopedia of evolutionary psychological science (eds Shackelford T, Weekes-Shackelford V ). Cham, Switzerland: Springer. Crossref, Google Scholar
23.
Antón SC, Potts R, Aiello LC . 2014 Evolution of early Homo: an integrated biological perspective. Science 345, 1236828. (doi:10.1126/science.1236828) Crossref, PubMed, Web of Science, Google Scholar
24.
Potts R . 2013 Hominin evolution in settings of strong environmental variability. Quat. Sci. Rev. 73, 1-13. (doi:10.1016/j.quascirev.2013.04.003) Crossref, Web of Science, Google Scholar
25.
Wynn JG . 2004 Influence of Plio-Pleistocene aridification on human evolution: evidence from paleosols of the Turkana Basin, Kenya. Am. J. Phys. Anthropol. 123, 106-118. (doi:10.1002/ajpa.10317) Crossref, PubMed, Web of Science, Google Scholar
26.
Maslin MA, Susanne S, Martin HT . 2015 A synthesis of the theories and concepts of early human evolution. Phil. Trans. R. Soc. B 370, 20140064. (doi:10.1098/rstb.2014.0064) Link, Web of Science, Google Scholar
27.
Owen RB et al. 2018 Progressive aridification in East Africa over the last half million years and implications for human evolution. Proc. Natl Acad. Sci. USA 115, 11 174-11 179. (doi:10.1073/pnas.1801357115) Crossref, Web of Science, Google Scholar
28.
Pereda M, Zurro D, Santos JI, Godino IB, Álvarez M, Caro J, Galán JM . 2017 Emergence and evolution of cooperation under resource pressure. Sci. Rep. 7, 45574. (doi:10.1038/srep45574) Crossref, PubMed, Web of Science, Google Scholar
29.
Smaldino PE, Schank JC, McElreath R . 2013 Increased costs of cooperation help cooperators in the long run. Am. Nat. 181, 451-463. (doi:10.1086/669615) Crossref, PubMed, Web of Science, Google Scholar
30.
Smaldino PE, Newson L, Schank JC, Richerson PJ . 2013 Simulating the evolution of the human family: Cooperative breeding increases in harsh environments. PLoS ONE 8, e80753. (doi:10.1371/journal.pone.0080753) Crossref, PubMed, Web of Science, Google Scholar
31.
Shen SF, Emlen ST, Koenig WD, Rubenstein DR . 2017 The ecology of cooperative breeding behaviour. Ecol. Lett. 20, 708-720. (doi:10.1111/ele.12774) Crossref, PubMed, Web of Science, Google Scholar
32.
Mesterton-Gibbons M, Dugatkin LA . 1992 Cooperation among unrelated individuals: evolutionary factors. Q Rev. Biol. 67, 267-281. (doi:10.1086/417658) Crossref, Web of Science, Google Scholar
33.
Green JP, Freckleton RP, Hatchwell BJ . 2016 Variation in helper effort among cooperatively breeding bird species is consistent with Hamilton's Rule. Nat. Commun. 7, 12663. (doi:10.1038/ncomms12663) Crossref, PubMed, Web of Science, Google Scholar
34.
Briga M, Pen I, Wright J . 2012 Care for kin: within-group relatedness and allomaternal care are positively correlated and conserved throughout the mammalian phylogeny. Biol. Lett. 8, 533-536. (doi:10.1098/rsbl.2012.0159) Link, Web of Science, Google Scholar
35.
Komdeur J . 1992 Importance of habitat saturation and territory quality for evolution of cooperative breeding in the Seychelles warbler. Nature 358, 493-495. (doi:10.1038/358493a0) Crossref, Web of Science, Google Scholar
36.
Rubenstein DR, Lovette IJ . 2007 Temporal environmental variability drives the evolution of cooperative breeding in birds. Curr. Biol. 17, 1414-1419. (doi:10.1016/j.cub.2007.07.032) Crossref, PubMed, Web of Science, Google Scholar
37.
Lin YH, Chan SF, Rubenstein DR, Liu M, Shen SF . 2019 Resolving the paradox of environmental quality and sociality: the ecological causes and consequences of cooperative breeding in two lineages of birds. Am. Nat. 194, 207-216. (doi:10.1086/704090) Crossref, PubMed, Web of Science, Google Scholar
38.
Jetz W, Rubenstein DR . 2011 Environmental uncertainty and the global biogeography of cooperative breeding in birds. Curr. Biol. 21, 72-78. (doi:10.1016/j.cub.2010.11.075) Crossref, PubMed, Web of Science, Google Scholar
39.
Cornwallis CK, Botero CA, Rubenstein DR, Downing PA, West SA, Griffin AS . 2017 Cooperation facilitates the colonization of harsh environments. Nat. Ecol. Evol. 1, 0057. (doi:10.1038/s41559-016-0057) Crossref, PubMed, Web of Science, Google Scholar
40.
Lukas D, Clutton-Brock T . 2017 Climate and the distribution of cooperative breeding in mammals. R. Soc. Open Sci. 4, 160897. (doi:10.1098/rsos.160897) Link, Web of Science, Google Scholar
41.
Roberts G . 2005 Cooperation through interdependence. Anim. Behav. 70, 901-908. (doi:10.1016/j.anbehav.2005.02.006) Crossref, Web of Science, Google Scholar
42.
Clutton-Brock T . 2002 Breeding together: kin selection and mutualism in cooperative vertebrates. Science 296, 69-72. (doi:10.1126/science.296.5565.69) Crossref, PubMed, Web of Science, Google Scholar
43.
Kokko H, Johnstone RA . 2001 The evolution of cooperative breeding through group augmentation. Proc. R. Soc. B 268, 187-196. (doi:10.1098/rspb.2000.1349) Link, Web of Science, Google Scholar
44.
Covas R, Doutrelant C, du Plessis MA . 2004 Experimental evidence of a link between breeding conditions and the decision to breed or to help in a colonial cooperative bird. Proc. R. Soc. B 271, 827-832. (doi:10.1098/rspb.2003.2652) Link, Web of Science, Google Scholar
45.
Baglione V, Canestrari D, Marcos JM, Griesser M, Ekman J . 2002 History, environment and social behaviour: experimentally induced cooperative breeding in the carrion crow. Proc. R. Soc. B 269, 1247-1251. (doi:10.1098/rspb.2002.2016) Link, Web of Science, Google Scholar
46.
Guindre-Parker S, Rubenstein DR . 2018 Multiple benefits of alloparental care in a fluctuating environment. R. Soc. Open Sci. 5, 172406. (doi:10.1098/rsos.172406) Link, Web of Science, Google Scholar
47.
Hawkes K, Coxworth JE . 2013 Grandmothers and the evolution of human longevity: a review of findings and future directions. Evol. Anthrop. 22, 294-302. (doi:10.1002/evan.21382) Crossref, PubMed, Web of Science, Google Scholar
48.
Sear R, Mace R . 2008 Who keeps children alive? A review of the effects of kin on child survival. Evol. Hum. Behav. 29, 1-18. (doi:10.1016/j.evolhumbehav.2007.10.001) Crossref, Web of Science, Google Scholar
49.
Kramer KL . 2014 Why what juveniles do matters in the evolution of cooperative breeding. Hum. Nat. 25, 49-65. (doi:10.1007/s12110-013-9189-5) Crossref, PubMed, Web of Science, Google Scholar
50.
Quinlan RJ . 2006 Human parental effort and environmental risk. Proc. R. Soc. B 274, 121-125. (doi:10.1098/rspb.2006.3690) Link, Web of Science, Google Scholar
51.
Jaeggi AV, Hooper PL, Beheim BA, Kaplan H, Gurven M . 2016 Reciprocal exchange patterned by market forces helps explain cooperation in a small-scale society. Curr. Biol. 26, 2180-2187. (doi:10.1016/j.cub.2016.06.019) Crossref, PubMed, Web of Science, Google Scholar
52.
Page AE et al. 2019 Testing adaptive hypotheses of alloparenting in Agta foragers. Nat. Hum. Behav. 3, 1154-1163. (doi:10.1038/s41562-019-0679-2) Crossref, PubMed, Web of Science, Google Scholar
53.
Murdock GP, White DR . 1969 Standard cross-cultural sample. Ethnology 8, 329-369. (doi:10.2307/3772907) Crossref, Google Scholar
54.
Barry H, Paxson LM . 1971 Infancy and early childhood: cross-cultural codes 2. Ethnology 10, 466-508. (doi:10.2307/3773177) Crossref, Google Scholar
55.
Kirby KR et al. 2016 D-PLACE: a global database of cultural, linguistic and environmental diversity. PLoS ONE 11, e0158391. (doi:10.1371/journal.pone.0158391) Crossref, PubMed, Web of Science, Google Scholar
56.
Dufour DL, Sauther ML . 2002 Comparative and evolutionary dimensions of the energetics of human pregnancy and lactation. Am. J. Hum. Biol. 14, 584-602. (doi:10.1002/ajhb.10071) Crossref, PubMed, Web of Science, Google Scholar
57.
Konner M . 2018 Nonmaternal care: a half-century of research. Physiol. Behav. 193, 179-186. (doi:10.1016/j.physbeh.2018.03.025) Crossref, PubMed, Web of Science, Google Scholar
58.
Colwell RK . 1974 Predictability, constancy, and contingency of periodic phenomena. Ecology 1, 1148-1153. (doi:10.2307/1940366) Crossref, Web of Science, Google Scholar
59.
Lima-Ribeiro MS, Varela S, González-Hernández J, de Oliveira G, Diniz-Filho JAF, Terribile LC . 2015 EcoClimate: a database of climate data from multiple models for past, present, and future for macroecologists and biogeographers. Biodivers. Inform. 10, 1-21. Crossref, Google Scholar
60.
Revell LJ . 2009 Size-correction and principal components for interspecific comparative studies. Evolution 63, 3258-3268. (doi:10.1111/j.1558-5646.2009.00804.x) Crossref, PubMed, Web of Science, Google Scholar
61.
Jackson JE . 2005 Quartimax Rotation. In Encyclopedia of biostatistics (eds Armitrage P, Colton T ). Hoboken, NJ: John Wiley & Sons. Google Scholar
62.
Dirks R . 1993 Starvation and famine: cross-cultural codes and some hypothesis tests. Cross-Cult. Res. 27, 28-69. (doi:10.1177/106939719302700103) Crossref, Google Scholar
63.
Murdock GP, Provost C . 1973 Measurement of cultural complexity. Ethnology 12, 379-392. (doi:10.2307/3773367) Crossref, Google Scholar
64.
Page AE et al. 2016 Reproductive trade-offs in extant hunter-gatherers suggest adaptive mechanism for the Neolithic expansion. Proc. Natl Acad. Sci. USA 113, 4694-4699. (doi:10.1073/pnas.1524031113) Crossref, PubMed, Web of Science, Google Scholar
65.
Murdock GP, Morrow DO . 1970 Subsistence economy and supportive practices: cross-cultural codes 1. Ethnology 9, 302-330. (doi:10.2307/3773028) Crossref, Google Scholar
66.
Ringen EJ, Duda P, Jaeggi AV . 2019 The evolution of daily food sharing: a Bayesian phylogenetic analysis. Evol. Hum. Behav. 40, 375-384. (doi:10.1016/j.evolhumbehav.2019.04.003) Crossref, Web of Science, Google Scholar
67.
Barry H, Schlegel A . 1982 Cross-cultural codes on contributions by women to subsistence. Ethnology 21, 165-188. (doi:10.2307/3773435) Crossref, Google Scholar
68.
Low BS . 1994 Pathogen intensity cross-culturally. World Cult. 8, 24-34. Google Scholar
69.
Olson DM et al. 2001 Terrestrial ecoregions of the world: a new map of life on earth: a new global map of terrestrial ecoregions provides an innovative tool for conserving biodiversity. BioScience 51, 933-938. (doi:10.1641/0006-3568(2001)051[0933:TEOTWA]2.0.CO;2) Crossref, Web of Science, Google Scholar
70.
Heldstab SA, Isler K, Burkart JM, van Schaik CP . 2019 Allomaternal care, brains and fertility in mammals: who cares matters. Behav. Ecol. Sociobiol. 73, 71. (doi:10.1007/s00265-019-2684-x) Crossref, Web of Science, Google Scholar
71.
Duda P, Zrzavý J . 2016 Human population history revealed by a supertree approach. Sci. Rep. 6, 29890. (doi:10.1038/srep29890) Crossref, PubMed, Web of Science, Google Scholar
72.
Duda P, Zrzavý J . 2019 Towards a global phylogeny of human populations based on genetic and linguistic data. In Modern human origins and dispersal, words, bones, genes, tools: DFG center for advanced studies series (eds Sahle Y, Reyes-Centeno H, Bentz C ), pp. 331-359. Tübingen, Germany: Kerns Verlag. Google Scholar
73.
Minocher R, Duda P, Jaeggi AV . 2019 Explaining marriage patterns in a globally representative sample through socio-ecology and population history: a Bayesian phylogenetic analysis using a new supertree. Evol. Hum. Behav. 40, 176-187. (doi:10.1016/j.evolhumbehav.2018.11.003) Crossref, Web of Science, Google Scholar
74.
Blows MW, Brooks R . 2003 Measuring nonlinear selection. Am. Nat. 162, 815-820. (doi:10.1086/378905) Crossref, PubMed, Web of Science, Google Scholar
75.
Myers RH, Montgomery DC, Anderson-Cook CM . 2016 Response surface methodology: process and product optimization using designed experiments, 4th edn. Hoboken, NJ: John Wiley & Sons. Google Scholar
76.
Bürkner PC . 2017 brms: An R package for Bayesian multilevel models using Stan. J. Stat. Softw. 80, 1-28. (doi:10.18637/jss.v080.i01) Crossref, Web of Science, Google Scholar
77.
Hadfield JD, Nakagawa S . 2010 General quantitative genetic methods for comparative biology: phylogenies, taxonomies and multi-trait models for continuous and categorical characters. J. Evol. Biol. 23, 494-508. (doi:10.1111/j.1420-9101.2009.01915.x) Crossref, PubMed, Web of Science, Google Scholar
78.
McElreath R . 2016 Statistical rethinking: a Bayesian course with examples in R and stan. Boca Raton, FL: CRC Press. Google Scholar
79.
Bürkner PC, Charpentier E . In press. Modelling monotonic effects of ordinal predictors in Bayesian regression models. Br. J. Math. Statist. Psychol. (doi:10.1111/bmsp.12195) Web of Science, Google Scholar
80.
Gelman A, Carlin JB, Stern HS, Dunson DB, Vehtari A, Rubin DB . 2013 Bayesian data analysis. Boca Raton, FL: CRC Press. Crossref, Google Scholar
81.
McShane BB, Gal D, Gelman A, Robert C, Tackett JL . 2019 Abandon statistical significance. Am. Stat. 73, 235-245. (doi:10.1080/00031305.2018.1527253) Crossref, Web of Science, Google Scholar
82.
Gelman A, Hwang J, Vehtari A . 2014 Understanding predictive information criteria for Bayesian models. Stat. Comput. 24, 997-1016. (doi:10.1007/s11222-013-9416-2) Crossref, Web of Science, Google Scholar
83.
Ember CR, Skoggard I, Ringen EJ, Farrer M . 2018 Our better nature: does resource stress predict beyond-household sharing? Evol. Hum. Behav. 39, 380-391. (doi:10.1016/j.evolhumbehav.2018.03.001) Crossref, Web of Science, Google Scholar
84.
Glowacki L, von Rueden C . 2015 Leadership solves collective action problems in small-scale societies. Phil. Trans. R. Soc. B 370, 20150010. (doi:10.1098/rstb.2015.0010) Link, Web of Science, Google Scholar
85.
Botero CA, Gardner B, Kirby KR, Bulbulia J, Gavin MC, Gray RD . 2014 The ecology of religious beliefs. Proc. Natl Acad. Sci. USA 111, 16 784-16 789. (doi:10.1073/pnas.1408701111) Crossref, Web of Science, Google Scholar
86.
Schiefenhovel W . 1989 Reproduction and sex-ratio manipulation through preferential female infanticide among the Eipo, in the Highlands of Western New Guinea. In The sociobiology of sexual and reproductive strategies (eds Rasa AE, Vogel C, Voland E ), pp. 170-193. New York, NY: Chapman and Hall. Google Scholar
87.
Firman RC, Rubenstein DR, Moran JM, Rowe KC, Buzatto BA . 2020 Extreme and variable climatic conditions drive the evolution of sociality in Australian Rodents. Curr. Biol. 30, 691-697. (doi:10.1016/j.cub.2019.12.012) Crossref, PubMed, Web of Science, Google Scholar
88.
Peters MK et al. 2016 Predictors of elevational biodiversity gradients change from single taxa to the multi-taxa community level. Nat. Commun. 7, 1-11. Crossref, Web of Science, Google Scholar
89.
Tallavaara M, Eronen JT, Luoto M . 2018 Productivity, biodiversity, and pathogens influence the global hunter-gatherer population density. Proc. Natl Acad. Sci. USA 115, 1232-1237. (doi:10.1073/pnas.1715638115) Crossref, PubMed, Web of Science, Google Scholar
90.
Griesser M, Drobniak SM, Nakagawa S, Botero CA . 2017 Family living sets the stage for cooperative breeding and ecological resilience in birds. PLoS Biol. 15, e2000483. (doi:10.1371/journal.pbio.2000483) Crossref, PubMed, Web of Science, Google Scholar
91.
Rubenstein DR, Abbot P . (Eds.) 2017 Comparative social evolution. Cambridge, UK: Cambridge University Press. Crossref, Google Scholar
92.
Bourke AF . 2011 Principles of social evolution. Oxford, UK: Oxford University Press. Crossref, Google Scholar
93.
Burkart JM, van Schaik C, Griesser M . 2017 Looking for unity in diversity: human cooperative childcare in comparative perspective. Proc. R. Soc. B 284, 20171184. (doi:10.1098/rspb.2017.1184) Link, Web of Science, Google Scholar