https://orcid.org/0000-0001-6324-6938
Abstract
In birds, extrapair paternity (EPP) constitutes an alternative mating strategy, with potentially important fitness consequences for both males and females and their offspring. Several factors have been identified that can influence the occurrence of EPP, but the role of environmental variability has so far received relatively little attention. Using long-term data set from a wild population of the blue tit (Cyanistes cearuleus), we assess the importance of ambient temperature in modulating the levels of extrapair paternity. Here, we showed that the variability of local thermal conditions affects the occurrence of EPP. Specifically, we found that the probability of EPP increased with rising variability in ambient temperature experienced by females prior to egg laying. This pattern is consistent with an idea of plastic female responses to unpredictable environments. Our results suggest that extrapair mating may represent an adaptive behavioral strategy to compensate for the potential negative effects of unstable environmental conditions.
INTRODUCTION
Given that over 90% of avian species are socially monogamous with biparental care, birds exhibit striking levels of genetic polygamy where individuals can seek for extrapair matings. A series of influential literature reviews of avian studies on extrapair paternity (EPP) (Birkhead and Møller 1995; Griffith et al. 2002; Garcia-Gonzalez et al. 2015; Brouwer and Griffith 2019) showed that the frequency of EPP (i.e., the proportion of nests having at least one extrapair offspring) varied widely among species from 0% in the common loon (Gavia immer) (Piper et al. 1997) to 85% in the superb fairy wren (Malurus cyaneus) (Colombelli-Négrel et al. 2009). Brouwer and Griffith (2019) pointed out that the amount of EPP can be highly variable between species, populations or even within a population (Brouwer and Griffith 2019). Variation in environmental conditions is one of the factors proposed to explain inter- and intra-specific variation in EPP among birds (Birkhead and Møller 1995; Griffith et al. 2002; Brouwer and Griffith 2019), through its effects on the balance of costs and benefits of extrapair matings (Schmoll 2011; Botero and Rubenstein 2012). Considerably less attention was given to short-term, local, and proximate causes of variation in the prevalence of EPP. This omission is particularly glaring, as ecological feedback may eventually translate into selective pressures that shape the distribution and maintenance of EPP in some populations. Although trends in average eco-climatic variables have usually received more research attention, recently growing interest was also focused on their variability and predictability (Botero and Rubenstein 2012). Unpredictable climatic conditions can influence the occurrence of EPP through several mechanisms. First, environmental uncertainty can lead to fluctuations in the availability of resources such as food and territory. Female birds may engage in EPP to access additional resources from extrapair mates, which can be crucial for successful reproduction (Gray 1997) and help offset the costs associated with unpredictable conditions. Second, unpredictable environmental conditions can affect the timing and availability of potential mates (Dunn and Møller 2019). Breeding windows may become shorter or less synchronized due to climate fluctuations. For instance, relatively unfavorable climatic conditions may limit the movement activity during the female fertile period and thus reduce the opportunity for extrapair mating. This can influence the opportunities for pair bonding and increase the likelihood of extrapair mating, as females may encounter more males outside their social pair during these periods of disruption. Third, in uncertain environments, females may use EPP as a form of “bet-hedging” (Garcia-Gonzalez et al. 2015). By having offspring with multiple males, females may seek extrapair mating to increase the genetic diversity of their offspring and potentially improve their fitness in changing or unpredictable environments, via a form of bet-hedging to increase the genetic diversity of their broods (Garcia-Gonzalez et al. 2015). Thus, females enhance the chances that at least some offspring may possess the necessary diversity of alleles required to adequately cope with environmental stochasticity (for more reviews on the effect of these variables on the occurrence of EPP, see Birkhead and Møller 1995; Griffith et al. 2002; Garcia-Gonzalez et al. 2015; Brouwer and Griffith 2019). Increasing the focus on climate variability seems especially crucial in the light of an ongoing global climate emergency caused by the greenhouse effect and rising ambient temperatures (Dunn and Møller 2019). Evidence is accumulating that the recent rapid climate change has important consequences for birds (Bouwman and Komdeur 2006; Dunn and Møller 2019; Bailey et al. 2022). For example, changes in ambient temperature and precipitation patterns can affect migration behavior and timing of breeding (Dunn and Møller 2019; Halupka et al. 2020), which can alter the availability of potential mates and, consequently, the likelihood of extrapair matings. Additionally, habitat changes caused by climate change can negatively affect the social structure of bird communities. For instance, the study of Garvin et al. (2006) showed that EPP prevalence in common yellowthroats (Geothlypis trichas) was higher in the colder season of the two study years. In a coal tit (Periparus ater) population EPP was more common among second broods, which are known to experience more stressful environments (Dietrich et al. 2004). In bluethroats (Luscinia s. svecica), when the temperature was low during the peak of the female fertile period EPP occurred less frequently (Johnsen and Lifjeld 2003).
However, little is known about how birds respond to unpredictable short-term temporal fluctuations in weather conditions. The prevalence of such adverse weather events is expected to increase under most scenarios of global warming (Seneviratne et al. 2021), potentially intensifying selective pressures resulting from unpredictable weather. To bridge this gap, we aimed to quantify the effects of such variability by studying whether temporal variability in local ambient temperature is associated with the occurrence of EPP in blue tits (Cyanistes caeruleus). We focus on the temperature due to its pronounced variability and the well-documented effects of ambient temperature changes on blue tit’s fitness-related traits (Arct et al. 2022a). The reproductive behavior of the blue tit has been extensively studied, which makes it an ideal species to address questions of plasticity in the reproductive strategies (Charmantier et al. 2004; Stenning 2018). Moreover, its phenology can be easily monitored, particularly in populations breeding in artificial cavities such as nest boxes. Furthermore, even though blue tits are monogamous, extrapair paternity is common in this species—typically, EPP occurs in 31–65% of broods and approximately 15% of all nestlings are sired by extrapair males (Delhey et al. 2003; Arct et al. 2013; Santema and Kempenaers 2021). However, the relative role of local weather conditions in determining patterns of the occurrence of EPP in blue tits still remains unknown. Thus, we predicted that females that breed under greater variability in ambient temperature over the breeding period should be more likely to produce extrapair offspring.
METHODS
Study area
Our research spans the years 2008–2016 and is part of a long-term study conducted in a nest-box breeding population of the blue tits on the island of Gotland, Sweden (57°03ʹ N, 18°17ʹ E). In our population, females lay a single clutch per season, and each year, almost all breeding attempts are recorded, and as many reproducing individuals as possible are captured, for a more detailed description of the study site and general field procedure (for details, see Drobniak et al. 2022).
Genetic and paternity analyses
We took a blood sample (ca. 20 µL) from nestlings on the second day after hatching, and from adults when they were caught by mist-nets or inside the nest boxes while feeding 14-day-old nestlings, DNA was extracted from blood samples with Chelex following the procedure described in Griffiths et al. (1998). In the following EPP analyses, the number of microsatellite loci used to determine parentage varied between years. In 2009–2010 and 2012–2015, the paternity was identified using five microsatellite loci (see details Arct et al. 2013). In 2011, we used 15 microsatellite loci (Arct et al. 2022b), for a more detailed description of parentage assignment in blue tits, see Arct et al. (2013, 2017). In our population, each year, on average, 30% of nests contain at least one EPY (range 14–45%). In our population, a relatively high proportion of females seek extrapair paternity; however, we have a small proportion of extrapair young in the nests—7.09% (see also Arct et al. 2013, 2022b). For this reason, we treated EPP as a binary response, as we did in other previous studies (Arct et al. 2022b).
Temperature data
We acquired temperature data using the website of the Swedish Meteorological and Hydrological Institute (https://www.smhi.se/en). We obtained data from the meteorological station located in Hemse (56.92°N, 18.15°E; approximately 10 km north from the study area). Raw climate data were used to compute the following descriptive measures of the local variation in ambient temperature for a period of 14 days prior to the laying of the first egg for each clutch:
mean ambient temperature (°C) calculated within those 14 days;
temperature variability (i.e., variance of average daily temperatures across those 14 days, expressed in °C2).
Statistical analysis
To analyze the relationship between local thermal conditions and the occurrence of extrapair offspring (EPO) in the nest, we used a generalized linear mixed model (GLMM) with a binomial error distribution and a logit-link function. The response variable was absence/presence of EPO (0/1) in the nest (i.e., the linear predictor expresses the logged odds ratio of observing EPP in females at the brood level). We ran the model using the glmer function from the lme4 package, ver. 1.1-31 (Bates et al. 2015) in R, ver. 4.2.1 (2022). The model included average mean temperature and temperature variability which both were of primary interest. However, due to the fact that some other variables may significantly affect EPP we also entered into this model additional terms such as social male’s/female’s age, laying date, and clutch size. Parental age was treated as categorical factor, and remaining variables were continuous covariates. All covariates were standardized across years with a mean of 0 and a standard deviation of 1. The year of study and female identity were used as categorical random factors. There was no problem with multicollinearity among explanatory variables (all VIF ≤ 1.71).
RESULTS
We found that the probability of EPP in blue tit pairs was not affected by mean ambient temperature but was significantly related to temperature variability in the period of 14 days prior to egg laying (Table 1). The probability of EPP at the brood level increased with increasing variability in ambient temperature (Figure 1). Specifically, the probability of EPP occurring in broods increased from a little below 25% in less variable thermal conditions to 50% in more variable (less predictable) temperatures. Moreover, the probability of EPP tended to be influenced by clutch size and female age (Table 1).
Results of GLMM model with absence/presence (0/1) of extrapair offspring in the nest as a response variable, mean temperature, temperature variability, laying date, and clutch size as covariates (all centered and standardized), and female and male age as categorical variables
| Term | Estimate | Standard error | z value | P |
|---|---|---|---|---|
| Intercept | −1.21 | 0.29 | −4.15 | <0.001 |
| Male age Female age | 0.02 0.55 | 0.28 0.29 | 0.06 1.86 | 0.948 0.063 |
| Laying date Clutch size | 0.00 0.30 | 0.18 0.16 | 0.01 1.93 | 0.996 0.054 |
| Mean temperature | −0.18 | 0.18 | −1.01 | 0.314 |
| Temperature variability | 0.33 | 0.15 | 2.13 | 0.033 |
| Random factors | ||||
| Female identity | 0.26 | |||
| Year of study | 0.00 |
| Term | Estimate | Standard error | z value | P |
|---|---|---|---|---|
| Intercept | −1.21 | 0.29 | −4.15 | <0.001 |
| Male age Female age | 0.02 0.55 | 0.28 0.29 | 0.06 1.86 | 0.948 0.063 |
| Laying date Clutch size | 0.00 0.30 | 0.18 0.16 | 0.01 1.93 | 0.996 0.054 |
| Mean temperature | −0.18 | 0.18 | −1.01 | 0.314 |
| Temperature variability | 0.33 | 0.15 | 2.13 | 0.033 |
| Random factors | ||||
| Female identity | 0.26 | |||
| Year of study | 0.00 |
Random factors were female identity and the year of study. Significant terms (P < 0.05) are in bold. Sample size included 275 broods from 267 females collected during 8 study years.
Results of GLMM model with absence/presence (0/1) of extrapair offspring in the nest as a response variable, mean temperature, temperature variability, laying date, and clutch size as covariates (all centered and standardized), and female and male age as categorical variables
| Term | Estimate | Standard error | z value | P |
|---|---|---|---|---|
| Intercept | −1.21 | 0.29 | −4.15 | <0.001 |
| Male age Female age | 0.02 0.55 | 0.28 0.29 | 0.06 1.86 | 0.948 0.063 |
| Laying date Clutch size | 0.00 0.30 | 0.18 0.16 | 0.01 1.93 | 0.996 0.054 |
| Mean temperature | −0.18 | 0.18 | −1.01 | 0.314 |
| Temperature variability | 0.33 | 0.15 | 2.13 | 0.033 |
| Random factors | ||||
| Female identity | 0.26 | |||
| Year of study | 0.00 |
| Term | Estimate | Standard error | z value | P |
|---|---|---|---|---|
| Intercept | −1.21 | 0.29 | −4.15 | <0.001 |
| Male age Female age | 0.02 0.55 | 0.28 0.29 | 0.06 1.86 | 0.948 0.063 |
| Laying date Clutch size | 0.00 0.30 | 0.18 0.16 | 0.01 1.93 | 0.996 0.054 |
| Mean temperature | −0.18 | 0.18 | −1.01 | 0.314 |
| Temperature variability | 0.33 | 0.15 | 2.13 | 0.033 |
| Random factors | ||||
| Female identity | 0.26 | |||
| Year of study | 0.00 |
Random factors were female identity and the year of study. Significant terms (P < 0.05) are in bold. Sample size included 275 broods from 267 females collected during 8 study years.
Relationship between the probability of extrapair paternity and temperature variability. Presented are trends (black line) with 95% confident intervals (gray polygons) predicted from the model and raw data (gray points).
DISCUSSION
Over the past few decades, a number of well-established hypotheses have been proposed to explain the evolution of female extrapair mating (e.g., good genes or compatible genes) (Griffith et al. 2002; Arct et al. 2015; Brouwer and Griffith 2019). Our results provide the first empirical evidence for the effects of short-term temporal fluctuations in thermal conditions on the long-term trends in the occurrence of EPP in a wild population of blue tits. This pattern is consistent with an idea of plastic female responses to unpredictable environmental conditions.
Our results indicate that the probability of EPP in blue tits was not influenced by the mean ambient temperature but was significantly impacted by temperature variability shortly before egg laying. Specifically, the probability of EPP occurring in broods increased from a little below 25% in less variable thermal conditions to 50% in more variable (less predictable) temperatures. This result is consistent with a comparative analysis showing a relationship between the variability and predictability of annual climatic cycles and the occurrence of extrapair paternity among birds on a macroevolutionary scale (Botero and Rubenstein 2012). Moreover, these results are in line with other studies suggesting that weather conditions may explain at least part of the variation seen in the prevalence EPP within species (Johnsen and Lifjeld 2003; Bouwman and Komdeur 2006). Specifically, the study of Johnsen and Lifjeld (2003) on bluethroats (Luscinia s. svecica) showed that relatively low temperatures during the peak of the fertile period were associated with lower proportions of EPP in the brood. Similarly, in the socially monogamous snow bunting (Plectrophenax nivalis), the occurrence of EPP was positively associated with the average minimum temperature during the female fertile period (Hoset et al. 2014). On the other hand, Bouwman and Komdeur (2006) showed that in the reed bunting (Emberiza schoeniclus), the lower proportion of EPO was related to higher daily minimum temperatures and more rainfall during the peak fertile period of the female.
The ultimate mechanistic foundations of all these patterns may actually originate at the level of offspring fitness. Extrapair paternity has been suggested to provide indirect genetic benefits to females by producing offspring superior in terms of fitness compared to within-pair siblings (Griffith et al. 2002; Pryke et al. 2010; Arct et al. 2015). Many studies failed to provide evidence of universal fitness advantage of extrapair offspring—but it is possible that extrapair mating behavior may represent an adaptive behavioral strategy to compensate for possible negative effects of poor environmental conditions. In such cases, the expected indirect benefits would manifest themselves only in certain conditions (a form of genotype-by-environment interaction), much like a bet-hedging strategy (Garcia-Gonzalez et al. 2015). For instance, in the coal tit (Periparus ater), EPO were more likely to recruit to the local breeding population than their maternal half-siblings only when they originated from second broods, when overall performance had significantly declined (Schmoll et al. 2005). In the common yellowthroat (Geothlypis trichas), EPO exhibited a stronger cellular immune response in comparison to WPO, but this was only observed in the colder of the two study years (Garvin et al. 2006), so again the superior performance of EPO in that study coincided once more with environmental conditions that were relatively unfavorable (Garvin et al. 2006). Our previous experimental study in the same blue tit population revealed that EPO mounted a higher immune response in comparison to WPO, but this was only apparent among offspring from experimentally enlarged broods (i.e., where offspring experienced nutritional restriction and elevated intra-brood competition) (Arct et al. 2013). Thus, females producing more EPY could be tapping into this environment-dependent source of hidden genetic variance, hedging their bets in less predictable conditions, and providing at least some of their offspring with increased fitness prospects.
Even though EPP was related to the variability of local thermal conditions, the correlative nature of our study calls for further work to distinguish between effects arising due to thermal conditions and other potential confounding factors. Thus, experimental and longitudinal studies would be needed to establish a causal relationship and better understand the underlying mechanisms driving this correlation. Moreover, our study represents broader-scale environmental variation since we obtained data from a local meteorological weather station. Thus, our data may not fully capture finer-scale environmental variation in thermal conditions. Indeed, while the weather station data provide valuable insights into the macro-environment, we recognize that for a more comprehensive understanding of the finer-scale environmental factors experienced by territorial birds like blue tits, additional locally collected environmental data may be beneficial in future research (Hinks et al. 2015).
To our knowledge, our study provides the first evidence of a significant influence of the variability of local thermal conditions on the occurrence of extrapair paternity at the brood level in a short-lived socially monogamous bird population. Although we were focused on the effects of local weather conditions on extrapair paternity, this type of effects may be important for other female mating tactics, for example, initial mate choice and divorce. We argue that a temporal perspective should be considered in any study investigating the role of the environment on the prevalence of extrapair paternity in other socially monogamous populations.
Acknowledgment
We thank the many students and field assistants that have contributed to the collection of the data. A special thanks to Edyta Podmokła and Blandine Doligez and her team. We also thank landowners and farmers from Gotland for their kind help in providing access to the study plots.
FUNDING
This work was supported by National Science Centre grant no. 2015/19/D/NZ8/01331 to A.A. Fieldwork was supported by the National Science Centre, Poland, grant no. 2012/07/D/NZ8/01317 to S.M.D., Ministry of Science and Higher Education grants no. N N304 061140 to S.M.D. and 2 P04F 004 29 to M.C. and the State Committee for Scientific Research, Republic of Poland, grant no. 2 P04F 026 27 to A.D. and Jagiellonian University, Institute of Environmental Sciences statutory grant no. N18/DBS/000003.
AUTHOR CONTRIBUTIONS
Aneta Arct (Conceptualization [Lead], Data curation [Supporting], Formal analysis [Supporting], Funding acquisition [Lead], Investigation [Lead], Methodology [Lead], Project administration [Lead], Supervision [Lead], Writing—original draft [Lead], Writing—review & editing [Lead]), Szymon Drobniak (Conceptualization [Supporting], Data curation [Supporting], Formal analysis [Supporting], Writing—original draft [Supporting], Writing—review & editing [Supporting]), Lars Gustafsson (Conceptualization [Supporting], Funding acquisition [Supporting], Supervision [Supporting]), and Mariusz Cichon (Conceptualization [Supporting], Data curation [Supporting], Writing—original draft [Supporting])
ETHICAL APPROVAL
The data upon which this study is based was obtained following the Swedish guidelines for work on natural populations and under a permit from the Swedish Ringing Centre at the Museum of Natural History in Stockholm (permit no. RC712 to S.M.D.).
CONFLICT OF INTEREST
The authors have no competing interest to declare.
DATA AVAILABILITY
Analyses reported in this article can be reproduced using the data provided by Arct et al (2023).
