Urban environments are expanding worldwide with an unprecedented speed (United Nations, 2014); environmental change associated with urbanization is being considered as one of the main current threats to biodiversity (Turner et al., 2004). However, urbanization also represents a potential source of selection and a new avenue to study evolutionary processes (Donihue and Lambert, 2014; Isaksson, 2015; Alberti et al., 2017; Hendry et al., 2017).

For wildlife, urban habitats differ in many respects from rural habitats, including microclimate, food abundance, pollution, abundance of exotic species, and predation risk (Luniak and Pisarski, 1982; Marzluff et al., 2001; Gaston, 2010; Gil and Brumm, 2014). Many of these factors induce stress, defined in a broad sense as changes away from an organisms' physiological homeostasis that emerge in response to a change in the environment (sensu Badyaev, 2005; Tuomainen and Candolin, 2012). This is especially the case for factors that represent, or are associated with increased perceived predation risk, such as elevated disturbance due to the presence of humans, pet mammals, and cars (Buchanan and Partecke, 2012). The obvious hypothesis based on this known disturbance or perceived threat is that urban animals would have an enhanced stress response than rural-dwelling animals. However, the results are mixed. For example, some studies reveal higher corticosterone levels (avian stress hormones) in urban compared to rural bird populations (Schoech et al., 2007; Fokidis et al., 2009; Zhang et al., 2011), whereas other studies found the opposite pattern or no significant differences (Partecke et al., 2006; French et al., 2008; Abolins-Abols et al., 2016).

Nevertheless, whatever the direction of the effects, they are likely to influence aspects of behavior, physiology, and life history of urban-dwellers in their coping response to the urban environment (Tuomainen and Candolin, 2012; Sprau and Dingemanse, 2017). Consequently, stress-responses, i.e., the behavioral and physiological means by which animals cope with environmental stress, play an important role in local adaptation (Badyaev, 2005). Hence, comparing stress-responses between urban- and rural-dwelling populations can shed light on the mechanisms underpinning adaptation to urban habitats. For example, predatory attack is most likely one of the highest stresses that a prey animal can experience, and this is irrespective of whether the animal is habituated to the urban environment. Behavioral responses to such threats are commonly referred to as distress behaviors (Møller and Ibáñez-Álamo, 2012). Previous research has shown that distress behaviors displayed during handling are parts of the same phenotypic syndrome that encapsulates an entire suite of behavioral and physiological traits that individual display in order to adaptively respond to stressful situations (Koolhaas et al., 1999, 2010; Carere and van Oers, 2004; Coppens et al., 2010; Carvalho et al., 2013; Class et al., 2014; Kluen et al., 2014). Responses can be categorized into different “coping styles,” varying along a continuum from “proactive” to “reactive” copers, and they relate to the way organisms deal with stressors. Summarizing, proactive individuals are more aggressive, explorative, neophilic, and risk-prone than reactive individuals (Carere et al., 2010).

Accordingly, birds displaying a high frequency of distress behaviors represent proactive individuals. We focus in this paper on three distress behaviors easily recorded during handling: distress calling, handling aggression, and breath rate. Distress calling rate has previously been related to proactivity: in black-capped chickadees Poecile atricapillus, distress calling rate was positively related to exploration rate, which is one of the main characteristics of the proactive coping style (Guillette and Sturdy, 2011). In siskins Carduelis spinus, individuals uttering more often distress calls, displayed also bolder behaviors in front of a novel object (Mateos-González and Senar, 2012; Pascual and Senar, 2014). In the Japanese quail Coturnix japonica, distress calling was positively related to the excretion of corticosterone metabolites, implying that high levels of distress calling are related to proactivity (Niall Daisley et al., 2005). Distress calling has therefore been used as a proxy of proactivity in several studies (Andersen, 2012; Pascual and Senar, 2014; Thorsteinsen, 2015; Richardson et al., 2016). Aggression has also been regarded as a typical response of proactive copers to stressful situations (Koolhaas et al., 1999; Carere et al., 2010), and pecking rate during handling is classically used to compare level of proactivity among individuals (Brommer and Kluen, 2012; van den Brink et al., 2012a,b; Class et al., 2014; Kluen et al., 2014; Dubuc-Messier et al., 2017). Breath rate has also been suggested as an indicator of coping style, with breath rates being higher in proactive individuals than in reactive ones (Carere and van Oers, 2004; van Oers and Carere, 2007; Torné-Noguera et al., 2014; Charmantier et al., 2017). Additionally, breath rate has been found to correlate positively to exploration rate, which as previously stated, is one of the main characteristics of the proactive coping style (Charmantier et al., 2017). The three behaviors have also been found to be heritable (Koenig et al., 1991; Brommer and Kluen, 2012). Therefore, these three distress behaviors can be used as an accurate proxy of coping style and their study may allow a broader understanding of the role of coping style mediating processes of urbanization in birds.

Recent work has found that distress behaviors during handling differ between urban and rural population of birds. For example, an interspecific comparison revealed that urban birds showed higher frequencies of distress calling when handled than their rural counterparts (Møller and Ibáñez-Álamo, 2012). Aggression, another typical response to stressful situations (Koolhaas et al., 1999), has also been found to differ in inter-specific comparisons between urban and rural individuals during handling (Møller and Ibáñez-Álamo, 2012). At the intra-specific level, urban birds have also been shown to display different breath rates compared to their rural counterparts during handling (Torné-Noguera et al., 2014; Abolins-Abols et al., 2016; Charmantier et al., 2017). Yet, this relationship may differ between species. To date, studies comparing distress behaviors between urban and rural bird populations (Møller and Ibáñez-Álamo, 2012; Torné-Noguera et al., 2014) have lacked sufficient replication to draw firm conclusions regarding the relationship between urbanization and avian distress behavior (Evans et al., 2009).

The main aim of the present paper was to investigate whether urban and rural populations of the great tits (Parus major) differ consistently in their behavioral response to an acute perceived threat, assayed during capturing and handling (Møller and Nielsen, 2010; Markó et al., 2013; Class et al., 2014; Kluen et al., 2014; Dubuc-Messier et al., 2017). To acquire a robust insight into the relationship between urbanization and distress behavior, we assayed three different distress behaviors (distress calling rate, pecking rate, and breath rate) in seven replicate pairs of urban and rural/forest populations across Europe. Previous work on this species has revealed many differences between urban and rural great tits in terms of morphology (Horak et al., 1995; Senar et al., 2014; Biard et al., 2017), physiology (Andersson et al., 2015; Salmón et al., 2016; Toledo et al., 2016), genetics (Björklund et al., 2010; Riyahi et al., 2015), life-history (Berressem et al., 1983; Schmidt and Einloft-Achenbach, 1984; Isaksson and Andersson, 2007; Hedblom and Soderstrom, 2012; Bailly et al., 2015; Vaugoyeau et al., 2016; Charmantier et al., 2017), population dynamics (Horak and Lebreton, 1998), and behavior (Slabbekoorn and Peet, 2003; Salaberria and Gil, 2010; Riyahi et al., 2017). Urban great tits have been found to be more explorative and less neophobic than rural tits (Tryjanowski et al., 2016; Charmantier et al., 2017; Riyahi et al., 2017), to display shorter flight initiation distances (Møller et al., 2015), and to show a higher problem solving performance than rural birds (Preiszner et al., 2017). Based on this growing body of literature, which suggests that a proactive coping strategy fits urban conditions, we predict that urban great tit populations show a higher distress calling rate, pecking rate and breath rate compared to their rural conspecifics.

The selected urban areas had significantly higher urbanization scores than the selected rural areas [F_{(1, 13)} = 96.08, p < 0.0001] based on the satellite data, confirming that our assignments were appropriate.

Distress calling rate showed a low and non-significant between-year repeatability within individuals (Table 2). However, repeatabilities of pecking and breath rates were of higher magnitude, and were similar between rural and urban populations (Table 2). Distress calling rate and pecking rate were positively correlated within populations (Table 3) but distress calling and pecking rates did not correlate significantly with breath rate (Table 3), suggesting that breath rate was independent from the two other traits. None of the between-population correlations were statistically significant (Table 3).

Distress calling rate was higher in birds from urban compared to rural habitats (Figure 1, Table 4). There was a non-significant trend toward an interaction between sex and habitat on distress calling rate (Table 4). Though the difference in distress calling rate between urban and rural areas tended to be stronger in females compared to males, urban great tits of both sexes had higher distress calling rates compared to their rural conspecifics. The interaction between locality and habitat could not be tested because models incorporating this fixed effect did not converge.

In the full model for pecking rate, there was no significant effect of habitat, age or their interaction (habitat: LRT = 2.75, p = 0.10; age LRT = 0.00, p = 0.98; habitat × age: LRT = 0.29, p = 0.59). Both locality and population were significantly related with great tit pecking rate (locality: LRT = 7.24, p = 0.01; population: LRT = 2.41, p < 0.01). In addition, males displayed higher pecking rates than females (sex: LRT = 4.63, p = 0.03). However, we also found a non-significant trend for an interaction between sex and habitat on pecking rate (sex × habitat: LRT = 3.34, p = 0.07). A post-hoc analysis separating the sexes, revealed that there was an effect of habitat on pecking rate in males (LRT = 4.79, p = 0.03, Figure 2, Table 5) but not in females (LRT = 0.01, p = 0.95). The interaction between locality and habitat was non-significant (LRT: 0.34, p = 0.95) in the model for males indicating that the higher pecking rate of urban than rural males was consistent across localities (Figure 2).

Breath rate did not differ between urban and rural habitats when controlling for sex, age, locality, population, and observer effects (Table 6). The interaction between locality and habitat was also non-significant (LRT < 0.001, p = 0.99), indicating that this lack of difference was consistent across localities.

The present study reveals consistent differences in distress calling rate between urban and rural populations of great tits, across multiple European localities, with urban birds having, in general, a higher distress calling rate than their rural conspecifics. In line with this finding, pecking rate showed similar tendencies, but only for males. By contrast, we did not find a general difference in breath rate between urban and rural birds.

Distress behaviors displayed during handling have been hypothesized to reflect the coping style of the different individuals in the presence of an acute stressful situation (Brommer and Kluen, 2012; Class et al., 2014; Kluen et al., 2014; Class and Brommer, 2016). Among the three measured distress behaviors, only distress calling rate showed a clear habitat effect. The higher distress calling rate of urban great tits compared to their rural counterparts is consistent with findings of previous studies that compared bird species (Møller and Ibáñez-Álamo, 2012). Given that high distress calling rates have been related to a proactive coping style (Guillette and Sturdy, 2011; Pascual and Senar, 2014), our results support the view that urban birds are more proactive than rural birds.

Because distress calling and pecking rates were correlated, we expected similar habitat differences for the two traits. Indeed, pecking rate was related to urbanization, but only for males. This sexual difference has previously been observed (Markó et al., 2013), and is congruent with the general observation that males are more aggressive than females. Under this general view, the higher aggression of urban male great tits could be boosted by intense intra-specific competition for a limited number of territories and a potentially higher breeding density in urban areas (Chamberlain et al., 2009). In line with this finding, elevated levels of male territorial aggression have been found in urban song sparrows Melospiza melodia, compared to rural male conspecifics. However, this pattern is not necessarily supported in other species (Newman et al., 2006; Bókony et al., 2010; Atwell et al., 2014; Hasegawa et al., 2014). Furthermore, bird species living in urban landscapes have been shown to peck in lower frequencies in response to handling than those inhabiting rural habitats (Møller and Ibáñez-Álamo, 2012). Yet, Møller and Ibáñez-Álamo (2012) did not discriminate between males and females, which hampers a straightforward comparison with our findings. Taken together, these inconsistent results do not allow assessing whether pecking rate is more related to aggression and the latter trait that differ between habitat types, or if it is per se a distress behavior that is subjected to selection. In any case, both aggression and behavior upon handling have been regarded as typical indicators of the same proactive coping style (Koolhaas et al., 1999; Carere et al., 2010; Class et al., 2014; Kluen et al., 2014; Dubuc-Messier et al., 2017). Accordingly, our data on great tit pecking rate suggests that urban males are more proactive than rural males. The interaction between habitat and locality was not significant, indicating that the pecking rate differences among habitats was consistent across localities.

In songbirds, breath rate has been proposed as an indicator of acute stress resulting from a predatory attack, and more generally, as an indicator of stress sensitivity and coping style (Carere and van Oers, 2004; van Oers and Carere, 2007; Torné-Noguera et al., 2014). Indeed, some studies found higher breath rates in urban great tits compared to their rural counterparts (Torné-Noguera et al., 2014; Charmantier et al., 2017). Our study suggests that this finding does not represent a general pattern. In other species, such as dark-eyed juncos Junco hyemalis, urban birds even had a lower breath rate than their rural conspecifics (Abolins-Abols et al., 2016). Our study further reveals that breath rate did not correlate with any of the other behaviors in great tits. This is surprising, since a positive association between pecking rate and breath rate has previously been documented in the closely related blue tit Cyanistes caeruleus (Brommer and Kluen, 2012). However, breath rate has been shown to be affected by seasonal, annual and other ecological factors, hence it may be especially prone to micro-climatic and habitat variation which may mask links to both general urbanization and other behaviors (Torné-Noguera et al., 2014; Charmantier et al., 2017).

The relatively low repeatability of distress calling, potentially due to the low number of repeats per individual (Dingemanse and Dochtermann, 2013; Markó et al., 2013), points to a conservative measure of differences among habitats. Despite the low repeatability, there existed strong differences in distress calling among rural and urban individuals, what points to important differences among habitats, suggesting that distress calling may have evolved in response to urban-induced selection pressures. The between-year repeatability estimates for pecking and breath rate that we recorded, were generally higher than repeatabilities for other behaviors reported in literature (Bell et al., 2009). Given that male pecking rate was different across the habitats and that the repeatability was high, this particular trait seems part of the “urban behavioral phenotype.” This finding is also supported by a recent study indicating a non-random distribution of behavioral types along urban gradients, but little behavioral plasticity in response to within-individual variation to urbanization (Sprau and Dingemanse, 2017).

The combined datasets from different European projects across the distribution of great tits enabled us to conduct a cross-population comparison over a large geographic scale. It entailed study-specific differences in capture time, trapping methods, and observer effects. Nevertheless, we were able to detect consistent differences between habitats, implying that effects reported should be conservative. Altogether, our data suggest that urban great tits, especially males, have a more proactive coping strategy when dealing with stressful conditions than rural birds. This finding is in line with studies showing that urban great tits are more explorative and less neophobic than their rural counterparts (Tryjanowski et al., 2016; Charmantier et al., 2017; Riyahi et al., 2017), and display shorter flight initiation distances (Møller, 2008, 2012), which represent other aspects of the proactive coping strategy. Findings on the great tit are also in line with data from other bird species, for which urban populations have also been found to display proactive behaviors (Evans et al., 2010; Carrete and Tella, 2017), but these studies compared few populations (Miranda et al., 2013). Our study included 14 populations from a large geographic area and revealed consistent results across localities, allowing to generalize the impact of urbanization on great tit coping styles.

This proactive strategy of urban great tits could be advantageous in the urban environment. Proactive individuals, for instance, more rapidly discover and use new food resources (van Overveld and Matthysen, 2010), and are less afraid of novel objects or environments than reactive birds (Tryjanowski et al., 2016; Charmantier et al., 2017; Riyahi et al., 2017), which could help urban dwellers to take profit of the new feeding opportunities that a city can provide (e.g., Fisher and Hinde, 1949). Our study therefore implies that urban birds may adapt to human disturbance by a process of local adaptation, as suggested previously (Partecke et al., 2006; Sol et al., 2013). Future research should investigate whether the differences in distress behaviors reported here have an underlying genetic base, hence are due to local adaption caused by divergent natural selection, or originate from phenotypic plasticity.

Behavioral data was recorded during ringing operations. Capture and ringing was performed under individual ringing permits delivered by the ringing office within each country.

The study was conceived by JCS, LG, VT, CB, GM-R, and CI. The data were collected by VT, CB, GM-R, PaS, JR, PhS, ND, AC, VD, and HN. The statistical analyses were conducted by JCS and LG. The manuscript was written by JCS and LG with input from all other co-authors. Contributed funding and materials: JCS, VT, CB, GM-R, PhS, ND, AC, CI.