Population differences in the structure and coloration of great tit contour feathers

Introduction

Contour feathers, which cover most of the avian body, provide aerodynamic shape, insulation, and surface for visual signals. Plumage is a crucial insulative layer that helps birds in maintaining water and temperature homeostasis (Stettenheim, 2000). Within species, variation in the total contour feather mass has been found both within and among populations as a consequence of a seasonal acclimatization process, presumably related to changes in thermal insulation (Swanson, 1991; Saarela, Klapper & Heldmaier, 1995). Furthermore, populations experiencing varying winter conditions have been found to differ in their thermal conductance and metabolic adjustments (Dawson et al., 1983; Swanson, 1993; Broggi et al., 2004). Contour feather structure is a plastic trait, which varies according to the environmental, nutritional and physiological conditions experienced during molt (Broggi et al., 2011; Vágási et al., 2012). However, although feather structure appears to play an important role in thermal insulation (Middleton, 1986), population variation in contour feather structure and thermal conductance remains unclear; but see also Wolf & Walsberg (2000).

Birds strongly rely on visual traits for social communication and plumage is the main trait involved in such information exchange (Savalli, 1995; Hill & McGraw, 2006). Plumage coloration results from the deposition of pigments in feathers (melanins, carotenoids, and porphyrins) and differences in feather microstructure, as well as a combination of both. Differences in plumage coloration have been intensively studied with respect to individual quality (Hill & McGraw, 2006). In particular, carotenoid-based coloration has been studied as a condition-dependent signal. Because carotenoids cannot be synthesized de novo by animals and need to be acquired (Fox, 1976), individuals may be constrained by limited access to such compounds (Olson & Owens, 1998). Carotenoids also play a role as antioxidants and immune-modulators, and individuals balance their physiological use with the signalling properties, making carotenoids good candidates as honest signals of individual quality (Pérez-Rodríguez, 2009). Besides pigments, plumage coloration can arise as a consequence of changes in feather microstructure (structural coloration) producing blue, green, purple and iridescent coloration (Finger, Burkhardt & Dyck, 1992; Prum, 2006). The final appearance of plumage coloration often results from the interaction between pigmented and structural colours (Shawkey & Hill, 2005; D'Alba et al., 2014), as well as other factors such as the shape of the feathers (Badyaev & Landeen, 2007).

Besides serving a signalling function, the development of colourful feathers can have a physical influence on feather microstructure. For example, deposition of melanin can directly affect the physical properties of the feather by improving resistance to abrasion (Burtt, 1986; Bonser, 1995; Roulin et al., 2013) and decrease bacterial and lice-chewing degradation (Kose & Møller, 1999; Gunderson et al., 2008; Burtt, 2009); but see also Grande, Negro & Torres (2004). Recent studies also suggest that populations may adaptively increase feather melanization when exposed to unfavourable conditions for the plumage (Peele et al., 2009). Furthermore, structurally-based coloration such as in iridescent feathers is known to impair plumage hydrophobicity (Eliason & Shawkey, 2011). Therefore, understanding microstructure variation in colourful contour feathers requires the consideration of the different kinds of pigments and structural coloration involved.

Feather development costs may arise from different selective pressures derived from varying ecological, social or physiological circumstances. Honest advertisement models posit that sexually-selected traits are costly to produce, maintain or bear, with brightly coloured feathers being classic examples of such traits. The cost of pigmented feathers is often ascribed to the value of the pigment itself, whereas the physiological cost of producing the pigmented feather and the survival implications of the general appearance of the plumage (e.g. increased predation risk) should also be considered as important costs (Hill & McGraw, 2006).

Plumage is replaced by periodic molts as it wears and deteriorates over time. Environmental and physiological conditions during feather growth are known to affect feather quality (Strochlic & Romero, 2008; Butler, Leppert & Dufty Jr, 2010; Moreno-Rueda, 2010; Pap et al., 2013). Furthermore, molting is energetically costly (Hoye & Buttemer, 2011) and molt speed is known to adversely affect feather structure (Vágási et al., 2012) and the expression of certain plumage ornaments (Vágási, Pap & Barta, 2010). Feather production is a demanding process in terms of time and resources, and molting individuals may be exposed to trade-offs with other costly activities such as reproduction (Bensch et al., 1985; Siikamäki, Hovi & Rätti, 1994), migration (de la Hera, Pérez-Tris & Tellería, 2009) or molting speed (Dawson et al., 2000). Indeed, few bird species do overlap breeding and molting on a regular basis, particularly in seasonal environments (Dawson, 2008). Thus, the final characteristics of feathers may depend on the balance between the available energy, the requirements for plumage development (Butler, Rohwer & Speidel, 2008) and the different functions that feathers serve. Additionally, as individuals are exposed to different environmental and feeding conditions across their geographical range, selective pressures and/or constraints upon different functions of the birds' plumage may change accordingly.

The great tit Parus major L. is a resident passerine distributed across Eurasia, with populations subjected to very different ecological conditions such as seasonality and food availability (Sanz, 1998). This has led to maladaptive life-history strategies at the borders of their distribution (Rytkönen & Orell, 2001), where great tits are often confronted with time and nutritional constraints affecting molting phenology and speed (Nilsson & Svensson, 1996). Previous studies have shown that great tits from different populations differ in their feather structure, and these differences are likely determined by nutritional constraints (Broggi et al., 2011). We studied population variation in contour feather structure and coloration (carotenoid-based and structural) among three wild European great tit populations located in the southern, middle, and northern portions of the species distribution range, which are exposed to very different environmental and ecological conditions, particularly during the nonbreeding season. We investigated the relationship between different contour feather traits across populations to find out whether feather traits vary in concert or independently, and in accordance with the different ecological/environmental circumstances encountered.

Material and Methods

We captured 61 wild great tits from three locations: 25 in Oulu, Finland (65°N, 25°30′E); 12 in Lund, Sweden (55°40′N, 13°25′E); and 24 in Barcelona, Spain (41°23′N, 2°9′E), hereafter referred to as the northern, central, and southern populations, respectively. Birds from the northern and central populations were sampled from January to March 2001 and birds from the southern population were sampled from February to March 2002. Oulu study area consists of mid-boreal forests with winters characterized by mean temperatures of approximately −8 °C, minimum day length of less than 4 h, and permanent snow cover. Lund study area consists of mixed forests of pine and deciduous trees, with mean winter temperatures of approximately 0 °C, minimum day length of 7 h, and nonpermanent snow cover. The southern study site in Barcelona consists of mixed forests of pine and oaks, winters with a mean temperature of approximately +8 °C, minimum day length of 9 h, and an absence of snow cover.

Birds were captured using baited funnel traps (Senar et al., 1997), and a few yellow contour feathers were plucked from an area on the right side of the breast, between the shoulder and the breast black stripe of each individual. All feathers were stored under equal conditions (i.e. dry and dark) for subsequent analyses in the laboratory. Feather structure analyses in the central and northern populations were undertaken in 2004, whereas the feather structure from the southern population and all colour measurements were undertaken in 2005 in the Natural History Museum facilities in Barcelona. Age (adult, N = 31, or yearling, N = 30) and sex (30 males and 31 females) were determined, sensu Jenni & Winkler (1994).

All procedures were approved by the ethical committee of the University of Oulu (097/04), Malmö/Lund Animal Care Committee (M126-00), and the Departament de Medi Ambient, Generalitat de Catalunya (2002).

Results

Feather structure variables and colour components varied significantly among great tit populations (Tables 1, 2). General contour feather structure was studied by means of a PCA, with a first factor explaining 57.6% of the variance and an eigenvalue of 3.45. Densities of plumulaceous and pennaceous barbs (0.72 and 0.75, respectively), and plumulaceous and pennaceous barbules (0.76 and 0.72, respectively) were positively loaded in the first component of the PCA, whereas feather length (−0.86) and the proportion of plumulaceous barbs (−0.70) were negatively loaded. In summary, high positive values of feather structure correspond to short dense feathers with a low proportion of plumulaceous barbs.

Feather structure varied significantly across populations (Fig. 1, Table 3): both northern and southern birds differed from the middle range population by having short and dense feathers, with lower proportion of plumulaceous barbs (post-hoc Tukey's test, both P < 0.001; Table 1). Feathers from southernmost and northernmost birds did not differ from each other (post-hoc Tukey's test, P = 0.98; Fig. 1, Table 1). Neither age or sex, nor their interactions, had any effect on feather structure (Table 3). Feather structure correlated negatively with plumage hue, after correcting for the number of feathers used in measuring the colour, so that birds with a more greenish plumage (higher hue) had less dense but longer feathers (Fig. 2, Table 3). Chroma, brightness, and UV reflectance were not affected by feather structure (Table 3). Plumage brightness and UV reflectance were higher in the southern than in the northern population (pairwise comparison, P < 0.05; Fig. 3). Birds from the central population had intermediate values of brightness, not differing from the other two populations (pairwise comparison, both P > 0.05) but differing from the northern population in UV reflectance (pairwise comparison, P = 0.020). Chroma and hue did not differ among populations (Table 3). Males had higher values of chroma, brightness, and UV reflectance than females, whereas hue variation was independent of sex (Table 3).

Discussion

We found differences between great tit populations in contour feather structure and coloration. Feather structure did not change according to winter severity because the central population had significantly longer feathers with higher percentage of plumulaceous barbs than the two marginal populations. Considering that longer contour feathers with more plumulaceous barbs have lower thermal conductance, and better insulating properties than shorter feathers with a higher proportion of pennaceous barbs (Dove et al., 2007), our results suggest that thermal insulation does not drive countour feather differences among the populations studied. Furthermore, because great tits from southern and northern populations exhibited contour feathers with similar microstructure, both populations living at the distribution margins, and therefore exposed to suboptimal environmental conditions (Sanz, 1998), may be constrained in developing optimal feather structure.

Broggi et al. (2011) suggested that birds from the central population could be less constrained during feather molt, thereby developing higher quality feathers than those from northern regions. Our results suggest that southern populations could also be constrained during the molting period because they developed feathers of a structure similar to that of northern populations. Time constraints for molting in the southern population may arise from the higher prevalence of double brooding in great tit populations at low latitudes (Sanz, 1998), or the harshness of summer conditions (Hemborg, Sanz & Lundberg, 2001). Vágási et al. (2012) recently showed a causal link between molt speed and contour feather structure by experimentally increasing molt rate of caged house sparrows Passer domesticus L., which developed feathers with similar characteristics to the ones we found in the two marginal populations of great tits (short, dense feathers with low percentage of plumulaceous barbs). Molting late in the season, as observed in populations at higher latitudes (Holmgren, Jönsson & Wennerberg, 2001), is usually compensated for by accelerating molting rate (Dawson, 2004), which can in turn decrease feather quality (Dawson et al., 2000; de la Hera et al., 2009). Similarly, great tit breeding success (Sanz, 1998) and yolk carotenoid composition in pied flycatchers Ficedula hypoleuca P. (Eeva et al., 2011) present nonlinear latitudinal patterns similar to those we found for feather structure, which are claimed to result from time constraints (Sanz, 1998) and the mismatch between laying time and caterpillar availability (Rytkönen & Orell, 2001; Eeva et al., 2011). Thus, great tits from the northern and southern populations appear to grow suboptimal feathers, which may be related to constraints on the length and/or the access to nutrients during the molting period, compared to birds from the central population. Alternatively, birds from the southern population may be released from the selective pressure for high insulation capacity, and therefore grow an adequate plumage adapted to milder winter conditions.

Despite recent demographic studies suggesting that the northern population is a ‘sink’ (Karvonen et al., 2012), Broggi et al. (2005) showed that northern great tits locally adapt their winter metabolism. However, they appear to be unable to develop a highly insulative plumage structure in line with previous results showing higher thermoregulatory costs for great tits from the northern population (Broggi et al., 2004). These results suggest that selective pressure for an optimal feather structure is weak, or otherwise constrained by other more important traits (e.g. timing of breeding) (Eeva, Veistola & Lehikoinen, 2000; Rytkönen & Orell, 2001).

Variation in the pattern in feather coloration was mostly unrelated to structure, and only hue exhibited a significant relationship to feather structure. Independent of the population of origin, hue increased with the inferred quality of the contour feathers (negative values of feather structure; Fig. 2). Because carotenoid-based hue is related to the ability to acquire food (Senar, Figuerola & Pascual, 2002; Senar et al., 2008) and to general body condition (as inferred from ptilochronology) (Senar, Figuerola & Domènech, 2003), contour feather structure may be a reliable indicator of individual quality in all populations. Higher values of hue are also found in great tit populations inhabiting good quality habitats (Ferns & Hinsley, 2008), suggesting that poorer nutritional condition could constrain both signalling properties and the structural quality of feathers, thus giving no support for a trade-off between carotenoid-based coloration and the structure of great tit feathers.

In accordance with previous studies, we found that yellow contour feathers of great tits were sexually dichromatic, with males having higher values for chroma, brightness and UV (Ferns & Hinsley, 2008; Isaksson et al., 2008) but not hue. Population of origin did not explain the variation in hue or chroma, although we found a decrease in contour feather brightness and UV with increasing latitude. Thus, molt speed does not appear to affect contour feather brightness and UV, as it does with feather structure, in contrast to reports in other species (Serra et al., 2007; Griggio et al., 2009). Higher values of contour feather brightness and UV in the southern population could be the result of a stronger sexual selection pressure in this population, which may be related to higher population density (Irwin, 2000; Forsman & Mönkkönen, 2003) or higher parasite pressure at lower latitudes (Møller, 1998). However, it should be considered that inter-individual variation in plumage coloration is often considerable (i.e. as a result of sex, age or season) (Figuerola & Senar, 2005). Although we controlled for several of these potentially confounding factors, further studies including more populations at different latitudes would be required to properly interpret the latitudinal pattern found in the present study.

In summary, great tit contour feather structure and coloration differs among the studied populations. The different feather traits do not generally vary in concert, although some patterns of co-variation emerge. The results suggest that feather structure could result from constraints during molting among populations at the distribution margins, whereas coloration may depend on other factors, such as a latitudinal decrease on the strength of sexual selection. The present study also shows that, except for hue, other signalling aspects of carotenoid-based and structural coloration are independent of feather microstructure. Great tits from different populations prioritize the development of certain feather characteristics over others, although the different traits do not appear to interact with each other. Experimental manipulations of feather coloration and/or structure in captivity would be required to further study covariation patterns and possible constraints on the development of feather traits.