Naturwissenschaften (2014) 101:929–938 DOI 10.1007/s00114-014-1234-7
ORIGINAL PAPER
Do feather-degrading bacteria actually degrade feather colour? No significant effects of plumage microbiome modifications on feather colouration in wild great tits Staffan Jacob & Léa Colmas & Nathalie Parthuisot & Philipp Heeb
Received: 11 June 2014 / Revised: 28 July 2014 / Accepted: 28 August 2014 / Published online: 17 September 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Parasites are known to exert selective pressures on host life history traits since the energy and nutrients needed to mount an immune response are no longer available to invest in other functions. Bird feathers harbour numerous microorganisms, some of which are able to degrade feather keratin (keratinolytic microorganisms) and affect feather integrity and colouration in vitro. Although named “feather-degrading” microorganisms, experimental evidence for their effects on feathers of free-living birds is still lacking. Here, we tested whether (i) keratinolytic microorganisms can degrade feathers in vivo and thus modify the colour of feathers during the nesting period and (ii) whether feather microorganisms have a long-term effect on the investment in colouration of newly moulted feathers. We designed treatments to either favour or inhibit bacterial growth, thus experimentally modifying plumage bacterial communities, in a wild breeding population of great tits (Parus major). Our analyses revealed no significant effects of the treatments on feather colours. Moreover, we found that differences in bacterial exposure during nesting did not significantly affect the colouration of newly moulted feathers. Our results suggest that significant feather degradation obtained during in vitro studies could have led to an overestimation of the potential of keratinolytic microorganisms to shape feather colouration in free-living birds. Communicated by: Alexandre Roulin S. Jacob : L. Colmas : N. Parthuisot : P. Heeb Laboratoire Évolution et Diversité Biologique (EDB), UMR 5174 Centre National de la Recherche Scientifique (CNRS), Ecole Nationale de Formation Agronomique (ENFA)–Université Paul Sabatier, 118 Route de Narbonne, 31062 Toulouse, France Present Address: S. Jacob (*) Station d’Ecologie Expérimentale du CNRS à Moulis, USR2936, 2 route du CNRS, 09200 Saint-Girons, France e-mail: [email protected]
Keywords Feather colouration . Feather-degrading microorganisms . Keratinolytic microorganisms . Microbiome . Sexual selection . Visual communication
Microorganisms are widespread and constitute the major part of the earth’s biomass (Gilbert et al. 2012; McFall-Ngai et al. 2013). Parasitic microorganisms can lead hosts to allocate energy and nutrients to immune responses at the expense of other functions such as signalling and reproduction (Clayton and Moore 1997; Schmid-Hempel 2011). As a result, parasiteinduced changes in host phenotypes can have major consequences for host survival and fitness (Clayton and Moore 1997; Schmid-Hempel 2011). Bird feathers are inhabited by numerous microorganisms. Some of them are keratinolytic—that is, able to degrade keratin (Burtt and Ichida 1999; Gunderson 2008; RuizRodríguez et al. 2009). Since keratin constitutes more than 90 % of feather mass (Onifade et al. 1998; Burtt and Ichida 1999; Gunderson 2008), keratinolytic microorganisms (also named “feather-degrading” microorganisms) are expected to degrade feather keratin and thus exert on birds selective pressures by decreasing feather integrity and affecting flight efficiency and thermoregulation (Burtt and Ichida 1999; Gunderson 2008). Moreover, Shawkey et al. (2007) showed in vitro that the first effect of a bacterial keratinolytic activity on feathers consists in an alteration of feather keratin structure that led to a decrease of feather ultraviolet (UV) reflectance and an increase in brightness. Since feather colouration is often involved in communication processes in birds (Hill and McGraw 2006), this suggests that microorganisms could play a role in the evolution of plumage signals involved in visual communication (Shawkey et al. 2007; Ruiz-Rodríguez et al. 2009; Burtt et al. 2011; Ruiz-De-Castañeda et al. 2012). Several studies on living birds provided correlative support for the hypothesis that microorganisms on bird plumage affect
930
feather colouration. Shawkey et al. (2007) found a positive correlation between the brightness of male eastern bluebirds’ (Sialia sialis) structural feather colouration and bacterial loads on their feathers. In contrast, carotenoid-based plumage colouration in house finches (Carpodacus mexicanus) and plumage brightness in female bluebirds were found to be negatively correlated to the densities of feather-degrading bacteria (Shawkey et al. 2009; Gunderson et al. 2009). Feather colour saturation was also found to be negatively correlated to the number of bacterial species on female great tit feathers (Kilgas et al. 2012a). These correlational studies suggest that feather microorganisms can degrade feather keratin in vivo and consequently alter feather colouration, as found with in vitro experiments (Shawkey et al. 2007; Gunderson 2008). However, the causal link between feather colouration and keratinolytic microorganisms still needs to be demonstrated in freeliving birds (Gunderson 2008). An alternative hypothesis to explain the negative correlations found between colouration and feather microorganisms (Shawkey et al. 2007; Shawkey et al. 2009; Kilgas et al. 2012a) is that birds could modify their investment in feather colouration during moult depending on their exposure to microorganisms. During moult, birds face trade-offs between producing new feathers and maintaining thermoregulation, flight and immune functions (Vézina et al. 2009; MorenoRueda 2010). Moreover, carotenoid pigments invested in feather colouration are no longer available for immune system functioning (Olson & Owens 1998; Aguilera & Amat 2007; but see Navara & Hill 2003). If bacteria present on feathers interact with the host immune function, an allocation trade-off between immune response and feather colouration would lead to negative correlations between feather bacterial densities and components of plumage colouration as found in previous studies (Shawkey et al. 2009; Gunderson et al. 2009; Kilgas et al. 2012a). To date, only one experimental study has investigated the effects of keratinolytic bacteria in vivo on feathers of captive northern cardinals (Cardinalis cardinalis), and European starlings (Sturnus vulgaris), and found no significant degradation of feathers by Bacillus licheniformis (Cristol et al. 2005). However, captive birds could have invested greater amounts of time and/or energy in plumage sanitation since they do not face the same trade-offs as free-living birds do (Metcalfe and Monaghan 2013). Moreover, previous studies examining feather degradation were performed using only strains of B. licheniformis as a representative of the diverse community of keratinolytic microorganisms (Lucas et al. 2003). Degradation of feather keratin by keratinolytic microorganisms involves two steps, each one requiring a different keratinase enzyme (Yamamura et al. 2002; Gunderson 2008). Consequently, synergetic effects are
Naturwissenschaften (2014) 101:929–938
likely to occur when microorganisms are present in communities on wild birds (Gunderson 2008). In agreement with this idea, Lucas et al. (2005) found that feathers with richer keratinolytic assemblages were more degraded in vitro. Moreover, conditions of high humidity and temperature used during in vitro experiments do not occur on plumage of living birds (Cristol et al. 2005). In this study, we investigated whether experimental modifications of the plumage microbiome of wild breeding great tits affected feather colouration. By spraying different solutions into randomly allocated nests, we either favoured or inhibited bacterial growth in active nests during the breeding season (Jacob et al. 2014). These treatments led to modifications of both total and keratinolytic cultivable bacterial densities in the nests and, by contamination through contact between birds and their nests, also modified adult feather bacterial loads (Jacob et al. 2014). We first tested the hypothesis that bacterial keratinolytic activity on feathers alters colouration in free-living great tits. Since alteration of feather colouration as a result of keratinolytic activity is due to the degradation of feather microstructure (Shawkey et al. 2007), we expected that feather colouration should be affected sooner than other aspects of feather integrity (e.g. missing barbules; Cristol et al. 2005). As a result, if keratinolytic bacteria degrade feathers on freeliving birds, we should observe differences of feather colouration between birds exposed to different densities of keratinolytic microorganisms on their plumage. According to the results of Shawkey et al. (2007), we expected that the yellow and UV reflectance of feathers might decrease as a result of keratinolytic degradation of feather structure, whereas feather brightness might increase. Second, we investigated whether modifications of feather microorganisms during the breeding season led to changes in investment in feather colouration during moult. To answer this second question, birds were captured the autumn following the experiment, after moult completion. Since moulting is the replacement of old feathers by new ones, we expected to observe an overall increase in feather colouration between reproduction and autumn captures. Moreover, we expected that birds would differ in their investment in colouration of new feathers depending on the microbial pressures faced during reproduction. Finally, males and females have often been found to differ in their exposure to nest microorganisms and reproductive strategies (Gosler 1993; Saag et al. 2011; Kilgas et al. 2012b). Females are expected to be more exposed to nest microorganisms and to harbour higher bacterial loads on feathers (Saag et al. 2011; Jacob et al. 2014), and consequently might experience more pronounced effects of the treatments. We thus investigated whether treatment effects differed between sexes by testing for the treatment by sex interaction on feather colouration.
Naturwissenschaften (2014) 101:929–938
931
Material and methods
Adult sampling and measurements
Experimental design
Great tit adults were trapped in the nest boxes about 10 days after the eggs hatched (54 females and 44 males, 45 in 2011 and 53 in 2012; Jacob et al. 2014) after having washed hands with 70 % ethanol. We directly took ten feather samples twice from each individual at a standardized position close to and above the left leg. As for nest material samples, one sample was placed in phosphate-buffered saline (PBS) and the other in PBS+glycerol. We measured tarsus length to the nearest 0.01 mm using a calliper, body mass with an electronic balance (±0.01 g) and wing length with a ruler (±0.1 mm). We found no significant differences in adult tarsus length, wing length and body mass between the treatments (general linear mixed models with nest as a random factor; tarsus length: F2,53 =1.34, p=0.27; wing length: F2,53 =1.86, p= 0.17; body mass: F2,53 =2.74; p=0.08). To measure feather colouration, we sampled approximately 15 feathers from each of five different body parts of the adults: yellow breast, black breast stripe, green back, black crown and white cheeks (Fig. 1; Galván 2010). This sampling did not lead to any visible changes in colour patches after manipulation. Since females showed reduced black breast stripes due to incubation and sexual dimorphism, analyses of feathers from this body part were performed only for males. Feathers were kept in small paper envelopes in the dark until we did spectrophotometric analyses. In order to investigate long-term effects of the treatments, we captured great tits that had been exposed to the different treatments using mist nets during the following autumn (between September and December), after moult had occurred (15 females and 16 males). At that time, we also measured tarsus length, body mass and wing length and took feather samples for spectrophotometric analyses.
The study was performed during the 2011 and 2012 reproductive seasons on a great tit population breeding in nest boxes close to Toulouse, France (43° 39′ N, 1° 54′ E). In the winter, old nest material was removed from the nest boxes and the insides of the boxes were scraped with a hard brush. Nest boxes were visited daily from the beginning of March to detect the beginning of nest building. We randomly assigned the nests to three treatments as presented in Jacob et al. (2014). First, to favour the bacterial growth in the nests, we used TSB (tryptic soy broth, 40 mg/ L in sterilized distilled water, Sigma), a liquid general growth medium for heterotrophic microorganisms that is commonly used in microbiology. Second, nisin, in association with EDTA, a bacteriostatic solution used for food conservation (7 g nisin (900 IU/mg; B&K Technology Group) in 50 mM EDTA; Harris et al. 1992; Economou et al. 2009), was used to inhibit bacterial growth in the nests. TSB and nisin were diluted in water, and humidity can favour microbial growth (Cook et al. 2005; Wang et al. 2011). Consequently, we used water as a control in order to have similar humidity levels in the three treatments, making between treatment differences in host phenotype a result of effects of TSB and nisin solution on bacterial communities. After carefully removing the eggs or the nestlings, one solution (TSB, nisin or water) was sprayed (mean volume 1.7±0.02 mL) in the centre of each nest cup every 2 days from the beginning of nest building until the nestlings fledged. The total number of times we sprayed these solutions per nest was 16.6±0.3 (mean±SE) with no significant difference between treatments (Kruskal-Wallis sum rank test: Χ2 =4.02, df=52, p=0.13). During incubation, nests were treated only on days 1, 5 and 9 after the start of incubation in order to limit the risks of nest desertion. A total of 54 nests (25 in 2011, 29 in 2012) were included in our study (17 nests in the TSB treatment, 17 in nisin and 20 in control), and they did not differ significantly in laying date (Χ2 =3.85, df=52, p=0.15) and clutch size (Χ2 =2.60, df=52, p=0.27). To measure the effects of the treatments on nest bacterial communities, we collected two samples of nest material using sterilized tweezers at a standardized position in the centre of the nest cup at day 9 of incubation (for details, see Jacob et al. 2014). Samples were kept on ice in the field and stored at −20 °C until lab analyses. All sampling and manipulations were made after systematically washing hands and material with 70 % ethanol in order to avoid cross contamination. All manipulations were performed according to French legislation, and permits were obtained from DREAL and CRBPO (ringing permit no. 565).
Microbial analyses As detailed in Jacob et al. (2014), we used culture-based and culture-independent techniques to measure, respectively, the density and composition of bacterial communities in the nests and on the feathers. Briefly, to estimate the densities of bacterial communities, we grew them on tryptic soy agar (TSA; Møller et al. 2009; Czirják et al. 2010; Jacob et al. 2014), a general medium allowing the growth of heterotrophic bacteria. Keratinolytic bacterial densities were estimated with feather meal agar (FMA), a medium containing only keratin as carbon source (Møller et al. 2009; Czirják et al. 2010; Jacob et al. 2014). Bacterial density values were log-transformed to follow normality. We used ARISA (automated ribosomal intergenic spacer analysis) to measure bacterial community composition (Ranjard et al. 2000). We amplified highly variable regions of the bacterial and fungal ribosomal operons and measured the length of the amplified fragments by sequencing to obtain profiles composed of several peaks, each peak corresponding
932
Naturwissenschaften (2014) 101:929–938
Fig. 1 Reflectance spectra at five locations on great tit plumage. The line represents the smoothed spectra used to calculate colour variables, and the grey area corresponds to the variance in the averaged spectra
to an operational taxonomic unit (OTU; for details, see Jacob et al. 2014).
Spectrophotometric analyses The colour of sampled feathers was measured using a USB2000 spectrophotometer with a DH-2000 deuterium-halogen lamp and OOIBase32 software (Ocean Optics, Inc., Dunedin, FL). Feathers were placed in stacks on a gloss-free black paper so as to resemble natural feather arrangement (Shawkey et al. 2007). The probe was placed at 90° to the feather surface, kept at 2 mm from feathers using a probe holder. Reflectance spectra were expressed comparatively to a white standard disc (type WS; Labsphere, Congleton, UK). For each body part, we measured three spectra, one after each arrangement of the feather stack. Using 20 birds measured three times for each body part and intraclass correlation coefficient (icc function, irr R package), we found that the repeatability of spectral measurements was high for all body parts (all r>0.80). Using the pavo package in R, we averaged the threereplicated reflectance spectra for each of the five body parts of each individual (Fig. 1). After smoothing the averaged spectra (procspec function in pavo R package, smoothness parameter=0.08 for yellow breast, green back and white
cheeks; smoothness=0.2 for black breast stripe and black crown), we computed plumage brightness, UV chroma and yellow chroma following Shawkey et al. (2007). We calculated brightness, the total amount of light reflected by feathers, as the mean reflectance value over 300 to 700 nm, the range of bird-visible light (Jacobs 1981; Cuthill et al. 2000). To measure the proportion of light reflected in the UV range visible to birds, UV chroma was calculated as the proportion of reflectance between 300 and 420 nm compared to total reflectance over the whole bird-visible spectrum. For the yellow breast and green back, we calculated yellow chroma as the proportion of reflectance between 515 and 700 nm compared to total reflectance between 420 and 700 nm, indicative of the influence of light absorbance by feather carotenoids. Yellow chroma is involved in sexual selection in great tits by revealing body condition (Senar et al. 2008) and has been found correlated to feather bacteria (Kilgas et al. 2012a), and both brightness and UV chroma are modified during in vitro degradation of feathers by keratinolytic bacteria (Shawkey et al. 2007).
Statistical analyses We estimated individual body condition through the regression of body mass on tarsus length (body mass=4.19+5.89×tarsus
Naturwissenschaften (2014) 101:929–938 Table 1 Analyses of factors affecting the colour variables from the five feather patches in adult great tits at the end of reproduction (feather brightness, yellow chroma and UV chroma ). Final models after backward selection are shown. p values
ORIGINAL PAPER
Do feather-degrading bacteria actually degrade feather colour? No significant effects of plumage microbiome modifications on feather colouration in wild great tits Staffan Jacob & Léa Colmas & Nathalie Parthuisot & Philipp Heeb
Received: 11 June 2014 / Revised: 28 July 2014 / Accepted: 28 August 2014 / Published online: 17 September 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Parasites are known to exert selective pressures on host life history traits since the energy and nutrients needed to mount an immune response are no longer available to invest in other functions. Bird feathers harbour numerous microorganisms, some of which are able to degrade feather keratin (keratinolytic microorganisms) and affect feather integrity and colouration in vitro. Although named “feather-degrading” microorganisms, experimental evidence for their effects on feathers of free-living birds is still lacking. Here, we tested whether (i) keratinolytic microorganisms can degrade feathers in vivo and thus modify the colour of feathers during the nesting period and (ii) whether feather microorganisms have a long-term effect on the investment in colouration of newly moulted feathers. We designed treatments to either favour or inhibit bacterial growth, thus experimentally modifying plumage bacterial communities, in a wild breeding population of great tits (Parus major). Our analyses revealed no significant effects of the treatments on feather colours. Moreover, we found that differences in bacterial exposure during nesting did not significantly affect the colouration of newly moulted feathers. Our results suggest that significant feather degradation obtained during in vitro studies could have led to an overestimation of the potential of keratinolytic microorganisms to shape feather colouration in free-living birds. Communicated by: Alexandre Roulin S. Jacob : L. Colmas : N. Parthuisot : P. Heeb Laboratoire Évolution et Diversité Biologique (EDB), UMR 5174 Centre National de la Recherche Scientifique (CNRS), Ecole Nationale de Formation Agronomique (ENFA)–Université Paul Sabatier, 118 Route de Narbonne, 31062 Toulouse, France Present Address: S. Jacob (*) Station d’Ecologie Expérimentale du CNRS à Moulis, USR2936, 2 route du CNRS, 09200 Saint-Girons, France e-mail: [email protected]
Keywords Feather colouration . Feather-degrading microorganisms . Keratinolytic microorganisms . Microbiome . Sexual selection . Visual communication
Microorganisms are widespread and constitute the major part of the earth’s biomass (Gilbert et al. 2012; McFall-Ngai et al. 2013). Parasitic microorganisms can lead hosts to allocate energy and nutrients to immune responses at the expense of other functions such as signalling and reproduction (Clayton and Moore 1997; Schmid-Hempel 2011). As a result, parasiteinduced changes in host phenotypes can have major consequences for host survival and fitness (Clayton and Moore 1997; Schmid-Hempel 2011). Bird feathers are inhabited by numerous microorganisms. Some of them are keratinolytic—that is, able to degrade keratin (Burtt and Ichida 1999; Gunderson 2008; RuizRodríguez et al. 2009). Since keratin constitutes more than 90 % of feather mass (Onifade et al. 1998; Burtt and Ichida 1999; Gunderson 2008), keratinolytic microorganisms (also named “feather-degrading” microorganisms) are expected to degrade feather keratin and thus exert on birds selective pressures by decreasing feather integrity and affecting flight efficiency and thermoregulation (Burtt and Ichida 1999; Gunderson 2008). Moreover, Shawkey et al. (2007) showed in vitro that the first effect of a bacterial keratinolytic activity on feathers consists in an alteration of feather keratin structure that led to a decrease of feather ultraviolet (UV) reflectance and an increase in brightness. Since feather colouration is often involved in communication processes in birds (Hill and McGraw 2006), this suggests that microorganisms could play a role in the evolution of plumage signals involved in visual communication (Shawkey et al. 2007; Ruiz-Rodríguez et al. 2009; Burtt et al. 2011; Ruiz-De-Castañeda et al. 2012). Several studies on living birds provided correlative support for the hypothesis that microorganisms on bird plumage affect
930
feather colouration. Shawkey et al. (2007) found a positive correlation between the brightness of male eastern bluebirds’ (Sialia sialis) structural feather colouration and bacterial loads on their feathers. In contrast, carotenoid-based plumage colouration in house finches (Carpodacus mexicanus) and plumage brightness in female bluebirds were found to be negatively correlated to the densities of feather-degrading bacteria (Shawkey et al. 2009; Gunderson et al. 2009). Feather colour saturation was also found to be negatively correlated to the number of bacterial species on female great tit feathers (Kilgas et al. 2012a). These correlational studies suggest that feather microorganisms can degrade feather keratin in vivo and consequently alter feather colouration, as found with in vitro experiments (Shawkey et al. 2007; Gunderson 2008). However, the causal link between feather colouration and keratinolytic microorganisms still needs to be demonstrated in freeliving birds (Gunderson 2008). An alternative hypothesis to explain the negative correlations found between colouration and feather microorganisms (Shawkey et al. 2007; Shawkey et al. 2009; Kilgas et al. 2012a) is that birds could modify their investment in feather colouration during moult depending on their exposure to microorganisms. During moult, birds face trade-offs between producing new feathers and maintaining thermoregulation, flight and immune functions (Vézina et al. 2009; MorenoRueda 2010). Moreover, carotenoid pigments invested in feather colouration are no longer available for immune system functioning (Olson & Owens 1998; Aguilera & Amat 2007; but see Navara & Hill 2003). If bacteria present on feathers interact with the host immune function, an allocation trade-off between immune response and feather colouration would lead to negative correlations between feather bacterial densities and components of plumage colouration as found in previous studies (Shawkey et al. 2009; Gunderson et al. 2009; Kilgas et al. 2012a). To date, only one experimental study has investigated the effects of keratinolytic bacteria in vivo on feathers of captive northern cardinals (Cardinalis cardinalis), and European starlings (Sturnus vulgaris), and found no significant degradation of feathers by Bacillus licheniformis (Cristol et al. 2005). However, captive birds could have invested greater amounts of time and/or energy in plumage sanitation since they do not face the same trade-offs as free-living birds do (Metcalfe and Monaghan 2013). Moreover, previous studies examining feather degradation were performed using only strains of B. licheniformis as a representative of the diverse community of keratinolytic microorganisms (Lucas et al. 2003). Degradation of feather keratin by keratinolytic microorganisms involves two steps, each one requiring a different keratinase enzyme (Yamamura et al. 2002; Gunderson 2008). Consequently, synergetic effects are
Naturwissenschaften (2014) 101:929–938
likely to occur when microorganisms are present in communities on wild birds (Gunderson 2008). In agreement with this idea, Lucas et al. (2005) found that feathers with richer keratinolytic assemblages were more degraded in vitro. Moreover, conditions of high humidity and temperature used during in vitro experiments do not occur on plumage of living birds (Cristol et al. 2005). In this study, we investigated whether experimental modifications of the plumage microbiome of wild breeding great tits affected feather colouration. By spraying different solutions into randomly allocated nests, we either favoured or inhibited bacterial growth in active nests during the breeding season (Jacob et al. 2014). These treatments led to modifications of both total and keratinolytic cultivable bacterial densities in the nests and, by contamination through contact between birds and their nests, also modified adult feather bacterial loads (Jacob et al. 2014). We first tested the hypothesis that bacterial keratinolytic activity on feathers alters colouration in free-living great tits. Since alteration of feather colouration as a result of keratinolytic activity is due to the degradation of feather microstructure (Shawkey et al. 2007), we expected that feather colouration should be affected sooner than other aspects of feather integrity (e.g. missing barbules; Cristol et al. 2005). As a result, if keratinolytic bacteria degrade feathers on freeliving birds, we should observe differences of feather colouration between birds exposed to different densities of keratinolytic microorganisms on their plumage. According to the results of Shawkey et al. (2007), we expected that the yellow and UV reflectance of feathers might decrease as a result of keratinolytic degradation of feather structure, whereas feather brightness might increase. Second, we investigated whether modifications of feather microorganisms during the breeding season led to changes in investment in feather colouration during moult. To answer this second question, birds were captured the autumn following the experiment, after moult completion. Since moulting is the replacement of old feathers by new ones, we expected to observe an overall increase in feather colouration between reproduction and autumn captures. Moreover, we expected that birds would differ in their investment in colouration of new feathers depending on the microbial pressures faced during reproduction. Finally, males and females have often been found to differ in their exposure to nest microorganisms and reproductive strategies (Gosler 1993; Saag et al. 2011; Kilgas et al. 2012b). Females are expected to be more exposed to nest microorganisms and to harbour higher bacterial loads on feathers (Saag et al. 2011; Jacob et al. 2014), and consequently might experience more pronounced effects of the treatments. We thus investigated whether treatment effects differed between sexes by testing for the treatment by sex interaction on feather colouration.
Naturwissenschaften (2014) 101:929–938
931
Material and methods
Adult sampling and measurements
Experimental design
Great tit adults were trapped in the nest boxes about 10 days after the eggs hatched (54 females and 44 males, 45 in 2011 and 53 in 2012; Jacob et al. 2014) after having washed hands with 70 % ethanol. We directly took ten feather samples twice from each individual at a standardized position close to and above the left leg. As for nest material samples, one sample was placed in phosphate-buffered saline (PBS) and the other in PBS+glycerol. We measured tarsus length to the nearest 0.01 mm using a calliper, body mass with an electronic balance (±0.01 g) and wing length with a ruler (±0.1 mm). We found no significant differences in adult tarsus length, wing length and body mass between the treatments (general linear mixed models with nest as a random factor; tarsus length: F2,53 =1.34, p=0.27; wing length: F2,53 =1.86, p= 0.17; body mass: F2,53 =2.74; p=0.08). To measure feather colouration, we sampled approximately 15 feathers from each of five different body parts of the adults: yellow breast, black breast stripe, green back, black crown and white cheeks (Fig. 1; Galván 2010). This sampling did not lead to any visible changes in colour patches after manipulation. Since females showed reduced black breast stripes due to incubation and sexual dimorphism, analyses of feathers from this body part were performed only for males. Feathers were kept in small paper envelopes in the dark until we did spectrophotometric analyses. In order to investigate long-term effects of the treatments, we captured great tits that had been exposed to the different treatments using mist nets during the following autumn (between September and December), after moult had occurred (15 females and 16 males). At that time, we also measured tarsus length, body mass and wing length and took feather samples for spectrophotometric analyses.
The study was performed during the 2011 and 2012 reproductive seasons on a great tit population breeding in nest boxes close to Toulouse, France (43° 39′ N, 1° 54′ E). In the winter, old nest material was removed from the nest boxes and the insides of the boxes were scraped with a hard brush. Nest boxes were visited daily from the beginning of March to detect the beginning of nest building. We randomly assigned the nests to three treatments as presented in Jacob et al. (2014). First, to favour the bacterial growth in the nests, we used TSB (tryptic soy broth, 40 mg/ L in sterilized distilled water, Sigma), a liquid general growth medium for heterotrophic microorganisms that is commonly used in microbiology. Second, nisin, in association with EDTA, a bacteriostatic solution used for food conservation (7 g nisin (900 IU/mg; B&K Technology Group) in 50 mM EDTA; Harris et al. 1992; Economou et al. 2009), was used to inhibit bacterial growth in the nests. TSB and nisin were diluted in water, and humidity can favour microbial growth (Cook et al. 2005; Wang et al. 2011). Consequently, we used water as a control in order to have similar humidity levels in the three treatments, making between treatment differences in host phenotype a result of effects of TSB and nisin solution on bacterial communities. After carefully removing the eggs or the nestlings, one solution (TSB, nisin or water) was sprayed (mean volume 1.7±0.02 mL) in the centre of each nest cup every 2 days from the beginning of nest building until the nestlings fledged. The total number of times we sprayed these solutions per nest was 16.6±0.3 (mean±SE) with no significant difference between treatments (Kruskal-Wallis sum rank test: Χ2 =4.02, df=52, p=0.13). During incubation, nests were treated only on days 1, 5 and 9 after the start of incubation in order to limit the risks of nest desertion. A total of 54 nests (25 in 2011, 29 in 2012) were included in our study (17 nests in the TSB treatment, 17 in nisin and 20 in control), and they did not differ significantly in laying date (Χ2 =3.85, df=52, p=0.15) and clutch size (Χ2 =2.60, df=52, p=0.27). To measure the effects of the treatments on nest bacterial communities, we collected two samples of nest material using sterilized tweezers at a standardized position in the centre of the nest cup at day 9 of incubation (for details, see Jacob et al. 2014). Samples were kept on ice in the field and stored at −20 °C until lab analyses. All sampling and manipulations were made after systematically washing hands and material with 70 % ethanol in order to avoid cross contamination. All manipulations were performed according to French legislation, and permits were obtained from DREAL and CRBPO (ringing permit no. 565).
Microbial analyses As detailed in Jacob et al. (2014), we used culture-based and culture-independent techniques to measure, respectively, the density and composition of bacterial communities in the nests and on the feathers. Briefly, to estimate the densities of bacterial communities, we grew them on tryptic soy agar (TSA; Møller et al. 2009; Czirják et al. 2010; Jacob et al. 2014), a general medium allowing the growth of heterotrophic bacteria. Keratinolytic bacterial densities were estimated with feather meal agar (FMA), a medium containing only keratin as carbon source (Møller et al. 2009; Czirják et al. 2010; Jacob et al. 2014). Bacterial density values were log-transformed to follow normality. We used ARISA (automated ribosomal intergenic spacer analysis) to measure bacterial community composition (Ranjard et al. 2000). We amplified highly variable regions of the bacterial and fungal ribosomal operons and measured the length of the amplified fragments by sequencing to obtain profiles composed of several peaks, each peak corresponding
932
Naturwissenschaften (2014) 101:929–938
Fig. 1 Reflectance spectra at five locations on great tit plumage. The line represents the smoothed spectra used to calculate colour variables, and the grey area corresponds to the variance in the averaged spectra
to an operational taxonomic unit (OTU; for details, see Jacob et al. 2014).
Spectrophotometric analyses The colour of sampled feathers was measured using a USB2000 spectrophotometer with a DH-2000 deuterium-halogen lamp and OOIBase32 software (Ocean Optics, Inc., Dunedin, FL). Feathers were placed in stacks on a gloss-free black paper so as to resemble natural feather arrangement (Shawkey et al. 2007). The probe was placed at 90° to the feather surface, kept at 2 mm from feathers using a probe holder. Reflectance spectra were expressed comparatively to a white standard disc (type WS; Labsphere, Congleton, UK). For each body part, we measured three spectra, one after each arrangement of the feather stack. Using 20 birds measured three times for each body part and intraclass correlation coefficient (icc function, irr R package), we found that the repeatability of spectral measurements was high for all body parts (all r>0.80). Using the pavo package in R, we averaged the threereplicated reflectance spectra for each of the five body parts of each individual (Fig. 1). After smoothing the averaged spectra (procspec function in pavo R package, smoothness parameter=0.08 for yellow breast, green back and white
cheeks; smoothness=0.2 for black breast stripe and black crown), we computed plumage brightness, UV chroma and yellow chroma following Shawkey et al. (2007). We calculated brightness, the total amount of light reflected by feathers, as the mean reflectance value over 300 to 700 nm, the range of bird-visible light (Jacobs 1981; Cuthill et al. 2000). To measure the proportion of light reflected in the UV range visible to birds, UV chroma was calculated as the proportion of reflectance between 300 and 420 nm compared to total reflectance over the whole bird-visible spectrum. For the yellow breast and green back, we calculated yellow chroma as the proportion of reflectance between 515 and 700 nm compared to total reflectance between 420 and 700 nm, indicative of the influence of light absorbance by feather carotenoids. Yellow chroma is involved in sexual selection in great tits by revealing body condition (Senar et al. 2008) and has been found correlated to feather bacteria (Kilgas et al. 2012a), and both brightness and UV chroma are modified during in vitro degradation of feathers by keratinolytic bacteria (Shawkey et al. 2007).
Statistical analyses We estimated individual body condition through the regression of body mass on tarsus length (body mass=4.19+5.89×tarsus
Naturwissenschaften (2014) 101:929–938 Table 1 Analyses of factors affecting the colour variables from the five feather patches in adult great tits at the end of reproduction (feather brightness, yellow chroma and UV chroma ). Final models after backward selection are shown. p values
