Beak coloration of starling (Sturnus unicolor) males depends on the length of their throat feathers

Beak colouration of starling (Sturnus unicolor) males depends on the length of their 1 throat feathers 2 3 Manuel Azcárate-García1, Magdalena Ruiz-Rodríguez1, Cristina Ruiz-Castellano1, 4 Silvia Díaz-Lora2, Gustavo Tomás1, Manuel Martín-Vivaldi2,3 & Juan José Soler1,3 5 6 7 1 Departamento de Ecología Funcional y Evolutiva, Estación Experimental de Zonas 8 Áridas (CSIC), Almería. Spain. 9 2 Departamento de Zoología, Facultad de Ciencias, Universidad de Granada, Granada. 10 Spain. 11 3 Unidad asociada (CSIC): Coevolución: cucos, hospedadores y bacterias simbiontes. 12 Universidad de Granada, 18071-Granada, Spain. 13 14 15 16 17 Corresponding Author: 18 Manuel Azcárate-García 19 ADDRESS: Estación Experimental de Zonas Áridas: Ctra. de Sacramento s/n, La 20 Cañada de San Urbano, 04120, Almería (Spain) 21 TLF: (+34) 660058398 22 E-MAIL: mazcarategarc[email protected] 23 24 25 Crossref Funding Data Registry 26 This work was supported by the currently named Ministerio de Ciencia, Innovación y 27 Universidades and European (FEDER) funds (CGL2013-48193-C3-1-P, CGL2017-28 83103-P, CGL2017-89063-P). MAG was financed by a predoctoral contract (BES-2014-29 068661) from the Spanish Ministerio de Economía y Competitividad, and GT by the 30 Ramón y Cajal Programme. 31 32 Acknowledgments 33 We thank Natalia Juárez García-Pelayo and Carmen Soler Zamora for the help in 34 capturing birds and collecting data from the field. The research group benefits from 35 facilities, including and apartment, provided by the city hall of Guadix where a small lab 36 to quickly proceed the samples was installed. 37 38 Ethical Note 39 We performed the study following the relevant Spanish national (Decreto 40 142/2013, 1 de octubre) and regional guidelines. The ethics committee of the Spanish 41 National Research Council (CSIC) approved the protocol, and the Consejería de Medio 42 Ambiente de la Junta de Andalucía, Spain, provided all the necessary permits for nest and 43 birds manipulations (Ref: SGMN/GyB/JMIF). The protocols adhered to the ASAB/ABS 44 Guidelines for the Use of Animals in Research. The authorization for carrying out the 45 experimental work on animals in the field was granted by the Dirección General de 46 Gestión del Medio Natural of the Consejería de Medio Ambiente of Junta de Andalucía. 47 48 Author contributions 49 Conceived and designed the experiments: JJS, MRR, GT and MMV. Fieldwork: 50 MAG, JJS, MRR, GT, CRC and SDL. Analysed the data: MAG and JJS. Contributed 51 reagents/materials/analysis tools: substantial contribution from all authors. MAG wrote 52 the first version with supervision of JJS and MRR. All authors substantially contributed 53 to final version. 54 55 Data accessibility 56 Data used in this paper can be found in CSIC Institutional Repository, with the 57 accession numbers <xxxxxxxx>. 58 59 1 Lay summary 1 2 The use of signals to indicate the individual quality is widespread in nature. However, 3 although most species show more than one signal, the relationships between different 4 signals have almost never been experimentally studied. Here, we demonstrated that the 5 experimental reduction of throat feathers length conditioned the beak colouration of 6 spotless starling males at the time of reproduction. Our results are the first experimental 7 evidence of two sexually dimorphic traits being related in natural conditions. 8 2 Beak colouration of starling (Sturnus unicolor) males depends on the length of their 9 throat feathers 10 Abstract 11 12 Within the context of complex sexual signalling, most research has focussed on exploring 13 the associations between several signals and/or their relationships with different proxies 14 of individual quality. However, very few studies have focused on checking whether the 15 expression of one signal is conditioned by the expression of the others. Here, by 16 experimentally shortening the throat feathers of male spotless starlings (Sturnus 17 unicolor), we evaluated the influence of this trait on the colour expression of the beak 18 base. In addition, we tested the relationship between these two sexually dimorphic 19 characters with traits indicating individual quality such as body condition and colour 20 reflectance at the wavelength related to carotenes in the tip of the beak. Our results show 21 that the colouration of the beak base in males, but not in females, is positively related to 22 body condition and to the length of ornamental throat feathers. Moreover, the 23 experimental shortening of throat feathers in males had a negative effect on the blue 24 chroma intensity of their beak base one year after manipulation. These results support for 25 the first time a causal link between the expression of two sexually dimorphic characters, 26 which is essential to understand their functionality in a multiple signalling framework. 27 28 Keywords: Beak colour, Body condition, Interacting signals, Multiple signals, 29 Ornamental feathers length, Sexually dimorphic characters. 30 31 32 3 Introduction 33 Animals use a wide array of signals to inform about their phenotypic or genetic 34 quality to conspecifics in social interactions, in contexts such as mate choice or 35 competition for resources (Kokko 2003, Andersson and Simmons 2006, Kraaijeveld et 36 al. 2007, Lyon and Montgomerie 2012, Edward 2015). In contexts of sexual selection, 37 males typically possess multiple traits that may convey independent information to 38 receivers (Møller and Pomiankowski 1993). Although most research on the evolution and 39 function of signals has focused on single traits, the importance of studying these 40 characters within the theoretical and more realistic framework of multiple signals has 41 been highlighted (Candolin 2003, Hebets and Papaj 2005). Different characters might, for 42 instance, imply multiple or redundant messages (i.e., information), or might be more 43 efficient in particular environments or in stimulating particular sensory channels (Møller 44 and Pomiankowski 1993, Candolin 2003, Hebets and Papaj 2005). Hebets and Papaj 45 (2005) developed a framework of testable hypotheses for explaining the evolution and 46 functioning of multiple signals. They highlighted (i) the importance of considering 47 complex signals and the unit of character selection; (ii) that complex signals include 48 several characters that function together, either facilitating the transmission (e.g., using 49 different sensory channels) or reinforcing transmitted information to receivers (i.e., 50 redundant information); and (iii) that individual signals or components of complex signals 51 do not necessarily function independently, but may interact in a functional way. 52 Most research on complex signalling has focussed on exploring the association 53 between several signals (Perrier et al. 2002, Bro-Jorgensen and Dabelsteen 2008, Mason 54 et al. 2014, Chaine and Lyon 2015, Girard et al. 2015), or between signals and different 55 proxies of fitness including phenotypic quality (Balmford et al. 1992, Martin and Lopez 56 2009), mating success (Møller and Pomiankowski 1993) and efficacy of signal 57 4 transmission in different environments (Endler and Houde 1995). Even though the study 58 of the interactions (i.e., associations) between different signalling characters is essential 59 to know individual or complex signals functioning, it is one of the least explored areas 60 within the field of signal evolution. The study of signal interactions has the potential to 61 shed light on signal functioning because, for instance, detecting a positive association 62 would suggest that transmitted information is redundant or complementary. Moreover, a 63 negative association would indicate that a trade-off between signalling characters exists, 64 while the absence of association between different signals would suggest that they convey 65 different information to receivers (Candolin 2003, Hebets and Papaj 2005). In most 66 instances, inter-signal interaction occurs when the presence of one signal or a signal 67 component alters the response of the receiver to a second signal or component by 68 amplifying or conditioning the information provided by each other. 69 Interactions between signals may also occur when the production of one signal 70 influences the cost of production of another signal (Johnstone 1996, Candolin 2003). In 71 this case, independently of the transmitted information, the phenotypic expression of one 72 signal impinges on the resulting phenotype of the other signal. Signals are typically costly 73 to produce (Hasson 1994, Salvador et al. 1996), to maintain (Ruiz-Rodríguez et al. 2015), 74 or to show (i.e. social cost; Tibbetts and Dale 2004), and the expression of signals or 75 signal components may be traded-off against each other. On the one hand, there could be 76 a trade-off between two signals (e.g. by using the same resources as carotenes), so that a 77 lesser expression of one increases the expression of the other (Andersson et al. 2002). On 78 the other hand, it is also possible that the expression of one signal reduces the average 79 costs due to social interactions (Morales and Velando 2018) and, thus, facilitates or 80 enhances the expression of other signals. For instance, ornaments that develop before 81 reproduction and function in social contexts others than sexual (e.g. intra-sexual), and 82 5 could serve to stablish social hierarchy, may reduce agonistic social interactions and 83 mitigate subsequent energetic costs. Saved energy could thus enhance the production of 84 other sexual ornaments during courtship or reproduction and, therefore, the expression of 85 ornaments developed before and during reproduction could be positively related. This 86 might be the case of certain plumage characteristics of birds that reduce social costs 87 before reproduction (Senar et al. 2000), and thus, could boost the expression of other 88 sexually selected traits, such as song or other similar flexible dynamics traits, that are 89 exclusively expressed during reproduction (Badyaev et al. 2002, Mason et al. 2014). 90 Detecting evidence supporting the hypothesis that the expression of one signal is 91 conditioned by the expression of other signals can be challenging. A main reason is that 92 sexual signals are typically condition-dependent (e.g. Saino et al. 1997, Velando et al. 93 2006, Soler et al. 2008). Thus, detecting positive or negative associations between the 94 expression of different signals is not enough to infer causation. Rather, this hypothesis 95 should be tested in experimental frameworks where the modification of one signal causes 96 or explains the phenotypic expression of other signals. As far as we know, this hypothesis 97 has been tested experimentally only once by Henderson et al. (2018), who manipulated 98 plumage colouration of house finch (Haemorhous mexicanus) males before reproduction 99 and detected an effect on male investment in song under captivity conditions. However, 100 the effect was dependent on experimentally modified social context (feather colouration 101 of neighbours) and, thus, it is not completely clear that the detected effects were 102 exclusively caused by costs associated to plumage colouration. Here, we go a step further 103 and look for experimental evidence supporting the hypothesis in the wild in spotless 104 starlings (hereafter starlings, Sturnus unicolor). 105 Starlings are semi colonial and sexually dimorphic birds, with males showing 106 elongated throat feathers (Hiraldo and Herrera 1974, Lezana et al. 2000) and conspicuous 107 6 yellow beak with blue coloured basal part (Navarro et al. 2010). These two sexually 108 dimorphicsecondary sexual traits could harbour different kinds of information or, at least, 109 information at different time scales. The apical part of these feathers is quite flexible, and 110 males exhibit them very conspicuously during the entire year in social interactions, 111 including courtship (Aparicio et al. 2001, Ruiz-Rodríguez et al. 2015). In addition, these 112 feathers honestly reflect the phenotypic quality of individuals (Lezana et al. 2000, López-113 Rull et al. 2007, Gil and Culver 2011, Ruiz-Rodríguez et al. 2015). On the other hand, 114 during mating and reproduction (from February to July in our study area), the otherwise 115 black coloured beak of starlings turn to yellow colouration in both sexes, while its basal 116 part turn to blue in males and to pink in females (Cramp 1998) (Fig.1). Beak colour in 117 starlings is a sexually dimorphic and dynamic trait that likely reflects antioxidant capacity 118 (Navarro et al. 2010) and, accordingly, previous studies found that the yellow colour of 119 the beak is related to the level of carotenoids and vitamin A in the plasma in both sexes 120 (Navarro et al. 2010). The moult of throat feathers occurs in September-October (Veiga 121 and Polo 2016), thus far before the reproductive period. Therefore, it is likely that these 122 feathers serve to stablish social hierarchies within the population during the whole year, 123 allowing to reduce agonistic interactions and to mitigate its associated costs (Andersson 124 1994). If that was the case, the length of the throat feathers could play an important role 125 during the non-breeding season by affecting the acquisition and allocation of resources, 126 which could be reflected in the intensity of beak colouration in starling males. Length of 127 throat feathers can be easily manipulated (see Material and Methods), so the hypothesis 128 that the expression of one signal (length of throat feathers) determines the expression of 129 the other (beak colouration) can be experimentally tested. 130 We manipulated the length of the throat feathers of males by cutting-off 131 approximately the half-distal portion of the feathers before reproduction, and explored its 132 13 factor (i.e. dependent variable), and experimental treatment as the categorical predictor. 286 The date of first and last captures, as well as number of days between captures were 287 included as continuous independent variables in the statistical models. In addition, we 288 checked whether the experiment did affect length of throat feathers after moult, by 289 carrying out repeated-measures ANOVAs. In this model, the feather length, at first and 290 last captures, was the dependent variables (repeated measures), the experimental 291 treatment was the categorical predictor, and the number of days between captures was the 292 continuous independent variable. Residuals of all statistical models were plotted and 293 visually checked for normality. All analyses were performed with Statistica V13 (Dell-294 Inc. 2015). 295 296 Results 297 The blue chroma and the yellow-red chroma of the starlings’ beak-base are 298 negatively related in both males (Beta(SE) = -0.68(0.08), F1,54 = 74.64, p < 0.001 ) and 299 females (Beta(SE) = -0.92(0.03), F1,100 = 933.58, p < 0.001). Moreover, the base (400-300 475 nm and 570-700 nm) and tip (450-570 nm) beak colours were not significantly 301 associated in males (blue400-475: Beta(SE) = 0.02(0.13), F1,54 = 0.03, p = 0.867; yellow-red 302 570-700: Beta(SE) = -0.08(0.11), F1,54 = 0.49, p = 0.487), but a tendency (positive for blue 303 and negative for yellow-red chroma) was detected in females (blue400-475: Beta(SE) = 304 0.14(0.07), F1,100 = 3.92, p = 0.050; yellow-red570-700: Beta(SE) = -0.14(0.07), F1,100 = 305 3.77, p = 0.055). 306 Body condition was positively and negatively related to intensity of blue and 307 yellow-red colouration of males’ beak, respectively (Table 1, Fig. 3). Neither the 308 brightness of males’ beak nor the length of their throat feathers were related to body 309 condition (Table 1). In females, none of these variables predicted body condition (Table 310 1). Similarly, length of throat feathers of males, but not that of females, was positively 311 14 and negatively related to blue and yellow-red colour intensity of males’ beak, respectively 312 (Table 1, Fig. 3). Beak brightness did not predict length of throat feathers of males or 313 females (Table 1). Thus, the blue colour intensity of males’ beak co-varied with the length 314 of throat feathers, which might inform females on the phenotypic quality (body condition) 315 of males. 316 Importantly, the experimental shortening of throat feathers in males provoked a 317 reduction in the intensity of the blue, but no other, colouration of their beaks (measured 318 one or two years after manipulation of throat feathers) (Table 2, Fig. 4). Moreover, the 319 experimental manipulation did not affect body condition or the length of throat feathers 320 in subsequent captures (Table 2). Neither date of first and second capture nor time 321 between the two captures did explain additional significant proportion of variance (results 322 not shown). These results suggest a direct link between length of throat feathers and beak 323 colouration of males, which is independent of the association of both characters with 324 phenotypic condition of males. 325 326 Discussion 327 Our main results are that (i) intensity of colouration of the beak base of spotless 328 starling males, but not that of females, was positively related to body condition; (ii) beak 329 colouration of males was positively related to the length of their ornamental throat 330 feathers, and (iii) the experimental shortening of throat feathers in males had a negative 331 effect on the blue chroma intensity of the beak of males one year after manipulation. 332 Length of throat feathers and beak colouration are two sexually dimorphic traits that 333 reflect phenotypic quality of males (Aparicio et al. 2001, Navarro et al. 2010) and, thus, 334 our results demonstrate a direct connection between these two traits suggesting that they 335 may function as a whole in a multiple signalling framework. 336 15 Starlings have several known sexually dimorphic traits and are an appropriate 337 model system to explore functional interactions between sexual signals. Most studies on 338 sexual signals in this species are focussed on the length of throat feathers of males, which 339 predicts mating success (Aparicio et al. 2001), genetic heterozigosity (Aparicio et al. 340 2001), immune response (Gil and Culver 2011) and telomere length (Azcárate-García et 341 al. 2020). Bill colouration of the distal yellow part has also been studied as a sexually 342 selected trait of the species because it is related to carotenoid and vitamin A concentration 343 in the blood of males and females, but only during the mating period (Navarro et al. 2010). 344 Sexual differences are however more apparent at the basal part of the beak (Fig 1), and 345 we concentrated on this trait to experimentally explore the possible association with the 346 length of throat feathers. In agreement with the assumption that the blue colouration of 347 the basal part of the beak has a sexual-signalling function, we found that its blue-colour 348 intensity was positively related with both body condition and the length of the throat 349 feathers. Thus, exploring the interaction between these two traits is justified. 350 Length of throat feathers and beak coloration of starling males provide 351 information at different time scales. Black feathers are relatively static and would provide 352 information of the phenotypic condition and quality of males at the time of moulting 353 (Badyaev and Hill 2000, Hebets and Papaj 2005). Moreover, feather deterioration would 354 also provide information on feather quality and on ability of males reducing feather 355 degradation (Shawkey et al. 2007, Shawkey et al. 2009, Ruiz-de-Castañeda et al. 2012, 356 Ruiz-Rodríguez et al. 2015). Thus, length of throat feathers might even include different 357 kinds of information at a long-term scale. The beak colouration, however, should function 358 at a short time scale. Like for the colouration of other bare parts of birds, beak colouration 359 has the potential to change within weeks, days, hours, or even seconds (Iverson and 360 Karubian 2017). Thus, this kind of dynamic characters should be continuously evaluated 361 16 by receivers (Velando et al. 2006, Simons and Verhulst 2011, Dey et al. 2015). As far as 362 we know, associations between these two types of sexually dimorphic traits have never 363 been assessed. 364 Our results showed that male body condition at the time of mating was related to 365 blue colouration of the beak, but not to the length of throat feathers, suggesting that both 366 signals do not provide identical but, perhaps, complementary information. This could be 367 due to the fact that throat feather length would explain body condition of males at the 368 time of moulting, while beak colouration would be a more dynamic character that, 369 similarly to the colour of the legs of blue-footed boobies (Sula nebouxii) (Torres and 370 Velando 2007), shows individual condition at the time of capture. However, length of 371 throat feathers was positively related to the intensity of blue colouration of the beak and, 372 thus, it is possible that both traits convey redundant information to females. In agreement 373 with the possibility that these two traits transfer complementary information to females, 374 we experimentally showed a negative effect of length of throat feathers on the intensity 375 of the blue chroma of the beak base of males several months after the manipulation. We 376 know that colouration of the tip of the beak reflects the antioxidant capacity of starlings 377 (Navarro et al. 2010). The association between beak coloration and carotenoids’ 378 concentration in the blood has also been detected in other species (Faivre et al. 2003). We 379 did not measure concentration of carotenoids in the blood in this study and, thus, we 380 cannot explore whether this association exists for the blue coloration of the beak base of 381 males. Moreover, colour reflectance of the beak tip at the carotenoid wavelength, which 382 resulted positively related to carotenoid level in starlings (Navarro et al. 2010), was not 383 related to colouration of the base of the beak of males. Consequently, the colouration of 384 the beak base is unlikely conveying information on antioxidant capacity to females. Thus, 385 our experimental results should be interpreted as length of throat feathers functioning 386 17 during the non-reproductive period and determining phenotypic condition of males during 387 mating. 388 Like other signals operating in non-sexual scenarios such as parent-offspring 389 communication (Morales and Velando 2018), or sibling negotiation (Johnstone and 390 Roulin 2003, Soler and Avilés 2010), including those mediated by feather colourations 391 (Senar 2006), the length of throat feathers of males might serve to stablish some kind of 392 social hierarchy between males that reduce the probability of agonistic interactions 393 among individuals of different status (Rohwer 1975, Senar 1999, McGraw and Hill 2000). 394 Starlings moult throat feathers several months before reproduction, and males frequently 395 display these feathers while singing in high visible places during non-reproductive 396 periods (pers. obs.), which might have a functional significance in a context of social 397 interactions. In some bird species, probability of social aggression by conspecifics is 398 related to feather characteristics signalling bird status (Senar 1990, McGraw et al. 2007, 399 Chaine and Lyon 2008)). Moreover, aggressions are more common among individuals 400 showing similar status (Midamegbe et al. 2011), with individuals harbouring signals of 401 higher quality eliciting lower level of aggressiveness (Lopez-Idiaquez et al. 2016). The 402 experimental reduction of throat feathers lasts until the next moult period in autumn and, 403 thus, it is possible that starling males with longer throat feathers experienced lower rates 404 of social aggressions during the non-reproductive period. These costs can affect the 405 expression of other traits related to phenotypic condition, including immune responses 406 (Hawley et al. 2006), oxidative status (Galván and Alonso-Alvarez 2009) or the 407 expression of sexual signals (Møller et al. 2000). Although we have no data on probability 408 of aggression or social interactions in general in relation to length of throat feathers in 409 starlings, we think that social costs associated to the experimental reduction of length of 410 throat feathers during the non-breeding period is the most likely explanation for the 411 18 detected experimental effects on beak colouration during reproduction. However, this 412 mechanistic explanation deserves further research exploring for instance the expected 413 association between feather length and aggression during the non-reproductive period. 414 Whatever the mechanistic explanations, our experimental results strongly suggest 415 a causal link between expression of two sexually dimorphic traits in spotless starlings. As 416 far as we know, causal links between two sexually selected traits have only been detected 417 in another bird species, the house finch, a highly social species in which head and breast 418 feathers of males show great variability from red to yellow colouration (Henderson et al. 419 2018). Henderson et al. (2018) found that red-feathered males are more attractive and 420 sing more than yellow-feathered males but, when yellow males were housed with red 421 males, they sang more than when housed with equally unattractive yellow males. Thus, 422 males adapted their singing effort to the social environment (attractiveness) determined 423 by the plumage coloration of the social groups. Therefore, the detected link was 424 explained, not as a direct consequence of one of the traits, but indirectly by the social 425 environment in terms of level of attractiveness of neighbours, which was also 426 manipulated. Our experimental results therefore show a direct causal effect of length of 427 throat feathers on the expression of the colouration of the base of the beak of spotless 428 starling males, a trait that is only expressed during the reproductive period. 429 To conclude, we demonstrate for the first time a causal link between the 430 expression of two sexually dimorphic characters, which is essential to understand their 431 functionality in a multiple signalling framework. This type of interactions between 432 sexually selected signals might be widespread in nature and could be more easily detected 433 when considering signals that, like feather coloration or morphological traits, have 434 signalling functions in non-reproductive contexts. 435 436 19 References 437 438 Andersson, M. 1994. Sexual selection, Princeton, Princeton University Press. 439 Andersson, M. and Simmons, L. W. 2006. Sexual selection and mate choice. -Trends in ecology 440 & evolution, 21: 296-302. 441 Andersson, S., Pryke, S. R., Ornborg, J., Lawes, M. J. and Andersson, M. 2002. Multiple 442 receivers, multiple ornaments, and a trade-off between agonistic and epigamic 443 signaling in a widowbird. -American Naturalist, 160: 683-691. 444 Aparicio, J. M., Cordero, P. J. and Veiga, J. P. 2001. A test of the hypothesis of mate choice 445 based on heterozygosity in the spotless starling. -Animal Behaviour, 62: 1001-1006. 446 Azcárate-García, M., Ruiz-Rodríguez, M., Díaz-Lora, S., Ruiz-Castellano, C., Martín-Vivaldi, M., 447 Figuerola, J., Martínez-De La Puente, J., Tomás, G., Pérez-Contreras, T. and Soler, J. J. 448 2020. Ornamental Throat Feathers Predict Telomere Dynamic and Hatching Success in 449 Spotless Starling (Sturnus unicolor) Males. -Frontiers in Ecology and Evolution, 7. 450 Badyaev, A. V. and Hill, G. E. 2000. Evolution of sexual dichromatism: contribution of 451 carotenoidversus melanin-based coloration. -Biological Journal of the Linnean 452 Society, 69: 153-172. 453 Badyaev, A. V., Hill, G. E. and Weckworth, B. V. 2002. Species divergence in sexually selected 454 traits: Increase in song elaboration is related to decrease in plumage ornamentation in 455 finches. -Evolution, 56: 412-419. 456 Balmford, A., Rosser, A. M. and Albon, S. D. 1992. Correlates of female choice in resource-457 defending antelope. -Behavioral Ecology and Sociobiology, 31: 107-114. 458 Bro-Jorgensen, J. and Dabelsteen, T. 2008. Knee-clicks and visual traits indicate fighting ability 459 in eland antelopes: multiple messages and back-up signals. -Bmc Biology, 6: 9. 460 Candolin, U. 2003. The use of multiple cues in mate choice. -Biological Reviews, 78: 575-595. 461 Chaine, A. S. and Lyon, B. E. 2008. Intrasexual selection on multiple plumage ornaments in the 462 lark bunting. -Animal Behaviour, 76: 657-667. 463 Chaine, A. S. and Lyon, B. E. 2015. Signal architecture: temporal variability and individual 464 consistency of multiple sexually selected signals. -Functional Ecology: n/a-n/a. 465 Cramp, S. 1998. Cramp's the complete birds of the Western Palearctic, Oxford, Optimedia, 466 Oxford University Press. 467 Dey, C. J., Valcu, M., Kempenaers, B. and Dale, J. 2015. Carotenoid-based bill coloration 468 functions as a social, not sexual, signal in songbirds (Aves: Passeriformes). -Journal of 469 Evolutionary Biology, 28: 250-258. 470 Edward, D. A. 2015. The description of mate choice. -Behavioral Ecology, 26: 301-310. 471 Endler, J. A. and Houde, A. E. 1995. Geographic-variation in female preferences for male traits 472 in poecilia-reticulata. -Evolution, 49: 456-468. 473 Faivre, B., Gregoire, A., Preault, M., Cezilly, F. and Sorci, G. 2003. Immune activation rapidly 474 mirrored in a secondary sexual trait. -Science, 300: 103-103. 475 Galván, I. and Alonso-Alvarez, C. 2009. The expression of melanin-based plumage is separately 476 modulated by exogenous oxidative stress and a melanocortin. -Proc. R. Soc. Lond. B. 477 Gil, D. and Culver, R. 2011. Male ornament size in a passerine predicts the inhibitory effect of 478 testosterone on macrophage phagocytosis. -Functional Ecology, 25: 1278-1283. 479 Girard, M. B., Elias, D. O. and Kasumovic, M. M. 2015. Female preference for multi-modal 480 courtship: multiple signals are important for male mating success in peacock spiders. -481 Proceedings of the Royal Society B-Biological Sciences, 282: 10. 482 Gomez, D. 2006. AVICOL, a program to analyse spectrometric data, free program available 483 from the author at dodogo[email protected]. 484 Green, A. J. 2001. Mass/length residuals: measures of body condition or generators of spurious 485 results? -Ecology, 82: 1473-1483. 486 20 Hasson, O. 1994. Cheating signals. -Journal of Theoretical Biology, 167: 223-238. 487 Hawley, D. M., Lindstrom, K. and Wikelski, M. 2006. Experimentally increased social 488 competition compromises humoral immune responses in house finches. -Hormones 489 and Behavior, 49: 417-424. 490 Hebets, E. A. and Papaj, D. R. 2005. Complex signal function: developing a framework of 491 testable hypotheses. -Behavioral Ecology and Sociobiology, 57: 197-214. 492 Henderson, L. J., Brazeal, K. R. and Hahn, T. P. 2018. Plumage coloration and social context 493 influence male investment in song. -Biology Letters, 14: 4. 494 Hiraldo, F. and Herrera, C. M. 1974. Dimorfismo sexual y diferenciación de edades en Sturnus 495 unicolor Temm. -Doñana, Acta Vert, 1: 149-170. 496 Iverson, E. N. K. and Karubian, J. 2017. The role of bare parts in avian signaling. -Auk, 134: 587-497 611. 498 Johnstone, R. A. 1996. Multiple displays in animal communication: 'Backup signals' and 499 'multiple messages'. -Philosophical Transactions of the Royal Society B-Biological 500 Sciences, 351: 329-338. 501 Johnstone, R. A. and Roulin, A. 2003. Sibling negotiation. -Behavioral Ecology, 14: 780-786. 502 Kokko, H. 2003. The evolution of mate choice and mating biases. -Proceedings of the Royal 503 Society of London B, Biological Sciences, 270: 653-664. 504 Kraaijeveld, K., Kraaijeveld-Smit, F. J. L. and Komdeur, J. 2007. The evolution of mutual 505 ornamentation. -Animal Behaviour, 74: 657-677. 506 Lezana, L., Miranda, R., Campos, F. and Peris, S. J. 2000. Sex differentiation in the spotless 507 starling (Sturnus unicolor, Temminck 1820). -Belgian Journal of Zoology, 130: 139-142. 508 Lopez-Idiaquez, D., Vergara, P., Fargallo, J. A. and Martinez-Padilla, J. 2016. Female plumage 509 coloration signals status to conspecifics. -Animal Behaviour, 121: 101-106. 510 López-Rull, I., Celis, P. and Gil, D. 2007. Egg colour covaries with female expression of a male 511 ornament in the spotless starling (Sturnus unicolor). -Ethology, 113: 926-933. 512 Lyon, B. E. and Montgomerie, R. 2012. Sexual selection is a form of social selection. -513 Philosophical Transactions of the Royal Society B-Biological Sciences, 367: 2266-2273. 514 Martin, J. and Lopez, P. 2009. Multiple color signals may reveal multiple messages in male 515 Schreiber's green lizards, Lacerta schreiberi. -Behavioral Ecology and Sociobiology, 63: 516 1743-1755. 517 Mason, N. A., Shultz, A. J. and Burns, K. J. 2014. Elaborate visual and acoustic signals evolve 518 independently in a large, phenotypically diverse radiation of songbirds. -Proceedings of 519 the Royal Society B-Biological Sciences, 281: 9. 520 Mcgraw, K. J. and Hill, G. E. 2000. Carotenoid-based ornamentation and status signaling in the 521 house finch. -Behavioral Ecology, 11: 520-527. 522 Mcgraw, K. J., Medina-Jerez, W. and Adams, H. 2007. Carotenoid-based plumage coloration 523 and aggression during molt in male house finches. -Behaviour, 144: 165-178. 524 Midamegbe, A., Gregoire, A., Perret, P. and Doutrelant, C. 2011. Female-female aggressiveness 525 is influenced by female coloration in blue tits. -Animal Behaviour, 82: 245-253. 526 Møller, A. P., Biard, C., Blount, J. D., Houston, D. C., Ninni, P., Saino, N. and Surai, P. F. 2000. 527 Carotenoid-dependent signals: Indicators of foraging efficiency, immunocompetence 528 or detoxification ability? -Avian and Poultry Biology Reviews, 11: 137-159. 529 Møller, A. P. and Pomiankowski, A. 1993. Why have birds got multiple sexual ornaments. -530 Behavioral Ecology and Sociobiology, 32: 167-176. 531 Morales, J. and Velando, A. 2018. Coloration of chicks modulates costly interactions among 532 family members. -Behavioral Ecology, 29: 894-903. 533 Navarro, C., Pérez-Contreras, T., Avilés, J. M., Mcgraw, K. J. and Soler, J. J. 2010. Beak colour 534 reflects circulating carotenoid and vitamin A levels in spotless starlings (Sturnus 535 unicolor). -Behavioral Ecology and Sociobiology, 64: 1057-1067. 536 Perrier, C., De Lope, F., Møller, A. P. and Ninni, P. 2002. Structural coloration and sexual 537 selection in the barn swallow Hirundo rustica. -Behavioral Ecology, 13: 728-736. 538 21 Rohwer, S. 1975. Social significance of avian winter plumage variability. -Evolution, 29: 593-539 610. 540 Ruiz-Castellano, C., Ruiz-Rodriguez, M., Tomas, G. and Soler, J. J. 2019. Antimicrobial activity of 541 nest-lining feathers is enhanced by breeding activity in avian nests. -FEMS 542 microbiology ecology, 95. 543 Ruiz-Castellano, C., Tomás, G., Ruiz-Rodríguez, M., Martín-Galvez, D. and Soler, J. J. 2016. Nest 544 Material Shapes Eggs Bacterial Environment. -PLoS One, 11: e0148894. 545 Ruiz-De-Castañeda, R., Burtt Jr, E., González--Braojos, S. and Moreno, J. 2012. Bacterial 546 degradability of an intrafeather unmelanized ornament: a role for feather-degrading 547 bacteria in sexual selection? -Biological Journal of the Linnean Society, 105: 409-419. 548 Ruiz-Rodríguez, M., Tomás, G., Martín-Gálvez, D., Ruiz-Castellano, C. and Soler, J. J. 2015. 549 Bacteria and the evolution of honest signals. The case of ornamental throat feathers in 550 spotless starlings. -Functional Ecology, 29: 701-709. 551 Saino, N., Bolzern, A. M. and Møller, A. P. 1997. Immunocompetence, ornamentation, and 552 viability of male barn swallows (Hirundo rustica). -Proceedings of the National 553 Academy of Sciences of the USA, 94: 549-552. 554 Salvador, A., Veiga, P., Martín, J., López, P., Abelenda, M. and Puerta, M. 1996. The cost of 555 producing a sexual signal: testoterone increases tha susceptibility of male lizards to 556 ectoparasitic infestation. -Behav. Ecol, 7: 145-150. 557 Senar, J. 1990. Agonistic communication in social species: what is communicated? -Behaviour, 558 112: 270-283. 559 Senar, J. 1999. Plumage coloration as a signal of social status. Proceedings of the International 560 Ornithological Congress, 1999. BirdLife South Africa, pp. 1669-1686. 561 Senar, J. 2006. Color displays as intrasexual signal of agression and dominance In: HILL, G. E. 562 and MCGRAW, K. J. (eds.) Bird coloration: Function and evolution. Harvard University 563 Press, Cambridge, Massachusetts, pp. 87-136. 564 Senar, J. C., Polo, V., Uribe, F. and Camerino, M. 2000. Status signalling, metabolic rate and 565 body mass in the siskin: the cost of being a subordinate. -Animal Behaviour, 59: 103-566 110. 567 Shawkey, M. D., Pillai, S. R. and Hill, G. E. 2009. Do feather-degrading bacteria affect sexually 568 selected plumage color? -Naturwissenschaften, 96: 123-128. 569 Shawkey, M. D., Pillai, S. R., Hill, G. E., Siefferman, L. M. and Roberts, S. R. 2007. Bacteria as an 570 agent for change in structural plumage color: Correlational and experimental evidence. 571 -American Naturalist, 169: S112-S121. 572 Simons, M. J. P. and Verhulst, S. 2011. Zebra finch females prefer males with redder bills 573 independent of song rate-a meta-analysis. -Behavioral Ecology, 22: 755-762. 574 Soler, J. J. and Avilés, J. M. 2010. Sibling competition and conspicuousness of nestling gapes in 575 altricial birds: a comparative study. -PLoS ONE, 5: e10509. 576 Soler, J. J., Navarro, C., Pérez-Contreras, T., Avilés, J. M. and Cuervo, J. J. 2008. Sexually 577 selected egg coloration in spotless starlings. -American Naturalist, 171: 183-194. 578 Soler, J. J., Ruiz-Castellano, C., Figuerola, J., Martin-Vivaldi, M., Martinez-De La Puente, J., Ruiz-579 Rodriguez, M. and Tomas, G. 2017. Telomere length and dynamics of spotless starling 580 nestlings depend on nest-building materials used by parents. -Animal Behaviour, 126: 581 89-100. 582 Tibbetts, E. A. and Dale, J. 2004. A socially enforced signal of quality in a paper wasp. -Nature, 583 432: 218-222. 584 Veiga, J. and Polo, P. 2016. Estornino Negro–Sturnus unicolor. En: Enciclopedia Virtual de los 585 Vertebrados Españoles. Museo Nacional de Ciencias Naturales, Madrid, Salvador, A., 586 Morales, M. B. (Eds.). 587 Velando, A., Beamonte-Barrientos, R. and Torres, R. 2006. Pigment-based skin colour in the 588 blue-footed booby: an honest signal of current condition used by females to adjust 589 reproductive investment. -Oecologia, 149: 535-542. 590 22 591 29 Fig.4 643 644 First Second Capture 0.18 0.20 0.22 0.24 0.26 Blue Intensity Experimental Control 645 646 647 30 Annex 1 648 649 Table A1: Number of captures made for each of the recaptured spotless starling males 650 during the study and the number of times the treatment was applied to each male. 651 652 Ring Treatment Year 2015 Year 2016 Year 2017 Number of treatments 3256561 Control 2 1 3256564 Control 1 1 1 3256565 Control 1 1 1 3256567 Control 2 1 1 3256590 Control 2 1 1 3301955 Control 3 1 1 3368660 Control 2 1 3368681 Control 1 1 1 2 3369509 Control 1 1 1 2 3406027 Control 1 1 1 3418841 Control 1 1 1 3256556 Experimental 2 1 3285646 Experimental 2 1 3387759 Experimental 1 2 1 2 3387764 Experimental 2 1 1 3387774 Experimental 1 1 1 3387838 Experimental 1 1 1 3392095 Experimental 1 1 1 3428304 Experimental 2 1 3256564b Experimental 1 2 1 3256565b Experimental 1 1 1 3256567b Experimental 1 1 1 653