Age-related changes in the temporal structure of song in the common rosefinch Erythrina erythrina

Bird species that have visual differences in plumage coloration between young and older males may also differ in their song parameters. We have studied the difference in song rate between yearlings vs. older males in the common rosefinch Erythrina erythrina, a species with delayed plumage maturation. Yearlings sang 2.3 more strophes per minute compared to older males, and the breaks between strophes were 1.3 s shorter in young males. The higher singing rate of yearling males is probably a form of compensation for lower age (male quality), which may be important in finding a mate or avoiding aggression from older males.

Bird song is normally a characteristic of males and is a sexually selected trait characterized by extremely varying complexity and fulfils many functions including and mate attraction (Mennill et al. 2002). Females choose males according to various traits including certain song characteristics such as song rate, song types or repertoire size (Gil et al. 2001). Many studies suggest that variation in song characteristics is associated not only with individual fitness of singers but also with their age (Kipper and Kiefer 2010;Nemeth et al. 2012). In many bird species, plumage coloration is often related to age (Delhey and Kempenaers 2006;Mitrus 2006). Yearling individuals are usually duller than older males, this phenomenon being defined as delayed plumage maturation (Rohwer et al. 1980). Bird species that have visual age differences may also differ in their song parameters (Garamszegi et al. 2006;Balsby and Hansen 2010). For example, in saffron finches Sicalis flaveola pelzelni second year males sang shorter syllables and shorter and less versatile songs than older males (Benítez Saldívar et al. 2020) and similar relationships between age and songs were found in the whinchat Saxicola rubetra (Vaytina and Shitikov 2019). In general, many studies have shown that song repertoire sizes increase with age (Balsby and Hansen 2010;Kiefer et al. 2011).
The common rosefinch Erythrina erythrina is a species with a small song repertoire; it is short, including three to nine pure-tone elements (Martens and Kessler 2000). Males exhibit delayed plumage maturation; in their second year they have greyish-green heads and chests and in this plumage coloration are similar to females, whereas in older males, these parts of the plumage have a carmine colour (Cramp and Perrins 1994). The characteristic of song of this species, taking into account the age of the males, has been studied to a limited extent. The strophe parameters of yearlings and older males were generally similar, but with the difference that older males sang a longer strophe than yearlings (Parapura et al. 2018). However, in some species other characteristics of singing also play an important role in male characteristics and possibility of obtaining a breeding partner, for example, the song rate (Hanski and Laurila 1993). A higher song rate in such males is preferred by females and affects the rate at which males mate (Arvidsson and Neergard 1991).
In this paper, we studied whether selected singing parameters were related to male age in the common rosefinch. We expected older males to have a faster singing rate than young males, as the rate of singing is positively correlated with the quality of males (Brooks and Kemp 2001;Garamszegi et al. 2007). Our analysis included the tempo of singing, as well as the time of the average and minimum breaks between strophes, which are related to the rate of singing. The study area is located along the Kostrzyń river in east-central Poland protected as Natura 2000 "Kostrzyń Valley" network (52°11′06.7"N; 21°58′02.5"E). Within this area there are 44 kms of river bed with adjacent meadows, bushes, fish ponds and forests. There were 114-150 breeding pairs of common rosefinch, of which 38% were one-year-old males (Parapura 2018). The research was conducted only in good weather conditions in May-June 2015. Male songs were recorded using a Marantz PMD620 digital recorder connected to a microphone Sennheiser ME 66 with capsule K 6/6. Recorded songs were analysed using Audacity v. 1.3.9., and visual inspection of spectrograms. The date and time of the song's performance were also noted. Males were divided into two age classes depending on plumage: uncolored -yearling males, and fully colored -older males. Singing males were not marked in any way, thus in the analysis, to avoid pseudo-replication, we always used only one recording from a given place of male singing, which lasted at least 1 min. The total analysis covered 185 min of recordings (mean = 3 min 48 s, SD = 2 min 4 s, range = 1 min -9 min 28 s).
We used general linear models (GLM) to explore the effect of male age (factor: yearling vs. older male), day of season (1 = 1 May), and hour of recording (to full hour) (covariates) on the dependent variables: 1) song rate, 2) the average time of breaks between strophes in the song, 3) the minimum time of breaks between strophes in the song. The last two variables were subjected to a logarithmic transformation, after which their distribution was close to the normal distribution. In the analysis, we included variables describing the passage of the season and the hours of recording of males because these variables may be relevant to the parameters of birdsong (Gil and Gahr 2002). Explanatory variables were tested for multicollinearity by examining the Variance Inflation Factor (VIF) (Quinn and Keough 2002), however, the variables did not exceed the value of 1.08, so all of them were included in the model. Only those results with a probability of α ≤ 0.05 were assumed to be statistically significant. The analysis was performed in Statistica 12.0 (StatSoft Inc 2014).
In total, 53 recordings for 20 yearling males and 33 older males were obtained. Overall, song rate was 8.6 strophes/ minute (SE = 0.33, Range 3.8-14.6), average time of breaks between strophes was 6.0 s (SE = 0.26, Range 3.4-12.3 m), and minimum time of break between strophes was 2.9 s (SE = 0.25, Range 0.2-9.0). The GLM analysis showed that the number of strophes sung per minute depended on the age of the male (F 1.51 = 14.21, P < 0.001, Table 1) and was 2.3 strophes higher in yearling males (Fig. 1). The average time of breaks between strophes in the song also depended on the age of the male (F 1.51 = 5.14, P = 0.028, Table 1), the breaks being were 1.3 s shorter in young males (Fig. 2). However, we did not find factors that would significantly affect the minimum time of breaks between strophes in the song (Table 1).
We found large, but negative, effect sizes for agedependent variation in song rate, with yearlings singing more than 2 strophes/min than older males, and with over 1 s shorter breaks between strophes in younger males. Correlations between age and song in the common rosefinch support the hypothesis that song is a good age signal as it helps to separate yearlings from older birds (Mountjoy and

3
Lemon 1995; Nemeth et al. 2012). However, we expected the opposite trend, as it is generally believed that the rate of singing is positively correlated with the quality of males (Garamszegi et al. 2007). In general, a positive relationship between song rate and male age is observed (Welling et al. 1997;Murphy et al. 2008), but a higher singing rate has also been noted in young males (Garamszegi et al. 2007;Nemeth et al. 2012). These authors offered two explanations for their results. First, they argued that the higher singing rate of yearlings may be a compensation for the larger size of the repertoire in older males. Second, the higher singing rate may be a consequence of higher aggression in yearlings, the singing rate being related to more intense territorial defense. In our study, the second hypothesis is rather unlikely, because our many hours of observations did not show that young males were aggressive towards older males. In previous studies, we found that young males sing shorter strophes compared to older males (Parapura et al. 2018). It is likely that longer strophes are an important part of singing for mate selection (Wasserman and Cigliano 1991;Nolan and Hill 2004). Producing a large repertoire (longer strophes) can be costly due to investment in a larger vocal centre of the brain (Airey et al. 2000) or to immune system suppression (Møller et al. 2000). Perhaps, therefore, younger males are unable to sing longer strophes and try to compensate for this with a faster singing rate, although this may also have an adverse effect on their vocal organs (Deoniziak and Osiejuk 2016). Singing shorter strophes while maintaining a faster tempo can also be a message about the age of males, their lower experience, to avoid aggression from older males. Only some males of common rosefinch start breeding in their second calendar year, and some males may also play an auxiliary role in the territories occupied by older males (Payevsky 2008). We also recorded intensely singing juvenile males in the territories of older males, which did not show aggression towards the younger birds.
In conclusion, the higher singing rate of young male common rosefinches compared to older males is surprising, although such relationships have also been found in several other species. Probably, young males have weaker singing parameters, e.g., shorter strophes, and try to compensate for these imperfections by singing faster and avoid aggression from older, more experienced males thanks to the information transmitted by the song.
Author contributions AP and AG designed the study. AP collected field data. AP and AG performed statistical analyses and wrote the part of the manuscript. CM wrote the part of manuscript and gave comments on the manuscript. All authors read and approved the final manuscript.
Funding This study was financially supported by the Siedlce University of Natural Sciences and Humanities with Grant number 76/20/B (to A. Parapura, A. Golawski).

Declarations
Ethical note This study complies with current Polish laws. No permit from the Bio-Ethical Committee was needed.

Competing interests The authors declare no conflict of interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.