Many bird species, notably but not exclusively the songbirds (Passeriformes), use vocal sounds to fulfill intersexual interactions, such as self-advertising and stimulating mates (e.g., Kroodsma, 1976; Otter and Ratcliffe, 1993; Welling et al., 1995), or intrasexual interactions like social communication and acquiring or defending territory (Amrhein and Erne, 2006; Foote et al., 2008), or both. This behaviour is particularly prominent at dawn and for males in the breeding seasons (Slagsvold, 1996; Kunc et al., 2005; McNeil et al., 2005; Amrhein and Erne, 2006), and often turns into a collective phenomenon known as dawn chorus.
While the association of dawn chorus with the breeding seasons offers ontogenetic (i.e., sexual maturity) and mechanistic (i.e., hormone) explanations, the reasons why male birds engage in vocal activity vigorously at dawn appear to be due to, at least in part, certain proximate physical factors that are not mutually exclusive, and also carry functional perspectives. Lower temperatures and high humidity at dawn may restrict activity of both birds and their prey (Avery and Krebs, 1984), but also that of predators that otherwise would be attracted to bird calls. In contrast, lower wind speed and background noise levels in this period are more energetic-efficient and favourable for sound travel with consistency (i.e., the acoustic transmission hypothesis; Henwood and Fabrick, 1979; Cuthill and Macdonald, 1990; Brown and Handford, 2003; but see Dabelsteen and Mathevon, 2002). The limited light level at dawn prohibits diurnal birds from many daytime activities, such as foraging (i.e., the inefficient foraging hypothesis; Kacelnik, 1979; Kacelnik and Krebs, 1982) or social contacts with one another (Cuthill and Macdonald, 1990), but the latter can be achieved vocally (Kacelnik, 1979; Amrhein and Erne, 2006; Foote et al., 2008).
Temporal variation in dawn chorus may also be caused by proximate factors such as light availability that is usually, but not completely (see Kempenaers et al., 2010; Da Silva et al., 2015), related to time, weather, and vegetation (Berg et al., 2006). Overnight birds may have surplus energy from previous feeding to sustain the night time fast and a lower temperature, which helps to explain why birds can afford to delay foraging and instead engage in singing at dawn (i.e., energy storage stochasticity hypothesis; Hutchinson, 2002). This reasoning concerns more on the variability of participation, so individual variation in quality or status can be assessed (i.e., state-dependent), but does not directly address the sequence and variation at dawn singing among species (Cuthill and Macdonald, 1990; Berg et al., 2006). Masking of signals, such as by noise, carry additional cost and negatively impact avian reproductive success (Halfwerk et al., 2011). Simultaneous songs by various species may also interfere with each other and affect the transmission effectiveness, thus a temporal asynchrony to avoid signal overlapping should be favored (Cody and Brown, 1969; Ficken et al., 1974; Popp et al., 1985).
Species may differ in their sensitivity to light, therefore sense the light level effectively at slightly different time or place (Thomas et al., 2002; McNeil et al., 2005; Berg et al., 2006). This has been noted particularly for species in forested areas (e.g., Popp et al., 1985). In addition, singing attracts predators as well as potential mates, but a bird may lower its alert to potential hazards while singing. Thus, birds should be selected for to sing only if they are able to scan sufficiently the surrounding at a certain light level (Thomas et al., 2002). It is expected that the earlier a bird sings at dawn, the better it is able to detect the surroundings even at dim light (Hutchinson, 2002; Thomas et al., 2002; McNeil et al., 2005), which may be determined by a bird’s visual anatomical features, but also associated with the surrounding settings and light levels where a bird is situated.
Birds with larger eyes tend to have earlier dawn singing (Thomas et al., 2002; McNeil et al., 2005). This is supported by anatomical relationships that larger-eyed birds own higher numbers of light-sensitive visual cells on the retina, a thicker inner retinal layer, and more neuron synapses, all these indicating a better visual sensitivity and higher resolution for low light conditions (Hart, 2001a, 2001b; Garamszegi et al., 2002; McNeil et al., 2005). Forest layers, due to the amount of foliage overhead, can have different light levels, and birds perching at a higher place usually receive a richer light availability. Yet, the position chosen by a bird for signaling (i.e., songposts) may not be the same as, and often is higher than, that used for roosting and foraging (Hunter, 1980; Barker and Mennill, 2009). Most forest birds foraging at lower-layers or on the ground also tend to roost at higher places or even up to the canopy at night. In contrast, species adapted to constantly foraging at lower heights or on the ground may have evolved a higher visual sensitivity and acuity that allows them to sense better in dim light earlier (McNeil et al., 2005; Thomas et al., 2006).
Further, avian foraging relies heavily on visual cues, so different visual sensitivities may have evolved in species adapted to different diets (Hart, 2001b; Garamszegi et al., 2002; Fernández-Juricic et al., 2004). Compared to plant resources, prey are often vigilant, requiring rapid detection by birds, or cryptic in coloration or pattern, both require substantial light levels to detect. In case of insect swarms, aerial insectivores are generally equipped with a larger degree of movement sensitivity and resolution-related double cone cells for detecting prey in movement at fast speeds (Hart, 2001a). Thus, predatory birds should have better visual capacity and light sensitivity to effectively sesarch for prey, particularly for foraging within a forest setting (Thomas et al., 2002; McNeil et al., 2005).
We examined the pattern and seasonal variation of avian dawn chorus in a tropical montane forest in Taiwan, where related information is extremely sparse. We tested whether species richness and the assemblage composition affect dawn chorus, and prediced that a greater number of species engaging in dawn chorus prolong the overall chorus duration but shorten individual species singing lengths. We also tested the hypothesis that the sequence of a dawn chorus can be determined by eye size, foraging height, and diet of birds, and predicted that earlier dawn singing is associated with species of larger relative eye size. In addition, we predicted that birds primarily foraging at forest floor or lower heights will sing earlier, and predatory birds will have an earlier dawn singing than plant-eating species.