Introduction

The behaviour of nest building is widespread among animals, including insects, fish, reptiles, mammals and birds (Hansell 2000). The vast majority of birds construct some sort of nest to lay eggs and raise offspring. It is widely accepted that bird nests evolved mainly to provide not only a secure substrate and thermally optimal conditions for eggs and nestlings, but also to provide camouflage and/or defence from predators, moderate the microclimate for the clutch and incubating parents and to provide many other benefits for nestlings and parents (Collias and Collias 1984; McGowan et al. 2004; Mainwaring 2015; Maziarz et al. 2017; Glądalski et al. 2020; Lambrechts and Caro 2022), but also costs (Glądalski et al. 2018). The behaviour by which a bird builds and maintains a nest is caused by genetic, phenotypic and cultural characteristics of the parents (Slagsvold 1989; Hansell 2000; Álvarez and Barba 2011; Deeming 2013; Lambrechts et al. 2014). In the Parids, a nest usually consists of two main parts. The lining layer is mainly associated with thermoregulation during incubation and the first days after hatching. The moss layer is rather considered as a structural support and sanitary layer, especially crucial at the nestling stage, but it also may have some thermoregulation functions (Gosler 1993; Cruz et al. 2016; Stenning 2018; but see Järvinen et al. 2017a). A substantial number of avian species (including tits) tend to use bryophytes as nest construction material (with some anthropogenic additions in some urbanized areas, see Jagiełło et al. 2022), but repeatability of nest parameters among individual territories remains largely unexplored (Glądalski et al. 2021).

There is also a large amount of variation in nest size and shape not only across bird species as well as within taxa (Alabrudzińska et al. 2003; Heenan 2013; Mainwaring et al. 2014). No single factor can explain why nests are so variable and studies show that depth and size of a nest is affected by a wide variety of factors. For example secondary cavity nesters, like tits, may change the mass, depth and/or composition of the nest depending on availability of the artificial food (Mainwaring and Hartley 2009), nestbox size (Lambrechts et al. 2013; Deeming et al. 2019), light intensity (Podkowa and Surmacki 2017; Holveck et al. 2019), rapid change in ambient temperature (Deeming et al. 2012, but experimental manipulation of photoperiod and temperature did not influence nest size in captive Blue and Great Tits, see Lambrechts and Caro 2018), amount of rainfall, moisture, latitude, climate (Perez et al. 2020), parental quality, tit species (Lambrechts et al.2014) or predation pressure (Kaliński et al. 2014). Experiments on captive and wild tits also showed that replacement nests are smaller than the first nests (Lambrechts et al. 2012; Deeming and Mainwaring 2015). In addition Hanmer et al. (2017) found that Blue Tits Cyanistes caeruleus (but not Great Tits Parus major) had lighter nest in more urbanized areas of the same English town. But in general, different studies provide rather mixed support for differences in nest parameters related to urbanization (Reynolds et al. 2019) and some authors did not find any variation in the nest size or mass along the urbanization gradient in studied tit species, so the discussion continues (Glądalski et al. 2016; Lambrechts et al. 2017). Recent experimental evidence on captive Zebra Finch Taeniopygia guttata suggests that some birds tend to build larger nest if their first nest did not successfully produced fledglings (Edwards et al. 2020).

Studies examining characteristics of bird nests and aspects of nest-building behaviour are gaining more and more interest. When compared to other stages of bird reproduction the number of nest-studies is still smaller, but the number of publications is increasing in recent years (Mainwaring et al. 2014; Harnist et al. 2020). But still long-term/large-scale studies that illustrate the variation of parameters of tit nests or in different study areas over a sufficiently long period of time are rather scarce (Lambrechts et al. 2016a, b; O’Neill et al. 2018). The majority of reports refer to usually only 1–2 seasons of research, and inferring from such a small number of seasons can give incorrect or sometimes even opposite results. In this paper, we show variation in nest depth and nest mass during an 11-year nestbox study on Great and Blue Tits between two floristically and structurally contrasting study areas: an urban parkland and a deciduous forest. We also present repeatability of nest parameters (depth and mass) among selected individual territories between years in both tit species and the comparative analysis of nest depth constructed by young (yearling) and older females in both tit species. Our prediction is that there should be differences in depth and mass of the nests between Blue and Great Tits. We also predict that nests should differ in depth and mass among years, but not between study areas in both tit species.

Methods

This study was carried out during the years 2012–2022 in two structurally and floristically contrasting study areas: a rich deciduous forest (51°50′N, 19°29′E) and an urban parkland (51°76′N, 19°41′E), located 10 km apart. The present study is part of a long-term project of research into the breeding biology of secondary cavity nesters within around the city of Łódź, central Poland (Glądalski et al. 2016). Both study sites were supplied with standardized nestboxes with a removable front wall and all studied nests came from nestboxes of exactly the same dimensions of 30 (height) × 11 (width) × 11.5 (depth) cm and a 30 mm diameter entrance (located 20 cm from the bottom of the front wall) (Lambrechts et al. 2010; Glądalski et al. 2016). All nestboxes were hung on tree trunks using nails at a height of c. 3 m, and in both study sites, the distance between nestboxes was about 50 m.

The urban parkland study area (Botanic and Zoological Gardens in Łódź) was c. 80 ha (200 nestboxes) with artificially fragmented tree cover (Marciniak et al. 2007). The vegetation in this area was formed artificially for the purpose of plant and animal exposition, meaning that the tree cover is largely patchy with extensive flowerbeds and lawns. In the Botanic Garden tree patches make up a mosaic of different deciduous and coniferous trees, containing a considerable number of exotic and foreign species. In the parkland study site, the mean number of trees within 25 m-radius circles surrounding nestboxes was 43.5 (Glądalski et al. 2017). The forest study site was a 145 ha area in the interior of the Łagiewniki forest (in total up to c. 1250 ha), equipped with 300 nestboxes. The dominant tree species in the Łagiewniki forest are the oaks—Quercus petraea and Quercus robur (more than 50% of all trees), with hornbeam Carpinus betulus, birches Betula pendula, maples Acer spp., limes Tilia spp., being less numerous species. The average number of trees within a circular 25 m-radius surrounding of the nestbox in the forest study site was 130.7 (Glądalski et al. 2017).

Nest depth (cm), following Hansell (2000), Alabrudzińska et al. (2003) and Glądalski et al. (2016), was measured before the incubation stage (as the distance from the top rim of the nest cup to the nest base) with a steel ruler to the nearest 0.5 cm (during the short period removal of the front wall of the nestbox). Nest depth is also called in some studies nest height (Lambrechts et al. 2014) or nest thickness (Hurtrez-Boussès et al. 1999). The depth of 322 Great Tit nests (147 from the forest site and 175 from the parkland site) was measured (Table 1). The depth of 183 nests (69 from the forest and 114 from the parkland) was measured in Blue Tits. After the end of the breeding season all measured nests of both tit species were sampled. All collected nests were placed in a freezer (− 80 °C) for 24 h in order to kill non-parasitic and parasitic invertebrates by deep freezing. After freezing the nests were dried for another 24 h in a laboratory dryer at 60 °C and then the nests were weighed to the nearest 0.1 g (by Glądalski et al. 2016, 2021).

Table 1 The numbers of nests measured and sampled in a field
Full size table

Adult Great Tits (98) and Blue Tits (21) were captured (2012–2022, no retraps) during the breeding season (i) in the nestbox or (ii) in a standardized manner using ornithological mist-nets placed next to the nestbox when parental birds fed 7–15-day-old nestlings. In the latter case, two 5-m-long mist-nets positioned fixedly during each trapping session and a loudspeaker with a recording of tit voices was used. Sex was determined by the presence of the brood patch in the females. The age of the females in both tit species (yearling: 1 year after hatching and older: ≥ 2 years after hatching) was determined after Demongin (2016) and Jenni and Winkler (2020).

Student’s t-tests for independent samples were used to examine the difference between the depth of the nest between younger (yearlings) and older females in both tit species. Pearson’s linear correlation analyses and t-tests were conducted applying Statistica 13.3 (TIBCO Statistica® 13.3 2017).

Nest depth and weight were separately modelled in relation to species (Great Tits vs Blue Tits), site (urban park site vs forest site) and year (2012–2022) as factors using general linear mixed models. Nestbox id was used as a random effect, with the denominator degrees of freedom being approximated by the Satterthwaite method. The modelling was performed applying IBM SPSS v. 22 software (IBM SPSS Statistics 22 2013). For both the nest traits initial models included species as a factor to compare effects for Great Tits and Blue Tits, but we assumed that we would further analyse submodels for the species separately. In all initial models and submodels main effects and their first-order interactions were included. These models were simplified by removing non-significant interactions, but retaining all main effects (Crawley 2002).

We further analysed the consistency of nest parameters across all cases of nests constructed in particular nestboxes by Blue Tits and Great Tits separately. To this end, we applied repeatability computed as intraclass correlation based on mean square sums from the general linear model where individual nestboxes were treated as a grouping factor, controlling for the number of nesting attempts in particular nestboxes (Bańbura and Zieliński 1990; Zar 2014). Standard errors of intraclass correlation coefficients were calculated according to Falconer (1989).

Results

The nest depth did not differ significantly between young (yearlings) and older females (≥ 2 years after hatching) in Great Tits (mean1yah = 9.89 ± 0.39 (SE) cm, mean≥2yah = 9.90 ± 0.35 (SE) cm, df = 96, t = 0.023, p = 0.982), whereas it differed significantly between young (yearlings) and older females (≥ 2 years after hatching) in Blue Tits (mean1yah = 13.97 ± 0.57 (SE) cm, mean≥2yah = 11.72 ± 1.05 (SE) cm, df = 25, t = 2.064, p = 0. 0496).

In Great Tits, the smallest nests were 4.5 times shallower than the largest ones in terms of nest depth. The maximum nest depth found in Great Tits was 18.0 cm (2277 cm3) and minimum 4.0 cm (506 cm3) in the urban parkland, while it was 15.0 cm (1898 cm3) and minimum 4.5 cm (569 cm3), respectively, in the forest site (Fig. 1). The mean nest volume was 1310 ± 327 cm3 (SD) in the urban parkland and 1270 ± 281 cm3 (SD) in the forest. All the above averages were calculated irrespective of year.

Fig. 1
figure 1

Year-to-year variation in the mean nest depth (cm) in the parkland and forest study areas for Great Tits. Mean nest depth ± standard errors are given

Full size image

In Blue Tits the smallest nests were 3.5 times shallower than the largest ones in terms of depth. The maximum nest depth in Blue Tits was 21.5 cm (2720 cm3) and minimum 6.0 cm (759 cm3) in the urban parkland [mean 1746 ± 365 (SD) cm3] and in the forest maximum nest depth was 19.0 cm (2403 cm3) and minimum 6.5 cm (822 cm3) [mean 1667 ± 350 (SD) cm3] (Fig. 2).

Fig. 2
figure 2

Year-to-year variation in the mean nest depth (cm) in the parkland and forest study areas for Blue Tits. Mean nest depth ± standard errors are given

Full size image

In Great Tits, the lightest nests were 22 times lighter than the heaviest ones in weight. Maximum nest mass in Great Tits was 94.7 g and the minimum 5.2 g in the urban parkland, and it was 115.4 g and 9.9 g, respectively, in the forest (Fig. 3). In Blue Tits the lightest nests were 9.5 times lighter than the heaviest ones in weight. The maximum nest mass in Blue Tits was 97.3 g and the minimum 10.2 g in the urban parkland, whereas it was 86.8 g and 18.0 g, respectively, in the forest (Fig. 4).

Fig. 3
figure 3

Year-to-year variation in the mean nest mass (g) in the parkland and forest study areas for Great Tits. Mean nest depth ± standard errors are given

Full size image
Fig. 4
figure 4

Year-to-year variation in the mean nest mass (g) in the parkland and forest study areas for Blue Tits. Mean nest depth ± standard errors are given

Full size image

The depth of the nest in the general inter-species model was affected by the site and species (Blue Tits tend to build a deeper nest), with a significant interaction between the year and the study site (Table 2). In the first sub-model for Great Tits the nest depth differed between years and did not differed between sites (Table 2, Fig. 1), while in the sub-model for Blue Tits, there was a significant interaction between year and site (Table 2, Fig. 2).

Table 2 Summary of linear mixed models (main model and sub-models) of nest depth in relation to species, site and year (factors)
Full size table

A general inter-species model showed that the nest mass was affected by site and species factors (Blue Tits tend to build heavier nest) (Table 3). In the sub-model for Great Tits the nest mass differed between study sites (being heavier in the forest), but did not differ between years (Table 3, Fig. 3). In the second sub-model considering Blue Tits the nest mass did not differ between either study areas or years. In general, the nest depth was positively correlated with nest mass in Great Tits (r = 0.71, n = 321, p < 0.001) and Blue Tits (r = 0.66, n = 183, p < 0.001).

Table 3 Summary of linear mixed models (main model and sub-models) of nest weight in relation to species, site and year (factors)
Full size table

The nest mass from the same nestbox tended to be consistently similar in Great and Blue Tits at the within-species level (for Great Tits, repeatability: R = 0.21 ± 0.08 (SE), F76; 125 = 1.69, p = 0.005, for Blue Tits, repeatability: R = 0.26 ± 0.12 (SE), F39; 57 = 1.83, p = 0.018), while the nest depth did not tend to be similar in the same nestbox in either species (for Great Tits, repeatability: R = 0.12 ± 0.08 (SE), F76; 125 = 1.35, p = 0.068, for Blue Tits, repeatability: R = 0.01 ± 0.12 (SE), F39; 57 = 0.99, p = 0.511).

Discussion

We showed that the nest depth differed significantly between young (yearlings) and older females in Blue Tits and that younger Blue Tits tended to build deeper nests. Nest size and nest mass were very variable in both tit species. Blue Tits tended to build deeper and heavier nests than Great Tits. The nest depth was positively correlated with the nest mass in Great and Blue Tits. In Blue Tits an interaction between year and study site affected nest depth. By comparison, in Great Tits the nest mass (but not nest depth) differed between study sites, with nests being heavier in the forest, while the nest depth differed between years. The nest mass from the same nestbox tended to be consistently similar in Great and Blue Tits and the nest depth did not show such a tendency. In general, we did not find any obvious linear patterns through the years—and this might have been expected if for example climate change was a factor.

In general, a large amount of variation in tit nest characteristics is hypothesized to have an adaptive basis (Heenan 2013), but there is no complete set of factors that fully explain such diversity and some of the effects shown are difficult to interpret and not conclusive. For example, experimental provisioning with supplementary food during nest building shortened the time of nest building and led to construction of shallower nests in Blue Tits (Smith et al. 2013), but caused that females constructed larger nests in another study (Mainwaring and Hartley 2009). Also predation in a particular nesting season may cause a reduction in nest height in the next breeding season (Kaliński et al. 2014). O’Neill et al. (2018) showed in Blue Tits that while individual females tend to build rather similar nests across years and there was no correlation between female nest size and nest size build by their genetic or cross-fostered mother. The low heritability of the nest parameters in Blue Tits was also shown by Järvinen et al. (2017b), and the high repeatability of particular females’ nest sizes across years was shown by Sonnenberg et al. (2020). On the other hand, Slagsvold et al. (2013) suggested that social learning and coping may be partly responsible for nest parameters. Some studies suggest that birds may build larger nest if their first nest did not successfully produced fledglings (Edwards et al. 2020). Chapman et al. (2022) showed their preliminary results (from Scotland) considering Blue Tit nests (86 nestboxes, between 2016 and 2022) that are similar to our findings on Blue Tits (a poster during the 9th International Hole-Nesting Birds Conference). In their study, naïve (younger) females tended to build heavier nests (adding more structural layer: moss and grass) than more experienced (older) females, and older females added more insulation than young females (insulation layer was found in many tit studies to be positively correlated with breeding success (Glądalski et al. 2016)). Chapman et al. (2022) and our present results (from Scotland and Poland) support previous studies that nest characteristics may be flexible to some extent and may have a learned component (at least in that context for Blue Tits). But Lambrechts et al. (2012, 2016a, 2017) conducting their long-term studies in warm and temperature-stable areas of Mediterranean areas of southern France did not find associations between nest size and female age (yearling versus older females); therefore, it is possible that ambient temperature (or climate) of the area may affect female building behaviour. As suggested by Chapman et al. (2022), the next question would be how previous reproductive success of a Blue Tit impacts nest design. The nest heights in Great Tits did not differ between older and younger females, and it may be a result of the difference between clutch sizes of both tit species. Blue Tits as smaller (and lighter) species but nonetheless having larger clutch (and larger number of nestlings in the nest) than Great Tits may be more thermally sensitive and/or predator-sensitive. More experimental work will be required to investigate this effect in both tit species.

The size of the nestbox may also affect the size and the composition of the nest (Deeming et al. 2019). Lambrechts et al. (2016b) suggested that the vast majority of nestbox studies usually use small nest chambers (Schwegler design) that impose physical constraints on the full expression of the nest. Our nestboxes are larger than used by the Lambrechts’ team, and the nests of our tits are on average 2 × larger (603 cm3 vs 1300 cm3 for Great Tits and 695 cm3 vs 1700 cm3 for Blue Tits) compared to the study by Lambrechts et al.’s (2014) study. In Lambrechts et al. (2017), four nestbox types differing in chamber size were placed along an urbanization gradient. Different measures of nest size (nest volume, nest depth and nest mass) were larger/heavier in the larger box types, whatever the habitat type (street versus park). The authors suggest that there was always enough material other than moss (e.g. wood sticks, pine needles, grass, straw, roots) to fill up the nest chambers. Therefore, it may be important to conduct research studies using more than one type of nestboxes during the whole study. On the other hand, our present study used only one nestbox type. Perhaps species or environmental effects on nest size might change with nestbox size, accepting that it is more costly to build larger than smaller nests.

Our long-term data confirm that Blue Tit nests are heavier and deeper than Great Tits nests. Lambrechts et al. (2014) suggest that Great Tits have smaller nest because parents may require longer safety distance (so it may be anti-predator behavioural adaptation). This suggestion may be supported by Kaliński et al. (2014), who showed that the difference in average height of fresh nests between Blue and Great Tits became significantly smaller when a tube was added that elongated the entrance hole (after a season with a very high rate of nest predation by European Pine Marten Martes martes). Simultaneously, the larger nest in Blue Tits may be more optimal for larger clutches in this species compared to Great Tits (with smaller clutch and shallower nest). But if there was such a relation, Blue Tits with smaller broods should have smaller nests and Great Tits with larger clutch should have larger nests, but Glądalski et al. (2016) found no relation of that kind. Still more experimental work will be needed to investigate why species-specific effects on the height of the nest exist.

In our study, the mass of the Great Tit nest (but not nest depth) differed between study sites, being heavier in the forest. Variation in nest parameters between both study sites could have resulted from the use of different materials and/or by the use of different proportions of the same materials by tits (Britt and Deeming 2011). Álvarez et al. (2013) conducted a comparative study on Great Tits in four different Mediterranean habitats and showed differences in nest mass and moss proportions. The authors explain some differences between nest characteristics by difficulties in obtaining moss in Orange Citrus aurantium plantations. Exploring species diversity of moss in Great and Blue Tits nests, Glądalski et al. (2021) showed that the nests (parkland vs forest) were similar, but with more bryophyte species in the nests from the forest study area than from the parkland study area. But given that there is a difference in weight between the nests in the Great Tit (parkland vs forest) and not between the Blue Tit nests (parkland vs forest), the species composition of the nests should not affect the nest weight of Great Tits. On the other hand, Harnist et al. (2020) showed that nest mass changes in relation to nesting stages, so that fresh nest (at the end of nest building phase) is lighter than the nest weighed at the post-fledging stage (the reason are probably brood size-dependent remnants like epidermis leftovers). But also in this case there were no differences in Blue Tit nest mass (parkland vs forest), which suggests that post-fledging remnants do not explain the differences. It seems that there are two options. Firstly, females during nest-building may compact the moss to form a dense structure (it may also be related to predation and larger safety distance in the forest), or secondly, tits in the forest may use a larger amount of small twigs at the bottom of the nest, and that sometimes happens in Great Tit nests (own observations). More studies will be required to investigate this effect in Great Tits.

We showed that the nest mass from the same nestbox tended to be consistently similar in Great and Blue Tits and the nest depth did not tended to be similar in the same nestbox in both tit species. Sonnenberg et al. (2020) showed that the nest depth is characterised by high repeatability in tit females. Our data (including Glądalski et al. 2021) suggest that breeding territory does not affect the depth of the nest, although the composition of bryophyte species of the nest is affected by the site (parkland vs forest). This may suggest that depth is rather dependent on individual bird preferences, but the general mass of the nest may be partly related to territory. Consistency in traits of nests constructed in particular nestboxes may result from special properties of the microhabitat in which a nestbox is located differing from properties of other microhabitats. Settlement in particular nestboxes is probably non-random with respect to both microhabitat properties and individual properties of the established birds (age-, experience-, quality-dependent) that may be reflected in nest characteristics. In this context, a combination of some subtle factors that vary between nestbox microhabitats may produce a detectable level of consistency in nest characteristics. In other words individual birds may use similar mass of the components during nest building on a breeding territory, but depending on individual predispositions in relation to nest depth a female may build a nest that is more or less compact/dense.

In conclusion we showed that the large amount of variation in nest parameters between and within populations of both tit species may be a result of many factors like female age, breeding area, breeding season, individual predispositions or the size of the nestbox/natural hole.