Rapid environmental change driven by urbanisation offers a unique insight into the adaptive potential of wildlife as it can induce distinct selective pressures on urban dwelling organisms. Despite mounting evidence for urban-driven phenotypic differentiation across taxa, knowledge of the impact of urbanisation on vertebrate developmental rates and subsequent survival is very limited. Importantly, the role of selection on urban-driven body mass divergence in juvenile organisms remains poorly understood. We studied nestling development in a gradient of urbanisation set in Warsaw, Poland, in two nestbox breeding passerine species: great tits (Parus major) and blue tits (Cyanistes caeruleus). Each nestbox in the study system was assessed for surrounding percentage of Impervious Surface Area (ISA). Within these nestboxes, weight measurements of individual nestlings were collected during three breeding seasons and at regular intervals after hatching. Although asymptotic mass and growth rate were not directly affected by ISA in a subset of frequently measured nestlings, the age of fastest growth (inflection point) was delayed in blue tit nests surrounded by greater ISA. Across the entire dataset, nestling body mass was negatively affected by increasing ISA at days 5 and 10 after hatching for great tits, and at days 10 and 15 for blue tits, respectively. Concomitantly, offspring survival at days 5 and 10 decreased with increasing ISA for both species. An analysis of selection differentials performed for two contrasting levels of imperviousness (low and high ISA) revealed a significant positive association between mass at day 2 and survival at fledging. Importantly, the strength of selection for heavier nestlings at hatching was greater for great tits in the more transformed, high ISA environment. This study confirms the considerable impact of imperviousness -a proxy for urbanisation level- on offspring development, body mass and survival, and highlights increased selection on avian birth weight in a high ISA environment.

MATERIALS AND METHODS

2.1  Study sites

Data was collected for three field seasons between 2016 and 2018 in a gradient of urbanisation in the city of Warsaw, Poland. Five hundred Schwegler woodcrete nestboxes (type 1b with 32 mm entrance hole, suitable for great tits and blue tits) were erected in a 50 m grid in eight contrasted study sites representative of the urban mosaic: six were located within the city borders while two were exurban sites (Figure 1). The total number of nestboxes within each area varied from 21 to 110. While the monitoring of three sites (B, E and H) had already started in 2016, these and all other sites were monitored in 2017 and 2018.

2.2  Quantifying urbanisation

Urbanisation was quantified in a 100m radius around each nestbox in the study. This corresponds to literature-based estimates of parental foraging while chick feeding, assessed in blue tits to average 53.2 meters (±22.9 SD) in food poor (but natural) environments (Tremblay et al., 2004). In such food poor environments, birds were also reported to fly beyond 50 m from the nest in c. one-third of all foraging trips (Tremblay et al., 2004). An estimation of urbanisation in a 100m radius around each nestbox thus corresponds to a conservative estimate of the range of food foraging distance covered by parents of offspring developing in the nest. Within this radius, we estimated the proportion of Impervious Surface Area (ISA) in QGIS following Szulkin et al. (2020). Specifically, a 20 m-pixel resolution of ISA extrapolated via satellite imagery from 2015 (Copernicus Land Monitoring Services, https://land.copernicus.eu/sitemap) was used to define ISA around each nestbox. Such index, expressed as a percentage, included all built-up areas that replaced original natural cover or water surfaces with an artificial and usually impervious surface. These artificial surfaces include built-up areas (such as infrastructural networks and buildings) and other elements characterised by a long cover duration. For further details on the imperviousness index description, see https://land.copernicus.eu/sitemap.

2.3  Life-history data collection and nestling measurements

Starting from mid-March, nestboxes were inspected weekly to record the date of the first egg laid and clutch size. Hatching date was determined by visiting the nest one day before the expected hatching (12 days after the last egg of the clutch was laid) and around hatching date. Both laying and hatching dates were coded by setting the 1^st of April as day 1. Only first broods, defined as broods that started no later than 30 days after the very first brood in a given year and site (Van Balen, 1973), were included in the analyses. Nestlings were uniquely marked by toenails clipping or by using waterproof markers on their first measurement day. In 2016, nestlings were individually weighed every 2 or 3 days from hatching (day 1) until ringing (day 15 or exceptionally, day 14 or 16, when the brood could not be accessed on day 15). In 2017 and 2018, nestlings were weighed specifically on days 2, 5, 10 and 15. Mass was recorded to the nearest 0.1 g using digital scales (KERN pocket balance CM 150-1N). At each nestbox visit, individual survival status (survived = 1, dead = 0) and brood size (number of chicks alive in the nest) were also recorded.

In 2016, up to 50 µl blood samples were collected by puncturing the brachial vein of each 15 days old nestling. Blood samples were subsequently stored in 99.0% ethanol at +4° C until molecular sexing analysis. Finally, nestboxes were checked c. 25 days after hatching to determine fledging success for each individual nestling: individuals found dead in the nest and individuals that were not present in the nestbox had a fledging success of 0 and 1, respectively.

2.4  Extrapolating data on temperature at the nestbox level

Weather data over the three-year period were provided by the Polish Institute of Meteorology and Water Management (IMGW-PIB). Average daily temperature data were computed from Warsaw Okecie and Legionowo weather stations; the nearest sampling points for the study locations situated within and outside the city borders, respectively.

To estimate temperature experienced by growing nestlings at the nestbox level, average temperature was calculated at each nestbox across the specific periods of interest: for growth curve analyses, we used averaged temperature from day 1 to day 15 of nestling growth. For body mass and survival analyses, we used averaged temperature from day 1 to day 2, from day 2 to day 5, from day 5 to day 10, from day 10 to day 15, which corresponds to the intervals between mass measurements. Finally, for fledging success, we used averaged temperature from day 15 to day 25.

Data used for this publication include two main files:

1) GrowthParam2016_EvoApp2019_FlexParamCurve: contains the data used to test the effect of ISA on great tits and blue tits growth curve parameters. Statistical methods are described in detail in Section 2.6.1.

2) Survival_EvoApp2016_2017_2018: contains the data used to test the effect of ISA on great tit and blue tit nestling body mass (Section 2.6.2), survival (Section 2.6.3), and selection differentials (Section 2.6.4). 

Additional files were used to generate figures in R (ggplot):

3) CurvePlot_EvoApp_Corsini2020: data used to generate Figure 2.

4) selection_differential_coefficients_BT and selection_differential_coefficients_GT: combined data used to generate Figure 3.

Legend and specific information on each dataset are listed below:

dataset: GrowthParam2016_EvoApp2019_FlexParamCurve:

Note: only one year of data (2016 - high resoluted measurements)

dataset: Survival_EvoApp2016_2017_2018:

Sheet: Nestlings_EvoApp2019

Sheet: Broods_EvoApp2019

Note: empty cells are missing data

Daily temperature (in C) (from 2 meteorological stations: Warsaw and Legionowo. Legionowo data used for KPN and PAL).

Station: Climatic station (Waw = within Warsaw city borders, Leg = outside Warsaw city borders, KPN and PAL)

dataset: CurvePlot_EvoApp_Corsini2020

Sheet: Data

Note: the plot shows growth inferred at the brood level for 1 year of data (2016) and across 3 study sites (POL, CMZ and KPN).