Show more...
Abstract:
This archive contains two tables as well as a document with the detailed methods associated with the project:
“Individual quality overwrites carry-over effects across the annual cycle of a long-distance migrant”
Authors: Léandri-Breton, D.-J., K H. Elliott, A. Tarroux, B. Moe, W. Jouanneau, F. Amélineau, F. Angelier, P. Blévin, V. S. Bråthen, P. Fauchald, G. W. Gabrielsen, A. Goutte, C. Parenteau, S. Tartu, P. Legagneux, O. Chastel.
1) “tracking_immersion_data.csv” containing all positions and saltwater immersion collected from light-level geolocators and used to estimate the daily activity budget and daily energy expenditure of wintering black-legged kittiwakes (Rissa tridactyla).
Geolocation and immersion data: The studied population is a long-term monitored colony of individually-marked black-legged kittiwakes situated in Kongsfjorden, Svalbard (High Arctic Norway; 78° 54’ N, 12° 12’ E). From 2008 to 2020, we deployed light-level geolocators on adults captured at their nest. Devices were mounted on leg bands and measured light intensity every minute and recorded the maximum ambient light intensity every 5 or 10 min. They also measured saltwater immersion (i.e., conductivity; whether the logger was in contact with the seawater) every 3 or 30 s and stored the number of wet measurements within every 10 min period. The saltwater immersion data were standardized to be comparable among individuals and logger models, such as x st = x / x max, where x max is the maximum value over the 10-min summing period. Overall, and after filtering out incomplete tracks or tracks missing saltwater immersion data, we acquired 277 tracks from 181 different individuals covering 13 non-breeding seasons (fall 2008 to spring 2021).
Geolocation light data processing: To infer geographic positions, geolocator light data were processed according to the procedure based on the threshold method to estimate twilight events and developed for the SEATRACK international initiative (Bråthen et al., 2021). Geolocation-based-tracking is inherently prone to create lower accuracy locations in latitudes around equinox periods and cannot be used over periods of continuous daylight (or continuous night) at high latitudes. To fill these gaps in the tracks, and to reduce biases along the trajectories, missing locations were re-estimated by interpolation between known locations using an algorithm specifically developed for SEATRACK (Fauchald et al., 2019) and applied in kittiwakes (Léandri-Breton et al., 2021).
Variable descriptions: “track_id”: unique track identifier; “ring”: unique individual identifier; “timestamp”: date and time (GMT; format %Y-%m-%d %H:%M:%S); “lon”: longitude; “lat”: latitude; “conductivity”: conductivity (saltwater immersion); “std_conductivity”: standardized conductivity.
2) “carryover_effects_variables.csv” dataset containing all variables used for fitting piecewise structural equation models (PSEM) to assess direct and indirect causal effects between consecutive breeding stages in black-legged kittiwakes. These include variables relative to breeding success and phenology, migration phenology, wintering area and energetics and baseline corticosterone.
Breeding monitoring and experimental design: We monitored nest contents from 2008 to 2021 every two to six days to determine individual breeding success and phenology. We used the success of rearing at least one chick for 10 days after hatching (hereafter ‘breeding success’) as a proxy of the annual reproductive success and expressed as a binary variable (failure = 0, success = 1). During summers 2019 and 2020, we conducted a clutch removal manipulation to relieve parents from further reproductive allocation investment and assess potential carry-over effects while controlling experimentally for individual quality. The full clutch (one or two eggs) was removed to provoke the reproductive failure of 14 nests in 2019 and 12 nests in 2020. Expressed as a binary variable (experimental failure = 0, control/unmanipulated = 1).
Migration phenology and winter distribution: The timing of departure from the colony area in fall and return to the colony area in spring were estimated using Lavielle partitioning algorithm (see Léandri-Breton et al., 2021), over a 5-day running maximum of the saltwater immersion data indicating a behavioural transition between land use (dry) and continuously pelagic behaviour (wet). The wintering stage was defined for each track as the last date in fall and the first date in spring the individual crossed the 60°N, corresponding to the northernmost extent of the winter distribution (Léandri-Breton et al., 2021). The centre of the population’s winter range was defined as the centroid of the 95% utilization distribution kernel (UD), estimated over the population’s winter positions (from all tracks, n = 38 900 positions) and projected using a Lambert Azimuthal Equal Area coordinate system (Fig. 1, h=200 km, 50x50 km grid cells). Similarly, the centroid of each individual’s track was calculated from a 95% UD over the individual’s winter position. The distance from the population’s range was defined as the orthodromic distance between the population’s winter centroid and the centroid of each individual’s winter track.
Daily energy expenditure: We used the standardized saltwater immersion data (SSI) to build a time-activity budget based on the identification of three behaviours for each 10-min periods: sustained flying (SSI = 0%), resting on water (SSI ≥ 98%) and active foraging (0% < SSI < 98%) which is characterized by a succession of short flights and short swimming or shallow diving bouts (Jodice et al., 2003). To estimate the daily energy expenditure during the wintering stage, we combined the time-activity budget with the population-specific resting metabolic rate (Gabrielsen et al., 1988) and the activity-specific field metabolic rates for the kittiwake flying and foraging behaviours (Jodice et al., 2003).
Blood sampling and corticosterone: From 2019 to 2021, 92 kittiwakes equipped with a geolocator were recaptured during the pre-laying stage and sampled within 3 min (average sampling time: 129±29 sec, range: 72-180 sec) for 2 mL of blood from the brachial vein. Total baseline corticosterone concentrations (bound and free fractions) were measured in plasma via radioimmunoassay following the method validated by Lormée et al., (2003).
Variable descriptions: “id”: unique individual identifier; “track_id”: unique track identifier; “yr_dep”: start year of the annual track; “yr_ret”: end year of the annual track; “treatment.y1”: experimental treatment applied to the clutch in year_ x_ (experimental failure = 0, unmanipulated = 1); “treatment.y2”: experimental treatment applied to the clutch in year x + 1 (experimental failure = 0, unmanipulated = 1); “breed.success.y1”: proxy of breeding success in year x (0: failure; 1: success); “col.depart.y1”: date of departure from the colony area (GMT; format %Y-%m-%d %H:%M:%S); “dist.centro”: distance (in km) between the centroid of the population wintering range and the centroid of the individual wintering home range; “DEE.mean”: averaged daily energy expenditure (kJ/day); “col.arrival.y2”: date of arrival in the colony area in year x + 1 (GMT; format %Y-%m-%d %H:%M:%S); “lay.date.y2”: date of first egg laying in year x+ 1 (GMT; format %Y-%m-%d %H:%M:%S); “cort.y2”: corticosterone level in ng/mL measured in plasma in year x + 1; “breed.success.y2”: proxy of breeding success in year x + 1 (0: failure; 1: success).
3) “detailed_methods.doc” document containing all detailed methodology on data collection and analysis.