Most birds incubate their eggs to allow embryo development. This behaviour limits the ability of adults to perform other activities. Hence, incubating adults trade-off incubation and nest protection with foraging to meet their own needs. Parents can either cooperate to sustain this trade-off or incubate alone. The main cause of reproductive failure at this reproductive stage is predation and adults reduce this risk by keeping the nest location secret. Arctic sandpipers are interesting biological models to investigate parental care evolution as they may use several parental care strategies. The three main incubation strategies include both parents sharing incubation duties (“biparental”), one parent incubating alone (“uniparental”), or a flexible strategy with both uniparental and biparental incubation within a population (“mixed”). By monitoring the incubation behaviour in 714 nests of seven sandpiper species across 12 arctic sites, we studied the relationship between incubation strategy and nest predation. First, we described how the frequency of incubation recesses (NR), their mean duration (MDR), and the daily total duration of recesses (TDR) vary among strategies. Then, we examined how the relationship between the daily predation rate and these components of incubation behaviour varies across strategies using two complementary survival analysis. For uniparental and biparental species, the daily predation rate increased with the daily total duration of recesses and with the mean duration of recesses. In contrast, daily predation rate increased with the daily number of recesses for biparental species only. These patterns may be attributed to two independent mechanisms: cryptic incubating adults are more difficult to locate than unattended nests and adults departing the nest or feeding close to the nest can draw predators' attention. Our results demonstrate that incubation behaviour as mediated by incubation strategy has important consequences for sandpipers’ reproductive success.

Study sites

The study was conducted at 12 sites across the Arctic (Figure 1) during the summers of 2016 to 2018. Field sessions began in June at the southernmost sites and early July at the high-arctic sites.

Nest monitoring

At each site, we monitored the incubation behaviour of one to three species of sandpipers (seven species in total). Three species are uniparental (little stint Calidris minuta, Temminck’s stint Calidris temminckii and white-rumped sandpiper Calidris fuscicollis), three species are biparental (dunlin Calidris alpina, Baird’s sandpiper Calidris bairdii and semipalmated sandpiper Calidris pusilla), and one species, the sanderling (Calidris alba), exhibits a mixed strategy with nests incubated by either two or only one adult in the same population (Reneerkens et al. 2011, Moreau et al. 2018). All species lay a typical clutch of four eggs (rarely three or five) in a shallow nest scrape directly on the tundra’s surface (Reid et al. 2002).

Nests were located opportunistically by walking through suitable breeding habitats and flushing incubating birds, or by following birds back to their nests. Nests found with incomplete clutches were visited repeatedly during the following days to determine the exact date of incubation start. For complete clutches, the relative age of the nest (in days since initiation of incubation) was estimated (with a precision of 1-3 days) by floating one to three eggs of the clutch and using flotation curves (Mabee et al. 2006, Liebezeit et al. 2007). Expected hatch dates were then inferred from age estimates and average duration of incubation known for each species (Liebezeit et al. 2007).

Incubation behaviour and nest fate

In each nest, we placed a temperature probe (Flylead thermistor PB 5009 with 60cm cable) coupled to a data logger (Tinytag Plus2 TGP-4020; Gemini Data Loggers Inc., West Sussex, U.K.) to continuously record the nest temperature. This design is widely used to monitor shorebirds’ incubation patterns (Tulp and Schekkerman 2006, Smtih et al. 2012) as it discriminates between periods of incubation and incubation recesses. The temperature probes were fixed to wooden sticks and anchored into the ground in the centre of the clutch, with the top of the probe levelled with the top of the eggs so as to be in continuous contact with the brood patch of the adults during incubation. Data loggers recorded nest temperature (accuracy of measurements: 0.2°C; temperature range: -40 to +125°C) every minute during the full incubation period, lasting for ca. three weeks (data storage capacity: 22.2 days). Data loggers and the wires connecting it to the probe were buried or concealed using vegetation and substrate to avoid visual attraction of predators. Device deployment took approximately 5-10 minutes and all efforts were made to avoid leaving scent at the nest site to prevent attracting mammalian predators.

Data loggers were retrieved after the expected hatch date (unless the nest was still active and then nests were visited again before the end of the fieldwork session) and temperature records were extracted using the software Tinytag Explorer 5.0 (Gemini Data Loggers Inc., West Sussex, U.K.). The fate of each nest (depredated vs. hatched) was visually inferred on a temperature plot by two independent observers (OG and NM) according to the temperature pattern recorded during the last 24h of recorded incubation. A nest was considered depredated if the temperature suddenly dropped and permanently stayed at environmental temperatures (usually before the expected hatch date, see Supplementary material Appendix 1 Figure A1, Weidinger 2006) . A nest was considered successful if the temperature was steadily declining for 24h (±12h) within 2 days of the expected day of hatching (Tulp and Schekkerman 2006). In a preliminary step, we used field evidence of nests’ fate (e.g. presence of pipped eggs, hatched chicks or small eggshell fragments typical of hatched eggs in the bottom of the nest cup; Mabee 1997, Mabee et al. 2006) to validate that the method of fate determination based on the temperature record was trustworthy, but fate assignments as used in our analyses were only inferred from temperature records.

To describe incubation behaviour, we used the temperatures recorded shortly after the incubating adult had returned to the nest, after thermologgers were first deployed, until either the predation event or the beginning of the hatching event. At our arctic study sites, environmental temperature is always lower than the temperature of incubation (ca. 41°C), resulting in a drop in measured temperatures when the adult leaves the nest. We considered a recess (i.e., a period when eggs are not incubated) to start when the temperature dropped by ≥3°C below the median incubation temperature of a nest (measured over 24h periods) and to end when the temperature returned above this threshold (see Figure 1 in Moreau et al. 2018). Hence, all temperature profiles shorter than 24 hours were excluded from the analyses. For each nest and each day of monitoring, we calculated three components of incubation behaviour: the total duration of recesses (TDR), the number of recesses (NR) and the mean duration of recesses (MDR; equal to TDR/NR). Note that these components describe individual incubation behaviour for uniparental species only. For biparental species, they combine recesses from both incubating adults during their respective incubation bouts.

Data analysis

Temperature measurements were obtained from 714 nests across all seven Calidris species monitored on the 12 study sites (Table A1, Figure A2). We removed 104 nests when the thermologgers failed to record temperatures for at least 24 hours (e.g. due to technical malfunctioning). Furthermore, in some of the 610 remaining nests, the temperature probe had moved during the monitoring period (for instance because the nest was built in too soft substrate) and temperature profiles had a decreasing trend, which limited our ability to detect some recesses. To prevent these unreliable records from affecting our analyses, all days when thermologgers recorded a daily median temperature below 36°C (n = 68 nests) were also removed. This approach led to a filtered data set of 542 nests with exploitable nest temperature data.

We described the incubation behaviour of adults by the daily TDR, NR and MDR by averaging values over the entire monitoring period for each nest. To ensure that the averaged behaviour was representative, we only kept nests with a daily median temperature over 36°C for at least 20% of the monitoring period; this approach resulted in the removal of 16 additional nests from the remaining 542 nests. Moreover, an additional 17 nests were excluded from analyses carried on their fates (hatched vs. depredated), as 9 nests were abandoned after the beginning of the monitoring and 8 had unclear fate (i.e. the two observers disagreed on the nest’s fate). Then, only data sub-sets with more than one suitable recording per species/site/year combination were kept. As the following models (see next paragraphs) can handle right censoring, nests with unknown fates were kept in our models. This approach resulted in a dataset of 505 nests (with 208 hatched and 229 depredated; see second column in Table A1 and Figure A2).

References

Liebezeit, J. R. et al. 2007. Assessing the development of shorebird eggs using the flotation method: species-specific and generalized regression models. - The Condor 109: 32–47.

Mabee, T. J. 1997. Using Eggshell Evidence to Determine Nest Fate of Shorebirds. - Wilson Bull. 109: 307–313.

Mabee, T. J. et al. 2006. Using egg flotation and eggshell evidence to determine age and fate of Arctic shorebird nests. - J. Field Ornithol. 77: 163–172.

Moreau, J. et al. 2018. Discriminating uniparental and biparental breeding strategies by monitoring nest temperature. - Ibis 160: 13–22.

Reid, J. M. et al. 2002. Nest scrape design and clutch heat loss in Pectoral Sandpipers (Calidris melanotos). - Funct. Ecol. 16: 305–312.

Reneerkens, J. et al. 2011. Do Uniparental Sanderlings Calidris alba Increase Egg Heat Input to Compensate for Low Nest Attentiveness? - PLOS ONE 6: e16834.

Smtih, P. A. et al. 2012. Shorebird incubation behaviour and its influence on the risk of nest predation. - Anim. Behav. 84: 835–842.

Tulp, I. and Schekkerman, H. 2006. Time allocation between feeding and incubation in uniparental arctic-breeding shorebirds: energy reserves provide leeway in a tight schedule. - J. Avian Biol. 37: 207–218.

Weidinger, K. 2006. Validating the use of temperature data loggers to measure survival of songbird nests. - J. Field Ornithol. 77: 357–364.

 

This dataset contains the data about the incuabtion behvaiour. 

•    The nest_ID is a unique column that identifies each nest;
•    the species is written with acronymes (BASA: Baird's sandpiper; DUNL: dunlin; LIST: little stint; TEST: Temminck's stint; WRSA: white-rumped sandpiper; SESA: semipalmated sandpiper; SAND: sanderling);
•    the breeding_site corresponds to the field site with the same acronyms as in Figure 1;
•    the year corresponds to the year of the field season;
•    the pop column identify a species for a given year at a given site;
•    the strat gives the incubation strategy of the species (BI: biparental, UNI: uniparental, MIXED: mixed strategy);
•    the mean_TDR column gives the avegraged daily TDR for each nest; 
•    the mean_NR column gives the avegraged daily NR for each nest; 
•    the mean_NR column gives the avegraged daily MDR for each nest;
•    the Fate provides the information about the fate of the nest with 1 for predated nests and 0 for others (hatched or censored).

For any additional information, please, conatct us.