OPEN 3 ACCESS Freely available online PLos one

Do Uniparental Sanderlings Calidris alba Increase Egg Heat Input to Compensate for Low Nest Attentiveness?Jeroen Reneerkens1*, Kirsten Grond1, Hans Schekkerman2, Ingrid Tulp3, Theunis Piersma1'41 Centre for Ecological and Evolutionary Studies, Animal Ecology Group, University of Groningen, Groningen, The Netherlands, 2 SOVON, Dutch Centre for Field Ornithology, Nijmegen, The Netherlands, 3 Institute for Marine Resources and Ecosystem Studies, IJmuiden, The Netherlands, 4 Department of Marine Ecology, Royal Netherlands Institute for Sea Research, Texel, The Netherlands

AbstractBirds breeding in cold environments regularly have to Interrupt Incubation to forage, causing a trade-off between two mutually exclusive behaviours. Earlier studies showed that uniparental Arctic sandpipers overall spend less tim e incubating the ir eggs than blparental species, but Interspecific differences In size and ecology were potential confounding factors. This study reports on a within-species comparison o f breeding schedules and metal egg temperatures in uni- and biparental sanderlings (Calidris alba) in Northeast Greenland In relation to ambient temperature. We recorded Incubation schedules w ith nest temperature loggers in 34 sanderling clutches (13 uniparentals, 21 biparentals). The temperature o f a metal egg placed w ith in the clutch o f 17 incubating birds (6 uniparentals, 9 biparentals) was measured as an indicator o f the heat put Into eggs. Recess frequency, recess duration and total recess tim e were higher In uniparentals than in biparentals and positively correlated w ith ambient temperatures in uniparentals only. Uniparental sanderlings maintained significantly higher metal egg temperatures during Incubation than biparentals (1.4°C difference on average). Our results suggest that uniparental sanderlings compensate for the lower nest attendance, which may prolong the duration o f the incubation period and negatively affect the condition o f the hatchlings, by maintaining a higher heat flux Into the eggs.

C ita t io n : Reneerkens J, Grond K, Schekkerman H, Tulp I, Piersma T (2011) Do Uniparental Sanderlings Calidris alba Increase Egg Heat Input to Compensate for Low Nest Attentiveness? PLoS ONE 6(2): e16834. doi:10.1371/journal.pone.0016834

E d ito r : Kevin Mcgraw, Arizona State University, United States of America

R e c e iv e d September 8, 2010; A c c e p te d January 16, 2011; P u b lis h e d February 9, 2011

C o p y r ig h t : © 2011 Reneerkens et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

F u n d in g : This work was supported by an International Polar Year grant (851.40.072) by the Netherlands Organization for Scientific Research (NWO) to T. Piersma and J. Reneerkens, and by a research grant by the Schure-Beijerinck-Popping Fund (http://www.knaw.nl/cfdata/funding/funding_detail.cfm7orgid = 103) to I. Tulp and H. Schekkerman. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

C o m p e tin g I n te r e s t s : The authors have declared that no competing interests exist.

* E-mail: J.W.H.Reneerkens@rug.nl

Introduction

Incubation is an energetically dem anding phase o f avian reproduction [1-3] and this is particularly so in H igh Arctic b reeding birds o f relatively small size that lay their eggs in open nest cups on the ground, often only a few centim eters above the perm afrost [4—5]. A m bient tem peratures in the H igh Arctic are usually far below the m inim um o f 26"C that is requ ired for em bryonic developm ent [6], bu t w hen incubating adults forage they necessarily expose their clutch to am bient tem peratures unless a p a rtn e r takes over. T his results in a conflict betw een the dem ands o f incubating adults an d those o f their offspring [7]. T he trade-off is especially relevant in uniparentally incubating adults that can not delegate incubation to a partner. As shorebirds are too small to have enough stores to rely on during incubation [8], they need to forage away from the nest to m eet daily energy requirem ents [9-10], W hen spending m uch tim e away from the nest, such birds ru n the risk o f slowing down or even ceasing em bryonic developm ent [11].

In term itten t b reeding schedules cause em bryos to experience varying tem peratures that negatively affect their grow th rates, hatching condition an d hatching success [2,11-13] and lead to prolonged incubation periods w ith associated p redation risk [14— 17]. T h e incubating adult needs to pu t extra energy into reheating the eggs after a foraging bout [2]. Typically, partners that divide the incubation duties am ong themselves (hereafter: biparentals)

experience a reduced cost o f incubation [2] and their clutches will encounter less variation in egg tem perature than clutches th a t are incubated by a single adult only (hereafter: uniparentals).

T h e ra te o f egg cooling will determ ine how long birds can leave their clutch una ttended until the critical tem perature is reached below w hich em bryonic developm ent will be im paired [6], H ow fast eggs cool dow n depends, am ong o ther factors, on clutch size [13,18], the tem perature gradient betw een egg and the environ­m ent [10,19], and the insulative properties o f the lined nest cup [19-23]. All these factors m ay offer opportunities for the incubating bird(s) to lim it the ra te o f egg cooling. U niparen tal incubators m ight also be able to minim ise the extension o f the incubation period by m aintain ing a higher egg tem perature during incubation bouts [16,17]. In this study, we investigated two possible ways for uniparentals to enhance em bryo developm ent:(1) by adjustm ents o f the incubation schedules in relation to am bient tem peratures, and (2) by regulating egg tem perature. Because the tim e that can be spent away from the nest will also depend on the ra te with w hich food can be found, we also m easured daily a rth ropod abundance.

T h e consequences of, and adaptations to, un iparental incuba­tion can best be studied in direct com parison to a b iparental conspecific a t the same location an d tim e. T h e H igh Arctic b reeding sanderling (Calidris alba; Fig. 1) is the only sandpiper know n to exhibit bo th uni- an d b iparenta l incubation w ithin the same breeding population [24,25], bu t see [26], U niparen tal

PLoS ONE I www.plosone.org 1 February 2011 | Volume 6 | Issue 2 | e16834

Egg Heat Input and Nest Attendance in Sanderling

Figure 1. An individually color-ringed sanderling C a lid r is a lb a tries to lure away a researcher from its nest. Note the an tenna loop around the clutch o f four eggs in th e background. The PIT is glued to th e side o f th e ex tended green color-ring th a t is no t visible on this photograph. Photo by Jeroen Reneerkens. doi:10.1371/journal.pone.0016834.g001

incubation in sanderlings is thought to be the result o f m ates o f a pa ir dividing incubation betw een two clutches that are laid in rapid succession (‘double-clutching’) [27]. Because sanderlings have an alm ost invariable clutch size o f four eggs [28], pairs that raise two clutches w ould thus have the potential advantage o f a twice as large annual reproductive output. However, the associated uniparental incubation entails a higher energetic challenge due to constraints in time that can be spent off the nests [5] a n d /o r m ay have consequences for the developm ent and condition o f the offspring due to the in term ittent na ture o f incubation [11]. Given earlier between-species comparisons, we predicted that uniparental sanderlings w ould incubate their clutches less regularly. In this study, we com pare the nest attendance o f uni- and b iparental sanderlings with those o f o ther waders in Arctic, tem perate and tropical climates. Because prolonged incubation duration entails certain fitness costs [14,17], we also predicted that uniparen­tals com pensate a lower nest attendance by an increase in egg tem perature. A within-species com parison o f b reeding schedules and egg tem perature o f uni- and biparental parents m ight give insight into the evolutionary effects o f different ecological conditions on breeding systems o f sandpipers and avian reproduction in general.

Methods

Ethics StatementT h e ethical guidelines p rom oted by the Association for the Study

o f A nim al Behaviour were followed. T he G reenland Ecosystem M onitoring C oordination G roup at the N ational Environm ental R esearch Institute, A arhus Ehiiversity approved our detailed research proposals for our research activities in bo th 2007 and 2008. An exem ption to the ’Executive O rd er no. 7 o f 17 Ju n e 1992 from the M inistry o f Dom estic Affairs, N ature and Environm ent (Greenland H om e Rule Authority) as am ended by Executive O rder no. 16 of 5 O ctober 1999, to catch, (colour-)ring and take blood samples o f birds and to travel into the N ational Park o f N orth and N ortheast G reenland in the sum m ers 2007 and 2008 was obtained

via Zackenberg Ecological R esearch O perations (ZERO) at the National Environm ental R esearch Institute, Aarhus ETniversity.

Study area and birdsSanderling nests were studied betw een 6 Ju n e and 24 Ju ly 2007

and 30 M ay to 27 Ju ly 2008 in an area o f ca. 24 km~ at Zackenberg, N ortheast G reen land (74"30' N 2 T 0 0 ' W) [29]. Nests were located in a valley and on a south-facing m ountain slope up to 550 m a.s.l. [30], In b o th years 18% (six and seven in 2007 and 2008, respectively) o f all clutches were still incom plete w hen we found them . Because all those clutches were preyed upon before they w ould have hatched, we were unable to determ ine incubation periods. T h e nest locations were registered by G PS and m arked with a small plastic m arker o r stone pile ca. 10 m away from the nest. H atch ing date was predicted by egg flotation (± 2 days) [31] and start o f incubation defined as 22 days before estim ated hatch. W e used an incubation length of 22 days based on com m on practice a t Zackenberg. This value appeared to fit our observations well [32]. I f an incom plete clutch was found, we considered incubation to start at clutch com pletion, w hich takes four consecutive days starting on the first egg date. T h e length and w idth o f each egg in a clutch were m easured with calipers to the nearest 0.1 nini. Egg volum e was calculated using the form ula length * w idth2 * 0.52 [33]. C lutch volum e is the sum o f egg volumes for each clutch.

Adult birds were caught on the nest with a small clapnet placed over the nest cup that was triggered by the returning bird. All caught birds were provided with a num bered metal ring on one o f the tibia and plastic colored leg rings on the tarsi which m ade birds indi­vidually recognisable in the field. O ne o f the color-rings was extended (as a ‘flag’). Blood samples were taken for molecular sexing [34],

Identification o f uni- or biparental clutchesIn 2007, we used transponders (PIT) in addition to therm o-

loggers to help identify w hether nests were lini- or biparental. A 10 m m long x 1.1 nin i glass bead P IT was glued w ith plastic glue onto the extended color-ring o f sanderlings. Twenty-six nests were

PLoS ONE I www.plosone.org 2 February 2011 | Volume 6 | Issue 2 | e16834

Egg Heat Input and Nest Attendance in Sanderling

fitted w ith P IT detecting an tenna loops a ttached to a da ta logger to identify tagged individuals incubating them (Fig. 1). T h e loggers were p rogram m ed to register the presence an d identity o f a tagged sanderling on a nest every m inute. Days during w hich nest visits by the researchers resulted in incubating sanderlings tem porarily leaving the clutch were no t included in the analyses. T h e an tenna did no t always detect an incubating bird; the therm ologgers (see next subsection “ D eterm ination of incubation schedules”) regu­larly recorded periods with invariable tem peratures a round 40"C while at the same tim e no recordings were logged by the P IT loggers (Fig. 2). How ever, w hen the P IT s were detected, the therm ologgers always m easured tem peratures exceeding 36"C, so false positive identifications did no t occur (Fig. 2). Because PIT s were not always detected during incubation bouts, we only used the tem perature recordings to determ ine incubation schedules. Because we h ad indications that m am m alian predators easier found clutches incubated by PIT -tagged parents, we refrained from using them in 2008.

Clutches were considered to be incubated by two adult sanderlings (biparental) if two different birds (always a m ale and a female) were observed incubating those clutches. If only a single b ird was observed at a nest, the P IT in com bination with the therm ologger could indicate the presence of an additional incubating b ird if constant tem peratures above 36"C were m easured for several periods longer

than an hour while the single bird with transponder was not detected. In 2008, when we did no t use P IT ’s, we considered nine clutches uniparental, based on 4-8 nest visits (average 5.6) during which the same single incubating individual was encountered. O n the biparental nests the second attending adult was usually encountered during the second visit to a nest (average visit no. 2.4, range 1-4).

O f the 68 sanderling nests found in 2007 an d 2008, we obtained useful data on breeding schedules o f 34 clutches. M ost other clutches found either (1) fell victim to predation soon after discovery,(2) were found shortly before or at the day o f hatch or (3) h ad the therm ologger displaced. In 2007, 4 o f 14 clutches, and in 2008, 9 out o f 20 clutches were uniparental. Individuals are flexible in their strategy, as confirm ed by three individuals that adopted different breeding strategies in different years (unpubl. data). However, we never obtained useful da ta o f individual birds that were found incubating on two clutches w ithin o r betw een years. W e m onitored nest attendance for an average o f 6 ± 1.3 (SE) full days in uniparental clutches and 6.1 ± 0 .8 days in b iparental clutches.

Determination o f incubation schedulesIncubation schedules were determ ined using small w aterproof

therm ologgers (Tiny T ag, Gemini) th a t were placed n ear each nest and covered w ith soil and stones. A m easuring probe (2 x5 nini,

UNIPARENTAL

BIPARENTAL

S 20

fem a le

c lu tch te m pera tu re

0:00 04:00 08:00 12:00 16:00 20:00 0:00 04:00 08:00 12:00 16:00 20:00 0:00

Time (hours)

Figure 2. Two examples of an incubation pattern of 48 hours as registered by a tem perature probe and by the PITs. M easured tem pera tu res are depicted as a solid line, th e presence of a PIT tagged male (black) and fem ale (grey) as horizontal bars on top . The difference in incubation schedules with m ore recesses by the uniparental than the biparental clutch is clearly visible. Note th a t the PIT was no t always d etec ted w hen the tem pera tu re logger suggested an incubating bird (constant tem pera tu re around 40 C). doi:10.1371/journal.pone.0016834.g002

PLoS ONE I www.plosone.org 3 February 2011 | Volume 6 | Issue 2 | e16834

Egg Heat Input and Nest Attendance in Sanderling

tem peratu re range — 10"C to 50"C) was a ttached to the logger via th in electrical w ire an d fixed in the centre o f the nest cup betw een the four eggs with a toothpick-sized piece o f wood. T h e probe touched the b ro o d pa tch o f the incubating b ird bu t we have no indication that it ham pered its incubation. T h e loggers were p rogram m ed to take m easurem ents every m inute. After hatch, or w hen a clutch had been preyed upon, the loggers were collected and the da ta dow nloaded. S tart and end of incubation bouts were determ ined by inspection o f graphs o f tem peratu re against time (Fig. 2). In terpreta tion o f the tem peratu re graphs was straightfor­w ard because the relatively low am bient tem peratures resulted in clear an d im m ediate tem peratu re drops w hen nests were left unattended (Fig. 2). As soon as tem peratures were on the rise to levels h igher th an 36"C, incubation was p resum ed to have started. Occasional visual observations n ear nests o f birds th a t started or ended an incubation b ou t always confirm ed our in terpreta tion of the tem peratu re m easurem ent. In three nests the probe becam e displaced and did no t touch the b ird ’s b rood patch , resulting in tem peratu re graphs that were difficult to interpret. These recordings were no t included in our analysis.

Heat input into eggsA solid brass egg the size and shape o f an average sanderling egg

(35 x25 nini) was p laced in the nest cup and tem porarily replaced one o f the real eggs that was kept in a cup w ith cotton wool for the duration of the m easurem ent. A probe, a ttached to a small therm ologger (Tiny Tag), was m ounted in the core o f the m etal egg. M etal egg tem peratu re was recorded every 10 sec. T em p er­ature was recorded for a period o f a t least 20 m in continuous incubation w hich is sufficiently long to reach a stable tem perature p lateau due to the high heat conductance o f brass.

W e calculated m etal egg p lateau tem peratures during incuba­tion, by taking the average constant tem perature. T h e period during w hich tem perature was constant was defined to start when tem peratu re did no t increase for the first successive m easurem ent. T h e end was the m om ent w hen the b ird left the nest (Fig. 3). Four birds incubated for such short bouts that no stable tem perature

p lateau was reached. Because o f their h igher recess frequency (see Results), this occurred m ore often in uniparentals (three times) th an in biparentals (once). In those cases, the highest tem perature recorded was used in the analyses and regarded as a conservative m easurem ent (Fig. 3). W e perform ed the statistical tests bo th with or w ithout these m easurem ents, bu t since this did no t result in different qualitative outcom es o f the statistical analysis, these m easurem ents were included.

W e aim ed to m easure m etal egg tem peratures on days 7 and 14 of incubation (if the start o f incubation was not known it was estim ated on the basis o f egg flotation). I f nests were discovered after day 14, m easurem ents were m ade as soon as possible. Because nests were regularly discovered after day 7 o f incubation o r were preyed upon before day 14, we m anaged to m easure m etal egg tem perature on bo th day 7 and 14 in only two clutches. W e random ly selected only one brass egg tem perature m easurem ent for each o f those two clutches and used those in further analyses. W e used “m etal egg tem perature” as an indicator o f heat pu t into the eggs by the incubating bird, w hich can be com pared betw een individuals. W e realize that a m etal egg has a different heat conductance than a real sanderling egg and that m etal egg tem perature is also the result of heat loss to the environm ent. In our study we could not control heat loss from the m etal egg to the environm ent, bu t a single m etal egg and the same therm ologger were used throughout the study.

Ambient temperatureAir tem perature was m easured every five m inutes by therm is­

tors located at g round level at 20, 70, 150, 250 and 400 m altitude on the m ountain slope w here the sanderling nests w ere located. This slope was in the shade for several hours during the night resulting in lower night tem peratures despite the continuous daylight. T em peratures did no t differ significantly betw een the five locations (two-way A N O V A F = 0.003, d f = 4, P = 0.999) and we used the average tem perature across all locations and averaged per hour in the analyses. A m bient tem peratures varied betw een -1 .7 and 28.5"C (average 12.7±6.5). Precipitation occurred rarely and is therefore not considered as an additional factor.

" 38

U niparental Biparental0 10 20 30 40

41

40

39

38

37

36

Time since m easu rem en t (min)

Figure 3. Metal egg tem perature measurements in uni- and biparental clutches. Examples of tem pera tu re profiles during which a tem pera tu re p lateau was or was no t reached are show n in (A) and connected with a dashed line to the corresponding data poin t in (B). The final data are dep icted in (B) w here black solid circles indicate values of tem pera tu re profiles during which a plateau tem pera tu re w as reached and th e open circles w here it was not. doi:10.1371/journal.pone.0016834.g003

PLoS ONE I www.plosone.org February 2011 | Volume 6 | Issue 2 | e16834

Egg Heat Input and Nest Attendance in Sanderling

Arthropod availabilityIn 2007 we set up five transects o f eight-ten yellow plastic pitfall

traps (diam eter 10cm, dep th 8 cm), each 1 m apart, to obtain a m easurem ent o f food availability for sanderlings. T h e transects were located a t 20, 70, 150, 250 an d 400 m altitude in habitat types in w hich sanderlings were seen foraging an d close to where we found their nests. T he traps were filled up w ith w ater to ca. 4 cm below the top. H a lf a teaspoon of salt was added to each trap to p revent freezing and decom position o f the arthropods. A drop o f scentless detergent was added to each cup to b reak the surface tension. Between 19 Ju n e and 22 Ju ly 2007, arth ropods were collected daily betw een 17:00 an d 20:00 hrs and stored in 96% ethanol until analysis. Occasionally, some pitfall traps were destroyed by musk oxes Ovibos moschatus. T herefore, we express insect abundance per pitfall pe r day for all transects com bined. Invertebrates were identified up to order, and often also family level, and m easured to the nearest nini. D ry weights were estim ated from group specific length- dry weight relationships that were established based on subsamples o f the collected arthropods in 2007. M ites (Acari) an d springtails (Collembola) were considered too small to be im portan t as prey for sanderlings and were no t taken into account in the analysis.

Statistical analysesT o account for the hierarchical structure o f our da ta (e.g.

m ultiple hourly observations pe r day, m ultiple days pe r nest), analyses were perform ed using the m ultivariate multilevel m ixed- m odeling procedure in M Lw iN 2.15 [35]. This m ethod decom ­poses the total variance o f the dependent variables into three levels (between nests, days an d observations) by including ‘nest’ and ‘day’ as random effects (intercepts). T h e m odel was used to estim ate the effects o f am bien t tem perature and am bient tem perature squared (continuous fixed effects) on incubation schedules, bu t we did not account for the random variation in slope and intercept o f tem peratu re on the dependent variables. T h e fixed effects were m odelled as separate coefficients. A m bient tem perature squared was included in the m odel because we expected the three variables to rise until a peak at high am bient tem peratures, above which they w ould decrease again due to a decreased energy dem and of parents. Incubation stage was also included as fixed effect (both linearly and the second o rder polynomial), bu t was elim inated from the final m odel because it d id no t have significant effects. Incubation schedules were described by the dependent variables ‘total recess tim e’ (in m inutes pe r hour), ‘recess frequency’ (num ber o f recesses pe r hour) and average recess length (minutes pe r hour) for bo th uni- o r b iparen tal clutches. W e used periods o f an hour to study the effects o f am bient tem peratu re on incubation behaviour w ithin days. E arlier observations h ad shown that sanderlings regularly leave their clutches w ithin periods o f an hour.

T h e effect o f am bient tem perature on total recess tim e, recess frequency an d average recess length was investigated a t the level o f clutch, per day nested w ithin clutch and pe r hour nested within clutch an d day. All explanatory variables were centred on their averages. T o m eet norm ality assum ptions, recess tim e was arcsine transform ed and we took the natural logarithm s In (x+1) o f the num ber o f recesses hour *. All two-way interaction term s were included in the m odel initially an d then all insignificant interaction term s were rem oved sequentially (backward deletion) by decreas­ing P value. T he rem oval o f interactions did not change the significance o f m ain effects.

W e used an analysis o f covariance to test for the effects o f day of incubation, current am bient tem perature, clutch volume and num ber o f incubating adults on metal egg tem perature. All two- way interactions were initially included in the model. Non-significant

variables were rem oved from the model, starting with the interaction terms. All tests w ere two-tailed and threshold values o f P = 0.05 were applied throughout.

Results

Incubation schedulesBoth uni- an d biparentals regularly left their clutches u n attend­

ed (Figs 2 and 4). W hen we observed color-ringed sanderlings

100

D Uniparental I 1 Biparental

T3or

T3Ca»COc3

■eai+-«inaj Z

inQJtnin<VUCU

DC

* 20

05 10 15 20

F am b ien t ( C )

25 30

Figure 4. Average percentage of tim e that clutches were left unattended, recess frequency and recess duration of unipar­entals and biparentals. Uniparentals are indicated by grey boxes, biparentals by w hite boxes. For p resentation purposes the hourly am bien t tem pera tu res are catagorized per 5°C. The graphs are based on the raw data . The do ts indicate th e average, th e boxes encloses 50% of the data and th e error bars 95%. doi:10.1371/journal.pone.0016834.g004

PLoS ONE I www.plosone.org February 2011 | Volume 6 | Issue 2 | e16834

Egg Heat Input and Nest Attendance in Sanderling

during an off-bout they usually w ere foraging (cf. [10]), bu t we ocassionally also observed sanderling males leaving their nest to chase away o ther sanderlings in the vicinity, as was also observed by [36], U niparentals spent m ore tim e off their nest (42.1 ±6.3% ) th an biparentals (32.2 ± 14.5%) an d this difference alm ost reached 5% -statistical significance (Table 1). T h e average for biparentals is influenced by a single unusual clutch that was left una ttended for 68% o f the 72 hours o f available data. Excluding this value makes the tim e spent off the nest by biparentals 3 0 ± 1 1 .3 % , and the difference with uniparentals significant (Chi-square = 3.94, elf = 1, P = 0.047). T h e difference was b rough t about by the m ore frequent recesses as well as longer recess durations o f uniparental sanderlings com pared to b iparen ta l sanderlings (Table 1, see Fig. 2 for an example).

A m bient tem perature at the level o f hour positively affected recess frequency bu t did no t have an overall effect on total recess tim e and recess duration . T h e significant interactions betw een breeding strategy an d am bient tem perature at the level o f hour indicates that b reeding schedules o f uniparentals were affected by am bient tem peratures, w hereas they were no t in b iparentals (Table 1, Fig. 4). At lower tem peratures all three descriptors o f incubation patterns w ere similar betw een uni- and biparentals, bu t with increasing tem peratures uniparentals increased recess fre­quency and recess duration (Fig. 4). This resulted in longer periods during w hich uniparentals left their clutch una ttended a t higher am bient tem peratures (Table 1, Fig. 4). T h e interaction betw een

breeding strategy an d am bient tem perature was also significant on the level o f day for recess frequency, bu t no t for recess duration and total recess tim e (Table 1). Recess frequency pe r hour leveled off and slightly decreased w ith h igher am bient tem peratures (Fig. 4), indicated also by the significant effect o f squared am bient tem perature at the level o f hour (Table 1).

Metal egg temperatureD ay o f incubation (Ancova, F 151 = 0.323, P = 0.58), clutch

volum e (F161 = 0.495, P = 0.49) and am bient tem perature (Fi? i = 1.098, P = 0.31) did no t effect m etal egg tem peratures, bu t w hether clutches were incubated by one or two parents did (F i7 ,i = 6.563, P = 0.02). M etal egg tem peratures were on average 0.8"C higher in uni- com pared to biparentals (Fig. 3). I f the analysis is restricted to m easurem ents in w hich a tem perature p lateau was reached only, the average difference is m ore pronounced (1.4"C; F 13 t = 33.55, PCO.OOl).

Discussion

O u r da ta confirm earlier observations [30] th a t clutches o f sanderlings in northeast G reen land are incubated either by one or by two parents, although un iparen tal clutches were less com m on (13 out o f 34 nests). Recess tim e in the un iparen tal sanderlings (42.1%) was considerably h igher com pared to w hat was found in o ther un iparental shorebirds in arctic o r tem perate regions

Table 1. Multivariate model summaries examining total recess tim e (minutes hour 1), recess frequency (recesses hour q) and recess duration for uniparental relative to biparental (reference category) sanderling clutches.

R e s p o n s e v a r i a b le

F ix ed e f fe c t P (SE) X2 P

Total recess time

Intercept 36.1 < 0 .0 0 1

Breeding strategy 8.34 (4.51) 3.43 0.06

Tambient (hour) 0.20 (0.19) 1.13 0.29

Tambient2 (hour) -0.005 (0.007) 0.51 0.47

Breeding strategy * Tambient (hour) 2 .1 3 (0 .29) 5 3 .6 < 0 .0 0 1

Breeding strategy * Tambient2 (hour) - 0 . 0 2 (0 .01) 4 .7 3 0 .0 3

Recess frequency

Intercept 1 4 1 .2 < 0 .0 0 1

Breeding strategy 0 .4 3 (0 .05) 6 4 .2 < 0 .0 0 1

T a m b ie n t (hour) 0 .0 2 (0 .004 ) 2 9 .2 < 0 .0 0 1

T a m b ie n t (day) 0.00 (0.004) 0.00 1.00

T a m b ie n t2 (hour) - 0 .0 0 1 (0 .00) 32.1 < 0 .0 0 1

Breeding strategy * T a m b ie n t (day) 0 .0 2 (0 .005 ) 1 9 .9 < 0 .0 0 1

Breeding strategy * T a m b ie n t (hour) 0 .0 4 (0 .002 ) 2 6 1 .3 < 0 .0 0 1

Recess duration

Intercept 5 8 .5 < 0 .0 0 1

Breeding strategy 3.01 (0 .91) 1 0 .9 0 .001

T a m b ie n t (nest) - 1 .3 1 (0 .45) 8 .6 9 0 .0 0 3

T a m b ie n t (hour) 0.05 (0.06) 0.50 0.48

T a m b ie n t (nest) 0 .0 5 (0 .02) 6 .0 8 0.01

Breeding strategy * T a m b ¡ e n t (hour) 0 .3 0 (0 .10) 8 .6 6 0 .0 0 3

Significant factors are in bold. The final model after elimination of the non-significant interaction terms is shown, but note that non-significant main effects were kept in the model if their interaction terms were significant. doi:10.1371 /journal, pone.0016834.t001

PLoS ONE I www.plosone.org 6 February 2011 | Volume 6 | Issue 2 | e16834

Egg Heat Input and Nest Attendance in Sanderling

(Table 2). Also the recess tim e o f 32.2% in b iparen ta l sanderlings exceeded the range o f o ther arctic and tem perate b iparental shorebirds, bu t fell w ithin the range o f tropical b iparental shorebirds (Table 2). Recess tim e can be affected by several factors, such as energetic requirem ents o f birds, food availability, insulation of the nest cup and climatic conditions [10,37]. C om pared with m any o ther H igh Arctic regions, in northeast G reen land am bient tem perature is relatively high and air hum idity is low [38], leading to a relatively slow heat loss o f eggs, w hich m ay explain why the frequent and long recesses o f sanderlings recorded in our study were possible.

Despite the relatively high overall recess tim e in sanderlings, there was a clear difference betw een uni- and b iparen ta l clutches in recess tim e and the effect o f am bien t tem peratu re on breeding schedules. This reconfirm s th a t the trade-off betw een incubation and foraging is m ore severe for uniparentals because incubation

duties cannot be taken over by a p a rtn e r while foraging. By foraging during frequent bu t short bouts, an d by tim ing such recesses during the w arm er periods o f the day, uniparentals can probably m inim ize the variation in egg tem peratures, which increases energy use o f the em bryo and negatively affect hatchling phenotype [11]. Experim ental heating of nests resulted in higher nest a ttendance in b iparen tal sandpipers [20] an d un iparen tal tree swallows Tachycineta bicolor [39]. This suggests th a t incubation schedules are determ ined by the energetic trade-offs o f the incubating birds ra th e r than optim ization of the conditions for em bryonic developm ent. O u r correlative da ta suggest a different response to h igher tem perature, bu t this in pa rt reflects the within- day organization o f foraging recesses, concentrated in the w arm er m id-day hours.

Leaving the clutch una ttended during the w arm est periods o f the day is p robably in tended to keep suboptim al conditions for

Table 2. Average nest attentiveness, recess frequency and recess duration per 24 hrs o f uni- and biparentally incubating shorebirds in different climates.

B re e d in gs t r a t e g y C lim a te S p e c ie s

N es ta t t e n t iv e n e s s

R ec essf r e q u e n c y

R e c e ss d u r a t io n (m in ) M e th o d Ref

Uni Arc Calidris ferruginea 0.82 ±0.07 21.6±7.4 15.3 ±19.5 Tnest [10]

Uni Arc Calidris fuscicollis 0.83 25.1 10.5 camera [9]

Uni Arc Calidris melanotus 0.82 ±0.01 28.8 ±1.4 6.0 Tnest [20]

Uni Arc Calidris melanotus 0.85 ±0.01 1"nest/1"egg [50]

Uni Arc Calidris melanotus 0.83 ±0.06 29.9±9.2 9.8±9.0 1"nest [10]

Uni Arc Calidris minuta 0.81 ±0.08 26.5±4.4 9.4 ±8.6 1"nest [10]

Uni Arc Charadrius morinellus 0.90 ±0.03 1"nest [52]

Uni Arc Charadrius morinellus 0.83 1"nest [53]

Uni Temp Gallinago media 0.90 ±0.02 15.7± 6.1 8.7±1.9 1"nest [54]

Uni Arc Phalaropus fulicarius 0.87 ±0.02 7.2 ±1.0 1"nest [10]

Uni Arc Phalaropus fulicarius 0.81 6.6 camera [55]

Uni Temp Rostrathula bengalensis 0.84 obs [56]

R a n g e (e x c lu d in g th i s s tu d y ) 0 .8 1 -0 .9 0 1 5 .7 -2 9 .9 6 .0 -1 5 .3

Uni A rc Calidris alba 0 .5 7 ± 0 .1 6 4 .6 ± 14.1 7 .2 ± 2 .1 PIT/Tnest T h is s tu d y

Bi Arc Calidris alpina 0.98 ±0.00 6.0±1.0 1"nest/1"egg [51]

Bi Arc Calidris bairdii 0.97 ±0.00 6.0±1.0 1"nest/1"egg [51]

Bi Arc Calidris pusilla 1.00 ±0.00 1.5 ±0.5 PIT/Tnest [22]

Bi Arc Charadrius morinellus 0.91 ±0.03 Tnest [52]

Bi Arc Charadrius morinellus 0.96 obs [53]

Bi* Temp Vanellus vanellus 0.84 ±0.02 camera [57]

R a n g e (e x c lu d in g th i s s tu d y ) 0 .8 4 -0 .9 8 1 .5 -6 .0

Bi A rc Calidris alba 0.71 ± 0 .1 3 8 .7 ± 7 .2 5.1 ± 3 .8 PIT/Tnest T h is s tu d y

Bi Trop Charadrius alexandrinus 0.86 ±0.01 obs [58]

Bi Trop Charadrius peronii 0.67 obs [59]

Bi Trop Charadrius wilsonia 0.83 obs [60]

Bi Trop Vanellus coronatus 0.99 Tnest/Obs [61]

Bi Trop Vanellus lugubris 0.91 Tnest/Obs [61]

Bi Trop Vanellus melanopterus 0.89 Tnest/Obs [61]

R a n g e (e x c lu d in g th i s s tu d y ) 0 .6 7 -0 .9 9

Climate is subdivided into arctic ('Arc'), temperate ('Temp') or tropical (Trop'). In bold are the ranges of average values for uniparentals and biparentals in Arctic or temperate regions and for biparentals in tropical regions summarized. They are compared with the values for sanderling based on this study (also in bold). Abbreviations for recording methods refer to visual observation (obs), camera surveillance (camera), recording of nest (Tnest) or egg temperature (Tegg), or recording of presence of PIT-tagged parents (PIT).*Based on 20 clutches of which two were never attended by a male and were thus uniparental, and 10 others only attended by males during daytime. doi:10.1371 /journal.pone.0016834.t002

PLoS ONE I www.plosone.org 7 February 2011 | Volume 6 | Issue 2 | e16834

Egg Heat Input and Nest Attendance in Sanderling

-£ 20

Average daily Tambient(°C)

Figure 5. Relationship between daily mean am bient tem pera­ture and the average dry weight of arthropods collected per pitfall trap. The linear regression is statistically significant (Fq=31.2, PC0.0001, R2 = 0.486). doi:10.1371/journal.pone.0016834.g005

em bryonic grow th to a lim it, bu t also suggests that the possibility for sanderlings to raise a clutch alone depends to a certain degree on environm ental circum stances an d the condition o f the incubating bird. Indeed, three uniparental, bu t no biparental, sanderlings were observed to abandon their nests before hatch during severe w eather conditions with low am bient tem peratures, high w ind speed and snowfall (JR unpubl. obs. and see [26]). In a different study, un iparental sandpipers w ere shown to considerably decrease in body mass during periods with low tem peratures, while a b iparen ta l species did not [10].

In addition to the need to provide a favourable physical environm ent for the developing em bryo, foraging success m ay also determ ine incubation schedules, especially in un iparental sander­lings. A rthropod availability on the Arctic tun d ra is strongly influenced by daily am bient tem perature [40,41] an d this was confirm ed in our study (Fig. 5). T his indicates that foraging during the w arm er periods o f the day has bo th the advantage of a slower cooling o f the clutch and potential h igher food intake ra te by the foraging ad u lts . T h a t un iparental sanderlings increase recess frequency, recess duration and total recess tim e with am bient tem peratures at the level o f hour, bu t no t consistently for the three variables a t the levels o f day and nest (Table 1), m ay indicate that the birds need to forage daily for a m inim um period o f time regardless o f the daily average tem perature, bu t that if there is the possibility, the w arm est period w ithin days is chosen to leave the nest. An alternative, bu t no t m utually exclusive, explanation is that the birds com pensate for daily fluctuations in their energy balance with stored energy to avoid too m uch variation in incubation constancy betw een days [10].

A djusting b reeding schedules to am bient tem peratures probably helps un iparental sanderlings m inim ize fluctuations in egg tem peratures while still m eeting energetic dem ands. However, overall nest attendence o f un iparental sanderlings was 9.9% lower com pared to biparentals. Leaving a clutch unattended slows down em bryonic grow th and therefore lengthens the incubation period [6,12,15,42]. Starling Sturnus vulgaris clutches that w ere incubated

References1. W illiam s J B (1996) E n erg etics o f a v ian in cu b a tio n . In : C a re y C , ed . A v ian

E n erg etics a n d N u tr it io n a l E cology. N e w Y ork: C h a p m a n a n d H a ii. p p 3 7 5 ^ 1 5 .

by bo th parents, and hence received a h igher attendance, indeed had shorter total incubation periods [2]. A longer incubation period has negative im plications for curren t an d future reproduc­tive success arising from higher energetic costs for the parents and an increased probability o f clutch p redation [2,17,43,44], T he latter is a serious factor, as p redation rates on shorebird nest can be considerable in the H igh Arctic [45,46], A longer incubation period m ay also have negative consequences for the offspring, th rough a poorer condition a t hatch due to greater energy expenses during the longer incubation period [11,42,47,48],

W e argue that the h igher heat input into the eggs by uniparental th an b iparen ta l sanderlings, indicated by the h igher m easured m etal egg tem peratures, could be a way to com pensate for the effect o f lower incubation constancy on incubation duration, bu t future experim ents are needed to confirm our correlative data. By m aintain ing h igher incubation tem peratures, parents can effi­ciently shorten incubation duration and positively m odify offspring phenotype [16,39,49], M ainta in ing high egg tem peratures is likely to be energetically costly and can probably only be perform ed by experienced individuals in good physical condition, as suggested by a study on b iparen ta l king penguins Aptenodytes patagonicus that were experim entally p revented from sharing incubation [50], In the absence o f a helping partner, birds incubated longer th an control birds an d kept egg tem peratu re constant, bu t egg tem perature significantly decreased two days before it was finally abandoned.

W e have dem onstra ted that un iparen ta l birds keep a tighter b reeding schedule th an biparentals and, for the first tim e under na tu ra l conditions, that they pu t m ore body heat into their clutches. These possible adaptations m ust come w ith some costs, raising the question o f why some sanderlings incubate clutches w ithout the help o f a partner. D ouble-clutching w ould have the benefits o f a possibly twice as large reproductive success in a given breeding season. In the relatively benign clim ate in northeast G reenland, un iparen ta l incubation is apparently feasible, bu t possibly only by high quality individuals. T h e existence o f this double-clutching breeding system has, however, never been convincingly proven [25] an d if it w ould occur, it is possible that males and females profit unequally if extra-pair paternity occurs too. M olecular paren tage analysis is needed to gain m ore insight into the b reeding system o f sanderlings and to evaluate the costs and benefits o f un iparen ta l incubation under different environ­m ental conditions.

AcknowledgmentsT h e staff o f Z ackenberg ecological re sea rch opera tions (Z E R O ) is acknow ledged for prov id ing logistics a t Z ackenberg an d for allow ing us to w ork in their long-term m onito ring area. W e gratefully acknow ledge M areen Becking, K oos D ijksterhuis, J an n ik H ansen , Jo s H ooijm eijer, J o o p Ju k em a an d K ees de R ijk fo r assistance in the fieldwork. O u r colleagues N iels D ingem anse, M artijn H am m ers, M artijn van de Po l a n d Jo o st T in b erg en kindly gave statistical advice. W e also app recia te the useful com m ents o f two anonym ous referees.

Author ContributionsConceived and designed the experiments: JR KG HS IT TP. Performed the experiments: JR KG HS IT. Analyzed the data: JR. Contributed reagents/mate rials/analysis tools: JR HS IT TP. Wrote the paper: JR KG HS IT TP.

2. R e id J M , M o n a g h a n P , N a g e r R G (2002) In c u b a tio n a n d th e co st o f rep ro d u c tio n . In : D e e m in g D C , ed . A v ian In cu b a tio n . B eh av io u r, E n v iro n m e n t, a n d E vo lu tio n . O x fo rd : O x fo rd U n iv ers ity Press, p p 314—325.

PLoS ONE I www.plosone.org 8 February 2011 | Volume 6 | Issue 2 | e16834

Egg Heat Input and Nest Attendance in Sanderling

3. T in b e rg e n J M , W illiam s J B (2002) E nergetics o f in cu b a tio n . In : D eem in g D C , ed. A v ian In cu b a tio n . B eh av io u r, E n v iro n m e n t, a n d E vo lu tion .. O xford : O x fo rd U n iversity Press, p p 299—313.

4. P ie rsm a T , L in d s trö m Á , D re n t R H , T u lp I, J u k e m a J , e t al. (2003) H ig h daily en e rg y e x p e n d itu re o f in c u b a tin g sh o re b ird s o n H ig h A rc tic tu n d ra : a c irc u m p o la r study. F u n c t E co l 17: 356—362.

5. T u lp I, S ch ek k e rm an H , B ru inzee l L W , J u k e m a J , V isser G H , e t al. (2009) E n e rg e tic d e m a n d s d u r in g in cu b a tio n a n d chick rea r in g in a u n ip a re n ta l a n d a b ip a re n ta l sh o re b ird b ree d in g in th e H ig h A rctic . A uk 126: 155—164.

6. W e b b D R (1987) T h e rm a l to le ran ce o f av ian em bryos: a review . C o n d o r 89: 8 7 4 -8 9 8 .

7. C a re y C (2001) In cu b a tio n in e x trem e en v iro n m en ts . In: D e e m in g D C , ed. A v ian In cu b a tio n . B eh av io u r, E n v iro n m e n t, a n d E vo lu tion . O xfo rd : O x fo rd U n iversity Press, p p 2 3 8 -2 5 3 .

8. K laassen M (2003) R e la tio n sh ip s b e tw e en m ig ra tio n a n d b ree d in g stra teg ies in A rc tic b ree d in g b irds. B e rth o ld P , G w in n er E , S o n n en sch e in E , eds. A v ian M ig ra tio n . Berlin: Springer-verlag . p p 237—249.

9. C a r ta r R V , M o n tg o m erie R D (1985) T h e in fluence o f w e a th e r o n in cu b a tio n sched u lin g o f th e w h ite -ru m p e d sa n d p ip e r {Calidris ßiscicollis): a u n ip a re n ta l in cu b a to r in a co ld en v iro n m en t. B eh av io u r 95: 261—289.

10. T u lp I, S ch ek k e rm an H (2006) T im e a llo ca tio n b e tw een feed in g a n d in cu b a tio n in u n ip a re n ta l a rc tic -b re e d in g shoreb irds: en erg y reserves p ro v id e leew ay in a tig h t schedule . J A v ian B iol 37: 2 0 7 -2 1 8 .

11. O lso n C R , V leck C M , V leck D (2006) P e rio d ic coo ling o f b ird eggs reduces e m b ry o n ic g ro w th efficiency. P hysio l B io ch em Z oo l 79: 92 7 —936.

12. D ee m in g D C , F e rg u so n M W J (1991) P hysio logical effects o f in cu b a tio n t e m p e r a tu r e o n e m b r y o n ic d e v e lo p m e n t in r e p t i le s a n d b i rd s . In : D e em in g D C , F e rg u so n M , eds. E g g In cu b a tio n : Its Effects o n E m b ry o n ic D ev e lo p m en t in Birds a n d R eptiles. C a m b rid g e : C a m b rid g e U n iversity Press, p p 1 47-173 .

13. R e id J M , M o n a g h a n P, R u x to n G D (2000) T h e co n seq u en ces o f c lu tch size for in cu b a tio n cond itio n s a n d h a tc h in g success in starlings. F u n c t E co l 14: 56 0 —565.

14. B osque C , B osque M T (1995) N e s t p re d a tio n as a selective fac to r in theev o lu tion o f d ev e lo p m en ta l ra te s in a ltric ia l b irds. A m . N a t 145: 234—260.

15. H e p p G R , Folk T H , M an lo v e C H (2005) N e s t tem p e ra tu re , in cu b a tio n p e rio d a n d in v estm en t decisions o f in cu b a tin g w o o d ducks Aix sponsa. J A v ian Biol 36: 5 2 3 -5 3 0 .

16. H e p p G R , K e n n a m e r R A , J o h n s o n M H (2006) M a te rn a l effects in w o o d ducks: in cu b a tio n te m p e ra tu re influences in cu b a tio n p e rio d a n d n e o n a te p h en o ty p e . F u n c t E co l 20: 307—314.

17. M a rtin T E , A u e r SK , B assar R D , N iklison A M , L loyd P (2007) G eo g rap h ic v a ria tio n in av ian in cu b a tio n p e rio d s a n d p a re n ta l in fluences o n em b ry o n ic tem p e ra tu re . E vo lu tio n 61: 2 5 5 8 -2 5 6 9 .

18. T h o m so n D L , M o n a g h a n P , F u rn ess R W (1998) T h e d e m a n d s o f in cu b a tio na n d a v ian c lu tch size. Biol R e v 73: 293—304.

19. R e id J M , C ressw ell W , H o lt S, M e llan b y RJ, W hitfie ld D P , e t al. (2002) N est sc rap e design a n d c lu tch h e a t loss in p ec to ra l san d p ip e rs {Calidris melanotus). F u n c E co l 16: 3 0 5 -3 1 2 .

20. C ressw ell W , H o lt S, R e id J M , W h itfie ld D P , M ellan b y R J, e t al. (2004) T h e e n e rg e tic costs o f egg h e a tin g c o n s tra in in cu b a tio n a tte n d a n ce b u t do n o t d e te rm in e daily en erg y ex p e n d itu re in th e p ec to ra l sa n d p ip e r. B ehav E co l 15:4 9 8 -5 0 7 .

21. C ressw ell W , H o lt S, R e id J M , W hitfie ld D P , M ellan b y R J (2003) D o en e rg ed c d e m a n d s c o n s tra in in cu b a tio n sched u lin g in a b ip a re n ta l species? B ehav E co l 14: 9 7 -1 0 2 .

22. T u lp I (2007) T h e A rc tic pulse. T im in g o f b ree d in g in lo n g -d is tan ce m ig ra n t shoreb irds . G ro n in g en : P h D thesis U n iversity o f G ro n in g en , 260 p . A vailab le via: h t tp : / / i r s .u b . ru g .n l /p p n /3 0 4 4 0 1 8 2 X .

23. D A lb a L, M o n a g h a n P , N a g e r R G (2009) T h e rm a l b enefits o f n es t she lte r forin cu b a tin g fem ale eiders. J T h e rm B iol 34: 93—99.

24. T o m k o v ich P S , Soloviev M Y (2001) Social o rg an isa tio n o f sanderlings b ree d in ga t n o r th e rn T a im y r, S iberia . O rn ith o lo g ia 29: 125—136.

25. R en ee rk en s J , B enhoussa A , B o lan d H , C o llie r M , G ro n d K , e t al. (2009) Sand erlin g s usin g A frican -E u ras ian flyways: a rev iew o f c u rre n t know ledge. W a d e r S tudy G ro u p Bull 116: 2 -2 0 .

26. P ienkow ski M W , G re e n G H (1976) B reed in g b io logy o f S anderlings in n o r th ­east G reen lan d . B rit B irds 60: 16 5 -1 7 7 .

27. P a rm e lee D F , P ay n e R B (1973) O n m ultip le b ro o d s a n d th e b ree d in g s tra tegy o f A rc tic sanderlings. Ib is 115: 2 1 8 -2 2 6 .

28. P ie rsm a T , v a n G ils J A , W ie rsm a P (1996) F am ily S co lopac idae {sandpipers, sn ipes a n d p h a laro p es). In: de l H o y o J , E llio tt A , S a rg a ta lJ , eds. H a n d b o o k o f th e B irds o f th e W o rld , V ol.3 H o a tz in to A uks B arcelona : L y n x E dicions. p p 4 4 4 -5 3 3 .

29. M elto fte H , R a sc h M (2008) T h e study a re a a t Z ack en b erg . In: M elto fte H , C h ris te n sen T R , E lb e rlin g B, F o rc h h am m er M C , R a sc h M , eds. T e n y ea rs o f m o n ito rin g a n d resea rch a t Z ac k e n b erg R e se a rc h S ta tion , N o r th e a s t G re e n la n d , A dvances in E cological R e se a rc h V ol. 40. S an D iego: A cad em ic Press, p p 1 0 1 - 1 1 0 .

30. P ie rsm a T , M elto fte H , J u k e m a J , R en ee rk en s J , d e G oeij P, e t al. (2006) A single-year co m p a riso n o f tw o m eth o d s o f censusing b ree d in g R e d K n o t a n d S an d e rlin g in H ig h A rc tic G reen lan d . W a d e r S tu d y G ro u p Bull 109: 83—87.

31. L ieb eze it J R , S m ith PA , L a n c to t R B , S ch ek k e rm an H , T u lp I, e t al. (2007) A ssessing th e d e v e lo p m en t o f sh o re b ird eggs using th e f lo ta tio n m eth o d : Species- specific a n d gen e ra lized reg ression m odels. C o n d o r 109: 32—47.

32. S c h m id t N M , B erg T B , M elto fte B (2010) BioBasis, C o n c e p tu a l d esign a n d sa m p lin g p ro ce d u re s o f th e b io log ical m o n ito rin g p ro g ra m m e w ith in Z a ck en ­b e rg Basic. 13th ed itio n Roskilde: T h e N a tio n a l E n v iro n m e n ta l R esea rch In stitu te . D e p a r tm e n t o f A rc tic E n v iro n m en t.

33. R o m a n o ff AE, R o m a n o ff A J (1949) T h e av ian egg. N ew Y ork: W iley a n d sons Inc. 918 p.

34. R en ee rk en s J , P ie rsm a T , J u k e m a J , d e G oeij P , Bol A , e t al. (2005) S ex-ratio a n d b o d y size o f s a n d p ip e r chicks a t Z ack en b erg , n o r th -e as t G re e n la n d in 2003. W a d e r S tudy G ro u p Bull 106: 12 -1 6 .

35. R a sb a sh J , S teele F , B row ne W , Prosser B (2004) A u sers’ gu ide to M L w iN . L o n d o n : C e n tre fo r M ultilevel M o delling , In stitu te o f E d u c a tio n , U n iversity o f L o n d o n . 260 p.

36. P a rm e lee D F (1970) B reed in g b eh av io r o f th e san d erlin g in th e C a n a d ia n H ig h A rctic . L iv ing B ird 9: 9 7 -1 4 6 .

37. D ee m in g D C (2001) B eh av io u r p a tte rn s d u r in g in cu b a tio n . In: D ee m in g D C , ed. A v ian In cu b a tio n . B eh av io u r, E n v iro n m e n t, a n d E vo lu tion . O x fo rd : O x fo rd U n iversity press, p p 63—87.

38. M elto fte H , P ie rsm a T , Boyd H , M cC affery B, G a n te r B, e t al. (2007) Effects o f c lim ate v a ria tio n o n th e b ree d in g ecology o f A rc tic shoreb irds . M ed d elser o m G ro n lan d -B io sc ien ce 59: 1—48.

39. A rd ia D R , P é rez J H , C h a d E K , V oss M A , C lo tfe lte r E D (2009) T e m p e ra tu re a n d life history: e x p e rim en ta l h e a tin g leads fem ale tree swallow s to m o d u la te egg te m p e ra tu re a n d in cu b a tio n b eh av io u r. J A n im E co l 78: 4—13.

40. S ch ek k e rm an H , T u lp I P ie rsm a T , V isser G H (2003) M ech an ism s p ro m o tin g h ig h er g ro w th ra te in a rc tic th a n in tem p e ra te shoreb irds . O eco lo g ia 134: 3 3 2 -3 4 2 .

41. T u lp I , S ch ek k e rm an H (2008) H a s p rey availab ility for a rc tic b ird s a d v an ced w ith c lim ate ch an g e? H in d c a stin g th e a b u n d a n c e o f tu n d ra a rth ro p o d s using w e a th e r a n d seasonal v a ria tio n . A rc tic 41: 48—60.

42. A rd ia D R , P é rez J H , C lo tfe lte r E D (2010) E x p e rim e n ta l coo ling d u r in g in cu b a tio n lead s to red u c e d in n a te im m u n ity a n d b o d y co n d itio n in nestling tree swallows. P ro c R o y Soc B 277: 1 8 8 1 -1 8 8 8 .

43. T o m b re IM , E rik stad K E (1996) A n ex p e rim e n ta l study o f in cu b a tio n effort in h ig h -A rc tic b a rn a c le geese. J A n im E co l 65: 325—331.

44. L in d s tro m J (1999) E a rly d e v e lo p m en t a n d fitness in b ird s a n d m am m als . T re n d s E co l E vol 14: 3 4 3 -3 4 8 .

45. M c K in n o n L , S m ith PA N o l E , M a rd n J L , D oy le F I, e t al. (2010) L ow er p re d a tio n risk fo r m ig ra to ry b ird s a t h igh la titudes . Science 327: 326—327.

46. S m ith PA , W ilson S (2010) In tra se a so n a l p a tte rn s in sh o re b ird n est surv ival a re re la te d to n es t age a n d defence b eh av io u r. O eco lo g ia 163: 61 3 —624.

47. V leck C M , H o y t D F (1980) P a tte rn s o f m etab o lism a n d g ro w th in av ian em bryos. A m Z o o l 20: 4 0 5 -4 1 6 .

48. H a in sw o rth F R , M o o n a n T , V oss M A , Sullivan K A , W e a th e rs W W (1998) T im e a n d h e a t a lloca tions to b a la n c e conflicting d e m a n d s d u r in g in te rm itte n t in cu b a tio n b y yellow -eyed ju n co s. J A v ian Biol 29: 113—120.

49. P é rez J H , A rd ia D R , C h a d E K , C lo tfe lte r E D (2008) E x p e r im e n ta l h ea tin g reveals n est tem p e ra tu re affects n estling c o n d itio n in tre e swallow s {Tachycineta bicolor). B iol L e tt 23: 4 6 8 -4 7 1 .

50. G rosco las R , D ecro ck F , T h il M -A , F ayolle G , Boissery G , e t al. (2000) R efeed in g signal in fas tin g -in cu b a tin g k in g p engu ins: ch an g es in b e h a v io r a n d egg tem p e ra tu re . A m J P hysio l-R eg I n t C o m p Physiol 279: R 2 1 0 4 -R 2 1 1 2 .

51. N o r to n D W (1972) In cu b a tio n schedules o f fo u r species o f ca lid rid in e sandp ipers a t B arro w , A laska. C o n d o r 74: 164—176.

52. K a las J A (1986) In c u b a tio n schedules in d iffe ren t p a re n ta l ca re system s in the d o tte re l Charadrius morinellus. A rd e a 74: 185—190.

53. P u llia in en L, S aa ri E (1997) A tten d a n c e a t th e nest o f p o ly an d ro u s d o tte re l Charadrius morinellus in F in land . A rd e a 85: 67—71.

54. L ofald i L (1985) In cu b a tio n rh y th m in th e g re a t sn ipe Gallinago media. H o i E co l 8: 10 7 -1 1 2 .

55. M e h lu m F (1991) T h e in cu b a tio n b eh a v io r o f th e g ray p h a la ro p e Phalaropus ßilicarius o n S v a lb ard A rc tic O c e a n . F a u n a N o r Ser C C inc lus 1: 33—38.

56. K o m e d a S (1983) N e s t a tte n d a n ce o f p a re n t b ird s in th e p a in te d sn ipe {.Rostratula benghalensis). A uk 100: 4 8 -5 5 .

57. J o n g b lo e d F , S ch ek k erm an H , T eu n issen W (2006) D ivision o f in cu b a tio n effort in lapw ings. L im osa 79: 63—70.

58. A lrash id i M , K o sz to lán y i A , K ü p p e r G , G u th ill IG , J a v e d S, e t al. (2010) T h e in fluence o f a h o t e n v iro n m en t o n p a re n ta l co o p e ra tio n o f a g ro u n d -n es tin g sh o reb ird , th e K e n tish p lo v er Charadrius alexandrinus. F ro n t Z o o l 7: 1.

59. Y asue M , D e a rd e n P (2008) P a ren ta l sex ro les o f M a lay sian p lovers d u r in g te rr ito ry acq u is itio n , in cu b a tio n a n d ch ick -rearing . J E th o l 26: 99—112.

60. T h ib a u lt M , M cN eil R (1995) D ay -tim e a n d n ig h t- tim e p a re n ta l in v es tm en t by in cu b a tin g W ilsons p lovers in a tro p ica l e n v iro n m en t. C a n J Z o o l 73: 87 9 —886.

61. W a rd D (1990) In cu b a tio n tem p e ra tu re s a n d b eh a v io u r o f cro w n ed , b lack ­w in g ed a n d lesser b lack-w inged Plovers. A uk 107: 10—17.

PLoS ONE I www.plosone.org 9 February 2011 | Volume 6 | Issue 2 | e16834