assessing the development of shorebird eggs using the flotation method: species-specific and...

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors nonprofitpublishers academic institutions research libraries and research funders in the common goal of maximizing access tocritical research

ASSESSING THE DEVELOPMENT OF SHOREBIRD EGGSUSING THE FLOTATION METHOD SPECIES-SPECIFICAND GENERALIZED REGRESSION MODELSAuthor(s) JOSEPH R LIEBEZEIT PAUL A SMITH RICHARD B LANCTOTHANS SCHEKKERMAN INGRID TULP STEVE J KENDALL DIANEM TRACY ROBERT J RODRIGUES HANS MELTOFTE JULIE AROBINSON CHERI GRATTO-TREVOR BRIAN J MCCAFFERY JULIEMORSE and STEVE W ZACKSource The Condor 109(1)32-47 2007Published By Cooper Ornithological SocietyDOI httpdxdoiorg1016500010-5422(2007)109[32ATDOSE]20CO2URL httpwwwbiooneorgdoifull1016500010-5422282007291095B323AATDOSE5D20CO3B2

BioOne (wwwbiooneorg) is a nonprofit online aggregation of core research in thebiological ecological and environmental sciences BioOne provides a sustainable onlineplatform for over 170 journals and books published by nonprofit societies associationsmuseums institutions and presses

Your use of this PDF the BioOne Web site and all posted and associated contentindicates your acceptance of BioOnersquos Terms of Use available at wwwbiooneorgpageterms_of_use

Usage of BioOne content is strictly limited to personal educational and non-commercialuse Commercial inquiries or rights and permissions requests should be directed to theindividual publisher as copyright holder

ASSESSING THE DEVELOPMENT OF SHOREBIRD EGGS USINGTHE FLOTATION METHOD SPECIES-SPECIFIC AND

GENERALIZED REGRESSION MODELS

JOSEPH R LIEBEZEIT113 PAUL A SMITH

2 RICHARD B LANCTOT3

HANS SCHEKKERMAN4 INGRID TULP

4 STEVE J KENDALL5 DIANE M TRACY

6ROBERT J RODRIGUES

7 HANS MELTOFTE8 JULIE A ROBINSON

914CHERI GRATTO-TREVOR

10 BRIAN J MCCAFFERY11

JULIE MORSE12 AND STEVE W ZACK

1

1Wildlife Conservation Society Pacific West Office 219 SW Stark St Suite 200 Portland OR 972042National Wildlife Research Centre Carleton University 1125 Colonel By Drive Ottawa ON K1A 0H3 Canada3US Fish and Wildlife Service Migratory Bird Management 1011 E Tudor Rd MS 201 Anchorage AK 99503

4Institute of Marine Resources and Ecosystem Studies PO Box 167 1790 AD Den Burg The Netherlands5US Fish and Wildlife Service Arctic National Wildlife Refuge 101 12th Ave Room 236 Box 20 Fairbanks AK 99708

6 3865 Potter Road Fairbanks AK 997097LGL Alaska Research Associates Inc 1101 E 76th Avenue Suite B Anchorage AK 99518

8National Environmental Research Institute Department of the Arctic Environment PO Box 358 DK-4000Roskilde Denmark

9Ecology Evolution and Conservation Biology Program University of Nevada Reno Reno NV 8950110Prairie and Northern Wildlife Centre Canadian Wildlife Service 115 Perimeter Road Saskatoon

Saskatchewan S7N OX4 Canada11US Fish and Wildlife Service PO Box 346 Bethel AK 99559

12Alaska Cooperative Fish and Wildlife Research Unit Dept of Biology and Wildlife 209 Irving I BldgUniversity of Alaska Fairbanks AK 99775

Abstract We modeled the relationship between egg flotation and age of a developingembryo for 24 species of shorebirds For 21 species we used regression analyses toestimate hatching date by modeling egg angle and float height measured as continuousvariables against embryo age For eggs early in incubation we used linear regressionanalyses to predict hatching date from logit-transformed egg angles only For lateincubation we used multiple regression analyses to predict hatching date from both eggangles and float heights In 30 of 36 cases these equations estimated hatching date towithin four days of the true hatching date for each species After controlling for incubationduration and egg size flotation patterns did not differ between shorebirds grouped bymass ($100 g or 100 g) or taxonomy (Scolopacidae versus Charadriidae) Flotationprogressed more rapidly in species in which both adults incubate the clutch versus speciesin which only one adult incubates the clutch although this did not affect predictionaccuracy We also pooled all continuous data and created a generalized regressionequation that can be applied to all shorebird species For the remaining three species weestimated hatching date using five float categories Estimates of hatching date usingcategorical data were overall less accurate than those generated using continuous data(by 3ndash5 of a given incubation period) Our equations were less accurate than resultsreported in similar studies data collected by multiple observers and at multiple sites aswell as low sample sizes for some species likely increased measurement error To minimizeflotation method prediction error we recommend sampling in early incubation collectingboth egg angle and float height data in late incubation and developing site- and species-specific regression models where possible

Key words Charadriidae embryo age hatching date Scolopacidae

Evaluacion del Desarrollo de los Huevos de Aves Playeras Usando el Metodo de FlotacionModelos de Regresion Especie-Especıficos y Generalizados

Resumen Modelamos la relacion entre la flotacion del huevo y la edad de un embrionen desarrollo para 24 especies de aves playeras Para 21 especies usamos analisis de

Manuscript received 17 January 2006 accepted 28 September 200613 Corresponding author E-mail jliebezeitwcsorg14 Present address NASA Johnson Space Center 2101 NASA Parkway Mail Code SA5 Houston TX

77058

The Condor 10932ndash47 The Cooper Ornithological Society 2007

[32]

regresion para estimar la fecha de eclosion modelando el angulo del huevo y la altura deflotacion medidos como variables continuas contra la edad del embrion Para huevos alinicio del perıodo de incubacion usamos analisis de regresion lineal para predecir la fechade eclosion solo a partir de la transformacion logit de los angulos de los huevos Para laparte final del perıodo de incubacion usamos analisis de regresion multiple para predecirla fecha de eclosion a partir de los angulos de los huevos y de la altura de flotacion En 30de los 36 casos estas ecuaciones estimaron la fecha de eclosion con una exactitud decuatro dıas alrededor de la verdadera fecha de eclosion para cada una de las especiesconsideradas Luego de ajustar los modelos considerando la duracion de la incubacion y eltamano del huevo los patrones de flotacion no difirieron entre las aves playeras agrupadaspor peso ($100 g o 100 g) o taxonomıa (Scolopacidae versus Charadriidae) Laflotacion avanzo mas rapidamente en las especies en las cuales ambos adultos incuban lanidada que en las que solo un adulto incubo la nidada aunque esto no afecto la exactitudde la prediccion Tambien combinamos todos los datos continuos y creamos una ecuacionde regresion generalizada que puede ser aplicada a todas las especies de aves playeras Paralas tres especies restantes estimamos la fecha de eclosion usando cinco categorıas deflotacion En terminos generales las estimaciones de la fecha de eclosion usando datoscategoricos fueron menos exactas que aquellas generadas usando datos continuos (de 3a 5 para un perıodo de incubacion dado) Nuestras ecuaciones fueron menos exactas quelos resultados presentados por otros estudios Los datos recolectados por multiplesobservadores y en multiples sitios y el tamano de muestra reducido para algunas especiesprobablemente incrementaron el error de medicion Para minimizar el error de predicciondel metodo de flotacion recomendamos muestrear en el perıodo temprano de incubacionrecolectar datos del angulo del huevo y de la altura de flotacion al final de la incubacion ydesarrollar modelos de regresion especıficos para los sitios y las especies cuando seaposible

INTRODUCTION

Determining the stage of incubation of eggs inactive nests is an important component in manyavian studies where nest initiation or hatchingdates are of interest Knowledge of such datescan improve the accuracy of nest survivalestimates help to ascertain nest fate and reducethe need for frequent nest visits Minimizingdisturbance to incubating adults can reduce thechance of observer-related nest failure Esti-mates of embryo age are also useful in decidingwhen to capture incubating adults and nestlings(especially nidifugous young of shorebirds)establishing embryo age in toxicological andegg-swapping experiments and schedulingmanagement actions (eg mowing or tilling offields) to avoid harm to nesting birds (Grant1996 Brua and Machin 2000)

For more than half a century biologists haveused a variety of techniques to determine theage of developing embryos in eggs Most ofthese methods rely on the fact that eggs tend tolose mass at a relatively constant rate duringincubation to evaporative water loss andrespiration by the developing embryo (Wester-skov 1950 Ar and Rahn 1980) In additionevaporation rates per unit of egg surface aresimilar across different species with disparateegg sizes (Drent 1970) In some cases variableenvironmental or biotic conditions may in-

fluence embryo development making estimatesof embryo age difficult (Nol and Blokpoel1983) In general however rates of loss of eggmass are highly correlated with the develop-mental stage of the embryo

Sacrificing eggs so that embryos can beviewed is clearly the most accurate method ofdetermining incubation stage (Fant 1957)although this is not a practical alternative whenstudying threatened and endangered species orwhen there is a desire to estimate natural levelsof hatching success Estimating egg density canalso be used to age embryos although thistechnique requires very accurate measurementof weight which can be difficult under fieldconditions (OrsquoMalley and Evans 1980 Grant1996 Morrison and Hobson 2004) A thirdmethod of aging eggs is to shine a strong lightthrough the shell and observe the contentsWith this method the development of theembryo and the size of the air cell indicateembryo age (Hanson 1954 Weller 1956)However this candling technique does notwork well for aging embryos in all stages ofdevelopment or for observing contents wheneggs are darkly colored mottled or thick-shelled (Wooler and Dunlop 1980 Brua andMachin 2000)

Flotation of eggs one of the most commonmethods employed for estimating embryo de-

AGING SHOREBIRD EGGS USING THE FLOTATION METHOD 33

velopment relies on the fact that as an embryodevelops the specific gravity of the egg changesfrom greater than to less than that of water (ie1 g1 ml) Newly laid eggs sink to the bottom ofa column of water and as the embryos developeggs tip upward and eventually float on thesurface One can quantify this development bymeasuring (1) the angle between the horizontalplane and the longitudinal axis of the egg(2) the location of the egg within the watercolumn and (3) if the egg breaks the surfacethe height at which it floats Of all the methodsof aging eggs flotation has been the mostwidely employed with information availablefor at least seven families of birds (Hays andLeCroy 1971 Dunn et al 1979 Carroll 1988Fisher and Swengel 1991 Custer et al 1992Sandercock 1998 Brua and Machin 2000)

Within the shorebirds (Suborder Charadrii)egg flotation data have been published forrelatively few species (van Paassen et al 1984Sandercock 1998 Mabee et al 2006) Becauseof this paucity of information in the literaturewe were motivated to combine our existingindividual flotation data sets to conduct com-prehensive analyses that would not only pro-vide embryo age estimates for additionalshorebird species but also evaluate the overallefficacy of this method in accurately determin-ing hatching date In this paper we presentspecies-specific and generalized equations thatdescribe the relationship between embryo de-velopment and egg flotation by collating datafrom 24 shorebird species collected at 17 studysites These equations can be used to predictnest initiation or hatching date based on thefloat characteristics of an egg

Our specific objectives were to (1) developregression equations for estimating embryodevelopment in eggs of a variety of shorebirdspecies using the egg flotation technique(2) compare embryo development among spe-cies with different taxonomy incubation dura-tions and parental care and (3) summarize theerror associated with estimating hatching dateusing the flotation technique and recommendmethods to minimize this error

METHODS

STUDY SITES

We present egg flotation data from 24 shore-bird species collected at 17 study sites located

throughout the northern hemisphere (Fig 1)Latitude and longitude coordinates for eachstudy site are given in an Appendix publishedonline at 5httpwwwwcsorgmediafileLiebezeitetal2007_Appendixpdf4 Most sites(13 of 17) were located in arctic or subarctictundra in habitats ranging from lowlandmarshes to drier uplands The remaining siteswere at temperate latitudes and consisted ofrocky intertidal shorelines prairie uplandsshallow alkaline wetlands or mudflats

FIELD METHODS

We determined the float characteristics of eggsby individually immersing them in water ina transparent wide-mouthed container Weused lukewarm water from a thermos or watertaken directly from ponds or lakes We usuallyfloated two or more eggs from each clutch andwe floated some clutches several times through-out the breeding season Generally observersused a protractor to measure the angle betweenthe horizontal plane of the water and thelongitudinal axis of the egg (hereafter lsquolsquoegganglersquorsquo typically measured to within 65u)although a few experienced observers estimatedthis angle visually We also measured thevertical distance an egg floated above thesurface of the water (hereafter lsquolsquofloat heightrsquorsquotypically measured to 61 mm) using a clearplastic ruler For most species egg angle andfloat height were measured as continuous dataHowever for three species observers classifiedegg flotation parameters into categories(Fig 2) For eight species egg angle measure-ments were collected as continuous data whilefloat height measurements were collected ascategorical data We did not uniquely identifyeggs or examine changes in flotation scores ofindividual eggs over time

INDIVIDUAL SPECIES ANALYSES

To determine the relationship between eggflotation characteristics and embryo age weused data from nests where the date of clutchcompletion or hatching was known Nests werepresumed to have hatched on the date whennewly hatched chicks were observed either in orwithin a few meters of the nest or within 24ndash72 hr after star-cracked and pipped eggs wereobserved in the nest (exact number of hoursbased on shell-breaking and emergence timesreported in Poole et al [2003]) For nests with

34 JOSEPH R LIEBEZEIT ET AL

a known hatching date we estimated embryoage at the time the eggs were floated by back-calculating from the date of hatching (labelledas lsquolsquodays until hatchingrsquorsquo in all figures) Fornests with a known clutch completion dateonly we estimated the days until hatching usingsite-specific incubation durations if available ormore generic incubation durations provided inthe literature (Poole et al 2003) We determinedclutch completion dates for nests discoveredduring laying by revisiting them daily untilclutches were complete or by assuming birdslaid one egg per day until the standard clutchsize for that species was reached (typically foureggs) We defined the first day of incubation asthe date the final egg of the clutch was laidWhen more than one egg per clutch was floatedon a given visit we used the average angle andheight as the sample unit We did not use egg

angle data from nests with fewer than threeeggs as these nests could represent incompleteor partially depredated nests and thus bias ouranalyses We discarded data from individualeggs within a clutch that had abnormal scoresrelative to the rest of the clutch (1 ofobservations) In most cases these eggs wereeither infertile or had a dead embryo inside

We developed species-specific equations topredict embryo development and thus antici-pate the day eggs would hatch To do this wefirst pooled data from different study popula-tions for a given species two study populationswere pooled for three species and betweenthree and seven populations were pooled fornine species The remaining 12 species had datafrom only one site (Appendix) Second wepooled data across years (range 2ndash13 years) forall species For our analyses we used float data

FIGURE 1 Geographic locations of study sites where shorebird egg flotation data were collected between1993 and 2006 A 5 Kuparuk Oilfield Alaska USA B 5 Teshekpuk Lake Alaska USA C D 5 PrudhoeBay Alaska USA E 5 Barrow Alaska USA F 5 Pt Thomson Alaska USA G 5 Arctic National WildlifeRefuge Alaska USA H 5 Medusa Bay Taimyr Peninsula Siberia Russia I 5 Zackenberg Greenland J 5Big Quill Saskatchewan Canada K 5 Honey Lake California USA L 5 Kenai Fjord National ParkAlaska USA M 5 Brooks Alberta Canada N 5 Cape Espenberg Alaska USA O 5 Yukon DeltaNational Wildlife Refuge Alaska USA P 5 NE Enontekio Finland Q 5 Great Plain West Baffin IsCanada Latitude and longitude coordinates for each study site are available at 5httpwwwwcsorgmediafileLiebezeitetal2007_Appendixpdf4


from a single nest on a given day as our sampleunit Consequently we did not restrict our datato a single observation per clutch per seasonWe accepted a low level of pseudoreplicationbecause sample sizes of nests were sometimessmall and we wanted to explore generalpatterns in egg flotation across a large rangeof species

Continuous data We analyzed the data forearly (ie sinking eggs) and late (ie floatingeggs) incubation separately When individualeggs in the same clutch both sank and floatedwe classified the nest as being in late incubationAs we were interested in predicting hatchingdate (a measure of embryo development orage) our models are based on regressions ofembryo development on float characteristicsie we treated embryo age as the dependentvariable and float characteristics as the in-dependent variables Although this methodessentially requires rotating the axes by 90uthe resulting regression parameters have stan-dard errors describing variance in hatching dateas opposed to flotation characteristics More-

over this avoids under- or overestimating therate of development for individual nests ina data set where development may progressdifferently among nests (as seen in aging studiesusing molt scores Pimm 1976 Ginn andMelville 1983)

We employed linear regression with logit-transformed egg angles to capture the relation-ship between embryo development and eggangle We could not use logistic regression inthese analyses because the relationship betweenvariables was not sigmoidal when consideringembryo age as the independent variable Be-cause of their pyriform profile unincubatedshorebird eggs lie at the bottom of the watercontainer at a minimum angle of 20u Similarlyan egg can only rotate to a vertical position or90u as the embryo develops Thus 20u and 90uwere the lowest and highest egg angles thatcould be measured during our study To trans-form the egg angle measurements to logits wefirst converted them to proportions (P) byassigning a value of 0 and 1 to 20u and 90urespectively and interpolating the values for

FIGURE 2 Egg angle (solid line) and float height (dashed line) regressions using continuous data pooledfrom 21 (egg angle) and 20 (float height) shorebird species Illustrations of how the position of an egg in a floatcontainer changes through time are presented above the graph The numbers at the top of the figure refer tothe five combined egg anglendashfloat height categories


intermediate angles Eggs with observed anglesof 20u and 90u were first adjusted to 21u and 89uto avoid proportional angles of exactly 0 and 1which cannot be logit-transformed Proportion-al angles were transformed to logits using theformula

logit P ~ ln P= 1 Peth THORNfrac12

We then employed linear regression to predictthe number of days to hatching using thesetransformed egg angles

For eggs in late incubation (ie eggs thatfloated above the water surface) we usedmultiple regression to predict the number ofdays to hatching using two factors egg angleand float height For these analyses we useduntransformed data as there was no evidenceof a nonlinear relationship between embryo ageand either egg angle or float height during thelatter half of incubation Six species wereexcluded from the late-incubation single speciesanalyses because of low sample sizes (n 10Appendix)

Categorical data For species for which floatdata were collected categorically we firststandardized the data into five egg angle andfloat height categories as described in Alberico(1995 Fig 2) We related float category toembryo age graphically and generated a meanage for each float category to be used inpredictions

GROUP COMPARISONS

To combine data across species we firstgenerated a mean egg angle or float height perspecies per lsquolsquodayrsquorsquo (based on days until hatching)Using mean values reduced the influence ofmeasuring the same nest more than once withoutrestricting our coverage of the incubation periodBecause incubation duration varies widely with-in and among species due to factors such as eggsize breeding locale and parental behavior (Nol1986 Schamel and Tracy 1987) we standardizedincubation duration by expressing embryo age asproportion of the incubation period completedThis allowed us to compare relative differencesin the rate of embryonic development betweengroups of species that had different life historycharacteristics We also attempted to standard-ize float height scores across species by dividingthe float height of a given egg by that speciesrsquoaverage egg dimensions (ie length + width2

species values reported by Schonwetter [1967]Schekkerman et al [2004]) We found that thisdid not improve model fit and consequentlyreport only the results for the untransformedfloat heights

We developed and compared regressionmodels for species grouped by taxonomy(Scolopacidae vs Charadriidae after Thomaset al [2004]) by whether both adults shareincubation or only one adult incubates the eggs(hereafter lsquolsquobiparentalrsquorsquo vs lsquolsquouniparentalrsquorsquo spe-cies after Larsen et al [1996]) and by adultbody mass (large vs small) We categorizedlsquolsquosmallrsquorsquo shorebirds as 100 g and lsquolsquolargersquorsquoshorebirds as $100 g because the mean massof shorebirds is roughly 99 g (Larsen et al1996) We tested for significant differencesamong these regression models by determiningwhether the 95 confidence intervals of theregression coefficients overlapped

PREDICTIVE ABILITY AND ERROR

We estimated the predictive ability of ourregression equations by subtracting the actualembryo age from the predicted age for each neston a given day We generated the predicteddays until hatching for each nest by insertingthe transformed egg angle or float height datainto the appropriate regression models Becausethe regression equations represent an average ofall data there were a few instances where thepredicted hatching date for nests were pasthatching (ie days until hatching were 21)This occurred rarely and only when eggs floatedunusually high In these cases we set the daysuntil hatching as 0 For the categorical data weassigned the predicted age of an egg in a givencategory as the average of all known-age eggsthat were classified in that same category Thenas with the continuous data we subtracted theactual age from the predicted age to generateestimates of bias

For both the continuous and categorical datasets we generated the absolute mean deviation6 SE as a descriptive statistic for the error Theabsolute value reflects the amount of uncer-tainty (ie both under- and overestimation)researchers will have to consider when decidinghow confident they should be in using flotationscores Similarly to (van Paassen et al 1984) wereport 90th percentiles of error to gauge theaccuracy of our samples in age prediction Ifour samples are representative of the total


population researchers can expect 90 of theirnests to be aged with error less than or equal tothe reported values For categorical data wegenerated separate estimates of error for each ofthe five float categories

For figures depicting regressions for individ-ual species we placed days until hatching on thex-axis for ease of use even though the re-gression equations were calculated with embryodevelopment as the dependent variable Be-

TABLE 1 Sample size and coefficients for linear regressions (r2) of embryo age versus egg angle (earlyincubation age 5 a + blogit of proportional egg angle) and multiple regressions (R2) of embryo age versusegg angle and float height (late incubation age 5 a + bfloat height in mm + cegg angle in u) for 21 shorebirdspecies for which continuous data were collected Embryo age is measured as lsquolsquodays until hatchingrsquorsquo whichrefers to the absolute number of days for single species analyses and the proportion of the incubation periodfor all species combined

Species

Early incubation (sinking eggs) Late incubation (floating eggs)

na r2 a b na R2 a b c

Black-bellied Plover(Pluvialis squatarola)

15 076 22047 143 17 066 2460 225 2013

American Golden-Plover(Pluvialis dominica)

38 075 22017 141 51 069 21061 142 2004

Pacific Golden-Plover(Pluvialis fulva)b

11 079 22060 117 1 ndash ndash ndash ndash

Common Ringed Plover(Charadrius hiaticula)b


Piping Plover (Charadriusmelodus circumcinctus)b


Willet (Tringa semipalmatainornatus)b


Marbled Godwit(Limosa fedoa fedoa)b


Ruddy Turnstone (Arenariainterpres interpres)

14 073 21617 125 27 071 21645 336 003

Sanderling (Calidris alba) 17 080 21818 079 12 087 21569 364 001Semipalmated Sandpiper

(Calidris pusilla)69 059 21650 084 187 063 2522 147 2008

Western Sandpiper(Calidris mauri)

24 077 21727 092 23 081 831 170 2027

Little Stint (Calidris minuta) 50 049 21829 074 52 068 21422 239 c

Temminckrsquos Stint(Calidris temminckii)

42 059 21642 084 10 016 1650 23150 2036

Pectoral Sandpiper(Calidris melanotos)

57 044 21747 082 92 041 2729 123 2006

Dunlin (Calidris alpina) 141 066 21750 088 45 056 028 132 2015Curlew Sandpiper

(Calidris ferruginea)b


Stilt Sandpiper(Calidris himantopus)

10 050 21615 055 13 075 2616 119 2007

Buff-breasted Sandpiper(Tryngites subruficollis)

14 069 21996 168 32 080 2787 202 2008

Long-billed Dowitcher(Limnodromus scolopaceus)

36 061 21715 144 22 053 2466 106 2008

Red-necked Phalarope(Phalaropus lobatus)

76 059 21661 108 44 050 21299 251 002

Red Phalarope(Phalaropus fulicaria)

53 036 21529 073 54 056 2303 141 2010

All species 265 070 021 005 214 065 079 007 000

a For single species analyses n equals the number of float values across all nests and days A float value foreach nest on a given day was obtained by calculating an average score for all eggs floated All eggs in a nesthowever could be floated and scored on multiple occasions For the lsquolsquoall speciesrsquorsquo group n is the number offloat values after generating a mean value for each day (day until hatching) for each species

b Equations were not developed for these species during late incubation because of inadequate sample sizesc Equation for this species included only float height data as egg angle information was not recorded


cause multiple regression equations are difficultto display graphically we illustrate only therelationship between hatching date and floatheight (ie egg angle is not used) Bothparameters were used however in the equa-tions (Table 1) because doing so increased thepredictive power of the equations in seven of 14species that were analyzed using both methodsFor all figures the scale of the x-axis corre-sponds to the maximum incubation period fora given species based on mean literature values(Poole et al 2003) and the range of our data

SPSS 107 (SPSS 1999) was used for allanalyses Results are reported as x 6 SE andP-values 005 were considered significantunless otherwise noted

RESULTS

INDIVIDUAL SPECIES CONTINUOUS DATA

Early incubation Linear regressions describingthe relationship between embryo development

and logit-transformed egg angle for nests inearly incubation were generated for 21 speciesRegression coefficients (r2) were $080 for onespecies between 060 and 079 for 11 speciesand below 060 for the remaining nine species(Table 1) Of the 21 species three had 90th

percentile error values within 62 days and allother species had values of 62ndash4 days exceptthe Willet (Tringa semipalmata 654 daysTable 2) When all species were analyzedtogether the 90th percentile of the predictiveerror was equal to 15 of the incubation period(ie within three days of the actual nest age90 of the time if a species had a 20-dayincubation period Table 2)

Late incubation Multiple regression analysesrelating embryo development to egg angle andfloat height for nests in late incubation weregenerated for 14 species (Fig 3) For oneadditional species (Little Stint [Calidris min-uta]) we used linear rather than multipleregression to relate embryo development to

TABLE 2 Error statistics associated with aging eggs by floating them during early incubation for 21shorebird species Estimates of prediction error for days until hatching are for linear regressions of embryo ageversus the angle an egg floated in water (logit-transformed)

Species naRange of error

(days)bAbsolute mean deviation

6 SE (days)90th percentile of absolute

error (days)

Black-bellied Plover 15 22 4 12 6 03 20American Golden-Plover 38 25 3 15 6 02 26Pacific Golden-Plover 11 23 2 12 6 02 17Common Ringed Plover 14 24 3 19 6 02 27Piping Plover 67 24 5 16 6 01 34Willet 44 28 6 23 6 03 54Marbled Godwit 19 26 4 23 6 04 37Ruddy Turnstone 14 24 3 14 6 03 24Sanderling 17 22 2 09 6 01 15Semipalmated Sandpiper 141 27 4 12 6 01 25Western Sandpiper 24 22 2 11 6 01 17Little Stint 50 26 5 15 6 02 31Temminckrsquos Stint 42 26 3 14 6 02 30Pectoral Sandpiper 57 27 6 16 6 02 37Dunlin 141 24 3 12 6 01 24Curlew Sandpiper 32 25 4 18 6 15 38Stilt Sandpiper 10 22 3 13 6 03 25Buff-breasted Sandpiper 14 22 4 12 6 03 20Long-billed Dowitcher 36 23 6 13 6 02 31Red-necked Phalarope 76 25 4 19 6 01 38Red Phalarope 53 25 5 17 6 02 35All speciesc 265 2025 026 007 6 000 015

a For single species analyses n is the number of float values across all nests and days A float value for eachnest was obtained on a given day by calculating an average score for all eggs floated All eggs in a nesthowever could be floated and scored on multiple occasions For the lsquolsquoall speciesrsquorsquo group n is the number offloat values after generating a mean value for each day (day until hatching) for each species

b A negative value indicates an underestimate of agec The unit of error for the lsquolsquoall speciesrsquorsquo group is lsquolsquoproportion of the incubation periodrsquorsquo


float height as egg angles were not collected forthis species during late incubation R2 valueswere $080 for three species between 060 and079 for six species and below 060 for theremaining six (Table 1) Of the 15 species 10had 90th percentile error values of 62ndash4 daysand the remaining five species exceeded64 days (up to 66 days Table 3) When allspecies were analyzed together the 90th percen-tile of the predictive error was equal to 17 ofthe incubation period (ie within 34 days ofthe actual age 90 of the time if a species hada 20-day incubation period Table 3)

INDIVIDUAL SPECIES CATEGORICAL DATA

Of the three species for which categorical datawere collected all showed a strong positiverelationship between float category and days tohatching (Fig 4) Of the three species and fivefloat stages the 90th percentile of error for allpredictions fell within 62 days of the true agefor four species and category combinations(27 n 5 15 Table 4) All but two of theremaining species and category combinationshad 90 of their predictions within 64 days ofthe true age There was a significant difference

FIGURE 3 Linear regressions depicting the relationship between (1) days to hatching and logit-transformedproportional egg angle for nests in early incubation (solid lines) and (2) days to hatching and float height fornests in late incubation (dashed lines) See Figure 1 for study area locations and Table 1 for regressionequations for each species


among the five categories in the accuracy of ourestimates (x2

4 5 160 P 5 0003) with the errorlowest in float categories 2 and 3 and highest incategory 5 Category 5 likely had higher errorbecause it covered the largest portion of theincubation period When all species wereanalyzed together the 90th percentile of thepredictive error was equal to 20 of theincubation period (ie within 4 days of theactual age 90 of the time if a species had a 20-day incubation period Table 4)

GROUP COMPARISONS

There was overlap between the confidenceintervals of linear (ie early incubation) ormultiple (ie late incubation) regression pa-rameters for lsquolsquolargersquorsquo and lsquolsquosmallrsquorsquo shorebirdsand species belonging to the Scolopacidae andCharadriidae families indicating no significantdifferences between these groups However wefound that the slope of the linear regression ofembryo age versus the logit-transformed eggangles was steeper and the multiple regression

float height coefficient of embryo age versusfloat height and egg angle was higher foruniparental versus biparental species (anglecoefficient uniparental 5 007 6 001 bi-parental 5 005 6 000 slope coefficientuniparental 5 010 6 001 biparental 5 0066 001) After standardizing for incubationduration these results suggest that both eggangle and float height change less rapidly withembryo age in uniparental versus biparentalspecies Despite these differences in the modelsregression equations specific to uniparental andbiparental species reduced the 90th percentiles ofprediction error by a maximum of 2 of theincubation period or about 04 days

FACTORS AFFECTING HATCHINGDATE ESTIMATION

Estimates of hatching date are affected byfactors that increase measurement error orincrease the variation in embryo developmentIn this study measurement error was likelyincreased by collating data from different

FIGURE 3 Continued


locations as each location had different ob-servers and associated subtle differences in howeggs were floated To assess this influence wecompared prediction error values for speciessampled at a single site and species sampled atmultiple sites During early incubation we didnot see an increase in measurement error withthe number of sites sampled (mean 90th

percentile error values weighted by sample sizewere 35 vs 29 for species sampled at single vsmultiple sites respectively Table 2 Appendix)However in late incubation the mean 90th

percentile error value (weighted by sample size)for prediction error was lower for speciessampled at a single site (31) compared to thatof species sampled at five or more sites (43Table 3 Appendix) Some of this error may beattributed to lower sample sizes at the multiplesites (x lt 12 nests per species per site frommultiple sites vs 30 nests per species at singlesites) No such comparison is possible for the

categorical data because all species were sam-pled at a single site

To indirectly evaluate the effect of interan-nual variation in environmental conditions weconducted an analysis to determine how thenumber of years that data were collected fora species affected our error estimates Theaverage 90th percentile error value for speciessampled in multiple years at a single site was 31for early incubation and 29 for late incubationBecause there was only one species sampled ata single year and site we could not generate anyvalues to directly compare these numbersHowever these two values are lower than thelsquolsquoall speciesrsquorsquo 90th percentile errors of 35 and38 days for the early and late incubationperiods respectively (generated using the lsquolsquoallspeciesrsquorsquo proportion errors of 15 and 17 andassuming a mean incubation length of237 days for species sampled during earlyincubation [ie sinking eggs] and 225 days

TABLE 3 Error statistics associated with aging eggs by floating them during late incubation for 15shorebird species Estimates of prediction error for days to hatching are for multiple regressions of embryo ageversus the height and angle an egg floated


(days)bAbsolute mean

deviation 6 SE (days)90th percentile of

absolute error (days)

Black-bellied Plover 17 25 6 19 6 04 43American Golden-Plover 51 28 4 18 6 02 39Ruddy Turnstone 27 28 5 17 6 03 32Sanderling 12 23 2 11 6 02 21Semipalmated Sandpiper 187 27 6 19 6 01 38Western Sandpiper 23 23 4 16 6 02 30Little Stintc 52 26 4 17 6 02 33Temminckrsquos Stint 10 24 5 29 6 05 42Pectoral Sandpiper 92 28 9 24 6 02 46Dunlin 45 25 5 17 6 02 34Stilt Sandpiper 13 22 3 12 6 03 24Buff-breasted Sandpiper 32 24 4 17 6 02 28Long-billed Dowitcher 22 25 3 15 6 03 31Red-necked Phalarope 44 26 9 24 6 03 45Red Phalarope 54 29 10 38 6 03 66All speciesde 214 2028 025 008 6 000 017

a For single species analyses n is the number of float values across all nests and days A float value for eachnest on a given day was obtained by calculating an average score for all eggs floated All eggs in a nesthowever could be floated and scored on multiple occasions For the lsquolsquoall speciesrsquorsquo group n is the number offloat values after generating a mean value for each day (day until hatching) for each species

b A negative value indicates an underestimate of agec The equation for this species included only float height data as egg angle information was not recordedd The unit of error for the lsquolsquoall speciesrsquorsquo group is lsquolsquoproportion of the incubation periodrsquorsquoe In addition to the 15 species listed in the table the lsquolsquoall speciesrsquorsquo group includes data from six species for

which sample sizes were too small to generate individual species regression equations (Common RingedPlover Curlew Sandpiper Marbled Godwit Pacific Golden-Plover Piping Plover and Willet see Table 1 andthe Appendix [available online at 5httpwwwwcsorgmediafileLiebezeitetal2007_Appendixpdf4] forsample sizes)


for species sampled during late incubation [iefloating eggs]) This suggests that variation inenvironmental conditions among years has littleeffect on estimates of embryo age

We also investigated how the presence ofsubspecies affected measurement error Wesampled four subspecies of Dunlin (Calidrisalpina) and their 90th percentile error value wassimilar to that of other species with only a singlerace (early incubation 24 versus an average of29 for 20 other species late incubation 34versus an average of 37 for 14 other species)This suggests that potential life history differ-ences (eg incubation duration) among thefour races of Dunlin did not increase the errorassociated with estimating hatching date

DISCUSSION

Using continuous and categorical egg angle andfloat height data we created species-specificflotation relationships that allowed us to predictdays until hatching for early and late stages ofincubation for 24 species of shorebirds Theserelationships provided reasonable predictivepower with the most accurate predictions fromcontinuous data for early incubation followedby continuous data for late incubation andfinally categorical data throughout incubationOther studies have also found that egg angletends to be more reliable at predicting hatchingdate during early compared to late incubation(van Paassen et al 1984) This greater accuracy

TABLE 4 Error statistics associated with aging eggs by assigning them to five float categories for threespecies for which categorical data were collected Estimates of prediction error for days to hatching weregenerated by comparing the predicted age of an embryo in a given float category (calculated as the average ofall known-age nests that were classified in that category) with the true age of the embryo based on nestinitiation or hatching See Figure 2 for illustrations of an eggrsquos position during each float category

SpeciesaFloat

category nbRange of error

(days)cAbsolute mean



Black Oystercatcher 1 7 220 40 19 6 05 28(Haematopus bachmani) 2 10 240 30 16 6 05 31

3 9 240 20 18 6 04 324 2 210 10 10 6 00 105 10 215 25 11 6 02 15

All 38 240 40 15 6 02 30

Black-necked Stilt 1 9 210 10 02 6 02 10(Himantopus mexicanus)d 2 6 220 40 15 6 06 30

3 4 230 20 15 6 08 274 3 225 15 18 6 04 235 ndash ndash ndash ndash

All 22 230 40 10 6 03 25

American Avocet 1 42 260 20 08 6 02 20(Recurvirostra americana) 2 58 245 65 15 6 02 45

3 37 260 30 15 6 02 304 18 295 55 25 6 05 485 5 230 30 18 6 06 30

All 160 295 65 14 6 01 31

All species 1 58 2020 019 008 6 001 0152 74 2025 029 007 6 001 0203 50 2038 020 011 6 001 0204 23 2030 024 011 6 002 0195 15 2010 011 004 6 001 009

All 220 2038 029 008 6 000 020

a Data are listed as number of days for each species and proportion of the incubation period for all speciescombined

b For single species analyses n is the number of float values across all nests and days A float value for eachnest on a given day was obtained by calculating an average score for all eggs floated All eggs in a nesthowever could be floated and scored on multiple occasions For the multispecies analyses n is an average ofthe average float values for each species for each proportion of the incubation period

c A negative value indicates an underestimate of aged There were no data available for Black-necked Stilt category five


may be due to rapid changes in the angle of theegg that occur early in incubation Also accuratefloat data can be collected more easily at thistime because the egg touches the bottom of thefloat container and thus does not move aroundduring measurement

We also showed that taxonomy (Charadrii-dae versus Scolopacidae) and body mass of theshorebird (small versus large) had little effect onthe relationship between hatching date and eggangle or float height However we did find thategg angle and float height changed morerapidly for species with biparental incubationcompared to species with uniparental incuba-tion after controlling for differences in in-cubation duration This observation makesintuitive sense since biparental incubationtypically means that eggs are incubated more

per unit time (Larsen et al 1996) Despite thisdifference there was little benefit in applyingseparate regression models in terms of pre-diction accuracy These results allowed us togenerate a standardized equation to predicthatching date from egg angle and float heightmeasurements that can be applied to almost anyshorebird species As an aid to predictinghatching date for shorebird species not includedin our single-species analyses an interactiveversion of Figure 2 is available at 5httpwww2dmudk1_Viden2_Miljoe-tilstand3_naturbiobasismanualasp4 The availability ofthese equations is timely since the need todetermine the incubation age of embryos willlikely increase as researchers begin includingembryo age (or lsquolsquonest agersquorsquo) as a covariate innest survivorship models (Dinsmore et al 2002Jehle et al 2004 Nur et al 2004)

INDIVIDUAL SPECIES CONTINUOUS ANDCATEGORICAL DATA

There were several advantages to using bothlinear and multiple regression to analyze thecontinuous egg angle and float height dataFirst the approach allowed us to predict eggangles and float heights for periods of timewhen no raw data were available This is incontrast to earlier studies that classified flota-tion data into categories or provided a scatterplot of egg angle and float height relative toembryo development (sometimes with confi-dence intervals around points with more thanone observation Dunn et al 1979 Custer et al1992 Brua and Machin 2000) Second theapproach provided an objective method forpredicting and estimating error around hatch-ing dates such statistical assessments of anembryo aging method have seldom been done(but see van Paassen et al 1984 Sandercock1998) Finally this approach allowed us toassess how well regression equations mightapply to other species based on the degree ofoverlap of their respective confidence intervalsWe believe this is the first attempt to test fordifferences among species or species groupingswith regard to flotation equations

The categorical data provided a relatively lowlevel of accuracy for estimating hatching datescompared to the continuous data This result isnot surprising since categorical data dividesa continuous process into subjective classes(Walter and Rusch 1997) However in temper-

FIGURE 4 Relationship between days to hatchingand the combined egg angle and float heightcategories for three shorebird species See Figure 1for study area locations and Figure 2 for illustrationof the five float categories


ate and tropical areas where heat can causeasynchronous embryo development (Grant1982) estimates of hatching date from categor-ical data using the flotation method mayprovide comparable accuracy to estimates usingcontinuous data (JAR unpubl data) Categor-ical data may also be easier to replicateespecially when numerous observers are re-cording data

PREDICTIVE POWER

Our hatching date estimates tended to besomewhat less accurate than those of otherstudies (typically within 2ndash3 days van Paassenet al 1984 Sandercock 1998) The greateraccuracy of other studies might be partiallyexplained by the fact that these researcherscollected their data independently and studiedincubation duration at one study site Assuggested above we suspect that data collectedfrom multiple study sites and relatively lowsample sizes for some species likely increasedthe measurement error in this study Howeverour analyses indicate it is unlikely that our errorvalues were inflated by collecting data overmany years (ie seasonal variation) or acrosssubspecies Schamel and Tracy (1987) foundthat both seasonal variation and latitudeaffected Red Phalarope (Phalaropus fulicaria)incubation duration We did not evaluatewhether data collected across large geographi-cal areas affected prediction accuracy How-ever the effect of geography is likely smallbecause information for each study species wascollated from a relatively narrow (8u) latitu-dinal range and we used site-specific incubationdurations when available

In addition to sampling variation species-specific sources of variation may affect pre-diction accuracy Nol and Blokpoel (1983)found flotation measurements to be highlyvariable for Ring-billed Gulls (Larus delawar-ensis) whereas other researchers have foundless variability in other study species (Carroll1988 Walter and Rusch 1997 Brua andMachin 2000) Prediction accuracy may alsobe influenced by (1) variability in egg size(Westerskov 1950) (2) behavioral differences inparental care that produce inconsistent incuba-tion durations (Nol and Blokpoel 1983 Feld-heim 1997) (3) physiological factors such asdifferences in egg pore area shell thickness Ar

and Rahn (1980) and variability in gasconductance (Visser et al 1995) and(4) environmental factors associated with nestmicroclimate (Romanoff 1934) The impact ofthese factors seems negligible howevergiven our inability to differentiate flotationpatterns based on shorebird body mass ortaxonomy

RECOMMENDATIONS

Researchers wishing to use the flotation methodmay implement additional measures to over-come some of the inaccuracy inherent inestimating the age of embryos First werecommend using information on both eggangle and float height when estimating hatchingdates Together these measurements quicklyindicate whether an egg has recently been laidor is near hatching Second eggs from the samenest should be floated multiple times tocorroborate findings from the initial measure-ment This will ameliorate the effects of in-advertently using misleading data from eggsthat were not fertilized or had embryos thatfailed to develop Third we suggest floatingeggs when the nest is initially found as thisincreases the chances of floating eggs duringearly incubation when estimates are generallymore accurate Fourth we advise visiting thenest on a daily basis starting 2ndash4 days prior tothe expected hatching date to check for signs ofhatching (ie star cracks or pip holes) Thisconservative approach will assist in confirmingnest fates when hatching dates are overesti-mated Fifth when practical we advise de-veloping site-specific float regression equationsfor study species rather than using our generalfloat equations such site-specific equationsremove the inaccuracies inherent in our datacaused by multiple observers recording data atmultiple sites and account for geographic andseasonal variation Finally in situations whereit is important to minimize disturbance time(eg for colonial birds where predators mighttake eggs after adults have been flushed) weadvise collecting categorical data as thismethod is quicker and as we have showncan provide a reasonable level of predictiveaccuracy

ACKNOWLEDGMENTS

We thank the many field technicians that helpedcollect the data especially C Ashling J Boadway


K Boadway T Booms M Burle L Coelho NavesN Coutsoubos M Denega R Dickson C Fitzpat-rick D Fontaine R Hill J Juillerat F LebouardA Leist D Morgan R Neudorf G North GNorwood D Pearson K Pietrzak S Rangen JSelvidge S Thomas M Swaim H Swensen ATaylor G Thomas B Trask and H Woodward JJohnson kindly created Figure 1 Funding or logis-tical support was provided by the Alberta NorthAmerican Waterfowl Management Plan Centre BPExploration (Alaska) Inc the Canadian WildlifeService Carleton University ConocoPhillips AlaskaInc Disney Wildlife Conservation Grants to theWildlife Conservation Society Ducks UnlimitedCanada Eastern Irrigation DistrictndashAlberta theEnvironmental Protection Agency (Denmark) theInterdepartmental Recovery Fund (Canada) JayDow Sr Wetlands National Audubon SocietyGeorge Whittell Nevada Environmental Fund theNatural Sciences and Engineering Research Councilof Canada Neotropical Migratory Bird Conserva-tion Act Grants to the Wildlife Conservation SocietyNevada Agricultural Experiment Station NorthSlope Borough Wildlife Department National Sci-ence Foundation Grant DEB-9196050 to LewisOring a National Science Foundation GraduateFellowship to JAR Point Thomson Oil ProductionUnit Owners (Alaska) the Canadian Polar Conti-nental Shelf Project National Biological Service(now part of the US Geological Survey) a USGeological Survey Natural Resources PreservationProgram grant the US Fish and Wildlife Servicethe Wildlife Conservation Society (including privatedonors) and the Zackenberg Research StationDanish Polar Center We also thank R Gill LOring and K Wohl for supporting our field studiesand B K Sandercock for his detailed review of ourmanuscript Order of authorship (after first three)was based on the number of species and number ofnests sampled

LITERATURE CITED

ALBERICO J A 1995 Floating eggs to estimateincubation stage does not affect hatchabilityWildlife Society Bulletin 23212ndash216

AR A AND H RAHN 1980 Water in the avian eggoverall budget of incubation American Zoolo-gist 20373ndash384

BRUA R B AND K L MACHIN 2000 Determiningand testing the accuracy of incubation stage ofRuddy Duck eggs by flotation Wildfowl 51181ndash189

CARROLL J P 1988 Egg-flotation to estimateincubation stage of Ring-necked PheasantsWildlife Society Bulletin 16327ndash329

CUSTER T W G W PENDLETON AND R WROACH 1992 Determination of hatching datefor eggs of Black-crowned Night-Herons SnowyEgrets and Great Egrets Journal of FieldOrnithology 63145ndash154

DINSMORE S J G C WHITE AND F L KNOPF2002 Advanced techniques for modeling aviannest survival Ecology 833476ndash3488

DRENT R M 1970 Estimating the error involved inusing egg density to predict laying dates Journalof Field Ornithology 58464ndash473

DUNN E H D J T HUSSELL AND R ERICKLEFS 1979 The determination of incuba-tion stage in starling eggs Bird-Banding 50114ndash120

FANT R J 1957 Criteria for ageing pheasantembryos Journal of Wildlife Management 21324ndash328

FELDHEIM C L 1997 The length of incubation inrelation to nest initiation date and clutch size indabbling ducks Condor 99997ndash1001

FISHER I J AND S R SWENGEL 1991 A guide foraging Sandhill Crane eggs Wildlife SocietyBulletin 19494ndash497

GINN H B AND D S MELVILLE 1983 Moult inbirds BTO guide 19 British Trust for Ornithol-ogy Tring UK

GRANT G S 1982 Avian incubation egg temper-ature nest humidity and behavioral thermoreg-ulation in a hot environment OrnithologicalMonographs 30

GRANT M C 1996 Predicting the hatching dates ofCurlew Numenius arquata clutches Wader StudyGroup Bulletin 8053ndash54

HANSON H C 1954 Criteria of age of incubatedMallard Wood Duck and Bob-white Quaileggs Auk 71267ndash272

HAYS H AND M LECROY 1971 Field criteria fordetermining incubation stage in eggs of theCommon Tern Wilson Bulletin 83425ndash429

JEHLE G A A YACKEL ADAMS J A SAVIDGEAND S K SKAGEN 2004 Nest survival estima-tion a review of alternatives to the Mayfieldestimator Condor 106472ndash484

LARSEN T T A SORDAHL AND I BYRKJEDAL1996 Factors related to aggressive nest pro-tection behaviour a comparative study ofHolarctic waders Biological Journal of theLinnean Society 58409ndash439

MABEE T J A M WILDMAN AND C B JOHNSON2006 Using egg flotation and eggshell evidenceto determine age and fate of arctic shorebirdnests Journal of Field Ornithology 77163ndash172

MORRISON R I G AND K A HOBSON 2004 Useof body stores in shorebirds after arrival on high-arctic breeding grounds Auk 121333ndash344

NOL E 1986 Incubation period and foragingtechnique in shorebirds American Naturalist128115ndash119

NOL E AND H BLOKPOEL 1983 Incubation periodof Ring-billed Gulls and the egg immersiontechnique Wilson Bulletin 95283ndash286

NUR N A L HOLMES AND G R GEUPEL 2004Use of survival time analysis to analyze nestingsuccess in birds an example using LoggerheadShrikes Condor 106457ndash471

OrsquoMALLEY J B AND R M EVANS 1980 Varia-tions in measurements among White Pelican eggsand their use as a hatch date predictor CanadianJournal of Zoology 58603ndash608

PIMM S 1976 Estimation of the duration of birdmolt Condor 78550


POOLE A F P STETTENHEIM and F B GILL

[EDS] 2003 The birds of North America lifehistories for the 21st century The Academy ofNatural Sciences Philadelphia PA and TheAmerican Ornithologistsrsquo Union WashingtonDC

ROMANOFF A L 1934 Study of artificial incubationof game birds Cornell University Agricul-ture Experimental Station Bulletin 6161ndash39

SANDERCOCK B K 1998 Chronology of nestingevents in Western and Semipalmated Sandpipersnear the Arctic Circle Journal of Field Orni-thology 69235ndash243

SCHAMEL D AND D M TRACY 1987 Latitudinaltrends in breeding Red Phalaropes Journal ofField Ornithology 58126ndash134

SCHEKKERMAN H I TULP K M CALF AND

J J DE LEEUW 2004 Studies on breed-ing shorebirds at Medusa Bay Taimyr insummer 2002 Alterra Wageningen The Nether-lands

SCHONWETTER M 1967 Handbuch der OologieVol 1 Akademi-Verlag Berlin

SPSS 1999 SPSS for Windows version 107 SPSSChicago

THOMAS G H M A WILLS AND T SZEKELY2004 A supertree approach to shorebird phy-logeny BMC Evolutionary Biology 428

VAN PAASSEN A G D H VELDMAN AND A JBEINTEMA 1984 A simple device for determi-nation of incubation stages in eggs Wildfowl35173ndash178

VISSER G H E C ZEINSTRA F VAN GASTERENAND A J BEINTEMA 1995 Gas conductanceand metabolism of shorebird eggs variationwithin and between species Respiration Physi-ology 99273ndash281

WALTER S E AND D H RUSCH 1997 Accuracyof egg flotation in determining age of CanadaGoose nests Wildlife Society Bulletin 25854ndash857

WELLER M W 1956 A simple field candler forwaterfowl eggs Journal of Wildlife Management20111ndash113

WESTERSKOV K 1950 Methods for determining theage of game bird eggs Journal of WildlifeManagement 1456ndash67

WOOLER R D AND J N DUNLOP 1980 The use ofsimple measurements to determine the age ofSilver Gull eggs Australian Wildlife Research7113ndash115


ASSESSING THE DEVELOPMENT OF SHOREBIRD EGGS USINGTHE FLOTATION METHOD SPECIES-SPECIFIC AND

GENERALIZED REGRESSION MODELS

JOSEPH R LIEBEZEIT113 PAUL A SMITH

2 RICHARD B LANCTOT3

HANS SCHEKKERMAN4 INGRID TULP

4 STEVE J KENDALL5 DIANE M TRACY

6ROBERT J RODRIGUES

7 HANS MELTOFTE8 JULIE A ROBINSON

914CHERI GRATTO-TREVOR

10 BRIAN J MCCAFFERY11

JULIE MORSE12 AND STEVE W ZACK

1

1Wildlife Conservation Society Pacific West Office 219 SW Stark St Suite 200 Portland OR 972042National Wildlife Research Centre Carleton University 1125 Colonel By Drive Ottawa ON K1A 0H3 Canada3US Fish and Wildlife Service Migratory Bird Management 1011 E Tudor Rd MS 201 Anchorage AK 99503

4Institute of Marine Resources and Ecosystem Studies PO Box 167 1790 AD Den Burg The Netherlands5US Fish and Wildlife Service Arctic National Wildlife Refuge 101 12th Ave Room 236 Box 20 Fairbanks AK 99708

6 3865 Potter Road Fairbanks AK 997097LGL Alaska Research Associates Inc 1101 E 76th Avenue Suite B Anchorage AK 99518

8National Environmental Research Institute Department of the Arctic Environment PO Box 358 DK-4000Roskilde Denmark

9Ecology Evolution and Conservation Biology Program University of Nevada Reno Reno NV 8950110Prairie and Northern Wildlife Centre Canadian Wildlife Service 115 Perimeter Road Saskatoon

Saskatchewan S7N OX4 Canada11US Fish and Wildlife Service PO Box 346 Bethel AK 99559

12Alaska Cooperative Fish and Wildlife Research Unit Dept of Biology and Wildlife 209 Irving I BldgUniversity of Alaska Fairbanks AK 99775

Abstract We modeled the relationship between egg flotation and age of a developingembryo for 24 species of shorebirds For 21 species we used regression analyses toestimate hatching date by modeling egg angle and float height measured as continuousvariables against embryo age For eggs early in incubation we used linear regressionanalyses to predict hatching date from logit-transformed egg angles only For lateincubation we used multiple regression analyses to predict hatching date from both eggangles and float heights In 30 of 36 cases these equations estimated hatching date towithin four days of the true hatching date for each species After controlling for incubationduration and egg size flotation patterns did not differ between shorebirds grouped bymass ($100 g or 100 g) or taxonomy (Scolopacidae versus Charadriidae) Flotationprogressed more rapidly in species in which both adults incubate the clutch versus speciesin which only one adult incubates the clutch although this did not affect predictionaccuracy We also pooled all continuous data and created a generalized regressionequation that can be applied to all shorebird species For the remaining three species weestimated hatching date using five float categories Estimates of hatching date usingcategorical data were overall less accurate than those generated using continuous data(by 3ndash5 of a given incubation period) Our equations were less accurate than resultsreported in similar studies data collected by multiple observers and at multiple sites aswell as low sample sizes for some species likely increased measurement error To minimizeflotation method prediction error we recommend sampling in early incubation collectingboth egg angle and float height data in late incubation and developing site- and species-specific regression models where possible

Key words Charadriidae embryo age hatching date Scolopacidae

Evaluacion del Desarrollo de los Huevos de Aves Playeras Usando el Metodo de FlotacionModelos de Regresion Especie-Especıficos y Generalizados

Resumen Modelamos la relacion entre la flotacion del huevo y la edad de un embrionen desarrollo para 24 especies de aves playeras Para 21 especies usamos analisis de

Manuscript received 17 January 2006 accepted 28 September 200613 Corresponding author E-mail jliebezeitwcsorg14 Present address NASA Johnson Space Center 2101 NASA Parkway Mail Code SA5 Houston TX

77058

The Condor 10932ndash47 The Cooper Ornithological Society 2007

[32]


INTRODUCTION










METHODS

STUDY SITES



FIELD METHODS





















GROUP COMPARISONS











Species




15 076 22047 143 17 066 2460 225 2013


38 075 22017 141 51 069 21061 142 2004












14 073 21617 125 27 071 21645 336 003




24 077 21727 092 23 081 831 170 2027



42 059 21642 084 10 016 1650 23150 2036


57 044 21747 082 92 041 2729 123 2006





10 050 21615 055 13 075 2616 119 2007


14 069 21996 168 32 080 2787 202 2008


36 061 21715 144 22 053 2466 106 2008


76 059 21661 108 44 050 21299 251 002


53 036 21529 073 54 056 2303 141 2010

All species 265 070 021 005 214 065 079 007 000






RESULTS










error (days)












GROUP COMPARISONS





FIGURE 3 Continued



















DISCUSSION



SpeciesaFloat






3 9 240 20 18 6 04 324 2 210 10 10 6 00 105 10 215 25 11 6 02 15

All 38 240 40 15 6 02 30



All 22 230 40 10 6 03 25


3 37 260 30 15 6 02 304 18 295 55 25 6 05 485 5 230 30 18 6 06 30

All 160 295 65 14 6 01 31

All species 1 58 2020 019 008 6 001 0152 74 2025 029 007 6 001 0203 50 2038 020 011 6 001 0204 23 2030 024 011 6 002 0195 15 2010 011 004 6 001 009

All 220 2038 029 008 6 000 020














PREDICTIVE POWER




RECOMMENDATIONS


ACKNOWLEDGMENTS




LITERATURE CITED













































INTRODUCTION










METHODS

STUDY SITES



FIELD METHODS





















GROUP COMPARISONS











Species




15 076 22047 143 17 066 2460 225 2013


38 075 22017 141 51 069 21061 142 2004












14 073 21617 125 27 071 21645 336 003




24 077 21727 092 23 081 831 170 2027



42 059 21642 084 10 016 1650 23150 2036


57 044 21747 082 92 041 2729 123 2006





10 050 21615 055 13 075 2616 119 2007


14 069 21996 168 32 080 2787 202 2008


36 061 21715 144 22 053 2466 106 2008


76 059 21661 108 44 050 21299 251 002


53 036 21529 073 54 056 2303 141 2010

All species 265 070 021 005 214 065 079 007 000






RESULTS










error (days)












GROUP COMPARISONS





FIGURE 3 Continued



















DISCUSSION



SpeciesaFloat






3 9 240 20 18 6 04 324 2 210 10 10 6 00 105 10 215 25 11 6 02 15

All 38 240 40 15 6 02 30



All 22 230 40 10 6 03 25


3 37 260 30 15 6 02 304 18 295 55 25 6 05 485 5 230 30 18 6 06 30

All 160 295 65 14 6 01 31

All species 1 58 2020 019 008 6 001 0152 74 2025 029 007 6 001 0203 50 2038 020 011 6 001 0204 23 2030 024 011 6 002 0195 15 2010 011 004 6 001 009

All 220 2038 029 008 6 000 020














PREDICTIVE POWER




RECOMMENDATIONS


ACKNOWLEDGMENTS




LITERATURE CITED















































METHODS

STUDY SITES



FIELD METHODS





















GROUP COMPARISONS











Species




15 076 22047 143 17 066 2460 225 2013


38 075 22017 141 51 069 21061 142 2004












14 073 21617 125 27 071 21645 336 003




24 077 21727 092 23 081 831 170 2027



42 059 21642 084 10 016 1650 23150 2036


57 044 21747 082 92 041 2729 123 2006





10 050 21615 055 13 075 2616 119 2007


14 069 21996 168 32 080 2787 202 2008


36 061 21715 144 22 053 2466 106 2008


76 059 21661 108 44 050 21299 251 002


53 036 21529 073 54 056 2303 141 2010

All species 265 070 021 005 214 065 079 007 000






RESULTS










error (days)












GROUP COMPARISONS





FIGURE 3 Continued



















DISCUSSION



SpeciesaFloat






3 9 240 20 18 6 04 324 2 210 10 10 6 00 105 10 215 25 11 6 02 15

All 38 240 40 15 6 02 30



All 22 230 40 10 6 03 25


3 37 260 30 15 6 02 304 18 295 55 25 6 05 485 5 230 30 18 6 06 30

All 160 295 65 14 6 01 31

All species 1 58 2020 019 008 6 001 0152 74 2025 029 007 6 001 0203 50 2038 020 011 6 001 0204 23 2030 024 011 6 002 0195 15 2010 011 004 6 001 009

All 220 2038 029 008 6 000 020














PREDICTIVE POWER




RECOMMENDATIONS


ACKNOWLEDGMENTS




LITERATURE CITED




























































GROUP COMPARISONS











Species




15 076 22047 143 17 066 2460 225 2013


38 075 22017 141 51 069 21061 142 2004












14 073 21617 125 27 071 21645 336 003




24 077 21727 092 23 081 831 170 2027



42 059 21642 084 10 016 1650 23150 2036


57 044 21747 082 92 041 2729 123 2006





10 050 21615 055 13 075 2616 119 2007


14 069 21996 168 32 080 2787 202 2008


36 061 21715 144 22 053 2466 106 2008


76 059 21661 108 44 050 21299 251 002


53 036 21529 073 54 056 2303 141 2010

All species 265 070 021 005 214 065 079 007 000






RESULTS










error (days)












GROUP COMPARISONS





FIGURE 3 Continued



















DISCUSSION



SpeciesaFloat






3 9 240 20 18 6 04 324 2 210 10 10 6 00 105 10 215 25 11 6 02 15

All 38 240 40 15 6 02 30



All 22 230 40 10 6 03 25


3 37 260 30 15 6 02 304 18 295 55 25 6 05 485 5 230 30 18 6 06 30

All 160 295 65 14 6 01 31

All species 1 58 2020 019 008 6 001 0152 74 2025 029 007 6 001 0203 50 2038 020 011 6 001 0204 23 2030 024 011 6 002 0195 15 2010 011 004 6 001 009

All 220 2038 029 008 6 000 020














PREDICTIVE POWER




RECOMMENDATIONS


ACKNOWLEDGMENTS




LITERATURE CITED















































Species




15 076 22047 143 17 066 2460 225 2013


38 075 22017 141 51 069 21061 142 2004












14 073 21617 125 27 071 21645 336 003




24 077 21727 092 23 081 831 170 2027



42 059 21642 084 10 016 1650 23150 2036


57 044 21747 082 92 041 2729 123 2006





10 050 21615 055 13 075 2616 119 2007


14 069 21996 168 32 080 2787 202 2008


36 061 21715 144 22 053 2466 106 2008


76 059 21661 108 44 050 21299 251 002


53 036 21529 073 54 056 2303 141 2010

All species 265 070 021 005 214 065 079 007 000






RESULTS










error (days)












GROUP COMPARISONS





FIGURE 3 Continued



















DISCUSSION



SpeciesaFloat






3 9 240 20 18 6 04 324 2 210 10 10 6 00 105 10 215 25 11 6 02 15

All 38 240 40 15 6 02 30



All 22 230 40 10 6 03 25


3 37 260 30 15 6 02 304 18 295 55 25 6 05 485 5 230 30 18 6 06 30

All 160 295 65 14 6 01 31

All species 1 58 2020 019 008 6 001 0152 74 2025 029 007 6 001 0203 50 2038 020 011 6 001 0204 23 2030 024 011 6 002 0195 15 2010 011 004 6 001 009

All 220 2038 029 008 6 000 020














PREDICTIVE POWER




RECOMMENDATIONS


ACKNOWLEDGMENTS




LITERATURE CITED














































RESULTS










error (days)












GROUP COMPARISONS





FIGURE 3 Continued



















DISCUSSION



SpeciesaFloat






3 9 240 20 18 6 04 324 2 210 10 10 6 00 105 10 215 25 11 6 02 15

All 38 240 40 15 6 02 30



All 22 230 40 10 6 03 25


3 37 260 30 15 6 02 304 18 295 55 25 6 05 485 5 230 30 18 6 06 30

All 160 295 65 14 6 01 31

All species 1 58 2020 019 008 6 001 0152 74 2025 029 007 6 001 0203 50 2038 020 011 6 001 0204 23 2030 024 011 6 002 0195 15 2010 011 004 6 001 009

All 220 2038 029 008 6 000 020














PREDICTIVE POWER




RECOMMENDATIONS


ACKNOWLEDGMENTS




LITERATURE CITED



















































GROUP COMPARISONS





FIGURE 3 Continued



















DISCUSSION



SpeciesaFloat






3 9 240 20 18 6 04 324 2 210 10 10 6 00 105 10 215 25 11 6 02 15

All 38 240 40 15 6 02 30



All 22 230 40 10 6 03 25


3 37 260 30 15 6 02 304 18 295 55 25 6 05 485 5 230 30 18 6 06 30

All 160 295 65 14 6 01 31

All species 1 58 2020 019 008 6 001 0152 74 2025 029 007 6 001 0203 50 2038 020 011 6 001 0204 23 2030 024 011 6 002 0195 15 2010 011 004 6 001 009

All 220 2038 029 008 6 000 020














PREDICTIVE POWER




RECOMMENDATIONS


ACKNOWLEDGMENTS




LITERATURE CITED





























































DISCUSSION



SpeciesaFloat






3 9 240 20 18 6 04 324 2 210 10 10 6 00 105 10 215 25 11 6 02 15

All 38 240 40 15 6 02 30



All 22 230 40 10 6 03 25


3 37 260 30 15 6 02 304 18 295 55 25 6 05 485 5 230 30 18 6 06 30

All 160 295 65 14 6 01 31

All species 1 58 2020 019 008 6 001 0152 74 2025 029 007 6 001 0203 50 2038 020 011 6 001 0204 23 2030 024 011 6 002 0195 15 2010 011 004 6 001 009

All 220 2038 029 008 6 000 020














PREDICTIVE POWER




RECOMMENDATIONS


ACKNOWLEDGMENTS




LITERATURE CITED





















































PREDICTIVE POWER




RECOMMENDATIONS


ACKNOWLEDGMENTS




LITERATURE CITED













































LITERATURE CITED












































assessing the development of shorebird eggs using the flotation method: species-specific and...

Documents