INCUBATION BEHAVIOR AND BODY MASS OF FEMALE GREATER SNOW GEESE

The Condor 97:993-1001 0 The Cooper Ornithologicl Society 1995 INCUBATION BEHAVIOR AND BODY MASS OF FEMALE GREATER SNOW GEESE AUSTIN REED Cndin Wildlife Service, 1141 Route de I Eglise, Ste-Foy, Quebec, Gl V 4H5. Cnd R. JOHN HUGHES AND GILLES GAUTHIER Dkprtment de Biologie nd Centre d Etudes Nordiques, Universitk Lvl, Ste-Fey, Q&bee, GlK 7P4, Cnd Abstrct. Nest ttentiveness nd behvior of incubting femle Greter Snow Geese (Anser cerulescens tlnticus) ws monitored on Bvlot Islnd. NWT in 1990-l 991. Femles incubted bout 94% of the time through much bf the incubtion period but nest ttentiveness declined to 91% t pproximtely 6 dys before htch then rose to close to 100% in the lst l-4 dys. Femles took recesses throughout the 24 dy incubtion period, but frequency incresed in the third week of incubtion then decresed mrkedly in the lst l- 4 dys. During the ltter hlf of the incubtion period, individul femles took n verge of 5-7 recesses/dy during which they remined close to the nest, nd most devoted more thn 90% of the time to feeding vigorously on sprouting grminoids. Men recess durtion ws constnt t bout 15-16 min, except for the lst l-4 dys when it incresed to ~30 min. Femles cptured during lte incubtion hd reltively high body mss (X = 2,158 g) which represented loss of only 17% in mss during incubtion nd showed tht body reserves hd not been depleted. Retining some body reserves through to htch my increse survivl of femles nd my enhnce their cpbility of cring for the young during brood rering. In comprison with other lrge geese, Greter Snow Geese showed low rte of nest ttentiveness nd took frequent recesses during which they fed intensively; they relied more hevily on food obtined in the breeding re (s opposed to reserves lid down during migrtion) nd lost smller proportion of their initil mss during incubtion. This strtegy ppers to be imposed by the lengthy migrtion to the high Arctic, which reduces the mount of stored reserves vilble for the demnds of nesting. Key words: Anser cerulescens tlnticus; Arctic; body condition; Bylot Islnd; Greter Snow Goose: incubtion behvior. INTRODUCTION Most geese of the Northern Hemisphere breed in the Arctic. The constrints imposed by this choice of breeding re include the requirement to undertke lengthy flights from southerly stging nd wintering res, short summer seson in which to crry out nesting nd brood rering, nd the reltively low biomss of rctic grminoid plnts which constitute the summer diet (Owen 1980, Afton nd Pulus 1992, Mnseu nd Guthier 1993). Egg lying occurs before new growth of vegettion hs begun nd when food vilbility is low, mking lrge demnd on the nutrient reserves of the femle just before she begins incubtion. Only femles incubte nd the requirement to mintin sufficiently wrm nest environment for embryonic development nd to provide protection ginst egg predtors Received 30 June 1995. Accepted 11 July 1995. limits opportunities to feed (Thompson nd Rveling 1987). The stressful nture of this period ws evident in studies of Lesser Snow Geese (Anser cerulescens cerulescens) breeding on the west cost of Hudson By where femles lost more thn 30% of initil body mss during incubtion (Ankney nd Mclnnes 1978) nd some individuls ctully strved just before htch (Hrvey 197 1). In ll species of geese studied to dte, incubting femles spend more thn 90% of their time on the nest nd lose substntil proportion of their initil body mss (Owen 1980, Thompson nd Rveling 1987) but there re importnt interspecific differences in incubtion behvior nd rtes of body mss loss. Thompson nd Rveling (1987) ttributed these differences minly to body size: smller species such s Brnt (Brnt bernicl) incubte bout 93% of the dy, the remining time being spent in severl short recesses during which they feed intensively 19931

994 A. REED. R. J. HUGHES AND G. GAUTHIER (Thompson nd Rveling 1987, Mdsen et l. 1989). In contrst, lrger species such s Gint nd Western Cnd Geese (B. cndensis mxim nd B.C. mojitti) (Cooper 1978, Aldrich nd Rveling 1983,MurphyndBog 1989)ndEmperor Geese (Anser cngic) (Thompson nd Rveling 1987) spend more time on the nest (97-99%) nd tke fewer but longer recesses during which less thn hlf of the time is devoted to feeding. Thus, lrger species generlly hve low food intke during incubtion, rely lmost entirely on body reserves ccumulted erlier on spring stging res (Ankney nd McInnes 1978, Thoms 1983) nd tend to lose lrger proportion of their body mss during incubtion thn do smller species (Thompson nd Rveling 1987). Mny lrge geese (e.g., lrger rces of Cnd Geese) migrte shorter distnces nd nest further south in comprison to smll geese (smll rces of Cnd Geese, Brnt). A lower mssspecific metbolic rte nd the bility to store lrge nutrient reserves llow lrge geese to dopt strtegy mximizing relince on stored reserves nd high rte of nest ttentiveness (Afton nd Pulus 1992). One of the lrgest subspecies, the Greter Snow Goose (Anser cerulescens tlnticus), does not fit the pttern described by Thompson nd Rveling (1987). This lrge goose migrtes long distnces nd nests t very high ltitudes (Guthier et l. 1992) nd initites incubtion with only moderte body reserves (Choiniere nd Guthier, in press). When weighing femle Greter Snow Geese cptured in lte incubtion during n erlier study (Hughes et l. 1994), we found men body mss higher thn expected bsed on studies of other lrge geese. We therefore undertook this study to (1) further document body mss loss by incubting femles nd (2) exmine whether behviorl ptterns could explin the mintennce of reltively high body mss through incubtion in this unique stock of lrge geese breeding in the high Arctic. STUDY AREA AND METHODS The study re ws glcil vlley ner the westem extremity of the southwest plin of Bylot Islnd (73 N, 8O W), Northwest Territories. The southwest plin is the most importnt nesting loction for Greter Snow Geese, ccommodting pproximtely 15% of the breeding popultion (Reed nd Chgnon 1987, Reed et l. 1992); our study site supported one of the highest nesting concentrtions on tht plin. Most geese, in- eluding those observed in 1991 nd those weighed in 1990-l 99 1, nested on low ridges seprting wet polygon depressions in generlly flt, lowlnd plin. The dry ridges supported willow (Slix rctic) nd vriety of forbs nd grminoids wheres the depressions were dominted by the grminoids Dupontijsheri, Crex qutilis vr. stns, nd Eriophorum spp. (Hughes et l. 1994b, Guthier et l. 1995). Predtors of Snow Goose eggs included, in pproximte order of importnce, Arctic Foxes (Alopex lgopus), Prsitic Jegers (Stercorrius prsiticus), Long-tiled Jegers (S. longicudus), Glucous Gulls (Lrus hyperboreus) nd Common Rvens (Corvus corx); only Arctic Foxes were potentil thret to dult geese. There ws dylight over the entire 24-hr period during incubtion. Geese were cptured (1989-91) on the nest, during the lst nine dys of the pproximtely 24-dy incubtion period (Guthier nd Reed, unpubl. dt), with modified bow net trp triggered t distnce from blind (n = 11 femles in 1989, 19 in 1990, 15 in 1991). In 1992 smple of 12 femles ws shot on the nest to determine the mss of ingest. Geese were weighed with spring blnce to the nerest 50 g. The min body of our dt on incubtion behvior comes from observtions mde from blind overlooking n expnse of lowlnd polygon depressions occupied by severl dozen nesting pirs in 1991. We recorded incubtion behvior for 24-hr periods t four dtes during incubtion: 16-17, 22-23, 26-28 June nd l-3 July 1991 (pek of htch ws 9 July). During ech of the first two dtes, we covered n uninterrupted 24- hr session using combintion of visul obser- vtion nd time-lpse (1 imge min- ) video recording (8 mm Cnon E57). The two remining dtes were ech covered by two 12-hr sessions (12:00-24:00, 24:00-12:00, EST) over 48-hr period, using only visul observtions. During visul observtions, we kept nests (n = 8 for the first two dtes, n = 13 for the lst two dtes) under continul surveillnce, recording times of deprture nd return of incubting femles to the nerest minute. For femles on recess we recorded their pproximte distnce from the nest t rndomly selected point in time during the recess; this yielded 4-28 records/femle for estimting individul men distnces. We determined the proportion of time spent feeding during recesses on the lst two dtes by recording the instntneous behvior of femles every 10 set for the entire recess (up to 1,260 set) or for

INCUBATION BEHAVIOR OF GREATER SNOW GEESE 995 periods distributed rndomly within the recess (80-300 set, mostly 250 set). We lso recorded pecking rtes (time for 25 pecks) which were converted to pecks/min. Men rtes for individul femles during ll but the first dte were determined using 3-l 3 (mostly 10) records per femle per recess. We compred nest ttentiveness, recess durtion nd recess frequency mong the four observtion dtes nd mong times of dy (four periods: 00:0@-06:00,06:00-12:00, 12:00-18:00, 1 S:OO-24:OO). Nest ttentiveness ws determined s the percentge of totl time spent on the nest by dy/period. For recess durtion nd frequency, only complete (both strt nd end recorded) recesses were retined nd ech ws ssigned to one of the four periods bsed on its midpoint in time. Incubtion bout length ws not compred becuse mny incubtion bouts extended beyond the beginning or end of the observtion sessions. Thus, the durtion of some of the longest bouts could not be determined. For comprisons between dtes, we restricted our nlysis to dt recorded between 07:OO nd 19: 30 (dt were incomplete for other time periods, minly for the first two dtes). Similrly, for comprisons between times of dy, we restricted our nlysis to the lst two dtes. Anlyses were conducted using two-wy nlysis of vrince (ANOVA) with femles considered s fctor to control for between-femle differences in incubtion behvior. Multiple posteriori comprisons were conducted using Tukey s Honestly Significnt Difference test. Pecking rtes t different dtes were compred using Kruskll-Wllis nlysis of vrince becuse of heterogenous vrinces mong dtes. In 1990, dditionl observtions (6 hr on 6 July nd 18 hr on 7-8 July) were conducted on 12 incubting femles during the lst l-4 dys before htch, phse of the incubtion period not covered by the 1991 observtions. We used liner regression to exmine the reltionship between body mss nd the number of dys before htch during the lst nine dys of incubtion. Only the body msses of femles with known htch dte (n = 36) were used. RESULTS INCUBATION BEHAVIOR Becuse of equipment limittions or poor lighting, video recording yielded only five records of recess durtion nd none of incubtion bouts. Thus, visul observtions provided most of the dt. Men nest ttentiveness mong the femles observed in 1991 verged 93.0% cross both the four observtion dtes (07:OO nd 19:30) nd the four time periods (dtes 3 nd 4). Between 07:OO nd 19:30, individul femles (n = 13) incubted for 88 to 98% of the time. Men nest ttentiveness ws constnt during the first three dtes, rnging from 93.8 to 94.3%, but declined to 91.0% on 1-3 July (F = 4.46, 3, 25 df, P = 0.012) (Fig. 1). The pproximte men number of dys into incubtion for the femles under observtion t these four dtes were two (n = 7 femles with known htch dte), nine (n = 8), 13 (n = lo), nd 18 (n = 10) respectively. Nest ttentiveness vried mong times of dy during the lst two observtion dtes (F = 4.14, 3, 84 df, P = 0.009) (Fig. 2). Femles were less ttentive (90.6%) between 12:OO nd 18:OO thn erlier in the dy (94.3-94.7%). Recess durtion ws constnt sesonlly (F = 1.39,3, 109 df, P = 0.250) (Fig. 1). Although the test mong times of dy suggested temporl difference in recess durtion (F = 2.96, 3, 137 df, P = 0.035), the subsequent multiple comprison did not revel ny mong-period differences (P L 0.05) (Fig. 2). Overll men recess durtion ws pproximtely 15-16 min cross both the four dtes nd four times of dy. Recess frequency incresed sesonlly (F = 6.58, 3, 23 df, P = 0.002) (Fig. 1). Although recesses occurred t ll times of dy, they were more frequent lter in the dy, prticulrly between 12:00 nd 18:00 (F = 3.83, 3, 84 df, P = 0.013) (Fig. 2). Bsed on the most complete records, fter 26 June, individul femles took two to eight recesses per dy, with men (*SE) of 5.3 (ko.5) on 26-28 June nd 6.9 (ko.3) on l-3 July. Men durtion of complete incubtion bouts recorded between 07:OO nd 19:30 for ll dtes ws just under 3 hr (X = 173.0 min, SE = 8.1, n = 68). Some incubtion bouts were, however, in progress t the strt nd termintion of ech observtion session, preventing determintion of their durtion; few were known to exceed the mximum observed durtion of 8.8 hr including two > 10 nd > 12 hr respectively on 26 June. The dditionl observtions of 12 femles in 1990 (6-8 July) were unique in two respects: firstly they were conducted lter in incubtion thn those of 1991 (fter ~20 dys of incubtion versus men of 5 18 dys) nd secondly, they re-

996 A. REED, R. J. HUGHES AND G. GAUTHIER b veled higher men rte of nest ttentiveness (98.8% versus 93.0% in 1991) which ws the result of nine femles tking no recesses during 24 hr of observtion nd the three others tking totl of six long recesses (X = 34.3 min, SE = 9.0). f b b 3 b T t BEHAVIOR DURING INCUBATION RECESSES b A While off the nest, femles devoted most of their time to feeding. On 26 June, the men proportion of time spent feeding by seven femles ws 88% 1 (rnge 63-969/o). The femle which fed the lest ws not ccompnied by her mte. For seven femles monitored between l-3 July, men feeding time ws 8 1% (rnge 33-100%). On both dtes, most femles (5 nd 4 respectively) spent >90% of the time feeding. The reminder of the time ws minly devoted to lert behvior. 7 Most feeding occurred in polygon depressions Ill-l T where femles grzed vigorously on emerging shoots of grminoids. Men ( f SE) pecking rtes (pecks/min) were 133 (& 11) on 22-23 June (n = 6 femles), 131 (+ 1) on 26-28 June (n = 6) nd incresed to 153 (+8) on l-3 July (n = 9) (Kruskll-Wllis, P = 0.052). One exceptionlly low rte observed on July 3 (60 pecks/min) ws excluded. Occsionlly femles fed on willow ct- kins, pecking t lower rte of 64 (k 18) pecks/ min (n = 3). Csul observtions in 1990 showed tht some femles lso fed on the flowers of Oxytropis mydellin, leguminous plnt common in the study re. During recesses femles usully remined close to their nests. More thn 90% of ll records (n = 183) were within 20 m of the nest (rnge l- 80 m), lthough csul observtions indicted few cses of distnces > 100 m. The men distnce from the nest (26 June-3 July, n = 13 femles ) ws 13.7 m (SE = 1.4). 16-17 22-23 26-28 1-3 JUNE JULY DATE FIGURE 1. Nest ttentiveness (A), recess durtion (B), nd recess frequency (C) of Greter Snow Geese in reltion to dte, Bylot Islnd 1991. Verticl brs represent stndrd errors. Smple sizes (femles) re the followine: 16-l 7 June (n = 7). 22-23 June (n = 8). 26-28 June & = 13), nd I-3 July (n = 13). Columns with different letters re significntly different (P < 0.05), Tukey s HSD. - BODY MASS OF INCUBATING FEMALES From 2 to 8 July 199 1, 15 incubting femles were cptured on the nest nd weighed, including seven whose incubtion behvior hd been monitored. Body mss ws lso vilble for 30 femles cught in 1989-l 990. Men mss did not differ between yers (F = 0.449, 2, 42 df, P = 0.641) nd the pooled men ws 2,158 g (SE = 3 1, n = 45). Femles whose htch dtes were not known exctly were ssumed to be in the ltter stge of incubtion becuse ll birds were cptured within the sme period in ech yer nd

INCUBATION BEHAVIOR OF GREATER SNOW GEESE 997 T b C 1 b ri TIME OF DAY (h) FIGURE 2. Nest ttentiveness (A), recess durtion (B), nd recess frequency (C) of Greter Snow Geese in reltion to time of dy, Bylot Islnd 199 1. Verticl brs represent stndrd errors. Smple sizes (femles) re the following: 00:00-06:00 (n = 12) 06:0&12:00 (n = 12) 12:00-l 8:00 (n = 13), nd 18:00-24:00 (n = 13). Columns with different letters re significntly different (P < 0.05) Tukey s HSD. 1590 t-. 1200 II -9-9 -7-5 -5-4 -3-2 -1 0 DAYS BEFORE HATCH FIGURE 3. Reltionship between body mss nd the number of dys before htch, during the lst nine dys of incubtion, for incubting Greter Snow Geese t Bylot Islnd 1989-1991: Y = 2257-22X, r* = 0.06, P = 0.146, n = 36. htching t Bylot Islnd is highly synchronous (G. Guthier nd A. Reed, unpubl. dt). Body mss of femles with known htch dte ws not correlted with the number of dys (O-9) remining before htch (Fig. 3). Becuse geese were cptured fter returning from forced recess n In (while the trp ws being set), their body msses included the weight of ny recently consumed food. We could not directly correct for tht possible bis becuse they were relesed live. However, for 12 incubting femles collected during lte incubtion (O-7 dys before pek htch dte) in 1992, the men mss (*SE) of ingest in the b esophgus nd proventriculus ws 7.5 g (f2.7) 1 in the gizzrd 2 1.7 (f4.4), nd in the intestine 55.1 (k9.1). The totl mount of ingest in the i esophgus, proventriculus nd gizzrd ws 136 g in 10 femles nd 52 nd 62 gin the two others. The men body mss of the 12 femles ws 2,144 g (SE = 51). n - DISCUSSION 00X& 06-12 12-18 18-24 With no periods of drkness, femle Greter Snow Geese on Bylot Islnd dopted regulr pttern lternting incubtion bouts with recesses over the full 24-hr cycle. The verge durtion of recesses (15-16 min) is shorter thn the overll men (19 min) clculted for five species of geese by Afton nd Pulus (1992). Short recesses re dvntgeous by reducing the probbility of egg loss to predtors (Inglis 1977, Thompson nd Rveling 1987, Spns et l. 1993) nd by re- I

998 A. REED, R. J. HUGHES AND G. GAUTHIER ducing het loss from the eggs (Afton nd Pulus 1992). The lck of vribility in recess durtion of Greter Snow Geese over the seson (except for the lst l-4 dys of incubtion-see below) or throughout the 24 hr cycle suggests tht mbient tempertures hd little influence on recess length. The number of recesses per dy did, however, increse sesonlly (Fig. l), nd becuse recess durtion remined stble, totl recess time per dy incresed through ll but the lst few dys of incubtion. An increse in totl recess time (or decrese in nest ttentiveness) s incubtion progressed ws observed for Cnd Geese (Aldrich nd Rveling 1983) nd for Brnt (Spns et l. 1993) but not for Emperor Geese (Thompson nd Rveling 1987); for Cnd Geese, the incresing time off nest ws ttributed to deteriorting body condition nd the need to increse time spent feeding, wheres Emperor Geese were thought to be mintining dequte body condition to be ble to remin highly ttentive. During the lst l-4 dys of incubtion nest ttentiveness incresed substntilly, lthough the infrequent recesses were longer. Incresed t- tentiveness in the lst few dys of incubtion hs been reported for Emperor Geese (Thompson nd Rveling 1987) nd for Gint Cnd Geese (Cooper 1978) nd is probbly the result of incresing communiction between the embryos nd the femle. The femles tht tke recesses during this period my be those in poorest condition nd therefore in gretest need of food. Regrdless of their condition, once off the nest they my remin wy longer becuse eggs dvnced in incubtion cool less quickly (Afton nd Pulus 1992). Femles on recess, by remining close to the nest nd devoting most of their time to feeding with rpid peck rte, were probbly minimizing energy expenditure nd mximizing food intke. Towrd the end of most recesses the femles hd swollen necks, indicting tht the esophgus ws full. Apprecible quntities of food (280 g, on verge) were found in the digestive trct of breeding femles collected during incubtion t the sme study site in 1989-I 990 (Guthier 1993) nd in 1992 (this study). Most of the food consumed by femles during incubtion ws newly sprouting grminoid folige nd willow ctkins which hve high nutritive vlue (Guthier 1993). Thus, femles completely filled their esophgi 5-7 times dy nd by retining tht high qulity food in the digestive system for extended periods (usully 13 hr) were enhncing nutrient ssimiltion (Owen 1980, Prop nd Vulink 1992). In Brncle Geese (Brunt leucopsis) feeding on grminoids, prolonging food retention beyond the pproximtely 2 hr norml retention time to 3-4 hr, substntilly incresed orgnic mtter digestibility (Prop nd Vulink 1992). On verge, femles devoted 93% of their time to incubting eggs. In ddition to ensuring proper embryonic development, such constncy undoubtedly served to protect the eggs from predtors (Thompson nd Rveling 1987). On severl occsions, the regulr pttern of nest ttentiveness ppered to hve been djusted in response to predtory ctivity. Femles on recess lmost invribly returned immeditely to the nest when Prsitic Jegers, Common Rvens or Arctic Foxes pproched. All femle geese remined on the nest while foxes were in the colony nd none took recesses for t lest 1 hr fter deprture of the foxes. Although body mss of the seven femles whose behvior hd been monitored ws not correlted with their rte of nest ttentiveness, nor with the durtion of their recesses, the lightest of them (1,800 g) hd the lrgest clutch (7 eggs), the longest verge recess durtion (21 min), nd the lowest percentge of time spent on the nest (88%). This suggests tht she begn incubtion with low level of body reserves nd ws djusting her schedule to increse the intke of nutrients. Despite the reltively low levels of her nutrient reserves nd her reduced ttentiveness to the nest, she ws suc- cessful in htching her eggs. The two lightest geese weighed in 1989 (1,480 nd 1,700 g) lso htched their eggs nd the lightest of them ws known to hve survived t lest four months s she ws observed stging 3,000 km to the south in October. From smple of severl hundred nests over six full sesons of observtion in this colony, no deths of incubting femles from exhustion or strvtion hve been recorded (G. Guthier nd A. Reed, unpubl. dt) which contrsts with the sitution reported for Lesser Snow Geese in western Hudson By (Hrvey 1971, Ankney nd McInnes 1978). The femles we weighed in lte incubtion, becuse they were cptured fter forced recess, my hve contined lrger-thn-verge quntity of recently ingested food. Tht bis in mss

INCUBATION BEHAVIOR OF GREATER SNOW GEESE 999 is probbly not greter thn the mximum vlues of ingest found in the nterior portion of the digestive trct of femles collected while incubting (i.e., 52 nd 62 g in two of 12 femles). The men body mss of 2,158 g (SE = 31, n = 45, 1989-91) in lte incubtion, even fter ccounting for bout Xl-60 g of newly ingested food, indictes tht nutrient reserves hd not been entirely depleted, from lbortory studies (Boismenu et l. 1992), it is known tht fsting Greter Snow Goose weighing 2,100 g hs sufficient reserves to survive > 12 dys on verge. The men body mss of femles during erly incubtion (dy 3) in the sme popultion in 1989-1990ws2,480g(SE= 55,n= ll)(choiniere nd Guthier, in press). Assuming tht body mss is lost t more rpid rte during the first few dys of incubtion thn lter on, we estimte totl mss loss over the full incubtion period of 430 g or 17% (Tble 1). Mss loss must occur minly in erly incubtion becuse we detected no reltionship between body mss nd the number of dys pre-htch during the lst nine dys of incubtion (though there ws slightly negtive slope). The pproximte period of 9-4 dys before htch ppers to be chrcterized by lower nest ttentiveness nd n incresed frequency of recesses which my llow geese to mintin their body condition, perhps in preprtion for the very lst dys of incubtion when nest ttentiveness pproches 100%. Our results indicte pttern rther different from tht described for closely relted Lesser Snow Geese in Hudson By. Femle Lesser Snow Geese fed only 65% of their time wy from the nest nd fed with pecking rtes < 100 pecks/min (Hrwood 1977). They left the nest infrequently during incubtion nd lost, on verge, >30% of their body mss with mny femles strving t the nest during lte incubtion (Ankney nd McInnes 1978). Almost ll of the reserves used to sustin femles through incubtion cme from stores remining fter reltively short migrtions from spring stging res in southern Mnitob nd the Jmes By/Hudson By lowlnds (Ankney nd McInnes 1978, Alisusks nd Ankney 1992). Thus, despite their smller size, Lesser Snow Geese initited incubtion with similr body mss to Greter Snow Geese, but by relying lmost entirely on stored reserves they lost mss more rpidly nd terminted incubtion in poorer condition (Tble 1). TABLE 1. Nest ttentiveness, feeding ctivity, nd loss of body mss of Greter Snow Geese nd other northern geese during incubtion. Species/subspecies Western Cnd Goose Greter Snow Goose Lesser Snow Goose4 Emperor Goose Cckling Cnd Goose Blck Brnt Min- Mss (9) % tes t strt Time feed- % of incu- incu- ing/ Mss btion bting dy loss 4,300 91.5 14 21 2,5902 93.0 83 11 2,530 91.1 n.. 32 2,230 99.5 1 21 1,387 93.6 71 21 1,140 89.6 95 11 Dt fro,m Thompson nd Rveling (1987), except for I Greter Snow F;;cnx& (this study), nd 4 Lesser Snow Goose (Ankney nd McInnes 2 Adjusted to dy one of incubtion from vlue of 2,480 g on dy three, published by Cboini&e nd Guthier (in press), ssummg dily rte of body mss loss during the first three dys twice tht occurring over the rest of the incubtion period. Averge of 2628 June (73 min) nd l-3 July (93 min). Nest ttentiveness, feeding intensity, nd body mss loss of Greter Snow Geese is compred with tht of other northern nesting geese in Tble 1. Thompson nd Rveling (1987) in compring Western Cnd Geese, Emperor Geese, Cckling Cnd Geese (Brunt cndensis minim) nd Brnt, suggested tht the lrger species devoted more time to incubtion, less to feeding, nd lost proportiontely more body mss thn the smller species. Using the Emperor Goose s n exmple, they rgued tht the medium nd lrge sized species, by remining on the nest lmost continuously, were ensuring mximl protection of the nest ginst both mmmlin nd vin predtors. This is possible becuse lrger species hve more reserves nd expend less energy per unit of mss. As corollry of this, lrger species generlly tke < 3 recesses per dy (lrge rces of Cnd Geese [Cooper 1978, Aldrich nd Rveling 19831, Emperor Geese [Thompson nd Rveling 19871, White-fronted Geese Anser lbifrons [Stroud 19821, Pink-footed Geese A. bruchyrhynchus [Inglis 19771) wheres the diminutive Brnt tkes 6-14 recesses per dy (Thompson, cited in Thompson nd Rveling 1987, Mdsen et l. 1989, Spns et l. 1993). The geese we studied did not fit the pttern between body size nd nest ttentiveness suggested by Thompson nd Rveling (1987). Despite their lrge body size, Greter Snow Geese took frequent recesses, devoted considerble time to feeding, nd lost reltively smll proportion of their body mss, resembling in those respects

1000 A. REED, R. J. HUGHES AND G. GAUTHIER much smller species such s Brnt nd Cckling Cnd Geese. Greter Snow Geese must use lrge proportion of the nutrient reserves ccumulted on the spring stging re in southern Quebec during the 2,500-3,000 km flight to the high Arctic (Guthier et l. 1992) nd must rely on food obtined on the breeding grounds to meet nutrient demnds during pre-lying nd lying (Guthier nd Trdif 1991, Guthier 1993). With smller imported nutrient reserves to drw on during incubtion, reltive to other lrge geese, they must continue feeding regulrly through incubtion. This requirement is pprently met by lternting short recesses, the durtion of which probbly represents the time required to fill the esophgus, with incubtion bouts long enough to mximize ssimiltion of nutrients. In the high Arctic this pttern is unhmpered by periods of drkness. By feeding close to their nests during recesses, they re cpble of defending their nests ginst predtors, nd by feeding intensively mintin level of body condition which ensures their own survivl through incubtion. By strting the brood-rering period in good condition, they my lso enhnce survivl of their goslings in the dys following htch. ACKNOWLEDGMENTS This project ws funded by the Arctic Goose Joint Venture (Cndin Wildlife Service funds) nd by grnts to G. Guthier from: Supply nd Services Cnd nd Environment Cnd (Contrct KA313-8- 5545-08-A), the Nturl Science nd Engineering Reserch Council of Cnd, the Fonds pour l Formtion de Chercheurs et l Aide l Recherche of the Minis&e de 1 Eduction du Quebec, nd Indin nd Northern Affirs Cnd. We re especilly grteful to the Polr Continentl ShelfProject for vluble logistic support. H. Boyd nd E. Reed ssisted with behviorl observtions in 1990 nd 1991 respectively. N. Hmel, G. Bourss, P. Sucy, J. Beulieu, M. Slthe, nd Y. Poirier helped cpture nd weigh the geese. H. Boyd provided ssistnce nd dvice throughout the study. We re grteful to D. Bog, M. Petersen, nd J. Sedinger for constructive comments on erlier drfts of the mnuscript. LITERATURE CITED, AXON, A. D., AND S. L. PAULUS. 1992. Incubtion nd brood cre, p. 62-108. In B.D.J. Btt, A. D. Afton, M. G. Anderson, C. D. Ankney, D. H. Johnson, J. A. Kdlec, nd G. L. Krpu [eds.], Ecology nd mngement of breeding wterfowl. Univ. of Minnesot Press, Minnepolis. ALDRICH, T. W., AND D. G. RAVELING. 1983. Effects of experience nd body weight on incubtion behvior of Cnd Geese. Auk 100:67C-679. ALISAUSKAS, R. T., AND C. D. ANKNEY. 1992. Sprint hbitt use nd diets of midcontinent dult Lesser Snow Geese. J. Wildl. Mne. 56:43-54. ANKNEY, C. D., AND C. D. MAC&&. 1978. Nutrient reserves nd reproductive performnce of femle Lesser Snow Geese. Auk 95:459-47 1. BOISMENU, C., G. GAUTHIER, AND J. LAROCHELLE. 1992. Physiology of prolonged fsting in greter snow geese Chen cerulescens tlntic. Auk 109: 511-521. Cr-rotmti, L., AND G. GAUTHIER. In press. Energetits of reproduction in femle nd mle Greter Snow Geese. Oecologi. COOPER, J. A. 1978. Thehistory nd breeding biology of the Cnd Goose of Mrshv Point. Mnitob. Wildl. Monogr. 61. GAUTHIER, G. 1993. Feeding ecology of nesting Greter Snow Geese. J. Wildl. Mnge. 57:216-223. GAUTH~ER, G., J.-F. Groux, AND J. BBDARD. 1992. Dynmics offt nd protein reserves during winter nd spring migrtion in Greter Snow Geese. Cn. J. Zool. 70:2077-2087. GAUTH~ER, G., R. J. HUGHES, A. REED, J. BEAULIEU, AND L. ROCHEFORT. 1995. Effect of grzing by Greter Snow Geese on the production of grminoids t n rctic site. J. Ecol. 83:653-664. GAUTHIER, G., AND J. TARDIF. 1991. Femle feeding nd mle vigilnce during nesting in Greter Snow Geese. Condor 93:701-711. HARW~OD, J. 1977. Summer feeding ecology of Lesser Snow Geese. J. Wildl. Mnge. 41:48-55. HARVEY, J. M. 1971. Fctors ffecting Blue Goose nesting success. Cn. J. Zool. 49:223-234. HUGHES, R. J., A. REED, AND G. GAUTHIER. 1994. Spce nd hbitt use by Greter Snow Goose broods on Bylot Islnd. Northwest Territories. J. Wildl. Mnge. 58:536-545. HUGHES, R. J., G. GAUTHIER, AND A. REED. 1994b. Summer hbitt use nd behviour of Greter Snow Geese Anser cerdescens. Wildfowl 45:4944. INGLIS, I. R. 1977. The breeding behviour of the Pink-footed Goose: behviourl correltes ofnesting success. Anim. Behv.25:747-764. -EN, J., T. BREGNBALLE, AND F. MEHLUM. 1989. Study of the breeding ecology nd behviour of the Svlbrd popultion of Light-bellied Brent Goose Brnt bernicl hrot. Polr Reserch 7: l- 21. MANSEAU, M., AND G. GAUTHIER. 1993. Interctions between Greter Snow Geese nd their rering hbitt. Ecolon, 74:2045-2055. MURPHY, A. J., &D. A. BOAG. 1989. Body reserve nd food use by incubting Cnd geese. Auk 106:439-446. OWEN, M. 1980. Wild geese of the world. Btsford Ltd., London. PROP, J., AND T. VULINK. 1992. Digestion by brncle geese in the nnul cycle: the interply between retention time nd food qulity. Functionl Ecology 6:180-189. REED, A., AND P. CHAGNON. 1987. Greter snow geese on Bylot Islnd, Northwest Territories, 1983. J. Wildl. Mnge. 51: 128-l 31. REED, A., H. BOG, P. CHAGNON, AND J. HAWKINGS.

INCUBATION BEHAVIOR OF GREATER SNOW GEESE 1001 1992. The numbers nd distribution of Greter Snow Geese on Bylot Islnd nd ner Jungersen By, Bffin Islnd, in 1988 nd 1983. Arctic 45: 115-119. SPAANS, B., M. STOCK, A. ST. JOSEPH, H. H. BERG- MANN, AND B. S. EBBINGE. 1993. Breeding biology of Drk-bellied Brent Geese in Timyr in 1990 in the bsence of rctic foxes nd under fvourble wether conditions. Polr Reserch 12: 117-130. STROUD, D. A. 1982. Observtions on the incubtion nd post-htching behviour of the Greenlnd White-fronted Goose. Wildfowl 33:63-72. THOMAS, V. D. 1983. Spring migrtion: the prelude to goose reproduction nd review of its implictions, p. 73-8 1. In H. Boyd [ed.], First Western Hemisphere Wterfowl nd Wterbirds Symposium. IWRB nd Cndin Wildlife Service, Ottw. THOMPSON, S. C., AND D. G. RAVELING. 1987. Incubtion behvior of emperor geese compred with other geese: interctions of predtion, body size, nd energetics. Auk 104:707-716.