LouIs LESAGE AND GILLES GAUTHIER 1

The Auk 114(2):229-241, 1997 GROWTH AND ORGAN DEVELOPMENT IN GREATER SNOW GOOSE GOSLINGS LuIs LESAGE AND GILLES GAUTHIER 1 Ddpartement de bilgie and Centre d'dtudes nrdiques, Universitd Laval, Sainte-Fy, Quebec G1K 7P4, Canada AlaSTRACT.--The tissue allcatin hypthesi states that functinal maturity and rapid embrynic grwth are incmpatible the tissue level. This culd explain why preccial birds, which have mre mature tissues at hatching, grw mre slwly than altricial birds. We evaluated this hypthesis in Greater Snw Geese (Chen caerulescens atlantica), which nest in the high arctic where the grwing seasn is very shrt. We examined grwth patterns and dry-matter cntent (an index f tissue maturatin) f varius tissues, and the accumulatin f fat in 176 gslings cllected frm hatch n Bylt Island, Nrthwest Territries, t their staging area at Cap-Turmente, Quebec (1 t 110 days). The mass-specific grwth cnstant (K = 0.093) f gslings was amng the highest f all preccial birds, including ducks. Gslings fledged at nly 68% f adult mass, a lw value cmpared with ther species. The timing and rate f grwth differed amng tissues, indicating majr shifts in the allcatin f prtein during grwth. Grwth rates fr bdy mass, bdy ash, and ttal bdy prtein were mderate. Leg muscles and digestive rgans had an early and rapid grwth rate. Breast muscles had ne f the highest grwth rates but started t grw very late. Fat accumulatin began after fledging, frcing gslings t start suthward migratin with very little fat reserves. In early-grwing tissues (digestive rgans and leg muscles), water cntent was lw at hatch, peaked befre fledging, and decreased thereafter. This cntrasts with the typical pattern in birds f peak values at hatch fllwed by a mntnic decline during grwth. The high dry-matter cntent f tissues at hatch culd be an adaptatin t increase thermgenesis f gslings in cld water. A strng inverse relatinship between expnential grwth rate and functinal maturity was fund in breast muscles but was absent in early-maturing tissues. Eclgical factr seem mre imprtant than embrynicnstraints in explaining fast grwth rates in geese. Received 26 February 1996, accepted 6 Nvember 1996. BIRDS HAVE VERY FAST GROWTH RATES, their pecially imprtant in cnstrainingrwth rates yung grwing abut twice as fast as mammals because the leg muscles usually make up the f equal size (Bj6rnhag 1979). In his reviews f largest skeletal muscle mass at hatching (Rickgrwth in birds, Ricklefs (1973, 1979a) nticed lefs 1979b). that the grwth rate cnstant is inversely related In the Anatidae, grwth rate varies cnsidert adult bdy mass and decreases with the pre- ably amng species, althugh it generally is ccity f the species. Birds that hatch with a mre rapid than in ther preccial birds f similarger prprtin f mature tissues (i.e. prec- lar size (Ricklefs 1973). High seasnal fd cial species) tend t grw mre slwly than availability and shrt breeding seasns, tw thse that hatch in a mre immature state (i.e. characteristics f the envirnments used by altricial species; Lightbdy and Ankney 1984, anatids, are sme factrs explaining their rapid Sedinger 1986). Accrding t the tissue- grwth (Ricklefs 1973, 1983; Aubin et al. 1993). allcatin hypthesis, grwth rate is inversely The largest anatids, geese (Anserinae), are prerelated t the degree f functinal maturity be- ccial birds whse legs are prprtinally larger cause mature muscle functin (i.e. capacity fr than thse f ducks, an adaptatin t their mre thermgenic heat prductin), which requires terrestrial lifestyle. Yung geese have welldifferentiated tissues, is incmpatible with em- develped legs and are highly mbile at hatchbrynic grwth, which requires undifferentiated ing. Yet, Bj6rnhag (1979) fund that geese have tissues (Ricklefs 1979a). The size and degree f higher grwth rates than several ther preccial maturity f leg muscles in yung birds are es- birds, including ducks. Aubin et al. (1986) als reprted that Lesser Snw Geese (Chen caerulescens caerulescens) had a mass-specific grwth Address crrespndence t this authr. E-mail: cnstant almst as high as that f Cmmn gilles.gauthier bi.ulaval.ca Terns (Sterna hirund), a parentally fed preccial 229

230 LESAGE AND GAUTHIER [Auk, Vl. 114 species with smaller legs than geese. Therefre, n the nrth shre f the St. Lawrence estuary the fast grwth rates f geese, in spite f their (47ø04'N, 70ø47'W), 45 km dwnstream f Quebec large, well-develped legs at hatching, appears City. Geese use tw ecsystems n the refuge, the agt be a challenge t the tissue-allcatin hypthriculture lands f the castal plain and the tidal marsh esis. dminated by American bulrush (Scirpus americanus). Cap-Turmente is a majr staging area fr Greater Because mst geese are lng-distance mi- Snw Geese cming frm Bylt Island as well as thse grants, newly fledged gslings als face the cming frm ther breeding clnies scattered acrss challenge f accumulating fat reserves fr the the Canadian arctic islands. The site is heavily used suthward migratin at the same time as they by geese frm late September t early Nvember. are cmpleting grwth. The early nset f fall Frm the St. Lawrence estuary, geese migrate an adin the arctic impses an verlap f these tw ditinal 1,000 km t their wintering areas alng the energy-demanding prcesses. The prblem f cast f the mid-atlantic states (Gauthier et al. 1992). assimilating enugh nutrients t meet these re- Cllecting methds.--we cllected three gslings in quirements is cmpunded by the declining as many nests during the hatching perid n 5 and 6 July (median hatching date f 3 July in 1993; n = 367 quality f arctic fd plants in late summer nests). Frm 15 July t 6 August, we randmly sht (Manseau and Gauthier 1993). Migratin is a 40 gslings frm wild families (ne per family) within critical perid fr the survival f yung geese, 4 km f the base camp. Frm 9 t 14 August, fur especially fr thse that d nt cmplete their banding drives were carried ut at fur different sites grwth in time. In yung Barnacle Geese (Branta within 8 km f the camp, just befre the fledging peleucpsis), mrtality during migratin between rid. In each capture, we sacrificed 24 gslings (12 Svalbard and Sctland (4,000 km) is high and in- males and 12 females). Within each sex, we randmly versely related t bdy mass near fledging selected six early-hatched and six late-hatched gs- (Owen and Black 1989). lings. Early and late-hatched gslings were defined as The bjectives f this study are t examine the thse >2 days r <2 days, respectively, than the median age f gslings caught in each banding drive (see pattern f grwth and maturatin f varius rbelw fr aging methd). Finally, n 22 August, we gans, and the prcess f fat accumulatin in sht an additinal seven gslings (all f them were fly- Greater Snw Gse (Chen caerulescens atlantica) ing, althugh their estimated ages shwed that they gslings frm hatching in the arctic until arrival tended t be late-hatched birds). at a temperate staging area, three mnths later. During the fall migratin, we captured gslings Greater Snw Geese are an interesting case be- with rcket nets at Cap-Turmente. The rcket nets cause they are large geese that have ne f the were installed in hayfields regularly used by geese. In nrthernmst nesting areas and shrtest brd- three captures (5, 19 and 20 Octber), we sacrificed 10 rearing seasns in Nrth America (Owen 1980). randmly selected individuals (five f each sex) amng the juveniles caught (42 t 45 juveniles were This prvided an pprtunity t evaluate Rickcaught each time). lefs' (1979a) tissue-allcatin hypthesis in a Birds were sexed by clacal inspectin. Mrphmetspecies breeding under extreme eclgical cn- ric measurements were taken accrding t Dzubin and ditins. Cch (1992) shrtly after cllectin by the same b- METHODS Study area. Greater Snw Gse gslings were cl- lected in tw different areas in summer and fall 1993. The first study area was lcated at the Bylt Island Natinal Wildlife Refuge, Nrthwest Territries (73ø08'N, 80ø00'W). The site is the mst imprtant nesting clny f Greater Snw Geese (abut 15% f the ttal ppulatin; Reed et al. 1992). The study area (50 km 2) was in the lwlands f a glacial valley and is characterized by wet plygn tundra cvered by a rich graminid cver (see Gauthier et al. 1995). The grwing perid extends frm early July t late August and the 24-h daylight perid frm early May t early August. The secnd study area was lcated 3,000 km farther suth at Cap-Turmente, a Natinal Wildlife Refuge server n all birds. Culmen, head, tarsus, and sternum lengths were measured with a caliper (_+ 0.1 mm). Lengths f the bdy, the 9th primary, and the lngest tail feather (when feather tips had emerged frm the sheath) were measured with a ruler (+ 1 mm). Birds were weighed (+ I g) with an electrnic field balance. Bdy cmpsitin analysis.--in feathered gslings, we remved feathers with a sheep shears and discarded them. The fllwing tissues were remved and weighed immediately n an electrnic balance (-+ 0.1 g): left breast muscles (pectralis and supracracideus), left leg muscles (including all muscles riginating alng the femur r the tibitarsus), esphagus (with prventriculus), gizzard, liver, and intestine. The length f the small intestine and caeca was measured (+ I mm), and the cntents f digestive tracts were remved, weighed, and subtracted frm bdy mass. The carcass, with all previus rgans

April 1997] Grwth in Greater Snw Gse Gslings 231 remved, als was weighed and then frzen. Frzen carcasses were passed thrugh a meat grinder three times and a 100-g sample f the hmgenate was retained. All tissues were then frzen in plastic bags until further analyses. Water cntent was determined by freeze-drying fr cntrl, early-hatched yung was used because it crrespnded mre clsely t the grwth f wild birds in the Lindhlm et al. study. Age f gslings cllecte during banding drives and shrtly after fledging was estimated using the length f the 9th primary (mm). The relatinship between age tissues t cnstant mass. Dried tissues were then and 9th-primary length was established n a sample grund t pwder. Fat was determined by weighing ( f knwn-age birds, i.e. gslings that were tagged in +_ 0.0001 g) duplicate 1-g dried samples in cellulse the nest at hatching and recaptured at banding in the thimbles and extracting them during 45 min in a same year. This was a mre accurate methd than the Rafatec apparatus (Randall 1974) using petrleum previus ne because it was established n birds that ether as slvent. Extracting time was verified t ensure ttal fat recvery (see Gauthier et al. 1992). Ttal bdy fat was calculated by adding the amunt f fat experienced the same grwth cnditins as the birds that we cllected. Because grwth f the 9th primary is linear (Byd and Maltby 1980, Lindhlm et al. 1994), determined in all tissues. Abdminal fat was absent we estimated the age (days) f gslings captured in in gslings n Bylt Island but was present during the fall at Cap-Turmente. When present, abdminal fat was als remved during the autpsy and its fresh banding drives r sht after fledging by the fllwing linear equatin (r 2 = 0.62, n = 102, P < 0.001; SE f estimate 1.7 days): mass added t ttal bdy fat because fat usually accunts fr >97% f the mass f this tissue. Prtein f age = 22.81 + (0.0991. 9th-primary length). (5) muscles and viscera was estimated using the lean (i.e. fat-free) dry mass, and carcass prtein using the ashfree lean dry mass because carbhydrate and ash are a negligible fractin f sft tissues (Drbney 1982). T- This methd culd nt be used with the yunger gslings cllecteduring the summer because n knwnage gslings were recaptured befre the banding petal bdy prtein and bdy water were the summatin rid. f prtein and water cntent f all bdy tissues (ex- Age f gslings caught at Cap-Turmente in fall was cluding feathers that were discarded), respectively. estimated by the number f days elapsed between the Bdy ash (again, excluding feathers) was estimated by incinerating a 3-g sample f dried, hmgenized carcass tissues in a muffle furnace at 550øC fr 12 h. Aging tnethds.--the exact age f cllected birds was unknwn (except fr thse cllected at hatching) and median hatching date n Bylt Island (3 July) and the capture date. The use f ther aging methds (i.e. 9thprimary length r bdy size) was nt pssible because grwth f mst bdy parts is cmpleted by then. Even thugh the methd based n median hatching date is therefre had t be estimated. Even thugh hatching less accurate, the relative errr n age is small in is highly synchrnus (87% f the 367 nests hatched ver a 7-day perid), aging birds by the difference in the number f days between cllectin date and the peak f hatching was t inaccurate in yung birds. Indeed, we fund a large variability in the develp- 3-mnth-ld gslings, which are almst fully grwn. Althugh the use f differentechniques t age birds may intrduce sme variability, it shuld nt affect cmparisns amng varius bdy cmpnents. Data analysis.--the sigmid grwth curve f each ment and size f individuals cllected n the same tissue was mdeled using the lgistic equatin (Rickdate. We chse t use a multivariate index f bdy size lefs 1983): t age gslings cllected befre the banding drives as recmmended by Gilliland and Ankney (1992). We applied the principal cmpnent equatin (PC1) develped by Lindhlm et al. (1994) in captive gslings at Bylt Island in 1991 based n culmen, head, and tarsus length (all in mm). This equatin was: PC1 = 0.335 sculmen + 0.335 shead + 0.334 starsus, (1) where sculmen = (culmen -29.61)/ 10.39, (2) shead = (head 68.01)/ 18.40, and (3) starsus = (tarsus -53.17)/ 17.67. (4) PC1 explained 99% f the variatin in the riginal measurements (Lindhlm et al. 1994). Age f individu- als was then btained frm the curve f Lindhlm al. relating age and PC1 in captive gslings. The curve et M(t) - -- (6) 1 + be - c t where a is the asymptte f the mass (g), b is the natural lgarithm f the rati f asympttic t initial mass, K is a cnstant prprtinal t the rate f mass increase (grwth rate), and t is age (days). These equatins were calculated using a nnlinear least squares estimatin technique. The lgistic equatin generates fixed-shape curves, in cntrast mre general grwth mdels, such as the Richards equatin, which allw shape t vary (Ricklefs 1983). The Richards equatin intrduces a 4th parameter, m, that cntrls the shape f the curve. When tn = 2, the Richards equatin is reduced t the lgistic equatin, whereas when tn appraches 1, it is reduced t the Gmpertz equatin (Ricklefs 1983). When we applied the Richards mdel t ur data, we fund

232 LESAGE AND GAUTHIER [Auk, Vl. 114 TABLE 1. Parameters frm lgistic grwth equatins (see Methds) fr bdy cnstituents and rgans (lean dry mass; ash-free lean dry mass fr carcass) in male (n = 84) and female (n = 89) Greater Snw Gse gslings. Adult masses are frm the perid f minimum bdy mass in the annual cycle (early March) utside the breeding seasn (Gauthier et al. 1992). K b a Adult mass (g) Males Females Males Females Males Females Males Females Bdy cnstituents Bdy mass 0.096 0.090 20.0 14.9 2,397 2,270 Bdy prtein a 0.094 0.089 23.1 17.2 514 472 Bdy ash a 0.123 0.127 100.0 94.0 88.5 79.2 Bdy water 0.124 0.117 31.7 21.7 1,400 1,312 Bdy rgans Breast muscles b 0.263 0.275 1.31 x 105 1.87 x 105 52.6 50.4 Leg muscles b 0.174 0.173 93.5 62.4 23.9 22.2 Esphagus 0.138 0.161 16.1 15.2 4.9 4.5 Gizzard 0.338 0.285 83.2 43.1 32.0 29.9 Intestine 0.146 0.156 14.8 18.7 19.2 18.6 Liver 0.110 0.102 14.0 11.9 13.8 14.0 Carcass 0.102 0.096 34.0 23.8 363.0 328.7 a Excluding bdy feathers. b Muscles frm left side nly. c Values frm geese cllected in April and May alng the St. Lawrencestuary (G. Gauthier unpubl. data). 2,666 2,391 561 516 109 c 91.5 c 64.4 57.8 25.7 21.1 34.3 34.0 15.2 c 13.8 c values f rn clse t, r higher than 2 (e.g. rn = 1.74 fr bdy mass). Hwever, we did nt retain the Riching f six females fitted with satellite transmitters (J.-E Girux pers. cmm.) indicated that deards equatin in ur final analyses because: (1) in sev- parture f geese frm Bylt Island fr the fall eral cases the nnlinear regressin algrithm did nt migratin ccurred arund 24. August 1993 cnverge t a slutin, a cmmn prblem when ap- (range 20 t 28 August), less than 10 days after plying the Richards equatin t variable data sets such the first gslings had been bserved t fly (15 as grwth f individual rgans (R. Ricklefs pers. cmm.); (2) the imprvement in the r 2 f the mdel August). The first gse families were seen at was very lw (-<1.5%) cmpared with the lgistic the Cap-Turmente staging area arund 1 Octmdel; and (3) three-parameter equatins simplify ber (A. Reed pers. cmm.), althugh arrivals ccmparisns f grwth curves amng different rgans curred thrughut mst f Octber (pers. bs.). and with ther studies. The prprtin f dry matter in lean tissues was At Cap-Turmente, the age f cllected gslings ranged frm 94 t 109 days (n = 30). used as an index f the functinal maturity f tissues Sex differences.--adult male Snw Geese are (Ricklefs 1979b, Ricklefs et al. 1994). The relatinship abut 12% heavier than females. All bdy cmbetween the prprtin f water in tissues and age pnents f males are larger than females, with was analyzed using plynmial regressins. We als differences ranging frm 10% in liver t 19% in examined the relatinship between the expnential grwth rate (the first derivative f the lgistic equa- bdy ash (Table 1). Althugh asympttic values tin) and the prprtin f dry matter in lean tissues f grwth equatins were higher in males than using linear regressin (Ricklefs 1983, Ricklefs et al. in females fr all bdy cmpnents except liver, 1994). The level f significance was set at P < 0.05 fr sex differences were smaller in asympttic valall statistical tests. ues than in adults (e.g. 6% difference fr bdy mass, 13% fr bdy ash). Despite these differ- RESULTS ences, grwth rates (as judged by K) f bdy cmpnents were similar between sexes (Table The age f gslings cllected between hatching and the banding perid ranged frm 1 t 34 days (n = 43). Gslings cllected during banding drives were between 29 and 45 days f age (n = 96), and thse cllected at fledging were 39 t 46 days f age (f = 43 days, n = 7) and capable f flying. Visual bservatins and track- 1). Grwth cnstants f bdy mass, bdy prtein, bdy water, leg muscles, gizzard, and liver were slightly higher in males than females, whereas the reverse was true fr bdy ash, breast muscles, esphagus, and intestine. Because f these small and incnsistent differences between sexes during grwth, data frm males

April 1997] Grwth in Greater Snw Gse Gslings 233 3000 2000' 1000' 0 10 20 30 40 S0 90 Age (days) // K L = 0.093 100 110 FIC. 1. Grwth curve fr bdy mass in Greater Snw Gse gslings. Open symbls = males; slid symbls = females; circles = gslings btained by shting; triangles = gslings caught in banding drives; squares = gslings caught in rcket nets. K L = grwth cnstant btained frm the lgistic equa- tin. and females were pled fr the cmparisn f grwth patterns amng bdy cmpnents. Bdy cnstituents.--mean bdy mass f gslings increased frm 80 g at hatching t 2,332 g in the St. Lawrence estuary, 110 days later (Fig. 1). Bdy prtein had a grwth pattern mst similar t bdy mass (Fig. 2A), althugh it was slightly delayed (50% f asympttic mass attained three days later; Table 2). The skeletal cmpnent (bdy ash) fllwed a similar grwth pattern, althugh its grwth was further delayed cmpared with bdy mass (50% f asympttic mass attained seven days later; Table 0. Age (days) ß Bedy m Bdy ash B dy prtein B dy water Gizzad FIG. 2. Grwth curves (% f asympttic mass) fr (A) bdy cnstituents and (B) rgans (prtein cntent) in Greater Snw Gse gslings. Bdy mass curves shwn fr reference. F = fledging age; S = age upn arrival at Cap-Turmente staging area. 100 TABLE 2. Timing f grwth f bdy cnstituents and rgans in Greater Snw Gse gslings. Fledging estimated at 43 days f age. Age (days) % f asympttic at 50% f value at fledging asympttic value Bdy cnstituents Bdy mass 76 30 Bdy prtein 71 33 Bdy ash 69 37 Bdy water 87 27 Bdy rgans Breast muscles 41 44 Leg muscles 95 19 Esphagus 97 18 Gizzard 100 13 Intestine 97 18 Liver 88 24 Carcass 76 34 2), and it eventually grew mre rapidly (grwth cnstant abut 36% higher; Fig. 2A). Increase in bdy water was slightly ahead f bdy mass (50% grwth attained three days earlier; Table 2) and was als faster (by abut 30%). Grwth f these cnstituents was nt cmpleted at fledging (43 days; prprtin f asympttic mass attained at fledging ranged frm 69 t 87%; Table 2). Fat was the cnstituent fr which accumula- tin was mst delayed. At hatching (0 t 2 days), vitellum accunted fr at least 4% f the bdy mass and was cmpsed f 85% f fat. This fat was rapidly catablized by gslings sn after hatching, and during mst f the grwing perid gslings were extremely lean (Fig. 3). Fat increased frm 1.5 + SE f 0.2 g (n = 9) at 10 days f age (0.6% f ttal bdy mass) t 12.0 + 0.5 g (n = 51) arund fledging at 40-45 days f age (0.7% f bdy mass). Accumulatin f fat must be very rapid thereafter because gslings cllected n the staging area 3,000 km farther suth had 281 + 15 g f fat (n = 30) at abut 100 days f age (12% f bdy mass). Organs.--The tw lcmtry muscle grups, leg and breast muscles, shwed dramatically different timing and pattern f grwth (Figs. 2 and 3). Leg muscles grew rapidly in the first days (50% f asympttic mass reached 11 days befre bdy mass; Table 2) and had almst cmpleted their grwth when gslings fledged arund 43 days (Table 2). In cntrast, breast muscles were the last rgan analyzed t initiate grwth, reaching nly 50% f asympttic mass near fledging. The asymptte estimated by the

234 LESAGE AND GAUTHIER [Auk, Vl. 114 500 / / ß 400 - ---r----'-; " "-;' ',,;,/,.,, 3. Leg muscles,,,.. --' 20- t - KL= 50-. 40 - reast muscles B 0.172 10- ß K L = 0.2 I 25-20- 5-5- KL = 0,149 0 0 10 20 30 50 1 110 Age (days) RG. 3. Bdy fat and grwth rves fr prtein cntent (lean d mass) f leg muscles, breast muscles (left side nly), and intestine f Greater Snw Gse gslings. en symbls = males; slid symbls = females; circles = gslings btained by shting; triangles = gslings caught in banding drives; squares = gslings caught in rcket nets. K L = grwth cnstant btained frm the lgistic equatin. grwth curve was nly 84% f adult size fr breast muscles cmpared with 99% fr leg muscles (Table 1), which suggests that significant grwth f breast muscles cntinued after three mnths. Despite the late nset f its develpment, the grwth cnstant fr breast muscles (K = 0.268) was much higher than that fr leg muscles (K = 0.172; Fig. 3), suggesting a very fast grwth f the breast. The grwth f all digestive rgans (esphagus, gizzard, intestine and liver) ccurred early in the develpment f gslings (Fig. 2B). The develpment f the gizzard, which was the largest muscle at hatching, fllwed a trajectry similar t that f the legs and was extremely early and rapid. The gizzard was the earliest r- gan t develp (50% f asympttic mass reached 17 days earlier than bdy mass) and had the highest grwth cnstant (K = 0.294). Develp- ment f the esphagus and intestine als ccurred early (50% f asympttic mass reached 12 days earlier than bdy mass; Table 2) and relatively rapidly (K = 0.143 and 0.149, respectively; Fig. 3). The liver was the last digestive rgan t develp althugh its grwth was still relatively early (50% f asympttic mass reached six days ahead f bdy mass; Table 2) and its grwth mderate (K = 0.104). Because f their early develpment, these fur rgans had reached 88 t 100% f their asympttic mass at fledging (Table 2). The remaining sft-bdy tissues (carcass) shwed a grwth pattern very similar t that fr bdy prtein (Tables 1 and 2). Water cntent.--water cntent differed bth in abslute values and pattern f tempral variatin amng the varius rgans. The patterns f change in water cntent with age fell int tw grups. In breast muscles and carcass, water cntent declined steadily thrughuthe grwth perid frm peak values at hatching r shrtly thereafter. In cntrast, water cntent f leg muscles, gizzard, intestine, and ttal bdy increased during grwth, peaked shrtly befre fledging, and declined thereafter as shwn by the significant plynmial regressins (Fig. 4). The largest change in water cntent ccurred in the breast muscles, with values decreasing frm an average f 82% sn after hatching t 71% at arund 110 days f age (lean mass basis; Fig. 4). Water cntent als was relatively high at hatching, r sn after, in the carcass (75-78%), but declined t lw values at 110 days (Fig. 4). Water cntent was lwest in the leg muscles (72%) and intestine (73%) at hatching but was relatively high at 110 days (75 and 77%, respectively) cmpared with the ther rgans (Fig. 4). Water cntent f the gizzard was especially variable in yung gslings. The tissue-allcatin hypthesis predicts an inverse relatinship between expnential grwth rate and functinal maturity f tissue. Using the prprtin f dry matter in lean tissues as an index f maturity, we fund n such relatinship in leg muscles, gizzard, and intes- tine, all f which are rgans that develped early (Fig. 5). In cntrast, the predictin was upheld in breast muscles, carcass, and ttal bdy prtein. Hwever, fr carcass and especially ttal bdy prtein, several pints were high abve the predicted line. These were mstly gslings with bth a high prprtin f dry matter in their tissue and a high expnential grwth rate, i.e. very yung gslings (Figs. 4 and 5).

April 1997] Grwth in Greater Snw Gse Gslings 235 85 8O 75 Leg R 2 -- 0.549 P < 0.001 Breast R2= 0.705 P < 0.001 70 --O I I I I I I I I I I I I g 85 8O 75 øz O' 70 65 6O 82 8O 78 76?Oø Gizzard R 2 -- 0.271 P< 0.001 I I I I I I R Carcass R 2 = 0.639 P<0.001 Intestine R 2 = 0.284 P< 0.001 I I Ttal bdy R2 = 0.655 74 72 7O -O 0 0 0 0-68 I I I I I I 0 20 40 60 80 1 O0 120 I I I I I I 0 20 40 60 80 100 120 Age (days) Age (days) FIG. 4. Water cntent (water mass/lean mass; ash-free lean mass fr carcass and ttal bdy) f leg muscles (left side nly), gizzard, carcass, breast muscles, intestine, and ttal bdy f Greater Snw Gse gslings. Plynmial regressins (including nly significant terms) are as fllws: leg (y = -1.15 ß 10-8 x 4 + 2.91-10 6 x 3 _ 2.48-10 -4x 2+ 7.70-10 3x+ 0.71); gizzard(y= -2.47.10 5x 2+3.44.10 3x+ 0.69); carcass(y- -5.43-10 6 X 2 d- 0.78)j breast (y = -1.16-10 3 x + 0.83); intestine (y = -8.39-10 9 X 4 d- 2.08 ' 10 6 X 3 1.77. 10 4 x 2 + 5.85-10 -3 x + 0.72); ttal bdy (y = 2.77-10 7 X 3 5.42. 10 5 x 2 + 2.49.10 3 x + 0.75). Water cntent (y) is expressed as a rati, nt a percentage, in the equatins.

236 LESAGE AND GAUTHIER [Auk, Vl. 114 0.15 Q Leg 0.25 = Breast 0.10, - _ 'ø.. 0.15 / ( a.. O P < 0.001 0.10 ; 0.05 e 0.05 O.f,:. -0.05-0.05 e -0.10-0.10 0.15 0.10 0.05 O 0.00 x 20 25 30 35 10 15 20 25 30 35 40 0.20 Intestine 0.15 t O R 2 = 0.015 P= 0.104 0.10 % 0.05 0.00-0.05-0.05-0.10 I I I 15 20 25 30 35 40 18 20 22 24 26 28 30 32 0.10 0.05 0.00 Gizzard R 2 = 0.001 ø P= 0.705 ½: 0 0 0 0 ø 0½3O Carcass OR 2=0.488 c cb P <0.001 ø'...;, % c% ø ø O0 0 0 0.10 0.05 0.00 O O ø Ttalbdy ^ ^ R 2--0.555 øø < 0.001 % -' ' -0.05 8 I I I I I i -0.05 I I I I I I 20 22 24 26 28 30 32 20 22 24 26 28 30 Dry matter cntent (%) Dry matter cntent (%) FIG. 5. Expnential grwth rate (first derivative f lgistic equatin) pltted against dry matter cntent f lean tissues (100- % f water; ash-free basis fr carcass and ttal bdy), the latter an index f tissue maturity, in Greater Snw Gse Gslings. Significant linear regressins: carcass (y = -0.793 x + 0.226); breast (y = -0.354 x + 0.283); ttal bdy (y = -0.646 x + 0.186).

April 1997] Grwth in Greater Snw Gse Gslings 237 DISCUSSION Cmparisns f grwth rates amng species are dependent upn the accurat estimatin f grwth parameters and the equatins used. Grwth equatins are sensitive t the inclusin f fully grwn yung in the data set (Bradley et al. 1984, Sedinger 1992). Lack f data pints near the asymptte may cause an underestimatin f K and an verestimatin f the asymptte value. When we applied the lgistic equatin t bdy mass data frm 0 t 50 days f age, the estimated parameters (K = 0.092, a = 2,350; bth sexes) were similar t thse ver the 110-day perid (Table 1), with asympttic values differing by less than 1%. Hwever, when we used the Gmpertz equatin, an equatin frequently used (Ricklefs 1983), the grwth cnstant f Greater Snw Geese calculated ver 0 t 50 days (K = 0.045) was underestimated cmpared with the 110-day perid (K = 0.062), whereas the asympttic mass was verestimated (0-50 days: a = 3,099; 0-110 days: a = 2,362). Because many studies fit grwth curves nly until fledging, sme grwth cnstant values may be biased dwnward when the Gmpertz equatin is used. 1986; in bth cmparisns, K fr Greaters was Patterns fgrwth.--because grwth rate is in- adjusted using K 2 = K [M2/M1]-0'22 t acversely related t bdy mass, allmetric equa- cunt fr size differences). The discrepancy with tins have been develped t allw interspecific Aubin et al.'s value is surprising althugh it is cmparisns. Ricklefs (1979a) described the re- pssible that his K values are verestimated belatinship between K (fr lgistic grwth) and cause f the statistical technique used t calcu- M (asympttic mass) in birds with the fllw- late the grwth equatin in this study (Aubin et ing equatin: K = 1.10M -ø'34. Using an adult al. 1986). bdy mass f 2,525 g (minimum winter mass; Even thugh the grwth rate f Greater Snw Table 1), this yields a predicted K value f 0.077 Geese appears similar t ther, smaller geese, fr Greater Snw Geese. The K value that we b- Greater Snw Geese fledge at the same age as tained in gslings (0.093) is higher than the pre- Lesser Snw Geese (43 days), but five days dicted value, which is surprising because the yunger than Cackling Geese. This is achieved equatin is largely based n altricial birds that by fledging at a lwer percentage f adult mass; shuld have faster grwth rates than preccial Greater Snw Geese fledge at 68% f adult bdy birds like geese. When cmpared with preccial birds such as gallifrms, geese als have a higher grwth rate. Fr instance, if we scale the grwth cnstant f Snw Geese t a bird the size f a quail (115 g; using the equatin K 2 = K [M2/M ].34), we btain a K f 0.266 cmpared with 0.106 fr Japanese Quail (Cturnix cturnix japnica; Ricklefs 1979b). This is in agreement with Ricklef's tissue-allcatin hypthesis, which states that birds with larger legs (the mst mature muscles at hatching) shuld have slwer grwth rate (leg muscles accunt fr 10% f bdy mass in Snw Geese at fledg- ing cmpared with 17% in Japanese Quail; Ricklefs 1983). Hwever, this predictin n lnger hlds within the family Anatidae because Snw Geese, which have bigger legs than ducks (Sedinger 1986), generally grw faster than ducks. Sedinger (1992) estimated the relatinship between adult bdy mass and grwth cnstant (Gmpertz equatin) within the Anatidae as: K = 0.31M -ø'22. Accrding t this equatin, the predicted Gmpertz K fr a bird the size f a Greater Snw Gse (0.055) is indeed lwer than the value that we btained (0.062). Althugh the waterfwl equatin may be biased dwnward by the inclusin f K values calculated n incmplete data sets (i.e. nly until fledging; see abve), this effect prbably is cunterbalanced by the inclusin f several studies n captive waterfwl that generally have faster grwth than in wild birds (Sedinger 1992, Lindhlm et al. 1994). The grwth cnstant fr Greater Snw Geese is similar t that fr Cackling Canada Geese (Branta canadensis minima; 0.072 vs. 0.074, respectively; Sedinger 1986) but lwer than that fr Lesser Snw Geese (0.100 vs 0.151; Aubin et al. mass (1,853 g), whereas Lesser Snw Geese fledge at 79%, and Cackling Canada geese at 89% f adult mass (Sedinger 1986). The shrter summer perid experienced by the Greater Snw Gse in its high breeding latitude may have favred fledging at an earlier stage f develpmenthan in the ther species. All these cmparisns suggest that eclgical factrs culd be mre imprtanthan embrynic cnstraints determinin grwth rates within the Anatidae. Selectin fr rapid grwth is strng in arctic-nestingeese because the very shrt summer frces gslings t grw as quickly

238 LESAGE AND GAUTHIER [Auk, Vl. 114 as pssible t reach fledging stage in time fr the fall migratin. Such a high grwth rate in a herbivruspecies is surprising. Indeed, geese are amng the very few species f birds whse yung rely entirely n plants t btain all f the nutrients required fr their grwth (Aubin et al. 1993). Other herbivrus birds have a diet that includes animal fd during grwth because plants are prer in prtein than are animal tis- sues, and grwth requires large amunts f prtein (Sedinger 1992, Mss et al. 1993). Hwever, the fast grwth rate in geese is favred by the 24-h daylight f arctic regins during the summer, which allws fr cntinuus feeding. The grwth cnstant we measured (K = 0.093) was lwer than that f gslings raised under artificial cnditins n Bylt Island, i.e. handreared, early hatched gslings with fd supplement (K = 0.168; Lindhlm et al. 1994). This difference further suggests that fd availability limits grwth rate in geese, either thrugh a limitatin in quality (Manseau and Gauthier 1993, Lindhlm et al. 1994) r quantity (Gauthier et al. 1995). Grwth f individuals rgans.--the grwth f all prtein synthesized in the bdy is in the breast muscles alne. Thus, grwth f this large muscle mass mnplizes mst f the resurces devted t grwth by gslings arund fiedging time. Delaying develpment f these muscles until the end f grwth allws gslings t maximize allcatin f internal resurces t the grwth f large supplying rgans such as the gizzard and legs at a yunger age. In return, the very fast grwth rate f breast muscles near fiedging is pssible because supplying rgans have almst reached their adult size. Grwth rate and functinal maturity.--accrding t the tissue-allcatin hypthesis, functinal maturity and rapid embrynic grwth are incmpatible at the tissue level (Ricklefs et al. 1994). The hypthesis predicts an inverse relatinship between expnential grwth rate and maturity at the tissue level. The prprtin f dry matter in muscles is assumed t be a gd index f maturity because it increases with the accumulatin f cntractile prteins and ther functinal elements as tissues mature (Ricklefs 1979b, Ricklefs et al. 1994). A strng inverse relatinship between expnential grwth rate and maturity was nly fund in the breast muscles. The relatinship individual rgans shwed very cntrasting patterns in Greater Snw Geese. Bdy rgans can be divided in three grups accrding t their grwth pattern (Fig 2). The first grup, the supplying rgans (Lilja 1981), is cmpsed f leg muscles and digestive rgans (esphagus, gizzard, intestine, and liver) that are characterized by an early and rapid grwth sn after hatching. Investment in supplying rgans is maxials was present in the carcass and ttal bdy prtein, but there was cnsiderable variatin in yung gslings. In cntrast, tissues that develped early and were mst mature at hatching (leg muscles, gizzard, and intestine) shwed n inverse relatinship between grwth rate and maturity. These tissues grew rapidly even mized early in develpment because gslings thugh their water cntent was relatively lw. need mature legs in rder t find fd and ma- These results again suggesthat embrynic cnture digestive rgans t prcess fd efficiently. straints are nt the primary factr limiting The rapid grwth f the gizzard is especially impressive; between 8 and 16 days f age, 40% f all prtein synthesized in the bdy is in this sle rgan (estimated frm Fig. 2). In additin t its essential digestive functin, this rgan als culd play a rle in thermgenesis f yung gslings because it is the largest muscle at hatchgrwth rate in geese. Hwever, cnstraint n grwth at the rganism level (Ricklefs 1979a) culd be imprtant in geese. There is evidence that nutrient intake in grwing gslings is cnstrained by gut capacity (Sedinger and Raveling 1988, Manseau and Gauthier 1993, Piedbeuf 1996). When fd quality declines, gslings are ing (Sedinger 1986). The secnd grup f rgans nt able t prcess a larger quantity f fd t t develp includes bnes (bdy ash) and car- maintain their nutrient intake, a cmmn phecass (verall bdy prtein), which are character- nmenn in herbivres. ized by a mderate grwth rate. The increase in water cntent f several tis- Finally, the breast muscles shw a unique sues frm lw values at hatching cntrasts with grwth pattern. It is the last rgan t develp the typical pattern in birds f a decline in water but des s at a very fast rate, which cncen- cntent frm maximum values at hatching when trates the resurce requirements fr the develpment f these muscles in a shrt span f time. Indeed, frm 40 t 54 days, mre than 65% f tissues are least differentiated (Ricklefs 1979b, Ricklefs and White 1981). Hwever, the pattern that we bserved is nt unique because ther

April 1997] Grwth in Greater Snw Gse Gslings 239 studies als have reprted an initial increase f water cntent in grwing tissues f waterfwl fat cntent f the bdy during the days fllwing hatching. This crrespnds t a transitin (Clay et al. 1979, Aubin et al. 1986, Sedinger perid when ylk reserves are mbilized t sus- 1986). This culd be a cnsequence f the preccial mde f develpment in anatids. In rder tain metablism while gslings becme cmpetent in finding fd (Clay et al. 1979). After this t be functinal at hatching, supplying rgans initial decline, the increase in fat cntent during must reach an advanced stage f maturatin in the egg, which increases their dry matter cntent at hatching. Hwever, this explanatin is at dds with the fact that preccial gallifrms such grwth was extremely slw, as bserved in ther gse species (Aubin et al. 1986, Sedinger 1986). Thus, the grwth perid was characterized by a very lw level f fat (<1% f bdy mass). The as quail generally have maximum water cntent absence f fat accumulatin befre fledging is values at hatching fllwed by the typical pattern f linear decline (Ricklefs 1979a). surprising given the early nset f migratin in this species (i.e. shrtly after fledging). It ap- As an alternative explanatin, we prpse pears that all ingested nutrients largely were althat a much greater requirement fr thermgen- lcated t prtein prductin rather than t fat esis in newly hatched anatids cmpared with gallifrms explains the high dry matter cntent f sme their tissues at hatching. Unlike gallifrms, anatids enter the water in the first hurs after hatching. This increases their metablic requirements cnsiderably because in cld envirnments, the rate f heat lss is much faster in accumulatin at fledging. By the time geese arrived at the Cap- Turmente staging area, 4 t 5 weeks after leaving Bylt Island, gslings had accumulated large amunts f fat. Althugh sme fat may be stred in the few days fllwing fledging, this suggests that gslings leaving Bylt Island fr water than in air (Schmidt-Nielsen 1991). The migratin (ca. 10 days after fledging) cannt fly capacity fr thermgenesis directly related t the dry matter cntent f tissues in newly lng distances. Gslings will need t stp frequently, at least in the first part f the migrahatched Eurpean Starlings (Sturnus vulgaris) tin, t accumulate fat stres, and this shuld and quail (Ricklefs and Webb 1985, Chi et al. 1993). Sme results in waterfwl are cnsistent with this hypthesis. The metablic scpe (defined as the difference between peak and minimum V02; Chi et al. 1993) is abut 2.2 ml impse a slw rhythm f migratin n their parents (family grups migrate tgether). This culd explain the lng fall migratin (4 t 5 weeks) f Greater Snw Geese in cntrast with the shrt spring migratin (<1 week; Gauthier O2/g/h in newly hatched (1-3 days ld) Greater and Tardif 1991). Gslings must therefre have Snw Geese (Ratt6, LaRchelie, and Gauthier access t high-quality feeding areas during miunpubl. data) and 2.8 ml O2/g/h in 1-day-ld gratin in rder t gather enugh energy and Cmmn Eiders (Smateria mllissima; Steen et al. 1989), values twice as high as in newly hatched quail (1.2 ml O2/g/h; Chi et al. 1993). Visser and Ricklefs (1993) als reprted much higher peak metablic rates in the Anatidae cmpared with ther preccial birds. The requirement fr high metablic rate culd have favred a high dry matter cntent in tissues at hatching while maintaining the ptential nutrients t cmplete the grwth f sme tissues (e.g. breast muscles), cver the cst f migratin, and arrive n the fall staging area with a surplus f fat. The small fat stres f gslings at the nset f migratin prbably reduce their ability t cpe with harsh envirnmental cnditins encuntered during migratin, such as bad weather and pr fd availabilit and may increase their mrtality. fr rapid grwth. The subsequent increase in water cntent f rgans during grwth wuld ACKNOWLEDGMENTS ccur because new relatively immature tissues wuld be added t these rgans thrugh cell This study was funded by the Natural Sciences and Engineering Research Cuncil f Canada, a team grant prliferatin. In cntrast, water cntent peaks frm the Fnds pur la Frmatin de chercheurs et early in rgans that are t immature at hatch- d'aide la recherche f the Quebec Gvernment, Ening t cntribute t thermgenesis (e.g. breast virnment Canada (Arctic Gse Jint Venture), and muscles). a Department f Indian and Nrthern Affairs student Fat dynamics.--as in ther waterfwl species grant t L.L. We als received generus lgistical sup- (Ankney 1980, Sedinger 1986), newly hatched prt frm the Plar Cntinental Shelf Prject. We Greater Snw Geese shwed a rapid decline in thank Julien Beaulieu, Anne-Marie Culmbe, Rch

240 LESAGE AND GAUTHIER [Auk, Vl. 114 Gu rin, Jack Hughes, Michel Lalibert, Martin Lambert, Denis Lepage, G rald Picard, and Jean-Pierre Tremblay fr their assistance in the field and in the lab; and Jean B dard, Lehr Brisbin, Jr., Jacques Larchelie, and Rbert Ricklefs fr cmments n an earlier draft f this paper. We als thank the Pnd Inlet Hunters and Trappers Assciatin fr allwing us t wrk n Bylt Island. L.L. was supprted by a schlarship frm the Natin Hurnne Wendat. LITERATURE CITED ANKNEY, C. D. 1980. Egg weight, survival, and grwth f Lesser Snw Gse gslings. Jurnal f Wildlife Management 44:174-182. AUdeN, A. E., E. H. DUNN, AND C. D. MACINNES. 1986. Grwth f Lesser Snw Geese n arctic breeding grunds. Cndr 88:365-370. AUBIN, A. E., A. DZUBIN, E. H. DUNN, AND C. D. MACINNES. 1993. Effects f summer feeding area n gslings grwth in Snw Geese. Ornis Scandinavica 24:255-260. BfiSRNHAG, G. 1979. Grwth in newly hatched birds. Swedish Jurnal f Agricultural Research 9:121-125. BOYD, H., AND L. S. MALTBY. 1980. Weights and primary grwth f Brent Geese Branta bernicla multing in the Queen Elizabeth Islands, N.W.T., Canada, 1973-1975. Ornis Scandinavica 11:135-141. BRADLEY, D. W., R. E. LANDRY, AND C. t. COLLINS. PIEDBOEUF, N. 1996. Qualit nutritive des sites 1984. The use f jacknife cnfidence intervals with the Richards curve fr describing avian grwth patterns. Bulletin f the Suthern Califrnia Academy f Sciences 83:133-147. CHOI, I.-H., R. E. RICKLEFS, AND g. E. SHEA. 1993. Skeletal muscle grwth, enzyme activities, and the develpment f thermgenesis: A cmparisn between altricial and preccial birds. Physilgical Zlgy 66: 455-473. d'alimentatin des isns de la grande ie des neiges: est-il avantageux d'utiliser des sites d jh brut s? M.S. thesis, Universit Laval, Sainte-Fy, Quebec. RANDALL, E. L. 1974. Imprved methd fr fat and il analysis by a new prcess f extractin. Jurnal f the Assciatin f Official Analytical Chemists 57:1165-1168. REED, g., H. BOYD, P. CHAGNON, AND J. HAWKINGS. CLAY, D. L., I. L. B ISBIN, JR., AND K. A. YOUNGSTROM. 1992. The numbers and distributin f Greater 1979. Age-specifi changes in the majr bdy cmpnents and calric values f grwing Wd Ducks. Auk 96:296-305. Snw Geese n Bylt Island and near Jungersen Bay, Baffin Island, in 1988 and 1983. Arctic 45:115-119. DROBNEY, g.d. 1982. Bdy weight and cmpsitin changes and adaptatins t breeding in Wd Ducks. Cndr 84:300-302. DZUBIN, A., AND E. COOCH. 1992. Measurements f geese: General field methds. Califrnia Waterfwl Assciatin, Sacrament, Califrnia. GAUTHmR, G., J.-E GIROUX, AND J. B DARD. 1992. Dynamics f fat and prtein reserves during winter and spring migratin in Greater Snw Geese. Canadian Jurnal f Zlgy 70:2077-2087. GAUTHIER, G., R. J. HUGHES, A. REED, J. BEAULIEU, AND L. ROCHEFORT. 1995. Effect f grazing by Greater Snw Geese n the prductin f graminids at an arctic site (Bylt Island, N.W.T., Canada). Jurnal f Eclgy 83:653-664. GAUTHIER, G., AND J. TARDIE. 1991. Female feeding and male vigilance during nesting in Greater Snw Geese. Cndr 93:701-712. GILLILAND, S. G., AND C. D. ANKNEY. 1992. Estimating age f yung birds with a multivariate mea- sure f bdy size. Auk 109:444-450. LIGHTBODY, J.P., AND C. D. ANKNEY. 1984. Seasnal influence n the strategies f grwth and devel- pment f Canvasback and Lesser Scaup ducklings. Auk 101:121-133. LILJA, C. 1981. Pstnatal grwth and rgan develpment in the gse (Anser anser). Grwth 45:329-341. LINDHOLM, A., G. GAUTHIER, AND A. DESROCHERS. 1994. Effects f hatch date and fd supply n gsling grwth in arctic-nesting Greater Snw Geese. Cndr 96:898-908. MANSEAU, M., AND G. GAUTHIER. 1993. Interactins between Greater Snw Geese and their rearing habitat. Eclgy 74:2045-2055. MOSS, g., A. WATSON, g. A. PARR, I. g. TRENHOLM, AND M. MARQUISS. 1993. Grwth rate, cnditin and survival f Red Gruse Lagpus lagpuscticus chicks. Ornis Scandinavica 24: 303-310. OWEN, M. 1980. Wild geese f the wrld. B. T. Batsfrd, Lndn. OWEN, M., AND J. M. BLACK. 1989. Factrs affecting the survival f Barnacle Geese n migratin frm breeding grunds. Jurnal f Animal Eclgy 58:603-618. RICKLEFS, R. E. 1973. Patterns f grwth in birds. II. Grwth rate and mde f develpment. Ibis 115:177-201. RICKLEFS, g. E. 1979a. Adaptatin, cnstraint, and cmprmise in avian pstnatal develpment. Bilgical Reviews f the Cambridge Philsphical Sciety 54:269-290. RICKLEFS, R. E. 1979b. Patterns f grwth in birds. V. A cmparative study f develpment in the Starling, Cmmn Tern and Japanese Quail. Auk 96:10-30. RICKLEFS, R. E. 1983. Avian pstnatal develpment. Pages 1-83 in Avian bilgy, vl. 7 (D. S. Farner, J. R. King, and K. C. Parkes, Eds.). Academic Press, New Yrk.

April 1997] Grwth in Greater Snw Gse Gslings 241 RICKLEFS, R. E., R. E. SHEA, AND I.-H. CHOI. 1994. f breeding waterfwl (B. D. J. Bart, A.D. Aftn, Inverse relatinship between functinal maturity M. G. Andersn, C. D. Ankney, D. H. Jhnsn, and expnential grwth rate f avian skeletal J. A. Kadlec, and G. L. Krapu, Eds.). University muscle: A cnstraint n evlutinary respnse. f Minnesta Press. Minneaplis. Evlutin 48:1080-1088. SEDINGER, J. S., AND D. G. RAVELING. 1988. Fraging RICKLEFS, R. E., AND T. WEBB. 1985. Water cntent, behavir f cackling Canada Gse gslings: Imthermgenesis, and grwth rate f skeletal plicatins fr the rles f fd availability and muscles in the Eurpean Starling. Auk 102:369- prcessing rate. Oeclgia 75:119-124. 376. STEEN, J. B., H. GRAY, B. BORCH-IOHNSEN, AND G. W. RICKLEFS, R. E., AND S.C. WHITE. 1981. Grwth and GABRIELSEN. 1989. Strategies f hmethermy in energetics f the Sty Tern (Sterna fuscata) and eider ducklings (Smateria mllissima). Pages 361- Cmmn Tern (S. hirund). Auk 98:361-378. 370 in Physilgy f cld adaptatin in birds (C. SCHMIDT-NIELSEN, K. 1991. Animal physilgy: Ad- Bech and R. E. Reinertsen, Eds.). Plenum Press, aptatin and envirnment. Cambridge University New Yrk. Press, Cambridge, United Kingdm. VISSER, G. H., AND R. E. RICKLEFS. 1993. Develpment SEDINGER, J. S. 1986. Grwth and develpment f f temperature regulatin in shrebirds. Physi- Canada Gse gslings. Cndr 88: 169-180. lgical Zlgy 66:771-792. SEDINGER, J. S. 1992. Eclgy f prefiedging waterfwl. Pages 109-117 in Eclgy and management Assciate Editr: M. E. Murphy