The Condor92:904-9 12 Q The Cooperomitholc&.al society1990

PATTERNS AND VARIABILITY OF GROWTH IN THE YELLOW-EYED ’

YOLANDA VAN HEEZ& Departmentof Zoology,University of Otago,P.O. Box 56, Dunedin,

Abstract. Growth of Yellow-eyed Penguin( antipodes) chicks was monitored at six breeding areas over one season,and at two breeding areas over three seasonsto describethe growth pattern and determine the degree of interseasonaland interlocality variation in growth rates.Growth parametersmeasured were weight, bill depth, Ripperarea, and lengthsof culmen, skull, foot, central tail feathers,wing (humerus), and flipper. Data were fitted to logisticcurves which were comparedusing a weightedleast squarestechnique from the Maximum Likelihood Program (Ross et al. 1980). Observed seasonaland geo- graphical variations in growth rates were small and appeared to reflect variations in food supply. Variations in growth rates of moxphometricparameters were small or absent, with few differencesin final fledgingsize. Dimensions with invariable growth rates were those that were smaller than average at maturity. Skull and foot + claw were least variable, implying high priority in the allocation of resourcesduring development.Foot + claw was first to reach asymptotic size, followed by wing and llipper (area and length). This pattern of growth is consistentwith that of other penguin .Feet may grow quickly to allow heat loss until the flippers can function as thermal windows. Bill dimensionsdid not reach adult size by the end of the fledgingperiod. Key words: Yellow-eyedPenguin; Megadyptes antipodes; growth; growth rates; growth pattern.

INTRODUCTION scribe the growth pattern of the Yellow-eyed Ricklefs (1973) hypothesizedthat the growth rate Penguin, to determine whether growth rates are of a speciesis determined within narrow limits variable, and if so, the magnitude of such vari- set by adult body size and the rate of develop- ations. The following questions were asked: (1) ment of mature functions, suchas flight and ther- Do growth rates vary according to locality and moregulation. He identified a number of influ- season?(2) How much do growth rates vary? (3) encesthat can result in variations in growth rates Are some morphometric parameters more vari- within a population: variation in diet quality, able than others? (4) Do variations in growth quantity of food, temporal pattern of food avail- rates result in variations in fledging sizes?Rel- ability, and temperature (Ricklefs 1983). Vari- ative growth rates of one- and two-chick broods, ability in these conditions may arise in relation and first- and second-hatched chicks are dealt to locality, season,habitat, weather, brood size, with in van Heezik and Davis (1990). egg size and composition, position in hatching sequence,age and experience of the parents, and STUDY POPULATION individual variability in the quality of parental Yellow-eyed are the fourth largest of care. the penguins (Stonehouse 1976), and are endem- Few studies, particularly on penguins, have ic to the southeasterncoast of South Island, New monitored growth over more than one locality Zealand, Stewart Island, and in the sub-Antarctic or seasonwith the purpose of identifying the ex- region, the Auckland Islands and Campbell Is- tent of intraspecific variation in growth rates. land. Breeding areasare located in areasof coast- The aim of this study was to examine and de- al forest or scrub. Yellow-eyed Penguins are se- cretive nesters,with nest sites aboveground, but isolated from one another (Darby, pers. comm.). I Received26 January 1990. Final acceptance28 The breeding cycle on the Otago coast has been June 1990. described by Richdale (1957). Two eggsare laid * Presentaddress: Percy Fitzpatrick Institute of Af- rican Ornithology,University of CapeTown, Ronde- in September to October, and hatching occursin bosch7700, CapeTown, SouthAfrica. the first 3 weeks of November. The duration of YELLOW-EYED PENGUIN GROWTH 905 the guard phase,when one adult is always present ample, monitoring chick growth at 34 nests at at the nest, is about 6 weeks, ending in the last Nugget Point required 12-13 hr, and 14 nests at week of December. Full juvenile feathering is Boulder Beach required 7-8 hr. Breeding areas attained in February, and chicks leave the breed- are small, and up to 250 km apart. In the post- ing area and go to seain the last week of February guard phaseschicks wander from the nest site so and first week of March, about 105 days after that considerable time was spent searching for hatching. Fledging period is usedhere to describe chicks in the dense scrub. the time from hatching to departure from the I weighed chicks with a range of Pesola bal- breeding area to the sea. Adults continue to feed ances (300 g + 2 g, 600 g + 10 g, 1 kg f 20 g, chicks right up till their departure, after which 2 kg + 50 g, 5 kg + 100 g, 10 kg +- 100 g) and parental input ceases(Richdale 1957). made the following measurements with Vernier calipers(accurate to + 0.5 mm): bill length (length METHODS of culmen), bill depth and flipper length (after Scolaro et al. 1983), skull length (from the bulge I identified variations in growth rates by mea- of the occipital condyle to the tip of the culmen), suring a number of parametersdescribing growth: tarsus plus middle toe and claw (= foot + claw, weight (WT), biIl length (BL), bill depth (BD), after Warham 1972), tail (center rectrices), and skull length (SK), and foot + claw (FC), tail (T), wing (from the middle of the joint at the proximal wing (W), and flipper length (FL). Although end of the humerus to the elbow). Flipper area weight comparisons are most frequently used to was traced out onto a card. Individual chicks in examine intraspecific variation, either as fledging two-chick nests were marked with cotton flipper weightsor as a fitted curve, body weight has been tagsuntil their flippershad developedto the point found to respond most sensitively to changesin where steelor aluminum bands could be applied. nutritional status (LeCroy and LeCroy 1974, I estimated dates of hatching (accurate to + 1 Schreiber 1976, Wingham 1984), making it dif- day) according to the following distinct and con- ficult to distinguish inherent changesfrom strict- sistent characteristics:on the f&t day chicks may ly nutritional effects(Ricklefs 1968). A more cau- be wet, eyes are shut, and fragments of eggshell tious approach is the comparison of fitted curves remain in the nest bowl. On the second day the of morphometric measurements, which appear eggshell has been removed from the nest bowl to respond more conservatively to changes in and the eyes are partially open. By the third day, nutritional status(Gavin0 and Dickerman 1972, the eyes are fully open, although the third eyelid LeCroy and LeCroy 1974, Schreiber 1976, is over the eye, but by the fourth day this eyelid Wingham 1984). has assumed the normal position. After 5 days, nestlings show the following behaviors; biting at THE 1983-l 984 BREEDING SEASON nesting material, sneezing, wing stretching, tail I located nestsat six breeding areason the south- wagging, and defecating over the nest edge (Sed- eastern coastof South Island, New Zealand, Shag don 1990), and on the seventh day secondary Point (45”2O’S, 170”40’E), Alfred and Cicely down first appears on the scapulars (Richdale Beach (45”5O’S, 170”40’E), Boulder Beach 1957). Loss of down and development of feath- (45”5O’S, 170”30’E), Green Island (45”5O’S, ering was noted, and dates of departure were 170”2O’E), and Nugget Point (46”2O’S, 169’4O’E) recorded (accurate to & 9.3 days). during the incubation period in October and ear- ly November. I visited each breeding area weekly except Boulder Beach, which I visited twice THE 1984-1985 AND 1985-1986 weekly, and Green Island, to which visits could BREEDING SEASONS only bc made when the opportunity arose. During the 1984-1985 and 1985-1986 breeding I made a total of seven visits to Green Island, seasons I followed growth at the two largest at 2-week intervals at the beginning and the end mainland breeding areas; Boulder Beach and of the breeding season,and two visits at 4-week Nugget Point. The measurements listed above intervals during December and January. More were made at weekly intervals. Flipper area was frequent visits to all breeding areaswere not pos- not measured. Hatching dates, departure dates sible due to the dispersed distribution of nests (in 1985-l 986 only), and development of plum- and the density of the nesting habitat. For ex- age were recorded. 906 YOLANDA VAN HEEZIK

TABLE 1. Time taken for Yellow-eyedPenguin chicks to reach 90% of the asymptotic value.

Weight 80 25 Bill length 15 Bill depth 40 :: 15 Skull 47 10 0 20 40 60 80 100 120 Foot + claw 25 :: 10 wing 40 12 Days after hatching Flipper length :: 40 10 Flipper area 45 63 18 FIGURE 1. Logisticgrowth curvesfitted to all mor- phometric parametersof Yellow-eyed Penguinchicks: A-flipper length; B-skull length; C-foot + claw; D-wing length;E-tail; F-bill length;G-bill depth. development is expressedas the time taken to reach 90% of the asymptotic value then foot + claw was found to reach this size fastest,followed CURVE FITTING by wing and flipper lengths (Table 1). Growth I fitted growth data to logistic and Gompertz rates of bill length and weight were the slowest. curves and compared curves using a weighted Variation also exists in the range of days over least squares technique from the Maximum which all chicks reach 90% of their asymptotic Likelihood Program, MLP (Ross et al. 1980, see value. Weight varied the most, followed by flip- also Ricketts and Prince 198 1, Gales 1987). This per area, whereas skull, foot + claw, and flipper fits curves for each data set and then pools the length varied the least. data to fit a single curve. The change in the re- A comparison of mean adult and chick (within sidual sum of squares is then treated for signif- 2 weeks before fledging) sizes showed that bill icance by analysis of variance. If this test for dimensions and wing length continued to grow parallelism shows nonsignificant differencesbe- after fledging (Table 2). Foot + claw size of adults tween curves, then the model with common pa- is significantly smaller than that of chicks. Some rameters should be preferred. If a significant dif- individuals measured as fledglings were caught ferenceis indicated, then individual curvesshould again and measured as juveniles (the juvenile be used. I made single and multiple comparisons stage extends from departure from the natal of growth curves using parallel curve analysis, breeding area as a fledgling until the following comparing both the shape and the scale of the molt); all (n = 14) of these had longer bill curves. If more than four curves were compared dimensions than they did as chicks, 77% (n = at one time, significance was set at the 1% level. 13) had longer skull and flipper lengths, 92% (n The logistic equation fitted marginally better = 12) had longer wing lengths, whereas 62% (n than the Gompertz curve in the majority of com- = 13) had smaller foot + claw measurements. parisons, so all analyses were done using the lo- The first feathers to appear were the central gistic equation. Log equations were also found tail rectrices, followed by the development of to best describe biomass, culmen, and flipper feathers on the ventral flippers and rump, and growth of Adelie Pygoscelisadeliae, Chinstrap finally the head and breast feathers (Table 3). P. antarctica, and Gentoo P. Papua penguins Timing of plumage development in the 1983- (Volkman and Trivelpiece 1980), and Adelie 1984 seasondiffered between breeding areas;de- Penguin biomass (Taylor and Roberts 1962), and velopment of feathers on the rump, head, and weight, feet, flipper, cuhnen length, and bill depth breast was earlier by about 4 days, and chicks of Blue Penguins minor minor (Gales attained full feathering about a week earlier at 1987). Boulder Beach (BB) than at Nugget Point (NP) (BB: 88.1 f 7.0 days, = 12 February 1984, n = RESULTS 14, NP: 95.7 +- 4.1 days, = 2 February 1984, n PATTERN OF GROWTH = 27, t = 4.40, df = 39, P c 0.001). Since the Certain morphometric parameters developed lengths of the fledging periods were not signifi- more rapidly than others (Fig. 1). If the speedof cantly different (BB: 102.8 f 4.3 days, = 24 Feb- YELLOW-EYED PENGUIN GROWTH 907

TABLE 2. Comparativemorphometric parameters of adult and fledglingYellow-eyed Penguins. Significance is as follows:P > 0.05 (ns),P -C0.05 (3, and P < 0.001 (*“).

Adult Chick n R SD n f SD P Bill length(mm) 119 55.3 1.89 38 53.8 1.66 *** Bill depth(mm) 121 19.6 0.89 38 18.0 0.66 ***

Skull length(mm) 118 142.3 4.36 37 142.9 3.97 llS Foot + claw length(mm) 110 135.1 4.17 :3 137.7 4.38 *** Wing length(mm) 112 104.0 4.07 101.9 2.82 * Flipper length(mm) 115 229.1 5.84 37 227.7 6.24 ns Flipper area(mm2) 62 93.1 4.39 28 94.1 5.22 ns

ruary 1984, n = 14, NP: 102.2 f 3.7 days, = 24 F= 4.13, P < 0.05). Only when thesetwo seasons February 1984, n = 26, t = 0.462, df = 38, P > were compared to 1985-l 986 did differencesoc- 0.05), chicks at Boulder Beach were fully feath- cur(1983-1984vs. 1985_1986;df=2,638,WT; ered 7-2 1 days before departure, whereasat Nug- F = 27.3, P < 0.001, BL; F = 5.84, P < 0.01, get Point chicks were feathered for only 2-10 BD; F = 7.57, P < 0.001, W; F = 5.82, P < days before leaving the breeding area. 0.01, FL; F = 11.24, P < 0.001; 1984-1985 vs. 1985-1986; df = 2, 444, WT; F = 12.89, P < INTRASPECIFIC VARIATION 0.001, W; F = 6.82, P < 0.01, n; F = 0.04, P IN GROWTH RATES < 0.01). Comparison of growth rates over three seasonsat No one seasoncaused most of the variation at two breeding areas. A simultaneous comparison Boulder Beach. Only skull length did not vary in of growth rates over three seasonswas made on the three-seasoncomparison of growth rates (df growth data from Nugget Point and Boulder = 3, 603, WT; F = 9.5, P < 0.01, BL; F = 5.3, Beach (Fig. 2), and then paired comparisons of P < 0.01, BD; F = 2.4, P < 0.05, Fc; F = 3.2, growth rates were made between seasonsto de- P < 0.05, T; F = 3.5, P < 0.01, W; F = 4.0, P termine whether morphomettic variation was the < 0.01, FL, F = 29.5, P -c 0.01). The two most result of one or more seasons.At Nugget Point, different seasons were 1983-1984 and 1984- skull, foot + claw, and tail lengths were the only 1985, from which comparison all parameters ex- parametersnot to showdifferences in growth rates cept skull showedsignificant differences in growth in the three-seasoncomparison (df = 3,79 1, WT; rates (df = 2, 542, WT, F = 10.7, P -c 0.001, F = 14.3, P < 0.01, BL; F = 2.7, P < 0.05, BD; BL; F = 6.2, P < 0.01, BD; F = 4.7, P < 0.01, F=4.2,P<0.05,SK;F=l.l,ns,FC,F=1.4, FC; F = 5.32, P < 0.01, T; F = 5.49, P < 0.01, ns, T, F= 1.4, ns, W, F= 4.7, P < 0.01, FL; F W; F = 4.88, P < 0.01, n; F = 60.32, P < = 5.7, P < 0.01). Growth rates compared be- 0.001). When growth rates from the 1985-1986 tween 1983-1984 and 1984-1985 did not vary seasonwere compared to those from 1983-1984, for all parameters except bill length (df = 2,644, only weight (F = 9.74, P < O.OOl), bill length (F

TABLE 3. Plumagedevelopment of Yellow-eyedPenguins at two breedingareas during the 1983-l 984 season. Fl = daysat whichfully feathered,F2 = fledgingperiod. ns = P > 0.05; ** = P i 0.01.

Nugget Point Boulder aeach 1983-1984 n x WY@ SD n x (days) SD P Tail 38 40.2 2.08 14 39.1 2.03 ns Ventral flippers 36 54.6 3.52 14 52.6 4.07 ns Rump 39 56.1 3.64 14 53.0 3.00 ** Head 35 68.9 3.85 14 65.1 4.06 ** Breast 35 68.6 2.96 14 64.7 5.59 t* Fl 27 95.7 4.10 14 88.1 7.0 ** F2 26 102.2 3.70 14 102.8 4.30 ns F2 - Fl 26 6.6 4.30 14 14.6 6.10 ** 908 YOLANDA VAN HEEZIK

(a) (b)

- 6-

1 0 20 40 60 60 100 120

Days after hatching FIGURE 2. Logistic weight curves fitted to all Yellow-eyed Penguin chicks during three breeding seasons: 1983-1984 (solid line); 1984-1985 (broken line); 1985-1986 (punctuatedline), at two breedingareas: (a) Nugget Point and (b) Boulder Beach.

= 5.52, P < 0.01) and flipper length (F = 3.43, t = 1.396, W; t = 1.201, FL; t = 1.572, P > 0.05). P c 0.0 1) differed, and when compared to those At Nugget Point SK, FC, and W were signifi- from 1984-1985, only weight (F = 4.58, P < cantly smaller in the 1985-1986 season(df = 5 1, O.Ol), wing (F = 7.63, P c O.OOl), and flipper SK, t = 3.590, FC; t = 2.270, W; t = 4.535, P < length (F = 8.33, P < 0.001) varied. 0.05), but there was no difference between the At Nugget Point and Boulder Beach there was other parameters (df = 5 1, BL, t = 1.727, BD; t a significant difference between fledging weights = 1.177, T; t = 0.793, FL; t = 1.456, P > 0.05). and recovery rates as juveniles over the three Interlocality comparisonduring the 1983-I 984 seasons(Table 4). Fledgling sizeswere compared breeding season. A two-by-two comparison of between the first and third seasonsat both breed- growth in weight at five breeding areas showed ing areas (final fledging size data were not col- no significant differences in growth rates except lected during the 1984-l 985 season).At Boulder between Nugget Point and Shag Point (df = 2, Beach no differences were found between final 572, F = 7.8, P < 0.01) and Nugget Point and sizes of all parameters (df = 30, BL, t = 0.154, Boulder Beach (df = 2,877, F = 10.9, P < 0.01). BD; t = 0.338, SK, t = 0.915, FC; t = 0.639, T; Fledging weightsdiffered between breeding areas

TABLE 4. Pledgingweights, recovery of juveniles, and number of nestsfor two Yellow-eyed Penguinbreeding areas over three seasons.x2 statisticcalculated following multiple comparisonfor proportions(Zar 1984).

1983-1984 1984-1985 1985-1986 P

Nugget Point Pledgingweights ;p) 6.2 4.1

(SP; R = 5.33, n = 3, AC; K = 5.70, n = 6, BB; K = 5.5, n = 15, GI; K = 5.65, n = 16, NP, K = 6 6.2, n = 41; ANOVA: f = 4.88, df = 79, P < 2 Y 5 0.0 1). Growth ratesat all five breeding areaswere also compared simultaneously (Fig. 3). Those ; 4 breeding areas with chicks of unusually high or ._ol3 low fledging weights were suspectedof contrib- I * uting more to overall variation in growth rates, 1 and were removed in a stepwisefashion in 0 I to determine where the variation lay. The com- 0 20 40 60 80 100 120 parison of all five localities resulted in the largest Days after hatching number of parameters showing significant dif- ferences(6/8) (df = 8, 1,058, WT; F = 4.4, P < FIGURE 3. Logistic weight curves fitted to all Yel- O.Ol,BD;F=4.2,P

and bill dimensions in Blue Penguins, although penguin species, with feet and flipper growing bill dimensions only varied when significant dif- fastest in Jackass (Cooper 1977), Galapagos ferences in weight growth occurred, indicating (Boersma 1976), Adelie, Chinstrap, Gentoo food supply as a proximate factor. Availability (Volkman and Trivelpiece 1980), and Blue pen- of food was reflected in more rapid growth of guins (Gales 1987). Galapagos Penguins Spheniscus mendiculus Growth rates of some dimensions of Yellow- (Boersma 1976) and Black-naped Terns Sterna eyed Penguins varied more than others. Re- sumatruna (Hulsman and Smith 1988) at some sourcesin growing birds are allocated at any one locations, and was responsible for considerable time to the growth of the components with the differencesin all aspectsof chick growth among currently highest functional priority, with due four colonies of Thick-billed Murres Uria Iomvia regard for future needs (O’Connor 1977). Hence, (Gaston et al. 1983). Lack of intraspecific differ- differencesin the degreeto which biomass influ- ences in growth rates of pygoscelidswas treated enced growth rates of dimensions, as well as the as being indicative of an abundant food supply speed at which some parts of the body attain (Volkman and Trivelpiece 1980). Variation in their asymptotic value, may be interpreted in growth rates of birds from Shag Point may have light of functional priority, with some features been due to a diet consisting mainly of fish with having a function that contributes proportion- low oil content (van Heezik 1990), although the ately more towards determining the successful sample size was low during the second half of outcome of growth and development, and there- the growth period because all but three chicks fore varying less in response to differing nutri- were killed by predators at about 6 weeks of age. tional conditions. The similarity of growth rates of most parame- The most consistently invariable parameters ters from chicks at other breeding areas suggests of Yellow-eyed Penguin growth were skull length that although characteristicsof these areas may and foot + claw. At Nugget Point, tail growth differ, they do not vary enough to have an effect also remained invariable and at Boulder Beach, on growth rates. bill depth and tail. At Nugget Point final fledging Rates of many developmental processesin sizesof skull, foot + claw, and tail measurements birds are closely linked to each other, but are were smaller in the 1985-1986 season, when largely independent of the nutritional state of the chicks were starving (van Heezik and Davis young (Ricklefs 1968). In Yellow-eyed Penguins, 1990). Therefore, parameters showing invariant growth rates of most morphometric parameters growth rates were those with most variable final varied when all three seasons were compared dimensions, suggestingthat the pattern of de- simultaneously but not as much as weight. Weight velopment of these parameters is more impor- also varied more than development of other pa- tant than their final size. rameters in JackassPenguins Spheniscusdemer- Skull growth is likely to be closely related to sus(Williams and Cooper 1984), Brown Pelicans brain development, and should develop rela- Pelecanus occidentalis (Schreiber 1976), and tively independently of fluctuations in body Common Terns Sterna hirundo (LeCroy and weight. The foot of the developing chick has two LeCroy 1974). Growth rates of biomass of puf- functions; the locomotory function enablesyoung fins (Fratercula cirrhata and F. corniculata) var- chicks to orientate themselves towards their ied in response to feeding conditions (Wehle sourceof food, and when older, engagein begging 1983), and in gannets (&fonts serrator), weight chases.Rapid foot and flipper growth in pygos- plus certain morphometric measurementsvaried celid penguins was attributed to the need to es- much more than others (Wingham 1984). Fastest cape predators and begin begging chases upon growing dimensions were feet, wing, and flipper reaching creche age (Volkman and Trivelpiece lengths, followed by bill depth, flipper area, skull, 1980), although Gales (1987) suggeststhat in less weight, and bill length in that order. The three colonial noncreching temperate species,such as fastestgrowing parameters,as well as skull length Blue Penguins, fast foot and flipper growth re- also showed least variability in time taken to sults from the need to maintain thermal balance. reach 90% of the asymptote, indicating priority In the large-bodied temperate-nesting Yellow- in the allocation of resources.Maximum flipper eyed Penguin, the feet showed the fastestgrowth length was achieved before flipper area. This pat- of all measured parameters, with maximum size tern of growth is consistent with that of other attained at the same time that chicks achieve YELLOW-EYED PENGUIN GROWTH 911 homeothermy (about 28 days), and soon alter ma, P. J. Seddon,J. T. Derby, E. Speirs,K. Nerdhi, chicks are first observed to seek shade (about 2 1 and G. Court for commenting on the manuscript. This study was financed by an Internal Affairs Wildlife days) (Richdale 1957). Large corpulent chicks 8,_holarship covered with thick down have been observed to become heat stressed, seeking out streams or LITERATURE CITED marshy ground to stand in (P. Seddon 1990). The BOERSMA,P. D. 1975. Adaptations for Galapagos ventral flipper is unlikely to function effectively Penguinsfor life in two different environments, p. as a heat window until it acquires feathersrather 101-l 14. In B. Stonehouse[ed.], The biology of than down, since premolt down is very thick and penguins.Macmillan Press, London. ~ERsMA, P. D. 1976. An ecologicaland behavioural has been shown to be a better insulator in still study of the GalapagosPenguin. Living 15: air or a gentle wind than feathersin both Gentoo 43-93. and Chinstrap penguins (Taylor 1986). Feather Coopsa, J. 1977. Energetic requirements for growth development on the ventral surfaceof the flipper of the JackassPenauin. Zool. Afr. 12:20l-2 13. in the Yellow-eyed Penguin beginswhen the chick FROST, P.G.H., W. R. S&ranm, ANDP. J. GREENWOOD. 1975. Arterio-venous heat exchange systems in is about 55 days old. Until then feet may have the JackassPenguin Spheniscusdemersus. J. Zool. an important function as heat dissipators. Both (Lond.) 175231-241. Galapagos (Boerma 1975) and Jackasspenguins GALES,R. P. 1987. Growth strategiesin Blue Pen- (Frost et al. 1975) also gain or lose heat by reg- guins Eudwtula minor minor. Emu 87:2 12-2 19. GA&N, A. J.;-G. CELWDELAINE, AND D. G. NOBLE. ulating the blood flow to the feet and underside 1983. The arowth of Thick-billed Murre chicks of the flippers. at colonies G Hudson Strait: inter- and intra-col- Variations in biomass and final size affected ony variation. Can. J. Zool. 61~2465-2475. survival of fledglings after leaving the natal col- GAVINO, G. T., AND R. W. DICKERMAN. 1972. Nest- ony. Data on diet collectedduring the years span- ling development of Green Herons at San Blas, Nayarit, Mexico. Condor 74~72-79. ning the 1984-1985 and 1985-1986 seasons HULSMAN,K., AND G. Sr+nrn. 1988. Biology and showed diet quality differed significantly be- arowth of the Black-naped Tern Sterna suma- tween thesetwo seasons,the differencebeing most trana: an hypothesisto explain the relative growth pronounced at Nugget Point (van Heezik and ratesof inshore, offshoreand pelagicfeeders. Emu 88~234-242. Davis 1990). The 1985-1986 season of poor LECROY, M., AND S. LECROY. 1974. Growth and growth in biomass resulted in lower growth rates fledging in the Common Tern (Sterna hirundo). of some morphometric parameterswhich reached Bird-Banding 45:326-340. a normal final size but took longer to do so, and MAUNDER, W. J. 1971. Elements of New Zealand’s unchangedgrowth rates of other parameters, that climate. In J. Gentilli [ed.], Climates of and New Zealand. Vol. 13. World survey of cli- ended up slightly smaller in size, i.e., fledglings matology. Elsevier, Amsterdam. with smaller heads and feet, and lower body O'CONNOR. R. J. 1977. Differential growth and body weights. The figure of 21% given by Richdale composition in altricial passer&s. Ibis 119:1471 (1942) as an average expected rate of recovery 166. RICHDALE,L. E. 1942. A comprehensive history of of first-year birds, was achieved only by birds the behaviour of the Yellow-eyed Penguin(Mega- from the first two seasonsat Nugget Point and dyptesantipodes) Hombron and Jacquinot. Un- the first seasononly at Boulder Beach (Table 4). published manuscript, Hocken Library, Dunedin, Almost negligible recovery of birds from the New Zealand. 1985-1986 season indicated chicks fledging at RICHDALE,L. E. 1957. A population study of pen- guins. Oxford Univ. Press, London. less than 5 kg have little chance of survival, al- R&&t-m, C., ANDP. A. Pan&. 1981. Comparisons though it is difficult to assessthe importance of ofgrowtbofalbatrosses.OmisScand. 12:120-124. small size and weight relative to the impact of RICKLE~~,R. E. 1968. Patterns of growth in birds. fledging during a season of poor food supply. Ibis 110:419-451. RICKLEFS.R. E. 1973. Patterns of growth in bids: II ACKNOWLEDGMENTS Growth rate and mode of development. Ibis 115: 177-201. I thank L. S. Davis and J. B. Jillett for their supervision R~CKLEFS,R. E. 1983. Avian postnataldevelopment, during this study. C. Ialas provided assistanceduring p. l-83. In D. S. Famer, J. R. King, and K. C. the first, and E.A.H. Speirs during the third seasonof Parkes [eds.], Avian biology. Vol. 3. Academic fieldwork. S. Neil1 provided accessto Alfred and Cicely Press, New York. Beach, and D. Kemp to the Moeraki breeding area. Ross,G.J.S., D. HAWKINS,R. D. JONES,R. A. KEMProN, B. E. Niven advised on statistical procedures,and J. F.B. LAUCKNER, R. W. PAYNE, AND R. P. Wnrra. Clough drew the figures.Many thanks to P. D. Boers- 1980. MLP: maximum likelihood program. Ex- 912 YOLANDA VAN HEEZIK

perimental Station, Rothamsted, United King- VAN Hnnzx, Y. M. 1990. Seasonal,geographical and dom. age-relatedvariation in the diet of the Yellow-eyed Sctmnmna, R. W. 1976. Growth and development Penguin Megadyptesantipodes. N.Z. J. Zool. 17: of nestling Brown Pelicans. Bird-Banding 47: 19- 205-215. 39. VAN Hammc, Y. M., AND L. S. DA-. 1990. Effects S~~LA~~,J.A.,M.A.HALL,ANDI.M.X~M~N~~. 1983. of food variability on growth rates, fledging sizes The (Spheniscusmagellani- and reproductive successin the Yellow-eyed Pen- cm): sexing adults by discriminant analysis of guin (Megadyptesantipodes). Ibis 132:354-365. morphometric characters.Auk 100:21l-224. VOLKMAN,N. J., ANDW. T~~vap mtn. 1980. Growth SEDDON,P. J. 1988. Patterns of behaviour and nest- in pygoscelidpenguin chicks. J. Zool. (Lond.) 191: site selection in the Yellow-eyed Penguin (Mega- 521-530. dyptes antipodes). Ph.D.diss., Univ. of Otago, WARHAM,J. 1972. The Fiordland Dunedin, New Zealand. Eudyptespachyrhynchus. Ibis 116:l-27. SEDDON,P. J. 1990. Behaviour of the Yellow-eyed Wnrnn, D.H.S. 1983. The food, feeding and devel- Penguin chick. J. Zool. (Lond.) 220:333-343. opment of young Tufted and Homed puffins in SEDDON, P. J., AND L. S. DAVIS. 1989. Nest-site se- Alaska. Condor 851427-442. lection by Yellow-eyed Penguins.Condor 9 1:653- WILLIAMS,A. J., AND J. COOPER. 1984. Aspects of 659. - - - the breeding biology of the Jackass Penguin STONEHOUSE,B. 1976. The generalbiology and ther- Spheniscusdemersus. Proc.V Pan-Afr. Omithol. mal balancesof penguins.Annu. Rev. Ecol. Syst. Congr.:841-853. :131-196. WINGHAM, E. J. 1984. Breeding biology of the Aus- TAYLOR.J.R.E. 1986. Thermal insulationofthe down tralasian Gannet Moms serrator fGrav) at Motu and feathers of pygoscelidpenguin chicks and the Karamarama, Hauraki Gulf, New Zealand II. unique properties of penguin feathers. Auk 103: Breeding successand chick growth. Emu 84:21l- 160-168. 224. TAYLOR,R. H., AND H. S. ROBERTS.1962. Growth ZAR, J. H. 1984. Biostatisticalanalysis. Prentice-Hall, of Adelie Penguin (Pygoscelisadeliae) (Hombron Englewood Cliffs, NJ. and Jacquinot) chicks. N.Z. J. Sci. 5:19 1-197.