INFORMATION TO USERS
This material was produced from a microfilm copy of the original document. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the original submitted.
The following explanation of techniques is provided to help you understand markings or patterns which may appear on this reproduction.
1. The sign or “target" for pages apparently lacking from the document photographed is "Missing Page(s)". If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting thru an image and duplicating adjacent pages to insure you complete continuity.
2. When an image on the film is obliterated with a large round black mark, it is an indication that the photographer suspected that the copy may have moved during exposure and thus cause a blurred image. You will find a good image of the page in the adjacent frame.
3. When a map, drawing or chart, etc., was part of the material being photographed the photographer followed a definite method in "sectioning" the material. It is customary to begin photoing at the upper lefi hand corner of a large sheet and to continue photoing from left to right in equal sections with a small overlap. If necessary, sectioning is continued again — beginning below the first row and continuing on until complete.
4. The majority of users indicate that the textual content is of greatest value, however, a somewhat higher quality reproduction could be made from "photographs" if essential to the understanding of the dissertation. Silver prints o f "photographs" may be ordered at additional charge by writing the Order Department, giving the catalog number, title, author and specific pages you wish reproduced.
5. PLEASE NOTE: Some pages may have indistinct print. Filmed as received.
Xerox University Microfilms 74-17,752
BOERaa, P. Dee, 1946- THE GALAPAGOS PENGUIN: A STUDY OF ADAPTATIONS FOR LIFE IN AN UNPREDICTABLE ENVIRONNENT.
The Ohio State University, Ph.D., 1974 Ecology
University Microfilms, A XEROX C o m p a n y , An n Arbor, Michigan
© Copyright by
P. Dse Boersnia
1974
THIS DISSERTATION HAS BEEN MICROFILMED EXACTLY AS RECEIVED. Tl-iE GALAPAGOS PENGUIN: A STUDY OF ADAPTATIONS
FOR LIFE IN AN UNPREDICTABLE ENVIRONMENT
DISSERTATION
Presented in Partial Fulfillment of the Requirements for
Tlie Degree Doctor of Pliilosophy in the Graduate
School of The Chic State. University
P. Dee Boersma, B.S.
Hie Ohio State University
1974
Reading Coimnittee: Approved B y
Paul A. Colinvaux
Jerry F, Dov.nhower
Rodger D. Mitchell Department of Zoology ACKNOWLEDGEMENTS
I am deeply indebted to Dr. Paul Colinvaux for guidance and
encouragement throughout this study. During my years as a graduate
student, his knowledge, dedication and support were invaluable. I
am also indebted to Drs. Jerry Downhower, Sheldon Lustick, Rodger
Mitchell and Tony Peterle for their ideas, criticism and friendship.
Madhukar Golhar provided statistical advice and the data were
analyzed at The Ohio State University Computer and Information
Science Center.
This research was financed by the National Science Foundation under Grant GB-2906.SX, Principal Investigator, Dr. Paul Colinvaux.
The author was supported by a University Fellowship during portions
of the research.
Personnel and scientists at the Charles Darwin Station, Lina
A, Golden Cachalot and many residents of the Galapagos Islands provided assistance during this research. David Day, Dwayne
Maxwell and Godfrey Merlin made general observations on penguin breeding and surface water temperatures while I was not in the
islands. Nancy Jo, Chris Kjolhede and Sally Cloninger accompanied me as field assistants at different times. Without their deter
mination and effort this study would not have been possible. Tlie duPont family took Sally Cloninger and myself on their yacht to
Elizabeth Bay, Isabela and deserve special thanks. I am also grateful to the many people who either live or have passed through the Galapagos Islands and contributed to this research. VITA
November 1, 1946... Born-Mt. Pleasant, Michigan
1968, 1969...... Spring field work, Research Assistant, Central Michigan University Museimi. Collecting and study of Florida birds.
1969...... B.Sc., Cum Laude, Central Michigan University
1969...... Board of Trustee Scholarship, Central Michigan University
1969...... UTio's UTio in American Colleges and Universities
1969...... National College Registary
1969...... Member, Presidential Task Force on Women's Rights and Responsibjlities
1969-197 3...... University Fellow, T]ie Ohio State University, Columbus, Ohio
1970-197 6 ...... Trustee, Central Michigan University
1970-1974...... Teaching and Research Associate, The Ohio State University, Columbus, Ohio
1970, 1971, 1972... Summer field work. The Galapagos Islands, Ecuador
1971 ...... IVho's IVho in American Politics
197 2 ...... Winter field work. The Galapagos Islands, Ecuador
197 3...... Representative to United States of American Department of State National Foreign Policy Conference
1973...... United States Advisor to United Nation Commission of the Status of Women VITA CONTINUED
CONTRIBUTIONS AND PUBLICATIONS
A Matter of Simple Justice. U.S. Government Printing 1-33 (collective authorship).
In press: Adaptations of Galapagos penguins for life in two different environments. ^ B. Stonehouse, The Biology of the Penguin. Macmillan, London.
Adaptations of the Galapagos penguin to an unpredictable environment. Paper presented October 9, 1973 at the American Ornithologists' Union.
The Ecologist and the Galapagos Penguin. 24 min., color 16 imm film, Department of Photography and Cinema, The Ohio State University, Production Associate.
Breeding Strategies of Seabirds in the Galapagos Islands, A.A.A.S. Galapagos Symposium, February, 1974. TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS...... ü
VITAE ...... iv
LIST OF T A B L E S ...... viii
LIST OF FIGURES...... x
LIST OF P L A T E S ...... xiii
INTRODUCTION ...... 1
PROCEDURE...... 21
NATIJR.U HISTORY OF THE GALAPAGOS PENGUIN...... 30
Distribution and Abundance ...... 30 Beliavior...... 43
The daily routine...... 43 Courtship behavior ...... 44 Head movement...... 48 Pecking...... 53 P r e e n i n g ...... 54 L o c o motion...... 55 Vocal responses...... 61 Nesting...... 65
Sexual and Age Dimorphism in Galapagos Penguins...... 66
Plumage...... 66 S i z e ...... 70
Movement and Mortality ...... 82
Movement of males and females...... 86 Movement of y o u n g ...... 87 TABLE ÜL CONTENTS (CONTINUED)
Breeding Activity ...... 90
Pair bonding...... 90 E g g s ...... 91 Incubation...... 9A C h icks...... 94 Synchrony in egg laying and hatching...... 109 Frequency of breeding...... 110
Molting...... 115 F o o d ...... 125 Predation and D e a t h ...... 129 Breeding Seasons...... 132 Breeding Success ...... 133
ECOLOGY OF GALAPAGOS PENGUINS...... 135
Food as the Limiting Factor on the Population...... 135 Further Test of the Mypotliesis that Food Limits the Population...... 149 Advantages of Sexual and Age Dimorphism...... 164 Maximizing Fitness and Minimizing the Cost of Reproduction ...... 172
M o l t i n g ...... 172 Pair b o n d s ...... 175 Synchrony...... 176 Egg shape...... 177 Incubation...... 179 Asynchrony in hatching ...... 180 A clutch size of t w o ...... 181 Reproductive cost m o d e l ...... 186
CONCLUSION: THE STRATEGIES OF GALAPAGOS PENGUINS AND THEIR RELATIONSHIP TO OTHER ORGANISMS ...... 1G6 PLATES ...... 203
LITER.ATURE C I T E D ...... 211 LIST OF TABLES
Table Page
1 Three expected characteristics of organisms which live in unpredictable and predictable environments. . . ^
2 English and official names of Galapagos Islands. • • • 10
3 Surface water temperatures at Pta. Espinosa, Fernandina, Cabo Douglas, Fernandina and Elizabeth Bay, Isabela between July 1970 and September 1972 . . . 13
4 Tlie number of penguins counted on each of 22 surveys made at Pta. Espinosa,Fernandina ...... 40
5 A comparison of body parts in male, female and juvenile Galapagos penguins ...... 69
6 Banding and recapture data for 432 Galapagos p e n g u i n s ...... 85
7 Recapture data on Galapagos penguins banded July-September 1 9 7 1 ...... 88
8 Recapture data on 423 Galapagos penguins of knoim age or sex...... 89
9 Dimensions of Galapagos penguin eggs and a comparison of the egg dimensions using a modified t test where variance may beunequal ...... 93
10 Body part measurements in cm for Galapagos penguins . . 108
11 Weights in grams of Galapagos penguins during the molt . , ...... 119
12 A comparison of weight in male, female and juvenile Galapagos penguins during three stages of the m o l t ...... 120 LIST OF TABLES (CONTINUED)
Table Page
13 The number of molting Galapagos penguins observed at Pta, Espinosa, Fernandina during months in 1971 and 1972 ...... 124
14 Comparison of growth rates of 95 Galapagos penguin chicks in one and two chick nests over the entire study from June 1971 until October 1972, using an analysis of covariance...... 137
15 Weights of male, female and juvenile Galapagos penguins during each of the life cycle divisions. . . . 141
16 Weight loss per day in Galapagos penguins confined to the nesting site during incubation...... 147
17 Comparisons of growth rates of 93 Galapagos penguin chicks in different breeding seasons...... 156
18 Comparison of growth rates of 93 Galapagos penguin chicks at different locations in the same breeding s e a s o n ...... 159
19 A comparison of body parts in mated and unmated male and female Galapagos penguin ...... 166 LIST OF FIGURES
Figure Page
1. Tlie Galapagos Islands located 600 miles off the coast of Ecuador ...... 4
2. Isabela and Fernandina, breeding grounds of the Galapagos penguin...... 6
3. Surface circulation in October and names of the major current systems (IVyrtlci, 1 9 6 6 ) ...... 17
4. Measurements taken on Galapagos penguins during b a n d i n g ...... 25
5. Hie number and location of Galapagos penguins counted around Isabela and Fernandina from September 4 through 11, 1970 ...... 31
6. The number and location of Galapagos penguins counted around Isabela and Fernandina from August 13 through 19, 1 9 7 1 ...... 33
7. Isabela and Fernandina divided into 8 sections. The number of Galapagos penguins counted during the 1970 and 1971 surveys in each section IS given 35
8. The bill length of 403 Galapagos penguins...... 71
9. Tlie bill width or 403 Galapagos penguins...... 73
10. The flipper length of 282 Galapagos penguins...... 75
11. The toe nail length of 309 Galapagos penguins...... 77
12. Flipper length of 32 juvenile Galapagos penguins . . . ^0
13. Toe nail length of 49 juvenile Galapagos penguins. . . 33 LIST OF FIGURES (CONTINUED)
Figure Page
14. The bill length for a kno\vn age Galapagos penguin chick is plotted against the age of the chick for 149 observations on over 60 chicks...... 96
15. The bill width for known age Galapagos penguin chick is plotted against the age of the chick for 147 observations on over 60 chicks...... 98
16. Tlie flipper length for a known age Galapagos penguin chick is plotted against the age of the chick for 195 observations on over 60 chicks...... 100
17. The oil gland feather length for a k n o m age Galapagos penguin chick is plotted against the age of the chick for 107 observations on over 60 chicks...... 102
18. The foot length of a knomi age Galapagos penguin chick is plotted against the age of the chick for 127 observations on over 60 chicks...... 104
19. Tlie weight of a known age Galapagos penguin chick is plotted against the age of the chick for 215 observations on 93 chicks...... 106
20. Breeding of Galapagos penguins at 7 locations in 1970, 1971, and 1972 ...... 111
21. Galapagos penguin chicks hatching at 4 locations on Fernandina and Isabela...... 113
22. Reconstruction of breeding in Galapagos penguins. . . .116
23. The number of feeding frenzies seen at different surface water temperatures...... 126
24. The mean weights of male, female, and juvenile Galapagos penguins during each of the life cycle divisions...... 143 LIST OF FIGURES CCONTINUCD]
Figure Page
25. Mean surface water temperatures at Pta. Espinosa, Fernandina for the first and second half of each month...... 153
26. A model illustrating the foraging area available to Galapagos penguins nesting at point A and B and swimiing a distance r or less from the nest. . . . 161
27. A model of the energy investment of male and female Galapagos penguins during reproduction...... 187
28. The percentage of survival for knoivn age eggs of Galapagos penguins...... 190
29. Tlie percentage of survival for known age Galapagos penguin chicks through time for two breeding seasons ...... 192 LIST OF PLATES
Plate Page
1. Adult male Galapagos penguin...... 203
2. Individual and age differences in plumage of Galapagos penguins ...... 205
3. Molting juvenile Galapagos penguin ...... 209 INTRODUCTION
Ecology has often been defined as "the study of animals and plants in relation to their habits and habitats" (Elton , 1927).
Implicate in the definition is the idea that the environment effects the organism. Speculation about the effects of the environment on the organism is not new. Darwin comments on the regulating effect of climate on organisms in his book On the Origin of Species while
Matthew in 1915 discusses evolution in relationship to the environ ment. Henderson in 1913 wrote a book devoted to pointing out how the environment was an ideal abode for life. Henderson (1913
cited in Colinvaux 1973) made it respectable to again study the environment. More recently Tinbergen (1965). Levins (1968) and
Slobodkin and Sanders (1969) have examined respectively the importance of the environment on behavior, evolution and diversity.
From an organism's viewpoint, the environment can be divided
into two types : those places which are predictable and those which
are unpredictable. By predictable I mean that there is an established pattern in the fluctuations of the parameters which an organism needs to survive which occurs with low variability. Places which
are highly seasonal as well as tropical regions may be considered
predictable for many organisms. However, if there is some parameter
in a tropical environment which a species needs to exploit for survival such as the flowering of a certain tree the environment may be unpredictable for that species. If the variance around a mean value is relatively high for some vital parameter and its value can not be determined spatially or temporally then the envir onment is unpredictable for that species. "New" environments
(Slobodkin and Sanders, 1969) are considered unpredictable environ ments since the organism can not determine when or where they will occur nor how long the environment will persist. "Severe" environ ments (Slobodkin and Sanders, 1969) are harsh, unusual or fleeting environments which can also be defined as unpredictable. Of course, whether an environment is predictable or unpredictable must be related to the particular organism's life history (MacArthur and
Connell, 1966). kTiat is predictable to a copepod which is small and short lived must be different than what is predictable to a deer which is large and long lived.
Predictable environments are inhabited by species which have the characteristics of a stable population: more restricted niche
(specialized), and a low intrinsic rate of increase (Colinvaux, 1973;
Hutchinson, 1951). Species with these qualities are referred to as equilibrium species. In contrast, species with opportunistic, fugitive or pioneer strategies (Margalef, 1968; Slobodkin and Sanders,
1969) inhabit unpredictable environments and the numbers of individuals are independent of density. Opportunistic species have a wider niche and a high intrinsic rate of increase (Table 1). Table 1. Three expected characteristics of organisms which live in unpredictable and predictable environments.
Type of environment Unpredictable Predictable (New, severe, unstable, (Stable, predictable) ______low diversity)
Type of Species Opportunistic Equilibrium (Fugitive, Pioneer) (Persistent)
Characteristics of Species 1. Density Independent Density Dependent (Population (Population is fluctuates and is constant and does curbed by accident. not fluctuate widely.)
2. Unspecialized Specialized (Wide ecological (More restricted niche, broad niche, narrow tolerance limits, tolerance limits, poor competitor) good competitor)
3. High Intrinsic Rate Low Intrinsic Rate of Increase of Increase (high r value) (low r value) Figure 1. The Galapagos Islands located 600 miles off the coast
of Ecuador. o
C> Figure 2. Isabela and Fernandina, breeding grounds of the
Galapagos penguin. XABO MARSHALL
CABO DOUGL
FERNANDINA
ISABELA
PTA. MORENO
CALETA IGUANA VI LLAMIL Species with equilibrium strategies may be found in both predictable and more rarely unpredictable places but species with opportunistic strategies are only found in transient habitats (Colinvaux, 1973).
The purpose of this study is to test if one species which inhabits an unpredictable environment has the strategies of an " opportunistic species. Furthermore, I wished to examine how the life cycle of an organism is adapted to survival in an unpredictable environment.
The Galapagos Islands (Fig. 1) are an ideal test site because of the characteristics of the land and water habitats. For the last eight years Dr. Paul A. Colinvaux has been involved in research to determine the past climatic history of the Galapagos Islands
(Golinvaux, 1968, 1969, 1972). His analysis of the strategraphy, pollen and fossil water plants present in cores taken from El Junco,
San Cristobal (Table 2) suggest that the weather of the Galapagos has not been stable. Colinvaux (1972) concludes;
That the climate of the Galapagos and the eastern Pacific Ocean has fluctuated broadly in synchrony with glacial and postglacial climatic events in the northern hemisphere. Drought on the Galapagos during a time of glacial advance suggests that the intertropical convergence was then north of the geographic equator at all seasons of the year. Tliat the climate of the Galapagos has never been wetter than now strongly suggests that the stable inversion which is a feature of the climate of the region has been present at least as commonly as at present for more than 48,000 yr. Increasing warmth during the postglacial hypsithermal interval seems to have affected the Galapagos even though the general pattern of atmospheric circulation must have been similar to the present one.
Goodman (1972] analyzing patterns in a core taken by Colinvaux from the Crater Lake on Genovesa Island concluded that pollen shifts in the core also indicated that the climate was drier and perhaps warmer than at present. Such broad climatic changes must have caused differences not only on land but also in the water environment.
It is quite probable that climatic changes on land in moisture and temperature were accompanied by increasing water temperature and perhaps changes in current patterns and productivity. Certainly the long term climatic record indicates that the environment of the Galapagos Islands is unpredictable.
The present day environment of the Galapagos is characterized by short term variability. In appearance the islands look barren and unworldly with black lava dominating the landscape instead of vegetation. Along the coasts of Isabela and Fernandina the black outline is modified more by guano than by vegetation. The low-lying coastal regions of the Galapagos are deserts. Inland temperatures may exceed 50°C during the day and at night go below sea temperature of 21°C. At sea level diurnal temperature changes average 5°C on the windward side and 8 to 10°C on leeward positions (Palmer and
Pyle, 1966).
The variability of rainfall is aptly demonstrated by Palmer and
Pyle's (1966) figures which showed 3.55 cm of rainfall and 141.9 cm Table 2. English and official names of Galapagos Islands. Names in common use by inhabitants are underlined.
ENGLISH NAME OFFICIAL NAME OTHER NAMES
Abingdon Pinta
Albemarle Isabela
Barrington Santa Fe
Bindloe Marchena
Charles Santa Maria
Chatham San Cristobal
Culpepper
Duncan
Espanola
Indefatigable Santa Cruz
San Salvador Santiago
Rabida
Narborough Fernandina
North Seymour Seymour
South Seynjour
Genovesa
Wenman of rainfall at the same station in 1950 and 1953 respectively. On the average, rainfall is less than 75 cm per year (Wiggins and
Porter, 1971]. During the 13 months I lived in the islands, the variability in rainfall was dramatic. During June and August 1971 no precipitation was observed at Pta. Espinosa, Femandina, but between June and August 1972 precipitation occurred over 15 times at the same location. Even the vegetation reflects the difference in rainfall. Between June and September 1970 and 1971 trees were completely bare and Isabela and Fernandina had a decidedly gray and brown appearance. During this same period in 1972, Fernandina and
Isabela, with green grasses and trees in leaf, appeared green.
Two seasons are recognized in the Galapagos Islands: a hot and rainy season and a cool and dry season. Tlie cool and dry season generally is between June and December and is locally knoim as the garua season. The garua is a fine mist which particularly covers the higher elevations but may descend to the ocean. Thoroughly wetting the vegetation, the garua is uncomfortable by human standards.
During some years the garua season may never arrive. On Femandina in 1970 and 1971 from June to August some cloud cover was common in the morning but by afternoon the sky was clear. In June to
September 1972, the pattern waS radically different with cloud cover the rule instead of the exception, and in June, the garua was often noticed. Like the garua season, the hot and rainy season is highly variable. In 1684 rivers were reported on some islands which indicates high rainfall, while in other years there was almost none
(Thornton, 1971). Seasons seems to be highly variable (Slevin, 1959)
and often may be shortened, lengthened or even skipped.
The water e.i/ironment is not unlike the land environment. It
is also unpredictable. Chlorphyll a values used to indicate primary production have little relationship to season or from month
to month (Maxwell, 1973).
Harris (1969a) states that surface water temperatures have
little seasonal variation and that surface water temperatures around
Fernandina and western Isabela rarely exceed 22°C. Between February and March 1972 and between June and September 1972 surface tempera tures were always above 24°C and reached 2S°C around Fernandina and
Western Isabela. During this study a low of 18°C and a high of 28°C were recorded at Pta. Espinosa, Femandina demonstrating that
Harris' (1969a) view that surface water temperatures around
Fernandina and Western Isabela rarely exceed 22°C is incorrect
(Table 3). Although there is no indication that surface water temperatures vary seasonally they are certainly not constant and vary as much as 14°C. Abbott (1966) reports a high temperature of
27.8°C and a low of 1S.4°C at Wreck Bay, San Cristobal between 1958 and 1963. Temperatures taken at the Charles Darwin Research Station,
Santa Cruz, also followed the same pattern of wide variation and ranged from a high of 28.8°C to a low of 20.2°C between June 1971 and
June 1972. Fluctuation in surface water temperature is a character istic of the water environment. Surface water temperatures at Pta. Espinosa, Fernandina, Cabo Douglas, Femandina and Elizabeth Bay, Isabela between July 1970 and September 1972. All tempe? atures are given in degrees Celsius. N=the numb^er of water temperatures taken in each time period, X=the mean surface water temperature and SE=the standard error of the mean in degrees Celsius.
PTA. ESPINOSA
NX ±SE
July 1970 2 21.5 .3 August 1 22
July 1-15, 1971 9 20.6 .1 July 16-30 3 20.7 .2 Sept 1-15 2 20.9 .1 Oct 1-15 1 22.0 Oct 15-30 1 22.5 Dec 1-15 2 21.2 .1
Jan 16-30 1972 2 24.7 .6 Feb 1-15 3 25.8 1.0 Feb 16-28 3 27.1 .2 March 1-15 4 23.0 .02 March 16-30 2 23.0 .03 April 1-15 2 26.6 .2 April 16-30 3 26.6 .3 May 1-15 4 25.0 .3 May 16-30 2 25.2 .03 June 1-15 2 25.8 .3 June 16-30 3 25.2 .3 July 1-15 4 25.2 .4 July 16-30 13 24.2 .1 Aug 1-15 7 24.2 .2 Aug 16-30 7 23.1 .3 Sept 1-15 15 21.2 .1 Sept 16-30 20 22.3 .1
CABO DOUGLAS
July 18, 19 July 19 17
Sept 25, 22.2 Sept 27 23^1 Table 3. Surface water temperatures (continued).
ELIZABETH BAY
Date
July 27, 1971 20.6 July 29 20.4 Sept 6 18.0 Sept 7 18.0
March 1, 1972 26.0 June 25 25.4 July 28 23.5 Sept 30 22.8 Observations of feeding behavior in seabirds also demonstrate
the aberration in the water environment. Seabirds such as the
Blue-footed booby (Sula nebouxii~),Galapagos penguin (Spheniscus mendiculus), Audubon's shearwater ÇPuffinus Iherminieri), Noddy
tern (Anous stolidus). Brown pelican (Pelecanus occidentalis) or a
combination of these species are often seen feeding together in
dense concentration. Birds are constantly diving, and the water
boils with activity. This activity is called a'feeding frenzy'.'
Usually a feeding frenzy will last for 30 minutes, but it is not
uncommon for them to last an hour or more. In a few hours another may occur. I have observed the water when the birds are diving,
and the concentration of fish is enormous. I actually scooped fish
from the water with my hands during one feeding frenzy, and in
another the water was blackened by the fish. One frenzy lasted
over five hours before the birds dispersed. Thus, feeding frenzies
appear to be a sign that food is extremely abundant. No feeding
frenzies were seen in June and July 1972 at Pta. Espinosa, F e m a n
dina; a few were seen in August 1972. In contrast, frenzies
occurred almost daily from July to September 1971. Food abundance
reflected by feeding frenzies appear variable and support the
belief that the water environment is unpredictable.
Currents in the Galapagos are complex and fluctuations are
still not completely understood. Abbott (1966) reviews the current
patterns in the Galapagos, but recent data has led to further refinements (Malone, 1968; Jones, 1969; Love, 1972; Wyrtki, 1966,
1967). Maxwell (1973) presents an comprehensive discussion and should
be consulted for a more detailed account of the current patterns.
Generally, the warm waters of the North Equatorial Current correspond to the hot and rainy season (December-June) while the
Penj Current and the South Equatorial Current are present during the garua season (July-November). Malone (1968) states that the
Cromwell Current which is an undercurrent is embedded in the South
Equatorial Current which is strongest from July to November. The deep Cromwell Current has water temperatures between 11°C and 20°C and up wells around Femandina and Isabela where it is forced to surface when it hits the islands (Jones, 1969). The North
Equatorial Current appears to have temperatures greater than 25°C and the Peru Current and the South Equatorial Current may have cooler surface water temperatures of less than 2S°C but above 22°C.
Surface water temperatures of currents have been surmised by relating the current patterns with the surface water temperature charts in Love (1972). Figure 3 shows the currents (Wyrtki, 1966).
Currents like seasons may be variable in both time and duration.
Although there is extensive literature on many species of penguins (Yeates, 1968; Warham, 1963; Stonehouse, 1953; Richdale,
1951, 1957; Sladen, 1958; Ainley and Schlatter, 1972), very little is known about the Galapagos penguin. Research on the Galapagos penguin in the wild is practically non-existent but the species has been observed in captivity. Townsend (1927) observed behavior and Figure 3. Surface circulation in October and names of the major current systems [Wyrtki, 1966). The Cromwell Current, which is an undercurrent, is embedded in the South Equatorial Current and flows east. These major currents may shift periodicially to the north or south effecting different longitudes. When the Cromwell Current is shifted to the north or south of the islands, it is no longer forced to upwell. ITie cold waters around Isabela and Fernandina are due to the upwelling of this deep undercurrent. OCrOP.ER lUPiACn crnrcKTS \ r / 'VI
—- < — <-— _ <-— - - y ^ — " ^ 0 1 Soclh F<;u< \ i\\ T/ / Vhi ' — > - ^ * ; ; — > 'f recorded the length of the molt in this species while Sumner (1934) studied the ability of tlie Galapagos penguin to capture cryptic colored fish. Even a book has been written on the Galapagos penguin (Bronson, 1931) but it is a fictional account. The literature on Galapagos penguin is restricted to morphological accounts such as Sundevali's (1871) original description of the
Galapagos penguin or to limited observations on distribution, behavior and numbers of penguins sighted (Snodgrass and Heller,
1904; Gifford, 1913; Ridgeway, 1897; Swarth, 1931; Beebe, 1924;
Fisher and Wetmore, 1931). Murphy (1936) reviews the early infor mation on the species and indicates that this species is restricted to the Archipelago de GoIon and is predominately on Western Isabela and Fernandina Island. Eisenmann (1956) did find a Galapagos penguin in Panama but this record is suspect since the bird might have been transported by boat. The first account of breeding in the Galapagos penguin that appeared in the literature (Gouffer,
1957) described a nest containing downy young which was discovered at Pta. Espinosa, Fernandina in late August 1954 and was followed by another nesting record in July 1960 of eggs at Pta. Espinosa,
Fernandina (Bailey, 1962). Gursory accounts of Galapagos penguin
(Eibl-Eibesfeldt, 196 0; Thornton, 1971) have not added any new information but have perpetuated old myths such as the Galapagos penguin is the "smallest of all penguins" instead of the smallest in the genus Spheniscus or that "its presence is testimony to the effects of the cold Humboldt Current sweeping through this tropical archipelago" instead of the Cromwell Current which upwells around the western islands. Tlie actual numbers of Galapagos penguins have been estimated at 5000 individuals by Brosset (1963) and Ziswiler
(1967), 500 by Peterson (1963) and 1500 with 500 breeding pairs by
Leveque (1963). Notes on the breeding season indicate the Galapagos penguin has an extended breeding period (Harris, 1969). Previously,
Lack (1950) suggested they did not breed in the cool dry season while Leveque (1964) concluded they were a cool season breeder.
14 penguins were banded by Leveque in 1962 but knowledge of this
species of penguin has been superficial, contradictory and
speculative. PROCEDURE
This study was conducted over a two year period with 350 days spent in the islands. Four visits were made with the following spacing: June 1970-September 1970, June 1971-September 1971,
January 1972-March 1972 and June 1972-October 1972. During this time, 169 days were spent living at Pta. Espinosa, Femandina, and
38 days at Elizabeth Bay, Isabela. Tlie rest of the time was spent surveying other colonies, traveling between islands, and living at the Charles Darwin Station on Santa Cruz.
Two population surveys were made, one in 1970 and the other in
1971, from boats traveling close to the shore of Isabela and
Femandina. In 1970, the craft was a sailing-motor yacht while in
1971 a local Ecuadorian fishing boat was used. The boats traveled less than a H mile off the coasts and individual penguins were counted to determine distribution, relative abundance, population size, breeding activity and the proportion of molting and juvenile penguins. When it was not possible for the fishing boat to get close to the shore, a Mark II Zodiac rubber boat with an 18 HP motor was used. Using 7x35 and 10x40 binoculars it was possible to see penguins from a quarter of a mile off the coast. At least two individuals were always looking for penguins.
During population surveys and other boat trips, likely breeding
21 areas were visited. Two visits or more were made to each of these points as well as to other unnamed areas. Isabela sites include
Villamil, Caleta Iguana, Caleta Webb, Pta. Moreno, Coastline
Elizabeth Bay, Islands Elizabeth Bay, Bahia Urvina, Caleta Tagus,
Pta. Tortuga, Caleta Black, Bahia Banks, Pta. Vincente Roca, Pta.
Albemarle, Cabo Marshall, Bahia Cartago and Fernandina sites include Cabo Douglas, Cabo Hammond, Pta. Mangle and Pta. Espinosa.
During visits to these places, notes were kept on number of penguins seen, breeding activity (i.e. courtship, eggs, chicks or fledgings), nest sites, and breeding activity of other species.
Penguins were observed and nests were checked throughout the day while working at Elizabeth Bay, Isabela and Pta. Espinosa,
Femandina. During midday no observations were usually kept since it was necessary to retreat from the sun. Early morning watches and nest checks were occasionally kept, but most of the routine observations were taken during the late mornings and afternoon.
Evening watches were made and behavior and interactions recorded.
Throughout the study voluminous notes were kept on behavior as well as selection of mates, body measurements and plumage. Penguins were tentatively sexed, and later the sex was confirmed by observa tion of copulation, dissection, and mate switching. Nest checks were made every day and sometimes more often. Since is was possible to miss a nest change if two changes had occurred since the last check, guano was also examined. Fluid, white guano indicated that the penguin had just recently returned from foraging while yellowish white guano meant that the penguin had returned from foraging about six hours previously. hTien the penguin remained on land for longer periods, the guano became yellower. After three days there was little if any guano, and it was dark green. At breeding areas other than Pta. Espinosa, Fernandina and Elizabeth
Bay, Isabela, all known nest sites were checked and others sought, but repeated checks were limited by the length of my visits.
Records were kept on the breeding activity, nest site location, nesting material and the individual on the nest. Adults at a few chosen nest sites were weighed throughout the breeding season. Eggs were measured for length and width with vernier calipers. The interval between eggs and whether the egg was the first or the second laid was recorded.
432 individuals were banded to observe behavior, movement, mortality, fidelity of mates, frequency of breeding, molting and nest changes. Only penguins which were known or were of specific interest were marked. Nesting pairs, molting birds, chicks of banded adults, juveniles and birds keeping company were usually banded.
161 males, 149 females, 57 juveniles, 51 chicks, and 14 adults which were not sexed were banded at the following locations on Fernandina:
137 at Pta. Espinosa, 34 at Cabo Douglas, 5 at Cabo Hammond, 4 at
Pta. Mangle. On Isabela 248 were marked at Elizabeth Bay, 1 at
Caleta Iguana, and 3 along the coastline at Elizabeth Bay. Tlie bands were a modified design of the Adelie bands used by
Sladen (1958). The band length was changed to 28 mm and it was
marked with 3 numbers and the letter A. The return address stajnped
on the band was: Darwin Station, Galapagos, Ecuador. Bands could
be read in good light from 60 feet with 7x10 binoculars, but
usually it was necessary to get closer. Although many more penguins
could have been banded, the author feels it is necessary to keep
banding to a minimum. Banding should be used as a tool to study
a species, but bands should never be applied indiscriminately, nor
should it be the goal to band every bird. Scientific study must
not be a license to mark all individuals, since marking is a
traumatic experience and will effect behavior and probably other
unknom parameters. Because banding is a traumatic experience, as much information as possible was taken at each banding, and penguins
were not repeatedly handled. General plumage condition, injuries
and facial coloration were recorded. Tlie bill was measured with vernier calipers taking the length of the exposed culmen (L) and
the width or the depth of the bill at the nasal openings (W)
as shown in Figure 4. The right flipper was measured by pressing
a ruler from the elbow joint to the tip. In some cases the width
of the flipper and the thickness of the flipper at the point of
attachment were taken with vernier calipers. The right middle toe nail was measured using vernier calipers from the point of
attachment to the tip. The right foot was measured from the heel Figure 4. Measurements taken on Galapagos penguins during banding. BILL
I
FLIPPER
OIL GLAND FEATHER
I—
FOOT TOE NAIL joint to the longest point. The oil-gland feathers (Figure 4) were measured using vernier calipers from the base to the longest
feather. The longest tail feather was measured using a millimeter
ruler pushed at right angles into the tail feathers and flush
against the bird.
All penguins were weighed when banded and a few banded birds were weighed periodically throughout the study. One of two spring balances graduated either to 5kg or 3kg in 25 gram stages was used
to record weights to the nearest 10 grams. Chicks were measured
in the same way as adults and weighed on spring balances graduated
from 100 grams to 2kg. Scales were graduated in 10 gram and 25 gram
stages. Chicks were weighed at least once daily and measured at
least once every seven days. Adults and chicks were suspended for weighing by placing a leather belt underneath their flippers and
around their bodies or by putting them in burlap or plastic bags.
The weight of the weighing apparatus was always subtracted from the weight read. The balances were checked frequently and adjusted when necessary.
General weather conditions and air and water temperatures were
kept throughout most of the study. Surface water temperatures were
taken approximately h of a mile from shore from fishing boats, in
the Zodiac or off the point at Pta. Espinosa with a portable, battery-powered multi-channel thermistor-therometer or with a pocket therometer. Temperatures were taken at a depth of 28
approximately 30 centimeters. Surface water temperatures taken from
shore were recorded at high tide at a depth of 30 cm at locations
which were exposed to the open ocean and not sheltered by reefs.
Air temperatures were taken with the same equipment. The pocket
thermometer was graduated in degrees fahrenheit while the
therraistor-thermometer was graduated in 1/10 degrees celsius. A maximum-minimum thermometer graduated in degrees C was used to
record variation in temperature inside and above nest sites.
Feeding conditions of seabirds were recorded as number of
foraging penguins and the frequency of feeding frenzies. If
groups of small fish were seen from the Zodiac, this was recorded
as well. A diary of weather, seabird behavior and general obser vations was kept throughout the study.
A visual record of research was made. For still pictures
a 35mm Konica was used, and for motion pictures Bolex 8mm and
16mm cameras were used. A 16mm documentary film produced by
Sally Cloninger illustrates the process of this research. The
still pictures and 8mm film were used to analyze behavior,
document growth changes, and sexual and facial pattern
differences. Sound was recorded on a Uher 4000L tape-recorder
at I h I.P.S.
In order to determine if there were differences between such parameters as sex, age, location, season, and movement, standard
statistical tests such as Chi square, analysis of variance, and t tests were used. Statistical analysis was performed using MANOVA programs and HMD Biomedical Computer Programs at the Instructional
Education Research Computer Center at The Ohio State University.
Only penguins which were paired and both mates were known were used to determine differences between the sexes. NATURAL HISTORY OF THE GALAPAGOS PENGUIN
Lamont Cole (1954) recognized the value in life history studies when he wrote "the total life history pattern of a species has meanings in terras of its ability to survive...". If the natural history of a species is unknown so are the effects of evolution.
Before the effects of an unpredictable environment can be determined on the Galapagos penguin a knowledge of its natural history is essential.
Distribution and Abundance
Surveys from fishing boats were made to determine the distri bution of the penguin population for two consecutive years, as described in the methods section. Figures 5 and G show che results of the surveys and regions which were not surveyed from a fishing boat are indicated. These locations were either checked on foot or were checked by going close to the coastline in a 14 foot rubber boat. The dates of the survey, although not identical (September 4 to 11, 1970 and August 13 to 19, 1971) are within a month of each other. Penguins in both years were breeding during the survey.
Isabela and Fernandina were arbitrarily divided into eight areas by easy geographical points of reference. Figure 7 shows the number of penguins counted in these areas during the 1970 and 1971 survey. The numbers of penguins counted in both surveys and in each
30 Figure 5 . The number and location of Galapagos penguins counted around Isabela and Fernandina from September 4 through 11, 1970. In the enclosed areas are the number of Galapagos penguins counted in that locale. Slashes along the coastline indicate coastline checked by small boat, dots equal areas checked on foot and no marks are areas counted from a fishing boat. A total of 1589 penguins were counted: 843 around Isabela and 746 around Fernandina. V118
FERNANDINA
ISABELA
.363
'168
23 Figure 6. The number and location of Galapagos penguins counted around Isabela and Fernandina from August 13 through 19, 1971. In the enclosed areas are the number of Galapagos penguins counted in that locale. Slashes along the coastline indicate coastline checked by- small boat, dots equal areas checked on foot and no marks are areas counted from a fishing boat. A total of 1888 penguins were counted: 1191 around Isabela and 697 around Femandina. FERNANDINA
ISABELA
275 EI.IZABETy,
298
CALETA Figure?. Isabela and Fernandina divided into 8 sections. The number of Galapagos penguins counted during the 1970 and 1971 [Z of data on Figures 5 and 6] surveys in each section is given. An asterisk indicates that the section was not counted completely. Area 1970 1971 1 381 2 36 66 3 83 32 PTA. a i.b i :m a r l e k $86 2l8* 5 75 15 6 • l44 239 7 $68* 619 8 51* 298
2 ? FERN'AI
ISABELA
BAHIA
ELIZABETH
CALETA iGUAl area are somewhat similar. The surveys probably are a good
index to indicate where penguins are found and their comparative
abundance.
Northern and Cas to in Femandina and the Elizabeth Bay area of
Isabela have the greatest number of penguins. The consistency of relative numbers of penguins in each area from survey to survey would suggest that higher numbers reflect more penguins living in an area and low numbers few inhabitants. The lack of Galapagos penguins in some areas can be explained by the coastal terrain.
Areas 2, 3, the northerly part of area 6, and from Caleta Iguana,
Isabela, southeast to Cabo Rosa, Isabela, are characterized predom inately by steep cliffs. Penguins can not land on steep terrain and consequently, their numbers are low where cliffs and bluffs dominate the coastline. iVhere there are beaches or places of gradual incline around Fernandina and western Isabela, breeding groups of penguins are found.
The eastern side of Isabela and many other islands in the
Galapagos such as Santiago, Pinzon, Santa Cruz, Plazas have been surveyed for penguins. Although penguins have been sighted at places other than Fernandina and northern and western Isabela, their numbers are low. Most of the sighting, by myself, tourists and fisherman, of Galapagos penguins on islands other than Fernandina and Isabela have been lone individuals or small groups of penguins.
I personally have seen penguins on Santiago, Santa Cruz and Pinzon. Personal reports from fishermen, tourist boat personnel and inhabi tants indicate that they had seen penguins at Santa Maria, Eden,
Plazas, and Espanola. Tlie penguin at Santa Cruz and the one at
Pinzon were both juveniles. One of the three penguins I saw at
Santiago was diseased and a skull was collected at Pinzon where the juvenile was previously seen. It is my impression that most of the
Galapagos penguins seen on islands other than Isabela and Fernandina are young penguins probably under three years of age. It may be that vagrant Galapagos penguins [penguins seen on islands other than Fernandina and Isabela) have a high mortality rate.
Breeding groups of Galapagos penguins are exclusively confined to Fernandina and to Isabela on the northern, western and southern sides. Even though a diligent search has been made at sites on eastern Isabela such as Cabo Marshall, Bahia Cartage, and Villamil, nesting penguins have not been found. Other islands have been searched for breeding groups or traces of possible breeding but
Galapagos penguins have never been found breeding. Questioning of fishermen and an inhabitant of Santiago for nine years substantiated the view that the penguins do not breed on islands other than
Isabela and Femandina. It seems possible to conclude that the m o d e m penguin population breeds only on Isabela and Fernandina.
It has not been easy to arrive at a satisfactory estimate of the size of the present penguin population. The head count of penguins does show that more than 1,500 Galapagos penguins exist. but it does not tell how many penguins there are. First, it is necessary to know how many penguins were missed to arrive at a total population estimate. To determine what sighting one penguin means in ternis of actual numbers of penguin in an area, a survey of the Pta. Espinosa, Fernandina area was conducted. From June to
September, 1972, 22 penguin counts were made between 3:00 P.M. and
6:00 P.M. employing the same methods as population surveys. These surveys were more thorough since sightings were made on foot and from the rubber boat instead of from the larger fishing boats.
Banded and unbanded penguins were counted (Table 4). Resident penguins were considered to be banded penguins that were known to have bred in the area or had been seen at Pta. Espinosa three or more times. By dividing the number of banded penguins seen by the number of resident penguins and finding the average we could determine how many resident penguins are usually seen. On any one survey an average of 22% of the banded, resident penguins were seen. This means that for every banded penguin seen, four are not seen. By using this same method in two surveys around the three islands in Elizabeth Bay, Isabela, it was found that for every banded penguin seen, seven were missed. If the chance of seeing individual penguins is similar to the chance of seeing banded resident birds, then the results of the population surveys may be used to assess the approximate size of the Galapagos penguin popula^ Table 4, The number of penguins counted on each of 22 surveys made at Pta. Espinosa, Fernandina. The mean percentage of resident banded penguins sighted for any of the 22 surveys was 21.7%.
Number of Banded Total Number of Number of Penguins Resident Penguins Penguins Counted Counted IvTiich Were at Pta. Espinosa Date of on Each Survey Banded on Survey on Each Survey Survey
9 54 June 25, 1972 79 13 54 June 26, 1972 72 9 54 June 27, 1972 49 4 54 July 2, 1972 63 5 54 July 8, 1972 63 10 56 July 13, 1972 69 7 56 July 17, 1972 106 9 57 July 20, 1972 108 11 57 July 21, 1972 95 16 57 July 22, 1972 81 9 57 July 24, 1972 67 14 57 July 31, 1972 59 10 57 Aug. 1, 1972 110 20 57 Aug. 2, 1972 96 9 58 Aug. 4, 1972 78 19 58 Aug. 6, 1972 38 11 59 Aug. 9, 1972 44 22 67 Sept. 4, 1972 33 14 67 Sept. 10, 1972 74 31 67 Sept. 13, 1972 69 18 67 Sept. 22, 1972 39 15 67 Sept. 23, 1972 Since the two population surveys were conducted most of the
time from fishing boats, and counts were continuous instead of
from 3:00 P.M. to 6:00 P.M. when penguins are most likely to be on
shore, consideration should be given to the reduced ability to see penguins during the two surveys. Perhaps a reasonable estimate is
for every penguin seen, nine are missed. Tlius, every penguin
sighted equals ten penguins of the total population.
As the population survey maps show [Figs. 5 and 6), Isabela
and Fernand:'na were not completely surveyed on either trip.
However, all areas of Fernandina and all breeding areas of Isabela
were surveyed at least once during the two surveys. Since popula
tion numbers did not change drastically in any of the areas
surveyed from year to year, numbers that are missing from one
survey may be estimated by using known numbers from the other
survey. In 1970, 2,310 penguins should have been counted and in
1971, 2,281 penguins should have been counted. If each penguin
seen actually means nine were missed, then around 21,300 penguins
should have been present in 1970 and 22,810 in 1971. For a lower
population value perhaps we should assume that for every penguin
counted four were missed, which was the result of the surveys at
Pta. Espinosa. By using these two procedures to find the range
of the Galapagos penguin population, an estimate of between 11,000
and 23,000 penguins can be made.
As previously pointed out, the limited literature on Galapagos 42y/V3 penguins is filled with sighting of penguins and estimates of the population. Since sighting of penguins is directly related to contact with humans, and estimates of the population are subjective guesses, it is not surprising that a range in numbers and locations of Galapagos penguins exist.
The Daily Routine. Galapagos penguins sleep on land, forage during the day coming ashore periodically for short periods, and return to the land between 4:00 P.M. and dusk for the night. They return to the sea the next morning between 5:00 A.M. and 7:00 A.M.
This general pattern is modified due to weather and food.
Unemployed penguins, during the breeding period, or all adults when not breeding, first arrive below their nest site or on land where they spend the night beginning at 4:00 P.M. The greatest influx of penguins occurs at sunset when they arrive in groups.
These groups are composed largely of non-resident penguins which seem to be just passing through, since they are not regularly observed. Non-breeding, resident birds change their daily routine prior to the onset of breeding by spending more time on shore than paired penguins or non-resident birds. This is presumably because they are seeking mates.
Breeding or mated Galapagos penguins may modify their daily routine. Before eggs are laid, both birds are frequently found in the nesting site. They may stay continuously in the nest site, come ashore early, or stay on shore later in the morning to remain in
the nest site. Throughout incubation one of the pair will be
present continuously. This changes the daily rhytlim of breeding
individuals since many penguins are on shore during the whole day
or the greater part of it. Once the chicks get large enough neither
parent will remain at the nest site. The daily routine for parents
feeding large chicks is very similar to non-breeding penguins.
Penguins come ashore to feed chicks shortly before or at sundown
depending on their foraging success. The daily routine of Galapagos
penguins is similar to other species of penguins as reported by
Richdale (1951) and van Zinderen Bakker (1971).
Courtship Behavior. Courtship behavior only occurs between
adult birds which are developing or have developed pair bonds. All
terms for behavior patterns were taken from Richdale (1951) and
will be used in the same context unless otherwise stated. Mutual
preening is when two birds are preening each other. Richdale (1951)
differentiates the mutual preen and the kiss preen. From my obser-^
vations these behaviors grade into one another and often in the
Galapagos penguin what starts out as mutual preening ends as kiss preening or vice versa. For this study, I will refer to both as mutual preening. It commonly occurs between mated pairs and has been exclusively seen at the beginning of the breeding season before
eggs are laid or in the nesting site during breeding. In order to mutual preen penguins must be standing relatively close together. Individuals often start preening their mate after they have been preening themselves. After a brief pause, one penguin may start to preen the other’s head or neck and the other penguin either reciprocates or continues to preen itself. I have only observed mutual preening between mated pairs or between individuals disposed toward each other. This activity seems to have pair bonding value and is not necessarily followed by more intense love habits.
Richdale (1951) indicates it is a common minor behavior in all species. Warham (1963) observed it in the Rockhopper penguins
(Eudyptes chrysocome). This behavior in the Galapagos penguin appears largely confined, or at least more frequent, during breeding.
During June to August 1972, when penguins were not breeding, it was never observed. Mutual preening also occurs between parents and chicks where parents will preen chicks. On one occasion an adult male preened a juvenile molting to adult plumage. Since breeding had just commenced when this observation was recorded it might only be a rare case of any over-zealous male.
Flipper patting and copulation:
Flipper patting (called "the arms act" by Richdale, 1951) is when a penguin vibrates its flippers rapidly against the body of another penguin. IVhile the penguin is vibrating its flippers, it moves toward the back of the other penguin and then leans over the penguin forcing it downward. As it leans against the other penguin its bill vibrates side to side and its flippers pat up and down rapidly against the other penguin. A Galapagos penguin which is flipper patted is partly forced downward from the weight of the other penguin. The penguin which is now prone pointed its bill upward and moves it side to side so it appears to vibrate. The penguin on top continues to flipper pat and vibrates its bill so that the base of its bill is touching the end of the other penguin's bill.
At this point the top individual treads on the other's back and moves its tail to the side while the prone penguin lifts its tail upward. Both penguins continue to vibrate bills. Copulation may take place, but often the cloacas do not meet, and the top penguin slips off and walks to the side of the prone bird with its head bowed downward. Both penguins often shake their tails and swallow afterward. In two observations where copulation was successful, the top penguin (a male) stayed in position from 30 seconds to over a minute. As this description indicates, flipper patting frequently leads to copulation. This is not peculiar to the Galapagos penguin
(Sladen, 1958). However, flipper patting does not always lead to
copulation, and often flipper patting is terminated before the patted penguin becomes prone. Unmated penguins which had not been obsen^ed previously standing together were never seen flipper patting.
Flipper patting seems to have pair bonding value and may serve to position the female for successful copulation. Only males, identi fied by plumage, flipper patted, and in one case semen was actually seen being released. Flipper patting is common in other species of penguins and seems to have the same function (Richdale, 1951).
Copulation was seen both on land and in the water. Although copulation in the water can not be observed as readily as on land, males seem to flipper pat and have the same posture as on land. On land there are more attempted than successful copulation. However, no data is available for copulations in the water. The frequency of copulations in each environment is unknom, but I saw more on land. Considering copulations were probably more likely to be seen on land, the frequency in the mediums may be relatively equal. In one copulation seen on land semen was observed running do\m the female and onto her cloaca. In the water the semen would have probably been lost. This suggests copulations may be more successful on land.
Bill dueling is when two penguins stand breast to breast, shake heads and bills from side to side in short movements, causing the bill tips to hit against each other. Bill dueling last no more than
IS seconds and never has been seen continuously for more than five seconds. Bill dueling has only been seen when two penguins are very close together. The penguins rock backward standing chest to chest and rapidly hit bills. It has been seen in both mated and unmated
individuals.
When bill dueling is seen in unmated pairs the events leading up to the behavior are somewhat the same: a penguin is standing on land when another penguin lands very near, or a penguin will be standing and another penguin lands and walks up to the first penguin, or a penguin will be standing when another moves closer. Bill dueling does not always follow these events; but, if neither penguin moves, which may indicate they are equally motivated to attack and flee, bill dueling seems to always follow. After bill dueling, the penguins move away from each other.
Mated penguins like unmated penguins may bill duel following the same events. Bill dueling most commonly follows when one of the pair moves closer to the other. After bill dueling a mated pair does not move apart. Instead, they may stand close toeach other and appear relaxed. Sometimes the birds' eyes close as they stand touching or nearly touching each other. At other times flipper patting may follow bill dueling. Bill dueling between mated pairs may serve to relieve aggression in a harmless way. While in unmated penguins, it may lead to pecking, if the birds do not move apart.
An added function of bill dueling may be keeping the pair bond intact since mated pairs are commonly seen uniting against another penguin which has gotten too close. Bill dueling has been seen on land and in the water and has been reported for other species
(Downes e^ , 1959; van Zinderen Bakker, 1971).
Head Movement. Head movements have been observedwhen two or more penguins come together. These movements have not been pre viously reported in the literature. Head movement is the term used to describe the behavior of moving the head and bill in half circles.
The bill of the penguin is pointed slightly downward and is sharply moved upward to the right until it is pointed at about a 140° angle
from the ground. The bill then is jerked downward and upward to
the other side until the bill is again at 140° angle from the ground but displaced to the left approximately 90°. The white chin of the penguin doing the head movement is exposed on the upward movement to the observing penguin or penguins. Although this is not the meaning of the head movement, it resembles the human response of moving the head to point in a direction and saying, "hey, let's go this way". Instead of continuing to point in one direction with the head as a human with his arms full might, the penguin moves the head back and up and to the other side as if to see if everyone is following and then back again as if again pointing the direction.
There may be a series of these movements. "Head movements" are reported in other species of penguins (Sladen, 1958) but no other species seems to have this elaborate behavior, nor has it been reported to function as it does in the Galapagos penguin.
All Galapagos penguins except chicks make head movements. This behavior has been observed both on land and in the water. It always follows the same pattern of one penguin moving in close proximity to another. It occurs most frequently when individuals are within pecking range. IVhen a penguin has been intruded upon it flattens the feathers on the top of the head and raises the neck feathers. Next the penguin points its bill toward the intruder with
the chin low enough that the underpart is not visible. The neck may or may not be outstretched. As the established penguin either becomes more intruded upon or gains courage, it stretches out its neck so that it is closer to the intruder. Its bill is always pointed at the head area. Usually the established penguin stretches
its neck and tips its head from side to side as the intruder starts making head movements. The established penguin tips its head as if
it is following the head movement. Once the established penguin
assumes the posture of pointing the bill and flattening the cro\m
feathers, the intruder starts making head movements. As the intruder makes head movements, sometimes the established bird will venture
closer. The intruder generally continues head movements and moves
away. If the intruder moves away, no pecking occurs. If the
intruder instead of moving away remains, pecking by the established penguin usually follows. Sometimes the intruder makes no head movements, but instead postures and points its bill toward the
established penguin. If the intruder threatens back, yelling by
both may occur, and this act seems to be a high intensity threat, van Zinderen Bakker (1971) describes what appears to be a similar
behavior in the Gentoo penguin (Pygoscelis papua) as a high
intensity threat. If the birds part after a serious threat with or without bill dueling and yelling, males may donkey bray. Yelling
and donkey braying will be described under vocal responses. Observations have shouTi that penguins which do not give the head movement and are threatened are pecked. Therefore, head movements appear to be an appeasement response to a threat display. I have never seen a penguin pecked while doing head movements even though frequently the threatening penguin comes closer and may actually touch the penguin making the head movement.
Head movement is almost always seen when penguins swim up to each other. When penguins arrive or land and walk up to other penguins, they usually start head movement. It may be a greeting response which allows penguins to get close enough to recognize individuals since they are very myopic (Murphy, 1936). Although penguins as well as other seabirds have been shown to recognize individuals vocally (Thompson and Emlen, 1968; Beer, 1969), auditory cues are not always available. Head movement appears to allow penguins to come in close proximity without harm to either individual.
This is advantageous to individuals which come ashore in small landing areas which are already occupied. In the water it allows penguins to come together which may be important to individual survival in a species which exploits a densely packed resource such as a school of fish. Although this behavior has been described between two individuals, it can occur between three or four indivi-
Mated pairs and unmated individuals react differently after head movement. Mated pairs are seen in head movement after coming close together. After the head movement, they may move off with the penguin making head movements preceding the other. Also, they may bill duel, flipper pat or preen. Unmated individuals may bill duel, yell, and peck. The penguin which makes the head movements in unmated pairs also initiates the departure, but is not necessarily followed by the others. Unlike mated penguins, if one of the penguins does not move away, pecking will follow.
The question remains why should Galapagos penguins be aggres sive toward each other and yet have evolved an appeasement response to reduce the chance of physical harm? Without aggression, pair bonds would be threatened. Any individual could court another's mate or impinge upon the nest site. As Richdale (1957) and Sladen
et al. (1968) have shown, reproductive success increases with age and fidelity from year to year. Without aggression, pair bonds might weaken and individual reproductive success would probably be lowered. Some aggression or tolerances toward other penguins is beneficial, however, pair bond formation, recognition, social gregariousness and foraging make it necessary for individuals to come together. Head movement seem to allow close proximity without physical damage to individuals. It is suspected that other species of penguins have a comparable behavior pattern, although it probably does not take the form of head movement since head movement is not mentioned in the literature. Pecking. Pecking can only take place when individuals come
close together and, naturally, it can be physically damaging. Before a Galapagos penguin is pecked, it is always warned by a threatening gesture. IVhen a threatened penguin does not move away or modify its behavior, it is usually pecked. Since Galapagos penguins usually do react to being threatened, pecking is rarely observed.
Two typical examples of pecking follow. An adult threatened another adult which reacted with a brief sequence of head movement.
The threatened penguin then turned around while the penguin that had threatened it then pecked. It seemed that the intruder did not move away quickly enough. In another instance, a juvenile landed on the coast and was threatened. It did some head movement which was not as stereotyped as the adults. It stretched its neck and acted like it was looking around. It was promptly pecked. In general, juveniles do not seem to do head movement as often when threatened as adults and do not move off as rapidly as adults. It may be that head movement improves with age, or that the knowledge of what is threatening is learned. Regardless, juveniles seem to be pecked more often than adults.
Pecking between siblings is common. As is generally believed, the younger chick is pecked by the older. One chick observed was missing feathers at the back of the head because of pecking by an older sibling. Miy chicks are pecked by siblings other than to force them to move was not determined. Galapagos penguins have many body movement behavior patterns
that are similar to other penguins (Boswall, 1973; Warham, 1958). van Zinderen Bakker [1971) described Gentoo penguin behavior using
McKinney (1965) classifications as a guide, van Zinderen Bakker*s
(1971) descriptions of body movements in Gentoo penguins aptly fit
Galapagos penguins, but minor differences in postures and the fre
quencies of behavior seem to exist. Consequently, it would be
redundant to describe similar behavior patterns in Galapagos pen
guins, so only differences between the species will be mentioned.
Following van Zinderen Bakker's (1971) organization, the tail wag
always follows excretion and accompanies preening, landing and even
floating on the water. It appears to be more common in the
Galapagos penguin than in the Gentoo penguin. Both-wing-stretch
in the Galapagos penguin is represented by a slightly different
posture. The neck is stretched parallel to the ground like in the
Gentoo penguins, but the head and bill -are pointed upward, and
sometimes a jaw stretch will also take place. The jaw stretch in
Galapagos penguins is not restricted to the standing posture and
may occur when the penguin is prone. Again, as in the both-wing-
stretch, the bill and head are thrust upward.
Preening. Cleaning movements appear to be identical in the
Galapagos and Gentoo penguins. Preening occurs in the water as well
as on land, and probably more than three hours are spent preening
each day. Preening keeps feather acting as an effective insulation 55 layer. It not only keeps feathers compact and eliminates any spaces exposing the skin, but it also must hamper algae growth. Therefore, the significance of preening may be to keep the plumage in good condition. Social preening described previously as mutual preening does occur in the Galapagos penguin, but it was not ob served in the Gentoo penguin (van Zinderen Bakker, 1971). Washing occurs not only after entering the sea as in the Gentoo penguin, but also frequently before landing especially in the evening and while at sea. Galapagos penguins sleep in two prone positions like the Gentoo with flippers either under the body or alone the side, but, unlike the Gentoo, the Galapagos penguin was never seen with its bill tucked under a flipper.
Locomotion. The Galapagos penguins like other species of penguins have five modes of locomotion: walking, tobogganing, hopping and jumping, porpoising, and swimming. Galapagos penguins do not walk with their flippers as extended, nor do they sway as much as other penguins. This is probably due to their size result ing in greater balance. IVhen they walk rapidly their flippers are held further out from the body, undoubtedly to maintain balance.
Snodgrass and Heller (1904) give a good account of walking and jumping in the Galapagos penguin. Unlike other species of penguin the Galapagos penguin has never been seen approaching another penguin in a manner to indicate an ecstatic display would follow.
This is probably because nest sites are not exposed, and Galapagos penguins generally do not have to maneuver between other penguins to reach their nest sites.
Tobogganing, although a common practice in some species like the Adelie penguin (Pygoscelid adelie) (Sladen, 1958), was only seen a few times while observing the Galapagos penguin. Galapagos penguins attempted to toboggan to get away when they were frightened on a sand beach or on relatively smooth lava. Of course, the rarity of this behavior in Galapagos penguins undoubtedly results from the terrain.
Hopping and jumping is commonly seen in the Galapagos penguin.
They jump or hop across places where they can not walk such as fissures, crevices and from rock to rock to climb up or do;m.
Unless the penguin was frightened, jumping or hopping was preceded by stretching of the neck forward and a surveying of the land with the bill near or touching the ground where the bird would move.
Other accounts can be found in Beebe (1924).
The Galapagos penguin is only infrequently seen porpoising.
Groups of approximately 25 to 50 penguins have been seen porpoising twice over distances of approximately % of a mile. Both times the penguins were chasing a school of small fish from behind. Individual penguins have been observed porpoising a few times when feeding in a large group and darting after fish. By porpoising, the penguin comes to the surface for a split second and then continues its forward motion with little interruption. Porpoising seems to be an adaptation to maintain maximum speed and still be able to breathe.
The reason porpoising is not seen more frequently is probably because many times the penguins feed on schools of fish that are relatively stationary. By diving in on the school from all sides, penguins can easily herd the fish into a clump making maneuver ability, not porpoising, essential for catching prey. Porpoising would only be advantageous to maximize forward speed. When penguins were surprised by the rubber boat and motor coming close to them while at the surface, many dove underwater and then porpoised. This is logical if porpoising is used to maximize forward speed. In places where predation by leopard seals (Hydrurga leptonyx) or water predators are high, porpoising for other species of penguins may be an adaptation to avoid predation as well as to secure prey.
Swimming on the surface is the most common behavior between dives. Three postures are seen. The head and bill are under water, but the top of the head and back are slightly exposed in one posture which is used when swimming along the shore. The posture resembles a human paddling along the surface with fins and mask. The penguin may raise its head to look around and call, but over half of the time, the bird is looking underwater. Then with a powerful downstroke the penguin totally submerges. Another posture is the floating posture where most of the body is underwater and only the head, a little of the back , and the tail appear above the surface. Tlie penguin employes this method while resting on the surface. During surface swimming this posture is commonly seen between dives, before
landing, when penguins are in large groups foraging, when surveying some activity like a boat, and when looking and calling. Another posture is described in Boersma (1974).
Van Zinderen Bakker (1971) states that "only rarely do Gentoo penguins swim on the surface of the water". However, as previously pointed out, the Galapagos penguin generally swims on the surface between dives. If van Zinderen Bakker's (1971) observations and conclusions are correct, there must be some striking differences in swimming behavior of Gentoo and Galapagos penguins. Although it seems possible and likely that the species have differences in the frequency of porpoising resulting from predator and prey differences, it seems incredible that the amount of time the species spend on the surface of the water would be so different. If van
Zinderen Bakker (1971) is correct, and the Gentoo rarely swims on the surface, then the Gentoo must swim exclusively underwater and porpoise. Van Zinderen Bakker (1971) points out that the Gentoo penguin is normally not present for 12 hours, and he assumes they are at sea. This means, according to van Zinderen Bakker (1971), that the Gentoo must be swimming and porpoising for most of the
12 hours since it "rarely" swims on the surface. This is absurd since it would be impossible for the Gentoo penguin to have enough energy to maintain this kind of activity. No birds or mammals are known to fly regularly for 12 hours in the air or water without stopping except on migration, and then individuals show weight losses. Gentoo penguins must, like the Emperor penguins
ÇAptenodytes forsteri) (Kooyman e^ , 1971) and the Galapagos penguins, rest on the surface of the water between dives. Surface swimming and resting in all species of diving organisms must be the most common activity while the organism is in the water, since diving requires a considerable amount of energy and individuals must recover between dives.
All penguins swim and dive. Dives are of short duration probably for all species of penguins. Kooyman et al. (1971) found
Emperor penguins dove for less than one minute in over half of the dives monitored. Galapagos penguins likewise usually remain sub merged for less than a minute. One dive which was timed lasted 1 minute 19 seconds, but this is a long dive, and most are less than
30 seconds. Scholander (1940) reports forced submersions of Gentoo and Macaroni penguins (Eudyptes chrysolophus) of seven and five minutes and gives the impression this is near the maximum possible diving duration for the birds. The Galapagos penguins probably have a similar tolerance for diving which is much shorter than the
18 minute dive reported for an Emperor penguin (Kooyman ejt al_., 1971).
The shorter diving time of the Galapagos penguin is probably linked to its smaller size.
Swimming posture and movement have been described for penguins by Kooyman e£ a]^. (1971) and briefly by van Zinderen Bakker (1971) .
Cine films were made of the Galapagos penguin swimming, and little can be added to Kooyman et al's (1971) description of the Emperor
penguin swimming underwater.
On the surface of the water, the swimming movements of the
Galapagos penguin is modified. The feet still trail with the soles facing upward acting as rudders. The flippers, instead of almost touching on the upstroke as they do underwater, remain submerged, and the back is partly exposed in the posture previously described.
Tlae flipper cuts the water in a movement resembling the wing move ment in flying birds. The flippers on the upstroke are tilted forward and brought forward slightly. On the do\mstroke they are rotated so that the leading edge of the flipper is tipped upward to reduce resistance. On the domstroke the flippers are rotated so that the leading edge is angled downward to give maximum thrust forward. The flippers are brought down and then forward. The movement is similar to the butterfly stroke of swimmers or the stroke of flying birds. The difference between surface swimming and underwater swimming seems to be only in the degree of vertical movement in the flippers, but not in the pattern of movement. In underwater swimming the flippers move through approximately 130°, while on the surface swimming, they may move only 5 or 10° depending on the desired speed. Penguins are not really flightless, although penguins do not fly in the air, they do fly through water at speeds in excess of 10 m.p.h. Vocal Responses. The Galapagos penguin makes a number of vocal responses. Mien penguins are at sea or are standing on land, they commonly make a call sounding like a drawn out haw. Snodgrass and
Heller (1904) mention that the Galapagos penguin utters a sort of grunt and this is probably describing the haw call. The haw call,
I think, sounds more like a sigh than a grunt, or as Toimsend (1927) describes it, a soft throaty whoo. After a penguin has given the call, it is frequently answered with a similar call, and soon the calling individuals come together. IVhen large groups of penguins are foraging together, the haw call is heard every IS seconds or more. The space between penguins is reduced; and when the penguins are foraging in groups, after the call, they become more compact and, at time, resemble a floating raft. IVhen the haw call is given on land, the individual either leaves and joins another bird in the water, or another penguin arrives and joins the penguin already on the land. I have frequently seen a known individual use the haw call and then have the mate suddenly appear. My assistants and myself have made the call causing a lone penguin to show interest by answering and swimming toward us. In many instances we continued the call, and the penguin would land and approach us. One penguin actually came and stood at our feet. The haw call is used to locate other penguins, and it appears that mated pairs can recognize each other's call.
Another distinctive call is a bray which sounds much like a donkey braying. Snodgrass and Heller (1904) nuted an elongated braying call resembling Ma-a-a-ah, The bray that they describe
occurs at the end of the call, The bray is preceded by two to four
introductory notes made by collapsing the chest. It resembles the
sound humans make when sucking in air. The donkey bray is heard
predominately during the breeding season. IVhen the Galapagos pen
guin is not breeding, the bray is not normally heard. The frequency
of donkey brays is highest immediately prior to egg laying. After
eggs are laid, the frequency of penguin brays is reduced. Only males, both mated and unmated, have been seen donkey braying.
Penguins bray both on land and in the water. I have watched males
quietly staying with females below their nest site when a bray is
heard. The mated male frequently starts braying in response. I
have even induced a male to donkey bray by braying. However, I have not found it possible to induce braying outside of the breeding period. The donkey bray appears to be a call which advertises the male's availability.
Donkey braying was observed in a slightly different context when two penguins were captured for temperature regulation. One bird was molting and, after testing, was released in a small tidal pool where he voluntarily stayed. The next day another male was
captured and introduced to the same pool. After much threat behavior
(i.e. posturing) and yelling, the molting male donkey brayed. The
call seemed to indicate intolerance towards the other male which was
in an area he seemed to be defending. Donkey braying may be used to indicate when an area is occupied by a male as well as to advertise
availability.
The posture during donkey braying appears to be the same as in
the Blackfooted penguin ÇSpheniscus demersus) (Roberts, 1940). The
donkey bray is the ecstatic attitude in the Galapagos penguin.
Yelling occurs after bill dueling or when a penguin is ap proached closely by another penguin. The one male in the tidal pool
frequently bent his body toward the introduced male and yelled.
The penguin which yells does not move away. It appears that yelling
is a call made by adults when they would like another penguin to move away, and do not intend to move themselves. As previously mentioned, yelling is a high intensity threat and seems to be made
in frustration and hostility.
The courtship bray is much like the donkey bray and similar to the yell in vocal quality. The posture is different from the donkey bray. In the donkey bray, the male points his head toward the sky and raises his flippers as his chest pulsates and two to four notes are made. Then the flippers are raised up and the back and neck come forward. The bray is emitted as the flippers are brought down.
In the courtship bray, the male and female point their bills toward the ground. The head is bent upward while the neck is bent down ward. The posture is similar to the posture of other penguins during the next relief ceremony (Richdale, 1951; Roberts, 1940). The courtship bray is made at next reliefs as well.
Throbs are also made by both males and females. They have only been noted between mated individuals and are most commonly heard
from the nesting sites. Throbs are heard after courtship brays or when one penguin of a mated pair jostles the other by moving. The quality of the sound would lead to the belief that the throbs
indicate contentment; but I think when the penguin has been dis turbed, it throbs because it feels some aggression. The throb may be a way to reassure each bird that all is well. From throbs the birds may go into a courtship bray. Both the courtship bray and the throb may indicate both fear and attraction and may be a type of displacement activity when the penguins would normally make a threat or appeasement response.
Juvenile penguins do not. donkey bray, courtship bray or throb.
This indicates that these calls are breeding calls. The haw call and yells that juveniles make do not always have the same quality as the adult calls. Since no adults have been heard with the same distinctive quality, it must be that the calls are perfected with
learning and time.
Chicks make a peeping call and with a little practice individual chicks can be distinguished by the scientist. At night or in the day, when it was impossible for the chicks to see the adult, the
Galapagos penguin chicks have been seen becoming suddenly active in the nest site or running to the opening of the nest site after its 65 parent or parents donkey bray or courtship bray. Previously, when
another adult makes a vocal response, the chick did not move or become alert. It seems that chicks can recognize their parents vocally.
Thompson and Emlen (1968) found that adult penguins can
recognize their own chicks at 17 days old. When a chick which
fledged early was introduced into a nest where the adult Galapagos penguin was incubating an addled egg, the adult did not peck the
chick which was begging and following the adult. The adult did not, however, care for or feed the chick. It appears that not only
can Galapagos chicks recognize adults, but also adults can
identify their own chicks.
Nesting. Many species of penguins build elaborate nests
and have much behavior associated with nest building. Murphy
(1936) and Penney (1968) note that returning Adelies spend a great
amount of time gathering and stealing stones. Other penguins build no nests, such as the Emperor penguin which houses its egg on its
feet. Members of the genus Spheniscus nest in cavities (Rand,
1950). Many dig their nest site while others use natural features
such as caverns.
The Galapagos penguin does not build an elaborate nest,
although it does dig nesting burrows where it is possible, such as
at Elizabeth Bay, in the tuft. Two cup shaped nests made of sargassum have been found, but this is a rarity. Galapagos flightless cormorants, however, regularly make their nests of sargassum. Objects used in the nest site often appear to be more symbolic than functionary. Mangrove leaves, feathers, bones, sticks and twigs are the most common objects found in nest sites. They all are materials which are readily available. The sticks and twigs are mangrove or Bursera while the leaves are either red (Rhizophora mangle] or white (Lagunculavia racemosa] mangrove, depending on what is around. Algae, sea fans, shells and land plants like
Desmodium and Cacabus miersii are more rarely encountered. Objects are placed in the nest in no apparent order. Usually, the nest site is dug out enough by the penguin or has sufficient old guano on the lava that the surface is fairly smooth and flat. In only a few nesting sites where the lava was not smooth and flat was there enough nesting material present to help prevent the eggs from cracking on the rough lava. One nest site where the male switched mates had an abundance of red mangrove leaves both seasons although the previous nest site of the female had only a stick and a leaf.
This may mean males are responsible for the nesting material.
Sexual and Age Dimorphism in Galapagos Penguin
Plumage. There are two plumage patterns in the Galapagos penguins. Both male and female adults have similar plumage patterns of black and white bellies. Tlieir breast is flecked with black feathers, and the side of the head has a semicircle ribbon of white
which extends from the back of the eye around under the chin and
to the back of the other eye (Plate 1). More detailed accounts are
found in Murphy (1936).
There are general characteristics of the plumage which differ
between the sexes. Tliroughout the study only known males or birds
which appeared to be males were observed donkey braying, flipper
patting, or in the uppermost position in copulation. These same
penguins had bolder markings and more pink around the base of the
manibles and eyes. Both sexes have white chin feathers which in
some individuals may be mottled. Although not previously recognized, males have wider and more conspicuous white chins than females, and,
generally, they are less mottled. The pectoral bands of males, as
well as the facial markings, are usually more distinct than the
female bands.
Individual penguins can be recognized, and females and males
can generally be distinguished by facial characteristics. At the
base of the bill, Galapagos penguins are unfeathered. Immediately
after molting, the base of the bill to the back of the eye is
covered with small, white feathers. These white feathers are soon
lost exposing distinctive facial markings. Each penguin can be
recognized by the pattern of marking on this unfeathered area (Plate
2). Patterns on the left side of the head may differ from those
on the right. 68
The general characteristics of the facial pattern differs between the sexes. The unfeathered region is predominately black with some pink. The pink color seems to be present only during breeding. IVhen the penguins are not breeding, the pink patches of both males and females are whitish. Although this characteristic is
only generally true, males have more pink than do females. Females usually have some pink, but these patches of skin are smaller and may be limited to a spot near the eye. Some females have extensive pink patches while others have little or none. The pattern and amount of pink is highly variable between individuals, but considered with other characteristics these facial patterns can be used to distinguish the sexes.
Penguins are considered "juveniles" once they have fledged the nest. The greatest differences between juvenile and adult plumage are in coloration, markings and facial patterns (Plate 3). After their first molt, juveniles possess adult plumage but the most striking difference in plumage is the juvenile's gray appearance.
Wherever the adult has black plumage, the juvenile Galapagos penguin has gray. Juveniles do not have facial and body bands but the side of the head does have a white cheek patch which becomes less dis tinct with age. The gray back and the white cheek patch are the major distinguishing characteristics of the juvenile. Other differences include the lack of an unfeathered facial region. In
juveniles the skin behind the mandibles and around the eyes is Table 5. A comparison of body parts in male, female and juvenile Galapagos penguins. The number of Galapagos penguins measured, the mean and the standard error of the mean of the body part, and the F value of the comparison are given. *** = P<.001.
SEXUAL DIMORPHISM
0 ^ 9 KN O M PAIRS N ( f n Q X± (SE] mm (f X ± (SE) mm Q F Bill Length 93 83 58.18±(.35) 53.87±(.20)*' 111.59*** Bill Width 93 84 19.94±(.36] 16.65±(.10) 69.41*** Flipper Length 79 71 118.74±(.75) 114.09±(.74) 19.28*** Toe Nail 81 70 16.21±(.13) 15.0S±(.ll) 40.56*** Weight 33 34 211.88±(3.79J 187.79±(2.98) 25.09***
( f JUVENILES NO' N J X ± (SE) mm (f X± (SE) mm J F Bill Length 158 52 58.38±(.23) 54.87±(.43) 56.21*** Bill Width 158 51 19.84+C.22) 16.49±(.18) 68.13*** Flipper Length 126 32 118.66±(.55) 117.28±(1.30) 1.18 Toe Nail 134 49 16.41±(.10) 16.06±(.14) 3.40 Weight 61 12 218.26±(2.85) 190.08+(7.06) 15.54***
Q JUVENILES N 0 N J X ± (SE) mm Q X ± (SE) mm J F Bill Length 160^ 52 53.99±(.20) ^ 54.87±(.43) 4.17 Bill Width 161 51 16.68+(.07) 16.49±(.18) 1.34 Flipper Length 119 32 115.25±(.58) 117.281(1.30) 9.39 Toe Nail 114 49 15.09±(.09) 16.05±(.14) 31.19*** Weight 51 12 187.02±(2.30) 190.081(7.06) .28 covered with gray feathers. Tlie bill is normally darker in color than the adult's and lacks the lighter shading on the lower mandible.
The feet are gray with a whitish or lighter pattern while adults have black feet. The eye color is dark gray in juveniles and gets lighter with age until it becomes pinkish.
Size. Unlike many species of birds (Selander, 1966; Amadon,
1959), penguins are not obviously sexually dimorphic. The plumage of the different sexes is similar, and only astute observation makes it possible to distinguish between males and females. Re search on other species of penguins has shown the sexes can be distinguished by size, weight, measurements of body parts and behavior (Warham, 1972: Richdale, 1951; O'Brien, 1940; Kinsky, 1960;
Sladen, 1958; Stonehouse, 1970, 1971). The Galapagos penguin is not any different.
Body parts which were measured such as bill length (L), bill width (W) and flipper length (F), toe nail (TN) and weight (OT) showed differences between adult males, adult females and juveniles.
A histogram of the measurements can be found in Figures 8 to 11.
The males were significantly larger (P<.001) than the females in all five body measurements (Table 5). It is apparent, therefore, that the sexes are dimorphic in plumage, behavior and size.
Juveniles (Table 5) have a significantly (P<.001) smaller bill length, bill width, and are lighter than males, but juveniles have significantly larger bill lengths, flipper lengths and toe nails Figure 8. Tlie bill length of 403 Galapagos penguins. The X axis is divided into the following 5 groups: males, females, juveniles, sex undetermined adults and chicks. The Y axis gives the bill length in mm. Tlie * symbolizes a penguin with the respective bill length. Mien more than 10 penguins fall into any one class nine * are printed followed by the number in the class. The mean bill length, standard deviation and the number of penguins (N) in each of the 5 classes are listed under the symbol for the group. POO.nnr. 7P 0 .ono 76 0.00 0 7 4 0 .0 0 0
700 .non
600 .000
600.000 5P0.OO0
56 0 .n o n 3 r O:! X- O % " =r -l f. 2 0.000 :•/: * V: !( -r *5^*% ;;c 1,?^ 500.000 A P n . nn o :X » ::: :% 460.000 z 400.000 y 300.000 3 6 0 . n o n 340 .n o n d 3 0 0 .0 0 0 NUMBER CLASS 2F 0.000 m 760.000 o " Q JUV. UNO CHICKS 0 ,o 3C.C13 54'- .711 5'1 .1:7 0.0 77.671 3 1.16 0 . 17“ . 5 7 . 11 . Figure 9. The bill width of 403 Galapagos penguins. The X axis is divided into the following 5 groups: males, females, juveniles, sex undetermined adults and chicks. The Y axis gives the bill width in mm. The * symbolizes a penguin with the respective bill width. IVhen more than 10 penguins fall into any one class nine * are printed followed by the number in the class. The mean bill width, standard deviation and number of penguins in each of the 5 classes are listed under the symbol for the BILL ^ WIDTH rmmj
o r\) '-o -ji o -n o •-* o > Ji ~.i -T o I-* > .J- -J o ~> lo w '.n JD O '.n O ji O vn O Oi O .3 -n O .n Q .1 O ^ O ■J’ O '-H O
ggg333ggEg33g§3g3Eg3gg33 OO ODD 0 00 00-3 000^)00 3 00 0 0 0 0
"LL ! -o V) Z_ ->}. ^ TfogLif■• 5 îîH ^ îti III >-C_ Vr 5 * * Lg^Lii" o II lii '=^
7\ ! LO Figure 10. The flipper length of 282 Galapagos penguins. The X axis is divided into the following 5 groups: males, females, juveniles, sex undetermined adults and chicks. The Y axis gives the flipper length in mm. The * symbolizes a penguin with the respective flipper length. V.Tien more than 10 penguins fall into any one class nine * are printed followed by the nuiriber in the class. The mean flipper length, standard deviation and number of penguins in each of the 5 classes are listed under the symbol for the group. 14?.non 140.000 138 .000 136.000 134.000 133.000 I130.000 138.000 I 136.000 *4* i- 124.000 *16**4 * O 123.000 11* *** 120.000 **** * * :X * *16 * * * * * * * * * * z 118.000 ********* 13*********14**** ÜJ 116.000 * * * * * * * * * 1o * ** * * * * * * 12 * * * « z 114.000 * * * * * * * * * 13*********? 3* * * * 113.000 * * ****** * *, * * * * * **.*13* * 110.000 108.000 106,000 104.000 ÛL 102 .000 100.000 QP .OOO 06 .000 u! _ 04.OOO
90,000 NUMBER IN CLASS (7 Q JUV. MF^M 0,0 18.086 1 1 ^ 2 5 2 i n . 381118. WND- S OFV 0,0 6,855 6,386 7.371 0.0 N .130. 119, 33.____ Figure 11. The toe nail length of 309 Galapagos penguins. Tlie X axis is divided into the following 5 groups: males, females, juveniles, sex undetermined adults and chicks. The Y axis gives the toe nail length in mm. The * symbolizes a penguin with the respective toe nail length. IVhen more than 10 penguins fall into any one class nine * are printed followed by the number in the class. The mean toe nail length, standard deviation and number of penguins in each of the 5 classes are listed under the symbol for the group. s .
Q . c
iT) < c
: t : : . SI
: ïl HSÎ : :
*-**#*. 4!^ -J!- * * * * -I!- # # *. -JÎ . 0 : 2 ^ : ÆmwwMïi-Æî-îï
O C C C C C G c. c c c c c c o c c c c c o c c c c c c c e c ggë0gggggig§èigipggS?Sëg£è?gg 9 ?• s s a & 13 sp s ;gïï Î- ÿ a s 5 3-3 ïvift a s ;« a g 2 a 3 s
i ! : I (ujai) H19N31 HIVN 301 than adult females. They do not differ from females in bill width and weight. At first, these results might appear curious, but it is not surprising because juveniles of both sexes were measured and considered together. If the adult sexes are dimorphic as the observations and data has indicated, then it is not surprising that juveniles also show sexual dimorphism. Therefore, differences in dimensions of body parts between adults and juveniles indicate that juveniles are also dimorphic. By taking all the measurements of juveniles of both sexes and comparing them to adults of one sex, the sexual differences are obscured.
Bill length and bill width of juveniles are significantly smaller than bill size of adult males. Bill lengths of juveniles are slightly longer [P<.05] than the female bill lengths. However, the juvenile bill is shorter than the male bill. The bill width of juveniles is similar to the bill width of females and significantly
(P<.001) shorter than the bill width of males. These data imply that bill width is a secondary sexual characteristic which develops sometime after male penguins acquire adult plumage.
Flippers of juveniles were slightly longer (P<.05) than the flippers of females and not significantly different from male flippers. This data is not meaningful because many juveniles were measured during molting when flippers are swollen (Fig.12 ).
Measurements of swollen flippers bias the data which indicates that juvenile flippers are longer than they actually are. Toe nail Figure 12. Flipper length of 32 juvenile Galapagos penguins.
Molting juveniles had longer flippers due to swelling.
Consequently, flipper lengths above 118 mm artificially
inflate the mean flipper length for normal (non-molting)
juvenile penguins. NUMBER IN CLASS
T1 c Tt m
m
3 3 differences, however, are not biased by molting and are similar to males and longer than toe nails of females. Tlie bimodal nature of the histogram suggests that two normal distributions which overlap are involved, one for each sex (Figure 13). Therefore, toe nails of male and female juvenile penguins are sexually dimorphic as are toe nails of adults. Juveniles have long toe nails similar to adult males, but Figure 13 does show that some juveniles have longer toe nails than adult males, suggesting that toe nail length depends on age. Younger penguins (Richdale, 1951) have longer toe nails than older penguins which have worn their toe nails down. Presumably, toe nails grow while the penguin is a juvenile and a young adult, so that longer toe nails are perhaps under-represented in this sample.
Juveniles are similar in weight to females, while adult males are heavier than juveniles, suggesting that increased weight is a secondary sexual characteristic like bill width. Because weights of molting juveniles were considered in the sample, juvenile weights have a higher variance than weights of males or females. Weights ' of molting birds would tend to increase the mean weight of juveniles.
Movement and Mortality
Penguins were banded to determine how males, females, juveniles and chicks differ in mortality and in faithfulness to a location
(Table 6). Galapagos penguins banded during the June to September
1971 season were scored with a plus if they were sighted between Figure 13. Toe nail length of 49 juvenile Galapagos penguins.
The mean toe nail length of adult males and adult
females is marked. The bimodal nature of these
data indicate that juveniles are sexually dimorphic. n u m b e r in e a c h c l a s s
_î______Ï______i_ Table 6* Banding and recapture data for Galapagos penguins. = August 1970, Tg = July • ■ September 1971 f To = January - ! ELIZABETH HAMMOND mangle TOTAL r, Tq 'OTAL ^ TOTAL TOTAL N U M B E R T t T, Ty Tl Tj' Ty T, total r, rs to ta l Tj Tq (T BANDED 7 ;s7, 14 15 Î88 3 ' 3' ' 1 I ' '1 16 0 ■ ! i 'V 1 ’ l'’ ( recaptured ). (I 8) (17) 13) 1(4 8)1 (1) (0) i(D|(0) |(0)|,o) ,(0), I 1 (114) ^ B. I 1 3 22 |s6 1 1 1 1 2 1 I 1 t 152 18 2 0 1 90 1 1 I 1 2 1 I I I I I -1 1 1 1 5 3 ( 0 REC. ) ,,, ( 1 6>(?1 ) (9)I«7)' (1) (0) '(1)' 1 '( 1 )*(45) (1 0) (I 3)^68)^ (0) (0) '(C)'(O) '(0)' ,0) 1(0)1 (117) 1 1 ) 1 1 JUV. B. 12 2 1 1 2 8 5 7 1)4 , r f i ' .. ( JUV. REC.) (9) (0) ' (9 >1 (2) |( 2)1 (0) ,<0),,o (0) (0)j(4)| I I I I 1 , (1 5) CHICK B.. 9 5 ,14 I 1 1 5 3 1 1 ( c h ic k REC.) (2) ( 1 ) '(3)' ' '(14) (0) '(14)' (0) '(0)' ' ' I I 1 ' 1 1 1 1 1 ' , 1 SEX UNO. ADULT 9 1 1 1 1 1 ( UNO. REC.) 1 I , 1 1 1 1 1 I I 1 1 TOTAL BANDED 4 32 TOTAL recaptured 2 63 January and March, 1972, and with another plus if they were sighted between June and October, 1972, The pattern of pluses (recaptures) and minuses (not recaptui'ed) was compared to a binomial distribution for each group, and then groups were compared to each other using a chi square test (Table 7). There is no reason to assume penguins should occur in a binomial distribution, but it is useful to know whether recaptures depart from chance. Movement of male and female. The pattern of movement for males and females was not significantly different (Table 7) indicating that recaptures of both sexes were similar in direction, but males and not females departed significantly (P<.001) from chance. Al though both males and females follow the same pattern of movement, males were recaptured less frequently than females. This means that males move more than females or have a higher mortality rate than females. If males are more abundant than females, as has been found for some other penguins (Stonehouse, 1970; Richdale, 1951, 1957) males may have trouble acquiring matps which might force them to move more than females. Observations on the Galapagos penguin population showed that unemployed penguins during the breeding season were males and not females. Five banded males were unmated for a breeding season while females, once they had bred, were never seen unmated. Unmated males may travel from place to place in search of mates which would explain the higher disappearance of males. Females, in contrast, probably because they are mated, show greater fidelity to a location than males which are presumably more prone to wander in search of mates. Throughout the study period, it was apparent that nonbreeding birds of both sexes would disappear for short periods of time. Both sexes frequently return to the place where they have bred even when not breeding, indicating that they have some affinity for the location. Movement of young. Penguins which were banded when they were juveniles or chicks (Table 7) were combined to test if their movement was a chance occurrence and if the young moved similarly to adult males and females. Tlie juvenile and chick recaptures depart significantly (P<.001) from chance and are significantly different (P<.001) from the recaptures of males or females which indicates that juvenile penguins have a higher mortality and wander more than adults. Young penguins compared to adults (Table 8) are half as likely to be recaptured. Furthermore, comparing the movement of recaptured adults with the movement of recaptured young indicates that young move more than adults (Table 7, P<.001). Two penguins which were banded as juveniles were recaptured, but no chicks were recaptured in the same time period. Of fourteen adults banded by Leveque (1962), three have been recaptured. The 3 recaptured were males, and all were banded where they were recaptured. This suggests that adults may have a higher survival rate than young. It may be that Recapture data on Galapagos penguins banded July-September 1971. The following indicates + = recaptured, - = not recaptured, ?2 = January-February 1972, T? = June-September 1972. * = P<.05, ** = P<.01, *** = P<.001. The number of adult males, adult females, juveniles and penguins banded as chicks in each time period is given according to whether they were recaptured or not recaptured. The observed values of each group were compared to a binomial distribution and to other age and sex groups. The Chi squared value and degrees of freedom are given. TIME OBSERVED DISTRIBUTION PERIOD VALUES COMPARISON x2 df c f q j C To (f : binomial 28.29*** 3 + + ^ ^ 2 0 0 : binomial 5.58 9 17 11 16 ! ) ( ' ■ binomial 13.30** 1 10 10 2 0 e g o ^ ^ 9 3 ^ 12 ^ 30 41.40*** 3 Juvenile/Chick combined y J __ 44.72*** 3 + +\.4 adult :jToung 80.94*** 1 - Table g. Recapture data on 423 Galapagos penguins of îînovn age or sex. Percentage of recaptured penguins vas computed from penguins banded from July 1971 to Karcli 1972. 77'e banding location and the percentage recaptured is given. The number (N) of penguins which moved, their breeding status, banding location and recapture location(s) in seciuence are given for each of the four groups. INOSA E LIZ. BAY . penguin „ S’hic^h^ MATED N % REC. % REC. %REC. BANDING . B ANDED LOCATIONS PTA. EllZ. PTA 8 7-5 8 6.3 8 6 . 7 3 0 YES ' ESPINOSA BAY ESPINOSA c T ? 1 CABO E llZ . CABO DOUGLAS BAY DOUGLAS YES 1 PTA. ESPINOSA ELIZ, BAY 7 8 . 6 82.1 2.1 ? 1 CABO PTA. CABO 9 1 53 3 8.1 DOUGLAS ESPINOSA DOUGLAS CABO DOUGLAS CABO J U V 5 7 6 6.3 1 4.8 3 1 7 26.7 DOUGLAS ESPINOSA esA' nosa r i DOUGLAS esA' n o s abay 2 1 .4 C H I C K 53 36.8 32.7 17.6 ESPINOSA z ’ ESPINOSA ’ young female had just completed first molt to adult plumage vhen banded y- X the Galapagos penguin like the young Royal penguin ÇEudyptes chrysolophus) (Garrick and Ingham, 1970) changes locations and has a higher mortality rate early in life. Some individuals have moved at least 80 miles and one penguin moved 40 miles in less than a month. Although the data is limited recaptures of juveniles at Elizabeth Bay and Pta. Espinosa (TdbJe 6) indicate juveniles may move away from Elizabeth Bay which has the greatest concentration of breeding pen guins. 'Hiis may be a result of adult intolerance to young since we would predict young should migrate to the best foraging regions. Breeding Activity Pair bonding. The Galapagos penguins frequently mates for more than one season, but not neceesarily for life. Of 79 Galapagos penguins which mated for two or more breeding seasons, nine birds chose different mates (11%). Of these five were females and four were males. One female and three males switched partners after their mates disappeared (these presumably died since they were never seen again). Three females switched mates when their old mate was still available. One male whose female partner mated with another male stayed unmated for two breeding seasons. Another male, which changed mates, courted a new female whose mate had died after his previous mate was mated. Before the female finally paired with this male and laid eggs, she kept company with two other males. The history of one female which changed mates was not known because her previous mate was unbanded. Nevertheless, pair bonds seem to be quite lasting, and females seem to choose their mates. This data supports an earlier premise that males may be more abundant than females since males and not females may be nnmated. The length of the pair bond seems to affect the length of time spent on land prior to egg-laying. Galapagos penguins which were mated for the first time spent at least a few days on land prior to egg-laying. One male spent 14 days in and out of his nesting site before his mate laid eggs. The female spent a few days in the site before she laid eggs. In contrast, three previously mated penguins were seen only one or two days in advance of egg-laying. One pair was never seen in the nest site during the day prior to the laying of the first egg. In general, males occupy the nesting site before the first egg is laid, donkey braying in the early morning and evening. Shortly before egg-laying both the male and female can be found in the nest site during the day as well as at night. Females frequently stay on the nest the whole period of egg laying. In 6 out of 13 nests the female stayed the entire egg-laying period, and in 6 of the remaining 7 she left for merely one day. In only 1 nest out of 13 did the female leave for more than one day (in this nest the female was equal to her weight at the end of molting). This female's unusual egg-laying period lasting 6 days was undoubtedly a result of insufficient food. Eggs. A clutch of two eggs was laid in all 15 nests which were discovered before or on the day the first egg appeared. First and second eggs were laid 3 to 4 days apart in all 15 nests and in five of the 15 nests an egg was later lost. Throughout the study nests were found with only one egg or one chick which presumably reflects a previous loss ancj only rarely, if ever, a clutch of one. Two eggs seem to be the normal clutch. The laying of replacement eggs appears to be rare since only one bird was known to lay a replacement egg. The first egg this penguin laid had a very thin shell which immediately broke when incubation was attempted. This was the only egg seen with these characteristics. Ten days later she laid another egg and three days later her third egg to complete the clutch. Other penguins which lost an egg a few days after their clutch was completed did not lay another egg. One individual which lost one egg five days after it was laid neither replaced it ncr the one slie lost seven days later Other species of penguins appear to replace the first egg if it is lost soon after laying (Taylor, 1962). Replacement egg laying, however, appears to be unusual for Galapagos penguins. The Galapagos penguin lays eggs of unusual shape (Table 9). First eggs are significantly longer (P<.05) than second eggs while second eggs are significantly wider (P<.05) than first eggs (Table 9). Eggs which did not hatch (addled eggs) were measured. It was not determined why the eggs were addled, or if they were fertile. The dimensions of addled eggs (Table 9) were compared to first and second eggs (Table 9). Addled eggs are significantly shorter(P<.05) than first eggs and significantly wider (P<.1) than first eggs. Addled eggs are not significantly different from second eggs. Thus, Table 9. Dimensions of Galapagos penguin eggs and a comparison of the egg dimensions using a modified t test where variance may be unequal. F = first egg laid, S = second egg laid, A = addled egg, * = P<.1, ** = P<.05, *** = P<.005. DIMENSIONS OF GALAPAGOS PENGUIN EGGS CLASS N X LENGTH! (SE] mm X WIDTH! ÇSE) mm X VOLUME m^ F 74 62.51+C.27) 47.88±(.14) 7.16 S 65 61.74+(.30) 48.62±(.20) 7.30 A 27 58.77±(2.10) 50.371(1.92) 7.46 COMPARISON OF THE DIMENSIONS OF FIRST, SECOND, AND ADDLED EGGS D£ t Value F length : S length 137 -1.88*' F width ;: S width 137 3.01*' F length ; A length 26 -1.73*' F width ;; A width 26 1.26* S length :; A length 26 -1.38 S width ;; A width 26 .88 it appears that addled eggs are second eggs. Incubation. Incubation commences with the laying of the first egg so that chicks hatch asynchronously. Both sexes incubate and make nest changes frequently which spreads the cost of incubation and allows maximum foraging time for each penguin. The only time both penguins are in the nesting site together is in the evening or early morning when penguins do not forage. Nest changes for both males and females vary between one and ten days. The mean nest stay for males is 1.90 days with a standard error of the mean of ±.09 days for 106 observations. Nest stays for females are slightly longer with a mean value of 2.00 days for 115 observations and a standard error of the mean of ±1.07 days. Incubation exceeds 38 days and 2 zoo records (Roberts, 1940; Richdale, 1957) report between 39 and 42 days. Chicks. During the last days the egg is incubated, the chick can be heard peeping within the egg. Usually the next day the egg is pipped and the chick continues to vocalize. After a day the chick frees itself from the egg. In one case the chick was free of the egg after 15 minutes once it started to cut the shell with its egg tooth. Another chick took over 8 hours to complete the process. First and second eggs hatch 2 to 3 days apart. Until the chicks are partly grown, nest stays remain the same. For two weeks the chicks are brooded; which may indicate that the chicks can not therraoregulate effectively at a young age. After this period the adult must remain at the nest site to protect the chicks from preda tors. The main natural predators are the Sally lightfoot crab (Grapsus grapsus), the Rice rat ÇOryzomis nesoryzomis narboroughi) and the Galapagos snake (Dromicus selvini, Dromicus dorsalis). On Fernandina there may be a third species of snake yet undescribed. If an adult remains in the nest until the chick is well grown, these predators are impotent since an adult penguin is a formidable opponent. Pr e d a t i o r ^ ^ g # M 0 ^ ^ g ^ s e d in more detail under predation and death. O n c ^ ^ ^ ^ ^ B ^ ^ ^ ^ ^ ^ ^ ^ ^ x i m a t e l y 30 days of age, neither adult have only one chick, adult ma’^B B B ^ ^ ^ H j ^ B ^ ^ H ^ ^ ^ ^ B a l l y : while with two adul wc^^^^^NN^^^^^^^^^^^moughout study. The general growth parts such as bill length, bill width, flipper, o 5 ^ P ® ^ ^ a t h e r , and foot for known age chicks can be found in Figures 14 to 18. Body parts grow at differ ent rates and different times. All body parts measured reached at least the mean adult female size at the time of fledging (Figure 18); and flippers and feet of some chicks were as large as the flippers and feet of male adults. This must be a reflection of sexual dimor phism even in chicks for some characteristics. Table 10 shows the means and variance for various body parts of males, females, and juveniles. Some body parts such as the foot grows to adult size thermoregulate effectively at a young age. After this period the adult must remain at the nest site to protect the chicks from preda tors. Tlie main natural predators are the Sally lightfoot crab (Grapsus grapsus), the Rice rat (Oryzomis nesoryzomis narboroughi) and the Galapagos snake (Dromicus selvini, Dromicus dorsalis). On Fernandina there may be a third species of snake yet undescribed. If an adult remains in the nest until the chick is well grown, these predators are impotent since an adult penguin is a formidable opponent. Predation will be discussed in more detail under predation and death. Once the chicks reach approximately 30 days of age, neither adult remains at the nest. In nests that have only one chick, one adult may stay with the chick occasionally;while in nest with two chicks adults never remain. Chicks were weighed and measured throughout this study. The general growth rates for different body parts such as bill length, bill width, flipper, oil gland, feather, and foot for known age chicks can be found in Figures 14 to 18. Body parts grow at differ ent rates and different times. All body parts measured reached at least the mean adult female size at the time of fledging (Figure 18); and flippers and feet of some chicks were as large as the flippers and feet of male adults. This must be a reflection of sexual dimor phism even in chicks for some characteristics. Table 10 shows the means and variance for various body parts of males, females, and juveniles. Some body parts such as the foot grows to adult size Figure 14 . The bill length for a k n o ™ age Galapagos penguin chick is plotted against the age of the chick for 149 observations on over 60 chicks. The observed value = 0 was compared to a parabola. KTien observed and predicted values coincide a B is given while P = predicted values. Mean adult bill lengths are indicated. PLOT OF OBSERVED AND PREDICTED VALUES. 3.750 5.250 6.750, . 1 - 0O 60BU 00BPOO_ ° 0 ( y . . 2 . 2 5 0 . __ 3.750 CR*PH SCAIE EXTENDS F D.fOCO TO 6.4000 ILL LENGTH (C/Tl.) Figure 15. Tlie bill width for known age Galapagos penguin chick is plotted against the age of the chick for 147 observations on over 60 chicks. The observed value = 0 was compared to a parabola. kTien observed and predicted values coincide a B is given while P = predicted values. Mean adult bill widths are indicated. (UJO) HiaiM 1118 S2£VI Oi SZS9T0 wcyj SQN3iX3 31»3S Hdvas ri— ,,6-0 - « = .0...------ >9-l = 0'' 56-l=p 0 0 0 g od 00 000 «•I m 0 0 0 ooa 0 000 o 0 ... 00 0 d 900 0 CO 00 0 d03 000 0 0 0 d dOO 0 — 0 — 00— od‘do‘^“ ~o'~o------OOdO 0 0 dd 0 03 0 0 •samvA a3i3io3dd qnv o3Aa3sno do iood • Figure 16. Tlie flipper length for a kno\m age Galapagos penguin chick is plotted against the age of the chick for 195 observations on over 60 chicks. The observed value = 0 was compared to a parabola. When observed and predicted values coincide B is given while P = predicted values. Mean adult flipper lengths are indicated. F OBSERVED AND PREDICTED VALUES. ------■ 6-300 1 o _— . . . . . ------in ------z 0 0 0 BO 0 0 - W - — g ...... ° ° " : : r “ . .. _... d'-M»». _ • ^ !—------0.300 ■ E EXTENDS FROM FLIPPER LENGTH jfe/7). j Figure 17. The oil gland feather length for a kno\m age Galapagos penguin chick is plotted against the age of the chick for 107 observations on over 60 chicks. The observed value = 0 was compared to a parabola. hTien observed and predicted values coincide B is given while P = predicted values for the parabola. Mean adult oil gland feathers are indicated. PLOT OF ObSERVCD AND PREDICTED VALUES. ■ ■ 0.150 O.A50 0.750 1.050' 1.350 ! -0.000 0.300 0.600 0.700 1.200 ------ ■ ‘ op® ...... ■ ...... — -Q DP ° d ° ■ ° ... . . ^.o.oo« ...... • 0 ° 0 V d P=' “ - ° ...... ° p' 0 . 0 ...... - ...... O O P 0 0 c ’l 1...... 30.000 Ù T °— ^ ...... \ ' 10.000 ...... OIL GLAND FEATHER (cm.) Figure 18. The foot length of a known age Galapagos penguin chick for 127 observations on over 60 chicks. The observed value = 0 was compared to a parabola. hTien observed and predicted values coincide B is given; while P = the predicted value for the parabola. Mean adult foot lengths are indicated. PLOT OP CbSlRVtO *N0 PREOICTtO VALUES. 3.750 5.250 6.750 O -- “ 1/1 » - 2 #'00 0 . 0 E EXTEND^ FROM , 2.5500 T FOOT LENGTH (CfT}. ) Figure 19. The weight of a knovm age Galapagos penguin chick is plotted against the age of the chick for 215 observations on 93 chicks. The observed value = 0 was compared to a parabola. IVhen observed and predicted values coincide B is given, while P = the predicted value for the parabola. Mean adult weights are indicated. PLOT OP OBSERVED AND PREDICTED VALUES. 00 OBOBOO 0 0 - aoo;ooo-..... — izoo'.doo...... iaoo. H SCALE EXTENDS f w e i g h t " IN "GRAMS Table 10. Body part measurements in cm for Galapagos penguins. H = height, FT = foot length, F(L] = total flipper length, F(W) = flipper width, F(TH) = flipper thickness, F*(W) = flipper width molting, F*(TH) = flipper thickness molting, TF = tail feather, OG = oil gland feather. X = the mean value and N = the number of penguins in the sample. MEASUREMENTS OF BODY PARTS IN GALAPAGOS PENGUINS H FT F(L) FCW) FCTH) F* (W) F*(TH) TF OG Males: X 22.90 9.68 14.80 2.38 .82 2.77 .93 N 76 35 92 66 66 35 31 X .93 .37 .68 .18 .09 .38 .10 X .12 .06 .07 .02 .01 .06 .02 Females: X 21.32 9.06 14.10 2 ^ # .84 2.76 .86 N 59 41 69 50 48 - 35 31 X .92 .34 .71 .20 .24 .44 .11 X .12 .05 .08 .03 .03 .07 .02 Both Sexes: X 2.41 1.34 N 28 29 X .16 .25 X .03 .05 Juveniles: X 21.72 N 45 X l.OI X 1.51 early in the chick stage, while the oil gland feathers do not start growing until late in the chick stage and do not reach adult size until just before fledging. The growth patterns of flippers and body mass (weight) do not change from season to season. Figure 19 illustrates the general growth pattern (weight increase) of chicks. Most of the body parts reach adult size shortly after 30 days, which presumably means that in a bad year when food was scarce chicks which fledged early were equipped to survive if they located food. IVhen food is available, early fledging will be detrimental to survival since mass and feathers are still increasing. The oil gland feathers (Fig. 17) do not become fully grown until around 55 days. Insulation and food problems would undoubtedly hamper survival in young chicks which fledge early. Synchrony in egg laying and hatching. During the June to September 1972 season, local breeding groups were synchronous in egg laying, but the entire population was not in synchrony. Penguins in Elizabeth Bay, Isabela, bred around the middle of August, while penguins at Pta. Espinosa, Fernandina, bred in late August and September. Penguins at Cabo Douglas, Fernandina, were just starting to court in September 1972. These changes in the degree of synchrony of the entire population do not seem unusual. In 1970 and 1971 penguins appeared fairly synchronous with eggs at all locations being laid in September, but between January and March 1972 the same locations were in different stages of reproduction. Figure 19 gives the known breeding periods over the last 2 years at five locales. Examining one location for egg laying dates shows that the penguins are fairly s>'nchronous at a particular location. At Pta. Espinosa, Fernandina, 50 eggs were laid over a 35 day period in August and September, 1972. Twenty of the 39 eggs were laid in an 11 day period from August 25 to September 4, 1972. Hatching dates in July and August, 1971 were highly s>Tichronous for four different locations. Not only were breeding colonies s>Tichronous but the entire population was in synchrony. Eighty percent of a sample of 88 chicks hatched within a ten day period (Fig. 21). The mean hatching date is 9.82 days, if July 19 is considered day one. The standard deviation of the sample is 12.84 days. Using a z test, there is a 95% chance of hatching after the fourth day and a 93% chance of hatching before the 15th day. Frequency of Breeding. In June 1971, 74 pairs of breeding penguins were banded. In January 1972, 48% of the 74 pairs bred and in August and September 1972, 50% bred. During a 15 month period, 74% of the penguins bred at least twice and 24% bred all three seasons surveyed. Of the 74 pairs, 11 skipped a breeding season because of molting. This means 15% of the pairs did not breed because one or both of the pair was molting. Four pairs were molting in January when they were to breed and 7 in September 1972 when they were to breed. No pair was knowTi to skip twice because of Figure 20. Breeding of Galapagos penguins at 7 locations in 1970, 1971, and 1972. ••• = eggs, ----= chicks, /// = no breeding, m = molting, and c = courtship. Overlapping lines indicate another influx of breeding at the location. ESPINOSA iiniiiiiiiiiiiiiniiiiiiwiiiinm. ...II .V/////17 CABO ' DOUGLAS C aSO HAMvlOND M a n g l e E L IZ A ^H BAY PTA. ALBEMARLE TAGUS COVE W f 'm 'a 'm ’J Ij 'a I’S' o ' n ' D 'j ' F'm 'a Im ' j l j U • S •© ' N ' D ' F' M'A'' m 'J"'J"'A' ‘s ' 0 ' N I Figure 21. Galapagos penguin chicks hatching at 4 locations on Fernandina and Isabela. The sample size was 88 chicks and the location, number and date of chicks hatching is shown. Cabo Hammond Pta. Espinosa Cabo Douglas I I Elizabeth Bay CJl —I a O' c- 5 . 1 M . EL I i I- Si- ■a o molting, however, 3% of the penguins did not breed because they died. Since :i:olting requires at least a month to gain weight and lose feathers, successful reproduction requires at least 4 months. Thus, a foLU'th of the penguin population is breeding at the near maximum rate of twice a year which is theoretically possible. Why did 76% of the banded group not breed every five or six months? Approximrii ely 25% did not breed because they were molting or had died and 4% which did not breed switched mates. Why the other 37% did not i; eed each season is not known, but some probably were getting oady to molt or just completed the molt and were not detecti ' Others might have been underweight or in poor physical conditj V- which could prevent them from taking a reproductive risk. Breeding is not only forgone by individuals but sometimes by the entire •, -pulation. In 1970 and 1971, penguins were breeding during May to -:ptember (Fig. 22) but in 1972 breeding never commenced between. March and August 1972. Thus, breeding appears to be flexible for in.Mviduals as well as for the entire population. No set breeding period can be defined and instead breeding becomes a fairl) rratic and unpredictable event. Molting. Once an individual increases its feeding and has gained weight, it comes ashore to molt. The period of the molt is considered to be from the day the penguin first stays ashore until it re-enters the water with new plumage (Richdale, 1957). Wlaen Galapagos Figure 22. Reconstruction of breeding in Galapagos penguins. Observations of breeding are shown on the first figure and from these observations breeding activity is reconstructed. -- eggs, chicks, /// no activity, and C = courtship. Overlapping lines indicate simultaneous activity. '"--.zrrlllh EGGS • CHICKS X /X //N O ACTIVITY C CO’JflTSKi? penguins which showed signs of molting were captured, they were weighed. Of course, it was difficult to know if the penguin had just come ashore for the first time since Galapagos penguins move around and rest during the day on shore. Three divisions of molting were recognized, although it is possible to distinguish five (Penney, 1967). The first was before any feather loss, the second after feathers were dropped on parts of the back and belly and the third after feathers were completely lost and new feathers had started to grow. Tables 11 and 12 give the weight gains and losses for males, females and juveniles in these three categories. Weights of juveniles were not significantly different from males or females (P<.05). Because of the sampling technique, the first category probably underestimates weight gained, while the third category underestimates weight lost. In general, approximately 400 grams are gained for molting, and 600 grams are lost. As much as 1600 grams over the normal weight has been gained by a juvenile and 750 grams by an adult male. The pattern of feather loss and growth of new feathers is similar in all individuals of the Galapagos penguins. IVhen a penguin is about ready to lose all of its feathers and replace them, its plumage looks bro\mish, and the penguin look exceptionally fat. After a few days on shore the contour feathers start to stand out so the bird looks even fatter. These feathers are loosely attached to the new feathers (see Richdale, 1957, for further details). As Weights in grams of Galapagos penguins during the molt. N = number in sample, X = the mean weight in grams and SE = the standard error of the mean. Males are always heavier than females. MALE N X ± (SE) Average weight 61 2183 (28) Weight before molt 6 2572 (119) After some feather 6 2422 (225) New feathers growing 4 1728 (156) FEMALE N X ± (SE) grams Average weight 50 1870 (23) Weight before molt 2 2285 (350) After some feather 3 1927 (IIC) New feathers growing 7 1676 (59) JUVENILE N X ± (SE) Average weight 12 1901 (71) Weight before molt 10 2655 (124) After some feather 18 2222 (59) New feathers growing 14 1721 (55) Table 12. A comparison of weight in male, female and juvenile Galapagos penguins during three £tages of the molt. * = P<.05, ** = P<.01, and *** = P<.001. N = number in sample, X = mean weight in grams, SE = the standard error of the mean in grams. 9 /JUVENILE N 9 N.l X± (SE) grams 9 X ± (SB) grams J F Weight before molt 2 10 228.50±(3.S0) 265.50±(12.55) 1.660 Weight after feather loss on back S front 3 18 192.67+(11.78) 222.22±(5.93) 3.68 Weight after shed feathers 7 14 167.14±(5.87) 172.141(5.52) .32 C f /JUVENILE N cf N.l X± (SE) grams d X± (SE) grams J F Weight before molt 6 10 257.17±(11.87) 265.501(12.35) .20 Weight after feather loss on back S front 6 18 242.17±(22.49) 222.221(5.93) 1.52 Weight after shed feathers 4 14 172.75±(15.64) 172.141(5.22) .00 the old feathers stand out more and look scruffy, they become loose and will come out in handfuls if the bird is touched. Feathers first fall off the middle of the back and belly. The flippers become swollen and nearly double in size. (See Table 10 for flipper dimensions during molting.) The penguin is very easily bruised and the flippers will bleed if they are bumped. As the molt pro gresses, the patch of feathers lost continues to expand exposing a white downy layer of feathers. Next, feathers are lost on the legs and around the neck followed by the flippers and head. The head and tail feathers are the last to be lost. Patches of feathers may adhere to the neck or tail or head so that some individuals take on a distinguished or odd appearance. One penguin looked as if it had a beard while others appeared to be wearing hats. During the molt, the penguin preens itself and often shakes. IVhen it shakes, feathers float through the air like when a t o m down pillow is tossed in the air. IVhen old feathers are first lost, the white do;m shows through, but within a few days the black tips of the new feathers can be seen. New feathers first grow in where old feathers are lost. The behavior of molting penguins is different from non-molting penguins. In general, the birds fast, avoid the water and seek out a microclimate such as their nesting site in which to molt. For a detailed account of molting, its significance, and its thermoregulatory effects see Boersma (1974). Galapagos penguins remain on shore for the molt from 10 to 15 days. The length of time spent on land may be partly a function of fat reserves. One penguin returned to the water after ten days. He still had head and neck feathers to molt, but his weight was only 1405 grams. This bird which was commonly seen before the molt was never seen again and presumied dead. Other penguins have been sighted for aging with old feathers still attached. These birds may be forced to return to the sea earlier than normal because of low fat reserves. The average weight loss of males is 455 grams while it is 181 grams for females and 199 grams for juveniles. The average weight loss of males is probably inflated since one of the males died. However, the differences are probably realistic in showing that males lose more weight than juveniles and females during the molt. Since males are still heavier than females after the molt, weight di morphism is not due solely to differences in the fat layer. Males are actually heavier than females when both have little or no fat layer. Because males are larger, they must require more energy to maintain themselves, but because of their size, they must also be able to withstand longer periods of starvation than females or juveniles. This ability to undergo greater weight loss may be why males have a lower mortality rate than females. Four times penguins have been found dead, and it has been possible to assess the condition of the plumage. Three died just after completing the molt with the stomach and intestines empty. The fourth penguin had an injured flipper and an empty stomach and intestines. The injury probably caused the penguin to starve to death. Because of the long period of fasting, it is not surprising that immediately after molting would be a time of high mortality, and would tend to be higher for females and juveniles, I would suspect that mortality would be highest in juveniles since they lack the adult's experience in gathering food. Molting in Galapagos penguins is not equally distributed throughout the year. Molting is most often seen just prior to, and during, the breeding season. One female laid an egg a week after molting while other individuals did not start breeding for a month or more. The blue-black plumage and sleekness of breeding Galapagos penguins attests that molting is a prelude to breeding. Galapagos penguins which molted during the breeding season were either juveniles or adults which did not breed during that season. Table 13 shows the numbers of molting penguins seen at Pta. Espinosa in different months. Molting was observed more than once for 13 penguins: 4 were adults, 8 were juveniles and one was of undetermined age when banded. Molting occurred as frequently as after five months. The longest period was 12 months after the previous molt. It would seem likely this individual molted once during the 12 month period which was not observed. The mean time between molts for the other 13 observations was 6.54 months and 124 Table 13. The number of molting Galapagos penguins observed at Pta. Espinosa, Fernandina during months in 1971 and 1972. CO g3 3 < 3 3 Ll) i 5 ^ O 2 r - ^ 0 ) g 0 ) a i ^ a GO < % (/) 3 y CD o 00 CD - 3 C/) the standard error of the mean was ±.21. Galapagos penguins, on the average, molt twice a year. Perhaps such a high frequency of molting maintains the plumage in prime condition and decreases the energy required to thermoregulate (Boersma, 1974). Food I have only observed the Galapagos penguin feeding on fish between 10 and 150 mm long. However, it seems possible that Galapagos penguins may feed on items other than fish since they often forage with Audubon’s Shearwaters which eat crustace. Food items which adults regurgitated to young were always fish. Mullet, sardines (identified by Dr. Ted Cavender) and fry are food items and on one occasion I observed a penguin eating a reef fish approximately 50 mm wide and 120 mm long. Tliis was the only time a penguin was seen manipulating a prey item above the water. Galapagos penguins were commonly seen in large groups at sea when water temperatures were below 23°C but only on three occasions when water temperatures were above 23°C (Figure 23). Surface water temperatures had been previously higher and had dropped when the two feeding frenzies at 25°C and 27°C were seen. Penguins were not seen in groups except during periods when feeding frenzies were quite common. Therefore, large foraging groups of twenty or more penguins and feeding frenzies are both associated with cooler surface temperatures. Figure 22 . The number of feeding frenzies seen at different surface water temperatures. Surface temperatures are in degrees Celsius. Only one feeding frenzy at a water temperature was recorded on a single day although more than one commonly occurred. Of 25 feeding frenzies, 22 occurred at surface water temperatures below 24°C. UJ N =25 I - 10 I 5 " r\J n -4- XL -+■ n 19 20 21 22 23 24 25 26 27 28 29 SURFACE WATER TEMPERATURES ( 0° ) 128 Feeding frenzies and larger groups of penguins are seen when water temperatures are low but it seems unlikely that water temperature alone is the cause for the change in behavior. A more logical hypothesis is that the resource patterning has changed. My daily records showed that 8 times when feeding frenzies and large groups of penguins were seen I saw large groups of small fish. Comparable numbers of fish were not seen when individual or a pair of penguins foraged together. Furthermore the behavior of the foraging birds changes, indicating the possibility of resource changes. Penguins commonly follow the coastline when foraging in small groups. For example, one penguin repeatedly attacked a school of yellow striped fish which were approximately 140 mm in length in the opening of a tidal pool. In contrast, groups of penguins are also seen away from the coastline. This does not mean that individuals are never seen at sea because they are, but groups are not seen hunting along the coastline. In summary, the frequency of feeding frenzies and the behavior of foraging penguins change in relation to surface water temperature. Maxwell (1974) found that • primary productivity is correlated with low water temperatures. Thus, feeding frenzies and group foraging are associated with higher productivity which probably reflects more abundant food. Changes in surface water temperature, primary productivity and feeding frenzies are likely results of variations in the currents of the Galapagos. Shifts in the Cromwell Current, which upwells around Fernandina (Maxwell, 1974) is probably the cause of variation in resources. Predation and Death Predation has been considered a major regulator of some popu lations (Mech, 1966; Caughley, 1970). However, there is disagree ment on the actual effect of predation on a population. Tliere are many possible predators of Galapagos penguins. On land hawks, introduced cats and introduced dogs are possible predators of adults. Young and eggs may be preyed upon by additional predators such as snakes, rats, and crabs. In the water sharks, fur seals, sea lions, and killer whales are all sufficiently large enough to prey on penguins. Observations show that the possible predators are in fact taking penguins. Nancy Jo, a research assistant for the Charles Dantfin Station, reported seeing a dog eating an adult penguin on Isabela. I have seen a cat stalking and investigating nesting sites along the cliffs at Tagus Cove, Isabela. David Balfour, an owner of the tourist boat the Golden Cachalot, reported a hawk capturing a penguin chick which was ready to fledge. The chick had an aber rant behavior pattern since it stood at the opening of its nesting site. (All chicks which I have observed huddle at the back of the nesting site.) I found evidence of a hawk eating on a carcass of an adult penguin but whether this adult was captured by a hawk or just eaten after it died is not knoim. Adults and juveniles with large gashes out of the posterior part of their bodies which probably are results of shark attacks have been observed. There is no indication that fur seals or sea lions attack penguins. Young sea lions often swim up to penguins standing on shore and sniff the penguins which walk out of their reach with no apparent con cern. Because penguins stand close to the water unless they are breeding, land predators would have difficulty capturing them. IVhen adults are breeding, the inexcessibility of their nesting sites should make predation unlikely. In the water the penguin's maneu verability and speed must make capture difficult. Although there are many possible predators of Galapagos penguins, the recapture rate of adults banded at Pta. Espinosa, Fernandina from June, 1971 to March, 1972 was 89%. Such a high recapture rate for adult penguins demonstrates that predation is insignificant as a population regula tor. Furthermore, such meager observations of predations suggest that predation is an unlikely event. Eggs and chicks encounter a different set of predators from adults. Rats, snakes and crabs all have been knoim. to take young and eggs. However, I have never observed either eggs or chicks taken before the nest is deserted. In most nesting sites a crab is present and I have seen them feed on the guano and alga in the nesting area. Rats are frequently seen running on the lava and even in nesting sites at night. The one nesting site where a rat was frequently observed fledged two chicks. After nests are deserted eggs frequently disappear. 5% of the eggs which were deserted in 1971 disappeared. I have noted that chicks like eggs frequently disappear. A .hick may appear perfectly healthy and the next nest check be gone without a trace. Two chicks less than a week old were present in one nest without an adult. Their weights were very low, less than the 60 graan weight at hatching. A crab dragged one of the young off, pulled the insides out the cloaca and ate it. Such an example illustrates that predators take eggs or chicks which are already deserted, or not attended. Tliese predators are insignificant as population regulators since they are only acting as scavengers and eating what would soon die. Two factors which may contribute to death by starvation and undoubtedly weaken the penguin are disease and parasites. Only one disease was recognized. This disease causes lesions on and around the bill which are filled with a whitish growth. Bills become swollen and often deformed. The nictitating membrane is clouded and sometimes a white nodule is present. Because of the appearance of the nodules and lesions, I suspect it is a fungal disease. Death sometimes results because the penguin stops feeding and stands on shore. About 3 times as many juveniles as adults have been sited with the disease. Out of 462 birds which were banded only one was captured with the disease. Perhaps 50 individuals over the entire study have been seen with the diseasejindicating it is rare. Mites and nemotodes are found on and in penguins. One nemotode was excreted by a breeding female. It is doubtful that these parasites cause death, but they undoubtedly weaken the bird. Harper et a^. (1967) suggest that parasites can be contributing factor in population fluctuations. At some critical period such as prior to the mo]t in Galapagos penguins, the parasites may sufficiently weaken an individual to hamper weight gains. The major cause of death, which will be discussed in greater detail subsequently, is lack of food at some critical point prior to the molt. If penguins do not gather enough resources to molt properly, death will result. Resource limitation regulates the population by limiting the number of successful breeders, and the number of successful molts. Breeding Seasons The few observations of breeding in the Galapagos penguin (Harris, 1969b)have led to the belief that the penguin has an ex tended breeding period. Figure 22 shows all known observations of breeding and a reconstruction of the breeding period. The recon struction of breeding was made by taking an observation such as Gifford's (1913) account of a female containing well developed eggs and assuming that the eggs would have been laid and chicks fledged. By knowing the length of incubation and the length of the chick stage the breeding period can be determined. If the breeding periods for all years are collapsed to determine the length of the breeding season, it is apparent that Galapagos penguins have been found breeding in every month of the year. This does not mean that the penguins breed year round or that the breeding period is extended. It only shows there is wide variation in breeding. In June of 1970 penguins had chicks while in June of 1972 there was no breeding activity. Unpredictability in the timing of breeding is not restricted to the Galapagos penguin. Swallow-tailed gulls (Creagrus forcatus) (Snow and Snow, 1967) have also been found to have considerable variation in the time of breeding. Not only do breeding periods fluctuate from year to year but the time of breeding may vary from location to location. Figure 20 shows how in some periods penguins breeding at sites around Fernandina and Isabela were nearly synchronous ; while in other seasons penguins breeding at these same sites were not synchronous. Thus, breeding seasons can be characterized by their unpredictabil ity both in time and location. Breeding Success Breeding periods do not always end in success. There was moderate success during the June to September 1970 breeding season when some chicks fledged. In contrast, the June to September 1971 season was tremendously successful with an average of 1.5 chicks fledged per nest that was found with eggs. This may be a slight over estimate of the success of the breeding periods since nests that failed in the beginning of the egg stage were not considered. All nests failed in the Decmmber to March 1972 season, and all nests failed in the August to October 1972 season. (The failure of the 134 August to October 1972 season was confirmed by David Day, a guide on a tourist boat. hTien I left in early October, half of the nests had already failed.) Breeding success is not constant and, like the breeding period, is highly variable. Natural catastrophes do play a small role in breeding success. Many nests were covered with water due to high tides in February and March 1972 as well as in October 1972. Many nests would have failed because of the tides but the nests had already been deserted before the tides occurred. At least two chicks were forced to fledge early due to moderately high tides in February 1972. They weighed only 1200 grams and were not recaptured suggesting that they died. Nest location in other species of seabirds has been shown to effect nesting success (Coulson, 1968) . Tenaza (1971) found Adelie penguins were more successful if they nested in the center of the colony rather than on the periphery. He also found older birds nested in the center and young penguins on the outside. Perhaps older Galapagos penguins have nest sites which are less prone to natural catastrophes such as mechanical breakage of eggs or tidal flooding. Less than 5% of the eggs laid in any season were broken or eaten. The changes in breeding success from total failure to 1.5 chicks per nest can not be attributed to tidal flooding, breakage of eggs, or predation. In order to understand the variation in breeding success we must understand why nests are deserted. ECOLOGY OF GALAPAGOS PENGUINS The total life history pattern of a species has meanings in terms of its ability to survive; thus, with a knowledge of the natural history of the Galapagos penguin on hand, we can precede to ask fundamental questions about adaptations to an unpredictable environment. IVhat limits the population? IVhy are Galapagos penguins found breeding only on Isabela and Fernandina? Why are they sexually and age dimorphic? What are their strategies for survival? The next sections will attempt to understand the pen guins' problems and adaptations for life in the Galapagos. Food as the Limiting Factor on the Population The food available to a penguin is difficult to measure directly; but resources can be assessed indirectly by measuring weight. The ability of breeding adults to maintain weight and nestlings to survive and gain weight reflects the extent to which food limits the population. For example, if food were limiting the population, then differences in growth rates or weight changes would be expected between lone and two-chick nests. We might further expect the oldest chick in a two-chick nest to grow faster than a younger sibling. Such as assumption would be based on the knowledge that older chicks might be more vigorous and better 135 competitors than their younger siblings for the limited food. Several other studies of penguin chick growth have been done. Ainley and Schlatter (1972) summarized most of the previous studies. Tliey found that growth patterns for Adelie chicks in one and two chick broods are similar, but the chick in a one chick brood is heavier than a chick in a two chick brood. Like Galapagos penguins, Adelie penguins commonly hatch two chicks and then later lose a chick. In Galapagos penguins lone chick nests commonly start as 2 chick nests, but one dies. Yellow-eyed penguin chicks (Megad)q3tes antipodes) (Richdale, 1957) grow at similar rates in nests of two. Weights on all available known aged chicks (93) at Pta. Espinosa, Fernandina and Elizabeth Bay, Isabela were analyzed using a MANOVA analysis of covariance where day was the covariant. Table 14 gives the results of the analysis for one and two chick nests at the two locations for three breeding periods. Although there are significant differences between growth rates of chicks, some breeding periods show less dramatic differences than others. This is probably a result of small sample size for breeding periods. The weights of one chick in the January-March 1972 period may also bias the data since this is the only chick which died of starvation just prior to fledging. The weights of this chick are not used after day 20 for the other comparisons on growth rates of one and two chick nests because its growth pattern departs from the norm. If the data is analyzed for Pta. Espinosa, Fernandina and Comparison of growth rates of 93 Galapagos penguin chicks in one and two chick nests over the entire study from June 1971 until October 1972, using an analysis of covariance. 0 = oldest chick in a nest of two, Y = youngest chick in a two chick nest, and L = long chick. * = F< .05, ** = P<.01, and *** = P<.001. Lone chicks grow faster than oldest chicks and oldest chicks grow faster than youngest chicks COMPARISON OF GROWTH RATES OF CHICKS IN DIFFERENT SIZE NESTS Number of chicks Number of in sample______observations F Pta. Espinosa 70 6.72* Pta. Espinosa 52 Pta. Espinosa 70 Pta. Espinosa 26 Pta. Espinosa 52 Pta. Espinosa 26 Elizabeth Bay 33 104 Elizabeth Bay 30 101 Elizabeth Bay 33 104 Elizabeth Bay 17 42 Pta. Espinosa, Elizabeth Bay 174 Pta. Espinosa, Elizabeth Bay 153 Table 14. Comparison of growth rates of chicks in different size nests (continued]. Number of chicks Number of Location Chick in sample______observations Slope Pta. Espinosa, Elizabeth Bay 0 38 174 34.06 Pta. Espinosa, Elizabeth Bay L 20 68 37.05 Pta. Espinosa, Elizabeth Bay Y 35 153 31.81 Pta. Espinosa, Elizabeth Bay L 20 68 37.05 Elizabeth Bay, Isabela, ignoring breeding periods it becomes apparent that at Pta. Espinosa lone chicks grow significantly faster (P<.01) than oldest or youngest chicks. The faster growth of oldest chicks (P<.01) compared to youngest chicks further indicates food may be limited. The same results hold for Elizabeth Bay except that no significant difference in the weights of the oldest and youngest chicks in two chick nests were found. An analysis of the data for the entire study shows that lone chicks grow significantly faster (P<.01) than chicks in 2 chick nests and that oldest chicks grow significantly faster (P<.01) than the youngest chick in a nest of two. These differences in growth support the hj'pothesis that food is a limited resource for the chicks. Food may be limited for chicks but is it limited for adults? If it i^ reproduction and molting are costly activities for adults and its effect should be seen during these periods. Weights were taken on both sexes of adults and juveniles. Individual penguins were weighed and placed into one of ten different divisions recog nized in the reproductive and non-reproductive cycle. Tlie ten divisions are: 1) before the molt or before much feather loss; 2) after molting has begun (large patches of feather lost on back and belly); 3) after feathers are shed and new feathers are growing; 4) after eggs are laid; 5) the middle of the incubation periods (10 to 25 days after eggs are laid); 6) before eggs hatch (26 to 40 days); 7) one to ten days after chicks are hatched; 8) 10 to 50 days after chicks have hatched; 9) weights immediately prior to nest failure; 10) non-reproductive or non-molting period. Table 15 and Figure 23 give the results of the penguins weighed in each of these classes. There are tremendous gains and losses of weight for molting (1). If food were superabundant we might expect that at the conclusion of molting adult weight would be similar to the normal non-reproductive weight. This is not true as Figure 24 shows. However, it may be that food is sufficiently abundant that penguins can gain the lost weight easily. This does not seem to be true since molting occurs prior to egg laying (4), and as the graph indicates, both sexes have not returned to their average non- reproductive weights (10). Furthermore, unlike other species of penguins which commonly remain on land until all feathers are lost (Penney, 1967; Richdale, 1957) the Galapagos penguin may return to the water with head and neck feathers still attached. Two male penguins which were captured as they came out of the water with head and neck feathers attached were 1600 grams and 1750 grams respectively. The wide variation in the time spent on land molting (Boersma, 1974) may reflect individual differences in gathering limited resources. Autopsies of 3 adult penguins as previously mentioned, suggested that starvation after molting was the cause of death. Many penguins banded as chicks or juveniles were seen during the same study period. Four to seven months later, 31 of 99 Table 15. Weights o£ m£le, female and juvenile Galapagos penguins during each of the life cycle divisions, X = mean, X = standard error of the mean, N = sample size. Max = maximum weight, and Min = minimum weight. (See text and Figure 24 for further explanation.) WEIGHTS OF GALAPAGOS PENGUINS DURING 10 STAGES IN THE LIFE CYCLE 1 2 3 4 5 6 7 8 9 10 o'’ X 2571.7 2421.7 1727.5 1993.8 2135.0 2096.1 2142.0 2068.6 1924.0 2182.6 X 118.7 224.9 156.4 27.7 28.1 44.7 23.4 54.2 53.1 28.5 N 6 6 4 21 56 18 43 7 5 61 Max 2820.0 2050.0 2000.0 2300.0 2880.0 2600.0 2510.0 2330.0 2120.0 2670.0 Min 2000.0 1550.0 1400.0 1800,0 1780.0 1800.0 1850.0 1850.0 1800.0 1750.0 ?) X 2462.0 2020.0 1676.4 1718.7 1730.0 1660.9 1826.8 1767.8 1673.3 1869.6 X 75.0 125.1 40.4 32.2 15.1 28.7 26.9 50.2 17.6 23.4 N 5 4 11 30 45 11 31 14 3 50 Max 2620.0 2300.0 1910.0 2100.0 2000.0 1800.0 2100.0 2200.0 1700.0 2350.0 Min 2250.0 1750.0 1420.0 1380.0 1500.0 1500.0 1600.0 1450.0 1640.0 1450.0 Table 15. Weights of Galapagos penguir.'. during 10 stages in the life cycle (continued). 1 2 3 10 J. X 2655.0 2222.2 1721.4 1900.8 X 123.5 59.3 55.2 70.6 N 10 18 14 12 Max 3500.0 2700.0 2100.0 2300.0 Min 2200.0 1800.0 1350.0 1480.0 Figure 24. The mean weights of male female (o--), and juvenile (x ) Galapagos penguins during each of the life cycle divisions. Males are always heavier than females and juveniles which consist of both males and females are intermediate in weight between adult males and females. Weight is gained for the molt (1) and lost during the fasting period (2 and 3). Weight is gained during the beginning stages of incubation (4, 5) but the rate of weight gain falls prior to hatching (6) while adults lose weight before fledging [8]. The weight of adults at the time of nest failure is comparable to the weight after molting suggesting a lack of resources cause nesting failure (9). Non- reproductive penguins (10) are heavier than breeding individuals. N O iiV g n ^N i 0003 145 juveniles and chicks banded were recaptured. During this recapture period not all of the penguins would have molted but at the end of a year all individuals would have molted at least once. In the last study period, a year after banding, only 4 penguins banded as molting juveniles were seen. Such differences in recaptures after molting suggest that molting is associated with high mortality. High mortality and slow recovery of weight after the molt suggests that resources for adults and juveniles are limited at least during this period in their natural history. During reproduction the weights of males and females are always below the non-reproductive weight (Fig. 24). Weight is gained by both sexes early in the incubation period (5) but is lost late in the incubation period (6). The average weight late in the incubation period for successful breeders may be slightly higher than what is shown since penguins whose nests failed are included in this average. Thus, for successful breeders some weight may be gained throughout incubation. When chicks are hatched (7), the adults have gained some weight since late incubation (6). As the chicks grow and their demand for food increases adults lose weight. Tliis indicates that food is in limited supply for adults with chicks since they do not maintain their weight. Thus, food appears limited for chicks as well as breeding adults. Out of 302 nests which were observed during the study at least 194 failed. Reports from David Day, a guide on the Lina A, indicated that all nests failed in October 1972 which means that out of 302 nests 242 failed or 80% of the nests with eggs or chicks failed. Wliat causes nest failures? Of the few nests where adults were weighed everyday before the nest failed, the weights of adults were lower than for nests with individuals which continued to incu bate. The adult weights at the time of nest failure (9) are nearly as low as weights of penguins after molting. Such low body weights suggest that nest desertation occurs when adults are unable to gather sufficient food. If adults did not desert their nest, they would continue to lose weight and eventually die. Incubation periods can also be used to deliniate the availabil ity of food and the success of individual foraging. lŸhen adults were daily relieving each other at the nest, no nest failures were seen. Galapagos penguins at Elizabeth Bay were more likely to change daily or more frequently than daily than those penguins at Pta. Espinosa. Although nests were checked three times or more at Pta. Espinosa only one pair was found that changed more often than daily compared with at least 10 pairs at Elizabeth Bay. This dif ference can be attributed to the more successful foraging at Elizabeth Bay which will be discussed later. The average weight loss for males and females during incubation (Table 16] is slightly higher the first day on land since not all the food has been digested. For each additional day on land, it costs the penguin approximately 50 grams to maintain itself. For everyday that a Table 16, Weight loss per day in Galapagos penguins con£ined_to the nesting site during incubation. N = sample size, X = mean weight loss, and SE = standard error of the mean. IŸEIGHT LOSS IN GRAMS MALES FEMALES BOTH SEXES N X ± (SE) N X ± (SE) N r ± (SE) 1 day on land 11 67.27±(3.S9) 7 52.86±(6.66) 2 days on land 4 46.2S±(5.57) 3 days on land 1 50.0 penguin sits on eggs, it looses not only foraging time and potential energy but also stored energy. Tliis means that if a Galapagos penguin normally gains 50 grams of weight everyday, then incubation costs them 100 grams per day. Richdale (1957) states incubation is a recovery period but although incubation is less costly than feeding chicks it is still expensive. Foregoing reproduction is really the only recovery period available. As we might expect if food is scarce, no nests succeeded where one of the pair was gone frequently more than 3 days. One penguin remained on the nest for 8 days necessitating a weight gain of 400 grams to maintain its previous weight. It returned to the nest after 3 days but the nest failed subsequently when the mate did not return the next day. If food is at all scarce, large weight gains are probably unlikely to be accomplished in one day so that the individual must continue to forage a number of days while its mate is losing weight. If resources are low, the pair can be trapped in a vicious circle which can only lead to nest failure or death of the individual. Of course if food becomes more abundant so that the individual can gain weight in shorter foraging periods the nest may not fail. The high nest desertation in Galapagos penguins indicates that resources are limited. The limited food supply has been shown to effect the growth of chick molting and breeding success. Tlius, food probably limits the population by regulating the recruitment rate of reproductive individuals and of the number of surviving molting adults. Further Tests of the Hypothesis that Food Limits the Population Galapagos penguins breed only on Isabela, excluding the east side, and Fernandina. IVliy are there not penguins breeding on other islands? From the human point of view, the land habitat on the eastern side of Isabela and on other islands appears to be suitable for nesting and landing. Although there do not appear to be any differences in the land habitat which would make breeding unsuitable there are differences in the water habitat. Cold surface water temperatures coincide with the breeding distribution of the Galapagos penguin. The coldest surface water temperatures are around Isabela, exclusing the east side, and Fernandina (Harris, 1969a; Maxwell, 1974). On the eastern side of Isabela at Cabo Marshall surface water temperatures were 26°C on September 30, 1972 and 23°C at Pta. Espinosa, Fernandina on the same date. Differences like the above seem to be representative (Maxwell, 1974) of surface water temperatures. Feeding frenzies which probably indicate more abundant food are seen more frequently at low surface temperatures (Fig. 23). The coldest water of the Galapagos may be the only areas with sufficient resources to support breeding penguins. Thus, breeding distribution may bo explained by the availability of resources. It is generally accepted that the breeding season is deter mined by environmental changes (Ashmole, 1963, 1971; Orians, 1961; Lack, 1968; and Perrins, 1970). Breeding cycles in seasonally pre dictable environments tend to be fixed and seem to have evolved in relationship to food abundances (Lack, 1954; Pitelka, 1958; Richdale, 1963; Nelson, 1966; Carrick and Ingham, 1967). The Rockhopper penguins' breeding season has been roughly correlated with water temperature and latitude (IVarham, 1972). This is not surprising since surface water temperature and latitude may be closely corre lated with increased productivity and food. Simmon (1967 and 1970) shows that the Brown booby (Sula leucogaster) in tropical areas exhibits a direct response to increasing food abundances. On the savanna (Moreau, 1950) breeding has been correlated with the increases in vegetation and insects which are a result of the rainy season. Penguins which live in predictable environments such as Antarctic regions have a fixed annual cycle (Carrick and Ingham, 1967). This is not surprising since Antarctic waters are noted for their rich and dependable seasonal food supply (Knox, 1970; Raymont, 1963). IVhere the climate is not as seasonal, penguins such as the Northern blue penguin (Eudyptula minor) and the Black-footed penguin (Kinsky, 1960; Davis, 1955) have extended breeding periods with some variation from year to year. Tlie literature has depicted the water environment of tropical areas as stable, predictable and non-seasonal (Murphy, 1936; Harris, 1969a) but such conclusions were undoubtedly a result of lack of data and understanding of the water environment. The water environ ment of the Galapagos has already been portrayed as unpredictable. Recent studies have shown by sampling phytoplankton production (Owen and Zeitzschel, 1970) that the eastern, tropical Pacific has seasonal changes and the pattern of primary production and phyto plankton along the coastal Peruvian waters (Guillen et a^., 1967) are seasonal but unpredictable. Marine environments in many, if not all, tropical areas are probably unpredictable places for food (Schreiber and Ashmole, 1970). The massive die-off of sea birds in "el Nino" years along the coast of Peru (Hutchinson, 1950) indicates that the pattern of resources in these marine environments is un predictable. If breeding periods have evolved in relationship to food in predictable environments might breeding periods likewise have evolved to occur with greater food abundances in unpredictable environments? If this model is true we would expect that Galapagos penguins should breed when food is most available. We have suggested that feeding frenzies are an index to food abundances and they occur more fre quently at low surface water temperatures. Furthermore, Maxwell (1974) correlated higher primary productivity with low surface water temperatures. Thus, surface water temperatures can be used as a reflection of food abundances. If breeding distribution is a reflection of suitable resources which can be traced by lower surface water temperatures then breeding periods should also be characterized by low surface water temperatures. We might further expect that if food is more abundant at lower temperatures then breeding periods should occur when surface water temperatures are low. Thus, we would predict penguins should breed year round if the surface water temperatures are cold year round or seasonally if the surface water temperatures are seasonally low. IVhen mean surface water temperatures and the breeding period of Galapagos penguins are plotted for each month (Figure 25) the result indicates that breeding occurs when surface water temperatures are below 24°C. The results of a t test (Figure 25) show that surface water temperatures are significantly colder (P<.001) when Galapagos penguin breed than when they do not breed. The mean surface water temperature during breeding is 21.67°C compared to a mean surface water temperature of 25.18°C when penguins are not breeding. Breeding either does not commence or ceases when surface water temperatures rise. As predicted breeding periods are associated with low surface water temperatures. Productivity values further support the hypothesis that food is the cause of breeding and breed's failure. Maxwell (1974) found chlorphyll 2, values of 3.37 ml/ra^ at Elizabeth Bay in July 1972 just prior to breeding. Chlorophyll & values were all below 1.25 ml/m^ when the penguins were not breeding from February to July 1972. Figure 25. Mean surface water temperatures at Pta. Espinosa, Fernandina for the first and second half of each month. The breeding period for Galapagos penguins is indicated. Dots represent eggs present, ___ chicks and /// no breeding activity. The water temperature was signi ficantly lower during breeding (P<.001) than when the penguins were not breeding. Mean surface water Standard Sample temperature deviation size t statistic breeding 21.67 1.25 31 5.93*** not breeding 25.18 1.93 I53 28 27° o o 2 0°- o 0 0 o o 2 5 ° o o I 2 4 ° Oo 23 ° o ^ oo QO O i 22° 21° § 2(f- s is to OÇOOM CO ««o 000 § JJAS J ASO NDJ FMAMJ J A SON 70 71 72 MONTHS The chlorophyll a value at Pta. Espinosa in late January prior to nest failures in February 1972 was so low it could not be detected. The consistency of food abundances should be linked to breeding success. Surface water temperatures above 22.5°C preceded the total nesting failure of two breeding periods (Figure 25). During the two breeding periods which were successful surface water tempera tures were below 22.S°C. The most successful period between June and September 1971 had surface water temperatures consistently below 22°C. The chlorophyll & value for September 1971 at Pta. Espinosa was 2.37 ml/ra^ which shows high productivity. Feeding frenzies were the most common from June to September 1971 when on the average 1.5 chicks were fledged per nest. Therefore, breeding success as well as breeding periods are dependent on abundant food The growth rates of chicks is consistent with the view that food was less abundant during breeding periods when surface water temperatures were higher. Table 17 shows that there were signifi cant differences (P<.01) in the growth of oldest chicks at Pta. Espinosa, Fernandina during the breeding periods of July to Septem ber 1971 and January to March 1972. There were also significant differences in the growth of the youngest chicks in the breeding periods of July to September 1971 and July to September 1972. The growth of chicks in the most successful breeding season of 1971 were always faster than for comparable chicks in other periods. Comparisons of growth rates of 93 Galapagos penguin chicks in different breeding seasons. 0 = oldest chick in nest of two, Y = youngest chick in nest of two, and L = lone chick. * = P<.05, ** = P<.01 and *** = P<.001. The rate of growth was significantly slower during breeding periods (P<.001 and P<.01) when warm surface water temperatures were recorded. COMPARISON OF GROWTH RATES OF CHICKS IN DIFFERENT BREEDING SEASONS Nuiriber of Number of chicks in iple Location Chick vations Season £ 3 Pta. Espinosa 0 42 Jul-Sep 1971 32.47 43.53*** 2 Pta. Espinosa 0 44 Jan-Mar 1972 22.03 3 Pta. Espinosa Y 15 Jul-Sep 1971 23.34 .00 2 Pta. Espinosa Y 13 Jan-Mar 1972 24.28 3 Pta. Espinosa L 14 Jul-Sep 1971 29.24 .43 1 Pta. Espinosa L 6 Jul-Mar 1972 24.56 23 Elizabeth Bay 0 37 Jul-Sep 1971 31.23 .79 9 Elizabeth Bay 0 19 Jul-Sep 1972 25.08 21 Elizabeth Bay Y 33 Jul-Sep 1971 23.65 7.15** 9 Elizabeth Bay Y 19 Jul-Sep 1972 13.26 13 Elizabeth Bay L 4 Jul-Sep 1971 26.59 .02 4 Elizabeth Bay L 7 Jul-Sep 1972 25.15 The breeding distribution, breeding period and breeding success have all been shown to be dependent on increased food abundances. Furthermore, increased food abundances are reflected by feeding frenzies and low surface water temperatures which are more produc- A1though there is no data to test the following model it is interesting to speculate on the cause of fluctuation in food abun dances reflected by changes in surface water temperatures and the number of feeding frenzies. Unfortunately, little research has been done on the temperatures of the currents (Sibert, 1971) which bathe the Galapagos, but some general ranges can be compiled. Oceano graphers disagree on the surface temperatures of currents (Malone, 1968; Wyrtki, 1963). But by using IVyrtki's (1965) current patterns. Love's (1972) charts, and temperature data from this study, a rough index to currents and their surface temperatures can be compiled. The coldest surface water temperatures (22°C or below) are a result of the Cromwell Current upwelling. The Cromwell Current with its nutrients is undoubtedly the current which supports the penguin population. None of the other currents in the Galapagos are as cold. Surface water temperatures of 22°C to 24°C may be a result of a heavy influence of the Peru Oceanic and the South Equatorial Currents, while surface water temperatures above 25°C correspond to the North Equatorial Current or the Equatorial Counter Current, depending on the season. If the Cromwell Current is shifted to the North or South of Fernandina and Isabela, then it would not upwe11 and surface water temperatures should be higher. A shift in the Cromwell Current, we might predict, would mean reduced resources and either a post ponement of breeding or breeding failure. The extensive breeding failure and lack of breeding when surface water temperatures are high does suggest a change in the Cromwell Current. Food abundances are not similar between locations. An analysis of covariance shows that chicks at Elizabeth Bay grow significantly faster (P<.001) than chicks at Pta. Espinosa (Table 18). Only lone chicks grow at similar rates. The differences in the slope of growth between the breeding areas is a reflection of the discrepancy in food abundance between Elizabeth Bay and Pta. Espinosa. Chicks raised at Elizabeth Bay grow faster demonstrating that resources in Elizabeth Bay must be either more abundant or more efficiently exploited "han resources at Pta. Espinosa. What are the possible explanations for these food differences? Currents which bring food to Pta. Espinosa and Elizabeth Bay may be different. If currents were different, they would be likely to have dissimilar resources associated with them. This explanation is highly suspect since the cold water temperatures are associated with the Cromwell Current at both locales. Furthermore, currents normally have a wide influence, and it is unlikely that two separate currents would consistently effect places less than 50 miles apart. Comparison of growth rates of 93 Galapagos penguin chicks at different locations in the same breeding season. 0 = oldest chick in a nest of two, Y = youngest chick in a nest of two, and L = lone chick. * = P<.05, ** = P<.01, and *** = P<.001. Chicks at Elizabeth Bay grow significantly faster than chicks at Pts. Espinosa. COMPARISON OF GROIVTH RATES OF CHICKS AT DIFFERENT LOCATIONS Number of Number of in sample Location Chick vations; Season Slope F 23 Elizabeth Bay 0 85 Jul-Sep 1971 34.99 7.16** 3 Pta. Espinosa 0 42 Jul-Sep 1971 32.47 21 Elizabeth Bay Y 82 Jul-Sep 1971 33J# 7.08** 3 Pta. Espinosa Y 39 Jul-Sep 1971 30.54 13 Elizabeth Bay L 35 Jul-Sep 1971 1.00 2 Pta. Espinosa L 20 Jul-Sep 1971 3& J # 57 Elizabeth Bay all 200 Jul-Sep 1971 34.47 6.95** 8 Pta. Espinosa all 102 Jul-Sep 1971 32.48 80 Elizabeth Bay all 247 Entire Study 34.33 48.84**' 13 Pta. Espinosa all 164 Entire Study 28.84 Perhaps the patterning of resources or the exploitation of the resources is different. Foraging vactics are likely to be the same so that to account for food differences, resources at Elizabeth Bay must be more abundant or be patterned differently to allow greater foraging success. Similar resources could be patterned differently to allow greater hunting success, but it is not only unlikely to occur, but also resources (i.e. fish) in such a pattern would be selected against. Thus, if such a pattern ever occurred, it would quickly fade away. An alternate hypothesis is that penguins may forage more efficiently at Elizabeth Bay which would result in faster growth rates of chicks with similar resources available at Pta. Espinosa and Elizabeth Bay. Horn (1968) presents a model which shows that if a species is exploiting food which is unpredictable in time and space, energy is saved by nesting in the center of the food source. Elizabeth Bay is unique since it has three islands located in the Bay. Penguins which nest on the islands are nesting in the center of the food resource. There must be some critical foraging dis tance beyond which it is not profitable to forage and return to these nests. Birds which live at A such as the islands in Elizabeth Bay (Fig. 26) can cover twice as much area and be r or less distance from the nest. Penguins nesting at B can cover only half of the foraging area available to penguins nesting at A and be only r distance away from the nest. Chicks which are raised at A are Figure 26. A model illustrating the foraging area available to Galapagos penguins nesting at point A and B and swimming a distance r or less from the nest. Point A represents an island while B represents a point on the coastline. Penguins nesting at point A have twice as much foraging area r distance from the nest than penguins nesting at point B. co/15 likely to grow faster than chicks raiseu at C since tlielr parents have twice the area available for hunting at r foraging distance. Point B is representative of penguins which nest along the coast of Isabela and Fernandina. In order for the model to be applicable, however, there should be equal numbers of penguins exploiting the foraging area at both places. Elizabeth Bay is much more densely populated so that the added advantage of increased foraging area at r distance would only apply if the clumped resources (i.e. fish) are so abundant that once the clump is found, it is for all practical purposes unlimited. Observations on feeding frenzies and groups of penguins seen for aging together (groups of over 200 have been seen on several occasions) indicate that food is probably highly clumped and each clump is probably unlimited from a predator's viewpoint. Increased numbers of penguins then do not effect the food available for each individual. Another explanation is that the pattern of resources is differ ent at Elizabeth Bay; in other words, Elizabeth Bay is much richer than Pta. Espinosa. It is possible that the Cromwell Current influences the Elizabeth Bay area to a greater extent and provides it with more nutrients and resources. It may be that more nutrients are forced to the surface over a greater area because of the shallow ness of the Bay. Or, water may be equally rich but Elizabeth Bay may act as a funnel and concentrates the fish on which the penguins feed. Unfortunately, data are not available to determine which, if any, of these hypotheses are valid. Tests for Chlorophyll a (Maxwell, 1974) at Elizabeth Bay have shown high productivity, and fishermen have said that red tides are more common in Elizabeth Bay than elsewhere in the islands. I have seen three extensive red tides at Elizabeth Bay and only one very minor one at Pta. Espinosa. Since I have spent much more time at Pta. Espinosa this suggests that Elizabeth Bay is generally more productive. In conclusion, chicks at Elizabeth Bay grow faster than Pta. Espinosa (Table 18) probably because resources are more abundant and greater foraging area is available close to the nest. Advantages of Sexual and Age Dimorphism Why a species should be sexually dimorphic has been pondered by Darwin and is still of considerable interest because it indicates the sexes have different constraints in maximizing fitness (Fisher, 1930). Many species have unbalanced sex ratios where males out number females (Lack, 1954; French, 1959, Bellrose ejt , 1961). In these species there is presumably strenuous competition for females and thus, char, eristics which would enhance the probabil ity of a male contributing genetically to the population are selected. Sexual dimorphism is the result of such a selection pressure. Perhaps male penguins, which are heavier and larger than females, have the greatest likelihood of acquiring and keeping mates. If this assumption is valid, then mated male Galapagos penguins would be expected to be heavier and larger than nonmated males. Table 19 shows that there is no significant difference (P<.05) between males which are mated and those not mated, except in weight. Nonmated males are heavier than males which were mated, but weight was taken when the penguins were not breeding or molting. The lower weight of mated males which was taken after reproduction reflects the energy cost of reproduction, since mated males were weighed usually after breeding. Nonmated males are not different from mated males in body dimensions which means there is no differ ence in sexual dimorphism between the groups. Females must not select males for their size or weight, which indicates that intra- sexual selection (Fisher, 1930) is currently not contributing to the development of sexual dimorphism. There must be other factors such as experience and age which the females are selecting over the sexual dimorphism of males per se. Trivers (1972) points out that where parental investment in the young is equal, as it is in Galapagos penguins, sexual selection should operate similarly on the sexes. Sexual dimorphism in Galapagos penguins can not be explained by the ability to acquire mates, although it may have some influence. It is difficult to conclude if sexual or natural selection is the cause of sexual dimorphism (Selander, 1972). If sexual differences are not a result of increasing male fitness by raising the A comparison of body parts in mated and unmated male and female Galapagos penguins. * = P<.05, ** = P<.fl, and *** = P<.001. Mated and unmated males differ only in weight. The higher weight of mated males reflects the cost of reproduction. The longer and narrower bills of unmated females may be characteristic of younger SEXUAL DIMORPHISM BETWEEN MATED AND UNMATED GALAPAGOS PENGUINS MATED cf to UNMATED cf M/N UM/N X±[SE] mm M X±(SE) mm UM F Bill Length 144 34 58.26+(.23) 58.92±(.48) 1.571 Bill Width 144 34 19.83±(.24) 19.92±(.20) .034 Flipper Length 107 22 117.95±(.69) 118.54±(1.22) .135 Toe Nail 104 29 16.35±(.12) 16.63±(.21) 1.162 Weight 41 20 213.93±(3.23) 227.15±(5.21) 5.052* MATED 9 to UmiATED Q M/N UM/N X±(SE) mm M X±(SE) mm UM F Bill Length 133 27 53.75±(,38) 55.17±(.22) 7.247** Bill Width 134 27 16.76±(.08) 16.26±(.18) 7.088** Flipper Length 96 23 115.45±(.64) 112.44±(1.44) .465 Toe Nail 91 23 15.18±(.10) 14.76±(.24) 3.314 Weight 41 10 186.95±(2.54) 187.30+(5.62) .004 probability of mating, then perhaps sexual differences are a result of selection for adaptations in feeding. Numerous species have been found which are sexually dimorphic with the sexes either exploiting different resources or exploiting resources in a different manner (Cade, 1960; Kilham, 1965; Selander, 1966; Storer, 1966; Newton, 1967; Schoener, 1968; Jackson, 1970). Perhaps penguins also are dimorphic because males and females have been selected which exploit different food resources. Certainly the advantage in increasing available food by enlarging the foraging area and reducing competition for food where food is limited by spacing can be seen. It is easy to understand how sexual dimorphism could arise in feeding as a by-product of where the sexes spend their time. For example, male vireos (Vireo olivaceus) spend time in the higher portions of trees (Williamson, 1971) where they are singing. It is not surprising that males would feed near singing points and females forage in the same plane as the nest. Males that feed like females would waste energy getting to their singing posts, be more likely to lose their territory, and might hinder reproductive success by competing with the female for food. Pen guins, however, do not seem to have a behavior trait linked with their foraging area. Sexes forage together in the same area and at the same time. Consequently, there is no obvious mechanism which would result in a partitionment of the foraging area. The Galapagos penguin may, as previously mentioned, have two different types of resource patterns to exploit. One is patchy and superabundant and the other patchy but not abundant. During a feeding frenzy when the penguins forage in groups the water boils with fish, and many species of birds cluster to feed on them. When resources are superabundant but unpredictable in time and space Ci.e. when the penguins are feeding on schools of fish), the prob lem is in locating the food. Once the food is located, there appears to be a superabundance of fish with all kinds of birds for aging in the same area. Tlie selection or benefit of sexual dimor phism in feeding for Galapagos penguins appears undefendable in this situation. When the small schooling fish, such as mullet and sardines do not seem to be available, the penguins forage individually, in pairs, or in very small groups. Most commonly foraging is in pairs with a male and a female present. Tliey seem to scan the bottom and the coastline looking for food which is presumably patchy and never very abundant. Reef fish which tend to space out (Smith and Tyler, 1972) and small groups of fish are probably encountered. If males and females exploited these resources differently, competition would be reduced and the niche would be expanded to support more individuals. Western Grebs (Aecpmophorus occidentalis) are sexually dimorphic (Rand, 1952; Palmer, 1962) with males having bigger and heavier bills. Selander (1966) has suggested that there may be sexual differences in the resources they exploit. Ainley (1970) found male and female Adelie penguins which are sexually dimorphic do feed on different sizes of krill. Males took larger prey items than females. Mien resources are patchy and not abundant selection for male and female Galapagos penguins to take different size prey items might be expected. Sexual dimorphism in Galapagos penguins, thus, may be a result of selection for males and females which take different size prey items. If sexual dimor phism is a result of males and females taking different prey items we might expect bill differences but not necessarily size differ ences. Sexual dimorphism in size could suggest that males and females are foraging at different depth. Since the sexes forage together in pairs when food is limited, foraging differences in depth may be beneficial. As we have pointed out, such foraging differences appear unlikely in feeding frenzies. Although foraging differences by depth are possible, there are not data available to test whether foraging differences in space actually occur. However, differences in prey selection between male and female Galapagos penguins doo< seem likely since other penguins which are sexually dimorphic in bill size take different size prey. Like other birds many species of penguins have a state of immature plumage. Juvenile penguins can be recognized by a lack of adult marking and more drabb1y colored heads. For example, juvenile Emperor penguins (Aptenodytes forsteri) have a semicircular patch on the side of the neck of white or light yellow while in the adults it is orange or dark yellow (Budd, 1968). In the genus Spheniscus, the immatures of all species have some differences from adults in markings, but none are reported to have different body coloration (Alexander, 1963; Johnson, 1965). The Galapagos penguin appears to be the only species of penguin where the immature differs from the adult in coloration. It does not seem unreasonable for juveniles to have different facial markings. Murphy (1936) speculated that facial patterns were important in species recognition. If this is true, then perhaps the immature head plumage allows adults to distinguish immatures, resulting in reduced aggressive and pair bonding acti vity toward the juvenile. Both adults and juvenile penguins would spend less time in extraneous behavior and thus, save energy. From a purely natural selection argument, one can see how the ability to distinguish juveniles by a different plumage pattern could be advantageous in terms of survival for both adults and juvenile penguins. If this assumption is true, why are there three species of penguins which have similar plumage as juveniles and adults? It is hard to understand why these species would not be benefited by adult and juvenile differentiation. There are two major explana tions to be considered. First, perhaps these three species breed at a younger age than other species, so that any penguin after fledging is a possible mate. In this instance, there would be no advantage for adults behaving differently toward juveniles. However, this explanation seems unlikely since most seabirds do not breed in the first year (Ashmole, 1963; Wynne-Edwards, 1955). Another pos sible explanation is that the adult coloration is advantageous and any deviation in plumage of juveniles of these particular species would hamper juvenile survival. The Chin-strap penguin (Pygoscelis antarctica) may fit into this category since it has an intermediate plumage between the Adelie penguin and Gentoo penguin (Murphy, 1936). Any distinction in the juvenile chin-strap plumage from the adult’s might hamper species recognition. None of these explanations are conclusive, but it would appear possible that these three species of penguins have plumage constraints that the other species are lacking. Perhaps minor changes in plumage for the chin-strap. Northern blue or white flipper (Eudyptula albosignata) penguin results in lower survival. Another unusual problem which needs explanation is if in eleven species of penguins juveniles differ from adults in head plumage, why should the juvenile Galapagos penguin also differ in back color as well as head plumage? If most species of penguins can distinguish juveniles from adults by facial plumage, why should the Galapagos penguin need an added cue? The plumage of immature Rockhopper penguins may be duller than adults (Alexander, 1963) but Warham (1972) did not mention this fact. Juvenile Black-footed, Humboldt and Galapagos penguins all lack body markings, but only the Galapagos penguin has a different colored back. The Galapagos penguin is more sedentary than any other species. They do not migrate and unlike other species the adults and juveniles are regularly seen at breeding locations throughout the year. The added identification cue of the gray back may function in reducing aggressive and pair bonding activity toward juvenile Galapagos penguins which unlike other species are constantly associated with adults even during the breeding period. Maximizing Fitness and Minimizing the Cost of Reproduction In an environment where resources are at sometime limited, individuals will be selected which survive periods of dearth and leave offspring. Under such selective pressures various strategies and behaviors become evident. In this section we will examine some of the selection advantages of the Galapagos penguin's natural history. Molting. If food is superabundant or equally available we would expect molting to occur at anytime. Since every species of penguin has a distinct molting period, food is probably not always superabundant or equally available. If food is limited we might expect molting to correspond to an increase in available food. Otherwise, it would not be possible for the individual to gather more food without dramatically increasing foraging time or effi ciency of foraging. Their foraging time can not be readily expanded since they do not forage at night. Furthermore, it is unlikely that the individual can become more efficient in capturing food. Thus, in order to gain weight, food must be more abundant. In seven species of penguins (Garrick and Ingham, 1967) molting coincides with the autumn plankton bloom (Hart, 1942). The Adelie penguin breeds when phytoplankton is at its peak (Hart, 1942) and molts before its decrease. Species of subantarctic penguin (Garrick and Ingham, 1967) have molting periods which coincide with autumn plankton blooms and breeding periods which occur when phyto plankton populations are at their highest (Hart, 1942). Other species of penguins such as the Black-footed penguin breed in early spring and in early fall (Rand, 1960) when plankton is at its highest (Raymont, 1963). In seasonal and predictable environments it is not surprising that breeding seasons would be clearly defined, and that penguins would molt after breeding on a second peak of productivity. Thus, Adelie chicks are present when food is most abundant (i.e. at the peak of plankton populations). The increase in weight for the molt and the fledging of chicks for Adelie as well as for other penguins occurs at the next highest plankton peak. Fledging chicks at times of higher than normal food resources should increase their chances of survival. Furthermore, fasting during decreased resources will keep adults from competing with their young for reduced resources while they are learning foraging techniques. Nonbreeding adult penguins as we might expect. molt before breeding adults which are feeding chicks. In season ally predictable environments molting after breeding results in the maximum utilization of the food resource. If molting is before breeding, then penguins will not be able to exploit the resources during fasting. The length of the resources available for breeding will be reduced and fledging will occur when resources are dimin ished. Since most species of penguins do inhabit a seasonally predictable environment, it is logical that penguins molt after breeding (Richdale, 1957; Rand, 1960; Kinsky, 1960; Sladen, 1958; Stonehouse, 1953; Garrick and Ingham, 1967). The Galapagos penguin molts before breeding. Since the patterning of food is unpredictable and sometimes chicks can not be fledged before food fails, molting must occur before breeding. If Galapagos penguins molt after breeding, not only chicks, but also many adults would die should the food supply dwindle because of shifting currents. Galapagos penguins which molted after breeding would be quickly lost from the population. The strategy of molting before breeding results in only adults which successfully molt taking a reproductive risk. Non-reproductive penguins molt during the breeding period so although they are prevented from breeding during that breeding period they can molt when resources are abundant and are not lost from the population. With experience and age non-reproductive individuals may molt early and be able to Molting may be one w a y that penguins can assess whether it is advantageous to breed. If food is still available after molting, it may be that the Cromwell Current is likely to remain. If food is not abundant after molting, then breeding should be delayed. If Galapagos penguins bred immediately when the waters first become cold and productive, individuals would have to wait until after breeding to molt which would lead to massive deaths if the current shifted. Therefore, the molting of the Galapagos penguin is adapted to the unpredictable food supply of the Galapagos waters. Molting as a prelude to breeding may also be advantageous in conserving energy since the plumage will be in peak condition dur ing breeding. Pair bonds, Seabirds tend to keep the same mate from breeding season to breeding season (Richdale, 1957; Lack, 1968), Yellow eyed penguins which breed with the same mate were more likely to succeed than those of comparable ages which switched mates. Royal penguins which changed mates often failed to breed (Garrick and Ingham, 1970), These data indicate that long term pair bonds increase reproductive success allowing for a more rapid onset of breeding than if individuals had to find a new mate which was ready to breed. In Galapagos penguins long term pair bonds may decrease the energy expended in courtship as indicated by the 4 observations of reduced time spent in the nest site before egg laying. Along with other cues males and females stimulate each other to be in reproductive condition at approximately the same time (Lehrman, 1965). Mated pairs of Galapagos penguins often molt together in their nesting site which in an unpredictable environment may be beneficial in synchronizing reproduction in the pair. In an un predictable environment, however, we might expect to find a greater number of mate switches because there is no fixed cycle. The benefits of a long term bond may be decreased since Galapagos penguins are not seasonal breeders and both mates are not always likely to be ready to breed at the same period depending on when they molt. Other species of penguins do seem to switch mates less frequently (Warham, 1963; Richdale, 1957). Synchrony. Although responses to food can explain the gross patterns of reproduction and the differences in synchrony between locales, why are some groups of Galapagos penguins slightly differ ent in their reproductive timing than others at the same location? Part of the answer may be that individuals with long term pair bonds tend to be ready to breed faster than others. This does not explain why penguins nesting on the large island in Elizabeth Bay had newly hatched chicks in September 1972, while the penguins on the medium island had only eggs. This same pattern was apparent in all breeding periods. At Pta. Espinosa the largest group of penguins always bred ahead of penguins at other nearby locales. This asynchrony can not be explained readily by the length of the pair bond or by the food supply. The early breeding of larger groups may be a result of social interaction and facilitation with in the group, Tlie Darling effect (1938) or social stimulation has been used to explain why large seabird colonies are more synchronous than smaller colonies. Individuals which are stimulated by the breeding activities of others may be able to respond faster to changes in the environment (Crook, 1964) and breed more rapidly. Thus, the larger groups of penguins at the large island in Elizabeth Bay and at Pta. Espinosa may breed earlier because of social stimu lation. For some species, earlier breeders have been shoim to be more successful than late breeders (Nettleship, 1971; Brown, 1967) which would genetically reinforce responses to other individuals. It has been suggested (Ashmole, 1962; Harris, 1969b) that synchrony is caused by predators. Since the Galapagos penguin has few predators, it does not seem likely that predation pressure has caused synchrony. Synchrony of breeding in the Galapagos penguin can be understood as a response to resources and other individuals. Egg shape. Virtually no research on the significance of egg • shape has been done. But it is well known that egg shape is characteristic for a species and that egg shape is dependent on the need for passage down the oviduct (Lack, 1968). First eggs of Adelie penguins are slightly larger than second eggs (Yeates, 1968). Penguins of the genus Eudyptes in contrast always lay larger second eggs (Warham, 1973). Galapagos penguins, using Reid's (.1965) formula for calculating volume, lay larger second eggs. Warham (1973) reported that it was the larger second eggs which he be lieves produced the one chick penguins of the genus Eudyptes fledge. It is the second eggs of Galapagos penguins which are more likely to be addled. This is contrary to what we would expect if addled eggs had an equal likelihood of being first or second eggs. Addled eggs then, do not occur by chance as first or second eggs. One explanation of why addled eggs are second eggs is second eggs may not be fertilized, however, why second eggs would be less likely to be fertile is not apparent. Birds can store sperm for long periods of time. Elder and Weller (1954) found domestic mallards can lay fertile eggs after being isolated from the drake for 30 days. It is unlikely that Galapagos penguins can not store sperm for a few days. In the domestic chicken fertility is highest two or three days after copulation (Sturkie, 1954). Perhaps egg shape is the critical factor and the wider egg decreases the chance of embryo survival. This seems unlikely since if egg shape were the critical factor, natural selection would have selected against the shape of the second egg, and one would expect the shape of the first egg to be produced. A third possibility is that Galapagos penguins make a choice after the two eggs are laid to incubate only one egg, which is always the first egg. There is no advantage to this strategy since the second egg would be a waste of energy. Furthermore, I have seen adults incubate only one egg, but it has always occurred well into the incubation stage, llow could the adult distinguish between first and second eggs and only incubate the first? Although this is possible, it seems improbable. Miy should a penguin neglect one egg when it costs no more to incubate one than two? Vdiy shouldn't both eggs be incubated and a decision made to raise one or two chicks after hatching? Another alternative is that the food resources of the adult female are different and the second egg is not as rich in nutrients and so has greater likelihood of dying. Mien the female makes and lays the first egg in approximately 25 hours [Taylor, 1971), she has the benefit of stored nutrients and proteins she has accumulated over some period of time. After laying the first egg, the protein and nutrient reserve is decreased. If proteins and nutrients are in limited supply, the second egg may be slightly deficient resulting in decreased viability. If eggs are relatively inexpensive to produce and nutrients and proteins are not limited, this explanation is almost as unlikely as the rest. However, half of the female Galapagos penguins did forage after laying the first egg suggesting, unlike other species of penguins which do not forage during egg laying, that nutrients and protein may be limited. Incubation. There are two ways to minimize the cost of incubation. One way is to shorten the incubation period. Lack (1968) shows there is a relationship between the length of the incubation period and the fledging period. Reducing the incubation period also reduces the fledging period. Faster growth of chicks is the result of reducing the incubation period. Therefore, once slow growth proves to be advantageous, which it is for Galapagos penguins which at times are resource limited, slow growth in the chick stage must be accompanied by slow growth in the egg stage. Physiologically, it appears impossible to switch from fast develop ment of the chick in the egg to slow development and growth of the chick to fledging. Tlie incubation period of the Galapagos penguin has probably evolved to insure the shortest incubation period which still allows slow growth of young. Since there are constraints on the length of the incubation period, the cost of incubation can not be reduced by shortening incubation. A second way to reduce the cost of incubation is by dividing the cost equally between the sexes, then the cost per individual can be minimized. Shared incubation therefore, is an adaptation to minimize the cost of reproduction. Both male and female Galapagos penguins incubate the eggs and share the cost. Asynchrony in hatching. In all species of penguins except the Galapagos penguin, incubation begins after the second egg is layed (Warham, 1973; Sladen, 1958; Richdale, 1957). Hatching of chicks occurs nearly synchronously (Warham, 1963, 1958; Richdale, 1957). Although eggs are in general layed 3 or 4 days apart (Warham, 1963, in press) young hatch within 24 hours of each other. The Galapagos penguin begins incubation immediately after the first egg is layed and chicks hatch consequently 2 to 4 days apart. In an environment where resources are unpredictable asynchronous hatching allows two chances to raise a young. If food is not available the first chick hatched may starve to death but the second chick, hatching later, may survive if food becomes more abundant. If food is only marginal the first chick may survive while its younger sibling dies. If food is adequate both may survive. If food was not available and Galapagos penguin chicks hatched synchronously like other penguin chicks both would die. In an environment which is predictable hatching can evolve to the peak food abundance. In an unpredictable environment, the strategy evolved must allow greater flexibility. A clutch size of two. Clutch size for nidicuolous species as Lack's (1954) hypothesis states, has been evolved through natural selection "to correspond with the brood-size from which, on the average, most young survive...the limit normally being set by the amount of food which the parents can collect for their brood." To clarify what "on the average'- means the hypothesis can be restated: Clutch size or the number of eggs laid for a nidicuolous species has evolved through natural selection to be equal to the maximum number of young it is sometimes possible to raise. Cody (1966) noted exceptions where instead of maximizing r, adults attempt to increase K- In either, adults are maximizing the number of sur viving young produced. Clutch size in the Galapagos penguin can be examined utilizing this hyi^othesis, but before proceeding. replacement eggs should be mentioned. Replacement eggs and clutches are an adaptation to maximize reproductive output. If birds did not replace lost eggs, then whenever eggs perish, the genetic contribution is proportionately lowered. By replacing lost eggs a genetic contribution can still be made, if the succeeding repro ductive effort is successful. Replacement eggs and clutches will be favored when the replacement of the eggs would be likely to increase the adult's overall genetic contribution (i.e. the risk and cost of further reproduction must not decrease the future reproductive contribution any more than normal reproduction when mortality is considered). The high degree of breeding synchrony may indicate that the chances of breeding success and individual genetic contribution in an unpredictable environment are greatest for those penguins which breed almost immediately once food becomes available. This means that the longer the penguin waits to breed, the lower the probab ility of reproductive success. If this is true, eggs or chicks should never be replaced unless they are lost almost immediately. Observations on replacement eggs support this conclusion. The laying of replacement eggs does demonstrate it is possible for Galapagos penguins to lay 3 eggs. I have seen adults covering 3 eggs and they seem to be able to incubate the eggs properly. Why then, is the clutch size of the Galapagos penguin two instead of three? If three eggs were laid the adults would have a chance of raising three instead of two yotmg. Fledging three young in stead of two might be expected to increase individual fitness. It has been shown that at both Elizabeth Bay and Pta. Espinosa lone chicks grow faster than chicks in two chick nests. Furthermore, in nests of two, older chicks grow faster than younger chicks suggest ing food may be limiting. It follows that clutches of three would grow even slower chan clutches of two. Such slow growth probably would make it highly unlikely for adults to ever fledge three chicks. Adults show weight losses before chicks fledge and thus, the added cost of a third chick could jeopardize adult survival and reduce fitness. A third chick may be so costly that there is high selection against adults who attempt this strategy with the result being death to the adult or the entire brood. It seems logical that in an environment where food is limited for chicks, an added chick would reduce the fitness of the other chicks and consequently, decrease adult fitness. The unlikelihood of fledging three chicks coupled with the probable loss of fitness must be great enough to make raising three chicks nearly profitless. Perhaps it is not the cost of chicks which makes laying three eggs unprofitable but the cost of laying a third egg. Lack (1968) suggests that the long interval between eggs reflects "that food is sparse for the female around the time of laying, but may be comparatively easy to obtain for tiny young." Lack (1968) is contending that the interval between eggs is some indication of the cost of eggs. If eggs are costly, then we would expect that females would lay one egg, leave the nest to forage and then return in three days to lay her second egg. Since female Galapagos penguins only leave for one day or not at all in 12 out of 13 nests during egg laying, eggs must not be very costly. It is possible that nutrients and not energy might be limiting. Regardless, the egg laying interval is probably more likely to reflect physiological rather than cost constraints. Female domestic chickens require approximately 52.8 grams of food per egg (Clayton, personal communication). Fisher (1972) indicates the cost for many birds is somewhat higher. Galapagos penguin eggs are not much larger than a hen egg. Consequently, a generous estimate of the cost of an egg is 100 grams for the 60 gram egg. During incubation adults lose 50 grams per day sitting on eggs. They replace this loss and gain weight in the beginning stages of incubation suggesting the cost of an egg is small. Clutch size is more likely to be limited by the number of chicks possible to rear than the food available during egg production. The question of why lay two eggs and not just one can be readily understood. If an adult's overall genetic contribution can be increased by raising two chicks instead of one, this trait should be selected. In 1970 and 1971, two chicks were fledged in many nests, indicating that in some breeding periods it is possible to fledge two chicks. Because the cost of laying an egg is inexpensive, the penguin's maximum genetic contribution (increased fitness) can be made without greatly increasing reproductive effort. Thus, clutch size in Galapagos penguins seems to be adapted to provide the largest number of young that the adult can sometimes fledge without sufficiently increasing the chances of mortality which would hamper future reproduction. If a clutch size of two is adaptive, why don't all breeding adults raise two chicks? In breeding seasons where food becomes a limiting factor, adults may gain maximum fitness by only raising one chick. If two chicks were attempted, both chicks might starve, or their slower growth rate would lengthen the fledging period increasing the probability of a change in resource availability leading to death. Even if one of the chicks in a two chick nest died half v/ay through the chick period, the growth period of the remaining chick might be lengthened sufficiently so that fledging could not occur before the pattern of resources changed. If chicks from a two chick nest did fledge, in a marginal food year they would probably be lighter than lone chicks at fledging and have a higher mortality. Even in good years, it is more costly to rear two chicks at Pta. Espinosa than at Elizabeth Bay where they grow faster. Raising two chicks in certain breeding periods conceivably may have a lower fitness value than rearing one chick. Raising two chicks will only be advantageous if chicks from two chick nests have a combined better chance of survival than a chick from a lone chick nest. Furthermore, this value must be high enough to compen sate for the added cost of raising two chicks. Adults which raise two chicks must be more fit than those that raise one. In seasons where food is abundant, an increased fitness for adults which raise two chicks would be more likely than in seasons with marginal food. A mechanism which would allow adults to raise two or one young depending on the fitness value for lone or two chick nests in a particular period might be expected to arise. Reproductive cost model. Food availability determines the time of breeding and breeding success. Based on the knowledge that breeding is a cost to the individual, a model of reproductive cost can be developed. It is common knowledge that organisms are more likely to desert young or eggs early in the reproductive cycle rather than later (Southern, 1959). Early desertation is to be expected since as the reproductive cycle progresses the adults have invested more energy and should be less likely to terminate their investment. Adults should further be expected to take greater and greater risks as their young reach fledging, since their investment or reproductive risk is maximized and the probability of increased fitness is also maximized. Figure 27 represents a model of parental investment for seabirds and depicts that adults should desert eggs before chicks and young chicks before older chicks. Incubation is a cost but feeding chicks according to adult weight loss (Table 15) is more costly. This model (Fig. 27) illustrates that desertation Figure 27. A model of the energy investment of male and female Galapagos penguins during reproduction. As pairs invest more energy and get closer to the fitness payoff, desertation of nests should be less likely. CUMULATIVE PARENTAL INVESTMENT B Y S E X (UNITS OF ENERGY) m > should be more likely to occur in the egg stage or early after hatching before the reproductive cost becomes high, Desertation immediately prior to fledging should be rare. As predicted, eggs of Galapagos penguins are more likely to be deserted than chicks (Figures 28 and 29). Tlie data conforms to the model and shows as the reproductive investment of adults increases, the rate of reproductive failure decreases. Observa tions on egg failure (Figure 28) depicts that most eggs are de serted between approximately 12 and 20 days. After 20 days the desertation rate decreases and at around 30 days there is little desertation. For a few days after the eggs hatch, chicks do not die because they can still live on their food reserves. Adelie chicks can survive 6.4 days in an incubator without food (Reid and Bailey, 1967). From two days until around 12 days mortality in young chicks appears equal to egg mortality during the beginning stages. At around 15 days, mortality levels off and the chances of death for chicks remains constant (Fig. 29). In yellow-eyed penguins, mortality of chicks was highest in the first week after hatching (Richdale, 1957). Just prior to fledging, mortality of chicks is slightly reduced suggesting adults are taking greater reproductive risks to fledge the chick or chicks. Data on deserta tion of eggs and chicks of Galapagos penguins conform to the model that desertation should occur before parental energy invest ment is high. Figure 28. Tlie percentage of survival for kno\m age eggs of Galapagos penguins. X = eggs laid in the January to March 1972 breeding period. • = eggs laid in the August to October 1972 breeding period. The number of eggs in the sample (N) = 39, the correlation coefficient (r) = .98, slope (b) = 1.11 and the constant (m) = -.02. The rate of egg loss is greater than the rate of chick loss. IOOt S ®°+ o N =39 .96 œ 404 Z) LO 2 0 - 10 20 30 40 AGE OF EGG IN DAYS Figure 29. The percentage of survival for known age Galapagos penguin chicks through time for two breeding seasons. 0 - June to September 1971 breeding period at Pta. Espinosa and Elizabeth Bay and • = the January to March 1972 breeding period at Pta. Espinosa. N = the number of chicks in the sample each season and b = slope. The rate of chick loss is less than the rate of egg loss. Mortality is generally higher at the beginning of the chick stage than before fledging. Vo SURVI VA L OF CHICKS > 3 Three critical decision making periods seem to occur in the reproductive cycle. Tlie first decision is to begin reproduction and produce eggs, second is to incubate the eggs and third, to raise one or two chicks. Tlie weight of the adults can be used (Table 15) to determine if breeding will continue. Weight prior to nest de sertation was always below the normal weight and was similar to the weight after the molt. Available resources and the physical condi tion of the individual may serve as the feedback mechanism in making the decision. As long as the pair can maintain or gain weight incubation will continue. Differences in the timing of desertation appear to reflect individual and pair differences in foraging ability. Once a pair has managed to sit on eggs for 20 days and not lose weight, the rate of desertation decreases demonstrating they can continue to pay the price of incubation. When chicks hatch an increment is added to the reproductive cost. Once again, the parents must make a decision whether to continue to breed, and if so, how many chicks should be reared. The condition of the breeding birds (weight) which reflects food availability to some extent, serves as the feedback mechanism. If the parents are losing weight one chick will be abandoned, and if this is not sufficient to reduce the reproductive cost allowing the adult to maintain or gain weight, both chicks will be deserted. If fitness of one and two chick nests varies from breeding period to breeding period, as previously suggested, a food related feedback mechanism is ideal to determine how many chicks can be successfully reared. High desertation or low reproductive effort can only be used as a successful tactic if the adults are long-lived and will live to reproduce another time. Mertz (1971) showed that the California Condor (Gymnogyps californianus) population would undergo great changes if adult survival changed. Longevity of adults is vital to a species that has low reproductive effort. I have shown that the Galapagos penguin has experienced total breeding failures as well as moderate nesting successes. By knowing the longevity of the species a better understanding of the effect of nesting failures on the population can be gained. Richdale (1957) found that yellow-eyed penguins do live over 17 years. It is not known how long Galapagos penguins live, but three individuals which are still breeding are known to be 11 years of age. The reproductive success of this species can be very low, and the population will still remain stable considering that individuals breed twice a year, lay two eggs and live long periods. Many seasons of almost total reproductive failure will not seriously hurt the population. Thus, individual adults can afford many reproductive failures and still replace themselves. If one breeding period in five results in one ox two offspring reaching reproductive age for a pair, the adults considering mortality have replaced themselves. In an unpredictable environment many breeding attempts are necessary to insure a breeding success. Longevity allows for numerous tries at repro duction. THE STRATEGIES OF GALAPAGOS PENGUINS AND THEIR RELATIONSHIP TO OTHER ORGANISMS An organism's natural history and adaptations such as sexual dimorphism and erratic breeding have evolved in relationship to the environment. Living in an unpredictable environment, the Galapagos penguin might be expected to have the characteristics attributed to an opportunistic species (Table 1). What are the strategies of the penguin? Penguins feed on a variety of fish and crustaceans (Ainley, 1970). Confined more to a size range of fish than to certain species, the Galapagos penguins, like other species of penguins, are not specialized. Unlike other seabirds, penguins forage at a variety of places and depths eating whatever prey is available. The food and foraging habits of Galapagos penguins demonstrate that compared to other seabirds, they are generalists and unspecialized (see section on Food). Galapagos penguins, therefore, have at least one characteristic ascribed to an opportunistic species (Table 1). Most species of seabirds resemble Antarctic penguins, poss essing breeding cycles which are fixed and predictable (Garrick and Ingham, 1967). Breeding once a year and laying one or two eggs, seabirds have low intrinsic rates of increase (Wynne-Edwards, 1955), a characteristic attributed to an equilibrium species. The 196 Galapagos penguin does not have a fixed annual cycle like other seabirds. Breeding and molting more frequently, the Galapagos penguin lays 4 to 6 eggs during an average year. This breeding strategy can be explained as an adaptation to an unpredictable environment (see Food as the Limiting Factor on the Population). Laying approximately 3 times the number of eggs as other seabirds, this species may be described as having a high intrinistic rate of increase. It has been shown that the Galapagos penguin does possess two characteristics of an opportunistic species: it is unspecialized and has a high R value. A population controlled by density inde pendent factors is the final characteristic ascribed to an oppor tunistic species. Unfortuately, no data exists which demonstrates whether the Galapagos penguin population is under density dependent or independent control. However, Goodman's (1972) evidence showing that a Red footed booby colony on Genevosa Island has remained stable over several thousand years does provide insight. As previously mentioned, the Galapagos penguin has reproductive strategies similar to the Red footed booby. Both feed on fish, have reproductive failures due to lack of food, and live less than 100 miles apart. The rather similar strategies of Red footed boobies and Galapagos penguins living in the same unpredictable environment may well result in the penguin population like the Red footed booby population being stable. Stable numbers would result if the population is under density- dependent control. The evidence suggests that the Galapagos pen guin is food limited. Breeding periods, breeding location, breeding success, molting periods and molting success are all determined by the amount of resources available. However, for the population to be under density dependent control, the number of penguins ex ploiting the food supply must effect individual foraging success. If food occurs in small energy packets which are consumed when found they will not exist for the next foraging penguin. Increased numbers of penguins consuming the spaced resources would decrease individual foraging success. It appears this food pattern may exist when penguins hunt individually. A feedback mechanism between the density of penguins and food may be operating on the reciuitment rate of the population. Thus, the Galapagos penguin population may be under density dependent control. Unlike populations normally thought to be under density depen dent control, breeding success for Galapagos penguins is erratic. In seasonally predictable environments nesting success for seabirds varies but in most seasons some young are fledged (Langham, 1971; Richdale, 1957). In contrast, the Galapagos penguin, Galapagos Albatross (personal communications, Mike Harris), Red footed booby (Nelson, 1968), Brown pelican and Flightless cormorant living in an unpredictable place sometimes experience total reproductive failure. Thus, total nesting failure is common for seabirds living in an unpredictable environment. 199 The difference between seabirds living in an unpredictable environment such as the Galapagos penguin and other seabirds is apparent, but their similarities are even more striking. Wynne- Edwards (1955) argues that seabird populations are under density dependent control and that the low recruitment rate of seabirds is a consequence of a series of adaptations such as a minimum clutch, failure to replace lost eggs, prolongation of the period of parental care, reproductive cycle exceeding one year, deferred maturity and social breeding habits. The Galapagos penguin breeds more fre quently than annually. However, if the cycle is viewed in terms of fledging young, then reproductive cycles often exceed one year. Such similarities in adaptations either imply that all seabirds arc living in comparable environments or that the environment makes little difference in determining the natural history of an organism. The shortened leproductive cycle, irregular breeding and erratic nesting success of the Galapagos penguin are differences in breeding strategies associated with an unpredictable environment. The similarities in breeding strategies in seabirds living in unpre dictable and predictable environments may be a result of adapta tions to common problems. IVhether they feed on plankton or fish, all seabirds have parallel problems in locating food. Phytoplankton populations are not equally distributed (Strickland e^ al_., 1969) . The location of these patches is constantly changing even where the waters are very productive (Strickland ^ al., 1969) . Consequently, their position in time and space becomes unpredictable. Tlie distribution pattern of sardines and other organisms which feed on phytoplankton popu lations (Strickland e^ , 1969) also becomes unpredictable. If a predator feeds on sardines instead of phytoplankton, the distance between clumps of food and the time necessary to find the unpre dictable resource both increase. Even if food is superabundant, there is still a minimum distance which must be traveled and a minimum time to travel that distance to locate food. During repro duction an increase in abundance of food may not enhance the food available for offspring if the adult must spend time traveling to and from the food source. The distribution of the prey may limit reproduction to one or two offspring since gathering sufficient resources for more young may be detrimental to young or adult survival. Nesting failures are a common phenomenon during reproduction. Although a breeding group may fledge a few individuals, many nests are deserted. Lack of resources appear to be a cause of nesting failure in seabirds (Hutchinson, 1950; Nelson, 1968; Simmon, 1967). Therefore, nesting failures indicate that food occasionally may be limited. Because of the distribution and abundance of resources, seabirds may be food limited at some point during their life cycle. The importance of resources in breeding and molting for Galapagos penguins has been demonstrated and it would seem likely that other seabirds have similar constraints. Thus, the problem of exploiting a resource which is temporally or spacially unpredictable may cause similar reproductive strategies. These six strategies are 1) longevity, 2) small clutch size, 3) failure to replace lost eggs or young, 4) prolonged period of parental care, 5) long successful reproductive cycle and 6) deferred maturity. Numerous organisms have the same 6 reproductive strategies as seabirds. Predatory organisms living in the sea exploit the same pattern of resources as seabirds so it is not surprising that whales, dolphins and seals have the same six adaptations. Land organisms such as the puma, hyena, condor, and vulture also exploit resource patterns similar to seabirds. Because their prey is scattered and its location changeable in time and space, we might expect these organisms to have the same traits. The recurrence of these reproductive strategies in diverse organisms may suggest that these six adaptations are characteristic of organisms which exploit a resource pattern which is unpredictable in time and space. The Galapagos penguin combines the characteristics of an opportunistic species such as a high r value and a broad niche with the six adaptations required to exploit a resource which is difficult to locate. It is presumably one of many species which is not entirely an opportunistic or equilibrium species. An organism's resources, and the strategies it has evolved to exploit an environ ment, may often result in a unique mixture of characteristics. Dwelling on isolated islands in the vast Pacific, the Galapagos penguin is an opportunist taking advantage of an unpredictable environment. Its adaptations to survive depend on the vagrancy of the Cromwell Current suggesting its species history may be short. The population is dependent on the upwelling of the Cromwell Current but the probability is high that current patterns will eventually change. If currents shift and resource patterns are altered the Galapagos penguin population may be unable to survive. Tlie remoteness and isolated habitat of the Galapagos penguin may protect it from humans but the restricted nature of the habitat may also result in its extinction. Plate 1. Adult male Galapagos penguin - Individual and age differences in plumage of Galapagos penguins. (A) A week old Galapagos penguin chick showing the beginning of the white cheek patch around the ear. (B) Young downy chick with white cheek patch. (C) Juvenile penguin with the white cheek patch becoming less distinct. (D) Adult Galapagos penguin. (A) (B) (C) CD] Plate 2 (Continued). (E) Adult female Galapagos penguin. (F) Adult female Galapagos penguin with large area of unfeathered black skin. (G) Adult male Galapagos penguin showing large unfeathered pink and black skin. (H) Adult female Galapagos penguin with unfeathered black skin. (ED (F) (GD (HD Plate 3. Molting juvenile Galapagos penguin. Pectoral and facial bands are lacking on juvenile Galapagos penguins...... ^ m m ■Jkid'ÆI wmmMmmi LITERATURE CITED A.bbott, D. P, 1966. Factors influencing the zoogeographic affinities of Galapagos inshore marine fauna, p. 108-122. ^ R. I. Bowman (ed.). The Galapagos. Univ. Calif. Press, Berkeley. Ainley, 0. G. 1970. Sexual size dimorphism in Adelie penguins. Auk 114:267-271. Ainley, D. G. and R. P. Schlatter. 1972. Chick raising ability in adelie penguins. Auk 89(3):559-566. Alexander, W. B. 1963. Birds of the ocean. Putnam's Sons, New York. 306p. Amadon, D. 1959. The significance of sexual differences in size among birds. Proc. Amer. Phil. See., 103:531-536. Ashmole, N. P. 1962. The black noddy Anous tenvirostris on Ascension Island. Ibis 103b:235-273. 1965. The regulation of numbers of tropical oceanic birds. Ibis 103b:458-473. 1971. Seabird ecology and the marine environment. p. 223-286. ^ I. D. S. Earner and J. R. King (eds.), Avian biology. Bailey, A. M. 1962. Nesting of the Galapagos penguin and the Galapagos sooty gull. Condor 64:159-161. Beebe, W. 1924. Galapagos world's end. Putnam's Sons, London and New York. 443p. Beer, C. G. 1969. Laughing gull chicks: recognition of their parents' voices. Science 166:1030-1032. Bellrose, P. C., T. G. Scott, A. S. Hawkins, and J. B. Low. 1961. Sex ratios and age ratios in North American ducks. Bull. Illinois Nat. Hist. Surv. 27:391-474. Boersma, P. D. Adaptations of Galapagos penguins for life in two different environments. ^ B. Stonehouse (ed.), Biology of Penguins. MacMillan, London. In press. Bronson, W. S. 1931. Paddlewings, the penguin of the Galapagos. MacMillan, New York. 106p. Brosset, A. 1963. La reproduction des oiseaux de mer des iles Galapagos en 1962. Alauda 31:81-109. Brown, R. G. B. 1967. Breeding success and population growth in a colony of herring and lesser black-backed gulls, Larus argentatus and JL. fuseus. Ibis 109:502-515. Budd, G . M. 1968. Population increase in the king penguin, Aptenodytes patagonica at Heard Island. Auk 85(4):689-690. Cade, T, J. 1960. Ecology of the peregrine and gyrfalcon populations in Alaska. Univ. Calif. Publ. Zocl. 63:151-290. Carrick, R. and S. E. Ingham. 1967. Antarctic sea-birds as subjects for ecological research. Japan Antarctic Res. Exped., Sci. Rapt. Spec. Issue, 1:151-184, 1970. Ecology and population dynamics of Antarctic seabirds, p. 505-525. ^ M. W. Holdgate (ed.), Antarctic ecology. Vol. I. Academic Press, London and New York. Caughley, G. 1966. Mortality patterns in mammals. Ecology 47: 906-918. Clayton, P. Poultry Science Department, The Ohio State University (personal communication). Cody, M. J. 1966. A general theory of clutch size. Evol. 20: 174-184. Cole, L. C. 1954. The population consequences of life history phenomena. The Quarterly Rev. Biol. 29:103-137. Colinvaux, P. A. 1968. Reconnaissance and chemistry of the lake and bogs of the Galapagos Islands. Nature 219:590-594. Colinvaux, P. A. 1969. Paleolimnologicpi investigations in the Galapagos Archipelago. Mitt. Internat. Verein. Limnol. 17: 126-130. Colinvaux, P. A. 1972. Climate and the Galapagos Islands. Nature 240:17-20. Colinvaux, P. A. 1973. Introduction to Ecology, John Wiley and Son, New York. 62Ip, Couffer, J. C. 1957. Nest of the Galapagos penguin. Condor 59: 399. Coulson, J. C. 1968, Differences in the quality of birds nesting in the centre and on the edge of a colony. Nature 217: 478-479. Crook, J. H. 1964. The evolution of social organization in the weaver birds (Ploceinae). Behavior Suppl. 10. Darling, F. P. 1938. Bird flocks and the breeding cycle. Cambridge. Davis, P. H. 1955. The South African pilchard (Sardinops ocellata) bird predators, 1953-4. Dept, of Commerce and Industries 18:1-32. Downes, N. C., E. H. M. Ealey, A, M. Gwynn and P. S. Young. 1959. The birds of Heard Island. A.N.A.R.E. Report, Ser. b 1:1- 135. Eibl-Eibesfeldt, I. 1960. Galapagos. MacGibbon and Kee, London. 192p. Eisenmann, E. 1956. Galapagos penguin in Panama. Condor 58:74. Elder, W. H. and M. W. Weller. 1954. Duration of fertility in the domestic mallard hen after isolation from the drake. Journal of Wildlife Management 18:495-502. Elton, C. S. 1927. Animal Ecology, Macmillan, New York. 209p. Fisher, A. K. and A. Wetmore. 1931. Report on birds recorded by the Pinchot expedition of 1929 to the Caribbean and the Pacific Islands. Proc. U. S. Nat. Mus. 79(10):l-66. Fisher, H. 1972. The nutrition of birds, p. 431-469. In D. Famer and J. R. Kind (eds.). Avian Biology. Academic Press, New York and London. Fisher, R, A. 1930. The genetical theory of natural selection. Clarendon Press, Oxford. 291p. French, N. R. 1959. Life history of the black rosy finch. Auk 76:159-180. Gifford, E. W. 1913. Expedition of the California Academy of Sciences to the Galapagos Islands 1905-1906. Proc. Calif. Acad. Sci. 2(4):1-132. Goodman, D. 1972. Tlie paleoecology of the Tower Island Bird Colony: A critical examination of complexity stability theory. Ph.D. Thesis, The Ohio State University. Guillen, 0., B. R. Mendiola and R. I. Rondan. 1971. Primary productivity and phytoplankton in the coastal Peruvian waters, p. 157-186. In J. P. Costlow (ed.), Fertility of the sea. Vol. I. Gordon and Breach, New York, London and Paris. Harper, G. R., R. 0. Klataske, R. J. Robel and M. F. Hansen. 1967. Helminths of greater prairie chickens in Kansas. Jour. Wildlife Management 31(2):265-269. Harris, M. P. 1969a. Breeding seasons of sea-birds in the Galapagos Islands. J. Zool., London. 159:145-165. Food as a factor controlling the breeding of Puffinus Iherminieri. Ibis 111:139-156. Hart, T. J. 1942. Phytoplankton periodicity in Antarctic surface waters. Discovery Repts. 21:263-348. Horn, H. S. 1968. The adaptive significance of colonial nesting in the Brewer's blackbird (Euphagus cyanocephalus). Ecology 49 (4):682-694. Hutchinson, G. E. 1950. Survey of contemporary knowledge of bio geochemistry. 3. The biogeochemistry of vertebrate excretion Bull. /uner. Mus. Nat. Hist. 96:1-554. Hutchinson, G. E. 1951. Copepodology for the ornithologist. Ecology 32:571-577. Jackson, J. A. 1970. A quantitative study of the foraging ecology of downy woodpeckers. Ecology 51(2):318-323. Johnson, A. W. 1965. The birds of Chile and adjacent regions of Argentine, Bolivia and Peru. Vol. I. Platt Establicimentos Graficos, Buenos Aires. 398p. Jones, J, H. 1969. Vertical mixing in the equatorial undercurrent. Ph.D. Thesis, Univ. of Calif., San Diego. 68p. Kilham, L. 1965. Differences in feeding behavior of male and female hairy woodpeckers. Wilson Bull. 77:134-145. Kinsky, F. C. 1960. The yearly cycle of the northern blue penguin ÇEudyptula minor novachollandiae) in the Wellington Harbour area. Records of the Dominion Museum 3(3): 145-218. Knox, G. A. 1970. Antarctic marine ecosystems, p. 69-96. ^ M. W. Holdgate (ed.), Antarctic ecology. Vol. I. Academic Press, London and New York. Kooyman, G. L., C. M. Drabsk, R. Eisner, and W. B. Campbell. 1971. Diving behavior of the emperor penguin, Aptenodytes forsteri. Auk 88:775-795. Lack, D. 1950. Breeding seasons, Galapagos. Ibis 92:268-278. 1954. The natural regulation of animal numbers. Clarendon Press, Oxford. 343p. 1968. Ecological adaptations for breeding in birds. Methuen, London. 409p. Langham, N. P. E. 1971. Chick survival in terns (Sterna spp.) with particular reference to the common tern. J. Anim. Ecol. 41(2):385-395. Lehrman, D. S. 1965. Interaction between internal and external environments in the regulation of the reproductive cycle of the ring dove, p. 335-380. P. A. Beach (ed.), Sex and behavior. Wiley and Sons, New York. Leveque, R. 1962. Bird ringing in the Galapagos Islands. Ring 32:126-127. 1963a. Le statut actuel des vertebres rares et menaces de l'archipel des Galapagos. Terre Vie 110:397-430. ______. 1963b. The status of some rarer Galapagos birds. International Council for Bird Preservation 9:96-98. 1,964, Breeding biology of Sphenisciis, p. 615. ^ A. L, Thomson (ed.), A new dictionary of birds. McGraw-Hill Co., New York. Levins, R. L. 1968. Evolution in changing environments. Princeton University Press, Princeton, N. J. 120p. Love, C. M. 1972. Eastropac Atlas Vol. I. U. S. Gov., Washington, D. C. MacArthur, R. H., and J. H. Connell. 1966. The biology of populations. John Wiley and Son, New York. 200p. Malone, T. 1968. Oceanography of the eastern equatorial Pacific, p. 14-29. ^ A broad inquiry into the biological oceano graphy of the eastern tropical Pacific. Stanford, oceano graphic expedition, unpublished manuscript. Margalef, R. 1968. Perspectives in ecological theory. Univ. of Chicago Press, Chicago. Hip. Matthew, W. D. 1915. Climate and evolution. Ann. N. Y. Acad. Sci. 24:271-318. Maxifell, D. C. Marine primary productivity in the Galapagos Islands. Ph.D. Thesis, The Ohio State University (in press). McKinney, F. 1965. The comfort movements of the Anatidae. Behavior 25(1-2):120-220. Mech, D. L. 1966. The wolves of the Isle Royale. U. S. Gov. Printing Office, Washington. 21Op. Mertz, D. B. 1971. The mathematical demography of the California condor populations. Am. Natur. 105(945):437-453. Moreau, R. E. 1950. The breeding seasons of African birds. I. Land birds. Ibis 92 223-267. Murphy, R, C. 1936. Oceanic birds of South America. Amer. Mus. Nat. Hist., New York. 1210p. Nelson, J. B. 1966. The breeding biology of the gannet (Sula bassana) on the bass rock Scotland. Ibis 108:584-626. 1968, Galapagos Islands of birds. William Morrow and Co., New York. 33Sp. 1969. The breeding ecology of the red-footed booby in the Galapagos. J. Anim. Ecol. 38:81-198. Nettleship, D. N. 1971. Breeding success of the common puffin ÇFratercula artica L.) on different habitats at Great Island, Newfoundland. Ecol. Mon. 42:239-268. Newton, F. 1967. The adaptive radiation and feeding ecology of some British finches. Ibis 109:33-98. O ’Brien, P. J. 1940. Some observations on the breeding habits of and general characteristics of the white flippered penguin (Eudyptula albosignata Finsch). Rec. Cant. Mus. 4(6):311-324. Orians, G. H. 1961. The ecology of blackbird (Agelaius) social systems. Ecol. Monogr. 31:285-312. Owen, R. W. and B. Zeitzschel. 1970. Phytoplankton production; seasonal change in the oceanic eastern tropical Pacific. Marine Biology 7:32-36. Palmer, C. E. and R. Pyle. 1966. The climatological setting of the Galapagos, p. 93-99. ^ R. I. Bovmian (ed.), The Galapagos. Univ. Calif. Press, Berkeley. Palmer, R. S. 1962. Handbook of North American birds. Vol. I. Loon through flamingos. Yale Univ. Press, New Haven. 567p. Penney, R. L. 1967. The molt in the adelie penguin. Auk 84(1): 61-71. 1968. Territorial and social behavior in the adelie penguin, p. 83-131. In 0. L. Austin, Jr. (ed.), Antarctic bird studies. Vol. 12. American Geophysical Union, Washing ton, D. C. Perrins, C. M. 1970. The time of birds' breeding seasons. Ibis 112:242-255. Peterson, R. T. 1963. The birds. Life Nature Library. Time Inc., New York. 192p. Pitelka, F. A. 1958. Timing of molt in steller jays of the Queen Charlotte Islands, British Columbia. Condor 60:38-49. Rand, A. L. 1952. Secondary sexual characters and ecological competition. Fieldiana-Zoology 34:65-70. Rand, R. W. 1950. Tlie government guano islands: birds and seals. Union of South Africa Dept, of Agriculture Rept. 1-15. 1960. The biology of guano-producing sea birds. The distribution, abundance and feeding habits of the cape penguin, Spheniscus deniers us off the south-western coast of the Cape Province. Union of South Africa Dept. Comm, and Indust. Div. Fish. Invest. Rept. 41:1-28. Raymont, J. E. G. 1963. Plankton and productivity in the oceans. Vol. 18. Pergamon Press, Oxford. 660p. Reid, B. E. 1965. The adelie penguin (Pygoscelis adeliae) egg. N. Z. Jour. Sci. 8:503-514. _ and C. Bailey. 1967. The value of the yolk reserve in adelie penguin chicks. Rec. Dominion Mus. 5(19): 185-193. Richdale, L. E. 1951. Sexual behavior in penguins. Univ. Kansas Press, Lawrence. 315p. 1957. A population study of penguins. Oxford Univ. Press, Oxford. 195p. 1963. Biology of the sooty shearwater Puffinus griseus. Proc. Zool. Soc. London 141:1-117. Ridgway, R. 1897. Birds of the Galapagos Archipelago. Proc. U. S. Nat. Mus. 19:459-670. Roberts, B. 1940. The breeding behavior of penguins with special reference to Pygoscelis papus (Forster). British Graham land expedition 1934-1937. Scientific Repts. 1(3):195-254. Schoener, T. W. 1968. The anolis lizard of Bimini; resource partitioning in a complex fauna. Ecology 49:704-726. Scholander, P. F. 1940. Experimental investigations on the respiratory function of diving mammals and birds. Hralradets Skrifter; Norske Vikenskaps Akad. Oslo 22:1-131. Schreiber, R. W. and N. P. Ashmole. 1970. Seabird breeding seasons on Christmas Island, Pacific Ocean. Ibis 112:363-394. Selander, R. K. 1966. Sexual dimorphism and differential niche utilization in birds. Condor 68:113-151. 1972. Sexual selection and dimorphism in birds, p. 181-281. In B. Campbell (ed.), Sexual selection and the descent of man 1871-1971. Aldine, Chicago. Sibert, J. 1971. Some oceanographic observation in the Galapagos Islands. Am. Zool. 11(3):405-408. Simmon, K. E. L. 1967. Ecological adaptations in the life history of the brown booby at Ascension Island. Living Bird 6:187- 212. 1970. Ecological determinants of breeding adaptations and social behavior in two fish-eating birds, p. 37-77. In^ J. H. Crook (ed.), Social behavior in birds and mammals. Academic Press, London and New York. Sladen, W. J. L. 1958. The pygoscelid penguins. F.I.D.S. Rept. 17:1-97. ______, R. Le Resche, and R. Wood. 1968. Antarctic avian population studies 1967-1968 (penguin skua banding). Antarc tic Jour. U. S. 3(6):247-249. Slevin, J. R. 1959. The Galapagos Islands; A history of their exploration. Occ. Pap. Calif. Acad. Sci. 25:1-150. Slobodkin, L. B. and H. L. Sanders. 1969. On the contribution of environmental predictability to species diversity, p. 82- 95. ^ Bookhaven Symposia in Biology, No. 22, U. S. Dept. Commerce, Springfield, Virginia. Smith, C. L. and J. C. Tyler. 1972. Space resource sharing in a coral reef fish community, p. 125-170. ^ B. B. Collette and S. A. Earle (eds.). Ecology of coral reef fishes. Nat. Hist. Mus. Los Angeles County Sci. Bull. 14. Nat. Hist. Mus., Los Angeles. Snodgrass, R. E. and E. Heller. 1904. Papers from the Hopkins- Stanford Galapagos expedition, 1898-99. Proc. Wash. Acad. Sci. 5:231-372. Snow, D. W. and B. K. Snow. 1967, The breeding cycle of the swallow-tailed gull Creagrus fureatus. Ibis 109:14-24. Southern, H. N. 1959. Mortality and population control. Ibis 101:429-436, Stonehouse, B. 1953. The emperor penguin Aptenodytes forsteri 1. Breeding behavior and development. F.I.D.S. Rept, 23: 1-81. _. 1970. Geographic variation in gentoo penguins (Pygoscelis papua). Ibis 112:52-57. 1971. The Snare Islands penguin Eudyptes robustus. Ibis 113(1):l-7. Storer, R. W. 1966. Sexual dimorphism and the food habits in three North /anerica’^ accipiters. Auk 83:423-456. Strickland, J. D. H., R. W. Eppley and B. Rojas De Mendiola. 1969. Phytoplankton populations, nutrients and photosynthesis in Peruvian coastal waters. Institute del Mar Del Peur Bull. 2(1):1-45. Sturkie, P. D. 1954. Avian physiology. Comstock Publishing Associates, Ithaca, New York. 766p. Sumner, F. B. 1934. Does "protective coloration" protect? - Results of some experiments with fishes and birds. Proc. Nat. Acad. Sci., 20:559:563. Sundevall, E. J. 1871. On birds from the Galapagos Islands. Proc. Zool. Soc., London 124-130. Swarth, H. S. 1931. The avifauna of the Galapagos Islands. Occ. Pap. Calif. Acad. Sci. 18:5-299. Taylor, R. 1962. The adelie penguin Pygoscelis adeliae at Cape Royds. Ibis 104(2): 176-204. Taylor, T. G. 1971. How an eggshell is made. Scientific Am. 222(3):88-95. Tenaza, R. 1971. Behavior and nesting success relative to nest location in adelie penguins (Pygoscelis adeliae). Condor 73: 81-92. Thompson, D. H. and J. T. Emlen. 1968. Parent-chick individual recognition in the adelie penguin, Antarctic Jour. U. S. 3(4):132, Thornton, I. 1971. Darwin’s Islands: A natural history of the Galapagos. Natural History Press, Garden City, N. Y. 322p. Tinbergen, N. 1965. Animal behavior. Time-Life Books, New York. 200p. Townsend, C. H. 1927. The Galapagos penguin in captivity. Auk 46:211-213. Trivers, R. L. 1972. Parental investment and sexual selection, p. 137-179. ^ B. Campbell (ed.) Sexual selection and the descent of man 1871-1971. Aldine, Chicago. van Zinderen Bakker, E. M., Jr. 1971. A Behaviour analysis of the Gentoo penguin (Pygoscelis papua Forster). p. 251-272. In E. M. van Zinderen Bakker Sr. and R. A. Pyer (eds.) Marion and Prince Edward Islands. Report on the South African Biological and Geological Expedition 1965-1966. Warham, S. 1958. The nesting of the little penguin Eudyptula minor. Ibis 100(4):605-616. 1963. The Rockhopper penguin, Eudyptes chrysocome. at Macquarie Island. Auk 89:86-105. 1972. Breeding seasons and sexual dimorphism in Rockhopper penguins. Auk 89:86-105. ______. 1973. Egg-size dimorphism in Eudyptes penguins. 91st A.O.U. Meeting, Provincetoim, Massachusetts. Wiggins, I. and D. M. Porter. 1971. Flora of the Galapagos Islands. Stanford Univ. Press, Stanford. Williamson, P. 1971. Feeding ecology of the Red-eyed vireo (Vireo olivaceus) and associated foliage-gleaning birds. Ecol. Monogr. 41 (2): 129-152. Wynne-Edwards, V. C. 1955. Low reproductive rates in birds, especially sea birds. Acta XI Int. O m . Congr. Basel 1954: 540-547. Wyrtki, K. 1963. The horizontal and vertical field of motion in the Peru Current. Bull. Scripps Institution of Oceanography 8(4):313-346. Wyrtki, K. 1965. Surface currents of the eastern tropical Pacific Ocean. Bull. Inter-Amer. Trop. Tuna Comm. 9:271-304. Oceanography of the eastern equatorial Pacific Ocean, p. 33-68. ^ H. Barnes (ed.), Oceanogr. Mar. Biol. Ann. Rev. G. Allen and Unwin, London. Circulation and water masses in the eastern equatorial Pacific Ocean. International Jour. Oceanography and Limnology 1(2): 117-147. Yeates, G. W: 1968. Studies on the adelie penguin at Cape Royds 1964-65 and 1965-66. New Zealand Jour, of Marine and Fresh water Research 2(3):472-496. Ziswiler, V. 1967. Extinct and vanishing animals. Longmans, Green and Co., London.