foraging behavior of peruvian boobies sula variegata carlos b
TRANSCRIPT
FORAGING BEHAVIOR OF PERUVIAN BOOBIES Sula variegata IN NORTHERN PERU: AN ANALYSIS OF INTERSEXUAL FORAGING SEGREGATION AND MARINE HABITAT USE
Carlos B. Zavalaga
A Dissertation Submitted to the University of North Carolina Wilmington in Partial Fulfillment
of the Requirements for the Degree of Doctor of Philosophy
Department of Biology and Marine Biology
University of North Carolina Wilmington
2008
Approved by
Advisory Committee
Joanne Halls Lawrence Cahoon
Frederick Scharf Michael McCartney
Steven Emslie
Chair
Accepted by
________________________ Dean, Graduate School
ii
“.... Hundreds of birds [Peruvian Boobies] seemed to strike the water at every
instant, and even within a few feet of our boat. The bewildering effect is to be imagined rather than described; the atmosphere “cloudy” with birds,
the surface of the sea broken and spattering from falls of animate drops and speckled with reappearing birds; the confused sounds of whirring
wings and unremitting splashes.
Robert E. Coker (1919).
Habits and economic relations of the guano birds of Peru.
iii
TABLE OF CONTENTS
INTRODUCTION...........................................................................................................................................................................v
ACKNOWLEDGEMENTS.......................................................................................................................................................viii
LIST OF TABLES......................................................................................................................................................................... ix
LIST OF FIGURES.......................................................................................................................................................................x
CHAPTER 1: SEXUAL SIZE DIMORPHISM IN PERUVIAN BOOBIES: VOICE AND SIZE AS TRAITS
FOR SEX RECOGNITION .........................................................................................................................................................1
ABSTRACT.................................................................................................................................................................................2
INTRODUCTION.......................................................................................................................................................................3
METHODS..................................................................................................................................................................................4
Birds and measurements ...................................................................................................................................................4
DNA analysis..........................................................................................................................................................................4
Statistical Analyses..............................................................................................................................................................5
RESULTS....................................................................................................................................................................................6
DISCUSSION .............................................................................................................................................................................8
REFEENCES ...........................................................................................................................................................................10
CHAPTER 2: AT-SEA MOVEMENT PATTERNS AND DIVING BEHAVIOR OF PERUVIAN BOOBIES
SULA VARIEGATA : SEXUAL SEGREGATION BY FORAGING HABITAT IN A RICH MARINE
ENVIRONMENT?.......................................................................................................................................................................12
ABSTRACT...............................................................................................................................................................................13
INTRODUCTION.....................................................................................................................................................................14
METHODS................................................................................................................................................................................17
Study site ...............................................................................................................................................................................17
Description of dataloggers...............................................................................................................................................17
Capture of birds...................................................................................................................................................................18
Handling of data and analyses.......................................................................................................................................20
Statistical analysis..............................................................................................................................................................23
RESULTS..................................................................................................................................................................................25
Timing of departures and duration of feeding trips..................................................................................................26
Diving behavior....................................................................................................................................................................27
At-sea activities....................................................................................................................................................................28
Flight ground speeds and sinuosity of paths .............................................................................................................29
Movement patterns .............................................................................................................................................................32
Foraging areas.....................................................................................................................................................................36
Diet...........................................................................................................................................................................................40
iv
DISCUSSION ...........................................................................................................................................................................41
Peruvian Boobies as predators of anchovetas.........................................................................................................41
Site-specific foraging behavior.......................................................................................................................................42
Foraging behavior of Peruvian Boobies in comparison to tropical boobies....................................................43
Sex-specific foraging behavior.......................................................................................................................................44
REFERENCES........................................................................................................................................................................47
CHAPTER 3: EFFECTS OF OCEANOGRAPHIC FEATURES AND WIND CONDIT IONS ON THE
FORAGING MOVEMENTS OF PERUVIAN BOOBIES SULA VARIEGATA: A COMPARISON BETWEEN
AN INSHORE AND OFFSHORE ISLAND .........................................................................................................................52
ABSTRACT...............................................................................................................................................................................53
INTRODUCTION.....................................................................................................................................................................54
METHODS................................................................................................................................................................................56
Study site ...............................................................................................................................................................................56
Dataloggers and handling of birds................................................................................................................................57
Handling of birds.................................................................................................................................................................57
Oceanographic data...........................................................................................................................................................58
Data handling and analysis.............................................................................................................................................58
Effects of wind on flight direction and speed.............................................................................................................59
Statistical analysis..............................................................................................................................................................60
RESULTS..................................................................................................................................................................................63
Effects of bathymetry and oceanographic features.................................................................................................63
Effects of wind ......................................................................................................................................................................70
DISCUSSION ...........................................................................................................................................................................74
Effects of SST and chla concentration ........................................................................................................................74
Bathymetric features..........................................................................................................................................................77
Wind patterns .......................................................................................................................................................................78
REFERENCES........................................................................................................................................................................80
v
INTRODUCTION
The Peruvian Booby (Sula variegata) is a member of the Sulidae family (Order Pelecaniformes), which
comprises three extant species of gannets and seven species of boobies. Sulids have a wide distribution,
occurring from temperate to tropical regions. Gannets are sexually monomorphic, but females are larger
than males in boobies. This female-based size dimorphism is uncommon within the seabird taxa, being
also found in skuas, jaegers and frigatebirds (Fairbairn and Shine 1993). The extent of body weight
dimorphism varies widely among booby species, ranging from 15% in Red-footed Boobies Sula sula to
38% in Brown Boobies Sula leucogaster (Nelson 1978). Sulids are strictly marine birds feeding on fish
and squid that are captured by plunge-dives.
The Peruvian Booby is one of the most representative and abundant endemic seabirds of the
Humboldt Current. Although this species breeds on islands and coastal cliffs from Punta Pariñas in Peru
(4o40’S, Harrison 1985) to Isla Mocha in Chile (38o22’S, Schlatter and Simeone 1999), the bulk of the
population occurs along the Peruvian coast (Nelson 1978). The population size of Peruvian Boobies is
estimated at 3.5 - 4 millions of birds off Peru (Birdlife International 2008), with major breeding colonies
located on islands of the central and northern coast (Nelson 1978). The abundance of this species has
varied considerably over the past several centuries as a consequence of human disturbance, guano
harvesting and commercial fishing (Duffy 1994). Together with the Guanay Cormorant (Phalacrocorax
bougainvillii) and the Peruvian Pelican (Pelecanus thagus), the Peruvian Booby is a member of the so-
called guano producing seabird and as such, has played an important role in the Peruvian economy
between the 1850s and 1950s (Duffy 1994). Peruvian boobies display
Peruvian Boobies feed on several species of pelagic fish, but the Peruvian anchovy or anchoveta
(Engraulis ringens) is the main prey consumed (Jahncke and Goya 1998). The anchoveta is the keystone
species in the trophic chain of the coastal upwelling system of the Humboldt Current. This fish captures
most of the primary and secondary production in the ecosystem, and simultaneously supports different
species of fish, marine vertebrates and one the world’s largest commercial fisheries. The distribution and
abundance of anchovetas are strongly influenced by seasonal (predictable) and interannual
(unpredictable) changes of oceanographic conditions (Bakun 1987), which in turn affect the foraging
behavior and productivity of their predators. The extremes of this environmental variability are exhibited
vi
during warm and cold episodes associated to El Niño and La Niña events, respectively. Life-history traits
exhibited by Peruvian Boobies such as enlarged broods and high breeding frequency are tuned to cope
with the environmental uncertainty (Nelson 1978). However, the links between the foraging behavior of
Peruvian Boobies and variations of oceanographic features remain unexplored.
The coastal upwelling ecosystem of the Peruvian coast offers unique conditions to study the
foraging ecology of seabirds in relation to the marine environment. The seasonal and interannual regimes
of the oceanographic conditions, coupled to the large-scale anchoveta fisheries, should have a strong
impact in the movement patterns, foraging success and productivity of several anchoveta-predator
species, including Peruvian Boobies. The application of advanced electronic technology to free-ranging
seabirds in recent years has allowed biologists to examine, with unprecedented accuracy, the
relationships between foraging behavior and morphology, conditions of the environment (review in Burger
and Shaffer 2008). In this study, I used GPS dataloggers and dive meters to characterize the foraging
behavior of breeding Peruvian Boobies on two islands in northern Peru. This information was used to test
several hypotheses of foraging behavior in relation to sexual size dimorphism, niche segregation and
marine habitat use.
In Chapter 1, I examined the extent of sexual size dimorphism of Peruvian Boobies and
determined a reliable method to identify females and males in the field. In Chapter 2, I described the
foraging behavior and compared it with what is known from other tropical boobies Likewise, I evaluated
whether females and males exploit different foraging habitats as a result of size dimorphism, and
identified foraging behavior differences of birds breeding in an inshore and offshore island. Finally, in
Chapter 3, I examined the marine habitat use by overlaying the birds’ foraging areas to remote-sensing
data of oceanographic conditions and bathymetry, as well as wind patterns.
REFERENCES
Bakun A (1987) Monthly variability in the ocean habitat off Peru as deduced from maritime observations,
1953 to 1984. In: Pauly D, Tsukayama I (eds) The Peruvian anchoveta and its upwelling ecosystem:
three decades of change. ICLARM Studies and Reviews 15. Instituto del Mar del Peru (IMARPE),
Callao, Peru; Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ) GMBH, Eschbom,
vii
Federal Republic of Germany; and International Center for Living Aquatic Resources Management
(ICLARM), Manila, Philippines, p 46-74.
Birdlife International (2008) Birdlife International American Bird Conservancy workshop on seabirds and
seabird-fishery interactions in Peru. RSPB, Sandy, UK.
Burger AE, Shaffer SA (2008) Application of tracking and data-logging technology in research and
conservation of seabirds. Auk 125: 253-264.
Duffy DC (1994) The guano islands of Peru: the once and future management of a renewable resource.
Birdlife Conservation Series 1:68-76.
Fairbairn J, Shine R (1993) Patterns of sexual size dimorphism in seabirds of the Southern Hemisphere.
Oikos 68: 139-145.
Harrison P (1985) Seabirds, an identification guide. Houghton Mifflin Company, New York.
Jahncke J, Goya E (1998) Las dietas del guanay y del piquero peruano como indicadores de la
abundancia y distribucion de anchoveta. Boletín Ins. Mar Perú 17:15-34.
Nelson JB (1978) The Sulidae: Gannets and Boobies. Oxford University Press, Oxford.
Schlatter RP, Simeone A (1999) Status and conservation of Chilean seabirds. Estudios Oceanológicos
18: 25-33.
viii
ACKNOWLEDGEMENTS It would have not been possible to complete this thesis without the help, collaboration and advice of
several persons including professors, friends, family, fishermen and boobies. I want to express my
gratitude to all of them for their support in the past five years.
I would like to thank Dr. Steve Emslie for his support on this project and for inviting me to pursue a
doctorate degree immediately after obtaining my Master degree here at UNCW. He provided suggestions
during the course of this research, reviewed initial drafts of the thesis, and supported me with travel
expenses and materials. He was also very supportive with all kind of requirements I had, including a new
computer and specialized software for data analysis. Steve traveled with me to the Galapagos Islands to
track Nazca boobies (data presented elsewhere) and worked together in the field. During those weeks on
the islands, I appreciated Steve not only as an admirable ornithologist, but also as a friend.
I want to express my gratitude to Dr. Giacomo Dell’Omo from T echnosmart who provided a lot of GPS
dataloggers and equipment for free! Giacomo taught me everything I know about the use of small
electronic loggers and seabird tracking. He traveled and worked with me on the Peruvian guano islands
and gave a thorough revision of this thesis.
I am grateful to Gina Mori and Scott Taylor for their help in the field, and to the island guards Rolando
Balbín and Armando Nieto for their hospitality. “El Puma” lent us his deep cycle batteries when ours
quickly discharged days after starting the fieldwork. I wouldn’t’ have been able to recharge the GPS
loggers and obtain bird tracks without their help. Loretta Baglietto carried out the difficult task of getting
the permits to work on the islands and exporting the blood samples to Canada. Thanks to T. Birt for his
help with laboratory work, and T. Maness for her advice related to molecular sexing of boobies. I thank
Amanda Kahn for her friendship and advice in completing all the graduation forms at time.
I received financial support for travel expenses to Peru from the Graduate School, the Graduate Student
Association and the Ralph W. Brauer Fellowship at UNCW. I thank PROABONOS for permission to work
on the islands (CARTA N 186-2007-AG-PROABONOS-GO/DE). Collection and exportation of blood
samples for DNA analysis was possible with permits issued by the Peruvian Institute of Natural
Resources, Ministry of Agriculture - INRENA (011352-AG-INRENA and 143-2007-INRENA-IFFS-DCB).
I thank my thesis committee Dr. J, Halls, Dr. L. Cahoon, Dr. F Scharf and Dr. M. McCartney for their
revisions and comments. Dr. Halls gave multiple suggestions and ideas to improve the GIS analysis.
Finally, I am grateful to all members of my family for their support during all these years.
ix
LIST OF TABLES
Table Page
1. Sexual differences of morphological characters of breeding Peruvian Boobies 6
2 Sample size at the bird, trip and data point level used in the analysis of foraging data of breeding
Peruvian Boobies 25
3 Sex-specific differences in the number of trips in a day, trip length and diving behavior of Peruvian
Boobies 30
x
LIST OF FIGURES
Figure Page
1. Morphometric differences in (a) body weight, (b) culmen length, and (c) wing chord
of Peruvian Booby females (N = 24) and males (N = 25) breeding on Islas Lobos de
Tierra and Lobos de Afuera, Perú 7
2. Breeding Peruvian Booby Sula variegata with two medium-sized chicks
at Isla Lobos de Tierra 19
3. Distribution frequency of instantaneous movement speeds of Peruvian Boobies 22
4. Frequency distribution in the departure time of breeding Peruvian Boobies 26
5. Frequency distribution of dive depths in female and male Peruvian Boobies 28
6. Sinousity index at the three sections of the foraging tracks of Peruvian Boobies
(beginning = outbound, middle = feeding, ending = outbound) on (a) Isla Lobos de Tierra
in December 2006 and (b) lsla Lobos de Afuera in December 2007 29
7. Straight (a) and looping (b) foraging tracks of two breeding Peruvian Boobies 32
8. Foraging tracks of females (blue tracks) and male (red tracks) Peruvian Boobies
breeding at Lobos de Tierra in 2006 (a) and Lobos de Afuera in 2007 (b) 34
9. Association between the duration of the foraging trip and the maximum
foraging range and total travelled distance of Peruvian Boobies 36
10 Frequency distribution of standardized location of dives in relation to the
farthest point in the foraging trip of Peruvian Boobies 37
11. Frequency distribution of dive locations in relation to the colony from
female and male Peruvian Boobies 38
12. Frequency distribution of the dive locations in relation to the nearest point
to the mainland of breeding female and male Peruvian Boobies 39
13. Foraging areas of female (white), male (grey) breeding Peruvian Boobies
instrumented with GPS dataloggers and a depth/temperature tags at
(a) Isla Lobos de Tierra in 2006 and (b) Isla Lobos de Afuera in 2007 39
14. Location of Isla Lobos de Tierra and Isla Lobos de Afuera in northern Peru 56
xi
15. Flow chart for the spatial analysis of marine habitat use of Peruvian Boobies
breeding at Isla Lobos de Tierra in 2006 and Lobos de Afuera in 2007 61
16 (a) Overlay of Peruvian Booby dive position fixes and bathymetry charts
around Isla Lobos de Tierra. (b) Frequency distributions of bathymetric data
in the foraged area in comparison to the available and crossed areas 63
17. (a) Overlay of Peruvian Booby dive position fixes and bathymetry charts
around Isla Lobos de Afuera. (b) Frequency distributions of bathymetric data
in the foraged area in comparison to the available and crossed areas 64
18. (a) Overlay of Peruvian Booby dive position fixes and SST charts around Isla Lobos
de Tierra. (b) Frequency distributions of SST data in the foraged area in
comparison to the available and crossed areas 65
19. (a) Overlay of Peruvian Booby dive position fixes and SST charts around Isla Lobos
de Afuera. (b) Frequency distributions of SST data in the foraged area in
comparison to the available and crossed areas 66
20. (a) Overlay of Peruvian Booby dive position fixes and chla charts around Isla Lobos
de Tierra. (b) Frequency distributions of chla data in the foraged area in
comparison to the available and crossed areas 67
21. (a) Overlay of Peruvian Booby dive position fixes and chla charts around Isla Lobos
de Afuera. (b) Frequency distributions of chla data in the foraged area in
comparison to the available and crossed areas 68
22. Observed and expected flight direction relative to wind direction of Peruvian Boobies
from Isla Lobos de Tierra in December 2006 during (a) outbound and (b) inbound flights 70
23. Observed and expected flight direction relative to wind direction of Peruvian Boobies
from Isla Lobos de Afuera in December 2007 during (a) outbound and (b) inbound flights 71
24. Mean flight speed (km·h-1) of Peruvian Boobies relative to wind direction on
(a) Isla Lobos de Tierra and (b) Isla Lobos de Afuera 72
CHAPTER 1: SEXUAL SIZE DIMORPHISM IN PERUVIAN BOOBIES: VOICE AND SIZE AS TRAITS FOR SEX RECOGNITION
2
ABSTRACT Females of the Sula boobies (Sulidae) are larger than males, but the degree of this dimorphism varies
widely among species. This scenario offers an excellent opportunity to test hypotheses on the origin and
evolution of sexual size dimorphism (SSD); however, sexual differences in the phenotypes have been
difficult to recognize in some species of boobies. We studied adult Peruvian Boobies (Sula variegata) on
two islands in northern Peru to determine their sex using DNA-based techniques. The genetic analysis
was used to assess the extent of SSD in this species and to validate the use of voice and morphometric
characters as a reliable method for sexing adult Peruvian boobies in the field. Female Peruvian Boobies
were 19% heavier and their culmens and wings were 3 - 4 % larger than males. The sex of 92% of the
birds (N = 24 females and 25 males) could be correctly classified using the discriminant function: D = -
39.57 + 0.0085(BW) + 0.2961(CL) + 0.6254(WC), where BW = body weight in grams, CL = culmen length
in cm, and WC = wing chord and in cm. A value of D > 0 classifies an individual as a female and D < 0 as
a male. Vocalizations unequivocally discriminated sexes of Peruvian Booby adults. Whistles were
performed exclusively by males (25/25 of cases), whereas grunts or goose-like honk vocalizations were
performed only by females (24/24 of cases). Our results show that calls provide a fast, reliable, and
inexpensive method for sexing adult Peruvian boobies in the field. In comparison to other Sula boobies,
the female-larger SSD of Peruvian Boobies is intermediate. Because the degree of SSD in seabirds may
be correlated to marine productivity (females larger than males in the impoverished waters of the tropics),
the intermediate SSD of Peruvian Boobies may reflect a transitional stage of evolution toward
monomorphism following colonization of the rich Humboldt Current, rather than an equilibrial SSD related
to the selective environment of this ecosystem.
3
INTRODUCTION Sexual size dimorphism (SSD) varies widely among the ten extant species of the Sulidae (boobies and
gannets). The three species of gannets, Morus spp. and their close relative Abbott’s Booby (Papasula
abbotti) are nearly monomorphic, whereas females of the Sula boobies are significantly larger than males
(Nelson 1978). The body mass disparity between sexes in Sula boobies ranges from 15% in Red-footed
Boobies (S. sula, Weimerskirch et al. 2006) to 38% in Brown Boobies (S. leucogaster, Lewis et al. 2002).
The origin and maintenance of the female-larger SSD in seabirds can be explained by two mechanisms
which are not mutually exclusive: sexual selection (e.g., Catry et al. 1999) and ecological segregation
(e.g., Weimerskirch et al. 2006). The wide range of SSD in sulids provides an excellent opportunity to test
these hypotheses (Lewis et al. 2002, Weimerskirch et al. 2006, Zavalaga et al. 2007), but SSD is poorly
characterized in some species. Sex in sulids can be determined by DNA analysis (Redman et al. 2002,
Lewis et al. 2005, Weimerskirch et al. 2006, Maness et al. 2007) or by sexual differences in behavior and
phenotypes (Gilardi 1992, Zavalaga et al. 2007). For instance, the sex of adult Blue-footed Boobies (S.
nebouxii) can be unmistakably assigned by the birds’ calls, the extent and shape of the inner pigmented
ring of the iris, and body size (Nelson 1978, Zavalaga et al. 2007). However, differences in these features
are absent or are more difficult to recognize in other booby species.
Very little information exists regarding sex-related morphometry of Peruvian Boobies (Sula
variegata). Based on data from 14 males and nine females, Murphy (1936) reported that the bill, wing, tail,
and tarsus were 4 – 5% larger in females. The only information on sexual body mass disparity comes from
comparison of four known-sex adults (Nelson 1978), where females were 14% heavier than males. The
voice appears to differ by sex (males whistle and female grunt), although the pitch intensity and amplitude
of the calls is less marked than in some other booby species (Nelson 1978). The reliability of morphology
and vocalizations in determining the sex of adult Peruvian Boobies has not yet been thoroughly
evaluated. In this study, we used a DNA-based technique to determine the sex of adult Peruvian Boobies
and then use this information to assess the extent of SSD in this species and to validate the use of voice
as a reliable method for sexing Peruvian boobies in the field.
4
METHODS We studied Peruvian Boobies on two islands in Northern Perú: Lobos de Tierra (6o24’S, 80o51’W)
between 22 and 30 December 2006, and Lobos de Afuera (6o57’S, 80o41’W) between 10 and 17
December 2007. Lobos de Tierra is located only 65 km north of Lobos de Afuera, and for the purpose of
this study, we pooled the data from the two islands because we did not expect to find any regional
morphological differences between the two populations. Further, genetic analysis of a 540 base pair
segment of the mitochondrial control region in 110 Peruvian Boobies from throughout their range did not
reveal significant genetic differences between these two colonies, or between these colonies and others
1000 km to the south of these islands (Taylor et al. unpublished data).
Birds and measurements
We captured 14 and 35 birds on Lobos de Tierra and Lobos de Afuera, respectively. On both islands,
Peruvian Boobies nested in dense groups (ca. 2 - 3 nests·m-2) of 800 – 1000 nests. We selected birds
rearing medium-size chicks and removed them from their nests using a hook attached to a 3-m pole,
placed gently around their necks. We recorded calls during restraining and while taking measurements,
as birds initially struggled and vocalized, recording vocalizations as either of two types: whistles or grunts.
We captured individuals early in the morning (0530 – 0800 h), before their first feeding trip, so the body
weight recorded represents the minimum daily weight. Birds were weighed with a Pesola® spring scale to
the nearest 50 g. Using calipers (accuracy of 0.01 mm) and a ruler (accuracy 1 mm), we measured
exposed culmen and wing chord, respectively. Measurements were performed by the same person (CBZ)
throughout. The brachial vein of each individual was pricked using a lancet to obtain three drops of blood.
These were absorbed on filter paper and stored in vials in 70% ethanol. This entire procedure took less
than 5 min per individual. Blood samples are archived at Queen’s University, Ontario, Canada.
DNA analysis
We extracted DNA from blood using a standard proteinase-K phenol/chloroform technique (Friesen et al.
1996). We sexed birds using primers 2550F and 2178R developed by Fridolfsson and Ellegren (1999) for
molecular sexing of non-ratite birds. Using these primers, fragments of two lengths are amplified from
5
female DNA (female birds are the heterogametic sex) and fragments of a single length are amplified from
male DNA. PCR amplifications were conducted in 25 µL of a cocktail containing 10 mM Tris pH 8.0, 3.5
mM MgCl2, 0.4 µM of each primer, 50 mM KCl, 0.2 mM dNTPs, and 0.5 units of Taq DNA polymerase
(Qiagen, Mississauga). The temperature profile consisted of an initial denaturizing period at 94°C for 3
min followed by annealing for 1 min at 47°C and extension for 1 min 30 sec at 72°C. This was followed by
34 cycles of 94°C for 45 sec, annealing at 47°C for 1 min and extension at 72°C for 1 min and 30 sec.
PCR products were subjected to electrophoresis and visualized on 2% agarose gels. All samples were
analyzed without knowledge of the suspected sex of the bird.
Statistical Analyses
For sex-specific comparisons of mean values of morphological characters we used t-tests. A discriminant
analysis (Manly 2005) was used to separate sexes based on combinations of more than one
morphological character. Statistical Analysis Systems (SAS Institute 2004) was used for all statistical tests
(a = 0.05). Means are expressed ± SD.
6
RESULTS On the basis of DNA analysis, the birds caught on Lobos de Tierra and Lobos de Afuera islands
comprised nine males and five females, and 16 males and 19 females, respectively. Analyses of
morphological characters revealed that female Peruvian Boobies were 19% heavier and their culmens
and wings were 3 - 4 % larger than males (Table 1). However, female and male distributions overlapped
(Fig. 1): If a measurement fell either in a range of 1300 - 1450 g for body mass, 9.19 – 9.94 cm for culmen
length, or 39.1 – 40.5 cm for wing chord, then the measurement could not be used alone to sex the bird.
Table1. Sexual differences of morphological characters of breeding Peruvian Boobies on Isla Lobos de
Tierra (N =14 birds) and Lobos de Afuera (N = 35), Perú. Means were compared using an independent
sample t– test. Mean values are expressed ± SD and range in parentheses. Dif = difference in percentage
(F – M)/M x 100, where F is female measurement and M is male measurement.
All morphometric characters were significantly selected by the discriminant analysis (Wilks’ Lambda =
0.269, F3,45 = 36.78, P < 0.0001) according to the following unstandardized canonical discriminant
equation:
D = -39.57 + 0.0085(BW) + 0.2961(CL) + 0.6254(WC)
where BW = body weight (g), CL = culmen length (cm), and WC = wing chord (cm). A value of D > 0
classifies an individual as a female and D < 0 as a male. This equation correctly classified 92% of males
Females (N = 24) Males (N = 25) Dif (%) t P
Body weight (g) 1543 ± 105
(1300 – 1675)
1290 ± 75
(1200 – 1450)
19 9.65 < 0.001
Culmen length
(cm)
9.61 ± 0.25
(9.19 – 10.14)
9.18 ± 0.34
(8.47 – 9.94)
4 4.95 < 0.001
Wing chord (cm) 40.25 ± 0.57
(39.1 – 41.5)
39.0 ± 0.68
(38.0 – 40.5)
3 7.03 < 0.001
7
and 92% of females. The average of the discriminant scores (group centroid) for females was 1.56 and for
males -1.50. No birds with scores higher than 0.0098 (females) or with scores below -1.3035 (males) were
misclassified. Thus, for greatest accuracy, we recommend that this method should not be used to identify
the sex of individuals whose discriminate scores falls between this range of values.
Vocalizations unequivocally discriminated sexes of Peruvian Booby adults. Whistles were
performed exclusively by males (25/25 of cases), whereas grunts or goose-like honk vocalizations were
performed only by females (24/24 of cases).
Figure 1. Morphometric differences in (a) body weight, (b) culmen length, and (c) wing chord of Peruvian
Booby females (N = 24) and males (N = 25) breeding on Islas Lobos de Tierra and Lobos de Afuera,
Perú. Boxes and whiskers represent the middle 50% and middle 75% of the data, respectively.
Females Males1100
1200
1300
1400
1500
1600
1700
Bod
y W
eigh
t (g)
Females Males8.0
8.4
8.8
9.2
9.6
10.0
10.4
10.8
Cul
men
Len
gth
(cm
)
Females Males37.6
38.4
39.2
40.0
40.8
41.6
Win
g C
hord
(cm
)
(a) (b) (c)
8
DISCUSSION Our results show that calls provide a fast, reliable, and inexpensive method for sexing adult Peruvian
Boobies in the field. The marked difference in voice between sexes is unmistakable and easily
recognizable by the human ear: high-pitched whistles in males, louder trumpet-like grunts in females.
Dimorphic calls are also given by Blue-footed, Masked (S. dactylatra) and Nazca Boobies (S.
granti)(Anderson 1993, Zavalaga et al. 2007), but call differences are ambiguous (at least to the human
ear) in Red-footed and Abbott’s Boobies, and the three species of gannets (Nelson 1978). This inter-
specific pattern of vocalizations accords with the molecular phylogeny of sulids, where the five species of
boobies with distinct vocal sexual differences are more closely related to each other than to Red-footed or
Abbott’s Boobies, or the three species of gannets (Friesen and Anderson 1997). The proximate
explanation for the sexually dimorphic calls is the anatomical difference of the syrinx between adult males
and females (Murphy 1936).
The Peruvian Booby is a relatively small sulid, only slightly larger than Brown and Red-footed
Boobies (Nelson 1978). In comparison to other Sula boobies, the degree of SSD of Peruvian Boobies is
intermediate between the highly dimorphic Blue-footed and Brown boobies, and the less dimorphic
Nazca, Masked, and Red-footed Boobies. The disparity in size between sexes in Peruvian Boobies found
on the Lobos Islands was large enough to separate the sexes in 92% of the cases by simultaneously
using body mass, culmen length and wing chord; however, regional variation in measurements occurs in
several species of sulids (Nelson 1978, Ropert-Coudert 2005), and the accuracy of our discriminant
equation may not necessarily be the same for other locations.
In a comparative study of SSD of 99 populations of seabirds (involving 33 species) in the
Southern Hemisphere, Fairbairn and Shine (1993) found that the degree of SSD was correlated with
marine productivity. Males were much larger than females in areas with high carbon fixation, whereas the
reverse pattern was observed in unproductive waters of the tropics. They postulated that this pattern
could be explained by three main hypotheses: (1) the extra body reserves of larger females would buffer
low food conditions of the impoverished waters of the tropics during incubation, (2) the scarcity and patchy
distribution of food and the more elusive behavior of prey in the tropics may favor small body size of
males by enhancing aerial agility, and (3) the relatively low nest density of tropical seabirds may reduce
9
nest defense and consequently the selective pressure for larger size in males is weak. Peruvian Boobies
appear to have evolved under the unpredictable, but generally enriched, upwelling waters off the coast of
Peru and northern Chile. In comparison to other seabirds, they show a significant degree of SSD, but
unlike their tropical congeners, they (1) have short feeding trips, (2) nest in colonies around which food is
relatively abundant for has been historically abundant, except during strong ENSO years), and (3) nest in
dense colonies where nest defense is likely important. Why then is the female-larger SSD exhibited by
Peruvian Boobies maintained in the highly productive, essentially extra-tropical waters of the Humboldt
Current? A possible explanation is that Peruvian Boobies have recently arisen from a tropical ancestor
with a high degree of SSD, and that the observed SSD reflects this ancestry and recent selection reducing
the SSD. The molecular phylogeny of sulids seems to support this hypothesis as Blue-footed and
Peruvian Boobies are recently diverged sister species (Friesen and Anderson 1997) that hybridize in
areas of sympatry (Ayala 2006, Figueroa and Stucchi 2008). Thus, the lower dimorphism of Peruvian
Boobies (19%) in comparison to Blue-footed Boobies (31%) may reflect an intermediate stage of evolution
toward monomorphism following colonization of the rich Humboldt Current, rather than an equilibrial SSD
related to the selective environment of this ecosystem.
10
REFERENCES Anderson DJ (1993) Masked Booby (Sula dactylatra), In: The Birds of North America Online (A. Poole,
ed.). Cornell Lab of Ornithology, Ithaca, NY. Retrieved from the Birds of North America Online:
http://bna.birds.cornell.edu/bna/species/073.
Ayala L (2006) Apparent hybridization between Blue-footed (Sula nebouxii) and Peruvian (Sula variegata )
Boobies on Lobos de Tierra islands, Peru. Marine Ornithology 34: 81-82.
Catry P, Philips RA, Furness RW (1999) Evolution of reversed size dimorphism in skuas and jaegers. Auk
116: 158-168.
Fairbairn J, Shine R (1993) Patterns of sexual size dimorphism in seabirds of the Southern Hemisphere.
Oikos 68: 139-145.
Figueroa J, Stucchi M (2008) Possible hybridization between Peruvian Booby (Sula variegata) and Blue-
footed Booby (Sula nebouxii) in Lobos de Afuera Islands, Peru. Marine Ornithology 36: 75-76.
Fridolfsson AK, Ellegren H (1999) A simple and universal method for molecular sexing of non-ratite birds.
Journal of Avian Biology 30: 116-121.
Friesen VL, Anderson DJ (1997) Phylogeny and evolution of the Sulidae (Aves: Pelecaniformes): a test of
alternatives modes of speciation. Molecular Phylogeny and Evolution 7: 252-260.
Friesen VL, Montevecchi WA, Baker AJ, Barret RT, Davidson WS (1996) Population differentiation and
evolution in the Common Guillemot (Uria aalge ). Molecular Ecology 5: 793-805.
Gilardi JD (1992) Sex-specific foraging distribution of Brown Boobies in the eastern tropical Pacific.
Colonial Waterbirds 15: 148-151.
Lewis S, Benevenuti S, Dall’Antonia L, Griffiths R, Money L, Sherrat TN, Wanless S, Hamer KC (2002)
Sex-specific foraging behavior in a monomorphic seabird. Proceedings Royal Society London B.
269: 1687–1693.
Lewis S, Schereiber EA, Daunt F, Schenk GA, Orr K, Adams A, Wanless S, Hamer KC (2005) Sex-
specific foraging behaviour in tropical boobies: does size matter? Ibis 147: 408-414.
Maness TJ, Westbrock MA, Anderson DJ (2007) Ontogenic sex ratio variation in Nazca Boobies ends in
male-biased sex ratio. Waterbirds 30:10-16.
Manly BFJ (2005) Multivariate statistical methods: a primer. Chapman and Hall/CRC Press, Boca Raton,
FL.
Murphy RC (1936) Oceanic Birds of South America Vol 2. MacMillan Company, New York, NY.
Nelson JB (1978) The Sulidae: gannets and boobies. Oxford University Press, Oxford, UK.
Redman K, Lewis S, Griffiths R, Wanless S, Hamer KC (2002) Sexing northern gannets from DNA,
morphology and behavior. Waterbirds 25: 230-234.
Ropert-Coudert Y , Grémillet D, Gachot-Neveu H, Lewis S, Ryan PG (2005) Seeking dimorphism in
monomorphic species: the lure of the gannet’s mask. Ostrich 76: 212-214.
SAS Institute. 2004. SAS/STAT User’s guide Version 9.1. SAS Institute, Cary, NC.
11
Zavalaga CB, Benevenuti S, Dall’Antonia L, Emslie S (2007) Diving behavior of Blue-footed Boobies (Sula
nebouxii) in northern Peru in relation to sex, body size and prey size. Marine Ecology Progress
Series 336: 291-303.
12
CHAPTER 2: AT-SEA MOVEMENT PATTERNS AND DIVING BEHAVIOR OF PERUVIAN BOOBIES Sula variegata : SEXUAL SEGREGATION BY FORAGING HABITAT IN A RICH
MARINE ENVIRONMENT?
13
ABSTRACT Sex differences in body size and morphology are responsible for habitat segregation in several seabird
species inhabiting temperate-polar and tropical regions. This hypothesis has been poorly documented in
species exploiting highly productive coastal upwelling systems, where the predictability and abundance of
food resources is usually higher than in other ocean basins. The at-sea movement patterns and diving
behavior of the sexually dimorphic Peruvian Booby (Sula variegata, females 19% heavier than males)
were studied on two Peruvian islands (Lobos de Tierra and Lobos de Afuera) located within the
boundaries of the Humboldt Current, one of the most productive marine ecosystems in the world. The
foraging behavior of Peruvian Boobies was characterized, compared it with what is known from other
tropical sulids, and evaluated to test if females and males exploit different foraging habitats as a result of
size dimorphism. Peruvian Boobies foraged during daylight hours only, 1 - 3 times a day (median 2 trips)
of short duration (median 1.8 h, max 5 h). Overall, 92% of the total foraging time was spent flying, and
consequently sitting on the water (6%) and diving (2%) were activities of short duration. They fed
exclusively on anchovies (Engraulis ringens), which were captured in shallow dives (median 2.5 m, max
8.8 m), and with a median high rate of 11 dives/h (max 37 dives/h). The foraging range varied between
4.5 and 68 km (median 25 km), whereas the total distance traveled in the foraging path ranged from 14 to
179 km (median 69 km). Foraging areas (identified by dive events) were located 1 – 67 km from the
colonies (median 21 km). There were no sex-specific differences in 13 of 15 variables involving timing of
foraging, movement patterns, at-sea activities, home range and foraging areas. However, females dived
slightly deeper and spent a larger proportion of the foraging time sitting on the water. These results
suggest that (1) the foraging behavior of Peruvian boobies markedly contrast with that of tropical boobies
probably as a result of the proximity and predictability of food sources, high energetic demands of the
brood (up to 4 chicks), and higher prey encounter in the Peruvian coastal upwelling system, and (2) the
lack of spatial segregation between females and males may be related to the attraction of birds from both
sexes to conspicuous multi -species feeding aggregations that are recurrently formed in areas close to the
booby colonies. Once the foraging patches are localized, females dive slightly deeper as a result of
passive mechanisms associated to a heavier mass.
14
INTRODUCTION Foraging site fidelity is higher in temperate-polar species because areas of enhanced marine productivity,
at least at a meso or larger scale (>100 km), are more predictable in space and time than in tropical
waters (Hunt and Schneider 1987, Hyrenback et al. 2002, Weimerskirch 2007). Within the Sulidae family,
there are marked differences between the foraging range of tropical boobies and temperate gannets. The
maximum foraging distance of the Red-footed (Sula sula) and Masked Booby (S. dactylatra) seems to be
restricted by the capacity of birds to forage only during daylight hours, to avoid sitting on the water during
the night, and to the dependence of diurnal sub-surface predators, such as tunas and dolphins, that bring
booby’s prey to the surface (Au and Pitman 1986, Ballance et al. 1997, Jaquemet et al. 2004).
Conversely, temperate gannets usually travel further distances from their colonies because they can
extend the duration of the feeding trip by spending the night at sea (Hamer et al. 2001, Grémillet et al.
2004, Adams and Navarro 2005). In comparison to tropical species, the foraging behavior of temperate-
polar marine birds has been more extensively studied (review in Weimerskirch 2007). However, tracking
studies for seabirds inhabiting low-latitude upwelling regions are poorly documented.
The most productive coastal upwelling ecosystem in the world occurs within the boundaries of the
Humboldt Current off the west coast of Peru and Chile. The annual primary production in this region is
estimated in 613 – 2,285 gC/m2 (Barber et al. 1985, Daneri et al. 2000), which converges mainly in a single
species: the Peruvian anchovy (Engraulis ringens) or anchoveta. The marine productivity of the cold
Humboldt Current is unpredictably interrupted by warm episodes linked to El Niño Southern Oscillation
(ENSO). Low anchoveta availability during ENSO events leads to drastic changes in the local marine
community, usually reflected in low breeding success and high mortality rates of the anchoveta predators
(Arntz and Fahrbach 1996). The anchoveta is a keystone species in this ecosystem, supporting one of the
world’s largest single-species fisheries (Pauly et al. 2002) and a high diversity of marine life, including 12
endemic species of seabirds (Crawford et al. 2006). The Peruvian Booby (Sula variegata) is recognized
as one of the most abundant seabird of the Humboldt Current system (Jahncke et al. 1998, Weichler et al.
2004), with a population size in Peru of approximately 3.5 - 4 millions birds (Birdlife International 2008).
They feed almost exclusively on anchovetas (Tovar and Guillén 1988, Jahncke and Zileri 1998, Janhcke
and Goya 2000), which they obtain by rapid, shallow plunge dives (Nelson 1978). The dependence on this
prey is reflected by population crashes during ENSO events (Duffy 1983a) and overfishing (Jahncke et al.
15
2004). One of the most remarkable life-history traits of Peruvian Boobies is their high rate of reproduction
as they have the largest average clutch size (up to 4 eggs) within the Sulidae (Nelson 1978). During years
of high food supply, Peruvian Boobies can raise up to four chicks per brood (Murphy 1936, pers. obs.),
and hypothetically are able to breed twice in a year (Nelson 1978). This breeding trait contrasts with the
low fecundity of most seabirds, particularly with tropical pelagic species, and has probably evolved under
the rich environment of the Humboldt Current to compensate the poor reproductive output and adult
mortality during strong ENSO events (Murphy 1936, Nelson 1978). Different aspects of seabird foraging
behavior reflect the predictability, distribution and abundance of food resources (Weimerskirch 1997), and
therefore, we speculate that the foraging behavior of Peruvian boobies must differ substantially from their
tropical counterparts.
The Peruvian Booby is a relatively small sulid, with females 19% heavier and 3 - 4% larger than
males (Chapter 1). In comparison to other Sula boobies, the degree of sexual size dimorphism of
Peruvian Boobies is intermediate between the highly dimorphic Blue-footed (S. nebouxii) and Brown
Boobies (S. leucogaster), and the less dimorphic Nazca (S. granti ), Masked, and Red-footed Boobies
(Nelson 1978). In a comparative study of sexual size dimorphism of 99 populations of seabirds (involving
33 species) in the Southern Hemisphere, Fairbairn and Shine (1993) found that the degree and direction
of size dimorphism between sexes was correlated with marine productivity. Males were much larger than
females in areas with high carbon fixation, whereas the reverse pattern was observed in unproductive
waters of the tropics. One of the hypotheses that explain this female-based size dimorphism is that the
scarcity and patchy distribution of food and the more elusive behavior of prey in the tropics may favor
small body size in males by enhancing aerial agility (Fairbain and Shine 1993). This hypothesis has
received support in the Red-footed Booby (females 15% heavier than males), where males fly faster and
occupy different foraging areas than females (Weimerskirch et al. 2006). The degree of size disparity
between female and male Peruvian Boobies is higher than in Red-footed Boobies, suggesting that the
dimorphism may be advantageous in reducing intersexual food competition by ecological niche
partitioning. Nevertheless, at-sea spatial segregation between sexes has not yet been examined in low-
latitude upwelling regions, where the structure of prey distribution and multi -species flock assemblages
differ from the conditions in the tropics. For instance, multispecies-flock formation in the tropics are
usually catalized by sub-subfarce predators like tunas that bring prey to the surface (Au and Pitman 1986,
16
Jaquemet et al. 2004), whereas the large feeding aggregations of seabirds in the Peruvian coastal
upwelling area are formed generally by other seabirds. The frequency and duration of these assemblages
can only be supported by the patchy distribution and sufficiently abundant schools of anchoveta (Duffy
1983b). Likewise, these multi-species flocks occur relatively close to the seabird breeding colonies (Duffy
1983b), where commuting birds can be attracted to feeding frenzies by local enhancement (Davoren et al.
2003). Thus, female and male Peruvian boobies may join existing feeding aggregations and consequently
forage in the same areas.
In this paper, I report the first use in Peru of precision GPS logging technology in conjunction with
depth meter tags to study the at-sea movement patterns and diving behavior of Peruvian boobies from
Isla Lobos de Tierra (inshore) and Isla Lobos de Afuera (offshore). The main goals were (1) to evaluate
the foraging behavior in a rich upwelling region and compare it with what is known from other tropical
boobies, (2) to examine whether females and males exploit different foraging habitats as a result of size
dimorphism, and (3) to identify foraging behavior differences of birds breeding in an inshore and offshore
island. We hypothesize that under the rich Humboldt Current (1) the distances traveled to the feeding area
will be shorter and the degree of foraging site fidelity higher in Peruvian than in tropical boobies, and (2)
no spatial segregation between sexes will be detected despite the sexual size dimorphism in Peruvian
Boobies because birds may potentially locate foraging areas by visual clues from multi-species flocks
foraging in areas close to the colonies.
17
METHODS
Study site
The foraging behavior of breeding Peruvian Boobies was studied on two islands in northern Peru: Lobos
de Tierra (LT; 6o24’S, 80o51’W) between 22 and 30 December 2006, and Lobos de Afuera (LA; 6o57’S,
80o41’W) between 10 and 17 December 2007. A description of the topography and fauna of LT and LA is
given by Zavalaga et al. (2007) and Figueroa and Stucchi (2008), respectively. Briefly, LT is an inshore
inland (area = 1426 ha) located 15 km from the continent, whereas LA includes two offshore islands
(Independencia and Cachimbo, total area = 236 ha) located 61 km to the nearest point on the mainland
and 65 km south of LT. Both islands lie over the continental shelf and within the boundaries of the cold
waters of the Humboldt Current. More specifically, the study plots were located approximately 500 m west
of the island’s dock on LT, and 200 m south of the navy meteorological station on LA. On both islands,
Peruvian Boobies nested in several dense groups (ca. 2 nests·m-2) of 100 – 3,000 nests each. During the
study period, the breeding population (estimated form direct counts) was 2,000 – 2,500 nests on LT and
5,500 – 6,000 on LA (Isla Independencia only). T he majority of nests contained half-grown chicks (75 % of
the body covered with downy feathers). Modal brood size of studied birds was 2 chicks on LT (range: 1 –
4, N = 14 nests) and 3 on LA (range: 1 – 4, N = 37 nests).
Description of dataloggers
The movement patterns of Peruvian Boobies were determined by using two types of GPS dataloggers
(www.technosmart.eu) equipped with an integrated antenna and a 1 Mb flash memory, but that differed in
size and recording time interval. The GiPSy-1 (33 g; 5 x 3 x 1 cm) was set to take one position fix every 10
s for approximately 48 h. The GiPSy-2 (14 g, 4.5 x 2.2 x 0.7 cm) had a time interval between fixes set at 1
s for approximately 14 h. The GPS battery lifespan was long enough to record more than one feeding trip
per deployment. The accuracy of both loggers was < 10 m in > 95% location fixes. Overall, two thirds of
birds were instrumented with the light GiPSy-2, whereas all other birds were equipped with the GiPSy-1.
The instruments were protected in two heat-sealed waterproof polyethylene bags (1 g). After logger
18
retrieval, the data were downloaded to a laptop computer using the dedicated GiPSy software
(www.technosmart.eu).
The diving behavior and sea-surface temperature at foraging sites were recorded with bullet-
shaped archival tags (5g; 1.1 x 3.2 cm Lotek LTD1110, www.lotek.com), set to collect data every second
for approximately 18 h. The pressure sensor recorded depths with an accuracy of 0.3 m (depth range: 0 –
30 m), whereas the external temperature was measured with an accuracy of 0.1oC (temperature
operational range -5 to +35oC). Data were downloaded using Tag Talk 1100 software (www.lotek.com).
Capture of birds
For logger deployment, a total of 14 and 35 known-sex Peruvian Boobies raising medium-sized chicks
were captured between 0530 and 0700 h at LT and LA, respectively. Birds were randomly selected from
the second or third raw of nests from two neighboring breeding groups (ca. 100 nests each) at LT and one
breeding group at LA (ca. 3000 nests). Birds were removed from their nests using a hook attached to a 3-
m pole, placed gently around their necks. The GPS dataloggers were deployed using four strips of
waterproof Tesa tape wrapped around four central tail feathers. The depth/temperature tags were
attached to a numbered metal band with three small plastic cable ties. Attachment of the loggers was
completed in < 5 min., and birds were released near their nests. Twelve birds at LT and 24 at LA were
equipped with a GPS-logger in conjunction with a depth/temperature tag (Fig. 2) in order to identify
foraging areas from the location fixes of dive events. T he rest of birds carried a GPS only. The GPS
dapataloggers, the depth/temperature tags and their accessories (two plastic bags, 4 fourstrips of tesa
tape, metal band and three plastic cable ties) added a maximum total weight to each bird of 30 g (with
GiPSy-2) or 50 g (with GiPSy-1). The range of body weight recorded from all tagged birds was 1200 –
1675 g, and therefore the GiPSy-1 and GiPSy-2 loggers accounted for a maximum 1.8% - 4.2% of the
bird’s body weight (mode = 2%), respectively, below the 5% limit recommended by the Ornithological
Council Guidelines (www.nmnh.si.edu/BIRDNET/GuideToUse, October 2008). The sex of tagged birds was
recognized by their vocalizations (goose-like honks in females and whistles in males, Chapter 1).
Because birds may compensate for any increase in foraging effort by their mates (Paredes et al. 2005),
only one member of the pair was equipped to avoid possible data pseudoreplication. T he tagged boobies
19
were marked with a short-lasting dye (rhodamine B) on the chest and head for easy sighting of birds in a
dense colony (Fig. 2). Boobies were recaptured 12 – 48 h later (82 % of birds recaptured between 10 - 14
h) to retrieve the devices, weighed with a spring Pesola scale to the nearest 25 g, and induced to
regurgitate by holding them upside down and pressing gently on their bellies until all the food has passed.
The identification and number of prey was assessed in situ because the majority of the regurgitations
were undigested.
Figure 2. Breeding Peruvian Booby Sula variegata with two medium-sized chicks at Isla Lobos de Tierra.
This bird was tagged with a GPS datalogger attached to the tail feathers with 4 strips of TESA tape, and
with a depth/temperature archival tag tied to a metal band with 3 plastic cable ties. The adult booby was
also marked with a small red dot of rhodamine-B in the chest.
GPS-logger
Depth/Temp-logger
20
Handling of data and analyses
Depth/temperature loggers.- The number, time, depth, duration and profile of dives as well as sea surface
temperature (SST) at foraging sites were calculated for each trip using Multitrace MT-dive software
(Jensen Software Systems, Kiel, Germany). All dives with depths < 0.5 m were excluded because they
may represent bathing splash immersions. Peruvian Booby dives were usually short and shallow (see
results) in relation to the sampling interval of 1 sec, leading to possible errors on the estimations of dive
parameters. To improve the accuracy of the calculations, an additional point was inserted between two
recorded consecutive points by using the interpolation option in Mt-Dive. To estimate the SST at the
foraging sites, we selected the minimum SST for each dive event.
GPS-dataloggers.- The spatial data were mapped and analyzed using the ArcGIS 9.2 Geographic
Information Systems (ESRI Inc., Redlands, CA). The timing of foraging trips, flight speed, sinuosity index,
trip length, at-sea activities and movement patterns were calculated from the longitude/latitude points. T he
position fixes were projected in the Universal Transverse Mercator (Zone 17M) coordinate system for all
spatial analysis. The ArcGIS extensions, Hawth’s Analysis Tools (http://www.spatialecology.com/htools) and
Xtools Pro 5.2 (Data East LLC 2007, http://www.xtoolspro.com) were also applied for estimations of
bearings, distances and areas.
The duration of the foraging trip was defined as the interval between the departure time from the
nest to the landing time on the island. Although the majority of birds landed on their nests after foraging, a
small proportion landed away from the colony, returning to their nests several hours after landing. Thus,
using the nest as a central place for foraging would overestimate the duration of some trips. Departure
and return times for each trip were identified by visual inspection of all data points in the track. Once the
departure and arrival times were established, all positional fixes on land were excluded for subsequent
analysis of spatial orientation, home range and foraging areas.
Instantaneous speeds were calculated from the distance and time elapsed between two
consecutive position fixes in a foraging track. However, an inspection of the frequency distribution of these
records revealed a clear cut-off value, with speeds <10 km/h represented by birds resting on the sea
surface or diving, and speeds > 10 km/h corresponded to flying birds (Fig. 3). A similar bimodal
21
distribution of speeds has been found in other species of Sulids (Grémillet et al. 2004, Weimerskirch et al.
2005). Thus, for the calculations and analysis of flight speeds we used values > 10 km/h.
Path sinuosity was defined as the ratio of the total distance traveled in 1-min interval to the
straight line distance in that interval (every 60th fix for GiPSy 2 and 6th fix for GiPSy-1). A sinuosity index
close to one indicates high path linearity . To examine the directionality of the tracks at different sections of
the foraging trip, the sinuosity index was calculated in three stages of equal duration: beginning (outbound
commute), middle (feeding), and ending (inbound commute, Hyrenbach et al. 2002).
The proportion of time flying in relation to the total duration of the feeding trip was estimated from
the cumulative time of all data points with speeds > 10 km/h. Consequently, the proportion of time on the
water (sitting, diving) was estimated from speeds < 10 km/h. The number of dives and dive duration was
determined from the Lotek depth/temperature tags, and hence, it was possible to separate the time
interval for sitting on the sea surface from the total time spent diving. Thus, at-sea activities were finally
categorized into three different activities: flying, sitting on the sea surface and diving.
To describe and compare the movement patterns between sexes and islands, individual tracks
were reduced to four main components: (1) the azimuth angle between the nest and the outmost foraging
point, (2) the maximum foraging range, defined as the straight line distance between the nest location and
the farthest position fix in the route, (3) the cumulative distance travelled in the foraging path, and (4) the
nearest distance of the foraging track to the mainland. To determine whether Peruvian Boobies dived
continuously along the foraging path or fed at specific areas, a standardized index of dive locations was
calculated in relation to the point of the maximum foraging range as follows:
SDLij = [(Dmax j – Ddivej)/Dmax j],
where SDLij = is the standardized distance of the ith dive in a jth foraging trip, Dmax j = maximum foraging
range of the trip j, and Ddive i = distance of the dive location i from the nest. A value close to 0 indicated
that dives occurred close to the maximum foraging range, whereas a value close to 1 indicated that dive
locations were close to the island. If birds dived continuously, then we predicted a uniform frequency
distribution of distances. Conversely, if boobies fed at specific areas before returning to the island, then
the distribution would be skewed, with the majority of dives located around the farthest point in the route.
22
Figure 3. Distribution frequency of instantaneous movement speeds of Peruvian Boobies Sula variegata ,
breeding at Isla Lobos de Tierra in December 2006 (N = 143,601 fixes) and Lobos de Afuera in December
2007 (N = 156,343 fixes). Speeds were derived from all GPS fixes outside the breeding colony.
0
2
4
6
8
0 8 16 24 32 40 48 56 64 72 >80
Movement speed (km/h)
Dis
tribu
tion
frequ
ency
(%
)
20062007
GPS loggers with depth/temperature loggers.- The depth/temperature tags were synchronized with the
GPS-loggers on the island to the nearest 1 sec using the satellite-read time read from a conventional
handheld GPS. Thus, all dives were geo-referenced and the locations used to identify foraging areas. The
extent of the feeding areas was assessed using the fixed kernel density estimation. This procedure not
only tempers the effects of spatial autocorrelation of multiple dive position fixes from an individual (Wood
et al. 2000), but it gives a more accurate estimation of the foraging areas than the minimum convex
polygon calculated from peripheral data points. The kernel analysis estimated a 95% contour probability
polygon from the pooled dive position fixes of females and males at each island. The bandwidth or search
radius employed for the calculation of the 95% contour polygons was determined with the least square
cross validation algorithm (LSCV, Seaman and Powell 1996) using the Animal Movement Analysis
extension (Hooge and Eichenlaub 1997) in ArcView 3.3 (ESRI). The contour polygons around LT
23
intersected land masses because the majority of dives were located in waters close to the mainland.
Thus, for calculations of foraging areas, the intersected polygons on land were removed from the analysis.
Statistical analysis
The data set compiled for each bird included multiple sequential data values within a trip and usually more
than one trip during deployment time, and therefore, they cannot be considered independent. To
circumvent data autocorrelation at the bird level, generalized mixed linear models were applied with
restricted maximum likelihood estimations (REML) for comparisons of linear foraging variables between
females and males, between LT and LA, and the interaction between sex and island. In this model, sex
and island were defined as fixed factors, whereas bird identity was categorized as a random factor.
Continuous variables were tested for homogeneity of variance using the Levene’s test before using
generalized linear models. If the variances were not homogenous or the data were not normally
distributed, then non-parametric tests were used. Thus, the Kolmogorov-Smirnov test (abbreviated K-S)
was performed for comparisons of distributions, and the Mann-Whitney U-test to compare the median of
two unmatched samples with a low number of replicates. The logistic regression analysis was selected for
data with a binary response (e.g. number of trips per day, 1 trip = 1, > 1 trip = 0). Categorical data were
analyzed by the Chi-square test.
Angular data were examined with circular statistical tests (Batschelet 1981). More specifically, the
mean directional vector (r) for each data set (males, females, LT and LA) was estimated to measure
bearing dispersion, with r values close to one indicating that headings were highly concentrated to a
specific location. For birds with multiple trips, the bearing value of each trip was averaged, and the mean
bearing for each bird used for the directional analyses. The Rayleigh’s uniformity test was performed to
calculate the probability of the null hypothesis that the birds’ bearing were uniformly distributed at all
directions, i.e., birds did not flight to specific or preferred locations to forage. To evaluate the difference in
the mean bearing between sexes and islands, we used the Watson’s U2 test. The Watson-Williams
pairwise F test was selected to examine the directional persistence of individuals on two subsequent trips.
Statistical Analysis Systems (SAS Institute 2004) was used for all statistical tests of linear
variables. Circular statistical analyses were performed using Oriana 2 software (Kovac Computing
24
Service). Means are expressed ± 1 s.d. and differences were deemed to be significant at P < 0.05 and
marginally significant at 0.1 < P < 0.05.
25
RESULTS All birds returned to their nests after 0.5 - 2 days of tag deployment with 100% recapture. None of the
loggers were lost, but four GPS-loggers failed to record data (three from LT and one from LA) because
they were waterlogged. All depth meter/temperature tags recorded data. Thus, only foraging data from
three birds equipped with depth/temperature dataloggers only, 12 birds with GPS dataloggers only, and
33 birds with GPS in conjunction with depth/temperature loggers were obtained. The number of
instrumented individuals, trips and data points for each gender as well as island and logger categories are
given in Table 2.
Table 2. Sample size at the bird, trip and data point level used in the analysis of foraging data of breeding
Peruvian Boobies Sula variegata at Isla Lobos de Tierra in December 2006 and Isla Lobos de Afuera in
December 2007. Sample size at each category derived from three birds equipped with depth/temperature
dataloggers only, 12 with GPS dataloggers only, and 33 with GPS with depth/temperature loggers.
Isla Lobos de Tierra 2006 Isla Lobos de Afuera 2007
Females Males Females Males
Depth/temp dataloggers
Number of birds 3 9 15 9
Number of trips 5 13 23 12
Number of data points 265 511 443 243
GPS-dataloggers
Number of birds 4 7 19 15
Number of trips 8 10 32 21
Number of data points (outside the colony) 60,746 82,855 102,889 53,454
GPS + depth/temp loggers
Number of birds 3 7 15 8
Number of trips 5 10 23 12
Number of data points 265 407 443 228
26
Timing of departures and duration of feeding trips
Departures from the colony occurred at any time from sunrise (0600 h local time) to two hours before
sunset (1900 h, Fig. 4). Most of the birds did not leave the colony immediately before or after dawn (0600
h), but departed mainly between 0900 -1000 h and between 1500 - 1600 h. There were no sex-specific
differences in the time of departures (K-S, D = 0.20, P = 0.766). The bimodal distribution of the departure
time was the result of birds having single or multiple trips in a day (51% of birds had double, 44% single,
and 5% triple trips, N = 43, Fig. 4). The number of trips in a day was similar between sexes and locations
(Table 3). Birds that departed before noon had greater chances of having multiple trips per day (Logistic
regression, odd ratio = 1.93, Wald-χ2 = 6.59, P = 0.01). Peruvian Boobies never spent the night at sea,
but always returned to the island before nightfall (1900 h). From a sample of 11 birds at LT with GPS-
loggers > 24-h, 64% of them attended their nests overnight, whereas the rest spent the night on the
island, but 0.1 – 5.5 km from the nest.
Figure 4. Frequency distribution in the departure time of breeding Peruvian Boobies Sula variegata,
categorized by the number of trips that birds undertook in a day. Data from Isla Lobos de Tierra in 2006
and Isla Lobos de Afuera in 2007 were pooled (N = 43 trips).
0
5
10
15
20
25
30
Time of departure (local time hh)
Per
cent
age
of th
e to
tal n
umbe
r of t
rips Triple
DoubleSingle
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
27
Trips were usually short (median = 1.86 h, range = 0.4 - 5), but significantly longer in birds from
LT than LA. The level of significance for the interaction between sex and study site was marginal,
indicating that females spent more time at sea than males at LT, but not at LA (Table 3).
Diving behavior
Overall, 88% (N = 1462) of the immersions occurred at depths < 4 m and lasted < 6 sec, with median
depth and duration of 2.5 m (max. = 8.81 m) and 4 sec (max. = 19.7 sec), respectively. Dive depth and
duration was significantly correlated: Duration(sec) = 1.98 + 0.9 x depth(m) (Regression, r2 = 0.40, n =
1462, P < 0.0001). The dive depth differences between sexes and locations approached significance (P =
0.06, Table 3), with females diving slightly deeper than males (depths > 5.5. m only attained by females,
Fig. 5), and birds attaining greater depths on LT than on LA (Table 3).
Two types of immersions were recognized from the dive profiles: V-shaped dives, when the bird
attained the maximum depth returning immediately to the surface, and U-shaped dives, when the bird
spent some seconds at the maximum depth before returning to the surface. V-dives were more common
(92% of all dives, n = 1462) than U-dives. No sexual differences in the U/V ratio were significant (Table 3).
However, birds from LA performed a higher number of U-dives in relation to V-dives than birds from LT
(Table 3).
The total number of dives in a trip was similar between sexes (Table 3), but birds from LT dived
2.5 times more frequent in a trip (median = 43 dives, range = 19 – 68) than birds from LA (median = 17
dives, range = 5 – 43, Table 3). When the total number of dives in a feeding trip was divided by the
duration of the trip, the number of dives per hour was similar between locations and between males and
females (median = 13 dives/h, range 3 – 37, Table 3). Likewise, the number of dives per hour estimated
from the time interval between the first and the last plunge in a trip was similar between sexes and
locations (median = 29.4 dives/h, range = 8.4 – 66.19, Table 3).
28
Figure 5 Frequency distribution of dive depths in female and male Peruvian Boobies Sula variegata,
breeding on Isla Lobos de Tierra in Dec 2006 and Isla Lobos de Afuera in Dec 2007 (N = 1462 dives
pooled data).
0
2
4
6
8
10
12
14
16
18
Dive depth (m)
Per
cent
age
of th
e to
tal n
umbe
r of
div
es FemalesMales
0 1 2 3 4 6 7 8 95 8
At-sea activities
Within a foraging trip, Peruvian boobies spent a high proportion of time flying (92%, N = 50 trips). Sitting
on the sea surface was usually rare (6% of the time), whereas the time spent diving only represented 2%
of the total trip duration. Birds from LT spent a significantly larger proportion of time sitting on the water,
and consequently a smaller proportion of time flying than birds from LA (Table 3). The percentage of time
sitting on the water was marginally significant between sexes and for the interaction of sex and islands
(Table 3), with females spending almost twice the amount of time resting on the sea surface than males
on LT, but not on LA (Table 3).
29
Flight ground speeds and sinuosity of paths
The average traveling speed ranged from 39 to 46 km/h, with burst speeds between 90 and 139 km/h.
There were no sexual differences in flight speeds, but birds from LA flew significantly faster than birds
from LT (Table 3).
There were not significant differences in the sinuosity index between sexes and islands (Table 3),
and the paths of birds tended to be direct when birds commuted to foraging areas and returned to the
colony. The sinuosity index was significantly lower and less variable during the beginning (outbound,
mean =1.49 ± 1.23, N = 50 birds) and ending (inbound, mean = 1.48 ± 1.1, N = 50) than during the middle
(feeding, mean = 1.80 ± 1.54, N = 50) section of the foraging trip (Fig. 6), indicating strong directionality in
the travel paths. No multiple interaction effects between sex, island and path section were significant (all
interactions P > 0.20).
Figure 6. Sinousity index at the three sections of the foraging tracks of Peruvian Boobies Sula variegata
(beginning = outbound, middle = feeding, ending = outbound) on (a) Isla Lobos de Tierra in December
2006 and (b) lsla Lobos de Afuera in December 2007. Box plots depict the 5, 10, 25, 50, 75, 90 and 95
percentiles of the distributions.
(b) (a)
30
Table 3. Sex-specific differences in the number of trips in a day, trip length and diving behavior of Peruvian Boobies Sula variegata breeding at Isla
Lobos de Tierra in 2006 and Isla Lobos de Afuera in 2007. Descriptive statistics are expressed by the mean ± 1 SD and range in parentheses.
Isla Lobos de Tierra (2006)
Isla Lobos de Afuera (2007)
Main Effect statistics
Variable
Females
Males
Females
Males
Sex
Location
Sex x Location
interaction
Number of trips per day
1.6 ± 0.55
(1 – 2)
1.5 ± 0.53
(1 – 2)
1.72 ± 0.67
(1 – 3)
1.46 ± 0.51
(1 – 2)
χ2 = 0.54 (1)
P = 0.46
χ2 = 0.006 P = 0.93
Trip length (h)
3.36 ± 0.55 (2.56 – 4.27)
2.46 ± 0.65 (1.54 – 3.45)
1.61 ± 1.05 (0.38 – 4.92)
1.88 ± 0.83 (0.69 – 4.03)
F1,41 = 1.51 P = 0.226
F1,41 = 17.5 P < 0.001
F1,41 = 3.80
P = 0.06 Proportion of time flying (%)
81 ± 9
(73 – 95)
90 ± 6
(78 – 98)
94 ± 6
(79 – 99)
94 ± 8
(68 – 99)
F1, 29 = 2.87 P = 0.101
F1, 29 = 11.8 P = 0.002
F1, 29 = 3.52 P = 0.071
Proportion of time sitting on the water (%)
15 ± 9 (3 – 25)
8 ± 6
(1 – 20)
4 ± 6
(0 – 20)
4 ± 7
(1 - 31)
F1, 41 = 3.22
P = 0.08
F1, 41 = 13.3 P < 0.001
F1, 41 = 2.96 P = 0.093
Dive depth (m)
2.79 ± 1.46 (0.56 – 7.51)
2.63 ± 1.07 (0.57 – 5.37)
2.64 ± 1.10 (0.59 – 8.81)
2.21 ± 0.96 (0.56 – 5.20)
F1, 32 = 3.81
P = 0.06
F1, 32 = 3.38
P = 0.08
F1, 32 = 0.63
P = 0.43 Dive duration (sec)
4.08 ± 1.36 (2 – 16.02)
4.28 ± 1.43
(2 – 19)
4.65 ± 2.08 (2 – 19.72)
4.12 ± 1.44
(2 – 10)
F1, 32 = 0.15
P = 0.69
F1, 32 = 0.01
P = 0.97
F1, 32 = 1.51
P = 0.22 Number of dives in a trip
53 ± 13.47 (31 – 68)
39.3 ± 13.9
(19 – 63)
19.26 ± 9.94
(5 – 43)
18.69 ± 7.61
(9 – 39)
F1, 32 = 3.54
P = 0.07
F1, 32 = 48.3
P < 0.001
F1, 32 = 2.48
P = 0.12 U/V dives ratio
0.055
0.080
0.088
0.151
χ2 = 2.11(2) P = 0.147
χ2 = 4.57 P = 0.03
Total dive rate (dives/h)
15.15 ± 3.53 (9.74 – 19.29)
15.78 ± 5.18 (6.78 – 25.17)
14.19 ± 8.87 (3.11 – 37.45)
10.44 ± 5.33 (3.61 – 22.76)
F1, 32 = 0.36
P = 0.55
F1, 32 = 1.48
P = 0.23
F1, 32 = 0.50
P = 0.48
31
Table 3 continued
Foraging dive rate (dives/h)
22.97 ± 4.61 (18.47 – 29.8)
28.17 ± 7.23 (16.67 – 42.4)
35.57 ± 18.3 (8.39 – 66.19)
30.71 ± 14.3 (11.21 – 55.7)
F1, 32 = 0.10
P = 0.75
F1, 32 = 2.23
P = 0.15
F1, 32 = 1.24
P = 0.27
Flight speed (km/h)
39.3 ± 3.5
(10 – 139.5)
40.9 ± 2.6
(10 – 111.3)
44.8 ± 3.3
(10 – 128.8)
45.7 ± 2.9 (10 – 90.2)
F1,41 = 1.23
P = 0.27
F1,41 = 26.3 P < 0.0001
F1,41 = 0.20
P = 0.66
Sinuosity index
1.69 ± 1.35
(1.02 – 15.73)
1.66 ± 1.26
(1.01 – 15.65)
1.59 ± 1.43 (1 – 16.64)
1.48 ± 1.16 (1.01 – 15.7)
F1, 41 = 0.74 P = 0.394
F1, 41 = 0.92 P = 0.344
F1, 41 = 0.84 P = 0.364
Maximum foraging distance (km)
31.1 ± 12
(14.3 – 50.5)
29.8 ± 12.1 (16.4 – 48.4)
23.9 ± 17.7 (4.5 – 67.9)
29.3 ± 14.5 (9.9 – 58.6)
F1,41 = 0.04
P = 0.84
F1,41 = 0.74
P = 0.39
F1,41 = 0.30
P = 0.58
Total foraging path (km)
106.9 ± 19.2 (72.6 – 126.9)
89.7 ± 24.1 (58 – 124.7)
65.6 ± 44.3
(14.9 – 178.7)
76.6 ± 36.4 (22.8 – 175)
F1,41 = 0.21
P = 0.65
F1,41 = 5.35 P = 0.026
F1,41 = 1.09
P = 0.30 (1) Logistic regression, (2) Chi-square. All other statistical tests using REML Generalized mixed models.
32
Movement patterns
Foraging tracks of Peruvian Boobies were characterized by either straight or looping routes. Straight
tracks were more common (72% of all trips, N = 71) than loops and consisted of parallel outbound and
inbound flights with feeding activities usually at the farthest end point of the route (Fig. 7a). Looping tracks
were wider and birds fed at more than one occasion throughout the trip (Fig. 7b). Individuals tended to fly
in the same direction in consecutive trips both at LT (Watson-William pairwise F-test, F1,12 = 0.018, P =
0.894) and at LA (Watson-William pairwise F-test, F1,38 = 0.601, P = 0.443), with a bearing difference of <
45o between two consecutive trips in 43% (N = 7 birds) and 63% (N = 19) of cases at LT and LA,
respectively.
Figure 7. Straight (a) and looping (b) foraging tracks of two breeding Peruvian Boobies Sula variegata,
equipped with GPS dataloggers in tandem with depth meters at Isla Lobos de Tierra in December 2006.
White dots indicate dive locations and arrows the flying direction. Dive events in straight routes occurred
at the farthest end point of the trip, whereas boobies in looping routes dived at more than one location
during the feeding trip. Inset boxes depict details of the path sinuosity.
(a)
33
(b)
The mean foraging direction of females and males was similar at LT (Watson’s U2-test, U2 = 0.097, 0.5 >
P > 0.2, Table 2, Fig. 8a) and LA (Watson’s U2-test, U2 = 0.099, 0.5 > P > 0.2, Table 2, Fig. 8b).
Nevertheless, on LT the directionality of male tracks was lower (r = 0.39) than female tracks (r = 0.74).
There were differences in the orientation of foraging tracks between islands. T he majority of trips from LT
were widely dispersed in an arc from the northeast and southeast of the island (Fig 8a). Although the
orientation of the tracks on LT was significantly different from a random angular distribution (Rayleigh’s
test, Z = 5.31, P = 0.004), the directionality of tracks was relatively low (mean vector r = 0.543) due to the
large variance of bearings (99% C.I. for the mean = 20 – 113o). The orientation of bird’s tracks from LA
was also significantly different from a uniform distribution (Rayleigh’s test, Z = 35.9, P < 0.001). Birds
preferentially travelled to the south and southwest of the island (Fig. 8b), but in contrast to LT the
directionality of tracks was higher (mean vector r = 0.824) and with a smaller angular variance (99% C.I.
for the mean = 176 - 201o).
The total distance traveled in the foraging path and the mean maximum foraging range were also
similar between females and males (Table 3). Overall, 78% of the trips on LT (N = 18) had a destination
34
point at mainland inshore waters (< 1 km from the coastline), often flying parallel to the shoreline for
several kilometers (Figure 8a). Boobies from LA never approached to the mainland (the shortest distance
of any foraging bird to the mainland was 68 km). Birds from LT covered significantly longer path distances
(median = 106 km, range = 58 - 127) than birds from LA (median = 63 km, range = 15 – 179, Table 3).
Because there were no significant differences in the mean radial distance between birds from LT (median
= 31 km, range = 14 – 50.5, Table 3) and birds from LA (median = 21 km, range = 4.5 – 68 km, Table 3),
the marked differences in foraging path distance between islands was attributed to the exploratory
behavior of boobies from LT along the mainland coastline.
Figure 8. Foraging tracks of females (blue tracks) and male (red tracks) Peruvian Boobies Sula variegata ,
breeding at Lobos de Tierra in 2006 (a) and Lobos de Afuera in 2007 (b).
(a)
35
(b)
The duration of the foraging trips was significantly correlated to the total foraging path (Regression,
Foraging path (km) = 5.89 + 35.36 x trip length (h), r2 = 0.89, P < 0.001, Fig. 9) and maximum foraging
distance (Regression, Max distance (km) = 3.31 + 11.87 x trip length (h), r2 = 0.64, P < 0.001, Fig. 9).
36
Figure 9. Association between the duration of the foraging trip and the maximum foraging range and total
travelled distance of Peruvian Boobies Sula variegata breeding at Isla Lobos de Tierra (LT) in December
2006 and Isla Lobos de Afuera (LA) in December 2007.
0
40
80
120
160
200
0 1 2 3 4 5 6
Trip length (h)
Dis
tanc
e (k
m)
Maximum from the colony LT
Total traveled path LT
Maximum from the colony LA
Total traveled path LA
Foraging areas
Dive were not evenly distributed along the foraging path (pooled data, K-S, D = 0.88, P < 0.001), but
Peruvian boobies preferentially foraged close to the farthest point in the route before returning to the
colony, with approximately 50 - 60% of dives (N = 1343) located at a distance equivalent to <10% of the
maximum travelled distance (Fig. 10). Thus for example, 50 - 60% of the dives from a bird that reached a
maximum distance of 30 km from the colony were located < 3 km from the route end point.
The mean distance of dive locations in relation to the colony was not significantly different
between islands and between sexes (median = 21.3 km, range = 0.98 – 67.1, Table 2, Fig 11a and 11b).
Nevertheless, the spatial distribution was bimodal for birds from LA as a result of temporal changes in
foraging behavior (Fig. 11b). Between 10 and 12 December 2007, the foraging areas were located < 25
km from LA, whereas dives that occurred between 13 and 17 December were located 35 – 67 km from the
island.
37
Figure 10. Frequency distribution of standardized location of dives in relation to the farthest point in the
foraging trip of Peruvian Boobies Sula variegata breeding at Isla Lobos de Tierra in 2006 and Isla Lobos
de Afuera in 2007. The standardized dive locations were calculated as: SDLij = [(Dmax j – Ddivej)/Dmax j], where
SDLij = is the standardized distance of the ith dive in a jth foraging trip, Dmax j = maximum foraging distance
of the trip j, and Ddive i = distance of the dive location i from the nest. A value close to 0 indicates that dives
occurred close to the maximum foraging distance, whereas a value close to 1 indicated that dive locations
were close to the island.
0
20
40
60
80
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Standardized dives locations in relation to the maximum distance in the trip
% o
f the
tota
l num
ber
of d
ives
LT
LA
Females from LT dived significantly closer to the mainland shoreline (median 1.66 km, range = 0.27 –
10.93) than did males (median= 8.91 km, range = 0.91 – 38.87 km, K-S, D = 0.51, P < 0.0001), with
approximately 25% and 55% of the total number of female dives (N= 265) located <1 km and < 2 km from
the shore (Fig. 12), respectively. The frequency distribution of dive distance to the mainland for males
from LT was bimodal (Fig. 12) as a result of two males feeding > 22 km from the mainland.
The contour maps revealed that dives of females from LT were enclosed in an area of 392 km2
whereas the area visited by males was 1456 km2. The foraging area of males and females on LT
overlapped, with 74% of the female’s area intersected by the male’s area, and 20% of the male’s area
overlapped by the female’s area (Fig. 13a). Likewise, the extent of foraging areas of females (1401 km2)
38
and males (1398 km2) from LA was very similar, with 45% of the female’s area overlapped by the male’s
area, and 45% of the male’s area included within the female’s area (Fig. 13b).
The SST was significantly colder around LA in December 2007 (mean = 18.5 ± 1.07oC, range =
17 – 22.6) than LT in December 2006 (mean = 19.4 ± 0.64oC, range = 18.3 – 20.5, REML, F1,32 = 4.54, P =
0.04). There were no significant differences in the SST between sexes (REML, F1,32 = 0.03, P = 0.856) or
the interaction of the two factors (REML, F1,32 = 1.07, P = 0.309).
Figure 11. Frequency distribution of dive locations in relation to the colony from female and male Peruvian
Boobies Sula variegata equipped with a GPS- datalogger in tandem with a depth/temperature tag at (a)
Isla Lobos de Tierra in December 2006 and (b) Isla Lobos de Afuera in December 2007.
0
10
20
30
40
0 10 20 30 40 50 60 70
Distance from the colony (km)
% o
f the
tota
l num
ber o
f div
es
MalesFemales
0
10
20
30
0 10 20 30 40 50 60 70
Distance from the colony (km)
% o
f the
tota
l num
ber
of d
ives
MalesFemales13 - 17 Dec10 - 12 Dec
(a) (b)
39
Figure 12. Frequency distribution of the dive locations in relation to the nearest point to the mainland of
breeding female and male Peruvian Boobies Sula variegata equipped with GPS datalogger in conjunction
with a depth/temperature tag at Lobos de Tierra in December 2006.
0
10
20
30
40
0 4 8 12 16 20 24 28 32 36
Distance of dives to the nearest point on the mainland (km)
% o
f the
tota
l num
ber o
f div
es
MalesFemales
Figure 13. Foraging areas of female (grey), male (white) breeding Peruvian Boobies Sula variegata
instrumented with GPS dataloggers and a depth/temperature tags at (a) Isla Lobos de Tierra in 2006 and
(b) Isla Lobos de Afuera in 2007. The female and male overlapped foraging area is show in dashed lines.
(a) (b)
40
Diet
At the time of recapture of tagged birds, only a total of five and 19 regurgitations were recovered on LT
and LA, respectively. At both islands, birds fed exclusively on Peruvian anchovies. There were no
significant differences in the number of fish per regurgitation between females (median = 6, range = 1 –
13, N = 13) and males (median = 5, range = 1 – 12, N = 11, Mann-Whitney U-test, χ2= 0.42, df = 1, P =
0.51).
41
DISCUSSION The possibility of any handicapping effects of the instruments cannot be ruled out because systematic
evaluations of foraging behavior between manipulated and unmanipulated birds could not be completed.
However, the duration of the foraging excursions (usually used as a proxy of foraging effort) of the tagged
birds (mean = 1.8 h, range = 0.5 – 5) was comparable to other observations from unmanipulated Peruvian
Boobies. For instance, trip length varied between 1 – 2 h on Isla Mazorca in central Peru (Duffy 1983b),
and averaged 2 h on LT (Duffy 1987). Dive duration of instrumented birds here (mean = 4 sec) was similar
to that recorded from direct observations of unequipped birds in Peru (mean = 3.69, Duffy 1987) and Chile
(mean = 3.89 sec, Duffy 1987). To minimize possible adverse effects of the tags, the loggers were
retrieved usually 12 h after attachment. Under this procedure, all equipped birds were recaptured in their
nests and no desertions were observed in subsequent days.
Peruvian Boobies as predators of anchovetas
Since historical times, Peruvian boobies have been one of the most conspicuous predators of anchovetas
(Coker 1919, Vogt 1942, Tovar and Guillén 1988, Jahncke and Goya 2000). The strong dependence by
boobies on this shoaling fish is evinced by the fact that, unlike their guano bird relatives the Guanay
Cormorant (Phalacrocorax bougainvillii) and the Peruvian Pelican (Pelecanus thagus), Peruvian Boobies
hardly ever switch to alternative prey (Jahncke and Goya 2000), abandoning their nests and dispersing to
other areas when anchovetas are not available. As an anchoveta specialist, Peruvian Boobies have
adjusted their foraging strategies to the movement patterns of their prey. For example, the majority of
Peruvian Boobies did not depart after daybreak, but waited until 0900 – 1000 h to initiate their first feeding
trip of the day. Departures decreased before noon and increased again > 1400 h. This pattern is linked to
the pronounced diel feeding and migration behavior of anchovetas, which disperse during the night and
noon to feed solitarily, but tend to form dense schools by mid-morning and mid-afternoon (Mathisen
1989). The same timing of foraging has been observed in other aerial-visual predators of anchovetas such
as Guanay Cormorants (Zavalaga and Paredes 1999) and Blue-footed Boobies (Zavalaga 2004).
42
Peruvian boobies are plunge-divers that forage only during daylight hours. They are shallow
divers, with the majority of dives < 4 m (max = 9 m). Their diving capabilities contrast with those of the
sympatric Blue-footed Boobies, which attain depths of 2 – 7 m (max = 22 m) when feeding on anchovetas
(Zavalaga et al. 2007). Anchovetas remain in deep water strata during daylight hours, usually between 2
and 30 m (Ganoza et al. 2000, Mathisen 1989), and consequently Peruvian boobies would have access to
fish only from the upper strata of the shoal. Any slightly downward movements of anchoveta (below 4 m)
would deprive boobies of their main food source. However, it has been reported that Peruvian Boobies
can also feed in multi-species seabird aggregations (Duffy 1983b), occasionally with dolphins that are
able to bring anchovetas close to the surface during feeding frenzies (Duffy 1987). The extensive use of
V-shaped dives in relation to U-shaped immersions is also related to the schooling behavior of
anchovetas because rapid V-dives more efficiently depolarize anchoveta schools (i.e. forcing the fish to
act independently), making individual fish more vulnerable to predation (Zavalaga et al. 2007).
Site-specific foraging behavior
There were clear site-specific differences in the boobies’ foraging behavior. Unlike Peruvian Boobies from
LA, birds from LT approached to the mainland coast, spent more time at sea, sat on the water surface for
longer periods of time, and dived slightly deeper. However, birds from LT and LA were tracked at different
years with different oceanographic and food conditions and therefore, the effects of site and year could no
be clearly separated. Regardless of this confounding effect, Peruvian Boobies showed a flexible foraging
strategy, being able to exploit inshore and offshore waters of the Peruvian coast.
As observed in other seabird species, the foraging behavior of Peruvian boobies from LT may
have been influenced by its proximity to the continental land masses (Hamer et al. 2001, Garthe et al.
2007, Steinfurth et al. 2008). Birds from LT exploited coastal waters of the mainland and, in several cases,
traveled parallel to the coast in search of food. As a result of this exploratory behavior, birds extended
their feeding trips and the total cumulative distance traveled. Nevertheless, the topographic features of the
coastline per se cannot explain the site-specific differences in the other foraging variables. Rather, there
is strong evidence suggesting that the higher foraging effort of boobies from LT can be attributed to poor
food conditions in December 2006. First, Peruvian boobies visited warmer waters around LT in December
43
2006 than around LA in December 2007. Under warm conditions, anchovetas congregate close to the
mainland and remain in deep waters (Jordán 1971). Second, hydroacoustic surveys undertaken two
months after this study revealed that the anchovy biomass were two times lower around LT in December
2006 (1.2 million metric tons, IMARPE 2007) than LA in December 2007 (2.3 million MT, IMARPE 2008).
Third, the seabird population (including other anchoveta-predator species such as Blue-footed-Boobies,
Guanay cormorants and Peruvian Pelicans) was at least an order of magnitude larger at LT than LA
during the study period (Zavalaga, unpub. data) leading to higher levels of competition for food around the
colonies (Lewis et al. 2001).
Foraging behavior of Peruvian Boobies in comparison to tropical boobies
Peruvian boobies made multiple short trips per day, dived at relatively high rates (mean = 11 dives/h,
range = 3 – 37), foraged close to their colonies (mean = 20 km, range = 5 – 68), and returned to the same
foraging areas in consecutive trips. In compariosn, breeding Masked, Red-footed and Brown Boobies
from the tropics make only one trip in a day, departing from their colonies immediately before and after
dawn to increase the time availability for foraging (Lewis et al. 2004, Weimerskirch et al. 2008). The
higher frequency of foraging excursions by Peruvian Boobies may be related not only to the proximity of
food to the breeding areas, but also to the higher energetic demands of the brood (up to 4 chicks) in
comparison to their tropical relatives (broods of 1 chick only). Second, the diving rate of Peruvian Boobies
was 3 – 5 times higher than that of Brown (mean = 3.8 dives/h, Lewis et al. 2004) and Red-footed Boobies
(mean 2.4 dives/h, Lewis et al. 2004), suggesting a lower prey encounter rate in the tropics than in the
Peruvian coastal upwelling. Third, the foraging site fidelity was higher and the distance traveled to the
foraging areas smaller in Peruvian Boobies than in Red-footed (4 - 114 km, Weimerskirch et al. 2005b),
Masked (5 – 245 km, Weimerskirch et al. 2008) and Nazca Boobies (approximately 65 km, Anderson and
Ricklefs 1987) indicating a higher predictability and proximity of food sources around Peruvian Boobies
colonies, at least in a small spatial (< 100 km) and temporal scale (< 1 week).
Surprisingly, although the recurrence to visit the same feeding sites in consecutive trips was
similar between Peruvian Boobies and temperate gannets (Hamer et al. 2007), the distance traveled to
foraging areas was greater and the frequency of trips was smaller in the temperate species. The
44
maximum foraging range of Northern (540 km, Hamer et al. 2000), Australasian (450 km, Wingham 1985)
and Cape Gannets (242 km, Grémillet at al. 2004) indicated that, although at higher latitudes the
distribution of prey is predictable at a meso or large scale (100 – 1,000 km), it is located far away from the
bird’s colonies.
Peruvian Boobies dived deeper (mean = 2.5 m, max = 9 m) than Brown boobies (mean = 0.8,
max = 3.8 m, Lewis et al. 2005, Yoda and Kohno 2008) and Red-footed boobies (mean = 0.87, max = 2.4
m, Weimerskirch et al. 2005a,b; Lewis et al. 2005), and attained depths similar to Masked Boobies (mean
= 2.2 m, max = 5 m, Weimerskirch et al. 2008). Although Peruvian Boobies are heavier (1200 – 1675 g)
than Brown (850 – 1550 g, Nelson 1978) and Red-footed Boobies (800 - 1210 g, Nelson 1978), the dive
capabilities were not necessarily associated to body size because Masked boobies are much heavier
(1500 – 2350 g, Nelson 1978) than Peruvian Boobies. Rather, interspecific variations in dive depths of
Peruvian and tropical boobies seem to reflect the behavior and vertical distribution of prey. Tropical
boobies feed mainly on flying fish and squid (Asseid et al. 2006, Weimerskirch et al. 2008), which can be
siezed in ‘aerial dives’ or near the sea surface after being chased by sub-surface predators such as tuna
and dolphins (Jaquemet et al. 2004). Conversely, Peruvian Boobies feed on anchovetas, a shoaling fish
that usually occurs in the upper 2- 30 m of the water column (Mathisen 1989) and thus birds need to
submerge some meters below the surface to capture them.
Sex-specific foraging behavior
The results of the present study demonstrate that there was no sex-specific spatial segregation in
Peruvian Boobies, despite females being 19% heavier and 3 – 4% larger than males (Chapter 1). Females
from LT foraged in coastal waters to the east (< 11 km from the mainland), whereas males fed farther
offshore (< 38 km from the mainland) without a specific directionality, but the small number of tracked
birds (seven males vs. four females) on this island may induce one to falsely reject the null hypothesis
(Type I error), particularly because the preferences to visit offshore waters was only recorded in two
males. When the sample size was substantially increased at LA (15 males, 19 females), no spatial
segregation was observed. Additionally, there were no sex-specific differences in 13 out 15 other foraging
variables measured; however, females dived slightly deeper and spent a larger proportion of the foraging
45
time sitting on the water. Thus, it seems that chick-rearing Peruvian Boobies do not exhibit sex-specific
preferences by foraging area, but ecological differentiation between sexes may occur underwater once
the anchoveta schools have been detected.
There is an extensive list of evidence demonstrating that sex differences in body size and
morphology are responsible for habitat segregation in several vertebrate taxa (review in Ruckstuhl and
Neuhaus 2005). In seabirds, sexual size dimorphism leads to competitive exclusion of feeding territories
(González-Solís et al. 2000), differential use of feeding habitats (Shaffer et al. 2001, Phillips et al. 2004) or
partitioning of parental roles (Clarke 2001, Gilardi 1992). In some species of sulids such as the Red-
footed Boobies from the Indian Ocean, the sex-specific differences in foraging behavior has been
attributed to differences in body size, with smaller males traveling shorter distances, diving shallower, and
having a higher manoeuvrability to catch agile prey such as flying fish (Weimerskirch et al. 2006). In the
highly dimorphic Brown booby from the central Pacific Ocean, males also forage close to the colony
(Gilardi 1992). Conversely, data from another population of Brown Booby from the eastern Pacific Ocean
showed that there are no differences between sexes in dive depths and it was the small male that foraged
farther distances from the colony (Lewis et al. 2005). Unexpectedly, monomorphic seabird species forage
at different locations, dive at different depths and spent different amount of time on the sea surface (Lewis
et al. 2002, Peck and Congdon 2006). Thus, the mechanisms associated with habitat segregation have
not been clearly identified and in some cases the conclusions are contradictory suggesting that there are
factors other than body size that explain foraging habitat segregation between sexes. These factors may
include local environmental conditions, interactions with con-specifics and with other species, prey
behavior and distribution, sex-specific energetic demands of breeding adults and stage of the breeding
cycle (e.g. incubating vs. chick rearing birds).
The highly productivity waters of the Humboldt Current catalyze the association of multi -species
aggregations in conspicuous feeding frenzies within sight of Peruvian Booby colonies and nearby areas
(Duffy 1983b, Zavalaga et al. 2007). The predictability and the frequency of these aggregations contrast
with that of tropical waters, where patches of prey availability for seabirds are rare and unpredictable
(Ballance et al. 1997, Jaquemet et al. 2004, Mills 1998, Anderson and Ricklefs 1987). Duffy (1983b) found
that 99% of the 28 species of seabirds feeding in the Humboldt Current foraged in flocks and these large
groups persisted for 2 – 3 h or more when feeding on anchovetas. The seabird feeding flocks in Peru can
46
only be supported by huge shoals of anchovetas that usually occur in patches (Jordán 1971, Mathisen
1989, Ganoza et al. 2000). We hypothesize that female and male Peruvian boobies are attracted to these
multi -species aggregations when they leave the colony or are traveling in search for food, and
consequently feed in the same areas. Once the foraging patch is localized males and females plunge dive
over the shoal using the momentum of the fall to gain depth (Ropert-Coudert et al. 2004), with heavier
females diving slightly deeper than males (Weimerskirch et al. 2006, Zavalaga et al. 2007). After feeding,
females may remain longer resting on the sea surface because the cost of obtaining flight after diving are
higher in larger individuals (Weimerskirch et al. 2000, Lewis et al. 2005).
If food competition is the driving force for disruptive selection of the sexes, then it is expected that
feeding-niche segregation of Peruvian Boobies would occur during El Niño years when anchovetas are
deeper and more dispersed (Arntz and Fahrbach 1996). Likewise, tracking of incubating adults are
necessary to complement the results from chick-rearing birds because the foraging behavior of seabirds
may differ at different stages of the breeding cycle (Phillips et al. 2004, Weismerskirch et al. 2005b).
Preliminary results here from LT in December 2006 suggest that sexual segregation may emerge under
poor food conditions, an hypothesis worthy of additional investigation.
47
REFERENCES Adams NJ, Navarro RA (2005) Foraging of a coastal seabird: flight patterns and movements of breeding
Cape gannets Morus capensis. - Afr. J. Marine Sci. 27: 239–248.
Anderson DJ, Ricklefs RE (1987) Radio-tracking Masked and Blue-footed boobies (Sula spp.) in the
Galápagos Islands. National Geographic Research 3:152-163.
Arntz WE, Fahrbach E (1996) El Niño: experimento climático de la naturaleza. Mexico: Fondo de Cultura
Económica.
Asseid BS, Drapeau L, Crawford RJM, Dyer BM, Hija A, Mwinyi AA, Shinula P, Upfold L (2006) The food
of three seabirds at Latham Island, Tanzania, with observations on foraging by masked boobies
Sula dactylatra . Afr. J. Mar. Sc. 28(1):109-114.
Au DWK, Pitman RL (1986) Seabird interactions with dolphins and tunas. Condor 88: 304-317.
Ballance LTR, Pitman RL, Reilly SB (1997) Seabird community structure along a productivity gradient:
importance of competition and energetic constraint. Ecology 78:1502-1518.
Barber RT, Chavez FP, Kogelschatz JE (1985) Biological effects of El Niño. Boletin ERFEN (14):3-29.
Batschelet E (1981) Circular statistics in biology. - Academic Press, New York.
Birdlife International (2008) Birdlife International American Bird Conservancy workshop on seabirds and
seabird-fishery interactions in Peru. RSPB, Sandy, UK.
Burger AE, Shaffer SA (2008) Application of tracking and data-logging technology in research and
conservation of seabirds. Auk 125: 253-264.
Clarke JR (2001) Partitioning of foraging effort in Adélie Penguins provisioning chicks at Béchervaise
Island, Antarctica. Polar Biology 24:16-20.
Coker RE (1919) Habits and economic relations of the guano birds of Peru. Proc. U.S. Nat/. Mus. 56:449-
511.
Crawford RJM, Goya E, Rouz J-P, Zavalaga, CB (2006) Comparison of assemblages and some life-
history traits of seabirds in the Humboldt and Benguela systems. African Journal of Marine Science
28: 553-560.
Daneri G, Dellarossa V, Quiñones R, Jacob B, Montero P, Ulloa O (2000) Primary production and
community respiration in the Humboldt Current System off Chile and associated oceanic areas.
Mar. Ecol. Prog. Ser. 197:41-49.
Davoren GK, Montevecchi WA, Anderson JT (2003) Search strategies of a pursuit-diving marine bird and
the persistence of prey patches. - Ecol. Monogr. 73: 463-481.
Duffy DC (1983a) Environmental uncertainty and commercial fishing: effects on Peruvian guano birds.
Biological Conservation 26: 227-238.
Duffy DC (1983b) The foraging ecology of Peruvian seabirds. Auk 100:800-810.
Duffy DC (1987) Aspects of the ecology of Blue-footed and Peruvian boobies at the limits of their ranges
on Isla Lobos de Tierra, Perú. Colonial Waterbirds 10:45-49.
48
Fairbairn J, Shine R (1993) Patterns of sexual size dimorphism in seabirds of the Sourthern Hemisphere.
Oikos 68:139-145.
Figueroa J, Stucchi M (2008) Las aves de la Isla Lobos de Afuera (Peru) en la primavera del 2008.
Ornitología Neotropical 19: 377-390.
Ganoza F, Castillo PR, Marín D (2000) Variaciones estacionales en la distribución y biomasa de
anchoveta entre 1983 y 2000. Bol Ints Mar Peru 19(1-2):157-177.
Garthe S, Montevecchi WA, Chapdelaine G, Rail J, Hedd A (2007) Constrasting foraging tactics by
northern gannets (Sula bassana ) breeding in different oceanographic domains with different prey
fields. – Mar. Biol. 151: 687-694.
Gilardi JD (1992) Sex-specific foraging distributions of Brown Boobies in the Eastern Tropical Pacific.
Colonial Waterbirds 15:148-151.
González-Solís J, Croxall JP, Wood AG (2000) Sexual dimorphism and sexual segregation in foraging
strategies of Northern Giant Petrels, Macronectes halli, during incubation. Oikos 90:390-398.
Grémillet D, Dell’Omo G, Ryan PG, Peters G, Ropert-Coudert Y, Weeks SJ (2004) Offshore diplomacy or
how seabirds mitigate intra-specific competition: a case study on GPS tracking of Cape gannets
from neighbouring colonies. – Mar. Ecol. Prog. Ser. 268: 265-279.
Hamer KC, Phillips RA, Wanless S, Harris MP, Wood AG (2000) Foraging ranges, diets and feeding
locations of gannets Morus bassanus in the North Sea: evidence from satellite telemetry. - Mar.
Ecol. Prog. Ser. 200: 257-264.
Hamer KC, Phillips RA, Hill JA, Wanless S, Wood AG (2001) Contrasting foraging strategies of gannets
Morus bassanus at two North Atlantic colonies: foraging trip duration and foraging area fidelity. -
Mar. Ecol. Prog. Ser. 224: 283-290.
Hamer KC, Humphreys EM, Garthe S, Hennicke J, Peters G, Grémillet D, Phillips RA, Harris MP,
Wanless S (2007) Annual variation in diets, feeding locations and foraging behaviour of gannets in
the North Sea: flexibility, consistency and constraint. Mar. Ecol. Prog. Ser. 338:295-305.
Hooge PN, Eichenlaub B (1997) Animal movement extension to Arcview, version 1.1. US Geological
Survey, Anchorage.
Hunt Jr GL, Schneider DC (1987) Scale dependent processes in the physical and biological environment
of seabirds. In: Croxall, J.P. (Ed.), Seabirds: Feeding ecology and role in marine ecosystems.
Cambridge University Press, Cambridge, pp. 7-41.
Hyrenbach KD, Fernández P, Anderson DJ (2002) Oceanographic habitats of two sympatric North Pacific
albatrosses during the breeding season. Mar. Ecol. Prog. Ser. 233:283-301.
IMARPE - Instituto del Mar del Perú (2007) Crucero 0702-04 de evaluación hidroacústica de recursos
pelágicos BIC Jose Olaya Balandra, BIC SNP-2, L/P IMARPE VI, Informe Ejecutivo Tacna -
Tumbes, del 22 Febrero al 8 Abril 2008. Available online from http://www.imarpe.gob.pe.
IMARPE - Instituto del Mar del Perú (2008) Crucero 0802-04 de evaluación hidroacústica de recursos
pelágicos BIC Jose Olaya Balandra, BIC SNP-2, L/P IMARPE V, Informe Ejecutivo, Puerto Pizarro
49
Tumbes-Los Palos (Tacna), del 26 Febrero al 12 Abril 2008. Available online from
http://www.imarpe.gob.pe.
Jahncke J, Zileri D (1998) Estudios sobre dieta de piqueros como indicadores de la estructura por tallas
de los stocks de anchoveta en el mar peruano. Boletín Instituto del Mar del Peru 17(1):47-54.
Jahncke J, Ayala L, Mendoza C (1998) Abundancia relativa y distribución de aves marinas frente a la
costa peruana y su relación con la abundancia de anchoveta. Crucero BIC Humboldt 9808-09, de
Piura a Lima. - Inf. Inst. Mar Perú 141: 85-95.
Jahncke J, Goya E (2000) Responses of three booby species to El Niño 1997-1998. - Waterbirds 23: 102-
108.
Janhcke J, Checkley DM, Hunt Jr GL (2004) Trends in carbon flux to seabirds in the upwelling system:
effects of wind and fisheries on population regulation. Fishery Oceanography 13(3):208-223.
Jaquemet SM, Le Corre M, Weimerskirch H (2004) Seabird community structure in a coastal tropical
environment: importance of natural factors and fish aggregation devices (FADs). Mar. Ecol. Prog.
Ser. 268:281-292.
Jordán R (1971) Distribution of anchoveta (Engraulis ringens) in relation to the environment. - Invest.
Pesq. Barcelona 35: 113-126.
Lewis S, Benvenuti S, Dall’Antonia L, Griffiths R, Money L, Sherratt TN, Wanless S, Hamer KC (2002)
Sex-specific foraging behavior in a monomorphic seabird. Proc R Soc Lond B 269:1687-1693.
Lewis S, Sherrat TN, Hamer KC, Wanless S (2001) Evidence of intra-specific competition for food in a
pelagic seabird. Nature 412: 816-818.
Lewis S, Schreiber EA, Daunt F, Schenk GA, Wanless S, Hamer KC (2004) Flexible foraging patterns
under different time constrains in tropical boobies. Anim Behav 68:1331-1337.
Lewis S, Schreiber EA, Daunt F, Schenk GA, Orr K, Adams A, Wanless S, Hamer KC (2005) Sex-specific
foraging behavior in tropical boobies: does size matter? Ibis 147:408-414.
Mathisen OA (1989) Adaptation of the anchoveta (Engraulis ringens) to the Peruvian upwelling system, p.
220-234. In D. Pauly, P. Muck, J. Mendo and I. Tsukayama (eds.). The Peruvian upwelling
ecosystem: dynamics and interactions. ICLARM Conference Proceedings 18, 438 p. Instituto del
Mar Del Perú (IMARPE), Callao, Perú.
Mills K (1998) Multispecies seabird feeding flocks in the Galápagos Islands. Condor 100:277-285.
Murphy RC (1936) Oceanic Birds of South America. American Museum of Natural History, New York.
Nelson JB (1978) The Sulidae: Gannets and Boobies. - Oxford University Press, Oxford.
Paredes R, Jones IL, Boness DJ (2005) Reduced parental care, compensatory behavior and reproductive
costs of thick-billed murres equipped with dataloggers. Anim. Behav. 69:197-208.
Pauly D, Christensen V, Guénette S, Pitcher TJ, Sumaila UR, Walters CJ , Watson R Zeller D (2002)
Towards sustainability in world fisheries. Nature 418: 689-695.
Peck DR, Congdon DC (2006) Sex-specific chick provisioning and diving behavior in the wedge-tailed
shearwater Puffinus pacificus . J. Avian. Biol. 37:245-251.
50
Phillips RA, Silk JRD, Phalan B, Catry P, Croxall JP (2004) Seasonal sexual segregation in two
Thalassarche albatross species: competitive exclusion, reproductive role specialization or foraging
niche divergence? Proc. R. Soc. Lond. B. 271:1283-1291.
Ropert-Coudert Y, Grémillet D, Kato A, Ryan PG, Naito Y, Le Maho Y (2004) A fine-scale time budget of
Cape gannets provides insights into the foraging strategies of coastal seabirds. Anim behav 67:985-
992.
Ruckstuhl KE, Neuhaus P (2005) Sexual segregation in vertebrates: ecology of the two sexes. Cambridge
University Press, UK.
Seaman, D. E. and R. A. Powell (1996) An evaluation of the accuracy of kernel density estimators for
home range analysis. Ecology 77:2075-2085.
Shaffer SA, Weimerskirch H, Costa DP (2001) Functional significance of sexual dimorphism in Wandering
albatross, Diomedea exulans . Functional Ecology 15:203-210.
Steinfurth A, Vargas FH, Wilson RP, Spindler M, Macdonald DW (2008) Space use by foraging
Galápagos penguins during chick rearing. Endamg, Species Res. 4:105-112.
Tovar H, Guillén V (1988) Comportamiento alimenticio del piquero Sula variegata , ave guanera peruana.
Boletin de Lima 60:85-90.
Vogt W (1942) Informe sobre la saves guaneras. Boletin de la Compañía Administradora de Guano
18(3):3-129.
Weichler T , Garthe S, Luna-Jorquera G, Moraga J (2004) Seabird distribution on the Humboldt Current in
northern Chile in relation to hydrography, productivity, and fisheries. ICES Journal of Marine
Science 61:148-154.
Weimerskirch H (2007) Are seabirds foraging for unpredictable resources? Deep-Sea research II 54:211-
223.
Weimerskirch H, Guionnet T, Martin J , Shaffer SA, Costa DP (2000) Fast and fuel efficient? Optimal use
of wind by flying albatrosses. Proc. Royal. Soc. Lond. B. 267:1869-1874.
Weimerskirch H, Le Corre M, Ropert-Coudert Y , Kato A, Marsac F (2005a) The three-dimensional flight of
red-footed boobies: adaptations to foraging in a tropical environment? - Proc. R. Soc. Lond. B Biol.
Sci. 272: 53-61.
Weismerskirch H, Le Corre M, Jaquemet S, Marsac F (2005b) Foraging strategy of a tropical seabird, the
red-footed booby, in a dynamic marine environment. - Mar. Ecol. Prog. Ser. 288: 251-261.
Weimerskirch H, Le Corre M, Ropert-Coudert Y, Kato A, Marsac F (2006) Sex-specific foraging behaviour
in a seabird with reversed sexual dimorphim: the red-footed booby. Oecologia 146:681-691.
Weimerskirch H, Le Corre M, Bost ChA (2008) Foraging strategy of masked boobies from the largest
colony in the world: relationship to environmental conditions and fisheries. Mar. Ecol. Prog. Ser.
362:291-302.
Wingham EJ (1985) Food and feeding range of the Australasian gannet Morus serrator (Gray). Emu
85(4):231-239.
51
Wood AG, Naef-Daenzer B, Prince PA, Croxall JP (2000) Quantifying habitat use in satellite-tracked
pelagic seabirds: application of kernel estimation to albatross locations. J. Avian Biol. 31:278-286.
Yoda K, Kohno H (2008) Plunging behaviour in chick-rearing Brown boobies. Ornithol. Sci. 7:5-13.
Zavalaga CB, Paredes R (1999) Foraging behaviour and diet of the guanay cormorant. S. Afr. J. Mar. Sci.
21: 251-258.
Zavalaga CB (2004) Testing the feeding-niche partitioning hypothesis in the sexually dimorphic Blue-
footed Booby. Master dissertation. University of North Carolina Wilmington, North Carolina, USA.
Zavalaga CB, Benvenuti S, Dall;Antonia L, Esmlie SD (2007) Diving behavior of Blue-footed Boobies Sula
nebouxii in northern Peru in relation to sex, body size and prey type. Mar. Ecol. Prog. Ser. 316-291-
303.
52
CHAPTER 3: EFFECTS OF OCEANOGRAPHIC FEATURES AND WIND CONDITIONS ON THE FORAGING MOVEMENTS OF PERUVIAN BOOBIES Sula variegata: A COMPARISON
BETWEEN AN INSHORE AND OFFSHORE ISLAND
53
ABSTRACT The Humboldt Current system is characterized by high marine productivity that varies with unpredictable
interannual changes in oceanographic conditions. The intraannual changes are more predictable and
during the austral summer cold water masses with enhanced productivity are confined in a narrow
onshore strip, which causes aggregations of anchoveta (Engraulis ringens) close to the mainland. The
spatial distribution of the anchoveta predators should be linked to the seasonal movement of their prey.
To test this hypothesis, the foraging movements of chick-rearing Peruvian Boobies (Sula variegata), an
anchoveta specialist, were examined in relation to concurrent remote-sensed data on sea-surface
temperature and chlorophyll concentration using geographic information systems (GIS) software.
Likewise, the effects of bathymetry and wind direction on foraging movements were evaluated. Birds were
tracked using global positioning system (GPS) dataloggers and depth/meter archival tags on an inshore
(Lobos de Tierra, LT) and offshore (Lobos de Afuera, LA) island in northern Peru. The foraging areas of
Peruvian Boobies were not randomly distributed, but were concentrated to the east and south of LT and
LA, respectively. Peruvian Boobies from LT were strongly influenced by the prevailing oceanographic
conditions, feeding in shallow waters (< 20 m) and areas with higher sea surface temperature and
chlorophyll concentration than other potential foraging areas not used by the birds. Conversely, birds from
LA foraged over the shelf break and slope (100 – 2500 m) in areas with similar oceanographic conditions
compared to those expected from a random distribution. Bathymetric features and wind patterns may
have been the proximate cause for the observed foraging movements of Peruvian Boobies from LA, but
food distribution was probably the ultimate cause. Areas of enhanced productivity close to the mainland
were never visited by boobies. We speculate that competition with birds from LT or social attraction to
feeding flocks around LA may preclude birds to visit areas close to LT. Likewise, we suggest that the
enlarged brood size (up to 4 chicks) force adults to forage more than once per day, restricting the foraging
range of adults and consequently constraining the birds to visits other more productive areas close to the
mainland.
54
INTRODUCTION Several studies have documented that physical and biological processes at distinct spatial and temporal
scales influence the habitat use of marine top predators (Piatt et al. 1989, Lowry et al. 2000, Hyrenbach et
al. 2002, Godley et al. 2002). Foraging areas of high-latitude marine vertebrates are generally associated
with ocean fronts (Hyrenbach et al. 2002, Shaffer et al. 2006), eddies (Hunt and Schneider 1987,
Weimerskirch et al. 2004), upwelling areas (Hatch et al. 2000, Biuw et al. 2007, Raya-Rey et al. 2007) and
tidal zones (Irons 1998), where marine productivity is usually higher than in the surrounding area.
Physical and biological features in these regions concentrate prey, which in turn provide enhanced
feeding opportunities for many marine predators. Bathymetric features also play an important role in the
aggregation of marine predators as the presence of submarine canyons, seamounts or banks disrupt
water flow and circulation, thereby causing upwelling of nutrient-rich waters to the surface. Thus, marine
birds and cetaceans concentrate over shelf breaks, continental slopes and shallow-water topographies
where the abundance of macro-zooplankton is usually elevated (Yen et al. 2004). Wind direction and
strength may have a major influence on the spatial distribution of marine top predators either by driving
upwelling processes (Baylis et al. 2008) or by decreasing locomotion costs of volant seabirds (Jouventin
and Weimerskirch 1990, Spear and Ainley 1997, Adams and Navarro 2005). The occurrence of recurrent
wind patterns on breeding colonies may be used by seabirds to predetermine flight orientation after
departure from their nests (Grémillet et al. 2004).
The Peruvian Booby (Sula variegata) is a suitable species for examining how hydrographic,
bathymetric and atmospheric factors influence foraging behavior. This species is an important top
predator of the marine community of the Humboldt Current system along the coast of Peru and northern
Chile. This environment is heterogeneous, characterized by a high marine productivity, with marked inter-
annual variations in oceanographic conditions (Bakun 1987, Brainard and McLain 1987, Morón 2000).
The dynamics of hydrographic features affect the abundance and distribution of the keystone species in
the ecosystem: the Peruvian anchovy or anchoveta (Engraulis ringens). This schooling fish is the main
prey consumed by Peruvian Boobies (Tovar and Guillén 1988, Jahncke and Zileri 1998) and the target
species of other marine predators (Majluf 1989, Espinoza 2000) and commercial fisheries (Pauly et al.
2000). Ship-based surveys have revealed that Peruvian Boobies are one of the most abundant resident
seabirds observed at sea, and that their highest densities occur at onshore waters close to the mainland
55
(Duffy 1983, Weichler et al. 2004). Foraging areas of breeding birds are not randomly distributed around
their colonies, but located in specific areas that are frequently visited within a small spatial and temporal
scale (Chapter 2). Although the selection of specific zones for foraging seems to be strongly influenced by
the presence of multi -species feeding flocks (Duffy 1983) and occasionally to fishing vessel activities
(Weichler et al. 2004), the association of Peruvian Booby distribution with physical features of the
environment is less clear (Weichler et al. 2004).
Because Peruvian Boobies feed a lmost exclusively on anchovetas, it is expected that their
foraging movements are strongly linked to the distribution of their prey. Although concurrent information
about distribution and abundance of anchovetas is usually difficult to obtain, their habitat use as well as
the biotic processes that affect their distribution and abundance are well-documented (Muck et al. 1989,
Mathisen 1989, Ganoza et al. 2000). Therefore, oceanographic variables can be used as proxies to
predict anchoveta occurrence. For example, anchovetas inhabit waters masses with temperatures of 13 –
23 oC, depths of 0 – 100 m, and exhibit a very coastal distribution during warm conditions (Guitiérrez et al.
2007). Given that Peruvian Boobies forage in the upper 4-m of the water column (Chapter 2.), remotely-
sensed satellite data such as sea-surface temperature (SST) and chlorophyll-a (chla) concentration can
be used to discern regions of enhanced ocean productivity and, consequently, top predator presence
(Grémillet et al. 2008).
This research was conducted on two islands in northern Peru that differed in their proximity to the
continent: Lobos de Tierra (inshore, 15 km) and Lobos de Afuera (offshore, 61 km). Lobos de Afuera is
the only offshore breeding site of Peruvian Boobies within its distribution range and therefore it is an
atypical island that is not influenced by the presence of land masses. This geographical difference
allowed the evaluation of booby foraging movements under different bathymetric features and
oceanographic conditions. The obje ctive of this study was to determine whether abiotic factors influence
the foraging movements of Peruvian Boobies attending half-grown chicks. We examined the marine
habitat use of birds by overlaying their foraging areas to remote-sensed data of sea-surface temperature,
chla concentration, bathymetry, and wind patterns . Our primary questions were whether boobies (1) only
forage in areas with low ocean temperatures and high productivity, (2) exclusively use waters over the
continental shelf where most of the upwelling processes occur, and (3) are influenced by wind patterns in
their spatial distribution.
56
METHODS
Study site
Peruvian Boobies were studied on Isla Lobos de Tierra (LT; 6o24’S, 80o51’W) from 22-30 December
2006, and Isla Lobos de Afuera (LA; 6o57’S, 80o41’W) from 10-17 December 2007 (Fig. 14). LT is an
inshore island with an area of 1426 ha (9 x 3 km) and located 12 km from the nearest landmark on the
mainland. It supports an important population of Blue-footed Boobies (Sula nebouxii, ∼ 105 breeding
pairs), Peruvian Pelicans (Pelecanus thagus , ∼ 105 individuals), and Peruvian Boobies (∼ 104 individuals,
Zavalaga unpub. data). LA (236 ha) is smaller than LT and supports a total population of the same three
species above of ∼ 105 individuals (Figueroa and Stucchi 2008). The shortest distance between the LA
and the mainland is 61 km. During the study period, Peruvian Boobies consumed exclusively anchovetas,
and pairs were attending broods of 2 - 3 half-grown chicks (Chapter 2).
Figure 14. Location of Isla Lobos de Tierra and Isla Lobos de Afuera in northern Peru. Track (black lines)
and dive locations (filled circles) are given for one trip of bird 240 (grey) and two trips of bird 249 (white).
57
Dataloggers and handling of birds
Thirty-six chick-rearing Peruvian Boobies (12 at LT and 24 at LA) were equipped with GPS dataloggers
and depth-meter archival tags. The two data sets were combined to determine movement patterns and to
identify foraging areas. T wo types of GPS dataloggers (www.technosmart.eu) were used: the GiPSy-1 (33
g; 5 x 3 x 1 cm), programmed to take one position fix every 10 s for approximately 48 h, and the GiPSy-2
(14 g, 4.5 x 2.2 x 0.7 cm) with a time interval between fixes at 1 s for approximately 14 h. The battery
lifespan was long enough to record at least one complete foraging trip (mode = 2 trips per deployment).
The accuracy of both loggers was < 10 m in > 95% location fixes. The instruments were protected in two
heat-sealed waterproof polyethylene bags (1 g) and attached to the four central tail feathers with four
strips of waterproof Tesa tape. Dive events were recorded with Lotek LTD1110 archival tags (5g; 1.1 x 3.2
cm, www.lotek.com), programmed to collect data every second for approximately 18 h. These loggers
were also equipped with a temperature sensor, but these data were not used in this study. The tags were
attached to a metal band with three small plastic cable ties. The pressure sensor recorded depths with an
accuracy of 0.3 m (depth range: 0 – 30 m). Because the depth tags were synchronized with the GPS
dataloggers, it was possible to geo-reference the dive events and therefore to identify the foraging areas
by the location of dives (Fig 14). The GPS and the depth/temperature data were downloaded to a laptop
computer using the GiPSy (www.technosmart.eu) and Tag Talk 1100 (www.lotek.com) software,
respectively. The GPS, the depth tags and their accessories weighed 30 g with GiPSy 2 and 50 g with
GiPSy 1, accounting for approximately 1.8 % - 4,2% of the bird’s body weight (mode 2%), similar the
proportion reported in other booby studies (Weimerskirch et al. 2005, Zavalaga et al. 2007).
Handling of birds
Birds were randomly selected from the second or third row of nests from a breeding group of ca. 100
nests at LT ca. 3000 nests at LA . Only one member of the pair was selected and removed from the nest
using a hook placed gently around their necks and attached to a 3-m pole . The tagged boobies were
marked with a temporary dye (rhodamine B) on the chest and head for easy sighting in a dense colony.
Boobies were recaptured 12 – 48 h later (82 % of birds recaptured between 10 - 14 h) to retrieve the
58
devices. All deployed birds returned to their nests and no evidence of nest desertion was recorded during
the study period.
Oceanographic data
Marine habitat use by Peruvian Boobies was analyzed by using three types of oceanographic data: SST,
chla concentration, and bathymetry. Remote sensing data of these variables were downloaded from the
National Oceanographic and Atmospheric Administration Coast Watch Program
(http://coastwatch.noaa.gov). The grid data of SST and chla were collected during daytime hours by the
MODIS (Moderate Resolution Imaging Spectroradiometer) sensor on Aqua’s NASA satellite with a cell
size resolution of 0.025 degrees (approximately 2.7 x 2.7 km). For LT and LA, the weekly composite
averages of SST and chla from 17 to 25 December 2006, and 19 – 26 December were selected. These
time intervals were the closest available to the study period (the mismatch was < 9 days). We attempted
to analyze concurrent environmental-foraging data at 1-day levels, but daily SST and chla data were not
available for most of the study periods due to cloud cover. Bathymetry data were extracted from
ETOPO2v2 Global Gridded 2-minute Database (approximately 3.5 x 3.5 km cell size).
Data handling and analysis
The spatial data were mapped and analyzed using the ArcGIS 9.2 Geographic Information Systems
software (ESRI Inc., Redlands, CA). Foraging tracks, dive positions (Fig. 14) and grid data of
environmental variables were projected in the Universal Transverse Mercator (Zone 17M) coordinate
system. Dives events were identified by immersions > 0.5 m (dives shallower than this threshold value
may represent bathing splashes) and extracted for each trip using Multitrace MT-dive software (Jensen
Software Systems, Kiel, Germany). Foraging areas were identified by the location fixes of dives. Data on
male and female birds were pooled for further analysis as there were no significant sex-specific
differences in the spatial distribution of Peruvian Boobies tracked in this study (Chapter 2).
The foraging information was represented by vector data (points, lines and polygons). To overlay
dive locations and tracks to the oceanographic data, the vector features were converted into raster files
with grid cell size similar to that of oceanographic data (2.7 or 3.5 km, see above). Peruvian Boobies
59
foraged 1 – 3 times in a day, repeatedly diving in localized areas (Chapter 2). The reduction of several
clustered diving points into one grid cell tempered the effects of autocorrelation and thus dive and track
data were analyzed at the foraging trip level. The Raster Calculator of the Spatial Analyst Extension in
ArcGIS was used for data queries as well as cell-by-cell operations of oceanographic and bird grid data.
The null hypothesis of habitat selection was that the conditions of the oceanographic variables within the
bird diving areas (areas foraged) were similar to those in non-diving areas (areas available or crossed),
i.e., birds did not select any particular oceanographic characteristics (e.g. low SST or high levels of Chl-a)
for foraging. The available foraging area was defined by the area enclosed in a circle centered on the
colony and with a radius equivalent to the maximum foraging distance (the farthest dive location). The
radial distance from the colony was 50.5 and 67.1 km on LT and LA, respectively (Chapter 2). The
crossed area was the area traveled by the birds when commuting between the colony and the feeding
grounds. The crossed surface area was calculated by subtracting the diving areas from the area
delineated by the bird’s complete track.
A flow chart of the complete spatial analysis of the oceanographic data in the diving and available
areas is given in Figure 15. The end products of this analysis were raster attribute tables that contained
the number of grid cells for each of the dataset classes (e.g. 10 cells of 19oC, 15 cells of 20oC, 45 cells of
21oC, and so on). These tables were subsequently used to create frequency distributions of the
oceanographic data within the diving areas and within available, foraged and crossed areas.
Effects of wind on flight direction and speed
Instantaneous ground flight speeds of boobies were calculated from the longitude/latitude points using the
ArcGIS extension Hawth’s tools (Hawth's Analysis Tools for ArcGIS available at
http://www.spatialecology.com/htools). First, the departure and return times for each foraging trip were
identified by visual inspection of all data points in the track. Once the departure and arrival times were
established, all positional fixes on land were excluded for the analysis. Only one trip was randomly
selected per bird to reduce data pseudoreplication. Speeds and bearing were calculated from the distance
and time elapsed between two consecutive position fixes in a foraging track (1 sec for GiPSy-2 and 10
sec for GiPSy-1). For analysis of flight speeds, only values >10 km/h were selected because speeds < 10
60
km/h represented other at-sea activities (Chapter 2). Data of wind direction concurrent to the study period
were obtained from the long-term meteorological dataset of the Peruvian Navy Hydrographic Service
Station (PNHSS) located on LA. Wind direction was measured at 15 – 20 m above sea level and recorded
daily at 0100, 0700, 1300 and 1900 h. An analysis of wind conditions measured simultaneously on LT
(Zavalaga et al. 2008) and LA (obtained from PNHSS) in January and February 2003 revealed that wind
direction on LT (circular mean = 146 ± 41o, n = 54) and LA (circular mean = 141 ± 30o, n = 124) was
similar (Watson-William F-test, F1,106 = 0.341, P = 0.56), and therefore, we assumed that wind direction
measured on LA during the study period could be extrapolated to LT.
The effects of wind on bird flight direction and speed were evaluated in both the outbound and
inbound paths. The outbound path was defined as the section between the location of the colony and the
first dive, whereas the inbound path corresponded to the distance between the location of the last dive
and the colony. For each trip, flight direction relative to wind direction (keeping in mind that flight direction
and wind direction are reversed 180o) was calculated between consecutive points on the track using the
wind direction that corresponded to the date and time interval of the trip. The resulting flight bearing was a
continuous variable (0 – 180o), which was classified into five categories according to Spear and Ainley
(1997), slightly modified for a continuous variable: (1) flight into headwinds (difference between bird
course and wind course of 0o to 20o), (2) flight across headwinds (difference of 20.1o to 60o), (3) flight
across wind (difference of 60.1o to 120o), (4) flight across tailwinds (difference of 120.1o to 160o), and (5)
flight with tail winds (difference of 160.1o to 180o). If flight direction was not related to wind direction then
the proportion of bearing records in each of the five directions would be 1:2:3:2:1, respectively. Significant
deviation from these ratios would indicate a response in flight direction to wind direction (Spear and Ainley
1997).
Statistical analysis
The Kolgomorov-Smirnov (K-S) two-sample test for distributions of continuous variables (Sokal and Rohlf
1995) was used to compare the frequency distribution of oceanographic data between the foraged and
the available/crossed areas. The χ2 goodness-of-fit was used to test flight direction preferences.
Differences between the flight and wind direction were calculated using circular statistics (Batschelet
61
1981) in Oriana software version d.02C (Kovach Computing Science, Pentreath, UK). All statistical tests
were considered significant at P < 0.05.
62
Figure 15. Flow chart for the spatial analysis of marine habitat use of Peruvian Boobies (Sula variegata)
breeding at Isla Lobos de Tierra in 2006 and Lobos de Afuera in 2007.
Oceanographic data (SST, Chla or
Bathymetry) ASCII file
Convert to grid ESRI
Raster file, cell size:
2.7 km (SST, Chl-a), 3.5 km (Bathym.)
(X, Y) Colony location text file
from GPS
Convert to shapefile
Colony point feature shapefile
Buffer 50.5 km for LT 67.7 km for LA
Raster dive file, cell size:
2.7 km (SST, Chl-a), 3.5 km (Bathym.)
(X, Y) Dive locations text file from GPS
Convert to shapefile
Dive point feature shapefile
Convert to grid
Circle polygon feature shapefile
Convert to grid
Raster circle file, cell size:
2.7 km (SST, Chl-a), 3.5 km (Bathym.)
Sum Raster
Calculator
Raster file oceanographic data
within circle
Export raster
database attribute
table
dbf file with number of grid cells at each oceanographic
data class
Sum operator Raster
Calculator
Raster file oceanographic data of
diving areas
Export raster
database attribute
table
dbf file with number of grid cells at each oceanographic
data class
Raster track file, cell size:
2.7 km (SST, Chl-a), 3.5 km (Bathym.)
(X, Y) tracks text file from
GPS
Convert to shapefile
Tracks polyline feature shapefile
Convert to grid
Boolean connector
Raster Calculator
Raster file oceanographic data of tracks without diving
areas
Export raster database
attribute table
dbf file with number of grid cells at each oceanographic
data class
Available Foraged Crossed
63
RESULTS There were geographical differences in the spatial distribution of Peruvian Boobies. Birds from LT
preferentially foraged to the east of the island on inshore waters close to the mainland, whereas birds
from LA visited waters to the southwest of the island and never approached the mainland coastline. Both
islands lie at the edge of the continental shelf (< 100 m, Figs. 16a and 17a). Remote-sensed data
revealed that during the study period both islands were surrounded by relatively cold (13 – 22oC, Figs. 18a
and 19a) and productive waters (0.38 – 118 mg·m -3 chla, Figs. 20a and 21a). Overall, for any given year,
the oceanographic conditions within the available zone were similar between islands (K-S, P > 0.4).
However, water masses were significantly warmer in December 2006 (median = 20oC, n = 92) than
December 2007 (median = 18oC, n = 175, K-S test, D = 0.738, P < 0.001). Likewise, the concentration of
chla was significantly higher in 2006 (median = 2 mg·m-3, n = 87 grids) than in 2007 (median = 1 mg·m-3, n
= 101, K-S test, D = 0.425, P < 0.001).
Effects of bathymetry and oceanographic features
Birds from LT foraged only in waters over the continental shelf, in waters shallower (median 17 m) than
those crossed (median 33 m) or expected from a random distribution (median 305 m, Fig 16b).
Conversely, boobies from LA foraged over the continental slope (median 185 m). The foraged habitat was
significantly shallower than that crossed (median 491 m), but slightly deeper than that available (median
157 m, Fig. 17b).
The SST of water masses occupied by boobies from LT was not different from that in the
available zone, but was significantly warmer (median 20.2oC) than in the areas which birds only crossed
(median 19.9oC, Fig. 18b). On LA birds dived in waters with SST similar to the available and crossed
areas (Median = 18.7oc, Fig. 19b). The concentration of Chla in the diving areas of birds from LT (median
3.47 mg·m -3) was higher than that expected from a random distribution (2.64 mg·m -3) and similar to the
productivity of the commuting zones (Fig. 20b). Conversely, there was no selection of habitat by birds
from LA as SST and Chla concentration were similar between the diving areas and the available and
crossed zones.
64
Figure 16. (a) Overlay of Peruvian Booby Sula variegata dive position fixes and bathymetry charts from
remote-sensing around Isla Lobos de Tierra in December 2006. (b) Frequency distributions of bathymetric
data in the foraged area in comparison to the available (circle around the island with a radius of 50.5 km,
upper panel) and crossed areas (lower panel).
(a)
(b)
0
20
40
60
80
100
0 400 800 1200 1600 2000 2400 2800 3200
Seafloor depth (m)
Fre
quen
cy (%
)
ForagedAvailable
D(41, 395) = 0.574 P < 0.001
0
10
20
30
40
0 20 40 60 80 100 - 609
Seafloor depth (m)
Fre
quen
cy (%
)
ForagedCrossed
D(41, 83) = 0.3582 P < 0.01
65
Figure 17. (a) Overlay of Peruvian Booby Sula variegata dive position fixes and bathymetry charts from
remote-sensing around Isla Lobos de Afuera. (b) Frequency distributions of bathymetric data in the
foraged area in comparison to the available (circle around the island with a radius of 67.1 km, upper
panel) and crossed areas (lower panel).
(a)
(b)
0
5
10
15
20
25
30
35
40
45
0 400 800 1200 1600 2000 2400 2800 3200 3600 4000
Seafloor depth (m)
Fre
quen
cy (%
)
ForagedAvailability
D(90, 989) = 0.269 P < 0.01
0
5
10
15
20
25
30
35
0 400 800 1200 1600 2000 2400 2800 3200 3600 4000
Seafloor depth (m)
Fre
quen
cy (%
)
ForagedCrossed
D(90, 233) = 0.183 0.01 < P < 0.05
66
Figure 18. (a) Overlay of Peruvian Booby Sula variegata dive position fixes and sea surface temperature
(SST) from remote-sensing around Isla Lobos de Tierra between 17 and 25 December 2006. (b)
Frequency distributions of SST data in the foraged area in comparison to the available (circle around the
island with a radius of 50.5 km, upper panel) and crossed areas (lower panel).
(a)
(b)
0
10
20
30
40
50
60
12 13 14 15 16 17 18 19 20 21 22 23 24 25
SST (oC)
Fre
quen
cy (
%)
ForagedAvailable
D (72,744) = 0.089P = 0.7
0
10
20
30
40
50
60
12 13 14 15 16 17 18 19 20 21 22 23 24 25
SST (oC)
Fre
quen
cy (%
)
ForagedCrossed
D(72,88) = 0.2530.01 < P < 0.05
67
Figure 19. (a) Overlay of Peruvian Booby Sula variegata dive position fixes and sea surface temperature
charts from remote-sensing around Isla Lobos de Afuera between 19 and 26 December 2007. (b)
Frequency distributions of SST data in the foraged area in comparison to the available (circle around the
island with a radius of 67.1 km, upper panel) and crossed areas (lower panel).
(a)
(b)
0
5
10
15
20
25
30
35
12 13 14 15 16 17 18 19 20 21 22 23 24 25
SST (oC)
Freq
uenc
y (%
)
ForagedAvailable
D(129,1758) = 0.107P = 0.15
0
5
10
15
20
25
30
35
12 13 14 15 16 17 18 19 20 21 22 23 24 25
SST (oC)
Fre
quen
cy (
%)
ForagedCrossed
D(129, 233) = 0.146P = 0.06
68
Figure 20. (a) Overlay of Peruvian Booby Sula variegata dive position fixes and chlorophyll-a
concentration (chla) from remote-sensing around Isla Lobos de Tierra between 17 and 25 December
2006. (b) Frequency distributions of chla data in the foraged area in comparison to the available (circle
around the island with a radius of 50.5 km, upper panel) and crossed areas (lower panel).
(a)
(b)
0
5
10
15
20
25
0 2.5 5 7.5 10 12.5 15 17.5
Chlorophyll-a (mg m -3)
Fre
quen
cy (%
)
ForagedAvailable
D (67, 728) = 0.271P < 0.01
0
5
10
15
20
25
0 2.5 5 7.5 10 12.5 15 17.5
Chlorophyll-a (mg m-3)
Fre
quen
cy (%
)
ForagedCrossed
D(67, 81) = 0.191P = 0.17
69
Figure 21. (a) Overlay of Peruvian Booby Sula variegata dive position fixes and chlorophyll-a
concentration (chla) charts from remote-sensing around Isla Lobos de Afuera between 19 and 26
December 2007. (b) Frequency distributions of chla data in the foraged area in comparison to the
available (circle around the island with a radius of 67.1 km, upper panel) and crossed areas (lower panel).
(a)
8b)
0
5
10
15
20
25
30
35
40
0 2.5 5 7.5 10 12.5 15 17.5
Chlorophyll-a (mg m-3)
Fre
quen
cy (%
)
ForagedAvailable
D (129, 1758) = 0.111P = 0.1
0
5
10
15
20
25
30
35
40
0 2.5 5 7.5 10 12.5 15 17.5
Chlorophyll-a (mg m-3)
Fre
quen
cy (
%)
ForagedCrossed
D (129, 233) = 0.056P = 0.9
70
Effects of wind
In both years, wind blew predominantly from the southeast sector (Mean2006 = 147 ± 12o; Mean2007 = 159
± 18o), within a narrow range (mean vector r2006 = 0.977; r2007 = 0.951), and with a mean speed of 26 ± 5.6
km·h-1 in 2007 and 32.5 ± 9.4 km·h-1 in 2006. Wind direction had a significant effect in the orientation of
bird flights (Figs. 22-23). However, the bird’s response to wind differed between islands. Boobies from LT
departed from the colony usually across the wind (41% of records), across headwinds (22%) and with
tailwinds (20%). After feeding, the majority of the boobies returned to the colony across tailwinds (39%),
with some individuals flying into headwinds (19%) and across headwinds (20%). Thus, Peruvian Boobies
flew more often across head winds when travelling to their feeding areas (Figs. 9a and 10a) and usually
across tailwinds when returning to the colony (Figs. 22b and 23b).
The response of boobies from LA to wind direction was more straightforward. When heading to
the feeding grounds, birds preferentially flew into head winds (25%), across headwinds (45%), and across
winds (25%), whereas heading of inbound flights was reversed: travelling across (21%), across tail (51%)
and with tailwinds (24%). Thus, birds from LA predominantly flew into head winds during outbound paths
and with the wind during inbound paths.
The flight speed was on average 43.9 ± 3.79 km·h-1, with burst speeds up to 139 km·h-1. Mean
flight speed of birds from both islands progressively increased from 39 - 41 km·h-1 with headwinds to 61 -
65 km·h-1 with tailwinds (Fig. 24).
71
Figure 22. Observed and expected flight direction relative to wind direction of Peruvian Boobies from Isla
Lobos de Tierra in December 2006 during (a) outbound and (b) inbound flights.
(a)
0
10
20
30
40
50
Head Accross head Across Across tail Tail
Fre
quen
cy (%
)
ObservedExpected
χ2 = 16.62, df = 3
P < 0.001
(b)
0
10
20
30
40
50
Head Accross head Across Across tail Tail
Fre
quen
cy (%
)
ObservedExpected
χ2 = 27.53, df = 3
P < 0.001
72
Figure 23. Observed and expected flight direction relative to wind direction of Peruvian Boobies from Isla
Lobos de Afuera in December 2007 during (a) outbound and (b) inbound flights.
(a)
0
10
20
30
40
50
Head Accross head Across Across tail Tail
Fre
quen
cy (%
)
ObservedExpected
χ2 = 67.64, df = 3
P < 0.001
(b)
0
10
20
30
40
50
60
Head Accross head Across Across tail Tail
Fre
quen
cy (%
)
ObservedExpected
χ2 = 84.18, df = 3
P < 0.001
73
Figure 24. Mean flight speed (km·h-1) of Peruvian Boobies (Sula variegata) relative to wind direction on (a)
Isla Lobos de Tierra in December 2006 and (b) Isla Lobos de Afuera in December 2007.
0
20
40
60
80
Flight direction relative to wind direction (o)
Mea
n fli
ght s
peed
± 1
s.d
.
0 20 40 60 80 100 120 140 160 180
Head Across head Across Across tail Tail
(b)
0
20
40
60
80
Flight direction relative to wind direction (o)
Mea
n fli
ght s
peed
± 1
s.d
.
0 20 40 60 80 100 120 140 160 180
Head Across head Across Across tail Tail
74
DISCUSSION This study highlights the influence of abiotic factors on the foraging movements of chick-rearing Peruvian
Boobies. Oceanographic conditions within the study area were typical of coastal upwelling systems, with
high levels of marine productivity and relatively cold waters that favor the presence of anchovetas
(Gutiérrez et al. 2007). Water masses were distributed heterogeneously in a coarse meso scale (10 - 100
km) in a narrow onshore strip, with areas of enhanced productivity concentrated alongshore. Here,
anchovetas are packed in dense schools, being highly vulnerable to their predators. This spatial
asymmetry of food conditions during the summer is seasonally predictable (Muck et al. 1989, Segura
2000), so it is expected that birds from different islands will forage in zones close to the mainland.
Peruvian Boobies from LT responded to oceanographic features by feeding in areas of higher marine
productivity and warmer waters than expected, whereas the environmental conditions occupied by birds
from LA were similar to those available or crossed. Although foraging areas of Peruvian Boobies from LT
were located close to the mainland, birds from LA did not approach the continental coastline, thus
indicating that birds from and inshore and offshore island have different foraging constraints.
Effects of SST and chla concentration
The results from LT suggest that the excursions of Peruvian Boobies to the east can be explained in terms
of the prevailing oceanographic conditions in December 2006. Remotely-sensed data showed that areas
of enhanced marine productivity were located at inshore waters east of LT and delimited by water masses
of 20oC. Similar oceanographic conditions in the same geographical area were observed in December
2007 (Figs. 6a and 8a). Although 17oC represents the optimal mean temperature of the anchoveta, they
can still survive in waters up to 21oC (Gutiérrez et al. 2007). This preferred feeding zone by Peruvian
Boobies is linked to the seasonal horizontal movement of anchovetas. During the austral summer or El
Niño Southern Oscillation (ENSO) events, when the sea surface temperature increases, there is an
onshore displacement of the 17oC and 20oC isotherms, a decrease in the size of the anchoveta habitat,
and an increase of school densities close to the mainland (Muck et al. 1989). The seasonal dynamics of
sea-surface temperature suggest that Peruvian Boobies can take advantage of the predictable conditions
75
of the environment (except during ENSO events) to locate food at a coarse-meso scale. The consistent
association between the summer onshore aggregation of anchovetas and the observed eastward
movement of breeding Peruvian Boobies from LT is supported by observations from other studies. Duffy
(1987) observed that breeding Peruvian Boobies returned from the east of LT after feeding. Ship-based
observations indicated that higher densities of Peruvian Boobies occurred onshore during the summer
than during the winter (Jahncke et al. 1998). Blue-footed Boobies feeding on anchovetas also foraged
east of LT during the summer of 2003 (Zavalaga et al. 2008).
Unlike at LT, the southwest distribution of Peruvian Boobies from LA was not influenced by
hydrographic features. Birds occupied water masses with SST and chla concentration similar to those in
the available or crossed areas, indicating that the links between oceanographic features and bird’s
foraging behavior is not always correlated. The low sea surface temperatures (<19oC) within the whole
available area and high levels of chla (median 1.23 mg·m-3) suggest that anchovetas may have been
equally distributed around LA, and therefore, boobies may have foraged at any cardinal sector around the
island. Regions of high chla concentration occurred within the foraging range of the boobies, along
inshore waters east of LA, with hot spots of productivity located northwest (Fig. 8a). Why didn’t Peruvian
Boobies from LA visit inshore areas of high productivity located northwest of the island? We provide some
non-mutually exclusive explanations for this question.
Competition for food with seabirds from LT.- The population size of seabirds breeding on LT is at least 4-5
times higher than LA, with Blue-footed Boobies, Peruvian Pelicans and Peruvian Boobies being the most
abundant. Because Peruvian and the sympatric Blue-footed Boobies from LT prey upon anchovetas in
areas close to the mainland (this study, Zavalaga et al. 2008), inter and intraspecific competition for
feeding areas may be intense in the proximity of the island. The fact that the areas occupied by Peruvian
Boobies from both islands did not overlap, despite the capacity of birds from LA to occupy 68% of the
available foraging area of LT, confirms that boobies from both islands avoided feeding in the same areas.
Because birds from both islands were tracked at different years, it can be argued that in any given year
the foraging areas may have widely overlapped. However, Blue-footed Boobies tracked simultaneously on
LT and LA in December 2007 occupied different areas for foraging (Chapter 2.), supporting the concept of
spatial segregation. Such a separation of foraging areas supports the predictions of the hinterland model
76
(Cairns 1989) which propose that neighboring colonies do not overlap in potential foraging range and that
colony size is a function of the feeding areas.
Restricted foraging range.- To avoid interference competition with other anchoveta predators close to LT,
Peruvian Boobies from LA may have travelled to inshore waters east and southeast of the island to areas
of cold waters and enhanced productivity (Figs. 6a and 8a), but they did not explore water to the east and
did not venture to a distance away of > 67 km. During the breeding season, Peruvian Boobies are obligate
central-foragers (Orians and Pearson 1979), alternating multiple short trips in a day with periods in the
nest. This foraging strategy may be related to the capacity of Peruvian Boobies to sustain a brood of 2 – 3
chicks (occasionally 4 chicks), the largest brood within the Sulidae (Nelson 1978). The high energetic
demands of an enlarged brood may force adults to feed at least twice per day (Chapter 2), restricting the
foraging range to a few kilometers from the colony. Based on the Peruvian Booby average flight speed of
44 km·h-1, to the necessity to fly and feed in 13 h of daylight (Peruvian Boobies are diurnal foragers), and
an average of 92% of their foraging time spent on flight (Chapter 2), the maximum foraging distance to
accomplish two trips in a day (maximum trip duration of 3.25 h, assuming that the mate is also foraging
twice a day) would be 65.78 km [(44 x 3.25 x 0.92)/2], very close to the observed maximum foraging
distance of 67.1 km. Thus, to be able to forage twice per day, Peruvian Boobies are limited to forage
within a radius of 66 km. The predictable food sources to the east and southeast are located > 90 km from
the island, far from the maximum range of Peruvian Boobies. A comparative study of foraging range of
breeding Peruvian Boobies with different brood size are necessary to test this hypothesis.
Prevailing food conditions during the study period.- The observed foraging asymmetries in boobies from
LA versus LT can also be attributed to factors other than the heterogeneity of the marine environment. It is
possible that the observed skewed distribution of boobies could be the result of the prevailing conditions
during the limited study period. Ephemeral prey aggregations may have been present southwest of the
island, but then may have been depleted or moved to other sites in subsequent weeks. Anchovy
aggregations are highly mobile, even during a single day (Bertrand et al. 2004), thus conditioning the
bird’s movements. For example, using data sets from December 2007 also used here (Chapter 2) found
77
that the mean foraging range of Peruvian Boobies decreased from 40 to 20 km in less than a two-day
interval.
Social behavior.- “Cultural foraging patterns” (Grémillet et al. 2004) may also play a role in the birds’
foraging asymmetries because birds tend to aggregate in the same feeding areas by sighting multi -
species flocks in conspicuous feeding frenzies around their colonies (Camphuysen and Webb 1999, Duffy
1983). They also may be using the direction of returning bird groups as information centers (Clode 1993)
or returning to areas where they previously experienced successfully foraging (Davoren et al. 2003).
Peruvian Boobies are also occasionally attracted to the anchoveta purse-seine fishing vessels
(Weichler et al. 2004). Flocks of hundred of boobies are seen feeding within the encircled school when the
net is hauled in. The extent of the interaction between Peruvian Boobies and anchoveta fisheries is
unknown and deserve further research to examine how fishing activities influence the at-sea spatial
distribution of birds.
Bathymetric features.
This study demonstrates that the foraging range of Peruvian Boobies is restricted not only to waters over
the continental shelf (depths < 100 m), but it also extended to waters over the shelf break and slope (100 –
2500 m). Deep oceanic waters (> 2500 m) were never visited by Peruvian Boobies. Birds from LA may
have occupied shallow waters to the east, but unexpectedly they preferred to forage over deeper waters.
Ship-based surveys along the Peruvian coast indicated that Peruvian Boobies fed generally in neritic
waters over the continental plateau and slope, and that the density of birds was correlated to anchovy
abundance (Jahncke et al. 1998). In a meso spatial scale (100s of km), static physical features such as
the bathymetry of the continental shelf and slope are good proxys for Peruvian Booby locations because
high levels of ocean productivity are generally observed over the continental shelf (Farias et al. 2004) as
well as the continental shelf-breaks and slopes (Croll et al. 1998, Yen et al. 2004). At a lower spatial scale
the ultimate cause of the foraging asymmetries would be the prevailing conditions of food (Guinet et al.
2001).
78
Wind patterns
Wind direction had a major impact on flight direction of Peruvian Boobies. Birds used across or across
headwinds when heading to their feeding grounds, and across tailwinds when returning to their colonies.
The preference for side winds has also been observed in other booby species (Spear and Ainley 1997,
Weimerskirch et al. 2005) and seems to be related to their style of flight and wing morphology. Boobies
display relatively low wing loading (Hertel and Balance 1999) and are considered glide-flappers, i.e., they
alternate flapping with short periods of gliding (Spear and Ainley 1997). Across winds may represent an
optimal wind direction to efficiently perform glide-flap flights (Weimerskirch et al. 2004) and an
intermediate point between high energetic costs by flying with headwinds (Adams and Navarro 2005) and
low rates of prey encounter by moving fast with tailwinds (Spear and Ainley 1997). Although flying with the
wind may be disadvantageous when searching for food, it is favorable when birds return to their nests with
heavy food loads because it increases ground flight speeds (Adams and Navarro 2005). By using across
tailwinds during the inbound flights, Peruvian Boobies may increase mean ground flight speeds by 38 -
53% in relation to headwinds, resulting in substantially decreased flight costs.
Because southeasterly winds were recurrent in the study area, we expected that Peruvian
Boobies would preferentially forage south of the islands to take advantage of tailwinds when returning to
their colonies. This prediction appears to be only partly supported. Birds from LA foraged extensively
southwest of the island and returned with across tail and tailwinds (75% of records). On the other hand,
birds from LT were oriented to the east of the island and, although half of the birds returned with across
tailwinds, the other half did not show any orientation preferences during the inbound flights. We speculate
that the orientation of birds from LT to the east was dictated by food conditions rather than wind direction.
The results from this study suggest that the marine habitat use of breeding Peruvian Boobies is influenced
by the proximity of the colonies to the mainland. The seasonal dynamics of oceanographic conditions in
the Humboldt Current systems is the most cyclic and predictable of the large components of variability.
The consistent onshore displacement of cold upwelled waters during the summer leads to the aggregation
of anchovetas and their predators in waters close to the mainland. Peruvian Boobies at inshore colonies
may take advantage of the food distribution predictability driven by oceanographic changes, but birds from
79
offshore islands are constrained by other ecological factors. Thus, habitat use patterns can also be
interpreted in the context of life-history limitations. The restricted foraging range of Peruvian Boobies
seems to be associated with the high energetic demands of an enlarged brood, forcing the adults to
perform multiple trips per day and consequently, to forage relatively close to the coast. Thus, birds from
offshore islands cannot have access to inshore areas of enhanced productivity, relying mainly on other
abiotic or biotic factors to locate prey such as wind patterns, bathymetry, social attraction to multi-species
feeding flocks and commercial fishing boats. Additional research is needed to examine how brood size
influences foraging movements of Peruvian Boobies and how changes in foraging effort affect breeding
success.
Peruvian Boobies exhibited a high flexibility in foraging behavior, being able to exploit either
coastal waters of < 15 m deep or waters on the continental break and slope (100 – 2500 m). This wide
range of habitat use expose boobies to interact with both the commercial and the coastal inshore fishery.
The effects of the fishing activities on the spatial distribution of boobies deserve further evaluation
because competition for the same feeding/fishing areas between fishers and birds can be intense.
Commercial fishing vessels are currently tracked through a satellite vessel monitoring system for the
entire industrial fleet (Bertrand et al. 2007), and fishing areas can be overlaid to the booby foraging zones.
The degree of overlapping may provide new insights about the direction and magnitude of the interactions
between the commercial fisheries and boobies. This monitoring program can be extended to the artisanal
fishery, as high levels of Peruvian Booby mortality due to entanglement in fishing nets have been recently
detected at inshore waters off Peru (J. Alfaro, personal communication).
80
REFERENCES Adams NJ, Navarro RA (2005) Foraging of a coastal seabird: flight patterns and movements of breeding
Cape gannets Morus capensis Afr. J. Mar. Sc. 27:239-248.
Bakun A (1987) Monthly variability in the ocean habitat off Peru as deduced from maritime observations,
1953 to 1984. In: Pauly D, Tsukayama I (eds) The Peruvian anchoveta and its upwelling ecosystem:
three decades of change. ICLARM Studies and Reviews 15. Instituto del Mar del Peru (IMARPE),
Callao, Peru; Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ) GMBH, Eschbom,
Federal Republic of Germany; and International Center for Living Aquatic Resources Management
(ICLARM), Manila, Philippines, p 46-74.
Batschelet E (1981) Circular statistics in biology. - Academic Press, New York.
Baylis AMM, Page B, Goldsworthy SD (2008) Effect of seasonal changes in upwelling activity on the
foraging locations of a wide-ranging central-place forager, the New Zealand fur seal. J. Can. Zool.
86:774-789.
Bertrand S, Bertarnd A, Guevara-carrasco R, Gerlotto F (2007) Scale-invariant movements of fishermen:
the same foaraging strategy as natural predators. Ecological Applications 17:331-337.
Bertrand S, Díaz E, Ñiquen M (2004) Interactions between fish and fisher’s spatial distribution and
behaviour: an empirical study of the anchovy (Engraulis ringens) fishery of Peru. ICES Journal of
Marine Science 61:1127-1136.
Biuw M, Boehme L, Guinet C, Hindell M, Costa D and others (2007) Variations in behavior and conditions
of a Southern Ocean top predator in relation to in situ oceanographic conditions. Proc. Nat. Acad.
Sci. USA 104:13705-13710.
Brainard RE, McLain DR (1987) Seasonal and interannual subsurface temperature variability off Peru,
1952 to 1984. In: Pauly D, Tsukayama I (eds) The Peruvian anchoveta and its upwelling ecosystem:
three decades of change. ICLARM Studies and Reviews 15. Instituto del Mar del Peru (IMARPE),
Callao, Peru; Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ) GMBH, Eschbom,
Federal Republic of Germany; and International Center for Living Aquatic Resources Management
(ICLARM), Manila, Philippines, p 14-45.
Cairns DK (1989) The regulation of seabird colony size. American Naturalist 134:141-146.
Camphuysen KCJ, Webb A (1999) Multi -species feeding associations in north sea seabirds: jointly
exploiting a patchy environment. Ardea 87:177-198.
Clode D (1993) Colonially breeding seabirds: predatory or prey? Trends Ecol. Evol. 8:336-338.
Croll DA, Tershy BR, Hewitt RP, Demer D, Fiedler PC, Smith SE, Amstrong W, Popp JM, Kiekhefer T,
Lopez VR, Urban J (1998) An integrated approach to the foraging ecology of marine birds and
mammals. Deep-Sea Res. II 45:1353-1371.
Davoren GK, Montevecchi WA, Anderson JT (2003) Search strategies of a pursuit-diving marine bird and
the persistence of prey patches. Ecol. Monogr. 73: 463-481.
Duffy DC (1983) The foraging ecology of Peruvian seabirds. Auk 100:800-810.
81
Duffy DC (1987) Aspects of the ecology of Blue-footed and Peruvian boobies at the limits of their ranges
on Isla Lobos de Tierra, Perú. Colonial Waterbirds 10:45-49.
Espinoza P (2000) Interacción trófica merluza-anchoveta; existe realmente impacto por predación?
Boletín Instituto del Mar del Perú 19:15-20.
Farias L, Graco M, Ulloa O (2004). Temporal variability of nitrógeno cycling in continental-shelf sediments
of the upwelling ecosistema off central Chile. Deep-sea Res. II 51:2491-2505.
Figueroa J, Stucchi M (2008) Las aves de la Isla Lobos de Afuera (Perú) en la primavera del 2008.
Ornitología Neotropical 19: 377-390.
Ganoza F, Castillo PR, Marín D (2000) Variaciones estacionales en la distribución y biomasa de
anchoveta entre 1983 y 2000. Boletín Instituto del Mar Perú 19:157-177.
Godley BJ, Richardson S, Broderick, AC, Coyne MS, Glen F, Hays GC (2002) Long-term saltellite
telemetry of the movements and habitat utilization by green turtles in the Mediterranean. Ecography
25:352-362.
Grémillet D, Dell’Omo G, Ryan PG, Peters G, Ropert-Coudert Y, Weeks SJ (2004) Offshore diplomacy or
how seabirds mitigate intra-specific competition: a case study on GPS tracking of Cape gannets
from neighbouring colonies. Mar. Ecol. Prog. Ser. 268: 265-279.
Grémillet D, Lewis S, Drapeau L, van Der Lingen CD, Huggett JA, Coetzee JC, Verheye HM, Daunt F.
Wanless S, and Ryan PG (2008) Spatial match-mistach in the Benguela upwelling zone: should we
expect chlorophyll and sea-surface temperature to predict marine predator distributions? J Applied
Ecol. 45:610-621.
Guinet C, Dubroca L, Lea MA, Goldsworthy S, Cherel Y, Duhamel G, Bonadonna F, Donnay J-P (2001)
Spatial distribution of foraging in female Antarctic fur seals Arctocephalus gazelle in relation to
oceanographic variables: a scale -dependent approach using geographic information systems. Mar
Ecol. Prog. Ser. 219:251-264.
Gutiérrez M, Swartzman G, Bertrand A, Bertrand S (2007) Anchovy (Engraulis ringens) and sardine
(Sardinops sagax) spatial dynamics and aggregation patterns in the Humboldt Current ecosystem,
Peru, from 1983-2003. Fish. Oceanogr. 16:155-168.
Hatch SA, Meyers PM, Mulcahy DM, Douglas DC (2000) Seasonal movements and pelagic habitat use of
murres and puffins determined by satellite telemetry. Condor 102:145-154.
Hertel F, Balance LT (1999) Wing ecomorphology of seabirds from Johnston Atoll. Condor 101:549-556.
Hunt GL Jr, Schneider DC (1987) Scale dependent processes, prey abundance, and the foraging ecology
of seabirds. In: Croxall JP (ed) The feeding ecology of seabirds and their role in marine
ecosystems. Cambridge University Press, Cambridge, p 7-41.
Huyer A (1983) Coastal upwelling in the California Current system. Prog. Oceanogr. 12:259-284.
Hyrenbach KD, Fernández P, Anderson DJ (2002) Oceanographic habitats of two sympatric North Pacific
albatrosses during the breeding season. Mar. Ecol. Prog. Ser. 233:283-301.
Irons DB (1998) Foraging area fidelity of individual seabirds in relation to tidal cycles and flock feeding.
Ecology 79:647-655.
82
Jahncke J, Zileri D (1998) Estudios sobre dieta de piqueros como indicadores de la estructura por tallas
de los stocks de anchoveta en el mar peruano. Boletín Instituto del Mar del Perú 17:47-54.
Jahncke J, Ayala L, Mendoza C (1998) Abundancia relativa y distribución de aves marinas frente a la
costa peruana y su relación con la abundancia de anchoveta. Crucero BIC Humboldt 9808-09, de
Piura a Lima. - Inf. Inst. Mar Perú 141: 85-95.
Jouventin P, Weimerskirch H (1990) Satellite tracking of wandering albatrosses. Nature 343:746-748.
Majluf P (1989) Reproductive ecology of South American fur seals in Peru. In: Pauly D, Muck P,
Tsukayama I (eds) The Peruvian upwelling ecosystem: dynamics and interactions, ICLARM
Conference Proceedings 18. Instituto del Mar del Perú (IMARPE), Callao, Perú; Deutsche
Gesellschaft für Technische Zusammenarbeit (GTZ) GMBH, Eschbom, Federal Republic of
Germany; and International Center for Living Aquatic Resources Management (ICLARM), Manila,
Philippines, p 332-343.
Morón O (2000) Características del ambiente marino frente a la costa peruana. Boletín Instituto del Mar
del Perú 19:179-204.
Muck P, Zafra de Moreno A, Aranda C (1989) The seasonal dynamics of sea surface temperature and its
impact on anchoveta distribution off Peru. In: Pauly D, Muck P, Tsukayama I (eds) The Peruvian
upwelling ecosystem: dynamics and interactions, ICLARM Conference Proceedings 18. Instituto del
Mar del Perú (IMARPE), Callao, Perú; Deutsche Gesellschaft für Technische Zusammenarbeit
(GTZ) GMBH, Eschbom, Federal Republic of Germany; and International Center for Living Aquatic
Resources Management (ICLARM), Manila, Philippines, p 33-44.
Nelson JB (1978) The Sulidae: Gannets and Boobies. Oxford University Press, Oxford.
Lowry LF, Burkanov VN, Frost KJ, Simpkins MA, Davis, R, DeMaster DP, Suydam R, Springer A (2000)
Habitat use and habitat selection by spotted seals (Phoca largha ) in the Bearing Sea. Can. J. Zool.
78:1959-1971.
Orians GH, Pearson NE (1979) On the theory of central palce foraging. In Horn DJ, Mitchell R, Stair GR
(eds). Analysis of ecological systems. Ohio State University Press, Columbus, p 155-177.
Pauly D, Christensen V, Guénette S, Pitcher TJ, Sumaila UR, Walters CJ , Watson R, Zeller D (2002)
Towards sustainability in world fisheries. Nature 418: 689-695.
Piatt JF, Wells JL, MacCharles A, Fadeley BS (1989) The distribution of seabirds and fish in relation to
ocean currents in the southeastern Chukchi Sea. In: Montevecchi WA, Gaston AJ (eds) Studies of
high-latitude seabirds. 1. Behavioral, energetic, and oceanographic aspects of seabirds feeding
ecology. Ocassional Paper No 34. Canadian Wildlife Service, p 21-31.
Raya-Rey A, Trathan P, Pütz, Schiavini A (2007) Effect of oceanographic conditions on the winter
movements of rockhopper penguins Eudyptes chrysocome chrysocome from Staten Island,
Argentina. Mar. Ecol. Prog. Ser. 330:285-295.
Segura M (2000) Distribución horizontal de la anchoveta peruana con relación a variables ambientales en
el período 1986-2000. Boletín Instituto del Mar del Perú 19:55-82.
83
Shaffer SA, Trembaly Y, Weimerskirch H, Cott D, Thompson DR, Sagar PM, Moller H, Taylor GA, Foley
DG, Block BA, Costa DP (2006) Migratory shearwaters integrate oceanic resources across the
Pacific Ocean in an endless summer. Proc. Nat. Acad. Sci. USA 103:12799-12802.
Sokal RR, Rohlf FJ (1995) Biometry. WH Freeman and Company, New York.
Spear LB, Ainley DG (1997) Flight behaviour of seabirds in relation to wind direction and wing
morphology. Ibis 139:221-233.
Tovar H, Guillén V (1988) Comportamiento alimenticio del piquero Sula variegata , ave guanera peruana.
Boletín de Lima 60:85-90.
Weichler T, Garthe S, Luna-Jorquera G. Moraga J (2004) Seabird distribution on the Humboldt Current in
northern Chile in relation to hydrography, productivity, and fisheries. ICES J. Mar. Sci. 61:148-154.
Wemerskirch H, Le Corre M, Jaquemet S, Potier M, Marsac F (2004) Foraging strategy of a top predator
in tropical waters: great frigatebirds in the Mozambique Channel. Mar. Ecol. Prog. Ser. 275:297-
308.
Weimerskirch W, Le Corre M, Roper-Coudert Y, Kato A, Marsac F (2005). The three-dimensional flight of
red-footed boobies: adaptations to foraging in a tropical environment? Proc. R. Soc. B 272:53-61.
Yen PPW, Sydeman WJ, Hyrenbach KD (2004) Marine bird and cetacean associations with bathymetric
habitats and shallow-water topographies: implications for trophic transfer and conservation. J. Mar.
Sys. 50:79-99.
Zavalaga CB, Benvenuti S, Dall’Antonia L, Emslie SD (2007) Diving behavior of Blue-footed Boobies Sula
nebouxii in northern Peru in relation to sex, body size and prey type. Mar. Ecol. Prog. Ser. 316-291-
303.
Zavalaga CB, Benvenuti S, Dall’Antonia L, Emslie SD (2008) Foraging areas of breeding blue-footed
boobies Sula nebouxii in northern Peru, as determined by direction recorders. J. Avian Biol. 39:405-
412.