Alca torda Atlantic Canada by

Alca torda Atlantic Canada  by
Offshore Movements and Behaviour of the Razorbill (Alca torda) in
Atlantic Canada
by
Travis Christopher Randolph Clarke
BSc., Memorial University of Newfoundland, 2004
This thesis is submitted in partial fulfillment
of the requirements for the degree of
Masters of Science
in the graduate academic unit of biology
Supervisor:
Antony Diamond, PhD, Biology, and Forestry and
Environmental Management
Examining Board:
Simon Courtenay, PhD, Biology
Barry Monson, PhD, Math and Statistics
This thesis is accepted by the Dean of Graduate Studies
THE UNIVERSITY OF NEW BRUNSWICK
January, 2009
© Travis C.R. Clarke, 2009
To My Wife
ii
Abstract
The majority of our knowledge concerning the Razorbill comes from studies
conducted during the breeding season when the bird is on or near land. I deployed three
different types of electronic devices on Razorbills in order to gain insight into their
movements and behaviour while they were away from the breeding colony. The first
type, satellite transmitters, were deployed on Razorbills (n = 8) at Machias Seal Island
(MSI), New Brunswick to explore relationships between Razorbill distribution and 4
oceanographic features (sea surface temperature, chlorophyll concentration, bathymetric
depth and the slope of the ocean floor). Razorbill density was correlated with bathymetric
depth and slope but no significant relationship with sea surface temperature or
chlorophyll concentration was found, suggesting that the birds were more closely
associated with physical characteristics of the ocean floor than with the effects of
upwelling and mixing. The second type, a geo-locator, was deployed on a breeding
female adult (n = 1) from MSI to assess diving behaviour. Diving behaviour was different
in the breeding season, when the bird concentrated most of its diving in the afternoon,
than in the non-breeding season, when both the morning and afternoon were used.
Diving rarely occurred at dusk, dawn or night, and dives never exceeded 40 m. The third
type, radio transmitters, were deployed on Razorbills from MSI (n = 14), Corossol Island,
Quebec (n = 10) and the Gannet Islands, Labrador (n = 15) to test the hypothesis that
Razorbills from breeding sites outside of the Bay of Fundy contribute to the large
(~10,000 individuals in 2008) wintering population south of Grand Manan. During flights
conducted from January – March, 93% of the transmitters deployed at MSI, 60% of those
deployed at Corossol Island, and 21% of those deployed at the Gannet Islands were
iii
detected. While it is possible that some transmitters failed, the asymmetric representation
of the three breeding population suggests that the Bay of Fundy is an important wintering
area for some of the largest breeding sites in North America but also raises the question
of whether there are other such congregations in the area.
iv
Acknowledgements
Thanks first and foremost to my supervisor Dr. Tony Diamond for supplying constant
advice, helping to secure funding and equipment and for all the time spent reviewing my
thesis. Thanks also to my committee members Dr. Graham Forbes and Dr. John Chardine
for providing advice and direction and helping to secure funding.
Dr. Jim Berry’s veterinary expertise was an invaluable component of this project and his
assistance is greatly appreciated.
The Razorbill is equally at home in the water, on the land and in the air and as such this
project succeeded only with the hard work and help of many folks in all three
environments. Thanks to Marie-Paule Godin, Raphaël Lavoie, Gabrielle Bouchard, Alex
Bond, Robin MacNearney, Emily Hines, Kirk Hart, Peter Bohan and Colin Buhariwalla
for help in the field. Thanks to Andy Patterson, Peter Wilcox, Les Hamil and Daryl
Holwell for transportation to and from Machias Seal Island and the Gannet Islands and
thanks to pilots Klaus and Peter from the Atlantic Charter Company. Thanks also to the
lighthouse keepers at Machias Seal Island for all their help and hospitality.
Help with data management, analysis and interpretation was kindly provided by Myriam
Barbeau and Michel Melancon.
Thanks to my lab mates for their ideas, questions, enlightening discussions and genuine
interest in my work.
Finally, I would like to thank my wife, Nelly, who has tolerated my summer absences for
far too many years. You’ve been a constant source of support and inspiration.
Funding and equipment for this project was provided by the Canadian Wildlife Service
CWS (thanks to Scott Gilliland, Greg Robertson and Jean-Francois Rail), the Atlantic
Cooperative Wildlife Ecology Research Network, the University of New Brunswick,
Science Horizons and the Northern Scientific Training Program.
v
Table of Contents
Dedication………………………………………………………………………… ii
Abstract…………………………………………………………………………… iii
Acknowledgments………………………………………………………………... v
Table of Contents………………………………………………………………… vi
List of Tables……………………………………………………………………... viii
List of Figures………………………………………………………………….…. ix
Chapter 1 – General Introduction………………………………………………… 1
References…………………………………………………………... 5
Chapter 2 - The Effects of Oceanographic Conditions on the Movements
of Razorbills (Alca torda) in the Bay of Fundy………………………
Abstract………………………………………………………………
Introduction…………………………………………………………..
Methods………………………………………………………………
Study area……………………………………………………..
Satellite telemetry………………………………………….….
Oceanographic data…………………………………………...
Quality control of PTT locations……………………………..
Data treatment and analysis…………………………………...
Results………………………………………………………………..
Transmitter performance……………………………………...
General movements…………………………………………...
Spatial patterns in oceanographic features……………………
Discussion……………………………………………………………
Transmitter performance……………………………………..
General movements…………………………………………..
Spatial patterns in oceanographic features……………………
Tables………………………………………………………………...
Figures………………………………………………………………..
References……………………………………………………………
6
6
6
7
7
8
9
10
11
13
13
14
14
16
16
17
18
23
28
36
Chapter 3 - Case Study of Razorbill (Alca torda) Diving Behaviour………….…
Abstract………………………………………………………………
Introduction…………………………………………………………..
Methods………………………………………………………………
Geo-locators…………………………………………………..
Diving behaviour……………………………………………...
Light-based locations…………………………………………
41
41
41
43
43
43
44
vi
Results………………………………………………………………
Diving behaviour……………………………………………..
Non-breeding season……………………………………….…
Breeding season………………………………………………
Light-based locations…………………………………………
Discussion……………………………………………………………
Seasonal differences…………………………………………..
Diving behaviour……………………………………………..
Tag performance……………………………………………...
Tables……………………………………………………………….
Figures…………………………………………………………….…
References…………………………………………………………...
44
44
46
46
47
48
48
50
54
57
60
63
Chapter 4 – Wintering Razorbills (Alca torda) in the Bay of Fundy…………….
Abstract……………………………………………………………...
Introduction……………………………………………………….…
Methods……………………………………………………………..
Study area…………………………………………………….
Radio transmitters………………………………………….…
Flights………………………………………………………...
Results…………………………………………………………….…
Discussion…………………………………………………………...
Tables……………………………………………………………….
Figures…………………………………………………………….…
References…………………………………………………………..
66
66
66
68
68
68
69
70
71
77
78
80
Chapter 5 – Summary………………………………………………………….… 83
Appendix I – Razorbills equipped with Satellite Transmitters and identified
after transmitter failure………………………………………….…. 86
Appendix II – Tukey’s HSD Post-hoc Analysis for Razorbill dives…………….. 87
Appendix III – Damaged Radio Transmitters……………………………………. 90
vii
List of Tables
Table 2.1: Details of Argos transmitter classes and summary of locations used
from each class……………………………………………………….
Table 2.2: Summary of data from satellite transmitters attached to 8 adult
Razorbills from Machias Seal Island, NB……………………………
Table 2.3: Monthly breakdown of the area covered by each zone pictured in
Figure 2.3…………………………………………………………….
Table 2.4: Temporal (monthly) and spatial (zone) parameters (Mean ± SD) for
Razorbill locations associated with depth and angle of the ocean
floor…………………………………………………………………..
Table 2.5: Temporal (monthly) and spatial (zone) parameters (Mean ± SD) for
Razorbill locations associated with sea surface temperature and
chlorophyll concentration…………………………………………….
Table 3.1: Dive data for five daily intervals from the non-breeding season and
breeding season for an adult Razorbill from Machias Seal Island,
New Brunswick………………………………………………………
Table 3.2: Results of the three ANOVAs summarizing the diving behaviour of
an adult female Razorbill in 2006/7………………………………….
Table 3.3: Values of diving parameters characterizing the interactions between
diving depth and time of day appearing in Figures 2.2 and 2.3.
Values correspond to average percent of daily dives at depth at given
time of day……………………………………………………………
Table 4.1: Details of flights undertaken to locate Razorbills equipped with radio
transmitters near Grand Manan from Jan to March 2008…………….
Appendix I
Table 1: Results of Tukey’s HSD post-hoc analysis showing significant
distribution of diving depths achieved during the non-breeding
season with respect to time of day……………………………………
Table 2: Results of Tukey’s HSD post-hoc analysis showing significant
distribution of times of day during the breeding season with respect
to depths achieved…………………………………………………….
Table 3: Results of Tukey’s HSD post-hoc analysis showing significant
distribution of diving depths achieved during the breeding season
with respect to time of day……………………………………………
Table 4: Results of Tukey’s HSD post-hoc analysis showing significant
distribution of times of day during the breeding season with respect
to depths achieved…………………………………………………….
viii
23
24
25
26
27
57
58
59
77
87
88
88
89
List of Figures
Figure 2.1: Razorbill breeding colonies in the Bay of Fundy and all Razorbill
locations from the entire duration of the PTT deployment……………….
Figure 2.2: Platform Terminal Transmitter 100 (PTT-100) abdominal implants
manufactured by Microwave Telemetry, Inc. Satellite Transmitter……...
Figure 2.3: Sub-divided monthly Razorbill distribution in the Bay of Fundy region
from June to October 2007.………………………………………………
Figure 2.4: Detailed of voltage and body temperature for Razorbill 65613, which was
presumed dead in early August when it appeared on Head Harbour
Island, Maine……………………………………………………………..
Figure 2.5: Map of bathometric depth in the Bay of Fundy region and main Razorbill
colonies…………………………………………………………………...
Figure 2.6: Depth associated with changing areas of Razorbill monthly
distribution.……………………………………………………………...
Figure 2.7: Mean angle of the slope of the continental shelf associated with
increasing Razorbill density………………………………………………
Figure 2.8: Map of the slope of the ocean floor in the Bay of Fundy region and main
Razorbill colonies………………………………………………………...
Figure 2.9: Mean monthly sea surface temperatures associated with increasing
Razorbill density………………………………………………………….
Figure 2.10: Figure 2.10: Chlorophyll concentration associated with changing areas
of monthly Razorbill distribution…………………………………………
Figure 3.1: LTD2400_50 geo-locator manufactured by Lotek Wireless, Inc. attached
to a plastic coiled band by two plastic zip-ties…………………………...
Figure 3.2: The use by an adult female Razorbill of each depth category throughout
the day in the non-breeding season……………………………………….
Figure 3.3: The use by an adult female Razorbill of each depth category throughout
the day in the breeding season……………………………………………
Figure 3.4: Map of Bay of Fundy region and the center of the locations of a geolocator deployed on an adult female Razorbill from the Machias Seal
Island breeding colony from 4 July 2006 to 7 July 2007…………………
Figure 4.1: Atlantic Canada showing the Gannet Islands, Machias Seal Island and
Corossol Island. Also included are Postville, where a bird from the
Gannet Islands was shot, and Grand Manan……………………………...
Figure 4.2: Map of the Grand Manan region, New Brunswick and the area around
Old Proprietor Shoal, where Razorbills were monitored…………………
ix
28
29
30
31
32
33
33
34
35
35
60
61
61
62
78
79
Appendix I
Figure 1: Photo of a male Razorbill (ID = 65598) highlighting the antenna of a
satellite transmitter, taken in July 2007, approximately 9 months after
last contact with the PTT. Transmission period was 116
days………………………………………………………………………. 86
Figure 2: Photo of a female Razorbill (ID = 65605) highlighting the antenna of a
satellite transmitter, taken in July 2008, approximately 20 months after
last contact with the PTT. Transmission period was 131
days………………………………………………………………………. 86
Appendix II
Figure 1: Two radio transmitters highlighting damage suffered after being deployed
on an adult Razorbill……………………………………………………... 90
x
Chapter 1 – General Introduction
The Alcidae are a diverse family of marine pursuit-diving birds. The family,
found within the order Charadriiformes, consists of seven tribes and 23 species. Alcids
occupy the same ecological niche in the Northern hemisphere as penguins do in the south
and are found throughout Arctic, sub-Arctic and boreal waters of North America, Europe
and Asia. They have compact streamlined bodies and short wings but unlike penguins
(Spheniscidae), the alcids are strong fliers as well as skilled divers. Alcids are prime
examples of K-selected species as defined by a number of characteristics such as slow
maturity (non-breeding status can last up to five years) and low fecundity (production of
a single egg in most species), high annual survival rate among adults and a long life span
(Hipfner and Chapdelaine 2002).
The Razorbill (Alca torda) is a large, robust alcid, generally weighing between
500 and 890g with a wingspan of 63-68cm (Hipfner and Chapdelaine 2002). Like most
alcids its head, neck, back, tail and upper wing are black while its underparts are white.
The Razorbill can be distinguished from similar-looking alcids such as the Common
(Uria aalge) and Thick-billed (Uria lomvia) Murres by its heavier, laterally compressed
bill, and the thin white line running from the base of the bill to each eye (Hipfner and
Chapdelaine 2002).
The North American distribution of the Razorbill extends from the Arctic waters
off western Greenland, down the Labrador coast to the boreal waters off Atlantic Canada
and the United States (Gaston and Jones 1998). However, in spite of this large range, the
total breeding population of the Razorbill in North America is only 38,000 breeding pairs
1
(Hipfner and Chapdelaine 2002), making the Razorbill the least abundant species of auk
in North America (Chapdelaine et al. 2001).
Historically, heavy persecution from humans for eggs, meat, and feathers, reduced
North American Razorbill populations across much of their northwest Atlantic breeding
range early in the twentieth century (Hipfner and Chapdelaine 2002). Currently, in North
America the Razorbill is protected by the Migratory Birds Convention Act and there are
no open hunting seasons (Chapdelaine et al. 2001). Some Razorbills, however, are
unintentionally mistaken for murres and shot during the annual “turr” hunt in
Newfoundland (Gaston and Elliot 1991) but despite this incidental take, the Razorbill
appears to have recovered substantially in most North American breeding sites
(Chapdelaine et al. 2001).
Like all seabirds, the majority of our knowledge concerning the Razorbill comes
from studies conducted during the breeding season. It is during this time, when the
Razorbill is on or near land and is engaged in socializing, incubating and brooding
(Hipfner and Chapdelaine 2002) that it can be studied most easily. Studies of seabirds
outside the breeding season often involve harsh winter conditions, isolated locations and
large distances and as such, it is very difficult to gather information regarding this aspect
of seabird ecology. Current knowledge of Razorbill, and all seabird, wintering locations
and migration routes is generally derived from one of the following methods; 1) Band
recoveries and re-sights, 2) counts or observations (from land or boat) or 3) attaching
electronic devices, which collect or transmit data.
Banding data can provide useful insight into seabird migration by identifying
dispersers. This applies largely to birds that are banded in one colony and then re-sighted
2
in other locations, often other breeding colonies (Lloyd and Perrins 1977; Lavers 2007).
But despite the value of the information that band recoveries and re-sights provide, the
information comes with limitations. Such data cannot identify when a bird arrived in a
particular area, how long it stayed, what route the individual followed to get there, or
what factors influence the bird’s relationship with the area in question.
The second method, observational studies and casual observations, has provided
additional insight into Razorbill migration (Gaston and Jones 1998; Huettmann et al.
2005). The advantage of observational studies is that, in general, large flocks of birds can
be identified and the flock size estimated. This gives the researcher some idea as to the
size of the population in the area at that time and allows them to attach a quantity with
each encounter. The disadvantage of observational studies is that, aside from species, the
identity of the birds is unknown. The observers cannot confirm where the flock came
from, how long it has been in the area, which breeding colony the birds are from, etc.
Furthermore, there is always the possibility that the same flocks are counted multiple
times and that many birds likely avoid detection by being out of range of the observer.
This thesis focuses on the third method, the attachment of electronic devices, to
explore the offshore movements and behaviour of the Razorbill. These techniques, rarely
if ever used on Razorbills, are expected to record and transmit information, thus
providing detailed information on the behaviour of individuals birds.
Through the use of bird-borne devices this thesis attempts to gain insight into this
relatively unknown area of Razorbill ecology, which can then be used to build on existing
knowledge from traditional studies. This thesis is presented in Articles format as outlined
by the University of New Brunswick School of Graduate Studies regulations and
3
guidelines for the preparation of graduate theses. Chapters 2, 3 and 4 are intended for
separate submission to journals after completion of this thesis.
4
References
Chapdelaine, G., A.W. Diamond, R.D. Elliot and G.J. Robertson. 2001. Status and
population trends of the Razorbill in eastern North America. Canadian Wildlife
Service Occasional Paper 105.
Gaston, A.J. and R.D. Elliot. 1991. Studies of high-latitude seabirds. 2. Conservation
biology of Thick-billed Murres in the Northwest Atlantic. Canadian Wildlife
Service, Occasional Paper 69.
Gaston, A.J. and I.L. Jones. 1998. The Auks. Oxford University Press, Oxford, U.K.
Hipfner, J.M. and G. Chapdelaine. 2002. Razorbill (Alca torda). In The Birds of North
America, No. 635 (A. Poole and F. Gill, eds.). The Birds of North America, Inc.,
Philadelphia, PA.
Huettmann, F., A.W. Diamond, B. Dalzell and K. Macintosh. 2005. Winter distribution,
ecology, and movements of Razorbills Alca torda and other auks in the outer Bay
of Fundy, Atlantic Canada. Marine Ornithology 33: 161-171.
Lavers, J.L., I.L. Jones and A.W. Diamond. 2007. Natal and breeding dispersal of
Razorbills (Alca torda) in eastern north America. Waterbirds 30: 588-594.
Lloyd, C.S. and C.M. Perrins. 1977. Survival and age at first breeding in the Razorbill
(Alca torda). Bird-Banding 48: 239-252.
5
Chapter 2: The Effects of Oceanographic Conditions on the Movements of
Razorbills (Alca torda) in the Bay of Fundy
Abstract
Little is known about Razorbill movements while they are away from the
breeding colony or the factors that influence their distribution. I deployed Argos satellite
transmitters on Razorbills (n = 8) breeding at Machias Seal Island from June 2006 to
December 2006 to assess Razorbill movements while away from the breeding colony and
determine the effects of four oceanographic characteristics on Razorbill spatial
distribution. The Razorbills maintained a close distance to shore and were not observed
outside of the Bay of Fundy. No correlation was observed between Razorbill distribution
and either sea surface temperature or chlorophyll concentration, however there is
evidence that Razorbill distribution is influenced by bathymetric depth and the slope of
the ocean floor. The Razorbills in this study were associated with factors that influence
upwelling and mixing but not with the effects themselves. Transmitter failure was
confirmed in two cases and suspected in others.
Introduction
The total breeding population of the Razorbill (Alca torda) in North America is
38,000 breeding pairs (Hipfner and Chapdelaine 2002), making it the least abundant
species of auk in North America. Because of low numbers, a remote distribution and
largely inaccessible nest sites (most Razorbills nest under boulders or on cliffs), the
Razorbill is often excluded or under-represented in studies involving seabird behaviour
and distribution (Piatt and Nettleship 1985; Wanless et al. 1988a; Wanless et al. 1988b;
Huettmann 2000). A few European studies of seabird associations with oceanographic
6
features have included the Razorbill (Alca torda islandica), suggesting that foraging
primarily occurs in shallow water (20-40m) and is associated with prey density
(Camphuysen and Webb 1999; Skov et al. 2000).
Little is known regarding how the distribution of Razorbills from North American
colonies is correlated with oceanographic features and aside from general geographic
descriptions, there is little information regarding the movements of Razorbills and the
factors that drive habitat selection, especially while away from the breeding colony
(Hipfner and Chapdelaine 2002). Despite the importance to conservation planning (Jones
2001) many oceanographic influences on seabirds are only recently becoming apparent as
our ability to track individuals develops (Rey et al. 2007). This study, following the
suggestion of Hipfner and Chapdelaine (2002), relies on satellite telemetry to characterize
the movements of Razorbills from a specific North American region. The movements of
Razorbills from the Machias Seal Island (MSI) breeding colony in the Bay of Fundy were
then used to analyze spatial distribution in relation to four oceanographic factors; 1) Sea
surface temperature, which has been linked to the distribution of many seabird species
(Huettmann 2000), 2) bathymetric depth; many Razorbills are noted to winter in shallow
coastal waters (Camphuysen and Webb 1999; Huettmann et al. 2005), 3) chlorophyll
concentration, an indicator of primary productivity (Feldman 2008) and 4) the slope of
the ocean floor, which highlights productive areas such as shelf-breaks and oceanic fronts
(Franks 1992; Yen et al. 2004).
Materials and Methods
Study area. Fieldwork was conducted over two days, 22-23 June, during the
2006 breeding season on MSI, NB (Fig 2.1). MSI (44 º60'N, 67º00'W) is a federal
7
Migratory Bird Sanctuary managed by the Canadian Wildlife Service located at the
junction between the cold waters of the Bay of Fundy and the warmer waters from the
Gulf of Maine (Diamond and Devlin 2003). MSI is approximately 9.5 ha, lies 18.8 km
southwest of Grand Manan and is composed primarily of bedrock with a large central
patch of vegetation (Bunin and Boates 1994). The interior of the island is composed of
various non-woody plants (there are no trees on the island) and the exterior portions of
the island are composed of boulders (where Razorbills nest) cobblestone beaches and
wide expanses of bare rock (Diamond and Devlin 2003).
Razorbills were snared with noose carpets placed in Razorbill loafing areas.
Traps were monitored by two researchers at all times so birds had little opportunity to
injure themselves while struggling to escape. After being removed from the trap,
Razorbills were placed in a cloth bag, weighed, measured and banded. Only birds
identified as breeders (by the presence of a brood patch) were used in this study.
Razorbills were sexed using a discriminant function (Grecian et al. 2003).
Satellite telemetry. Satellite transmitters (n = 8) were Platform Terminal
Transmitter 100 (PTT-100) abdominal implants (Fig 2.2) manufactured by Microwave
Telemetry, Inc. (Maryland, U.S.A.). The transmitters were programmed on a 6-hour
on/72-hour off duty cycle and were expected to last between 2 and 2.5 years. The
transmitters measured approximately 3.5 x 4.5 x 1 cm, had a 25 cm antenna and weighed
22 g (3 to 4% of body mass). Satellite data were collected by the ARGOS Data
Collection System. In addition to latitude and longitude PTTs also transmit the internal
temperature of the bird and the voltage of the transmitter. Variation in temperature or
8
battery voltage is indicative of bird mortality and transmitter failure respectively (Hatch
et al. 2000).
Surgery was performed by veterinarian Dr. Jim Berry and followed the procedure
outlined in Korschgen et al. (1996), using isoflorane as anesthetic. Transmitters were
sterilized at a local hospital and stored in sterile packaging until the time they were to be
placed in the bird. Surgery was conducted at the field site in as sterile a location as
conditions allowed.
Two incisions were made through the body wall. The primary incision was made
just below the sternum into the abdominal cavity and the second incision on the bird’s
dorsal side, just above the rump (Korschgen et al. 1996). The second incision was just
large enough for the diameter of the antenna to fit through. After implanting the
transmitter and closing the incisions, birds were held in a 60x47x36 cm box to recover
from the procedure. This generally took about 20 minutes until the Razorbill assumed a
standing posture and would move around the box. The box was then carried into the
colony and the lid was removed with the opening pointed towards the ocean. The
researcher remained hidden behind the box so as to hide their presence from the bird
when it left the box.
Oceanographic data. Bathymetric depth, surface chlorophyll concentration and
sea surface temperature were used to determine the environmental and oceanographic
characteristics of the areas where the Razorbills were monitored. Both sea surface
temperature and chlorophyll concentration data were obtained for each month from June
to November 2006.
9
Chlorophyll monthly composite maps (9 km resolution) were obtained from
SeaWiFS (Sea-viewing Wild Field-of-View Sensor). Monthly sea surface temperature
data (4 km resolution) were obtained from the Moderate Resolution Imaging
Spectroradiometer (Physical Oceanography DAAC 2008). Bathymetry data were
obtained from the Bedford Institute of Oceanography (D. Greenburg, personal
communication, 2007) and coastline coverage was obtained from the NOAA Coastline
Extractor (Signell 2008). Bathymetric data were used to infer the slope of the ocean floor
using the spatial analyst function in ArcMap 9.2 (Environmental Systems Research
Institute 2005).
Quality control of PTT locations. The locations of the PTTs are determined by a
Doppler shift as the satellites pass over the transmitters. The location of the PTT is
transmitted and, depending on a number of factors, an estimate of accuracy for each PTT
location is calculated. An estimate of accuracy requires that four messages be received
from the PTT but a number of other factors including pass duration, the number of
successful plausibility checks, topography of the land (if applicable), speed of the PTT
and the stability of the transmitter oscillator can influence whether an accuracy can be
calculated or not. Accordingly there are various location classes to be applied to each
location reported for the PTTs and a single PTT can have locations distributed over
several different classes over the course of its life. Transmitter locations that successfully
transmit four signals and meet other standards of transmission are assigned location
classes of 1, 2 or 3 (Table 2.1) and are considered the most reliable. Location classes of 0,
A, B or Z indicate that some criterion associated with accuracy estimates has not been
met (Table 2.1).
10
All classes of locations were used in the analysis; however certain specific
locations were rejected when they proved to be unreliable. Unreliable locations were
identified by suggesting an unrealistic speed between two points (Weimerskirch et al.
1993). Thus any point that required a PTT to move in excess of the supposed Razorbill
flight speed of 58 km/h (Benvenuti et al. 2001) was considered questionable and removed
from the dataset. All locations for each bird, regardless of class, were then plotted and
compared to the locations chronologically before and after to identify speeds that
exceeded 58 km/h (Table 2.1). All invalid locations (class Z) were excluded from the
analysis. The largest numbers of erroneous locations were from class 0 (23.1% rejected).
Only 4.9%, 1.0% and 1.1% of the locations were rejected from classes 1, 2 and 3
respectively and 9.7% of the class A locations and 7.0% of the class B locations were
rejected. Although Argos does not guarantee the accuracy of classes A and B, satellite
transmitters deployed on captive grey seals have shown class A locations to be nearly as
accurate as class 1 (Vincent et al. 2002). Class B locations, although less accurate than
class A, are accurate within 4,815 m at the 68% percentile (Vincent et al. 2002). This is
comparable to the oceanographic data used in the analysis which ranged in resolution
from 4 to 9 km2 and thus, class B locations were included in the analysis.
Data treatment and analysis. Estimates of Razorbill distribution were obtained
using the fixed kernel procedure (Worton 1989) and the bandwidth of each kernel was
determined using least squares cross validation (Silverman 1986). This technique first
involves the generation of a utilization distribution (a description of how much time an
animal spends in any place) for the sample. The utilization distribution places a
probability density (termed “kernel”) over each observation in the sample (Seaman and
11
Powell 1996). Each kernel has a value of 1 at its centre and decreases to 0 at the edge.
Therefore, observations that are close to the area being evaluated will contribute more to
the estimate than observations that are further away (Seaman and Powell 1996). This
process is repeated for each observation. Areas where kernels overlap are averaged to
give a single density for that location. Thus, areas within the distribution where
observations are more clustered will have a higher density estimate than areas where
observations are sparse.
Within each month locations for all Razorbills were pooled in order to maintain
50 locations per month, the minimum recommended sample size for the fixed-kernel
estimator using least squares cross (Millspaugh et al. 2006). A utilization distribution was
calculated for June to October. Distributions were not calculated for November because
there were not enough locations to meet minimum requirements (n=6). After the
utilization distributions were estimated contours representing 25%, 50%, 75% and 95%
of the utilization distribution were calculated for each month. The end result was a map
of Razorbill home-range divided into four zones reflective of decreasing utilization (Fig
2.3).
Zone 1 covers the smallest area representing 25% of the utilization, zone 2 the
second 25%, zone 3 the third 25% and zone 4 covers the next 20%. When combined, all
4 zones account for 95% of the utilization distribution. The innermost regions (zones 1
and 2) where utilization was highest were considered to represent key feeding areas while
the peripheral areas (zone 3 and 4) where utilization was lowest were considered areas
with little foraging activity (Rey et al. 2007).
12
PTT and oceanographic data were mapped with ArcGIS 3.2 (Environmental
Systems Research Institute 2005). To identify habitat selection the map of sub-divided
Razorbill home-range was overlaid with maps for monthly sea surface temperature,
monthly chlorophyll, bathymetric depth and slope of the ocean floor. Oceanographic
conditions found in each month and zone was then compared to look for indications that
Razorbill distribution is being influenced. Means are presented ± SD.
Results
Transmitter performance. All eight PTTs stopped transmitting prematurely
(Table 2.2). The transmitters were expected to last a minimum of 1 year (Microwave
Telemetry, Inc., PTT-100 Field Manual); however, the first PTT transmission was lost
after transmitting for only 4 days. Transmissions were lost from one or two transmitters
at a time until 14 December 2006 when the last transmission was received. Although
mortality was suspected initially, Razorbills with satellite implants were later seen
throughout the 2007 breeding season at MSI, and one was seen in 2008 (Appendix I).
Transmitter failure was distinguished from bird mortality using criteria from Meyers et
al. (1998) where transmitter failure is recognized by sudden loss of the signal without
concomitant loss in body temperature or battery voltage.
All transmitters reported a consistent temperature except for one (ID = 65613)
whose internal temperature began to drop when the PTT suddenly began transmitting on
land on 13 Aug after if had flown from the vicinity of MSI to the coast of Maine (Fig 2.4)
at a mean speed of 4.3 km/h. Similarly, signals from PTTs indicated a consistent voltage
for all transmitters with the exception of Razorbill 65613 whose PTT suffered a voltage
drop beginning on 20 Oct (Fig 2.4). Locations received for this PTT after 1 August were
13
excluded from subsequent analyses. Observations from 1 August to 13 August (the first
appearance on land) although in the ocean, were very close to the shore. These points
were excluded from the analysis because the accuracy of the location classes associated
with these points could not remove the possibility that they may have been transmitting
from land.
General movements. Razorbills remained in the Bay of Fundy along the New
Brunswick and Maine coast for the entire time that signals were received (Fig 2.1). On
average the birds kept within 5 km of land with the exception of two birds, which
occupied average distances of 6.9 ± 2.5 and 12.2 ± 9.8 km from shore (Table 2.2).
Eighty-eight percent of PTT locations within 5 km of Grand Manan were from Razorbill
65606 (Table 2.2) while locations from within 5 km of known Razorbill breeding
colonies ranged from 3.6% for Razorbill 65614 to 65.5% for Razorbill 65613. Of all 105
Razorbill locations within 5 km of a breeding colony, 74 were from the area around
Freeman Rock (44º27'N 67º32'W) and Pulpit Rock (44º33'N 67º28'W; Fig 2.1). All birds
had an average traveling speed less than 5 km per hour (Table 2.2). Razorbill 65606
made the fastest flight between two points while covering 1.16 km at 38.6 km per hour.
The fastest flight between days and thus the best estimate of maximum traveling speed
was 1.04 km per hour and occurred when Razorbill 65605 covered a distance of 79.9 km
over a span of 76.6 hours.
Spatial patterns in oceanographic features. The largest home-range distribution
occurred in July, covered an area of 3,553.5 km 2 (Fig 2.3, Table 2.3) and was based on 7
of the 8 Razorbills (Table 2.2). The distribution for June was almost as large, covering
an area of 3416.2 km2 (Fig 2.3, Table 2.3). The total area for the August, September and
14
October distributions covered smaller areas (1518.9, 1804.0 and 2653.6 km 2 respectively;
Table 2.3) and were also based on from fewer birds (Table 2.2). Regardless of month,
zone 1 (which accounts for the densest 25% of the utilization distribution) always
represented less than 25% of the area of the total distribution indicating that a large
amount of utilization (25%) occurred in a small area (3.7 to 7.1%; Table 2.3). The same
trend is occurs in zone 2, which accounts for the second densest 25% of the utilization
distribution (Table 2.3).
The Bay of Fundy reaches depths of up to 380 m (Fig 2.5), although Razorbills
remained in areas ranging in from 2 to 251 m. The monthly average depth used for each
zone was much more variable in the peripheral areas than in core areas (Table 2.4; Fig
2.6). Average depths for zone 4, the most peripheral area, ranged from 22.5 m to 72.6 m
while monthly averages for zone 1 occupied a range from 41.6 m to 54.2 m (Table 2.4).
Regardless of month, Zone 1, which represents areas of highest utilization, were
associated with a narrow range of 12.6 m of mean depths while in zone 4 mean depth was
spread over a range of 50.1 m.
Similar to depth there was a decrease in the range of monthly averages for slope
across zones of increasing utilization (Fig 2.7). The slope of the ocean floor in zone 4
ranged from an average of 20.0º in July to 39.0º in August, representing a spread of 19º
(Table 2.4). Zone 1 occupied a much narrower range from 27.5 º to 30.9º, a spread of only
3.4º (Table 2.4). Zones 2 and 3 occupied intermediate ranges of 8.7 º and 10º respectively
(Table 2.4; Fig 2.7). The core areas of Razorbill distribution (zones 1 and 2) always
overlap some landmass (Fig 2.7) where the ocean floor rises at a steeper angle to meet
shoreline (Fig 2.8).
15
Considering each month separately, there was little variation in sea surface
temperature across different zones and variation between months was likely a reflection
of changing seasonal temperature (Table 2.5; Fig 2.9). June experienced particularly cold
sea surface temperatures (9.3 to 9.6ºC) after which the temperature decreased from an
average of 14.5ºC in July to 12.0ºC in October (Table 2.5).
Excluding July chlorophyll concentration values for zone 1 were all near 1
mg/m3, however, three of the five months experienced very little (less than 2.0 mg/m 3)
variation in chlorophyll concentration across different zones (Table 2.5; Fig 2.10).
Standard deviations for chlorophyll concentration were high (Table 2.5) indicating a large
overlap in values across areas of differing Razorbill utilization.
Discussion
Transmitter performance. Failure of Argos tags is not a new phenomenon.
Hays et al. (2007) summarized Argos satellite tag failure from a number of studies, in
which internal malfunction such as battery failure, aerial breakage and animal mortality
were all identified as causes of signal loss. In addition many transmissions have been lost
from Argos satellite tags for unknown reasons (Meyers et al. 1998; Polovina et al. 2004).
Two Razorbills used in this study were identified in later years after contact with
the transmitters were lost (Table 2.2; Appendix I) and although Razorbill 65613
eventually died, mortality is not suspected to be the primary factor contributing to loss of
the other PTT signals. Similar seabird studies involving abdominal implants in Common
Murres, Thick-billed Murres and Tufted Puffins (Fratercula cirrhata) indicate that
mortality as a result of the procedure is most likely to occur between 9 days (Meyers et
al. 1998) and one month (Hatch et al. 2000) after implantation. Only 1 signal (ID =
16
65597) from this study was lost in less than a month (Table 2.2) although no signs of
voltage or temperature loss were observed to confirm mortality. This Razorbill also
remained mobile and traveled an average minimum of 10 km per day during the 4 days
that transmissions were received.
General movements. No transmissions were received from outside the Bay of
Fundy (Fig 2.1) and although birds in this study were not tracked for the entire year, the
pattern is consistent with Huettmann et al.’s (2005) suggestion, as well as results from
Chapter 4 of this thesis, that the population of Razorbills breeding on MSI is resident
year-round. If so, it is the only known resident population of this species.
No long-distance migration was observed in this sample such as occurs in
Razorbills migrating south from the Gannet Islands, Labrador into Newfoundland waters
(Chapdelaine 1997). Thus, the speeds associated with this sample are likely more
representative of wintering birds than of migrating Razorbills. The highest traveling
speed observed in this study was 1.04 km/h. Razorbills leaving the Gannet Islands in
August have been recovered off eastern Newfoundland (~260 to 660 km away) in
October and November (Chapdelaine 1997) but lack of daily activity budgets preclude
direct comparisons relating to traveling speeds.
The maximum flight speed observed was 38.6 km/h, which is less than the
Razorbill flight speed (58 km/h) calculated by Benvenuti et al. (2001) to reconstruct
foraging patterns. In this study maximum flight speed was calculated by assuming the
Razorbill made a direct flight of 1.16 km from one point to another without stopping.
Had this Razorbill rested at any point or flown in anything except a straight line, 38.6
km/h would be an under-representation of the top speed.
17
Five of the 8 Razorbills were tracked within 5 km of Freeman Rock (44 º27´N
67º32´W) and Pulpit Rock (44º33´N 67º28´W), two small, Razorbill colonies on rocky
islands along the coast of Maine, United States (Fig 2.1). Freeman Rock supports 50 pairs
of Razorbills and Pulpit Rock a further 10 pairs (Chapdelaine et al. 2001). Neither
banding nor band reading takes place in either colony for fear of disturbance (Linda
Welch, personal communication); however the areas of highest Razorbill utilization
calculated for June, July and September suggest a possible relationship between
Razorbills from MSI and the area around Freeman Rock and Pulpit Rock (Fig 2.6).
Spatial patterns in oceanographic features. Generation of the monthly
utilization distributions was influenced more by some birds than others. Four Razorbills
contributed less than 30 locations each to the analysis while the remaining four
contributed over 50 location each (Table 2.2). Razorbill 65606 contributed 216 locations,
which was more than twice the number of locations available from any other bird in the
sample. Also, as mentioned earlier, monthly utilization distributions receive
representation from fewer birds as time progresses (Fig 2.3). The October distribution for
example only included locations from 4 of the 8 PTTs. Thus, discussion of these results
comes with the caveat that there is biased representation of individual birds in the sample.
The Bay of Fundy reaches depths of 380 m, however this study indicates that this
sample of Razorbills are associated with waters that are 40 to 55 m deep (Table 2.5) and
all but 1 Razorbill stayed, on average, within 7 km from shore (Table 2.2). In the past
Razorbills have been described as an offshore species (Tuck 1961; Brown 1986) but data
from this study more closely resemble descriptions of Razorbills as coastal (Camphuysen
and Webb 1999; Hipfner and Chapdelaine 2002), and is consistent with Huettmann et
18
al.’s (2005) suggestion that the Razorbill be viewed more as a component of the coastal
ecosystem than as a pelagic species. Water depth is an important habitat variable for
marine birds and inner shelf areas are important to subsurface foragers, which account for
higher bio-mass and energy uptake in these areas than surface foragers (Schneider 1997).
Continental shelves are characterized by complex water flows, which generate fronts and
eddies which aggregate prey close to the surface for diving predators (Yen et al. 2004).
Dall’Antonia et al. (2001) reported that Razorbills in Iceland never dived below 32 m
even though sea depth in their study area reached 60 m and Razorbills are capable of
diving up to 120 m (Piatt and Nettleship 1985). This suggests that Razorbills are capable
of provisioning themselves with prey from shallow depths, perhaps by targeting fronts
and eddies. Furthermore, in 2007, a single Razorbill (see Chapter 3) spent a year in the
Gulf of Maine without diving below 40 m. Kessel (1979) proposed that mixing regimes
associated with characteristic depths influenced seabird abundance over continental
shelves. This study suggests that Razorbills in the Bay of Fundy are most closely
associated with waters around 40 to 55 m deep.
Continental shelf slopes and shelf-breaks are productive habitats (Yen et al.
2004), which often support high numbers of marine predators (Croll et al. 1998). The
ocean floor in the Bay of Fundy ranges from areas that are relatively flat (0 to 10 º) to
areas such as the the Grand Manan channel and Grand Manan Basin where drop offs
reach slopes as steep as 60º. The densest areas of the monthly home-range estimates were
centered over areas near the coast such as the southwest corner of Grand Manan (Fig 2.6)
where the ocean floor rises at steep angles ranging from 20 to 50 º to form the island (Fig
2.8). Furthermore, the highest densities of Razorbill were found in areas where the
19
average angle of the ocean floor ranged from 27.5 in September to 30.9 in July; a spread
of only 3.4º, whereas in the most peripheral areas, the average slope ranged from 20.0 in
July up to 39.0 in August, a difference of 10º.
Tidally-driven upwellings around islands are often sites of enhanced primary and
secondary production (Franks 1992). Common Murres and Cassin’s Auklets
(Ptychoramphus aleuticus) have been associated with shelf-breaks in response to the
large numbers of euphausiids, which are also a noted food source for Razorbills
(Huettmann et al. 2005; Clarke et al. 2008). There is also an historical herring (Clupea
sp.) spawning ground near the southwest corner of Grand Manan and along the Maine
coast of the Grand Manan Channel (Chenoweth et al. 1989), which may influence
Razorbill density. Huettmann et al. (2005) found no relationship between the occurrences
of schooling fish and Razorbills, however their data are from December-February.
Chlorophyll concentration from phytoplankton is used as an indicator of primary
production in marine systems (Feldman 2008), has been linked to seabird distributions
(Ainley et al. 1998) and is associated with the occurrence of euphausiids (Ressler et al.
2005), which form a portion of Razorbill diet in the Bay of Fundy (Huettmann et al.
2005; Clarke et al. 2008). With the exception of July, our data suggests that areas of
highest Razorbill utilization are most closely associated with low chlorophyll
concentrations (Figure 2.7). However, July, September and October experienced very
little variation across areas of differing Razorbill utilization, which suggests that
Razorbills may not have been reacting to chlorophyll concentration or that the spatial
scale used in this study was not large enough to detect any influence that chlorophyll
concentration may have on Razorbill distribution. Furthermore, large standard deviations
20
associated with chlorophyll concentration indicate overlaps in monthly values across
different zones making evidence of selection inconclusive. Nonetheless, it is worth noting
that a possible association between increasing Razorbill utilization and decreasing
chlorophyll concentrations may be a response to high turbidity, which may make feeding
difficult for visual feeders such as Razorbills. Henkel (2006) reported that Brandt’s
Cormorants (Phalacrocorax penicillatus), which employ the same pursuit-diving strategy
as Razorbills, were less likely to appear in turbid waters, often caused by suspended
organic matter such as phytoplankton. Razorbills in the Bay of Fundy feed their chicks,
and presumably themselves, predominantly on herring (Bond 2007; Clarke et al. 2008).
As such they probably rely on acute vision to locate and pursue fish in the water column,
which would likely be diminished by increased turbidity.
It is possible that Razorbills are responding to prey species that are in turn
responding to some other oceanographic feature, such as tidally-driven factors
(Jovellanos and Gaskin 1983), which may not be correlated with chlorophyll density.
Chlorophyll concentration is also not correlated to the intensity of the purse seine Herring
fishery in the Gulf of Maine, likely because herring do not feed directly on phytoplankton
(Friedland et al. 2006). Thus, Razorbills feeding on herring would not necessarily make
use of areas with high chlorophyll concentrations.
As expected, sea surface temperature varied greatly among months (Fig 2.2),
however the variation across kernels is less than 1°C for all months (Table 2.5) and it is
unlikely that the birds can respond to such a small change in temperature. Sea surface
temperature has been associated with fish recruitment (North 2005), negatively correlated
with adult Razorbill survival (Sandvik et al. 2005; Lavers 2008), and is likely a major
21
factor in defining habitat for Atlantic Puffin (Fratercula arctica), Common Murre and
Thick-billed Murre (Huettmann and Diamond 2006). These studies however were
conducted at a larger spatial scale, and relationships between Razorbill density and sea
surface temperature may become evident only across larger spatial scales. Juvenile
herring, the main component of Razorbill diet in the Bay of Fundy, are most abundant at
cooler temperatures (8 – 12°C) found deeper in the water column (Reid et al. 1999) and
thus any associations between Razorbill distribution and temperature may be evident at
deeper depths.
Habitat selection was evident for this sample of Razorbills in relation to both
bathymetric depth and slope. Both are physical, relatively unchanging properties of the
ocean floor and are determinants of tidally-driven upwelling and depth-related mixing.
There was no association between Razorbill utilization and sea surface temperature and
associations between bird utilization and chlorophyll concentration are tentative. Both of
these factors change throughout the year and are influenced by upwelling and mixing.
The overall implication is that this sample of Razorbills was more closely associated with
factors that influence upwelling and mixing than they are with the effects themselves.
22
Tables
Table 2.1: Details of Argos transmitter classes and summary of locations used from each class.
Class
Accuracy
Meaning
Number used
Number
rejected
3
Less than 150m
These classes were obtained
86
1
with
at
least
4
messages
2
Between 150 and 350m
106
1
1
Between 350 and 1000m
59
3
0
The upper limit is greater
30
9
than 1000m
A
No Estimate
The location was obtained
68
7
with 3 messages
B
No Estimate
The location was obtained
121
9
with 2 messages
Z
Invalid Locations
The location process failed
0
15
23
24
F
F
F
F
M
M
F
M
65590
65605†
65606
65613
65614
65589
65597
65598†
22/06/06
22/06/06
22/06/06
22/06/06
22/06/06
23/06/06
23/06/06
23/06/06
Start of
deployment
(dd/mm/yy)
14/10/06
30/10/06
2/11/06
14/12/06
19/08/06
15/09/06
26/06/06
16/10/06
115
131
134
40*
59
85
4
116
2.0 ± 1.9
6.9 ± 2.5
4.1 ± 2.5
2.2 ± 1.9
1.4 ± 1.5
3.0 ± 2.7
1.4 ± 1.4
12.2 ± 9.8
Last message Transmission
Mean
received
period
distance
(dd/mm/yy)
(d)
from land
(km)
155.4
927.6
932.2
234.6
291.2
378.7
40.1
402.0
3.3 ± 7.6
2.8 ± 4.4
3.9 ± 5.6
3.8 ± 5.8
3.0 ± 4.5
1.5 ± 2.6
1.8 ± 1.7
4.0 ± 7.0
Min.
Mean travel
distance speed ± SD
covered
(km h-1)
(km)
8
61
216
87
56
28
4
10
Number of
locations
used in
analysis
37.5
8.2
38.9
0.0
1.8
10.7
0.0
0.0
% of locations
within 5 km of
Grand Manan
25.0
36.1
4.6
65.5
3.6
32.1
25.0
20.0
% of locations
within 5 km
of a breeding
colony
* Transmission continued from 134 days after mortality was assumed at day 40. All calculations reflect only the first 40 days
† Both Razorbills were seen in subsequent breeding seasons after transmitter failure occurred. Razorbill 65598 was observed throughout the 2007 breeding
season and Razorbill 65605 was observed in July 2008 (see Appendix I).
Sex
Razorbill
ID
Table 2.2: Summary of data from satellite transmitters attached to 8 adult Razorbills from Machias Seal Island, NB.
Table 2.3: Monthly breakdown of the area covered by each
zone pictured in Figure 2.3. Total numbers of PTT
locations used in kernel estimates appear underneath the
associated month.
Month
Zone
Area*
% of total area*
2
(km )
June
1
241.5
7.1
(n = 74)
2
571.7
16.7
3
1168.7
34.2
4
1434.3
42.0
Total
3,416.2
July
(n =170)
1
2
3
4
Total
August
(n = 82)
129.9
191.9
1034.8
2196.9
3,553.5
1
2
3
4
3.7
5.4
29.1
61.8
67.2
4.4
103.1
6.8
213.2
14.0
1135.4
74.8
Total
1,518.9
September
1
105.2
5.8
(n = 70)
2
276.9
15.3
3
456.0
25.3
4
965.9
53.5
Total
1,804.0
October
1
186.9
7.0
(n = 60)
2
242.8
9.1
3
400.9
15.2
4
1823.0
68.7
Total
2,653.6
* Portions of the Razorbill distributions appearing over
landmasses in Figure 2.3 have been excluded
25
26
Month
June
July
August
September
October
Zone
52.1 ± 35.9
22.5 ± 34.5
35.3 ± 36.9
70.5 ± 32.9
72.6 ± 40.3
4
54.6 ± 22.3
39.3 ± 31.2
43.5 ± 36.5
51.0 ± 29.9
55.9 ± 37.2
3
2
52.7 ± 34.6
50.4 ± 38.0
60.0 ± 36.0
50.6 ± 32.7
47.8 ± 39.4
Depth
(m)
44.2 ± 19.3
41.6 ± 18.6
46.4 ± 34.4
53.3 ± 15.5
54.2 ± 31.4
1
21.0 ± 8.0
20.0 ± 6.3
39.0 ± 9.9
26.7 ± 12.2
29.7 ± 14.5
4
23.5 ± 10.5
31.4 ± 17.7
29.2 ± 13.8
24.8 ± 12.5
33.5 ± 16.0
3
28.0 ± 13.9
31.7 ± 14.3
30.2 ± 16.0
23.0 ± 12.6
31.0 ± 16.1
2
Angle of the ocean floor
(θ)
30.6 ± 14.9
30.9 ± 14.6
30.3 ± 14.0
27.5 ± 15.1
29.3 ± 15.3
1
Table 2.4: Temporal (monthly) and spatial (zone) parameters (Mean ± SD) for Razorbill locations associated with depth and angle of the ocean
floor.
27
Month
June
July
August
September
October
Zone
9.6 ± 1.0
14.6 ± 0.8
11.7 ± 0.0
12.8 ± 0.3
12.2 ± 0.1
4
9.3 ± 1.0
14.3 ± 0.8
12.0 ± 0.2
12.8 ± 0.3
12.1 ± 0.2
3
9.4 ± 0.7
14.4 ± 0.5
12.1 ± 0.3
12.7 ± 0.3
12.0 ± 0.2
2
Sea surface temperature
(0C)
9.6 ± 0.8
14.5 ± 0.6
13.1 ± 1.1
12.7 ± 0.3
12.1 ± 0.4
1
6.5 ± 10.3
6.4 ± 8.6
10.7 ± 12.5
2.2 ± 2.6
2.0 ± 3.1
4
6.5 ± 10.2
5.4 ± 2.9
3.2 ± 3.2
1.3 ± 1.1
1.9 ± 1.2
3
4.1 ± 5.1
6.9 ± 4.3
3.9 ± 3.4
1.2 ± 1.0
1.9 ± 0.5
2
Chlorophyll Concentration
(mg/m3)
1.3 ± 2.3
6.6 ± 4.3
2.7 ± 3.2
1.0 ± 1.0
1.9 ± 0.6
1
Table 2.5: Temporal (monthly) and spatial (zone) parameters (Mean ± SD) for Razorbill locations associated with sea surface
temperature and chlorophyll concentration.
Figures
Figure 2.1: Razorbill breeding colonies in the Bay of Fundy and all Razorbill locations (n=470) from the
entire duration of the PTT deployment.
28
Figure 2.2: Platform Terminal Transmitter 100 (PTT-100) abdominal implants manufactured by Microwave
Telemetry, Inc. Satellite Transmitter.
29
Figure 2.3: Sub-divided monthly Razorbill distribution in the Bay of Fundy region from June to October
2007. Note: n = # of Razorbills used in analysis. Details of area breakdown can be seen in Table 2.3.
30
Figure 2.4: Detailed of voltage and body temperature for Razorbill 65613, which was presumed dead in early
August when it appeared on Head Harbour Island, Maine. Transmissions were received until early December.
31
Figure 2.5: Map of bathymetric depth in the Bay of Fundy region and main Razorbill
colonies. Bathymetry data were obtained from the Bedford Institute of
Oceanography (D. Greenburg, personal communication, 2007).
32
80.0
70.0
Depth (m)
60.0
June
July
Aug
Sept
Oct
50.0
40.0
30.0
20.0
10.0
0.0
Zone 4
Zone 3
Zone 2
Zone 1
Figure 2.6: Depth associated with changing areas of Razorbill monthly distribution. Standard deviations have
been summarized in Table 2.5 for clarity.
45.0
Slope of the ocean floor
40.0
35.0
June
July
Aug
Sept
Oct
30.0
25.0
20.0
15.0
10.0
5.0
0.0
Zone 4
Zone 3
Zone 2
Zone 1
Figure 2.7: Mean angle of the slope of the continental shelf associated with changing areas of monthly
Razorbill distribution. Standard deviations have been summarized in Table 2.5 for clarity.
33
Figure 2.8: Map of the slope of the ocean floor in the Bay of Fundy region and main Razorbill colonies.
Slope was interpreted from bathymetry data were obtained from the Bedford Institute of
Oceanography (D. Greenburg, personal communication, 2007).
34
Sea Surface Temperature (oC)
16.0
14.0
12.0
June
10.0
July
8.0
Aug
Se pt
6.0
O ct
4.0
2.0
0.0
Zo ne 4
Zo ne 3
Zo ne 2
Zo ne 1
Figure 2.9: Mean monthly sea surface temperatures associated with changing areas of monthly Razorbill
distribution. Standard deviations have been summarized in Table 2.5 for clarity.
Chlorophyll Concentration (mg/m 3)
12.0
10.0
8.0
June
July
6.0
Aug
Se pt
O ct
4.0
2.0
0.0
Zo ne 4
Zo ne 3
Zo ne 2
Zo ne 1
Figure 2.10: Chlorophyll concentration associated with changing areas of monthly Razorbill distribution.
Standard deviations have been summarized in Table 2.5 for clarity.
35
References
Ainley, D.G., S.S. Jacobs, C.A. Ribic and I. Gaffney. 1998. Seabird distribution and
oceanic features of the Amundsen and southern Bellingshausen Seas. Antarctic
Science 10: 111-123.
Benvenuti, S., L. Dall’Antonia and P. Lyngs. 2001. Foraging behaviour and time
allocation of chick-rearing Razorbills Alca torda at Græsholmen, central Baltic
Sea. Ibis 143: 402-412.
Black, A.L. 2002. Foraging area characteristics of Arctic Terns (Sterna paradisaea) and
Common Terns (Sterna hirundo) breeding on Machias Seal Island. M.Sc. thesis.
Department of Biology, University of New Brunswick, New Brunswick.
Black, A.L., L.I. Minich and A.W. Diamond. 2004. Machias Seal Island 1995-2004
progress report. Atlantic Co-operative Wildlife Ecology Research Network,
Fredericton, N.B.
Bond, A.L. 2007. Patterns of mercury burden in the seabird community of Machias Seal
Island, New Brunswick. M.Sc. thesis. Department of Biology, University of New
Brunswick, New Brunswick.
Brown, R.G.B. 1986. Revised atlas of eastern Canadian seabirds. Halifax: Canadian
Wildlife Service, Bedford Institute of Oceanography. 110 pp.
Bunin, J.S. and J.S. Boates. 1994. Effects of nesting location on breeding success of
Arctic Terns on Machias Seal Island. Canadian Journal of Zoology 72(10): 18411847.
Camphuysen, C.J. and A. Webb. 1999. Multi-species feeding associations in North Sea
seabirds: jointly exploiting a patchy environment. Ardea 87: 177-198.
Chapdelaine, G. 1997. Pattern of recoveries of banded Razorbills (Alca torda) in the
Western Atlantic and survival rates of adults and immatures. Colonial Waterbirds
20(1): 47-54.
Chapdelaine, G., A.W. Diamond, R.D. Elliot and G.J. Robertson. 2001. Status and
population trends of the Razorbill in eastern North America. Canadian Wildlife
Service Occasional Paper 105.
Chenoweth, S.B., D.A. Libby, R.L. Stephenson and M.J. Power. 1989. Origin and
dispersion of larval herring (Clupea harengus) in coastal waters of Eastern Maine
and Southwestern New Brunswick. Canadian Journal of Fisheries and Aquatic
Science 46: 624-632.
36
Clarke T.C., M.F. Godin and A.W. Diamond. 2008. Machias Seal Island 1995 – 2007
progress report. Atlantic Co-operative Wildlife Ecology Research Network,
University of New Brunswick, Fredericton, New Brunswick.
Croll, D.A., B.R. Tershy, R.P. Hewitt, D. Demer, P.C. Fiedler, S.E. Smith, W.
Armstrong, J.M. Popp, T. Kiekhefer, V.R. Lopez and J. Urban. 1998. An
integrated approach to the foraging ecology of marine birds and mammals. DeepSea Research II 45: 1353-1371.
Dall’Antonia, L., G.A. Gudmundsson and S. Benvenuti. 2001. Time allocation and
foraging pattern of chick-rearing Razorbills in northwest Iceland. The Condor
103: 469-480.
Diamond, A.W. and C.M. Devlin. 2003. Seabirds as indicators of changes in marine
ecosystems: Ecological monitoring on Machias Seal Island. Environmental
Monitoring and Assessment 88: 153-175.
Environmental Systems Research Institute. 2005. ArcView GIS 9.2. Redlands,
California.
Feldman, G.C. 2008. SeaWiFS Project. NASA Goddard Space Flight Center. Available
at: http://oceancolor.gsfc.nasa.gov/cgi/level3.pl. Last accessed 7 May, 2008.
Franks, P.J.S. 1992. Sink or swim: accumulation of biomass at fronts. Marine Ecology
Progress Series 82: 1-12.
Friedland, K.D., J.E. O’Reilly, J.A. Hare, G.B. Wood, W.J. Overholtz and M.D. Cieri.
2006. Environmental preferences of Atlantic herring under changing harvest
regimes. Northeast Fisheries Science Center Reference Document 06-18. U.S.
Department of Commerce, NOAA. National Marine Fisheries Service, Northeast
Region, Northeast Fisheries Science Center, Woods Hole, Massachusetts.
Grecian, V.D., A.W. Diamond and J.W. Chardine. 2003. Sexing Razorbills (Alca torda)
breeding at Machias Seal Island using discriminant function analysis. Atlantic
Seabirds 5: 73-81.
Hatch, S.A., P.M. Meyers, D.M. Mulcahy and D.C. Douglas. 2000. Performance of
implantable satellite transmitters in diving seabirds. Waterbirds 23: 84-94.
Hays G.C., C.J.A. Bradshaw, M.C. James, P. Lovell and D.W. Sims. 2007. Why do
Argos satellite tags deployed on marine animals stop transmitting? Journal of
Experimental Marine Biology and Ecology 349: 52-60.
Henkel, L.A. 2006. Effect of clarity on the distribution of marine birds in nearshore
waters of Monterey Bay, California. Journal of Field Ornithology 77: 151-156.
37
Hipfner, J.M. and G. Chapdelaine. 2002. Razorbill (Alca torda). In The Birds of North
America, No. 635 (A. Poole and F. Gill, eds.). The Birds of North America, Inc.,
Philadelphia, PA.
Hooge, P.N., W. Eichenlaub and E. Solomon. 1999. The animal movement program.
USGS, Alaska Biological Science Center.
Huettmann, F. 2000. A descriptive model of environmental determination of winter
seabird distribution in the Canadian North Atlantic. In Environmental
determination of seabird distribution in the northwest Atlantic. Pd.D thesis.
University of New Brunswick, New Brunswick. Pp 213-292.
Huettmann, F. and A.W. Diamond. 2000. Seabird migration in the Canadain northwest
Atlantic Ocean: moulting locations and movement patterns of immature birds.
Canadian Journal of Zoology 78: 624-647.
Huettmann, F and A.W. Diamond. 2006. Large-scale effects on the spatial distribution of
seabirds in the Northwest Atlantic. Landscape Ecology 21: 1089-1108.
Huettmann, F., A.W. Diamond, B. Dalzell and K. Macintosh. 2005. Winter distribution,
ecology, and movements of Razorbills Alca torda and other auks in the outer Bay
of Fundy, Atlantic Canada. Marine Ornithology 33: 161-171.
Jones, J. 2001. Habitat selection studies in avian ecology: a critical review. The Auk 118:
557-562.
Jovellanos, C.L. and D.E. Gaskin. 1983. Predicting the movements of juvenile atlantic
herring (Clupea harengus harengus) in the SW Bay of Fundy using computer
simulation techniques. Canadian Journal of Fisheries and Aquatic Science 40:
139-146.
Kessel, B. 1979. Avian habitat classification for Alaska. Murrelet 60: 86-94.
Korschgen, C.E., K.P. Kenow, A. Gendron-Fitzpatrick, W.L. Green and F.J. Dein. 1996.
Implanting intra-abdominal radiotransmitters with external whip antennas in
ducks. Journal of Wildlife Management 60: 132-137.
Lavers, J.L, I.L. Jones, A.W. Diamond and G.J. Robertson. 2008. Annual survival of
North American Razorbills (Alca torda) varies with ocean climate indices.
Canadian Journal of Zoology 86: 51-61.
Meyers, P.M., S.A. Hatch and D.M. Mulcahy. 1998. Effect of implanted satellite
transmitters on the nesting behavior of murres. The Condor 100: 172-174.
38
Millspaugh, J.J., R.M. Nielson, L. McDonald, J.M. Marzluff, R.A. Gitzen, C.D.
Rittenhouse, M.W. Hubbard and S.L. Sheriff. 2006. Analysis of resource
selection using utilization distributions. Journal of Wildlife Management 70: 384395.
North, A.W. 2005. Mackerel icefish size and age differences and long-term change at
South Georgia and Shag Rocks. Journal of Fish Biology 67: 1666-1685.
Physical Oceanography DAAC. 2008. AVHRR OceansPathfinder Global Equal-angle
Best SST (NOAA/NASA). NASA JPL Physical Oceanography DAAC, Pasadena,
CA. Jet Propulsion Laboratory. Available at http://poet.jpl.nasa.gov/. Last
accessed on 7 May 2008.
Piatt, J.F. and D.N. Nettleship. 1985. Diving depths of four alcids. The Auk 102: 293297.
Polovina, J.J., G.H. Balazs, E.A. Howell, D.M. Parker, M.P. Seki and P.H. Dutton. 2004.
Forage and migration habitat of loggerhead (Caretta caretta) and olive ridley
(Lepidochelys olivacea) sea turtles in the central north pacific ocean. Fisheries
Oceanography 13: 36-51.
Reid, R.N., L.M. Cargnelli, S.J. Griesbach, D.B. Packer, D.L. Johnson, C.A. Zetlin,
W.W. Morse and P.L. Berrien. 1999. Essential fish habitat source document:
atlantic herring, Clupea harengus, life history and habitat characteristics. U.S.
Department of Commerce, NOAA. National Marine Fisheries Service, Northeast
Region, Northeast Fisheries Science Center, Woods Hole, Massachusetts.
Ressler, P.H., R.D. Brodeur, W.T. Peterson, S.D. Pierce, P.M. Vance, A. Røstad and J.A.
Barth. 2005. The spatial distribution of euphausiid aggregations in the Northern
California Current during August 2000. Deep-Sea Research II 52: 89-108.
Rey, A.R., P. Trathan, K. Pütz and A. Schiavini. 2007. Effect of oceanographic
conditions on the winter movements of Rockhopper Penguins Eudyptes
chrysocome chrysocome from Staten Island, Argentina. Marine Ecology Progress
Series 330: 285-295.
Sandvik, H., K.E. Erikstad, R.T. Barrett and N.G. Yoccoz. 2005. The effect of climate on
adult survival in five species of North Atlantic seabirds. Journal of Animal
Ecology 74: 817-831.
Schneider, D.C. 1997. Habitat selection by marine birds in relation to water depth. Ibis
139: 175-178.
Signell, R. 2008. Coastline Extractor. NOAA/National Geophysical Data Center, Marine
Geology and Geophysics Division.
Available at http://rimmer.ngdc.noaa.gov/coast/. Last accessed on 7 May, 2008.
39
Silverman, B.W. 1986. Density estimation for statistics and data analysis. Chapman and
Hall, London, UK.
Skov, H., J. Durinck and P. Andell. 2000. Associations between wintering avian
predators and schooling fish in the Skagerrak-Kattegat suggest reliance on
predictable aggregations of herring Clupea harengus. Journal of Avian Biology
31: 135-143.
Tuck, L.M. 1961. The murres: their distribution, populations and biology – a study of the
genus Uria. Canadian Wildlife Service Monograph Series No.1.
Vincent, C., B.J. McConnell, V. Ridoux and M.A. Fedak. 2002. Assessment of Argos
location accuracy from satellite tags deployed on captive grey seals. Marine
Mammal Science 18: 156-166.
Wanless, S., M.P. Harris and J.A. Morris. 1988a. The effect of radio transmitters on the
behaviour of Common Murres and Razorbills during chick rearing. The Condor
90: 816-823.
Wanless, S., J.A. Morris and M.P. Harris. 1988b. Diving behaviour of guillemot Uria
aalge, puffin Fratercula arctica and razorbill Alca torda as shown by radio
telemetry. Journal of Zoology London 216: 73-81.
Weimerskirch, H., M. Salmolard, F. Sarrazin and P. Jouventin. 1993. Foraging strategy
of Wandering Albatross through the breeding season: A study using satellite
telemetry. The Auk 110: 325-342.
Worton, B.J. 1989. Kernel methods for estimating the utilization distribution in homerange studies. Ecology 70: 164-168.
Yen, P.P.W, W.J. Sydeman and K.D. Hyrenbach. 2004. Marine bird and cetacean
associations with bathymetric habitats and shallow-water topographies:
implications for trophic transfer and conservation. Journal of Marine Systems 50:
79-99.
40
Chapter 3 – Case Study of Razorbill (Alca torda) Diving Behaviour
Abstract
Little is known about Razorbill diving behaviour, especially outside of the
breeding season. A geo-locator was deployed on an adult female Razorbill breeding at
Machias Seal Island, New Brunswick from 4 July 2006 to 7 July 2007. Depth in the water
column was sampled every 2 minutes for 82 days spaced throughout the year-long
deployment. Analysis indicated that diving patterns differed between the breeding season
and non-breeding season and diving is much more frequent in the morning and afternoon
than at periods of lower light intensity. The depths achieved by this Razorbill (17.9 ± 6.8
meters) are in agreement with other studies of diving behaviour although peak diving
activity occurred primarily in the morning and afternoon, whereas Razorbills from more
northerly colonies made only moderate use of these times of the day.
Introduction
The Razorbill is a strong swimmer and wing-propelled diver. While on the water,
it swims by paddling its feet, but while diving short, quick wing flaps propel the bird
through the water. Razorbill dives are typically V-shaped and the birds may deviate from
the straight descent and ascent in order to pursue prey (Benvenuti et al. 2001; Paredes et
al. 2008).
Diving behaviour is one of the least studied aspects of Razorbill ecology and it is
primarily through the attachment of devices such as capillary depth gauges or electronic
data loggers to the birds that their underwater abilities and behaviour can be measured
and recorded. However, most studies of Razorbill diving behaviour that are using these
methods are concerned only with foraging trips during the breeding season (Wanless et
41
al. 1988b; Harris et al. 1990; Dall’Antonia et al. 2001; Benvenuti et al. 2001; Watanuki
2006).
Little work has actually been done concerning the diving behavior of the
Razorbill, especially at the southern extent of its range. Piatt and Nettleship (1985) used
alcid by-catch from gill nets deployed around Newfoundland to gain insight into alcid
diving abilities. They identified what is generally regarded as the maximum Razorbill
diving depth of 120 to 130 m. The emerging picture, however, is that Razorbills achieve
such depths only on extremely rare occasions and the majority of Razorbill dives are
rarely beyond 40 m (Barrett and Furness 1990; Benvenuti et al. 2001; Dall’Antonia et al.
2001; Paredes et al. 2008).
At the Gannet Islands, Labrador, Paredes et al. (2008) found that in the breeding
season there were sex differences in Razorbill diving behaviour; male birds had a
tendency to dive deeper than females and dive patterns of both sexes were influenced by
prey behaviour.
Bird foraging strategies can differ significantly from colony to colony (Benvenuti
et al. 2001) and it is likely that Razorbills at the southern extent of their North American
range will employ different foraging strategies than those from northern and European
colonies due to different ecological factors. Furthermore, almost no detailed information
is available on Razorbill diving behaviour outside the breeding season. This study
describes the first account of a year-long deployment of a geo-locator on a Razorbill from
Machias Seal Island, New Brunswick.
42
Methods
Geo-locators. In July, 2006, four adult Razorbills at the Machias Seal Island
breeding colony in New Brunswick (44°30´N, 67°06´W) were equipped with
LTD2400_50 series data loggers (Fig 3.1), although only one tag was recovered (see
Results section). The tags were manufactured by Lotek Wireless, Inc. and mounted on
coiled plastic bands with glue and plastic zip-ties. The plastic band was attached to the
leg of the Razorbills. The geo-locators had a diameter of 11mm, a length of 36mm and a
mass of 8g (1.14% of the total body mass). Unlike similar devices which record dive
profile, these tags were programmed to activate every 4 days for a 24-hour period. During
this time the tags systematically sampled the temperature, light and depth in the water
column in two-minute intervals for a total of 60,481 samples taken per tag over the
course of the deployment. In addition, the tags also recorded the maximum depth
achieved each day. Data were archived in the tags until recovered.
Diving behaviour. To assess diving behaviour, each sample collected by the tag
was categorized according to season, time of day and depth in the water column.
Time of day: In order to distinguish day from night, the tag uses a particular light
threshold that, when crossed, indicates sunrise and sunset. Mid-day was then calculated
as the point halfway between sunrise and sunset and each sample was categorized as
either dawn, morning, afternoon or night.
Dawn and dusk were then defined as including the 30 minutes before and after
sunrise and sunset. Morning and afternoon were defined as the time between dawn and
mid-day, and between mid-day and dusk respectively. Night was defined as the entire
time following dusk and preceding dawn.
43
Season: Each sample was considered a part of either the breeding season or nonbreeding season. Using observations of breeding behaviour from Machias Seal Island
from the time the bird was tagged in July 2006, all observations gathered from 10 July to
20 August 2006 and from 10 May to 6 July 2007 were considered part of the breeding
season. All observations from 27 August 2006 to 7 May 2007 were considered a part of
the non-breeding season.
Depth: Any sample that was taken at a depth greater than 20cm was considered a
dive. Depths shallower than 20 cm were considered to be at the surface. Depths were
categorized as 0.2 to 1.9 m, 2 to 4.9 m, 5 to 14.9 m, 15 to 24.9 m and 25 to 35 m.
Days: Twenty days were sampled in the breeding season and 62 days in the nonbreeding season.
Tests for significant differences between different diving variables were
conducted using analysis of variance (ANOVA). For differences found significant at P =
0.05, Tukey’s post-hoc test was used to determine which levels within the factors were
significantly different. Means are presented as ± SD.
Light-based locations. The geo-locator produces estimates of latitude and
longitude based on light levels recorded daily by the tag. The raw light data are used to
estimate sunrise and sunset in Greenwich Mean Time which is then used to calculate
latitude and longitude. Data were downloaded from the geo-locator using “Tag Talk
1100”, a Lotek manufactured software.
Results
Diving behaviour. Of the four geo-locators that were deployed in 2006, only one
was recovered. In 2007 two Razorbills were seen wearing the plastic band without the
44
geo-locator attached indicating that the attachment had failed and the third Razorbill was
not observed after the geo-locator was deployed in 2006. The fourth tag was originally
attached to an adult Razorbill taken from its nest on 4 July 2006. One year later, the tag
was observed on the Razorbill loafing on a rock near the same nest in late June and early
July 2007. Attempts to catch the Razorbill on this loafing rock with noose carpets and
noose poles for approximately two weeks were unsuccessful, but on 7 July, the geolocator was found near the base of the rock, after the bird slipped it off its leg.
The Razorbill was a breeding female adult with two grooves across its mandible,
indicating an age of at least four years (Jones 1988). The bird was nesting in a rock
burrow on the northwestern shore of the island and was brooding a chick at the time it
was equipped with the geo-locator in 2006. Both adult and chick were observed in the
burrow during the remainder of that breeding season. The chick was last seen on 25 July
and was presumed to have fledged within a week after the last encounter.
The total time between the geo-locator deployment and recovery was 368 days.
Maximum daily depth achieved during the deployment ranged from 2.6 meters to 38.9
meters, with a mean of 17.9 ± 6.8 meters.
Initially, the ANOVA for diving behaviour was run with season (S), day (D), time
of day (T) and depth (H). The response variable tested was the mean time spent at a
certain depth per time of day (indicated by proportion of records at each depth). The
ANOVA showed significant interaction between season, depth and time of day (P <
0.0001). There was also a significant interaction between depth and time of day (P <
0.0001) however, due to the influence of season on depth and time of day, performing a
post-hoc analysis on the interaction was impractical.
45
To deal with this complication, the data were divided into two datasets, one each
for the breeding season and non-breeding season. This allowed statistical analysis of the
interaction between time of day and depth at the cost of being able to compare the effects
of season.
Non-breeding season. During the non-breeding season most diving activity took
place during the morning and afternoon where about one-quarter of time was spent diving
in each interval (Table 3.1). Almost no time was spent diving at dawn, dusk and night.
Deepest dives (around 16 m) in the non-breeding season occurred in the morning and
afternoon: few dives at other times went below 10 m (Table 2.1).
The ANOVA showed a significant interaction between time of day and depth (p <
0.0001; Table 3.2). The Razorbill used the three shallowest depth classes more than the
deeper ones; however, post-hoc analysis indicated that this difference was significant
only during the morning and afternoon (Appendix II). At dawn, dusk and night, the
seldom-used periods of the non-breeding season (Table 3.1; Fig 3.2), depths did not differ
significantly (Appendix II).
In the mornings, the 0.2 – 1.9 and the 5 – 14.9 m depths were used most
frequently (17-18% of dives) and in the afternoons a similar proportion of dives were
between 5 and 14.9 m (Table 3.3).
Breeding season. Diving activity during the breeding season peaked in the
afternoon during which the Razorbill spent nearly 20% of its time diving (Table 3.1).
Although a large portion of dusk (23.2 ± 24.7) was spent diving, this amounted to an
average of only 12.1 ± 10.3 minutes per day (Table 3.1). Unlike the non-breeding season,
here the afternoon was clearly the time of day that was most commonly used for diving
46
(Fig 3.3). The Razorbill spent less than 9% of the morning diving, noticeably less than in
the non-breeding season; diving at dawn and night were rare during the breeding season
(Table 3.1).
Like the non-breeding season, deepest dives (9-10 m) took place during the
morning and afternoon. Breeding season maximum dives were always shallower than
maximum dives from the same time of the day in the non-breeding season (Table 3.1).
The breeding season ANOVA revealed a significant interaction between time of
day and depth (p < 0.0001; Table 3.2), which indicates that the Razorbill used different
depths at different times of the day. Like the non-breeding season, post-hoc analysis
indicated that this difference was significant only during the morning and afternoon
(Appendix II). The three shallow depth classes (0.2 – 14.9 m) were used more often than
the deeper depth classes and, although not significantly different from each other in terms
of the how much use they received, they were as a group targeted more often than the
deeper depth categories during the morning and afternoon (Appendix II). There was no
significant difference in depths at dawn, dusk and night (Table 3.1; Fig 3.2).
Light-based locations. The ability of the tag to approximate a location was far
too imprecise to reflect the fine-scale movements of the Razorbill within the Bay of
Fundy. Distances between successive locations were so large that the bird would have to
sustain an average speed of 16.6 ± 15.9 km per hour for the entire day to cover them.
Much more precise data from Razorbills equipped with satellite transmitters suggested a
maximum movement speed of 1.04 km per hour for a bird over several days (see Chapter
2), but none of the minimum speeds calculated between successive points generated by
the geo-locator were this low. All points were densely clustered around the Gulf of Maine
47
and over the landmasses of Nova Scotia, New Brunswick and northern Maine.
Calculating the movements of this Razorbill was not possible, but the center of the point
cluster was calculated using ArcView 3.3 (Environmental Systems Research Institute
2002) and was determined to be 43°28´10N 67°3´10W (Fig 2.4).
To help determine the precision of the geo-locator, all points that were obtained
over land, and thus known to be unrealistic Razorbill locations, were isolated and the
shortest distance to the coast was calculated for each point. On average, these locations
were 12.4 ± 21.1 km from the ocean.
Discussion
The data presented here represent the first year-long deployment of an archival
tag on a breeding Razorbill and the most complete picture of a Razorbill’s yearly diving
behaviour to date. This is especially true for data regarding the non-breeding season,
which has received very little attention in the literature. Almost all current knowledge of
Razorbill diving behaviour, including that used for comparison in this discussion comes
from the breeding season when the bird is returning frequently to its nesting site and thus
most accessible. It is also important to note that all results in this study are based on a
sample of one bird and individual differences in foraging behaviour is a common
observation. Further study is required before any conclusions regarding Razorbill diving
behaviour can be drawn from these findings. Nonetheless, a discussion of this individual
and how it compares to current Razorbill knowledge will make an insightful, if
contingent, comparison.
Seasonal differences. Although the non-breeding season data (Fig 3.2) followed
much the same basic pattern as the breeding season data (Fig 3.3), there were some
48
notable exceptions. In the non-breeding season the Razorbill distributed its diving effort
almost equally between morning and afternoon, whereas in the breeding season most of
the diving effort was concentrated in the afternoon. It is possible that during the breeding
season certain obligations in the colony such as brooding or attending large social
gatherings may have limited diving activities to certain times in the afternoon. This bird
did raise a chick in the 2006 breeding season and it is possible that its mate may have
concentrated its diving efforts in the morning while this individual was brooding the
chick. Different foraging patterns for male and female Razorbills has been observed at
the Gannet Islands, Labrador (Paredes et al. 2008) although further studies involving both
pair members are needed to determine if this pattern is typical of the MSI population.
Mean maximum dives were deeper in the non-breeding season than in the
breeding season at all times of day (Table 3.1); and the Razorbill never dived deeper than
25 m and rarely below 15 m in the breeding season, but these depths were achieved much
more commonly in the non-breeding season (Table 3.3). Although depths used in both
breeding and non-breeding seasons are comparable to depths achieved in studies
involving breeding season behaviour (Harris et al. 1990; Benvenuti et al. 2001;
Dall’Antonia et al. 2001), this is the first case where Razorbill diving behaviour has been
shown to differ between the breeding and non-breeding seasons. The Bay of Fundy
reaches depths of 380 m and depths of 80 to 120 m are characteristic of the area around
Machias Seal Island so it is unlikely that dives did not exceed 40 m due to a narrow range
of available depth.
A nesting seabird can forage only so far from the breeding colony before it has to
return to the nest to feed its chick or incubate. Dall’Antonia et al. (2001) reported that
49
Razorbills in Iceland, foraging 29.0 ± 17.3 km away from their breeding site, invested a
daily average of 2.3 ± 0.8 hours of flight in 1997 and 1.2 ± 0.8 hours in 1998 in their
foraging trips. Benvenuti et al. (2001) reported that Razorbills in Denmark invested an
average of only 14.0 ± 12.8 minutes of flight time in their diurnal foraging trips and 38.5
± 20.1 minutes in their nocturnal trips; foraging distance ranged between 1 and 35 km.
Assuming such short flights, it is reasonable to assume that a nesting Razorbill would
minimize its foraging radius and concentrate the bulk of its breeding season dives in
shallower waters closer to Machias Seal Island. This Razorbill certainly made deeper
dives in the non-breeding season than it did in the breeding season (Table 3.1); however
it is unknown whether this reflects a shift in the prey it was exploiting or exploitation of
the same species in deeper water after ties to the breeding site, and the shallower waters
surrounding it, are severed. Juvenile Atlantic herring in the Gulf of Maine are abundant in
surveys at depths of 15 – 135 m in summer and depths of 30 – 90 m at other times of the
year (Reid et al. 1999). If, for energetic reasons, the Razorbill is exploiting herring at
shallower ranges of these distributions then deeper dives would likely be required outside
of the breeding season. Studies of Razorbill diet outside the breeding season are needed
to explore this further.
Diving Behaviour. This Razorbill concentrated the majority of its diving efforts
in the daylight hours with a negligible amount of diving activity at night and very little at
dawn and dusk. Although dusk and dawn experienced an apparent increase in diving
activity when compared to night, these times are not significantly different from each
other.
50
This is a distinctly different pattern than reported by Benvenuti et al. (2001)
where breeding Razorbills in Denmark concentrated foraging during early morning
(02:00 to 05:00) and evening (20:00 to 22:00) and made only moderate use of daylight
hours. Similarly, Razorbills from Iceland are reported to follow at least two different
foraging patterns, one where birds foraged primarily around midnight and a second using
the morning (05:00 to 09:00) and afternoon (around noon; Dall’Antonia et al. 2001). The
MSI bird most resembled the morning/afternoon pattern observed in Iceland, diving very
little at night.
Whether this is typical of the entire MSI population or not requires further study;
however one possible explanation for the discrepancy, as noted by Dall’Antonia et al.
(2001), is the increased day length experienced at higher latitudes during the breeding
season. Pursuit divers such as Razorbills are visual hunters so, assuming that diving is an
indication of food acquisition, the use of daylight hours would be expected. Birds
breeding farther north in the summer will experience more daylight, which may in turn
affect their foraging behaviour.
During the breeding season, the female Razorbill from MSI concentrated the bulk
of its dives in the afternoon and the male presumably concentrated its effort in the
morning. Both Razorbill adults are involved in brooding chicks and shift changes
generally coincide with feedings (Hipfner and Chapdelaine 2002), however any sexlinked differences in diving behaviour are apparently different than other North American
colonies. Paredes et al. (2008) reported that, during the breeding season, females
concentrated about half of their dives in the morning and males at noon and early
afternoon and proposed that male presence at the breeding site at times when the chick is
51
most at danger, would potentially benefit the fitness of both parents and consequently
determine foraging schedules. Paredes et al. (2008) hypothesized that increased
intraspecific kleptoparasitism at key feeding times and interloping Razorbills as possible
causes of this behaviour. MSI however has a much lower rate of kleptoparasitism than
the Gannet Islands (Lavers and Jones 2007) and is unlikely to experience the level of nest
site competition seen at the Gannet Islands due to the increased number of floaters that
generally visit larger colonies (Paredes et al. 2008). MSI supports approximately 600
Razorbill pairs (Grecian et al. 2003) while the Gannet Islands is the largest breeding
colony of Razorbills in North America with an estimated 9,800 breeding pairs
(Chapdelaine et al. 2001)
If the behaviour of male Razorbills at MSI is being modified by the necessity to
defend the nest site, it is most likely in response to marauding Herring Gulls (Larus
argentatus), which prey on seabird eggs and chicks (Pierotti and Good 1994; personal
observation). Gull predation at MSI is a new occurrence (Grecian (2005) considered the
Razorbill colony at MSI to be predator free) and Razorbills have been observed chasing
Herring Gulls away from nest sites (personal observation). No diurnal pattern in gull
predation has been reported at MSI so any effects on Razorbill foraging behaviour cannot
be determined. In any case, if Razorbill foraging patterns are influenced by the need to
defend the nest site then Razorbills breeding at different colonies likely employ different
foraging patterns reflective of different threats.
Barrett and Furness (1990) suggested that dive depths reflect prey distribution
more than Razorbill diving ability. However it is difficult to discuss diving patterns in
relation to prey species because little is known about Razorbill diet outside the breeding
52
season. Adult Razorbills in the Bay of Fundy are suspected to feed primarily on Atlantic
herring (Clupea harengus; Bond 2007). Long-term chick provisioning data at Machias
Seal Island indicate that chick diet is composed predominantly of juvenile herring (Clarke
et al. 2008) and stable-isotope analysis indicates a similar adult diet (Bond 2007).
Stomach contents of twenty-one wintering Razorbill adults collected near Grand Manan
in January of 1998 and 1999 consisted largely of euphausiids/krill (Meganyctiphanes
norvegica), crustaceans, Three-spined Stickleback (Gasterosteus aculeatus) and Atlantic
herring although stable-isotope analysis of these birds suggested a herring diet
(Huettmann et al. 2005). In the 2006 and 2007 breeding seasons, Razorbills at Machias
Seal Island provisioned their chicks primarily with Herring and Sandlance (Ammodytes
americanus; Clarke et al. 2008) so the diving data obtained for the breeding season likely
reflect the Razorbill’s exploitation of these prey items. Juvenile and adult herring perform
vertical migrations moving up in the water column when light intensity is low with
highest activity occurring just after sunrise and before sunset (Reid et al. 1999). This
Razorbill spent very little time diving at dusk, dawn and night, when herring are nearest
the surface, possibly due to reduced visual acuity in reduced light. However the high
levels of diving activity observed in the morning and afternoon may correspond to the
timing of the herring vertical migration that occurs just after sunrise and before sunset.
Surprisingly, the deepest depth achieved by this Razorbill was 38.9 meters, which
did not approach the maximum depths reported by Piatt and Nettleship (1985) who
collected 9 Razorbills as by-catch in gill nets fixed on the ocean floor; 1 bird was trapped
in gear at 70 to 80 m and another at 110 to 120m. These depths greatly exceed the range
achieved in this study in which no dives below 40 m were recorded, but post-hoc analysis
53
suggested that the shallowest depths (less than 15 m) were targeted much more than
depths over 15 m. It is important to note that bird recoveries from nets come with the
caveat that it is possible for birds to become entangled and drown while the net is being
lowered into the water column or while it is being retrieved (Piatt and Nettleship 1985).
Such an occurrence would exaggerate the diving potential of any recovered birds.
Although it did not reach the maximum depths reported buy Piatt and Nettleship
(1985), the depths used by this Razorbill are in agreement with other studies of Razorbill
diving behaviour as well as with distribution of herring, its primary prey species, which
migrate vertically in the water column in response to changing light intensity (Reid
1999). Harris et al. (1990) reported maximum diving depths for 11 Razorbills from 14 to
32 m, and Benvenuti et al. (2001) reported that 50% of dives were within 15 m and dives
beyond 40 m were rare. Dall’Antonia et al. (2001) reported that 70% of dives were within
15 m and that dives beyond 35 m were rare.
Although the conclusion of Piatt and Nettleship (1985) that Razorbills’
physiological capabilities allow them to achieve depths of 120 m has been confirmed by
observations from a submersible (Jury 1986), the Razorbill in this study spent an entire
year without going below 40 m, consistent with the 4 other studies of diving behaviour in
Razorbills.
Tag performance. It is not unusual for externally-mounted devices deployed on
marine animals to become detached (Harris et al. 1990; Newman et al. 1999; Hays et at.
2007) however this style of attachment proved to be a particularly unreliable method of
attaching geo-locators to our sample of Razorbills. It is possible that the Razorbills
particular style of walking was partially responsible for the loss of geo-locators.
54
Excessive shuffling through the breeding colony on its tarsus may have weakened the
plastic zip ties by constantly grinding them over the rocks. Radio transmitters mounted on
plastic bands that were recovered from Razorbills showed considerable damage from
being ground over the rocks (Appendix III). Unfortunately, leg attachment is likely the
least invasive method of attachment for a diving bird. Glue or subcutaneous attachments
would not last for a year-long deployment (Newman et al. 1999) and a backpackmounted device would likely cause substantial hydrodynamic drag as the bird was
swimming (Bannasch et al. 1994). Common Murres equipped with externally-mounted
devices do provision their chicks less although their mates compensate for this reduction
(Wanless et al. 1988a; Hamel et al. 2004). This Razorbill bred successfully in 2006 but
behaviour modifications due to the tag are unknown. Wanless et al. (1988a) reported no
effects of externally-mounted devices on Razorbill behaviour.
Although the geo-locator provided precise depth data, the light data were much
too imprecise to reflect the small-scale movements of the Razorbill. In order for the tag to
be able to determine movements of the bird, the geo-locator has to be moved to some
point where the distance traveled is great enough to result in a change in the time of
sunrise and sunset. The Razorbill remained in a very small area during the year that the
geo-locator was deployed and given that the tag sampled light data in 2-minute intervals
it is unlikely that the bird’s movements around the Bay of Fundy resulted in detectible
differences in sunrise from day to day. Similarly, Begg and Reid (1997) reported that
Razorbills foraging on the Irish Sea Front were more spatially restricted than other
foraging seabirds including the Common Murre. In this regard the device is more suited
to studies in which tagged individuals are migrating over large distances rather than small
55
local movements; however this study was undertaken without prior knowledge that at
least some Razorbills from Machias Seal Island are likely a resident population within the
Bay of Fundy (see Chapters 2 and 4) as appears to be the case with this individual.
It is possible that short dives less than 2 minutes are under-represented in this
analysis. Unlike data-loggers which can be programmed to activate and measure diving
profile when submerged (Dall’Antonia et al. 2001) the geo-locators in this study sample
in 2-minute intervals throughout a 24-hour period and record the bird’s depth in the water
column at that point. It is generally accepted that a bird can be expected to carry a
package only about 5% of its body weight (Bannasch et al. 1994) and although this geolocator could have been programmed to sample at a rate that would reflect diving profile,
the limited internal memory of the device would not have been sufficient for a year-long
deployment.
This deployment of a geo-locator on a breeding Razorbill has indicated several
behaviour patterns. The most novel finding is that, on a seasonal scale, the Razorbills
diving behaviour in the non-breeding season was markedly different from the breeding
season. During the non-breeding season the birds use of daylight hours for diving was
near equal in the morning and afternoon, whereas in the breeding season most diving
occurred in the afternoon. Also, this Razorbill dived deeper in the non-breeding season
than in the breeding season. On a daily scale, this Razorbill concentrated the vast
majority of its diving effort in the morning and afternoon, unlike foraging patterns from
several northern European colonies. Dives never went below 40 m and most dives did not
go below 15 m, in agreement with most of the literature.
56
Tables
Non-breeding Season
Table 3.1: Dive data for five daily intervals from the non-breeding season (27 Aug 2006
to 7 May 2007; n = 64 days) and breeding season (8 May 2007 to 6 Aug 2007; n = 20
days) for an adult Razorbill from Machias Seal Island, New Brunswick.
Mean percent
of daily
interval spent
at surface
Mean percent
of daily
interval spent
diving
Mean time
spent diving
per daily
interval (min)
Mean
maximum
depth (m)
Breeding Season
Mean percent
of daily
interval spent
at surface
Mean percent
of daily
interval spent
diving
Mean time
spent diving
per daily
interval (min)
Mean
maximum
depth (m)
Dawn
Mean ± SD
Morning
Mean ± SD
Afternoon
Mean ± SD
Dusk
Mean ± SD
Night
Mean ± SD
93.7 ± 10.1
74.0 ± 10.4
76.9 ± 14.3
96.0 ± 9.2
99.9 ± 0.3
6.3 ± 10.1
26.0 ± 10.4
23.1 ± 14.3
4.0 ± 9.2
0.1 ± 0.3
3.8 ± 6.0
80.5 ± 32.5
71.9 ± 44.5
2.4 ± 5.5
0.9 ± 2.3
5.0 ± 3.8
16.0 ± 6.7
15.9 ± 7.7
6.1 ± 6.1
1.7 ± 1.3
93.0 ± 12.8
91.7 ± 9.2
80.8 ± 11.9
76.8 ± 24.7
99.9 ± 0.3
7.0 ± 12.8
8.3 ± 9.2
19.2 ± 11.9
23.2 ± 24.7
0.1 ± 0.3
4.4 ± 7.8
37.2 ± 39.7
87.0 ± 49.3
12.1 ± 10.3
0.4 ± 1.2
3.9 ± 2.7
9.8 ± 4.3
9.9 ± 5.1
5.0 ± 3.4
1.0 ± 0.9
Note: Lotek tags distinguish day from night using a particular light threshold that, when
crossed, indicates sunrise and sunset. Dawn and dusk were then defined as including the
30 minutes before and after sunrise and sunset. Mid-day was then calculated as the point
halfway between sunrise and sunset and was used to define morning and afternoon as the
time between dawn and mid-day, and between mid-day and dusk respectively. Night was
defined as the entire time following dusk and preceding dawn.
57
Table 3.2: Results of the three ANOVAs summarizing the diving behaviour of an adult female Razorbill in
2006/7. Note: primary factors of interest appear in bolded text. Situations where calculation of the F ratio
is not possible due to division by zero are symbolized by “X”.
Source of Variation
Degrees
of
Freedom
Mean
Square
F ratio
denom.
F ratio
P value
ANOVA 1 – Both seasons
Time of Day (T)
Season (S)
Depth (H)
Time of Day*Season (TS)
Time of Day*Depth (TH)
Season*Depth (SH)
Time of Day*Season*Depth (TSH)
Day(Season) (D(S))
Time of Day*Day (Season) (TD(S))
Depth*Day(Season) (HD(S))
Time of Day*Depth*Day(Season)
(THD(S))
Residual (e)
4
1
4
4
16
4
16
82
328
328
1312
0
81.51
1.916
34.945
6.7
10.411
1.799
2.217
0.816
0.698
0.520
0.368
TD(S)
D(S)
HD(S)
TD(S)
THD(S)
HD(S)
THD(S)
e
e
e
e
116.8
2.3
67.2
9.6
28.3
3.5
6.0
X
X
X
X
< 0.0001
0.1332
< 0.0001
< 0.0001
< 0.0001
0.0081
< 0.0001
ANOVA 2 – Non-breeding season
Time of Day (T)
Depth (H)
Day (D)
Time of Day*Depth (TH)
Time of Day*Day (TD)
Depth*Day (HD)
Time of Day*Depth*Day (THD)
Residual (e)
4
4
62
16
248
248
992
0
99.71
37.52
0.76
12.59
0.65
0.51
0.361
TD
HD
e
THD
e
e
e
153.4
73.6
X
34.9
X
X
X
< 0.0001
< 0.0001
ANOVA 3 – Breeding season
Time of Day (T)
Depth (H)
Day (D)
Time of Day*Depth (TH)
Time of Day*Day (TD)
Depth*Day (HD)
Time of Day*Depth*Day (THD)
Residual (e)
4
4
20
16
80
80
320
0
25.57
11.99
0.98
4.22
0.85
0.54
0.39
TD
HD
e
THD
e
e
e
30.1
22.2
X
10.8
X
X
X
< 0.0001
< 0.0001
58
< 0.0001
< 0.0001
Table 3.3: Values for diving parameters characterizing the interactions between diving depth and time of
day appearing in Figures 2.2 and 2.3. Values correspond to average percent of daily dives spent diving at
depth at given time of day.
Depth (m)
0.2 to 1.9
2 to 4.9
5 to 14.9
15 to 24.9
25 to 35
Non-breeding Season
Dawn
Morning
Afternoon
Dusk
Night
1.6 ± 3.1
18.0 ± 11.6
12.0 ± 10.7
0.6 ± 1.4
0.5 ± 1.3
0.6 ± 1.7
14.4 ± 9.2
10.3 ± 6.4
0.4 ± 1.2
0.1 ± 0.3
0.5 ± 1.1
17.2 ± 11.9
16.0 ± 12.5
0.8 ± 3.4
0.0 ± 0.2
0.0 ± 0.3
2.4 ± 4.2
3.4 ± 5.0
0.1 ± 1.0
0.0 ± 0.0
0.0 ± 0.0
0.3 ± 0.0
0.4 ± 1.7
0.2 ± 1.9
0.0 ± 0.0
Breeding Season
Dawn
Morning
Afternoon
Dusk
Night
1.7 ± 3.1
6.0 ± 6.9
22.0 ± 17.0
4.8 ± 6.9
0.3 ± 0.8
1.7 ± 4.2
10.2 ± 10.1
20.4 ± 10.7
3.6 ± 3.6
0.1 ± 0.4
0.6 ± 1.8
8.9 ± 12.5
15.9 ± 15.0
2.8 ± 5.2
0.0 ± 0.0
0.0 ± 0.0
0.3 ± 1.1
0.7 ± 1.8
0.0 ± 0.0
0.0 ± 0.0
0.0 ± 0.0
0.0 ± 0.0
0.0 ± 0.0
0.0 ± 0.0
0.0 ± 0.0
59
Figures
Figure 3.1: LTD2400_50 geo-locator manufactured by Lotek Wireless, Inc. attached to a
plastic coiled band by two plastic zip-ties
60
20.0
% of dives per day spent at depth
18.0
16.0
14.0
Depth (m)
12.0
0.2 to 1.9
2 to 4.9
10.0
5 to 14.9
15 to 24.9
8.0
25 to 35
6.0
4.0
2.0
0.0
Dawn
Morning
Afternoon
Dusk
Night
Mean % of dives per day spent at depth
Figure 3.2: The use by an adult female Razorbill of each depth category throughout the
day in the non-breeding season. Standard deviations (Table 2.3) omitted for clarity.
25.0
20.0
Depth (m)
15.0
0.2 to 1.9
2 to 4.9
5 to 14.9
10.0
15 to 24.9
25 to 35
5.0
0.0
Dawn
Morning
Afternoon
Dusk
Night
Figure 3.3: The use by an adult female Razorbill of each depth category throughout the
day in the breeding season. Standard deviations (Table 2.3) omitted for clarity.
61
Figure 3.4: Map of Bay of Fundy region and the center of the locations of a geo-locator
deployed on an adult female Razorbill from the Machias Seal Island breeding colony from
4 July 2006 to 7 July 2007.
62
References
Bannasch, R., R.P. Wilson and B. Culik. 1994. Hydrodynamic aspects of design and
attachment of a back-mounted device in penguins. Journal of Experimental
Biology 194: 83-96.
Barrett, T. and R.W. Furness. 1990. The prey and diving depths of seabirds on Hornøy,
North Norway after a decrease in the Barents Sea capelin stocks. Ornis
Scandinavica 21: 179-186.
Begg, G.S. and J.B. Reid. 1997. Spatial variation in seabird density at a shallow sea tidal
mixing front in the Irish Sea. ICES Journal of Marine Science 54: 552-565.
Benvenuti, S., L. Dall’Antonia and P. Lyngs. 2001. Foraging behaviour and time
allocation of chick-rearing Razorbills Alca torda at Græsholmen, central Baltic
Sea. Ibis 143: 402-412.
Bond, A.L. 2007. Patterns of mercury burden in the seabird community of Machias Seal
Island, New Brunswick. M.Sc. thesis. Department of Biology, University of New
Brunswick, New Brunswick.
Chapdelaine, G., A.W. Diamond, R.D. Elliot and G.J. Robertson. 2001. Status and
population trends of the Razorbill in eastern North America. Canadian Wildlife
Service Occasional Paper 105.
Clarke T.C., M.F. Godin and A.W. Diamond. 2008. Machias Seal Island 1995 – 2007
progress report. Atlantic Co-operative Wildlife Ecology Research Network,
University of New Brunswick, Fredericton, New Brunswick.
Dall’Antonia, L., G.A. Gudmundsson and S. Benvenuti. 2001. Time allocation and
foraging pattern of chick-rearing Razorbills in northwest Iceland. The Condor
103: 469-480.
Environmental Systems Research Institute. 2002. ArcView GIS 3.3. Redlands,
California.
Gaston, A.J. and I.L. Jones. 1998. The Auks. Oxford University Press, Oxford, U.K.
Grecian, V.D. 2005. The effects of physical and biological parameters on the breeding
success if Razorbills (Alca torda L. 1758) on Machias Seal Island, N.B., in 2000
and 2001. M.Sc. thesis, Department of Biology, University of New Brunswick,
Fredericton.
Grecian, V.D., A.W. Diamond and J.W. Chardine. 2003. Sexing Razorbills (Alca torda)
breeding at Machias Seal Island using discriminant function analysis. Atlantic
Seabirds 5: 73-81.
63
Hamel, N.J., J.K. Parrish and L.L. Conquest. 2004. Effects of tagging on behavior,
provisioning, and reproduction in the Common Murre (Uria aalge), a diving
seabird. The Auk 121: 1161-1171.
Harris, M.P., H. Towll, A.F. Russell and S. Wanless. 1990. Maximum dive depths
attained by auks feeding young on the Isle of May, Scotland. Scottish Birds 16:
25-28.
Hays G.C., C.J.A. Bradshaw, M.C. James, P. Lovell and D.W. Sims. 2007. Why do
Argos satellite tags deployed on marine animals stop transmitting? Journal of
Experimental Marine Biology and Ecology 349: 52-60.
Hipfner, J.M. and G. Chapdelaine. 2002. Razorbill (Alca torda). In The Birds of North
America, No. 635 (A. Poole and F. Gill, eds.). The Birds of North America, Inc.,
Philadelphia, PA.
Huettmann, F., A.W. Diamond, B. Dalzell and K. Macintosh. 2005. Winter distribution,
ecology, and movements of Razorbills Alca torda and other auks in the outer Bay
of Fundy, Atlantic Canada. Marine Ornithology 33: 161-171.
Jones, P.H. 1988. Post-fledging wing and bill development in the Razorbill Alca torda
islandica. Ringing & Migration 9: 11-17.
Jury, J.A. 1986. Razorbill swimming at depth of 140 m. British Birds 79: 339.
Lavers and Jones. 2007. Impacts of intraspecific kleptoparasitism and diet shifts on
Razorbill Alca torda productivity at the Gannet Islands, Labrador. Marine
Ornithology 35: 1-7.
Newman, S.H., J.Y. Takekawa, D.L. Whitworth and E.E. Burkett. 1999. Subcutaneous
anchor attachment increases retention of radio transmitters on Xantus’ and
Marbled Murrelets. Journal of Field Ornithology 70: 520-534.
Paredes, R., I.L. Jones, D.J. Boness, Y. Tremblay and M. Renner. 2008. Sex-specific
differences in diving behaviour of two sympatric Alcini species: Thick-billed
Murres and Razorbills. Canadian Journal of Zoology 86: 610-622.
Piatt, J.P. and D.N. Nettleship. 1985. Diving depths of four alcids. The Auk 102: 293297.
Pierotti, R.J. and T.P. Good. 1994. Herring Gull (Larus argentatus), The Birds of North
America Online (A. Poole, Ed.) Ithaca: Cornell Lab of Ornithology; Retrieved
from the Birds of North America Online:
http://bna.birds.cornell.edu/bna/species/124 doi:10.2173/bna.124
64
Reid, R.N., L.M. Cargnelli, S.J. Griesbach, D.B. Packer, D.L. Johnson, C.A. Zetlin,
W.W. Morse and P.L. Berrien. 1999. Essential fish habitat source document:
atlantic herring, Clupea harengus, life history and habitat characteristics. U.S.
Department of Commerce, NOAA. National Marine Fisheries Service, Northeast
Region, Northeast Fisheries Science Center, Woods Hole, Massachusetts.
Wanless, S., M.P. Harris and J.A. Morris. 1988a. The effect of radio transmitters on the
behaviour of Common Murres and Razorbills during chick rearing. The Condor
90: 816-823.
Wanless, S., J.A. Morris and M.P. Harris. 1988b. Diving behaviour of guillemot Uria
aalge, puffin Fratercula arctica and razorbill Alca torda as shown by radiotelemetry. Journal of Zoology London 216: 73-81.
Watanuki, Y., S. Wanless, M. Harris, J.R. Lovvorn, M. Miyazaki, H. Tanaka and K. Sato.
2006. Swim speeds and stroke patterns in wing-propelled divers: a comparison
among alcids and a penguin. Journal of Experimental Biology 209: 1217-1230.
65
Chapter 4 – Wintering Razorbills (Alca torda) in the Bay of Fundy
Abstract
In the past, large flocks of Razorbills, often exceeding the local breeding
population, have been observed wintering off the island of Grand Manan, New
Brunswick however the origin of these birds is unknown. Thirty-nine radio transmitters
were deployed on Razorbills breeding at Machias Seal Island, New Brunswick, Corossol
Island, Quebec and the Gannet Islands, Labrador to determine if these individuals were
spending a portion of the winter in the Bay of Fundy, near Grand Manan. From January –
March 2008 21 frequencies were monitored near the island; 13 from Machias Seal Island,
6 from Corossol Island and 2 from the Gannet Islands. Consequently, I conclude that
Razorbills from all three breeding sites contributed to the 2007/08 wintering population
in the Bay of Fundy.
Introduction
Flocks of wintering Razorbills have been reported in a number of North American
locations including Labrador, the Gulf of Maine, southern Greenland, southern Nova
Scotia, and on the Hamilton Banks (Gaston and Jones 1998). Large flocks of Razorbills
are also commonly sighted near Grand Manan, New Brunswick during winter months
and Huettmann et al. (2005) reported that there were upwards of 50,000 Razorbills
wintering in the Bay of Fundy in 1997/8 and 1998/9. This translates to ~65% of the total
North American breeding population (~38,000 pairs; Chapdelaine et al. 2001) which
prompted the Huettmann et al. (2005) to suggest that the bay is an important wintering
area for a significant portion of the North American population.
66
Such a large influx of Razorbills into the Bay of Fundy during the winter months
would almost certainly require a contribution from the two largest Razorbill breeding
areas, the Gulf of St. Lawrence and the Gannet Islands, Labrador. The combined North
Shore of the Gulf of St. Lawrence supports approximately 15,200 pairs while the Gannet
Islands support the largest single colony of Razorbills in North America (~9800 pairs;
Chapdelaine et al. 2001).
The breeding population at Machias Seal Island (MSI) is also a likely contributor
to the wintering population described by Huettmann et al. (2005). MSI is the largest
Razorbill breeding colony (592 pairs; Grecian et al. 2003) in the Gulf of Maine and is
only 18.8 km from Grand Manan. Furthermore, Razorbills from the breeding population
at MSI have an established relationship of exchange with the Gannet Islands. Nine
Razorbills banded as chicks at MSI were re-sighted in later years on the Gannet Islands,
Labrador (Lavers 2006), suggesting that both populations come into contact with each
other at some time and place.
Such a highly concentrated wintering population in one location implies that
much of the North American population can be affected by events within the Bay of
Fundy. Razorbills wintering in the Bay of Fundy region may be subject to the effects of
paralytic shellfish poisoning toxins (Haya et al. 2003), eco-tourism, coastal development
(Ronconi and St. Clair 2002), continuing aquaculture activities and development which
have the potential to affect the ecosystem by influencing distribution patterns of both
seabird prey and local gull populations (Ronconi and Wong 2003). The potential for oiled
birds is another major concern. The oil port of Saint John has a frequent spill record
(Lock et al. 1994) and the proposed LNG terminal(s) would also take tankers, each of
67
which typically transports approximately 125,000 – 140,000 m3 of liquid natural gas
(Jacques Whitford Environment Limited. 2004) through inshore waters.
This study was undertaken to test the hypothesis of Huettmann et al. (2005) that
Razorbills from breeding sites in Labrador, Quebec and New Brunswick and Bay of
Fundy contribute to the large wintering population in the Bay of Fundy.
Methods
Study area. In July and August 2007, 39 Razorbills were equipped with a small
radio transmitter. Fifteen transmitters were deployed on Razorbills from the Gannet
Islands, Labrador (53056’N, 56032’W), ten on birds from the Corossol Island migratory
bird sanctuary (50005’N, 66023’W), Quebec, and 14 at the MSI migratory bird sanctuary,
New Brunswick (44030’N, 67006’W; Fig 4.1).
The Gannet Islands is the largest breeding colony of Razorbills in North America
with an estimated 9,800 breeding pairs, while the numerous breeding colonies scattered
along the north shore of the Gulf of St. Lawrence contain an estimated 15,200 pairs
(Chapdelaine et al. 2001). MSI supports about 600 Razorbill pairs (Grecian et al. 2003)
and represents the closest breeding colony to the wintering area described by Huettmann
et al. (2005).
Adult breeding birds were snared in noose-mats on loafing ledges and each bird
was measured, banded, equipped with a radio transmitter and released back into the
breeding colony. Breeding status was assessed by the presence of a brood patch; only
birds with a brood patch were equipped with a transmitter.
Radio transmitters. Radio transmitters were a custom design manufactured by
Holohil Systems Ltd., Ontario. Plastic coiled bands were attached to the transmitters with
68
heavy monofilament line prior to being coated with epoxy during the production process.
This band was snapped around the tarsus of the Razorbill when it was captured in the
field. The transmitters had a 27 cm whip antenna and a mass of 6.7g. The predicted
lifespan of the device was 12 months with an expected range of 1 to 2 km. The
Razorbill’s particular mode of walking (shuffling on its tarsus) was expected to facilitate
loss of the transmitters after one or two years due to the constant grinding of the plastic
band over rocks in the breeding colony.
Flights. Five flights were conducted between 4 Jan and 3 March to search for the
39 transmitter signals around the island of Grand Manan. Flights were conducted in a
twin-engine Piper Seneca PA-34-200 flying up to 2.5 km from the coast of Grand Manan
at 100 to 200 m above sea level. One three-element yagi antenna was mounted to the
exterior of the aircraft and a two-element antenna was kept inside the aircraft and used
through the fiberglass hull. Flights ran clockwise around the island with detours to survey
Old Proprietor Shoal (Fig 4.2), where Huettmann et al. (2005) reported large flocks of
wintering alcids, and MSI (44060’N, 67000’W), the nearest Razorbill breeding site.
During the flights, one researcher would constantly monitor for bird frequencies
using a Model R1000 telemetry receiver (Communications Specialists, Inc.) while the
second researcher and pilot would look for flocks of seabirds. Razorbills were
distinguished from other birds primarily by body size, and on occasion colour. Although
turbulence was an occasional hindrance, binoculars were used to help identify species
whenever possible. Rafting birds were rarely startled by the plane, which allowed
multiple flyovers and prolonged observation. Estimates of Razorbill numbers were then
69
noted. Flights were conducted in early afternoon and would last between 35 min and 65
min.
During the flights one transmitter was left at the airport near the runway and its
frequency was monitored until it was no longer audible. This transmitter was monitored
during the departure and return to the airport to determine transmitter range.
Results
The range of the transmitter left at the airport varied between 1 and 1.5 km while
on land but its range generally increased by 1 or 2 km when detected from the plane. The
greatest distance achieved was between 4 and 4.5 km. Of the 39 transmitters deployed,
21 (53.8%) were detected during flights and one was lost before the flights began. This
transmitter was deployed on a bird from the Gannet Islands that was shot near Postville,
Labrador and reported in early October 2007 (See Discussion).
Thirteen of the 14 transmitters (92.9%) deployed at MSI were detected, while
only 6 out of 10 (60% )of those deployed at Corossol Island were detected. Excluding the
one lost transmitter, 2 of the 15 (13.3%) transmitters deployed at the Gannet Islands were
detected (Table 4.1).
On two of the five flights (29 Jan and 12 Feb) large flocks of Razorbills were
observed rafting on Old Proprietor Shoal. These rafts were very extensive and contained
approximately 8,000 to 10,000 individuals (Table 4.1). All transmitter frequencies
monitored during the flights were detected south of Grand Manan on Old Proprietor
Shoal (Fig 4.2) and to a lesser extent, near the southwest corner of the island.
Transmitters were never detected along the northern coast of Grand Manan or in the area
of the MSI bird sanctuary, and birds were rarely seen in these locations.
70
On the last flight, 3 March, no frequencies were detected and very few Razorbills
were seen: approximately 25 Razorbills near MSI and 50 on Old Proprietor Shoal.
Several other species were identified during flights. Herring Gulls (Larus
argentatus) and Black-backed Gulls (L. marinus) were commonly seen, the largest flock
being ~100 individuals on 12 Feb. Small groups of up to 5 Common Eiders were seen
along the south and east coast of Grand Manan and a flock of ~100 birds was seen on 3
March. No murres were seen during the flights although they may have been represented
in small flocks of alcids (less than 50 individuals), which were to far away too be
identified to species.
Discussion
As pointed out by Huettmann et al. (2005), the large numbers of Razorbills
observed off Grand Manan during winter can be explained only by an influx of birds
from outside the Bay of Fundy. The nearby colony at MSI can account for only about 600
pairs (Grecian 2005) and other colonies in the region on the Yellow Murre Ledges and
along the Maine coast sustain a further 650 or so pairs (Chapdelaine et al. 2001). This
investigation not only confirms Huettmann et al.’s (2005) deduction but also establishes a
direct link between the Grand Manan wintering population and the two largest Razorbill
breeding areas in North America. All three breeding populations involved in this study
were represented in the wintering population in Jan – Feb 2008 and although it is still
unknown how much of the wintering population came from the respective breeding sites,
it is certain that some individuals from Corossol Island, the Gannet Islands and MSI did
spend a portion of the Jan – Feb 2008 winter near the island of Grand Manan.
71
Fewer of the transmitters deployed at the Gannet Islands and Corossol Island were
detected than those deployed at MSI. It is possible that the missing individuals were
simply absent from the immediate area during the flights. On occasions when large rafts
of Razorbills were observed, small flocks of 10 to 50 birds were also encountered up to 2
or 3 kilometers away from the main flock. It is possible that these small flocks may
account for some of the missing frequencies. Considering a larger spatial scale, in the
past Razorbills have been sighted in the Gulf of Maine on the Nantucket Shoals and on
Georges Bank (Chapdelaine 1997; Hipfner and Chapdelaine 2002). Whether wintering
Razorbills in the Gulf of Maine move about as one population or multiple large flocks is
still unknown, however the existence of multiple wintering flocks in the Gulf is a
possibility that might account for frequencies that were not in the range of the flights and
thus went undetected.
It is also possible that undetected frequencies belonged to transmitters that were
damaged, lost or removed by the bird, however neither seems likely given the
asymmetric representation of the three breeding colonies during the flights. Razorbills
were often seen wearing radio transmitters in all three breeding colonies in 2007 and
aside from occasionally stepping on the antenna, no sign of irritation was observed
(personal observation; Raphaël Lavoie personal communication). Furthermore, Wanless
et al. (1988) reported that a Razorbill fitted with an externally-mounted device did not
differ significantly from control birds in its colony attendance or chick provisioning
behaviour. One Razorbill from MSI did lose its transmitter after the flights had taken
place and it had returned to the breeding colony in 2008 (Appendix III). The transmitters
were designed to detach at some point after the study so as not to burden the bird, but
72
premature detachment in the Gannet Island population and not the MSI population seems
unlikely as all transmitters were of the same design and material.
One drawback of radio telemetry is the prerequisite that a Razorbill be in flight
while its frequency is being scanned thus presenting another potential source for missed
frequencies during the flights. Radio frequencies are hampered by water and as a result,
any attempts to monitor a radio attached to a Razorbill with its legs submerged will have
a high chance of failure. The alternative to a leg-mounted package was to equip the
Razorbill with a backpack mounted between the scapulas but this option was rejected
because a dorsal attachment of the device would likely cause substantial turbulence as the
bird was swimming (Bannasch et al. 1994). Placing the transmitter on the leg, beyond the
point of the bird where diameter is largest, was necessary to help keep flow disturbances
to a minimum.
The annual murre hunt in Newfoundland and Labrador is a well-documented
source of Razorbill mortality (Elliot 1991), which may contribute to the asymmetric
representation of the three breeding colonies during the flights. A large portion of
recovered Razorbill bands from Labrador and the Gulf of St. Lawrence are from birds
that have been shot in the Newfoundland and Labrador fall murre hunt as they are
migrating (Chapdelaine 1997). Razorbills from MSI would not be exposed to this source
of mortality because there is no murre hunt in the Bay of Fundy. Lavers et al. (2008)
reported that Razorbills at MSI had a much higher adult and pre-breeder survival rate
than the Gannet Islands and suggested hunting or environmental differences as a cause.
Indeed, one Razorbill from this study, equipped with a transmitter at the Gannet Islands
on 9 August, was shot near the town of Postville, Labrador in early October 2007
73
(55°15’N, 59°30´W). Although this bird was removed from our summary in Table 4.1,
other Razorbills from our Gannet Islands sample may have been shot and not reported.
The reverse migration of this individual is also noteworthy. Although still well
within the range of the species, this individual was much farther north than would be
expected at that time of year and had flown at least 260 km along the Labrador coast in a
northwest direction (Fig 4.1). Reverse migration has been recorded in a number of
migratory bird species and is believed to be an important cause of vagrancy (Thorup
2004) but I am not aware of published accounts of the phenomenon in alcid species.
The maximum estimate during the 2008 flights, 10,000 individuals, was 20.3% of
Huettmann et al.’s (2005) reported a maximum count of 49,153 Razorbills on 23 Jan
1998. Also in 2008 no more than a few dozen Razorbills were observed off Long Eddy,
which Huettmann et al. (2005) described as a main area of concentrated feeding activity.
Such a large congregation of wintering birds from multiple breeding colonies
suggests that the Bay of Fundy may serve as an important site for dispersal among
colonies. Razorbill dispersal between breeding colonies has been identified through band
re-sighting (Lavers et al. 2007) and this study suggests that the mixing of breeding
populations off Grand Manan during winter months would provide a likely opportunity
for this to take place. MSI, for example, is a relatively small island (9.5 ha) with a limited
amount of rock piles to serve as Razorbill habitat. Mingling with other breeding
populations during the winter may promote dispersal to sites such as the Gannet Islands,
where breeding habitat is not limited (Lavers 2008).
It is certain from our results that a portion of the birds from the MSI breeding
colony contributed to the wintering population near Grand Manan in Jan – Feb 2008. In
74
fact, of the three breeding populations included in this study, the transmitter frequencies
from the local MSI Razorbills were most commonly observed during the aerial surveys.
This implies that Razorbills that breed in the Bay of Fundy may spend more time in the
Bay of Fundy than birds migrating from other breeding populations. This may even be an
indication that Razorbills breeding at MSI are a resident population. The Bay of Fundy
contains commercial stocks of herring (Townsend 1992) for wintering Razorbills to
exploit and the lack of pack ice would reduce pressure to migrate south. Although
wintering Razorbills are seen further south of the Bay of Fundy in small numbers (Gaston
and Jones 1998), their origin is still unknown. A resident population in the Bay of Fundy
would suggest that large numbers of Razorbills seen at more southerly locations in winter
likely originated from more northerly breeding populations than the Bay of Fundy, and
would suggest “leapfrog” migration in which northern breeders winter further south than
southern breeders, as is known in many other birds (Boland 1990).
This investigation has confirmed a direct connection between three major
Razorbill breeding sites and the wintering population observed on Old Proprietor Shoal in
the Bay of Fundy. These findings highlight some important management issues for the
three Razorbill populations (and likely others). Knowing where a wintering seabird
spends its time is a crucial factor in assessing any potential threats to the species and
implementing any sort of management strategy. Identifying how much time the New
Brunswick population, as well as the North American population as a whole, spends in
the Bay of Fundy will help identify how vulnerable the species is to potential threats in
the Bay of Fundy. Such a high concentration of Razorbills from multiple breeding
colonies suggests that population managers should have an interest in the effects of
75
anthropogenic factors within the Bay of Fundy which likely affect the entire Razorbill
population of North America.
76
Departure Time
Flight Duration
Frequencies Detected
Machias Seal Island
Corossol Island
Gannet Islands
Max. Razorbill Count
General Weather Conditions
13:05
40 min
3
0
0
No Count
Foggy/Calm
5
1
1
No Count
Clear/Calm
12 Jan
11:40
35 min
4 Jan
5
3
2
~8000
Clear/Calm
12:02
51 min
29 Jan
5
3
0
~10,000
Clear/Calm
11:47
65 min
12 Feb
0
0
0
50
Clear/Calm
12:21
55 min
3 March
13
6
2
Total # of
individuals
from each
breeding colony
Table 4.1: Details of flights undertaken to locate Razorbills equipped with radio transmitters near Grand
Manan from Jan to March 2008. Transmitters were deployed at Machias Seal Island (n = 14), Corossol
Island (n = 10) and the Gannet Islands (n = 14).
Tables
77
Figures
Figure 4.1: The location of the Gannet Islands, Machias Seal Island and Corossol Island
in eastern Canada. Also included are Postville, where a bird from the Gannet Islands was
shot, and Grand Manan.
78
Figure 4.2: Map of the Grand Manan region, New Brunswick and the area around
Old Proprietor Shoal, where Razorbills were monitored. Coastal shapefiles
obtained from NOAA Coastline Extractor (Signell 2008).
79
References
Bannasch, R., R.P. Wilson and B. Culik. 1994. Hydrodynamic aspects of design and
attachment of a back-mounted device in penguins. Journal of Experimental
Biology 194: 83-96.
Boland, J.M. 1990. Leapfrog migration in North American shorebirds: intra- and
interspecific examples. The Condor 92: 284-290.
Chapedlaine, G. 1997. Pattern of recoveries of banded Razorbills (Alca torda) in the
Western Atlantic and survival rates of adults and immatures. Colonial Waterbirds
20: 47-54.
Chapdelaine, G., A.W. Diamond, R.D. Elliot and G.J. Robertson. 2001. Status and
population trends of the Razorbill in eastern North America. Canadian Wildlife
Service Occasional Paper 105.
Elliot, R.D. 1991. The management of the Newfoundland turr hunt. In Studies of high
latitude seabirds. Vol. 2. Conservation biology of the Thick-billed Murre in the
Northwest Atlantic. Edited by A.J. Gaston and R.D. Elliot. Canadian Wildlife
Service, Ottawa, Ont. pp. 29-35.
Gaston, A.J. and I.L. Jones. 1998. The Auks. Oxford University Press, Oxford, U.K.
Grecian, V.D. 2005. The effects of physical and biological parameters on the breeding
success if Razorbills (Alca torda L. 1758) on Machias Seal Island, N.B., in 2000
and 2001. M.Sc. thesis, Department of Biology, University of New Brunswick,
Fredericton.
Grecian, V.D., A.W. Diamond and J.W. Chardine. 2003. Sexing Razorbills (Alca torda)
breeding at Machias Seal Island using discriminant function analysis. Atlantic
Seabirds 5: 73-81.
Haya, K., J.L. Martin, S.M.C. Robinson, J.D. Martin and A. Khots. 2003. Does uptake of
Alexandrium fundyense cysts contribute to the levels of PSP toxin found in the sea
scallop, Placopecten magellanicus? Harmful Algae 2: 75-81.
Hipfner, J.M. and G. Chapdelaine. 2002. Razorbill (Alca torda). In The Birds of North
America, No. 635 (A. Poole and F. Gill, eds.). The Birds of North America, Inc.,
Philadelphia, PA.
Huettmann, F., A.W. Diamond, B. Dalzell and K. Macintosh. 2005. Winter distribution,
ecology, and movements of Razorbills Alca torda and other auks in the outer Bay
of Fundy, Atlantic Canada. Marine Ornithology 33: 161-171.
80
Huettmann, F. and A.W. Diamond. 2000. Seabird migration in the Canadian northwest
Atlantic Ocean: moulting locations and movement patterns of immature birds.
Canadian Journal of Zoology 78: 624-647.
Jacques Whitford Environment Limited. 2004. LNG marine terminal/multi-purpose pier
project: Environmental impact statement prepared for Irving Oil Limited.
Fredericton, NB.
Lavers J.L. 2006. Razorbill (Alca torda) survival, dispersal, and age of first breeding on
Machias Seal Island, New Brunswick, 1995-2005. In Bond, A.L., A.L. Black,
M.F. McNutt and A.W. Diamond. 2006. Machias Seal Island 1995-2005 progress
report.
Lavers, J.L, I.L. Jones, A.W. Diamond and G.J. Robertson. 2008. Annual survival of
North American Razorbills (Alca torda) varies with ocean climate indices.
Canadian Journal of Zoology 86: 51-61.
Lavers, J.L., I.L. Jones and A.W. Diamond. 2007. Natal and breeding dispersal of
Razorbills (Alca torda) in eastern north America. Waterbirds 30: 588-594.
Lock, A.R., R.G.B. Brown and S.H. Gerriets. 1994. Gazetteer of marine birds in Atlantic
Canada: An atlas of seabird vulnerability to oil pollution. Canadian Wildlife
Service, Atlantic Region.
Ronconi, R.A. and C.C. St. Clair. 2002. Management options to reduce boat disturbance
on foraging black guillemots (Cepphus grylle) in the Bay of Fundy. Biological
Conservation 108: 265-271.
Ronconi, R.A. and S.N.P. Wong. 2003. Abundance estimates and changes in seabird
numbers of the Grand Manan Archipelago, New Brunswick, Canada. Waterbirds
26:462-472.
Signell, R. 2008. Coastline Extractor. NOAA/National Geophysical Data Center, Marine
Geology and Geophysics Division.
Available at http://rimmer.ngdc.noaa.gov/coast/. Last accessed on 7 May, 2008.
Thorup, K. 2004. Reverse migration as a cause of vagrancy. Bird Study 51: 228-238.
Townsend, D. 1992. Ecology of larval herring in relation to the oceanography of the Gulf
of Maine. Journal of Plankton Research 14: 467-493.
Wilson, R.P., H.J. Spairani, N.R. Coria, B.M. Culik and D. Adelung. 1990. Packages for
attachment to seabirds: what color do Adelie Penguins dislike least? Journal of
Wildlife Management 54: 447-451.
81
Wilson, R.P. and M.T. Wilson. 1989. A peck activity record for birds fitted with devices.
Journal of Field Ornithology 60: 104-108.
Wanless, S., M.P. Harris and J.A. Morris. 1988. The effect of radio transmitters on
behaviour of Common Murres and Razorbills during chick rearing. The Condor
90: 816-823.
82
Chapter 5 - Summary
One of the most consistent trends throughout this thesis has been the strong
association between the Razorbills breeding at Machias Seal Island and the Bay of
Fundy. Almost all MSI birds equipped with radio transmitters were monitored near
Grand Manan in the winter, which suggests these Razorbills are a resident population.
Furthermore, none of the eight birds equipped with satellite transmitters were observed
outside of the Bay of Fundy and although tracked for only a portion of the year, this
sample was monitored in late summer and fall and remained in the Bay of Fundy long
after the breeding season was complete. The importance of the bay is evident not only
for the local breeding population, but for breeding colonies across North America as well.
This thesis has confirmed a relationship between the two biggest Razorbill breeding sites
in North America and the Bay of Fundy, which along with a resident population from
MSI, raises management issues concerning the vulnerability of the population to the
anthropogenic factors within the Bay of Fundy.
The deployment of a geo-locator on a breeding adult Razorbill has resulted in the
first detailed description of Razorbill diving behaviour in the non-breeding season, which
was distinctly different from that of the breeding season. The diving behaviour of this
individual was also different from Razorbills breeding at more northerly colonies, which
perform the majority of their dives at different times of the day.
My findings suggest that the distribution of this sample of Razorbills was
influenced by depth and slope of the ocean floor while chlorophyll concentration and sea
surface temperature, which are generally considered to be indicators of primary
productivity, did not influence Razorbill distribution. The implication is that Razorbill
83
distribution may be influenced more by factors that drive upwelling and mixing than they
are with the effects themselves. Seabird movements are certainly driven by some degree
by the behaviour of their prey species, however my findings suggest that Razorbill
distribution may be equally driven by features where prey are likely to occur as with the
prey itself.
There is an apparent discrepancy between the distribution of Razorbills equipped
with satellite transmitters and the area south of Grand Manan where large flocks of
Razorbills were seen during winter flights in 2008. Razorbills equipped with satellite
transmitters made little use of the Old Proprietor Shoal while they were monitored,
however it is important to note that data used in chapters 2 and 4 are from different times
of the year. It is still unclear when Razorbills from outside of the Bay of Fundy arrive in
the area and it is possible that local birds use a wider range within the bay until the arrival
of other populations. It is also apparent from chapter 2 that Razorbill distribution can
vary at different times of the year (Fig 2.6) and it is certainly possible that whatever
factors make the Old Proprietor Shoal such an attractive location for Razorbills do not
come about until late fall or early winter. The exact reason for the attraction to Old
Proprietor Shoal is one of many aspects of Razorbill behaviour outside the breeding
season that has yet to be fully understood. Analysis of Razorbill distribution from
December to May is still required to complete our understanding of how the MSI
population responds to oceanographic features throughout the year.
The emerging
picture is that Razorbill diving behaviour and foraging strategy is variable from colony to
colony and if this is any indication, comparisons with how other North American
84
colonies are correlated with oceanographic features may help identify differences or
similarities in such strategies.
The asymmetric representation of the three breeding colonies during the winter
flights suggests that there are other large concentrations of Razorbills elsewhere. It
seems clear that Razorbills breeding at MSI were remaining near Grand Manan, however
any alternate wintering locations that are used by birds from more northerly locations
may be identified through satellite telemetry.
This thesis has identified a common behaviour among three breeding populations,
and likely exists for others, associating them with the wintering area in the Bay of Fundy,
but has also highlighted specific diving behaviours not yet reported in other breeding
populations, thereby enhancing our knowledge of the subtle differences between breeding
colonies. I have also supplied insight into the movements and behaviour of the Razorbill
at one of its most southerly breeding colonies, thereby allowing comparisons with
northern populations, where the majority of studies concerning this species have taken
place.
85
Appendix I – Razorbills equipped with Satellite Transmitters and identified after
transmitter failure (see Table 2.2)
Figure 1: Photo of a male Razorbill (ID
= 65598) highlighting the antenna of a
satellite antenna taken in July 2007,
approximately 9 months after last
contact with the PTT. Transmission
period was 116 days.
Figure 2: Photo of a female Razorbill
(ID = 65605) highlighting the antenna of
a satellite antenna taken in July 2008,
approximately 20 months after last
contact with the PTT. Transmission
period was 131 days.
86
Appendix II – Tukey’s HSD Post-hoc Analysis for Razorbill dives (Chapter 3)
Tukey’s HSD post-hoc test was conducted to determine which periods of the day
the Razorbill was focusing its diving activities and which depths were most commonly
visited. Separate analysis was conducted on the breeding season and non-breeding
season data to test the interactions between diving depth and time of day. Results in the
following tables that are covered by the same letter are not considered to be significantly
different from one another. The formula for Tukey’s Honest Significant Difference
(HSD) is HSD =q√(MSe/n).
Non-breeding Season
Mean squared error (MSe) of the F ratio denominator = 0.361
Degrees of freedom in the F ratio denominator = 320
Number of samples (n) in each mean = 62
Number of treatments = 5
Alpha level = 0.05
Critical Value (q) (0.05, 320, 5) = 3.89
HSD = 3.89 √(0.361/62) = 0.30
Table 1: Results of Tukey’s HSD post-hoc analysis showing significant distribution of
diving depths achieved during the non-breeding season with respect to time of day.
Time of Day
Depth Categories
Dawn
a
25 to 32
a
15 to 25
a
5 to 15
a
2 to 5
a
0.2 to 5
Morning
a
25 to 32
a
15 to 25
b
2 to 5
c
0.2 to 5
c
2 to 5
Afternoon
a
25 to 32
b
15 to 25
c
2 to 5
c
0.2 to 5
d
2 to 5
Dusk
a
25 to 32
a
15 to 25
a
5 to 15
a
2 to 5
a
0.2 to 5
Night
a
25 to 32
a
15 to 25
a
5 to 15
a
2 to 5
a
0.2 to 5
87
Table 2: Results of Tukey’s HSD post-hoc analysis showing significant distribution of
times of day during the breeding season with respect to depths achieved.
Time of Day
Depth Categories
0.2 to 2
a
Night
a
Dusk
a
Dawn
b
Afternoon
c
Morning
2 to 5
a
Night
a
Dusk
a
Dawn
b
Afternoon
b
Morning
5 to 15
a
Night
a
Dusk
a
Dawn
b
Morning
b
Afternoon
15 to 25
a
Night
a
Dusk
ab
Dawn
b
Morning
b
Afternoon
25 to 35
a
Night
a
Dusk
a
Dawn
a
Afternoon
a
Morning
Breeding Season
Mean squared error (MSe) of the F ratio denominator = 0.39
Degrees of freedom in the F ratio denominator = 320
Number of samples (n) in each mean = 21
Number of treatment levels = 5
Alpha level = 0.05
Critical Value (q) (0.05, 320, 5) = 3.89
HSD = 3.89 √(0.39/21) = 0.53
Table 3: Results of Tukey’s HSD post-hoc analysis showing significant distribution of
diving depths achieved during the breeding season with respect to time of day.
Time of Day
Depth Categories
Dawn
a
25 to 32
a
15 to 25
a
5 to 15
a
2 to 5
a
0.2 to 5
Morning
a
25 to 32
a
15 to 25
a
0.2 to 5
ab
5 to 15
b
2 to 5
Afternoon
a
25 to 32
a
15 to 25
a
5 to 15
b
0.2 to 5
b
2 to 5
Dusk
a
25 to 32
a
15 to 25
a
5 to 15
a
2 to 5
a
0.2 to 5
Night
a
25 to 32
a
15 to 25
a
5 to 15
a
2 to 5
a
0.2 to 5
88
Table 4: Results of Tukey’s HSD post-hoc analysis showing significant distribution of
times of day during the breeding season with respect to depths achieved.
Time of Day
Depth Categories
0.2 to 2
a
Night
a
Dawn
a
Dusk
a
Morning
b
Afternoon
2 to 5
a
Night
a
Dawn
a
Dusk
b
Morning
c
Afternoon
5 to 15
a
Night
a
Dawn
a
Dusk
b
Morning
c
Afternoon
15 to 25
a
Night
a
Dawn
a
Dusk
a
Morning
a
Afternoon
25 to 35
a
Night
a
Dawn
a
Dusk
a
Morning
a
Afternoon
89
Appendix III – Damaged Radio Transmitters
The radio transmitters pictured in the figure below were both attached to adult
Razorbills from Machias Seal Island. The transmitter on the right is one of the actual
devices used in Chapter 3 of this thesis and was developed by Holohil Systems Ltd. It
was deployed on an adult Razorbill at Machias Seal Island on 14 July 2007, monitored
south of Grand Manan on 4 Jan, 12 Jan and 29 Jan and later recovered in May 2008 at
Machias Seal Island. Presumably, the plastic band became so worn from being ground
over the rocks of the breeding colony that it eventually cracked. The epoxy covering of
the radio also exhibited extensive wear and the whip antenna had been broken.
The transmitter on the left was developed by Advanced Telemetry Systems Inc.
and was used as part of a pilot project in 2006. It was deployed for approximately 2
months during which the plastic band became almost completely worn through.
Figure 1: Two radio transmitters highlighting damage suffered after being deployed on
an adult Razorbill.
90
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising