1
2009 Annual Meeting of the
UK Student Chapter of the
Society for Marine
Mammalogy Meeting,
St Andrews
Abstracts
Time
Event
Wed 4th FEB
19:00 – 23:00
Icebreaker, Sailing club
Thu 5th FEB
10:00 - 10:45
10:45 - 11:00
Opening welcome talk - Ian Boyd
Foraging specialization influences movement and site fidelity in a top marine
predator, the killer whale - Andy Foote
Tea / coffee - Poster session
- A fully automated whistle classifier - Marjolaine Caillat
- Intrinsic and extrinsic forcing in life histories: patterns of growth and stable
isotopes in male Antarctic fur seal teeth – Nora Hanson
- Passive Acoustic Monitoring of beaked whales using towed arrays – Jeppe
Balle
- Isolation and Development of Corneal Cell Cultures from Marine Mammals –
Vittoria Elliott
Lunch
Workshop “Paper writing” – Dr Patrick Miller
11:00 - 11:30
11:30 - 12:30
12:30 - 14:00
14:00 - 16:00
Fri 6th FEB
09:30 - 11:00
11:00 - 11:30
11:30 - 12:30
15:00 - 15:30
Conservation Genetics of Marine Mammals – Prof. Rus Hoelzel
Tea / coffee - Poster session
- Individual behavioural differences and site fidelity of postpartum female grey
seals on two island breeding colonies - Ross Culloch
- A Method of Identifying Signature Whistle Types from Field Recordings of
Wild, Unrestrained Bottlenose Dolphins (Tursiops truncatus) – Stephanie
King
- An assessment of the foraging behaviour of bottlenose dolphins interacting
with a bottom trawl fishery - Alice Mackay
Lunch
Photo-ID for beginners: from mug shots to abundance estimates – Dr Ben
Wilson
Tea / coffee - Poster session
15:30 - 17:00
Workshop “Field techniques - Photo ID” – Dr Ben Wilson
12:30 - 14:00
14:00 - 15:00
Sat 7th FEB
09:30 - 10:30
Working title: “Why seals are better than dolphins (official title TBD)”- Dr
Callan Duck
Individual discrimination in sperm whale codas – Ricardo Antunes
Tea / coffee - Poster session
- Spatial modelling of minke whale foraging habitats in the St. Lawrence
Estuary, Canada: a maximum entropy approach – Mike Tetley
- Designing a shipboard line transect survey to estimate cetacean abundance
off the Azores Archipelago, Portugal - Cláudia Faustino
- Searching for a deeper meaning: Maximum dive depth in pinnipeds - Theoni
Photopolous
- Fine scale population differentiation in selected vs. neutral markers in
common dolphins off the coast of Portugal - André Moura
Lunch
Discussion groups
Closing words, choosing next years venue, and new volunteers for the
10:30 - 10:45
10:45 - 11:15
11:15 - 12:15
12:15 - 13:30
13:30 - 15:00
15:00 - 16:00
2
St Andrews – 5 to 7 February 2009
th
th
committee
Foraging specialization influences movement and site fidelity in a top
marine predator, the killer whale
Andrew D. Foote1, Tiu Similä2, Gísli A. Víkingsson3, Peter T. Stevick4
1 Institute of Biological and Environmental Sciences, University of Aberdeen, Lighthouse Field
Station, Cromarty, Ross-shire, IV11 8YJ, UK.
2 Wild Idea, Box 181, 8465 Straumsjøen, Norway.
3 Marine Research Institute, Program for Whale Research, PO Box 1390, 121 Reykjavík, Iceland.
4 Hebridean Whale and Dolphin Trust, 28 Main Street, Tobermory, Isle of Mull, PA75 6NU, UK.
Movement and site fidelity have important consequences for evolution of population
structure and therefore conservation and management of species. In this study
photographs of naturally marked killer whales collected from sites across the
Northeast Atlantic are used to estimate fidelity to sampling locations and movement
between locations, expressed as transition probabilities, pt, using maximum likelihood
methods. High transition probabilities suggest there is high inter-annual site fidelity to
all locations and movement between the spawning and wintering grounds of Atlantic
herring. We also found movement between the Northern Isles, Scotland and East
Iceland. An association network indicates that killer whales predating seals around the
Northern Isles, Scotland are linked to the community of killer whales that follow the
Icelandic summer-spawning herring, suggesting seasonal prey switching by some
individuals.
3
St Andrews – 5 to 7 February 2009
th
th
A fully automated whistle classifier
M. Caillat2, P. White1, D. Gillespie², J. Gordon², M. Lonergan²
1 School of Engineering Sciences, University of Southampton;
2 SMRU, University of St Andrews
It is now established that using passive monitoring acoustic to detect marine mammals
is a very useful tool which complements the use of visual monitoring. The software
detector PAMGUARD has been developed and enhanced over the last 5 years as an
aid for conducting passive acoustic monitoring. It aims to provide an easy to use
flexible platform; allowing the user to configure the software for the monitoring
system being used and to create a display suitable for their particular needs.
The automated identification of the species detected is a facility which can be of
importance in some acoustic monitoring applications. Currently, for most species,
visual observation is the only real-time method of species confirmation; although
anecdotal evidence suggests that experienced operators can also successfully identify
species from whistles. Several studies have considered the creation of an automated
whistle classifier but to-date all published schemes require the manual extraction of
whistles contours and so are neither real-time nor fully automated. The need for
manual whistle extraction is a consequence of the fact that whistle contour extraction
schemes commonly fail, for example by generating whistle contours with artificial
breaks. Computer algorithms, such as those used in image processing, only realise
moderate success at repairing such breaks in the whistle contours. Manual methods
remain the only reliable technique to extract full whistle contours. The whistle
classification methods previously proposed use the features obtained from complete
whistle contours for species identification. These semi-automatic methods are able to
achieve quite good classification performance which demonstrates that species
identification from whistles is indeed feasible. This paper describes a fully automated
species classification tool, based on different philosophy for whistle classification.
The PAMGUARD whistle detector provides a fully automated, but imperfect, system
for the extracting whistles and provides the basis of this work. The PAMGUARD
whistle extraction scheme, is subject to the same limitations as previous methods; it
generally fails to extract complete whistles. However the method used here does not
rely upon the extraction of complete whistles, but processes only whistle fragments.
The classification is based on the statistical distribution of the features extracted from
many fragments, as opposed to the features extracted from single whistles.
The study uses data from 15 different species of small to medium sized odontocetes,
grouped within a single geographic area. The database is imbalanced since there are
some species for which there is much data and other specifies for which the data is
very much more limited. Where possible the algorithm is trained on a subset of
recordings and tested on a set of different recordings. The error rates of whistle
classification vary from species to species (partly because of the data imbalance) with
some error rates as low as 10%. The system provides an output in the form of
probabilities of the current whistles belonging to each of the species, with these
probabilities being updated in real-time.
4
St Andrews – 5 to 7 February 2009
th
th
Intrinsic and extrinsic forcing in life histories: patterns of growth
and stable isotopes in male Antarctic fur seal teeth
Nora N. Hanson1, Michael I. Bird2, Christopher M. Wurster2, Keith Reid3 & Ian L.
Boyd1,
1 Sea Mammal Research Unit, University of St. Andrews, St. Andrews, Fife, KY16 8LB, Scotland,
UK;
2 School of Geography & Geosciences, University of St. Andrews, St. Andrews, Fife, KY16 9AL,
Scotland, UK;
3 British Antarctic Survey, Natural Environment Research Council, High Cross, Madingley Road,
Cambridge CB3 OET, UK (Current address: CCAMLR, P.O. Box 213, North Hobart, Tasmania 7002,
Australia)
Life-time records of the trophic sources of carbon and nitrogen and of growth can be
generated from biogenic structures that show accretionary growth - such as tree-rings,
mollusc shells and the teeth of some animals. These can be combined among
individuals to provide long-term estimates of the environmental conditions
encountered by the population. Both intrinsic (i.e. ontogenetic) and extrinsic (i.e.
environmental) factors are important in modulating variation in growth and the
apparent dietary sources of C and N. We used the canine teeth of a large marine
predator, the male Antarctic fur seal (Arctocephalus gazella) from South Georgia, to
investigate both sources of variation. Substantial ontogenetic shifts occurred in both
δ13C and δ15N values, indicating a change in the trophic sources of C and N as the
animals age. Over the 40-year period from 1964 to 2005, and after statistical
reduction of ontogenetic variation, we also detected long-term declines in δ13C and
δ15N values and in annular tooth growth, indicating that the population has become
more dependent on energy from a lower-trophic level. The time series of δ13C values
was also inversely correlated with sea surface temperatures in the region although any
causal relationship remains elusive. Our analyses suggest that both intrinsic and
extrinsic sources of variation, and their interaction, must be considered from such
time series data; failure to do so could results in a biased analysis and spurious results.
5
St Andrews – 5 to 7 February 2009
th
th
Passive Acoustic Monitoring of beaked whales using towed arrays
Jeppe Dalgaard Balle1
1 Sea Mammal Research Unit, University of St Andrews, St Andrews, KY16 8LB, UK
Beaked whales remain amongst the most poorly studied mammals. Mass strandings
linked to military sonar exercises have increased the need to better understand this
group and to devise improved mitigation procedures.
Beaked whales are extremely difficult to sight at sea, which hampers attempts to study
them, and makes operational mitigation difficult. Passive acoustic monitoring could
improve detection efficiency. Continuous recordings were made at a sampling rate of
192 kHz from 2 or 4 element towed hydrophone arrays during joint visual/acoustic
surveys in the Bahamas, Canaries and Azores, and in conjunction with monitoring of
bottom-mounted hydrophones at the AUTEC range and shore based visual tracking in
the Canaries and Azores. A beaked whale click detector and classifier was developed
within Rainbow Click and PAMGUARD. This was run in real time and on
recordings to detect beaked whale click trains. Two species of beaked whale were
encountered visually and detected acoustically: Mesoplodon densirostris and Ziphius
cavirostris. Acoustic detections correlated well with visual sightings. Target motion
analysis of bearings to sequences of clicks suggests a maximum detection range of
approximately three kilometers and preliminary results indicate that clicks can be
identified to species.
Fieldwork in the Canaries and Azores through the spring and summer of 2008 has
focused on determining the effect of distance and hydrophone depth on the probability
of detection. These early results suggest that passive acoustic monitoring could play
an important role in improving the detection of these animals and thus facilitate
population surveys, photo-id studies and, potentially, real time detection for
mitigation.
Further work using towed arrays in the Alboran Sea is scheduled to begin in 2009.
Analysis of off-shore CODA data has been initiated, but is still on a very preliminary
stage.
6
St Andrews – 5 to 7 February 2009
th
th
Isolation and Development of Corneal Cell Cultures from Marine
Mammals
Vittoria Elliott1
1 Biological and Biomedical Sciences, University of Durham, Durham DH13LE
Free swimming marine mammals are inherently difficult to capture for nondestructive sampling. Although biopsy darting has been used for the application of
genetic methods, cell culture generally requires sufficiently active and sterile cells to
promote proliferation. Whilst unfortunate, marine mammal stranding can be a source
of material for the establishment of cell cultures. In order to capitalize on these
opportunities it is necessary to overcome certain fundamental difficulties associated
with taking tissue from a recovered cadaver.
Eyes represent a sealed element protected from environmental contaminants in which
the integrity of the cells are preserved en route to the strand location. The corneal
limbus is also a source of stem cells which can increase the in vitro proliferative
response of epithelial cells. Techniques were developed and adapted to develop
intermediate term epithelial and fibroblast cultures from corneal cells.
7
St Andrews – 5 to 7 February 2009
th
th
Individual behavioural differences and site fidelity of postpartum
female grey seals on two island breeding colonies
R.M. Culloch1,2, P.P. Pomeroy2 and S.D. Twiss1
1
2
Department of Biological Science Laboratories, University of Durham, South Road, Durham DH1 3LE.
Sea Mammal Research Unit, Gatty Marine Laboratory, University of St. Andrews, Fife KY16 8LB
It is now recognised that non-human animals have personalities; where personality is
defined as individual behavioural differences that are consistent over time and/or
contexts. Such behavioural differences are likely to affect the way an individual
interacts with its environment (including conspecifics), potentially resulting in fitness
consequences. In order to investigate the ecological significance of personality in a
long-lived wild animal, the grey seal, behavioural observations of postpartum females
were carried out on two contrasting island breeding colonies during 2007 and 2008;
(1) North Rona, which is an undisturbed site where breeding areas are unaffected by
the tide, and (2) Brownsman, Farne Islands, which is subject to regular human
disturbance, and is affected by the tide.
Cross-site analyses of data collected in 2007 identified ‘resting’, ‘alert’, ‘presenting’
and ‘nursing’ as the behaviours that explained the majority of variation between
individuals (66%). The females that spent most time ‘alert’ were found on Brownsman.
Preliminary analysis of site fidelity showed that approximately 33% and 66% of known
females returned and pupped on the study sites at North Rona and Brownsman,
respectively. The higher site fidelity on the more disturbed site is of particular interest,
especially given that the more ‘alert’ females were found on Brownsman. We discuss
whether levels of alertness reflect a simple response to disturbance, or constitute part of
a behavioural type that are more likely to raise a pup successfully on a ‘disturbed’
colony.
8
St Andrews – 5 to 7 February 2009
th
th
A Method of Identifying Signature Whistle Types from Field
Recordings of Wild, Unrestrained Bottlenose Dolphins (Tursiops
truncatus)
Stephanie L. King1, Vincent M. Janik1, Laela S. Sayigh2, Randall S. Wells3 &
Volker B. Deecke1
1
Sea Mammal Research Unit, School of Biology, University of St Andrews, St Andrews, Fife, KY16
8LB, U.K.
2
Biology Department, Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA.
3
Chicago Zoological Society, c/o Mote Marine Laboratory, Sarasota, FL 34236 USA.
Bottlenose dolphins use individually distinctive recognition signals, termed ‘signature
whistles’ to maintain group cohesion. Signature whistle types are therefore a
biologically relevant category within the bottlenose dolphin vocal repertoire. To date
they could only be identified from isolated animals, in cases were they were emitted
repeatedly by a restrained or free-ranging animal clearly separated from others. We
investigated the temporal delivery of whistles to see whether this parameter holds
information that can be used to identify signature whistles in wild, unrestrained
animals. Recordings from focal follows of bottlenose dolphins from Sarasota Bay,
Florida, USA were analysed using ARTwarp, a technique that combines an adaptive
resonance theory neural network with dynamic time warping, and which has been
shown to categorize signature whistles correctly (Deecke & Janik 2006). Signature
whistles were identified by comparing ARTwarp categories to the signatures gained
from focal animals during short capture-release sessions. We found that signature
whistles tended to be emitted in sequences and that this can be used to identify them.
Signature whistles were found to make up all ARTwarp categories in which at least
75% of the whistles comprising that category occurred within a 1-10 second time
frame. Non-signature whistles did not show this degree of co-occurrence in time but
were occasionally produced in rapid succession. This was, however, over a much
shorter (< 1 second) temporal scale which differed significantly from the delivery
pattern of signature whistles. Our method of identifying signature whistles using
sequence analysis successfully identified the signature whistles of 50% of all the
animals that emitted their own signatures more than twice during a follow. Using the
75% threshold, no non-signature whistles were incorrectly identified as signatures
whistles.
9
St Andrews – 5 to 7 February 2009
th
th
An assessment of the foraging behaviour of bottlenose dolphins
interacting with a bottom trawl fishery
Alice I. Mackay1 and Peter C. Stephenson2
1 Sea Mammal Research Unit, Gatty Marine Lab., University of St Andrews, KY16 8LB, UK
2 Western Australian Fisheries and Marine Research Laboratories, PO Box 20, North Beach, WA
6920, Australia.
Numerous accounts exist globally of foraging associations between bottlenose
dolphins (Tursiops sp.) and bottom trawl fisheries. Dolphins have been observed to
feed from the discards of such fishing operations as well as directly from the codend
of the net. However observations of foraging events (e.g. searching, pursuit and
capture) are limited, with most information on foraging behaviour restricted to
inferences made from the surface activity of dolphins in the vicinity of trawl nets.
Here we present data on the underwater interactions between bottlenose dolphin and a
bottom trawl fishery and the first quantification of individual foraging behaviour
inside such nets. Underwater video recordings were collected in 2005 and 2006 as
part of projects conducted by the Department of Fisheries Western Australia to assess
and mitigate dolphin interactions with the Pilbara Finfish Trawl Interim Managed
Fishery. The fishery is prosecuted in the Pilbara region of northwest Australia, in
water depths of 50 – 100m. Video cameras were deployed at 4 different positions
inside trawl nets, orientated either towards the vessel or the codend. Over 60 hours of
video footage were examined and results are presented on the presence and behaviour
of bottlenose dolphins relative an exclusion grid, as well as possible reasons for how
this bycatch mitigating device is reducing dolphin bycatch in this fishery. We discuss
these results in relation to future bycatch mitigation strategies in this fishery.
10
St Andrews – 5 to 7 February 2009
th
th
Individual discrimination in sperm whale codas
Antunes, R.1, Schulz, T.2, Gero, S.2, Whitehead, H.2, Rendell, L.1
1 Sea Mammal Research Unit, University of St. Andrews, St. Andrews, UK
2 Department of Biology, Dalhousie University, Halifax, Canada
Long term social aggregations of animals may lead to the development of behavioural
strategies that are based on individualized relationships within groups, potentially
leading to the development of mechanisms that allow recognition of individuals.
Female and immature sperm whales live in long term stable social units in tropical
and temperate waters. Some authors have suggested that within these units
menopausal females might have a special role assisting kin, caring for calves and
acting as repositories of information that enhances group fitness. Also within social
units individuals have preferred associations and avoidances. Differential interaction
among unit members allows for an individual discrimination system. Sperm whales
rely on the emission of stereotyped series of clicks termed codas for communication.
studies have demonstrated the sharing and production at similar rates of different coda
types between individuals which not support the idea of individual distinction by
differences in coda repertoire. However, variations within coda type could potentially
carry information on animals identity allowing both group and individual level
information to be transmitted. This hypothesis was tested using a dataset of codas
assigned to individual whales by their inter-pulse intervals from a social unit off
Dominica. From the initial 315 codas having between 3 and 10 clicks, classified into
15 types using k-means clustering, only those in three types had enough clicks to be
used in the analysis. Using discriminant analysis it was possible to discriminate
between individual whales with lower error rates than the ones obtained from a
randomly assigning identities to codas. The discrimination performance was not the
same for all coda types suggesting differential functionality between coda types.
These findings agree with the prediction that preferred associations among sperm
whale unit members, which would require the ability to discriminate among
individuals, can be transmitted by codas.
11
St Andrews – 5 to 7 February 2009
th
th
Spatial modelling of minke whale foraging habitats in the St.
Lawrence Estuary, Canada: a maximum entropy approach
Tetley M.J.1, Tscherter U.2 & Mitchelson-Jacob E.G.3
1 School of Ocean Sciences, Bangor University, Wales
2 Ocean Research and Education Society ORES, Canada)
3 Centre for Applied Marine Sciences, Bangor University, Wales
The St. Lawrence Estuary, Canada, is a dynamic and varied environment providing
an important foraging habitat for a number of cetacean species. Minke whales
(Balaenoptera acutorostrata) which occur in the confluence area between the main
Estuary and the Saguenay Fjord have been observed to utilise a range of
specialised foraging behaviours. The aim of this study was to determine whether the
underlying environment of the area was important in promoting the use and
distribution of the strategies observed. The study compared the distribution of whales
recorded on dedicated surveys (2002-2007) with a suite of co-occurring bathymetric
and remotely sensed oceanographic variables within a geographical information
system (GIS). Multivariate principal component analysis (PCA) was used to
determine the covariance of variables affecting three primary behavioural strategies
(Type I, Type II and Generalist). The loadings of components 1 and 2 indicated
differences in the environmental distribution of the strategies tested. Distributions
were modelled within a geographic space using a maximum entropy (MAXENT)
approach to determine areas of potential high habitat suitability within the wider
Estuary. JACKKNIFE analysis showed that bathymetric variables were more
influential in modelling the distribution of lunge feeders whilst the models of those
employing a herding strategy were greater influenced by oceanographic parameters.
The findings of the study are considered important with respects to developing
dynamic marine spatial planning strategies, particularly in intensive whale watching
areas such as the St. Lawrence Estuary.
12
St Andrews – 5 to 7 February 2009
th
th
Designing a shipboard line transect survey to estimate cetacean
abundance off the Azores Archipelago, Portugal
Cláudia Faustino1, Len Thomas2, Mónica A. Silva3 and Tiago A. Marques2
1 SMRU, University of St.Andrews, UK
2 CREEM , University of St.Andrews, UK
3 IMAR, Universidade dos Açores, Portugal
Management schemes dedicated to the conservation of wildlife populations rely on
the effective monitoring of population size, and this requires the accurate and precise
estimation of abundance. The accuracy and precision of estimates are determined to a
large extent by the survey design. Line transect surveys are commonly applied to
wildlife population assessments in which the primary purpose of a survey design is to
ensure that the critical distance sampling assumptions are met.
Little information is available regarding cetacean abundance in the Archipelago of the
Azores (Portugal). This study aims to design a line transect shipboard survey that
allows the collection of data required to provide abundance estimates for such species.
Several aspects must be taken into consideration when designing a survey to estimate
cetacean abundance. This is an iterative process, and there is a constant trade off
between the logistic constraints and the desired statistical robustness. Information on
this process is provided to aid policy makers and environmental managers, such as the
criteria used for the choices made when defining the elements of a survey design.
Three survey effort scenarios are provided to illustrate the range of possibilities
between statistical robustness and logistic/ management restrictions. A survey is
designed for the more economical scenario (L=5000Km), although the second
scenario is the one recommended to be implemented (L=17,600Km) given it provides
robust estimates of abundance (CV ≤0.2).
13
St Andrews – 5 to 7 February 2009
th
th
Searching for a deeper meaning: Maximum dive depth in pinnipeds
Theoni Photopoulos1
1 SMRU, University of St.Andrews, UK
Diving marine mammals occupy a three-dimensional environment therefore their
movement, space use and foraging ecology should ideally be treated as threedimensional problems. Choosing an appropriate response variable, which will
adequately represent an animal’s use of the water column, is challenging yet central to
understanding what governs this aspect of marine mammal distribution. Commonly
the behaviour of marine predators and their prey is not directly observable and we rely
on measuring proxies for information on behaviour and resource-availability. These
proxies are selected from the information made available via telemetry. Technological
developments in wildlife telemetry in the last two decades have made it possible to
record animal behaviour and synchronous in situ physical features of the marine
environment. In this ongoing analysis, maximum dive depth is used as an indicator of
space use in the water column by grey seals (Halichoerus grypus). The question
addressed in this talk considers analytical approaches to investigating the relationship
between maximum dive depth and environmental covariates recorded onboard
instrumented animals.
14
St Andrews – 5 to 7 February 2009
th
th
Fine scale population differentiation in selected vs. neutral markers
in common dolphins off the coast of Portugal
André Moura1
1 Biological and Biomedical Sciences, University of Durham, Durham DH13LE
Population differentiation and speciation without limitation of gene flow remain a
contentious issue in evolutionary biology. Even though some such cases are well
documented in the literature, there is still widespread belief that unless gene flow is
constrained by some external geographic barrier, population differentiation and
speciation cannot occur. In this respect, cetaceans are an ideal study model. Given
their widespread distributions and large dispersal abilities, under a strict geographical
model of differentiation they should exhibit large panmitic populations spread over
large distances, and little population substructure. However, it has been proposed that
different selective pressures on a regional and local scale should promote
differentiation in several traits, while allowing for neutral traits to spread across subpopulations. In fact, cetaceans have diverged from a single land ancestor into a large
number of species in a relatively short time, even though their ecology is incompatible
with the most accepted differentiation model. In the absence of geographic barriers,
such diversity might be better explained by the adaptation to different ecological
niches. In this project, we intend to further test this hypothesis by looking at the fine
scale population genetics of common dolphins in a heterogeneous environment of
neutral versus functional markers. Our preliminary results show that selection is
detectable in these populations, and that markers under selection show a strong
population differentiation as opposed to neutral markers. One such example refers to a
marker associated with a gene affecting coloration. We hypothesize that the
population structure revealed by this marker is related to coloration differences
observed in the field that might possess some type of selective advantage.
15
St Andrews – 5 to 7 February 2009
th
th