Content uploaded by Karl Norling
Author content
All content in this area was uploaded by Karl Norling
Content may be subject to copyright.
Reçu le 16 février 2010 ; accepté après révision le 11 mai 2010.
Received 16 February 2009; accepted in revised form 11 May 2010.
Cah. Biol. Mar. (2010) 51 :
A live video observatory reveals temporal processes at a
shelf-depth whale-fall
Adrian G. GLOVER
1
, Nicholas D. HIGGS
1
, Philip M. BAGLEY
2
, Ralph CARLSSON
3
, Andrew J. DAVIES
4
,
Kirsty M. KEMP
5
, Kim S. LAST
6
, Karl NORLING
3,7
, Rutger ROSENBERG
3
, Karl-Anders WALLIN
3
,
Björn KÄLLSTRÖM
8
and Thomas G. DAHLGREN
9,10
*
(1)
Zoology Department, The Natural History Museum, Cromwell Rd., London SW7 5BD, UK
(2)
OceanLab, Unversity of Aberdeen, Newburgh, Aberdeenshire, UK
(3)
Sven Lovén Centre for Marine Sciences, Kristineberg, University of Gothenburg, SE 450 34 Fiskebäckskil, Sweden
(4)
School of Ocean Sciences, University of Wales, Bangor LL59 5AB, UK
(5)
Institute of Zoology, Zoological Society of London, London NW1 4RY, UK
(6)
Scottish Association for Marine Science, Scottish Marine Institute, Oban, Argyll PA37 1QA, UK
(7)
Norwegian Institute for Water Research, Gaustadalléen 21, NO-0349 Oslo, Norway
(8)
The Maritime Museum and Aquarium, SE 414 59 Göteborg, Sweden
(9)
Department of Zoology, University of Gothenburg, Box 463, SE 405 30 Göteborg, Sweden
(10)
Uni Environment, Postboks 7810, N-5020 Bergen, Norway
* Corresponding author: thomas.dahlgren@uni.no
Abstract: There have been very few studies of temporal processes at chemosynthetic ecosystems, even at relatively more
accessible shallow water sites. Here we report the development and deployment of a simple cabled video observatory at ≈
30 m water depth in Gullmarsfjorden, Sweden. The camera provides a live video feed to the internet of faunal activity at
the experiments, which to date have included 5 separate whale-fall deployments. Our data suggest that the time to
decomposition of small cetacean carcasses at shelf-depth settings is considerably slower than at deep-sea sites. We have
also provided a new methodology for the deployment of low-cost live video observatories at up to 30 m water depth, which
can be used both for research and outreach activities.
Keywords: Skeletonization
l Taphonomy l Porpoise l Carcass l Scavengers l Bacterial mat l Corps
2 LIVE VIDEO OBSERVATORY OF A SHELF-DEPTH WHALE-FALL
Introduction
The study of temporal processes at chemosynthetic
ecosystems is severely hampered by a lack of sampling
resolution. In the majority of studies, inferences have to be
made from sampling that is rarely less than several months
apart, and often several years. Observations made during
submersible or ROV operations, and from repeat visits,
suggest that chemosynthetic ecosystems such as hydro -
thermal vents are highly dynamic at a range of temporal
scales (e.g. Sarrazin et al., 1997; Shank et al., 1998). A
crucial tool to better understanding temporal change is the
development of cabled observatory systems at these
ephemeral habitats. Several large deep-water projects are
currently being constructed (e.g. Neptune Canada, 2009)
and new methods to use video data are explored (e.g.
Aguzzi et al., 2009). Here we show that small-scale inshore
observatories installed at diver-accessible depths can also
teach us much about the development of chemosynthetic
ecosystems as well as illustrate the power of continuous
observational marine data for both science and outreach.
When whales die, they sink to the seafloor and become
food for scavenging animals (Smith & Baco, 2003).
Microbial decomposition can also lead to the formation of
a chemosynthetic ecosystem at these sites with similarities
to both hydrothermal vents and seeps (Smith et al., 1989).
At shelf-depth whale-falls, rather little is known about the
feeding behavior of the scavenging communities, the
development of the chemosynthetic bacterial mats, and the
influence of external drivers such as tides, currents and
photoperiod. Studies at shallow-water hydrothermal vents
(Dando et al., 1995) and seeps (Sahling et al., 2003) suggest
that whilst there are few (or no) specialist macrofaunal
animals at these sites, sulphide-oxidising bacterial mats are
abundant. The aims of our pilot project is to test the
hypotheses that 1) shelf-depth whale-falls are consumed by
a typical shallow-water scavenging fauna over a period of
several weeks and 2) bacterial mats form over the carcass
and remain until the bones are consumed or dispersed by
currents. These observational data are also useful in
understanding the taphonomic processes that occur at shelf-
depths, which should assist paleontological studies of
whale-falls (e.g. Dominici et al., 2009), and macro -
evolutionary studies of whales themselves (Gatesy &
O’Leary, 2001).
Materials and methods
Observatory site and design
The Kristineberg cabled video observatory is located on the
west coast of Sweden at a site centered on 58°15.31’N-
11°26.95’E which is approximately 1 km from the Sven
Lovén Centre for Marine Sciences (Kristineberg), a marine
laboratory operated by the University of Gothenburg
(Figure 1). The experiments have been run in depths from
6-30m at this site. The observatory in its current form
consists of a stainless steel tubular frame with basal
dimensions of approximately 1 x 1 m. Attached to the frame
looking down vertically is a SubSea anodized aluminum
camera housing (1000 m rated) containing an Axis 211
network camera connected to the shore using 100m of
ethernet cable (S-F/UTP 4P CAT6 LSZH+PE). The ethernet
cable is threaded into the housing using a SubConn wet
connector. Power for the camera is provided down the
ethernet cable using PoE (power over ethernet). A separate
cable with 30 volts provides power to a ROS LED
SmartLIGHT II attached to the side of the frame pointing
obliquely at the experiment and providing continuous light
for real time video as well as time laps recordings. The two
cables running from the observatory are threaded into a
protective plastic hose, which runs over the seafloor
approximately 100m to an island where they are connected
into the Kristineberg Observatory Node (KON) which
constitutes an insulated and heated wooden hut which is
operational in all weathers. The KON is connected with 400
V AC power and 24-fibre optic cable which is run
underwater to the main laboratory and fed into the local
network.
The video stream from the Axis 211 is sent to a server
and encoded into a windows media stream and mirrored at
a streaming server at the University of Gothenburg for
public access and education/outreach projects (http://uw-
observatory.loven.gu.se). Authenticated client access can
Figure 1. (a) map with arrow indicating observatory location on the Swedish west coast (Skagerrak), (b) outline of observatory: 1 -
Axis 211 IP camera, 2 - LED light, 3 - carcass, 4 - power and PoE cable, 5 - Kristineberg Observatory Node (KON, server and power),
6 - fibre cable to shore, (c) one of the first frames from experiment #1 showing feeding on the skin by Hyas araneus, (d) the website
(accessible by scientists) for archiving stills and footage, organised by time/date.
Figure 1. (a) carte indiquant la position de l’observatoire sur la côte suédoise (Skagerrak), (b) schéma de l’observatoire : 1 - Caméra
vidéo, 2 - éclairage (Led), 3 - carcasse, 4 - alimentation, 5 - relais de l’observatoire, 6 - fibre optique, (c) une des premières images de
l’expérimentation 1 montrant Hyas araneus se nourissant sur la peau, (d) le site web (accessible aux scientifiques) d’archivage des
données, gérées par heure et date.
A.G. GLOVER, N.D. HIGGS, P.M. BAGLEY et al. 3
Table 1. Whale-fall experiments at the Kristineberg Observatory located at 58°15.31’N-11°26.95’E.
Tableau 1. Expérimentations sur des carcasses de baleines à l’observatoire de Kristineberg, localisé à 58°15.31’N-11°26.95’E.
Exp. Carcass / bait Depth Camera Study period Taphonomy Notes
system
1 P. phocoena 30 m Oblique 7/7/08 - 29/7/09 Mobile scavengers remove Camera flooded
(intact, 30 kg) (23 d) ~90% skin and ~10% flesh after 23 d
after 23d
2 P. phocoena 30 m Oblique 5/11/08 - 12/12/08 Mobile scavengers remove Camera flooded
(intact, 30 kg) (38 d) ~90% skin and ~30-60% after 38 d
flesh after 38 d
3 P. phocoena 30 m Oblique 5/2/09 - 22/3/09 Mobile scavengers remove Barnacle fouling after 24 d
(intact, 30 kg) (46 d) ~40% skin and ~20-40% Camera flooded after 46 d
flesh after 46d.
4 P. phocoena 23 m Vertical, 19/5/09 - 11/8/09 Mobile scavengers remove Fouling cleaned by
(intact, 20 kg) IP camera (85 d) 100% skin and 100% flesh SCUBA once per month
after 85d
5 B. acutorostrata 6 m Vertical, 3/9/09 - ongoing Bones intact after 90d; Fouling cleaned by
(fin bones) IP camera bacterial mat present SCUBA once per month
also be made directly to the camera by scientists to adjust
settings as well as view a higher-bandwidth MJPEG stream.
An additional server records a still image every 10 minutes
onto an archive. This archive is accessible by scientists
wishing to collect data at any time. The current observatory
design is as described above, earlier versions of the
observatory differed slightly in the model of the camera,
housing and position of the camera (oblique rather than
vertical view) (Table 1).
Experiments
Including the experiment ongoing at the time of writing
there have been 5 whale-fall experiments at the observatory
(Table 1). The first experiment (#1) was put down after
installation of the observatory in July 2008 and consisted of
a ≈ 30kg harbour porpoise (Phocoena phocoena).
Experiments #2-3 were exactly the same as the first,
experiment 4 was a smaller porpoise (20 kg) and
experiment 5 consisted of the fin bone of a minke whale
(Balaeonoptera acutorostrata). All experimental material
was collected from dead cetacean strandings. Experiments
#4 and #5 were studied using a camera mounted vertically
rather than obliquely. The first 3 experiments were
terminated after failure of the camera at 23 days, 38 days
and 46 days respectively. Experiment #4 lasted 85 days
after which the carcass was completely gone. Experiment
#5 is ongoing and this camera system (Axis 211 inside a
SubSea housing) has provided over 12 months of
continuous data without failure.
Data collection and analysis
During the experiments, live observations of the video
stream could be made by any of the scientists connected to
the website from anywhere in the world, including using
portable devices that can play MJPEG (e.g. Apple iPhone).
Species observed on the carcass actively feeding or
associated with the remains were identified to the lowest
taxonomic level (Figure 2) both from live video and time
laps video capture. The first attempt at analysis consisted of
collating images over a 24 hr period (1 image every 4
hours) from the archive every 3 days, to examine both daily
and weekly-scale events. This was used on all experiments
to provide preliminary observational data on changes at the
whale-falls. Secondly, for experiment 4 (the longest-last-
ing) a time-lapse movie was made using 1679 images from
the archive, each image being 1 hr apart. This created a
movie 1.08 mins long played at a speed of 25fps
(www.youtube.com/theuwobservatory). Finally, as an
exemplar for a future, larger study we analysed in detail the
images from experiment #4 every 4 hours over an 85 day
period (337 images) recording the number of mobile
scavengers, quantity of bacterial mat (1-5 scores) and %
cover of skin, flesh and bone. Significant events were
recorded in an event log.
Results
Technical issues
Corrosion of the camera housing (ROS CE-X36), pan and
tilt (ROS PT-10-FB) and control box caused a failure of the
observatory in experiments #1, #2 and #3 after 23 days, 38
days and 46 days respectively. The corrosion problem was
cured by experiments #4 and #5 which used a more robust
housing and sacrificial anodes on the steel frame. The
second major problem was fouling, in particular barnacles
on experiment 3 during spring. These were first observed as
cyprid larvae attaching to the lens, then metamorphosing
into adults over a period of ~20 days in March 2009.
Regular cleaning of the lens (once per month) by divers has
prevented any serious fouling problem in experiments #4
and #5.
Mobile scavenger community
The dominant mobile scavengers found feeding on the
carcasses at 23m and 30m depth (exp1-4) were spider crabs
Hyas araneus (Linnaeus, 1758), edible crabs Cancer
pagurus (Linnaeus, 1758), hermit crab Pagurus bernhardus
Linnaeus, 1758, seastars Marthasterias glacialis Linnaeus,
1758 and Asterias rubens Linnaeus, 1758. Associated with
but not obviously feeding on the carcasses were the scallop
Pecten maximus Linnaeus, 1758 and fish species including
Atlantic cod Gadus morhua Linnaeus, 1758, horse
mackerel Trachurus trachurus Linnaeus, 1758, goldsinny
Ctenolabrus rupestris Linnaeus, 1758, and sea scorpion
Myoxocephalus scorpius Linnaeus, 1758. Harbour seals,
Phoca vitulina Linnaeus, 1758 were observed on several
occasions exploring the observatory, camera and
experiment area (but not feeding). At 6m depth the
scavenging community of was dominated by A. rubens, P.
bernhardus, whelk Nassarius reticulatus (Linnaeus, 1758),
seastar Henricia sanguinolenta (O.F Müller, 1776) and
shore-crab Carcinus maenas Linnaeus, 1758. Fish observed
at this shallower experiment include eelpout Zoarces
viviparus (Linnaeus, 1758), wrasse Labrus mixtus
Linnaeus, 1758, pipefish Syngnathus acus Linnaeus, 1758
and flounder Platichthys flesus (Linnaeus, 1758). In total,
18 species were identified associated with the whale-falls
from the video observatory.
Temporal trends
Decomposition of the carcass was studied in detail for
experiment #4 over a ~3 month period (Table 1 & Fig. 2).
4 LIVE VIDEO OBSERVATORY OF A SHELF-DEPTH WHALE-FALL
During month 1, the skin was slowly removed by the steady
action of a small number of mobile scavengers, with
bacterial mat recorded forming on the skin surface after
about 12 days. During month 2, the rate of skin removal,
flesh exposure and bone exposure increased rapidly, with
almost all the skin removed at the beginning of month 3. In
the final month, the flesh was eaten and the bones exposed,
separated and dispersed by currents as the carcass broke up.
Bacterial mat was mainly recorded on the surface of the
skin, and in experiment #5 we noted it forming on bones
after a few days exposure. Preliminary data from
experiments 1-3 is consistent with this pattern although
these experiments did not last long enough for the flesh to
be removed. A nocturnal cycle in fish abundance was
clearly detected around all experiments, although the fish
were not in general observed feeding but presumed to be
attracted by the light, with some exceptions (e.g. Fig. 2). At
experiment #2, the carcass was observed to float (held in
place approx 50cm above the seafloor by the line to the
ballast) for approximately 2 days after implantation, before
sinking back to the seafloor, still largely intact.
Discussion
In general, the carcasses at all deployed experiments were
consumed by generalist mobile scavengers already well-
known from the study areas. There was a notable absence
of scavengers (lysianassid amphipods, hagfish Myxine
glutinosa Linnaeus, 1758 and unidentified sharks) that have
been recorded at deeper-water experiments (125 m) in near-
by Kosterfjord, Sweden (Glover et al., 2005; Dahlgren et
al., 2006). The fish present at the carcasses was not
observed eating to any extent that could significantly
contribute to the decomposition. The planctivore fish
A.G. GLOVER, N.D. HIGGS, P.M. BAGLEY et al. 5
Figure 2. Temporal processes at a 30 m whale-fall at the Kristineberg Observatory. Video frames show the state of the carcass after
0, 16, 33, 52, 67, and 83 days, arrows highlighting observed feeding events by scavengers. Graphs from top, percentage of bone/flesh/skin
remaining on the carcass over the 85 day period, abundance of mobile scavengers in the video frame, quantity (1-5 score) of bacterial
mat.
Figure 2. Evolution temporelle sur une carcasse de baleine par 30 m de fond à l’observatoire de Kristineberg. Les images vidéo
montrent l’état de la carcasse après 0, 16, 33, 52, 67 et 83 jours, les flèches montrent les épisodes de nutrition par des charognards.
Schémas de haut en bas, pourcentage d’os, de chair et de peau restant sur la carcasse après 85 jours, abondance des charognards sur la
vidéo, quantité (échelle arbitraire de 1 à 5) du biofilm bactérien.
species frequently observed was not included in the species
list Previous studies of porpoise decomposition at Atlantic
abyssal (Jones et al., 1998) and bathyal (Kemp et al., 2006)
depths also show strikingly different scavenging
communities, with a much more dominant scavenging role
being undertaken by fish (e.g. grenadier Coryphaenoides
armatus Hector, 1875) compared to dominance by
crustaceans (e.g. spider crabs Hyas araneus) at our shallow
sub-littoral sites.
In addition to the differences in species composition, our
data suggest a significantly slower rate of decomposition
(in terms of time taken to skeletonize) in shallow shelf
environments compared to deep-water. At bathyal and
abyssal Atlantic depths, carcasses were skeletonized within
5-10 days (Jones et al., 1998; Kemp et al., 2006) compared
to almost 2 months at our observatory sites. Our data
suggest that the slow removal of the skin by scavengers and
development of possibly toxic bacterial mat prevents rapid
consumption of the flesh in this initial period, lasting about
1 month. These data are supported by previous observations
of an intact, bacterial-mat coated, pilot whale carcass at
30m depth in Kosterfjord (Dahlgren et al., 2006), but more
research is needed to further understand the processes
behind the slow removal of cetacean skin at shallow depths.
Once the skin has been removed and the flesh exposed,
removal rates of the carcass increase dramatically over the
second month and for that period are closer to those
recorded in the deep sea (Figure 2). Final skeletonization
and dispersal of the bones took over 3 months, with
apparently limited opportunity for colonization by bone-
eating worms (Osedax mucofloris) as the remaining bones
were small and rapidly dispersed by currents, as previously
hypothesized (Glover et al., 2008).
Other studies of shallow-water carcass scavenging are
surprisingly rare, but include forensic studies of pig
remains in Canada (Anderson & Hobischak, 2004) and a
study with fish-bait in Antarctica (Smale et al., 2007). The
pig experiments showed similarly quite slow rates of flesh
consumption, while the Antarctic study showed large
aggregations of nemertean worms (Parbolasia corrugatus
McIntosh, 1876) or lysianassid amphipods with few other
scavengers present. At the pig experiment, it was reported
that the carcasses were buoyant for up to 28 hours after
being deposited on the seafloor, after which they became
negative again and decomposition accelerated (Anderson &
Hobischak, 2004). We observed a similar pattern at
experiment #2, where the carcass was buoyant for approxi-
mately 2 days, before sinking again, still intact. This is not
unexpected since previous studies suggest that marine
mammal carcasses initially float under certain oceano-
graphic conditions, and depending on the nutritional status
of the animal when it died (Schäfer, 1972). However, our
limited data suggest that intact negatively-buoyant
carcasses are likely to be present on the seafloor at 30m
shelf depths.
In terms of a chemosynthetic community, the only
observations were of a thin, flocculent and fragile layer of
bacterial mat forming on the skin after about 10 days, and
on exposed bones after 2-3 days, as already reported for a
30m depth pilot whale in Sweden (Dahlgren et al., 2006).
Some data collected from experiment #5 suggest that the
bacterial mat return at a later stage covering these bones.
This has also been documented at carcasses from large-
bodied whale species in the deep sea (Smith & Baco, 2003)
as well as at a 125m carcass in Swedish waters (Glover et
al., 2005). It remains to be seen if these very shallow
bacterial mats are home to new species of mat-eating
worms, as has been recorded at deeper whale-falls
(Wiklund et al., 2009).
Our study has provided the first detailed time-series for
whale-fall succession in a shelf-depth setting. Our data
suggests much slower (months rather than days) rates of
scavenging and decomposition in shallow-waters compared
to bathyal and abyssal depths. As at shallow vents and
seeps, there is apparently an absence of specialist whale-
fall species at depths of 30m or less, although a single
occurrence of Osedax mucofloris has been recorded at 30m
in Kosterfjord (Dahlgren et al., 2006). We have also
provided a new technological method to study experiments
in real-time in the marine environment, helping to
overcome the substantial pedagogic challenge of
understanding processes in the sea.
Acknowledgements
We are grateful to Sparbanksstiftelsen Väst for assistance in
funding this project. We thank the crew at R/V Arne
Tiselius and R/V Oscar von Sydow at the Sven Lovén
Centre for Marine Sciences, Kristineberg for assistance
during deployment of cables and camera systems. We are
also grateful to Jens Bjelvenmark, Pia Norling and Erik
Selander for help during dive operations. NDH is funded by
the Natural Environment Research Council, UK and TGD
is funded by the Swedish Research Council.
References
Aguzzi J., Costa C., Fujiwara Y., Iwase R., Ramirez-Llorda E.
& Menesatti P. 2009. A novel morphometry-based protocol of
automated video-image analysis for species recognition and
activity rhythms monitoring in deep-sea fauna. Sensors, 9:
8438-8455.
Anderson G.S. & Hobischak N.R. 2004. Decomposition of
carrion in the marine environment in British Columbia,
Canada. International Journal of Legal Medicine, 118: 206-
209.
6 LIVE VIDEO OBSERVATORY OF A SHELF-DEPTH WHALE-FALL
Dahlgren T.G., Wiklund H., Kallstrom B., Lundalv T., Smith
C.R. & Glover A.G. 2006. A shallow-water whale-fall
experiment in the north Atlantic. Cahiers de Biologie Marine,
47: 385-389.
Dando P.R., Hughes J.A. & Thiermann F. 1995. Preliminary
observations on biological communities at shallow hydro -
thermal vents in the Aegean Sea. Geological Society, London,
Special Publications, 87: 303-317.
Dominici S., Cioppi E., Danise S., Betocchi U., Gallai G.,
Tangocci F., Valleri G. & Monechi S. 2009. Mediterranean
fossil whale falls and the adaptation of mollusks to extreme
habitats. Geology, 37: 815-818.
Gatesy J. & O’Leary M.A. 2001. Deciphering whale origins
with molecules and fossils. Trends in Ecology & Evolution, 16:
562-570.
Glover A.G., Källström B., Smith C.R. & Dahlgren T.G. 2005.
World-wide whale worms? A new species of Osedax from the
shallow north Atlantic. Proceedings of the Royal Society B,
272: 2587-2592.
Glover A.G., Kemp K.M., Smith C.R. & Dahlgren T.G. 2008.
On the role of bone-eating worms in the degradation of marine
vertebrate remains. Proceedings of the Royal Society B, 275:
1959-1961.
Jones E.G., Collins M.A., Bagley P.M., Addison S. & Priede
I.G. 1998. The fate of cetacean carcasses in the deep sea:
observations on consumption rates and succession of
scavenging species in the abyssal north-east Atlantic Ocean.
Philosophical Transactions of the Royal Society of London B,
265: 1119-1127.
Kemp K.M., Jamieson A.J., Bagley P.M., McGrath H., Bailey
D.M., Collins M.A. & Priede I. G. 2006. Consumption of
large bathyal food fall, a six month study in the NE Atlantic.
Marine Ecology-Progress Series, 310: 65-76.
Neptune 2009. Nepture Canada: transforming ocean science.
Sahling H., Galkin S.V., Salyuk A., Greinert J., Foerstel H.,
Piepenburg D. & Suess E. 2003. Depth-related structure and
ecological significance of cold-seep communities - a case
study from the Sea of Okhotsk. Deep-Sea Research Part I-
Oceanographic Research Papers, 50: 1391-1409.
Sarrazin J., Robigou V., Juniper S.K. & Delaney J.R. 1997.
Biological and geological dynamics over four years on a high-
temperature sulfide structure at the Juan de Fuca Ridge
hydrothermal observatory. Marine Ecology-Progress Series
153: 5-24.
Shank T.M., Fornari D.J., Von Damm K.L., Lilley M.D.,
Haymon R.M. & Lutz R.A. 1998. Temporal and spatial
patterns of biological community development at nascent
deep-sea hydrothermal vents (9°50’ N, East Pacific Rise).
Deep-Sea Research Part II-Topical Studies in Oceanography
45: 465-515
Smith C.R. & Baco A.R. 2003. Ecology of whale falls at the
deep-sea floor. Oceanography and Marine Biology-An annual
review, 41: 311-354.
Smith C.R., Kukert H., Wheatcroft R.A., Jumars P.A. &
Deming J.W. 1989. Vent fauna on whale remains. Nature, 341:
27-28.
Wiklund H., Glover A. G. & Dahlgren T. G. 2009. Three new
species of Ophryotrocha (Annelida: Dorvilleidae) from a
whale-fall in the North-East Atlantic. Zootaxa, 2228: 43-56.
A.G. GLOVER, N.D. HIGGS, P.M. BAGLEY et al. 7
