Content uploaded by Mads Peter Heide-Jørgensen
Author content
All content in this area was uploaded by Mads Peter Heide-Jørgensen
Content may be subject to copyright.
Rate of increase and current abundance of humpback whales in
West Greenland
M.P. HEIDE-JØRGENSEN*, K.L. LAIDRE*+, R.G. HANSEN*, M.L. BURT#, M. SIMON£, D.L. BORCHERS#J. HANSEN$, K. HARDING$,
M. RASMUSSEN&, R. DIETZBAND J. TEILMANNB
Contact e-mail: mhj@ghsdk.dk
ABSTRACT
Aerial line transect surveys of the density of humpback whales (Megaptera novaeangliae) conducted off West Greenland eight times between 1984
and 2007 were used to estimate the rate of increase on the summer feeding ground. Only surveys in 1993, 2005 and 2007 had enough sightings to
construct independent density estimates, whereas the surveys in 1984–85 and 1987–89 had to be merged and treated as two surveys. The annual
rate of increase was 9.4% yr–1 (SE = 0.01) between 1984 and 2007. This rate of increase is higher than the increase estimated at the breeding grounds
in the West Indies, but is of the same magnitude as the observed rate of increase at other feeding grounds in the North Atlantic. A matrix model
based on observed life history parameters revealed that the theoretical growth rate of a humpback whale population ranged between 1 and 11%.
This confirms that the observed growth in West Greenland is within the plausible values. The survey in 2007 was used to make a fully corrected
abundance estimate including corrections for whales that were submerged during the passage of the survey plane. The line transect estimate for
2007 was 1,020 (CV = 0.35). When the estimate was corrected for perception bias with mark-recapture distance sampling (MRDS) methods, the
abundance increased to 1,505 (0.49). A correction for availability bias was developed based on time-depth-recorder information on the time spent
at the surface (0–4m). However, used directly this correction leads to a positively-biased abundance estimate and instead a correction was developed
for the non-instantaneous visual sighting process in an aircraft. The resulting estimate for 2007 was 3,272 (CV = 0.50) for the MRDS analysis. An
alternative strip census estimate deploying a strip width of 300m resulted in 995 (0.33) whales. Correction for perception bias resulted in 991 (0.35)
whales and corrected for the same availability bias as for the MRDS method resulted in a fully corrected estimate of 2,154 (0.36) humpback whales
in West Greenland in 2007.
KEYWORDS: HUMPBACK WHALE; ABUNDANCE ESTIMATE; SURVEY-AERIAL; SATELLITE TAGGING; WEST GREENLAND;
MARK-RECAPTURE; DISTANCE SAMPLING
example, annual increases of 11% from 1970 to 1988
(Sigurjónsson and Gunnlaugsson, 1990) and 12% during
1986 and 2001 (Pike et al., 2009) around Iceland, 5.5% in
the Gulf of Maine (Barlow and Clapham, 1997) and 9.4% in
the Western North Atlantic (Katona and Beard, 1990) have
been observed or estimated. Until now, no estimates of
changes in abundance have been developed for the West
Greenland feeding ground.
Aerial surveys for common minke (Balaenoptera
acutorostrata) and fin whales (Balaenoptera physalus) have
been conducted at regular intervals in West Greenland since
1984. Estimates of abundance of humpback whales from
these surveys have only been presented for 2005 (Heide-
Jørgensen et al., 2008) mostly due to the low number of
sightings in the previous years.
In this study the aerial survey data from 1984 to 1993 were
re-examined and used to construct a time series of the
relative abundance of humpback whales using eight surveys
from 1984, 1985, 1987, 1988, 1989, 1993, 2005, and 2007.
These estimates are then used together with recent
abundance estimates to estimate the rate of increase of
humpback whales on the West Greenland feeding ground
since 1984. The observed rate of increase is compared to a
theoretical model of the plausible range of growth based on
J. CETACEAN RES. MANAGE. 12(1): 1–14, 2012 1
INTRODUCTION
Humpback whales (Megaptera novaeangliae) undertake
long migrations between high latitude, productive feeding
grounds during summer and warmer oligotrophic
mating/breeding grounds at low latitudes during winter
(Kellogg, 1929; Norris, 1967). The main breeding grounds
in the North Atlantic are located in the West Indies and the
feeding grounds are primarily located in northern Norway,
around Iceland, in West Greenland, in eastern Canada, and
in the Gulf of Maine (Stevick et al., 2003).
The large catches of North Atlantic humpback whales
during the commercial whaling époque nearly exterminated
the population and as an effect commercial whaling of
humpback whales has been banned since 1955 (Smith and
Reeves, 2002). To document the recovery of such long-lived,
slowly reproducing migratory species long time series of
abundance estimates covering the distributional range of the
population is needed. Such time series of abundance have
been collected in most of the core areas and there seem to be
a general increase in the population. In the West Indies the
instantaneous rate of increase between 1979 and 1993 has
been estimated at 3.1% (Stevick et al., 2003).
Increases in abundance of humpback whales have also
been detected at several of these feeding grounds. For
*Greenland Institute of Natural Resources, Box 570, DK-3900 Nuuk, Greenland.
+Polar Science Center, Applied Physics Laboratory, University of Washington, 1013 NE 40th Street, Box 355640, Seattle, WA 98105-669, USA.
#RUWPA, The Observatory, Buchanan Gardens, University of St Andrews, KY16 9LZ.
£ Greenland Climiate Research Centre, Greenland Institute of Natural Resources, Kivioq 2, 3900 Nuuk, Greenland.
$University of Gothenburg, Department of Marine Ecology, Box 461, SE-405 30 Gothenburg, Sweden.
&Húsavik Research Center, University of Iceland, Hafnarstétt 3, 640 Húsavik, Iceland.
bAarhus University, Department of Bioscience, Frederiksborgvej 399, DK-4000 Roskilde, Denmark.
2HEIDE-JØRGENSEN et al.: HUMPBACK WHALES IN WEST GREENLAND
Table 1
Effort and sightings distributed by year and strata that are comparable between years for the aerial surveys of West Greenland. Only effort and sightings in
Beaufort sea state <5 is included.
Year/strata Effort (km) Area (km2) Transects Effort/area Sightings Mean pod size (SE) Sighting rate
1984
1: 71°20–70°N 491 24,516 5 0.0200
2: 70°–68°30’N 435 17,872 3 0.0243
3A: 68°30–67’N inshore 224 14,913 3 0.0150
3B: 68°30–67’N offshore 735 19,305 7 0.0381
4A: 67°–66’N inshore 442 9,446 5 0.0468
4B: 67°–66°N offshore 398 8,311 6 0.0479
5A: 66°–65°N inshore 174 6,431 3 0.0271
5B: 66°–65°N offshore 644 10,900 7 0.0591
6: 65–64°N 2,145 17,107 15 0.1254 3
7: 64–63°N 699 11,122 7 0.0628 1
8: 63°–62°N 410 11,748 4 0.0349 1
Sum 6,797 151,671 65 0.0448 5 2.14 (0.27) 0.00074
1985
1: 71°20–70°N 791 24,516 7 0.0323
2: 70°–68°30’N 321 17,872 2 0.0180
3A: 68°30–67’N inshore 337 14,913 4 0.0226
3B: 68°30–67’N offshore 424 19,305 4 0.0220
4A: 67°–66’N inshore 444 9,446 5 0.0470 1
4B: 67°–66°N offshore 462 8,311 7 0.0556
5A: 66°–65°N inshore 829 6,431 9 0.1289 2
5B: 66°–65°N offshore 1,156 10,900 12 0.1061 1
6: 65–64°N 1,007 17,107 7 0.0589 3
7: 64–63°N 298 11,122 3 0.0268
8: 63°–62°N 772 11,748 6 0.0657
Sum 6,841 151,671 66 0.0451 7 2.14 (0.27) 0.00102
1987
1A: 71°30’–69°15’N 1,915 14,779 13 0.1296
1B: Disko Bay and Vaigat 729 5,358 11 0.1361
2: 69°15’–67°N 1,153 39,883 7 0.0289
3: 67°–64°15’N 1,417 42,400 8 0.0334 4
4: 64°15’–60°40’N 1,673 25,165 9 0.0665 1
5: 60°40’–58°45’°N 1,118 16,518 8 0.0677 2
Sum 8,005 144,103 56 0.0556 7 1.9 (0.14) 0.00087
1988
1A: 71°30’–69°45’N 703 24,560 10 0.0286
1B: Disko Bay and Vaigat 404 13,876 12 0.0291
2A: 69°45’–68°N 820 29,228 5 0.0281
2B: 68°–66°30’N 1,077 19,488 10 0.0553
3: 66°30’–64°15’N 1,399 41,660 9 0.0336 7
4: 64°15’–60°45’N 648 50,742 6 0.0128 2
5: 60°45’N–58°45’N 605 34,283 8 0.0176
Sum 5,656 213,837 60 0.0265 9 1.1 (0.14) 0.00159
1989
2A: 69°45’–68°00’N 428 29,228 4 0.0146
2B: 68°–66°30’N 836 19,488 5 0.0429
3: 66°30’–64°15’N 706 41,660 11 0.0169 1
4: 64°15’–60°45’N 1,218 50,742 19 0.0240 2
5: 60°45’–58°45’N 72 34,283 2 0.0021
Sum 3,260 175,401 41 0.0186 3 2.7 (0.7) 0.00092
1993
1A: 71°30’–69°45’N 138 25,130 5 0.0055
1B: Disko Bay and Vaigat 392 13,110 8 0.0299
2A–C: 69°45’–68°00’N 1,635 15,160 0.1078
2B–C: 68°–66°30’N 94 15,700 5 0.0060
3 offshore: 66°30’–64°15’N 185 26,680 2 0.0069 1
3 coast: 66°30’–64°15’N 828 23,100 10 0.0358 6
4 offshore: 64°15’–60°45’N 348 24,320 4 0.0143
4 coast: 64°15’–60°45’N 2,341 27,410 29 0.0854 9
5 offshore: 60°45’–58°45’N 436 18,450 6 0.0236 1
5 coast: 60°45’–58°45’N 881 14,920 11 0.0590 3
Sum 7,140 178,850 75 0.0399 20 3.2 (0.60) 0.00280
2005
CF: 59°–58°N 293 11,523 4 0.0254
CW: 67°30’–64°N 1,958 74,798 30 0.0262 4
Disko Bay 556 12,312 12 0.0452 1
SG: 61°–59°N 1,106 19,491 19 0.0567 4
SH: 68°30’–67°30’N 577 15,669 7 0.0368
SW: 64°–61°N 1,968 29,781 31 0.0661 13
Sum 6,458 163,574 103 0.0395 22 8.3 (0.38) 0.00340
Cont.
J. CETACEAN RES. MANAGE. 12(1): 1–14, 2012 3
life history observations from North Atlantic and North
Pacific humpback whale populations.
MATERIAL AND METHODS
Construction of abundance estimates for 1984 and 1985
Aerial surveys of the West Greenland banks north of 62°N
were conducted in June–July 1984 and 1985 (Figs 1a and 1b).
East-west going transects separated by two nautical miles
were chosen randomly and were flown in a twin-engine high
winged Partenavia Observer P68 at a target altitude and
speed of 183m (600ft) and 160km hr–1 (100 knots),
respectively. Three observers participated and the right front
observer also acted as data recorder. Distance to sightings was
estimated with Suunto inclinometers and was together with
information on size of humpback whale groups recorded on
tape recorders. The number of sightings from the surveys in
1984 and 1985 were too low to develop reliable detection
functions. Instead the detection function from the surveys in
1987–1989 was used with a left truncation at 200m to take
into account the effects of the flat windows used in the 1984–
85 surveys (cf. Richard et al., 2010).
Construction of abundance estimates for 1987–89 and
1993
Aerial line transect surveys covering the West Greenland
banks were completed in July–August 1987–1989 and 1993
(Figs 1c to 1f) and were conducted with a twin engine
Partenavia Observer P68 with two observers in rear seats
with bubble windows and one observer in the right front seat
with a flat window. Information on size of humpback whale
groups and declination angle to sightings measured with
Suunto inclinometers were recorded.
Due to the low number of sightings, a common detection
function was developed for the surveys between 1987 and
1989. These surveys all used the same aircraft, the same
target altitude (229m or 750ft), same speed (160km hr–1) and
in some cases, the same observers. The surveys were also
completed in weather conditions that were similar between
years. The survey in 1993 had a sufficient number of
sightings to develop an independent detection function.
Construction of abundance estimate for 2005
An aerial survey in 2005 covering most of West Greenland
(Fig. 1g) essentially used the same aircraft and techniques as
previous surveys and the details of the survey were presented
in Heide-Jørgensen et al. (2008). The survey provided
several sightings of large groups (>10 whales) which caused
problems for the line transect estimation. Instead a line
transect estimate for all groups <10 whales was derived and
added to a strip census estimate of all groups >10 whales
(discussed in detail in Heide-Jørgensen et al., 2008).
Construction of abundance estimates for 2007
An aerial line transect survey of humpback whales in West
Greenland was conducted between 25 August and 30
September 2007. The survey platform was a Twin Otter, with
long-range fuel tank and two pairs of independent observers
all with bubble windows. Sightings and a log of the cruise
track (recorded from the aircrafts GPS) were recorded on a
Redhen msDVRs system that also allowed for continuous
video recording of the trackline as well as vertical digital
photographic recordings. Declination angle to sightings was
measured with Suunto inclinometers. Target altitude and
speed was 213m and 167km hr–1, respectively.
Survey conditions were recorded by the primary observers
at the start of the transect lines and whenever a change in sea
state, horizontal visibility and glare occurred. The survey was
designed to systematically cover the area between the coast
of West Greenland and offshore (up to 100km) to the shelf
Table 1 cont.
Year/strata Effort (km) Area (km2) Transects Effort/area Sightings Mean pod size (SE) Sighting rate
2007
1: Uummannaq Fjord 191 8,404 3 0.0227
2: 71°30’–69°45’N 502 22,631 5 0.0222
3: Disko Bay and Vaigat 532 14,653 9 0.0363
4: 69°45’–68°N 545 34,272 4 0.0159 1
5: 68°–66°30’N offshore 862 16,226 9 0.0531 3
6: 68°–66°30’N inshore 973 14,902 9 0.0653
7: 66°30’–64°N offshore 551 22,085 6 0.0249 2
8: 66°30’–64°N inshore 1,345 20,264 12 0.0664 5
9: 64°–62°N 998 20,334 12 0.0491 4
10: 62°–60°30’N 932 15,951 10 0.0584 3
11: 60°30–59°N 1,194 24,085 16 0.0496 2
14: coastal 67–66°30’N 45 189 6 0.2381 1
Sum 8,670 213,996 101 0.0405 21 1.5 (0.21) 0.00242
Table 2
Estimates of relative abundance of humpback whales in West Greenland.
Numbers in parenthesis indicate the coefficient of the variation. Photo-id
estimates from 1982 from Perkins et al. (1984; 1985) and from 1988–92
from Larsen and Hammond (2004). Aerial line-transect estimates from
1984–85 and 1987–93 from this study, from 2005 from Heide-Jørgensen et
al. (2008) and from 2007 from this study. The ship-based line transect
estimate is from Heide-Jørgensen et al. (2007). *=partial coverage.
Aerial line Ship-based line
Year transect abundance transect abundance Photo-id
1982 – – 271 (0.13)
1984 99 (0.46)* – –
1985 177 (0.44)* – –
1987 220 (0.62) – –
1988 200 (0.74) – –
1989 272 (0.75) – 357 (0.16)
1990 – – 355 (0.12)
1991 – – 376 (0.19)
1992 – – 566 (0.42)
1993 873 (0.53) – 348 (0.12)
2005 1,158 (0.35) 1,306 (0.42) –
2007 1,020 (0.35) – –
4HEIDE-JØRGENSEN et al.: HUMPBACK WHALES IN WEST GREENLAND
break (i.e. the 200m depth contour). Transect lines were
placed in an east-west direction except for south Greenland
where they were placed in a north-south direction. The
surveyed area was divided into 12 strata (Fig. 1h).
Conventional line transect abundance estimation for all
the surveys
Declination angles to sightings were converted to
perpendicular distance of the animal to the trackline from:
distance (m) = 213*tan(90-angle). Using conventional
distance sampling (CDS) methods, animal abundance in each
stratum was estimated by
where Ais the area of the stratum, Lis the total search effort
in the stratum, n is the number of unique groups detected in
the stratum by either observer and
μ
ˆwas the estimated
effective strip width of perpendicular distances to detected
groups and E
ˆ[s] was the estimated mean group size estimated
using a regression of log group size against estimated
detection probability (cf. Buckland et al., 2001).
Mark-recapture distance sampling correction for
perception bias for the 2007 survey
The search method deployed during the 2007 survey used an
independent observer configuration where the primary and
secondary observer teams acted independently of each other.
Detections of animals by the primary observer served as a
set of binary trials in which a success corresponded to a
detection of the same group by the secondary observer in the
same side of the aircraft. The converse was also true because
the observers were acting independently; detections by
secondary observers served as trials for the primary
observers. Analysis of the detection histories using logistic
regression allowed the probability that an animal on the
trackline was detected by an observer to be estimated, and
thus, abundance could be estimated without assuming g(0)
was one. These methods combine aspects of both mark-
recapture (MR) techniques and distance sampling (DS)
techniques and so they are known as mark-recapture distance
sampling (MRDS) methods (Laake and Borchers, 2004).
Although observers were acting independently,
dependence of detection probabilities on unmodelled
variables (called unmodelled heterogeneity) can induce
correlation in the detection probabilities. Laake and Borchers
(2004) and Borchers et al. (2006) developed estimators
ˆ
N=
n
2Lˆ
μ
ˆ
E[s]A
which assumed that detections were independent at zero
perpendicular distance only (called point independence
estimators) that are well suited for aerial surveys where no
responsive movements are expected.
The effects of the correlation in detections can be reduced
by modelling the effects of variables which cause the
correlation. Variables, additional to perpendicular distance,
can be included in the MRDS models using a model selection
criteria to select the best model. Detection probability was
estimated using the independent observer configuration
implemented in Distance 6.0 (Thomas et al., 2009).
Group abundance was estimated in each stratum using:
where w is the truncation distance, zi
–is a vector of
explanatory variables for group i(possibly including the
group size, si) and pˆ(zi
–) is the estimated probability of
detecting group iobtained from the fitted MRDS model.
Individual animal abundance is estimated by
The estimated mean group size in the stratum is given by
Strip census estimation of the survey in 2007
Most of the humpback whale sightings were made within
300m from the trackline and at relatively short distances. The
detection function dropped beyond 300m and it was therefore
decided to assume a constant probability of detecting a group
of humpback whales in a 300m strip on each side of the
aircraft. The mark-recapture line transect analysis indicates
that no variables other than distance and observer affect
detection probability (see later). Thus in addition to the CDS
estimates a strip census estimate was also obtained using a
simple arithmetic mean of the group size across all strata (s¯¯ ).
To correct for perception bias (p’) by the observers
Chapman’s (1951) modification of the Petersen estimator was
used to estimate group abundance within w= 300m of the
trackline (the ‘covered region’) over all strata:
ˆ
NG
=
A
2wL
1
ˆ
p(z i)
i=1
n
ˆ
N=
A
2wL
si
ˆ
p(z
i
)
i=1
n
ˆ
E[s]=
ˆ
N
ˆ
N
G
ˆ
NG.strip
=
(n1
+1)( n2
+1)
(m2
+1)
1
Table 3
Humpback whale abundance estimates in 2007 using CDS methodology showing the encounter rate (n/L), effective strip width (esw) and estimates for pod
size E[s], pod density DG, pod abundance NG, animal density Dand animal abundance N. Strata without sightings are not shown although the total densities
take all strata into account. CV are given in parentheses.
Stratum n/L (pods/km) esw (km) E[s] DG (pods/km2) NG (pods) D (whales/km2) N (whales)
4 0.0018 (0.81) 0.0030 (0.83) 101 (0.83) 0.0041 (0.84) 141 (0.84)
5 0.0035 (0.77) 0.0056 (0.79) 91 (0.79) 0.0078 (0.80) 127 (0.80)
7 0.0036 (0.96) 0.0058 (0.97) 129 (0.97) 0.0081 (0.98) 180 (0.98)
8 0.0037 (0.61) 0.0060 (0.64) 121 (0.64) 0.0083 (0.65) 169 (0.65)
9 0.0050 (0.38) 0.311 (0.19) 1.394 (0.12) 0.0081 (0.43) 164 (0.43) 0.0112 (0.44) 228 (0.44)
10 0.0021 (0.68) 0.0035 (0.71) 55 (0.71) 0.0048 (0.72) 77 (0.72)
11 0.0017 (0.60) 0.0027 (0.63) 65 (0.63) 0.0038 (0.64) 90 (0.64)
14 0.0223 (0.85) 0.0358 (0.87) 7 (0.87) 0.0500 (0.88) 9 (0.88)
Total 0.0022 (0.20) 0.0033 (0.33) 732 (0.33) 0.0046 (0.35) 1,020 (0.35)
J. CETACEAN RES. MANAGE. 12(1): 1–14, 2012 5
where nis the total number of sightings, n1and n2are the total
number of sightings by the primary and secondary observers
and m2is the number of sightings by both pairs of observers.
The abundance in stratum v(v= 4,5,7,8,9,10,11,14) was
estimated as follows:
where nvis the number of groups detected in stratum v, L vis
the total length of transect in stratum v, A vis the surface area
of stratum vand the combined detection probability for both
observers (p’) across all strata was estimated as follows:
The variance of pˆ ′, N
ˆG,vand N
ˆG(Chapman)= ∑vN
ˆG,vwas estimated
using a nonparametric bootstrap with transect as the
sampling unit. Transects were sampled with replacement,
separately in each stratum, until the total number of sightings
was at least as large as the original number of sightings in
the stratum (nv).
The mean group size s¯¯ and its coefficient of variation,
cv(s¯¯ ) was estimated across all strata and estimated individual
abundance and its CV was obtained by
and
Correction for availability bias of the survey in 2007
The above estimates of abundance from aerial surveys are
negatively biased if some animals were underwater and hence
undetectable during the passage of the plane. To correct for
this availability bias satellite-linked time-depth recorders
were deployed on five humpback whales off Central West
Greenland (Fyllas Bank 64°N, 52°W) in June–July 2000 to
estimate the probability of an animal being available for
detection. The satellite transmitters (SDR-T16) produced by
Wildlife Computers (Redmond, Washington) were fitted with
a harpoon spear for attachment. The transmitter had a length
of 10cm and a diameter of 2.5cm and was sitting on the
outside of the whale while an anchoring spear of 14.5cm was
partly or fully inside the whale. The tags were programmed
to collect and summarize measurements of the time spent at
or above 4m depths in four 6hr periods and the data were
transmitted through Service Argos. The tags were deployed
from the stern of a MK II Zodiac powered by a 40 Hp engine.
ˆ
NG,v
=
nvAv
ˆ
p'2wLv
ˆ
p'=
n
ˆ
NG.strip
=
n
(n1
+1)( n2
+1)
(m2
+1)
1
ˆ
N(Chapman)
=ˆ
NG(Chapman)s
cv(ˆ
N(Chapman))=cv(ˆ
NG(Chapman))2+cv(s)2
A person fixed with a harness deployed the transmitter with
a 6.8m aluminum pole (diameter 33mm).
As humpback whales are available for more than an
instant during aerial surveys and some whales may even be
seen ahead of the plane, the probability that an animal is
available is not simply the probability that it is available at a
randomly-chosen instant in its dive cycle. McLaren (1961)
derived an equation, used by others, including Barlow et al.
(1988) for estimating the average probability that an animal
is available (at the surface) at least some of the time within
a time interval of length t:
Pr (available) = (s+t)/(s+d)
where sis the average time the whale is at the surface, dis
the average time it is below the surface and tis the window
of time the whale is within visual range of the observers.
However, this equation is inappropriate if tis not very small
relative to d, as is clear by noting that when t>d the
probability is greater than 1. A more appropriate estimator
of the probability that an animal is available within time t
was provided by Laake et al. (1997):
where E[s] is the average time the whale is at the surface, E[d]
is the average time it is below the surface and tis the window
of time the whale is within visual range of the observers.
It was assumed that the whales were available for
detection when within 4m of the surface and the times spent
at above and below this measurement from 7 June through
18 July from the satellite-linked time-depth-recorders were
used to estimate this probability.
Abundance (corrected for availability bias) was then
estimated as
with estimated CV
Construction of time series
A time series of indices of relative abundance of humpback
whales was constructed from previous photo ID mark-
recapture studies and from aerial and ship-based surveys
presented previously (Heide-Jørgensen et al., 2007; Larsen
and Hammond, 2004), re-analysed in this study (Heide-
ˆ
a=
E[s]
E[s]+E[d]
+
E[d](1 et/E[d])
E[s]+E[d]
ˆ
Nc
=
ˆ
N
ˆ
a
cv(ˆ
N
c)=cv(ˆ
N)2+cv(ˆ
a)2.
Table 4
MRDS point independence model fitted to the data from 2007 survey.
Distance sampling model Mark recapture model AIC ΔAIC
Uniform Petersen 205.34 0
Half Normal: Distance Distance 296.03 90.69
Hazard rate: Distance Distance 296.55 91.21
Half normal: Distance Distance + Observer 292.97 87.63
Hazard rate: Distance Distance + Observer 293.49 88.15
Table 5
Number of sightings seen by each observer and the number of duplicates
(seen by both) during the 2007 survey. The total column shows the number
of sightings seen by observer 1 plus observer 2 minus sightings seen by
both.
Primary Secondary
Pod size observer observer Seen by both Total
1 14 11 10 15
2 4 1 1 4
3 1 1 1 1
5 1 1 1 1
Total 20 14 13 21
Jørgensen et al., 2008; Larsen, 1995; Larsen et al., 1989) or
presented for the first time here. The trend in abundance or
instantaneous rate of increase (Nt= Noert) was estimated by
weighted (weight = 1/cv(Nt)2) regression through the log
transformed estimates of relative abundance (Nt) with jack-
knifed standard error.
Population dynamics model
An age based Leslie-matrix model was created (Caswell,
2001; Leslie, 1945; 1948) using life-history data obtained
from literature (Barlow and Clapham, 1997; Clapham, 1992;
Gabrielle et al., 2001; Mizroch et al., 2004). This model was
used to calculate the growth rate at a stable age structure as
the dominant positive eigenvalue of the matrix. The matrix
only projects female individuals, and due to this, the fertility
used is half of that reported in the literature, since there is no
evidence of a strongly biased sex ratio at birth.
RESULTS
Construction of estimates of relative abundance
In all years, the aerial surveys covered the coastal areas of
West Greenland from 60°N (in 1984 and 1985 from 62°N)
to 70°N with the maximum effort between 62° and 66°N
(Figs 1a–h). The total survey effort however ranged between
3,260 and 8,670km (Table 1). The average ratio between
survey effort and stratum area was 0.04 (SD = 0.01).
However this fluctuated in the first five years between 0.02
and 0.06, but remained constant around 0.04 after 1989. The
seven abundance estimates were not significantly correlated
with the survey effort (p= 0.42). There was an increasing
trend in sighting rate in the aerial surveys with r= 0.06 (CV
= 0.28, r2= 0.69) for the period 1984 to 2007.
The combined detection function for humpback whales
for the surveys in 1987–89 was fitted with a half-normal
function with a left truncation at 200m to construct a
detection function for the surveys in 1984–85 that used flat
windows. The sample size was 10 and the effective search
width was 587m (CV = 0.37) (Fig. 2a). The distribution of
perpendicular distances to the 15 humpback whale sightings
were combined for the surveys in 1987–1989 and a half-
normal model was selected to fit the sightings distance data
(Fig. 2b). The effective search width was estimated at 708m
(CV = 0.20). The survey in 1993 had 18 sightings that were
fitted to the half-normal model to derive an effective search
width of 503m (CV = 0.43, Fig. 2c). A simple mean of the
group sizes was used for each of the years.
In 2005, 22 sightings within the truncation distance of
3km were used for deriving a half-normal detection function
model with an effective search width of 664m (CV = 0.12,
Fig. 2d), similar to that found in previous years (see Heide-
Jørgensen et al., 2008). A regression of log group size against
estimated detection probability was used to estimate mean
group size across all strata.
In 2007, the distribution of perpendicular distances of
sightings shows some sightings close to the trackline
indicating the absence of a blind spot for observers beneath
the plane (Fig. 2e). However, in the distributions for both
observers there was a peak in sightings between 200–250m
after which detection declined substantially. In 2007 all
6HEIDE-JØRGENSEN et al.: HUMPBACK WHALES IN WEST GREENLAND
Table 6
Humpback whale abundance estimates in 2007 using MRDS methodology showing the encounter rate (n/L), estimates for pod size E[s], pod density DG, pod
abundance NG, whale density Dand whale abundance N. Strata without sightings are not shown although the total densities take all strata into account. CV’s
are given in parentheses.
Stratum n/L (pods/km) DG (pods/km2)NG (pods) D (whales/km2) N (whales) E[s]
4 0.0018 (0.81) 0.0040 (0.90) 136 (0.90) 0.0040 (0.90) 136 (0.90) 1.00 (00.0)
5 0.0035 (0.77) 0.0075 (0.86) 122 (0.86) 0.0125 (0.96) 203 (0.96) 1.67 (0.21)
7 0.0036 (0.96) 0.0078 (1.03) 173 (1.03) 0.0157 (1.03) 346 (1.03) 2.00 (00.0)
8 0.0037 (0.61) 0.0080 (0.72) 163 (0.72) 0.0080 (0.73) 163 (0.73) 1.00 (00.0)
9 0.0050 (0.38) 0.0108 (0.54) 220 (0.54) 0.0238 (0.60) 484 (0.60) 2.20 (0.34)
10 0.0021 (0.68) 0.0046 (0.78) 74 (0.78) 0.0046 (0.78) 74 (0.78) 1.00 (0.24)
11 0.0017 (0.60) 0.0036 (0.71) 87 (0.71) 0.0036 (0.71) 87 (0.71) 1.00 (00.0)
14 0.0223 (0.85) 0.0482 (0.93) 9 (0.93) 0.0489 (0.94) 9 (0.94) 1.00 (00.0)
Total 0.0022 (0.20) 0.0045 (0.47) 985 (0.47) 0.0068 (0.49) 1,505 (0.49) 1.53 (0.14)
Table 7
Proportion of time spent at surface (0–4m) for four humpback whales
instrumented on Fyllas Bank in June 2006.
Whale Date 6 hr period Percentage time at 0–4m
21809 8/6/2000 03–09 47.92
20158 7/6/2000 03–09 19.80
20158 8/6/2000 03–09 25.59
31.10
21801 10/6/2000 09–15 37.17
21801 20/06/2000 09–15 42.51
21802 10/6/2000 09–15 34.35
21802 17/6/2000 09–15 68.42
21802 18/6/2000 09–15 71.75
21802 22/6/2000 09–15 32.04
47.71
21801 10/6/2000 15–21 33.52
21801 14/6/2000 15–21 26.57
21801 15/6/2000 15–21 40.67
21801 16/7/2000 15–21 34.94
20160 9/6/2000 15–21 26.53
21802 14/6/2000 15–21 37.73
21802 17/6/2000 15–21 57.77
21802 19/6/2000 15–21 39.58
37.16
21801 9/6/2000 21–03 31.79
21801 11/6/2000 21–03 26.35
21801 14/7/2000 21–03 44.44
21801 18/7/2000 21–03 42.62
20158 5/6/2000 21–03 48.89
20158 7/6/2000 21–03 30.72
21802 16/6/2000 21–03 57.64
21802 23/6/2000 21–03 35.30
39.72
Average All days all whales 09–21 41.68
SD 14.24
n 14.00
SE 3.81
CV 0.09
J. CETACEAN RES. MANAGE. 12(1): 1–14, 2012 7
sightings were within 500m from the trackline which is very
different from the distribution in 2005 where most sightings
were beyond 500m. The difference is due to a combination
of a different type of survey planes and observer instruction
in 2007 to concentrate on covering the trackline. Both hazard
rate and half normal functional forms were considered for
the 2007 distribution of sightings, but based on AIC the half-
normal model was chosen. The effective search width was
311m (CV = 0.19). The survey region in the 2007 survey
included an area of 213,996km2with 8,670km tracklines
covered in Beaufort sea states less than 5 (Fig. 1h and Table
1). The group sizes varied between 1 and 5 whales and all
the 21 humpback whale sightings were seen in strata 4 to 11
with the exception of one sighting in stratum 14.
Trends in abundance
The uncorrected estimates from the aerial surveys are smaller
than the estimates from the photo identification study except
for 1993 where the survey abundance estimate was about
twice the estimate from the photo ID study (Fig. 3). It is
however not straightforward to compare the estimates as the
aerial surveys covered a much larger area and they are not
corrected for the time the whales were not available at the
surface to be seen by the observers. The aerial survey
estimate from 2005 (1,158 95% CI 595–2,255) is similar to
a ship-based line transect survey in 2005 (Fig. 3).
The time series of aerial line transect surveys provides an
index of the changes in relative abundance (i.e. uncorrected
for perception and availability bias) of humpback whales in
West Greenland from 1984 through 2007 (Table 2). If it is
assumed that the bias remains constant, the rate of increase of
humpback whales on the feeding ground in West Greenland
can be estimated. The abundance estimates from 1984–1985
and 1987–1989 used the same detection function and were
therefore averaged for the purpose of estimating the rate of
increase. The overall exponential rate of increase from 1984
to 2007 was 0.09 or 9.4% per year (SE = 0.01, p = 0.010).
Current abundance
The CDS estimate of 1,020 (CV = 0.35) humpback whales
for 2007 does not include animals that were submerged or
missed by the observers (Table 3). Both the conventional DS
model and the MRDS models were fitted to the data without
truncation. The final MRDS model included a term for
observer in the MR model (Table 4). This indicated that the
secondary observers had a much smaller probability of
detection on the trackline than the primary observers (Table
5); 0.66 (CV = 0.43) for the primary observers compared to
0.22 (CV = 0.76) for the secondary observers (Fig. 4). The
estimate for both observers combined was 0.73 (CV = 0.34).
The abundance of humpback whales was 1,505 animals (CV
= 0.49; 95% CI 581–3896) when using MRDS methods to
correct for perception bias (Table 6).
Data on surface time obtained from the satellite-linked time-
depth-recorders indicate that humpback whales in West
Greenland spend on average 42% (CV = 0.09) of their time
during daylight periods (09–21hr) at depths <4m (Table 7). In
the relatively productive waters of West Greenland, 4m is
probably the maximum depth to which humpback whales can
be reliably detected on the trackline from an aircraft passing
at 213m altitude. Humpback whales are known to have long
dive cycles with average dive times lasting several minutes
and with average time spent at the surface (<4m) mostly lasting
>40 seconds (Winn and Reichley, 1985). Both the dive time
and the at-surface-time are considerably longer than the
average time the whales are visible from an aircraft. In this
survey the time between first sighting of the whales and the
time when the whales passed abeam was on average 3.21s
(CV = 0.38). If the probability of detecting a whale at the
surface given the observation time of 3.21s and the ratio
between dive and surface times is compared to an
instantaneous correction of whales at the surface then the most
severe positive bias can be expected for short durations of
surfacings and dives (Fig. 5). For surface times >30s the
positive bias from using an instantaneous correction of
availability ranges between 7 and 15% for observation times
between 2 and 7s, or 10% for an average 3.21s observation
period. This positive bias can be eliminated by increasing the
availability correction factor to 0.46. Applying this correction
to the MRDS estimate gives a fully corrected abundance
estimate of 3,272 (CV = 0.50, 95% CI 1,300–8,233) humpback
whales in West Greenland in 2007.
The Chapman estimate of perception bias was 0.98 (CV
= 0.03) and correcting for this bias results in an abundance
of 995 (0.33) humpback whales in 2007 from the strip census
analysis (Table 8). In comparison the CDS estimate was
1,020 (0.35) and the MRDS estimate was 1,528 (0.51).
Further correction of the strip census analysis with â46%
(CV = 0.09) gives an estimate of 2,154 (CV = 0.36, 95% CI
1,087–4,270) humpback whales corrected for whales that
were submerged during the passage of the plane or a slightly
lower but more precise estimate than the MRDS estimate.
Table 8
Humpback whale estimates in 2007 using strip census methodology and estimated detection probability pˆ ′ = 0.98 (cv = 0.03) with esw = 300m showing the
encounter rate (n/L) and simple estimate of pod size s¯¯, pod density DG, pod abundance NG, animal density D, and Nanimal abundance. Strata without sightings
are not shown. CV’s are given in parentheses.
Stratum n/L (pods/km) s¯¯ DG (pods/km2) NG (pods) D (animals/km2) N (animals)
4 0.002 (0.81) 0.003(0.81) 105 (0.81) 0.004 (0.82) 149 (0.83)
5 0.004 (0.77) 0.006 (0.77) 94 (0.77) 0.008 (0.78) 134 (0.78)
7 0.004 (0.96) 0.006 (0.96) 134 (0.96) 0.009 (0.97) 190 (0.97)
8 0.003 (0.75) 0.005 (0.75) 100 (0.75) 0.007 (0.77) 143 (0.77)
9 0.004 (0.47) 1.42 (0.16) 0.007 (0.47) 136 (0.47) 0.010 (0.49) 193 (0.49)
10 0.002 (0.68) 0.004 (0.68) 57 (0.68) 0.005 (0.70) 81 (0.70)
11 0.002 (0.60) 0.003 (0.60) 67 (0.60) 0.004 (0.62) 96 (0.62)
14 0.002 (0.85) 0.037 (0.85) 7 (0.85) 0.053 (0.86) 10 (0.86)
Total 0.002 (0.22) 0.003 (0.29) 700 (0.29) 0.005 (0.33) 995 (0.33)
8HEIDE-JØRGENSEN et al.: HUMPBACK WHALES IN WEST GREENLAND
Table 9
Life history data used to calculate plausible growth rates for North Atlantic humpback whales.
Lower CI Average Upper CI Geographical region Reference
Fertility (females) 0.20 0.21 0.22 North Atlantic Barlow and Clapham (1997)
Age at sexual maturity 6.4 5.9 5.4 North Atlantic Clapham (1992)
Calf survival 0.797 0.805 0.813 North Pacific Gabriele et al. (2001); Zerbini et al. (2010)
Juvenile survival 0.797 0.895 0.995 Estimated
Adult survival 0.954 0.984 0.995 North Pacific Mizroch et al. (2004)
Growth rate 0.9964 1.0578 1.1070 Calculated
Population dynamics
Age at first parturition is reported in decimal numbers in the
literature and was included in the age based matrix by adding
partial fertility at age 5 (Upper 95% CI and average models
in Table 9) or 6 (Lower CI model, based on the 95% CI for
the individual life history traits used) corresponding to the
deviation from the closest higher integer, i.e. 60% fertility at
age 5 (Upper) and 6 (Lower) for the CI models and 10%
fertility at age 5 for the average model. Calf survival was
multiplied by the fertility to obtain the chance of birth and
survival to age 1. Due to uncertain data in the literature,
juvenile survival (up to an age of first parturition of 5 or 6,
depending on model) was set as the average of calf and adult
survival in the average model, as the same value as calf
survival in the Lower CI model and as the same value as
adult survival in the Upper CI model. These widely ranging
numbers were used to avoid under- or over-estimation of the
extreme lambdas. The effect of juvenile survival was tested
within the average model where juvenile survival was
stepwise changed from calf survival values to the adult
survival values (0.8 to 0.96) which consequently affected the
growth rate linearly from 3% to 8% with all other parameters
kept constant. Survival estimates and fertility affected the
theoretical growth rates in a linear fashion whereas earlier
age of first parturition increased the growth exponentially
(Fig. 6). Estimates of the longevity of the whales had
relatively little effect on the theoretical growth rate.
DISCUSSION
Humpback whales have generally been protected in the North
Atlantic since 1955 although a low level of exploitation (total
catch 1955–85; 24) continued in West Greenland until 1985
(IWC, 2003). After 1985, they were completely protected
although a few whales were taken as bycatch in fishery
operations (total 1986–2001; 7, IWC, 2003). Considering this
low level of exploitation and the fact that the number of
humpback whales have clearly increased on their breeding
ground (i.e. the West Indies) and feeding grounds in other
areas of the North Atlantic, it is not surprising that the
abundance on the West Greenland feeding ground has also
increased. The detected increase is considerably larger than
the increase of 3.1% per year observed in the West Indies
(Stevick et al., 2003). However, it is of the same magnitude
as some of the estimates of increase from other North Atlantic
feeding grounds (Katona and Beard, 1990; Pike et al., 2009;
Sigurjónsson and Gunnlaugsson, 1990).
The analysis of the dynamics of a hypothetical humpback
whale population in the North Atlantic shows that the
observed growth in West Greenland is within the upper range
of plausible growth rates based on an age structured model
with life history parameters from observed populations of
humpback whales. Both the age at first parturition and
subadult survival had a profound effect on the dynamics of
the population and population specific determination of these
life history parameters is required to narrow the range of
plausible growth rates. The values used in the model were
from the Gulf of Maine (Clapham, 1992), an area considered
to be part of the range of the western North Atlantic
humpback whale breeding population that also is found in
West Greenland.
The use of upper and lower CI models should not be
interpreted as the 95% CI of population growth, since it is
based on the assumption that all life history traits are at their
own individual 95% CI border values. This leads to an over-
and under-estimation for the possible 95% CI for the whole
population growth since the probability of all life history
traits to be at their maximum/minimum values at the same
time is low. The matrix model does not discern between calf
survival for first time mothers and experienced mothers,
something that can have significant impact on other mammal
species (for example rabbits (Rödel et al., 2009) and
cheetahs (Durant et al., 2004)). A recent study on Hawaiian
humpback whales also show that larger females attract more
male suitors (Pack et al., 2009), which could have a
significant impact on young female fertility rates.
The estimates of humpback whale abundance derived
from the photo-identification study in West Greenland in
1989–1993 (Larsen and Hammond, 2004) may provide a
correct magnitude of the occurrence of humpback whales in
the areas where the photo-identification work was
concentrated at that time. However, the photo-identification
work covered a smaller area of West Greenland than the
aerial surveys and it is reasonable to expect that an increasing
humpback whale population will also expand its distribution.
Satellite tracking studies in 2001 and 2002 demonstrated that
some humpback whales do not spend time within the area
used for the photo-identification study (Heide-Jørgensen and
Laidre, 2007; GINR, unpubl. data). In recent surveys
humpback whales were found more widely in West
Greenland than in previous surveys and there are now
frequently records of observations far north in West
Greenland (e.g. in Uummannaq 71°N; GINR, unpubl. data).
If detection probability varies with distance within the first
300m (and the CDS and MRDS analyses strongly suggest it
does), then the strip transect estimate is negatively biased
because it neglects heterogeneity due to distance. If some
animals at distance zero are missed (and the MRDS analysis
suggests that this is the case), then the CDS estimate is
negatively biased. If the detection function does in reality
initially increase with distance from the transect line, the
MRDS estimator of abundance might be positively biased,
because while the MR component of the model allows this,
the CDS component does not (i.e. the CDS detection
function is monotonically decreasing) – see Fig. 4. While it
is difficult to say whether or not the MRDS estimate of
abundance is positively biased, it is probable that both the
strip transect and CDS estimates are negatively biased.
The best estimate of the abundance of humpback whales
in 2007 was 3,299 whales, with a relatively large coefficient
of variation (0.57). Even the lower bound of this estimate
(1,170 whales) is substantially higher than any previous
estimates. The estimate is based on a visual aerial line
transect survey that covered a larger part of West Greenland
than in previous surveys. However coverage was still partial
with poor coverage west of Disko Bay and humpback whales
were often observed at the westernmost point of the transects
indicating that the West Greenland feeding ground may
extend over deeper water (>200m) west of the shelf area into
areas not covered in any of the surveys.
The observed rate of increase and the estimates of current
abundance of humpback whales on the summering ground
in West Greenland change the status of this stock and allows
for the resumption of a low level of harvesting which was
abandoned in 1985.
ACKNOWLEDGEMENTS
We thank the observers Finn Christensen, Arne Geisler, Anita
Gilles and Werner Piper for their involvement and
persistence during the 2007 survey. The North Atlantic
Marine Mammal Commission is acknowledged for
organizing the 2007 TNASS survey, which the survey
presented here was part of. The Greenland Institute of
Natural Resources provided funding for the survey and the
Vetlessen Foundation supplied additional funding for
purchasing the recording equipment.
REFERENCES
Barlow, J. and Clapham, P.J. 1997. A new birth-interval approach to
estimating demographic parameters of humpback whales. Ecology 78(2):
535–46.
Barlow, J., Oliver, C.W., Jackson, T.D. and Taylor, B.L. 1988. Harbor
porpoise, Phocoena phocoena, abundance estimation for California,
Oregon, and Washington: II. Aerial surveys. Fish. Bull. 86(3): 433–44.
Borchers, D.L., Lake, J.L., Southwell, C. and Paxton, C.G.M. 2006.
Accommodating unmodelled heterogeneity in double-observer distance
sampling surveys. Biometrics 62: 372–78.
Buckland, S.T., Anderson, D.R., Burnham, K.P., Laake, J.L., Borchers, D.L.
and Thomas, L. 2001. Introduction to Distance Sampling: Estimating
Abundance of Biological Populations. Oxford University Press, Oxford,
UK. vi+xv+432pp.
Caswell, H. 2001. Matrix Population Models. Construction, Analysis and
Interpretation. 2nd ed. Sinauer Associates, Inc, Sunderland,
Massachusetts, USA. i–xxii+722pp.
Chapman, D.G. 1951. Some properties of the hypergeometric distribution
with applications to zoological censuses. Univ. Calif. Publ. Statist. 1:
131–60.
Clapham, P.J. 1992. Age at attainment of sexual maturity in humpback
whales, Megaptera novaeangliae. Can. J. Zool. 70(7): 1,470–1,72.
Durant, S.M., Kelly, M. and Caro, T.M. 2004. Factors affecting life and
death in Serengeti cheetahs: environment, age and sociability.
Behavioural Ecology 11(11–22).
Gabrielle, C.M., Straley, J.M., Mizroch, S.A., Baker, C.S., Craig, A.S.,
Herman, L.M., Glockner-Ferrari, D., Ferrari, M.J., Cerchio, S., von
Ziegesar, O., Darling, J., McSweeney, D., Quinn, T.J.I. and Jacobsen, J.K.
2001. Estimating the mortality rate of humpback whale calves in the
central North Pacific Ocean. Can. J. Zool. 79: 589–600.
Heide-Jørgensen, M.P., Borchers, D.L., Witting, L., Laidre, K.L., Simon,
M.J., Rosing-Asvid, A. and Pike, D.G. 2008. Estimates of large whale
abundance in West Greenland waters from an aerial survey in 2005. J.
Cetacean Res. Manage. 10(2): 119–30.
Heide-Jørgensen, M.P. and Laidre, K. 2007. Autumn space-use patterns of
humpback whale (Megaptera novaeangliae) in West Greenland. J.
Cetacean Res. Manage 9(2): 121–26.
Heide-Jørgensen, M.P., Simon, M.J. and Laidre, K.L. 2007. Estimates of
large whale abundance in Greenlandic waters from a ship-based survey
in 2005. J. Cetacean Res. Manage 9(2): 95–104.
International Whaling Commission. 2003. Report of the Scientific
Committee. Annex H. Report of the Sub-Committee on the
Comprehensive Assessment of Humpback Whales. J. Cetacean Res.
Manage. (Suppl.) 5:293–323.
Katona, S.K. and Beard, J.A. 1990. Population size, migrations and feeding
aggregations of the humpback whale (Megaptera novaeangliae) in the
western North Atlantic Ocean. Rep. int. Whal. Commn (special issue) 12:
295–305.
Kellogg, R. 1929. What is known of the migration of some of the whalebone
whales. Smithsonian Institution. Annual Report of the Board of Regents,
1928: 467–94+2pls.
Laake, J. and Borchers, D. 2004. Methods for incomplete detection at
distance zero. pp.108–89. In: Buckland, S.T., Anderson, K.P., Burnham,
K.P., Laake, J., Borchers, D. and Thomas, L. (eds). Advanced Distance
Sampling. Oxford University Press, Oxford. 595pp.
Laake, J.L., Calambokidis, J., Osmek, S.D. and Rugh, D.J. 1997. Probability
of detecting harbour porpoise from aerial surveys: Estimating g(0). J.
Wildl. Manage. 61(1): 63–75.
Larsen, F. 1995. Abundance of minke and fin whales off West Greenland,
1993. Rep. int. Whal. Commn 45: 365–70.
Larsen, F. and Hammond, P.S. 2004. Distribution and abundance of West
Greenland humpback whales Megaptera novaeangliae. J. Zool., London.
263: 343–58.
Larsen, F., Martin, A.R. and Nielsen, P.B. 1989. North Atlantic Sightings
Survey 1987: report of the West Greenland aerial survey. Rep. int. Whal.
Commn 39: 443–46.
Leslie, P.H. 1945. On the use of matrices in certain population mathematics.
Biometrika 33: 183–212.
Leslie, P.H. 1948. Some further notes on the use of matrices in population
mathematics. Biometrika 35: 213–45.
McLaren, I.A. 1961. Methods for determining the numbers and availability
of ringed seals in the eastern Canadian Arctic. Arctic 14: 162–75.
Mizroch, S.A., Herman, L.M., Straley, J.M., Glockner-Ferrari, D.A., Jurasz,
C., Darling, J., Cerchio, S., Gabriele, C.M., Salden, D.R. and von
Ziegesar, O. 2004. Estimating the adult survival rate of central north
Pacific humpback whales (Megaptera novaeangliae). J. Mammal. 85(5):
963–72.
Norris, K.S. 1967. Some observations on the migration and orientation of
marine mammals. pp.101–25. In: Storm, R.M. (eds). Animal Orientation
and Navigation. Proceedings of the Twenty-Seventh Annual Biology
Colloquium. Oregon State University Press, Corvallis, OR. ix+134pp.
Pack, A.A., Herman, L.M., Spitz, S.S., Hakala, S., Deakos, M.H. and
Herman, E.Y.K. 2009. Male humpback whales in the Hawaiian breeding
grounds preferentially associate with larger females. Anim. Behav. 77:
653–62.
Perkins, J.S., Balcomb, K.C., Nichols, G., Hall, A.T., Smultea, M. and
Thumser, N. 1985. Status of the West Greenland humpback whale feeding
aggregation, 1981–1983. Rep. int. Whal. Commn 35: 379–83.
Perkins, J.S., Balcomb, K.C., Nichols, G.N., Jr. and DeAvilla, M. 1984.
Abundance and distribution of humpback whales (Megaptera
novaeangliae) in West Greenland waters. Can. J. Fish. Aquat. Sci. 41(3):
533–36.
Pike, D.G., Paxton, C.G.M., Gunnlaugsson, T. and Vikingsson, G.A. 2009.
Trends in the distribution and abundance of cetaceans from aerial surveys
in Icelandic coastal waters, 1986–2001. NAMMCO Sci. Pub. 7: 117–42.
Richard, P.R., Laake, J.L., Hobbs, R.C., Heide-Jørgensen, M.P., Asselin,
N.C. and Cleator, H. 2010. Baffin Bay narwhal population distribution
and numbers: aerial surveys in the Canadian High Arctic, 2002–04. Arctic
63(1): 85–99.
Rödel, H.G., von Holst, D. and Kraus, C. 2009. Family legacies: short- and
long-term fitness consequences of early-life conditions in female
European rabbits. J. Anim. Ecol. 78: 789–97.
Sigurjónsson, J. and Gunnlaugsson, T. 1990. Recent trends in abundance of
blue (Balaenoptera musculus) and humpback whales (Megaptera
novaeangliae) off west and southwest Iceland, with a note on occurrence
of other cetacean species. Rep. int. Whal. Commn 40: 537–51.
Smith, T.D. and Reeves, R.R. 2002. Report of the Scientific Committee.
Annex H. Report of the Sub-Committee on the Comprehensive
Assessment of North Atlantic Humpback Whales. Appendix 2. Estimating
historical humpback whale removals from the North Atlantic. J. Cetacean
Res. Manage. (Suppl.) 4: 242–55.
Stevick, P.T., Allen, J., Clapham, P.J., Friday, N., Katona, S.K., Larsen, F.,
Lien, J., Mattila, D.K., Palsbøll, P.J., Sigurjónsson, J., Smith, T.D., Øien,
J. CETACEAN RES. MANAGE. 12(1): 1–14, 2012 9
N. and Hammond, P.S. 2003. North Atlantic humpback whale abundance
four decades after protection from whaling. Marine Ecology. Progress
Series 258: 263–73.
Thomas, L., Laake, J.L., Rexstad, E., Strindberg, S., Marques, F.F.C.,
Buckland, S.T., Borchers, D.L., Anderson, D.R., Burnham, K.P., Burt,
M.L., Hedley, S.L., Pollard, J.H., Bishop, J.R.B. and Marques, T.A. 2009.
Distance 6.0 Release 2. Research Unit for Wildlife Population
Assessment, University of St. Andrews, UK. [Available at:
http://www.ruwpa.st-and.ac.uk/distance/].
Winn, H.E. and Reichley, N.E. 1985. Humpback whale – Megaptera
novaeangliae (Borowski, 1781). pp.241–73. In: Ridgway, S.H. and
Harrison, R. (eds). The Sirenians and Baleen Whales. Academic Press,
London and Orlando. xviii+362pp.
Zerbini, A.N., Clapham, P.J. and Wade, P.R. 2010. Plausible maximum rates
of increase in humpback whales. Mar. Biol. 157: 1225–36.
Date received: January 2010
Date accepted: October 2010
10 HEIDE-JØRGENSEN et al.: HUMPBACK WHALES IN WEST GREENLAND
Fig. 1a. Strata, survey lines and sightings (incl. off effort sightings) of
humpback whales in 1984.
Fig. 1b. Strata, survey lines and sightings (incl. off effort sightings) of
humpback whales in 1985.
Fig. 1c. Strata, survey lines and sightings (incl. off effort sightings) of
humpback whales in 1987.
Fig. 1d. Strata, survey lines and sightings (incl. off effort sightings) of
humpback whales in 1988.
J. CETACEAN RES. MANAGE. 12(1): 1–14, 2012 11
Fig. 1g. Strata, survey lines and sightings (incl. off effort sightings) of
humpback whales in 2005.
Fig. 1h. Strata, survey lines and sightings (incl. off effort sightings) of
humpback whales in 2007. Note that stratum 14 is inside coastal fjords.
Fig. 1e. Strata, survey lines and sightings (incl. off effort sightings) of
humpback whales in 1989.
Fig. 1f. Strata, survey lines and sightings (incl. off effort sightings) of
humpback whales in 1993.
12 HEIDE-JØRGENSEN et al.: HUMPBACK WHALES IN WEST GREENLAND
Fig. 2e. Distribution of humpback whale sightings at various distances from
the trackline during the survey in 2007. Data has been fitted to the hazard
rate function and the fitted curve shows the expected number of sightings.
The effective search width was 311m (CV = 0.19).
Fig. 2a. Distribution of humpback whale sightings at various distances from
the trackline during the surveys in 1987–89 with a left truncation at 200m
to allow the detection function to be applied to the surveys in 1984 and
1985 that used flat windows instead of the bubble windows that were
used in subsequent surveys. Data has been fitted to the half-normal model
and the fitted curve shows the expected number of sightings. The
sightings were truncated at 1,500m and the effective search width was
587m (CV = 0.37).
Fig. 2b. Distribution of humpback whale sightings at various distances from
the trackline during the surveys in 1987–89. Data has been fitted to the
half-normal model and the fitted curve shows the expected number of
sightings. The sightings were truncated at 1,500m and the effective search
width was 708m (CV = 0.20).
Fig. 2c. Distribution of humpback whale sightings at various distances from
the trackline during the survey in 1993. Data has been fitted to the half-
normal model and the fitted curve shows the expected number of
sightings. The sightings were truncated at 1500 m and the effective search
width was 503m (CV = 0.43).
Fig. 2d. Distribution of humpback whale sightings at various distances from
the trackline during the survey in 2005. Data has been fitted to the hazard
rate function and the fitted curve shows the expected number of sightings.
The effective search width was 1,506m (CV = 0.17) (see also Heide-
Jørgensen et al., 2008).
J. CETACEAN RES. MANAGE. 12(1): 1–14, 2012 13
Fig. 3. Trends in relative abundance of humpback whales in West Greenland
1982–2007. The exponential growth model is fitted to the estimates from
the aerial surveys. Details of the three abundance options from the ship-
based survey in 2005 are given in Heide-Jørgensen et al. (2007).
Fig. 4. Detection function plots for the MRDS analyses. Duplicate detections are indicated in the shaded areas; as
a number in the top plots and as a proportion in the middle plots. The points are the probability of detection for
each sighting given its perpendicular distance. The lines are the fitted models (in the pooled detection plot, the
line is a smooth function fitted to the points).
14 HEIDE-JØRGENSEN et al.: HUMPBACK WHALES IN WEST GREENLAND
Fig. 5. Estimation of the positive bias in instantaneous availability correction
factors compared to correction based on the probability of detecting a
whale given surface-dive patterns with 42% of time at surface and
average observation times of 2, 3.1 and 7 seconds.
Fig. 6. Changes in lambda (y-axis) due to changes in different life history traits (x-axis). Base values used for the life history
traits are not changed: Age of first parturition = 6, Fertility = 0.21, Calf survival = 0.805, Juvenile survival = 0.894,
Adult survival = 0.984, Max age = 100 years. Based on Barlow and Clapham (1997), Clapham (1992), Gabriele et al.
(2001) and Mizroch et al. (2004).