Biological parameters in a declining population of narwhals (Monodon monoceros) in Scoresby Sound, Southeast Greenland

,


Introduction
The narwhal (Monodon monoceros Linnaeus, 1758) is an important game animal in Greenland and is hunted by Greenlanders almost throughout its distribution (Hobbs et al. 2019).The main hunting product is the highly prized mattak (skin) of the whales, which reached a retail price of ~$75/kg in 2021 (E.Garde, personal observation, November 2021).The male tusk is also of economic value, and the meat is sometimes used for local consumption.The large demand and high prices for narwhal products makes narwhal by far the economically most valuable game animal in Greenland and is a significant cash income for the hunters (Flora et al. 2018;Nielsen and Meilby 2013).In addition to the economic value, the hunt of narwhals is also culturally significant with traditional utilization and trade through centuries (Reeves and Heide-Jørgensen 1994).The North Atlantic Marine Mammal Commission (NAMMCO) and the Canada-Greenland Joint Commission of the Conservation and Management of Narwhals and Belugas (JCNB) assess the sustainability of the hunt.The hunt in Greenland is managed by the Greenland Government and quotas in Southeast Greenland have regulated the narwhal hunt since 2011.
In East Greenland, narwhal hunting takes place from the communities of Tasiilaq and Ittoqqortoormiit located in Southeast Greenland (Fig. 1; Garde et al. 2019).No narwhal hunting takes place in Northeast Greenland (north of Scoresby Sound).The narwhal population that is available to the hunters from Ittoqqortoormiit spends the summer months traversing the Scoresby Sound fjord system and winters off the East Greenland coast, on the shelf southeast of the fjord mouth (Heide-Jørgensen et al. 2015).When the ice breaks from late June to early July, the whales enter Scoresby Sound along the southern coastline and stay in the large fjord system until the ice forms again in October or early November.Data from narwhals tagged with satellite transmitters show a consistent migration pattern between summer and winter grounds separating this population from others on the approximately 3000 km long coastline of East Greenland (Heide-Jørgensen et al. 2020a).A decreasing trend in abundance during the past decade has been identified for the narwhal population in Scoresby Sound (hereinafter the Scoresby Sound population) by the NAMMCO Scientific Committee based on a series of aerial surveys (NAMMCO 2019).The abundance dropped from 1991 narwhals (95% confidence interval (CI): 709-5590) in 2008 to 421 ) in 2016 (NAMMCO 2020).The most apparent reason for this decline is the level of catches that since the introduction of quotas by the government of Greenland in 2011 and to 2019 have ranged between 30 and 74 individuals annually with a mean of 50 (landed animals not including struck-but-lost whales; Garde et al. 2019; Table 1).From 1993 to 2010, before the introduction of quotas, the catches ranged from 6 to 93 with a mean annual catch of 30 narwhals.In 2017, the NAMMCO Scientific Committee recommended a catch level of not more than 10 narwhals (NAMMCO 2017) and in 2019 this was reduced to a recommendation of a moratorium on narwhal catches, implying that any continued catch level would be unsustainable (NAMMCO 2019).Despite this recommendation, the quota for 2020, 2021, and 2022 was set to 40, 25, and 20 whales (naalakkersuisut.gl),respectively, jeopardizing the long-term existence of this isolated population and the prospects of continued harvest of narwhals in the local communities (NAMMCO 2019).It is vital that the scientific advice for regional populations of narwhals is accepted and populations managed responsibly (Heide-Jørgensen et al. 2020b), especially because there is evidence for a lack of strong recovery in heavily exploited odontocete populations (Wade et al. 2012).A low resilience to overexploitation and a lack of recovery is explained by odontocete life history; a relatively old age at first reproduction and low calving rate.Overexploitation can lead to social disruption, fragmentation of social units, and loss of key individuals and cultural knowledge, which can result in a decrease in birth rates because of a deficit of adult females and (or) males and disruption of mating systems (see review by Wade et al. 2012).In addition, narwhals show extreme site fidelity to their summer grounds and the limited plasticity in movements makes it unlikely that they can re-colonize areas that they have been extirpated from (Heide-Jørgensen et al. 2015, 2020b).
Alongside hunting, the Scoresby Sound population may also be affected by ongoing changes in the marine environment.Suitable habitat for the whales seems to be declining as sea surface temperatures rise (Louis et al. 2020;Chambault et al. 2020).The Scoresby Sound population have recently been shown to forage in a narrow temperature range and a continued increase in sea water temperatures could reduce the feeding habitat available for the whales (Heide-Jørgensen et al. 2020a).Also, altered prey species availability, competition from migratory species, and potentially new diseases could render the current habitat less suitable for these cold-adapted whales (Chambault et al. 2020).A consequence of such accumulated effects caused by climate change could be population displacement Fig. 1.Map of Scoresby Sound fjord system with community (star) and locality (dot) names mentioned in the text (Mikkelsen et al. 2005).Numbers of biological samples collected at the different localities are shown in parentheses (n = 158).The location of Scoresby Sound fjord system (grey square), Kangerlussuaq fjord (red polygon), and Tasiilaq (red star) is shown on the overview map of Greenland in the inset.Map projection: WGS 84/UTM zone 24N.towards higher latitudes with colder sea water temperatures (Louis et al. 2020) or affect the body condition of the whales as a result of poor nutrition (Burek et al. 2008;Laidre et al. 2008).Even though a rise in sea water temperatures in Southeast Greenland have been documented (Alexander et al. 2018), the Scoresby sound population show no signs of either population displacement (Heide-Jørgensen et al, 2015, 2020a) or malnutrition (Heide-Jørgensen et al. 2014;Garde et al. 2015), and the anticipated effects of climate change on this small population of narwhals have not yet been detected.Concurrently with a reduction in sea ice caused by climate change, anthropogenic activities have been on the rise as an increasing number of cruise ships, sail boats, and other marine activities are present in the narwhal habitats (Reeves et al. 2014;Hauser et al. 2018).The anthropogenic activities in Scoresby Sound are, however, still at a low level with only a few tourist ships operating in the area from July through August and a minor number of local speed boats (<10) from Ittoqqortoormiit that are present in the fjord during the open water season.Narwhals are skittish animals, known to avoid humans, and disturbance caused by an increasing vessel fleet or other human activities, e.g., oil and gas exploration, can potentially have serious negative effects on the whales, as shown for various marine mammals including the narwhals (Richardson et al. 1995;Booth et al. 2020;Heide-Jørgensen et al. 2021;Tervo et al. 2021).Continued monitoring of life history parameters and population dynamics of this rapidly declining population is essential for future assessments of the consequences of these accumulated threats.
In this study, we estimated life history parameters and investigated the biological effects of a high harvest pressure on a small and isolated population of narwhals in Scoresby Sound, Southeast Greenland.We hypothesize that the level of exploitation is reflected in temporal changes in vital biological parameters as sex and age distributions, somatic growth, and reproduction.We examined this using data collected over a decade: biological samples from the Inuit hunt of narwhals (2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018)(2019), somatic measurements from satellite-tagged narwhals (2010-2019), and official hunters' records (2008-2019).The relationship between somatic parameters and narwhal swimming speed was examined to explain the consequences of the exploitation on the sex and age class composition of the population.

Sampling
Biological samples and sampling locations Biological samples and measurements from 158 narwhals (n = 64 females (41%); n = 94 males; Table 1) harvested by local Inuit hunters in Southeast Greenland were collected by the hunters and by researchers from the Greenland Institute of Natural Resources (GINR) from 2007 to 2019.Samples collected included eyes for age estimation, reproductive organs (ovaries and testes), hearts, and stomach contents.Measurements of the whales included body length, body mass, length of tusk, fluke width, circumference, and heart mass.Information on pregnancy was also recorded.Standard body length was measured in a straight line alongside the whale from the tip of the lower jaw to the notch of the tail flukes (Heide-Jørgensen and Teilmann 1994) and tusk length was measured in situ from the front edge of the upper jaw where the tusk protrudes to the tip of the tusk (Garde et al. 2015).Body mass of the whale was measured in the field, using a specially designed lift (from 2012 to 2019; Fig. 2).The maximum circumference of the whale was found by measuring the distance from the ventral side of the whale at the line of the umbilicus to the dorsal ridge and the measure was multiplied by two to get the full circumference.The heart was weighed on a scale on site.Samples collected from 2007 to 2010 (n = 88) were previously presented by Garde et al. (2015) and re-examined here to construct a time series from 2007 to 2019 (n = 70; 2011-2019).The majority of samples (58%; n = 92) were collected at Hjørnedal, a location within the Scoresby Sound fjord used as a hunting ground and camp site by local hunters from the settlement of Ittoqqortoormiit (Fig. 1).This is also the location of a GINR field station established in 2010.Another 39 samples (25%) were collected by hunters at unspecified locations within the large Scoresby Sound fjord system.The remaining samples were collected at Gåsefjord (n = 1), Syd Kap (n = 2), Kap Tobin (n = 6), three locations south of the entrance to Scoresby Sound (n = 16; Stewart Island, Turner Island, and Rømer fjord), and Kangerlussuaq (n = 2), a fjord further south on the East Greenland coast (Fig. 1).A young male taken in the Inuit hunt is being weighed.
Measurements from satellite-tagged whales and information from hunters' reports Information on sex (n = 69; 30% females) and measurements of body (n = 65; 29% females) and tusk length (n = 45) were obtained from whales tagged with satellite transmitters (n = 70) at the field camp in Hjørnedal from 2010 to 2019 (Heide-Jørgensen et al. 2015, 2020a).
For the period 2008-2019, information on sex of the whales (n = 420) and pregnancy (n = 112) was attained from the official catch reporting system, the Special Reports (Garde et al. 2019), which includes the hunter's own information on each narwhal taken (Garde et al. 2019).

Age and growth analyses
Age estimation and age structure Age estimation was done using eye lens nuclei and the aspartic acid racemization (AAR) technique.Laboratory work followed procedures described by Garde et al. (2007Garde et al. ( , 2010Garde et al. ( , 2012) ) and AAR age was estimated using the age equation specific for narwhals from Garde et al. (2015).Age structures were constructed using the AAR ages in years.

Somatic growth
Growth curves were fitted to the age-at-length and age-at-tusk length data using the von Bertalanffy growth model (eq.1; Table 2) for females and males separately: where L(t) is the length at age t, L max is the asymptotic length, L 0 is the length at time 0 (at birth), K is the growth rate constant, and t is time (age).
Body mass was fitted to the AAR ages using the Gompertz growth model (eq.2): where W(t) is the mass at age t, M max is the asymptotic mass, K is the growth rate constant, and t is time (age).Asymptotic body mass was found for the males.A sample size of five females was insufficient for prediction using the model.The Gompertz model was chosen for estimation of the mass-age relationship because of higher F values compared with the von Bertalanffy model (Garde et al. 2015).
The relation between body length and body mass was expressed by an exponential growth curve: where a is the intercept, b is the exponential constant, and x is the body length.
Prediction of body mass based on length and maximum circumference was estimated as follows: Log 10 ðmassÞ = a + b 1 log 10 ðlengthÞ + b 2 log 10 ðmaximum circumferenceÞ (4) where a is the intercept, and b 1 and b 2 are coefficients.

Sexual maturity, pregnancy, and menopause in female narwhals
Reproductive organs (ovaries) were collected for the purpose of estimating age (years), body length (cm) and body mass (kg) at sexual maturity, age at first pregnancy, and pregnancy rate.Ovaries were visually investigated by external examination for the presence of ovarian corpora (corpora lutea (CLs) and corpora albicans (CAs)), which are the yellow body present at pregnancy and the scars after pregnancies, respectively (Perrin and Donovan 1984).Garde et al. (2015) found no statistical difference in number of ovarian corpora of narwhal ovaries (n = 26) that had been examined both externally and internally after fixation in formaldehyde and CAs were therefore only detected as surface scars on the ovaries in this study.For most females (n = 41), both ovaries were collected, while only one ovary was attained for seven females.In these seven cases, the estimated number of pregnancies, based on the number of ovarian corpora in one ovary instead of two, were therefore a minimum.The ovaries were not collected from two females, but one had a fetus and the other a small calf, confirming that both females had been pregnant at least once."Several" CAs were recorded in the field report for a single female, which was subsequently interpretated as five CAs.In general, CAs regresses with age and in older females it is expected that a proportion of the CAs have disappeared completely due to this regression (Dabin et al. 2008).This results in a minimum estimate of CAs in older females.The females were defined as immature based on absence of ovarian corpora, mature females had either large mature follicles (≥10 mm in diameter) or at least one corpora, and pregnant females either had a fetus (and then also a CL) or only a CL if the fetus was not detected (Garde et al. 2015).Four females were determined as being mature based solely on their body length (Garde et al. 2015).Age (years; n = 38) and body length (cm; n = 50) at sexual maturity was predicted from age-and length-frequency distributions, when 50% of females were mature.Based on information on body mass from 16 females, an approximate body mass at sexual maturity was assessed.
Pregnancy rates were estimated as the proportion of pregnant (and hence mature) females to all mature females collected during 2007-2017.No samples from females were collected in 2018 and 2019.Menopause or post-reproductive lifespan, defined as common and prolonged female survival after the cessation of reproduction (Ellis et al. 2018), was evaluated based on ovarian activity for four females >60 years of age.

Male sexual maturity
Testes were collected for the purpose of estimating age (years), length of body (cm) and tusk (cm), and body mass (kg) at sexual maturity.Based on previous studies of male narwhal reproduction, it was assumed that males with testes mass ≤100 g were immature, >100-400 g were maturing, and >400 g were fully mature (Garde et al. 2015;Hay and Mansfield 1984).Age (years; n = 35), body length (cm; n = 37), and tusk length (cm; n = 36) at sexual maturity was predicted from age-and length-frequency distributions, when 50% of males were mature.Based on information on body length from seven males, an approximate body mass at sexual maturity was assessed.
Analyses of the temporal trend in growth and sex distribution Data for the analyses of the trend in growth and sex distribution originates from three sources: (1) biological samples from the hunt of narwhals collected by local hunters and researchers from the GINR, (2) data from animals tagged with satellite transmitters at the camp in Hjørnedal, and (3) from the hunters' Special Reports.The models only included data from landed whales within the Scoresby Sound fjord between 1 July to 31 December from 2008 to 2019.These catches are recognized as belonging to whales from the Scoresby Sound narwhal population (NAMMCO 2020).Whales landed in the spring prior to 1 July possibly belong to a putative stock summering north of Scoresby Sound and thus are not part of the stock hunted in summer in Scoresby Sound (NAMMCO 2021).
The effect of year on growth parameters (body length, tusk length, and body mass), sex distribution, and pregnancy were investigated using regression models.The fit of the response distribution was assessed using diagnostics plots and Akaike's information criterion (AIC) from distribution fit analysis output (function descdist, R package "fitdistrplus"; Delignette-Muller and Dutang 2015; R Core Team 2019), which uses a combination of kurtosis and squared skewness of the sample to locate an appropriate response distribution.No other model selection was performed due to the focus on temporal trends and lack of other explanatory variables.
The trend on body length was modelled as a beta distribution with a logit link, using body length (converted to range between 0 and 1) as the response and year as explanatory variable (function betareg, package "betareg").The temporal trend in body mass was modelled with a generalized linear model as a Gaussian response with an identity link using body mass (kg) as the response and year as the explanatory variable.Separate models were performed for males and females.The effect of year (explanatory variable) on tusk length was modelled as a Gaussian distribution with an identity link, where the tusk length (response variable) was square root transformed prior to the analysis to account for positive skewness in the distribution of data.Data on body length, body mass, and tusk length were from hunted whales and whales tagged with satellite transmitters collected by GINR.
The effect of year on the probability of catching a female, was modelled as a binomial response with a log link (function glm, base package) using occurrence of females as the response and year and data source (with two levels: GINR data (from biological samples and tagged narwhals) and the Special Reports), with an interaction term in between, as the explanatory variables.A similar logistic model was used to investigate the temporal trend in the probability of pregnant females, where the occurrence of pregnant females was used as the response and year and data source with an interaction term in between, were entered as the explanatory variables.This latter model included only sexually mature females.
Fluke width and heart mass in relation to body length and swimming speed The relationships between body length and fluke width and between body length and heart mass for the whales were assessed by linear regressions for females and males, respectively.These three somatic parameters (body length, fluke width, heart mass) are known indicators of swimming speed in other cetaceans (Fish 1998;Fish et al. 2014), but the relationship between these parameters and swimming speed has not been evaluated in narwhals.Linear regressions between fluke width and swimming speed were performed to show this relationship.
Horizontal speed was calculated for 13 narwhals tagged with Fastloc-GPS satellite transmitters in Scoresby Sound in 2015, 2017, and 2018(Heide-Jørgensen et al. 2015, 2021).A total of 51 829 distance measurements between positions with associated times were obtained from the whales from August through March.The Fastloc-GPS positions are considered fairly accurate (Tomkiewicz et al. 2010), and the speed between positions represents the horizontal movement of the whales and does not include the actual speed while diving.To minimize the effect of nonlinear underwater movements, only positions separated by less than 30 min were included.Furthermore, unrealistic speeds exceeding 5 m/s were excluded (1.75% of the data).

Narwhal diet
Stomach contents from narwhals were collected, inspected on site (Hjørnedal), and recorded to evaluate narwhal diet during summer in Scoresby Sound (2010-2019).The contents from 78% of the stomachs were drained and weighed.It was assessed whether a narwhal had been feeding recently by the amount and decomposition of the food items in the stomach.

Growth of body and tusk
Using the von Bertalanffy model, asymptotic body lengths of females (n = 49) and males (n = 66) were 406 cm and 463 cm, respectively (Table 2; Figs.4A, 4B).Asymptotic length was attained at ages of ∼23 years for females and ∼45 years for males.External length of the tusk (n = 62) reached asymptotic length at 197 cm and ∼56 years (Table 2; Fig. 4C).Asymptotic body mass predicted from the Gompertz model was 1496 kg for males (n = 23) attained at ages ∼60 years (Table 2; Fig. 4D).Body mass, body length, and age collected for five females ranged from 880 to 1000 kg, from 390 to 420 cm, and 34.6 years ± 2.8 SD to 107.7 years ± 7.7 SD, respectively.Body mass and length from 13 females with no matching age estimates ranged from 270 to 1120 kg and from 265 to 410 cm, respectively.During the study period, the largest female measured was 441 cm in length ( 2007).The largest male measured was 501 cm in length with a tusk of 193 cm (2016; Table 2).The maximum body mass for a male was 1560 kg (2019).The longest tusk measured 240 cm in length (2007).The youngest male with a tusk measurement was 3 years ± 1.6 SD and had a tusk of 7.5 cm in length and a body length of 283 cm.Two other males of 2 years of age with body lengths of 226 and 266 cm, had no information on tusk lengths probably because the tusks had not yet erupted.

Reproduction
Female sexual maturity at age, length, and mass Ovarian corpora were not present in the ovaries of females <9 years (Supplementary Fig. S1 1 ).Small follicles (>20; <1 mm) but no ovarian corpora were present in the ovaries of one female of 9.0 years ± 1.63 SD, indicating that this female was sexually maturing but had not yet reached full sexual maturity.Based on the presence of large follicles but no ovarian corpora, three females from 9.9 years ± 1.66 SD to 10.7 years ± 1.67 SD were assessed as being sexually mature but with no past pregnancies.The youngest pregnant female sampled was 9.5 years ± 1.64 SD and carried a 15 cm long fetus.No other ovarian corpora or follicles were present in the ovaries of this female, indicating that it was the first pregnancy.The second youngest pregnant female sampled was 12.7 years ± 1.73 SD, carrying a 24 cm long fetus, and besides a CL, one CA was present, indicating one previous pregnancy.As only one ovary was collected and examined from this female, the female could potentially have been pregnant more than twice.Fifty percent of the sampled females were sexually mature at 9 years and a body length of 340 cm (Figs.5A, 5B).Three females with body masses ≤500 kg were sexually immature and 13 females with body masses ≥610 kg were mature.It was assessed that females reach sexual maturity at a body mass ∼600 kg (Table 2).the female but the production of milk suggested the presence of a calf.One (2 mm) and six (1-5 mm) small follicles present in the ovaries of two females of 62.8 years ± 4.6 SD and 84.2 years ± 6.0 SD, respectively, implied reproductive activity.No current ovarian activity was recorded in a female of 83.7 years ± 6.0 SD, indicating menopause.No ovaries or other reproductive information was available for the oldest female in the sample of 107.7 years ± 7.7 SD.
Male sexual maturity at age, body and tusk length, and mass Males with testes mass ≤100 g (n = 18) were considered immature (Supplementary Fig. S1 1 ; Garde et al. 2015).They were all ≤12.5 years ± 1.7 SD of age, except for one male of 15.4 years ± 1.8 SD with testes mass of 66.1 g.Testes mass increased with age, from 100 g to ∼300 g corresponding to ages from 13.4 years ± 1.8 SD to 20.6 years ± 2.0 SD (n = 8), and once males reached ages >25 years, testes were >445 g (n = 9).One young male of 17.0 years ± 1.9 SD had a high testes mass of 480 g, and one old male of 79.7 years ± 5.7 SD had a low testes mass of 131.5 g.Based on information on testes mass and age in this study, male narwhals were assessed to be sexually immature until they reached ages of ∼13-15 years, after which they entered full sexual maturity at ages from ∼17-20 years.Fifty percent of males were found to be sexually mature at an age of 16 years and a body length of 380 cm (Figs.5C, 5D; Table 2).Of seven males with information on both body mass and testes mass, one was immature (350 kg; 14%), one was maturing (900 kg; 14%), and five were sexually mature (≥900 kg; 71%).Based on this information, it was assessed that males reach sexual maturity at ∼900 kg.The relationship between tusk length (cm) and testes mass (g; n = 36) Fig. 5. Age (years) and body length (cm) frequency distributions of mature females (A, B) and males (C, D).The figure shows that 50% of females were sexually mature at the age of 9 (A; n = 38) and a body length of 340 cm (B; n = 50) and that 50% of males were sexually mature at the age of 16 (C; n = 35) and a body length of 380 cm (D; n = 37).Solid black circles are immature and mature whales (A-D), and black Xs are maturing males (C, D).Sample sizes for females are 8 immature and 30 mature (A; age) and 11 immature and 39 mature (B; body length).Class intervals for females are 5 years and 20 cm body length.Sample sizes for males are 17 immature, 7 maturing, and 11 mature (C; age), and 15 immature, 8 maturing, and 14 mature (D; body length).Class intervals for males are 5 years and 50 cm body length.
showed that tusk length was ≤80 cm when testes mass was ≤100 g (n = 14; with body lengths ranging from 262 to 373 cm).When tusks reached lengths >120 cm, testes mass was ≥447 g (n = 12; with body lengths from 370 to 480 cm), except for one male with a tusk length of 123 cm and low testes mass of 293 g.Fifty percent of males were sexually mature when tusk length reached 100 cm.
The probability of catching a female decreased significantly (n = 636; p = 0.0015) from 2007 to 2019 (Fig. 7A).The two data sets, the data collected by the GINR (from biological samples and tagged animals) and the Special Reports from the hunters, were both consistent in showing the decreasing trend with no significant difference between the trends (p = 0.22).The two data sets were also consistent in regard to probability of catching a pregnant female (n = 158; p = 0.078) showing a decreasing, however not significant, trend with time (p = 0.61; Fig. 7B).

Swimming speed as a function of fluke width
Fluke width and heart mass increased linearly with body length and males reached larger fluke widths than females (Fig. 8A).The mean horizontal swimming speed of the whales for positions separated by <30 min was correlated with the fluke width of the whales, indicating that larger whales generally attained larger speeds when swimming (Fig. 8B).

Diet of the narwhals from Scoresby Sound
The diet of 32 narwhals (12 females; 20 males; 2011-2019) from Scoresby Sound consisted mainly of squids (Gonatus spp.), but also remains of polar cod (Boreogadus saida (Lepechin, 1774)) and Arctic cod (Arctogadus glacialis (Peters, 1872)) were regularly found in the narwhal stomachs, whereas remains of Greenland halibut (Reinhardtius hippoglossoides (Walbaum, 1792)) and northern shrimps (Pandalus borealis Krøyer, 1838) was occasionally detected.A small fish was found in the stomach of a female in 2014, which could have been a capelin (Mallotus villosus (Müller, 1776)).Capelins were found in the stomachs of four males in 2016 (n = 2) and 2017 (n = 2).The quantity of stomach contents ranged for females (n = 8) from 500 to 4400 g and for males (n = 17) from 0 to 3500 g.The most frequent food item was squid; some stomachs were full of fresh squids and a few whales had fresh squids in their mouth when landed.Three (two females and one male) out of five immature whales had fresh food remains in their stomachs containing the same prey items as the adults.The two remaining immature whales, both males, had otoliths from polar cod in their stomachs and one also had jaws from squids.The oldest female sampled had squids in the stomach.

Discussion
Estimation of life history parameters, e.g., rates of growth and reproduction, of exploited populations of wildlife provide diagnostic information about body condition and changes in density-dependent parameters that are essential to the management and conservation of species (Chivers 2009;Murphy et al. 2009).Growth in length and especially mass-at-length are robust indicators of body condition, ecosystem changes, and changes in carrying capacity (Castrillon and Nash 2020).In this study, there were no signs of reduced growth or deteriorating body condition of the narwhals in Scoresby Sound (Murphy et al. 2009).There were too few samples to detect changes in age at sexual maturity for both sexes and values were similar to estimates from Garde et al. (2015).
A decreasing trend, although not significant, in pregnant females of the Scoresby Sound narwhals was observed for the period 2007-2019 coinciding with a significant and steady decrease of females.These observations were based on data from both biological samples from a minor portion of the population and from the Special Reports covering all landed narwhals, which comprise a substantial proportion of the small Scoresby Sound population.In addition to the decreasing trend of females and pregnant females, there was an increasing under-representation of young males in the catches.All evidence suggests that there is a ∼1:1 sex ratio at birth of narwhal calves and population models predict that about 1/3 of the population should be immature whales in a stable population (Hay 1984).In the present study, the catches in recent years were mainly composed of old males.We have no evidence that old males were preferentially targeted by the hunters in recent years.On the contrary, if the harvest was selective, older and bigger animals with large tusks would have been preferred by the hunters throughout the study period because they represent a higher economic benefit but instead, young animals were taken earlier in the exploitation history implying opportunistic hunt instead of a selective hunt.Furthermore, the hunter gains ∼$3900 for the mattak of an average-sized adult narwhal (130 kg mattak at ∼$30/kg), which is considerably more than they obtain for the less valuable tusk sold for ∼$200/kg (a tusk ∼200-240 cm long weighs ∼3-6 kg; Garde et al. 2012).Although we have biological samples from only a portion of the total harvest, to compensate for low sample sizes we also analyse data from the hunters' own records, the Special The narwhals in Hjørnedal feed regularly on squid, shrimps, polar cod, and Arctic cod, but squid seems to be the most common prey item during the summer months.Since 2016, capelin have become increasingly common in the Scoresby Sound fjord system, where it has been observed and caught in the fjord by the local people (E.Garde, personal observation, August 2019).Capelin was found in the stomachs of narwhals in 2016 and 2017, and perhaps as early as in 2014, but were not among the prey items collected from 2010-2013and in 2015 (this study;Heide-Jørgensen et al. 2014).Nothing is known about prey densities in the East Greenland waters where the narwhals are found, but the stomach contents does not support a major shift to new prey items.No other habitat changes can be identified as drivers for the observed changes in life history parameters.
The asymptotic body lengths predicted in this study are similar to previous estimations of asymptotic lengths in narwhals, whereas the predicted tusk length in this study are somewhat higher compared with previous estimates, indicating that the sampled males are larger in recent years (Garde et al. 2015).Altogether, the body condition seems to be unchanged or even improved in the recent years and nothing suggests that the narwhals in East Greenland are nutritionally stressed.The age structures and growth patterns for this population indicate that the catches of males include increasingly older and larger animals and that a decreasing proportion of females and especially pregnant females are included in the catches.
There was no hunting of narwhals in Scoresby Sound before the establishment of the main hamlet (Ittoqqortoormiit) in 1925.Larger catches (>10 whales) were uncommon before 1979 but with increasing prices for the hunting products, catches have increased substantially.The official catch records only report on landed animals and do not include whales that were struck-but-lost before they could be retrieved, which is an additional 15%-20% of the hunt (Garde et al. 2019).The hunt for narwhals in Scoresby Sound is primarily an opportunistic open water hunt from fast moving speedboats with outboard engines, where hunters will target any spotted whale, or a hunt where single or groups of whales are driven into ∼50 m long set nets deployed from shore.These methods have been used for decades and nothing indicates that the hunters have changed their hunting practices in recent years (Garde et al. 2019).The most parsimonious explanation for the declining proportion of females and young males is the effect of hunting, where these age and sex categories are the easiest to catch.The most likely reason for this is the general slower mean swimming speed as documented for smaller whales, but other factors like experience of older whales, poor condition of females due to the nutritional demands of their calves (Hay 1984), and pregnancy, which reduces speed as shown for bottlenose dolphins (Tursiops truncatus (Montagu, 1821); Noren et al. 2011), may also play a role.The whales usually react to being chased, e.g., by killer whales or boats, by long dives or by moving towards shore (Breed et al. 2017;Heide-Jørgensen et al. 2021).Larger animals are capable of longer dives (Schreer and Kovacs 1997;Kooyman 1989) and faster swimming speeds than smaller conspecifics, as also documented here for the narwhals in Scoresby Sound.Larger animals may also be more experiencedlarge animals have been observed to change direction fast and often, and to dive underneath hunters' boats as a response to being chased (M.H.S. Sinding, personal observation, August 2019).Less experienced smaller whales will seek shelter along the coast, where they are more easily targeted by the hunters that also can drive the coastal whales towards whale nets.The effect is that older, larger and more experienced whales are more likely to escape the hunt, whereas younger, less experienced whales are the easiest to catch and will be first to be targeted by the hunt.Females are generally smaller than males and are sometimes accompanied by calves, which is probably slowing them down in a hunting situation, and they are therefore more likely to be caught.The effects are seen in the changes in the age and sex structure over the period with intensive hunting.The increased age of the females may have reduced the pregnancy rates either because older females do not give birth as frequently as younger females or because of an Allee effect in a small declining population, where reproductively active females are less likely to encounter males during the ovulation period (Drake et al. 2019).The combination of Allee effect and human-caused mortality, e.g., hunting, have the potential to substantially increase extinction risk for critically depleted populations (Wade and Slooten, preprint 2020).
In conclusion, the effects of continued overharvest of a small local population of narwhals are shown in the population dynamics as a decreased proportion of females, a decreasing trend in pregnancy rate, an overrepresentation of old males and a lack of calves and juveniles.It is hypothesized that the disappearance of young whales and females in the catches is a result of opportunistic hunting practices that initially target any spotted whale but eventually include the slow and less experienced whales.This is supported by the relation between speed and size that predicts that bigger whales are faster and thereby more difficult to catch.Not all trends are statistically significant, but the population implications of the trends are in any case detrimental.For exploited species and populations that are notoriously difficult to study and monitor, like High Arctic narwhals, even minor signals in behavior and life history parameters need to be taken seriously to avoid irreversible population declines.By the time statistically robust data are acquired the populations may be gone.

Note added in proof
The regression analyses of the temporal trends in growth, sex distribution, and pregnancy were performed again between acceptance of the manuscript (June 2021) and receiving the proofs (late December 2021).This was due to a change from "catch location" to "catch date" as a criterion for separating the Scoresby Sound narwhal stock and the putative northern stock (Table 1; NAMMCO 2021).This resulted in a larger data set; however, the changes did not affect the major conclusions.

Fig. 2 .
Fig. 2. A lift and scale specially designed for measuring the body mass of narwhals at the camp site in Hjørnedal.A young male taken in the Inuit hunt is being weighed.

Fig. 6 .
Fig. 6.Model outputs for the effect of year on body length (A; n = 75 females; n = 133 males), body mass (B; n = 128 females; n = 44 males), and tusk length (C; n = 128 males).Females are shown in red and males in black.Numbers indicate sample sizes of females (red) and males (black) per year; no number indicates that samples were not available.The solid line indicates a significant effect of year (p ≤ 0.05), the dashed line a nonsignificant effect (p > 0.05), and the shaded area marks the 95% confidence interval.

Fig. 7 .
Fig.7.Model outputs for the probability of catching a female (A; n = 636) and for the probability of catching a pregnant female (B; n = 158) as a function of year separately for the two sources of data: Greenland Institute of Natural Resources (GINR; blue) and hunters' Special Reports (orange).The solid line indicates a significant effect of year (p < 0.01), the dashed line indicates a nonsignificant effect (p > 0.05), and the shaded area marks the 95% confidence interval.There was no significant difference between the two data sets GINR and hunters' Special Reports in either case (p > 0.05).Blue numbers (n) are GINR data and orange numbers (n) are Special Reports data per year.No number indicates no available samples.

Table 1 .
(Garde et al. 2019)narwhal catches in the Scoresby Sound area retrieved from the Special Reports(Garde et al. 2019), biological samples collected from narwhals included in this study, females (%) of the biological samples, and biological samples (%) of the official narwhal catch numbers.
Note: Catches (n) landed in the Ittoqqortoormiit management area between 1 July and 31 December belong to the Scoresby Sound stock, and catches landed from 1 January to 30 June belong to the putative northern stock (NAMMCO 2021).*Nocatches for 2007 were reported in the Special Reports.
Note: Length measurements are in centimetres, body mass in kilograms, and age estimates in years.Standard error is indicated in parentheses.