My research interests ranges from animal population ecology to responsible wildlife tourism, including: Monitoring of Icelandic seal populations and management advice; Seal diet studies and interactions with fishing industry; Studies on animal behaviour; Anthropogenic interactions with marine mammals, such as effects of tourism on marine mammal ecology; Transferring of scientific knowledge to society.
Skills and Expertise
- Reykjavík, Iceland
- Head of Department Icelandic Seal Center
- I am currently Head of the seal research at the Icelandic Seal Center. I am working on projects related to pinniped ecology and ethology, equine science and wildlife tourism.
Research Items (22)
A grey seal census was conducted by aerial survey during the pupping period of 2017. Pups were counted three to five times in the main grey seal pupping areas. In addition, areas where grey seal pups have occasionally been observed were surveyed once. The peak of the pupping period varied from 2 October (Frameyjar in Breiðafjörður) to 24 October (Strandir). Based on the estimated pup production (1452; 95% CI= 1385-1529), the total grey seal population size was estimated to be 6269 (95% CI= 5375-7181) animals. Breiðafjörður was the most important pupping area in Iceland, with a total of 845 (CI 95%= 807-887) pups, corresponding to 58% of the total estimated pup production in 2017. Other important pupping areas were the northwest coast (Strandir and Skagafjörður) and the south coast (Öræfi and the island Surtsey). The population was approximately 32% smaller than when the first census was conducted in 1982 with an approximate total exponential growth rate (rest) of -0.01. The population estimate for 2017 corresponds to an increase of 49% since the last census in 2012. However, trend analysis for the period 2005–2017 revealed no statistically significant trend for the total population size since the current population size is close to the estimated population size of 2008/9 and slightly larger than the estimate of 2005. In 2017 the population size was larger than the governmental management objective for the size of the grey seal population of 4100 animals. However, according to the Icelandic red list for threatened populations, which is based on criteria put forward by IUCN, the grey seal population should at its current level be considered as “Vulnerable”.
Understanding ecological relationships between humans and marine predators is crucial for the implementation of sustainable management practices. Comprehensive estimation of pinniped diet is essential for assessing interaction with fisheries and often has an important conservational value. Due to uncertainty regarding the accuracy of methods traditionally used to estimate harbour seal (Phoca vitulina) diet it is necessary to improve analysis methods. We investigated the diet of harbour seals hauling out in an estuary area in north-western Iceland between May and August of 2010 and 2011 by genetic (molecular) analysis of prey in faeces using DNA metabarcoding. The results were compared to previously published results from morphological analysis. Our results showed that species consumed were mainly sandeels (Ammodytes sp.), flatfishes (Pleuronectidae), gadoids (Gadidae), herring (Clupea harengus) and capelin (Mallotus villosus). The results from molecular and morphological analyses were similar in regards to important prey species, but species diversity was lower in the morphological analysis and 38% of the samples included prey items that were unidentifiable in the morphological analysis. Notably, despite Atlantic salmon (Salmo salar), brown trout (Salmo trutta) and Arctic char (Salvelinus alpinus) availability in the study area, neither of the methods found evidence of salmonids in the harbour seal diet. Recently, a severe decline has been observed in the Icelandic harbour seal population. Since the main reason for culling harbour seals in Iceland is to reduce predation on salmonids, findings presented in this paper have essential conservation implications and suggest that culling needs to be reassessed.
There are both positive and negative impacts on wildlife associated with wildlife tourism. In Sweden, the endangered Arctic fox is subject to a growing tourist interest. In the Helags mountain region there are guided Arctic fox safari tours that provide visitors with information about the Arctic fox. A survey of five separate groups of visitors in the region revealed that knowledge about the status of Arctic foxes and awareness of the behavioral guidelines for Arctic fox encounters improved after participation in a safari tour and with increasing Arctic fox interaction. We propose a schematic model summarizing the diverse ways in which wildlife tourism affects wildlife and their relative importance for conservation. The Arctic fox population in Sweden is small and sensitive to disturbance, but the positive impacts of Arctic fox tourism seem to compensate for the negative and contribute to their conservation under the current level of tourism pressure.
- Oct 2017
- The 13th International Conference on Responsible Tourism Destinations.
Seal watching as a form of wildlife tourism is becoming increasingly popular worldwide. Behavioral changes caused by the presence of tourists could lead to negative consequences for seal welfare and may affect reproduction and survival. Therefore, managing seal watching activities to ensure future protection and conservation is important. Codes of conduct or guidelines for how to behave around animals are one way to regulate wildlife watching and are often easier and quicker to implement than laws. Codes explaining the consequences for wildlife if the code is not followed appeal to the moral obligation of tourists and thereby increase incentives to act appropriately. This study focused on analyzing the content of codes of conducts for seal watching. Codes of conducts (n=33) accessible on the internet during the time of study were analysed. Results show that in many areas where seal watching occurs there are no regulations or guidelines. The content and detail of the codes varied and the information was often insufficient to offer adequate protection of seals. Few of the codes were developed in cooperation with scientists or stated that the content was based on research. Further, a majority of the codes did not explain the consequences for wildlife if the code was not followed. More research on seals and the tourists watching them is needed to better understand the effects of tourism and how disturbance could be minimised. Meanwhile, developing an international code of conduct (with local additions) built on existing knowledge in the field, could be one option to increase protection and ensure conservation of these animals. The results presented in this paper could assist the development of such a code of conduct.
The effectiveness of interpretive signage as a means of modifying visitor behaviour to reduce negative impacts on wildlife was tested empirically at a seal watching site on Vatnsnes peninsula in North West Iceland. From July to September 2014, the actions of 2440 visitors were observed and their behaviour recorded. To test the importance of how interpretive information is presented, signs with either ontological (instructions without explanation) or teleological (instructions with explanation) information were positioned along the path towards the site. A control group, to which no signs were provided, was also observed. Our results show that the majority of the tested behaviour was influenced when signs were present and that under some conditions teleological signs were more effective than ontological. The type of visitor group was found to significantly influence behaviour, with families having the most intrusive behaviour compared to singles, couples or other groups. The findings of this study contribute to a better understanding of how interpretative signage can modify tourist behaviour to facilitate sustainable wildlife tourism. The use of teleological signs for managing wildlife tourism activities is recommended because they are more effective than ontological signs in terms of modifying the general visitor behaviour. In addition, signage and other management strategies should address the different needs and responses relevant to the nature of the tourist group visiting the site. Special focus should be placed on families when signs are designed because this group type showed the highest probability of causing disturbance at the site.
It is of critical importance to identify factors that affect harbour seal haul-out patterns to improve the accuracy of harbour seal censuses. In this study, haul-out patterns of harbour seals (Phoca vitulina) were investigated during different conditions at several major haul-out sites on Vatnsnes peninsula, NW Iceland (65°40′N and 20°48′W), over the 2008–2011 period. A seasonal haul-out pattern was detected among the seals, with the maximum number of seals on land found in July for most of the sites. Analyses of data for harbour seals on Vatnsnes indicate that the main pupping period occurs during late May to the beginning of June and moulting during late July to early August. Abundance at the sites increased with rising air temperature and decreased with increased windspeed and rising tides. However, no evidence that precipitation or cloud cover affected haul-out behaviour of the seals was detected. The diel haul-out pattern was investigated thoroughly in one of the haul-out sites and the results underlined the effect of tidal flucturation, air temperature and wind direction on the haul-out behaviour of harbour seals in the area. Results from this study can be used to improve the survey design when estimating the population size of harbour seals in Iceland and applied more broadly to the study of haul-out behaviour of harbour seals.
The effect of harbour seal predation on salmonids has been frequently debated, although interactions between these species have never been thoroughly investigated in Icelandic waters prior to this study. We investigated the diet of harbour seals in a salmon estuary in NW Iceland between 2009 and 2011, using hard part analysis from collected faeces. No evidence of seal predation on salmonids was found in the study. The reconstructed weight and estimated energy content of prey species showed that flatfish was the most important species group, followed by Ammodytidae. The species group found in the highest proportion of samples during the three years combined was also Ammodytidae (45% of the samples). Ammodytidae, flatfishes and capelin dominated by numerical occurrences. However, inter- and intra-annual variation was found regarding the diet.
Landselir (Phoca vitulina) voru taldir í nokkrum helstu landselslátrum á Íslandi í júlí, ágúst og september 2014; á Vesturlandi, Vestfjörðum og Norðurlandi vestra. Samanburður var gerður með því að telja úr Cessna yfirþekjuflugvél á hefðbundinn hátt og með því að telja á myndum teknum úr ómönnuðu loftfari (flygildi). Einnig var flogið yfir Vatnsnes með þyrilvængju Landhelgisgæslunar og landselir taldir af stafrænum myndum sem teknar voru úr þyrlunni. Samanburður á hefðbundnum talningum úr flugvél og talningum úr flygildi leiða í ljós að lítil munur sé á marktækni talningargagnna, en aftur á móti teljum við öryggi talningarmanna vera meira við nokun flygilda samaborið við notkun Cessnuflugvélar. Ljóst er þó að notkun flygilda er mun tímafrekari og Cessna yfirtekju flugvél kemst yfir stærri talningarsvæði á styttra tíma. Þess vegna er líklega ekki sparnaður við notkun flygildi. Samanborið við niðurstöður talninganna 2011 þegar stofnstærðarmat var gert, er fjöldi landsela í þeim látrum sem skoðuð voru nú (2014) yfirleitt mun minni. Þrátt fyrir síminnkandi veiðar landsela á þessum sem öðrum svæðum á Íslandi benda þessar niðurstöður til þess að landselsstofninn hafi fækkað um 30% árlega á tímabilinu 2011 – 2014, en 2011 var stofnin metin til 11-12.000 dýr. Gæta skal að því, að til þess að hægt verði að segja til um ástand landselsstofnsins í heild á Íslandi verður að telja oftar en einu sinni í hverju látri og fara yfir alla ströndina eða svo gott sem. Í ár var þessi kostur hinsvegar ekki inni í myndinni vegna skorts á fjármagni. Þar sem vísbendingar eru um það að landselsstofninn hafi minnkað mikið og sé nú langt undir viðmiðunarmörkum sem miða við stofnstærð ársins 2006 er mikilvægt að meta stofnstærð hans árið 2015, en þá eru liðin 4 ár frá síðustu sambærilegu talningu. English abstract Harbour seals (Phoca vitulina) were counted in some of the largest haul-out sites in Iceland between July and September 2014; in the West Iceland, the West fiords and the North West of Iceland. To compare different methods, the seals were counted from a Cessna aircraft and by using an unmanned aerial vehicle (UAV). Harbour seals on the shores of Vatnsnes were in addition counted once from a helicopter in co-operation with the icelandic coast guard. When comparing the counting methods, it is clear that both the methods works equally well to collect data, although flying a UAV can be considered as safer for the counting personal compared with counting from a Cessna aircraft. Further, using a UAV is more time consuming and consequentally the Cessna aircraft can cover a bigger area in a shorter time. Hence, over all, using the Cessna aircraft to count harbour seals in Icelandic condition can not be considered more expensive than by using a UAV. The results show that compared with counts from the same areas in 2011, the number of seals had decreased. In 2011, the icelandic harbour seal population was estimated to 11-12.000 seals in total. Despite decreasing harbour seal hunting in Iceland, the results from the present study indicate an annual decrease of around 30% during the period 2011 to 2014. It is important to underline that to find out the status of the harbour seal population as a whole, all haul-out sites should be counted and to get a significant result, each site should preferably be counted three times. This was however not possible in 2014, due to financial restrictions. Since there is an indication of a sever decrease in the harbour seal population in Iceland, it is of great importance to conduct a harbour seal count covering the whole island in 2015, when four years has passed since the last population estimate was carried out.
Due to increasing interest in wildlife tourism, there is a growing need to consider the balance between use and protection of wildlife. Mutual exchange and acceptance of research results between different academic disciplines, such as wildlife ecology and tourism research, has until recently been scarce. Absence of discipline-independent guidance on the management of wildlife tourism, in combination with a lack of knowledge-transfer from academia to society regarding how human impact can be reduced, may contribute to unintended disturbance of wildlife. Here we present a methodology, where use and protection constitute equal importance within wild animal watching, by showing how a synergetic gain of combining knowledge from different academic disciplines may occur and be implemented in order to decrease potential human disturbance on harbour seals (Phoca vitulina). Further, we suggest that improved transferal of interdisciplinary research from academia to industry increases understanding of the wildlife tourism industry and has the potential to change tourist behaviour and hence minimise disturbance of wild animals. We exemplify this possibility by combining results from two case studies derived from biology and tourism research. The aim of both was to study potential human disturbance on harbour seals (P. vitulina) during land based seal-watching. The combined findings indicate that more attention should be paid to understanding and communicating the types of tourist behaviour likely to cause distress.
An aerial grey seal survey was conducted in Iceland during the pupping season of 2012 with the aim of estimating the size of the grey seal population. From the 17th of September until the 29th of November, pups were counted one to four times in the main grey seal haul-out sites in Iceland (a total of 100 flying hours) and the total population size was estimated from the total number of pups. In the rookeries surveyed from the air the estimated peak of pupping varied from the 29th of September to the 31st of October. In addition, the peak of the pupping season in Bjarnarfirði and Drangatanga was estimated using age determinations of pups (in days). This peak was found to be the 5th of October. The highest number of pups were found in Breiðafjörður; on average 525 pups (90% CI: 496-553), on the shores of northwest Iceland; on average 322 pups (90% CI: 279-323), and in the south of Iceland; on average 133 pups (90% CI: 133-137) were found. In Faxaflói only 12 (90% CI: 10-13) pups were found on average. The results indicate that the number of grey seal pups born in 2012 were minimally 992 (90% CI: 918-1026), which is less than the years of 2005 and 2008 when grey seal censuses were carried out in Iceland using the same methods. The estimated population size for 2012 was 4200 (95% CI: 3400 – 5000) animals, corresponding to a yearly decrease of 5% (90% CI: 4%-6%) between 2005 and 2012. The estimated population size in 2012 is therefore the lowest since 2004, although the change in population size is not statistically significant since the turn of the century. This is due to the fact that censuses of the Icelandic grey seal population have been carried out to irregularly, and therefore the statistical power of the data is too low to detect a significant trend in population sizes. The guidance from Icelandic authorities for the Icelandic grey seal population has been set to a minimum criterion of 4100 animals and hence, the current population size in 2012 is close to that number. Calculations reveal a 44% chance that the population is <4100 animals and therefore we highlight the importance of taking this into consideration in regards of exploitation of the population.
Wildlife tourism in Iceland Visitors to Iceland come in search of unique nature-based experiences. Known as a land of geysers, volcanoes and glaciers, Iceland is also home to interesting species of wildlife that add significantly to the attraction of the country. In 2013 tourism for the first time became the largest export sector in Iceland, taking over from fisheries and aluminum (Oladottir 2014). In the same year, an Icelandic Tourist Board report (Oladottir 2013) predicted the country would receive one million visitors during 2020. Popular media, however, frequently claim that this number may be reached in 2014. This rapid increase in visitation is coupled with concerns about sustainability of the industry and its products, and how the associated issues and challenges should be managed successfully. It is well documented that tourism can negatively impact wild animals (e.g., Green and Giese 2004); nevertheless, the phenomenal growth in wildlife tourism in Iceland has developed without comprehensive policies or guidelines. Managing tourism to ensure positive visitor experiences, while also minimising disturbance to wildlife, is a field of on-going interest. Codes of conduct, designed as guidelines to govern tourist activities in wildlife habitat, are often developed in reaction to a particular local situation rather than as a result of long term strategic planning. Compared to government based regulations and legislations, more informal codes can be relatively quick to implement. The introduction of a code of conduct can have positive effects on wildlife (e.g. Wray et al. 2010). However, codes of conduct seldom eradicate negative impacts on wild animals (Quiros 2007; Duprey et al. 2008) partly because important factors that should be considered when preparing such codes are rarely elaborated. Codes of conduct and ethical principles In May 2014 The Wild North project released a series of codes of conduct designed to guide visitor interactions with whales, seals, birds and foxes across several Nordic countries, including Iceland. This paper examines these codes, through a case study of seal watching on the Vatnsnes peninsula in northern Iceland. The region supports a large population of harbour seals (Phoca vitulina), which haul out on skerries close to land, where they are readily observed by visitors to the region either on foot or by organized boat trips. This paper discusses the identified need for The Wild North codes of conduct, their history and anticipated implementation on the Vatnsnes peninsula. Based on a literature review and secondary data, we compare The Wild North codes with those used in other countries to gauge their strengths and weaknesses. Questioning the efficacy of codes of conduct, we ask: Are there better ways to manage wildlife tourism? Searching for better ways, a set of ethical principles designed to facilitate a more ecocentric approach to managing the interactions between visitors and wildlife in natural areas is explored (Burns, Moore and Macbeth 2011). The principles encourage visitors to recognise the intrinsic value of wildlife and develop a sense of moral obligation and moral reasoning toward their wildlife experience. The principles propose that management strategies work within a precautionary framework, acknowledge the interconnectedness between people and nature, and accept that wildlife belongs in nature. This requires managers to engage in a reflexive process with regard to their own ethical position to facilitate the practical application of an ecocentric approach. Understanding and effectively utilizing codes of conduct The study revealed that codes of conduct are commonly developed and used as an idealized way to guide tourism related interactions with wildlife. However, little attention appears to have been given to assessing their effectiveness. Consequently, we argue the need for greater attention to monitoring and evaluating the effectiveness of codes of conduct. Based on the ethical principles, we also suggest that codes of conduct with a more anthropocentric focus support the perception of wildlife as tourism objects; that is, for their use to people (an extrinsic, or instrumental value) which may not be the most sustainable way forward. In contrast, codes developed by those concerned primarily with protection of the animals tend toward discourse that may alienate the visitors and not encourage compliance behaviours. Arguing that an interdisciplinary approach, incorporating expertise from both biologists and tourism specialists, is needed to research, develop and monitor codes (Granquist and Nilsson 2013), we propose how codes of conduct might look if they start from a perspective that promotes the intrinsic value of wildlife to visitors. Applying ecocentric ethical principles to codes of conduct enables us to translate current thinking in wildlife tourism to real world applications. Our goal is a framework for the development of codes of conduct that maximise the positive experience for visitors to natural areas and minimize their disturbance of the wildlife. References Burns, G. L., Macbeth, J. and Moore, S. 2011 Should dingoes die? Principles for engaging ecocentric ethics in wildlife tourism management. Journal of Ecotourism, 10(3): 179-196. Duprey, N.M.T., Weir, J.S. and Würsig, B. 2008 Effectiveness of a voluntary code of conduct in reducing vessel traffic around dolphins. Ocean and Coastal Management, 51: 632-637. Granquist, S.M. and Nilsson, P.Å. 2013 The Wild North: Network cooperation for sustainable tourism in a fragile marine environment in the Arctic Region. In D. Müller, L. Lundmark, and R. Lemelin (eds), New Issues in Polar Tourism. Heidelberg: Springer. Pp 123-132. Green, R. and Giese, M. 2004 Negative Effects of Wildlife Tourism on Wildlife. In K. Higginbottom (ed), Wildlife Tourism: Impacts, Management and Planning. Altona: Common Ground Publishing. Pp 81-98. Oladottir, O.P. 2013 Tourism in Iceland in Figures, April 2013. Icelandic Tourist Board. Oladottir, O.P. 2014 Tourism in Iceland in Figures, April 2014. Icelandic Tourist Board. Quiros, A.L. 2007 Tourist compliance to a code of conduct and the resulting effects on whale shark (Rhincodon typus) behaviour in Donsol, Philippines. Fisheries Research, 84: 102-108. Wray, K., Espiner, S. and Perkins, H. 2010 Cultural Clash: Interpreting established use and new tourism activities in protected natural areas. Scandinavian Journal of Hospitality and Tourism Research, 10: 272-290.
- Feb 2014
The frequency of injuries on salmonids caused by harbour seals (Phoca vitulina) was investigated in five important salmon angling rivers in NW-Iceland: Víðidalsá, Gljúfurá, Vatnsdalsá, Miðfjarðará og Laxá á Ásum during the angling periods of 2009 and 2010. Anglers in the rivers were asked to report all types of wounds, scars and injuries found on the caught salmonids (Atlantic salmon (Salmo salar), brown trout (Salmo trutta) and arctic charr (Salvelius alpinus)), the size of the wound/scar and if it as due to seal teeth/claws, nets, other factors or if the reason for the wound was unclear. The results show an overall low frequency of injuries. 78 salmonids injured by seals were recorded in total for 2009 and 2010 combined, corresponding to 0.61% of the total number of caught salmonids in the rivers during the time of study. In 2009, 0.36% of caught Atlantic salmon were injured by seals, while no injured brown trout or arctic charr were recorded. In 2010, 0.92% of caught salmon were injured by seals, while only 0.07% of caught brown trout and 0.04% of caught arctic charr were injured. 62% of seal injured salmon were 2SW salmon (≥70cm). Further research in the area regarding the effects of harbour seals predation on salmonid is ongoing.
Rapidly increasing tourist flows to the Arctic North have put focus on the sustainable development of tourism in natural environments. The Wild North project shows how different stakeholders in a network across the Arctic North deal with the need for a development of sustainable use of wildlife in a tourism context by using interdisciplinary research, active cooperation, and exchange of knowledge. The wildlife focused in this international project consists of arctic foxes, whales, seals, and birds. The project demand came from the tourism industry and was taken up by public organizations and academies within nature protection and tourism management. The project has also the maintainability of a destination in focus together with its sustainability.
- Jan 2013
Selasetur Íslands og Veiðimálastofnun standa fyrir margvíslegum selarannsóknum á Íslandi. Eitt aðalrannsóknarsvæðið er Vatnsnes og Heggstaðarnes í Húnaþingi vestra og því er mikilvægt að fylgjast með fjölda og útbreiðslu sela á þessum slóðum. Eitt af markmiðum Selaseturs Íslands er einnig að fræða almenningi um seli og um þær selarannsóknir sem setrið stendur fyrir. Með þetta tvennt í huga hefur Selatalningin mikla farið fram árlega síðan 2007. Talningin byggir á því að vísindamenn með hjálp almennings telja seli á stóru svæði (55-100km strandlengja) á stuttum tíma. Niðurstöður sýna að heildarfjöldi sela á svæðinu hefur haldist nokkuð stöðugur árin 2007 - 2012. Töluvert færri selir sáust hinsvegar árið 2012, heldur en undanfarin ár, en frávikið var ekki marktækt. Ekki kom fram marktækur munur á selafjölda við Vatnsnes tímabilið 2007 – 2012 og eðalfjöldi (meðalfrávik) var 853,33 (207,045). Talningar á Heggstaðanesi hófust fyrst 2009 og á tímabilinu 2009 – 2012 var meðalfjöldi (meðalfrávik) 104,75 (13,985). Talningardagurinn hafði ekki marktæk áhrif á selafjölda. Mikilvægt er að hafa í huga að þessar talningar eru aðeins vísbending um lágmarksfjölda sela sem dvelja á þessu tiltæka svæði og geta einnig þættir eins og mismunandi veðurskilyrði þegar talið var skekkt niðurstöðurnar. Til að fá marktækari niðurstöður væri hægt að framkvæma talninguna oftar en einu sinni á hverju ári. Þátttakendur eru einnig misvanir talningarmenn og er það eitthvað sem ber að íhuga þegar niðurstöðurnar eru skoðaðar. Þessar talningarnar segja ekki heldur til um ástand selastofna í heild þar sem talið er á afmörkuðu svæði. Heildartalningar á selastofnunum við Íslandsstrendur fara fram um landið allt á nokkurra ára fresti.
A permanent herd of Icelandic horses with four stallions and their harems was studied for a total of 316 hours in a large pasture (215 ha) in May 2007 in Iceland. Interactions between stallions of different harems and other aspects of the horses' behaviour were studied. One stallion and nine horses were introduced into the pasture prior to the study to examine the reactions of the resident stallions to a newcomer. The stallions spent significantly less time grazing than other horses and were more vigilant. Home ranges overlapped, but harems never mixed. The stallions prevented interactions between members of different harems indirectly by herding. Generally, interactions between resident stallions were nonviolent. However, encounters with the introduced stallion were more aggressive and more frequent than between the other stallions. Here, we show that four harems can share the same enclosure peacefully. The social network seems to keep aggression at a low level both within the harems and the herd as a whole. We encourage horse owners to consider the feasibility of keeping their horses in large groups because of low aggression and because such a strategy gives the young horses good opportunities to develop normally, both physically and socially.
Earlier research indicates that stallions may supress interactions of their harem members, leading to less stable hierarchies and friendship bonds in harems compared to non-stallion groups. In this paper, the effect of the presense of a stallion on the social behaviour of mares was studied by comparing six harems containing stallions to four mixed sex groups not containing stallions. Both temporary and permanent harems were studied, giving the possibility to investigate the effect of group stability on social interactions.A significant linear hierarchy was found in all non-stallion groups that were used for comparison, while the hierarchies were only found to be linear in three of the six harems containing stallions (Landaus h′, p < 0.05). Aggression rate was lower (t-test, p < 0.05) and fewer friendship bonds (G-test, p < 0.0001) were found within the harems, compared to the groups without stallions. Stallions seldom intervene directly in interactions between harem members. Thus, our results give support to the hypothesis that stallions may suppress interactions of harem members, but in a more indirect way than with direct interference. In addition, our results give support for earlier findings that aggression rate may be affected by group stability. We found a higher aggression rate in the temporary harems compared to the permanent harems (Kruskal–Wallis, p < 0.05) and in the temporary non-stallion group compared to the permanent non-stallion group. The results have significance for further research on social structure of mammals, and may be applied in management of domestic animals.