Unmanned aerial vehicle footage of sexual interactions and atypical group sizes within harbour porpoises (Phocoena phocoena) around archipelago of Shetland, UK

Abstract

Studying social behaviour of mobile and cryptic marine mammals is challenging, particularly for small species like harbour porpoises whom in many areas tend to be challenging to initially spot and then gain repeated observations. Recently, there has been a rise in the number of observations of harbour porpoises occurring in atypical group sizes that are larger than the typical size of two to three individuals. This behaviour, in conjunction with the evolving availability and usability of Unmanned Aerial Vehicle (UAV) technology, is supporting and improving the capture of information on social behaviours. Around Shetland, Scotland, harbour porpoises are sighted from land via opportunistic and systematic surveys year-round with atypical group sizes also observed. To capture information on group size, opportunistic and systematically collected UAV footage were collated and analysed. This revealed sexual approach behaviours in typical group sizes of small groups (two to four individuals), as well as atypical group sizes of medium groups (five to nine individuals), large groups (10 to 19 individuals), and aggregations (≥20 individuals) occur in multiple bays around Shetland. This is the first insight into group sizes and behaviours for the region which was recently designated as an Important Marine Mammal Area (IMMA). Further insight into these social behaviours and group size (e.g. seasonality, site fidelity) could aid harbour porpoise conservation and management within the IMMA and elsewhere.

Introduction

Harbour porpoises (Phocoena phocoena) are the most abundant small cetacean in European Atlantic waters (Hammond et al., 2013). Abundance is estimated to be 338,918 within the North Sea assessment unit (Gilles et al., 2023). Although their relative abundance, population trajectory and other aspects related to harbour porpoise life histories remain largely unknown (Braulik et al., 2020). This is due to the challenges related to studying and collecting data on highly mobile, small, cryptic and elusive species which may ultimately hinder the development of effective conservation and management measures (Elliser and Hall, 2021; Gilles et al., 2023).

Despite the challenges in obtaining data, a considerable body of research has established important information about harbour porpoise life histories. At birth, they range from 60 to 75 cm in length and 3 to 9 kg in weight, with growth rapidly occurring within their first year of development and living to around 12 years of age (Learmonth et al., 2014; Lockyer, 1995; Lockyer and Kinze, 2003; Stepien et al., 2023). As adults, they have shown to display sexual dimorphism where mature females are slightly longer and heavier than mature males by around 8% and 20%, respectively (Galatius and Å, 2005). Within Scottish waters mean mature male size was 147 cm and mean mature female size was 158 cm in length (Learmonth et al., 2014). The sexual dimorphism is thought to be from females needing higher energy storage during increased energy demands within pregnancy and lactation periods (Lockyer, 1995). However, morphological dimensions of harbour porpoises can vary depending on the geographic location in which they reside (Lockyer, 1995; Lockyer and Kinze, 2003).

Harbour porpoises reach sexual maturity between 2 and 5 years (Lockyer and Kinze, 2003). Reproductive activity predominately occurs within the summer months in comparison to winter months, thought to be from increased activity with male testes as well as female harbour porpoises being in oestrus (Keener et al., 2018; Webber et al., 2023). Female gestation lasts around 9 to 11 months with birthing periods occurring the following spring and summer (Lockyer and Kinze, 2003). Calving then last 8 to 12 months followed by a period of oestrus causing female reproductive intervals to vary from 1 to 2 years (Lockyer and Kinze, 2003).

Although considerable information exists on harbour porpoise life histories, many aspects of their ecology remains poorly understood. This is largely due to their elusive nature which makes observations of harbour porpoises, and particularly their associated behaviours, to be rare and fleeting. A small number of more prolonged sightings have revealed details of their reproductive encounters and feeding tactics; Orbach et al., (2019) described female reproductive behaviours from field observations and video records, noting that females may often be unaware of the presence of males as female interactions to male reproductive behaviours included; passive (receptive), fluke lifts (evasive), dives (ambiguous), peduncle curl (ambiguous), change in direction (ambiguous), or body rolls (ambiguous). Webber et al., (2023) also described these female reproductive behaviours in which the most frequent responses were body rolls and fluke lifts, often occurring simultaneously. Collectively, these varied responses suggest that females may exhibit a range of behavioural states during sexual approaches from males, though their interpretation remains unclear. Keener et al. (2018) first described mating behaviours of wild harbour porpoise by capturing still imagery from the high vantage point of the Golden Gate Bridge (San Francisco, USA). The images showed males to display a distinct laterality towards the left side of females during sexual approaches of copulation attempts and displays. More recently, Webber et al. (2023) developed on these findings with an addition of Unmanned Aerial Vehicles (UAVs) footage collated from various locations around the Northern Hemisphere. From this, mating activity was seen to be highly energetic leading to splashy aerial behaviour, thought to be from a combination of female fluke lifts and male angled re-entry into the water. It also included footage from Shetland, United Kingdom (UK), which provided one of the first UAV-based observations of harbour porpoise sexual activity in this region and documented one of the first confirmed instances of male–male socio-sexual behaviour in harbour porpoises. This study provided new insights by documenting underwater mating interactions that had previously been difficult to quantify from surface-based observations and still imagery. Ortiz et al. (2021) utilised UAVs in Denmark to capture cooperative hunting behaviours and revealed that individuals carried out specific roles within prey capture. Developing from this research, Stedt et al. (2024) also identified six foraging techniques using UAVs in Denmark. These insights highlight how observations from height can help capture more complex social behaviours, which are typically challenging or not possible to observe from land- or boat-based platforms (Fiori et al., 2020; Ramos et al., 2023; Webber et al., 2023). UAVs have also successfully captured behavioural data for other cetacean species, including southern right whales (Eubalaena australis), Indo-Pacific humpback dolphins (Sousa chinensis), humpback whales (Megaptera novaeangliae), killer whales (Orcinus orca), Risso’s dolphins (Grampus griseus), and dusky dolphins (Lagenorhynchus obscurus) (Chung et al., 2022; Hartman et al., 2020; Orbach et al., 2020; Pirotta et al., 2021; Sprogis et al., 2023; Weiss et al., 2021). Given the collection of behavioural data from UAV aerial observations is still relatively new, many knowledge gaps related to group dynamics, social interactions, and behaviours remain especially for the harbour porpoise species (Ortiz et al., 2021; Palacino González 2019; Stedt et al., 2024).

Harbour porpoises are most commonly observed either alone or in typical groups sizes of small groups (two to three individuals) (Elliser et al., 2025, 2018; Jefferson et al., 2016; Haelters et al., 2011). However, in recent years, atypical group sizes such as aggregations (≥ 20 individuals) have been documented year-round though they are more frequently observed between August and April with some reports describing over 100 individuals. These have been recorded in several coastal regions across the species’ range, including the Salish Sea and the Port of Prince Rupert (Canada), as well as Longhope Bay (Orkney, UK) (Anderson et al., 2023; Dracott et al., 2022; Neave-Webb and Hetherington 2023). In the coastal waters of Shetland, UK, harbour porpoises are sighted year-round, and atypical group sizes including unusually large aggregations (>100) are occasionally reported (The Shetland Cetacean Group, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2003; Evans et al., 1996; Shetland Islands Marine Planning Partnership, 2021; WDC Shorewatch, unpublished data). Many of the sightings are from opportunistic observations from land, with limited systematic effort and data collection, likely leading to underreporting of sightings and therefore a lack of understanding towards the species in this area (Shetland Islands Marine Planning Partnership, 2021; Shucksmith 2017).

The waters around Shetland and Fair Isle have recently been designated as an Important Marine Mammal Area (IMMA), with harbour porpoises listed as one of the qualifying species (IUCN-MMPATF, 2024a). While a Special Area of Conservation (SAC) has been established on the west coast of Scotland in the Inner Hebrides and the Minches with harbour porpoises being a designated species, no similar designations currently exist for Shetland waters (NatureScot, 2024). Therefore, to further develop our understanding of the harbour porpoise within the IMMA, the aim of this study was to document and describe the occurrence of harbour porpoise behaviours and group sizes using UAV footage collected around coastal Shetland. This initial insight adds evidence to this area being important to this species, particularly towards atypical group sizes (e.g. medium and large groups, and aggregations) and for important social behaviours (e.g. mating). This information can be used as evidence to help support the identification and development of conservation measures around Shetland.

Materials and methods

Study site

Shetland is a subarctic archipelago in the North Atlantic. It is the northern most group of islands of the UK, that lies approximately 170 km from mainland Scotland and 80 km from Orkney (Hall et al., 2021) (Figure 1). Data collection was conducted on the mainland island of Shetland at the study sites of Gulberwick Bay, South Nesting Bay, Mousa Sound, and Quendale Bay (Figure 1).

Figure 1. Map of Shetland showing study sites of South Nesting Bay, Gulberwick Bay, Mousa Sound, and Quendale Bay (Shetland mainland) where UAV footage was captured, and the location of the Shetland and Fair Isle IMMA (IUCN-MMPATF 2024b).

Data collection

Following the characterisation of harbour porpoise mating behaviours by Keener et al. (2018), similar splashing events were observed opportunistically from shore around Shetland in 2021 (Karen Hall, personal observation). These preliminary observations prompted the deployment of UAVs to investigate and document the behaviours more systematically and from an aerial perspective.

UAV footage of harbour porpoise(s) was collected opportunistically and during targeted data collection efforts between 2019 and 2023 at the four study sites (Figure 1). Footage was captured by the co-author team (local professional UAV pilots or trained researchers from the University of Highlands and Islands Shetland) when harbour porpoise(s) were observed from land and conditions of no rain and light winds allowed for UAV flights. UAV footage was collected using DJI Mavic 2 Pro, Mavic 3 Cine, and Mini 3 Pro, with UAVs flown at least 20 m above sea level.

Data analysis

UAV footage was viewed using Adobe Premiere Pro (Version 25.2.3). UAV clips were initially screened to determine the species. Harbour porpoises were identified from the footage based on their characteristics of small bodies, blunt heads with lack of beak, small and triangular dorsal fin, dark grey back, and a lighter underbelly. In Shetland, there have been 15 species of cetaceans recorded all of which are considerably larger and morphologically distinct from harbour porpoises (Shetland Islands Marine Planning Partnership, 2021; Shucksmith, 2017). Therefore, harbour porpoises are the only regularly occurring small cetacean species and the likelihood of confusion with other species in UAV footage considered to be minimal.

UAV clips were then screened focusing on sexual behaviours (display and copulation attempts) in small groups (two to four), medium groups (five to nine), large groups (10 to 19), and aggregations (20 or more). UAV clips that had one porpoise individual with no apparent sexual behaviours were excluded from the final analysis.

Sexual behaviours were classified using the classification system described by Keener et al. (2018) in which approaches exhibited either: (1) displays (where male was not sufficiently proximate to copulate, involving a body roll to present ventrum to females) and (2) copulatory attempts (where the male was sufficiently proximate to the female to copulate, involving rapid, high-energy approaches in which males accelerated toward females). Individuals were considered to be sufficiently proximate during copulation attempts when they were observed either to make physical contact with the female or to approach closely (<1 m) without contact. Both copulatory attempts and displays were subcategorised based on whether or not the penis was extruded.

While harbour porpoises do have sexual dimorphism, it is not easily distinguishable in the field. Thus, the sex of individuals was assumed unless male genitalia were displayed. During encounters, the male was assumed to be the individual initiating the approach, and the female was the individual who was the target of the approach. Although females can also be identified by the presence of a dependent calf (Keener et al., 2018), and groups of harbour porpoises were observed in our footage, none of the individuals involved in sexual approaches appeared to have dependent calves.

Group size was estimated by manually counting the number of individuals visible within a UAV clip frame. To classify harbour porpoise group sizes, a tiered approach was adopted based on both the distribution of our data and consistency with previous studies (Anderson et al., 2023). The majority of sightings consisted of two to four individuals, which we defined as ‘small groups’, reflecting the typical social organisation reported for this species. Groups of five to nine individuals were classified as ‘medium groups’, representing slightly larger but relatively common groupings within our dataset. Groups of 10 to 19 individuals were defined as ‘large groups’, as these are less frequently observed and stand out from the typical pattern. Finally, groups of 20 or more individuals were classified as ‘aggregations’, following the definition of Anderson et al. (2023). This approach allows us to remain consistent with the established literature while also acknowledging the frequent occurrence of atypical group sizes in our study area.

Results

A total of 125 UAV clips around coastal Shetland was captured between 2019 and 2023 during the months of February, March, April, May, August, September, October, and November over 18 days (Supplementary Table S1). Forty-eight of these clips did not match our screening criteria (see ‘Methods’ section). Therefore, 77 clips captured during the months of February (n = 42), March (n = 14), May (n = 1), September (n = 14), October (n = 3), and November (n = 3) over 16 days featured harbour porpoises, giving 79 mins and 5 secs of footage suitable for analysis (Table 1, Supplementary Table S2).

Table 1. Summary of UAV footage analysis (please see Supplementary Table S1 for more information)

Group size

Small groups were captured in the footage in all 6 months of data collection (February [n = 23], March [n = 5], May [n = 1], September [n = 10], October [n = 2], and November [n = 2]) (Table 1, Supplementary Table S2). Medium groups were captured in 4 out of the 6 months of data collection (February [n = 11], March [n = 4], September [n = 4], and October [n = 1]) (Table 1, Supplementary Table S2). Large groups were captured in 3 out of the 6 months (February [n = 6], March [n = 4], and November [n = 5]) (Table 1, Supplementary Table S2). Aggregations were captured in 2 out of the 6 months (February n = 2], March [n = 1]) (Table 1, Supplementary Table S2). Of the footage collected, the average group size seen was seven individuals (Table 1, Supplementary Table S2). The maximum number of harbour porpoise seen in a group was 26 individuals which was recorded on two occasions in February 2018 and March 2023 at South Nesting Bay (Table 1, Figure 2a, Supplementary Table S2, Supplementary video).

Figure 2. (A) harbour porpoise aggregation of 25 individuals (South Nesting Bay), (B) harbour porpoise sexual approach classified as a display (Gulberwick Bay), (C) harbour porpoise sexual approach classified as a copulation attempt (South Nesting Bay), all via UAV footage.

Sexual approaches

A total of 13 clips contained one or more sexual approaches, relating to either copulation attempts or sexual displays across all sites (Table 1, Supplementary Table S2). These were filmed in 4 out of the 6 months with recordings in February (n = 6), March (n = 2), September (n = 3), and October (n = 2) (Table 1, Supplementary Table S2). Multiple sexual approaches could be recorded within a clip therefore, there were 23 sexual approaches within the 13 clips (Table 2, Supplementary Table S2). Fifteen were sexual displays (Table 2, Figure 2b, Supplementary Table S2, Supplementary video) and eight were copulation attempts (Table 2, Figure 2c, Supplementary Table S2, Supplementary video). Sexual displays were present in February (n = 6), March (n = 2), September (n = 5), and October (n = 2) (Table 2, Supplementary Table S2). Whereas copulation attempts were only recorded in February (n = 5), March (n = 1), and September (n = 2) (Table 2, Supplementary Table S2). It is important to note that nine of the 23 sexual approaches accounted for the activities of the same individual conducting multiple sexual approaches in a series over a short timespan (Table 2, Supplementary Table S2, Supplementary video).

Table 2. Summary of sexual approach analysis (please see Supplementary Table S2 for more information). *Multiple sexual approaches conducted by same individual in one clip

Details regarding orientation, sufficiently proximate and penis visibility were not always clear therefore uncertainty classifying sexual approaches were apparent. This meant out of the 23 occasions sexual approaches were recorded, five were unclear but still had other characteristics of sexual approaches to be included in the analysis (Table 2, Supplementary Table S2).

Discussion

The UAV footage collated as part of this short note highlights, for the first time, the use of multiple inshore bays around Shetland as sites where harbour porpoises occur in atypical group sizes of > 5 individuals and exhibit sexual approach behaviours.

Mating-related surface behaviours in harbour porpoises, characterised by Keener et al. (2018), have often been considered challenging to observe from land due to the species elusive and rapid surface activity (Gutiérrez-Muñoz et al., 2021; Teilmann et al., 2007). However, recent studies demonstrate that such behaviours can be documented regularly from land-based sites that have good vantage points (Elliser et al., 2025). Opportunistic sightings around Shetland (Karen Hall, personal observation, 2021) of such behaviours prompted further investigation using UAVs. The UAV footage in this study revealed that sexual approaches are indeed occurring around Shetland in multiple bays. The clear classification of behaviours in our footage aligned with findings from Keener et al. (2018) and Webber et al. (2023), including the distinction between copulation attempts and displays, left-sided lateralisation of sexual approach orientation, and the occurrence of aerial termination. However, in the analysed UAV footage, sexual approaches were not always clear due to difficulties in confirming details regarding sex of individuals, sufficiently proximate or not, penis extrusion or no extrusion, and differences in sexual approach orientation. Other clear visual cues related to sexual approaches such as body rolls, rapid, high-energy approaches, and splashing provided strong evidence of mating attempts. Thus, while clear and detailed classification was not always possible, the presence of some of these conspicuous details and cues allowed us to reasonably infer sexual behaviours on these occasions. These challenges were due to clarity in footage related to flight duration, recording duration, height, angle, or frame clarity. In addition, our understanding of harbour porpoise sexual behaviours is still relatively new and limited, and many aspects of their broader ecological and behavioural repertoire remain poorly understood. Thus, potential variation in sexual behaviours around Shetland may occur and may have contributed to the difficulty of behavioural classification. However, due to the confirmation of harbour porpoise sexual behaviours being consistent across the Northern Atlantic (Webber et al., 2023), it is more likely that the inherent complexities and limitations around capturing the full scope of harbour porpoise behavioural data with the technologies available are the major limiting factor to behavioural categorisation. Future improvements in UAV technology and protocols and associated video and image quality, as well as development on harbour porpoise behavioural ecology may alleviate some of the challenges related to behavioural categorisation.

Harbour porpoises can be challenging to detect and to reliably estimate group size from shore due to their distance from observer and detections diminishing with increasing wind speeds and sea states (Teilmann 2003). Despite these limitations, atypical group sizes have been documented by anecdotal land-based observations, but little research has been conducted, until now (The Shetland Cetacean Group, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2003; Evans et al., 1996; WDC Shorewatch, unpublished data). Here, the presence of larger groups enhanced the likelihood of shore-based detection, aiding UAV footage capture. The aerial perspective improved group size estimation and revealed that all sites had the presence of atypical group sizes, giving highlight to largest group sizes in South Nesting Bay. However, UAVs have limitations within frame capture at a given height, as well as challenges within locating, tracking, and maintaining focus on individuals to estimate group size. Additionally, from the footage it was apparent that individuals and groups move, merge, and split over time displaying a fission–fusion nature (Anderson et al., 2023). This behavioural fluidity made it difficult to obtain consistent group size estimates, as the composition of groups changed frequently. Even though UAVs can provide improved estimates, conservative counts of group size at a given UAV height may offer the most consistent method of true numbers (Álvarez-González et al., 2023; Brown et al., 2023).

The collation and analysis of UAV footage of harbour porpoises around Shetland highlights that they aggregate in these areas and conduct sexual approaches. Notably, South Nesting Bay recorded the highest number of sexual behaviours and aggregations during the month of February. However, this is subjected to a limitation of uneven effort both spatially and temporally. Additionally, UAV footage was only collected when harbour porpoises were first sighted by an observer and could be successfully recorded by a UAV pilot under favourable weather conditions (i.e. no rain, low wind, and minimal glare). As a result, it is challenging to determine any spatial and temporal relationship to aggregations and behaviours. Anecdotally, it appears that aggregations occur less frequently in the summer months (The Shetland Cetacean Group, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2003; Evans et al., 1996; Shetland Islands Marine Planning Partnership, 2021; WDC Shorewatch, unpublished data). Previously, reproductive activity in harbour porpoises was thought to occur predominantly in summer, based on increased testicular activity in males and females being in oestrus during this period (Lockyer and Kinze, 2003; Webber et al., 2023). More recent evidence, however, suggests that mating activity occurs throughout the year, with seasonal peaks in spring (April–June) and autumn (October–December) (Elliser et al., 2025; Keener et al., in press, 2018; Webber et al., 2023). While earlier perspectives emphasised summer as the primary mating period, these subsequent observations, including those documented in this study, suggest a broader temporal window for reproductive behaviour, highlighting that social and sexual interactions can occur year-round (Keener et al., 2018; Webber et al., 2023). However, mating activity does not translate to reproductive success as copulatory attempts may be unsuccessful due to factors such as timing of oestrus and testicular activity (Keener et al., 2018; Lockyer and Kinze, 2003; Webber et al., 2023) or successful execution of mating behaviours. Aggregational events may serve as key opportunities for socialisation and reproductive interactions, with evidence suggesting these occur preferentially in specific bays and are influenced by seasonal patterns (Anderson et al., 2023). Whilst not directly investigated here, this pattern is supported by previous studies that have identified recurring aggregations of harbour porpoises in specific areas around Shetland (Evans et al., 1996; The Shetland Cetacean Group, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2003). More recently, land-based sightings through opportunistic reports and WDC Shorewatch programme, which although unpublished and not yet formally analysed, appear to show that larger groups are more frequently recorded in these bays compared to other locations based on initial observations and field experience (Hall, unpublished; WDC, unpublished). Nevertheless, the role of these aggregations in facilitating such behaviours warrants further investigation to better understand their significance within harbour porpoise populations.

In other studies, harbour porpoise aggregations have been documented via land- and boat-based observational surveys and have been associated with multiple behaviours including foraging, travelling, mating, social, vessel approach, and fission–fusion in the Salish Sea (Anderson et al., 2023), and feeding opportunities and prey distribution in the Port of Prince Rupert, Canada (Dracott et al., 2022). Mating behaviours were observed more frequently in larger groups compared to smaller ones (Anderson et al., 2023), highlighting that while social and reproductive interactions are an important component of aggregations, other activities such as foraging also occurred (Dracott et al., 2022). Harbour porpoise spatial occurrence around Shetland has previously been linked to sandeel fishing grounds (Evans et al., 1993, 1996). We did not investigate the overlap between sites of aggregation and sandeel fishing grounds, though the drivers behind habitat usage would shed light on aggregations, important areas and may identify areas that may require additional protection. In other cetacean species, aggregations can be caused by a response to the presence of predators (Connor 2000; Dracott et al., 2022). Killer whales are frequent visitors to Shetlands coastline and have been observed to predate on harbour porpoises (Sutherland, 2024). However, the relationship between harbour porpoise aggregations and killer whale occurrence has not been investigated here.

In other cetacean species, cohesion between individuals has been ascribed through analysis of spatial proximity and associated social behaviours (Connor et al., 2000; Mann, 1999). In this study, UAV footage suggests that harbour porpoise aggregations may display these spatial definitions and social behaviours in individual cohesion. UAVs could potentially be used to explore other questions, which would provide valuable insights into their interactions and group dynamics in these coastal waters (Pedrazzi et al., 2025). However, while such analyses have been successfully applied to other cetacean species, such as Risso’s dolphins, they have not yet been fully explored for harbour porpoises (Hartman et al., 2020).

The presence of harbour porpoises in coastal areas leaves them vulnerable to anthropogenic impacts, as these regions tend to have higher level of human activities (Carlén et al., 2021; IAMMWG et al., 2015; Nielsen et al., 2021). The waters around Shetland and Fair Isle have recently been designated as an IMMA, meaning it has the potential to be delineated and managed for conservation designation (IUCN-MMPATF, 2024a). However, no specific management measures are in place. The development and implementation of effective management and conservation measures will require an understanding of critical habitat, underpinned by the understanding of a species’ spatial and temporal patterns of space use (Sveegaard et al., 2011). This initial analysis of this study adds to the growing evidence that the utility of UAVs for aerial view can support our understanding of harbour porpoise occurrence, behaviours and group sizes, helping to inform and support effective conservation and management measures. Further research is required to further explore the spatio- and temporal dynamics related to these atypical group sizes (such as seasonality, site fidelity, etc.), and the behaviours occurring within them, to ultimately improve understanding of harbour porpoise social and population dynamics, and how best to develop effective conservation.