Abstract
Shipping activity has increased greatly over the past 50 years, leading to growing concerns about potential harmful effects of increased shipping noise on marine mammals and the potential for physical injury due to collisions.
When originally commissioned, this project was designed to investigate the interactions between seals and vessels operating in the inner Moray Firth as part of a study into the cause of a particular type of mortality in harbour seals (Phoca vitulina), characterised by distinctive spiral lacerations. Such stranded carcasses are often referred to as corkscrew seals. Results of investigations of the wounds and associated pathological features of these, together with the spatial and temporal patterns of occurrence, had led to the conclusion that collisions with ducted propellers were the most likely cause of these wounds. These types of propulsion systems are widely used on tug boats and other work-vessels. Activity by both types of vessel was expected to increase in the inner Moray Firth during planned construction work associated with port developments.
In late 2014, observations of grey seal (Halichoerus grypus) predation on conspecifics which produced similar wounds provided a plausible alternative explanation of the likely cause of spiral lacerations. At that time there was insufficient evidence to rule out interactions with vessels as a possible cause of some of these mortalities, so the requirement to investigate the temporal and spatial overlap between seals and vessels remained.
A telemetry based study of the swimming behaviour of harbour seals was carried out in the Moray Firth to (1) allow a comparison of seal movements and vessel movements and (2) compare the densities of seals and shipping traffic to identify areas with high spatial overlap. A secondary aim was to provide data on seal movements between haulout areas, to identify connectivity between seal haulout sites in the Dornoch Firth Special Area of Conservation (SAC) and the potential construction site at Ardersier in the inner Moray Firth.
During the two year study in 2014 and 2015, 37 Ultra High Frequency (UHF)/GPS tags were deployed on harbour seals at two different sites at Ardersier and in the Dornoch Firth. Locations from the animals were interpolated to produce regularised locations at three minute intervals. Each location was assigned a status of ‘hauled out’ (when the animal was on land), or on a ‘trip’ (when the animal was at-sea). Each trip was associated with a specific departure and destination haulout site. Trips to and from each haulout site were then combined by smoothing all locations to produce a usage surface that was weighted by the population estimate derived from the number of seals counted at that haulout site during annual aerial surveys. All haulout sites were combined to produce estimated mean and associated 95% confidence interval usage maps on a 0.5 km x 0.5 km grid in the Moray Firth.
These maps were compared with shipping activity using Automated Information System (AIS) vessel tracking data from MarineTraffic.com. AIS data provided information on vessel movements within the Moray Firth over the same period as the seal tracking study, for all vessels using an operating AIS transmitter. An estimated mean shipping usage map was developed in a similar manner to the seal usage map. The maps were combined by multiplying seal density by ship density for each grid cell to provide a total number of minutes of seal/vessel co-occurrence per year in the inner Moray Firth. Annual co-occurrence is therefore the sum, over a year, of the number of times individual seal-ship combinations occur in a 0.5 km by 0.5 km grid cell multiplied by the time step, in this case five minutes.
An annual total of 130,395 minutes of seal-vessel co-occurrence was estimated for the study area. Although this value appears high, the study area encompasses 10,137 km2 and areas with a high number of co-occurrences were localised. Only four grid cells (1 km2) displayed co-occurrence values above 2,500 minutes per annum and these were in narrow channels through which seals travel to access their haulout sites. Offshore sites showed comparatively low co-occurrence rates even in areas with high levels of shipping traffic.
In comparison with previous estimates of seal-shipping spatial overlap throughout the UK, the estimates for the Moray Firth appear low. Although the co- occurrence maps do not show clear evidence of avoidance behaviour, areas of high seal and high shipping usage are not coincident.
There does not appear to be any obvious relationship between incidences of strandings of spiral lacerated (corkscrew) seals and areas where seal-shipping overlap is high.
These results are not conclusive, but it is clear that the number and location of corkscrew seal strandings within the Moray Firth are not proportional to the estimates of seal-shipping overlap. A previous analysis of seal shipping interactions around the UK coast also indicated that areas where high incidences of corkscrew strandings were located did not generally coincide with intense seal-shipping co-occurrence (Jones et al., 2015b). Some areas where corkscrew seal strandings were recorded had low or zero estimates of seal-ship co-occurrence.
Further analyses including data from other seal telemetry studies should be carried out to further test these conclusions. However, this study has provided a useful method for identifying areas of high co-occurrence and provides a method for highlighting areas of importance when considering the degree to which seals interact with shipping at a broad spatial scale.
Observations of movements of individual seals and vessels did not show any apparent responses; seals did not appear to react to close passing vessels; they neither moved towards nor away from them. A single seal was observed swimming directly to a vessel at anchor and remaining in close proximity (<80 metres) for 4 hours. The majority of seal-vessel close-passes occurred in the localised areas of high co-occurrence identified in the seal/ship density mapping exercise and appeared to be largely a result of a funnelling effect.