The movements of the grey seals in this study (Figs 1 and 2) ranged from large-scale, distant travel to shorter, smaller-scale return-trips to sea from haul-out sites. Long-distance travel (> 100 km) was generally consistent in direction and speed (75–100 km day–1), and the outward and return routes were often very similar (for example, male F3–92). An exception was male F3–91 which, although travelling within the same speed range, travelled in wide arcs out into the North Sea. Travel terminated at sites where other seals were known to haul out on land. Positive haul-out records indicated that most study seals hauled out some time after arrival at such sites. Such direct travel to, and arrival at, a distant haul-out site suggests both navigational ability and knowledge of that site. Although the travels of female F8–92 were, by far, the most distant, there was nothing about her condition or the deployment which singled her out from the other study animals. Indeed, her directed travel to sites where other seals are known to haul-out in Shetland and the Faroes suggests her navigational skills were good. The cues the study seals used to aid navigation are not known. However, the apparent ‘normal’ movements of blind female F6–92 suggest that visual cues are not essential.
The distant travel shown here, and in other studies (McConnell et al. 1992; Thompson et al. 1996a), indicates that grey seals at the Farnes, Orkney, Shetland, the Faroes, and even off the west coast of Ireland, are not ecologically isolated. Such geographical mixing has important consequences in modelling epidemiology, fishery interactions and population management. For example, local population control measures may have a reduced effect due to the interchange of seals from other regions.
Inference of foraging
A temporal view of the tracks (Fig. 3) shows that most movements consisted of a series of trips from haul-out sites to offshore areas. These areas were often very localized and visited repeatedly. We have no direct evidence of when feeding was taking place, but Thompson et al. (1991) directly observed grey seals making similar trips from the Farnes to specific offshore areas. The fact that the seals remained in these localized areas for periods of hours to days, they repeatedly dived to the seabed, together with the presence of pisciverous seabirds led Thompson et al. (1991) to infer foraging. We also infer that the similar behaviour observed in this study represents foraging. We do not, however, exclude the possibility that seals were also foraging near haul-outs or while travelling.
The proposal that local trips represent foraging is further supported by a consideration of the seals’ dive behaviour and diet. The dive depth data shown in Fig. 7 indicate that many dives were to, or near to, the seabed. This is consistent with diet information. Hammond & Prime (1990) and Hammond, Hall & Rothery (1994) analysed grey seal scats collected at the Farnes and estimated that sandeels (Ammodytidae), which spend part of their time burrowed in the seabed (Wheeler 1978), comprised up to 87% of their diet by weight. Further evidence for the role of sandeels when seals are foraging at these offshore areas is based on seabed type. Sandeels have a burrowing preference for a mix of gravel and sand, avoiding pure sand (Reay 1970) and this was the seabed type at all of the areas where we believe foraging was occurring (Fig. 5). Indeed, these foraging areas were often delimited by a transition to a sediment type that did not contain gravel. In this study we do not have any direct evidence of whether the study seals preyed on sandeels themselves (either while the sandeels were shoaling or by the seals disturbing them into the water column) or on other fish species associated with sandeels.
Distribution of foraging effort
The distribution of seal activity within the Farnes Box is shown in Fig. 4. Although the tracks varied with respect to duration, time of year and individual seal, this map does emphasize two key features of the tracks that persisted, to a large degree, across most study seals and through time. These features are the large amount of activity around the haul-outs at the Farnes (and to a lesser extent the Isle of May and Abertay) and the fact that activity at sea was not randomly distributed, but focused within certain localized areas.
We distinguish eight clusters of SAS locations where we believe foraging was taking place within the Farnes Box (Fig. 5). Although all these clusters were within areas containing gravel in the sediment, the gaps between some of the clusters showed no obvious discontinuity in sediment type. This may be due to an insufficient spatial resolution in the sediment type map, an inherent patchiness in sandeel distribution, or to predation on other species (e.g. gadoids) that may not have a preference for specific sediment types. Because the clusters of SAS locations reflect the specific individual foraging areas of the study seals, it is possible that a greater sample of animals would have resulted in a greater spread of clusters.
Return-trips were short (mean 2·33 days) and relatively close to haul-out sites (mean 39·8 km). The positive correlation between return-trip extent and duration was also found in harbour seals Phoca vitulina L. (Thompson et al. 1998). The general pattern of proximity of foraging to haul-out sites suggests that, although grey seals are capable of extended and distant travel, the impact of predation may be greater on inshore fisheries, particularly those close to seal haul-out sites, rather than on fisheries further offshore.
The duration, extent and regularity of return-trips varied considerably, both from seal to seal and through time. Reasons for this variability may be endogenous or external. Endogenous factors include the requirements to spend more time ashore at specific sites, and with sufficient energy reserves, during the breeding and moult seasons. The moult period (February and March) is not covered in this study due to detachment of the SRDLs. However, two seals did change the pattern of their movements towards the start of the breeding season (October). While subadult male A1–93 continued to forage off the Farnes, it changed its haul-out site from Abertay, a non-breeding haul-out site, to the Farnes. Male F3–91 initiated a series of long travel-trips to other breeding sites at Donna Nook and Scroby Sands. There is evidence that grey seals return to the same breeding site each year (Pomeroy et al. 1994; Twiss, Pomeroy & Anderson 1994) and this may be reflected in the changes in movements of adult male F3–93 towards the start of the breeding season.
External factors that may influence foraging patterns include changes in prey availability. This has been suggested (on an interannual scale) by Boyd (1996) for Antarctic fur seals Arctocephalus gazella Peters where the length of foraging trips varied with the changing availability of krill. Thompson et al. (1996b) showed that harbour seal foraging distribution changed with prey availability. In this study we have no direct data on prey availability. However, we suggest that the interruption of foraging trips by long-distance travel, observed in this study, was not due to prey depletion. On different dates, male F3–92 and female F8–92 left the Farnes to travel to Orkney and to Shetland. However, while each was travelling north, concurrently tracked seals continued to forage off the Farnes. The costs and benefits of long-distance travel for grey seals are not clear. In terms of cost, distant travel often took little more time than a return to a local haul-out plus time spent around that haul-out. Potential benefits include the exploration of new foraging areas and the possibility of opportunistic foraging en route.
Understanding the activity budgets of seals is essential in estimating their energy requirements. The activity classification we have presented here is a simple attempt to partition seals’ time into one of three classes: near haul-out (NH), travelling (FAS) and foraging (SAS). In short trips it was difficult satisfactorily to split the time spent travelling and foraging due to the temporal resolution and spatial error of the track data. In addition, the FAS/SAS split was affected by the choice of location travel rate threshold. Failure to distinguish a natural break in the frequency distribution of location travel rates may have been due in part to the smoothing effect of using interpolated 3-hour locations rather than primary locations, and also to the effect of location error. The two factors, spatial and temporal resolution of locations and choice of location travel rate threshold, may account for some of the individual and temporal variability in SAS activity (Fig. 6). The future incorporation of dive-type information, which may be indicative of foraging (Thompson et al. 1991), may increase the ability to identify foraging.
The adult males tracked over the breeding season spent more time in NH activity during this period. However the extent and timing of these increases varied and they may reflect the breeding status achieved during the previous and current season. No such trend was apparent in the subadult males.
It is of interest that the mean NH activity was as high as 40% (excluding the breeding season) while our proposed foraging areas were significantly further than 10-km offshore and only an average of 12% of time was spent in the SAS activity. Why so much time was spent near haul-out sites, and, indeed, why seals haul out on land at all, is a matter of some debate (Brasseur et al. 1996; Watts 1996). A possible explanation is that seals may be safer from predation, for example by killer whales Orcinus orca L., at or near a haul-out site. Alternatively, sufficient food may have been caught during offshore trips and the periods at or near haul-out sites may be periods of rest or social interaction. A third possibility is that we underestimate foraging activity near haul-out sites.
Persistent, localized foraging areas were used by seals that hauled out at the Farnes and Abertay. Can this pattern be quantified and generalized? The number of seals hauling out may be censussed by land or aerial counts and the pattern of haul-out behaviour may be monitored by studies such as this one, conventional VHF telemetry (Thompson 1989; Hammond et al. 1992) or a series of direct counts (Grellier, Thompson & Corpe 1996). In addition, the dynamics of interhaul-out site usage may be estimated by mark–recapture models based on photo-identification data (Hiby 1994). The incorporation of such data (which were not collected during this study) would allow the absolute intensity of foraging of these seals to be mapped. Additional data on diet would allow the predation pressure on different prey species to be mapped. While this may be appropriate in providing a static map of predation pressure, the predictive power of this approach depends on the temporal and spatial predictability of prey abundance and of the prey preference of seals across age, sex, and season. We suggest that this approach may have some predictive power at the Farnes for three reasons. First, the repeated use of the localized foraging areas off the Farnes by all but two study animals (female F8–92 and female A2–93) for most of the tracking periods is a persistent feature across animals and seasons. Secondly, diet studies (Hammond & Prime 1990) have shown that, although the importance of sandeels decreased in spring, the predominance of sandeels in the diet at the Farnes during the pupping season (October to December) persists across years. Thirdly, sandeels have a requirement for specific sediment types, the location of which is stable. Thus, we suggest that the patterns of movement and foraging observed in this study may persist through time and among seals that haul out at the Farnes. This may not, however, be the case in other geographical regions if grey seals forage on prey which have chaotic or transient distributions.
Data on haul-out site usage and foraging patterns also aid interpretation of scat analysis results. Material recovered from scats provides information biased towards the previous 1–2 days’ feeding (Prime & Hammond 1987), the approximate time taken for prey remains to pass through the gut. If foraging trips were long relative to this period, data from scats may not be representative of diet if the prey species consumed changed during a trip. However, most foraging areas identified here were not far in time from a haul-out site; the mean return-trip length was 2·3 days and 75% of return-trips were of three days or less.
The large majority of our sampled trips to sea ended at the Farnes. Five of the 14 seals used the Farnes exclusively for hauling out. Only one seal did not use the Farnes at all, after capture and release. This persistent use of the Farnes as the main haul-out site, combined with the relatively short return-trip duration, gives confidence that inferences about grey seal diet based on scat analysis are justified in this respect.