Current state of knowledge of effects of offshore renewable energy generation devices on marine mammals and research requirements

3 Spatial and Temporal Overlap

For a stressor to have any impact it is necessary, although not sufficient, for the stressor to overlap with the receptor, both in time and space. We thus first summarise the current state of knowledge of spatial, temporal and movement characteristics of UK marine mammal populations and the uncertainty around these. We follow this with a consideration of the spatial distribution of OREG activity and then highlight some locations that provide particular opportunities for answering general questions.

3.1 Marine mammal abundance and distribution

Estimates of the abundance of marine mammals are notoriously imprecise, and that makes the detection of changes, let alone the ascription of causes, difficult. In general, the precision of estimates of abundance is higher for larger populations. Seals are also easier to count than cetaceans (Taylor et al. 2007). The coefficient of variation ( CV) of estimates of the total size of UK seal populations are around
10-15% for large regions. The SCANS II survey of the European Atlantic and North Sea estimated harbour porpoise abundance over that area with a CV of 20%. With the exception of the intensively studied Moray Firth bottlenose dolphin populations, other cetacean species will have higher levels of uncertainty in abundance estimates, and therefore trends.

The uncertainty around estimates of absolute abundance and trends in population size will limit the detectability of the effects of any OREG developments on population size. The detection of small localised reductions in abundance is likely to require substantial additional monitoring efforts, both before and after construction.

There are few useful population estimates for cetaceans in UK waters and for several species there is only limited information on distribution and/or occurrence. Whereas some species are well distributed throughout the areas of interest ( e.g. harbour porpoise, minke whale, white beaked dolphin) others occur only sporadically or irregularly so there is uncertainty even about which species are present in sufficient numbers to be a cause for concern. Appendix 2 gives brief details of the seal and cetacean populations likely to be of interest to the OREG industry and considered in the licensing process.

3.1.1 UK seals

The Marine (Scotland) Act 2010 and Conservation of Seals Act 1970 require the Natural Environment Research Council ( NERC) to provide scientific advice to Scottish and UK Ministers concerning the conservation and management of seal populations. These tasks are carried out through annual meetings of NERC's Special Committee on Seals ( SCOS), using information mainly provided by SMRU. As well as examining the work directly funded by NERC for its use, the Committee also reviews other relevant projects, funded by the Scottish Government and other bodies, carried out by SMRU.

3.1.1.1 Haulout counts

Both grey and harbour seals haulout on land for periods throughout the year. Whilst the drivers and moderators of haulout behaviour (at least outside moult and breeding periods) are not well understood such counts are a useful, and readily obtained, index of local seal abundance. Harbour seal populations are surveyed using aerial photographic counts of individuals hauled out during their annual moult in August. While the intention has been to cover all locations where harbour seals haulout in a rolling five year cycle, the dramatic declines observed in many areas (Lonergan et al. 2007) have led to additional effort being concentrated in areas, such as the Orkney Islands, where the most rapid changes have been observed. Grey seals observed during these surveys are also recorded, and between 2007 and 2009 the surveys were extended to cover all areas where that species is known to haul out. In addition, there have been intensive studies of haulout counts at certain study sites, particularly in the Moray Firth (Mackey et al. 2008; Cordes et al. 2011).

With the exception of the Moray Firth ( e.g. Grellier, Thompson & Corpe 1996) and Kyle Rhea (Cunningham et al. 2009) there is little information on seasonal seal haulout distribution and numbers for any of the areas of interest. Planned telemetry work on harbour seals around Islay to examine movements into and out of tidal stream areas will help determine the extent of the area that may be affected by developments. Tailored monitoring/survey programmes for specific areas would allow development of seasonal haulout distribution maps and in conjunction with telemetry data will allow development of seasonal at sea habitat usage maps to determine the periods of maximum risk and the scales at which effects may operate (see HCON Research Gap below).

3.1.1.2 Pup counts

Grey seal pups spend their first few weeks ashore, so the main monitoring of that species around the UK is through aerial surveys of breeding colonies ( SCOS 2011). Total pup production at each colony is estimated from a series of surveys and used to generate abundance estimates from statistical population models.

3.1.1.3 Population modelling

Telemetry data can be used to directly scale up haulout counts to population estimates (Lonergan et al. 2011; Lonergan et al. 2012). Alternatively, demographic models can be used to allow for components of populations underrepresented in surveys.

Grey seal abundance is estimated by using state space models to extrapolate from the pup production estimates (Newman et al. 2009a). These require knowledge of demographic parameters such as survival rates and fecundity and an understanding of how these change with population density. The models produced very different estimates of abundance depending on whether density dependence was considered to affect fecundity or pup survival and were not able to select between those possibilities. The 2007-9 independent grey seal abundance estimate, generated from combining data from electronic telemetry tags with observations during the summer aerial surveys, resolved that issue (Lonergan et al. 2011).

Attempts have been made to build similar detailed models for harbour seal populations, though these have been hampered by the much more limited information that is available on harbour seal demography. Up until the 1980's, consideration of grey seals as competitors to fishermen led to culls and lethal sampling. These provided much of the currently available demographic data. The most similar dataset for harbour seals comes from the 1988 and 2002 phocine distemper epidemics (Harkonen & Heidejorgensen 1990; Harkonen et al. 2007). These data are mainly from Scandinavian animals and may not be representative of UK harbour seal populations.

Detailed models have also been constructed for aspects of local population dynamics, based on long-term observational studies (Cordes et al. 2011a; Matthiopoulos et al. 2011). In most cases a major limiting factor has been the shortage of background information. The lack of relevant information forces modellers to rely on intuition and judgement about what is plausible. Those assumptions and judgements are only testable by the collection of additional data so, in its absence, the validity of models' results remains uncertain. Attempts are being made to formalise the process of eliciting opinions from knowledgeable scientists (Lusseau et al. 2012), though it is hard to see how those can adequately fill gaps where there really are no data.

3.1.1.4 At-sea behaviour

Over the past 20 years, more than 200 grey and 200 harbour seals have been fitted with telemetry devices. In the past eight years, the technology has advanced from Argos satellite tags (infrequent, approximate locations) to GPS / GSM ^[1] tags (frequent, accurate locations plus high bandwidth channels to relay detailed behavioural data). The data holdings are summarised by (Russell et al. 2011)

In relation to tidal-OERG development in the north coast of Scotland and Orkney, McConnell et al. (in SMRU Ltd 2011) identified a number of data gaps. These included a lack of harbour seal movement and diving behavioural data in relation to high current regimes, especially in the Pentland Firth, and a lack of recent, high quality ( GPS/ GSM tags) adult and pup grey seal data in the same area. Since the report was completed (2011), some of these telemetry data gaps in areas of high current regimes have been filled:

date	reporting date	species	region	funder	comments
2011 ongoing	Jan 2013	harbour seals	Pentland Firth, Sound of Islay, Kyle rhea	SNH/ MS/ NERC	Describe movements and diving behaviour of harbour seals in relation to high tidal energy sites in PF, SoI & KR.
2013	June 2014	harbour seals	Islay and Jura	SNH/ MS/ NERC	Assess degree of movement into and out of the Sound of Islay to identify and if possible quantify the population at risk.
2010	2011	grey seal pups	Pentland Firth & Eday ( EMEC site)	MS	Describe movements and diving behaviour of grey seals during first year. Seals tagged at sites adjacent to tidal rapids.
2009 -2010	2011	grey seal pups	Anglesey and Ramsey (Wales)	WAG	Describe movements and diving behaviour of grey seals during first year. Seals tagged at sites adjacent to tidal rapids.

Simple descriptive summaries of the movement and dive data from grey seal pups have been presented (Thompson 2012b; Thompson 2012a). A detailed analysis of the harbour seal data has yet to be completed. Raw data from all the deployments detailed above have been incorporated in the at sea usage maps described in the next section.

Research gap
Title	Code	Details	Status	Reporting date
Behaviour of grey seal adults in relation to high current regimes in the Pentland Firth.	HCR	There will be significant tidal- OREG development in the Pentland Firth. There is a lack of adult grey seal movement and dive behaviour data in this region - especially in relation to areas of high current flows. GPS/ GSM tags will be deployed in this region to address this data gap.	Not funded	NA

3.1.1.5 At-sea usage

At-sea density of individuals may be estimated from haulout counts and haulout-specific foraging patterns using methods developed by (Matthiopoulos et al. 2004). Usage maps at 5km grid granularity have been prepared using all data up to the end of 2012 for both harbour and grey seals (Jones et al. 2011). This was a deliverable of Task MR5 (Characterisations of seal populations) under the MMSS/001/11 Research Project which reported in January 2013.

The usage maps present uncertainty in the form of upper and lower 95% confidence surfaces. Uncertainty can derive from a number of sources, but can be used to identify regions that are sparse in telemetry data. Such uncertainty would be reduced by strategic tagging of specific seal species and age classes in areas relevant to OREG developments.

Research gap
Title	Code	Details	Status	Reporting date
Telemetry studies targeted on specific areas to improve map confidence intervals.	TAG	In light of results of current telemetry studies (3.1.1.4) and results of MR5, targeted deployments on particular species and regions will improve confidence intervals on at sea distribution maps.	Not funded	NA

3.1.1.6 At-sea habitat preference

Whilst at-sea usage maps estimate usage density, they do not indicate why individuals form these distributions. Neither do they predict the consequences of any OREG-induced environmental change.

Research gap
Title	Code	Details	Status	Reporting date
Determine factors affecting UK grey and harbour seal habitat preference.	HAB	Using grey and harbour seal telemetry data, habitat preference will be assessed using a case-control strategy (Aarts et al. 2008). Abiotic variables ( e.g. depth, sediment type) will be used as candidate covariates.	Funded by MS & DECC	2014

3.1.1.7 At sea activity

The impact of overlapping OREG stressors and seals' density will be determined / moderated by the activity associated with the geographical area of overlap. Seal behaviour at sea can be conveniently divided into three main activity classes: resting, travelling and foraging. One of the metrics that discriminates foraging from travel is the speed of directed travel.

The rate of travel through an area of potential stress (rather than just the density of animals there) may affect the population level consequences of a local stressor. For example the population consequences of 10 seals each being exposed to one minute's exposure to a given level of piling noise may be less (or more) than one seal being exposed to 10 minutes of similar noise. In other words, the cumulative effect may be non-linear and dependent upon residence time.

Research gap
Title	Code	Details	Status	Reporting date
Map distribution and activity of UK seals	ACT	Behaviour using historical grey and harbour seal telemetry data will be classified into three states: resting (hauled-out or at the surface), travelling and foraging. To define these states we will develop existing state-space models based on track speed and tortuosity. The results will be used: 1. to generate usage maps distinguishing between foraging and travelling 2. to investigate changes in activity budgets resulting from at-sea developments 3. to identify core foraging areas. 4. modify usage maps to account for residence time	Funded by MS & DECC	2013

3.1.1.8 Meta-population structure

Telemetry data on the movements of both grey and harbour seals suggest that individual animals have preferred foraging areas or regions. For harbour seals such foraging areas are usually associated with haulout sites that are used throughout the year for resting, breeding and moulting. For grey seals such foraging areas may be hundreds of kilometres away from their favoured breeding locations and many grey seals spend most of the year well away from their breeding sites. The wide ranging movements of grey seals and the more localised movements of apparently resident harbour seals means that the two species will have different meta-population structures.

The Scottish Government has divided the coast into seven seal management regions (map available at http://www.scotland.gov.uk/Resource/Doc/295194/0112738.pdf). However, developing appropriate strategies for managing the localised disturbance effects of OREG developments requires an understanding of the structure of these meta-populations at a finer spatial resolution. An OREG stressor may produce a response whereby individuals move away (emigrate) from the source. A likely, measurable response is that lower numbers of seals will haul out locally. For individuals that move long distances ( e.g. grey seals) this response may be diluted geographically, to the extent that the response may not be detectable. For individuals that move less far ( e.g. harbour seals) the response may be more local and more detectable. The area needing examination to investigate such effects is likely to depend on the baseline patterns of movement of the animals as well as the location of the stressor source.

Research gap
Title	Code	Details	Status	Reporting date
Haulout connectivity of grey and harbour seals.	HCON	The network of movements between haul out sites will be mapped using grey and harbour seal telemetry data. We will generate a transition matrix, illustrating the probability of an animal originating from each haul-out moving to another haul-out or remaining at the haulout of origin. We will use telemetry data to parameterise these transition matrices. Uncertainty resulting from population size and number of animals tagged will result in confidence intervals surrounding these transition probabilities.	Funded by MS & SNH	2013

3.1.2 UK cetacea

Monitoring cetacean abundance is generally more difficult and expensive than it is for seals, because large areas of sea need to be surveyed to generate biologically useful estimates. Abundance estimates come from three general survey types - visual line transect surveys, acoustic surveys using passive acoustics to monitor odontocete abundance, and photo-id studies that are directed at closed populations. Electronic telemetry devices have been attached to cetaceans in other areas but not UK waters.

3.1.2.1 Coarse scale distribution

Less is known about the population status and distribution of most UK cetacea compared with UK seals. The primary synoptic information comes from the two SCANS surveys conducted in 1994 and 2005 (Hammond et al. 2002; SCANS-II 2008). These were large scale international collaborations that involved multiple ships and aircraft. Planning is underway for a third survey, hopefully to be carried out in 2015.

Throughout the UK, several organisations have also been conducting local surveys using acoustics, visual surveys including small scale line transect surveys and population studies using photo-id. Many or most of the visual sightings data series have been collated and standardised under the Joint Cetacean Protocol ( JCP), a collaborative project lead by the Joint Nature Conservation Committee ( JNCC). This project explored the potential of this information to supplement the SCANS data and examine localised effects. This project has been faced with two major constraints in its analyses: the differences in the data collection methods and formats, which require multiple assumptions and simplifications to be made, and computational complexity of fitting models to such a large and disparate dataset. The report of the third phase of the project is currently under review.

At present there has been no attempt to systematically collate or analyse acoustic data, and it seems unlikely that this will be achieved in the near future.

Photo-id studies have been conducted for only a few species, most notably for bottlenose dolphins, and here at least three UK based populations have been studied and abundance estimates have been produced (Pesante et al. 2008; Thompson et al. 2011)

The planned East Coast Surveillance Strategy currently in development by Marine Scotland, using various passive acoustic methods and aerial surveillance using high resolution cameras, will also provide useful data on the occurrence of cetaceans in the coastal region from St Abbs to Caithness.

Research gap
Title	Code	Details	Status	Reporting date
Review the utility of Joint Cetacean Protocol ( JCP)	JCP	1. Monitor and report on developments under the JCP and in particular where the tools being developed under the JCP analyses to address Favourable Conservation Status at the population level, are also developed in respect of the concerns at the smaller spatial scales of marine renewable development. 2. Monitor and report on the development of methods to combine existing acoustic and sightings data to best detect population trends. 3. Explore ways to generate probability of encounter estimates for specific OREG sites, and thus consider ways to define the "natural range" of cetacean species based on measures used for other species groups. 4. Explore ways to define optimal temporal and spatial scales at which cetacean density should best be examined in order to detect changes in density or distribution that are both statistically and biologically significant. 5. Examine existing baseline survey data, in order to assess how useful it is for determining changes in cetacean density or distribution and thereby to help refine data collection protocols to ensure that monitoring is fit for purpose.	Funded by MS	2013

3.1.2.2 Fine scale distribution and behaviour

There is limited information on the density of cetaceans in the coastal waters where wave and tidal OREG developments are planned. There are more extensive data from some wind OREG sites (but see above - JCP Research Gap is re-assessing value of JCP data) and in most cases, developers will be conducting local surveys.

At a smaller scale, presence and perhaps also local abundance of small cetaceans can be obtained by the deployment of passive acoustic detectors such as CPODs. Each of these devices contains a battery-powered hydrophone, processor and software to identify clicks produced by porpoises and dolphins. Most analyses currently report changes in the proportion of "click-positive minutes", which is taken as an index of relative abundance on the assumption that porpoises produce echolocation clicks most of the time. Other metrics ( e.g. waiting times and detection positive hours) may be more appropriate based on nature of devices, temporal autocorrelation issues and comparison with visual data ( e.g. Bailey et al. 2010, Thompson et al. 2010). CPODs are effective tools for assessing porpoise activity patterns at spatial scales of hundreds of square metres. The analysis of the resulting data remains problematic in terms of identifying the fine scale behaviour of individual porpoises since no simple way has been demonstrated to combine information from multiple devices or identify individuals.

Research gap
Title	Code	Details	Status	Reporting date
Estimation of harbour porpoise abundance from TPOD/ CPOD click detections	PTID1	Convert TPOD/ CPOD output (click-positive minutes) to an index of actual harbour porpoise density.	Funded by MS	2014

Until recently, fine scale behavioural studies have been more problematic. Although tagging of cetaceans using D-tags or Argos tags or equivalent has yielded useful information in other areas, no such tagging has been undertaken in the UK. However, recent OREG funded work has led to the development of towed hydrophone arrays and associated software that can track the movement of animals based on their echolocation clicks, opening up the possibility of examining fine scale foraging and movement patterns in specific targeted areas, such as those where OREG development are being planned. This technique has been used experimentally to track harbour porpoises in Ramsey Sound (Wales). However there is a need to extend this to other high current energy sites that may be exploited by tidal- OREG to provide sufficient data at appropriate resolution to allow us to describe porpoise behaviour in such habitats.

Research gap
Title	Code	Details	Status	Reporting date
Harbour porpoise behaviour in tidal rapids	PTID2	Use towed array hydrophone systems to detect and track the behaviour of vocalising harbour porpoises in the vicinity of tidal rapids associated with future tidal- OREG.	Funded by MS	2013

Whilst a large array of static hydrophones ( e.g. CPODs) may be used to investigate changes in large scale distribution over ranges of kilometres (Thompson et al. 2010b) they are of limited use in obtaining the fine scale movement data required to assess collision risk with tidal turbine arrays.

3.2 Possible indicator sites

The importance of baseline data for estimating impacts and changes resulting from the installation and operation of OREG devices means that these will be easiest to characterise in areas where there has previously been long-term and intensive monitoring. Concentration of effort at particular sites may provide more information overall than spreading resources evenly across all areas. The existence of previous studies however, is not in itself sufficient to identify appropriate locations for focussing research effect. How well an area represents the wider population, and the range of OREG technologies and other threats present are also important. It is also necessary that there is sufficient access to areas and populations to allow research to be carried out effectively and efficiently.

Long-term intensive studies are ongoing at the North Rona and Isle of May grey seal breeding colonies. These have been following and examining the reproductive success of individual females (Twiss et al. 2012). Similar studies of seabirds are also being carried out at the Isle of May ( e.g. Burthe et al. 2012).

The most intensively studied harbour seal population in Scotland is that within the Moray Firth (for example Cordes et al. 2011b). The long time series of behavioural and reproductive data from part of this population makes it an obvious candidate site for continued/further investigation of demographic processes and population responses to the impending large scale wind OREG developments. However, the Moray Firth population is not necessarily representative of the wider harbour seal population, e.g. while the population fell during the 1990s, that decline was much less than has been seen in other harbour seal populations in eastern Scotland. It is therefore important that additional sites are included in any programme of demographic studies of harbour seals. There are now too few harbour seals around the Firth of Tay Special Area of Conservation ( SAC) for much data to be gathered there. Orkney has seen steep reductions in harbour seal numbers, but still contains one of the largest populations and is an area where both tidal and wave OREG developments are concentrated, so that might be an appropriate area in which to focus efforts to understand impacts on that species.

Many cetacean species are very wide ranging and individuals are therefore only sporadically present in areas of interest to OREG. Of the two dozen or so species reported from UK waters, several can be reliably expected to be present in the areas currently under consideration for development. Specifically, porpoises and minke whales are likely to occur in the vicinity of any OREG developments in any marine area in Scotland, while bottlenose dolphins and white beaked dolphins could well be expected to occur fairly frequently in certain areas. Being the most abundant and widely distributed cetacean in UK waters, the harbour porpoise provides a useful 'model cetacean' species to study in respect of OREG developments.

Under Task MR7 of the MMSS/001/11 Research Project, fine scale harbour porpoise movement data will be collected using towed hydrophone technology. Actual and potential study sites include Orkney, Kyle Rhea, Sound of Islay and Bluemull Sound in Shetland. The Sound of Islay has been identified as a demonstrator tidal- OREG project by Scottish Government. There is clear potential for developing larger indicator sites to monitor porpoise density and seasonal movements in the Firth of Forth, Moray Firth, and west coast areas.