Estimates of Collision Risk of Harbour Porpoises and Marine Renewable Energy Devices at Sites of High Tidal-Stream Energy

Section 3: Overall Discussion and Implications for Collision Modelling

3.1 Key findings

1. We set out to visually and acoustically survey the Sound of Islay and the Kyle Rhea and surrounding waters for harbour porpoises in the summers of 2009 and 2010. To allow comparison, we used a research boat ( HWDT's Silurian) that has surveyed much of the rest of western Scottish waters for cetaceans over the preceding decade. However, using a survey vessel that cruises at speeds only marginally faster than the high energy tidal flows in the two tidal narrows presented the possibility of significant survey bias for both visual and acoustic recording methods. To resolve this, we adapted the trajectory of the conventional survey path with respect to the flow so that the boat crabbed across the flow at a rate equivalent to progress over the bottom. We tested this new method and successfully applied it to the Sound of Islay and Kyle Rhea. Survey effort was conducted for a total of 16 days and covered over 1300 km of sea.

2. Harbour porpoises were seen and acoustically detected in all areas surveyed. Of particular focus for this study were the areas of strongest tidal flow in the Sound of Islay and the Kyle Rhea because of the interest in these for commercial tidal stream arrays. Porpoises were seen and acoustically detected in both of these sites. However, rates of sightings or detections were an order of magnitude lower than surrounding waters, particularly the Sound of Jura and northern Sound of Sleat. For example, sightings in the Kyle Rhea narrows were around 0.01 km ^-1 whereas to the south in the Sound of Sleat: they ranged from 0.09 - 0.22 km ^-1. As with other studies, porpoise sightings were heavily influenced by sea conditions but this effect did not explain the low sightings rates in the tidal narrows because these areas, due to restricted fetch, were some of the calmest surveyed during this study (Tables 5 and 10). Acoustic detections showed a similar pattern with 0.2 click events.km-1 in the narrows and 5.41.km ^-1 in the open water to the south. Rates in the Sound of Islay were similar (sightings in narrows 0.01 km ^-1 vs. 0.08/km and 0.24/km to the south; acoustic detections 0.02 click events.km ^-1 versus 0.09 click events.km ^-1).

3. Other than demonstrating that porpoises use these areas and occur at lower densities than in adjacent waters it was not possible during the boat surveys to determine their activity (foraging, transiting etc.) while in these highly tidal areas. Nevertheless this finding considerably strengthens suggestions of low use of tidal sites derived from more wide-scale surveys on the west of Scotland (Embling et al., 2010 and Booth 2010) and contrasts markedly with observations of porpoises targeting tidal-steam sites in Wales (Pierpoint 2008).

4. Other marine mammal species were seen on the surveys. Of particular interest were the sightings in the potential tidal-energy areas. Most abundant were harbour and grey seals which were seen in high numbers both in the water and hauled-out in these locations. Both species were also seen in more open waters but at much lower rates. A minke whale, bottlenose dolphins and an otter were also observed in the tidal narrows. No basking sharks were seen.

5. In terms of cetaceans, these survey results indicate that porpoises (as well as other, less frequently encountered cetaceans) do not appear extremely abundant within tidal races in Kyle Rhea and the Sound of Islay, but do occur in them and especially in their close vicinity. It must be remembered that abundance does not equal importance because these areas are likely to be required for transit between other water masses. Bottlenose dolphins may be an exception because their abundance on the whole west of Scotland (Thompson et al., 2011) is very low so observing them in one of these narrows was both unlikely and potentially significant.

6. The use of moored acoustic porpoise detectors ( C-PODs) proved problematic both from the practical perspective of keeping them moored but also because they experienced high levels of noise that substantially reduced their detection capabilities when the tide was running. Deployments out of (but adjacent to strong flows) could be used to detect porpoises in the neighbouring tidal-streams and provide useful information on temporal patterns. Wide variations in temporal occurrences detected using PODs sited relatively close to one another suggested that multiple detectors in a variety of locations would be best to characterise porpoise use of even relatively small sites. In addition other methods such as visual observations should be used to validate these findings.

7. We also explored the use of C-PODs allowed to drift freely in the current as part of a location logging drogue system. These proved highly successful and rapidly revealed similar patterns of porpoise distribution to the more intensive boat based surveys. That said, because these devices drift with the current, interpretation of the resulting data requires more careful interpretation than towed acoustic methods. In addition to mapping porpoise detections the drifters also revealed spatial patterns in the distribution of the problematic ambient noise experienced by the moored PODs. It may therefore be possible to use this method to explore for areas most suitable longer term device moorings.

8. Finally, we tested a new acoustic array-based method to determine the depth of echolocating porpoises (Gordon et al., 2011). The advantage of this method is that it can be deployed from a drifting boat in flowing water and is therefore workable in tidal-energy sites where other methods would not be. The array was deployed in the relatively still waters of Loch Duich and then in the moving water in the northern Sound of Sleat. The method proved successful and porpoises were both detected and their diving depths resolved. Overall, the depth of origin of more than a thousand clicks were determined. Although this was a test of the method rather than a definitive trial, the dive depths of porpoises were surprisingly shallow for this species (mean call detection depth 10 m, maximum 28 m) despite the recordings being made in water than was between 50 and 100 metres deep. This suggests that diving behaviour data collected from other habitats may not be directly applicable to tidal energy sites such as the Kyle Rhea.

3.2 Implications for collision modelling

The encounter model developed for the Scottish wet renewables Strategic Environmental Assessment (Wilson et al., 2007) aimed to look at how frequently harbour porpoises would share the same time and space as operating tidal turbines. Given that we know that porpoises are likely to respond to turbines in some way, this was not a collision model but rather an opportunity to examine how often animals and machinery might encounter one another. Based on a number of assumptions, this model found that around 13 porpoises would encounter an operating turbine each year at close range unless it took avoiding or evasive action. Without other information of habitat preference then available, the model assumed that porpoise density in tidal-stream energy sites off western Scotland was equivalent to the density in the rest of their habitat. This present study was set up to test this equivalent-density assumption.

In both sites investigated, we found that porpoises were around an order of magnitude less abundant than surrounding waters. Furthermore because we used the same survey platform (Hebridean Whale and Dolphin Trust's Silurian) as used for other surveys on the west coast of Scotland it was then possible to compare our results with those over a much wider area equivalent to the SCANS survey block whose density was used in the original modeling study.

Combining all survey data between 2003 and 2010, observers on the Silurian sighted porpoises at a rate of between 0.039 - 0.058 porpoises.km ^-1. This rate was strikingly similar to the sighting rate found for all of the areas surveyed in this study combined (0.04 - 0.07 porpoises.km ^-1, SD 0.25-0.27). Thus when all of our survey effort was combined then it was broadly typical of the west of Scotland ( i.e. from the Kintyre Peninsula north to Cape Wrath and west to St. Kilda including the Minch, inner and outer Hebrides). However, our sighting rates in the tidal narrows (both areas: 0.01 porpoises.km ^-1, SD 0.10-0.19) were substantially less (an order of magnitude) than our most abundant areas and crucially an average or around five and a half times less than the average rate. If we are to add this to our encounter model then the rate of potential interaction falls substantially from 13 to around 1.8 to 3.25 porpoise-turbine "encounters" per year. However it must be emphasized that this number is not a collision rate, it is simply a rough indication of how often porpoises would encounter turbine blades if they took no action to approach or avoid them. Nevertheless this study suggests that, for the two sites at least, porpoise-turbine interactions are likely to be substantially rarer than if turbines had been deployed in other habitats.

Likewise, our preliminary investigations of porpoise vertical distribution using the suspended array suggested that the animals tended to use the upper layers of the water column in tidal areas. Accordingly, it is possible that these animals may encounter mid-water turbines less than otherwise expected. That said our investigations using the vertical array were mostly in waters adjacent to but not directly in key areas of interest to tidal stream developers. This is ironically because we couldn't find sufficient porpoises in fast flowing water for the method evaluation.

3.2 Implications for tidal-stream energy developments

1. Studying porpoises in moving water using conventional techniques is challenging and throughout this study we have had to substantially modify existing and develop new methods to gather useful data. Accordingly, it is very likely that the same will apply to monitoring around active turbines in these Scottish sites or elsewhere.

2. The low rates of porpoise detection in the two tidal-sites investigated means that potential rates of interaction between porpoises and tidal energy devices are likely much lower than would be experienced if these devices were sited elsewhere. That said, these sites are likely to be used as transit routes for animals living in waters either side of the narrows. Accordingly the day to day rate of interaction may be low but that is not to say that the number of individuals potentially passing through the area will be any lower than for more open water areas. Accordingly the opportunity for individuals to become familiar (learn, habituate etc) with both the site and the device(s) may be less.

3. The very low abundance of porpoises in the key sites of interest made it difficult to look in more detail at their behaviour in areas or fast moving water. In fact, this lack of sightings prompted us to cancel our plans for shore based observatory in favour of mobile acoustic methods. As a result it was impossible to determine what the animals' activities were in these sites (active foraging, simple transit etc). Accordingly, such low sightings rates will make monitoring of interactions between porpoises and tidal turbines difficult simply because of the low abundance creating low statistical power to detect patterns.

4. The profound differences in porpoise occurrence observed between these Scottish sites and the Welsh observations, in concert with 1) the suggestion of different temporal patterns of porpoise occurrence from PODs close to each other and 2) the surprisingly shallow dive depths documented, suggest that general patterns of porpoise ecology may not be easily applied to tidal-energy sites. It is clear that our understanding of this species in moving water needs further targeted study for generalities to be better established for this species. This is particularly true for tidal-energy developments in more open habitats such as waters to the west of Islay or Pentland Firth where studies such as those described here have not been performed but the eventual commercial scale developments are likely to take place.