PhD Project 4

An oceanic species in coastal habitats: Risso’s dolphin distribution in Scottish waters and implications for conservation management content

An oceanic species in coastal habitats: Risso’s dolphin distribution in Scottish waters and implications for conservation management

This project is based at SAMS UHI

This project will be supervised by Dr Nienke van Geel (SAMS UHI), Dr Denise Risch (SAMS UHI), Dr Gordon Hastie (University of St Andrews) and a supervisor from NatureScot.

Risso’s dolphins generally occupy offshore habitats beyond the continental shelf, where they forage on neritic and oceanic squid. However, in the UK, and off western Scotland in particular, the species also exhibits a coastal distribution. This distribution means that they overlap with anthropogenic activities, and increases the risks of disturbance.

Risso’s dolphins are protected under various national and international regulations, and are a Priority Marine Feature in Scottish seas. In 2020, a Nature Conservation Marine Protected Area was established for the species off the Isle of Lewis (North-east Lewis MPA), representing an area with persistent high densities, and evidence that it may be used as a calving and nursing area.

This study seeks to improve understanding of the coastal habitat use by Risso’s dolphins. Understanding why Risso’s dolphins occupy coastal waters, when they are there, and what they are doing is key to quantifying risk and the importance of these nearshore areas. The results of this inter-disciplinary research will allow regulators to better monitor Risso’s dolphins, within and outside the MPA, and ensure their long-term conservation in Scottish waters, thereby promoting the sustainable progression of industries and economic growth in increasingly urbanised coastal seas around Scotland.

In recent years, the COMPASS (2017-2022) and SAMOSAS (2020-2021) projects have collected year-round passive acoustic monitoring data in Scottish coastal (6x) and continental shelf waters (10x), respectively (FIGURE 1). Building on these existing large-scale multi-year datasets, the project will allow the successful student to explore research topics including:

  • Development of an automated Risso’s dolphin click detector and classifier
  • Spatio-temporal distribution of Risso’s dolphin in coastal habitats

Using the long-term acoustic data collected in and near the MPA, a Risso’s dolphin echolocation click detector and classifier will be developed using machine learning approaches implemented in existing infrastructure (e.g. ORCA-SPOT [1], BANTER [2]).

The classifier will be applied on the COMPASS and SAMOSAS datasets, and the spatial and temporal distribution of Risso’s dolphin in nearshore habitats quantified. These results will be modelled as a function of temporally variable environmental, oceanographic, and topographic covariates using statistical modelling approaches, such as Generalised Additive Models, to investigate what drivers are underpinning coastal habitat use [3].

  • Risso’s dolphins foraging behaviour in coastal habitats
  • Habitat and geographic variation in Risso’s dolphin echolocation behaviour

It is currently unknown whether these Scottish coastal habitats are used as foraging areas and if so, what constitutes their prey. The acoustic data will be analysed to assess the presence and temporal distribution of Risso’s dolphin foraging activity [4]. Dependent on the results of this analysis, simultaneous passive acoustic recordings and visual observations may be carried out. These observations may be accompanied by hydro-acoustic surveys to identify and map prey fields and investigate dolphin diving behaviour [5,6].

Habitat speciation and geographic population structuring may manifest in variation in the species' vocal repertoire, which in turn may affect classifier performance. National and international collaborations provide access to coastal and deep-sea datasets from other regions. These datasets will be analysed to investigate the presence of habitat/geographic differences in click characteristics and assess classifier performance across datasets [7].

While the student will be primarily based at SAMS-UHI, the student will have the opportunity to spend time at the University of St Andrews’ Sea Mammal Research Unit. Additionally, they will be expected to spend time at NatureScot to gain valuable experience in practical conservation and learn about relevant policy backgrounds in which their work is embedded.

The start date of this project is: 3 October 2022

Figure 1. Overview of COMPASS and SAMOSAS acoustic monitoring locations relative to the North-east Lewis MPA and ScotWind leasing options for offshore marine renewables developments.

COMPASS and SAMOSAS acoustic monitoring locations map

The candidate should have strong quantitative analysis skills and ideally background in acoustic data collection and analysis. Experience in applying machine learning approaches and programming skills (e.g., R, Python, MATLAB) are strong advantages. This project will combine of fieldwork and analysis of acoustic data; the candidate should therefore be equally comfortable to plan and conduct work from small boats and in remote locations, as well as with the quantitative analysis of large datasets.

[1] Bergler et al. (2019). Doi: http://doi.org/10.1038/s441598-019-17335-w 

[2] McCullough et al. (2021). Doi: http://doi.org/10.1121/10.0005512 

[3] Soldevilla et al. (2011). Doi: http://doi.org/10.3354/meps08927 

[4] Arranz et al. (2018). Doi: http://doi.org/10.1242/jeb.165209 

[5] Benoit-Bird & Moline (2021). Doi: http://doi.org/10.1002/lno.11855 

[6] Hastie et al. (2019). Doi: http://doi.org/10.1002/aqc.3017 

[7] Soldevilla et al. (2017). Doi: http://doi.org/10.1121/1.4996002  

Contacts and supervisory team for this project: content

Contacts and supervisory team for this project:

Project specific enquiries: Nienke.vanGeel@sams.ac.uk

General enquiries: Graduate School Office gradresearch@uhi.ac.uk

Supervisory team (click name to view research profile):

Dr Nienke van Geel, SAMS UHI, Oban.