Methods of Monitoring

1. Key Questions in Monitoring

Before designing a monitoring program, two key questions must be asked:

1. Why are you monitoring?

   • To understand spatial patterns and distributions

   • To detect changes in abundance or population status

   • To study life history parameters (birth rates, death rates, age structure)

   • To monitor health and pathology

2. What are you monitoring?

   • A whole population across its range

   • A specific area or habitat use

   • Population structure and behavioural ecology

Monitoring objectives affect which methods and metrics are most suitable.

2. Metrics Used in Marine Mammal Monitoring

3. Genetic and Biochemical Techniques

• Microsatellites & mtDNA → earlier population structure studies.

• Whole-genome sequencing → modern, fine-scale population resolution.

• Stable isotopes → reveal feeding ecology and trophic position.

• Telemetry & tagging → identify range, site fidelity, and movement.

• Biopsy sampling → allows DNA, contaminant, and hormonal analysis.

4. Visual vs Acoustic Detection

Visual Detection

• Advantages:

  • Direct counts → abundance estimation

  • Visual data on group size, age, sex, and ID

  • Relatively cheap and accessible

• Limitations:

  • Observer skill variability

  • Weather and sea state reduce detection

  • Some species respond to vessel presence (avoidance or attraction)

  • Group size estimation errors (especially large, social species)

Acoustic Detection

• Advantages:

  • Works day/night, all weather

  • Less biased by human error

  • Automated and continuous recording possible

  • Detects submerged animals

• Limitations:

  • Only detects vocal species

  • Social calling makes individual identification difficult

  • Expensive equipment

  • Background vessel noise can mask signals

  • Requires triangulation to estimate position

5. Observation Platforms

6. Case Studies

• Shetland Harbour Porpoise Surveys:
  Long-term land-based observations identified consistent hotspots on the north and east coasts.

• Aerial Surveys (e.g. SCANS Project):
  Conducted across North Atlantic and North Sea since 1994 → track trends in porpoise, dolphin, and minke whale abundance.

• Seal Surveys (Thermal + Visual):
  Harbour seals counted during moulting season (August) using thermal imaging and photography → effective for density and population trend estimation.

7. Photo-Identification (Photo-ID)

• Uses: Capture-recapture methodology to estimate population size, site fidelity, life history, and social structure.

• Requirements:

  • Distinct natural markings (e.g. dolphin dorsal fins, whale flukes)

  • Repeat surveys over time

• Limitations:

  • Some species lack visible marks

  • Changes in markings over time

  • Recognition errors

• Models:

  • Closed population → use simple capture-recapture

  • Open population → use robust/open models to account for immigration/emigration

8. Telemetry & Tracking

• Types:

  • VHF radio tracking (short-range)

  • Satellite tags / GPS tags (long-range, precise)

• Applications:

  • Movement ecology, home range estimation

  • Foraging and migration routes

  • Habitat use and overlap with human activity

• Example: Grey seals tagged in UK and Ireland revealed extensive inter-regional movement patterns.

9. Abundance Estimation Techniques

Relative Abundance

• Based on sightings per unit effort (SPUE).

• Cheaper but less robust — cannot compare across regions easily.

Absolute Abundance

• Requires systematic survey design and distance sampling:

  • Measure detection function based on perpendicular distance.

  • Calculate effective strip width (ESW) where detection probability = 1.

  • Correct for detection probability on the transect line (g(0)).

Core Formulae

Density (D)=n×sˉ2Lw\text{Density (D)} = \frac{n \times \bar{s}}{2Lw}Density (D)=2Lwn×sˉ​

where

• nnn = number of sightings

• sˉ\bar{s}sˉ = mean group size

• LLL = total transect length

• www = effective half-strip width

Abundance (N)=D×A\text{Abundance (N)} = D \times AAbundance (N)=D×A

where AAA = total area surveyed.

• Biases to consider:

  • Availability bias (animals underwater)

  • Perception bias (missed sightings)

  • Movement bias (avoidance or attraction)

• Precision & Accuracy:

  • Precision → confidence interval width

  • Accuracy → lack of bias in the mean estimate

10. Distance Sampling Example

• Transects designed to ensure equal sampling probability.

• Use program Distance to model detection curves.

• Sea state, observer height, species, and behaviour influence detectability.

11. Modelling & Habitat Analysis

• Density Surface Modelling (DSM):

  • Combines environmental covariates (depth, temperature, chlorophyll, etc.)

  • Predicts density in unsampled areas.

• Used to identify critical habitats and special conservation areas (SACs).

12. Life History and Health Monitoring

• Strandings & Necropsies:
  Identify cause of death (e.g., bycatch, disease, predation).

• Biomarkers:
  Blood, blubber, or faecal samples reveal stress, hormones, contaminants.

• Reproductive parameters:
  Pregnancy rates linked to prey abundance (e.g., krill-driven changes in baleen whales).

13. Examples of Population Trends

• Bottlenose dolphins (Moray Firth): Increasing numbers and range expansion.

• Harbour seals (North Sea): Decline in several colonies due to disease (e.g., Phocine Distemper Virus).

• Grey seals: Generally increasing, may be reaching carrying capacity.

• Common dolphins: Seasonal shifts and northward range expansion.

14. Monitoring Design Steps

1. Define objectives and spatial coverage

2. Gather baseline data

3. Assess population structure

4. Establish regular surveillance scheme

5. Conduct power analysis to detect trends

6. Combine distribution, abundance, and health data

7. Assess anthropogenic impacts (bycatch, noise, pollution)

8. Repeat every 6 years to detect long-term trends (EU reporting cycle)

15. Summary Table — Choosing Methods

Key Takeaway

Marine mammal monitoring is multi-method, combining visual, acoustic, genetic, and modelling tools.
Reliable monitoring requires:

• Clear objectives

• Consistent methodology

• Statistical correction for bias and precision

• Integration of biological, environmental, and anthropogenic data