Cymraeg | English

Work Package 2 - Oceanographic Modelling

Numerical modelling of larval transport

  • Dr. Peter Robins (Bangor University)
  • Dr. Simon Neil (Bangor University)

State-of-the-art, three-dimensional hydrodynamic modelling of the Irish and Celtic Seas has been coupled with particle tracking modelling in order to simulate shellfish larval transport and dispersal. In this way, a much greater insight into larval behaviour can be gained than from field observations alone. The hydrodynamic models are used to produce hindcast and forecast simulations of sea surface heights, tidal currents, residual currents and temperatures (see Figure 1). We can compare the model outputs against real field observations and, once the model is well validated, use it to predict future climate change projections.

Separate particle tracking models use the simulated velocities from the hydrodynamic model to advect and diffuse individual particles representing larvae. Millions of larvae-particles can be released at any given time from different spawning locations and tracked forward in time for a typical larval stage (e.g. 28 days). The particles are given biological traits, most important being vertical migration behaviour. Some organisms swim vertically in response to sensory cues, such as pressure, light, salinity, or temperature. Larval age and weight may also influence their migratory pattern. For example tidally-synchronised organisms (e.g. the shore crab Carcinus maenas) swim upwards during flood tides and downwards during ebb tides, in order to increase retention near the parent population in estuaries. Other species (e.g. the great scallop Pecten maximus) swim upwards during dusk and downwards during dawn, to minimise predation. Traits such as settling criteria and mortality can also be incorporated into the algorithm. Final stage dispersal probability distributions can then be mapped (Figure 2) which show us where larvae from one release site are likely to go to.

The models can be used to address important questions relating to larval transport:

  • How larvae travel and how they adapt their migratory characteristics to the local environment to promote survival and production
  • What is the likely connectivity between geographically distinct sub-populations
  • How climate change is likely to affect larval dispersal and connectivity in the future

Figure 1. Hydrodynamic model output of sea surface temperatures and depth-averaged velocities in the Irish Sea. The snap-shot shows a flooding tide during June 1990. Areas of stratification occur where temperatures are warmer in shallow coastal regions, such as in Cardigan Bay and in the western Irish Sea (where the western Irish Sea gyre generates anti-cyclonic residual circulation during summer). These residual flows are fundamental to larval dispersal.

Figure 2. Dispersal probability distributions for patches of 10,000 passive (no vertical migration) larvae-particles. Each panel shows dispersal probabilities after 28 days from releases in the Dyfi Estuary, Wales, for different release dates (April to September, 1990, using velocity output from the hydrodynamic simulation). The colour scale indicates the proportion of the patch that is in each 2 km * 2 km model cell. It is clear that as summer heating increases and density-driven residual currents develop, the dispersal patch increases.