Biophysical models have become the ‘go-to’ tool for predicting the dispersive
trajectories of planktic marine organisms, and are used to design Marine Protected
Areas (MPAs), identify pathways of invasion, and understand metapopulation dynamics
and biogeography. Yet, despite this relatively long history of development, continued
technological advancement and increased usage, models continue to often fail to predict
patterns in nature.
As biophysical models are able to accurately predict the dispersal of abiotic particles, it
is argued that it is how we incorporate larval behaviour (sensu vertical swimming) in
models that may be decoupling predictions of dispersal and species distribution
patterns. Yet, despite the recognised importance of vertical distribution/position to
dispersal by advection, especially in smaller organisms, there is currently no general
consensus of how, what, and when behaviours can and should be included in models,
perhaps because the drivers of larval behaviour are inherently complex and as yet, not
fully understood (Chapter 1). The typical approach is to parameterise behaviours as
‘rules’ based on laboratory observed responses to cues, but it this approach appropriate
given the complexity of larval decision-making in the presence of multiple cues in
nature in comparison to single cue responses in controlled environments? In Chapter 2 I
explored the movements larvae must undertake to achieve the vertical distribution
patterns observed in nature. Results suggest that behaviours are not consistent with
those described under the Tidal Vertical Migration (TVM) hypothesis, instead, showing
a need for swimming speed and direction to vary over the tidal cycle -- with upward
swimming needing to be 2.5x faster than downwards swimming and a change in
direction from upwards to downwards needing to occur around the midpoint of the
flood tide - and low model compatibility during the ebb tide. Next, I looked to identify
the environmental drivers of larval vertical distribution during the ebb tide, where the
model compatibility of Chapter 2 was low. Explored external drivers (density, salinity,
temperature, turbulence) explained very little of the vertical distribution patterns of the
larvae, however, results suggest differential usage of environmental cues based on
ontogenetic stage, and vertical distribution patterns observed differed from previous
observations of a similar species at a different location. Finally, I presented a
framework for assessing how behavioural parameterisation can influence dispersal
trajectories in marine systems (Chapter 4), comparing a novel approach of reverse
engineering larval swimming from in-situ observations (Chapter 2: REVM behaviours)
against simulations adopting passive dispersal, and particles attributed a tidal vertical
migration (TVM) signature. Results highlight how the implementation of behaviour
within biophysical models can lead to fundamentally different dispersal outcomes, and
specifically, that the inclusion of vertical migration behaviour is a mechanism that
significantly reduces dispersal distances, but depending on the approach to
implementation can lead to fundamental differences in dispersal direction.
This thesis makes significant steps towards improving the parameterisation of behaviour
within dispersal models by considering larval movement as a manifestation of
behaviour influenced by the larva’s in-situ environment. The methodologies and
analytical techniques designed and applied within the data chapters can be applied to
any species with a planktonic dispersal phase in any location, and provide an important
step towards improving the biological ‘realism’ of behavioural parameterisation in
dispersal models in the absence of an understanding of the complex drivers of active
larval movement
Date of Award | 2021 |
---|
Original language | English |
---|
Awarding Institution | |
---|
Supervisor | Antony Knights (Other Supervisor) |
---|
- Larval Behaviour
- Lagrangian Modelling
- Vertical Migration
- Biogeography
Mechanisms of movement in meroplankton: A primer for dispersal
James, M. K. (Author). 2021
Student thesis: PhD