Linking animal population dynamics to alterations in foraging behaviour
Noise from ships and wind farms can cause marine mammals to change behaviour. They may forage less efficiently if disturbed, or they may avoid dispersing through some areas. Although behavioural changes are sometimes observed for individual animals, little is known about how populations are affected. The aim of this study is to investigate how the dynamics and survival of the harbour porpoise population in the Inner Danish Waters (IDW) is affected by the combined effects of noise and by-catch using an individual-based population model.
Published on: Mar 3, 2016
Transcripts - Linking animal population dynamics to alterations in foraging behaviour
Figure 1. Harbour porpoises are caught in pound nets used in commercial fisheries
and subsequently equipped with telemetry tags in order to study their movement
In order to study how the porpoise population responds to
different noise scenarios it is important to simulate the
movements of undisturbed animals as realistically as possible.
We used a combination of correlated random walk (CRW) and
spatial memory behaviour to produce movement patterns that
strongly resembled those of satellite-tracked animals (Fig. 2). The
spatial memory behaviour allowed the simulated animals to
navigate back to food patches they had visited in the past if the
CRW behaviour was not sufficiently profitable.
Figure 4. The move per half-hour time step in the
absence of noise is represented by the vector A. It is
calculated from a combination of CRW behaviour
and a memory of the locations of previously visited
food patches. The effect of noise is represented by B.
The length of B is proportional to the sound
pressure level, which decreased linearly with the
distance to the source. The actual move taken is
represented by D.
200 400 600 800 1000 1200 1400
Figure 2. Simulations produced tracks with the same autocorrelation structure in
turning angles and distance moved per half-hour step d as observed in telemetry
data. The panel to the right shows a simulated track. Details on movement
simulations are provided in Nabe-Nielsen et al. 2013 (Oikos, doi: 10.1111/
Figure 6. Average yearly population sizes over 30 simulation years for scenarios
including existing wind farms (e), planned wind farms (p), large ships (s) and
different yearly by-catch rates (b) measured in number of super-individuals. Error
bars indicate ±1 SD. Five replicates were run for each scenario.
The study illustrates how a spatially explicit, behaviour-based
model can be used for investigating the relative impacts of
various management scenarios. This has not previously been
attempted for marine mammals. Rather than providing exact
predictions, this type of study could be used to produce
worst-case scenarios and to identify parameters that have a
particularly large impact on population survival, and where
further research is needed.
In the model the population dynamics
emerged from the balance between
recruitment and mortality. The risk of dying
increased as the animals’ energy levels
decreased. Their energy use increased when
they were lactating and during winter,
causing the average energy level and the
population size to fluctuate (Fig. 5). The
parameters that controlled the animals‘
chances of giving birth, their temperaturedependent energy use, when they should
mate etc. were obtained from literature.
Noise caused the simulated animals to
deviate from the behaviour they would
have had in the absence of disturbances
(Fig. 4). This caused them to forage less
efficiently which in turn caused them to
have an increased risk of dying.
Mean porpoise energy
Food in patches
-150 -100 -50
Figure 3. Simulation including existing wind farms in the
IDW and large ships along the main shipping routes.
Porpoises were simulated as super-individuals, each
representing several real female porpoises. The colour
of the food patches changed from yellow to orange and
green as the food level increased.
p=0.0012, r= 0.534
p=0.029, r= -0.375
-150 -100 -50
Although the population carrying capacity decreased when
introducing wind farms and ships, none of the shown scenarios
resulted in a population collapse. Only the inclusion of by-catch
rates of ~10 percent resulted in continuous population declines.
When introducing wind turbines and ships into the model it
caused the equilibrium population size to decrease approx. 10
percent (Fig. 6). This was based on the worst-case assumption
that post construction noise from turbines caused animal
densities to be strongly reduced at a distances of up to 300 m
from the turbine, and the results also depended on how fast
food replenished in the patches.
The energy levels of the simulated
porpoises depended on their ability to
find food and the time spent foraging.
As the spatial distribution of the
porpoises’ prey is unknown, we
simulated food as being distributed in
randomly dispersed, equal-sized
patches (Fig. 3). Patches located where
porpoises frequently occur in nature
could contain more food than the ones
in less used areas. In the model the
porpoises consumed the food they
encountered on their way. Subsequently
the food level increased logistically in
the patches where it had been depleted.
Noise from ships and wind farms can cause marine mammals to
change behaviour. They may forage less efficiently if disturbed,
or they may avoid dispersing through some areas. Although
behavioural changes are sometimes observed for individual
animals, little is known about how populations are affected. The
aim of this study is to investigate how the dynamics and survival
of the harbour porpoise population in the Inner Danish Waters
(IDW) is affected by the combined effects of noise and by-catch
using an individual-based population model.
Figure 5. Variations in energy and number of
super-individuals. Food and energy in relative
measures. Each simulation year started on 1 Jan.
The study was funded by The Environmental Group under the
Danish Environmental Monitoring Programme and by the
European Union under the 7th Framework Programme (project
acronym: CREAM, contract number PITN-GA-2009-238148,