220
~ Grid Ba,cd Modcllini!
Settlement of Myti/us
In contrast to Lanice, the settlement of Mytilus is characterized by episodic spatfalls. Although larvae of Mytilus occur each year, this "normal" supply of larvae
does not seem to be able to establish new mussel beds or to sustain existing mussel
beds (Zens et al. 1997). Therefore we neglect this "normal" recruitment. Major
spatfalls of Mytilus often, but not exclusively, occur after ice winters. Thus, we
assume that after an ice winter a spatfall will occur with a probability of
plceSpat=O.9, whereas in years without ice winter the probability is pSpat=O.05.
Similar to Lanice, the abundance of Mytilus within a cell cannot exceed the capacity K M • By specifying how KM depends on topographic height (Fig. 8.4.2), we
define the parts of a sandflat where Mytilus may in principle become established.
The outcome of a spatfall depends on the situation in a cell. If only Arenicola is
present (which we assume to be present everywhere implicitly), every second
spatfall will be successful (pSpatSucc=0.5), i.e. the abundance of Mytilus will be at
its maximum value K" in the next year.
If Lanice is present in a cell, but not Mytilus, a spatfall will be successful with a
probability of pSuccession=O.8. In TOPOGRID, this is the only rule to take succession into account' Finally, if Mytilus is already present in a cell, a spatfall will
irrespective of the actual abundance of Mytilus turn abundance to KM in the next
year. In addition. the "age" of the local mussel bed in the cell is reset to one.
Disturbance events
Two kinds of disturbance events are modelled: storms or extreme wind events, and
ice winters. Ice winters occur with a probability of plce=O.1 per year. As for
Lanice, the strength of an ice winter is irrelevant because even rather short periods
of temperatures near O°C will kill the entire population of Lanice in the intertidal
(L1ceMort = 0.95). As for Mytilus, we distinguish between "normal", "severe", and
"extremely severe" ice winters, which reduce the abundance of Mytilus in each cell
by MylceMort = 10, 30, or 80 % respectively. The three kinds of ice winters occur
with a relative frequency of 2: I : I.
Storms occur on average every second year. Again, we distinguish between different degrees, i.e. "normal", "severe", and "very severe", which occur with a
relative frequency of 4:2: I. These storms reduce Lanice by LStormMort = 10, 25,
or 50 % respectively, and Mytilus by MyStormMort = 15,25, or 65 %. It should be
noted that all these mortalities and frequencies only reflect intuitive assumptions
and are not based on hard, quantitative data. If the model were to be fully analysed
(which will not be done in this chapter), these model parameters in particular
would have to be varied systematically.
Dispersal of Myti/us
If sustained by successful spatfalls, mussel beds increase in size. To model this, we
use a filter algorithm in TOPOGRID in which the abundance of Mytilus in a certain
cell is calculated as a weighted mean of the abundances in that cell and its eight
neighbouring cells. The abundance of the focal cell is weighted with the factor
weight (=200), and the sum of the abundances is divided by divi (=1000). Without
~ Grid Ba,cd Modcllini!
Settlement of Myti/us
In contrast to Lanice, the settlement of Mytilus is characterized by episodic spatfalls. Although larvae of Mytilus occur each year, this "normal" supply of larvae
does not seem to be able to establish new mussel beds or to sustain existing mussel
beds (Zens et al. 1997). Therefore we neglect this "normal" recruitment. Major
spatfalls of Mytilus often, but not exclusively, occur after ice winters. Thus, we
assume that after an ice winter a spatfall will occur with a probability of
plceSpat=O.9, whereas in years without ice winter the probability is pSpat=O.05.
Similar to Lanice, the abundance of Mytilus within a cell cannot exceed the capacity K M • By specifying how KM depends on topographic height (Fig. 8.4.2), we
define the parts of a sandflat where Mytilus may in principle become established.
The outcome of a spatfall depends on the situation in a cell. If only Arenicola is
present (which we assume to be present everywhere implicitly), every second
spatfall will be successful (pSpatSucc=0.5), i.e. the abundance of Mytilus will be at
its maximum value K" in the next year.
If Lanice is present in a cell, but not Mytilus, a spatfall will be successful with a
probability of pSuccession=O.8. In TOPOGRID, this is the only rule to take succession into account' Finally, if Mytilus is already present in a cell, a spatfall will
irrespective of the actual abundance of Mytilus turn abundance to KM in the next
year. In addition. the "age" of the local mussel bed in the cell is reset to one.
Disturbance events
Two kinds of disturbance events are modelled: storms or extreme wind events, and
ice winters. Ice winters occur with a probability of plce=O.1 per year. As for
Lanice, the strength of an ice winter is irrelevant because even rather short periods
of temperatures near O°C will kill the entire population of Lanice in the intertidal
(L1ceMort = 0.95). As for Mytilus, we distinguish between "normal", "severe", and
"extremely severe" ice winters, which reduce the abundance of Mytilus in each cell
by MylceMort = 10, 30, or 80 % respectively. The three kinds of ice winters occur
with a relative frequency of 2: I : I.
Storms occur on average every second year. Again, we distinguish between different degrees, i.e. "normal", "severe", and "very severe", which occur with a
relative frequency of 4:2: I. These storms reduce Lanice by LStormMort = 10, 25,
or 50 % respectively, and Mytilus by MyStormMort = 15,25, or 65 %. It should be
noted that all these mortalities and frequencies only reflect intuitive assumptions
and are not based on hard, quantitative data. If the model were to be fully analysed
(which will not be done in this chapter), these model parameters in particular
would have to be varied systematically.
Dispersal of Myti/us
If sustained by successful spatfalls, mussel beds increase in size. To model this, we
use a filter algorithm in TOPOGRID in which the abundance of Mytilus in a certain
cell is calculated as a weighted mean of the abundances in that cell and its eight
neighbouring cells. The abundance of the focal cell is weighted with the factor
weight (=200), and the sum of the abundances is divided by divi (=1000). Without