Geoscience Reference
In-Depth Information
III. RESULTS
A. Consequences of Behavioural Movement on Size Spectra
The spatial distribution of primary producers strongly influenced the spatial
distributions of all consumer-size classes with all asymptotic distributions
qualitatively approximating the phytoplankton distribution. For prey seeking
and combined prey-seeking and predator-avoiding behaviours, a stationary
asymptotic distribution was attained, but for predator-avoiding behaviour, the
central spatial points exhibited persistent, bounded oscillations about a fixed
average and the outer spatial points approached asymptotic limits. To make
general comparisons between distributions for different behaviours, we com-
pared averaged distributions over the past 10 years of each run.
Comparisons of prey seeking and control distributions (
Figure 1
) for three
size classes ((60, 3 and 150 g) selected because the largest class feeds optimally
on the medium class, and the medium class feeds optimally on the smallest)
revealed that the largest class moved to the centre of the grid as a conse-
quence of the higher prey abundance there. This concentration of the largest
individuals led to reduced abundance of medium-sized individuals and
allowed a higher proportion of the smallest individuals to persist with the
largest ones. The areas vacated by the largest size class led to increased
abundance of the medium-size class and a concomitant reduction in the
abundance of the smallest size class. Comparison of predator avoiding and
control distributions revealed that the largest class moved away from the
centre to avoid larger predators. The presence of the largest size class away
from the centre led to changes in the abundance of the medium-sized and
smallest size classes, both top-down effects. When both prey seeking and
predator avoiding were allowed to interact, the distributions approximated
the sum of the independent effects of the two behaviours (
Figure 1
).
The spatially aggregated size spectra (
Figure 2
) masked considerable local
variability in the size spectra that resulted from the different behaviours.
When the size spectra were aggregated over the entire space, there was less
than 1% difference in both slopes and intercepts (
Table 2
). However, when we
compiled size spectra for regions closer to and farther from the centre of peak
biomass, differences in slopes were revealed. For the central region, different
behaviours led to differences of less than 1% in slopes, whereas differences
between behaviours led to difference in slopes of 5-20% in the outer region.
For the outer region, prey-seeking behaviour caused the slope to become
steeper (
1.46), whereas predator-avoiding behaviour caused the slope to
become shallower (
1.17). Combining prey-seeking and predator-avoiding
behaviour also caused the slope to become steeper (
1.28) relative to the
control spectrum (
1.22). While variations in size-spectra slopes were much
