I
am interested in describing and modeling patterns of spatial distribution of
pelagic fish species, in particular small-scale (<100 km) distribution patterns.
This is an interesting area of research. Small-scale distribution patterns determine
how a species of fish interacts with its predators and preys. Some predators
depend on the existence of fish aggregations of high density, and may not forage
efficiently if the same prey biomass is scattered in numerous smaller aggregations.
In addition, changes in spatial distribution at this scale have implications
for fisheries management. For example, some species of fish form dense aggregations
even when the total stock biomass is low. For these species, CPUE (catch per
unit effort) may not be a good indicator of stock biomass, because the fishery
can target those high-density aggregations obtaining high yields even when the
overall biomass of the stock is being depleted.
For
my thesis research, I am working with walleye pollock (Theragra chalcogramma).
Walleye pollock is the target of one of the largest fisheries in the world and
an important species in North Pacific and Bering Sea ecosystems. Although large-scale
distribution patterns are relatively well know from fishery data and stock surveys,
small-scale features have not been described in detail.
The first objective of
my research is to quantitatively describe walleye pollock spatial and temporal
distribution patterns in the Gulf of Alaska and Bering Sea. The description
will include
· quantification of scale dependent variability in biomass distribution
· inference and modeling of spatial autocorrelation of biomass distribution
· characterization of walleye pollock aggregations
· quantification of daily, seasonal and geographical differences in distribution
patterns
This description will be based on high-resolution acoustic data collected
by the Alaska Fisheries Science Center during walleye pollock stock surveys.
The
second objective is to produce a three-dimensional, dynamic simulation model
of walleye pollock distribution. A hierarchical simulation model is proposed
to generate initial two and three-dimensional static distributions of pollock
biomass. The model consists of a series of two dimensional geostatistical simulations
to obtain distributions of biomass and aggregation descriptors, coupled with
three-dimensional Boolean algorithms to simulate fish distributions at small
scales. Movement rules will be defined to simulate the dynamic behavior of pollock
aggregations, including movement, daily changes in aggregation location and
morphology and interaction among aggregations.
Image
from: Cohen, D.M., T. Inada, T. Iwamoto and N. Scialabba, 1990. FAO species
catalogue. Vol. 10. Gadiform fishes of the world (Order Gadiformes). An annotated
and illustrated catalogue of cods, hakes, grenadiers and other gadiform fishes
known to date.. FAO Fish. Synop. (125, Vol. 10):442 p.
Walleye
pollock (Theragra chalcogramma)
Echogram
representing acoustic data collected in the Bering Sea. The large masses near
the bottom are walleye pollock aggregations.
Example
of simulated horizontal distribution of walleye pollock acoustic density. The
simulation was carried out following a sequential gaussian aproach, and was
based on an exponential model variogram. Resolution: 0.5 x 0.5 nautical miles.
Sections
of the three dimensional distribution generated from the horizontal distribution
shown above. The acoustic density in each cell is fractioned in horizontal layers.
The proportions in each layer are assigned by a histogram of random numbers
from a beta distribution. The parameters of the beta distribution determine
the position of the biomass in the water column. The number of fish in each
cell was calculated using adult pollock target strength (-32 dB). Resolution:
0.5 nautical miles (horizontal) x 10 m (vertical).
Some
(very) preliminary modeling results..
To learn more about walleye pollock,
click here.
