US AMLR Stock Assessments
Stock assessments to estimate the status and productivity of
biological populations are a central tool in the management of the harvesting
of living resources. These methods are used to determine sustainable catch
levels that can be demonstrated to meet long-term management criteria.
The role of AMLR in assessing the sustainability
of the fisheries in the CCAMLR convention area is two-fold. First, AMLR
scientists participate with other CCAMLR member nation scientists in working
group discussion and review of assessed, exploratory, and research fisheries in
the CCAMLR convention area using existing quantitative tools. The eventual goal
of this work is to establish that all allowable fisheries removals for each
fished stock are sustainable. Second, AMLR scientists participate in the
development of new assessment frameworks that can incorporate additional
ecosystem and food web considerations and data into the formal models used to
establish catch limits.
Traditionally, fisheries stock-assessment is based on single-species
models, which for the most part consider the productivity of a fished stock in
isolation from the rest of its environment. The effect of the environment
enters traditional models indirectly through such parameters as natural
mortality (often estimated as a single constant representing an average value
over all ages and years) or recruitment (sometimes correlated with
environmental data but often estimated as an independent value or time series).
Conventional stock assessments use as data such measured
quantities as total catches, age- or length-structure of the catches, and
possibly other indices such as catch-per-unit-effort (CPUE) or non-fishery data
from research surveys, to estimate the parameters of a quantitative model for
the fishery system. These single-species frameworks for estimating model
parameters and derived quantities of interest have a well-developed statistical
foundation but ignore much of the ecological detail that would be required to
recognize ecological phase transitions such as regime change or shifts in the
scale of variability that could affect long term productivity or resilience.
Catch limits for the three most important (by weight and by
economic value) harvested species in the CCAMLR convention area, Antarctic krill
) and Antarctic toothfish (Dissostichus mawsoni)
using single-species models. The CCAMLR "precautionary principle"
defines these catch limits in terms of changes that could be reversible within
two or three decades if harvesting ceased.
Krill harvest targets are established using
the "Generalized Yield Model" (GYM), a simulation framework developed
over several decades. Toothfish catch limits in most assessed areas are set
using the "CASAL" integrated modeling framework, although limits for
exploratory or research fisheries in areas without formal assessments may be
set using GYM, comparative-CPUE, or similar approaches.
All of these modeling approaches are
simplifications (see Figures 1 and 2 below). The goal is to capture the important aspects of
a system in a quantitative way so that the likely effects of alternative
management options can be estimated and understood.
Figure 1. Distributions of krill lengths measured in net
trawls during 19 years of January and February surveys in four areas* of the
Figure 2. Model distributions of krill lengths for January
and February in four areas of the Antarctic Peninsula produced from a stock
For additional information on stock assessments prepared by the AERD, contact Doug Kinzey.
*The four AMLR study areas are Elephant Island (EI), Joinville, South and West, as shown below: