| Understanding CCAMLR's Approach to Management (Download Text) (Download Figures) |
|
3. CCAMLR’s Approach to Management [return to table of contents] The principal institutional elements of CCAMLR are the Commission (a policy-making and regulatory body) and the Scientific Committee (a scientific body providing management advice). This management advice is based on assessments conducted by the two working groups of the Scientific Committee. One of these, the Working Group on Ecosystem Monitoring and Management (WG-EMM), is primarily concerned with assessing and developing advice on the krill fishery, and analysing data from CEMP. The other, the Working Group on Fish Stock Assessment (WG-FSA), develops management advice on fisheries other than the krill fishery. It also assesses the incidental mortality of seabirds and interactions of longline fisheries with other non-target species, such as cetaceans. The advice from the working groups is submitted to the Scientific Committee, which may refine it by taking into account additional information available to the Committee. The management advice is then referred to the Commission for consideration. 3.1 Directed Scientific Research – Collection of Data for Assessment Purposes CCAMLR draws on data from four main sources:
The CCAMLR Convention Area is divided into statistical areas, subareas and divisions (Figure 8), internationally agreed and recognised by the Food and Agriculture Organisation of the UN (FAO), which is responsible for collecting and publishing world fishery statistics. The three statistical areas are: Area 48 (Atlantic Ocean sector), Area 58 (Indian Ocean sector) and Area 88 (Pacific Ocean sector). The boundaries of statistical subareas and divisions within these areas, which were decided on general oceanographic and biological grounds, incorporate areas thought to contain relatively discrete populations of some species. The reason for dividing the Convention Area into subareas and divisions is twofold:
The stock concept is therefore extremely important in the definition of discrete areas. Although most stocks in the Convention Area are still believed to be confined to specific statistical subareas or divisions, some are now thought to be distributed over two or more, or are straddling stocks as defined by UNIA. Examples of these are:
The acquisition and analysis of data from the entire geographical range of such stocks is crucial for assessment purposes, but can be difficult because of the historical definition of statistical areas and the Convention Area itself. Fisheries catch data are reported to CCAMLR for each of the subareas or divisions in the Convention Area. Most data are now reported in fine-scale format (1° longitude x 0.5° latitude by 10-day period) or even, in some fisheries, haul-by-haul. This means that, if required, either smaller or larger areas than the statistical subareas and divisions can be defined for assessment purposes. However, the subareas and divisions are still the basic units for management purposes.
Central to any management regime is the acquisition of high-quality data, some of which come from scientific sampling, but many come from commercial fishing activities. Scientific observers on board a vessel can provide detailed information on its fishing operations. This is a separate responsibility from checking on compliance with conservation measures. The CCAMLR Scheme of International Scientific Observation was first implemented in the 1992/93 fishing season. It was designed to gather information on fishing activities in the Convention Area, including details of vessel operations, biological data pertaining to the species caught, and incidental mortalities of non-target species. The scheme operates through bilateral agreements between CCAMLR Members to exchange observers (i.e. an observer of one Member serves on a vessel of another Member). The scientific observers must be nationals of the Member that designates them, but Members fishing are still obliged to report information from their fisheries at regular intervals. Nevertheless, the CCAMLR Scheme of International Scientific Observation is often the most effective means of obtaining reliable data and information from fisheries, and also of educating the crews of vessels in the use of measures designed to reduce the incidental mortality of seabirds. The presence of observers on board longline vessels of CCAMLR Members fishing for Patagonian toothfish in the Convention Area is mandatory. In 1995, the Commission endorsed the Scientific Committee’s recommendation that 100% coverage by observers should eventually become mandatory for all finfish fisheries in the Convention Area.
Abundance estimates are essential for assessing stock sizes. Two main types of survey are used to estimate the abundance of fish, krill and squid species: acoustic surveys and net surveys. Acoustic surveys use calibrated echosounders that transmit pings of high-frequency sound vertically down into the water column from a transducer mounted in the hull of a ship moving along a predetermined course. Sound is reflected back to the ship by the sea floor and by objects, such as fish, that are in the water. The difference in time between the sound being transmitted and its arrival back at the ship is used to estimate the depth of the seabed or the targets in the water. The proportion of the sound energy that is reflected is used to calculate the quantity of individual targets present in the water column. Different species have different acoustic characteristics but, although this helps to identify the source of the sound reflection, the best way is to sample the species in the water with nets. Electronic and data processing methods are used to integrate the total quantity of reflected sound so that the integrated signal is proportional to the density of animals along the course of the survey vessel. The absolute abundance of animals is then estimated by calibrating the echosounder with known targets, estimating the sound reflected by individuals of the species of interest, and scaling the density to the total area of the survey. Examples of acoustic surveys are the krill surveys by the international Biological Investigations of Marine Antarctic Systems and Stocks (BIOMASS) Program – FIBEX (First International BIOMASS Experiment) in 1981 and SIBEX (Second International BIOMASS Experiment) from 1983 to 1985; the US Antarctic Marine Living Resources (AMLR) Program conducted since 1988/89 in the Elephant Island/King George Island Area; and the Australian krill biomass survey in Division 58.4.1 in 1996. In a net survey, trawl or plankton nets are towed through the water, or along the bottom, for a measured distance. A commercial trawl has a large mouth opening and a coarse mesh, so it catches large fish. By contrast, plankton nets have a small mouth area and fine mesh; although in theory they can catch fish and krill of all sizes, they cannot be towed fast and larger individuals can get out of the way. Consequently, both types of net survey are useful for assessing a stock. Commercial nets collect information on the larger, breeding part of the stock, while plankton nets give information on juveniles that will become recruits to the fishery in the future. The total catch of each species divided by the area or volume fished gives estimates of the densities of animals in the trawled area. By carrying out such hauls at random sites, the mean density for the survey area can be estimated. Examples of net surveys are the UK demersal fish surveys around South Georgia since 1988/89, and the krill and fish surveys conducted in the Elephant Island/King George Island region by Germany since 1977/78 and by the USA since 1988/89. Acoustics allow a large area of ocean to be surveyed relatively quickly, but the information acquired still needs to be assessed alongside biological information derived from net catches. Nets provide detailed information about small areas, but net surveys are time-consuming.
Biological parameters – principally reproductive characteristics, growth curves and natural mortality rates – are key components in all the types of yield calculations outlined in section 3.2(ii). Information on these parameters is collected during both scientific surveys and commercial fishing operations. The growth curve of a sample of fish is usually estimated by measuring their lengths and weights and plotting them against age. Length and weight are quite straightforward to measure, but estimating age is much more difficult. In the case of fish, this is usually attempted by counting the rings in scales or otoliths (bones found in the ears). These rings are laid down regularly throughout life, much like growth rings in trees, though not necessarily annually. However, reliable counts are often hard to obtain, particularly in older animals, because individual rings are either difficult to distinguish or their annual deposition cannot be validated. For crustaceans, such as krill, this method cannot be used at all, because they moult their exoskeletons and have no hard parts (except in their eyes) that are retained throughout their lives. However, species such as krill, which have a short, once-yearly spawning season and a life span of six to seven years, often exhibit distinctive modes in their length frequencies. These can be linked to their age because individual krill born in the same year (cohorts) grow at similar rates and are distinguishable from groups born in other years. These cohorts constitute the ‘structure’ of a stock. The rate of natural mortality, which is the rate at which animals die from predation, disease, parasites or senescence, is a notoriously difficult parameter to estimate for exploited populations – Antarctic fish and krill are no exception. The difficulty is that, when a species is being fished, mortality due to fishing is impossible to distinguish from natural mortality simply by examining the stock structure. A variety of methods is used in fish stock assessments to estimate natural mortality rates, ranging from general methods that relate growth rates to natural mortality for a large number of species, through to methods that involve taking a random sample (before commercial fishing starts) of animals, whose ages are then estimated. In principle, the latter methods are preferred because they make direct estimates of natural mortality, but the sample of age readings must be representative of the stock, and the population itself must be unexploited and in equilibrium. However, it is usually difficult to fulfil either of these requirements. Moreover, as in most marine species, recruitment fluctuates widely from year to year, so the numbers of fish of each age are highly variable. As a consequence, estimates of natural mortality (M) rates for a species sometimes vary considerably. Typical examples are mackerel icefish, for which reported estimates of M range from 0.2 to 0.6 and krill, with a reported range of 0.6 to 1.2. Thus, the key biological parameters used in assessments are usually subject to considerable uncertainty. In deterministic assessment models, such as virtual population analysis and yield-per-recruit analyses, which are widely used in fisheries conventions around the world, this uncertainty is difficult to take into account; further work is needed to develop a more systematic approach to evaluating the effects of uncertainty on the results. In stochastic projections, some of the effects of uncertainty in the parameters are already incorporated in the analyses by using a different value for the biological parameters in each of the many simulations used in calculating the future states of the stock.
In addition to assessing the status of exploited stocks, CCAMLR monitors selected dependent species in CEMP as part of its ecosystem approach. This program has two broad aims: to detect and record significant changes in critical components of the ecosystem in order to provide information for conserving Antarctic marine living resources; and to distinguish between changes due to the harvesting of commercial species and changes due to environmental variability, both physical and biological. The Scientific Committee realised at the outset that monitoring the entire ecosystem would be impossible. It therefore selected species in a few key areas and the parameters that were most likely to reflect changes in the ecosystem and the availability of harvested species, especially krill. The inclusion of a species in the program is also based on its likely utility in indicating the state of some part of the ecosystem that may be affected by fisheries. In addition, other environmental parameters, such as hydrographic and sea-ice cover information, were selected to monitor trends in the physical environment. Monitoring selected species and evaluating the numerical and functional relationships between them and other components of the ecosystem contribute to the detection and recording of significant changes in critical components of the ecosystem (Aim 1). Monitoring prey species and measuring environmental parameters and the links between these and predators helps to distinguish between changes due to harvesting and changes due to environmental variability (Aim 2). The species, their biological parameters and sites at which they are monitored were chosen to meet specified criteria. Prey species were selected for their key positions in Antarctic ecosystems and their potential as harvestable resources. These were krill, the Antarctic silverfish Pleuragramma antarcticum, Euphausia crystallorophias (which replaces krill as a prey item in some regions of the High-Latitude Antarctic Zone), and early life stages of fish. Predator species were selected if they feed predominantly on the prey species identified, have a wide geographical distribution, and represent important ecosystem components. In addition, sufficient should be known of their biology and sufficient baseline data of the parameters to be monitored should exist to construct a scientific monitoring program. On the current list are crabeater and Antarctic fur seals; Adélie, chinstrap, gentoo and macaroni penguins; Antarctic and Cape petrels; and black-browed albatrosses (see Annex II). A core set of sites was chosen from three Integrated Study Regions (ISRs) (Figure 9), and a wide network of complementary additional sites (Figure 9) was proposed. Within the ISRs, sites were chosen so that researchers could distinguish between broad-scale and local-scale changes, and between changes in fished areas and non-fished areas. However, their position was also limited by logistics, including the presence of established bases and the availability of long-term datasets. The selection of ‘control’ sites was very difficult because the geographical scale of the changes to be studied was expected to be large, and the sites had to be outside such large areas but with comparable environmental and biological characteristics, and also be suitable for long-term monitoring. Several parameters are monitored for each predator species. The geographic and temporal scales over which these parameters are expected to reflect changes in the ecosystem varies from several weeks and close to monitoring sites (e.g. the duration of foraging trips, composition of chick diets) to annual or semi-annual and region-wide (the weight of birds arriving to breed, breeding success, population size). Monitoring methods for the environmental parameters of sea-ice cover, local weather and snow cover have already been agreed. Sea-ice and hydrographic conditions influence both the distribution, abundance, movement (‘flux’) and recruitment of krill (Figure 10), as well as the distribution, rate of survival over winter, time of arrival and access to breeding colonies of its predators, such as penguins. The parameters for monitoring environmental conditions and the condition of prey species are currently being refined and developed further. WG-EMM guides CEMP, particularly the design and coordination of research, acquisition of data by standard methods, and centralised storage and analysis. This is combined with a strong emphasis on empirical and modelling-based research, which both modifies the monitoring approach in line with methodological developments and creates a sound scientific background against which the effects of management options on the Antarctic ecosystem can be assessed. The final link in the monitoring scheme is a management mechanism to regulate marine harvesting. Field work and data acquisition for the program are carried out voluntarily by CCAMLR Members. The data they collect are sent to the CCAMLR Secretariat, which carries out standard analyses for consideration by WG-EMM. The Secretariat also collects and archives data acquired from remote-sensing initiatives – for example satellite-derived sea-ice and sea-surface temperature data. WG-EMM analyses these data to arrive at an annual ecosystem assessment. Trends in the monitored parameters and anomalous years are identified for each species and site, and explanations for these phenomena are sought by examining the monitored parameters of harvested species and the environment. Since the establishment of standard methods for monitoring these parameters in 1987, CCAMLR has collected data from over 80 combinations of site, species and parameter. For some series, data are available from the late 1950s, but most data series start in the mid-1980s when CEMP was initiated. |