CCAMLR

SeaBird is a generalised age- and/or stage-structured seabird population dynamics model that allows a great deal of flexibility in specifying the population dynamics, parameter estimation, and model outputs. The manual provides information on how to use SeaBird, including how to run it, how to set up the input files, descriptions of the population dynamics and estimation methods, and how to generate outputs. It also contains a brief overview of the technical specifications of the software, and examples of models using SeaBird. SeaBird is designed for flexibility. It allows the user to structure the modelled population in the way that best suits the available data. Depending on these data the user may which to specify the population structure using some or all of the following characteristics: age, life stage (e.g., immature or mature), sex, or behaviour (e.g., in any year mature birds may be classified as breeders or non-breeders). Interactions with fisheries can be modelled and the user can choose the sequence of events in a model year. A wide variety of types of data can be used. Estimation can be by maximum likelihood or Bayesian. As well as generating point estimates of the parameters of interest, SeaBird can calculate likelihood or posterior profiles and can generate Bayesian posterior distributions using Monte Carlo Markov Chain methods. SeaBird can project population status into the future under various alternative scenarios. SeaBird was designed to share many features and concepts with the fishery stock assessment model CASAL (Bull et al. 2005) and users of the latter program will find it easy to adapt to SeaBird. However, there are some important differences between the programs that are described.

ic life is developed in accordance with the rules of metrology, mathematical statistics and using the biocenological regularities of water objects. The basis of evaluating the stocks of water life was investigated through biocenotic conditionality in the areas of their equal probability in concentrations. The traditional method of squares in evaluating the stocks of water life was modernized with the use of probabilistic approach, the required knowledge of rules of statistical distributions of its specific concentration. Allocation of borders in the areas of probably equal concentrations of aquatic life applied in the proposed technique offer the mean integral values of probabilities. It is recommended to minimize errors of evaluation of the aquatic life resources using the statistical method of producing the average values if it is stated that casual component is more than two times higher of the regular component in a resulting error of evaluation of the average specific concentration of water life.

Quantitative method for describing krill mass congestions based on perennial observations using trawling and hydroacoustic data is proposed. Reliable evaluation of krill resources is provided with probability methods and spatial analyses requiring knowledge of statistical distribution rules applicable for natural habitat and equal probability concentrations. Spatial analyses of krill population densities have shown mixed rules of statistical distribution over its natural habitat. Data analyses procedures applied for trawling and hydro-acoustic sampling have to in compliance with the rules of statistical distribution and principles of metrology. Division of natural habitat of the Antarctic krill into regions is based on a principle of equal probability applied for congestions instead of the principle of equal proportions. Metrological features of evaluating population densities of krill are revealed. Standard measures for the population densities of krill in natural habitat are not available, thus it is principally impossible to structure systems and random errors in evaluating process, to define their impact on final evaluation of the resource. Special metrological principles are required and proposed for the correct evaluations based on reproducibility of statistic distribution parameters for krill and minimizing miscalculations in evaluations of congestions. Evaluations regime applied for the population density of krill should be defined by reliability and admissible error of estimated resource. Traditional concepts of observation system concerning the resources of krill should be further developed with applications of rules and parameters dealing with statistical distributions of population density, information about sources of errors, tools and methods of evaluation, standard techniques of minimizing errors in evaluations taking into account biological features of krill development. Methodical standards and uniform evaluation criteria for parameter evaluations of statistical distributions should be proposed to the Countries participating in the Antarctic Treaty as proceedings regarding minimal errors in the evaluation procedure. Reliable evaluation of krill resources requires advanced technical tools and observation systems. Population density values for krill should be calculated using Aitchison delta distribution. Primary statements regarding the krill in strategic planning for fisheries should include relevant biologically valid evaluations of population numbers for krill with the maintained reliability requirements and reasonable errors in evaluating the density of populations.

An updated version of the Spatial Multi-species Operating Model (SMOM) of krill-predator-fishery dynamics is described. This has been developed in response to requests for scientific advice regarding the subdivision of the precautionary catch limit for krill among 15 small-scale management units (SSMUs) in the Scotia Sea, to reduce the potential impact of fishing on land-based predators. The model includes krill as prey and four predator groups (penguins, seals, fish and whales) in each of 15 SSMUs. A number of updates have been made to the model such as linking krill growth rate to sea surface temperature, and these are described here. Moreover, the methodology used to condition the model using the WG-SAM set of reference observations for Area 48 (the SAM calendar) is described. Alternative combinations of model parameters essentially try to bound the uncertainty in, for example, the choice of survival rate estimates as well as the functional relationships between predators and prey. An example is given of how this Operating Model can be used to develop a management scheme which includes feedback through management control rules.

In this paper, we develop an ecosystem-based, precautionary management procedure for krill fisheries which draws together past experience in CCAMLR. It provides an empirical ecosystem assessment model, a decision rule for determining local scale catch limits based on a harvest strategy and a single-species assessment of yield, and a method for implementing the procedure. The decision rule for setting catch limits for a given harvest strategy has a straight forward expression of the target conditions to be achieved and the uncertainties that need to be managed and does not assume an understanding of predator-prey dynamics beyond that evident in the data. It is a natural extension of the current precautionary approach of CCAMLR for krill and can utilise existing datasets, including B0 surveys, local scale monitoring of krill densities, local-scale monitoring of predator performance, monitoring of predator foraging locations and time series of catches from the fishery. This procedure provides a common framework for inserting data, assessment methods and candidate modelling approaches for assessing yield. Consequently, its formalism means there is no need to undertake a staged approach in providing advice. The advice can be updated as improvements are made in any component of the procedure, including the provision of data, implementation of new assessment or projection models or a revision of the decision rule. This framework formalises the decisions that need to be made in dealing with an ensemble of food web models for providing suitably precautionary advice on how to spatially structure krill fisheries to account for the needs of predators. It provides the primary expectation for managing uncertainty, either by obtaining better estimates of parameters for the projection models and/or by altering the harvest strategy. Consequently, a preferred harvest strategy, which is initially untenable because of the uncertainties associated with its ecosystem impacts, could become a suitable option if its related uncertainties are reduced. Conceivably, the procedure outlined here could be used in a spatially-structured feedback management system that can ensure CCAMLR is able to respond to trends in the status of the ecosystem, including trends arising from climate change.

This paper details how FOOSA, an operating foodweb model for evaluating spatially-structured harvest strategies for krill, has been implemented within the EPOC modelling framework. It also shows how the parameters developed for use in FOOSA can be adapted for use in EPOC. The paper has three main parts – an outline of the structure of EPOC, consideration of the general FOOSA structure that needs to be implemented and a description of the implementation of FOOSA in EPOC. The latter section includes the methods used for implementing environmental variability, the krill population, generic predators, the krill fishery and the system for setting catch limits. The process of implementing FOOSA in EPOC has been a useful opportunity to consider the functions needed to represent different processes in a minimal realistic model. A number of revised functions are developed as options to reflect different dynamics that may be present in the krill-predator-fishery system in Area 48. Some of these functions and model structures have been generalised to enable more predators to be included in the food web and to provide flexibility in the number of stages of a predator consuming krill. An important step now in the implementation of FOOSA in EPOC is for this implementation to be reviewed by the developers of FOOSA.

We present a generalised spatially explicit Bayesian statistical catch-at-age population dynamics model (SPM) for developing and investigating plausible spatial movement models, and apply a preliminary development version of this model to Antarctic toothfish in the Ross Sea as an age and maturity state spatial movement model. SPM is an aggregate movement model suitable for use with large numbers of areas, and is implemented as a discrete time-step state-space model that represents a cohort-based population age structure in a spatially explicit manner. The model is parameterised by both population processes (i.e., ageing, recruitment, and mortality), as well as movement processes defined as the product of a set of preference functions that are based on known attributes of spatial location. SPM was designed to be flexible, allow for the estimation of both population and movement parameters based on local or aggregated spatially explicit observations, and optimised for speed. Model validation consisted of three types: implementation checking; development-driven unit tests; and comparative software evaluation. Comparisons with expected output from CASAL and movement processes coded in S+/R were essentially identical, and estimates of example parameters for models implemented in both CASAL and SPM gave essentially identical results. We have also developed a preliminary model for Antarctic toothfish in the Ross Sea and describe the spatial and population structure and processes, data, observations, and likelihoods used to estimate movement parameters. The model was a single sex model that categorised fish as immature, mature, or spawning. Observations included within the model were spatially explicit commercial catch proportions-at-age and CPUE indices. While we caution that model results are preliminary, we note that they appeared reasonable, and suggested immature fish were located in the southern Ross Sea on the continental shelf, mature fish were located on the continental slope, and spawning fish were located on the northern banks of the Ross Sea. The results also suggested that parameterising of movement based on latitude, depth and distance provided a significantly better fit to the observations than a model where depth was ignored. However, further development to the SPM model is required, including processes and observation classes to incorporate year class variability, stock recruitment relationships, tag-release and tag-recapture observations, and maturation state observations. Further, the current implementation of the MCMC algorithm in SPM is only partially complete, and there is some further work on parallelisation algorithms for MCMC that could be investigated. And, in order to address the questions of the adequacy of the Antarctic toothfish Ross Sea assessment model, SPM needs to be modified to allow simulation of observations from underlying movement parameters. Finally, once adequate models for Antarctic toothfish in the Ross Sea have been developed using SPM, the current assessment model (Dunn & Hanchet 2007) would need to be evaluated within a simulation-experiment in order to address current assessment model uncertainties.

Measures are developed which aim to summarise the quality of fishing event, catch, and biological sampling data from a fishing trip. In particular these measures aim to quantify the prevalence of position or time reporting errors, the diversity of catch, the extent to which catch data follow Benford's Law for the distribution of the first significant digit, whether length-frequency data have been collected as expected, and the reliability of length-weight measurements. Individually these measures can assist in assessing which data from a trip should be used in an assessment, and can also guide how these data can best be used.The quality of tag data is hard to assess. A methodology is developed to use data quality measures for other data sets to group trips on the basis of their overall data quality. Ongoing development of this method is intended to provide a consistent basis for selecting the tagging data set that is fitted in an assessment model.The data quality measures illustrate sometimes substantial variation in the quality of particular data sets from different trips in the Ross Sea Antarctic toothfish fishery. Cluster analyses suggests two groups of trips, one of which can tentatively be considered to have better data. Tags released by trips in this latter group have been recaptured at a higher rate than tags released by the other group of trips.

nnovative multivariate statistical modelling techniques make it possible to generate spatially comprehensive species distribution layers from discontinuous biological data, by fitting complex and scale-dependent relationship between species abundance and available environmental data. The resulting species-specific distribution layers have many potential applications. We apply one such method, called BRT (Boosted Regression Trees), to data on the distribution of Oithona similis, a small cyclopoid copepod which is abundant through much of near-surface waters of the Southern Ocean. A large dataset (>19 000 records) of abundances of O. similis were measured during the SCAR Southern Ocean Continuous Plankton Recorder (SO-CPR) Survey. We demonstrate that it is possible to obtain a relationship between both the abundance and the probability of presence of O. similis and the long-term, broad-scale environmental conditions of the location where the CPR sample was taken. These fitted relationships were used to estimate abundances of O. similis through the Southern Ocean. We present a number of methods for investigating the robustness of the prediction. (1) Non-spatial cross validation tested the relationship against data withheld from the fitting. We found that the data withholding of data from the model fitting must be done on a tow by tow basis as there is significant within-tow correlation. (2) Spatial cross-validation withheld data from particular geographic regions from the fitting process and used these subsequently to test the predictive accuracy of the model. These cross-validation methods showed that the fitted relationships explained 28–38% of the total variance in abundance (depending on method of cross-validation). The area under the ROC for the model predicting presence was 0.77 indicating good discrimination between presence and absence. (3) Spatially-resolved measures were used to test how well the environmental space of the predicted area was spanned by the environmental characteristics associated with the biological samples. This method was applied to the individual environmental data layers singly and to multivariate space defined by all environmental data layers together. The multivariate statistic was used to create a “mask” which excluded from prediction those geographic areas of the Southern Ocean where environmental conditions were not well represented by the SO-CPR sample locations.