CCAMLR Science, Volume 24 (2023)
The purpose of CCAMLR Science is to publicise the science done in CCAMLR, especially the science supporting management of marine living resources, and to raise the profile of CCAMLR within the international scientific community. A renewed effort following a pause since 2016 has led to a focus on special topics based on submissions to working groups and approved by the Scientific Committee beginning with Volume 24, which will comprise papers relating to the development of the krill management approach.
Proportional recruitment summarises the variability of new individuals entering a population over time. Two parameters characterising proportional recruitment, the mean and standard deviation of the interannual proportion of juveniles in the population, are important inputs to the generalised yield model (Grym) when the proportional recruitment option is being used to set fishery catches. The Grym is a simulation framework that can define the amount of fisheries catch that is considered precautionary as defined by decision rules. It is currently under consideration by CCAMLR for managing catches of Antarctic krill. This study calculated proportional recruitment of krill from seven data sources in Subarea 48.1 representing research trawl surveys, fishery observer data and predator diets. Krill length-frequency distributions provided values of proportional recruitment from each of these data sources using a range of alternative upper length bounds (‘thresholds’) from 30 to 44 mm for defining juveniles. All datasets tracked the same interannual peaks and troughs in proportional recruitment. Proportional recruitment parameters calculated using the alternative thresholds from the same datasets varied widely. Across all data sources and thresholds, the interannual mean proportional recruitment of krill varied from 0.02 to 0.76 with standard deviations varying from 0.03 to 0.3. The choice of length threshold had a larger effect on the proportional recruitment parameters than differences among datasets. The potential importance of size selectivity in krill samples, especially if smaller bounds on the juvenile length threshold are assigned, could require adjusting observed frequencies for the lower selectivity of smaller individuals. These results highlight the importance of deciding which upper length bound and which data source(s) to use to identify juveniles in calculating the parameters to be supplied to the Grym. |
The objective for this research was two-fold: (i) to provide updated estimates of the biomass and distribution of krill in the Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR) Statistical Area 48, and (ii) to develop knowledge on the marine environment essential for the implementation of an adaptive management system for Antarctic krill. Survey design followed the transects of the CCAMLR 2000 Krill Synoptic Survey of Area 48 and of national surveys performed in the South Atlantic sector of the Southern Ocean by the People’s Republic of China, the Republic of Korea, Norway, the United Kingdom and the United States of America. The survey also focused on high krill-density areas and employed state-of-the art methods and technology. The future management system will need standardised acoustic data from fishing vessels to be collected, processed and reported in near real-time as a measure of the available prey field. This information can be integrated with finer-scale knowledge of krill predator feeding strategies and updated through specific scientific studies at regular (multi-year) intervals. To aid such implementation and to encourage the development of future management tools, the survey took place during the austral summer of 2018/19. The work was coordinated by Norway and involved collaborative international efforts of six survey vessels provided by the Association of Responsible Krill harvesting companies and Aker BioMarine AS, the People’s Republic of China, the Republic of Korea, Norway, Ukraine and the United Kingdom. This paper reports on the main research priorities and data collection performed on board the Norwegian research vessel Kronprins Haakon in January and February 2019, and the land-based predator work carried out between November 2018 and February 2019. |
While the influence of spatial scale in ecology is well established, few studies have evaluated the impact of spatial scale on hotspot analysis of biological resources. Using data obtained from the KRILLBASE-ABUNDANCE database, this study aimed to investigate the effects of spatial scale on hotspot analysis of Antarctic krill (Euphausia superba) density distribution in the Antarctic Peninsula. Krill density data from 1929 to 2018 were interpolated at 10-year intervals into 10 spatial scales as follows: 10' × 10', 20' × 20', 30' × 30', 40' × 40', 50' × 50', 1° × 1°, 2° × 2°, 3° × 3°, 4° × 4° and 5° × 5°. Linear, logarithmic, exponential, power-law and parabolic functions were used to determine the relationship between spatial scale and krill density distribution in the Antarctic Peninsula region. Additionally, variations in centroid and area of hotspots at various spatial scales were analysed. The results revealed a strong scaling relationship between spatial scale, number of patches and indices of krill density. The hotspot area increased with an increase in the coarseness of the spatial scale and the calculated location of the centroid showed that the hotspot locations were markedly affected by the spatial scale of analysis, with coarser spatial scales resulting in larger spatial shifts in centroid location. Thus, based on the KRILLBASE-ABUNDANCE database, it is not recommended to use a spatial scale greater than 1° × 1° to identify the local spatial pattern for hotspot analysis of krill density. The effects of spatial scale on other types of krill density data (e.g. catch data, acoustic data from research surveys or acoustic data from fishing vessels) remain to be investigated. |
The Commission for the Conservation of Antarctic Marine Living Resources has recently decided to re-adopt the generalised yield model (GYM) as an assessment tool to progress the management of Antarctic krill (Euphausia superba). To this end, the GYM has been re-implemented in the open source R package Grym. This document describes the model parameters required to conduct a krill stock assessment with Grym and, where possible, provide guidance as to how these parameters can be estimated. |
| Maschette, D., S. Wotherspoon, P. Ziegler, S. Thanassekos, K. Reid, S. Kawaguchi, D. Welsford (2023), Grym: A new open-source implementation of the generalised yield model for flexible stock assessments |
Grym is an open-source implementation of the generalised yield model in R that provides greater transparency and extensibility (Maschette et al., 2023; Wotherspoon and Maschette, 2020). This paper describes an extension of the Grym to permit modelling of multiple fleets within a season, allowing the Grym implementation to model more complex fishery behavior and evolving fisheries practices. Results from the extension are compared with existing analyses from the Grym. The model can include vessels using different or identical selectivities with an example presented for Patagonian toothfish. |
Antarctic krill (Euphausia superba) is a keystone species in the Southern Ocean food web, and, as such, it is crucial to effectively manage the krill fishery to ensure its long-term sustainability. Setting precautionary catch limits for krill relies on sampling and population modelling. Krill stock projections are developed with the generalised yield model (GYM), which provides an assessment for stock status under current harvesting scenarios and various levels of uncertainties. One of the fundamental components of the GYM is the simulation of recruitment. De la Mare (1994) presents a proportional recruitment model for estimating krill recruitment based on length-frequency distributions collected from field surveys. The de la Mare (1994) function uses estimates of the mean and variance of proportional recruitment from survey data to determine the scaling of natural mortality and the distribution of random recruits that reproduce the observed mean and variance estimates. Here we evaluated de la Mare’s (1994) proportional recruitment function and found that for large variations in recruitment the function does not accurately reproduce the observed mean proportion of recruits and its variance. The deficiencies within the de la Mare (1994) function were reviewed and two alternative methods were provided, which can support a wider range of values and possible scenarios, such as years of extremely low recruitment. |