No projection is done for none of these stocks.
3.5.1. Short, medium and/or long-term projections
Not relevant.
3.5.2. Are projections deterministic or stochastic?
Not relevant.
3.5.3. How is recruitment simulated in the projection/ (historical geometric mean, using S/R model etc)
Not relevant.
3.5.4. How is stock growth simulated (e.g. exponential survival equation)?
Not relevant.
3.5.5. How are biological parameters projected (stochastically, mean of last 3 years etc)
Not relevant.
3.5.6. What reference points are used in the projections?
Not relevant.
3.5.7. Harvest control rules (HCRs) and management strategy evaluation (MSE): does the stock have a pre-defined HCR? If so, please specify.
None of these stocks is managed by HCR.
3.5.8. Has this rule been agreed with all stakeholders?
Not relevant.
3.5.9. Has the rule been simulation tested using MSE? If so please describe methods and outcomes
Not relevant.
3.5.10. Is the rule robust to uncertainties within the fishery system?
Not relevant.
3.5.11. Do you have an estimate of virgin biomass, if so what is it, how was it derived and how reliable is it? Roundnose Grenadier Vb, VI, VII, XIIb
Estimates of biomass at the beginning of the fishery are compromised by the lack of information on discards and efforts. Exploratory assessments have shown the initial estimates are highly sensitive on the assumptions made about discards distribution and fishing efforts (WGDEEP, 2009). Russian data from the 1970s-80s estimated stock biomass to be between 400 000 and 700 000 tons. Some estimates of virgin biomass were based upon survey data and swept area method. Gordon and Hunter (1994) used this method to estimate this biomass in an area bound by latitudes 53°N - 62°N and longitudes 5°W - 15°W. This area includes the continental slope of Scotaland and Ireland from the Porcupine Bankj to the Faeroes and the offshore banks (Rockall, George Blight, Hatton, Lousy, Bill Bailey…). The total area of the seabed between 5000 m and 2000 m countours was estimated 247,000 km². Based upon estimated catch efficiency of 40 to 50 % (i.e. 40 to 50 % of fish present on the trawl path would be caught) the following biomass estimated were derived from Scottish Association of Marine Science (SAMS) survey data: 1,200 x 103 t (900 x 103 - 1,600x 103) for a catch efficiency of 40%; 990 x 103 t (730 x 103 - 1,240x 103) for a catch efficiency of 50%. As stated by the authors, these values should be considered with caution. Based on these estimated biomass, maximum sustainable yields of 13,000 to 17,000 tonnes were estimated. On average, since the early 1990s, annual landings were about 13,000 tonnes and are estimated to correspond to an overexploitation of the stock (WGDEEP, 2010)
Other species
For black scabbardfish, greater forkbeard, Portuguese dogfish and leafscale gulper shark, tThere are no estimate of the virgin biomass for this species. However, exploratory assessments using DeLury and Schaefer production model provided some estimated of virgin biomass and carrying capacity (Basson et al. 2002), but are likely to be strongly dependent upon model assumptions.
3.6. Assessment packages/programs used (e.g. FLR, CEDA, ASPIC, Lowestoft XSA etc)
This section is relevant only to Roundnose Grenadier only in Vb , VI, VII, XIIb as the other stocks are not quantitatively assessed.
3.6.1. Were any technical problems encountered, were these resolved and if so how?
Roundnose grenadier is a long-lived species. The current separable VPA packages VPA95 and FLR are limited to 25 age groups. This problem has not been solved mostly because the code is complex (FLR) or compiled (VPA95).
Past assessment were carried out using CEDA package (Basson et al., 2002).
Current assessment is carried out using a surplus production model form the FLR FLBayes library. The distribution of this package online being problematic (due to erratic website), the library and example code is available from http://code.google.com/p/wgdeep-rng/ .
3.6.2. Were the packages/programs used suitable for use by scientists with little or no experience of them?
Separable VPA is a quite straightforward approach with few parameters to use although selectivity-at-age is generally poorly defined. The VPA95 suite is easy to use and provides a single output files containing biomass at age, population numbers, fishing mortalities and residuals.
SepVPA using the FLR package provides FLStock objects, one of the standard output format of the FLR package. Using this routine requires some knowledge in R programming. Documentation on SepVPA is poor. Using the same initial parameters as for VPA95 does not provide exactly the same results. SepVPA does not provide residuals of adjustment as well. This routine as part of the R environment can easily be implemented into scripts which is convenient for running several assessments in batch (such as in the case of quantifying uncertainties).
Experts in R will have few difficulties to use SepVPA mostly due to the poor documentation. Scientists with little or no experience especially in programming may prefer using VPA95.
The Surplus production model is easy to implement and its code natively provides MSY indicators.
3.6.3.If not, how could they be improved?
The sepVPA manual should be more documented and the sepVPA code should be reviewed against VPA95 code to understand why results are different.
3.6.4. Were the assessment diagnostics fit for purpose? If not how could they be improved?
No assessment diagnostic is performed for roundnose grenadier in Vb, VI, VII, XIIb.
3.6.5. Did you receive any training in the use of the assessment packages/programs?
No.
Share with your friends: |