GREATER SILVER SMELT

Argentina silus


Technical report
Published by

Marine and Freshwater Research Institute, Iceland

Published

7 June 2024

General information

Greater silver smelt is a rather small silvery bathypelagic species that can form large schools close to the seafloor mainly at depths >500 m. In Icelandic waters it can live to around 26 years old. Juveniles tend to aggregate in shallower depths. Greater silver smelt mainly feed on zooplankton (e.g., euphausiids, amphipods, and copepods) or small nekton (e.g., squids, jellyfish, or fish).

The Fishery

Fleets

Since 1996 between 20 and 40 trawlers have annually reported catches of greater silver smelt in Icelandic waters (Table 1). The trawlers participating in the greater silver smelt fishery also target redfish (Sebastes norvegicus and S. mentella) and to lesser extent Greenland halibut and blue ling. Number of hauls peaked in 2010, but the number of hauls has decreased since then in line with lower total catches. In most years over 50% of the greater silver smelt catches are taken in hauls where the species is more than 50% of the catch (Table 1).

Targeting and mixed fisheries issues in the fishery in Icelandic waters

Mixed fisheries issues: species composition in the fishery

Redfish spp. (golden redfish (Sebastes norvegicus) and deepwater redfish (S. mentella)) are the main bycatch species in the mixed fishery encompassing greater silver smelt. Other species of lesser importance are Greenland halibut, blue ling and ling. Other species than these rarely exceed 10% of the bycatch in the greater silver smelt fishery in Icelandic waters (Table 2).

Table 2: Greater silver smelt in Icelandic waters. Proportional bycatch species composition where greater silver smelt was more than 50% of the total catch in a haul.
Year Golden redfish Deepwater redfish Greenland halibut Ling Blue ling Other
1997 1.41 79.28 0.00 6.80 7 5.39
1998 5.23 77.49 0.00 3.51 7 6.65
1999 4.09 79.80 0.00 2.72 6 7.55
2000 4.92 70.88 0.16 0.34 10 13.74
2001 22.69 55.05 4.50 0.52 1 16.10
2002 17.32 73.92 0.44 1.19 4 3.13
2003 38.44 51.24 0.44 0.05 5 4.83
2004 24.87 68.68 0.68 0.12 1 4.80
2005 15.40 69.88 4.22 1.42 3 6.08
2006 28.80 59.79 1.44 0.88 1 8.14
2007 11.90 71.20 5.93 0.32 6 4.63
2008 26.66 60.84 2.76 1.21 5 3.30
2009 20.14 64.62 3.20 0.19 8 3.99
2010 15.96 63.74 2.03 0.87 6 11.05
2011 13.20 66.41 2.18 0.36 5 13.01
2012 8.79 67.30 1.33 0.24 8 14.82
2013 9.54 63.91 4.61 0.15 9 12.63
2014 2.46 78.28 2.83 0.26 5 10.68
2015 12.58 64.07 4.67 0.23 4 14.53
2016 10.88 73.54 5.45 0.22 3 7.14
2017 2.93 85.63 1.57 0.24 3 6.77
2018 4.68 87.66 2.05 0.05 2 3.99
2019 7.77 81.15 1.84 0.55 2 7.03
2020 5.58 87.46 1.69 0.12 1 4.20
2021 11.55 72.26 5.80 0.28 1 8.66
2022 5.68 83.95 3.95 0.22 3 2.90
2023 10.64 58.88 21.43 0.28 2 6.83

Spatial distribution of catches through time

Spatial distribution of catches in 1995–2022 is presented in Figure 3 and Figure 4. Most of the catches have been from the southern edge of the Icelandic shelf. In the period, a gradual relative increase is seen in the western area and a gradual decrease in the southeastern area (Figure 3).

Figure 3: Greater silver smelt. Catch distribution and proportions by area according to logbooks.

Figure 4: Greater silver smelt. Spatial distribution of catches as reported in logbooks.

Table 3: Greater silver smelt. Landings (tonnes) records from the Icelandic directorate of Fisheries and Greenland.
Year Section 5.a Section 14.b Total
1988 241 0 241
1989 8 0 8
1990 113 0 113
1991 246 0 246
1992 657 0 657
1993 1526 0 1526
1994 756 0 756
1995 586 0 586
1996 881 0 881
1997 3935 0 3935
1998 15242 0 15242
1999 6681 0 6681
2000 5657 0 5657
2001 3043 0 3043
2002 4960 0 4960
2003 2680 0 2680
2004 3645 0 3645
2005 4482 0 4482
2006 4769 0 4769
2007 4227 0 4227
2008 8778 0 8778
2009 10828 0 10828
2010 16428 0 16428
2011 10516 0 10516
2012 9289 0 9289
2013 7155 0 7155
2014 6344 4 6348
2015 6058 23 6081
2016 5646 16 5662
2017 4344 666 5010
2018 4035 425 4460
2019 3209 2 3211
2020 3775 27 3802
2021 4140 15 4155
2022 6886 28 6914
2023 5268 0 5268

Data available

In general, sampling from commercial catches is considered representative of the stock, as one of the requirements of owning a fishing license for greater silver smelt is the retention of scientific samples (Table 4). Samples were only obtained from bottom trawls. The sampling does seem to cover the spatial and temporal distribution of catches. The sampling coverage in 2023 is shown in Figure 5, but in recent years there has been a large decline in sampling with an increase in 2023.

Figure 5: Greater silver smelt. Fishing grounds in 2023 as reported in logbooks and positions of samples taken from landings (asterisks).

Landings and discards

Landings by Icelandic vessels are given by the Icelandic Directorate of Fisheries. Discarding is banned in Icelandic waters, and currently there is no available information on greater silver smelt discards. It is however likely that unknown quantities of greater silver smelt were discarded prior to 1996.

Length compositions

Table 4 gives the number of samples and measurements available for calculations of catch in numbers of greater silver smelt in Icelandic waters. Length distributions from autumn survey and commercial samples are presented in Figure 6 and Figure 7, respectively. Length distributions from the autumn survey are rather stable, with 2023 being close to the long-term average (Figure 6).

Figure 6: Greater silver smelt. Length distribution from the autumn survey (grey area) and mean length distribution (black line).

Figure 7: Greater silver smelt. Relative length distributions from Icelandic commercial bottom trawl catches.

Age compositions

Table 4 gives the number of samples and measurements available for calculations of catch in numbers of greater silver smelt in Icelandic waters. Age distributions from the autumn survey in Figure 8 and estimated as catch in numbers are given in Figure 9.

Table 4: Greater silver smelt. Summary of sampling intensity and overview of available data.
Year No. length samples No. length measurements No. otolith samples No. aged otoliths
1997 48 4992 1447 1059
1998 148 15559 6964 889
1999 58 4163 2180 82
2000 27 2968 1011 113
2001 10 489 245 17
2002 21 2270 360 127
2003 63 5095 425 0
2004 34 997 225 84
2005 49 3708 772 0
2006 29 4186 616 525
2007 14 2158 285 272
2008 44 3726 1768 1387
2009 53 5702 1746 1574
2010 134 16353 3370 3120
2011 63 6866 1953 1774
2012 43 4440 1492 603
2013 47 4977 710 704
2014 39 4709 350 340
2015 11 1275 221 217
2016 45 5879 285 283
2017 29 3466 430 416
2018 12 1437 185 181
2019 10 1250 40 40
2020 12 1905 130 130
2021 14 1301 215 214
2022 8 603 165 165
2023 26 2598 439 436

Figure 8: Greater silver smelt. Age distributions in proportions from the Icelandic autumn survey.

Figure 9: Greater silver smelt. Greater silver smelt. Catch in numbers by age (proportion).

Weight at age

Biological data from the spring survey, autumn survey, and commercial catches were combined to analyse growth. Von Bertalanffy growth curves were fitted and plotted within a series of time periods, including 2016-2019, 2011-2015, 2006-2010, 2001-2005, 1994-2000, and prior to 1994 to increase sample sizes for estimating each curve. The exponential length-weight relationship is extremely consistent across periods. In general, there is very little variation between periods, although females can be seen to grow larger sizes than males.

Maturity at age and natural mortality

Estimates of maturity ogives of greater silver smelt in 5.a were presented at the ICES 2020 meeting for both age and length (ICES 2020) using data collected in the Icelandic autumn survey (See stock annex for details). Males tend on average to mature at a slightly higher age or at 6.5 compared to 5.6 for females but at a similar length as females 35.3 cm. Most of the greater silver smelt caught in commercial catches in Icelandic waters are mature.

No information exists on natural mortality of greater silver smelt in Icelandic waters.

Catch, effort and Survey Data

Catch per unit of effort and effort data from commercial fisheries

At WKDEEP 2010 a glm cpue series was presented (WKDEEP 2010, GSS-05), however because of strong residual patterns the group concluded that the glm-cpue series was not suitable to use as an indicator of stock trends. The cpue is not considered to represent changes in stock abundance as the fishery is mostly controlled by market factors, oil prices and quota status in other species, mainly redfish.

Icelandic survey data

The Icelandic spring groundfish survey, which has been conducted annually in March since 1985, gives trends on fishable biomass of many exploited stocks on the Icelandic fishing grounds. In total, about 550 stations are taken annually at depths down to 500 m. The survey area does not cover the most important distribution area of the greater silver smelt fishery in Icelandic waters and is therefore not considered representative of stock biomass. The survey may be indicative of recruitment; however, the data have not been explored in sufficient detail to be used for this purpose.

The Icelandic groundfish autumn survey was commenced in 1996 and expanded in 2000. A detailed description of the autumn groundfish survey is given in the stock annex (ICES 2020) for greater silver smelt in Icelandic waters. The survey is considered representative of stock biomass of greater silver smelt since it was expanded in 2000. Figure 11 gives the most recent catch quantities and locations of surveys. Due to a strike in 2011 the autumn survey was cancelled after about one week of survey time. Greater silver smelt is among the most difficult demersal fish stocks to get reliable information on from bottom-trawl surveys. This is in large part because most of the greater silver smelt caught in the survey is taken in few but relatively large hauls. This can result in very high indices with large variances particularly if the tow-station in question happens to be in a large stratum with relatively few tow-stations. For example, survey indices in 1999, 2014, and 2021 are especially high in comparison with survey indices from adjacent years (Figure 10). No substantial changes in spatial distribution are seen in general in Figure 12 & Figure 13.

Figure 10: Greater silver smelt. Indices calculated from the Icelandic spring survey (lines and shaded area) and from the autumn survey (dots). Vertical lines and shaded area represent +/- 1 standard error.

Figure 11: Greater silver smelt. Abundance and distribution of greater silver smelt in the spring survey (SMB) in 2024 and in the autumn survey (SMH) in 2023.

Figure 12: Greater silver smelt. Estimated survey biomass in the autumn survey by year from different areas (upper figure) and as proportions of the total (lower figure).

Data analyses

Landings and sampling

Spatial distribution of catches in Icelandic waters has not changed markedly in recent years and fishing for greater silver smelt in the NW area seems to have reduced (Figure 2 and Figure 3). Landings of greater silver smelt increased rapidly from 2007 to 2010 when they peaked at around 16 000 tonnes, since then they have decreased to around 5430 tonnes in 2023 (Figure 4 and Table 3). The decrease in catches is the result of increased vigilance by the managers to constrain catches to those advised and also lesser interest by the fleet in the stock. Mean length of greater silver smelt in catches has been rather stable from 2005 in the range of 37 – 43 cm (Figure 7). However, there was a slight increase in mean length in 2012 which can also be seen in recent years (Figure 7 and Figure 8). A similar continuous downward trend in mean age in the commercial catches is also observed. Mean ages from catches have been variable from 2000 in the range 8 – 14 years, with relatively high mean ages recently. The reason for these changes is not known as there is no marked difference in the spatial distribution of the fishery; however, reduced fishing pressure may be a factor.

Surveys

As mentioned above, greater silver smelt is a difficult species to survey in trawl surveys and the indices derived from the both the spring and autumn surveys have high CVs. Occasional spikes in the indices without any clear trend characterize the spring survey biomass indices. The only thing that can be derived from the spring survey is that the biomass indices (total and >25 cm), in 1985–1993 and again from 2002 to 2023 are at a higher level than in 1994–2001. The juvenile index (spring survey) has a very high peak in 1986 but then hardly any juveniles are detected in the survey in 1987 to 1995. Since 1998 there have been several small spikes in the recruitment index (Figure 10).

The observed trends in the biomass indices from the autumn survey have a considerably different trend than those observed in the spring survey (Figure 10). According to the autumn survey, biomass increased more or less year on year from 2000 to 2008 but then decreased in 2009 and 2010. The total biomass index in the autumn survey showed slight variations until 2014 when the index increased to the highest value observed, and thereafter has been relatively stable but with high variability. In 2023, the index reached the highest value in the timeseries.

There is a clear gradient in mean length of greater silver smelt with depth, larger fish being in deeper water, and therefore the spring survey, which is conducted at shallower depths, is not considered representative of the stock.

Analytical assessment

In 2020 a model of greater silver smelt in Icelandic and Greenlandic waters (ICES areas 5.a and 14) developed in the Gadget framework (Globally applicable Area Disaggregated General Ecosystem Toolbox, see http://www.hafro.is/gadget for further details) was benchmarked for the use in assessment (ICES 2020). This year, Gadget 3 was used instead of Gadget 2. Gadget 3 is the same model except that it uses template model builder (TMB) which allows it to utilize TMB’s automatic differentiation procedures producing models that can be optimized faster and that can use R optimizers (rather than Gadget 2’s inbuilt optimizers).

Data used and model settings

Data used for tuning and model settings used in the Gadget model are described in more detail in the stock annex (ICES 2020).

Diagnostics

Observed and predicted proportions by fleet

Overall fit to the predicted proportional length and age–length distributions is close to the observed distributions, except for a small peak of small-sized fish (Figures 13-16). This peak does not shift from year to year and therefore is considered due to high catchability in aggregations of small fish rather than cohorts in recruitment peaks. These peaks are likely absent from commercial data due to the requirement of fishing at >400 m depth.

Figure 13: Greater silver smelt. Fitted proportions-at-length from the Gadget model (black lines) compared to observed proportions in the autumn survey (vertical lines and points).