[an error occurred while processing this directive]

 

Hake semi-pelagic longline selectivity and evaluation of selectivity models for hook and line gear - 1997-1999. DG XIV, Ref. 96/062

Summary for non-specialists

The "semi-pelagic" longline used to fish the hake (Merluccius merluccius) consists of a monofilament mainline to which are attached monofilament gangions with hooks at regular intervals. The longline is fished on the bottom with weights (rocks) with the hooks lifted off the bottom with floats. This allows the hooks to fish the water column close to the bottom.

Although the hake is one of the most important resources in both Portugal and Spain, there is little information on hook and line fisheries. The main objectives of the first part of this study were the following:

  1. to study and compare the species composition and size distributions of catches of the semi-pelagic longline used for fishing hake (Merluccius merluccius) in Portugal (Algarve) and Spain (Cantabrian Sea, Bay of Biscay),
  2. to evaluate the size selective properties of the different types and sizes of hooks used in these fisheries,
  3. to evaluate the differences in efficiency (catch-per-unit-effort) between different sized hooks in terms of number and weight of the main species caught per longline unit, and
  4. to study the way the fish are hooked (i.e. position of the hook in the mouth, throat or stomach), and to determine if there are differences due to hook type and size.

The objective of the second part of the study was to use simulated data to evaluate graphical methods for determining the underlying selectivity curve and to compare a variety of models and methods for estimating selectivity parameters. The main results are as follows:

Experimental longline fishing in the Algarve (southern Portugal)

  1. Four hook sizes of SIAPAL brand hooks were used in the Algarve: numbers 10, 9, 7 and 5 (hook number 10 is the smallest, hook number 5 is the largest) where the overall maximum size (width x height) of the largest hook is 2.98 times that of the smallest.
  2. At least 32 species were caught (1997: n = 3757, 1998: n = 2791)
  3. Six species accounted for 91% of the catch in numbers, with the hake dominating the catch in numbers: Merluccius merluccius (42%), Galeus melastomus (21%), Micromesistius poutassou (8.5%), Etmopterus pusillus (7.2%), Benthodesmus elongatus (6.9%) and Scyliorhinus canicula (4.7%).
  4. Discarding rates were high (between 40 and 50% of the catch in numbers). The main reason for discarding was low or no commercial value. Discarding rates were highest for the smallest hooks.
  5. Overall catch rates in numbers per 100 hooks and in weight (kg per 100 hooks) decreased significantly with increasing hook size. Catch rates for the target species, Merluccius merluccius, also decreased with increasing hook size. The smallest hook (number 10), which is not used by commercial fishermen had significantly higher catch rates for hake but also had higher catch rates for undesirable species.
  6. Catch size frequency distributions for the main species were generally highly overlapped, with the four hook sizes catching a wide size range. No illegal (<27 cm) sized hake were caught. Hake ranged in size from 27 to 94 cm total length (TL) in 1997 and from 34 to 66 cm TL in 1998. Although no differences were found between distributions within each year, significant differences were found between years, with larger hake caught in 1997 (mean TL of 47.4 cm) than in 1998 (mean TL of 45.2 cm).
  7. Overall, 52% of the hake caught had swallowed the hook and 42% were mouth hooked. Hook size was a significant factor in the method of capture, with mouth hooking increasing with decreasing hook size and the percentage of hake that had swallowed the hook increasing with increasing hook size.

Experimental longline fishing in the Cantabrian Sea (Spain)

  1. Four sizes of Round hook were used: 1, 1/0, 3/0 and 5/0, where the number 1 hook is the smallest and the number 5/0 the largest. Size 3/0 is the most commonly used in the fishery.
  2. Longline fishing experiments took place on the continental slope in 1997 and at shallower depths on the continental shelf in 1998.
  3. Overall, four species accounted for 80% of the catch in numbers: blue whiting, Micromesistius poutassou (35%), horse mackerel, Trachurus trachurus (19%), hake, Merluccius merluccius (15%) and mackerel, Scomber scombrus (11%). In both years, a by-catch species dominated the catch: blue whiting in 1997 (54% of the total catch in numbers) and horse mackerel (38%) and mackerel (13%) in 1998.
  4. Catch rates for hake decreased with increasing hook size but not as significantly as in the Algarve. Important variability in catch rates per tub within days and over the course of the fishing experiments was found. This can probably be attributed to the patchy distribution of hake on the bottom and to within season variability in abundance.
  5. Catch rates for horse mackerel decreased with increasing hook size. However, the opposite was true for the blue whiting, where more fish were hooked in the body with the larger hook sizes.
  6. As in the case of the Algarve, size distributions were highly overlapped, with no significant differences between the distributions for the four hook sizes within each year. However, the hake in 1998 (mean total length of 50.7 cm) were significantly smaller than those in 1997 (mean of 52.6 cm TL).
  7. In contrast to the Algarve, most hake were hooked in the mouth (between 70 and 80% depending on hook size), with some significant differences due to hook size.

Simulations study

  1. Catch size frequency distributions were simulated for two species type: a large, fast growing pelagic and a relatively small, slow growing, long lived demersal species such as a sea bream. For both species types, simulations were carried out with both normal and logistic type selectivity. A total of 30 cases were simulated, covering a wide range of situations in terms of life history characteristics, overlapping, selectivity and fishing scenarios.
  2. Following the methodology of Millar (1995), plots of the proportions caught on the larger hook against fish size class were used to identify the underlying selectivity model.
  3. A variety of models/methods were used to estimate the selectivity parameters for each case. These included the Ishida (1962), Wulff (1986) and SELECT (Millar, 1992a) models/methods.
  4. In general, the preliminary Millar plots proved satisfactory and except for cases where the distributions where highly overlapped, it was possible to decide whether the underlying selectivity model was normal or logistic.
  5. None of the models/methods used to estimate the selectivity parameters was able to correctly estimate the selectivity parameters for highly overlapped distributions.

Hake (Merluccius merluccius) hook selectivity

  1. Millar plots, particularly those based on three-point running averages strongly indicated that the most appropriate hook selectivity model was the logistic.
  2. Eight different models were fitted to each of the four data sets (Algarve 1997, 1998; Bay of Biscay 1997, 1998) using the Wulff (1986) methods, with the parameters of the logistic model constant, proportional or linear functions of hook size.
  3. No satisfactory estimates were obtained due to the highly overlapped distributions and to the lack of fish at the extremes of the distributions: fish smaller that 30 cm TL and greater than 65 cm TL.
NOTE: More information is included in the following report which is available from the European Comission's Directorate General XIV or from the authors

Anon. (2000) Hake semi-pelagic longline selectivity and evaluation of selectivity models for hook and line gear. Ref. 96/062. Final Report. European Commission, DG XIV/C/1, Universidade do Algarve, UCTRA, Faro, 190pp.

The report is available in PDF under request
[an error occurred while processing this directive]