Comparative fixed gear selectivity studies in Portugal and Greece - 1997-1999. DG XIV, Ref. 96/065

Summary for non-specialists

Small scale, artisanal fisheries are of great importance in Portugal and in Hellas, involving the use of large numbers of small fishing vessels and employing significant numbers of fishers. These fisheries are characterised by the use of many different gears, with gill nets, trammel nets and longlines being the most important in both countries. The catch composition in different fishing operations will depend to a large extent on the species and size selection characteristics of the gear used. Other relevant factors include target species, the fishing grounds, the time of the year, and the fishing strategy. Because the same gear used under different circumstances may result in very different catches, the term "métier" is often used to distinguish particular fishing operations on the basis of the above combination of factors.

Despite the major socio-economic importance of small-scale fisheries in both countries, many aspects have not been comprehensively studied. In particular, there is a lack of information on the relative impacts of different gears that are used to catch the same species, in effect 'competing' for the resource. These impacts are reflected in differences between gears and mesh / hook sizes in terms of species composition, catch rates, discarding rates, size distributions of the catches, mortality rates and economic value of the catch. Information on the species and size selection of different static or fixed gears, gear competition and gear overlap is of vital importance for decision making concerning optimal harvesting strategies and the improved management and rational use of valuable demersal resources.

This study focused on two of the most important gears used in the Portuguese and Hellenic artisanal fisheries, longlines using small hooks and monofilament gill nets. Parallel experimental fishing trials at different fishing grounds (depths) and seasons were carried out with both gears in the Algarve (southern Portugal, Atlantic) and Cyclades (Hellas, Aegean Sea). The main objectives of the study were as follows:

A secondary objective was the analysis of the implications of this study for the management of multi-species, multi-gear demersal artisanal fisheries.

Fishing trials took place between July 1997 and June 1998 in the Algarve and between September 1997 and October 1998 in the Cyclades. The experimental fishing gears were constructed by local fishermen and were similar to those used locally. In both countries the longline mainline consisted of 1.1 mm diameter monofilament, the branch lines or gangions were of 0.5 mm diameter monofilament, 80 cm in length and were spaced approximately 1.7 m apart. In order to study size selectivity, four hook sizes of "Mustad" brand, round bent, flatted sea hooks (Quality 2316 DT, numbers 15, 13, 12 and 11) were used in both countries, covering the range of hook sizes normally used in small hook longlines. Four mesh sizes of monofilament gill nets were used in each country, with the choice of mesh size based on the gill nets used locally. Thus, gill nets of 22, 24, 26 and 28 mm nominal bar length were used in the Cyclades, whereas larger mesh sizes of 25, 30, 35 and 40 mm were fished in the Algarve. In both cases the monofilament was of 0.3 mm diameter and the hanging ratio in the floatline was 0.5 for all mesh sizes. Normal fishing practices were followed. In the case of the Algarve two fishing boats were used simultaneously whereas one chartered fishing boat was used to fish both gears at the same fishing grounds in the Cyclades. All fishing operations were accompanied by scientists on board who ensured that the catches were correctly separated according to mesh and hook size and who recorded the relevant data. All individuals were subsequently weighed, measured and classified as commercial, discard or for self-consumption.

The catch and size distribution data were analysed using a variety of techniques ranging from simple graphs of the catch size frequency distributions for each species and gear size to multivariate analysis techniques such as cluster analysis and non-metric multidimensional scaling for studying gear competition and gear overlap. To study the size selectivity of each gear a preliminary graphical analysis was used to select the most appropriate type of selectivity model for each species and gear. Indirect methods based on maximum likelihood were used to estimate the selectivity parameters of uni-modal models (for gill nets) and the logistic model (longline) from the catch size frequency distributions of different gear sizes. The relative selection of the two gears was evaluated for the most important species caught by both gears.

The main results are as follows:

1. Both gears caught a large number of species, 84 in the Algarve and 79 in the Cyclades, with gill nets accounting for more species than longlines. Significant numbers of the species caught by longline were also caught by gill nets, 75 and 83% respectively for the Algarve and the Cyclades.

2. Five species of sea breams (Diplodus vulgaris, 24%, Pagellus acarne, 21%, Spondyliosoma cantharus, 13%, Diplodus sargus, 11% and Pagellus erythrinus, 10%) accounted for 79% of the Algarve longline catches by weight. In contrast, gill net catches were dominated by Scomber japonicus (29% in number and 17% by weight), with 66 species accounting for 36% and 50% of the Algarve gill net catches in number and in weight respectively.

3. Cyclades longline catches were dominated in number by Serranus cabrilla (34%), Pagellus erythrinus (11%), Conger conger (10%), Diplodus annularis (7%) and Diplodus vulgaris (7%). In terms of weight, the most important species in the longline catches were Conger conger (15%), Pagellus erythrinus (13%), Serranus cabrilla (12%) and Diplodus vulgaris (9%). Gill net catches were dominated in number and in weight by Spicara maena (22% and 16%), followed by Pagellus erythrinus (16% and 11%), with 74 species accounting for 50% of the total catch in weight.

4. The overall catches with both gears were considerably greater in the Algarve (n = 19059, 2713 kg) than in the Cyclades (n = 4751, 533 kg). This may be due in part to differences in productivity between the Aegean and the eastern Atlantic, which may be reflected in decreased abundance and generally smaller sizes of the same species. 5) Catch rates generally decreased with increasing hook size in the Algarve whereas there was no clear pattern in the Cyclades. In both areas the smallest mesh sizes had the highest catch rates in numbers. Catch rates varied according to season, reflecting seasonal changes in species composition and abundances.

5. The results of multivariate analysis showed a clear separation of the two gears in terms of catch composition by number and by weight. Within the two main groups, subgroups generally based on season were also formed, especially in the case of Algarve longline catches. In both areas subgroups were also occasionally formed by gear size but the overall pattern was not clear.

6. Differences in dominance between gears, seasons and gear sizes were observed based on K-dominance curve analysis. Thus for example, dominance in number in the Algarve was more pronounced for gill nets, especially the largest mesh size and for the summer. This was due to the large catches of mackerel with the 40 mm mesh size in the summer fishing trials.

7. For species caught by both gears in sufficient numbers, hooks generally accounted for larger fish and a wider size range than nets. Thus, for the Algarve significant differences in the catch size frequency distributions of the two gears were found for all the species tested (Boops boops, Diplodus bellottii, Diplodus vulgaris, Pagellus acarne, Pagellus erythrinus, Spondyliosoma cantharus, Scomber japonicus and Scorpaena notata). For the Cyclades four out of five species tested had significantly different length frequency distributions (Diplodus annularis, Diplodus vulgaris, Pagellus erythrinus and Serranus cabrilla).

8. For species caught in both areas (e.g. Diplodus vulgaris, Pagellus erythrinus), those caught in the Cyclades were significantly smaller in size than the ones caught in the Algarve. While catches with gill nets are not comparable due to the different mesh sizes used, the same sizes of hooks were used in both areas. This suggests that there are real differences in the demographic structures between the Aegean and the eastern Atlantic fish populations that may reflect regional differences in productivity and/or fishing mortality.

9. In general, the four mesh sizes used in both areas caught relatively narrow length ranges and the modal lengths of the species gradually increased with increasing mesh size. In contrast, length frequency distributions for the four hook sizes for each species tended to be highly overlapped, with similar modal lengths or with only small increases in modal length with increasing hook size.

10. Preliminary plots of the catch length frequency distributions per species for evaluating the type of selectivity model clearly suggested uni-modal selectivity models for the gill net data and monotonic or logistic type selectivity curves for the longline data.

11. Gill net selectivity parameters were estimated using the SELECT model for 9 species in the Algarve (Boops boops, Diplodus bellottii, Diplodus vulgaris, Mullus surmuletus, Pagellus acarne, Pagellus erythrinus, Spondyliosoma cantharus, Scomber japonicus and Scorpaena notata) and 9 species in the Cyclades (Boops boops, Diplodus annularis, Mullus surmuletus, Pagellus erythrinus, Scorpaena porcus, Serranus cabrilla, Symphodus tinca and Trachurus mediterraneus). Although the log-normal model provided the best fits for the majority of the Cyclades species, normal-scale, gamma and log-normal were chosen for the different Algarve species.

12. Parameters of the logistic selectivity curve were estimated for 9 species caught in Algarve longlines (Boops boops, Diplodus bellottii, Diplodus sargus, Diplodus vulgaris, Pagellus acarne, Pagellus erythrinus, Spondyliosoma cantharus, Scomber japonicus and Scorpaena notata) and 6 species from the Cyclades longline trials (Conger conger, Diplodus annularis, Diplodus vulgaris, Pagellus erythrinus, Scorpaena porcus and Serranus cabrilla).

The results of this study show that these two static gears have a very different impact on the demersal resources in both areas, in terms of species composition of the catch, catch rates, and the sizes of the fish caught. Differences in fishing mortality rate are reflected in the observed size distributions of the catches of the two gears. This should be taken into account in the management of such multi-species, multi-gear fisheries. Together with information on fishing effort with different gears, the results of this study can be used to improve resource allocation and use by better licensing and control of effort.

The multivariate analysis carried out in this study proved a useful tool for the identification of métiers. Division of fishing effort and fishing mortality by métier is an important step in improved management of artisanal fisheries.

---

Anon. (1999) Comparative fixed gear selectivity studies in Portugal and Greece. Ref. 96/065. Final Report. European Commission, DG XIV/C/1, Universidade do Algarve, UCTRA, Faro, 134pp + annexes