Geographical patterns in abundance and population structure of Nephrops norvegicus and Parapenaeus longirostris ( Crustacea : Decapoda ) along the European Mediterranean coasts *

The Norway lobster Nephrops norvegicus and the deep-water pink shrimp Parapenaeus longirostris are two of the main target species of the commercial fisheries in the Mediterranean continental shelf and upper slope and have been the subject of important biological and fishery studies (e.g. Ardizzone et al., 1990; Maynou and Sardà, 1997; Maynou et al., 1996; Sardà, 1995, 1998a,b; Sardà and Lleonart, 1993; Tom et al. 1988; Tursi et al., 1998). Their population dynamics have been thoroughly studied and overexploitation of the N. norvegicus stocks has been reported for some areas of the Mediterranean (Abella et al., 1999; Jukic, 1971, 1974; Levi and Giannetti, 1973; Sardà and Abelló, 1984; Sardà, 1998c; Sardà et al., 1998), while in some other fishing grounds the species is SCI. MAR., 66 (Suppl. 2): 125-141 SCIENTIA MARINA 2002


INTRODUCTION
The Norway lobster Nephrops norvegicus and the deep-water pink shrimp Parapenaeus longirostris are two of the main target species of the commercial fisheries in the Mediterranean continental shelf and upper slope and have been the subject of important biological and fishery studies (e.g.Ardizzone et al., 1990;Maynou and Sardà, 1997;Maynou et al., 1996;Sardà, 1995Sardà, , 1998a,b;,b;Sardà and Lleonart, 1993;Tom et al. 1988;Tursi et al., 1998).Their population dynamics have been thoroughly studied and overexploitation of the N. norvegicus stocks has been reported for some areas of the Mediterranean (Abella et al., 1999;Jukic, 1971Jukic, , 1974;;Levi and Giannetti, 1973;Sardà and Abelló, 1984;Sardà, 1998c;Sardà et al., 1998), while in some other fishing grounds the species is SCI.MAR., 66 (Suppl.2): 125-141 SURVEY (1994SURVEY ( -1999)).P. ABELLÓ, J.A. BERTRAND, L. GIL DE SOLA, C. PAPACONSTANTINOU, G. RELINI and A. SOUPLET (eds.)Geographical patterns in abundance and population structure of Nephrops norvegicus and Parapenaeus longirostris (Crustacea: Decapoda) along the European Mediterranean coasts* exposed to a lower fishing pressure (Abella and Righini, 1998).Since the life histories of the two species show different characteristics, especially concerning their lifespan and growth rates, the study of their different responses to exploitation will be of valuable interest.

MEDITERRANEAN MARINE DEMERSAL RESOURCES: THE MEDITS INTERNATIONAL TRAWL
Nephrops norvegicus is a sedentary lobster which inhabits burrows constructed in muddy substrates throughout the continental shelf and upper slope of the north-eastern Atlantic and Mediterranean (Farmer, 1975;Chapman, 1980;Froglia and Gramitto, 1981).The species spends large periods of time inside the burrows, from which it emerges periodically giving rise to diurnal and seasonal fluctuations in the catches (Jukic, 1971(Jukic, , 1974;;Atkinson and Naylor, 1976;Chapman and Howard, 1979;Naylor, 1988).In some areas characterised by a wide continental shelf, such as the Adriatic Sea or close to the Ebro delta in the western Mediterranean, populations of N. norvegicus are usually found much shallower than in most other areas and have been the object of several biological and ecological studies (Gauss-Garady, 1912;Pesta, 1918;Karlovac, 1953;Abelló et al., 1988Abelló et al., , 2000;;Maynou and Sardà, 1997).The spatial component seems to have an important role in the definition of the species population structure.This is known to take place between different regions throughout their distribution area (Karlovac, 1953;Chapman and Howard, 1988;Tully and Hillis, 1995;Tuck et al., 1997).Patchiness in population structure characteristics has been detected and appears to be related with heterogeneity in the characteristics of the sediment as well as with variations in fishing effort (Fariña et al., 1994;Tully and Hillis, 1995;Maynou et al., 1996).
Parapenaeus longirostris is also one of the most important commercial crustaceans in the Mediterranean.The species inhabits muddy or sandymuddy bottoms on the upper slope at depths of 150-400 m.Its distribution area encompasses the Mediterranean Sea and the eastern North-Atlantic Ocean (Maurin, 1968;Holthuis, 1987;Ribeiro-Cascalho and Arrobas, 1987;Ardizzone et al., 1990;Sardà, 1995;Sardà et al., 1982;Levi et al., 1995).It shows a marked size-dependent distribution by depth, with small individuals being found at the edge of the continental shelf.Investigations on P. longirostris based in trawl surveys have been performed in the Adriatic Sea (Karlovac, 1953) and in some other areas of the Mediterranean (Bombace, 1972;Froglia, 1982;Ardizzone et al., 1990;Tom et al. 1988;D'Onghia et al., 1998;Lembo et al., 1999Lembo et al., , 2000;;Mori et al., 2000).Some investigations have also been performed along the NW coasts of Africa and off Portugal (e.g.Bravo de Laguna, 1985;Dos Santos, 1998;Ribeiro-Cascalho and Arrobas, 1987;Sobrino and García, 1994).
The aim of the present paper is to analyse the distribution pattern of both N. norvegicus and P. longirostris along the European Mediterranean coasts, as well as to study the differences in the size structure of their populations related to depth range and geographical sector.All samples were taken during daylight hours, with identical fishing gear, equal fishing procedures, all cruises being performed in the same season (spring), with the coverage of the entire depth range where the species is distributed (Bertrand et al., 2000(Bertrand et al., , 2002)).These characteristics of the sampling scheme (homogeneity of sampling gear, season and timing) facilitate the comparison of results among the so widely different geographical areas included in this study but, in the other hand, makes it impossible to analyse any seasonal change in spatial distribution or population structure that are likely to occur.

MATERIAL AND METHODS
The samples analysed originate from a total of 6336 hauls performed during day-light hours between 30 and 800 m in spring (May-June) 1994 to 1999 on board several research vessels within the framework of the European Union research project "MEDITS" (Bertrand et al., 2000(Bertrand et al., , 2002)).The surveys took place along the European coasts of the Mediterranean Sea from the Straits of Gibraltar to the Aegean Sea, and extended to Morocco since 1999.The sampling procedures were standardised according to a common protocol.The bottom trawl used had a 4 m vertical opening and a 20 mm codend mesh size.Tows were performed at a speed of 3 knots.A random sampling stratified by depth, with proportional allocation of tows taking into account the area of each depth interval and geographical sector was used.Further details on the survey methodology and defined geographical sectors can be found in Bertrand et al. (2000Bertrand et al. ( , 2002)).In order to analyse the distribution patterns, two depth intervals have been used: shelf (depth <200 m), and slope (depths comprised between 200 and 800 m).
The allocation of trawl stations by geographical sector and year (1994)(1995)(1996)(1997)(1998)(1999) can be found in Bertrand et al. (2000Bertrand et al. ( , 2002)).Total catch for both species in the samples was weighed and all individuals counted and measured (carapace length, CL, in mm).Abundance and biomass are presented for each year as mean number of individuals and kg • km -2 by depth stratum and geographical sector, respectively.
Size frequency distributions were reconstructed for each sector and depth stratum (shelf/slope).Mean size values have been estimated for every geographical sector and year.Correspondence analysis (Greenacre, 1984) has been applied to size frequency distributions in order to detect assemblages of samples based in the resemblance among their size structure.This is a useful multivariate method to summarise information when gradients, rather than strong groupings are found among samples (Badia and Do Chi, 1976;Bouchard et al., 1986).
In order to obtain an index of fishing pressure exerted on each sector, an analysis of the relative size structure using the length converted catch curve (Pauly, 1983) was performed.The analysis allowed to obtain an estimate of the instantaneous total mortality rate Z. Results for the different areas can be comparable if similar rates of natural mortality are assumed to occur in all the studied sectors.The results of the mentioned analysis can be useful for a better understanding of the causes of observed differences in demographic structure among sectors that can not be explained by environmental conditions.The analysis, however, was performed without distinction of sex.Considering that a single couple of Von Bertalanffy's growth parameters L inf and K were used for each species, this choice may have produced a light overestimation of the Z values obtained.For Nephrops norvegicus, values of L inf = 72 mm and K=0.17 were used.In the case of Parapenaeus longirostris, chosen values were L inf = 47 mm and K=0.49.

Distribution
The presence of N. norvegicus was detected in 2161 hauls (470 on the shelf, at depths shallower than 200 m, and 1691 on the continental slope, deeper than 200 m).By computing the percentage occurrence of N. norvegicus on the continental shelf and slope for all the geographical sectors considered (Table 1), it can observed that the highest proportions of N. norvegicus on the continental shelf were found in the Adriatic Sea, especially in the northern (91.5% of the occurrences) and central sectors, as well as in the E Ionian Sea and Argosaronikos.In the rest of the considered geographical sectors, N. norvegicus was mainly found on the continental slope, with the exception of the N Aegean Sea and the Gulf of Lions -Corsica sectors, where a fair proportion of occurrences was found on the shelf.
The highest values of abundance, both in number of individuals and in biomass, were located in the 200-500 m depth stratum, especially in the Catalan Sea, Gulf of Lions, E Corsica, NE and W Sardinia, NE and SW Adriatic, E Ionian Sea and N Aegean Sea (Tables 2 and 3).High densities were also located in the 500-800 m depth stratum, especially so in E Corsica, NE and W Sardinia, E Ligurian and N Tyrrhen-DISTRIBUTION OF NEPHROPS AND PARAPENAEUS 127   (1994)(1995)(1996)(1997)(1998)(1999).Not sampled strata are indicated by '*'.Values higher than 25 kg km -2 are presented in bold.ian Seas.In the Adriatic Sea, the highest densities were also located on the 200-500 m depth stratum, but important densities were also located on the shelf.Interannual fluctuations in recorded densities were high, but in general, the patterns were similar throughout the years within each geographical sector.

Population size structure
Figure 1 shows the mean size of N. norvegicus obtained in the different years in each geographical sector, separately for the shelf and slope.Data analysed refer to sample sizes larger than 20 individuals.Some increasing-decreasing trends can be observed between different geographically adjacent zones.Thus, a clearly decreasing trend in mean size of the individuals from the slope is found along the Spanish Mediterranean waters from the Alborán Sea towards the East, through Alicante and the Catalan Sea.Mean size clearly increases from the Catalan Sea to the Gulf of Lions and Corsica.The mean size in Sardinia is similar to that of Corsica.The Ligurian-Tyrrhenian Sea has slightly smaller sizes whereas mean size in the S Thyrrenian Sea increases slightly.There is also a clear decreasing trend from the S Thyrrenian Sea passing through the Sicilian Channel to NW Ionian Sea.In the Adriatic Sea, the smallest mean sizes are found in the N-Central Adriatic (Italy) and in the NE Adriatic (Croatia) and the largest ones in the SW Adriatic (Italy) and the SE (Albania).In Greek waters, an increasing trend in mean size can be found from the E Ionian Sea to the S Aegean Sea through Argosaronikos and the N Aegean Sea.
Correspondence analysis has been applied to the matrix of size frequency distributions by geographical sector.As an example, Figure 2 shows the results of the analyses for two representative years: 1997 and 1999.Only samples larger than 60 individuals were analysed.Samples and size classes are presented in relation to the first two inertia axes.Correlative size classes are strongly associated to the first inertia axis and are found along a gradient.Samples (geographical sectors) are also found associated to the gradient of sizes.Thus, some sectors are found associated to large sizes, such as Corsica and the N Aegean in 1997 and S Adriatic-Albania, N Aegean, Sardinia, Corsica and S Aegean in 1999.Other geographical sectors appear associated to small sizes, such as N Adriatic (Croatia and Italy (both shelf and slope)), Sicily Channel and Catalan Sea in 1997 and Catalan Sea, N Ionian, N Adriatic shelf and S Adriatic) in 1999.Clearly, areas such as the N Adriatic and the Catalan Sea, among others, although geographically isolated and quite distant, are close-by placed in the analyses, indicating that their size structure was similar.

Mortality estimates
The estimates of total mortality derived from the analysis of the length transformed catch curve (Table 4) are shown in Figure 3.Total mortality was found significantly (p<0.05) and negatively correlated with the mean size obtained in each of the geographical sectors (Fig. 4).High mortality related to small mean sizes was particularly evident in the Catalan Sea, whereas the inverse relationship was found in the Alborán Sea and Corsica.The estimated mortality rates seem coherent with the available information on fleet importance and exerted fishing pressure in each sector.For each sector, estimates of Z in different years have shown very small differences, making results more reliable and suggesting equilibrium situations related to fishing effort and stable population structures.Sectors such as the Alborán Sea, Corsica, Argosaronikos and N Aegean showed lower values of total mortality and these findings look consistent with the modest fishing pressure on Nephrops that, according to available information, is exerted in the mentioned areas.An increase from West to East in total mortal-ity (as well as in number of boats per unit area) is observed for the three contiguous sectors of the Spanish coast.The highest Z values were found in the Catalan Sea, in which a high fishing pressure on the species is recorded.This positive trend is very coherent with the negative trend observed in mean size along the three sectors (Fig. 4).For some sectors, the samples were too small and did not allow a proper estimation of Z.
Some contrasts found in some sectors between mortality rates and available information on exerted fishing pressure can be explained by a lacking of homogeneity inside them.This phenomenon may 132 P. ABELLÓ et al.  (1994)(1995)(1996)(1997)(1998)(1999) and geographical sectors studied.Lines indicate mean values (black line) and 95% confidence intervals (dotted lines).
depend on the environmental characteristics or on the spatial distribution of fishing pressure inside a sector.For instance, the northern portion of the Ligurian-North Tyrrhenian sector is characterised by a very narrow shelf.The shelf becomes wider towards the south (especially around the Tuscan archipelago area).In the southern portion, more suitable grounds for the species are present and they occupy larger surfaces.Inside the mentioned sector, many fleets from several ports do operate.These fleets, of quite different importance, exert their fishing pressure on the grounds positioned inside the sector, especially on those closer to the ports from which they come from.In consequence, some fishing grounds (or some sub-sectors) will be more highly exploited than others.

Distribution
A total of 2161 presences of P. longirostris (805 caught at depths shallower than 200 m, and 1356 deeper than 200 m) have been analysed.The percentage occurrence of P. longirostris on the continental shelf and slope for all the considered geographical sectors (Table 1), shows that the species was mainly found on the continental shelf in the N, Central and SE Adriatic Sea and in the E Ionian Sea-Argosaronikos, whereas in the rest of the considered geographical sectors, P. longirostris was mainly found on the continental slope, especially so in the Gulf of Lions-Corsica, Sardinia and S Aegean Sea.
Overall, very low densities were found in the western Mediterranean, whereas the highest values were found in the Central and Eastern Mediterranean.Bathymetrically, the highest values of abundance in number of individuals were located in the 100-200 m and 200-500 m depth stratum, especially off Sardinia, S Tyrrhenian, Sicilian Channel, E Sicily, NW and E Ionian Sea, Argosaronikos and Aegean Sea (Tables 5 and 6).Biomasses were, however, usually higher in the 200-500 m depth stratum than in the 100-200 m stratum, in accordance with the depth-related size trend found in this species.Interannual variability was high and there seems to be a general increasing trend in the abundance of the species throughout the study period (Table 5).

Population size structure
Figure 5 shows the mean size of P. longirostris obtained in the different years in each geographical sector, separately for shelf and slope subgroups.Mean sizes on the shelf were clearly smaller than on the slope.No clear-cut trends can be found among the different geographical zones, except for the fact that the smallest mean sizes were found in the waters around Sicily and in the Greek sectors.In the western Mediterranean and Adriatic Sea, the mean sizes on the slope were higher than in the previously mentioned sectors.

Mortality estimates
The estimates of total mortality derived from the analysis of the length transformed catch curve (Table 7) are presented for P. longirostris in Figure 7. Mortality rates for P. longirostris have shown very wide differences among sectors but also a high internal variability in each sector among years.This variability can be attributed more to imprecise estimates of Z due to the low quality of the samples (small sample size in some areas due to low densities) than to a real variability in natural mortality rates or in changes in fishing pressure in time.Total mortality was significantly (p<0.05) and negatively correlated with the mean size obtained in each of the geographical sectors (Fig. 8).High mortality related to small mean sizes was particularly evident in Sicily and the South Aegean, whereas the inverse relationship was found in the Ligurian and South Adriatic seas.
For several areas (Alicante, Catalan Sea, Gulf of Lions, Corsica, NE Adriatic-Croatia), the reduced size of the samples made impossible the estimation of mortality rates.The highest values of total mortality were found in the Sicilian channel sector.This DISTRIBUTION OF NEPHROPS AND PARAPENAEUS 137   looks reasonable considering the high level of effort targeted to P. longirostris that occurs in the area.These high values are consistent with the very low mean size observed in the same sector.In other areas, however, these consistencies do not apply.For instance, in the North-Central Adriatic, the mean size is rather high but this does not seem to be related to the fishing pressure at which the species is exposed in the area.This phenomenon may be explained by the very low catches obtained on the shelf, where generally small-sized individuals are concentrated.The analysis of the few samples considered representative of the demographic structure in these areas suggest medium to high values of mortality rates.Parapenaeus longirostris.In both species, the marked differences observed among geographical sectors appear to be due to a combination of environmental characteristics and differences in fishing effort among the areas, all these coupled with the very different life history strategies of the two species, N. norvegicus being a benthic burrowing, long-lived species with slow growth rates, and P. longirostris an epibenthic, short-lived species with faster growth rates.Fishing activities differentially affect the populations of the two species.
In the case of Nephrops, the above mentioned differences may be correlated to ecological characteristics of the populations, but they probably also reflect actual different exploitation rates among populations.Levels of dissimilarity in size structure among N. norvegicus populations have been attributed to several factors, among which the most important appear to be differences in sediment characteristics as well as in exploitation rates (Chapman and Howard, 1988;Tully and Hillis, 1995;Maynou et al., 1996;Maynou and Sardà, 1997).In the Adriatic Sea, differences in size structure between shallow and deep fishing areas have been usually considered a consequence of environmental factors rather than of different levels of fishing effort (Froglia and Gramitto, 1987).In the present study, areas with high exploitation rates, as estimated from their fishing mortalities (Sardà et al., 1998), such as the Catalan Sea, the northern Adriatic or the Ligurian Sea, have appeared closely associated to small sizes, whereas areas with low exploitation rates, such as the Alborán Sea, Corsica or Sardinia, appear associated to large sizes.As an example of this fact, the clearly decreasing trend in mean sizes found along the Spanish Mediterranean closely matches the increasing trend in fishing effort found throughout the three Spanish sectors.Thus, according to Spanish official figures, the number of trawlers working in the Alborán Sea sector (12753 km 2 of trawlable bottoms) was 175, with a mean length of 15.9 m, 195.0 HP and 61.7 GT, the number of trawlers in Alicante sector (15928 km 2 ) was 263, with a mean length of 17.4 m, 275.2 HP and 66.6 GT, and the number of trawlers in the Catalan Sea (16578 km 2 ) was 490, with a mean length of 17.2, 304.1 HP and 62.1 GT.This figures give a fishing intensity of 1.4, 1.7 and 3.0 trawlers per 100 km 2 , respectively for the Alborán, Alicante and Catalan Sea sectors.
Although differences in growth have been highlighted for N. norvegicus from different habitats in the same geographical area (Central Adriatic) (Froglia and Gramitto, 1987) and such differences could affect the total mortality estimates, in this paper a unique set of Von Bertalanffy parameters was adopted only as a baseline for a first analysis and comparison of total mortality in the different sectors.Notwithstanding the necessity of a deeper comparative evaluation of the growth in the different geographical sub-areas, present preliminary estimates evidenced a lower total mortality in some sectors, such as Alborán Sea, Corsica, Argosaronikos and N Aegean, where fishing pressure is also presumably lower when compared with other subareas.
For the analysis of the length converted catch curve, it is necessary that the available length frequency data be representative of the population.Data used for catch curve estimates of Z must represent an equilibrium status or stable-age distribution.This is generally obtained by reconstructing the mean size structure of the stock by pooling in a proper way samples collected all along the year and assuming that recruitment strength has varied little along the considered years.The MEDITS data, however, only furnish the size structure during a single sampling period of the year (spring).We have considered that this may not constitute a very important problem in the case of Nephrops norvegicus, because several cohorts are present in the catch and so, the decline in numbers with time can be analysed without major problems.In the case of Parapenaeus longirostris, as well as for any other short-lived fastgrowing species, the use of length-based methods involves a dynamic non-equilibrium situation.In these cases, it is necessary to adequately sample in both space and time, since biomass and numbers change greatly on a seasonal basis.Smaller time intervals for samples (at least one by season) should be needed to avoid missing important events in the life history of these species.The lack of samples from other seasons than spring can be critical in the present case.Notwithstanding the problems described above, the analysis was also done for P. longirostris.Results, however, have to be considered, specially for this species, only as indicative and must be handled with care.
It is probable that the von Bertalanffy growth parameters used here be unsuitable for properly describing the growth performance of some of the stocks included in the analysis.Growth and natural mortality rates can change depending on food availability, competition, etc.Moreover, suitability of fishing grounds can determine differences in size structure.All these aspects should invalidate the uncritical comparison of the Z estimates, especially in the case of very distant areas.
Parapenaeus longirostris was found widely distributed throughout the outer shelf and upper portion of the continental slope.Its biomass at the border of the shelf was however lower than on the upper slope where adults concentrate, as reported in other Mediterranean areas (Audouin, 1965).In some sectors of the western Mediterranean the presence of the species was always scarce.This can not only be attributed to a low sampling density in the depth range considered optimal for the species, since these depths were fully sampled, but to an intrinsic population behavior of the species probably linked to mesoscale oceanographical processes.
Concerning the analysis of the demographic structure of the population of Parapenaeus longirostris by area, the highest values of total mortality rates found in the Sicilian Channel look reasonable considering the high level of effort directed to this shrimp that takes place in the area (authors unpublished data).These high values are consistent with the observed low mean size in the sector.In other areas, however, these consistencies do not apply.For instance, in the North-Central Adriatic, the mean size is rather high but this is hardly to be related to the fishing pressure at which the species is exposed in the area.This phenomenon may be better explained by the very low catches obtained on the shelf, where generally small-sized individuals are concentrated.The analysis of the few samples considered representative of the species demographic structure in these areas suggest medium to high values of mortality rates.Similar mean sizes (the smallest ones) were observed in Sicily and Greek waters, where differences in fishing pressure are expected.This could be explained by the strength of the recruitment pulses, higher in these areas than in the others (authors unpublished observations).However, hydrological factors could also be evocated.Bombace (1972) and Levi et al. (1995) suggested a work hypothesis based on migration as a continuous flow from east to west supported by the role that water masses (Intermediate Levantine Waters) might play.
Different levels of exploitation appear to affect the population size structure of N. norvegicus in a much more marked way than in P. longirostris, since Nephrops population structure is less dependent on variations in recruitment success considering that more age classes constitute the species exploitable fraction.N. norvegicus is a long-lived, benthic, sedentary species with slow growth and mortality rates, whereas P. longirostris is a short-lived epibenthic species, much more mobile being a natant shrimp and with relatively high growth and mortality rates (Farmer, 1975;Ribeiro-Cascalho and Arrobas, 1987;Sardà, 1995).It can be hypothesised that, if fishing effort is of major importance, changes may also be detectable in the demographic structure of Parapenaeus longirostris.Interannual differences in recruitment strength are more likely to affect the population structure of P. longirostris, which is also more dependent on high growth and mortality rates than that of N. norvegicus, which, given its habits (burrowing behaviour) and population dynamics, is expected to show more constant characteristics in both population dynamics and size structure.
FIG. 1. -Mean sizes (CL, in mm) of Nephrops norvegicus obtained in 1994-1999 in the different geographical sectors.Triangle: shelf samples; open circle: slope samples.The black line follows the mean values of the slope populations.
Figure5shows the mean size of P. longirostris obtained in the different years in each geographical sector, separately for shelf and slope subgroups.Mean sizes on the shelf were clearly smaller than on the slope.No clear-cut trends can be found among the different geographical zones, except for the fact that the smallest mean sizes were found in the waters around Sicily and in the Greek sectors.In the western Mediterranean and Adriatic Sea, the mean sizes on the slope were higher than in the previously mentioned sectors.As for Nephrops, correspondence analysis was applied to the matrix of size frequency distributions by geographical sector.Figure6shows the results of the analysis for two representative years: 1997 and FIG. 5. -Mean sizes of Parapenaeus longirostris obtained in 1994-1999 in the different considered geographical sectors.Triangle: shelf samples; open circle: slope samples.The black line indicates the mean values of the slope populations, whereas the dotted line follows those on the shelf.

DISCUSSION
Different patterns are found concerning the geographical distribution and observed trends in population size structure of Nephrops norvegicus and 138 P. ABELLÓ et al.

TABLE 1 .
-Percentage of the total number of occurrences of N. norvegicus and P. longirostris on the continental shelf and slope in each of the geographical sectors analysed.

TABLE 3
. -Nephrops norvegicus: Mean biomass (kg km -2 ) estimated from the MEDITS trawl surveys per depth stratum, geographical sector and year

TABLE 4 .
-Total mortality estimates (Z) for Nephrops norvegicus in the different years and Mediterranean Sea subareas sampled, based on population size structure.sd= standard deviation; CV= coefficient of variation (%).
FIG. 3. -Total mortality (Z) estimates for Nephrops norvegicus in the different years

TABLE 6 .
-Parapenaeus longirostris: Mean biomass (kg km -2 ) estimated from the MEDITS trawl surveys per depth stratum, geographical sector and year

TABLE 7 .
-Total mortality estimates (Z) for Parapenaeus longirostris in the different years and Mediterranean Sea subareas sampled, based on population size structure.sd= standard deviation; CV= coefficient of variation (%).