The purple dye murex, Bolinus brandaris (Linnaeus, 1758), is a common muricid species that is extensively dispersed throughout the Mediterranean Sea, whereas its distributional range along the adjacent Atlantic Ocean is restricted mainly to the coasts of Portugal and Morocco (Poppe and Goto 1991, Macedo et al. 1999, Houart 2001). Probably as a consequence of accidental introduction of juveniles, B. brandaris has extended its distributional range northwards, and has already been recorded in Galicia (northwestern Spain) (Bañón et al. 2008). This species usually occurs in shallow waters of the sub-littoral zone (Dalla Via and Tappeiner 1981), but can also be found at 100 m (Muzavor and Morenito 1999), 150 m (Houart 2001) or even 200 m depth (Macedo et al. 1999, Malaquias 2007), inhabiting sandy, sandy-muddy and muddy bottoms (Macedo et al. 1999, Muzavor and Morenito 1999, Malaquias 2007).

The purple dye murex was harvested during the Roman Empire using baited wicker baskets (Spanier and Karmon 1987) in order to extract the purple dye that was greatly prized in the ancient Mediterranean (e.g. Reese 1980, 2010, Oliver 2015). Today, B. brandaris is fished for human consumption using diverse types of artisanal fishing gears (Poppe and Goto 1991, Houart 2001), mainly in Portugal (Vasconcelos et al. 2008a) and Spain (Martín et al. 1995, Tirado et al. 2002, Mallol et al. 2004), where this species is an important gastronomic resource and is highly valued in local seafood markets (Ramón et al. 2005, Vasconcelos et al. 2008a), but also in France (Bartolome 1985), Italy (Ramón and Amor 2001, Cecalupo et al. 2006) and Greece (Katsanevakis et al. 2011), and occasionally in Turkey (Ramón and Flos 2001) and Tunisia (Elhasni et al. 2013).

Knowledge on meat yield (weight and proportion of the raw edible content) of seafood products is important, both as general information for the consumers and as specific data for the seafood processing and marketing industry. However, despite its importance and the existence of specialized literature on seafood worldwide (e.g. FAO 1989), information on the meat yield of shellfish and specifically of gastropod species is still relatively scarce and restricted to a few commercial species (either fishery-exploited or aquaculture-produced species) (see compilation by Vasconcelos et al. 2009). In Portugal, the purple dye murex is fished along the Algarve coast (southern Portugal), mainly in the Ria Formosa lagoon. It is a greatly appreciated seafood with high demand and commercial value in local seafood markets, reaching prices of around 20-25 € kg^–1 for first sale (Vasconcelos et al. 2008a). In addition, the commercial value of B. brandaris has prompted a few studies to assess its potential as a candidate species for molluscan aquaculture in both Spain (Ramón and Flos 2001, Vela and Moreno 2004, Ramón et al. 2005) and Portugal (Vasconcelos et al. 2012a).

The present study reports the estimation of meat yield and analyses the influence of factors such as sex, size and reproductive status on the meat yield of B. brandaris from the Ria Formosa lagoon (southern Portugal). Besides improving the overall knowledge on the species biology, this kind of information is important to increase the awareness on the need to sustainably harvest B. brandaris, as well as to support a responsible commercialization and consumption of this locally important shellfish resource.

A total of 563 individuals of B. brandaris, with broad ranges in both SL (48.6-107.7 mm) and TW (4.6-79.7 g), were subjected to analyses for estimating MY (Table 1). In the overall samples, females (n=308) were significantly more abundant than males (n=255) (chi-square test: χ²=4.803, p<0.05), corresponding to a female-biased sex ratio of 1 M:1.2 F. Male and female B. brandaris showed similar mean shell length (males 72.7±7.4 mm SL; females 73.8±8.7 mm SL) and mean total weight (males 20.0±6.2 g TW; females 21.1±8.8 g TW), without statistically significant differences between sexes in either shell length (H=1.039, p>0.05) or total weight (H=0.326, p>0.05) (Table 1). Overall, average MY of B. brandaris was 40.5±6.1%, ranging from a minimum of 25.8% to a maximum of 56.1%. There were no statistically significant differences (H=0.667, p>0.05) in the average MY obtained from males (40.7±6.2%; range=25.8-55.3%) and from females (40.4±6.0%; range=27.2-56.1%) (Table 1).

The relationships established between specimen SL, TW and SpW in males and females of B. brandaris are presented in Figure 3. In both sexes, all relationships were highly significant (p<0.001) and invariably displayed high correlation coefficients (r=0.904 to 0.955), slightly higher in females (r=0.919 to 0.955) than in males (r=0.904 to 0.933) (Fig. 3). Overall, these relationships indicated that specimen size (in both SL and TW) is an excellent indicator of SpW, and consequently of raw MY. Regarding the relative growth between variables (isometry vs. allometry), only the relationship SL vs. TW in males showed isometric growth (b=2.872±0.086; range=2.703-3.040), while all remaining relationships showed positive allometries (b=1.124 to 3.711) that revealed that throughout ontogeny the SpW increased at a proportionally faster rate than SL and TW (Fig. 3). Regarding the comparison of relative growth between males and females, the relationship TW vs. SpW displayed similar growth between sexes (b_M=b_F: t-test=1.514, p>0.05), whereas statistically significant differences between males and females were detected in the relationships SL vs. TW (b_M<b_F: t-test=8.704, p<0.001) and SL vs. SpW (b_M<b_F: t-test=2.660, p<0.01), with females showing proportionally faster grow rates than males in both TW and SpW during ontogeny (Fig. 3).

The variation in MY as a function of the size of B. brandaris, with individuals of both sexes grouped into five SL size classes, is presented in Figure 4. Both males and females displayed highly significant differences in MY between size classes (males H=53.801, p<0.001; females H=56.595, p<0.001), clearly reflected in an evident trend of growing MY with increasing SL (Fig. 4). Similarly, the comparison of MY obtained from individuals below and above the MLS established for B. brandaris (MLS=65 mm SL) also showed highly significant differences in the MY obtained from individuals of both sexes below and above the MLS (males F=56.675, p<0.001; females H=27.004, p<0.001), reflecting much lower MYs in individuals below the MLS (males=34.4±4.7%; females=35.6±4.5%) than in individuals above the MLS (males=41.8±5.8%; females=41.0±5.9%) (Fig. 4). In both cases, i.e. with individuals grouped into five size SL classes and distributed below and above the MLS, there were no statistically significant differences (F or H, p>0.05) in MY between sexes within the same size class.

The monthly variation in the MY obtained from B. brandaris is presented in Figure 5. The MY of both males and females displayed highly significant monthly oscillation throughout the one-year study period (males F=37.583, p<0.001; females F=33.827, p<0.001), with both sexes reaching minimum yields in January (males 33.5±4.8%; females 34.2±3.7%) and maximum yields in June (males 49.4±2.6%; females 50.2±3.6%) (Fig. 5). Although the MY in both sexes exhibited a fairly similar and synchronous temporal trend, namely a marked decrease between December and January, followed by a clear increase from January to February, and another simultaneous sharp decrease between June and July, still some statistically significant differences in MY between sexes were detected in the monthly samples of October (females>males F=4.132, p<0.05), February (males>females F=6.860, p<0.05), March (females>males H=4.987, p<0.05) and September (males>females F=6.527, p<0.05) (Fig. 5).

The influence of diverse indicators of the reproductive status on the MY of B. brandaris is depicted in Figure 6. The monthly variation in MY in both sexes slightly reflected the fluctuation in the mean GI during late spring - early summer. Indeed, sharp declines in male GI in May-August (from a maximum of 5.0 to a minimum of 1.1) and in female GI in June-August (from 4.4 to a minimum of 1.4), reflecting the decrease in the proportion of ripe gonads and the subsequent increase in the proportion of partially spent and spent gonads, coincided with significant reductions in the MY of both males (from 45.1% in May to 37.4% in August) and females (from 50.2% in June to 37.0% in August) (Fig. 6). Similarly, the gonadosomatic index (GSI), indicative of gonadal maturation and inherent variation in gonad weight in both sexes, displayed abrupt decreases during the above period (males May-July; females June-July), reflecting gonad emptying during gamete release and female spawning. These decreases in both male GSI (from 29.4% in May to 12.2% in July) and female GSI (from 22.2% in June to 12.4% in July) were accompanied by marked reductions in MY in both males (from 45.1% in May to 38.2% in July) and females (from 50.2% in June to 39.9% in July) (Fig. 6). The sex-specific indices (male PI and female CGI) also exhibited the same general temporal trend of the previous bio-physiological indices. The PI, indicative of the copulative activity and consequent male gamete release, decreased continuously between the peak of copulation in May (12.5%) and September (8.5%), encompassing the period of marked reduction in male MY (from 49.4% in June to 37.4% in August). The CGI also displayed a sharp decrease between a maximum in May (7.8%) and a minimum in August (1.4%), indicating the deposition of egg capsules (oothecae) during the spawning season and consequent emptying of the capsule gland, which further contributed to the reduction in female MY recorded from June (maximum of 50.2%) to August (37.0%) (Fig. 6).

The average MY obtained from the purple dye murex (sexes combined) from the Ria Formosa lagoon was 40.5±6.1%, ranging between 25.8 and 56.1% during the study period. As stated above, information on the MY of gastropod species is still relatively scarce, and to the author’s best knowledge in the available literature there is only one other study reporting the MY of B. brandaris (42.5%) from the Adriatic Sea (Dalla Via and Tappeiner 1981). The comparison of the MY obtained from B. brandaris with that of other species of the family Muricidae, including some edible and/or commercially valuable muricids, is compiled in Table 2. It is worth stating that such comparisons in some muricids (e.g. B. brandaris, H. trunculus and S. haemastoma) should be cautiously interpreted because their edible content varies depending on the country where these species are used for human consumption. For instance, whereas in the northeastern Atlantic and northern Mediterranean the whole meat (foot and visceral mass) is ingested, in the southern Mediterranean only the foot is consumed. In general, the average MY of the purple dye murex from the Ria Formosa lagoon (40.5%) is higher than that obtained from other sympatric muricid species, namely H. trunculus, with values of 33.3% in Cyprus (Alyakrinskaya 2005), 33.6% in Slovenia (Dalla Via and Tappeiner 1981), 34.2% in Greece (Alyakrinskaya 2005) and 37.9% in Portugal (Vasconcelos et al. 2009), and S. haemastoma with values ranging from 22.8% in Tenerife (Alyakrinskaya 2005) to 34.1% in Majorca (Alyakrinskaya 2005) (Table 2). This higher MY of B. brandaris compared with H. trunculus and S. haemastoma is due to marked differences in their shells in terms of thickness, strength and ornamentation, and consequently, in their shell relative weight (e.g. Dalla Via and Tappeiner 1981, Alyakrinskaya 2005, Vasconcelos et al. 2016). For instance, the contribution of shell weight to TW in H. trunculus is almost 10% higher than that in B. brandaris (Dalla Via and Tappeiner 1981), whereas in S. haemastoma the contribution is over 10% higher than that in H. trunculus (Alyakrinskaya 2005).

The relationships established between B. brandaris size and the respective SpW revealed that both SL and TW are excellent indicators of SpW, and therefore of the raw MY that can be obtained from the purple dye murex. This information is most useful both for the seafood processing industry and for consumers, because the relationships SL vs. SpW and TW vs. SpW allow us to predict the edible content of B. brandaris as a function of the individual size (either in SL or TW), and thus estimate the MY that can be obtained from a batch of purple dye murexes with a certain size or weight range. Concerning relative growth, both males and females displayed positive allometries in the relationships SL vs. SpW and TW vs. SpW, revealing that during growth the edible content increased at a proportionally faster rate than SL and TW (i.e. the largest and heaviest individuals provide proportionally higher MYs), which in practice means that B. brandaris allocates more energy resources to somatic growth than to shell secretion throughout ontogeny. Inter-specific comparisons of the type of growth (isometry vs. allometry) are rather complicated due to high variability in shell morphology and structure between species and their variation during growth. Still, in B. brandaris the allometry coefficients of the relationships SL vs. SpW (males, b=3.607; females, b=3.711) and TW vs. SpW (males, b=1.169; females, b=1.124) are higher than those reported for other muricid species, such as Chicoreus ramosus (SL vs. meat weight, b=2.805) (Thapnu and Tantichodok 1991), Pinaxia coronata (SL vs. fresh tissue weight, b=2.801) (Udoh and Abiaobo 2014) and H. trunculus (SL vs. meat weight, b=3.557; TW vs. meat weight, b=1.118) (Vasconcelos et al. 2009). Concerning the comparison of relative growth between sexes, in the relationships SL vs. TW (males, b=2.872; females, b=3.198) and SL vs. SpW (males, b=3.607; females, b=3.711), females exhibited significantly higher allometry coefficients than males, indicating faster grow rates in both TW and SpW throughout ontogeny, which are probably due to particular features of the female reproductive system, namely a female-specific accessory sexual organ, the capsule gland that reaches a considerable size and weight during the spawning season (please see further details below about the influence of the reproductive status on the MY of B. brandaris).

The trend of increasing MY during the ontogeny of B. brandaris was further highlighted by grouping the individuals of both sexes into SL size classes. Indeed, highly significant differences in MY between size classes in both males and females constitute additional evidence of growing MY with increasing SL, i.e. smaller individuals yield significantly less edible content than larger individuals. Accordingly, individuals below the MLS legally stipulated for B. brandaris (<65 mm SL) provided significantly lower MYs (males=34.4%; females=35.6%) than individuals that achieved the minimum landing size (≥65 mm SL) (males=41.8%; females=41.0%), further reinforcing the importance of compliance with the species MLS by fishermen/harvesters, seafood traders and shellfish consumers. In fact, since commercially undersized specimens yield significantly lower edible content than individuals above the MLS, this should also help discourage their commercialization and consumption, in order to avoid unconsciously paying for shells at the price of meat, considering the high commercial value of the purple dye murex (which fetches prices around 20-25 € kg^–1 for first sale) (Vasconcelos et al. 2008a). Moreover, the existence of commercially undersized B. brandaris available in local shellfish suppliers and seafood markets justifies the need to improve regulation measures and reinforce inspection activities focusing on this artisanal fishery.

The MY of B. brandaris exhibited significant monthly variations and an analogous and synchronous temporal trend in both males and females during the study period, which were undoubtedly related to their reproductive status, as evidenced by the monthly oscillation in diverse bio-physiological indices calculated for both sexes. In fact, the most noteworthy monthly variation in the MY of both males and females was a significant decrease recorded from June to July, which was accompanied by a simultaneous reduction in most indices of reproductive condition, namely the mean GI and the GSI in both sexes, the male PI, and the female CGI. Chronologically, such sharp decreases in these indices correspond to the following sequential phases in the reproductive cycle of B. brandaris: peak of male copulative activity in May (PI); maturation peak in May (males) and June (females) followed by sharp decreases in both GI and GSI, which denote decreasing proportions of ripe gonads and consequent replacement by increasing proportions of partially spent and spent gonads (GI) as a consequence of gamete release (GSI) in both sexes; and female spawning peak between June and July (CGI). Overall, the sharp decreases in these bio-physiological indices reflect significant reductions in the size and weight of male and female gonads and in the female capsule gland, which are inevitably translated immediately into a substantial decrease in SpW and consequently a significant decline in the raw MY of the purple dye murex between June and July. Indeed, during the period of reproductive maturation, both male and female B. brandaris develop large gonads and females fill the capsule gland responsible for the formation of the egg capsules (oothecae) that encapsulate the developing eggs and embryos, an accessory sexual organ that also reaches considerable size during the spawning season (e.g. Amor et al. 2007, Vasconcelos et al. 2012b, Elhasni et al. 2013), and whose deflation and emptying after gamete release and spawning is noticeably reproduced as an abrupt decrease in the raw MY of the purple dye murex.

In certain fishing resources, the optimum harvesting season should coincide with the period of highest MY (e.g. Udoh and Abiaobo 2014), which allows the weight of the catches/landings and the respective commercial value and resulting profits to be maximized. In the present case, besides the general concerns related to the sustainability of the fishing/harvesting activity and the appropriate management of this fishing resource, data on the MY of the purple dye murex confirmed the importance and interest of accomplishing the MLS=65 mm SL legally stipulated for the catches of B. brandaris in the Ria Formosa lagoon (D.R. 1990, 2001), primarily by fishermen/harvesters, but also by seafood traders and shellfish consumers. In addition, the present data also confirmed the influence of the reproductive status on the edible content of the purple dye murex, with the highest MYs being recorded in the period of reproductive maturation (i.e. prior to the gamete release and the spawning peak). In this particular, it should be taken into account that the fishing/harvesting activity targeting B. brandaris is subjected to a closed season (1 May - 30 June) (D.R. 2010), which was deliberately implemented during the spawning season in order to protect the female broodstock and collective spawns (Vasconcelos et al. 2012b). For this reason, although the highest MYs of B. brandaris coincide with the closed season of the fishing/harvesting activity, this temporary prohibition of the catches should be rigorously fulfilled by fishermen/harvesters in order to ensure a balanced and sustainable long-term exploitation of this locally important fishing resource of the Ria Formosa lagoon.

Finally, the marked influence of the reproductive status on the MY detected in the present study prompted a comparative assessment of the spawning season and peak, including spawning events observed in the field and in the laboratory, of three sympatric muricid species (B. brandaris, H. trunculus and S. haemastoma) throughout their geographic distributional range (Table 3). This multi-species comparative approach compiles most relevant baseline information consulted in virtually all literature available on this subject, providing supporting data from a long-time period (1888-2015) and thus allowing analysis of long-term trends. In addition, this dataset comprises information gathered at locations covering a wide geographical area (Atlantic Ocean, Mediterranean, Adriatic, Tyrrhenian, Ionian and Aegean Seas) and including ecologically diverse habitats (e.g. coastal areas vs. coastal lagoons). Overall, 55 data records on studies of the spawning season and peak of those three muricid species, obtained through biological sampling and analyses, field and laboratory observations, were collected from 41 bibliographic sources published by 27 different authors (Table 3).

According to the information available in the literature, these gastropod species display slight variation in the timing of reproduction throughout their distributional range, namely in the spawning season and peak, most probably associated with the colder seawater temperatures in the Atlantic Ocean compared with the Mediterranean Basin. As a consequence, Mediterranean populations of these species show a lengthy spawning season, generally ranging from spring to summer (sometimes even with an unusual longer duration), as well as more than one spawning period during the year (Table 3). Gonadal ripening and spawning are apparently triggered by the progressive rise in seawater temperatures from late spring to early summer. This reproductive dynamics induces a significant increase in the volume and weight of the gonads and accessory reproductive organs, subsequently reflected in the species MY, as is clearly demonstrated in the present study.

Indeed, gastropod reproductive dynamics, including the seasonality of gonadal ripening and spawning, are regulated by environmental cues that allow reproduction to occur at the most favourable time (Wayne 2001, Harding et al. 2008, Hotchkiss et al. 2008), with seawater temperature and food availability controlling the timing of spawning and being responsible for inter-annual variations in the reproductive cycle (Sternberg et al. 2010). Accordingly, as for the sympatric B. brandaris, H. trunculus and S. haemastoma, similar reproductive patterns (i.e. spawning season coinciding with rising seawater temperatures) have been reported for several other muricid species worldwide, such as Hexaplex nigritus (Cudney-Bueno et al. 2008), Rapana venosa (Saglam and Duzgunes 2007, Harding et al. 2008, Saglam et al. 2009), Reishia clavigera (Tong 1988, Lee 1999), Tenguella musiva (Tong 1988) and Trophon geversianus (Cumplido et al. 2010).

In conclusion, the present study showed that MY of the purple dye murex increases during ontogeny, as revealed by the relationships established between specimen size and SpW and by significant differences in MY between size classes. In addition, MY exhibited significant monthly variation directly related to reproductive status, as evidenced by the calculation of bio-physiological indices. Overall, these findings have implications at diverse levels, including the management, regulation and inspection of the artisanal fishing/harvesting activity and the commercialization and consumption of this valuable seafood product. In addition to B. brandaris, the same applies to other muricid species exploited for human consumption, including H. trunculus and S. haemastoma.