The mollusc shell is a suitable biotope for the settlement and development of several groups of invertebrates. In molluscs, biofouling has mainly been described in bivalves such as Pinna bicolor (Keough 1984), Chlamys opercularis (Ward and Thorpe 1991), Crassostrea gigas (Duault et al. 2001), Venus verrucosa (Trigui El Menif et al. 2005), Flexopecten felipponei (Schejter and Bremec 2006), Lithophaga lithophaga (Trigui El Menif et al. 2007) and Pinna nobilis (Rabaoui et al. 2009). The associated fauna highlighted in these bivalve species belongs to several zoological groups: bryozoans, annelids, serpulid polychaetes, tunicates, sponges, crustaceans, ascidians, cnidarians and echinoderms.

Few studies have as yet dealt with shell fouling and its effects in gastropods. These include the study by Buschbaum and Reise (1999) reporting the effects of barnacle colonization on the shell of Littorina littorea and those of Bick (2006) and Vasconcelos et al. (2007) on polychaete fixation in the shells of Stramonita haemastoma and Hexaplex trunculus, respectively.

Few studies of the biology of S. haemastoma have been published, and they deal mainly with reproduction, spawning and intracapsular development (Beliste and Stickle 1978, Lahbib et al. 2011). The most studied aspect in S. haemastoma is the imposex phenomenon (Spence et al. 1990, Rilov 1999, Lemghich and Benajiba 2007, Lahbib et al. 2010, El Ayari et al. 2015), the sexual deformity associated with marine pollution by organotin compounds such as tributyltin (TBT) and its derivatives (Terlizzi et al. 2001). Recently, the effect of epibiotic barnacles on imposex in S. haemastoma sampled from Tunisian coasts was investigated (El Ayari et al. 2015).

Taking into consideration the well-known effects of biofouling that could possibly affect the biology of the basibiont, the present study aimed to identify species associated with the shell of S. haemastoma from Tunisian waters and to investigate whether this fouling could affect the relative growth and reproductive indices of this locally abundant gastropod species.

Specimens of S. haemastoma with a shell length of 20-80 mm were collected monthly (N=80-120) on a rocky bottom from June 2009 to May 2010. Sampling was performed by scuba diving at 11 m depth at a station located in the artificial channel linking the Bizerta lagoon with the Mediterranean Sea (Fig. 1). Seawater temperature and salinity were measured using a multi-parameter sounder. In the laboratory, fauna living on the shell external surface was collected by washing the shells using seawater from the same sampling station.

Comparison of relative growth and reproductive indices in S. haemastoma was investigated monthly in two groups: 40 to 60 gastropods non-fouled by barnacles (GnFB) and 40 to 60 gastropods fouled by barnacles (GFB). The position and number of barnacles per individual were determined after removing them from the shells of the hosts using a scalpel. Gastropods were then sacrificed by freezing at –20°C; shell length (SL), shell diameter (SD), and penis length (PL) were measured to the nearest 0.1 mm using a digital caliper. After thawing, the shell was broken using a bench vice and the soft part of the organism was carefully removed. Endobiotic fauna was collected after shell breakage, identified following macroscopic observation or under a binocular microscope, and then preserved in 70% alcohol.

In S. haemastoma, three gonad maturation stages were detected following macroscopic observations of the gonads in both sexes and of the capsule gland in females, following Ramón and Amor (2002) in the sympatric muricid Bolinus brandaris. Stage I (immature) corresponds to undifferentiated gonads from the underlying capsule gland in both sexes and females having an inconspicuous capsule gland; stage II (intermediate) gonads in both sexes are more developed and correspond approximately to one-third of the area of the digestive gland; and stage III (mature) males show a well-developed, light brown testicle corresponding to more than half of the area of the digestive gland, whereas females have a voluminous yellowish ovary and a large yellowish capsule gland.

The flesh wet weight (FwW) was recorded after removing the operculum. The female capsule gland was separated and weighed (CGwW). A cross-section was made on the coiled part of the organism, directly under the stomach, and was photographed to measure the gonad area (GA) and the area of the digestive gland–gonad complex (DGGA) using the software Image J 1.38 x. The shell dry weight (SdW) and the flesh dry weight (FdW) were recorded after drying them at 60°C for 3 days.

The reproductive condition of S. haemastoma was assessed using the following indices, expressed as percentages:

- Gonadosomatic index (GSI) = digestive gland-gonad complex dry weight/SdW (Lucas and Beninger 1985);

- Capsule gland index (CGI) = CGwW/FwW (Giménez and Penchaszadeh 2003);

- Gonad area index (GAI) = GA/DGGA (Poore 1973);

- Penial index (PI) = PL/SL (Vasconcelos et al. 2011).

Relationships between morphometric and ponderal variables (SL, SD, SdW and FdW) were established through regression analysis (Y=aX^b). Relative growth between variables was determined by comparing the regression slopes (b) using the Student t-test (Mayrat 1959). A chi-square test was employed to verify a balanced proportion (1M:1F) in the sex ratio of the samples. Comparisons between fouled and non-fouled gastropods were made through analysis of covariance (ANCOVA - model of homogeneity of slopes) using the software STATISTICA 10, with the covariable being always SL. Post-hoc pairwise comparisons were made using the Tukey test. Monthly variations in bio-physiological indices were analysed by one-way ANOVA (Kruskal-Wallis test). In all statistical analysis, significance level was considered for P<0.05.

The present study has shown that the shell of live S. haemastoma collected from Bizerta Channel is a suitable biotope for several invertebrate species. The most abundant epibiotic species were barnacles that totally covered the shell of several snails. Sedentary polychaetes and lithophagous bivalves were the most abundant endobiotic species. For polychaetes, a prevalence of cirratulids especially represented by Dodecaceria concharum was recorded. In S. haemastoma from the Spanish coast, the shell of live individuals sheltered, in addition to D. concharum, three other species, namely Dipolydora armata, Capitella minima and Spirobranchus polytrema (Bick 2006). Seven families of polychaetes found in the present study (Spionidae, Ciratulidae, Syllidae, Nereidae, Sabellidae, Sabellariidae and Serpulidae) were also reported in H. trunculus from Portuguese waters by Vasconcelos et al. (2007). These authors showed that the dorsal surface of the shell, mainly in the proximity of the apex, was generally more fouled by epibiotic polychaetes than the ventral surface, a pattern of colonization which is in agreement with the present results and that is probably related to the mode of locomotion of this snail. Indeed, it has been shown that S. haemastoma moves by crawling on hard substrate and by burrowing on soft substrate when the rocks become exposed during low tides (Papp and Duarte 2001). In both cases, the continuous abrasion of the ventral surface leads to an erosion of the settled epibionts, with the same finding being also recorded in H. trunculus (Vasconcelos et al. 2007).

Fouling is likely to be a function of time, food availability, seawater salinity and temperature (Wahl 1989, Sahu et al. 2013). In fact, it has been shown that salinity influences all physical-chemical variables of seawater and controls the larval dispersal and recruitment patterns of epi- and endobiotic species (Nair 1965), while seawater temperature influences chemical and biological interaction and regulates the growth of benthic organisms (Sahu et al. 2013). Consequently, the shells of smaller gastropods are usually less colonized by epifauna than those of larger gastropods, which also agrees with our observations. Abundant colonization could have serious and deleterious consequences on the basibiont. Indeed, Thieltges and Buschbaum (2007) and Buschbaum et al. (2007) showed that the presence of the worm Polydora ciliata on the shell of Littorina littorea facilitates barnacle fixation, leads to a gradual destruction of the shell, reduces the resistance against predators and reduces the fecundity and the growth of the gastropods. According to Lleonart et al. (2003), an infestation of 30% in the gastropod Haliotis spp. from the Australian coast by two spionid species (Boccardia knoxi and Polydora hoplura) leads to the mortality of half of the affected population.

For endofauna, we recorded the presence of L. aristata and R. rubia, with infestation rates of 10.1% and 0.1%, respectively. Trigui El Menif et al. (2007) isolated those two boring mytilids from the rock hosting L. lithophaga collected in the Bay of Bizerta. Rocellaria dubia was also found by the same authors in breaches dug in the shell of live Venus verrucosa. Tebble (1976) reported the presence of R. dubia in various substrates (sand, limestone, sandstone and dead mollusc shells) but not in live specimens. This bivalve, as well as the other lithophagous species, has a pair of pallial glands that secrete a calcium-binding mucoprotein for boring into calcareous organisms (Jaccarini et al. 1968). The continuous growth of the endobiont involves a growth in the volume of the breach causing a progressive perforation of the shell of the basibiont, which in some cases leads to its death. This finding was recorded by Simone and Gonçalves (2006) and by Trigui El Menif et al. (2006), respectively in Nodipecten nodosus infested by L. aristata and in V. verrucosa infested by Rocellaria dubia. With regard to our results, the infestation of S. haemastoma by L. aristata does not seem to act negatively on the gastropod at this level, because the extracted endobiont had a size of 17 mm in length and 6 mm in thickness, which does not exceed either the length or the thickness of the gastropod columella. Negative effects in the basibiont could probably occur with the progressive growth of the endobiont, since L. aristata can attain a length of 52 mm (Turner and Boss 1962). According to Simone and Gonçalves (2006), the geographical distribution of the L. aristata is probably limited to the Atlantic coasts. Moreover, Ávila et al. (2009) discovered L. aristata as a fossil form on the island of Santa Maria (Azores). The presence of this species in Tunisian coasts, in the same habitat as L. lithophaga (Trigui El Menif et al. 2006) and in the shell of S. haemastoma, lets us suppose that it is an invasive species coming from the Strait of Gibraltar.

A male-biased sex ratio was found in S. haemastoma from Bizerta Channel, which agrees with previous studies on this species that concluded that this could be due to either female mortality or imposex in some females (Rilov 1999). Usually, differences in growth between populations collected from different sites are associated with many factors, including environmental conditions, TBT pollution, prey type and parasitism (Crothers 1985). Another factor that seems to affect the growth of gastropods is biofouling, since non-fouled and fouled gastropods were collected at the same sampling site. Indeed, the annual growths in terms of SD and FdW were higher in non-fouled gastropods. In fact, biofouling is time-dependent (Wahl 1989), which means that the growth of S. haemastoma is also accompanied by an increase in number and size of epibiotic barnacles. In some cases the weight of the barnacles per gastropod host is equal to its own weight; this most probably has negative effects on locomotion of snails, which must move between different substrates in order to feed, avoid unfavourable environmental conditions and escape predation. Consequently, gastropods fouled by epibiotic barnacles spend more energy during their movements at the expense of their growth than non-fouled gastropods. Indeed, it has been demonstrated that epibiotic barnacles in L. littorea decrease the locomotion and consequently the growth of gastropods (Buschbaum and Reise 1999, Thieltges and Buschbaum 2007). Thieltges and Buschbaum (2007) also showed that epibiotic Crepidula fornicata reduces the growth and survival of blue mussels (Mytilus edulis). Recently, Lacoste et al. (2014) recorded that biofouling reduces the growth of pearl oysters (Pinctada margaritifera) in French Polynesia. Our results showed that non-fouled females of S. haeamstoma have heavier shells and flesh. In contrast, Lahbib (2004) found that males of Hexaplex trunculus have heavier shells and flesh than females, considering this as sexual dimorphism characterizing specimens collected from Menzel Jemil (northern Tunisia).

The study of the reproductive cycle of S. haemastoma from the Tunisian coast showed an acceptable agreement between the macroscopic classification of gonad maturation stages and three of the bio-physiological indices used (GAI, CGI and PI). In contrast, the GSI showed significant fluctuations that could not be directly related to the reproductive cycle of S. haemastoma. This index is probably influenced by other factors, such as variations in the weight of the digestive gland after food ingestion and also SdW, because it is sometimes difficult to entirely remove all epibionts from the basibiont shell. Therefore, based on this result it would be more suitable to follow the reproductive activity of S. haemastoma using macroscopic observations together with GAI, CGI and PI. The simultaneous increase in GAI and decrease in CGI indicate gonad maturation and the beginning of spawning in May, which were probably triggered by an abrupt increase in seawater temperature, from 17°C in April to 23.9°C in May, with a similar finding being recorded in B. brandaris from Tunisian waters (Abidli el al. 2012). Vasconcelos et al. (2008) and Elhasni et al. (2010) recommended the use of two indices (GAI and CGI) and considered them simple, practical and efficient for the routine assessment of reproductive activity in a sympatric muricid (H. trunculus). Abidli et al. (2012), Vasconcelos et al. (2012) and Elhasni et al. (2013) confirmed these findings in analogous studies with another sympatric muricid (B. brandaris). Usually, differences in the reproductive cycle are explained by variation in seawater temperature (Lahbib et al. 2009, Abidli et al. 2012). Another factor that seems to affect the reproductive activity of gastropods is imposex (Lahbib et al. 2009). Results gathered in this study further suggest that the reproductive cycle of S. haemastoma is also affected by biofouling. Furthermore, Vasconcelos et al. (2011) established the PI for the study of the reproductive activity of male H. trunculus. This index was also practical for the study of the reproductive activity of male S. haemastoma, but must be used cautiously at sites highly affected by TBT pollution, such as Bizerta Channel (Lahbib et al. 2011), because it has been shown that TBT contamination might increase male PL (Castro et al. 2007, Abidli et al. 2013).