Abstract
The endangered patellid limpet Patella ferruginea is a protandrous species endemic to the Mediterranean. Previous studies have shown that sex change occurs in the species when individuals reach a range of 4–6 cm shell length. In our study individuals were sexed and population structure, mean density, mean shell length and sex ratio determined for seven populations in Ceuta (North Africa, Strait of Gibraltar). Our results indicate that sex change appears to occur at smaller sizes when large individuals (presumably females) are rare. All of the populations had a male-biased sex ratio. The greatest differences between the number of males and females were recorded in those populations that had the largest females and the greatest densities of large individuals, suggesting that a shortage of this fraction of the population would affect the proportion of sex-changing individuals. The fact that P. ferruginea may exhibit environmentally mediated sex change could influence population management decisions.
INTRODUCTION
Sequential hermaphroditism is common among marine invertebrates (Bauer, 2000), and environmentally mediated sex-changing processes in molluscs have been widely described since the early 20th century (e.g. Gould, 1919; Coe, 1953; Hoagland, 1978; Charnov, 1982). In the case of limpets, some species such as Cymbula oculus have a relatively fixed timing of sex change (Branch & Odendaal, 2003), while in others the size at which sex change occurs is highly variable. Coe (1938), Hoagland (1978) and later Branch (1981) suggested the possibility that these differences in the size/age of sex change could be mediated by environmental factors, which was later confirmed by Wright (1989) and Warner, Fitch & Standish (1996).
Patella ferruginea is a patellid limpet species endemic to the Mediterranean. It is considered to be at serious risk of extinction (Espinosa, Rivera-Ingraham & García-Gómez, 2009; European Council Directive 92/43/CEE; MMAMRM, 2008) as a consequence of human harvesting (Aversano, 1986; Guerra-García et al., 2004; Moreno, 2004). It is presently distributed mainly on the coast of North Africa, from the Strait of Gibraltar to Tunisia, although some populations remain on Corsica (Laborel-Deguen & Laborel, 1991), Sardinia (Doneddu & Manunza, 1992; Porcheddu & Milella, 1991) and the southernmost coast of Spain (Espinosa, 2006). It is a protandrous hermaphrodite (Frenkiel, 1975; Espinosa, Rivera-Ingraham & García-Gómez, 2008a), which is the most common pattern in sex-changing invertebrates (Berglun, 1991). It spawns once a year (Frenkiel, 1975) and sex change has been reported to occur when the species reaches around 4–6 cm shell length (see Espinosa et al., 2006,, 2008a; Templado et al., 2006), but sex change can also occur when individuals exceed 6 cm (Espinosa et al., 2006). Although sex change in many invertebrate species is genetically controlled, high variability in the timing or size of change leads us to think that environmental factors may influence the process. Since conspecific interactions often influence the timing at which sex change occurs in sex-changing molluscs (e.g. Coe, 1953; Hoagland, 1978; Charnov, 1982; Wright, 1989; Warner et al., 1996; Collin et al., 2005), the main objective of the present study was to determine if sex change in P. ferruginea is determined by intrapopulation conditions.
MATERIAL AND METHODS
The present study was carried out with genetically homogenous Patella ferruginea populations (based on the results by M. Casu, G.A. Rivera-Ingraham, P. Cossu, T. Lai, D. Sanna, G.L. Dedola, R. Sussarellu, B. Cristo, M. Curini-Galletti, J.C. García-Gómez, F. Espinosa, unpubl.) present on the coasts of Ceuta (North Africa, Strait of Gibraltar). One of the populations was at a site known locally as ‘Parque del Mediterráneo’ (Fig. 1), which is an inaccessible area because it is privately owned. The other populations studied were at sites accessible to a greater or lesser degree and, in consequence, the limpets are subject to harvesting, which can reach maximum annual rates of 40% (Rivera-Ingraham, 2010). At the seven sites, the sex and sizes of individuals were recorded in Autumn (October to December) 2006–2009. At each site, the shell length (measured with callipers) and sex (determined by the nonlethal method of Wright & Lindberg, 1979) were (where possible) assessed for 10 individuals per 1-cm size class exceeding 2 cm. This process occurred in Autumn because that is when the highest percentage of individuals are sexually mature (Frenkiel, 1975).
Sampling sites. 1, Parque del Mediterráneo; 2, Chorrillo; 3, Foso de San Felipe; 4, Sarchal; 5, Desnarigado; 6, Dique de Levante; 7, Benítez.
At each site (all with similar shore inclination, 40–45°), a 100-m transect was laid down during low tide, running parallel to the coast, and all animals present along the 100 m of shore were counted and measured to the nearest millimetres using callipers. This census and the process of sexing individuals were in all cases carried out during the same month. For each area, mean density, total number of adult individuals, mean shell length and the density of all large individuals (>5 cm) were calculated. The sex ratio (male:female) was estimated by taking into account both the results from the sexing process and the population structure recorded on the transects. The sex ratio recorded for the Parque del Mediterráneo population (protected from harvesting) was considered as the expected overall value for the species for further analyses.
The use of the L50 (size at which 50% of the animals have changed sex) is considered to be the most statistically rigorous estimate of the mean ‘size at sex change’ of a population (Allsop & West, 2003). However, the size of the smallest second-sex individual (female in the case of P. ferruginea) was used here as the ‘size at sex change’, as done by Shapiro (1981), because the L50 could not be calculated with our data owing to the lack of larger individuals in some populations. Statistical analyses were carried out using SPSS 14.0. All data satisfied the assumptions for parametric analysis. Pearson's correlation and regression analyses by curve estimation were performed to demonstrate the association between the size at sex change and the aforementioned population parameters. The statistically significant regression model with the highest coefficient of determination (R2) was accepted as the optimal model for the relationship. The sex ratios recorded in harvested areas were statistically compared with the value obtained in Parque del Mediterráneo by means of a χ2 test.
RESULTS
In total, 326 individuals were sexed. At different sites, sex change occurred at quite different sizes (Fig. 2). The density and size composition of the Patella ferruginea populations also differed (Fig. 3). Parque del Mediterráneo and Chorrillo showed the highest densities (8.70 and 14.30 ind/m, respectively) and, along with the population present in the Foso de San Felipe area, the highest number of large individuals (>5 cm shell length).
Differences in the proportions of males, females and immature individuals of Patella ferruginea at the seven sampling sites.
Size structure of the Patella ferruginea populations.
The size at sex change was correlated with the density of large individuals (Fig. 4) (Pearson's correlation coefficient R=0.773; P<0.05). A total of 11 different regression models were performed through curve-estimation regression analysis and a logarithmic model provided the optimal fit (R2=0.813; P=0.026; Fig. 4). A scarcity of larger individuals thus appeared to produce sex change at smaller sizes. The smallest female found (3.8 cm) was located in Dique de Levante, which had 0.24 large individuals per metre of shoreline. Moreover, the largest minimum-sized female found was 6.40 cm and was located at Chorrillo, the area with the highest density of large individuals (2.98 ind/m). On the other hand, the size at sex change was not correlated with populations’ overall density (R =0.603; P=0.15) or mean shell size (R=0.477; P=0.28).
Logarithmic regression of population density on minimum female size of Patella ferruginea for each sampling site. Letters are correlated with the populations considered in the study: P, Parque del Mediterráneo; C, Chorrillo; F, Foso; S, Sarchal; D, Desnarigado; L, Dique Levante; B, Benítez.
All populations were male-biased (Fig. 5) and sex ratios deviated significantly from 1:1 (χ2 tests P<0.01). Moreover, the two populations with the highest density values and the largest minimum-sized female recorded (Parque del Mediterráneo and Chorrillo) also had the greatest differences between the number of males and females. As mentioned, the sex ratio value recorded for Parque del Mediterráneo (25:1) was considered as the expected value for the species, and a χ2 test indicated that the other populations (accessible and harvested) had significantly lower sex ratios (P<0.05) (total mean value of 13:1).
Sex ratio of each of the populations of Patella ferruginea.
DISCUSSION
Although our results are preliminary, the species seems to be adjusting sex change to local density of larger individuals. It is known that some patellogastropod limpets such as Lottia gigantea also delay the size/age at which sex change occurs, the overall density of individuals being the determining factor (Wright, 1989). However, our results agree with those reported by other authors studying fish, crustaceans and molluscs, who determined that sex change can be delayed in populations where large females are present (e.g. Coe, 1938; Hoagland, 1978). This is also in accord with the size-advantage hypothesis, which predicts that sex change occurs earlier for those populations with higher mortality rates or slower growth rates (review by Munday, Buston & Warner, 2006). Social control of sex change has also been reported for Crepidula norrisiarum (Warner et al., 1996) and Lottia gigantea (Wright, 1989). For the latter, several possible cues for sex change have been suggested: contact frequency between individuals, movement area available, food intake, growth rate, pheromonal information and communication by mucus traces left by individuals during foraging excursions (Wright, 1989).
Patella ferruginea individuals in populations with low density of larger individuals switch to female at smaller sizes. This would explain why the species has not become extinct despite intense harvesting targeting large individuals. Moreover, recruits can be found in populations where no large individuals are present. This was the case in some of the populations considered in the study (e.g. Sarchal) and is also true of Andalusian populations (Moreno & Arroyo, 2008). Although these recruits could arrive from nearby populations, it is possible that they are being produced within the same population thanks to the smaller females.
Sex ratios were male-biased for each of the populations. This type of deviation is expected in protandric hermaphrodite species (Allsop & West, 2004). Moreover, males have been reported to outnumber females in many limpets such as P. kermadecensis (3:1) (Creese, Schiel & Kingsford, 1990), L. gigantea (2–2.6:1) (Kido & Murray, 2003), P. vulgata (1.3:1) and P. ulyssiponensis (2.4:1) (McCarthy, Woosnam & Culloty, 2008). However, it was surprising that the population present at Parque del Mediterráneo (almost completely free from collection, with a good proportion of large individuals and recruitment levels; Fig. 3) showed such large differences between the number of males and females (25:1). The other populations (which are accessible and subject to harvesting) had significantly different sex ratios and were more balanced. For other patellid species such as Cymbula oculus, populations located in marine protected areas also showed sex ratios different from those recorded in populations subject to harvesting, but the number of males and females was more equal in the former (Branch & Odendaal, 2003).
An enhancement of the sex-change rate in those populations lacking larger individuals could explain the more balanced sex ratios recorded in accessible areas. This would compensate for a shortage of large females, which produce a considerably higher number of oocytes than smaller ones. In fact, one 8-cm female produces approximately the same number of oocytes as ten 6-cm females (Espinosa et al., 2006). However, it should also be taken into account that the high recruitment rates that occur in Parque del Mediterráneo could also contribute to the large differences between the number of males and females.
The fact that the size of sex change in P. ferruginea is being adjusted according to the density of large individuals may partially offset the negative impacts of the Allee effect, which is a major problem in this species (Espinosa et al., 2009). The lack of enforcement of the laws regarding collection and the fact that the intertidal habitat is often easily accessible results in high depletion rates of intertidal organisms. Harvesting rates mainly affect the largest individuals, which are the most conspicuous (Haedrich & Barnes, 1997; Rochet & Trenkel, 2003). Because P. ferruginea is protandrous, the largest individuals tend to be females, and even though the species is not used commercially, accessible populations are often subject to considerable population structure alterations because of harvesting (Roy et al., 2003; Espinosa et al., 2009; Rivera-Ingraham, 2010). However, the depletion of larger individuals may be partially eased by this sex-change plasticity.
Traditionally, the sex ratios of P. ferruginea populations have been calculated using the population structure and mean sex-change size recorded for the species in nearby areas (e.g. Espinosa et al., 2009). Consequently those populations lacking larger individuals (>5 cm) have been considered as ‘nonreproductive’ or ‘not viable’ (e.g. Templado & Moreno, 1997; Espinosa, 2009) and their recruits were thought to come from nearby ‘reproductive’ populations. In consequence, management programmes have been directed towards the conservation of ‘reproductive’ populations. Taking into account the present results, we suggest instead that all populations should be considered for conservation, as a lack of larger individuals does not necessarily compromise the viability of the population. The size at sex change or the sex ratio of any particular P. ferruginea population should not be inferred using information from other populations, as it seems to depend highly on intrapopulation parameters; each population should be studied and considered individually when proposing conservation measures.
The sensitivity of P. ferruginea individuals produces high mortality rates during adult individual translocations (Espinosa et al., 2008b) and long-term maintenance in aquariums (personal observation). This complicates the possibility of conducting experimental research, which is needed in order to corroborate the results described here. Our results are preliminary but, nevertheless, contribute to the knowledge of the reproductive biology of P. ferruginea. This subject and the way in which individuals are able to detect the presence of large individuals deserve further research.
ACKNOWLEDGEMENTS
The authors would like to express their gratitude to Jorge Francisco Marín Lora and Juan José Díaz Pavón for their help in the sampling process and to staff of Consejería de Medio Ambiente de Ceuta-OBIMASA for their support. Thanks also go to Dr Jose Manuel Guerra-García and four anonymous referees for their constructive comments on the original manuscript. This work was supported by a F.P.U. grant from the Spanish Ministry of Science (grant number AP2006-04220) awarded to G.A.R.-I. Additional funding was provided by Autoridad Portuaria de Ceuta.
